key: cord-0795653-e0o2lbbw
authors: Littler, Dene R.; Gully, Benjamin S.; Colson, Rhys N.; Rossjohn, Jamie
title: Crystal structure of the SARS-CoV-2 non-structural protein 9, Nsp9
date: 2020-06-09
journal: iScience
DOI: 10.1016/j.isci.2020.101258
sha: 525b12c03f6df45e0a06abdfff506f88949e91fd
doc_id: 795653
cord_uid: e0o2lbbw

Summary Many of the SARS-CoV-2 proteins have related counterparts across the Severe Acute Respiratory Syndrome (SARS-CoV) family. One such protein is non-structural protein 9 (Nsp9), which is thought to mediate viral replication, overall virulence and viral genomic RNA reproduction. We sought to better characterise the SARS-CoV-2 Nsp9 and subsequently solved its X-ray crystal structure, in an apo-form and, unexpectedly, in a peptide-bound form with a sequence originating from a rhinoviral 3C protease sequence (LEVL). The SARS-CoV-2 Nsp9 structure revealed the high level of structural conservation within the Nsp9 family. The exogenous peptide binding site is close to the dimer interface and impacted the relative juxtapositioning of the monomers within the homodimer. We have established a protocol for the production of SARS-CoV-2 Nsp9, determined its structure and identified a peptide-binding site that warrants further study to understanding Nsp9 function.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is comprised of a large single stranded positive polarity RNA genome that acts as messenger RNA after entering the host. The 5' two thirds of the genome encodes a large polyprotein that is translated into ORF1a and ORF1ab through host ribosomal frameshifting, with the remainder of the viral RNA encoding structural and accessory proteins within smaller ORFs. The viral proteins necessary for host cell infection such as the RNA-polymerase along with enzymes that facilitate RNA synthesis are largely contained within the SARS-CoV-2 polyproteins and are released by the action of two internally encoded proteases. The mature proteins thus released are known as non-structural proteins (Nsp) as they are not incorporated within the virion particles. Due to their degree of sequence conservation the enzymatic roles and essentiality of each of the Nsps within SARS-CoV-2 is likely to mimic the behavior of homologous proteins within previously studied coronaviruses such as SARS-CoV.

The development of therapeutic interventions against SARS-CoV-2 infection has focused on a number of approaches: vaccination strategies that target the structural spike glycoprotein of the envelope (Wrapp et al., 2020) but may also include a larger selection of viral proteins (Thanh Le et al. 2020 ); while small-molecule compounds have predominantly targeted two conserved viral enzymes, the main protease (Zhenming et al. 2020 ) (Yang et al., 2005) and the RNA-polymerase (Yan et al. 2020) . Nevertheless, some of the betacoronaviral non-structural proteins appear necessary for viral replication within SARS-CoV and influence pathogenesis (Frieman et al., 2012) . Despite their close homology between viruses, such non-structural proteins remain of interest as they may have conserved roles within the viral lifecycle of SARS-CoV-2 that could be susceptible to inhibition.

During infection of human cells, SARS-CoV Non-structural protein 9 (Nsp9 SARS ) was found to be essential for replication (Frieman et al., 2012) . Homologs of the Nsp9 protein have been identified in numerous coronaviruses including SARS-Cov-2 (Nsp9 COV19 ), human coronavirus 229E (Nsp9 HcoV ), avian infectious bronchitis virus (Nsp9 IBV ), porcine epidemic diarrhea virus (Nsp9 PEDV ) and porcine delta virus (Nsp9 PDCoV ).

Nsp9 SARS has been shown to have modest affinity for long oligonucleotides with binding thought to be dependent on oligomerisation state (Sutton et al., 2004) . Nsp9 SARS dimerises in solution via a conserved α-helical 'GxxxG' motif. Disruption of key residues within this motif reduces both RNA-binding (Sutton et al., 2004) and SARS-CoV viral replication (Frieman et al., 2012) . The mechanism of RNA binding within the Nsp9 protein family is not understood as these proteins have an unusual structural fold not previously seen in RNA binding proteins (Sutton et al., 2004) . The fold's Greek-key motif exhibits topological similarities with Oligonucleotide/oligosaccharide binding proteins (OB-fold) but such vestiges have proven insufficient to provide clear insight into Nsp9 function . As a consequence of the weak affinity of Nsp9 SARS for long oligonucleotide stretches it was suggested that the natural RNA substrate may instead be conserved features at the 3' end of the viral-genome (the stem-loop II RNA-motif) (Ponnusamy et al., 2008) . Furthermore, potential direct interactions with the co-factors of the RNA polymerase have been reported . However, it remains to be determined how the oligonucleotide-binding activity of Nsp9 proteins promote viral replication during infection.

