Single-Cell Virology: On-Chip, Quantitative Characterization of the Dynamics of Virus Spread from One Single Cell to Another

Figures

• 

  Figure 1. Cell pairing for unambiguous characterization of the spread of virus infection at the single-cell level. (a) Utilizing laminar flow in microchannels, a microfluidic device has been developed to capture two different cells in each nanowell. Vybrant DiI-labeled cells (red) and DiO-labeled cells (green) were infused into the two inlets of the channel at the same flow rate. Streamlines obtained by COMSOL simulation of the velocity field for the flow in the channel indicated that the two cell suspensions would not mix with each other. (b) Brightfield and fluorescence images show red cells captured on one side while green cells were captured on the other side of the same nanowell. Optimized condition: infusion rate was 100 nL/min; cell density was 8 × 105 cells/mL.

• 

  Figure 2. Different modes of PV spread observed by time-lapse imaging of cell pairs. Vybrant DiI (red) labeled cells were infected with PV-GFP (donor) and loaded simultaneously on the device with uninfected cells (recipient). The chip was imaged every 30 min over a period of 24 h. Selected images are shown. As virus replicated in donor cells, green fluorescence was detected as yellow because of the red-dye-labeled cell. As recipient cells became infected, these cells turned green. Lysis of infected donor cells caused the color to change from yellow to red as GFP leaked into the media. Three modes of spread were observed. (a) No spread: GFP signal increased in donor cells over time followed by lysis, indicated by the return of a red color to the donor cell, without any detectable GFP signal observed in recipient cells. (b) Lytic spread: GFP signal increased in donor cells over time followed by lysis, indicated by the return of a red color to the donor cell, followed later by observation of an ever-increasing GFP signal in recipient cells. (c) Non-lytic spread: GFP signal increased in donor cells over time; however, prior to lysis, indicated by the absence of red donor cells, GFP signal was…

• 

  Figure 3. Quantitative analysis of PV infection dynamics in donor cells. (a) Representative plots of the fluorescence intensity observed in each well normalized to the background fluorescence of the well as a function of time revealed unprecedented, between-cell variability in the outcome of infection. (b,c) Two models could be used to describe infection outcomes: infection model (b); infection and lysis model (c). Both models identify phenomenological parameters that can be used to describe each curve. (d) To distinguish between lytic and non-lytic spread, we represented each event in each well as the start time of replication in the recipient cell on the y-axis and the lysis time of the donor cell on the x-axis. A difference (y − x) greater than 1 h indicates lytic spread while a value less than 1 h indicates non-lytic spread. For donor cells that did not lyse by the end the experiment (24 hpi), 24.5 hpi was set as the lysis time here. To determine if infection dynamics in the donor cell influenced the mode of spread, distributions for experimental parameters measured in donor cells were plotted for each category of spread: no spread (- -); lytic spread (––); and non-lytic spread (––).…

• 

  Figure 4. Quantitative analysis of PV infection dynamics in recipient cells: comparison of lytic and non-lytic spread. (a) Distributions of the difference (recipient–donor) for start point. Distributions of the fold change (parameter in recipient cell/parameter in donor cell) for (b) maximum, (c) slope, and (d) infection time. In each case, lytic infections and non-lytic infections were treated independently. Non-lytic spread has the capacity to reach values of maximum higher than those observed for lytic spread in the same time period. Non-lytic spread appears to establish infection in recipient cells much faster than lytic spread. (n = 100 cell pairs).

• 

  Figure 5. Lytic spread is sensitive to the presence of an entry/uncoating inhibitor; non-lytic spread is not. Pocapavir is known to bind to the PV capsid, which interferes with entry or uncoating. We have evaluated the impact of the presence of pocapavir on establishment of PV infection in recipient cells after lytic or non-lytic spread. (a) Infected donor cells were added to uninfected recipient cells in the absence or presence of the indicated concentration of pocapavir. The impact of the drug on spread was evaluated. The area under each curve is proportional to the number of events observed. Lytic spread was sensitive to the presence of pocapavir. In order to determine how pocapavir (1 µM) impacted each parameter, we evaluated distributions for the following parameters: (b) maximum, (c) slope, (d) infection time, and (e) start point in the donor cells stratified by spread phenotype: no spread (- -); lytic spread (––); and non-lytic spread (––). Higher values of maximum and slope, and earlier start time of infection in donor cells that lead to lytic spread suggests that viruses that are resistant to the capsid inhibitor may have higher viral fitness. (n = 146 cell pairs). * p < 0.05.

• 

  Figure 6. Non-lytic spread between single cells uses a non-traditional mechanism of autophagy as observed in bulk culture. Non-lytic spread of enteroviruses uses secretory autophagy, which is dependent on LC3 but independent of Vps34 (target of SAR405). To assess the relationship of non-lytic spread observed between single cells to that observed at the population level in cell culture, we evaluated the requirement for Vps34 by monitoring spread in the presence of SAR405 or from donor cells in which LC3B had been depleted (LC3B siRNA). SAR405 exhibited no impact on spread. Loss of LC3B interfered selectively with non-lytic spread. (Control: n = 99 cell pairs; SAR405: n = 127 cell pairs; LC3B siRNA: n = 69 cell pairs).

• 

  Figure 7. Summary of virus-spread phenotypes observed in this study and corresponding implications. Primary infection is characterized by observation of yellow fluorescence signal in donor cells (stained red at t = 0). Lysis is marked by loss of yellow signal and return of red signal. Observation of a green signal from recipient cells before or after lysis of donor cells distinguishes between lytic and non-lytic spread. The absence of green fluorescence in recipient cells indicates failure of infection to spread. Governing factors, advantages or disadvantages, and future studies associated with each mode of spread are described.