PML plays both inimical and beneficial roles in HSV-1 replication

Generation and Properties of PML^−/− 1D2 Clone Derived from HEp-2 Cells.

PML is a family of seven isoforms. The largest, PML I, consists of nine exons (29–31). PML^−/− cell clones were generated from HEp-2 cells by transfection of clustered regularly interspaced short palindromic repeats [clustered regularly interspaced short palindromic repeats (CRISPR)/cas9 cassette] targeting exon 1 of PML (32–34). The procedure for drug selection and flow cytometry were both performed according to the manufacturer’s instructions and are briefly outlined in Materials and Methods. Of the numerous clones derived from single cells, we selected clone ID2 for further study.

Two series of experiments were done to verify the absence of PML in 1D2 cells. In the first, we took advantage of earlier studies showing that PML is amplified in cells exposed to IFN-β (26). Specifically, after overnight exposure to IFN-β, extracts of isolated nuclei or total cell lysates were solubilized, subjected to electrophoresis in a denaturing polyacrylamide gel, transferred to a nitrocellulose sheet, and reacted with rabbit polyclonal anti-PML antibody. As shown in Fig. 1A, the antibody to PML reacted with multiple protein bands present in HEp-2 cells. The antibody did not react with proteins present in lysates of 1D2 cells.

In the second series of experiments, parental HEp-2 or ID2 cells grown on four-well slides and exposed overnight to IFN-β were reacted with mouse monoclonal antibody to PML (Fig. 1 B, a and b) or with rabbit polyclonal antibody to PML (Fig. 1 B, c and d). HEp-2 cells exhibited typical intranuclear clusters of ND10 bodies containing PML. No such structures were observed in 1D2 cells. We did note that ID2 cells contained a ring of green fluorescent puncta observed in cells reacted with the mouse monoclonal antibody. These rings of fluorescent puncta in the perinuclear space were also present in parental HEp-2 cells but were overshadowed by the ND10 bodies. The presence of the fluorescent puncta in both HEp-2 and 1D2 cells suggests they are not derived from truncated PML proteins. Moreover, the observation that these puncta were observed in cells reacted with mouse monoclonal antibody, but not with the rabbit polyclonal antibody against PML, reinforces the conclusion that the antigens reacting with the monoclonal antibody are not derived from PML. We conclude that PML is not expressed in 1D2 cells.

Effect of IFN-β on Accumulation of Sp100 and Daxx mRNAs.

In the experiments described here, we examined the distribution of Sp100 and Daxx in parental HEp-2 cells and in the PML^−/− 1D2 clone. Here (Fig. 2), we report that exposure of both parental and 1D2 mutant cells to IFN-β enhanced the accumulation of Sp100 but had no significant effect on the accumulation of Daxx in either the parental HEp-2 or 1D2 cells. The procedures used in this study are described in Materials and Methods.

The Intranuclear Distribution of Sp100 and Daxx in Parental HEp-2 and in Clone 1D2 PML^−/− Cells.

In this series of experiments, parental HEp-2 or 1D2 cells were grown on four-well slides in the presence of growth medium alone or medium supplemented with IFN-β (1,000 U/mL). After 24 h, the cells were fixed and reacted with monoclonal antibody to PML and polyclonal antibody to Sp100 (Fig. 3A) or monoclonal antibody to Daxx and polyclonal antibody to Sp100. The cells were examined with the aid of a Zeiss confocal microscope. The results may be summarized as follows. First, in the absence of IFN-β, the accumulation of Sp100 was weak and largely perinuclear (Fig. 3 A, b, c, e, f, and B, b, c, e, f). There were few ND10 structures in parental HEp-2 cells reacted with antibody to PML and Sp100 (Fig. 3 A, c), and these were more readily seen in cells reacted with antibodies against Daxx and Sp100 (Fig. 3 B, c). As expected, the monoclonal antibody to PML reacted with an antigen localized in the perinuclear space of 1D2 cells (Fig. 3 A, d and e) in close proximity to Sp100. The structures containing Sp100 and Daxx in parental HEp-2 cells could not be differentiated from those detected in nuclei of 1D2 cells.

