Biomedical Engineering Reference
In-Depth Information
sequences necessary for subunit dimerization, nuclear localization, DNA binding,
and inhibitor binding. Subunit dimerization is mediated entirely by the dimerization
domain, a conserved 100 amino acid folded domain located within the carboxyl-
terminal third of the RHD. Both the dimerization domain and an amino-terminal
domain participate in target DNA binding. In the following section we describe
NF-
κ
B subunit dimerization. DNA binding is discussed in Section 2.4.
2.2.1
NF-
κ
B D
IMER
F
ORMATION
B family can associate with each other
to form various combinations of homo- and heterodimers. Such combinatorial asso-
ciations can broaden the spectrum of activity and diversify biological function.
Although theoretically 15 possible dimeric combinations can be generated, not all
of them have been demonstrated to be physiologically important. The p50/p65
heterodimer is by far the most ubiquitous, being seen in almost all cell types. The
p50/p50, RelB/p50, and RelB/p52 dimers are also known to be prevalent. However,
p65/p65, p50/p50, c-Rel/c-Rel, p65/c-Rel, and p50/c-Rel dimers are only found in
a limited subset of cell types. On the other hand there is no convincing evidence for
the presence of the p52/p52 or p52/p50 dimers. This is somewhat surprising if one
considers the close sequence similarity between p50 and p52 and the relative abun-
dance of the p50/p50 homodimer. Recent studies also point toward the existence of
the RelB/p65 and RelB/c-Rel heterodimers, but it is still not clear if the Rel-subunits
in these dimers interact through the dimer interface. The RelB/RelB homodimer is
yet to be detected in cells.
Cell-specific expression of the different Rel-subunits may help explain the
presence of the different dimeric forms. On the other hand, differential affinities
may also contribute to the varied distribution of the dimers. Although no direct
measurements of binding affinities between the different dimerization domains
have been reported, high-resolution crystal structures have provided some clues
about how the differential dimers might be generated. Structures of the dimer-
ization domains, and of the NF-
As mentioned above, members of the NF-
κ
B dimers bound to DNA, show that the dimer-
ization interface remains fairly conserved in both the free and DNA bound forms.
All the structures solved to date, with the exception of the RelB dimerization
domain, show similar overall features wherein each monomer assumes an inde-
pendent Ig-fold that packs against each other to form a dimer interface. The
variable dimerization specificities may then be explained by amino acid substi-
tutions within the dimerization domain. For example, Y267 in murine p50 is a
key residue involved in extensive hydrogen bonding and van der Waals contacts,
both of which aid in homodimer formation. This residue is replaced by pheny-
lalanine in p65 and c-Rel, which abolishes hydrogen bonding, thereby weakening
subunit association in p65 and c-Rel compared to p50. It is likely that the lack
of the Y267 hydrogen bonding network in p65 may explain why the p65
homodimer is weaker than the p50 homodimer. In the p50/p65 heterodimer,
however, D254 of p50 forms strong hydrogen bonds with N200 of p65, thereby
strengthening the p50-p65 dimer interface and making it a more stable het-
erodimer compared to the individual homodimers.
κ
