Unwinding mechanism of SARS-CoV helicase (nsp13) in the presence of Ca²⁺, elucidated by biochemical and single-molecular studies

Abstract

The recent outbreak of COVID-19 has created a serious health crisis with fatFal infectious viral diseases, such as Severe Acute Respiratory Syndrome (SARS). The nsp13, a helicase of coronaviruses is an essential element for viral replication that unwinds secondary structures of DNA and RNA, and is thus considered a major therapeutic target for treatment. The replication of coronaviruses and other retroviruses occurs in the cytoplasm of infected cells, in association with viral replication organelles, called virus-induced cytosolic double-membrane vesicles (DMVs). In addition, an increase in cytosolic Ca²⁺ concentration accelerates viral replication. However, the molecular mechanism of nsp13 in the presence of Ca²⁺ is not well understood. In this study, we applied biochemical methods and single-molecule techniques to demonstrate how nsp13 achieves its unwinding activity while performing ATP hydrolysis in the presence of Ca²⁺. Our study found that nsp13 could efficiently unwind double stranded (ds) DNA under physiological concentration of Ca²⁺ of cytosolic DMVs. These findings provide new insights into the properties of nsp13 in the range of calcium in cytosolic DMVs.

1. Introduction

Coronavirus infections have caused serious respiratory illnesses worldwide, resulting in significant health crises and fatal consequences. From the early 2000s outbreak of Severe Acute Respiratory Syndrome (SARS) to the Middle East Respiratory Syndrome (MERS) in 2016 and the current COVID-19 pandemic, coronaviruses have proven to be a major public health threat [[1], [2], [3]]. Despite significant efforts made to understand these viruses, there is still much to learn about them, and it remains to develop a treatment method to prevent progression to severe symptoms after infection.

The SARS-CoV gene is translated into two replicated polyproteins by ORFs 1a and 1b, which are then separated into multiple non-structural proteins (nsps) through a proteolytic process by proteinases. Among the nsps, nsp13 is a genetically well-conserved and a critical protein of the Replication and Transcription Complex (RTC) that plays a central role in the viral life cycle [[4], [5], [6]]. Previous studies have focused on revealing the characteristics of nsp13. The nsp13 is a processive 5′-3′ helicase of the SF1B family that unwinds double-stranded (ds) into single stranded (ss) DNA or RNA, using chemical energy released from ATP hydrolysis [5,7,8]. In particular, SARS-nsp13 and SARS-CoV2 have high conservation with only one amino acid difference (I570V), making them a promising target for drug development [9].

Various metal ions play a pivotal role in nucleic acid biology. Types of divalent cations are very sensitive and critical to the catalysis of nucleic acid editing enzymes based on one- or two-metal ion mechanism that hydrolyzes various nucleotides and ribonucleotide. Calcium is known to a catalytically inactive cofactor that does not support hydrolysis of nucleic acids, and even inhibits enzymatic activity in many nucleases and helicases. The SARS-CoV amplification cycle involves viral gene replication and transcription processes, which occurs on the cytosolic surface of endoplasmic reticulum (ER)-derived vesicles, so-called virus-induced cytosolic double-membrane vesicles (DMVs) [10,11]. Since the cytoplasmic DMVs are often associated with and derived from the ER, the concentration of calcium ions in DMVs is thought to be similar to the ER, which contains calcium ions in the range of ∼1-5x10⁻⁴ M [12]. In this context, we explore the possibility of utilizing Ca²⁺ as a cofactor for ATP hydrolysis by nsp13, which acts at the same condition as DMVs derived from the ER.

