week ending

PRL 105, 158101 (2010) PHYSICAL REVIEW LETTERS 8 OCTOBER 2010

DNA–DNA Interactions in Tight Supercoils Are Described by a Small Effective Charge Density
Christopher Maffeo,1 Robert Schöpflin,2 Hergen Brutzer,3 René Stehr,2 Aleksei Aksimentiev,1,*
Gero Wedemann,2,† and Ralf Seidel3,‡
1
Department of Physics, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801, USA
2
CC Bioinformatics, University of Applied Sciences Stralsund, 18435 Stralsund, Germany
3
Biotechnology Center Dresden, University of Technology Dresden, 01062 Dresden, Germany
(Received 12 May 2010; published 4 October 2010)
DNA-DNA interactions are important for genome compaction and transcription regulation. In studies of
such complex processes, DNA is often modeled as a homogeneously charged cylinder and its electrostatic
interactions are calculated within the framework of the Poisson-Boltzmann equation. Commonly, a charge
adaptation factor is used to address limitations of this theoretical approach. Despite considerable theoretical
and experimental efforts, a rigorous quantitative assessment of this parameter is lacking. Here, we
comprehensively characterized DNA-DNA interactions in the presence of monovalent ions by analyzing
the supercoiling behavior of single DNA molecules held under constant tension. Both a theoretical model
and coarse-grained simulations of this process revealed a surprisingly small effective DNA charge of 40% of
the nominal charge density, which was additionally supported by all-atom molecular dynamics simulations.

DOI: 10.1103/PhysRevLett.105.158101 PACS numbers: 87.14.gk, 82.37.Rs, 87.15.A
, 87.15.La

The large linear charge density is a fundamental prop-
erty of DNA which governs its biological function by
influencing DNA folding, packaging [1], pairing [2], and
interactions with other biological macromolecules [3]. In
order to develop meaningful quantitative models describ-
ing such systems and processes, a precise knowledge of the
interaction between two DNA molecules, which is mostly
of electrostatic origin, is mandatory. DNA electrostatics is
affected by surrounding counterions, which screen the
DNA charge on the scale of the Debye length
D . The
counterion cloud is mainly set by the interplay between
solute-ion electrostatic attraction and entropic repulsion,
for which the Poisson-Boltzmann (PB) equation provides a
mean-field description. For highly charged polymers such
as DNA, this approach can have considerable limitations
arising from the reduced structural detail with which the
DNA macromolecule is approximated and the assumption
of a continuous counterion density. To offset these prob-
lems, it is common practice to rescale DNA electrostatic
potentials with a charge adaptation factor. Values between
70% and 100% of the bare DNA charge density are typi-
cally used [4–6], but the correct parameterization is de-
bated [4]. Often such factors are indirectly obtained from
electrophoresis experiments [7], which are unrelated to
DNA-DNA interactions [8,9]. Direct experimental studies
of DNA-DNA interaction are rare and themselves limited,
e.g., to condensed DNA phases [10], short DNA
particles [11], or large distances [12,13]. Despite increas-
ingly sophisticated experiments, an unambiguous quanti-
tative assessment of DNA-DNA interactions has not been
achieved yet.
Here we address this issue by analyzing the ionic
strength-dependent supercoiling response of single DNA
molecules held under constant tension in magnetic
tweezers experiments [14–16]. An illustration of the ex-
periment is shown in Fig. 1(a). A DNA molecule is tethered
between a glass surface and a 1:0 m magnetic bead (see
supplementary material [17]). Nearby magnets allow us to
stretch and twist the attached molecule. Initially, the mea-
sured DNA end-to-end distance reduces only slightly upon
supercoiling. Once a critical supercoil density is reached,
the molecule buckles and the end-to-end distance LDNA
decreases linearly with the number of added turns N
[Fig. 1(b)] as the extrusion of a superhelical structure
absorbs the additional turns in the form of writhe [18].
While the torque within the DNA increases linearly before
the supercoiling transition, it remains constant afterwards
[14,19]. The slope dLDNA =dN in the superhelical phase
depends on the applied force and on the ionic strength of
the solution [Fig. 1(c)]. Several models have been devel-
oped to describe this dependence theoretically [20,21], but
a quantitative prediction of the slopes has not yet been
achieved [14,21] (see Fig. S1 [17]).
In order to provide an improved description of the super-
helical regime, we calculate the energy Esh
tot per added turn
to form an ideal DNA superhelix with superhelical radius
 and helical repeat length h in the absence of fluctu-
ations [4,5,21] [Fig. 1(a)], with the DNA charge as a free
parameter:
tot ¼ Epot þ Ebend þ EEstat :
Esh force DNA DNA
(1)
The value Eforce
pot denotes the potential energy change due to
shortening the DNA end-to-end distance against the
applied force F, EDNAbend the bending energy of the DNA
within the superhelix, and EDNAEstat the DNA-DNA electro-
static interaction energy. Per added turn, i.e., per super-
helical writhe, the DNA length within the superhelix grows
by [5] ½ð2Þ2 þ h2 =h ¼ dLDNA =dN, which equals the

