© 2000 Oxford University PressNucleic Acids Research, 2000, Vol. 28, No. 11
Early melting of supercoiled DNA topoisomers observed
Viktor Víglaský*, Marián Antalík1, Jozef Adamcík and Dušan Podhradský
P.J. Safarik University, Faculty of Sciences, Department of Biochemistry, Moyzesova 11, 041 54 Košice, Slovakia
and1Institute of Experimental Physics, Slovak Academy of Sciences, Watsonova 47, 043 53 Košice, Slovakia
Received January 21, 2000; Revised and Accepted March 31, 2000
We have used temperature gradient gel electro-
phoresis (TGGE) to measure the progress of local
denaturation in closed circular topoisomer DNA as a
function of temperature and superhelicity (σ). We
describe the versatility of this method as a tool for
detecting various conformational modifications of
plasmid DNAs. The early melting temperature of a
structural transition for any topoisomer is dependent
on the value of superhelicity. Supercoiled topo-
isomers represent a system of molecules that is
sensitive to changes in temperature. We show that
the topoisomer with the highest absolute value of
superhelicity melts earlier than topoisomers with
lower values. Thermal sensitivity of highly super-
coiled plasmids could play a biologically important
role in regulation of replication and expression in
cells under thermal stress. The estimated melting
temperature for plasmids with σ < –0.05 is very
significant because these temperatures for early
melting are below physiological temperatures.
The most common method for the study of DNA topological
properties is agarose gel electrophoresis (1). Unlike linear or
nicked circular DNA, whose mobility in agarose gels is primarily
determined by the molecular weight, the mobility of closed
circular DNA also varies with the linking number (Lk) (2).
These species are topological isomers of the molecule and are
generally called topoisomers.
The geometric distortion of a supercoiled molecule means
that the average shape and frictional properties of the molecule
are changed as a result of its topology. Electrophoresis separates
DNA molecules on the basis of size and compactness; smaller
and/or more compact molecules will migrate more rapidly
through the matrix of the gel under the influence of the electric
field. Within a certain range of the absolute value of specific
linkingdifference, σ < 0.05,the gel electrophoretic mobility
of a topoisomer of linking number Lk increases with the
magnitude of ∆Lk (the linking number relative to that for a
relaxed form) (3,4). The resolving power of agarose gels for
topoisomers is impressive, but unfortunately the range is
limited, so that more highly supercoiled species co-migrate as
a broad band. The major problem with one-dimensional gel
electrophoresis is the relatively limited range over which
migration is a function of topology. The traditional approaches
to the study of linear DNA melting proved ineffective in this
case. The resolution of topoisomers has been greatly improved
with the recent use of two-dimensional gel electrophoresis (5).
Information about conformational changes of separated
molecules is also offered by denaturing gradient gel electro-
phoresis (DGGE) and temperature gradient gel electrophoresis
(TGGE). The ability of DGGE to detect changes in plasmid
DNAs was used to study the effect of superhelical density on
melting in different topoisomers (6). Meanwhile, very little is
known about the formation of melted regions in supercoiled
DNA (scDNA). TGGE is easy to perform and allows sensitive
detection of the conformational stability of DNA under a wide
variety of conditions. The sample of DNA in TGGE experiments
moves through different denaturing conditions during electro-
phoresis (7). TGGE very sensitively separates covalently
closed DNA of different structure. On one side of the gradient
the biopolymers migrate as native molecules, on the other they
are denatured, while inbetween the whole transition curvemay
be recorded. Aside from technical ease, TGGE has other
attributes which can be exploited. The method allows one to
fractionate plasmids according to size and shape and to monitor
structural changes in the same experiment. Topoisomers in
TGGE are electrophoresed through an agarose gel which
contains a temperature gradient perpendicular to the direction of
theelectricfield.Wepresentanovelview onthe conformational
transitions of a system of topoisomers induced by temperature.