The sequence of Nsp9 homologues are conserved amongst betacoronoaviruses, suggesting a degree of functional conservation. Nsp9 COV19 exhibits 97% sequence identity with Nsp9 SARS but only 44% sequence identity with Nsp9 HCoV. The structure of the HCoV-229E Nsp9 protein suggested a potential oligomeric switch induced upon the formation of an intersubunit disulfide bond. Here, disulfide bond formation shifts the relative orientation of the Nsp9 monomers, which was suggested to promote higher-order oligomerisation (Ponnusamy et al., 2008) . The resultant rod-like higher order Nsp9 HCoV assemblies had increased affinity for the RNA oligonucleotides. Cysteine mutants of Nsp9 HCoV that are unable to produce the disulfide displayed reduced RNA-binding affinity (Ponnusamy et al., 2008) . The observation of a redox-induced structural switch of Nsp9 HCoV led to the hypothesis that Nsp9 HCoV may have a functional role in sensing the redox status of the host cell (Ponnusamy et al., 2008) . While the "redox-switch" cysteine responsible for oligomer formation in Nsp9 HCoV is conserved amongst different viral Nsp9 homologues the higher-order oligomers were not observed for Nsp9 SARS (Ponnusamy et al., 2008) . Because of these potential differences between Nsp9 proteins we sought to further characterise the nature of Nsp9 COV19.

The Nsp9 protein from SARS-CoV-2 (Nsp9 COV19 ) was cloned and recombinantly expressed in E. coli.

The expression construct included an N-terminal Hexa-His tag attached via a rhinoviral 3C-protease site.

Following Ni-affinity chromatography Nsp9 COV19 was further purified via size-exclusion chromatography to yield > 95% pure and homogeneous protein. Nsp9 COV19 eluted from gel filtration columns with the apparent molecular weight of a dimer suggesting that, as with other Nsp9 proteins, Nsp9 COV19 is an obligate homodimer.

The N-terminal tag was removed prior to any biochemical experiments via overnight digestion with precision protease, as reported for Nsp9 SARS (Sutton et al., 2004) .

The affinity of viral Nsp9 homologues for oligonucleotides has a range of binding affinities reported, some of which are dependent on oligomerisation state and nucleotide length, ranging from 20-400 μM (Zeng et al., 2018) . We therefore sought to assess the potential for Nsp9 COV19 to bind to fluorescently labelled oligonucleotides using fluorescence anisotropy. Preliminary experiments were performed under conditions similar to those previously identified for Nsp9 SARS (Sutton et al., 2004) . Oligonucleotide affinity was very limited under our assay conditions (Fig. 1) . Indeed, protein concentrations up to 200 µM of Nsp9 COV19 did not result in saturated binding and thus indicated an incredibly low affinity K D , or no affinity for these oligonucleotides at all under these assay conditions.

We next determined the structure of apo-Nsp9 COV19 ( Table 1 ). The apo-Nsp9 COV19 structure aligned closely to that of Nsp9 SARS (R.M.S.D of 0.57Å over 113 C α, Fig. 2A -C) . Like other Nsp9 homologues it exhibits an unusual fold that is yet to be observed outside of coronaviruses (Sutton et al., 2004) .

The core of the fold is a small 6-stranded enclosed β-barrel, from which a series of extended loops project outward (Fig 2A) . The elongated loops link the individual β-strands of the barrel, along with a projecting Nterminal β-strand and C-terminal α1-helix; the latter two elements make up the main components of the dimer interface ( Fig. 3A) . Two loops project from the open-face of the barrel: the β2-3and β3-4-loops are both positively charged, glycine rich, and are proposed to be involved in RNA-binding. The only protrusion on the enclosed barrel-side is the β6-7-loop; the C-terminal half of the β7-strand is an integral part of the fold's barrelcore but its other half extended outward to pair with the external β6-strand and create a twisted β-hairpin, cupping the α1-helix and interacting with subsequent C-terminal residues.

