Research ArticleCORONAVIRUS

Hallmarks of Alpha- and Betacoronavirus non-structural protein 7+8 complexes

See allHide authors and affiliations

Science Advances  03 Mar 2021:
Vol. 7, no. 10, eabf1004
DOI: 10.1126/sciadv.abf1004
  • Fig. 1 Candidate structures in agreement with the observed stoichiometries and topologies.

    (A) Replicase polyproteins pp1a (nsp1-11) or pp1ab (nsp1-16) undergo processing by two internal proteases and subsequently release nsps that assemble into the CoV replication transcription complex residing in double-membrane vesicles (DMVs) within the cell. (B) For the full-length heterotetramer, an isolated structure does not exist. However, from the larger SARS-CoV nsp7+8 hexadecamer (22) [Protein Data Bank (pdb) 2ahm], two conformer subcomplexes of nsp7+8 (2:2), T1 and T2, can be extracted. Both conformers constitute a head-to-tail interaction of two heterodimers by an nsp8:nsp8 interface. Notably, nsp8 in T1 is more extended, containing an almost full-length amino acid sequence (2 to 193), while in T2, the nsp8 N-terminal 35 to 55 residues are unresolved. (C) For the trimeric complexes, the only deposited structure is FIPV nsp7+8 (2:1) trimer (23) (pdb 3ub0).

  • Fig. 2 Mpro-mediated processing of precursor protein constructs.

    SDS–polyacrylamide gel electrophoresis (SDS-PAGE) analysis of Mpro (nsp5)-mediated processing and generation of CoV nsp7+8 complexes with authentic N and C termini from (A) polyprotein precursors nsp7-8 and (B) nsp7-11. (A) SDS-PAGE showing the purified Mpro (nsp5-His6): lanes 1, 5, 9, 13, and 17; nsp7-8-His6: lanes 2, 6, 10, 14, and 18; Mpro-mediated cleavage reaction: lanes 3, 7, 11, 15, and 19; enriched nsp7+8 complexes: lanes 4, 8, 12, 16, and 20. (B) SDS-PAGE showing the purified Mpro (nsp5-His6): lane 1; nsp7-8-9-10-11-His6: lane 2; Mpro-mediated cleavage reaction: lane 3. Lane M, marker proteins with molecular masses in kilodaltons indicated to the left. Black arrows on the right indicate the identities of proteins generated from precursor proteins by Mpro-mediated cleavage. Gray arrowheads indicate aberrant in vitro cleavage products of nsp8 as observed previously for SARS-CoV (24). +/– indicate the presence or absence of the respective proteins.

  • Fig. 3 Native MS of nsp7+8 complexes of seven CoVs representing five different CoV species.

    Representative mass spectra showing distinct nsp7+8 complexation patterns that were classified into the three groups A, B, and AB. Complex formation triggered by Mpro (M)–mediated cleavage of 15 μM nsp7-8-His6 or MERS-CoV nsp7-11-His6 precursors in 300 mM AmAc, 1 mM dithiothreitol (DTT) (pH 8.0). (A) SARS-CoV and (B) SARS-CoV-2 from group A forming nsp7+8 (2:2) heterotetramers (red), (C) FIPV and (D) TGEV from group B forming nsp7+8 (2:1) heterotrimers (blue), and (E) HCoV-229E and (F) PEDV from group AB forming both complex stoichiometries. (G) MERS-CoV, also from group AB, produced from an nsp7-11-His6 precursor, additionally results in several processing intermediates that allow for an estimation of relative cleavage efficiencies at different cleavage sites. All groups form nsp7+8 (1:1) heterodimers as intermediate state (green).

  • Fig. 4 Mass and oligomeric state of nsp7-8 precursors.

