R778 Dispatch

DNA topology: Topoisomerases keep it simple

Andrew D. Bates* and Anthony Maxwell†

The ability of type II DNA topoisomerases to perturb the
equilibrium distributions of DNA topoisomers is a
consequence of their ability to hydrolyse ATP. A sliding
mechanism of topoisomerase action has been
proposed to account for this phenomenon.
Addresses: *School of Biological Sciences, University of Liverpool, Life
Sciences Building, Crown Street, Liverpool L69 7ZB, UK. E-mail:
bates@liv.ac.uk. †Department of Biochemistry, University of Leicester,
University Road, Leicester, LE1 7RH, UK. E-mail: ony@le.ac.uk.
Current Biology 1997, 7:R778–R781
http://biomednet.com/elecref/09609822007R0778
DNA gyrase, which is the only type II enzyme known to
be capable of introducing supercoils into DNA, the
requirement for ATP is clear. The introduction of super-
coils into a relaxed DNA circle is energetically
unfavourable and it seems that, under the right condi-
tions, much of the energy of ATP hydrolysis can be con-
verted into DNA supercoiling energy [4–6]. Gyrase
performs this directional intramolecular reaction by
wrapping a segment of DNA around itself, and deliver-
ing a nearby T segment to the ATP-operated clamp with
a specific orientation [7].

© Current Biology Ltd ISSN 0960-9822 In the case of the eukaryotic type II enzymes or the
bacterial enzyme topoisomerase IV (topo IV), which are
One of the basic tenets of enzyme-catalysed reactions is specifically implicated in the unlinking of daughter
that an enzyme cannot alter the equilibrium between chromosomes after DNA replication [8,9], the energy
substrate and product — it can only accelerate the rate of requirement is not so clear. These enzymes do not wrap
interconversion between them. However, if the enzyme DNA, and they catalyse reactions that appear to be ener-
couples an energy-yielding reaction, such as the hydrolysis getically favourable, such as the relaxation of super-
of ATP, to the conversion of substrate to product, then an coiled DNA or the unlinking of catenanes, yet they too
apparently non-equilibrium mixture can be generated. It require the hydrolysis of ATP. It has been suggested
has long been a puzzle why many type II DNA that this energy input is needed to drive the enzyme
topoisomerases hydrolyse ATP while catalysing energeti- through an ordered series of conformational changes
cally favourable reactions, such as the relaxation of [10], in other words, to facilitate the opening and closing
supercoiled DNA. Recent experiments by Rybenkov et al.
[1] show that these enzymes are able to reduce the
topological complexity of DNA relative to equilibrium Figure 1
values, and ATP hydrolysis is invoked to explain their
ability to perturb the thermodynamic equilibrium.
(a)
DNA topoisomerases catalyse the interconversion of differ-
ent topological forms of DNA, including relaxed and super-
coiled, knotted and unknotted, and catenated (linked) and
unlinked DNA circles (Figure 1). They are classified into
two types: type I enzymes catalyse reactions involving tran-
sient single-strand breaks in DNA, and type II enzymes (b)
catalyse reactions involving transient double-strand breaks
[2]. The current mechanistic model for type II enzymes +
involves the binding of two segments of DNA: a G (gate)
segment, which is broken in both strands by the enzyme
with the formation of covalent bonds between active-site
(c)
tyrosines and 5′ phosphates in the DNA, and a T (trans-
ported) segment, which is captured by an ATP-operated
clamp and passed through the enzyme-stabilised break in
the G segment [3]. This passage of one double-stranded
segment through another, in either an intramolecular or Current Biology
intermolecular fashion, can accomplish all the reactions
illustrated in Figure 1. The reactions carried out by type II topoisomerases: (a) supercoiling/
relaxation; (b) catenation/decatenation; (c) knotting/unknotting. All these
transformations can be performed by the passage of one double-
Most reactions catalysed by type II enzymes require stranded DNA segment through another (double-headed arrows).
ATP hydrolysis. In the case of the bacterial enzyme

