As the DNA winds around the nucleosome core, the major and minor grooves are compressed on the
inside with complementary widening of the grooves on the outside of the curved duplex. Runs of AT base
pairs, which have an intrinsically narrow minor groove should be most favourably placed on the inside of
the curved segment while runs of GC base pairs should be more favourably aligned with minor grooves
facing outwards, where they are more accessible to enzyme cleavage. In practice, Drew and Travers meas-
ured the periodicities of AT and GC base pairs by cleavage with DNase I and found them to be exactly
out of phase and having a periodicity of 10.17 0. 5 bp.74,75This result was later confirmed by hydroxyl
radical cleavage, which avoids the steric constraints of DNase I.
2.6.2 Chromatin Structure
Chromatinis too large and heterogeneous to yield its secrets to X-ray analysis, so electron microscopy is the
chosen experimental probe (Section 11.5.1). At intermediate salt concentration (1 mM NaCl) the nucle-
osomes are revealed as ‘beads on a string’. Spherical nucleosomes can be seen with a diameter of 7–10 nm
joined by variable-length filaments, often about 14 nm long. If the salt concentration is increased to 0.1 M
NaCl, the spacing filaments get shorter and a zig-zag arrangement of nucleosomes is seen in a fibre 10–11nm
wide (Section 10.6, Figure 10.15). At even higher salt concentration and in the presence of magnesium, these
condense into a 30 nm diameter fibre, called a solenoid, which is thought to be either a right-handed or a
left-handed helix made up of close-packed nucleosomes with a packing ratio of around 40.
For the further stages in DNA condensation, one of the models proposed suggests that loops of these
30 nm fibres, each containing about 50 solenoid turns and possibly wound in a supercoil, are attached to a
central protein core from which they radiate outwards.^76 Organisation of these loops around a cylindrical
scaffold could give rise to the observed mini-band structure of chromosomes, which is some 0.84 m in
diameter and 30 nm in thickness. A continuous helix of loops would then constitute the chromosome.
These ideas are illustrated in a possible scheme (Figure 2.45).
It is clear from all of the relevant biological experiments that the single DNA duplex has to be continu-
ously accessible despite all this condensed structure in order for replication to take place. Some of the
most exciting electron micrographs of DNA have been obtained from samples where the histones have
been digested away leaving only the DNA as a tangled network of inter-wound superhelices radiating from
a central nuclear region where the scaffold proteins remain intact (Figure 2.46).
Even then, the most condensed packing of nucleic acid is found in the sperm cell. Here a series of
arginine-rich proteins called protamines bind to DNA, probably with their -helices in the major groove of
the DNA where they neutralise the phosphate charge, and so enable very tight packing of DNA duplexes.
Bacterial DNA is also condensed into a highly organised state (Section 10.6.2). In E. colithe genome
has 4400 kbp in a closed circle, which is negatively supercoiled. It is condensed around histone-like pro-
teins, HU and HI, to form a nucleoidand achieve a compaction of 1000-fold, which is followed by further
condensation into supercoil domains. Unlike chromosomal DNA in eukaryotes, there is some additional
negative supercoiling in prokaryotes that is not accounted for by protein binding.^77 This is probably a con-
sequence of the activity of the bacterial DNA gyrase, which is capable of actively introducing further
negative supercoiling, driven by hydrolysis of ATP. This whole process differs in several respects from
assembly of chromatin in eukaryotes.
There is no apparent regular repeating structure equivalent to the eukaryotic nucleosome although
short DNA segments of 60–129 bp are organised by means of their interaction with abundant
DNA-binding proteins.
There is no prokaryotic equivalent to the solenoid structure.
Bacterial DNA seems to be torsionally strained in vivoand organised into independently
supercoiled domains of about 100 kbp.
The establishment of DNA architecture in the bacterial chromosome has progressed through the analy-
sis of two types of structure. First, the interaction of a dimer of the HU protein from B. stearothermophilus
DNA and RNA Structure 69