Acidic pH Triggers Lipid Mixing Mediated by Lassa Virus GP

Abstract

Lassa virus (LASV) is the causative agent of Lassa hemorrhagic fever, a lethal disease endemic to Western Africa. LASV entry is mediated by the viral envelope glycoprotein (GP), a class I membrane fusogen and the sole viral surface antigen. Previous studies have identified components of the LASV entry pathway, including several cellular receptors and the requirement of endosomal acidification for infection. Here, we first demonstrate that incubation at a physiological temperature and pH consistent with the late endosome is sufficient to render pseudovirions, bearing LASV GP, non-infectious. Antibody binding indicates that this loss of infectivity is due to a conformational change in GP. Finally, we developed a single-particle fluorescence assay to directly visualize individual pseudovirions undergoing LASV GP-mediated lipid mixing with a supported planar bilayer. We report that exposure to endosomal pH at a physiologic temperature is sufficient to trigger GP-mediated lipid mixing. Furthermore, while a cellular receptor is not necessary to trigger lipid mixing, the presence of lysosomal-associated membrane protein 1 (LAMP1) increases the kinetics of lipid mixing at an endosomal pH. Furthermore, we find that LAMP1 permits robust lipid mixing under less acidic conditions than in its absence. These findings clarify our understanding of LASV GP-mediated fusion and the role of LAMP1 binding.

Keywords: Lassa virus; membrane fusion; single-particle fusion; virus entry.

Figure 1

Exposure to low pH at physiological temperature inactivates LASV GP. VSVΔG-GFP-GP viral pseudotypes were treated with either phosphate buffer at pH 6.0 or 7.0, or acetate buffer at pH 5.5 or 5.0 at the indicated temperature for (A) 5 min or (B) 30 min. Virus was subsequently brought back to neutral pH and used to infect Vero cells. Infectivity was determined by measuring GFP expression using flow cytometry. Infectivity was normalized to the percent of GFP-positive cells at the 21 °C pH 7.0 condition for each virus. Assays were performed in biological triplicate with new viral preps, and each biological replicate was performed in technical triplicate. Data are presented as the mean of three biological replicates and error bars represent the standard error of the mean. An asterisk indicates a statistically significant difference as compared to the case of pH 7 (p < 0.05) calculated using a one-way ANOVA. The lack of an asterisk indicates no statistically significant difference as compared to pH 7.

Figure 2

The binding of neutralizing and non-neutralizing antibodies to LASV GP after exposure to acidic pH at physiological temperature. VSVΔG-GFP-GP viral pseudotypes were treated with either phosphate buffer at pH 6.0 or 7.0, or acetate buffer at pH 5.5 or 5.0 at the indicated temperature for 30 min. Virus was subsequently brought back to neutral pH and assayed for binding by non-neutralizing (26.5E, 24.6C) and neutralizing antibodies (12.1F, 37.2D, 37.7H) by ELISA [32]. Data are presented as the mean of three biological replicates and error bars represent standard error of the mean. An asterisk indicates a statistically significant difference as compared to the case of incubation at 0 °C and pH 7 (p < 0.05.) calculated using a one-way ANOVA. The lack of an asterisk indicates no statistically significant difference.

Figure 3

LASV GP mediates lipid mixing between fluorescently labeled virions and a supported lipid bilayer. (A) Cartoon of pseudovirions formed with a VSV core, inactive HA0, and LASV GP, hemifusing to a planar lipid bilayer supported by a quartz microscope slide within a microfluidic chamber. Pseudovirions were flowed onto the bilayer and allowed to attach at neutral pH by way of the HA0–sialic acid interaction. DiD fluorescence dequenching arose due to single pseudovirions hemifusing to the supported bilayer upon the introduction of acidic pH. CF was included in the supported bilayer in order to provide a fluorescence indicator of pH. Fluorescence was detected using the evanescent field generated from the simultaneous total internal reflection of 642-nm and 532-nm lasers on a custom-built prism-based TIRF microscope. Movies were recorded at a 100 ms exposure time (10 Hz) for a minimum of 1400 frames. (B) Frames from a single DiD dequenching event at pH 5 acquired at the indicated time points, where Time = 0 was defined as a 50% decrease in CF fluorescence. (C) Representative fluorescence traces (CF, red; DiD, blue) are from a single dequenching event. The drop in CF fluorescence arises from the acidification of the buffer in the channel. The subsequent spike in DiD fluorescence arises from dequenching as DiD diffuses into the planar bilayer. The interval (t) between the 50% decrease in CF fluorescence and DiD dequenching was extracted for kinetic analysis.

Figure 4

LAMP1 increases the kinetics of GP-mediated lipid mixing. For each lipid mixing event, the interval of time between the 50% decrease in CF fluorescence and DiD dequenching was compiled into a histogram and fit to a gamma distribution by a least-squares algorithm. (A) The histogram for each pH considered, in the absence or presence of LAMP1, is shown with the gamma distribution fit overlaid in red. N indicates the number of individual lipid mixing events. (B) For each experiment, the number of rate-determining steps leading from the acidification of the flow cell to dequenching was determined through the gamma distribution fitting of each histogram in A. (C) The rate constants (k) determined from the gamma distribution fitting of each histogram in A. (D) The mean time to dequenching, defined as N/k for a gamma distribution. Error bars reflect the 95% confidence intervals determined in the fitting. An asterisk indicates a statistically significant difference as compared to the case of LAMP1 at pH 5 (p < 0.05.) calculated using a one-way ANOVA.