key: cord-280994-w8dtfjel
authors: Peng, Qi; Peng, Ruchao; Yuan, Bin; Zhao, Jingru; Wang, Min; Wang, Xixi; Wang, Qian; Sun, Yan; Fan, Zheng; Qi, Jianxun; Gao, George F.; Shi, Yi
title: Structural and biochemical characterization of nsp12-nsp7-nsp8 core polymerase complex from COVID-19 virus
date: 2020-04-23
journal: bioRxiv
DOI: 10.1101/2020.04.23.057265
sha: 
doc_id: 280994
cord_uid: w8dtfjel

The ongoing global pandemic of coronavirus disease 2019 (COVID-19) has caused huge number of human deaths. Currently, there are no specific drugs or vaccines available for this virus. The viral polymerase is a promising antiviral target. However, the structure of COVID-19 virus polymerase is yet unknown. Here, we describe the near-atomic resolution structure of its core polymerase complex, consisting of nsp12 catalytic subunit and nsp7-nsp8 cofactors. This structure highly resembles the counterpart of SARS-CoV with conserved motifs for all viral RNA-dependent RNA polymerases, and suggests the mechanism for activation by cofactors. Biochemical studies revealed reduced activity of the core polymerase complex and lower thermostability of individual subunits of COVID-19 virus as compared to that of SARS-CoV. These findings provide important insights into RNA synthesis by coronavirus polymerase and indicate a well adaptation of COVID-19 virus towards humans with relatively lower body temperatures than the natural bat hosts.

In the end of 2019, a novel coronavirus (2019-nCoV) caused an outbreak of pulmonary disease 33 in China (Zhu et al., 2020) , which was later officially named "severe acute respiratory syndrome viruses with a broad host-spectrum (Vicenzi et al., 2004) . Currently, a total of seven human-47 the interface domain ( Figures 1B and 3D ). The two nsp8 subunits display significantly different 164 conformations with substantial refolding of the N-terminal extension helix region, which 165 mutually preclude the binding at the other molecular context ( Figure 3C ). The importance of 166 both cofactor-binding sites has been validated by previous biochemical studies on polymerase, which revealed their essential roles for stimulating the activity of nsp12 168 polymerase subunit (Subissi et al., 2014) . Given the residue substitutions between SARS-CoV-2 and SARS-CoV polymerase subunits 182 albeit the high degree overall sequence similarity, we compared the enzymatic behaviors of the 183 viral polymerases aiming to analyze their properties in terms of viral replication. Both sets of 184 core polymerase complex could well mediate primer-dependent RNA elongation reactions templated by the 3'-vRNA. Intriguingly, the SARS-CoV-2 nsp12-nsp7-nsp8 complex displayed 186 a much lower efficiency (~35%) for RNA synthesis as compared to the SARS-CoV counterpart 187 ( Figure 4A ). As all three nsp subunits harbor some residue substitutions between the two 188 viruses, we further conducted cross-combination analysis to evaluate the effects of each subunit 189 on the efficiencies of RNA production. In the context of SARS-CoV-2 nsp12 polymerase 190 subunit, replacement of the nsp7 cofactor subunit with that of SARS-CoV did not result in 191 obvious effect on polymerase activity, whereas the introduction of SARS-CoV nsp8 subunit 192 greatly boosted the activity to ~2.1 times of the homologous combination. Simultaneous 193 replacement of the nsp7 and nsp8 cofactors further enhanced the efficiency for RNA synthesis 194 to ~2.2 times of that for the SARS-CoV-2 homologous complex ( Figure 4B ). Consistent with 195 this observation, the combination of SARS-CoV-2 nsp7-nsp8 subunits with the SARS-CoV 196 nsp12 polymerase subunit compromised its activity as compared to the native cognate cofactors, 197 among which the nsp8 subunit exhibited a more obvious effect than that for nsp7 ( Figure 4C) . 198 These evidences suggested that the variations in nsp8 subunit rendered a significantly negative 199 impact on the polymerase activity of SARS-CoV-2 nsp12. The non-significant effect of nsp7 200 on polymerase activity was quite conceivable as only one residue substitution occurred between 201 the two viruses ( Figure 2B ). In addition, we also compared the polymerase activity of different 202 nsp12 subunits in the same context of nsp7-nsp8 cofactors. Combined with either cofactor sets, 203 the SARS-CoV-2 nsp12 polymerase showed a lower efficiency (~50%) for RNA synthesis as 204 compared to the SARS-CoV counterpart ( Figure 4D ). This observation demonstrated that the 205 residue substitutions in nsp12 also contributed to the reduction of its polymerase activity, with 206 similar impact to the variations in the nsp8 cofactor. 207 208

