Mechanical transition in a highly stretched and torsionally constrained DNA.

Janusz Strzelecki1, ukasz Pepowski 1, Robert Lenartowski 2, Wiesaw Nowak 1,

Aleksander Balter 1
1

Institute of Physics, Nicolaus Copernicus University, Grudzidzka 5, 87-100 Toru,

Poland
2

Faculty of Biology and Environment Protection, Laboratory of Isotope and

Instrumental Analysis, Lwowska 1, 87-100 Toru, Poland

(Received
ABSTRACT
We show results of our high force (up to 1.8 nN) AFM force
spectroscopy measurements of a double stranded DNA. We have found that
the force spectra of torsionally constrained molecules display a small
plateau occurring at a force of approximately 1nN. This transition, not
reported before, is absent in molecules with rotational freedom. Based on
all-atom molecular dynamics simulations, we suggest that this plateau is a
result of reducing the diameter of a double helix through extreme
stretching. The simulation suggests that the molecule is forced into a form
resembling an underwound P-DNA, with bases protruding outside of
backbones. These results broaden our understanding of fundamental aspects
of DNA nanomechanics.
PACS numbers 87.15.La 87.14.gk 87.64.Dz

DNA is a key molecule of life and provides great hopes as a bionanotechnology

material [1]. Thus, since the very beginning of a single molecule force spectroscopy [2] this
all important biopolymer remains within focus of extensive nanomechanical research. DNA
was stretched and twisted in experiments involving optical/magnetic tweezers and atomic
force microscopy (AFM) [3-6]. Those experiments revealed that the DNA double helix can
be overstretched up to 70% of contour length when pulled. This transition manifests as a
force-specific plateau on a force - distance curve [7-8]. In majority of experiments the ends of
double helix may rotate along a long axis of DNA and then the plateau is observed at 65 pN
[9-10]. In torsionally constrained setting, when such rotation is not possible, the force is 110
pN [7-8, 11-12]. A second plateau, marking the full strands melting, was encountered at
approximately 300 pN [13-15]. Despite significant research effort, very few experiments
explored the DNA mechanics beyond 500pN limit. Such extreme deformation of double
stranded DNA (dsDNA) should be feasible, as single stranded DNA (ssDNA) was stretched
with forces as large as 2nN [13, 16]. To date, only stretching torsionally unconstrained DNA
with forces up to 1 nN was reported [13]. A more detailed study of DNA mechanics in this
high force regime, including stretching of torsionally constrained molecules is still missing.
This regime is biologically important. A recent research [17-18] suggests that some damage
acquired by genetic material during abnormal mitosis can result from extreme stretching and
rupturing of DNA when the chromosomes are pulled apart. Such damage can cause mutations
and trigger tumor initiation and progression. Additionally, extracellular DNA that strengthens
bacterial biofilm [19] is also exposed to significant deforming forces by the environment.
Knowing the behavior of double helix at its mechanical limits is also important in case of its
potential applications as a bionanotechnology building block [1].

In this paper, we report an experimental evaluation of dsDNA stretched up to nN

force range. We observed for the first time, to the best of our knowledge, a mechanical
transition exhibited as a third plateau in force spectra of torsionally constrained molecules. A
mechanism of the plateau origin supported by Steered Molecular Dynamics (SMD)
simulations is proposed.

A simple, and yet a very effective method of nonspecific molecule picking with an
untreated AFM tip allowed us to achieve a stretching force close to 2nN. We used the pUC18
plasmid (Fermentas, SD0051), linearized with the EcoRI enzyme (Fermentas, ER0271). To
check a possible influence of the DNA base pair sequence and a molecule length on the
observed effect, a lambda phage (Fermentas, SD0011) was studied as well using the same
protocol. SMD simulations have been performed using NAMD 2.8 code [20] and all-atom
CHARMM27 force field [21]. Simulations mimicked topologies of our AFM force
spectroscopy experiment - dragged atoms had constraints in all directions except the helical
axis, thus preventing rotation during stretching. A control set of simulations was also made,
with the pulling carbon atoms unconstrained and a free rotation around the helical axis [22].