Despite the extensive studies on the structure and unwinding activity of nsp13, the role of divalent cations, which are crucial cofactors for nsp13's unwinding activity, is poorly understood [7,[13], [14], [15], [16], [17], [18], [19], [20]]. Previous studies have focused mainly on Mg²⁺ as the cofactor for nsp13, but none have investigated the potential role of Ca²⁺ [[21], [22], [23]]. In this study, we employed a combination of biochemical and biophysical assays including polyacrylamide gel electrophoresis (PAGE), ATPase activity, and single-molecule FRET (smFRET), to investigate the unprecedented role of Ca²⁺, which is more physiologically relevant condition of viral replication. We found that nsp13 could efficiently unwind dsDNA at Ca²⁺ concentrations in cytosolic DMVs where rapid viral replication occurs. These findings provide a novel perspective on the components of helicase unwinding activity, which are essential for coronavirus replication and transcription in cytoplasmic DMVs.

2. Materials and methods

2.1. Protein expression and purification

The nsp13 (pHelA12) plasmid containing His tag was obtained from Prof. Dong-Eun Kim lab (Konkuk University). The plasmid was transformed into E. Coli BL21 (DE3) Star and grown in LB/Kanamycin at 37 °C until it reached an OD of 0.5. The protein was overexpressed by inducing with 0.5 mM IPTG for overnight at 18 °C. The cells were then harvested by centrifugation at 5,000×g and re-suspended in Buffer A (25 mM Tris-HCl (pH 6.8), 500 mM NaCl, 1 mM PMSF). Cells were lysed by sonication (Sonics & Materials, flat tip), followed by centrifugation at 30,000×g for 30 min. The supernatant was loaded onto a HisTrap HP column 5 ml (Cytiva), and elution was performed using Buffer B (25 mM Tris-HCl (pH 6.8), 500 mM NaCl, 500 mM Imidazole) with a gradual increase. The pure fractions were confirmed by SDS-PAGE, and concentrated with centrifugal filter (Amicon Ultra-15, 50K, Merck). Buffer was changed with Buffer C (25 mM Tris-HCl (pH 6.8), 200 mM NaCl, 30% glycerol). The protein was divided into small portions, frozen in liquid nitrogen, and stored at −80 °C. The concentration of the purified nsp13 was measured using Nanodrop (Thermo fisher scientific) at A280nm.

2.2. DNA oligo

All DNA oligo were purchased from Integrated DNA Technologies (IDT) or Macrogen. Amine-modified DNA was labeled with Cy3 NHS Ester or Cy5 NHS Ester (Cytiva). For substrates, the corresponding ssDNA was heated with their complementary strand in a T50 buffer (10 mM Tris-HCl (pH 8.0), 50 mM NaCl) for 3 min at 90 °C and slowly cooled down to room temperature for 3 h. If necessary, ligation was performed using T4 DNA ligase (New England Biolabs). Sequences and modification information of all DNA oligos are shown in Table 1 .

Table 1.

Oligo modification information.

Assay (Substrate name)	Oligo name	Sequence and modification
PAGE (60ss60ds-Cy5) in Fig. 1	60mer-Cy5	Cy5-TAA TAC GAC TCA CTA TAG GGA CAC AAA AAC AAA ATA ACA AGA AAA CAG AAC AAA TAA AAA
	120mer	TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT GTG TTT GTT TTG GTT TAT TTG GTG TTT TTA TTT GTT CTG TTT TCT TGT TAT TTT GTT TTT GTG TCC CTA TAG TGA GTC GTA TTA
ATPase (60ss60ds) in Fig. 2	Amine-60mer	Amine-TAA TAC GAC TCA CTA TAG GGA CAC AAA AAC AAA ATA ACA AGA AAA CAG AAC AAA TAA AAA
	120mer	TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT GTG TTT GTT TTG GTT TAT TTG GTG TTT TTA TTT GTT CTG TTT TCT TGT TAT TTT GTT TTT GTG TCC CTA TAG TGA GTC GTA TTA
smFRET (60ss18ds-Cy3, Cy5) in Fig. 3	Biotin-18mer-Cy5	Biotin-TGG CGA CGG CAG CGA GGC-Cy5
	78mer-Cy3	TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TT(Cy3)T TTT TTT TTT TTT GCC TCG CTG CCG TCG CCA
smFRET (60ds90ss-Cy3, Cy5) in Fig. 4	Cy5-60mer	Cy5-TTC AGA CTA AAC AAA TCA AAT ATC CAT AAC CAA TCA ACT CAA CTC ATC AAC CTT
	Phos-15mer-Cy3	Phosphate-TTT GTT TAG (Cy3)TCT GAA
	Phos-35mer-Amine	Phosphate-TTG ATG AGT TGA GTT GAT TGG TTA (Amine)TGG ATA TTT GA
	Biotin-100mer	Biotin-TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT GTG TTT GTT TTG GTT TAT TTG GTG TGG CGG AAG G