In the present investigation we examine the TGGE behavior
of an extensive family of topoisomers of plasmids pBR322
(4361 bp) and pUC19 (2763 bp) over a temperature range of
25–75°C. We find that a series of mobility transitions occurs as
linking number and electrophoresis temperature are systemati-
cally varied and reveal the fine structure of early melting in
closed circular DNA. These transitions are manifested as
cyclic variations in mobility with increasing temperature. The
marked differences in electrophoretic mobility between super-
coiled and relaxed DNA molecules made it possible to observe
the denaturation of each DNA topoisomer leading to topological
relaxation. Our results confirm others data (6) that such
transitions occur before the melting of linear molecules; the
transition width varies for different topoisomers. We have
shown that thermal stability varies as a function of superhelical
density for different topoisomers.
*To whom correspondence should be addressed. Tel: +421 95 62 235 82; Fax: +421 95 62 221 24; Email: email@example.com
Nucleic Acids Research, 2000, Vol. 28, No. 11
The degree of supercoiling and topological equilibrium in
the cell is strongly controlled by specific enzymes (DNA
topoisomerases and gyrases) that are capable of either adding
or subtracting supercoiled twists in DNA. The distribution of
topoisomers in the various supercoiled states and their biological
function in the cell are not understood.
MATERIALS AND METHODS
Chemicals and enzymes
All reagents and chemicals used in the experiments were
obtained from commercial sources. Plasmids pUC19 and
pBR322 and agarose type II no. A-6877 was purchased from
Sigma, calf thymus topoisomerase I was purchased from
Amersham Pharmacia Biotech and LB medium from Fluka
Biochemika. The plasmids were used to transform Escherichia
Culture of bacteria and plasmid isolation
Aliquots (250 ml) of exponentially growing E.coli cultured to
an optical density of 0.3–0.4 at 600 nm in LB medium supple-
mented with ampicilin (75 µg/ml) at 37°C were further incubated
in the presence of ampicilin (150 µg/ml) for 16 h. Plasmids
were obtained from these cultures according to the method of
Birnboim and Doly (8). The residual impurities were removed
from cloning vector pUC19 by the following cleaning steps:
equimolar phenol/chloroform extraction, chloroform and
isoamyl acohol extraction (24:1) and precipitation with
isopropyl alcohol and then with 2 vol of ethyl alcohol. The
precipitate of DNA was dried and dissolved in TE buffer
(10 mM Tris–HCl, 1 mM EDTA, pH 8.0). scDNA must be
prepared and handled with care because cleavage anywhere
within the entire circular molecule completely changes the
topology and releases the superhelical constraint.
A method modified from those previously described (2,9,10)
was used. An aliquot of 5 µg of plasmid DNA from a stock
solution (0.5 mg/ml) was incubated for 2 h in 40 µl of TOPO
buffer (5 mM dithiotreitol, 0.1 M Tris, 1 M KCl, 50 mM
MgCl2, 1 mM ethidium bromide) and 1 µl of topoisomerase I
from a stock solution (~5 U/µl) at 37°C. The reaction was
stopped with 2% SDS. Topoisomerase and ethidium bromide
were removed by the same procedure as was used for the
removal of impurities from plasmid DNA.
Two-dimensional gel electrophoresis
Electrophoresis was performed as described previously (11).
Briefly, the first dimension gel electrophoresis was conducted
in tube gels of 1.0% agarose in 0.5× TBE buffer. The tube gel
was prepared with a 40 × 0.4 cm (i.d.) glass tube, partially
constricted at the bottom. Electrophoresis was performed for
60 h at 100 V (2.5 V/cm) in an electrophoresis apparatus
maintained at a constant temperature (25°C) by a water incu-
bator/circulator. The seconddimensiongel electrophoresis was
conducted by placing the appropriately sliced tube from the
first dimension into the upper slot of a slab gel, also of 1%
agarose in the same buffer and containing chloroquine at a
concentration of 1.0 µM. The second dimension electro-
phoresis was conducted for 23 h at 5 V/cm and 25°C.
