Visualization of Calcium Ion Loss from Rotavirus during Cell Entry

Figures

• 

  FIG 1. Rotavirus structure, reconstitution, location of the S469C mutation in VP4, and structures of the Ca2+ sensor components. (A) Rotavirus structural components, including the outer coat proteins VP4 (red) and VP7 (yellow) and the DLP proteins VP6 (green) and VP2 (cyan). The double-layer particle (DLP) of rotavirus can be reconstituted into an infectious triple-layer particle (TLP) by the addition of VP4 and VP7 in the presence of Ca2+ (19). Chelating Ca2+ ions dissociates VP7 trimers, stripping off the outer layer of the TLP. The blue box indicates the VP4 spike, magnified in the larger box, showing the location of the S469C mutation (yellow and cyan spheres) and the VP5* “body” and VP8* “lectin” subdomains of VP4 (the “foot” subdomain of VP8* plus VP5* is not shown [see reference 8]). (B) Structure of DBCO-maleimide. (C) Structure of Rhod5F azide.

• 

  FIG 2. Infectivity and live-cell imaging of Atto 565 rcTLPs. (A) Focus-forming assay comparing infectivities of native RRV (TLP) with Atto 565-labeled rcTLPs. rcTLPs were generated from DLPs, VP7, and VP4 labeled with Atto 647N, Atto 488, and Atto 565 dyes, respectively. Specific infectivity is shown as particles per focus-forming unit from triplicate experiments in BSC-1 cells. Error bars indicate standard deviations (SD). Infectivity and SD (focus-forming units per milliliter) of virus in each sample are shown above the respective bars. (B) BSC-1 cells were infected at an MOI of 50 with Atto 565 rcTLPs and imaged for 30 min. Individual particles were tracked by the signal in the 642-nm channel (DLP) (see Materials and Methods); uncoating events were detected as particles with a roughly constant 642-nm signal that lose the 488-nm signal (VP7). A similar loss in the 561-nm channel (VP4) coincides with the loss of VP7. The trace is representative of hundreds of measurements: another example is shown in Fig. 3. The spike in the 488- and 561-nm traces at about 740 s (see the 740-s time point in panel C), well after the particle in question has lost intensity in these channels, comes from…

• 

  FIG 3. Kinetics of VP7 and VP4 loss and of DLP release from Atto 565 rcTLPs. (A) Traces for DLP (left) and VP7 and VP4 (right). Dashed vertical lines show the onset of VP7 loss (t1) and termination of VP7 loss (t2), the onset of VP4 loss (tVP4-1) and its termination (tVP4-2), and DLP release (t3). See the legend of Fig. 1B for details. (B) Distribution of time intervals between the onset of VP7 loss and VP4 loss (t1 − tVP4-1), the onset of VP4 loss and DLP release (t3 − tVP4-1), the onset of VP7 loss and DLP release (t3 − t1), termination of VP4 loss and DLP release (t3 − tVP4-2), and termination of VP7 loss and DLP release (t3 − t2) from 127 uncoating particles collected from 7 BSC-1 cells infected at an MOI of 50. (C) Schematic of the average times outlined in panel B demonstrating the timing and duration of VP7 loss (green bar), VP4 loss (blue bar), and overall DLP release (red bar), in relation to t1.

• 

  FIG 4. pHrodo Red rcTLPs. (A) Focus-forming assay comparing infectivities of native RRV (TLP) and pHrodo Red-labeled rcTLPs. rcTLPs generated from DLPs, VP7, and VP4 were labeled with Atto 647N, Atto 488, and pHrodo Red dyes, respectively. Specific infectivities are shown as particles per focus-forming unit from triplicate experiments in BSC-1 cells. Error bars indicate SD. Infectivity and SD (focus-forming units per milliliter) of the virus in each sample are shown above the respective bars. (B) BSC-1 cells were infected at an MOI of 50 with pHrodo Red rcTLPs and imaged for 30 min. Individual particles were tracked by the signal in the 642-nm channel (red trace) (see Materials and Methods). Uncoating events were detected as particles with a roughly constant 642-nm signal that lose the 488-nm signal (VP7). Uncoating occurred in the absence of any increase in 561-nm intensity (VP4). A trace of a different particle from the same experiment is shown in Fig. 7A. (C) Micrographs from all three channels of the rcTLP represented in panel B. Circular outlines highlight the particle in question in the 642-nm (DLP), 488-nm (VP7), and 561-nm (VP4) channels. The time interval between frames was 30…