    Native MS of nsp7-8 precursors (A to G) sprayed at 18 μM from 300 mM AmAc (pH 8.0) and 1 mM DTT. Dominant charge envelope is highlighted (blue box). Labeled are charge states and molecular mass. (C) Inset shows mass heterogeneity in FIPV nsp7-8. The experimental molecular weight Mexp of the precursors agrees with the sequence-derived theoretical Mtheo (table S2). Only FIPV nsp7-8 contained two mass species separated by ~110 Da. This heterogeneity was attributed to the precursor’s central nsp8 domain following Mpro processing. Assignment to an amino acid variation failed but potentially was the result of codon heterogeneity in the plasmid. Nevertheless, both forms behaved identically and we refrained from further optimization.

  • Fig. 5 Gas-phase dissociation reveals complex topology.

    CID-MS/MS product ion spectra (A and C) and dissociation pathways and topology maps (B and D) for HCoV-229E nsp7+8 (2:1) heterotrimers and (2:2) heterotetramers are shown. With increasing collisional voltage, protein complexes are successively stripped from their subunits revealing alternative dissociation pathways. The remaining dimeric species expose direct subunit interactions in the nsp7+8 complexes (gray boxes). Charge states are labeled. (A and B) The heterotetramers (2:2) undergo two consecutive losses, resulting in dimeric product ions of nsp7+8 (1:1) and nsp82. These products indicate that nsp7:nsp8 and nsp8:nsp8 have direct interfaces in heterotetramers. (C and D) HCoV-229E heterotrimers dissociate into the dimeric products nsp7+8 (1:1) and nsp72, indicating direct interfaces between nsp7:nsp8 and nsp7:nsp7 in heterotrimers. All CoV heterotrimers follow similar dissociation pathways; also, all CoV heterotetramers follow a common dissociation route, allowing a topological reconstruction of two distinct complex architectures (fig. S2).

  • Fig. 6 Chemical cross-linking of prepurified nsp7+8 complexes shows species-specific complex formations.

    MALDI-MS of nsp7+8 complexes from FIPV (A) and HCoV-229E (B) stabilized with 0.15% glutaraldehyde (GA) for 25 min at 4°C. Mass spectra are background-subtracted and mass species of interest are labeled according to their stoichiometry with symbols for nsp7 (yellow) and nsp8 (green). Peak areas calculated from Gaussian fits of the respective peak for 1:1 heterodimers (green), 2:1 heterotrimers (blue), and 2:2 heterotetramers (red). Masses are higher in cross-linked samples owing to additional glutaraldehyde molecules. Mass spectra were not calibrated. Each spectrum shown generated from three MALDI spots. Signals above 50,000 m/z, except for the HCoV-229E heterotetramer, are low abundant and likely due to over–cross-linking. (C) SDS-PAGE analysis of chemically cross-linked HCoV-229E, FIPV, and SARS-CoV nsp7+8 complexes; 5 µg protein of nsp7+8 complexes cross-linked with 10 μM BS3 at 37°C for 30 min. Lanes 1, 3, 5: nsp7+8 complexes not treated with BS3 (−); lanes 2, 4, 6: nsp7+8 complexes treated with BS3. Lane M, marker proteins; molecular masses in kilodaltons are indicated to the left. Black arrows indicate the different oligomeric states of the nsp7+8 complexes obtained by cross-linking.

  • Fig. 7 DLS and SAXS reveal oligomeric state of SARS-CoV-2 nsp7+8 at higher protein concentrations.

