Differential role of Sp100 isoforms in interferon-mediated repression of herpes simplex virus type 1 immediate-early protein expression

Abstract

Nuclear domains called ND10 or PML nuclear bodies contain interferon (IFN)-upregulated proteins like PML and Sp100. Paradoxically, herpes simplex virus 1 (HSV-1) begins its transcriptional cascade at aggregates of ND10-associated proteins, which in turn are destroyed by the HSV-1 immediate-early protein ICP0. While PML is essential in the formation of ND10, the function of Sp100 in the cells' defense against viral infection is unknown. In this study we investigated the potential antiviral effect of IFN-beta-induced Sp100. We found that IFN-beta treatment leads to a differential accumulation of four Sp100 isoforms in different cell lines. Using an HEK293 cell line derivative, 293-S, producing no detectable amounts of Sp100 even after IFN exposure, we analyzed individual Sp100 isoforms for their effect on HSV-1 infection. Sp100 isoforms B, C, and HMG, but not Sp100A, suppressed ICP0 and ICP4 early after infection. Isoforms B, C, and HMG suppressed expression from the ICP0 promoter in transient transfection, whereas Sp100A enhanced expression. Moreover, Sp100A localized in ND10, whereas the repressive isoforms were either dispersed within the nucleus or, at unphysiologically higher expression levels, formed new aggregates. The repressive activity was dependent on an intact SAND domain, since Sp100B bearing a W655Q mutation in the SAND domain lost this repressive activity and accumulated in ND10. Using RNA interference to knock down the repressive Sp100 isoforms B, C, and HMG, we find that they are an essential part of the IFN-beta-mediated suppression of ICP0 expression. These data suggest that repression by the Sp100 isoforms B, C, and HMG takes place outside of ND10 and raise the possibility that viral genomes at Sp100A accumulations are more likely to start their transcription program because of a more permissive local environment.

FIG. 1.

Induction of Sp100 isoforms in response to IFN-β treatment. (A) Structural and functional domains of Sp100 isoforms. (B) Induction of ISG15 and ISG54 mRNA was assessed by semiquantitative RT-PCR analysis. HEp-2 cells were treated with 500 U/ml of IFN-β for 18 h before total cellular RNA was extracted. (C) Levels of Sp100 isoforms were analyzed by semiquantitative RT-PCR from the same samples. (D) Induction of Sp100 isoforms in different cell lines. Levels of GAPDH mRNA were used to control an equal amount of mRNAs. To control DNA contamination we used RNA samples without RT (-RT). Asterisks indicate positions of GAPDH primers. M, molecular size marker.

FIG. 2.

Nuclear distribution of Sp100 protein. (A to F) HEp-2 cells grown on coverslips were transfected with plasmid constructs expressing mRFP-tagged NLS (A), Sp100A (B), Sp100B (C), Sp100C (D), Sp100HMG (E), or Sp100B-W655Q (F) isoforms and analyzed by IF assay with MAb PG-M3 (green) against endogenous PML. (G to L) HEp-2, HEK293, and 293-S cells grown on coverslips were treated with 1,000 U/ml of IFN-β for 18 h (J, K, and L) or left untreated (G, H, and I) and double-labeled with polyclonal Ab 1380 against endogenous Sp100 (green) and MAb PG-M3 against endogenous PML (red). Colocalization appears as a yellow signal. Note that we artificially increased the background on panels H and K to detect the lowest amount of Sp100.

FIG. 3.

Accumulation of Sp100 isoforms in HEp-2, HEK293, and 293-S cells. Whole-cell extracts (A) or nuclear fractions (B) of replicate HEp-2, HEK293, and 293-S cell cultures were stimulated with 500 U/ml of IFN-β for 18 h or left untreated as a control. 293-S cells were transfected with 2 μg of pSG5 vector DNA or constructs expressing Sp100A (A), Sp100B (B), Sp100HMG (HMG), or Sp100C (C). Note that sumoylation of Sp100 proteins, indicated by asterisks, is preserved in whole-cell extracts but not in nuclear fractions. WB, Western blot.

FIG. 4.

