Orthoparamyxovirinae C Proteins Have a Common Origin and a Common Structural Organization

Abstract

The protein C is a small viral protein encoded in an overlapping frame of the P gene in the subfamily Orthoparamyxovirinae. This protein, expressed by alternative translation initiation, is a virulence factor that regulates viral transcription, replication, and production of defective interfering RNA, interferes with the host-cell innate immunity systems and supports the assembly of viral particles and budding. We expressed and purified full-length and an N-terminally truncated C protein from Tupaia paramyxovirus (TupV) C protein (genus Narmovirus). We solved the crystal structure of the C-terminal part of TupV C protein at a resolution of 2.4 Å and found that it is structurally similar to Sendai virus C protein, suggesting that despite undetectable sequence conservation, these proteins are homologous. We characterized both truncated and full-length proteins by SEC-MALLS and SEC-SAXS and described their solution structures by ensemble models. We established a mini-replicon assay for the related Nipah virus (NiV) and showed that TupV C inhibited the expression of NiV minigenome in a concentration-dependent manner as efficiently as the NiV C protein. A previous study found that the Orthoparamyxovirinae C proteins form two clusters without detectable sequence similarity, raising the question of whether they were homologous or instead had originated independently. Since TupV C and SeV C are representatives of these two clusters, our discovery that they have a similar structure indicates that all Orthoparamyxovirine C proteins are homologous. Our results also imply that, strikingly, a STAT1-binding site is encoded by exactly the same RNA region of the P/C gene across Paramyxovirinae, but in different reading frames (P or C), depending on which cluster they belong to.

Keywords: Paramyxoviridae; overlapping genes; protein structure; viral evolution; virulence factor.

Figure 1

Architecture of the Tupaia paramyxovirus (TupV) genome and expression mechanisms of its P, V, and C proteins. (A) The upper panel shows a schematic representation of TupV genome, containing from the 3′ end to the 5′ end, the genes coding for the N, P, M, F, H, and L proteins. The middle panel shows the architecture of the mRNAs generated from the P gene by editing at the nucleotide 2821. The mRNA coding for the P protein is unedited, whereas that coding for the V protein has one G inserted at the editing site. Both mRNA contain the 3′ and 5′ UTR and are processed at both extremities with a CAP structure at the 3′ end and a polyA tail at the 5′ end. The lower panel shows the proteins expressed from these mRNAs. The P and V proteins are expressed by translation initiation at the first AUG codon of their respective mRNA, whereas the C protein is expressed by alternative translation initiation at the next downstream AUG codon from both mRNAs. (B) Excerpt of the mRNA coding sequence (of both P and V mRNAs) showing the first AUG initiation codon used for the translation of P or V proteins and the second AUG codon, located in the +1 frame, used for the translation for the C protein.

Figure 2

TupV C protein sequence analysis. (A) Multiple sequence alignment of Narmovirus C protein. Members of the Narmovirus genus and their UniProt accession number: TupV—Tupaia paramyxovirus—Q9WS38; NarV—Nariva narmovirus—B8XH61; BaV-1—Myodes narmovirus (Bank Vole virus-1)—A0A2H4PJ60; DenaV—Denalis virus—UQM99579; DeneV—Denestis virus—UQM99571.1; MossV—Q6WGM3; MeleV—Meleucus virus—UQM99623. (B) HCA (Hydrophobic Cluster Analysis) plot. (C) Secondary structure and disorder predictions. The upper part shows the secondary structure prediction from PSIPRED (the position of predicted helices is indicated by the orange cylinders with the residue numbers shown above). The middle part shows the predictions from individual predictors as obtained in the output of the D-score script. The lower part shows a consensus disordered prediction (D-score) calculated as described in [25]. The threshold to distinguish between ordered and disordered regions was set at 0.5.