    (A) Comparison of molar and mass concentration based on the molecular weight of the nsp7-8 polyprotein. Mass ranges for native MS and DLS/SAXS are indicated. (B) Two exemplary DLS plots (1 and 15 mg/ml) and (C) how the hydrodynamic radius (R0) develops with increasing protein concentration. Theoretical radii (R0,theo) of heterotetramer and hexadecamer candidate structures indicated (red dashed lines). Data points depicting R0 with increasing complex concentrations. Error bars show SD. (D) SAXS curves collected at 1.2 to 47.7 mg/ml; (E) radius of gyration (Rg), error bars correspond to SD; and (F) molecular weight (MW), error bars correspond to credibility interval, estimated from the SAXS data. Both plots stabilize with increasing concentration on values that are in agreement with the R0,theo of the T1 nsp7+8 (2:2) heterotetramer. (G) Fit of the curve computed from T1 tetramer (red line) to the SAXS data collected at 1.2 mg/ml (blue dots with error bars). (H) Fits to all SAXS data. Experimental data (blue with experimental errors) for SARS-CoV-2 nsp7+8 complexes fitted with a mixture of T1 and hexadecamer (red), or T1 and dimer of T1 (green).

  • Fig. 8 Candidate structures and sequence conservation.

    Candidate structures for nsp7+8 heterotetramer and heterotrimer are chosen based on experimental stoichiometry and topology in solution and exhibit similar binding sites in nsp8, BS I and BS II. (A) Two conformers of SARS-CoV nsp7+8 (2:2) heterotetrameric subcomplexes (pdb 2ahm), T1 (left) and T2 (middle), and FIPV nsp7+8 (2:1) heterotrimer (right, pdb 3ub0). For BS II, residue (res) numbers are given (see also fig. S3). Candidate complexes involving similar conserved residues (red) in the nsp8 BS II are shown here for (B) SARS-CoV T1 and (C) T2 as well as (D) the FIPV heterotrimer. (E) Sequence alignment of BS II contact sites displayed for seven CoVs. Specific contact sites (red) exhibit sequence conservation well in line with the complexation groups determined by native MS. (F) Unique heterotetramer contact shown in SARS-CoV T2 (G) replaced by a neighboring amino acid in a homology model of HCoV-229E. (H) Unique heterotrimer contact shown in FIPV heterotrimer structure. Insets show magnifications with contact distances. (I) Cryo–electron microscopy (cryo-EM) of nsp7+8+12+13 (1:2:1:1) polymerase complex (pdb 6xez) (18). (J) Zoomed-in view of nsp8a and nsp8b BS II and its amino acids in contact with nsp12 thumb domain (brown), nsp13.1, and nsp13.2 (blue).

  • Fig. 9 Proposed model for nsp7+8 complex formation.

    (A) For complexes of group A, heterodimers form via nsp8 BS I, which quickly dimerize via BS II into a heterotetramer. A theoretic route via a preformed nsp8 scaffold is unlikely to play a role in heterotetramer formation since no nsp7+8 (1:2) intermediates are observed for complexes of SARS-CoV or SARS-CoV-2. Moreover, nsp7 and nsp8 occupy neighboring positions in the replicase polyproteins, thus favoring their interaction (in cis) at early stages in the infection cycle (when intracellular viral polyprotein concentrations are low) over intermolecular interactions between different replicase polyprotein molecules as is also evident from the low dimerization ability of the precursors. (B) For group B complexes, we propose the formation of a heterodimer intermediate via nsp8 BS I or BS II and subsequent recruitment of a second nsp7, resulting in an nsp7+8 (2:1) heterotrimer. This is also supported by the relatively high peak fractions of heterodimers detected. Group AB complexes can use both complexation pathways. In line with this, the proteins also produce a relatively high heterodimer signal but, ultimately, prefer to form heterotetramers rather than heterotrimers.

Supplementary Materials

  • Supplementary Materials

    Hallmarks of Alpha- and Betacoronavirus non-structural protein 7+8 complexes

    Boris Krichel, Ganesh Bylapudi, Christina Schmidt, Clement Blanchet, Robin Schubert, Lea Brings, Martin Koehler, Renato Zenobi, Dmitri Svergun, Kristina Lorenzen, Ramakanth Madhugiri, John Ziebuhr, Charlotte Uetrecht

    Download Supplement

    This PDF file includes:

    • Tables S1 to S6
    • Figs. S1 to S3

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article