0031-9007=10=105(15)=158101(4) 158101-1 Ó 2010 The American Physical Society

 week ending
PRL 105, 158101 (2010) PHYSICAL REVIEW LETTERS 8 OCTOBER 2010

a b nominal charge density. The electrostatic potential along

600
N S one DNA double strand is obtained by integrating point

DNA length (nm)

charge potentials placed at the center line of the opposite
Magnets F
400 double strand. The DNA charge density  entering the
potential is adapted (see Fig. S2 [17]) to account for
h the cylindrical geometry (factor Rod ) and for deviations
200
DNA due to using solutions to the linearized PB equation (factor
2ρ 0 5 10 15 PB ). For an ideal infinite superhelix the electrostatic
Turns
potential is invariant along the DNA contour, and the
c electrostatic interaction energy per added turn is approxi-
70 1.00
70 Slope (nm) 0.55 mately the product of the calculated potential (see Fig. S2
60 CR
0.42
0.32
[17]) and the adapted charge of one of the double strands
60
50
[22], 1=2dLDNA =dN :
40
Slope (nm)

30 60 mM Na+ 1 dLDNA Z e
rðsÞ=
D

50
0 1
Force (pN)
40
30
CR= 0.42
20
0 1 2 3
Force (pN)
FIG. 1 (color). Dependence of DNA supercoiling on force and
salt concentration. (a) Experimental setup and superhelix pa-
rameters. (b) DNA supercoiling curves recorded in buffer con-
taining 170 mM Naþ at stretching forces of 0.25, 0.5, 1.0, 2.0,
3.0, and 4.0 pN (gray, light blue, dark blue, red, green, and dark
gray lines, respectively) for a 1.9 kbp long DNA molecule.
Continuous twisting was carried out at 0.5 Hz. Data were taken
at 300 Hz and smoothed to 20 Hz. (c) Slopes after buckling as a
function of force obtained from the supercoiling curves (taken in
part from previous measurements [16]) for Naþ concentrations
of 30, 60, 170, and 320 mM (gray, red, black, and blue circles,
respectively) together with the prediction from the theoretical
model (solid lines) calculated for a charge adaptation factor CR
of 0.42. Inset: Slopes for 60 mM Naþ (red circles) together with
theoretical predictions for different values of CR (lines).
slope of the supercoiling curves when neglecting fluctua-
tions. Eforce
pot is then given by
pot ¼ FdLDNA =dN:
Eforce
The bending energy per turn can be written as [5]
Estat ¼
EDNA kB TlB 2 ds; (4)
2 3 4 2 dN s rðsÞ
with  ¼ CR Rod PB and lB ¼ 0:7 nm being the
Bjerrum length in water. rðsÞ denotes the distance between
a point located at position s of the center line of one double
strand and a fixed point located at the center line of the
4 other double strand of the superhelix. Numeric integration
is performed over the whole center line s of the former
double strand. This simplified electrostatic potential yields
values similar to the full numerical solution of the PB
equation (see Fig. S3 [17]).
By combining Eqs. (1)–(4) we now can calculate Esh tot ,
and by minimization with respect to  and h one obtains
the slope dLDNA =dN and the DNA torque
¼ Esh tot =2 for
the energetically favored superhelix configuration.
We compared the predictions of our model with the
experimentally obtained slopes of our supercoiling curves
for forces between 0.25 and 4 pN [Fig. 1(b)] in buffers
containing 30–320 mM Naþ . Remarkably, the slopes at all
applied forces and ion concentrations are accurately
described [Fig. 1(c)]. However, achieving the agreement
required a substantial reduction of the DNA charge by
employing CR ¼ 0:42, independent of the salt concentra-
tion (see Fig. S4 [17]). Our theoretical model accurately
describes the slopes of recently available, independently
measured supercoiling curves [14] [Fig. 2(a)]. Since the
(2)
a b 30
70 50 mM
Torque (pN nm)