Second, the number of ND10 bodies was consistently higher in HEp-2 cells treated with IFN-β (Fig. 3 A, i and B, i). As shown in the enlarged boxed area (Fig. 3 B, i, Inset), the intranuclear fluorescent structures were relatively uniform in size and contained both Sp100 and Daxx.

Finally, in 1D2 cells treated with IFN-β, the distribution of antigens reacting with anti-PML monoclonal antibody (Fig. 3 A, i) was similar to that observed in untreated cells (Fig. 3 A, a). In the IFN-β treated cells, Sp100 and Daxx colocalized in structures that varied in size (Fig. 3 A, k and, l and B, k and l), but on average, they were significantly larger than the ND10 bodies observed in parental HEp-2 treated with IFN-β (Fig. 3 A, l and B, l).

The results of the studies presented here suggest that the colocalization of Sp100 and Daxx is not dependent on the presence of PML. However, the variability of the structures that contain Sp100 and Daxx suggests that the relatively uniform sizes of ND10 structures in parental HEp-2 cells are defined by PML. To determine the role of ICP0 in defining the size of the aggregates early in infection, a monolayer culture of parental HEp-2 or 1D2 cells was exposed to 10 pfu HSV-1 per cell. After 2 h, the cultures were fixed and reacted with monoclonal antibody to PML and polyclonal antibody to ICP0. Typical images shown in Fig. 4 A–C indicate that the ND10 bodies containing both PML and ICP0 are uniform in size in HEp-2 cells. In contrast, ICP0 aggregates formed in infected 1D2 cells varied considerably in size (Fig. 4 D–F).

The Role of PML in the Degradation of Sp100.

Studies published elsewhere have shown that ICP0 is an E3 ligase that, in conjunction with UbcH5a ubiquitin-conjugating enzyme, degrades PML and Sp100 (9–11). Mutations in the ring finger located in exon 2 of ICP0 abolish the ligase activity (11). The question posed here is whether the absence of PML hinders the degradation of Sp100. We report two series of experiments. In the first, parental HEp-2 cells or 1D2 cells grown on four-well slides were infected with 10 pfu HSV-1(F) or a ring finger (RF) mutant that is bereft of ubiquitin ligase activity. At 3 h after exposure to the virus, the cells were fixed and reacted with antibody to ICP0 and to Sp100, as described earlier.

The results may be summarized as follows: First, Fig. 5 A–C shows two 1D2 cells. The cell labeled A contains numerous aggregates of Sp100 differing in size and relatively few aggregates of ICP0. As shown in Fig. 5C, only a few of the aggregates of ICP0 (green) colocalized with the aggregates of Sp100 (red). Moreover, as apparent in Fig. 5C, the colocalized structures did not fully merge. In the cell labeled B, there are few prominent Sp100 bodies. The one prominent Sp100 body colocalized but did not merge with the ICP0 (green) aggregates.

Second, Fig. 5 D–F shows two parental HEp-2 cells. The cell labeled C shows no signs of being infected. Thus, Sp100 aggregates are prominent, and no ICP0 can be seen. In contrast, there was little Sp100 and very prominent ICP0 aggregates in small uniform structures consistent with the size of ND10 bodies in the cell labeled D. The data are consistent with the hypothesis that Sp100 had been degraded.

Finally, in parental (PML^+/+) HEp-2 cells infected with RF mutant (Fig. 5 J–L), ICP0 and Sp100 aggregated in small nuclear structures relatively uniform in size indistinguishable from classical ND10 bodies. In contrast, in 1D2 cells, there was no evidence of large-scale colocalization of ICP0 and Sp100. Although Sp100 formed aggregates differing in size, ICP0 appeared to diffuse in the nucleus (Fig. 5 G–I).

The results presented here suggest that in PML^−/− cells, the colocalization of ICP0 and Sp100 is a slow process and may not result in a total mixing of the two proteins. The paucity of Sp100 in Fig. 5C (cell labeled “B”) suggests Sp100 was degraded, possibly by massive accumulation of ICP0.

