A conserved motif in Henipavirus P/V/W proteins drives the fibrillation of the W protein from Hendra virus

Abstract

The Hendra (HeV) and Nipah (NiV) viruses are high-priority, biosafety level-4 pathogens that cause fatal neurological and respiratory disease. Their P gene encodes not only the P protein, an essential polymerase cofactor, but also the virulence factors V and W. We previously showed that the W protein of HeV (W^HeV) forms amyloid-like fibrils and that one of its subdomains, PNT3, fibrillates in isolation. However, the fibrillation kinetics is much faster in the case of the full-length W^HeV compared to PNT3, suggesting that another W^HeV region contributes to the fibrillation process. In this work, we identified the region spanning residues 2-110 (PNT1) as the crucial region implicated in W^HeV fibrillation. Through site-directed mutagenesis, combined with thioflavin T binding experiments and negative-staining transmission electron microscopy, we showed that a predicted cryptic amyloidogenic region (CAR) within PNT1 is the main driver of fibrillation and deciphered the underlying molecular mechanism. Using FTIR, we showed that PNT1 fibrils are enriched in cross β-sheets. Sequence alignment revealed conservation of the CAR across the Henipavirus genus and enabled the identification of a hitherto never reported pro-amyloidogenic motif. The ability to form fibrils was experimentally shown to be a common property shared by Henipavirus PNT1 proteins. Overall, this study sheds light on the molecular mechanisms underlying W^HeV fibrillation and calls for future studies aimed at exploring the relevance of the newly identified pro-amyloidogenic motif as a valuable target for antiviral approaches.

Keywords: FTIR spectroscopy; amyloidogenic regions and motifs; circular dichroism; fibrillation mechanisms; negative‐staining transmission electron microscopy; thioflavin T binding.

FIGURE 1

PNT1 drives W^HeV fibrillation through a mechanism involving α‐helices. Turbidity (a) and ThT fluorescence measurements (b) of W^HeV and W^NiV in the absence or presence of 5% or 10% (v/v) TFE after 120 min of incubation and without a pre‐incubation in oxidative conditions (n = 4). Asterisks above error bars denote significance compared to the protein‐free conditions. *p < 0.05, **p < 0.01, ***p < 0.0001. Circular dichroism (CD) spectra of W^HeV (c) and W^NiV (d) in the absence or presence of 5% or 10% (v/v) TFE. MRE, mean residue ellipticity. (e) NS‐TEM observations of W^HeV and W^NiV in the absence or presence of 5% or 10% (v/v) TFE after 120 min of incubation. Scale bar: 200 nm. (f) Schematic representation of W^HeV (blue) with its predicted amyloid hotspots (orange boxes) and cysteine residues (green circles). PNT1, PNT2, and PNT3 subdomains and the α‐MoRE are indicated. (g) Near‐UV spectra as obtained during CD experiments of W^HeV PNT1 unincubated (t₀) or incubated for 1 h or 72 h without TFE. (h) NS‐TEM observations of W^HeV PNT1 after 0, 1 h, and 72 h of incubation. Inset: higher magnification at t_72h highlighting the fibrillar nature of W^HeV PNT1. Scale bar: 200 nm.

FIGURE 2

The first 50 residues are responsible for W^HeV fibrillation and accommodate the cross β‐sheet structure observed in fibrils. (a) Schematic representation of the different HeV PNT1 variants generated and used in this work. Orange: CAR motif, green: artificially disordered sequence, red: polyalanine stretch, white: truncated regions. (b) Proportion of α‐helix in PNT1, 50φPNT1, and 50φPNT1CAR as obtained from the deconvolution of CD spectra as obtained on unincubated samples. (c) NS‐TEM micrographs after 30 min of incubation of PNT1 and 50φPNT1 (both at 40 μM) in the presence of different TFE concentrations. (d) ThT fluorescence measurements of PNT1 in the presence of increasing concentrations of TFE (1–10%) (n = 12). Asterisks above error bars denote significance compared to the condition containing only ThT. ****p < 0.0001. (e) IR absorption spectra and (f) respective second‐derivative IR absorption spectra of PNT1 (blue) and 50φPNT1 (green). Absorption wavenumbers of interest are indicated in blue (PNT1) and green (50φPNT1).

