Interferon-γ-inducible protein 16 (IFI16) is required for the maintenance of Epstein-Barr virus latency

Abstract

Background: Epstein-Barr virus (EBV) exhibits both lytic and latent (Lat. I, II, and III) phases in an infected individual. It's during the latent phase of EBV that all EBV-associated cancers, including Burkitt's lymphoma, nasopharyngeal carcinoma and lymphoproliferative disease arise. Interferon-γ-inducible protein 16 (IFI16) is a well-established innate immune sensor and viral transcriptional regulator involved in response to invading DNA viruses. During latency, IFI16 remains in the nucleus, in part bound to the EBV genome; however, neither its role in EBV lytic cycle or latency has been established.

Methods: Short interfering RNA against IFI16 and IFI16 overexpression were used to identify the role of IFI16 in the maintenance of EBV latency I. We also studied how induction of the lytic cycle affected IFI16 using the EBV positive, latently infected Akata or MUTU-1 cell lines. Akata cells were induced with TPA and MUTU-1 cells with TGF-β up to 96 h and changes in IFI16 protein were analyzed by Western blotting and immunofluorescence microscopy. To assess the mechanism of IFI16 decrease, EBV DNA replication and late lytic transcripts were blocked using the viral DNA polymerase inhibitor phosphonoacetic acid.

Results: Knockdown of IFI16 mRNA by siRNA resulted in enhanced levels of EBV lytic gene expression from all temporal gene classes, as well as an increase in the total EBV genome abundance, whereas overexpression of exogenous IFI16 reversed these effects. Furthermore, 96 h after induction of the lytic cycle with either TPA (Akata) or TGF-β (MUTU-1), IFI16 protein levels decreased up to 80% as compared to the EBV-negative cell line BJAB. Reduction in IFI16 was observed in cells expressing EBV lytic envelope glycoprotein. The decreased levels of IFI16 protein do not appear to be dependent on late lytic transcripts of EBV but suggest involvement of the immediate early, early, or a combination of both gene classes.

Conclusions: Reduction of IFI16 protein levels following lytic cycle induction, as well as reactivation from latency after IFI16 mRNA knockdown suggests that IFI16 is crucial for the maintenance of EBV latency. More importantly, these results identify IFI16 as a unique host factor protein involved in the EBV lifecycle, making it a potential therapeutic target to combat EBV-related malignancies.

Keywords: Epstein-Barr virus (EBV); Herpesvirus; Interferon-γ-inducible protein 16 (IFI16); Latency; Lytic cycle.

Fig. 1

Knockdown of IFI16 in EBV latency I Akata cells results in lytic cycle induction. a Immunoblot showing the efficient knockdown of IFI16 in Akata cells at 48 h post siRNA transfection. b & e Real-time qRT-PCR of IFI16 mRNA levels 48 h after transfection of siRNA (b) or IFI16 overexpression (e). mRNA levels were normalized against β–tubulin and expressed as relative amounts compared to siC (control) treatments. c & f Real-time DNA PCR of relative EBV genome abundance. Primers specific to the latent EBNA1 gene were used and the level of DNA was normalized against β-tubulin levels. d & g Real-time qRT-PCR for gene-specific primers for the four major classes of EBV genes (immediate early, early, late, and latent) 48 h after siRNA transfection (d) or IFI16 overexpression and TPA treatment (g). mRNA levels were normalized against β-tubulin mRNA levels and data are expressed as the relative amount as compared to the siC treatments. Results are represented as means ± SD of data from three independent experiments and statistical analysis was done with a Student’s t test. **, P < 0.01; ***, P < 0.001

