|
(1) Structural basis for helix bending in B-DNA
|
|
Bending is nearly always the result of roll between successive base pairs, seldom tilt.
|
|
Positive roll, compressing the wide major groove, is more common than negative roll, in which the narrower minor groove is compressed.
|
|
Observed bends in B-DNA are of three main types: (a) localized kinks (large positive roll at one or two discrete base steps), (b) three-dimensional writhe (positive roll at a series of successive steps), or (c) smooth curvature (alternation of positive and negative roll every half turn, with side-to-side zigzagging at intermediate positions). (a) and (b) are easier to accomplish than (c), and hence are more common.
|
|
Local writhe in a DNA helix produces macroscopic curvature only when the extent of writhe does not match the natural rotational periodicity of the helix. Endless writhe results in a straight helix, and indeed A-DNA can be regarded as a continuously writhed variant of the B form. Conversely, the bending effect of writhe can be amplified if it is repeated with the periodicity of the helix itself – that is, repeated alternation of writhed and unwrithed segments every ten base pairs, as with A-tract B-DNA.
|
|
|
|
(2) Pyrimidine-purine (Y-R) steps: C-A = T-G, T-A and C-G
|
|
Little ring–ring stacking overlap.
|
|
Polar N or O stacked over polarizable aromatic rings.
|
|
Y-R steps are natural fracture points for the helix. They can show (but are not required in every case to show) large twist and slide deformations, and bending mainly via positive roll, compressing the major groove.
|
|
|
|
(3) Purine-purine (R-R) steps: A-A = T-T, A-G = C-T, G-A = T-C and G-G = C-C
|
|
Extensive ring–ring overlap.
|
|
Base pairs tend to pivot about stacked purines as a hinge, with greater ring–ring separation at pyrimidine ends.
|
|
Tight stacking, with only minor roll, slide and twist deformations.
|
|
|
|
(4) Purine-pyrimidine (R-Y) steps: A-C = G-T, A-T and C-G
|
|
Behaviour in general like R-R steps, with extensive ring–ring overlap and tight stacking, with again only minor roll, slide and twist deformations.
|
|
|
|
(5) A-A and A-T steps, as contrasted with T-A
|
|
Especially resistant to roll bending, probably because of sawhorse interlocking of highly propellered base pairs, supplemented by inter-base-pair hydrogen bonds within grooves. In contrast, T-A is particularly weak and subject to roll bending.
|
|
A-tracts, defined as four or more consecutive AT base pairs without the disruptive T-A step, are especially straight and resistant to bending. Natural selection has apparently chosen short A-tracts for regions of protein–DNA contacts where bending is not wanted.
|
![[link]](/graphics/greenarr.gif)