Triatoma virus recombinant VP4 protein induces membrane permeability through dynamic pores

Abstract

In naked viruses, membrane breaching is a key step that must be performed for genome transfer into the target cells. Despite its importance, the mechanisms behind this process remain poorly understood. The small protein VP4, encoded by the genomes of most viruses of the order Picornavirales, has been shown to be involved in membrane alterations. Here we analyzed the permeabilization activity of the natively nonmyristoylated VP4 protein from triatoma virus (TrV), a virus belonging to the Dicistroviridae family within the Picornavirales order. The VP4 protein was produced as a C-terminal maltose binding protein (MBP) fusion to achieve its successful expression. This recombinant VP4 protein is able to produce membrane permeabilization in model membranes in a membrane composition-dependent manner. The induced permeability was also influenced by the pH, being greater at higher pH values. We demonstrate that the permeabilization activity elicited by the protein occurs through discrete pores that are inserted on the membrane. Sizing experiments using fluorescent dextrans, cryo-electron microscopy imaging, and other, additional techniques showed that recombinant VP4 forms heterogeneous proteolipidic pores rather than common proteinaceous channels. These results suggest that the VP4 protein may be involved in the membrane alterations required for genome transfer or cell entry steps during dicistrovirus infection.

Importance: During viral infection, viruses need to overcome the membrane barrier in order to enter the cell and replicate their genome. In nonenveloped viruses membrane fusion is not possible, and hence, other mechanisms are implemented. Among other proteins, like the capsid-forming proteins and the proteins required for viral replication, several viruses of the order Picornaviridae contain a small protein called VP4 that has been shown to be involved in membrane alterations. Here we show that the triatoma virus VP4 protein is able to produce membrane permeabilization in model membranes by the formation of heterogeneous dynamic pores. These pores formed by VP4 may be involved in the genome transfer or cell entry steps during viral infection.

FIG 1

VP4 sequence analysis. (A) Transmembrane helix (TMH) prediction performed using TopPred software as described in Materials and Methods. Two helices are predicted; the first one is from residues 5 to 25, and the second one is from residues 37 to 57. (B) Sequence alignment of the VP4 proteins from TrV and other dicistroviruses: cricket paralysis virus (CrPV) and Drosophila C virus (DCV). (C) Sequence alignment of the VP4 proteins from TrV and the picornaviruses human rhinovirus 16 (HRV16) and poliovirus (Polio). (B, C) Conserved residues are in a white font on a black background, and similar residues are framed.

FIG 2

Influence of protein expression on culture turbidity. The turbidity of cultures expressing VP4 (squares) and a control protein (filled circles) was measured after induction by monitoring the absorbance at 600 nm. Results for a bacterial growth control without heterologous protein expression (empty circles) were also measured.

FIG 3

Membrane binding assay. Proteins (MBP-VP4 and control MBP) were incubated with liposome preparations and loaded on the bottom of a sucrose gradient. After ultracentrifugation, gradients were fractionated in four different fractions (F1 to F4, from top to bottom) and analyzed by SDS-PAGE. Liposomes banded at F1. Molecular mass markers (in kDa) are indicated on the left.

FIG 4

Membrane permeabilization induced by MBP-VP4. (A) Time course of leakage after addition of MBP-VP4 or MBP at the indicated concentrations. The proteins were added at 0 s, and after 1,000 s, liposomes were disrupted by addition of Triton X-100. (B) Dose-dependent response of MBP-VP4 leakage activity and control MBP at the same concentrations. Data points correspond to the mean values of three independent measurements, and error bars represent standard deviations.

FIG 5

Characterization of the leakage activity induced by MBP-VP4 and pore type determination. (A) Solute entry experiment. The extent of solute entry was compared with the extent of solute leakage from 100 μM PA-PC (1:1 molar ratio) liposomes obtained using 25, 50, 100, and 200 nM MBP-VP4. (B) Sizing experiment. The release of compounds of different sizes from 100 μM PA-PC (1:1 molar ratio) liposomes was measured as a function of the protein concentration. Data points correspond to the mean values of three independent measurements, and error bars represent standard deviations.

FIG 6

Cryo-EM imaging. (A) Pore sizes observed by cryo-EM. PA-PC (1:1 molar ratio) liposomes at 500 μM were incubated with different protein concentrations. Membrane interruptions observed in digital micrographs acquired at different nominal magnifications ranging from ×40,000 to ×80,000 were clustered in three different groups: 6 to 12 nm, 13 to 20 nm, and >20 nm. (B to D) Cryo-EM micrographs of membrane disruption. Interruptions in the membrane (black arrows) correspond to the groups with pore sizes of 6 to 12 nm (B), 13 to 20 nm (C) and >20 nm (D).

FIG 7

Effect of different concentrations of Lyso-PC (LPC) and DAG on the leakage activity of MBP-VP4. The protein/lipid ratio was 1:1,000 (on a molar basis). Data points correspond to mean values from three independent measurements, and error bars represent standard deviations.