Figure 2
Equilibrated
mixture
Topo IV + ATP
Topo IV + ATP
Steady-state
mixture
Topo IV + ATP (Topo IV + ATP)
Current Biology
A schematic illustration of the reaction of topoisomerase IV with DNA
catenanes. The steady-state mixture of catenated and unlinked
products formed in the presence of topo IV and ATP contains a lower
proportion of catenanes than the equilibrium mixture formed by the
cyclisation of P4 DNA in the presence of pAB4 plasmid. Catenanes
are in red, and unlinked DNA circles are in green.
of the ATP-operated clamp that captures the T segment
and initiates the strand-passage process.
A number of recent studies have suggested that these
enzymes may be acting more specifically than previously
thought — for example topo IV has been shown to be more
active in decatenation than in relaxation reactions [11,12].
Recent work by Rybenkov et al. [1] now suggests an alter-
native explanation for the ATP requirement of these
enzymes. In previous studies, these workers developed a
system that allows equilibration of catenanes without the
requirement for enzymic catalysis, consisting of a nicked
plasmid (pAB4) and linear phage P4 DNA, which has long
cohesive ends [13]. Under conditions of equilibration, the
P4 DNA molecules circularise, trapping some of the pAB4
plasmids to form dimeric catenanes. The proportion of
catenated DNA at equilibrium can be determined using an
agarose gel. This system enabled Rybenkov et al. [1] to
compare samples of catenated DNA generated either by
equilibration or by the action of type II topoisomerases
(Figure 2). The fractions of catenated DNA formed were
shown to be very different in these two cases.
When a sample with twice the equilibrium concentration
of catenanes was treated with topo IV and ATP at a range
of enzyme:DNA ratios, the steady-state fraction of cate-
nanes in the sample was as much as 16-fold lower than the
value at equilibrium in the absence of enzyme, and this
same fraction was obtained when the initial substrate was
a mixture of unlinked circles at the same concentrations
(Figure 2). For both pAB4 and P4 DNA, the fraction of
knotted versus unknotted species was also measured
Dispatch R779
using an analogous method, and similarly found to be far
from the equilibrium value when enzyme and ATP were
present: 90-fold lower for pAB4, and 50-fold lower for P4.
With supercoiled pAB4 plasmid, the relaxation reaction
was studied and topo IV was found to generate a distribu-
tion of topoisomers with a narrower range of linking
numbers (Lk) than that generated by wheat-germ topoiso-
merase I, an ATP-independent DNA relaxing enzyme.
This means that topo IV can reduce the level of supercoil-
ing in a sample below the equilibrium level established by
thermal energy [14]. Similar experiments were carried out
with other type II enzymes from phage T2, Saccharomyces
cerevisiae, Drosophila melanogaster and human cells, and
although the magnitude of the effect varied from enzyme
to enzyme, all reduced the fraction of catenanes and knots
and narrowed the Lk distribution. In other words, all these
enzymes actively simplify DNA topology relative to the
complexity expected at equilibrium.
To comply with the laws of thermodynamics, the energy
of ATP hydrolysis must be invoked to drive these reac-
tions away from their equilibrium positions. With hind-
sight, it seems clear that coupled ATP hydrolysis should
enable type II topoisomerases to generate non-equili-
brated products, given the persuasive example of the
supercoiling reaction catalysed by DNA gyrase (see
above). The extreme nature of the gyrase case may have
blinded us to the rather more subtle effects of the other
type II enzymes. The experiments of Rybenkov et al. [1]
elegantly demonstrate these effects, and show that all type
II enzymes act with a specific directionality: gyrase in the
direction of increasing supercoiling complexity, and the
other enzymes in the direction of simplifying DNA topol-
ogy. This behaviour is consistent with our understanding
of the roles of these enzymes in vivo; gyrase maintains the
level of negative supercoiling in bacterial cells and specifi-
cally removes positive supercoiling generated by transcrip-
tion and replication, whereas the other type II enzymes
relax both positive and negative supercoiling, and unlink
daughter chromosomes during and after replication [15].
To envisage a mechanism that will explain this behaviour
is a more difficult matter. In order to perturb the equilib-
rium between catenated and unlinked DNA, for example,
the enzymes need to identify specifically a G segment and
a T segment that are located on two catenated circles in
preference to a pair of segments on unlinked molecules.
Rybenkov et al. [1] propose a model in which each enzyme
can bind to three sites on DNA. Initial binding to two seg-
ments of a DNA circle, the G segment (which will become
cleaved) and another distant site, might constrain a cate-
nated or knotted strand or a supercoiled region (destined
to become the T segment) within a shorter loop of the
circle, and thus make capture of the T segment more