2.3. PAGE-based unwinding assay

Unwinding reactions were carried out in a buffer solution containing 25 mM Tris-HCl (pH 6.8), 25 mM NaCl, 50 μg/mL BSA, 25 nM 60ss60ds-Cy5 DNA, and 500 nM nsp13, with the addition of either MgCl₂ or CaCl₂ at various specified concentrations. A 50 nM trap DNA, which was a non-labeled complementary 60 nt (nucleotide) ssDNA, was included in the reaction mixture. The reaction mixture was incubated at 37 °C for 30 min, after which the unwinding reaction was stopped by adding an equal volume of quenching buffer containing 100 mM EDTA, 20% glycerol, 0.4% SDS, and 4 U/mL proteinase K (New England Biolabs) and incubating at 37 °C for 10 min. The products were separated by non-denaturing polyacrylamide gel electrophoresis (PAGE) using a 12% gel, and the Cy5 fluorescence signal was imaged using a ChemiDoc system (Bio-Rad). The fluorescence intensity was quantified and analyzed using Image Lab software (Bio-Rad).

2.4. Setup of smFRET and unwinding assay

The single-molecule Förster resonance energy transfer (smFRET) setup was constructed based on a custom-built prism-type Total Internal Reflection Fluorescence (TIRF) microscopy system (IX 71, Olympus). Quartz slides and cover glasses were coated with polyethylene glycol (PEG) to prevent nonspecific binding, and assembled using double-sided tape and epoxy [24]. Fluorescently labeled DNA substrates (Table 1) were immobilized in T50 buffer using NeutrAvidin-biotin interaction. The green laser (532 nm, Coherent) was used to excite the fluorescent emission light, which was collected using an UplanSApo 60x objective lens (N.A. = 1.20, Olympus). The donor (Cy3) and acceptor (Cy5) fluorescence signals were separated using a dichroic mirror (660 nm cutoff) and recorded using an electron-multiplying charge-coupled device (EMCCD) camera (iXon Ultra 897, Andor) with 100 ms time resolution. The fluorescence signal was detected using the EMCCD and amplified by EM gain. The Cy3 and Cy5 signals were separated by Gaussian fitting and signals above the average background were selected. The peak positions of Cy3 and Cy5 channels were co-localized using a mapping algorithm developed in Interactive Data Language (IDL) software. The intensity was then extracted from the recorded video file. The FRET efficiency (E_FRET) was calculated using the equation E_FRET=(I_A-α∗I_D)/(I_A + I_D), where α, I_D, and I_A are the leakage correction factor and the signals of the donor (Cy3) and acceptor (Cy5), respectively.

For the unwinding assay, the immobilized DNA substrates were washed with imaging buffer (20 mM Tris-HCl (pH 6.8), 20 mM NaCl, MgCl₂ (or CaCl₂), 100 μg/ml BSA, 1 mg/ml Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid), 1 mg/ml glucose oxidase, 0.03 mg/ml catalase, and 1% (v/v) Dextrose) to remove free DNA. The reaction buffer containing nsp13 and ATP was injected into the chamber after approximately 6 sec of recording, and recording continued for 2 min. Alternatively, several images (∼100 shots) of different locations were recorded for 10 min after injection of the reaction buffer. All experiments were conducted at 37 °C. Data were analyzed using MATLAB and Origin software.

