A Structural Model for the PS3 Complex with DNA Response Elements
269
1998; Grossman et ai., 1998). A clue as to the importance of specific DNA binding
to p53 function may therefore lie in the fact that this interface is formed only
upon the association of multiple p53DBD moieties with DNA response elements.
Furthermore, it is exposed outside the DNA loop and is therefore easily accessible
in the complex. Similarly, the interactions between the p53DBD subunits involving the N-terminal fragments adjacent to the core domain (e.g., the proline rich
fragments) might be operative in creating a "signal", indicating that p53 docking
has occurred. On the other hand, it may be important in certain functions that
the interface be hidden in order to prevent certain proteins from binding. For
example, a protein involved in the p53 degradation pathway binding to the HlHI interface surface or to the proline-rich fragment of the N -terminus would be
effectively prevented from binding, thus increasing the lifetime of p53. This
might represent an alternative pathway for transcriptional regulation.
The binding motif for p53 appears to be unique among transcription factors
so far investigated. The closest similarity to currently reported structures is with
the lac repressor complex (Lewis et al., 1996). The implications for allosteric control, which has been suggested for p53 (Halazonetis et ai., 1993; Vojtesek et al.,
1995; Waterman et ai., 1995), may also share similarities with this system (Horton
et al., 1997). The possibility for allosteric control of p53 binding is implicit in the
location of the N- and C-termini of the p53DBD in Fig. 19.3a (and see also Fig.
19.8 of Nagaich et ai., 1997b) where the N-termini lie on the outside of the DNA
loop and the C-termini on the inside. The phasing experimental results show
clear differences in both DNA bending and twisting between the p53 core domain
and wild type p53 complexes, suggesting that DNA binding is moderated by
other regions of the p53 molecule. These termini can, in principle, interact either
with the DNA binding domain (DBD) or with the DNA, either within or flanking
the specific binding site.
Based on the structural model, it seems likely that the N-terminus interacts
with the p53DBD, thereby affecting the p53DBD-p53DBD interactions, whereas
the C-terminus is likely to interact directly with the DNA. Each N-terminus may
interact either with its own p53DBD or with an adjacent one, a reasonable postulate since the N-terminal region, although variable among species, is proline-rich
in the same region. In the model, this region is positioned in the vicinity of the
adjacent p53DBD, and such interactions could lead to an increased bend in the
bound DNA. Such a postulate might also help clarify the role of spacer DNA
between half-sites (Tokino et ai., 1994), since the N-terminus potentially can
form a bridge between p53 dimers separated by spacers up to 21 bp in length
(Cook et ai., 1995; Miner and Kulesz-Martin, 1997). The latter could therefore be
an aspect of indirect recognition for the binding specificity of p53. Finally, since
the N-termini are involved in downstream signal transduction pathways, structural information on them is important. In the present model, they are relatively
exposed and accessible to other downstream proteins after p53 is bound to DNA.
Functional roles of the C-termini are equally important and include moderating p53 DNA binding properties. This view is supported by a variety of studies
showing that the basic, 30-residue C-terminal region strongly influences the DNA
binding properties of p53 (Foord et ai., 1991; Anderson et ai., 1997), and that this
binding is further moderated by the charge state as determined by its acetylation
269
1998; Grossman et ai., 1998). A clue as to the importance of specific DNA binding
to p53 function may therefore lie in the fact that this interface is formed only
upon the association of multiple p53DBD moieties with DNA response elements.
Furthermore, it is exposed outside the DNA loop and is therefore easily accessible
in the complex. Similarly, the interactions between the p53DBD subunits involving the N-terminal fragments adjacent to the core domain (e.g., the proline rich
fragments) might be operative in creating a "signal", indicating that p53 docking
has occurred. On the other hand, it may be important in certain functions that
the interface be hidden in order to prevent certain proteins from binding. For
example, a protein involved in the p53 degradation pathway binding to the HlHI interface surface or to the proline-rich fragment of the N -terminus would be
effectively prevented from binding, thus increasing the lifetime of p53. This
might represent an alternative pathway for transcriptional regulation.
The binding motif for p53 appears to be unique among transcription factors
so far investigated. The closest similarity to currently reported structures is with
the lac repressor complex (Lewis et al., 1996). The implications for allosteric control, which has been suggested for p53 (Halazonetis et ai., 1993; Vojtesek et al.,
1995; Waterman et ai., 1995), may also share similarities with this system (Horton
et al., 1997). The possibility for allosteric control of p53 binding is implicit in the
location of the N- and C-termini of the p53DBD in Fig. 19.3a (and see also Fig.
19.8 of Nagaich et ai., 1997b) where the N-termini lie on the outside of the DNA
loop and the C-termini on the inside. The phasing experimental results show
clear differences in both DNA bending and twisting between the p53 core domain
and wild type p53 complexes, suggesting that DNA binding is moderated by
other regions of the p53 molecule. These termini can, in principle, interact either
with the DNA binding domain (DBD) or with the DNA, either within or flanking
the specific binding site.
Based on the structural model, it seems likely that the N-terminus interacts
with the p53DBD, thereby affecting the p53DBD-p53DBD interactions, whereas
the C-terminus is likely to interact directly with the DNA. Each N-terminus may
interact either with its own p53DBD or with an adjacent one, a reasonable postulate since the N-terminal region, although variable among species, is proline-rich
in the same region. In the model, this region is positioned in the vicinity of the
adjacent p53DBD, and such interactions could lead to an increased bend in the
bound DNA. Such a postulate might also help clarify the role of spacer DNA
between half-sites (Tokino et ai., 1994), since the N-terminus potentially can
form a bridge between p53 dimers separated by spacers up to 21 bp in length
(Cook et ai., 1995; Miner and Kulesz-Martin, 1997). The latter could therefore be
an aspect of indirect recognition for the binding specificity of p53. Finally, since
the N-termini are involved in downstream signal transduction pathways, structural information on them is important. In the present model, they are relatively
exposed and accessible to other downstream proteins after p53 is bound to DNA.
Functional roles of the C-termini are equally important and include moderating p53 DNA binding properties. This view is supported by a variety of studies
showing that the basic, 30-residue C-terminal region strongly influences the DNA
binding properties of p53 (Foord et ai., 1991; Anderson et ai., 1997), and that this
binding is further moderated by the charge state as determined by its acetylation