Apparatus for TGGE
The equipment was basicallysimilar to the device described by
Riesner (7). A temperature gradient was formed in a gel
perpendicular to the direction of the electric field. The gradient
was established on a copper plate adjacent to the electro-
phoretic apparatus by cooling and heating of the plate at
opposite sides with two independently circulating water baths.
The gradient of temperature was linear along the plate, as was
confirmed by measuring the temperature over the whole plate
with a thermistor.
The agarose gel was in 0.5× TBE buffer (40 mM Tris, 2 mM
of DNA topoisomers was performed in 1% agarose in the same
buffer as was used to prepare the gel. Before application of the
DNA sample it is advisable to pre-electrophorese the gel for
~10–15 h. This step helps to remove reactive charged
compounds and low molecular weight impurities from the gel
that may cause artifacts. Electrophoresis was started without a
temperature gradient. After 30 min the temperature gradient
was established. The electric field was switched off during
formation of the temperature gradient. The lower temperature
part of the apparatus was set up at 20°C and the higher part at
80°C. After temperature gradient stabilization the DNA
sample was applied to the starting slot of the gel. Before
electrophoresis 0.5 µg of DNA were dissolved in 40 µl of gel
loading buffer (10mM Tris, 1mM EDTA,pH 8.0,20% glycerol,
0.1% bromophenol blue). During electrophoresis the gels were
submerged in the selected buffer, the sample was run at a
constant 5.6 V/cm for 8 h and the electrophoretic buffer was
recirculated. The nucleic acids were visualized by staining
with the fluorescent intercalating dye ethidium bromide
(1.5 µg/ml) and destained in distilled water.
Comparison of two-dimensional gel electrophoresis and
Two-dimensional electrophoresis was carried out to determine
the distribution of pBR322 plasmids of different superhelical
densities at room temperature (~25°C). The second dimension
was performed in the presence of the intercalating agent
chloroquine (1 µM), which causes a decrease in twist (Tw).
This changes the twist of each DNA molecule by about four
turns. More than 24 discrete topoisomers could be resolved in
this gel, as shown in Figure 1. Clearly, the position of open
circular DNA (ocDNA) is not identical to the linear, super-
coiled or relaxed forms of DNA. The level of linear form was
estimated to be <1% of total plasmid.
Figure 2 shows the corresponding TGGE records of pBR322
topoisomers. The chloroquine concentration used in TGGE
experiments was 1 µM. Every band represents the mobility of
one topoisomer at different temperatures. The most intensive
band consists of ocDNA. The less intensive but well-separated
bands correspondtoscDNAs; their mobilities increase withthe
absolute value of writhe (Wr) at a temperature of ~35°C.
If the shape and charge of a DNA molecule remain constant,
mobility is a linearly increasing function of temperature. The
Nucleic Acids Research, 2000, Vol. 28, No. 11
structural transition of a topoisomer is seen as a continuous or
discontinuous band which is retarded or accelerated in the
temperature range of transition (12). It is possible to observe a
decrease and an increase in mobility at certain temperatures
characteristic for each topoisomer. We observe clear cut
transitions,but the transition curve for some of the topoisomers
intersect, making it difficult to identify unambiguously the
transition curves for all topoisomers.
To get rid of the ambiguity, we performed analogous experi-
ments with shorter scDNA preparations containing fewer
topoisomers under the same conditions. TGGE of pUC19 is
presented in Figure 3. Figure 3a presents the result of gel
electrophoresis for a DNA preparation containing only eight
topoisomers. In another electrophoretic experiment the DNA
was electrophoresed without a temperature gradient for 14 h at
4 V/cm and only after this procedure was the temperature
Figure 1. Two-dimensional gel electrophoretic analysis of topoisomers of pBR322. The position of nicked (oc) and linear (L) DNA are designated. (Left) Gel
photograph; (right) schematic representation of the gel photograph. The first dimension was performed in a 1% agarose gel at 2.5 V/cm in 0.5× TBE for 25 h. For
the second dimension, the gel was turned through 90° and run for another 22 h at 5 V/cm in 0.5× TBE containing 1 µM chloroquine.
