RNA helicase DDX3X modulates herpes simplex virus 1 nuclear egress

Figures

• 

  Fig. 1. HSV-1 redirects DDX3X to the nuclear rim at late time points. a HeLa cells grown on coverslips were either mock treated or infected with wild-type strain 17+ virus at an MOI of 5. The infection was stopped at 3, 6 or 9 h post-infection (hpi). Cells were fixed and processed for immunofluorescence (DDX3X: red; ICP4: green; Hoechst 33342: blue). Samples were imaged using STED microscopy. Scale bars represent 5 µm. The images represent three independent experiments. b Quantification of DDX3X relocalisation to the nucleus was performed by seeding HeLa cells on coverslips and mock treated or infected with wild-type strain 17+ virus at an MOI of 5. At 9 hpi, cells were fixed and processed for dual colour STED microscopy using anti-DDX3X (red) and anti-LBR (green) antibodies. In three independent experiments, twelve random whole cells (z-stacks) were imaged for both conditions. Next, images were processed with the Imaris software to calculate the amount of DDX3X found on a masked LBR channel. Scale bar represents 5 µm. ±Standard deviation of the means (bilateral Student T-tests; p = 0.0367).

• 

  Fig. 2. DDX3X modulates HSV-1 exit from the perinuclear space. HeLa cells were seeded on 6-well plates and treated with a control non-targeting NC1 or b siDDX3X for 48 h then infected with wild-type strain 17+ virus at an MOI of 5. At 12 hpi, cells were fixed and prepared for electron microscopy (see Methods). The images show whole cell views with zooms on A-, B- and C-nuclear capsids as well as viral particles in the cytoplasm and the cell surface. c Fourteen cells were randomly selected from each infected condition in two independent experiments (N1; N2). A-, B- and C-nuclear capsids as well as viral particles in perinuclear space (PNS), cytoplasm and extracellular space were counted. The percentages (±standard error of the means) refer to the relative amounts of nuclear capsids (% Nuclear capsids) or the total viral particles containing the viral genome (% DNA-filled particles, i.e., C-capsids, perinuclear virions, cytoplasmic viral particles, and extracellular virus). Bilateral Student T-tests were performed. *p < 0.05 and **p < 0.01.

• 

  Fig. 3. The DDX3X carboxyl tail DDX3X physically interacts with the HSV-1 nuclear egress protein pUL31. The interaction of DDX3X and pUL31 was assessed using co-IP. a HeLa cells were either mock-treated or infected with wild-type strain F virus (negative controls) or with HA-tagged pUL31 virus at an MOI of 5 for 12 hpi. Cells were lysed and HA-tagged pUL31 pulled down using anti-HA agarose beads. Total lysates and eluates were analyzed by WB by probing pUL31 (anti HA or pUL31) and DDX3X (anti-human DDX3 rabbit R648 polyclonal serum) as indicated to the left of the blots. b The reverse co-IP was also performed with HeLa cells either mock-treated or infected as above but by precipitating DDX3X (anti-human DDX3 rabbit R648 polyclonal serum and protein A agarose beads). The presence of DDX3X and pUL31 in the eluates were assessed by Western blot using anti-DDX3X R648 or anti-pUL31 as shown to the left of those blots. c Schematic map of DDX3X showing its different domains and motifs (adapted from103). IDR, NTE and CTE represent intrinsically disordered domains, N- and C-terminal extensions respectively. The helicase core (residues 182-544) represents the minimally active portion of DDX3X. All…

• 

  Fig. 4. DDX3X colocalises with large VP5 aggregates at the nuclear rim distinct from the viral replication compartments. HeLa cells seeded on coverslips were either a mock-treated or b–d infected wild-type strain 17+ virus at an MOI of 5 for 9 hpi. Cells were fixed and labelled for a, b DDX3X (red) and VP5 (green) c DDX3X (red) and ICP4 (green) or d gB (red) and VP5 (green). All nuclei were stained with Hoechst 33342 (blue). Cells were imaged using dual colour high-resolution STED microscopy. Scale bar represents 5 µm. The images are representative of three independent experiments.

