Human astrovirus capsid protein releases a membrane lytic peptide upon trypsin maturation

Abstract

The human astrovirus (HAstV) is a non-enveloped, single-stranded RNA virus that is a common cause of gastroenteritis. Most non-enveloped viruses use membrane disruption to deliver the viral genome into a host cell after virus uptake. The virus-host factors that allow for HAstV cell entry are currently unknown but thought to be associated with the host-protease-mediated viral maturation. Using in vitro liposome disruption analysis, we identified a trypsin-dependent lipid disruption activity in the capsid protein of HAstV serotype 8. This function was further localized to the P1 domain of the viral capsid core, which was both necessary and sufficient for membrane disruption. Site-directed mutagenesis identified a cluster of four trypsin cleavage sites necessary to retain the lipid disruption activity, which is likely attributed to a short stretch of sequence ending at arginine 313 based on mass spectrometry of liposome-associated peptides. The membrane disruption activity was conserved across several other HAstVs, including the emerging VA2 strain, and effective against a wide range of lipid identities. This work provides key functional insight into the protease maturation process essential to HAstV infectivity and presents a method to investigate membrane penetration by non-enveloped viruses in vitro. IMPORTANCE Human astroviruses (HAstVs) are an understudied family of viruses that cause mild gastroenteritis but have recent cases associated with a more severe neural pathogenesis. Many important elements of the HAstV life cycle are not well understood, and further elucidating them can help understand the various forms of HAstV pathogenesis. In this study, we utilized an in vitro liposome-based assay to describe and characterize a previously unreported lipid disruption activity. This activity is dependent on the protease cleavage of key sites in HAstV capsid core and can be controlled by site-directed mutagenesis. Our group observed this activity in multiple strains of HAstV and in multiple lipid conditions, indicating this may be a conserved activity across the AstV family. The discovery of this function provides insight into HAstV cellular entry, pathogenesis, and a possible target for future therapeutics.

Keywords: astrovirus; liposome; proteases; structural biology.

Fig 1

HAstV8 maturation process and constructs used in this study. (a) The maturation process of the HAstV8 capsid protein VP90 catalyzed by caspases and trypsin. (b) Recombinant protein constructs used in this study. Red asterisks indicate sites where amino acids recognized by trypsin (arginine and lysine) were mutated. (c) Cryo-EM maps of the immature HAstV8 VP70 and mature HAstV1 VP34/27/25 particles. Left, EMDB ID: EMD-5414. Right, EMDB ID: EMD-5413. (d) Crystal structures of the S-P1 core and P2 spike. Left, PDB ID: 5IBV. Right, PDB ID: 3QSQ. The S and P1 domains in the core are colored in blue and yellow, respectively. In the spike dimer, the two subunits are colored in different shades of red.

Fig 2

Expression and purification of the HAstV8 capsid protein VP90. (a) HAstV8 VP90^28-782 purified by Superdex-200 (S200) size exclusion chromatography. The protein was eluted with a major peak at an elution volume of 46.4 mL, indicating a large molecular weight assembly. A black arrow marks the peak elution of the blue dextran molecular weight standard indicating void volume, while another arrow shows the peak elution of ferritin of 440 kD. Inset—protein size and stability were confirmed by SDS-PAGE. (b) TEM micrographs of HAstV8 VP90^28-782. Rod-shaped particles were observed with a consistent diameter but variable lengths. Scale bar, 200 nm.

Fig 3

Liposome disruption by HAstV8 VP90. (a) Schematic diagram of the lipid disruption assay employed in this study. (b) HAstV8 VP90^28-782 was able to disrupt liposome only in the presence of trypsin. Liposomes were measured to collect a baseline signal for 0% disruption by averaging values at each timepoint (t = −5–0 min, not shown), then mixed with trypsin and HAstV8 proteins. The change in fluorescence signal was measured for 20 min in 30-s kinetic intervals (t = 0–20 min). All assays were concluded with the addition of 1% Triton X-100 detergent solution to determine the 100% dequenched signal by averaging the values at each timepoint (t = 20–30 min, not shown). Each test includes control conditions of liposome alone (top left), HAstV8 VP90^28-782 mixed with liposomes (top right), trypsin mixed with liposomes (bottom left), and the test condition of HAstV8 VP90^28-782 mixed with both trypsin and liposomes (bottom right). N = 3 technical replicates, with each replicate shown as a separate curve.

Fig 4

Liposome disruption activity of HAstV8 VP90 mapped to the P1 domain. (a) SDS-PAGE gels of purified HAstV8 VP90 domain constructs, purified VP70 particles, and other purified HAstV samples tested for lipid disruption with POPC liposomes. Protein samples were evaluated on separate SDS-PAGE gels and aligned based on the position of a standard protein ladder. All samples were visualized by Coomassie staining. (b) Liposome disruption assays. All samples were evaluated at a 200:1 lipid to protein (L:P) molar ratio. Fluorescence dequenching was measured in the same kinetic protocol as in Fig. 3b, with percentage disruption at timepoints of 0.5, 1, 2.5, 5, 10, and 25 min shown for clarity. Three technical replicates were performed.

Fig 5

Trypsin cleavage sites from amino acids 264-313 of the HAstV8 Core are essential for liposome disruption. (a) Amino acid sequence of P1, with β-sheets labeled with arrows and α-helices with blocks. Trypsin cleavage sites are highlighted in magenta, with four sites from aa264 to 313 (yellow) and eight sites from aa314 to 415 (orange). (b) The structure of the HAstV8 core with the 12 trypsin cleavage sites mapped to P1. The shell domain is in blue and P1 domain is in orange (block-1 region)/yellow (block-2 region). Lysine and arginine residues are highlighted by magenta sticks. (c) Liposome disruption by P1 mutants. Trypsin cleavage site removal in Core(P1-KR) and Core(P1-B1KR) resulted in the loss of liposome disruption activity. All mutant proteins were mixed with liposomes at the same 200:1 lipid-to-protein molar ratio.

Fig 6

Mass spectroscopy identification of HAstV8 liposome-associated peptides. (a) A schematic of the protein-associated liposome flotation assay. (b) Structural modeling of the liposome-associated peptides. (Left) Crystal structure of the core with aa300–313 highlighted in red and aa274–299 in pink. The shell domain is shown in blue, and the rest of the P1 region is colored in yellow and orange as in Fig. 5a. (Right) The predicted structure of aa274–313 with a hydrophobic core that could be exposed upon liposome insertion.

Fig 7

Liposome disruption activity of HAstV capsid proteins observed using different virus strains and different lipids. (a) Liposome disruption by HAstV1 and VA2 VP90. The L:P ratio and trypsin concentration were maintained from the HAstV8 test conditions. (b) Liposome disruption of HAstV8 core tested in different lipids.