In the majority of cases molecules picked with an AFM tip detached and were lost at
stretching forces below 0.5 nN. Nevertheless, patient molecule fishing allowed us to record
48 force curves (each resulting from different molecule) above this limit, with detachment
force beyond 1 nN (a maximum observed stretching force was 1.8 nN). Force curves obtained
showed characteristics either of torsionally constrained (rounded onset of overstretching
transition at 110pN, no hysteresis upon relaxation) or unconstrained DNA stretching (sharp
onset of overstretching at 65pN, significant hysteresis upon relaxation). Apart from the well
known overstretching and melting transitions also a third, never previously reported plateau

was present, but only if the stretched molecule was torsionally constrained (Fig. 1). This
feature appeared at very high extensions, when the molecule was stretched approximately 2.3
times its contour length and for forces approaching 1 nN. As can be seen in Fig. 1, this
plateau is very small when compared to the overstretching and melting transitions.

Force spectrograms of molecules that did not detach during stretching show that the
transition is reversible (Fig. 2(a)) and is present both in the stretching and relaxing curves
similarly to the overstretching and melting. This plateau was observed in each constrained
DNA force curve, obtained in independent experiments on different samples and in the
molecules of different length (Fig. 2(b)), provided the stretching force was large enough.
Repetitive stretching of the same molecule also marked its presence in each force
spectrogram (Fig. 3(a)). Additionally, when a portion of a stretched molecule detached and
was stretched again the transition appeared each time (Fig. 3(b)). This third plateau cannot
thus be considered just a minor detachment event or an instrumental artifact. In our opinion it
is an integral feature of DNA mechanics and reflects a forced conformational change within
the highly stretched double helix itself.

To directly compare data from different molecules we employed extension normalization

procedures [23] for 10 curves with molecular extension at 1100 pN (Fig. 4(a)). In order to
account for the influence of base pair sequence and a significant variation in the contour
length, additional measurements with the lambda phage DNA were made and force curves
were analyzed along with the curves obtained or pUC18 plasmid DNA. While analyzed
curves overlap perfectly within the overstretching and melting plateaus region, a significant
variation can be seen in the third plateau. It was observed for a certain range of forces (8201020 pN) rather than a specific value, as in the case of overstretching transition (Fig. 4(b)).

This force value does not seem to depend on the length of the stretched molecule. The
extension resulting from this transition is, on the other hand, clearly proportional to the
contour length (Fig. 4(a) inset). A rough estimate with a linear fit yields a 0.015 nm extension
per single base pair resulting from this plateau.

The third plateau never appeared when the stretched molecule was unconstrained.
This can be clearly seen on Fig 5, where normalization was made for force curves showing
characteristics of resulting from molecules with a free rotation. This implies that the origin of
this transition is inherently connected with the DNA torsional constrain.

In order to explain the mechanism behind this plateau we performed SMD simulations
for both constrained and unconstrained dsDNA. Those simulations show that when the DNA
cannot unwind, the stretched double helix reduces its diameter, forcing both strands into
contact. As a result, stacking is disrupted and bases start protruding outwards (Fig. 6). The
molecule adopts a shape similar to the previously reported P-DNA [24]. However, the
structure obtained through extreme stretching is underwound in comparison to the P-DNA
created by the extreme overtwisting with magnetic tweezers. Conversion of SMD force vs
extension curves relative to the contour length allowed for comparison with AFM force
spectroscopy data. We can see that the plateau coincides well with the transition to the
underwound P-DNA. This leads to conclusion that this origin of the third plateau is a sound
hypothesis.

SMD with a free rotation of stretched molecule leads to formation of the Zip-DNA
form, with an interchangeable base stacking, reported previously [25]. This explains the lack
of the third plateau in unconstrained molecules, as stretching induced helix diameter

reduction cannot take place in that case. Additionally, we can expect the force necessary to
cause the diameter reduction would depend on the arrangement of the DNA helix. On the
other hand, in AFM experiment with nonspecific molecule attachment a variation in pulling
angle or in separation of DNA strands is very likely. Those conditions can explain the fact
that the plateau did appear within a certain range of forces rather than for a specific one.

We did a rough estimate to test this model further. A height of a single helical pitch is
equal to

, with l being the helix length and r the helix radius. With 38 bp S-

DNA present after overstretching a helix length would be 26.6 nm. Based on AFM imaging
of stretched DNA [26] we can assume that before the transition the helix radius is
approximately 0.6 nm. Reducing the helix radius by 0.25 nm, which would correspond to the
average size (0.5 nm) of a single pyrimidine and purine base would result in elongation of
helical pitch by 0.18 nm and 0.0047 nm per single base pair. The extension of 0.015 nm per
single base pair obtained from the linear fit from the inset on Fig 4 is approximately three
times smaller.