100 mM
Slope (nm)

200 mM

2
60
500 mM 20
dLDNA 1 
¼
50
EDNA pkB T ; (3)
bend
dN 2 2 þ ðh=2Þ2 40
10
30
where p ¼ 50 nm denotes the bending persistence length,
20 0
kB the Boltzmann constant, and T the absolute tempera- 0 1 2 3 4 0 1 2 3 4
Force (pN) Force (pN)
ture. The term within square brackets describes the DNA
curvature. We calculate the electrostatic interaction energy FIG. 2 (color). Comparison of the predictions from the theo-
between the two DNA double strands of the superhelix as retical model with data from Ref. [14]. (a) Slopes after buckling
previously described [22,23]. DNA is approximated as a versus force for different Naþ concentrations. Circles represent
cylindrical, uniformly charged rod with 1.2 nm radius experimental data and solid lines the theoretical prediction for
and linear charge density CR , where CR is the adjust- CR ¼ 0:42. (b) Torque after buckling as calculated from force-
able charge adaptation factor and  ¼ 2e=0:34 nm is the extension data [14]. Colors and symbols are as in (a).

158101-2
 week ending
PRL 105, 158101 (2010) PHYSICAL REVIEW LETTERS 8 OCTOBER 2010

force dependence of slope and torque are linked [24], our a 30 mM 320 mM

model also reproduces the torque in the superhelical phase

as obtained from force-extension data [14] [see Figs. 2(b) b c
and S5 [17]]. The obtained CR is found to be rather

Slope (nm)
600 70 0.55
70 CR 0.42

DNA length (nm)

insensitive to other parameters of the model; e.g., changing 50 0.32

Slope (nm)
60
the weakly salt-dependent persistence length by 5 nm alters 400 30
50
CR only by 0:03. 0 1 2 3
Force (pN)
4

40
To exclude the possibility that the small effective charge
200 30
is an artifact due to neglected fluctuations, we carried out CR= 0.42
coarse-grained Monte Carlo simulations (Fig. 3) [25]. The 0 5 10 15
20
0 1 2 3 4
DNA is modeled as a chain of small cylindrical segments, Turns Force (pN)

with equivalent terms describing the potential, bending, d e

Torque plateau (pN nm)

and electrostatic energies as in the theoretical model [17]. 30 30

Torque (pN nm)

Additional terms account for the twisting and stretching
energy [6]. By performing simulations with different num- 20 20

