Lymphocyte gene expression and JC virus noncoding control region sequences are linked with the risk of progressive multifocal leukoencephalopathy

Abstract

Progressive multifocal leukoencephalopathy (PML)-derived noncoding control region (NCCR) sequences permitted greater early viral gene expression than kidney-associated NCCR sequences. This was driven in part by binding of the transcription factor Spi-B to unique PML-associated Spi-B binding sites. Spi-B is upregulated in developing B cells in response to natalizumab therapy, a known risk factor for PML. Naturally occurring JCV sequence variation, together with drug treatment-induced cellular changes, may synergize to create an environment leading to an increased risk of PML.

FIG 1

Spi-B gene expression is upregulated in CD34⁺ hematopoietic precursors and CD19⁺ B cells in response to long-term treatment with natalizumab. CD3⁺, CD19⁺, and CD34⁺ cells were isolated by immunomagnetic separation from PBMC samples from patients treated with natalizumab or normal donors. The remaining cells from the separation are labeled the negative fraction. Total RNA was isolated from each fraction of cells, and the Spi-B gene expression level was measured by qRT-PCR. Spi-B mRNA levels were normalized between samples by using the endogenous control PUM1 as a reference for the input template. Expression was normalized to levels in untreated CD19⁺ cells. The Spi-B mRNA level is expressed as a relative value calculated by using a standard curve.

FIG 2

Schematic depiction of viral NCCR, cloning strategy, and early gene expression. (A) The NCCRs from PML patient tissues that are classified as type II-S no-repeat archetype like, type I-R single TATA box repeat Mad-4 like, and type II-R single TATA box repeats with insertions are represented. Conserved sequence blocks that contain deletions are red, and TATA boxes are blue. Sites that bind Spi-B protein in electrophoretic mobility shift assays are yellow, and sites that did not bind protein are white. The Spi-B binding sites grouped by location in reference to TATA boxes are listed. GGAA core binding sites targeted for mutation analysis are bold, and the L38 3′ G that creates an L3-like site and abrogates Spi-B binding when mutated is underlined. (B) An intermediate destination plasmid termed a swap vector (pMad-1_SW) was generated by creating restriction enzyme sites (swap sites) that overlap the start sites for the T antigen (AgeI) and agnoprotein (KpnI) present on either side of the Mad-1 NCCR in the pM1_TC plasmid. Plasmids containing the PML-derived NCCR with the same restriction sites overlapping the start sites were generated by DNA 2.0 (pJ241:PML_swap). The pMad-1_SW and pJ241:PML_swap plasmids were digested with KpnI and AgeI restriction enzymes. The enzymatic products were separated by gel electrophoresis, and the destination vector and PML-derived NCCR inserts were purified from the gel slices. The PML-derived NCCR inserts were ligated into the pMad-1_SW destination vector, and the restriction sites were restored to the wild-type sequence by SDM. This cloning scheme allowed the insertion of the PML-derived NCCR sequences in frame with the T antigen and agnoprotein start sites within the Mad-1 coding sequence. (C) Plasmids encoding Mad-1 or archetype viral genome or the PML-derived NCCR–Mad-1 reporter genome were introduced into PDAs via nucleofection. Total RNA was isolated at the indicated time points, and T antigen mRNA expression (number of copies per nanogram of total RNA) was measured by qRT-PCR.

FIG 3

Early viral gene expression from patient and Spi-B-mutated sequences. Cells were nucleofected with wild-type or Spi-B-mutated PML variant NCCRs in the pMad-1 coding region plasmid. Total RNA was harvested, and T antigen mRNA expression (number of copies per nanogram of total RNA) was measured by qRT-PCR at the indicated time points. (A) Mutation of the L38 Spi-B binding site in the rituximab patient NCCR reduces early viral gene expression in PDAs. The red line represents a GG-to-CC mutation in the core of the L38 Spi-B binding site. The green line represents a G-to-A mutation at the 3′ end of the L38 Spi-B binding site, which results in the creation of an L3-like site (archetype) incapable of binding Spi-B protein. (B) Mutation of the L28 Spi-B binding site in the HAART patient NCCR reduces early viral gene expression, but mutation of the L4 Spi-B binding site does not. The red line represents a GG-to-CC mutation in the core of the L4 Spi-B binding site. The green line represents a GG-to-CC mutation in the core of the L28 Spi-B binding site. (C) Mutation of the L4 Spi-B binding site in the natalizumab patient NCCR reduces early viral gene expression in PDAs. The red line represents a GG-to-CC mutation in the core of the L4 Spi-B binding site. (D) Mutation of the L41 Spi-B binding site in the mycophenolate mofetil patient NCCR slightly increases early viral gene expression in PDAs. The red line represents a GG-to-CC mutation in the core of the L41 Spi-B binding site, which creates an NFI binding site.