Despite that there are amino acid substitutions in all three subunits of the core polymerase 210 complex between SARS-CoV-2 and SARS-CoV, none of these residues is located at the 211 polymerase active site or the contacting interfaces between adjacent subunits ( Figure 2B) , 212 suggesting these substitutions do not affect the inter-subunit interactions for assembly of the 213 polymerase complex. To test this hypothesis, we measured the binding kinetics between 214 different subunits of the two viruses by surface plasmon resonance (SPR) assays. Each 215 interaction pair exhibited similar kinetic features for the two viruses, all with sub-micromolar 216 range affinities ( Figure 5A and B). We also tested the cross-binding between subunits of the 217 two viruses, which revealed similar affinities for heterologous pairs as compared to the native 218 homologous interactions ( Figure 5C and D). See also Figures S1-S3 and Table S1 . 

The codon-optimized sequences of nsp7 and nsp8 were synthesized with N-terminal 392 6×histidine tag and inserted into pET-21a vector for expression in E. coli (Synbio Tec, Suzhou, 393 China). For the nsp7L8 fusion protein, the sequence was also codon-optimized for E. Coli 394 expression system and a 6×histidine linker was introduced between the nsp7 and nsp8 subunits 395 (Genewiz Tec, Suzhou, China). Protein production was induced with 1 mM isopropylthio-396 galactoside (IPTG) and incubated for 14-16 hours at 16 °C. Bacterial cells were harvested by 397 centrifugations (12,000 rpm, 10 min), resuspended in buffer A (20 mM HEPES, 500 mM NaCl, 398 2 mM Tris (2-carboxyethyl) phosphine (TCEP), pH 7.5, and lysed by sonication. Cell debris 399 were removed via centrifugation (12,000 rpm, 1h) and filtration with a 0.22 μm cut-off filter. 

An aliquot of 3 μL protein solution (0.6 mg/mL) was applied to a glow-discharged Quantifiol 419 1.2/1.3 holey carbon grid and blotted for 2.5 s in a humidity of 100% before plunge-freezing 420 with an FEI Vitrobot Mark IV. Cryo-samples were screened using an FEI Tecnai TF20 electron 421 microscope and transferred to an FEI Talos Arctica operated at 200 kV for data collection. The 422 microscope was equipped with a post-column Bioquantum energy filter (Gatan) which was 423 used with a slit width of 20 eV. The data was automatically collected using SerialEM software 424 (http://bio3d.colorado.edu/SerialEM/). Images were recorded with a Gatan K2-summit camera 425 in super-resolution counting mode with a calibrated pixel size of 0.8 Å at the specimen level. 426

Each exposure was performed with a dose rate of 10 e -/pixel/s (approximately 15.6 e -/Å 2 /s) and lasted for 3.9 s, resulting in an accumulative dose of ~60 e -/Å 2 which was fractionated into 30 428 movie-frames. The final defocus range of the dataset was approximately -1.4 to -3.4 μm. 429 430

The image drift and anisotropic magnification was corrected using MotionCor2 (Zheng et al., 432 2017). Initial contrast transfer function (CTF) values were estimated with CTFFIND4.1 (Rohou 433 and Grigorieff, 2015) at the micrograph level. Images with an estimated resolution limit worse 434 than 5 Å were discarded. Particles were automatically picked with RELION-3.0 (Zivanov et al., 435 2018) following the standard protocol. In total, approximately 1,860,000 particles were picked 436 from ~4,200 micrographs. After 3 rounds of extensive 2D classification, ~924,000 particles 437 were selected for 3D classification with the density map of SARS-CoV nsp12-nsp7-nsp8 438 complex (EMDB-0520) as the reference which was low-pass filtered to 60 Å resolution. After 439 two rounds of 3D classification, a clean subset of ~101,000 particles was identified, which 440 displayed clear features of secondary structural elements. These particles were subjected to 3D 441 refinement supplemented with per-particle CTF refinement and dose-weighting, which led to a 

The structure of SARS-CoV nsp12-nsp7-nsp8 complex (PDB ID: 6NUR) was rigidly docked 448 into the density map using CHIMERA (Pettersen et al., 2004) . The model was manually corrected for local fit in COOT (Emsley et al., 2010) and the sequence register was corrected 450 based on alignment. The initial model was refined in real space using PHENIX (Adams et al., 451 2010) with the secondary structural restraints and Ramachandran restrains applied. The model 452 was further adjusted and refined iteratively for several rounds aided by the stereochemical 453 quality assessment using MolProbity (Chen et al., 2010) . The representative density and atomic 454 models are shown in Figure. Images were taken using a Vilber Fusion system and analyzed with the Image J software. 