• 

  Fig. 5. VP5 fails to concentrate at the nuclear rim in the absence of DDX3X. a HeLa cells grown on coverslips were transfected with 25 nM of siDDX3X or non-targeting control siRNA (NC1) for 48 h. Cells were then infected wild-type strain 17+ virus at an MOI of 5 for 9 hpi. Fixed cells were labelled for DDX3X (magenta) and VP5 (red) and the nuclei stained using Hoechst 33342 (blue). This panel shows representative images of the different VP5 phenotypes observed and their 3D rendered illustrations (see Methods). The rendered images (Z-stacks acquired on a conventional confocal microscope). Scale bar represents 5 µm. b Eight random images (containing a total of about 200 cells) were analyzed for each condition to quantify the above VP5 phenotype. The bars indicate means and the standard error of the means. Bilateral Student T-tests, **p < 0.01; ***p < 0.001; NS: not significant. c Twenty random cells per phenotype (six cells for the less abundant small nuclear foci in siDDX3X treated cells) were analyzed on the Imaris software to compare the distribution of the VP5 signals (% intensity sum) in relation to their distance from the nuclear rim (µM). Each point represents the mean of three…

• 

  Fig. 6. DDX3X and the large VP5 nuclear aggregates are C-capsids dependent. a HeLa cells were mock-treated or infected at an MOI of 5 with the viral mutant ∆1-50 UL25 (only makes A and B nuclear capsids) or corresponding KOS wild-type virus. At 9 hpi, cells were fixed and processed for immunofluorescence (DDX3X: red; VP5: green; Hoechst 33342 stained nuclei: blue). Images were acquired on a conventional confocal microscope. The scale bar represents 5 µm. b Ten random images (~200 cells per condition) were used to quantify the different VP5 phenotypes. The percentages are means of three independent experiments along with their standard deviation (bilateral Student T-tests; ***p < 0.001).

• 

  Fig. 7. The presence of DDX3X on nuclear membranes is C-capsid dependent. a HeLa cells seeded on coverslips were either mock-treated or infected at an MOI of 5 with the ∆VP5 or ∆1-50 UL25 mutant viruses or the corresponding KOS wild-type virus. The cells were fixed at 9 hpi and labelled for DDX3X (red), pUL34 (green) and Hoechst 33342 stained nuclei (blue). Zoom boxes were added to highlight DDX3X and pUL34 localisation in each condition. All scale bars represent 5 µm. b Quantification of the nuclear pUL34 and DDX3X using 100 random cells per condition. Bars indicate means and the standard deviation of the means (bilateral Student T-tests; ***p < 0.001). The data represents three independent experiments.

• 

  Fig. 8. Large elongated VP5 foci at the nuclear periphery are enveloped C-capsids. a HeLa cells grown on coverslips prior to infection. Twenty-four hours later, cells were infected with the ∆US3, ∆UL31 or ∆UL34 mutant viruses or corresponding wild-type virus (F strain) at an MOI of 5. At 9 hpi the cells were treated for conventional confocal microscopy and labelled for DDX3X, VP5 and pUL34 (see Fig. S7 for IF images). At least 100 cells per condition were analyzed to quantify the intracellular VP5 distribution. The percentages are averages and standard deviations of the means of three individual experiments. Bilateral Student T-tests were performed and compared to wild-type virus (*p < 0.05, **p < 0.01 and ***p < 0.001). b HeLa cells were infected with wild-type strain 17+ virus at an MOI of 5 for 9 h. Next, cells were fixed and treated for electron microscopy. Several cells were imaged using by transmission electron microscopy. The images are representative of three independent experiments.

• 

  Fig. 9. DDX3X promotes optimal pUs3 virion incorporation. a HeLa cells were grown on 245 mm plates and treated with control non-targeting NC1 or siDDX3X for 48 h. Transfected cells were mock-treated or infected with wild-type strain 17+ virus. At 24 hpi, supernatant containing the mature virions were collected and concentrated at 60,000 x g. They were next analyzed on 5–20% gradient SDS-PAGE and Western blotting, using a panoply of antibodies targeting various viral tegument proteins. Total lysates of the same cells collected in lysis buffer and analyzed in parallel served as negative (mock cells) and positive (infected cells) antibody controls. DDX3X depletion in the cell and virions was confirmed by probing DDX3X in the same samples. b Image Lab version 5.0 was then used to quantify the Western blots (note that detection of the proteins was done on an instrument with a 4-log linear range, not on X-ray film). VP5, which served as a loading control, was used to normalize the data of the infected samples. The bars indicate means of three independent experiments and the standard error of the means (bilateral Student T-tests; *p < 0.05). pUS3 was the only statistically different tegument in…

• 

  Fig. 10. DDX3X physically interacts with pUS3. HeLa cells were either mock treated or infected with wild-type strain 17+ virus for 12 h. Cells were then fractionated into cytoplasm, nuclear matrix and nuclear membranes as detailed in Methods and a probed by WB for the GAPDH and LBR cytoplasmic and nuclear membrane markers respectively and b used to immunoprecipitate DDX3X using the R648 polyclonal serum and monitored for the presence of pUS3. NFA: no first antibody. In parallel, total lysates were loaded on the gels. Results represent three independent experiments.