Figure 3

Crystal structure of TupV C protein C-terminal domain (C_Δ53). (A) SEC-MALLS of TupV C_Δ53 and C_FL. The lines show the chromatograms monitored by differential refractive index measurements. The crosses indicate the molecular mass across each elution peak calculated from static light scattering and refractive index, and the numbers indicate the weight-averaged molecular mass (kDa) with standard deviations (the molecular mass calculated from the aa sequence are 17,807 Da for C_FL and 11,897 Da for C_Δ53, respectively). (B) Plot of the hydrodynamic radius measured by SEC as a function of the molecular mass measured by MALLS and RI. The blue and red circles show data taken from [42] for globular proteins in native or chemically unfolded forms, respectively. (C) Electron density map of one protomer in the asymmetric unit contoured at 1.8 s and stick representation of the TupV C_Δ53 protein in the crystal structure (PDB ID: 8BJW). (D) Cartoon representation of TupV C_Δ53 in the crystal. The N and C-terminal residues of C_Δ53 are indicated and helices are named B to F. (E) Close-up of the C-terminal domain showing residues conserved among Narmoviruses (see also Figure 2A). (F) Cartoon representation of the first mode of interaction between protomers within the asymmetric crystal unit. (G) Cartoon representation of the second mode of interaction between protomers within the asymmetric crystal unit. The orange protomer is in the same orientation as in panel (F). (H) Surface representation showing the assembly of three protomers within the asymmetric crystal unit by the two modes of interaction. The color is the same as in panel (F,G). (I) Surface representation in two orthogonal orientations of one asymmetric unit containing six protomers. The same three protomers of panel (H) are shown with the same color code, the protomer associated with the one in teal by the first mode of interaction is shown in olive, while the other two are shown in pale green. (J) Surface representation of the tubular assembly seen in the crystal packing. The same four protomers described in panels (F–I) are shown with the same color code, and all the others are shown in pale green. (K) Cartoon representation of the crystal packing showing the side-by-side assembly of the tubes described in panel (J) with the same four protomers using the same color code.

Figure 4

Superposition with the C-terminal domain of Sendai virus C protein. (A) Superposition in two orthogonal orientations of the C-terminal domain of TupV C protein with the C-terminal domain of SeV C protein complexed with the BRO-1-like domain of the protein Alix (PDB ID: 6KP3). TupV C is shown in orange, SeV C in green and the BRO-1-like domain of Alix in a grey ribbon. (B) Superposition in two orthogonal orientations of the C-terminal domain of TupV C protein with the C-terminal domain of SeV C protein in complex with the N-terminal domain of STAT1 (PDB ID: 3WWT). TupV C is shown in orange, SeV C in green, and STAT1 in a grey ribbon. (C) Superposition in two orthogonal orientations of TupV C (in orange) and SeV C (in 3WWT) taken from its complex with STAT1 (in green). (D) Close up of the superposition of TupV C and SeV C (in 3WWT) showing the orientation of conserved residues in both proteins. The labeled residues are those of TupV C protein. STAT1 is shown in grey ribbon. (E) Sequence alignment of Narmovirus and Respirovirus C proteins. Members of the Narmovirus genus and their UniProt accession number are the same as in Figure 2A, members of the Respirovirus genus are: SeV-Sendai virus-Q38KG9; HPIV1-human parainfluenza virus 1-Q8QT30; HPIV3-human parainfluenza virus 3-Q81077; BPIV1-bovine parainfluenza virus 3-P06164. The orange cylinders above the MSA indicate the location of helices in the crystal structure of TupV C. The black box indicates residues of SeV C that interact with STAT1.

