DNA and RNA Structure 27
2.2.3 A-DNA
Among the first synthetic oligonucleotides to be crystallised in the late 1970s were d(GGTATACC), an
iodinated-d(CCGG) and d(GGCCGGCC). They all proved to have A-type DNA structures, similar to the
classical A-DNA deduced from fibre analysis at low resolution. Several other oligomers, mostly octamers,
also form crystals of the A-structure, but NMR studies suggest that some of these may have the B-form in
solution. It is conceivable that crystal packing might especially favour A-DNA for octanucleotides.
The general anatomy of A-DNA follows the Watson–Crick model with anti-parallel, right-handed
double helices. The sugar rings are parallel to the helix axis and the phosphate backbone is on the outside
of a cylinder of about 24 Å diameter (Figure 2.16).
X-ray diffraction at atomic resolution shows that the bases are displaced 4.5 Å away from the helix axis
and this creates a hollow core down the axis around 3 Å in diameter. There are 11 bases in each turn of
28 Å, which gives a vertical rise of 2.5 6Å per base pair. To maintain the normal van der Waals separation
of 3.4 Å, the stacked bases are tilted sideways through 20°. The sugar backbone has skewed phosphate
ester bonds, and anti-periplanar conformations for the adjacent CO ester bonds. Finally, the furanose ring
has a C3-endopucker and the glycosylic bond is in the anticonformation (Table 2.3). As a result of these
features, the major groove of A-DNA is cavernously deep and the minor groove is extremely shallow, as
Figure 2.18 The minor groove hydration ‘ribbon’ in the dodecamer d(CGCGAATTCGCG). The inner and outer
water (1–9) spines define four fused hexagons that dissect the minor groove. Only 10 bp are shown
and terminal residues are numbered
(Adapted from V. Tereshkoet al.,J. Am. Chem. Soc., 1999, 121 , 3590–3595. © (1999), with permission
from the American Chemical Society)