bers of added turns, we obtained supercoiling curves very

similar to those observed experimentally [Fig. 3(b)].
By applying CR ¼ 0:42, the slopes from simulation, the-
ory, and experiments are in excellent agreement over the
entire range of applied forces and ionic strengths [Fig. 3(c)].
This provides strong, independent support for a small
effective DNA charge. Comparing torque values from the
simulations and theoretical model, we find good overall
agreement, except the values from simulations appear
globally 1.5 pN nm higher [Fig. 3(e)]. We attribute this
deviation to DNA fluctuations which are neglected in the
model (see Figs. S6–S8 [17]).
In order to obtain a microscopic verification for CR , all-
atom molecular dynamics simulations [26] were employed
to obtain the force between parallel DNA molecules at
different salt concentrations [17]. We obtained good agree-
ment with the force calculated according to Eq. (4) using
CR ¼ 0:42 and very poor agreement with CR ¼ 1:0
(Fig. 4). To test whether the value found for CR is specific
to DNA-DNA interactions or is a universal constant for
DNA electrostatics, we obtained the ion distributions
around isolated double-stranded DNA molecules. The ra-
dial ion distribution extended further from the DNA than
expected from PB theory with CR ¼ 0:42 and approached
the distribution for CR ¼ 1:0 at low ionic strength [see
Figs. S10 and S11(a) [17]]. Thus, the small value for CR is
specific to DNA-DNA interactions. At elevated ion con-
centrations, the mean electrostatic potential around the
DNA was too weak to create the ion distribution as ob-
served in simulation and as predicted by PB theory [see
Figs. S11(b) and S11(c) [17]]. This suggests that ion dis-
tributions are not only determined by electrostatics, but
that other effects such as correlations in the ion clouds, the
noncontinuum nature of the dielectric surrounding, and ion
exclusion can have a significant influence. This in turn is
likely to cause the low DNA-DNA interaction forces.
Additionally, deviations from the homogeneously charged
rod model, such as strongly localized charges at the phos-
phates and counterions entering the DNA grooves, may
reduce the interaction forces.
10 10
0 0
0 5 10 15 0 1 2 3 4
Turns Force (pN)
FIG. 3 (color). Coarse-grained Monte Carlo simulations of
DNA supercoiling. (a) Snapshots of the formed plectoneme at
1 pN and 8 turns in the presence of 30 and 320 mM monovalent
ions. (b) Simulated DNA supercoiling curves for 170 mM Naþ at
stretching forces of 0.25, 0.5, 1.0, 2.0, 3.0, and 4.0 pN for CR ¼
0:42 [colors are as in Fig. 1(b)]. Simulations reproduce the
experimentally observed abrupt buckling at the onset of the
plectonemic phase, which is accompanied by a torque overshoot
[see (d)]. Buckling is followed by a linear DNA length decrease
with added turns at constant torque. (c) Slopes after buckling
obtained from the simulated curves for Naþ concentrations of
30, 60, 170, and 320 mM [circles, colors as in Fig. 1(b)] and
corresponding predictions from the theoretical model for CR of
0.42 (solid lines). Inset: Slopes from simulated curves at 60 mM
Naþ for CR ¼ 0:42 (red dots) as well as for CR ¼ 0:32 and
CR ¼ 0:55 (gray dots with dashed lines). The prediction for
CR ¼ 0:42 is shown as a solid red line. (d) Torque during DNA
supercoiling for the curves shown in (b). (e) Torque after buck-
ling as obtained from the simulations (filled circles) and after
subtraction of 1.5 pN nm (open squares) together with the
corresponding predictions from the theoretical model for CR
of 0.42 (solid lines). Naþ concentrations and colors are as in (c).
By combining single-molecule experiments, theoretical
considerations, and coarse-grained and all-atom simula-
tions, we have shown that, within a cylinder approximation,
DNA-DNA interactions can be described only by a signifi-
cantly reduced DNA charge. Furthermore, we have pro-
vided a theory that accurately describes DNA supercoils
over a broad range of tension and ionic strength. Super-
coiling under tension is a unique way to confine and bring
two DNA molecules into close proximity in the absence of
interfering surfaces [8]. In contrast to previous topological
investigations of long DNA [6,12,13], our force-based
experiments allow a more reliable quantification of
DNA-DNA interactions since much smaller distances are