Figure 5

Structural alignments. (A) Superposition of TupV C (8BJW.pdb) and SeV C (taken from 3WWT.pdb) crystal structures. The non-superposable parts of TupV and SeV C are, respectively in light blue and light orange. (B) Superposition of the crystal structure of TupV C with an AlphaFold model of NiV C (aa 100–166). The averaged DDT score for this model of NiV C is 92.6, which indicates a highly reliable prediction. The additional C-terminal helix of NiV C is circled with a dotted line. (C) Superposition of the crystal structure of TupV C with an AlphaFold model of MeV C (aa 101–186). The averaged pLDDT score for this model is 86.0, also indicating a reliable prediction. The additional C-terminal helices of MeV C are shown and circled with a dotted line. (D) Superposition of (i) the helical domain of TupV C taken from the crystal structure (8BJW.pdb), (ii) the corresponding regions of SeV C taken from the crystal structure (3WWT.pdb), (iii) an AlphaFold model of MeV C and (iv) an AlphaFold model of NiV C. The N and C-termini and the name of the helices in TupV C (D–F) are labeled. (E) Superposition of the Alphafold models of the NiV and MeV C protein domains.

Figure 6

SEC MALLS experiments at different TupV C_D53 protein concentrations. 50 mL of protein solution were injected onto a Superdex 75 column at different initial concentrations: 3.3 mg.mL⁻¹ in dark red, 8.5 mg.mL⁻¹ in red, 8.8 mg.mL⁻¹ in blue, 20.0 mg.mL⁻¹ in dark blue and 23.5 mg.mL⁻¹ in black. The lines show the chromatograms monitored by refractive index and the crosses show the molecular masses calculated at each time from the light scattering intensity and refractive index. Numbers show the calculated weight average molecular masses (M_w).

Figure 7

Small-angle X-ray scattering experiments. (A) SEC-SAXS analysis of TupV C_Δ53 protein. The black line shows the scattering at zero angles (I(0)), which is proportional to both molecular mass and concentration, as a function of the frames recorded at regular time intervals. The red crosses show the radius of gyration calculated from the Guinier plots at different time intervals. (B) SEC-SAXS analysis of TupV C_FL protein. The black line and the blue crosses are as in panel (A). (C) Averaged scattering curves. The lines show the scattering profiles obtained for C_Δ53 (in red) and C_FL (in blue) by averaging the individual profiles recorded throughout the SEC elution peak shown in panels (A,B). (D) Guinier plots. The upper panel shows the Guinier plots for C_Δ53 (in red) and C_FL (in blue). The lower panels show the plots of the normalized residuals. (E) Normalized Kratky plots. (F) Comparison of TupV C_Δ53 protein in solution and in crystal. The experimental SAXS curve obtained for TupV C_Δ53 protein (black line) is compared to the theoretical curve calculated with CRYSOL for one monomer (green line) and one dimer (orange line) extracted from the crystal structure (shown in cartoon representations in green and orange, respectively). The curves were scaled in order to simply compare their shape and not to take into account differences in molecular mass.

Figure 8

Conformational ensemble modeling of TupV C_Δ53 protein from SAXS data. (A) Description of the initial ensemble, which is comprised of 10,000 conformers in which the regions corresponding to the helix B (in pink), the helix C (in purple), and the loop connecting them (in grey) were allowed to adopt random conformations. (B) Experimental SAXS curve for TupV C_Δ53 (in black) and fitted curve (in red) calculated for one representative ensemble of conformers selected with the program GAJOE (shown in panel (E)). (C) Rg distribution for an ensemble of 12 conformers of TupV C_Δ53. The red area shows the distribution for the initial ensemble of conformers. The black bars show the Rg distribution for selected ensembles of 12 conformers. (D) D_max distribution for an ensemble of 12 conformers of TupV C_Δ53. The red area shows the distribution for the initial ensemble of conformers. The black bars show the D_max distribution for selected ensembles of 12 conformers. (E) Representative ensemble of 12 conformers that reproduced the experimental SAXS curve. The N-terminal disordered part is shown in blue and the structured domain is shown in color from blue to red going from N to C-terminal.

