Single particle assay of coronavirus membrane fusion with proteinaceous receptor-embedded supported bilayers

Abstract

Total internal reflection microscopy combined with microfluidics and supported bilayers is a powerful, single particle tracking (SPT) platform for host-pathogen membrane fusion studies. But one major inadequacy of this platform has been capturing the complexity of the cell membrane, including membrane proteins. Because of this, viruses requiring proteinaceous receptors, or other unknown cellular co-factors, have been precluded from study. Here we describe a general method to integrate proteinaceous receptors and cellular components into supported bilayers for SPT fusion studies. This method is general to any enveloped virus-host cell pair, but demonstrated here for feline coronavirus (FCoV). Supported bilayers are formed from mammalian cell membrane vesicles that express feline aminopeptidase N (the viral receptor) using a cell blebbing technique. SPT is then used to identify fusion intermediates and measure membrane fusion kinetics for FCoV. Overall, the fusion results recapitulate what is observed in vivo, that coronavirus entry requires binding to specific receptors, a low-pH environment, and that membrane fusion is receptor- and protease-dependent. But this method also provides quantitative kinetic rate parameters for intermediate steps in the coronavirus fusion pathway, which to our knowledge have not been obtained before. Moreover, the platform offers versatile, precise control over the sequence of triggers for fusion; these triggers may define the fusion pathway, tissue tropism, and pathogenicity of coronaviruses. Systematically varying these triggers in this platform provides a new route to study how viruses rapidly adapt to other hosts, and to identify factors that led to the emergence of zoonotic viruses, such as human SARS-CoV and the newly emerging human MERS-CoV.

Keywords: Coronavirus; Fusion kinetics; Membrane fusion; Supported lipid bilayers.

Fig. 1

(Top) Illustration of the formation of a fAPN-bleb supported bilayer from cell blebs derived from BHK cells. (Bottom) Fluorescence images of fAPN-SB formation, corresponding to the above cartoon. (Left, t = 0) fAPN-blebs containing R18 adsorbed to glass substrate. Note that some larger blebs dominate the signal, but many smaller blebs are adsorbed as well. (Middle images) ∼100 s after the addition of BHK-liposome solution to adsorbed blebs. Note that the BHK-liposome solution is devoid of fluorescent label, thus all signal comes from release of R18 initially confined to the bleb vesicle before rupture. (Right, t = 300 s) Continuous supported bilayer observed 300 s after the addition of liposomes. These images are all taken under 40× magnification. The dark lines in each image are scratches intentionally made with a dissection tool that is used to find the focal plane of the bilayer. The continuous focus of this line throughout the rupture process indicates that the focal plane not change and that the uniform distribution of fluorescence at t = 300 s is due to mobility of fluorophores redistributed throughout the newly-formed planar bilayer.

Fig. 2

fAPN supported bilayer characterization and mobility. R18 fluorescence recovery after photobleaching in a fAPN supported bilayer (similar to t = 300 s in Fig. 1). The images correspond to the times for each color-coded arrow on the plot. The data are fit to curve (black line) to obtain the diffusion coefficient. At t = 0, the bilayer was bleached with a 561 nm laser beam. The diameter of the bleached area is ∼20 μm. The reported diffusion coefficient on the plots is averaged from several experiments. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

A comparison of FCoV binding to (top) empty vector-SB and (bottom) fAPN-SB. The fAPN-SB exhibits specific binding of the virus. Note that due to the random intercalation of R18 in the viral membranes, fluorescence quenching varies among the virions, so some viruses are dimmer than others in this image.

Fig. 4

Dual-labeling scheme of coronavirus for single particle fusion experiments that facilitate the capture of intermediate states. The viral membrane is labeled with a green-emitting, lipophilic fluorophore. The viral contents are labeled with a red-emitting fluorophore. The two leaflets of the membranes are distinguished by the thin white line. (top) Spike proteins bind to fAPN (purple) present in the supported bilayer (gray). (middle) A drop in pH triggers hemifusion between the viral membrane and supported bilayer, leading to the mixing of the outer leaflets of each and the formation of a stalk. (bottom) Collapse of the stalk into a fusion pore, which results in the release of viral contents. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Coronavirus hemifusion kinetics after pretreatment with trypsin. (a) Images of a single coronavirus hemifusion dequenching event. The system was acidified at t = 0 to pH 5.3. The color-coded frames correspond to the time points in the curve marked with arrows. The trace plots the fluorescence in a small 4 × 4 μm region around the virus. The spike in intensity is when the hemifusion event begins. The lag time leading up to this point is the hemifusion lag time between acidification and hemifusion. (b) Many events like those in (a) are cataloged and plotted as a cumulative distribution function and fit with equation (1) for several pH values. (c) Hemifusion rate constants over a range of pH. (d) The corresponding number of spike proteins, N, determined from the statistical analysis of data. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

(a) A single, dual-labeled coronavirus fusion event at pH 4.5. Green and red channel images of this virion undergoing hemifusion (green) and then pore formation (red). The arrows denote the virion being analyzed, which has both labels co-localized in one particle. In the red channel, a second particle is visible; however, during this span of time, it does not fuse. (b) The corresponding fluorescence intensity traces of the virion in (a). The lag time between hemifusion and pore formation for this virion is denoted by the black double-ended arrow. (c) Pore formation statistics at pH 4.5 taken from single particle fusion events, like those in (b). The bar graph shows the distribution of lag times between the hemifusion spike and the release of internal fluorescence for each dual-labeled fusing virus. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)