158101-3

PRL 105, 158101 (2010) PHYSICAL REVIEW LETTERS
a *aksiment@illinois.edu
b (2001).
60 mM
Force (pN / DNA turn)
20
170 mM
300 mM
15
10
5 780 (1998).
0 (1977).
2.5 3.0 3.5 4.0 4.5
DNA-DNA separation (nm)
FIG. 4 (color). All-atom molecular dynamics simulations of
the effective force between double-stranded DNA. (a) All-atom
model. The DNA atoms are depicted as red spheres, counter- and
coions as blue and purple spheres, respectively, and water as a
semitransparent molecular surface. The distance between the
DNA molecules was restrained by a harmonic potential, sche-
matically depicted as a spring. (b) Mean force between the DNA
molecules from molecular dynamics simulations (circles, solid
lines) alongside the theoretical predictions for CR ¼ 0:42
(dashed lines) and 1.0 (dotted lines) at 60 (red) and 170 mM
(black) bulk ion concentrations. Data for 300 mM (blue) are
reproduced from previous work [26].
achieved (see Figs. S4 and S7 and discussion [17]). The
surprisingly small DNA-DNA interactions result from a
complex interplay of a highly charged and structured mole-
cule with solvent molecules and ions. The simple effective
interaction potential will be an important contribution for
quantitative models of complex biomolecular systems
which cannot be treated with atomic detail such as DNA
packaging in chromatin and viruses but possibly also for
protein-DNA interactions. The findings reveal that particu-
lar caution is necessary when applying effective charge
parameters obtained from experiments that probe a differ-
ent physics, such as electrophoresis [22], resulting from a
complex interplay between hydrodynamics and electrostat-
ics [8,9,27].
We gratefully acknowledge K. Kroy, U. Keyser, and M.
Emanuel for helpful discussions. This work was supported
by Grants No. SE 1646/1-1 and No. SE 1646/2-1 from the
DFG to R. Seidel, the National Institutes of Health (R01-
HG003713 and PHS 5 P41-RR05969), the National Science
Foundation (PHY0822613), the Petroleum Research Fund
(48352-G6), and the TeraGrid (MCA05S028) to A. A. as
well as by Project No. mvb00007 of the North German
Supercomputing Alliance (HLRN).
158101-4
week ending
8 OCTOBER 2010
†
gero.wedemann@fh-stralsund.de
‡
ralf.seidel@biotec.tu-dresden.de
[1] D. Marenduzzo, E. Orlandini, A. Stasiak, D. W. Sumners,
L. Tubiana, and C. Micheletti, Proc. Natl. Acad. Sci.
U.S.A. 106, 22 269 (2009).
[2] A. A. Kornyshev and S. Leikin, Phys. Rev. Lett. 86, 3666
[3] Y. Kao-Huang, A. Revzin, A. P. Butler, P. O’Conner, D. W.
Noble, and P. H. von Hippel, Proc. Natl. Acad. Sci. U.S.A.
74, 4228 (1977).
[4] J. Ubbink and T. Odijk, Biophys. J. 76, 2502 (1999).
[5] J. F. Marko and E. D. Siggia, Phys. Rev. E 52, 2912 (1995).
[6] K. Klenin, H. Merlitz, and J. Langowski, Biophys. J. 74,
[7] J. A. Schellman and D. Stigter, Biopolymers 16, 1415
[8] S. van Dorp, U. F. Keyser, N. H. Dekker, C. Dekker, and
S. G. Lemay, Nature Phys. 5, 347 (2009).
[9] B. Luan and A. Aksimentiev, Phys. Rev. E 78, 021912
(2008).
[10] D. C. Rau and V. A. Parsegian, Biophys. J. 61, 246
(1992).
[11] X. Qiu, L. W. Kwok, H. Y. Park, J. S. Lamb, K. Andresen,
and L. Pollack, Phys. Rev. Lett. 96, 138101 (2006).
[12] A. Vologodskii and N. Cozzarelli, Biopolymers 35, 289
(1995).
[13] M. Hammermann, C. Steinmaier, H. Merlitz, U. Kapp, W.
Waldeck, G. Chirico, and J. Langowski, Biophys. J. 73,
2674 (1997).
[14] F. Mosconi, J. F. Allemand, D. Bensimon, and V.
Croquette, Phys. Rev. Lett. 102, 078301 (2009).
[15] D. Klaue and R. Seidel, Phys. Rev. Lett. 102, 028302
(2009).
[16] H. Brutzer, N. Luzzietti, D. Klaue, and R. Seidel, Biophys.
J. 98, 1267 (2010).
[17] See supplementary material at http://link.aps.org/
supplemental/10.1103/PhysRevLett.105.158101 for fig-
ures, discussion, and methods.
[18] T. R. Strick, M. N. Dessinges, G. Charvin, N. H. Dekker,
J. F. Allemand, D. Bensimon, and V. Croquette, Rep. Prog.
Phys. 66, 1 (2003).
[19] S. Forth, C. Deufel, M. Y. Sheinin, B. Daniels, J. P. Sethna,
and M. D. Wang, Phys. Rev. Lett. 100, 148301 (2008).
[20] J. F. Marko, Phys. Rev. E 76, 021926 (2007).
[21] N. Clauvelin, B. Audoly, and S. Neukirch, Biophys. J. 96,
3716 (2009).
[22] D. Stigter, Biopolymers 16, 1435 (1977).
[23] S. L. Brenner and V. A. Parsegian, Biophys. J. 14, 327
(1974).
[24] H. Zhang and J. F. Marko, Phys. Rev. E 77, 031916
(2008).
[25] A. V. Vologodskii and J. F. Marko, Biophys. J. 73, 123
(1997).
[26] B. Luan and A. Aksimentiev, J. Am. Chem. Soc. 130,
15 754 (2008).
[27] B. Luan and A. Aksimentiev, Soft Matter 6, 243
(2010).