The arrangement of monomers within Nsp9-dimers is well-conserved in different viruses and is maintained within Nsp9 COV19 (R.M.S.D of 0.66Å over 226C α compared to the dimeric unit of Nsp9 SARS ). The main component of the intersubunit interaction is the self-association of the conserved GxxxG protein-protein binding motif ( Fig. 3C ) that allowed backbone van der Waals interactions between interfacing copies of the Cterminal α1-helix (Hu et al. 2017) . Here Gly-100 of the respective parallel α1-helices, formed complementary backbone van der Waals interactions. These interactions were replicated after a full helical turn by Gly-104 of the respective chains, thereby forming the molecular basis of the Nsp9 COVID19 dimer interface (Fig. 3C ). The 2fold axis that created the dimer ran at a ~15 ° angle through the GxxxG motif allowing the 14-residue helix to cross its counterpart (Fig. 4A) , the N-terminal turns of the helix were relatively isolated, only making contacts with counterpart protomer residues. In contrast, the C-terminal portions were encircled by hydrophobic residues, albeit at a distance that created funnel-like hydrophobic cavities either side of the interfacing helices ( Fig. 3A) . Strands β1, β6 and the protein's C-terminus served to provide a ring of residues that encircled the paired helices. The first 10 residues of Nsp9 COV19 exchanged across the dimer-interface to form a strand-like extension of β1 that ran alongside β6 from the other protomer (Fig. 3A) . The interaction these strands made did not appear optimal, indeed the remaining four C-terminal residues projected sideways across the dimer interface, inserting between the two strands while contributing a hydrophobic backing to the main helix.

In a separate crystallisation experiment we determined the structure of Nsp9 COV19 that included the Nterminal tag together with a rhinoviral 3C protease sequence (termed 3C-Nsp9 COV19 ). The 3C-Nsp9 COV19 crystal form diffracted to 2.05 Å resolution in space group P4 3 22, and had 1 molecule within the asymmetric unit, with the dimer being created across the crystallographic 2-fold axis.

Unexpectedly the high-resolution structure of 3C-Nsp9 COV19 diverged from that of the apo-Nsp9 COV19 (R.M.S.D 0.86 Å for the monomer and 2.23 Å when superimposing a dimer). The 3C sequence folded-around either side of the paired intersubunit helices to fill two funnel-like hydrophobic cavities (Fig. 2D , 3B, 4C, D).

Namely, 3C residues LEVL, inserted into the opposing cavities either side of the dimer interface and ran parallel to the paired GxxxG motif. Moreover, the 3C sequence formed additional β-sheet interactions with the N-terminus of the protein from the other protomer (Fig. 3B) . To accommodate the 3C residues the N-terminal strand residues moved outward by ~1.6 Å (residues 6-10). This movement allowed the N-terminus to increase the number of β-sheet interactions it formed with β6'. The β-barrel core of the fold remained unchanged but the increase in interactions between β1 and β6' served to exclude the C-terminus, prompting residues 106-111 to condense into a bent extension of the α-helix (Fig. 4A, B) . The subtle structural changes near the interacting GxxxG motifs (Fig. 3D ) are amplified at the periphery of the dimer resulting in ~ 6 Å shift in the β-barrel core ( Fig. 4C) 

When comparing apo-Nsp9 COV19 with 3C-Nsp9 COV19 the point where the N-terminal interface strand diverges is near Leu-9 (Fig. 3D) . Within the apo form it makes van der Waals interactions with the sidechains of Met-101, Asn-33 and Ser-105; this latter serine is important as it immediately follows the conserved proteinbinding motif ( 100 GMVLGS 105 ), while also specifically interacting with Gly104' from the opposing protomer.

