NMPylation and de-NMPylation of SARS-CoV-2 nsp9 by the NiRAN domain

Figures

• 

  Figure 1.. NiRAN-mediated NMPylation of nsp9. (A) Cryo-EM structure of the SARS-CoV-2 RdRp-helicase complex bound to a non-native nsp9 with a two-residue extension at the N-terminus (21). Left. Overall structure of the complex; PDB: 7CYQ; nsp13 helicase is not shown. Proteins are shown as colored molecular surfaces (as shown in the key) and RNA as black cartoon. The color coding corresponds to the figures throughout this manuscript unless otherwise specified. Right. Zoom in on the active site of the NiRAN domain (AS2) with GDP-Mg2+ (lime carbon atoms and magenta sphere, respectively). Side chains of key conserved residues from the pre-AN, BN and CN motifs that were substituted in this work are shown as sticks. Four N-terminal residues of nsp9 (GSNN) are shown; the cryo-EM difference density for Gly and Ser residues from EMDB: 30504 is shown (gray mesh). Structural figures were prepared with Coot (61), UCSF ChimeraX 1.2 and PyMOL Molecular Graphics System, version 2.4.1, Schrodinger, LLC. (B) Conservation of residues at the N-terminus of nsp9 and four conserved NiRAN motifs in alpha-, beta-, gamma- and deltacoronavirus genera. (C) Mutations in the NiRAN active site and nsp9 N-terminal GS…

• 

  Figure 2.. Effects of nucleotides on nsp9 modification. (A) Nucleotide binding to the NiRAN active site. The Mg2+ ion is shown as magenta sphere and key NiRAN residues - as sticks. Left, ADP (cyan carbon atoms) from PDB:6XEZ. Right, RDP (lime carbon atoms) modeled in place of ADP; the C1′ cyano group of RDP could clash with the side chain of nsp12 His75. (B) Unlabeled natural NTPs compete with [α32P]-GTP for transfer to nsp9. (C) GTP analogs and PPi inhibit NMPylation, but ppGpp and RTP do so less effectively; all nucleotides were present at 0.5 mM. In (B) and (C), NMPylation efficiency was compared to that observed in the absence of competitors, set at 1, and is shown as mean ± SD (n = 3).

• 

  Figure 3.. Removal of the GMP moiety in the presence of pyrophosphate. (A) Top: Pyrophosphorolysis of the nascent product RNA (pRNA; red). The active site (marked by the position of the catalytic Mg2+; magenta sphere) consists of two sub-sites: the P-site (product; yellow) and the A-site (acceptor; cyan). Following nucleotide addition, the pRNA 3′ end is bound in the A-site in the pre-translocated state. Upon forward translocation, the 3′ end moves to the P-site and the incoming substrate NTP (blue) can bind to the A-site through base pairing with the acceptor base of the RNA template strand (black). The pre-translocated state is sensitive to pyrophosphorolysis, which generates the NTP product and a one-nt shortened pRNA. Bottom: SARS-CoV-2 RdRp transcription complex assembled on the 5′ Cy5.5-labeled hairpin, which comprises both the template and the product RNAs, is completely resistant to PPi even in the absence of NTPs. (B) GMP removal from nsp9 in the presence of PPi. (C) The hand-over assay in which pre-GMPylated nsp9 is incubated with WT or mutant nsp12 variants. Signal intensity was compared to that observed with nsp9 incubated with buffer (set at 1) and is shown as mean ± SD (n = 3).

• 

  Figure 4.. Effects of bisphosphonates on nsp9 modification. (A) Cryo-EM structure of the SARS-CoV-2 RdRp with PPi bound to the NiRAN domain AS2. PDB: 7AAP (39). (B) Inhibition of NMPylation by PPi and bisphosphonates; structures are shown below each compound. (C) Bisphosphonates do not promote the removal of NMP from modified nsp9. In (B) and (C), signal intensity was compared to that observed in the absence of added ligands (set at 1) and is shown as mean ± SD (n = 3).

• 

  Figure 5.. A model for nsp9 cycle; see text for details. AS1 and AS2 are shown as yellow circles, with the key catalytic residues indicated.