JC virus promoter/enhancers contain TATA box-associated Spi-B-binding sites that support early viral gene expression in primary astrocytes

Abstract

JC virus (JCV) is the aetiological agent of the demyelinating disease progressive multifocal leukoencephalopathy, an AIDS defining illness and serious complication of mAb therapies. Initial infection probably occurs in childhood. In the working model of dissemination, virus persists in the kidney and lymphoid tissues until immune suppression/modulation causes reactivation and trafficking to the brain where JCV replicates in oligodendrocytes. JCV infection is regulated through binding of host factors such as Spi-B to, and sequence variation in the non-coding control region (NCCR). Although NCCR sequences differ between sites of persistence and pathogenesis, evidence suggests that the virus that initiates infection in the brain disseminates via B-cells derived from latently infected haematopoietic precursors in the bone marrow. Spi-B binds adjacent to TATA boxes in the promoter/enhancer of the PML-associated JCV Mad-1 and Mad-4 viruses but not the non-pathogenic, kidney-associated archetype. The Spi-B-binding site of Mad-1/Mad-4 differs from that of archetype by a single nucleotide, AAAAGGGAAGGGA to AAAAGGGAAGGTA. Point mutation of the Mad-1 Spi-B site reduced early viral protein large T-antigen expression by up to fourfold. Strikingly, the reverse mutation in the archetype NCCR increased large T-antigen expression by 10-fold. Interestingly, Spi-B protein binds the NCCR sequence flanking the viral promoter/enhancer, but these sites are not essential for early viral gene expression. The effect of mutating Spi-B-binding sites within the JCV promoter/enhancer on early viral gene expression strongly suggests a role for Spi-B binding to the viral promoter/enhancer in the activation of early viral gene expression.

Fig. 1.

Diagram of potential Spi-B-binding sites outside the promoter/enhancer region within the JCV NCCR. The NCCRs from the PML-associated Mad-1 (a), Mad-4 (b) and the non-pathogenic archetype (c) JCVs are represented. The PML-associated promoter/enhancers contain 98 bp tandem repeats. A 19 bp deletion in the second repeat of Mad-4 (diagonal stripes) results in the loss of the second TATA box. The promoter/enhancer of the archetype variant contains two inserts (spotted) in a single 98 bp unit. Sites that bound Spi-B protein in EMSA assays are grey and sites that did not bind protein are white. Sites read on the late strand are named with an L and the sites read on the early strand are named with an E. (d) The Spi-B-binding site sequences are listed according to location in the pre-origin sequence, the promoter or the post-promoter sequence.

Fig. 2.

Spi-B binds sites present outside the promoter/enhancer within the JCV NCCR. EMSAs were performed using BJAB, KG-1a or PDA cell extract and biotin-labelled oligonucleotide probes for the JCV L2, L6, L8 Spi-B-binding sites (a) and the SV40 Spi-B-binding site as a positive control for Spi-B protein–DNA complex formation. Representative EMSAs using: BJAB cell extract and biotin labelled oligonucleotide probes for the SV40 (b) and JCV L1 (c) Spi-B-binding sites; or PDA cell extract for the L18 (d), E1 (e) and E2 (f) Spi-B-binding sites are shown. Biotinylated authentic binding site probe was incubated with cell extract alone or in combination with 200-fold excess unlabelled oligonucleotide competitor or Spi-B antiserum (b–f). Biotinylated mutant probe was incubated with cell extract to demonstrate specificity for the Spi-B-binding site core (b–f). The results presented are representative of three independent experiments.

Fig. 3.

Mutation of Spi-B sites outside the promoter/enhancer region does not affect early viral activity in PDAs. JCV T-antigen mRNA is represented as total copies per 50 ng of RNA along with sd (light grey). The number of T-antigen protein-positive cells was quantified on a per cell basis and is represented as a fold-change over the wild-type Mad-4 plasmid along with sd (dark grey).

Fig. 4.

Deletion of the second repeat in the Mad-4 promoter/enhancer reduces early viral activity in PDAs. The NCCR sequences for Mad-4 (a), single a–c–e Mad-4 (b) or L5 mutant a–c–e Mad-4 (c) are represented in a diagram form. JCV T-antigen mRNA is represented as 2×10⁵ copies per 50 ng RNA along with sd (d, light grey). The number of T-antigen protein-positive cells was quantified on a per cell basis and is represented as a fold-change over the wild-type Mad-4 plasmid along with sd (d, dark grey). An asterisk denotes the following statistically significant changes in value for mRNA: a–c–e, P = 0.0005; L5 mutant a–c–e, P = 0.0009. Two asterisks denote the following statistically significant changes in value for protein: a–c–e, P = 0.0012; L5 mutant a–c–e, P = 0.0008.

Fig. 5.

Point mutations converting an L5 Spi-B-binding site to an L3 Spi-B-binding site in the Mad-1 promoter/enhancer reduces early viral activity in PDAs. The NCCR for wild-type Mad-1 (b) and L5–L3 point mutants for the first (a) or second (c) repeats within the Mad-1 promoter/enhancer are represented in a diagram form. JCV T-antigen mRNA is represented as 2×10⁵ copies per 50 ng RNA along with sd (d, light grey). The number of T-antigen protein-positive cells was quantified on a per cell basis and is represented as a fold-change over the wild-type Mad-1 plasmid along with sd (d, dark grey). JC virus DNA is represented as 2×10⁶ copies per 50 ng total DNA along with sd (e). An asterisk denotes the following statistically significant changes in value of mRNA: L5–L3 mutant first repeat, P = 0.0006; L5–L3 mutant second repeat, P = 0.0009. Two asterisks denote the following statistically significant change in value of protein: L5–L3 mutant first repeat, P = 0.0076; L5–L3 mutant second repeat, P = 0.0059. An asterisk denotes a statistically significant change in JC virus DNA for the L5–L3 mutant first repeat, P = 0.01.

Fig. 6.

A point mutation converting the L3 Spi-B-binding site to an L5 Spi-B-binding site is sufficient to activate early viral gene expression from the archetype promoter in PDAs. The NCCR sequences for wild-type archetype (a) and the L3–L5 point mutant archetype (b) are represented in a diagram form. JCV T-antigen mRNA is represented as total copies per 50 ng RNA along with sd (d, light grey). Eight days post-nucleofection, indirect immunofluorescence was performed to visualize T protein-positive cells. In the merged images, DNA is shown in blue and T-antigen is shown in red (c). Mad-4, panel (i); archetype, panel (ii); and L3–L5 mutant archetype, panel (iii). Bar, 500 µm at ×20 magnification. The number of T-antigen protein-positive cells was quantified on a per cell basis and is represented as a fold-change over the archetype plasmid along with sd (d, dark grey). An asterisk denotes the following statistically significant changes in value of mRNA. L3–L5 mutant, P = 5.40×10⁻⁵. Two asterisks denote the following statistically significant change in value of protein, 2.08×10⁻⁶. JCV DNA is represented as total copies per 50 ng DNA along with sd.