probable (Figure 3). Although there is some evidence for
R780 Current Biology, Vol 7 No 12
the interaction of yeast topoisomerase II with two distant
sites on DNA [16], such an effect will not, of itself, explain
the perturbation of the equilibrium. It is therefore envis-
aged that the enzyme can slide along the DNA at the third
site, constraining the prospective T segment in a progres-
sively smaller region and making its capture increasingly
probable (Figure 3). This directional sliding would have to
be the energy-requiring step of the process.
This idea, however, is at odds with what we know of the
properties of the ATP binding and hydrolysing domains of
these enzymes. There is abundant evidence from studies
of gyrase and yeast topoisomerase II that binding of ATP
acts to close a clamp by domain dimerisation, trapping the
T segment, which is then passed through the G segment
[3,17,18]. Subsequent hydrolysis of the ATP re-opens the
clamp ready for a new reaction. It is difficult to imagine a
dual role for ATP hydrolysis both in clamp opening and
closing, and in driving the sliding of the topoisomerase on
DNA, but as there is no obvious alternative mechanism to
account for these phenomena, we are left with this intrigu-
ing model.
The work of Rybenkov et al. [1] also raises other interest-
ing questions. For example, which part of the enzyme
binds the postulated third DNA segment? It would have
been tempting to propose that this segment binds to the
carboxy-terminal domain, which does not have an identi-
fied role in the reaction mechanism of non-gyrase type II
enzymes. However, Rybenkov et al. [1] show that a
mutant form of yeast topoisomerase II that lacks the last
approximately 200 amino acids can also perturb the topo-
logical equilibria (actually in a more extreme fashion than
the full-length enzyme).
Does DNA gyrase have the same ability to generate non-
equilibrium topoisomer products? Although the answer is
obviously ‘yes’ in the case of the supercoiling reaction, it
would be interesting to compare the products of the ATP-
dependent and ATP-independent relaxation reactions of
the truncated form of the enzyme [19], and to examine the
(albeit rather inefficient) decatenation reaction of full-
length gyrase.
What is the free-energy requirement of the proposed
sliding reaction? Is the extent of perturbation of the equi-
libria dependent on the amount of free energy available
from ATP hydrolysis? It would be interesting to examine
the distribution of the products of type II topoisomerase
reactions in the presence of ATP analogues that yield
more or less free energy of hydrolysis than ATP, or using a
range of ATP:ADP ratios [4].
Perhaps the most important aspect of the work of
Rybenkov et al. [1] is that it has clearly established that all
the type II topoisomerases can perturb the equilibria of
Figure 3
G segment
T segment
Topoisomerase
ATP hydrolysis
Current Biology
The model proposed by Rybenkov et al. [1] for the ATP-driven
decatenation reaction of type II topoisomerases. The enzyme (blue)
binds to the G segment and another distant site, at which ATP
hydrolysis can drive the enzyme along the DNA, constraining the T
segment in a smaller loop and making its capture more probable. The
notch in the enzyme indicates the ATP-operated clamp and the arrows
indicate the ATP-dependent movement of DNA.
DNA topological forms in the direction consistent with
their biological role. How the enzymes achieve this
remains to be shown. It is clear that there are many more
complexities of these remarkable enzymes which remain
to be unravelled.
Acknowledgements
We would like to thank Tim Hammonds, Sotirios Kampranis, Stephen
Robertson and Melisa Wall for helpful discussions and comments on the
manuscript. AM is a Lister-Jenner Research Fellow.
References
1. Rybenkov VV, Ullsperger C, Vologodskii AV, Cozzarelli NR:
Simplification of DNA topology below equilibrium values by type
II topoisomerases. Science 1997, 277:690-693.
2. Bates AD, Maxwell A: DNA Topology. Edited by Rickwood D, Male D
Oxford: IRL Press; 1993.
3. Roca J, Wang JC: The capture of a DNA double-helix by an ATP-
dependent protein clamp: a key step in DNA transport by type II
DNA topoisomerases. Cell 1992, 71:833-840.
4. Westerhoff HV, O’Dea MH, Maxwell A, Gellert M: DNA supercoiling
by DNA gyrase. A static head analysis. Cell Biophys 1988, 12:157-
181.
5. Bates AD, Maxwell A: DNA gyrase can supercoil DNA circles as
small as 174 base pairs. EMBO J 1989, 8:1861-1866.
6. Cullis PM, Maxwell A, Weiner DP: Energy coupling in DNA gyrase: a
thermodynamic limit to the extent of DNA supercoiling.
Biochemistry 1992, 31:9642-9646.
7. Bates AD, O’Dea MH, Gellert M: Energy coupling in Escherichia
coli DNA gyrase: the relationship between nucleotide binding,
strand passage and DNA supercoiling. Biochemistry 1996,
35:1408-1416.
8. DiNardo S, Voelkel K, Sternglanz R: DNA topoisomerase II mutant
of Saccharomyces cerevisiae: Topoisomerase II is required for
segregation of daughter molecules at the termination of DNA
replication. Proc Natl Acad Sci USA 1984, 81:2616-2620.
 Dispatch R781

