Biology Reference
In-Depth Information
(Maul et al. 1993; Korioth et al. 1996). These findings indicate that PML-NBs may
actually be detrimental, not helpful, to viral infections. Thus a controversy exists as
to whether PML-NBs are pro-virus or anti-virus. Several recent studies from
multiple laboratories summarized below paint an incomplete yet quickly resolving
portrait of the nuclear events that precede HCMV IE gene expression and argue that
PML-NBs are not preferred sites for viral transcription but represent repressive
subnuclear domains that are sequentially dismantled during HCMV infection.
Most fluorescent images of PML-NBs give the impression that they are static
structures. However, the major constituent proteins of PML-NBs, PML, Sp100 and
especially Daxx, actually have a dynamic association with these structures, with
high rates of association and disassociation (Wiesmeijer et al. 2002; Everett and
Murray 2005). Using synchronized and polarized infections with HSV-1, an elegant
series of experiments (Everett and Murray 2005) showed that new PML-NBs are
formed de novo around infecting viral genomes (the expected result if PML-NB
proteins represent a cellular antiviral defense), and argue against viral genomes
migrating through the nucleus to sites of preformed PML-NBs (the expected result
if localization to these sites provided an advantage to the virus). The rapidity with
which Daxx enters and leaves PML-NBs makes it an obvious candidate for the ini-
tial cellular sensor of infecting viral genomes, and this appears to be the case for
HCMV. In cells in which the level of PML has been reduced by RNA interference,
there are no PML-NBs, and the Daxx and Sp100 proteins are diffusely distributed
throughout the nucleus. However, Daxx and Sp100 co-localize to form punctate
spots reminiscent of PML-NBs upon HCMV infection (Tavalai et al. 2006). The
newly synthesized viral IE2 protein is also found in these aggregates, implying that
transcriptionally active viral DNA is located there as well. This significant study
showed that in the absence of PML-NBs, Daxx and Sp100 can sense and apparently
migrate to infecting HCMV genomes (Tavalai et al. 2006). Any effects of Sp100 on
HCMV infection have yet to be established, but it is becoming increasingly clear
that Daxx inhibits HCMV infection, and is the very first PML-NB component
whose antiviral activities must be neutralized in order for HCMV to express its
immediate early genes.
At the low multiplicities presumed to mimic an in vivo infection, the viral pp71
protein, which is delivered from the tegument to the nucleus of infected cells
(Hensel et al. 1996), is required for immediate early gene expression and subse-
quent viral replication (Bresnahan and Shenk 2000). pp71 binds to Daxx through
two Daxx-interaction-domains, termed DIDs (Hoffman et al. 2002), and through
this interaction partially and transiently localizes to PML-NBs (Hoffman et al.
2002; Marshall et al. 2002; Ishov et al. 2002). Recombinant HCMVs expressing
DID-mutant pp71 proteins (and not wild type) have the same phenotype as the
pp71-null mutant, indicating that pp71 binding to Daxx is required for efficient
viral IE gene expression (Cantrell and Bresnahan 2005). This series of experiments
was important because it defined the role of a single function of the multifunctional
pp71 protein during viral infection by examining the phenotype of a recombinant
virus expressing a mutant pp71 protein, and because it identified Daxx as a critical
determinant of HCMV IE gene expression.
