Functional characterization of the Mediator subunit MED25

In this study a structure–function analysis has been employed to
analyze transcriptional regulation through the Mediator subunit
MED25. A relationship could be established between predicted
structural domains and functional characteristics of this protein.
Most critically the region responsible for interaction of MED25
with the Mediator was identified. Immunoprecipitation experiments
demonstrated that the so–called VWA domain (von–Willebrand A
domain, amino acids 1–290) is both sufficient and required for this
contact. Site–directed mutagenesis indicates that this binding
reaction involves the non–conserved loop SR2, which is protruding
from this domain. Based on the results of this analysis a model was
proposed, in which the primary contact is established by ionic
forces and is further stabilized by hydrophobic interactions. The
previously identified ACID domain was reported to bind to VP16.
Targeted mutagenesis of four different motifs in this region
impaired not only transcriptional activation through MED25 but also
led to reduced binding to VP16. In particluar a lysine–rich motif
is also present in two domains of PTOV1, a close homolog of MED25.
Noteworthy, K518 is not conserved in the PTOV1_B domain, which in
contrast to PTOV1_A and the ACID domain of MED25 does not bind to
VP16. This led to the hypothesis that K518 is critically involved
in the binding of VP16 to MED25. Furthermore it could be
demonstrated that MED25 contains an intrinsic transcriptional
activation capacity, which is localized in the region 290–715. This
indicates additional recruitment of other factors to promoters
through this region. Together with the Mediator binding VWA–domain
and the VP16–interaction domain this region might facilitate
transcriptional activation. A genome–wide screen showed
downregulation of c–Jun and FosB following overexpression of MED25.
Interestingly, expression of GSK3β, a downstream target of which is
cyclin D1, seems to be stimulated by MED25. Together with the
finding that overexpression of MED25 leads to activation of a p21
reporter, this raises the possibility that MED25 is involved in
cell cycle control. An overlap has been discovered by comparison of
MED25 target genes and genes identified previously as target for
the viral activator EBNA2. The close homology between the
activation domains of EBNA2 and VP16 implies a common mechanism of
transcriptional activation by these two viral proteins through
MED25. The involvement of MED25 in gene activation by viral
activators might indicate a role for this Mediator subunit in viral
transcription.