Within the 3C-Nsp9 COV19 structure the extraneous LEVL residues insert at this point (Fig. 5B) , the hydrophobic side chains clasp either side of Ser-105 and allowed its hydroxyl group to form backbone hydrogen bonds to the glutamate within the extraneous sequence (Fig. 5B) . Meanwhile the C-terminal Leucine from the extraneous residues inserted behind the α-helix of the other protomer (Fig. 5B ). Cumulatively these changes allow for a ~5° rotation of the protomer subunits about the 2-fold axis compared to apo-Nsp9 COV19 (Fig. 4C ).

Most residues involved in protein-binding within the hydrophobic cavity and the structural changes needed to accommodate them appear broadly conserved amongst other Nsp9 viral homologues (red highlights in Fig. 5A ). The main exception to this is Ser-105, which is a Tyrosine in the distantly related Nsp9 HCoV and Nsp9 PEDV proteins (Ponnusamy et al., 2008) , (Zeng et al., 2018) . However, the N-terminal interface β-strand in these homologues is known to be involved in interface re-organisation of the subunits (Ponnusamy et al., 2008) and thus denotes other structural differences at this site.

Here we describe the structure of the recombinantly expressed Nsp9 COV19 as part of a global effort to characterise the virus causing a current global pandemic. Nsp9 is important for virulence in SARS-CoV (Miknis et al., 2009) . It remains to be understood whether Nsp9 COV19 plays a similar role in SARS-CoV-2, however the 97 % sequence identity suggests a high degree of functional conservation. The CoV Nsp9 proteins are seemingly obligate dimers comprising a unique fold that associates via an unusual α-helical GxxxG interaction motif. The integrity of this motif is considered important for viral replication (Miknis et al., 2009 ), leading to a proposal that disruption of the unusual dimer interface impacts on RNA binding and function (Hu et al., 2017) .

Mutation of the same interaction motif in the porcine delta coronavirus Nsp9 PDCoV also disrupted nucleotide binding capacity (Zeng et al., 2018) .

We describe the ability to produce homogenous Nsp9 COV19 which purifies as an obligate dimer, consistent with other Nsp9 proteins. Our preliminary nucleotide binding assays brought into question the RNA binding capacity of Nsp9 COV19 . The structure of the Nsp9 COV19 showed conservation of the unique Nsp9 fold when compared with homologues from SARS (Sutton et al., 2004) . Indeed, the topological fold was conserved as was the Nsp9 specific α-helical GxxxG dimerisation interface. This -helical interface is encircled by hydrophobic residues but the interface includes considerable cavities as observed previously . We made a serendipitous discovery in our 3C-Nsp9 COV19 structure, whereby the hydrophobic cavity captured the 3C cleavage sequence LEVL. The extraneous residues were tightly bound on all sides within the site situating themselves proximal to the conserved GxxxG motif. Coordination of the 3C sequence induced changes within interfacing residues, serving to both restructure key structural elements and cause a modest shift in subunit orientation.

At this stage it is unclear whether the bound residues within our structure have any bearing on the physiological function of Nsp9 COV19 . In the first instance this would seem unlikely, however our sequence is that of a rhinoviral 3C protease site and the SARS-CoV main protease cleaves consensus sequences following an LQ sequence (Zhu et al., 2011) . The bound 3C-LE residues have hallmarks of the LQ motif and are proximal to the highly conserved GxxxG motif. There are no obvious structural features to preclude, or select for, a Glu to Gln substitution within our bound sequence. Notably the M pro cleavage sequence occurs at multiple points throughout the CoV genome as the majority of viral proteins are released by its activity, thus it remains possible that Nsp9 may associate with unprocessed viral polyproteins retaining them near the viral RNA. Within our structure Met-101 provides contacts with the bound valine sidechain but the presence of Asn-33 nearby may also accommodate residues such as lysine at this position.

Peptide-binding assays will need to be developed to rigorously assess if other sequences are preferred by this putative binding site. Indeed, Nsp9 may be part of the viral replication-transcriptase complex so we may have serendipitously identified a protein-protein interaction interface for another viral or host protein.

In summary we have established a protocol for the production and purification of SARS-CoV-2 Nsp9 protein.

We determined the structure of the Nsp9 COV19 and described the conservation of the unique fold and dimerisation interface identified previously for members of this protein family. We also determined structure of Nsp9 COV19 in complex with a 3C sequence, although the significance of this is yet to be established. The structures we describe here could potentially be utilised in drug screening and targeting experiments to disrupt a dimer interface known to be important for coronavirus replication.