However, experimental results indicate that 1/5 of overstretched torsionally

constrained DNA has P-DNA form [12], thus 1/5 of base pairs could not contribute to the
extension through the diameter reduction. Additionally, further bases protruding outside
backbones appear beyond overstretching transition [27] due to melting. Thus, when
comparing the extension per single base pair obtained from our experiment with a proposed
model, the actual value should not be calculated over all base pairs, but adjusted accordingly
at the very least by a factor of 0.8. Considering this, we can find a good agreement between
our results and this simple geometrical estimate.

In summary, we report for the first time the presence of a mechanical transition in a
highly stretched DNA with a torsional constraint. Based on SMD simulations we postulate

that the plateau observed in force spectra is caused by the double helix diameter reduction of
the strained molecule, leading to formation of structure similar to the underwound P-DNA.
Our results give a new insight into the understanding of DNA mechanics. The DNA proves to
be a remarkably resilient molecule. As long as the double helix does not unwind, it is not
completely denatured until a force approaching 1 nN is reached. Even after such extreme
stretching, this molecule may easily return to its relaxed conformation. Additionally, if the
hypothesis of the forced transition to an underwound P-DNA is correct, then a new
mechanism, besidest the full separation of strands or overtwisting, to expose nucleobases
seems to be possible. Thus, strong mechanical deformation could serve as a means to provide
an access to the genetic sequence. A similar mechanism of opening binding sites in globular
proteins through the mechanical unfolding has been postulated [28].

ACKNOWLEDGEMENTS
We thank Prof. Piotr E. Marszalek for stimulating our work in single molecule force
spectroscopy and Prof. Richard Lavery for DNA pdb files.
Janusz Strzelecki acknowledges grant Krok w przyszo (I and III edition) from Marszaek
of Kujawsko-Pomorskie Voivodeship and EU.

[1]

A. V. Pinheiro, D. Han, W. M. Shih, and H. Yan, Nature nanotechnology (2011).

[2]

S. B. Smith, L. Finzi, and C. Bustamante, Science 258, 1122 (1992).

[3]

C. Bustamante, S. B. Smith, J. Liphardt, and D. Smith, Current Opinion in Structural

Biology 10, 279 (2000).

[4]

C. Bustamante, Z. Bryant, and S. B. Smith, Nature (London), 423 (2003).

[5]

J. F. Marko and S. Cocco, Phys. World 16, 37 (2003).

[6]

C. Prvost, M. Takahashi, and R. Lavery, ChemPhysChem 10, 1399 (2009).

7

[7]

S. B. Smith, Y. Cui, and C. Bustamante, Science 271, 795 (1996).

[8]

P. Cluzel, A. Lebrun, C. Heller, R. Lavery, J. L. Viovy, D. Chatenay, and F. Caron,

Science 271, 792 (1996).

[9]

N. Bosaeus, A. H. El-Sagheer, T. Brown, S. B. Smith, B. kerman, C. Bustamante,

and B. Nordn, Proceedings of the National Academy of Sciences 109, 15179 (2012).

[10]

X. Zhang, H. Chen, S. Le, I. Rouzina, P. S. Doyle, and J. Yan, Proceedings of the

National Academy of Sciences 110, 3865 (2013).

[11]

D. H. Paik and T. T. Perkins, Journal of the American Chemical Society (2011).

[12]

J. Leger, G. Romano, A. Sarkar, J. Robert, L. Bourdieu, D. Chatenay, and J. Marko,

Physical review letters 83, 1066 (1999).

[13]

M. Rief, H. Clausen-Schaumann, and H. E. Gaub, Nature Structural Biology 6, 346

(1999).

[14]

H. Clausen-Schaumann, M. Rief, C. Tolksdorf, and H. E. Gaub, Biophysical Journal

78, 1997 (2000).

[15]

C. P. Calderon, W. H. Chen, K. J. Lin, N. C. Harris, and C. H. Kiang, Journal of

Physics: Condensed Matter 21, 034114 (2009).

[16]

T. Hugel, M. Rief, M. Seitz, H. E. Gaub, and R. R. Netz, Physical Review Letters 94,
48301 (2005).

[17]

A. A. Guerrero, M. C. Gamero, V. Trachana, A. Ftterer, C. Pacios-Bras, N. P. DazConcha, J. C. Cigudosa, C. Martnez-A, and K. H. M. Van Wely, Proceedings of the
National Academy of Sciences 107, 4159 (2010).