Fig. 2

Activation of EBV lytic cycle by TPA in latency I infected Akata cells. a Immunoblot of IFI16, EBV (Ea-D p52/50), and GAPDH in the EBV-positive Akata cells either untreated or treated with 30 ng/ml TPA for the indicated time. All bands were normalized to their respective level of GAPDH and the fold change was calculated relative to the uninduced sample. h.p.i., hours post induction. b EBV-negative BJAB cells were left uninduced or induced with 30 ng/ml TPA for the indicated time, followed by immunoblotting as done in A. c Real time DNA PCR for the relative EBV genome copy number. Primers specific to the latent EBNA1 gene were used and the level of DNA was normalized against β-tubulin. d & e Real time qRT-PCR of mRNA levels from uninduced to 96 h post TPA induction in Akata cells for IFI16 (d) or the four major gene classes of EBV genes in Akata cells (e). Total RNA was extracted and mRNA levels analyzed by real-time RT-PCR using gene-specific primers that were normalized against β-tubulin mRNA levels. All values are represented as the relative amount compared to the uninduced control (time zero). Results are presented as ± SD of data from at least three independent experiments and statistical analysis was done with a Student’s t test. **, P < 0.01; ***, P < 0.001; ns, not significant

Fig. 3

Immunofluorescence analysis of the effect of lytic cycle induction by TPA on IFI16. BJAB (a) or Akata (b) cells were treated with 30 ng/mL of TPA or left untreated (uninduced) for 96 h. Cells were harvested, plated onto 10-well slides, fixed with ice-cold acetone, thoroughly washed, and blocked for 30 min at room temperature. Slides were stained with primary antibodies against the EBV glycoprotein gp350/220 (green) to visualize infected cells undergoing lytic replication and anti-IFI16 (red). Cell nuclei were visualized by DAPI staining (blue). Magnification, 40X. Red boxes are enlarged in the far right panels with white arrows indicating IFI16 puncta and yellow arrows showing gp350/220 puncta. A representative experiment out of three is shown

Fig. 4

EBV lytic cycle induction by TGF-β in latency I infected MUTU-1 cells. MUTU-1 cells were either uninduced (time zero) or treated with TGF-β (5 ng/mL) for the indicated time. a Immunoblots for IFI16 (top), ZEBRA (BZLF1; middle), or GAPDH (bottom) following induction. All bands were normalized to GAPDH with the fold change calculated relative to the uninduced sample. b EBV-negative BJAB cells were treated with TGF-β similar to MUTU-1 cells or left untreated and immunoblotted for IFI16, ZEBRA, and GAPDH. c Real time DNA PCR for the relative EBV genome abundance. Primers specific to the latent EBNA1 gene were used and the level of DNA was normalized against the β-tubulin level. d Real time qRT-PCR for mRNA levels in uninduced to 96 h post TGF-β induction in MUTU-1 cells. Total RNA was extracted and mRNA levels analyzed by real-time qRT-PCR using gene-specific primers to EBV temporal class genes. mRNA values were normalized against β-tubulin mRNA levels. All values are represented as the relative amount compared to the uninduced (time zero) control. Results are presented as ± SD of data from three independent experiments and statistical analysis was done with a Student’s t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant

Fig. 5

TGF-β-induced lytic cycle in latency I infected MUTU-1 cells results in decreased levels of IFI16 protein. BJAB (a) or MUTU-1 cells (b) were induced with TGF-β (5 ng/mL) or left untreated for 96 h. 96 h after induction, cells were harvested and prepared for IFA as described previously (Fig. 3). Slides were stained with primary antibodies to IFI16 (red) and EBV glycoprotein gp350/220 (green). Cell nuclei were visualized by DAPI staining (blue). Magnification, 40X. Red boxes are enlarged in the far right panels with white arrows indicating IFI16 puncta and yellow arrows showing gp350/220. A representative experiment out of three is shown

Fig. 6

IFI16 protein reduction is not dependent on late lytic transcripts. a Akata cells were left untreated (lane 1), incubated with 100 μg/mL PAA for 96 h (lane 2), with 30 ng/mL TPA for 96 h (lane 3), or with 100 μg/mL PAA for 1 h followed by the addition of 30 ng/mL TPA for an additional 96 h (lane 4), then immunoblotted for anti-IFI16 or anti-GAPDH. b MUTU-1 cells were treated the same as in A, except with 300 μg/mL of PAA and with 5 ng/mL of TGF-β in place of TPA. b & d Real time DNA PCR was performed to analyze relative EBV genome abundance in Akata (b) or MUTU-1 (d) cell lines. Primers specific to the latent EBNA1 gene were used and the level of DNA was normalized against the β-tubulin level. Results are presented as ± SD of data from three independent experiments and statistical analysis was done with a Student’s t test. ***, P < 0.001; ns, not significant