2.5. ATPase activity assay

To assess the activity of ATP hydrolysis, an ATPase assay kit (NOVUS biologicals) was employed and the manufacturer's protocol was strictly followed. The amount of inorganic phosphate released by ATP hydrolysis was quantified using this kit. To initiate the reaction, a mixture containing 15 nM nsp13, 50 μM ATP, and 0.5 mM MgCl₂ (or CaCl₂) was added to a buffer solution consisting of 20 mM Tris-HCl (pH 6.8), 20 mM NaCl, and 10 nM 60ss60ds DNA. After the incubation period, the reaction was terminated based on the instruction specified in the kit manual. In the assay of Fig. 4C, the ATPase activity assays were repeated at various concentration of Ca²⁺ ranging from 0.001 to 10 mM to determine the degree of ATP hydrolysis by unwinding activity of nsp13. The extent of ATP hydrolysis was determined by measuring the absorbance at 620 nm, using 96-well clear low binding plates (NOVUS biologicals) and FlexStation3 (Molecular Devices).

Fig. 4.

The Ca²⁺-dependent unwinding activity of nsp13 driven by ATP hydrolysis. (A) Experimental schematics before (left) and after (right) unwinding by nsp13 and ssDNA mimicking the unwound product. In the partial duplex DNA substrate, Cy3 was placed 5 bp away from the 3′ end of the immobilized strand, and Cy5 was labeled at the 5′ end of the complementary strand to be unwound. (B) Representative histogram obtained 2 min after reaction. (C) Unwinding fraction and ATPase activity versus various Ca²⁺ concentrations after 10 min reaction. (D) Unwinding fraction using ATP or non-hydrolysable nucleotide analogs. (E and F) non-denaturing PAGE gel showing the unwinding activity as a function of ATP concentration at 0.5 mM and 10 mM Ca²⁺ and its quantification.

3. Results

3.1. In the presence of Mg²⁺ or Ca²⁺, the nsp13 exhibits unwinding activity

A PAGE assay was used to confirm that nsp13 has unwinding activity of purified nsp13 (Fig. 1 A) on DNA substrate with a 60 nt 5′ single-stranded (ss) overhangs in a 60 nt double-stranded regions (shown in the top left of Fig. 1B). All control lanes showed no unwinding activity of nsp13 (lanes 1–7), except for lane 2 where the DNA substrate was denatured by heating at 95 °C for 10 min. In the presence of 0.5 mM Mg²⁺, the DNA substrate was unwounded as a similar level as the control of lane 2 (lane 8). In the presence of 0.5 mM Ca²⁺, it was also unwound significantly (lane 9). The result was quite surprisingly given the fact that Ca²⁺ has been known to be a catalytically inactive ion for the two-metal ion catalysis [25], and previous studies did not observe the unwinding activity of nsp13 under Ca²⁺ conditions [[21], [22], [23]]. We speculated that the reason for the lack of activity might be the use of high concentration of Ca²⁺ and large amounts of trap DNA (over 20-fold excess). We used the minimal trap DNA to prevent inhibition of the enzymatic activity, which allowed us to observe the unwinding activity of nsp13 in the presence of Ca²⁺. Consistently, no unwinding activity was not clearly observed under Ca²⁺ conditions when a higher excess (10-fold excess) of trap DNA was added (Fig. 1D).

Fig. 1.

Unwinding activity of DNA unwinding of nsp13. (A) SDS-PAGE gel showing the purity of nsp13. (B) non-denaturing PAGE gel showing the degree of unwinding: Lane 1, DNA substrate alone (shown in the top left); Lane 2, DNA substrate heated for 10 min at 95 °C (shown in the bottom left); Lanes 3–7, each condition indicated at the top of the gel was added or subtracted; Lane 8, the reaction with MgCl₂; Land 9, the reaction with CaCl₂ The nsp13 was incubated with the DNA substrate as described in Materials and Methods. The unwinding products were quantified on the 12% native-PAGE. (C) Quantification of B. Error bars denote the standard error of the mean (SEM). (D) non-denaturing PAGE gel, run under the same conditions as in B but at a 10-fold higher trap DNA concentration than the DNA substrate.