• 

  FIG 5. Response to reduced pH of pHrodo Red rcTLPs on glass coverslips. pHrodo Red rcTLPs with DLP-647N, VP7-488, and VP4-pHrodo Red were imaged in the same Fluorobrite medium used for live-cell imaging (see Materials and Methods). Each field of particles was imaged at wavelengths of 642 nm (red), 562 nm (green), and 488 nm (blue) in a 15-step z-series (0.5-μm step size). Shown are maximum-intensity z-projections of an overlay of all three channels as well as individual channels. (A) Fluorobrite medium; (B) Fluorobrite medium supplemented with 500 mM sodium acetate (pH 5.2); (C) Fluorobrite medium with 1% DMSO and 100 nM Baf A1; (D) Fluorobrite medium with 1% DMSO and 100 nM Baf A1 supplemented with 500 mM sodium acetate (pH 5.2).

• 

  FIG 6. Infectivity and live-cell imaging of pHrodo Red rcTLPs in the presence of Baf A1. (A) Focus-forming assay comparing infectivities of pHrodo Red-labeled rcTLPs infecting BSC-1 cells in DMEM, DMEM with 100 nM Baf A1 and 1% DMSO, or 1% DMSO in DMEM (see Materials and Methods). Specific infectivity is shown as particles per focus-forming unit from triplicate experiments in BSC-1 cells. Error bars indicate SD. Infectivity and SD (focus-forming units per milliliter) are shown above the respective bars. (B) BSC-1 cells were incubated in medium containing 100 nM Baf A1 and 1% DMSO for 1 h at 37°C before infection at an MOI of 50 with pHrodo Red rcTLPs and imaging for 30 min. Individual particles were tracked by the signal in the 642-nm channel (red trace) (see Materials and Methods). Uncoating events were detected as particles with a roughly constant 642-nm signal that lose the 488-nm signal (VP7). Uncoating occurred in the absence of any increase in the 561-nm intensity (VP4). (C) Micrographs of all three channels of the rcTLP represented in panel B. The time interval between frames was 10 s.

• 

  FIG 7. Kinetics of VP7 loss and DLP release from pHrodo Red rcTLPs. (A) Traces for DLP (left) and for VP7 and pHrodo Red (right). Dashed vertical lines show the onset of VP7 loss (t1), termination of VP7 loss (t2), and DLP release (t3). See the legend of Fig. 1B for details. (B) Distribution of intervals between the onset of VP7 loss and DLP release (t3 − t1) in BSC-1 cells infected in the absence (top) or presence (bottom) of Baf A1. (C) Distribution of times between the termination of VP7 loss and DLP release (t3 − t1) in BSC-1 cells infected in the absence (top) or presence (bottom) of Baf A1. (D) Schematics of the average times outlined in panel B demonstrating the timing and duration of VP7 loss (green bars) and overall DLP release (red bars) in relation to t1 in the absence (top) or presence (bottom) of Baf A1.

• 

  FIG 8. VP5 CT digestion of VP4 Rhod5F and infectivity of Rhod5F rcTLPs. (A) Monomeric VP4 469C was sequentially labeled with DBCO-maleimide and Rhod5F azide as described in Materials and Methods. The resulting labeled VP4 Rhod5F was sequentially digested with chymotrypsin and trypsin (see Materials and Methods). Reducing sample buffer was added to a sample of undigested VP4 Rhod5F that was then boiled and subjected to analysis by SDS-PAGE. Reducing sample buffer was also added to two samples of the digested protein, one of which was boiled prior to SDS-PAGE; the other sample remained unboiled. Samples were then run on a 5-to-20% gradient SDS gel and imaged with UV light (right) prior to staining with Coomassie blue and imaging with transmitted light (left). (B) Focus-forming assay comparing infectivities of native RRV (TLP), 469C rcTLPs, and Rhod5F rcTLPs (see Materials and Methods). Specific infectivity is shown as particles per focus-forming unit from triplicate experiments in BSC-1 cells. Error bars indicate SD. Infectivity ± SD (focus-forming units per milliliter) for the virus in each sample are shown above the respective bars.

