superhelix with a radius of curvature of 43 Å. To achieve this, the major and minor grooves are compressed
on the inside of the curve and stretched on its outside. At the same time, the helix axis must change direction.
DNA curvature has also been examined in kinetoplast DNA from trypanosomatids. It provides a source
of openDNA mini-circles whose curvature is sequence-dependent rather than being enforced by covalent
closure of the circles. Such circles can be examined by electron microscopy and have 360° curvature for
about 200 bp. Such kinetoplast DNA has short adenine tracts spaced at 10 bp intervals by general
sequence. This fact led to solution studies on synthetic oligomers with repeated sets of four CA5–6T
sequences spaced by 2–3 bp. These behave as though they have a 20°–25° bend for each repeat, which led
to the simple idea that DNA bending is an inherent property of poly(dA) tracts (Figure 2.24a). In conflict
with this idea, poly(dA) tracts in the crystal structures of several oligonucleotides are seen to be straight.
What then is the real origin of DNA curvature?24,25
Richard Dickerson has examined helix bending in a range of B-form crystal structures of oligonu-
cleotides containing poly(dA) tracts and has concluded that poly(dA) tracts are straight and not bent and
that regions of AT base pairs exhibit a narrow minor groove, large propeller twist and a spine of hydra-
tion in the minor groove. He argues that DNA curvature results from the direct combination of two gen-
eral features of DNA structure.
- General sequence DNA writhes
- Poly(dA) tract are straight.
DNA and RNA Structure 39
Figure 2.24 Curved DNA illustrated by straight poly(dA) tracts (upper), consecutive tracts of writhing DNA of general
sequence (lower), and curved DNA (right) of alternating segments of linear poly(dA) and curved tracts
of general sequence