Figure 2. Temperature gradient gel of topoisomers of pBR322. The direction
of mobility of the sample is from anode (up) to cathode (down). The gels were
electrophoresed with the indicated temperature gradient in 0.5× TBE buffer
containing 1 µM chloroquine at 6 V/cm for 12 h.
Figure 3. Temperature gradient gel of topoisomers of pUC19. The electro-
phoresis conditions in (a) were the same as in Figure 2. The sample of DNA
was run for 9 h. The experiment in (b) was also performed under the same con-
ditions. However, the topoisomers were run at constant temperature (20°C)
without a temperature gradient at 4 V/cm for 14 h, then the temperature
gradient was switched on at 6 V/cm for 9 h. The numbers on both records
denote absolute values of writhing number.
Nucleic Acids Research, 2000, Vol. 28, No. 11
gradient applied and electrophoresis was continued for 9 h at
6 V/cm (Fig. 3b). Clearly, we get a much better separation of
the less supercoiled topoisomers. The comparison of two-
dimensional and TGGE under identical conditions enabled us
to identify all transitions, which are numbered in the figure.
The numbers denote the absolute value ofthe writhingnumber.
The difference between the mobility of DNA without writhe
(Wr ~ 0) and ocDNA is close to 0 and marked conformational
changes cannot be seen with a temperature increase (5,13).
The transitions for topoisomers 1–8 are very clear. We can
see that topoisomer 8 is the first to melt; followed by 7, then 6,
etc. Generally, from both figures (Figs 2 and 3) it is obvious
that the topoisomer with the highest absolute value of super-
helicity melts earlier than topoisomers with lower values.
Analysis of melting curves
The curves are evidently not sigmoidal, as in experiments with
a denaturing gradient (6). A sigmoid curve is characteristic for
two-state processes. For data from TGGE it is not advisable to
fit by means of a two-state mechanism (14,15). For a more than
two-state system it is impossible to determine strictly the
temperature of transition.
Therefore, to obtain more precise information on initiation
of the melting process, we determined the derivative of
mobility with respect to temperature separately for each topo-
isomer (Fig. 4). The temperature Ti,mat the first local
maximum of dµ/dT is an intrinsic property of the topoisomer
transition temperature. The peaks are clearly defined for most
supercoiled topoisomers of pBR322 with an absolute value of
Wr > 5. It is possible to estimate the transition width for various
topoisomers at Ti,m. It amounts to ~3–7°C. The accuracy of fit
analysis for supercoiled pUC19 is higher then for pBR322 for
the reason described above.
Figure 5 shows the plot of Ti,mversus corresponding super-
helicity σifor pUC19 and pBR322. It reflects a topological
unfolding (melting) of the DNA structure. From this plot it is
evident that Ti,mdecreases with increasing superhelicity.
We show in the present paper the effect of temperature on the
early melting of superhelical DNAs with different superhelicities.
The results were obtained by two-dimensional gel electro-
phoresis and TGGE. The results presented above demonstrate
that TGGE makes it possible to observe the melting of scDNA
caused by supercoiling. Since it is possible to observe this
transition for individual topoisomers, one can conclusively
prove that the previously observed early melting of scDNA
leads to transient states of supercoiled molecules with lower
values of negative writhe.
The so called ‘native’ negatively supercoiled plasmid DNAs
isolated from bacteria consist of a distribution of topoisomers.
One requirement for the investigation of DNA melting of
different topoisomers by TGGE is a good distribution of super-
coiled plasmids. This should cover the required range,
i.e. topoisomers of σ = 0–0.045, where the distribution of
topoisomers can be described by an ideal Gaussian curve at
room temperature (3). The electrophoretic resolution of DNA
topoisomers is more sensitive for topoisomers with lower
values of Wr. We obtained such preparations by treating the
original plasmids pUC19 and pBR322 with topoisomerase I in
the presence of ethidium bromide and the electrophoretic
buffer and gel contained the intercalating agent chloroquine,
which also affects twist (9,13). The average value of linking
number obtained on relaxation depends on the solution condi-
tions under which the relaxation took place. Specifically, the
change in twist caused by topoisomerase I and/or altered
Figure 4. The derivative of mobility with respect to temperature for each
topoisomer. dµ/dT for topoisomers of (upper) pUC19 and (lower) pBR322.