3.2. ATP hydrolysis-dependent unwinding activity by nsp13 in the presence of Ca²⁺

To test whether unwinding activity of nsp13 was driven by ATP hydrolysis with the help of Ca²⁺ as with Mg²⁺, we quantified ATP hydrolysis using a colorimetric ATPase assay kit (NOVUS biologicals) (Fig. 2 A and B). Although the result was slower than Mg²⁺ condition, a considerable degree of hydrolysis activity was shown even under Ca²⁺ condition (Fig. 2B). Since the ATP assays kit provides sensitivity for very low concentrations of ATP hydrolysis (e.g., 50 μM ATP in our case), a PAGE-based unwinding assay was used to measure a higher range of ATP hydrolysis (Fig. 2C). As a result, the ATP hydrolysis of nsp13 was saturated at ∼2 mM ATP in the presence of Ca²⁺ (Fig. 2D).

Fig. 2.

In the presence of Ca²⁺, nsp13 utilizes ATP hydrolysis in a wide range of ATP. (A) Schematics of the colorimetric ATPase assay. When ATP is decomposed into ADP and Pi by the hydrolysis reaction, the dye color changes from orange to green, which can be quantitatively measured by absorbance at 620 nm. (B) Absorbance at 620 nm as a function of time. (C and D) non-denaturing PAGE gel showing the unwinding activity as a function of ATP concentration and its quantification.

3.3. Single-molecule FRET assay for measuring unwinding activity by nsp13

For a more detailed analysis, a single-molecule FRET technique was used [24]. The DNA substrate labeled with Cy3 (green) on the 60 nt overhang strand and Cy5 (red) on the complementary strand was immobilized on the PEGylated quartz surface via NeutrAvidin-biotin interaction (Fig. 3 A). The unwinding reaction was carried out in the buffer (see materials and methods), containing 2 mM ATP, 0.5 mM Ca²⁺ (or Mg²⁺). FRET-time trajectories showed disappearance of Cy3 fluorescent signals (Fig. 3B) and the images of Cy3 and Cy5 channels also showed loss of Cy3 spots, but not Cy5 (Fig. 3D and E). This was because after dsDNA is unwound by nsp13, the Cy3-labeled strands diffused away, but the surface-immobilized Cy5-labeled strands remained on the surface. The comparison of Mg²⁺ and Ca²⁺ indicated that the rate of unwinding is about 7–10 times slower for Ca²⁺ compared to Mg²⁺ (Fig. 3C).

Fig. 3.

Single-molecule FRET shows that the unwinding activity of nsp13 was slower in the presence of Ca²⁺ compared to Mg²⁺. (A) Experimental schematics before (left) and after (right) unwinding by nsp13 and ssDNA mimicking the unwound product. In the partial duplex DNA substrate, Cy3 was placed 47 bp away from the 5′ end of the overhang strand, and Cy5 was labeled at the 3′ end of the complementary strand to be unwound. (B) Representative FRET time trace obtained under 2 mM ATP and 0.5 mM Mg²⁺ (top) or 0.5 mM Ca²⁺ (bottom). (C) Unwinding rate as a function of nsp13 concentration (1, 5, 10, 50, and 150 nM). (D) Individual spots representing single-molecules on the donor (Cy3) and acceptor (Cy5) channels. The most of the Cy3 spots disappeared while that of the Cy5 spots remained the same after the nsp13-induced unwinding reaction. (E) Quantification of single-molecule spots on the donor and acceptor channels.

