264
R. E. HARRINGTON et aI.
The structural model for the complex as formulated (Nagaich et ai., 1997b,
1999; Durell et ai., 1998) is fully consistent with solution experiments as noted.
This is summarized in Table 1 where the various experimental determinations
are compared with predictions of the molecular model. All experimental methods, sensitive to DNA bending, show that the response element DNA is bent in
the complex, although various methods provide somewhat different specific
bending angles. We believe that the bending angle of32-36°, based upon phasing
studies, (Nagaich et ai., 1999) is the most accurate value presently available for
the p53DBD-DNA tetrameric complex. The bending directionality for the DNA,
originally proposed and based upon chemical probes results (Nagaich et ai.,
1997b), is confirmed by the phasing studies. Finally, the importance of the CATG
tetrads as bending loci in the complex is supported by studies of bending as a
function of response element sequence (Nagaich et ai., 1997a) as well as by recent
cyclization studies (P. Balagurumoorthy, manuscript submitted).
2.4
DNA Twisting and Bending Directionality in the pS3-DNA Complex
Bending and twisting in the response element DNA upon complexation with both
p53DBD and wild type p53 was examined using A-tract-based phase sensitive
detection methods. A basic "three-segment" phasing construct (Zinkel and Crothers, 1987) was used to determine both bending directionality and the bend
angle. Overtwisting in the DNA was identified and quantitated using a "foursegment" construct (Niederweis and Hillen, 1993). Together, the two constructs
allow for a relatively precise determination of these important quantities
(Nagaich et ai., 1999). The predictions of the original model are confirmed by
these experiments: the bending is into the major groove in both the p53DBD and
wild type complexes, but both bending and twisting in the DNA is appreciably
greater in the wild type complex. A substantial increase in twist upon protein
binding accompanied by bending into the major groove has been reported previously only for the nucleosome (Luger et ai., 1997) and does not appear to be a
common feature of transcription complexes in which DNA bending is generally
accompanied by undertwisting. This has important implications for possible p53
interaction with chromatin structures since it suggests that the p53 tetramer
might bind to nucleosomal DNA sites without completely unraveling the nucleosome structure.
The fact that response element DNA bending in the nucleoprotein complex is
correlated with overtwisting in both types of complex is of special importance,
since molecular modeling suggested that DNA overtwisting occurs in a very specific manner with complex formation, and that local supercoiling in genomic
DNA (Liu and Wang, 1987) might therefore playa role in the regulation of specific sequence recognition (Durell et ai., 1998). The coupling of bending and
overtwisting may also have important implications in the interaction of p53 with
chromatin. That these effects are significantly larger in the wild type p53 complex
supports the notion that DNA binding in the wild type complex may be affected
by regions of the N- and C-termini. This appears consistent with the proposal,
based on other types of evidence, that p53 binding may be regulated by more
R. E. HARRINGTON et aI.
The structural model for the complex as formulated (Nagaich et ai., 1997b,
1999; Durell et ai., 1998) is fully consistent with solution experiments as noted.
This is summarized in Table 1 where the various experimental determinations
are compared with predictions of the molecular model. All experimental methods, sensitive to DNA bending, show that the response element DNA is bent in
the complex, although various methods provide somewhat different specific
bending angles. We believe that the bending angle of32-36°, based upon phasing
studies, (Nagaich et ai., 1999) is the most accurate value presently available for
the p53DBD-DNA tetrameric complex. The bending directionality for the DNA,
originally proposed and based upon chemical probes results (Nagaich et ai.,
1997b), is confirmed by the phasing studies. Finally, the importance of the CATG
tetrads as bending loci in the complex is supported by studies of bending as a
function of response element sequence (Nagaich et ai., 1997a) as well as by recent
cyclization studies (P. Balagurumoorthy, manuscript submitted).
2.4
DNA Twisting and Bending Directionality in the pS3-DNA Complex
Bending and twisting in the response element DNA upon complexation with both
p53DBD and wild type p53 was examined using A-tract-based phase sensitive
detection methods. A basic "three-segment" phasing construct (Zinkel and Crothers, 1987) was used to determine both bending directionality and the bend
angle. Overtwisting in the DNA was identified and quantitated using a "foursegment" construct (Niederweis and Hillen, 1993). Together, the two constructs
allow for a relatively precise determination of these important quantities
(Nagaich et ai., 1999). The predictions of the original model are confirmed by
these experiments: the bending is into the major groove in both the p53DBD and
wild type complexes, but both bending and twisting in the DNA is appreciably
greater in the wild type complex. A substantial increase in twist upon protein
binding accompanied by bending into the major groove has been reported previously only for the nucleosome (Luger et ai., 1997) and does not appear to be a
common feature of transcription complexes in which DNA bending is generally
accompanied by undertwisting. This has important implications for possible p53
interaction with chromatin structures since it suggests that the p53 tetramer
might bind to nucleosomal DNA sites without completely unraveling the nucleosome structure.
The fact that response element DNA bending in the nucleoprotein complex is
correlated with overtwisting in both types of complex is of special importance,
since molecular modeling suggested that DNA overtwisting occurs in a very specific manner with complex formation, and that local supercoiling in genomic
DNA (Liu and Wang, 1987) might therefore playa role in the regulation of specific sequence recognition (Durell et ai., 1998). The coupling of bending and
overtwisting may also have important implications in the interaction of p53 with
chromatin. That these effects are significantly larger in the wild type p53 complex
supports the notion that DNA binding in the wild type complex may be affected
by regions of the N- and C-termini. This appears consistent with the proposal,
based on other types of evidence, that p53 binding may be regulated by more