9. Adams DE, Shekhtman EM, Zechiedrich EL, Schmid MB, Cozzarelli

NR: The role of topoisomerase IV in partitioning bacterial
replicons and the structure of catenated intermediates in DNA
replication. Cell 1992, 71:277-288.
10. Liu LF, Liu C-C, Alberts BM: Type II DNA topoisomerases: enzymes
that can unknot a topologically knotted DNA molecule via a
reversible double-strand break. Cell 1980, 19:697-707.
11. Peng H, Marians KJ: Escherichia coli topoisomerase IV.
Purification, characterisation, subunit structure and subunit
interactions. J Biol Chem 1993, 268:24481-24490.
12. Ullsperger C, Cozzarelli NR: Contrasting enzymatic activities of
topoisomerase IV and DNA gyrase from Escherichia coli. J Biol
Chem 1996, 271:31549-31555.
13. Rybenkov VV, Vologodskii AV, Cozzarelli NR: The effect of ionic
conditions on the conformations of supercoiled DNA. II.
Equilibrium catenation. J Mol Biol 1997, 267:312-323.
14. Depew RE, Wang JC: Conformational fluctuations of DNA helix.
Proc Natl Acad Sci USA 1975, 72:4275-4279.
15. Wang JC: DNA topoisomerases: why so many? J Biol Chem 1991,
266:6659-6662.
16. Roca J, Berger JM, Wang JC: On the simultaneous binding of
eukaryotic DNA topoisomerase II to a pair of double-stranded
DNA helices. J Biol Chem 1993, 268:14250-14255.
17. Wigley DB, Davies GJ, Dodson EJ, Maxwell A, Dodson G: Crystal
structure of an N-terminal fragment of the DNA gyrase B protein.
Nature 1991, 351:624-629.
18. Ali JA, Jackson AP, Howells AJ, Maxwell A: The 43 kilodalton N-
terminal fragment of the DNA gyrase B protein hydrolyses ATP
and binds coumarin drugs. Biochemistry 1993, 32:2717-2724.
19. Kampranis SC, Maxwell A: Conversion of DNA gyrase into a
conventional type II topoisomerase. Proc Natl Acad Sci USA 1996,
93:14416-14421.