3.4. The unwinding activity of nsp13 is maximal from 0.2 to 1 mM Ca²⁺ and inhibited at higher concentrations

Next, we investigate the Ca²⁺ concentration sensitivity of unwinding activity by nsp13 at 2 mM ATP using the single-molecule FRET technique. The partial DNA duplex composed of a 60-bp dsDNA followed by a 90-nt ssDNA was used (Fig. 4A). In this experiment, the Cy3-labeled strands remained after the unwinding reaction since only the Cy3-labeled strands were immobilized on the surface, but the Cy5-labeled strands were released and diffused away. The single-molecule histogram of FRET efficiency (E) was shifted from a peak at E = ∼0.9 before the unwinding reaction (top of Fig. 4B) to a peak at E = ∼0.0 after unwinding reaction (bottom of Fig. 4B). The unwound fraction was then calculated based on the increase in the low FRET population as a function of Ca²⁺ concentration (Fig. 4C). The analysis showed the highest unwinding activity from 0.2 to 1 mM Ca²⁺ (Fig. 4C). The pattern of the unwinding fraction of nsp13 was very consistent with that of ATP hydrolysis by the unwinding activity of nsp13 at various Ca²⁺ concentrations (Fig. 4C). In Fig. 4D, we also measured the unwinding activity of nsp13 using non-hydrolysable nucleotide analogs (ATPγS and AMPPNP), demonstrating that the unwinding activity via Ca²⁺ is performed by ATP hydrolysis. Unwinding activity of nsp13 using Ca²⁺ was observed only under ATP where ATP hydrolysis occurs, but not non-hydrolysable nucleotide analogs. These data strongly indicate that the unwinding activity of nsp13 with Ca²⁺ is driven by the ATP hydrolysis. To examine the ATP-concentration dependence of activity, PAGE-based unwinding assays were performed at 0.5 and 10 mM Ca²⁺ (Fig. 4E). The results also confirmed that the unwinding activity of nsp13 was inhibited at high concentrations of Ca²⁺ (see pink and red curves) and saturated at concentrations higher than 2 mM ATP (Fig. 4F).

4. Discussion

The replication and transcription of the SARS coronavirus take place in virus-induced cytosolic DMVs [10,11], which have been recently shown to be connected with ER [[26], [27], [28]].

Assuming that the Ca²⁺ concentration in the ER (∼1-5 x 10⁻⁴ M) is higher than the normal cytosolic Ca²⁺ concentration (∼10⁻⁷ M) and that the ER and cytoplasm are connected, the 0.1–1 mM Ca²⁺ concentration used in the experiment is physiologically very reasonable for cytosolic DMVs [12]. The level of intracellular ATP are generally in the range of 0.5–5 mM, which agrees well with the range of concentrations used in our experiment [[29], [30], [31]].

Previous studies have shown that helicases are unable to unwind double-stranded DNA (dsDNA) or double-stranded RNA (dsRNA) in the presence of Ca²⁺ because Ca²⁺ inhibits ATP hydrolysis by helicases [32,33]. Most studies have therefore focused on unwinding activities on helicases in the presence of Mg²⁺. However, our study has found that nsp13 can utilize ATP hydrolysis to perform dsDNA unwinding activity even under conditions of Ca²⁺, which may be physiologically more relevant than Mg²⁺ to cytosolic DMVs. Unwinding activity of nsp13 under Ca²⁺ is ∼7–10 times slower than that of Mg²⁺, but the highest activity occurs at Ca²⁺ concentrations of 0.2–1 mM, which is sufficient for efficient dsDNA unwinding.

Despite the inhibitory effect of Ca²⁺ on helicase activity, our study highlights the importance of understanding that nsp13 is able to unwind dsDNA at low concentrations of Ca²⁺ because this is the biological milieu and directionally relevant to the replication and transcription processes of coronaviruses. In summary, our study provides insights into the utilization of Ca²⁺ as a cofactor by nsp13 during DNA unwinding, shedding light on the diverse spectrum of divalent cations involved in coronavirus replication and transcription.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.