Figure 9

Conformational ensemble modeling of TupV C_FL protein from SAXS data. (A) Description of the initial ensemble, which is comprised of 10,000 conformers in which the regions corresponding to the N-terminal intrinsically disordered region (NT-IDR in grey), the helix B (in pink), the helix C (in purple), and the loop connecting them (in grey) were allowed to adopt random conformations. (B) Experimental SAXS curve for TupV C_FL (in black) and fitted curve (in red) calculated for one representative ensemble of conformers selected with the program GAJOE (shown in panel (E)). (C) Rg distribution for an ensemble of five conformers of TupV C_FL. The red area shows the distribution for the initial ensemble of conformers. The black bars show the Rg distribution for selected ensembles of five conformers. (D) D_max distribution for an ensemble of five conformers of TupV C_FL. The red area shows the distribution for the initial ensemble of conformers. The black bars show the D_max distribution for selected ensembles of five conformers. (E) Representative ensemble of five conformers that reproduced the experimental SAXS curve. The N-terminal disordered part in shown in blue and the structured domain is shown in color from blue to red going from N to C-terminal.

Figure 10

In vivo studies. (A) Schematic representation of the RNA minigenome. The scheme drawn as a single strand of negative-sense RNA shows the positions of the Pol1 promoter (P Pol1) and terminator (T Pol1), of the NiV leader, trailer, gene start, and untranslated regions from the N gene (N GS and N UTR), of the gene end and untranslated regions from the L gene (L GE and L UTR) and of the Renilla luciferase, including four bases added to correct the length of the minigenome to be evenly divisible by six. Lengths in nt of the different elements are shown in brackets. (B) Schematic representation of the mini-replicon assay. Initially, the minigenome synthesis is controlled by the polymerase I promoter in the pPol I NiV-REN plasmid. The components of the NiV ribonucleoprotein complex (N, P, L) are expressed from helper plasmids under the control of polymerase II (CMV pr). Successful encapsidation of the minigenome RNA will lead to replication of the minigenome and transcription of the reporter gene under the control of NiV polymerase L. Renilla luciferase expression is then used to quantify NiV polymerase activity. A plasmid expressing firefly luciferase under the CMV promoter (CMV pr) is used as a normalization control. (C) Effect of NiV and TupV C protein on NiV mini-replicon assay. HEK293T cells were transfected with plasmids encoding Nipah viral N, P, and L proteins and one NiV minigenome encoding Renilla luciferase. A plasmid encoding Firefly luciferase for normalization. In the negative control (-L), the plasmid encoding NiV L was omitted. Increasing plasmid amounts of NiV C or TupV C were co-transfected as indicated, while the amount of transfected plasmid DNA was kept constant by the addition of the pCDNA3 vector. At 24 h after transfection, luciferase activities were measured and results obtained in the absence of C protein (NiV+EV) were set to 100% whereas results of the negative control (-L) were set to 0 %. For relative Firefly expression analysis, NiV+EV values were set to 100%. Individual values are represented by dots. Mean and SD from five independent experiments are indicated. **** p < 0.0001; ns not significant (One-way ANOVA, Dunnett Test). (D) Identification of interactions between TupV C protein and human proteins by TAP-MS. A total of 12 human proteins were captured with the TupV C bait, representing different molecular functions involved in several biological processes and localized in different cellular components.

Figure 11

Position of the STAT1 binding site in Orthoparamyxovirinae P or C proteins. The upper panel of the figure shows the alignment of the part of the C proteins encompassing helices D and E of TupV C for members of both clusters. The sequence alignment of MeV, NiV, and TupV C is taken from [22], and SeV C is aligned on the TupV C sequence according to the structural alignment shown in Figure 4D. This alignment is converted into a nucleotide alignment (middle panel), where the regions coding for the STAT1 binding site are highlighted in red (encoded in the reference frame in MeV and NiV) or in green (encoded in the +1 frame in SeV). The lower panel is a schematic representation of the proteins expressed from the P gene, showing the location of the STAT1 binding site either in P/V/W proteins (cluster 1) or in C protein (cluster 2).