Internal Dynamics of Superhelical DNA Roman Shusterman, 1 Tatyana Gavrinyov, 1 and Oleg Krichevsky 1,2, * 1 Physics Department, Ben-Gurion University, Beer-Sheva 84105, Israel 2 Ilse Kats Center for Nanoscience, Ben-Gurion University, Beer-Sheva 84105, Israel (Received 26 July 2007; published 6 March 2008) We present the first data on the temporal kinetics of monomer mean square displacements in DNA circles with defined degrees of superhelicity. The segmental dynamics of specifically labeled DNA plasmids with superhelical densities between 0 and 0:016 was assessed by fluorescence correlation spectroscopy. Introduction of superhelicity leads to progressively faster dynamics in the long time regime corresponding to the coil diffusion as observed previously by Langowski et al. [Biopolymers 34, 639 (1994)], but also in the short time range corresponding to the segmental motion within the DNA coil. DOI: 10.1103/PhysRevLett.100.098102 PACS numbers: 87.14.G, 87.15.H, 87.15.Ya The conformational dynamics of polymers is deter- mined by general polymer features such as bending rigid- ity, entropic elasticity, and hydrodynamic interactions between polymer segments. Circularized DNA is unique among other polymers in possessing an additional major property: superhelicity, characterized by linking number difference Lk which can be pictured as an amount of whole twist turns introduced into relaxed DNA prior to circularization. A corresponding intensive parameter, superhelical density, is defined as Lk=Lk 0 , where Lk 0 is the number of helical repeats of a relaxed double- helix. Superhelicity has a major effect on the equilibrium conformations and mechanical properties of DNA, and should greatly influence DNA dynamics. The equilibrium and stress-response behavior of superhelical DNA has been studied by a variety of experimental methods [1– 4] and is well understood both in the framework of simulations and analytical theory [5 –7]. However, the effect of superhelicity on the dynamics of DNA circles is known in much less detail. Most of the experimental progress was achieved by Langowski and co- workers using dynamic light scattering (DLS) which sup- plied data on the kinetics of rotational and translational diffusion of whole DNA circles, as well as on the dynamics of the dominant mode of their internal fluctuations [8 –10]. The internal dynamics of DNA in this approach is charac- terized by a single kinetic parameter, the ‘‘internal diffu- sion coefficient’’ D i , and by the relative amplitude of internal fluctuations. These parameters were measured for a DNA with different superhelicities [8 –10]. With the analytical theory still in development [11–14], much of the current understanding of the dynamics of superhelical DNA comes from computer simulations [5,10,15] whose detailed predictions are verified on the more coarse- grained data from DLS experiments. Here we report the first data on the stochastic segmental dynamics of superhelical DNA obtained with an experi- mental approach based on the fluorescence correlation spectroscopy (FCS) [16,17]. Whereas DLS measures the dynamics of the longest collective modes of DNA coil fluctuations, FCS applied to specifically labeled DNA re- veals the dynamics of a single labeled position (essentially a ‘‘monomer’’) in a DNA polymer [18,19]. Here we pre- sent data on the temporal dependence of monomer mean square displacements (MSD) hr 2 ti for 2686 bp long DNA circles with defined superhelicities Lk in 0 to 4 range ( 0:016 < 0). In the FCS approach, DNA circles labeled specifically with a single fluorophore each move freely in solution illuminated with a focused laser beam. The stochastic motion of labeled DNA segments in the nonuniform exci- tation field results in the fluctuations It in fluorescence intensity It. Therefore the autocorrelation function GthI0Iti of emission fluctuations reflects the dynamics of base pair motion. For independent point sources of fluorescence randomly moving in a Gaussian beam, Gt is directly related to MSD of the fluorophore [20]: Gt G 0 1 2 3 hr 2 ti w 2 xy 1 1 2 3 hr 2 ti ! 2 w 2 xy 1=2 ; (1) where ! w xy =w z is the aspect ratio of the sampling volume, w xy and w z define the dimensions of the sampling volume in lateral and transversal dimensions, respectively, and G 0 is the amplitude of the correlation function deter- mined from the behavior of Gt at short time scales. Then with calibrated w xy and !, the Eq. (1) can be used to convert the collected Gt into segmental MSD in a wide range of time and length scales [18]. We use pUC18 plasmids tagged specifically through the covalent attachment of prelabeled triple-helix forming oli- gonucleotides (TFO) according to the procedure developed in [21] with minor adjustments [20]. Topoisomers with defined superhelicities were prepared and gel separated by standard methods [22,23] modified to avoid exposure of DNA samples to UV and intercalating dyes [20]. Typically, we were able to obtain topoisomers from 6 bands in quantities sufficient for experiments: the first two bands correspond to relaxed and nicked circles and PRL 100, 098102 (2008) PHYSICAL REVIEW LETTERS week ending 7 MARCH 2008 0031-9007= 08=100(9)=098102(4) 098102-1 2008 The American Physical Society