FIGURE 3

A short amyloidogenic motif is responsible for W^HeV fibrillation. (a) ThT fluorescence measurements of PNT1, PNT1ΔCAR, and PNT1Ala after 30 min of incubation (n = 12). The asterisks denote a statistically significant difference (p < 0.0001). (b) Corresponding micrographs after 30 min of incubation of PNT1ΔCAR (top) and PNT1Ala (bottom). Scale bar: 500 nm. (c) ThT fluorescence measurements of W^HeV Δ19 after 90 min of incubation (n = 4). (d) NS‐TEM micrographs as obtained after 90 min of incubation for W^HeV, W^HeV Δ19 (at 25 μM), PNT1 (25 μM) and of an equimolar mixture of W^HeV Δ19 and PNT1 (both at 25 μM). Scale bar: 500 nm. (e) Fibril contour length measurements made on micrographs shown in (d). The number of measurements (n) is indicated.

FIGURE 4

Impact of charged and hydrophobic residues of the CAR and contribution of the CAR‐flanking α‐helices to PNT1 fibrillation. (a) ThT fluorescence measurements after 30 min of incubation of PNT1ΔAliph and PNT1ΔCharg compared to wild‐type PNT1. (b) Corresponding micrographs as obtained after 30 min of incubation. Scale bar: 500 nm. (c) CD spectra of unincubated PNT1ΔAliph and PNT1ΔCharg compared to wild‐type PNT1. MRE: mean residue ellipticity. (d) Proportion of α‐helix in PNT1ΔAliph and PNT1ΔCharg in the presence of 0–40% TFE as obtained from the deconvolution of CD spectra. (e) ThT fluorescence measurements after 30 min of incubation of 50φPNT1CAR in the presence of TFE (0–10%). (f) Corresponding micrographs made after 30 min of incubation. Scale bar: 500 nm. (g) ThT fluorescence measurements of PNT1, PNT1Δ9 (Δ9), PNT1Δ20‐29 (Δ20‐29), and PNT1Δ29 (Δ29) after 30 (t₃₀) or 120 min (t₁₂₀) of incubation. (h) Corresponding micrographs made at t₃₀ or t₁₂₀. Scale bar: 500 nm. The asterisks shown in panels (a), (e), and (g) indicate a statistically significant difference (p < 0.0001); ns: not significant.

FIGURE 5

Identification of a new consensus amyloidogenic motif in the P/V/W proteins of henipaviruses and fibrillation propensities of PNT1 subdomains from various henipaviruses. (a) Sequence alignment of the 110 first amino acids (PNT1 subdomain) of the seven major Henipavirus P/V/W proteins, along with the derived consensus sequence, as obtained using Jalview (Procter et al. 2021). The blue gradient is related to the amino acid conservation. The identified pro‐amyloidogenic motif is framed in red. Turbidity (b) and ThT fluorescence measurements (c) of the various Henipavirus PNT1 subdomains after 1 h of incubation (n = 4). The asterisks shown in panel (c) indicate a statistically significant difference (p < 0.0001). (d) Corresponding micrographs obtained after 1 h of incubation. Scale bar: 500 nm.

FIGURE 6

Proposed mechanism of fibrillation. The involvement of the three different N‐terminal regions in the fibrillation process is depicted using a gray color gradient: the 2–9 region shown to be fully dispensable is represented in white, the CAR is shown in dark gray to reflect its role as main driver of the process, and the 20–29 region exerting a synergistic effect with the CAR and strongly promoting nucleation is shown in light gray. The α‐helices are shown as partly transparent to illustrate that they are transiently populated in the unassembled form and likely also in the assembled form, contrary to the β‐strands that are stabilized in the form of β‐sheets in the core of the fibrils.