Temperature at the first local maximum of a topoisomerwith σiis denoted Ti,m.
Figure 5. A plot of Ti,mversus superhelicity σifor pUC19 (closed circle) and
pBR322 (open circle). It reflects a topological unfolding (melting) of the DNA
structure. The error in Ti,mdetermination increases with more relaxed
Nucleic Acids Research, 2000, Vol. 28, No. 11
conditions forces the writhe to change from its initial average
value to a new value, and hence the DNA will run with a
different mobility in the gel, as we can see in Figures 1–3.
Structures such as cruciforms and Z-form DNA, which are
unstable in relaxed or linear DNA, are formed cooperatively at a
threshold level of supercoiling (16). We have attempted to
select conditions where the values of superhelicity of topoisomers
are under this threshold.
The mobility difference between adjacent topoisomer bands
is a hydrodynamic property and is therefore determined as first
order by the associated difference in Wr. Changes in Wr
between successive topoisomer bands might not be accompanied
by changes in Tw (2,17). The fact that any non-linear mobility
behavior reflects a conformational change in the molecule
allows the study of topological changes in scDNA. The
decrease in mobility of an individual topoisomer (Figs 2 and 3)
during the early melting process is based on induction of
consecutive decoupling of weak interactions (hydrogen bonds
and stacking interactions) in the structure of DNA. The change
in temperature alters the helical repeat (h) of the DNA and
hence changes the value of Tw0(twist) and hence Wr. Next, the
weak interactions is observed at temperatures above 65°C. In
this case the covalently closed complementary interwound
strands cannot be completely separated to a larger distance.
This state is analogous to the denatured state of proteins
The nicked form of plasmid DNA (ocDNA) undergoes
partial or/and total denaturation of base pairs (helix to random
coil transition) similarly to linear DNA, resulting in a rapid
decrease in mobility at higher temperatures. Denaturation is
detected by electrophoresis as part of a band with a very low
mobility value. Nicked circular molecules are not necessarily
present in vivo, but may be formed by breakage of one strand
treatment with topoisomerase I. However, at temperatures >70°C
separation of ocDNA and relaxed scDNA without writhe is
possible by electrophoresis.
Under our conditions (see Materials and Methods) the
melting of linear pUC19 and pBR322 molecules starts at
temperatures >70°C. Consequently, the Ti,mvalues in Figure 4
for all topoisomers are below the melting temperature of linear
and ocDNA. Hence, what we observe is early melting of
scDNA. Our further experiments with others plasmids have
shown that Ti,mis independent of the sequence and length of
scDNA. It appears to be a function only of superhelicity. Local
denaturation is closely coupled to supercoiling in closed DNA.
However, a correlation based on existing models of melting of
scDNA between a calculated denaturation profile for an indi-
vidual topoisomer, the transitiontemperature and cooperativity
were not executed (18,19).
It is worthwile to estimate the Ti,mof the ‘higher’ topo-
isomers, which were not directly detected by TGGE, by linear
extrapolation, because highly supercoiled DNA topoisomers
are naturally occurring in cells (20–22). The estimated Ti,mfor
plasmids with σ < –0.05 is very significant because these
temperatures for earlymeltingare below physiological temper-
Many recent experiments have shown the importance of
DNA supercoiling and topoisomerase action in gene activa-
tion, replication and recombination (5,20–23). The topo-
isomers with higher negative values of σ (more condensed) are
more sensitive to structural changes induced by temperature.
