Structural characterization of NiV N 0 P in solution and in crystal.

Transcription

1 Supplementary Figure 1 Structural characterization of NiV N 0 P in solution and in crystal. (a) SAXS analysis of the N P 50 complex. The Guinier plot for complex concentrations of 0.55, 1.1, 1.6 and 2.4 mg.ml -1 shows no evidence for aggregation or intermolecular interactions. The averaged intensity at zero angle corrected for protein concentration (I(0)/C) value of 46 ± 1 kda is in good agreement with the theoretical molecular mass of the heterodimeric complex (45,613 Da). (b) Superposition of the experimental SAXS curve in black (2.4 mg.ml -1 ) with the theoretical curve (in red) back calculated for the averaged ab initio bead model (20 independent models; normalized spatial discrepancy (NSD) < 1.0) generated with DAMMIN 28 and DAMAVER 29. (c) The heterodimeric N P 50 complex has a bean shape in solution. The ab initio bead model generated from SAXS data (in grey) accommodates a single N P 50 copy from the crystal. (d) Comparison of the 2D 1 H- 15 N heteronuclear single-quantum coherence (HSQC) NMR spectra of free P 100 (in red) and of P 100 bound to N (in blue). (e) Electrostatic surface potential of NiV and RSV N (PDB code 2WJ8; ref. 5) proteins at ± 5 kte -1 : blue (basic), white (neutral) and red (acidic). The hypothetical RNA binding

2 site is indicated in NiV N, and bound RNA in the RSV complex is shown in orange. (f) Western blot using inhouse henipavirus specific rabbit anti-n antibody with cell lysates from Nipah infected Vero E6 cells lysed 48h post-infection (NiV, left lane) compared to uninfected cells (NI, right lane). Theoretical molecular mass of the NiV N protein is 58,168 Da.

3 Supplementary Figure 2 NiV P 100 is globally disordered in solution but contains fluctuating α-helical elements. (a,b). Molecular size. The hydrodynamic radius of 2.6 ± 0.5 nm measured by SEC (a) and the radius of gyration of 3.1 ± 0.5 nm measured by SAXS (b) are larger than expected for a compact domain of this size. SAXS profiles were recorded at 0.3, 0.5 and 0.6 mg.ml -1. (c) NMR spectroscopy. The poor chemical shift dispersion of amide 1 H resonances in the heteronuclear single quantum coherence (HSQC) NMR spectrum (Supplementary Fig. 1d) is typical of disordered proteins. The secondary structure propensity (SSP) parameter calculated from C α and C β secondary chemical shifts indicates the presence of fluctuating helices (red boxes). The position of helices in the N-terminal region of P in the crystal structure of the N P 50 complex is shown below. (d) The N 0 -binding region of Paramyxovirinae P contains two conserved motifs (Soyuz1 and Soyuz2) 23. Conserved residues are colored according to their properties: acidic residues in violet, basic in red, hydrophobic in blue, polar in green and glycine in orange. G10 and I17 (larger letters) were mutated into arginine in order to destabilize the N 0 -P complex.

4 Supplementary Figure 3 The crystallographic asymmetric unit contains three copies of the N P50 complex. (a) Initial plate cluster used for microseeding. (b) Typical crystal obtained for the N P50 from microseeding and used for data collection. (c) Overall view of crystal packing of the N P50 complex. The three copies of the complex in the asymmetric unit are colored in blue and red. The packing shows no indication of specific oligomerization of N in the crystal. (d,e). Experimental selenium SAD phased density map of P50 (d) and N0 (e) in the N P50 complex. The final model after refinement is superposed on the map at contour level 1 σ. Only one N0-P copy from the asymmetric unit is represented. (f,g) Final 2Fo-Fc difference electron density map of P50 (f) and N0 (g) in the N P50 complex contoured at 1 σ. (h). Comparison of the different copies of the N P50 complex present in the asymmetric unit. Two copies are well defined in the SAD phased map, whereas NNTD of the third one is less densely packed in the crystal and the corresponding electron density is ill-defined. (i) Structural overlay of the three copies of N. Slightly different orientations of secondary structure elements in NNTD suggest that this region of the protein is dynamic, at least in the absence of RNA.

