In the idealized A-form of duplex DNA, the major and minor grooves are more equal in width (Figure
10.1a). The major groove of the A-form is comparatively narrower than the B-form, so that the bases of the A-
form are more deeply buried within the body of the duplex and are consequently less accessible from that side.
In contrast, the minor groove of the A-form is widened and flattened relative to that of the B-form, so that the
bases and the sugars are more exposed and thus available for contact. Moreover, the sugars in the A-form have
the C3-endoconformation (Section 2.2.2), in contrast to the C2-endoconformation of the B-form, and they
consequently have greater exposed surface for non-polar interaction with proteins in the minor groove.^2
The different groove characteristics of the A- and B-forms arise from their distinctive helical geom-
etries: the helical axis of the idealized B-form is straight and the structure repeats every ten base pairs
(Figure 10.1) while the local axis of the A-form spirals in space and the helix repeats every 11 base pairs.
In the B-form, the centres of the base pairs lie close to the central helical axis, while in the A-form they are
displaced from it. Thus, when viewed along the length of the helical axis, the A-form appears to have a
central channel, while the B-form is densely packed. The A-form is favoured under dehydrating conditions
and by G:C rich sequences, and the transition between the A- and B-forms involves a sliding movement of
the bases and adjustments in the backbone torsion angles.3,4
The A- and B-forms represent idealized conformations on the basis of fibre diffraction studies of DNA
that have provided cylindrically averaged structural parameters (Section 2.3). More detailed stereochemical
information has been obtained from high-resolution X-ray crystal structures of short segments of synthetic
DNA. These structures show, for instance, that both grooves are well hydrated and contain intricately organ-
ized networks of hydrogen-bonding water molecules. Such bound water molecules are mostly displaced when
a specific protein complex forms, but a small proportion may be retained in the interface to fill gaps because
of imperfections in matching the surfaces (Section 10.4). An important finding from X-ray crystal struc-
ture analyses is that there are significant structural variations in duplex DNA, such that local structures often
hardly seem to resemble either of the idealized A- or B-forms, but instead something intermediate between
these forms (Section 2.3). The structural variations are related to the underlying DNA sequence itself
because of the stacking preferences of the bases. Rules relating DNA sequence to conformation have been
worked out qualitatively from mechanical principles,^1 and extended and quantified on the basis of chem-
ical principles^5 (Section 2.3).
The fact that sequence can impart structural variations of the DNA suggests that a sequence could be
recognized indirectly through its effects on the shape and deformability of the DNA (Section 10.4). As an
example of the type of conformational adjustment seen in protein–DNA complexes, the major groove is
often observed to become narrower when it engages an -helix to form a complex. This creates a gentle
in-plane curvature of the DNA towards the bound protein. A more extreme example of in-plane bending
occurs in DNA packaged by proteins in the chromosomes of nucleated cells (Figure 10.1d). The DNA
segment follows the surface curvature of the packaging protein, and here the bending of the DNA is asso-
ciated with periodic changes in the geometry of the base steps. Finally, it should be noted that a DNA-bind-
ing protein seldom amounts to a rigid surface upon which the DNA moulds its shape: in the process of
forming a complex with DNA, portions of the protein surface may also adjust their conformation. Thus,
recognition often involves a mutually induced fitof both protein and DNA (Section 10.4). Similar effects
also occur in the formation of most RNA–protein complexes (Section 10.9).
In cells, RNA is often found in single-stranded form, and may fold into compact structures, such as an
RNA hairpin (Figure 10.3a), especially if it contains regions of self-complementarity. The segment of RNA
that separates the complementary sequences forms a loop in which the bases are exposed. Such a loop often
acts as a target for protein recognition, since the unpaired bases are more readily accessible for contacts
with amino acid residues than those in paired duplex regions: see for example the interaction of the protein
U1 with an RNA hairpin (Figure 10.3a). RNA bases may also become exposed by internal bulges, which are
formed wherever extra bases are accommodated on one strand within a duplex RNA region.
Duplex regions of RNA resemble the A-form of DNA, and consequently these regions have a deep
major groove in which the bases are not readily accessible. Heteroduplexes of RNA and DNA, such as
those formed transiently when genes are transcribed by RNA polymerase, also have a structure similar to
386 Chapter 10