Epigenetic regulation of polyomavirus JC involves acetylation of specific lysine residues in NF-κB p65

Abstract

Progressive multifocal leukoencephalopathy (PML) is a fatal demyelinating disease caused by neurotropic polyomavirus, JC virus (JCV), a virus that causes lytic infection of CNS glial cells. After primary infection, JCV is controlled by the immune system but virus persists asymptomatically. Rarely, when immune function is impaired, it can reemerge to cause PML. The mechanisms of JCV persistence and reactivation are not well understood but our earlier work implicated epigenetic control by protein acetylation since histone deacetylase inhibitors such as trichostatin A (TSA) strongly stimulate JCV transcription. Since both TNF-α and TSA activate JCV transcription via the same unique NF-κB site in the JCV control region, we investigated a role for acetylation of NF-κB in JCV regulation. A site-directed mutagenesis strategy was employed targeting the known lysine acetylation sites of NF-κB p65: K218, K221, and K310. We individually mutated each lysine to arginine, which cannot be acetylated and retains a positive charge like lysine. K218R and K221R impaired transactivation of JCV early promoter transcription either alone or combined with TSA treatment or coexpression of acetyltransferase transcriptional coactivator p300 but K310R was largely without effect. Mutation of lysine to glutamine gives mutants with a negative charge like acetyllysine. However, K218Q and K221Q showed impaired activity and only K310Q showed enhanced transactivation. NF-κB acetylation can regulate several aspects of the process of activation including complex formation with IκB, translocation to the nucleus, and DNA binding and transcriptional activation. Cell fractionation studies revealed that the mutants had no defect in translocation to the nucleus whereas gel shift studies revealed reduced binding to the JCV NF-κB site. Thus, acetylation regulates NF-κB p65 activity toward JCV at the level of p65 binding to the JCV control region and activation of JCV transcription.

Keywords: JC Virus; Progressive multifocal leukoencephalopathy; Viral persistence; Viral reactivation.

Figure 1. Effect of p65 Lys to Arg mutants on transactivation of the JCV early promoter

TC620 cells were transfected with JCV_E-Luc reporter plasmid and NF-κB p65 expression plasmid (wild-type, K218, K221R or K310R) and then untreated or treated with 300 ng/ml TSA 24 h later. After a further 24 h, cells were harvested and assayed for luciferase activity as described in Methods. Activities were normalized to untreated controls. The bar represents one standard deviation. The histogram at the bottom shows quantification of the Western. The experiment was performed at least twice.

Figure 2. Effect of coexpression of p300 and p65 Lys to Arg mutants on transactivation of the JCV early promoter

TC620 cells were transfected with the reporter plasmid JCV_E-LUC and NF-KB p65 expression plasmid (wild-type, K218R, K221R or K310R) together with expression plasmid for p300 (0, 0.25 and 0.5 μg), harvested and assayed for luciferase activity. Activities were normalized to untreated controls. The bar represents one standard deviation. The histogram at the bottom shows quantification of the Western. The experiment was performed at least twice.

Figure 3. Effect of p65 Lys to Gln mutants on transactivation of the JCV early promoter

TC620 cells were transfected with JCV_E-Luc reporter plasmid and NF-κB p65 expression plasmid (wild-type, K218Q, K221Q or K310Q) and then untreated or treated with 300 ng/ml TSA 24 h later. After a further 24 h, cells were harvested and assayed for luciferase activity as described in Methods. Activities were normalized to untreated controls. The bar represents one standard deviation. The histogram at the bottom shows quantification of the Western. The experiment was performed at least twice.

Figure 4. Effect of coexpression of p300 and p65 Lys to Gln mutants on transactivation of the JCV early promoter

TC620 cells were transfected with the reporter plasmid JCV_E-LUC and NF-κB p65 expression plasmid (wild-type, K218Q, K221Q or K310Q) together with expression plasmid for p300 (0, 0.25 and 0.5 μg), harvested and assayed for luciferase activity. Activities were normalized to untreated controls. The bar represents one standard deviation. The histogram at the bottom shows quantification of the Western. The experiment was performed at least twice.

Figure 5. Effect of p65 Lys to Arg and Lys to Gln mutants on subcellular localization of p65

TC620 cells were transfected with NF-κB p65 expression plasmid (wild-type, K218R, K221R, K310R, K218Q, K221Q or K310Q), treated with and without TNF-α and then harvested for nuclear and cytoplasmic fractions as described in Materials and Methods. Western blots are performed for each fraction for p65 and purity and equal loading was assessed by Western for cytoplasmic α-tubulin and nuclear lamin. a. K to R mutants. b. K to Q mutants. The line between lanes 12 and 13 in each panel indicates that samples 1–12 and 13–20 were run in parallel on separate gels. The experiment was performed twice.

Figure 6. Effect of p65 Lys to Arg and Lys to Gln mutants on binding of p65 to the JCV NF-κB binding site

a. TC620 cells were transfected with NF-κB p65 expression plasmid (wild-type, K218R, K221R, K310R, K218Q, K221Q or K310Q) for 48 h and nuclear extracts harvested. An oligonucleotide corresponding to the JCV NCCR KB element was used to perform EMSA using nuclear extracts from untreated cells (Control) or cells transfected with expression vector for p65, untreated or treated with TSA overnight as described in Methods. For supershifts, antibody to p65 (α-p65) or nonimmune rabbit serum (NRS) was added as indicated. Competitor–excess unlabelled probe was added. Unbound probe (loaded alone in lane 1) runs near the bottom of the gel (position 4). b. Western blot for the expression of p65 is shown. The experiment was performed twice.