Python-based system for macromolecular structure solution

Mechanism of nucleic acid unwinding by SARS-CoV helicase

Biochemical characterization of a 555 recombinant SARS coronavirus nsp12 RNA-dependent RNA polymerase capable of copying 556 viral RNA templates

Shaping the flavivirus replication complex: It is 558 curvaceous!

The proximal 560 origin of SARS-CoV-2

MolProbity: all-atom structure 563 validation for macromolecular crystallography

The 566 species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and 567 naming it SARS-CoV-2

The crystal 569 structure of Zika virus NS5 reveals conserved drug targets

Features and development of

Structure of the RNA-dependent RNA polymerase from COVID-19 virus

Structural insights into bunyavirus 576 replication and its regulation by the vRNA promoter

Crystal structure of Zika virus NS5 RNA-dependent RNA polymerase

Structural basis for active site closure by the poliovirus 581

RNA-dependent RNA polymerase

Crystal structure of the RNA-584 dependent RNA polymerase from influenza C virus

United States

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

China: implication for infection prevention and control measures

Structure of the SARS-CoV nsp12 polymerase 595 bound to nsp7 and nsp8 co-factors

Quantifying the local resolution of 597 cryo-EM density maps

Discovery of an essential nucleotidylating activity associated with a newly 601 delineated conserved domain in the RNA polymerase-containing protein of all nidoviruses

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications 605 for virus origins and receptor binding

Bat flight and zoonotic viruses

Structural insight into arenavirus replication machinery

UCSF Chimera--a visualization system for exploratory research and 614 analysis

Structure of influenza A polymerase 616 bound to the viral RNA promoter

Structural insight into cap-snatching and RNA synthesis by 619 influenza polymerase

CTFFIND4: Fast and accurate defocus estimation from 621 electron micrographs

Transmission of 2019-nCoV infection from an 624 asymptomatic contact in Germany

Insights into RNA 626 synthesis, capping, and proofreading mechanisms of SARS-coronavirus

Viral RNA polymerase: a promising antiviral target 628 for influenza A virus

One severe acute respiratory syndrome 631 coronavirus protein complex integrates processive RNA polymerase and exonuclease activities

Coronaviridae and SARS-associated coronavirus strain HSR1

Clinical characteristics of 138 hospitalized patients with 2019 novel 638 coronavirus-infected pneumonia in Wuhan, China

Structural basis for the inhibition of the RNA-dependent RNA polymerase 641 from SARS-CoV-2 by Remdesivir. bioRxiv

Insights 643 into SARS-CoV transcription and replication from the structure of the nsp7-nsp8 hexadecamer

A genomic perspective on the origin and emergence of 646 SARS-CoV-2

Structure and function of the Zika virus full-length NS5 protein

MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron 652 microscopy

Epidemiology and cause of severe acute respiratory syndrome

People's Republic of China

A pneumonia outbreak associated with a new coronavirus of 658 probable bat origin

A novel coronavirus from patients with pneumonia in China

The coronavirus replicase

New tools for automated high-resolution cryo-EM structure determination in 665 RELION-3

The affinities between nsp12, nsp7 and nsp8 or nsp7L8 proteins were measured at room 473 temperature (r.t.) using a Biacore 8K system with CM5 chips (GE Healthcare). The nsp12 474 protein was immobilized on the chip with a concentration of 100 μg/mL (diluted by 0.1 mM 475 NaAc, pH 4.0), and the nsp7 protein was immobilized with a concentration of 50 μg/mL (diluted 476 by 0.1 mM NaAc, pH 4.5). For all measurements, the same running buffer was used which 477 consists of 20 mM HEPES, pH 7.5,150 mM NaCl and 0.005% tween-20. Proteins were pre-478 exchanged into the running buffer by SEC prior to loading to the system. A blank channel of 

The cryo-EM density map and atomic coordinates have been deposited to the Electron 494 Microscopy Data Bank (EMDB) and the Protein Data Bank (PDB) with the accession codes 495 EMD-30226 and 7BW4, respectively. All other data are available from the authors on 496 reasonable request. 497