Pathogenesis of progressive multifocal leukoencephalopathy--revisited

Abstract

Progressive multifocal leukoencephalopathy (PML) is a demyelinating disease of the central nervous system that is rare even though the proven etiological agent of PML, the polyomavirus JC (JC virus), is ubiquitous within the human population. The common feature of PML cases appears to be underlying immunosuppression, and PML has gained clinical visibility because of its association with human immunodeficiency virus and AIDS and its occurrence as a side effect of certain immunomodulatory drugs. A hypothesis has gained general acceptance that JC virus causes a primary infection in childhood and enters a latent state, after which immunosuppression allows viral reactivation leading to PML. Nonetheless, many important aspects of PML pathogenesis remain unclear, including the molecular bases of latency and reactivation, the site(s) of latency, the relationship of archetype and prototype virus and the mode of virus transmission within the body and between individuals. In this review, we will revisit these areas and examine what the available evidence suggests.

Figure 1.

JC virus (JCV) antibody seropositivity in the human population. JCV and BK virus (BKV) antibody titers from 4 well-rounded studies conducted during the last 3 decades are presented as bar graphs of percentage seropositivity against age. A, A 1983 study by Walker and Padgett [13] using a hemagglutination inhibition assay. B, A 2003 study by Knowles et al [14] using a hemagglutination inhibition assay. C, A 2009 study by Kean et al [15] using a recombinant VP1-based enzyme-linked immunoassay assay. D, A 2009 study by Egli et al [16] using a viruslike particle–based enzyme-linked immunoassay assay.

Figure 2.

Comparison of prototypical progressive multifocal leukoencephalopathy (PML)–type Mad-1 strain of JC virus (JCV) with archetypal CY strain. The DNA sequences of the noncoding control region (NCCR) of Mad-1 and CY are shown aligned together, with Mad-1 on top in boldface and CY below in regular face. Numbering is relative to the Mad-1 strain (GenBank NC_001699, formerly J02226) [26]. The capitalized CAT (nucleotides 5011–5013) corresponds to the reverse complement of ATG start codon for T antigen and demarks the end of the early region and the start of the NCCR. The region at the 5’ end of the NCCR (nucleotides 5014–5036) is identical for Mad-1 and CY and is very highly conserved in all strains of JCV. This area contains nucleotide 1 of the 5130–base pair (bp) JCV Mad-1 circular genome. The underlined sequence (nucleotides 5111–5129) contains the origin of DNA replication, including the GGA/GGC-rich element (nucleotides 5118–5132) and the A/T-rich element (nucleotides 15–29). The 23-bp element that is deleted in Mad-1 (compared with CY, between nucleotides 36 and 37) is indicated by a single asterisk and highlighted in green. The 66-bp element deleted in Mad-1 (compared with CY, between nucleotides 84 and 85) is indicated by a double asterisk and highlighted in blue. The sequence of the 23- and 66-bp deletions is shown at the bottom of the figure. The region containing the deletions (nucleotides 12–109) is duplicated in Mad-1; this is known as the 98-bp tandem repeat and is highlighted in yellow (nucleotides 12–207). The junction between the two 98-bp repeats is shown by brackets, highlighted in red. The 3’ end of the NCCR (nucleotides 208–276) is identical for Mad-1 and CY. The capitalized ATG (nucleotides 277–279) corresponds to the ATG start codon of agnoprotein and demarks the end of the NCCR and the start of the late region.