The identity of the peptide bound to Nsp9 COV19 raises follow-up questions that were not addressed within this study. Namely, it remains to be determined whether a physiological peptide, either of the same sequence or a 3C-peptide variant is also able to occupy this putative site. Further questions remain on whether this is a retention mechanism for preprocessed or post-processed polyproteins or another protein altogether and if so, what it binding affinity might be.

All methods can be found in the accompanying Transparent Methods supplemental file and available at https://data.mendeley.com/datasets/p8prnj8ckd/draft?a=754f9acc-31c3-4986-9e07-140f6bdd3d9f

The accession number for the atomic coordinates of the apo-Nsp9 COV19 with 3C-Nsp9 COV19 and associated diffraction data have been deposited at the protein databank (www.rcsb.org) with accession codes 6W9Q and 6WC1, respectively. Fluorescence polarization anisotropy assays were used to examine the possibility that Nsp9 COV19 could bind to labelled 17-mer and 10-mer single stranded oligonucleotides. The plot shows corrected anisotropy for each Nsp9 COV19 protein concentration, error bars represent the SD from the mean of triplicate measurements after 60 minutes incubation.

Cartoon representation of the monomeric units of a) apo-Nsp9 COV19 b) apo-Nsp9 SARS (Sutton et al., 2004) and c) a backbone alignment of the two structures. The COV19 structures are colored with β-strands in marine and the α-helix in wheat; the SARS structures are in teal and orange respectively. d) the bound peptide is highlighted in red.

Top-down views of the dimer interface highlighting the interaction helices for a) unbound Nsp9 COV19 in which the surface of the hydrophobic interface cavity is displayed labelled b) an equivalent representation of peptide-occupied 3c-Nsp9 COV19 dimer. c) Stick representation of the GxxxG protein-protein interaction helices at the dimer interface for apo-Nsp9 COV19 . d) C α backbone overlay of the Nsp9 COV19 interface in the apo and peptide-occupied states. The GxxxG motif residues are colored light purple.

A side-view of the N-and C-terminal structural elements at the dimer interface are shown for the a) apo-and b) peptide-occupied forms. c) Overlay of the Nsp9 COV19 dimer in the apo and peptide-bound forms indicating respective shifts in subunit orientation. The center-of-mass of the nonaligned subunit is depicted with a light-pink and dark-pink point respectively. d) Unbiased omit map contoured at 3.2σ near the hydrophobic cavities into which the exogenous bound peptide was refined. 

Cloning DAY -5 -Primer design and ordering Determine the cDNA sequence of the constructs you want to make. Design primers based on this sequence aiming for a 68oC annealing temperature. Add the LIC-specific extensions "cagggacccggt" to the 5' end of the fwd primer and "cgaggagaagcccggtta" to the 5' end of the rev primer. For example:

forward cagggacccggtaataatgaactgagtcctgtc reverse cgaggagaagcccggttattgaaggcgaactgtggcggc

Primers can be ordered from the Sigma Genosys website 0.025 ug quantity and desalting is sufficient. Delivery time can vary considerably allow plenty of time.

From glycerol stocks or a plate inoculate 250 -300 mL LB with 30ug/mL Kanamycin in a 1L flask with either DH5a or Nova Blue cells containing the pET-NKIb 3C/LIC empty plasmid. Grow to saturation overnight at 37oC 140rpm. Remove 5 uL of the PCR reaction and mix with 3 uL of 6x Loading dye. Run on a 0.8% Agarose gel and check the size of the bands. Ensure that they are correct (usually by looking for slight movements up/down compared to nearby samples).

Perform a Qiagen PCR-cleanup of the PCR reaction mixtures providing there is a single band only. If two or more bands occur a gelextraction may be necessary, but if they are clearly distinct these can be more easily separated by performing a larger number of PCRcolony screen

Treat the cut-vector and PCR'd insert with T4 polymerase in 1.5mL eppendorf tubes. The number of pg/pmol = (#of bp)x650. So for a 1000bp PCR-product we have 1000x650 = 650 000 pg/pmol = 650 ng/pmol which means for 0.2pmol of insert we require:

650ng/pmol x 0.2pmol = 130ng normal yields of Qiagen miniprep are 20ng/uL -100ng/uL so will require 8uL-1uL of PCR product.