[18]

N. J. Ganem and D. Pellman, The Journal of Cell Biology 199, 871 (2012).

[19]

W. Hu et al., PloS one 7, e51905 (2012).

[20]

J. C. Phillips et al., Journal of Computational Chemistry 26, 1781 (2005).

[21]

A. D. MacKerell Jr et al., The Journal of Physical Chemistry B 102, 3586 (1998).

[22]

See Supplemental Material for the AFM force spectroscopy experiment procedure and
SMD simulations details.

[23]

M. Rabbi and P. E. Marszalek, Cold Spring Harbor Protocols 2007, pdb. prot4900
(2007).

[24]

J. Allemand, D. Bensimon, R. Lavery, and V. Croquette, Proceedings of the National

Academy of Sciences 95, 14152 (1998).

[25]

A. Balaeff, S. L. Craig, and D. N. Beratan, The Journal of Physical Chemistry A

(2011).

[26]

M. Maaloum, A. Beker, and P. Muller, Physical Review E 83, 031903 (2011).

8

[27]

J. Van Mameren, P. Gross, G. Farge, P. Hooijman, M. Modesti, M. Falkenberg, G. J.

L. Wuite, and E. J. G. Peterman, Proceedings of the National Academy of Sciences
106, 18231 (2009).

[28]

R. Perez-Jimenez, A. P. Wiita, D. Rodriguez-Larrea, P. Kosuri, J. A. Gavira, J. M.

Sanchez-Ruiz, and J. M. Fernandez, Journal of Biological Chemistry 283, 27121
(2008).

FIGURE CAPTIONS

FIG. 1. Stretching non-specifically attached dsDNA using AFM force spectroscopy.

Untreated AFM tip was used to pick and stretch DNA molecules with a force reaching almost
2 nN. Representative force curves obtained for molecules of similar length but different in
torsional arrangement are shown. A small third plateau, absent in a torsionally relaxed DNA
is visible in a constrained molecule.

FIG. 2. A third plateau observed in a highly stretched constrained DNA. The transition
underlying this event is reversible, as it appears both during the stretching (upper black) and
the relaxation (lower red) of a single DNA molecule (a). A magnified fragment of a force
curve, showing the third plateau in detail is displayed in the inset. This plateau was observed
each time a force exceeding 800 pN was achieved, as can be seen in three representative
curves obtained for different molecules on different samples, with significant variation in
contour length (b).

FIG. 3 The third plateau is present in repeatedly stretched DNA molecule. The plateau
was observed in a series of multiple stretching of the same DNA molecule (a). Two curves
(upper black stretching, lower red relaxation) that show this transition both before and after a
minor detachment of a stretched molecule are also shown (b).

FIG. 4 Normalization of 10 force curves showing the third plateau and a histogram of
transition forces. Curves were normalized for the force of 1100 pN [9 pUC18 (black) and 1
lambda phage (bright green)]. A full overlap can be seen in overstretching and melting
transition areas (a). A dependence of plateau length on the molecule contour length

10

[expressed in number of base pairs (bp)] is shown in the inset [pUC18 (black square) and 1
lambda phage (bright green diamonds)]. A linear fit to this data (red solid line) gives an
estimate of 0.015 nm increase in length per base pair due to transition. This plateau appears
for a certain range of forces (820 -1020 pN) (b) instead of a specific one. A molecule length
does not seem to influence the value of those forces.

FIG. 5. Lack of the third plateau in torsionally relaxed DNA molecules. The figure shows
a set of 6 curves resulting from the stretching of unconstrained molecules, with the extension
normalized for force of 1100 pN. The plateau which was present in the constrained DNA is
not observed here.

FIG. 6. SMD simulations of highly stretched DNA show a possible mechanism of the
transition. The figure shows curves acquired from two 10 ns (grey) and one 100 ns (pink)
SMD simulations of poly(dA-dT) poly(dA-dT) and two 10 ns (green) and 100ns (blue)
poly(dG-dC) poly(dG-dC). The rotation of molecule during stretching was prevented. An
experimental force curve (black) is included as a reference. The extension is calculated in
reference to the initial B-DNA form length. Snapshots from different phases of the
simulation, marked by orange dots are also shown. A constrained stretched molecule is
forced to reduce its diameter as the extension increases and the base pairing and stacking is
disrupted. At the extension coincident with the plateau a structure resembling an underwound
P-DNA is formed, with base pairs protruding outside of backbones. The transition can thus be
attributed to the transition into this new form.

11

FIG. 1
Strzelecki
12

FIG. 2
Strzelecki

13

FIG. 3
Strzelecki

14

FIG. 4
Strzelecki

15

FIG. 5
Strzelecki

16

FIG. 6
Strzelecki
17