To quantify the pattern of decay of SP100 in the two cell lines, replicate cultures of HEp-2 or ID2 cells were transfected with FLAG-tagged SP100, as described in Materials and Methods. After 48 h, the cells were exposed to 10 pfu HSV-1(F) per cell. Cultures harvested at time of exposure to the virus (Mock) or at 3, 5, 7, or 10 h after infection were solubilized, subjected to electrophoresis in denaturing gels, and probed with antibody to Sp100 and β-actin. Fig. 6A shows the immunoblots of Sp100 and β-actin. Fig. 6B shows the results of a density scan normalized with respect to Sp100 present in mock-treated cells. The results show that under conditions used in these studies, the initiation of decay of Sp100 in 1D2 cells was delayed by at least 3 h, in contrast to the apparent initiation of decay of Sp100, which extrapolates to a time at or shortly after exposure of the cells to the virus.

The Replication of Wild-Type Virus Is Impaired in 1D2 Cells Infected at Low Ratios of Virus per Cell.

The degradation of PML mediated by ICP0 E3 ligase early after infection suggests that PML is inimical to the replication of HSV-1. To test this hypothesis, we exposed replicate cultures of parental HEp-2 or 1D2 cells to either 0.1 or 5 pfu HSV-1(F) per cell. The cells were harvested at intervals between 2 and 48 h in the case of cultures exposed to 5 pfu per cell, or between 2 and 72 h in the case of cultures exposed to 0.1 pfu virus per cell. As shown in Fig. 7B, the virus yields from cultures exposed to 5 pfu per cell were virtually identical. In contrast, the yields of virus obtained 48 and 72 h after infection of 1D2 cells with 0.1 pfu per cell were ∼100-fold lower than those obtained in parental HEp-2 cells exposed to the same ratio of virus per cell (Fig. 7A). The results suggest that PML may play both a supportive and inimical role in HSV-1 replication.

The Accumulation of ICP0 Is Impaired in 1D2 Cells Infected at Low Ratios of Virus per Cell.

One hypothesis that could explain the reduced accumulation of virus in 1D2 cells at low multiplicities of infection is a reduction in the synthesis of ICP0. To test this hypothesis, replicate cultures of HEp-2 or 1D2 cells were infected with either 10 or 0.1 pfu virus per cell. At intervals after infection shown in Fig. 8, the cells were harvested, solubilized, subjected to electrophoresis in denaturing gels, transferred to a nitrocellulose sheet, and probed with antibody to ICP0 and β-actin. The results shown in Fig. 8 indicate that the rates of accumulations of ICP0 in HEp-2 or 1D2 cells exposed to 10 pfu per cell were approximately equal. In contrast, the accumulation of ICP0 in 1D2 cells lagged significantly behind the accumulation of ICP0 observed in HEp-2 cells. The significance of the results rests on two considerations: Foremost, the synthesis and accumulation of ICP0 is dependent on the presence of PML, even though ICP0 ultimately degrades PML, and second, the delay in the degradation of Sp100, as shown in Fig. 5, is not a result of the lack of ICP0, as in those experiments, the cells were exposed to 10 pfu per cell.

Effect of IFN-β on the Replication of HSV-1(F) in Parental HEp-2 and in 1D2 Cells.

Earlier studies have shown that IFN-β is more effective in suppressing HSV-1 replication in primary murine fibroblast cultures than in sibling cells in which PML was knocked out. The question arose whether IFN-β had a similar effect in human cells, and in particular, in the parental HEp-2 and in PML^−/− 1D2 clone cells used in the studies reported here. In the experiments reported, replicate cultures of the two cell lines were exposed to 0.1 pfu HSV-1 per cell and then either incubated for 24 h without further treatment or maintained in medium containing IFN. The results are shown in Fig. 9. As expected from the results in Fig. 7, the yields of untreated 1D2 cells were lower than those obtained in parental HEp-2 cells. In IFN-β–treated cells, the yield of virus at 15 h was reduced 100-fold compared with those obtained in untreated infected cells. In contrast, IFN-β decreased the yield of virus in 1D2 cells 10-fold compared with yields obtained in infected untreated 1D2 cells. In essence, the results obtained in a human cell line paralleled those obtained in murine cells.

Cells and Viruses.

The sources and maintenance of HEP-2 and Vero cells were reported elsewhere (47). HSV-1(F) is the prototype strain used in our laboratory (48). RF-HSV (F) was a recombinant virus containing ring-finger domain deletion of both copies of ICP0 protein described elsewhere (12). All viruses were amplified in HEp-2 cells. Plaque assays were performed in Vero cells.