• 

  FIG 9. Cell binding by 469C and Rhod5F rcTLP. (A) BSC-1 cells were infected with rcTLPs with DLP-647N, VP7-488, and VP4 469C at an MOI of 10 for 1 h at 37°C (see Materials and Methods); washed in HNC; fixed for 10 min in 4% paraformaldehyde in HNC; and washed three more times for 5 min in HNC. The full volumes of 124 cells were imaged at 488, 561, and 642 nm as a z-series with a step size of 0.5 μm. Shown are representative maximum-intensity z-projections of micrographs of all three channels, pseudocolored red (642 nm), green (488 nm), and blue (561 nm), as well as an overlay of all three. The white outline shows the outer boundary of the cell; the yellow box highlights particles of interest on the cell. (B) Magnified image of particles highlighted in panel A and overlay of the three. White arrows indicate rcTLPs that have yet to lose the VP7-488 shell, showing that the unlabeled 469C rcTLPs lack the 560-nm Rhod5F signal. Other particles in the field of view represent uncoated DLP-647N particles. (C) BSC-1 cells were infected at an MOI of 1 for 1 h at 37°C with rcTLPs from the preparation used for images in panels A and B, sequentially labeled with DBCO-maleimide and Rhod5F azide (see…

• 

  FIG 10. Live-cell imaging of Rhod5F rcTLPs. (A) BSC-1 cells were infected at an MOI of 10 with Rhod5F rcTLPs and imaged for 30 min. Individual particles were tracked by the signal in the 642-nm channel (left) (see Materials and Methods). Uncoating events were detected as particles with a roughly constant 642-nm signal that lose the 488-nm signal (VP7) (right). A decrease in the 561-nm intensity (VP4 Rhod5F) (right) preceded the onset of VP7 uncoating. Another example from the same experiment is shown in Fig. 9A. (B) Micrographs of all three channels of the uncoating rcTLP represented in panel A.

• 

  FIG 11. Kinetics of VP7 and Ca2+ loss and DLP release from Rhod5F rcTLPs. (A) Traces for DLP (left) and for VP7 and VP4 Rhod5F (right). Dashed vertical lines show the onset and termination of VP7 loss (t1 and t2, respectively), the onset and termination of Ca2+ loss (tc1 and tc2, respectively), and DLP release (t3). (B) Distribution of intervals between the onset of VP7 and Ca2+ loss (t1 − tc1), the onset of Ca2+ loss and DLP release (t3 − tc1), the onset of VP7 loss and DLP release (t3 − t1), termination of Ca2+ loss and DLP release (t3 − tc2), and termination of VP7 loss and DLP release (t3 − t2), from 134 uncoating particles collected from 35 BSC-1 cells infected at an MOI of 10. (C) Schematic of the average times outlined in panel B showing the timing and duration of VP7 loss (green bar), Ca2+ loss (blue bar), and overall DLP release (red bar) in relation to t1.

• 

  FIG 12. Probability densities of 640- and 488-nm channel amplitudes from 469C and Rhod5F rcTLP binding assays. (A and B) Distribution of 640-nm amplitudes derived from particle detections as described in Materials and Methods for all 469C (A) and Rhod5F (B) binding assay samples. Blue and red arrows indicate the upper and lower limits for detection to be considered a single particle; signals weaker than those indicated by the blue arrow were taken as noise, and signals stronger than those indicated by the red arrow were considered aggregates or multiple unresolved particles. (C and D) Distribution of 488-nm amplitudes derived from particle detections for all 469C (C) and Rhod5F (D) binding assay samples (see Materials and Methods). The red arrow points to the cutoff, above which a particle was taken to be a VP7-488-coated rcTLP and below which it was considered uncoated.