Reagent Add PCR insert 0.2 pmol 10x buffer (NEBuffer #2) 2 uL 25 mM dATP 2 uL T4 polymerase (NEB 3U/uL) 0.8 uL MQ H2O Upto 20 uL Reagent Add Cut, gel-extracted vector 0.2 pmol 10x buffer (NEBuffer #2) 2 uL 25 mM dTTP 2 uL T4 polymerase (NEB 3U/uL) 0.8 uL MQ H2O Upto 20 uL 1) Incubate at 22oC for 30 minutes.

2) Denature the T4 polymerase by then incubating the reaction at 75oC for 10-20 minutes.

3) Spin down the evaporation.

Conjoin the vector and insert 1) Add 2 uL of the T4 treated insert to a 1.5 mL eppendorf tube 2) Add 1 uL of the T4 treated vector (~50ng/uL) 3) Incubate together at 22oC for 5 minutes.

MX2: a high-flux undulator microfocus beamline serving both the chemical and macromolecular crystallography communities at the Australian Synchrotron

Current developments in Coot for macromolecular model building of Electron Cryo-microscopy and Crystallographic Data

MX2: a high-flux undulator microfocus beamline serving both the chemical and macromolecular crystallography communities at the Australian Synchrotron

Current developments in Coot for macromolecular model building of Electron Cryo-microscopy and Crystallographic Data

Structural Analysis of

Porcine Reproductive and Respiratory Syndrome Virus Non-structural Protein 7alpha (NSP7alpha) and Identification of Its Interaction with NSP9

Rapid Progression to Acute Respiratory Distress Syndrome: Review of Current Understanding of Critical Illness from COVID-19 Infection

Structural basis for dimerization and RNA binding of avian infectious bronchitis virus nsp9

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Xds

Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix

Structure of Mpro 1 from COVID-19 virus and discovery of its inhibitors bioRxiv doi

Variable oligomerization modes in coronavirus non-structural protein 9

The epidemiology and pathogenesis of coronavirus disease (COVID-19) outbreak

The nsp9 replicase protein of SARS-coronavirus, structure and functional insights

The COVID-19 vaccine development landscape

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Structure of RNAdependent RNA polymerase from 2019-nCoV, a major antiviral drug target

Dimerization of Coronavirus nsp9 with Diverse Modes Enhances Its Nucleic Acid Binding Affinity

Peptide aldehyde inhibitors challenge the substrate specificity of the SARS-coronavirus main protease

Novel Coronavirus from Patients with Pneumonia in China

1/2 Σ i | I hkl, i -<I hkl > | / Σ hkl <I hkl > 2 R factor = ( Σ | |F o | -|F c | | ) / ( Σ |F o | ) -for all data except as indicated in footnote 3. 3 5% of data was used for the R free calculation Values

Transform the reaction into Novagen NovaBlue's 1) Add 10-15 uL of Nova Blue competent cells to 2ul of conjoined vector/insert mix.(Store the other 2ul at -20oC)

Incubate on ice for 30 minutes

Heat shock at 42oC for 30-40 seconds

Incubate on ice for 2 minutes

Add 250 uL of SOC media or LB without antibiotic

Allow the cells to recover by incubating at 37oC

Plate out on the appropriate antibiotic for the vector, leave overnight at 37oC

Setup a Taq sequencing mix such as: Reagent For 1 reaction For 25 reactions 10 x invitrogen Taq buffer ( -Mg2+) 2 uL 50 uL 50 mM invitrogen MgCl2 1 uL 20 uL 10 mM dNTPs 0.4 uL 10 uL Fwd primer (10 pmol/uL) 0.3 uL 7.5 uL Rev primer

Label 2 or 3 bacterial colonies from each samples plate

Aliquot 20 uL of the master PCR mix into each well of a 96-well thermowell tray

Take a sterile P10 tip and scrape half of each bacterial colony and put this into the 20 uL

