RecA. Perhaps, like the T7 ring helicase, the DNA conjugation pump might undergo asymmetric changes
within the ring, so that two non-equivalent states of the subunits cause the displacement of the DNA
duplex.
10.8.3 DNA Topoisomerases
Genomic DNA is packed densely and continuously undergoes superhelical strain during transcription,
replication and (occasionally) recombination. For example, in E. coli, replication involves a rate of unwind-
ing of 160 turns of duplex DNA per second: it can be envisaged that unwinding must be compensated at
comparable rates to avoid tangling. Organisms in all taxonomic domains have evolved enzymes that manipu-
late DNA topology, either by relaxing superhelical tension generated from transcription, replication and
recombination (a spontaneous process) or by introducing supercoiling (which requires an input of energy)
(Section 2.3.5). There are two principal classes of topoisomerases for these purposes: type I and type II.
The former break one strand of duplex DNA transiently and allow rotation about the unbroken strand before
the broken strand is re-ligated, so as to alter the linking number of the two strands. Type II enzymes break both
strands and pass one region of DNA through the gap, resulting in stepwise changes in linking number of two
units. The structure of topoisomerase II reveals a toroidal shape, and it is possible to imagine the DNA pass-
ing through the ring during strand exchange. For both type I and type II topoisomerases, strand cleavage
involves nucleophilic attack of the phosphodiester backbone by a tyrosine hydroxyl group, resulting in a cova-
lent bond between the enzyme and the 5end of the broken strand. Rotation of the 3end is followed by
re-ligation.
A unique sub-group of the type II topoisomerases are the gyrases of prokaryotes, which catalyse the
ATP-dependent negative supercoiling of double-stranded closed circular DNA, and they thus maintain the
bacterial genome under negative superhelical tension. In eukaryotes, there is little superhelical tension,
which may explain the requirement for DNA distortion at the promoter by TBP and the requirement for
helicase activity in one of the transcription factors (Section 10.3.6).
10.9 RNA–Protein Interactions and RNA-Mediated Assemblies
Just as in the recognition of DNA, many different protein structural motifs are used to recognize RNA.
RNA–protein complexes that have been characterized so far have tended to be more complicated than
protein–DNA complexes, since RNA structures are themselves more complex than double-stranded
DNA, and include duplex regions, pseudo-knots and single-stranded regions, such as bulges and hairpin
loops (Section 7.1).^55 Loops can be sub-divided into nearly a dozen different known classes of internal
loops that occur between adjoining duplex regions and external loops in hairpins.
RNA–protein complexes are divided roughly into two main classes, depending on the mode of RNA
binding: the groove-binding modeand -sheet binding, which uses the sheet to bind unpaired RNA
bases.^56 For groove binding, the -helix is a predominant secondary structural element for recognition. In
many RNA–protein complexes, backbone and base contacts are used with roughly equal frequency, in con-
trast to protein–DNA interactions where backbone contacts are dominant. Because of the intricate tertiary
structure of the RNA, the packing density of the RNA–protein interface is often not as great as in complexes
of protein with single- and double-stranded DNA. However, the recognition of RNA generally involves
more van der Waals contacts than hydrogen bonds.^57
Sequence-specific complexes of RNA and protein achieve their specificity through tight packing of com-
paratively non-polar surfaces. Thus, aromatic residues play key roles, probably by stacking on the unpaired
bases. The 2-hydroxyl group often protrudes from these RNA targets; it is more solvent-exposed than the
other sugar oxygen atoms,^57 and is therefore a target for hydrogen-bonding interactions. Induced fit occurs
frequently in protein–RNA recognition. For instance, assembly of almost all ribosomal protein–RNA
complexes involve induced fit for either the protein, or the RNA or for both.^58
416 Chapter 10