RNA Isolation and Real-Time PCR.

RNA extraction and assays were done as previously described (49). mRNA level of Sp100 and Daxx were measured by gene-specific primers: Sp100 forward primer, 5′-GCTCAGGACCCCAGATTGTAC-3′, and reverse primer, 5′-CTAATCTTCTTTACCTGACCC-3′, and Daxx forward primer, 5′-GACGGACATTTCCTCTTCCA-3′, and reverse primer, 5′-CGCCTCCATTGAAGGAAGTA-3′.

Transfection Plus Superinfection.

Parental HEp-2 and PML^−/− clone 1D2 cells were grown to 70–90% confluency in T150 flasks. Each flask was transfected with 18 μg of a plasmid DNA containing a CMV promoter-driven FLAG-tagged Sp100 ORF, using Lipofectamine LTX with Plus Reagent (Life Technologies, 15338100), according to manufacturer’s instructions. At 24 h after transfection, the cells contained in T150 flasks were split into T25 flasks. At 24 h after seeding, the mock-transfected and FLAG-tagged Sp100-transfected cells were mock infected or exposed to 10 pfu HSV-1(F) per cell. The cells were harvested and processed as described in Results.

Generation of PML^−/− Cell Line Using CRISPR.

A CRISPR kit targeting the first exon of PML gene was obtained from Origene (cat. no. KN200700). Parental HEp-2 cells were cotransfected with a gRNA vector plasmid targeting sequences 5′-CTCGGAGATCGGGCGGGTGC-3′ (150th–169th bp in exon 1; 4th–10th amino acid from the start codon) or 5′-GGCCTTCAGAGGGGGTCTCG-3′ (218th–237th bp in exon1; 26th–32nd amino acid from the start codon), and a donor vector containing a selection cassette expressing GFP–puromycin was inserted into the targeted site in human PML gene. Selected cells (GFP⁺ and puromycin resistant) were serially diluted to form single cell-derived colonies. The colonies were further screened for expression of PML by immunoblotting and immunofluorescence.

Immunoblot Analyses.

At indicated times, the cells were lysed, heat denatured, electrophoretically separated in denaturing gels, transferred to polyvinylidene difluoride membranes, and probed with the appropriate primary antibody, as previously described (49). The antibodies included monoclonal anti-PML (Santa Cruz, sc-966) (50), rabbit polyclonal anti-PML (Abcam, 79466), mouse monoclonal anti–β-actin (Thermo Fisher, MA5-15739), rabbit polyclonal anti-ICP0 (Lab stock), and mouse monoclonal anti-FLAG (Sigma, F1804). The membranes were then incubated with alkaline phosphatase-conjugated goat anti-mouse or goat anti-rabbit secondary antibody (Sigma) and visualized with BCIP/NBT Western Blotting Detection Reagent (GE Healthcare).

ImageJ Quantification and Statistics.

Image J software was used to quantify band intensity. FLAG-Sp100 levels were first normalized to β-actin. Degradation rates of FLAG-Sp100 were calculated by normalization to Sp100 in mock-treated cells. Error bars were calculated from the results of three independent quantifications of band intensities. Two-tailed P values were calculated using standard t test. To compare ICP0 expression level, the band intensity of ICP0 was first normalized to β-actin. For high multiplicity of infection, ICP0 expression level of all points examined was then normalized to that in HEp-2 at 24 h postinfection; for low multiplicity of infection, ICP0 expression level of all points examined was then normalized to that in HEp-2 at 48 h postinfection.

Immunofluorescence Staining and Confocal Microscopy.

Cells in four-welled slides were fixed and permeabilized, blocked, and then incubated with primary antibodies, as previously described (47). Those included antibodies anti-PML, as described earlier, rabbit anti-Sp100 (Abcam, ab43151); mouse anti-Daxx (Abcam, ab9091); and rabbit anti-ICP0 or mouse anti-ICP0, described elsewhere (47), at 4 °C overnight, rinsed three times with blocking solution, and subjected to reaction with fluorescein isothiocyanate-conjugated goat anti-mouse (Invitrogen) or/and Texas Red-conjugated goat anti-rabbit (Sigma) secondary antibodies. Images were taken at 100× magnification with a Zeiss confocal microscopy.