After scraping all colonies use a pipette to mix briefly and seal the reactions

Run using a program such as: 96oC for 10 minutes 96oC for 1 minute | 55oC for 1 minute | Repeat steps 25x 68oC for 1 minutes/kbp | 68oC for 10 minutes 10oC for ever 6) When finished add loading dye to each PCR reaction and run half on a 0.8%-1.0% Agarose gel. Check the size of each band (remember the sequencing primers will add approx 170 extra bp compared to the original reaction)

Inoculate overnights from the remaining half-spot of the positive colonies

From the overnights purify plasmids by either using Qiagen mini-preps or in a 96-well block format. Transform into an expression cell-line like BL21 (DE3) or Rosetta (DE3)

MX2: a high-flux undulator microfocus beamline serving both the chemical and macromolecular crystallography communities at the Australian Synchrotron

The severe acute respiratory syndromecoronavirus replicative protein nsp9 is a single-stranded RNA-binding subunit unique in the RNA virus world

Current developments in Coot for macromolecular model building of Electron Cryo-microscopy and Crystallographic Data

Macromolecular structure determination using Xrays, neutrons and electrons: recent developments in Phenix

Funding for the work originated from the Australian Research Council Centre of Excellence for Advanced Molecular Imaging. This research was undertaken in part using the MX2 beamline at the Australian Synchrotron, part of ANSTO, and made use of the Australian Cancer Research Foundation (ACRF) detector.Additionally, we thank Dr. Geoffrey Kong of the Monash Molecular Crystallisation Facility for his assistance with crystallographic screening and optimization. We thank A. Riboldi-Tunnicliffe for assistance with data collection and J. Whisstock and G. Watson for advice on the manuscript. 

The authors declare no conflict of interest.This article contains supporting information online.

The plasmid was transformed into E. coli BL21(DE3) cells which were grown in Luria Broth at 37°C until reaching an Absorbance at 600nm of ~1.0 before being induced with 0.5mM Isopropyl βd-1-thiogalactopyranoside for 4 hours.Cells were harvested in 20mM HEPES pH 7.0, 150mM NaCl, 20mM Imidazole, 2mM MgCl2 and 0.5mM TCEP and frozen until required.Lysis was achieved by sonicating the cells in the presence of 1mg of Lysozyme and 1mg of DNAase on ice.The lysate was then cleared by centrifugation at 10,000xg for 20 minutes and loaded onto a nickel affinity column.Bound protein was washed extensively with 20 column volumes of 20mM HEPES pH 7.0, 150mM NaCl, 0.5mM TCEP before being eluted in the same buffer with the addition of 400mM Imidazole.For His-tag removal samples were incubated with precision 3c protease overnight at 4°C.All samples were subjected to gel filtration (S75 16/60; GE Healthcare) in 20mM HEPES pH 7.0, 150mM NaCl before being concentrated to 50mg/mL for crystallization trials.

Nsp9COV19 crystallized in 2.0-2.2M NH4SO4 and 0.1M Phosphate-citrate buffer pH 4.0.Crystals of the His-tag samples grew with rectangular morphology in space group P4122, however if the His-tag was removed the crystals grew in space group P6122 form with hexagonal morphology.All diffraction data were collected at the Australian synchrotron's MX2 beamline at the Australian synchrotron (Aragao et al., 2018 ) (see Table 1 for details).Data were integrated in XDS (Kabsch, 2010) , processed with SCALA, phases were obtained through molecular replacement using PDB 1QZ8 ) as a search model. Subsequent rounds of manual building and refinement were performed in Coot and PHENIX .

To examine the RNA-binding affinity, an 18-point serial dilution (212-0 µM) of Nsp9COV19 was incubated with 1 nM 5'-Fluorescein labelled 17mer poly-U single-stranded RNA (Dharmacon GE, USA) or 10mer PolyT single-stranded DNA (IDT, USA) in assay buffer (20mM HEPES pH 7.0, 150mM NaCl, 2mM MgCl2) at room temperature.The assay was performed in 96-well non-binding black plates (Greiner Bio-One), with fluorescence anisotropy measured in triplicate using the PHERAstar FS (BMG) with FP 488-520-520 nm filters.The data was corrected using the anisotropy of RNA sample alone, then fitted by a one-site binding model using the Amax and KD were used as fitting parameters and nonlinear regression was performed using GraphPad Prism. Measurements were taken after 60 minutes incubation between protein and RNA.