Effect of IFN-β treatment on expression of IE ICP0. (A and B) Replicate HEp-2, HEK293, and 293-S cell cultures were stimulated with 1,000 U/ml of IFN-β for 18 h or left untreated as a control. Half of each replicate was mock infected or inoculated with 0.5 PFU of HSV-1 (strain 17+) per cell as described in Materials and Methods. After 3 h of infection, cells were harvested, and total cell extract and the nuclear fractions (30 μg/lane) were electrophoretically separated on denaturing polyacrylamide gels, transferred to nitrocellulose membranes, and probed with antibodies to ICP0 (A) or ICP4 (B). WB, Western blot. (C) HEp-2, HEK293, and 293-S cells were transiently cotransfected with ICP0 promoter-luciferase and stimulated 24 h later with 1,000 U/ml of IFN-β for 18 h or left untreated. Cells were harvested and fractionated on nuclear and cytoplasmic fractions. Firefly luciferase activity was measured in the cytoplasmic fractions and normalized to protein concentration. Responsiveness to IFN-β was calculated as the ratio of normalized luciferase activity of treated cells to that of untreated cells.

FIG. 5.

Effect of Sp100 isoforms on expression of IE HSV-1 proteins. (A) 293-S cells grown on coverslips were transfected with plasmid constructs expressing mRFP-tagged NLS (a), Sp100A (b), Sp100B (c), Sp100C (d), Sp100-HMG (e), or Sp100B-W655Q (f). Eighteen hours later, cells were inoculated with HSV-1 (strain 17+) as described in Materials and Methods. After 3 h of infection, cells were fixed and analyzed by IF assay with rabbit anti-ICP0 antibodies (green). (B) 293-S cells grown on coverslips were transfected with plasmid constructs expressing mRFP-tagged NLS (pmRT), Sp100A, Sp100B, Sp100C, Sp100-HMG, or Sp100B-W655Q (Sp100BQ). Eighteen hours later, cells were inoculated with HSV-1 (strain 17+) and analyzed by counting cells labeled for ICP0 or ICP4 in cells expressing the respective Sp100 isotype. At 3 h postinfection approximately 1,000 cells expressing red fusion protein were analyzed for the presence of ICP0 or ICP4. Bars indicate the numbers of those Sp100-expressing cells that exhibited ICP0 or ICP4 staining relative to the vector control (pmRT). (C) Western blot analysis of the nuclear fractions (30 μg/lane) from the above experiment were electrophoretically separated, transferred to nitrocellulose membranes, and probed with antibodies to ICP0 (top), ICP4 (middle), or Sp100 (bottom).

FIG. 6.

Effect of Sp100 isoforms on luciferase expression from the ICP0 promoter. (A) 293-S cells were transiently cotransfected with luciferase reporter plasmid under the ICP0 promoter, minimum thymidine kinase promoter-β-galactosidase reporter, and increasing amounts (0.5, 1, and 2 μg) of effector plasmids expressing HA-tagged Sp100 isoforms A, B, and C. (B) 293-S cells were transiently cotransfected with luciferase reporter plasmid under the ICP0 promoter and increasing amounts (0.1, 0.3, and 1 μg) of effector plasmids expressing HA-tagged Sp100 isoforms A, B, BQ (the W655Q mutant of isoform B), and HMG. Cells were harvested 48 h later, and luciferase activity was determined from the cytoplasmic fraction and normalized to protein concentration (top). Results are presented as percentages of luciferase activity of the pSG5 vector. Mean values (± standard deviation) are from three independent experiments. Nuclear fractions (30 μg/lane) were electrophoretically separated on denaturing polyacrylamide gels, transferred to nitrocellulose membranes, and probed with antibodies to Sp100 to assure similar levels of expression. WB, Western blot.

FIG. 7.

Effect of depletion of Sp100 isoforms containing SAND domains on IFN-β-mediated suppression of ICP0 expression. (A) Knockdown of Sp100 mRNA. Total RNA from HEp-2, and stable cell lines based on lentiviral infected HEp-2 cells, was extracted and subjected to semiquantitative RT-PCR for Sp100 isoforms and GAPDH. (B) Replicates of HEp-2 and stable cell lines expressing Sp100si-2 or NT2si were stimulated with 1,000 U/ml of IFN-β or left untreated as a control. Eighteen hours later cells were inoculated with 0.5 PFU of HSV-1 (strain 17+) per cell as described in Materials and Methods. After 3 h of infection, cells were harvested, and the nuclear fractions (30 μg/lane) were electrophoretically separated, transferred to nitrocellulose membranes, and probed with antibodies to ICP0 (top). Ezh2 was used as a loading control (bottom).