5

6 Supplementary Figure 4 Comparison of NNVs N 3D fold. (a) Conserved architecture of the N proteins: NiV, Nipah virus (Paramyxoviridae - Paramyxovirinae); RSV, respiratory syncytial virus (Paramyxoviridae - Pneumovirinae) (PDB code 2WJ8; ref 5); BDV, Borna disease virus (Bornaviridae) (PDB code 1N93; ref 19); VSV, vesicular stomatitis virus (Rhabdoviridae) (PDB code 2GIC; ref 7). Pairwise structural alignments led us to define four subdomains in the N core, namely N NTD1, N NTD2, N NTD3 and N CTD. The structures are similarly oriented and colored according to subdomains. VSV N contains an additional C-terminal subdomain (N CTD2 in purple) comprising three α-helices, which form the surface of binding of the C-terminal domain of P. (b-d). Superpositions of subdomains from the different proteins. The NiV N subdomains are shown in the same colors as in panel a and the corresponding subdomains from the other viruses are shown in grey: N NTD1 (b), N NTD3 (c) and N CTD (d). NNV N proteins differ mainly in the relative orientations of these three domains and the structure of the variable N NTD2 region. (e) Pairwise comparison of NiV N and human RSV N. Structure-based sequence alignment of NiV N (PDB code 4CO6) and RSV N (PDB code 2WJ8; ref 5). Secondary structure elements and residue numbering of NiV N are indicated above and secondary structure elements of RSV are indicated below.

7 Supplementary Figure 5 RNA grasp and hinge motions. (a) Normal mode analysis (NMA) and hypothetical hinge motion in NiV N. Elastic network NMA was carried out on the N 0 molecule to explore its motions 45. Representative models of the two lowest-frequency modes show motions of N NTD relative to N CTD with a hinge at the junction between the two domains (red circle) in agreement with a mechanism of closure of the RNA binding cavity. (b) The second lowest mode (mode 8) captures a rotation of N NTD, which agrees with the hypothetical closure of the protein around the RNA. The initial model (in grey) superposes with the crystal structure of NiV N 0 in blue. (c) After displacement along mode 8, the model (in wheat) superposes with RSV N structure taken from the N-RNA complex (in light blue). (d) Close-up of RSV N-RNA complex showing that helices α N6a and α N9 pack against the RNA molecule. G241 and G245 are shown in yellow. (e) Close-up of NiV N in the N 0 -P crystal structure (left panel) and in a hypothetical closed conformation (right panel). The RNA molecule (in grey) is positioned against N CTD as in RSV NC. Several residues of conserved motif 3 and conserved D 254 interfere with base 1 binding in the closed form.

8 Supplementary Table 1. Summary of conserved sequence motifs and their roles in the structure and function Motifs Residues Functional or structural roles Helix α N2 Hydrophobic core of NTD C-terminal of Helix α N5 (K178) Hydrophobic core of NTD ; Loop α N5/ α N6 RNA binding RNA binding and interaction (Arg192, Arg193, motif KYxQqxrx) Helix α N8 Hydrophobic packing against NTD core and with helix η C2 (Arg228) Helix α N9 Hinge region (Gly263) and RNA binding interaction (Tyr258) Helices α C1, η C1 and α C2 P binding Helix η C2 Interactions with helices α N8 and α N9 (contact between CTD and NTD) Loop η C2 /α C3 and helix α C3 Hydrophobic core of CTD Loop α C3 /α C4 RNA binding Helix α C3 Hydrophobic core of CTD and P binding

9 Supplementary Table 2. Summary of domain and subdomain localization in NiV N sequence Subdomains Residues NT ARM 1-45 N NTD N NTD N NTD N CTD CT ARM N TAIL

10