being its 3-neighbour, both in the opposing strand. This network of hydrogen bonds gives added rigidity
to the duplex and may explain why long runs of adenines are not found in the more sharply curved tracts of
chromosomes (Section 2.6.2), yet are found at the end of nucleosomal DNA with decreased supercoiling.
Why should the bases twist in this way?^16 The advantage of propeller twist is that it gives improved
face-to-face contact between adjacent bases in the same strand and this leads to increased stacking stabil-
ity in the double-helix. However, there is a penalty! The larger purine bases occupy the centre of the helix
so that in alternating purine–pyrimidine sequences they overlap with neighbouring purines in the opposite
strand. Consequently, propeller twist causes a clash between such pairs in adjacent purines in opposite
strands. For pyrimidine-(3→ 5 )–purine steps, these purine–purine clashes take place in the minor groove
where they involve guanine-N-3 and-N-2 and adenine-N-3 atoms. For purine-(3→ 5 )–pyrimidine steps,
they take place in the major groove between guanine-O-6 and adenine-N-6 atoms (Figure 2.21). There are
no such clashes for purine–purine and pyrimidine–pyrimidine sequences.
One of the consequences of these effects is that bends may occur at junctions between polyA tracts and
mixed-sequence DNA as a result of propeller twist, base pair inclination and base-stacking differences on
two sides of the junction (see below).
2.3.1.1 Electrostatic Interactions between Bases. There are two principal types of base–base inter-
action that drive the local variations in helix parameters described above and in Figure 2.19a–c. First, there are
repulsive steric interactions between proximate bases and sugars. They are associated with steric interactions
between thymine methyl groups, the guanine amino group and the configuration of the step pyrimidine–purine
(described as YR), purine–pyrimidine (described as RY) and RR/YY. Second, there are – stacking inter-
actions that are determined by the distribution of -electron density above and below the planar bases.
Chris Hunter has identified four principal contributions to the energy of – interactions between DNA
base–pairs:^17
(1) van der Waals interactions (designated vdWand vary as r–6)
(2) Electrostatic interactions between partial atomic charges (designated atom–atomand vary as r–1)
(3) Electrostatic interactions between the charge distributions associated with the -electron density
above and below the plane of the bases (designated
–
and vary approximately as r–5)
(4) Electrostatic interactions between the charge distributions associated with the -electron density
and the partial atomic charges (designated as atom–
, this is the cross-term of (2) and (3) and
varies as r–4)He has used these components to calculate the – interaction energies between pairs of stacked bases
and applied the results to interpret the source of slide, roll and helical twist, of propeller twist, and of a range
of other conformational preferences that are sequence-dependent. In addition, his calculations correlate
34 Chapter 2
Figure 2.21 Diagrams illustrating (a) clockwise propeller twist for a C-(3→ 5 )-G clash between guanines in the
minorgroove and (b) clockwise propeller twist for a G-(3→ 5 )-C sequence showing purines clashing
in the majorgroove