The thermal sensitivity of supercoiled plasmids could play a
biologically important role in regulation of replication and
gene expression under thermal stress. The excess free energy
associated with negative supercoiling of DNA may be utilized
in many cellular mechanisms. In general, processes that
require untwisting or writhing of DNA or which stabilize such
deformations are facilitated with negatively supercoiled as
compared to relaxed DNA. Examples of such processes
include the replication and transcription of DNA, which
require unwinding of the DNA helix, and the formation of
nucleosomes and other protein complexes on DNA, which
stabilize negative writhing of the helix (5,10,20–25).
This study was supported in part by grants nos 5053 and 6116
from the Slovak Grant Agency.
1. Mickel,S., Arena,V.J.andBauer,W.(1977)NucleicAcidsRes.,4, 1465–1482.
2. Depew,R.E. andWang,J.C. (1975)Proc.NatlAcad. Sci.USA, 72,4275–4279.
3. Lyamichev,V.I., Panyutin,I.G. and Frank-Kamenetskii,M.D. (1983)
FEBS Lett., 153, 298–302.
4. Pulleyblank,D.E., Shure,M., Tang,D., Vinograd,J. and Vosberg,H.P.
(1975) Proc. Natl Acad. Sci. USA, 72, 4280–4284.
5. Wang,J.C., Peck,L.J. and Becherer,K. (1983) Cold Spring Harbor Symp.
Quant. Biol., 47, 85–91.
6. Lyubchenko,Y.L. and Shlyakhtenko,L.S. (1988) Nucleic Acids Res., 16,
7. Riesner,D., Steger,G., Zimmat,R., Owens,R.A., Wagenhöfer,M., Hillen,W.,
Vollbach,S. and Henco,K. (1989) Electrophoresis, 10, 377–389.
8. Birnboim,H.C. and Doly,J. (1979) Nucleic Acids Res., 7, 1513–1523.
9. Keller,W. (1975) Proc. Natl Acad. Sci. USA, 72, 4876–4880.
10. Peck,L.J. and Wang,J.C. (1983) Proc. Natl Acad. Sci. USA, 80, 6206–6210.
11. Bauer,W.R. and Benham,C.J. (1993) J. Mol. Biol., 234, 1184–1196.
12. Horowitz,D.S. and Wang,J.C. (1984) J. Mol. Biol., 173, 75–91.
13. Rahmouni,A.R. and Wells,R.D. (1989) Science, 246, 358–363.
14. Creighton,T.E. (1990) Protein Structure: A Practical Approach.
Oxford University Press, Oxford, UK, pp. 225–255.
15. Pace,C.N. (1990) Trends Biotechnol., 8, 93–98.
16. Jaworski,A., Hsieh,W.T., Blaho,J.A, Larson J.E. and Wells,R.D. (1987)
Science, 238, 773–777.
17. Shore,D. and Baldwin,R.L. (1983) J. Mol. Biol., 170, 983–1007.
18. Kundu,S., Lahiri,A. and Thakur,A.R. (1998) Biophys. Chem., 75, 177–186.
19. Schaeffer,F., Yeramian,E. and Lilley,D.M. (1989) Biopolymers, 28,
20. Wang,J.C. (1971) J. Mol. Biol., 55, 523–533.
21. Bates,A.D. andMaxwell,A.(1993) DNATopology.OxfordUniversityPress,
New York, NY, pp. 83–103.
22. Shore,D. and Baldwin,R.L. (1983) J. Mol. Biol., 170, 957–981.
23. Zechiedrich,E.L., Khodursky,A.B., Bachellier,S., Schneider,R., Chen,D.,
Lilley,D.M. and Cozzarelli,N.R. (2000) J. Biol. Chem., 275, 8103–8113.
24. Rybenkov,V.V., Vologodskii,A.V. and Cozzarelli,N.R. (1997) J. Mol. Biol.,
25. Bagga,R., Ramesh,N. and Brahmachari,S.K. (1990) Nucleic Acids Res.,
e51 Download full-text
Nucleic Acids Research, 2000, Vol. 28, No. 11