DNA is a large and flexible polymer that has been selected by nature to transmit information. Its most common three-dimensional structure is represented by a double helix, that has been resolved for the first time in 1953 , but since then many different 3D structures have been described for DNA.
DNA is extremely flexible and polymorphic and can easily change its conformation to adapt to different interactions . Moreover it can also reach different topologies giving rise for instance to supercoiled DNA, formed because of the limited free rotation of the DNA domain flanking a replication or transcription complex. The importance of “unusual” or transient DNA structures is growing since recent studies of DNA topology, DNA supercoiling, DNA knotting and linking have shown that DNA must be considered an active molecule and not simply a passive substrate of “active” enzymes. In fact DNA topology-induced structural and geometric changes can drive, or at least strongly influence, the interactions between protein and DNA and so direct its own metabolism . On the other hand the unique self-recognition properties of DNA, determined by the strict rules of Watson-Crick base pairing, makes this material ideal for the creation of self-assembling predesigned nanostructures in a bottom-up approach. The construction of such structures is one of the main focuses of the thriving area of DNA nanotechnology, where several assembly strategies have been employed to build increasingly complex three-dimensional (3D) or two-dimensional (2D) DNA nanostructures.
A common limit involving both natural and artificial DNA structures concerns the difficulty to provide an atomistic description. Moreover the knowledge of the degree of freedom and of the accessible conformational space of these nanostructures remains limited. Structural data, primarily from X-ray diffraction studies, is sparse in comparison to the manifold configurations possible, and direct experimental examinations of DNA's flexibility still suffer from many limitations. In order to face these shortcomings, molecular dynamics (MD) is becoming an essential tool since it provides a detailed structural and dynamical description of the analyzed nanostructures [7,8]. However, given the speed with which MD studies of DNA have spread, the roots remain somewhat shallow: in many cases, there is a lack of deep knowledge about the foundations, strengths, and limits of the technique.
This workshop wants to put together researchers working on natural or artificial DNA nanostructures to discuss together how MD simulation can provide information useful to describe the structural-dynamical properties of these structures and at the same time to plan new ones. At the same time we want to discuss the limit of the techniques and how the simulative and experimental approaches can intersect one with the other to permit a detailed description of known and future DNA nanostructures.
The unique self-recognition properties of DNA, its high thermodynamic stability, and the ease by which it can be synthesized have in combination made this molecule one of the most efficient building blocks for the creation of predesigned self-assembling nano-structures. Consequently a large variety of precisely programmed two- or three-dimensional (2D or 3D) DNA nano structures have been presented during the past decade. Despite the success and the fast growing of this area of research several problems must still be solved since an user-defined artificial assembly of the DNA polymer requires a deep knowledge of the DNA itself that can be only obtained through a strict interconnection between experimental and simulative scientific communities working on the characterization of both natural and artificial DNA structures. The assembly of DNA nanostructures display similarities with protein folding in spite of chemical and structural differences. It has been in fact shown that at constant temperature, hundreds of DNA strands can cooperatively fold within minutes into complex nanoscale objects along nucleation-driven pathways influenced by the choice of sequences, strand lengths, and chain topology (Dietz et al., 2012). Folding occurs out of equilibrium, whilst unfolding occurs in apparent equilibrium at higher temperatures than those for folding. DNA strand can actually be folded in many different shapes that can be also able of entering live mammalian cells(Walsh, Yin et al. 2011) or to penetrate lipid bilayer mimicking the response and function observed in natural ion channels. In detail the Simmel and Dietz groups have been able to use a scaffolded DNA origami to create a stem that penetrated and spanned a lipid membrane, as well as a barrel-shaped cap that adhered to the membrane, in part via 26 cholesterol moieties. In single-channel electrophysiological measurements, they found similarities to the response of natural ion channels, such as conductances on the order of 1 nanosiemens and channel gating (Simmel Science 2012). A similar a stable DNA-based nanopore that structurally mimics the amphiphilic nature of protein pores and inserts into bilayers to support a steady transmembrane flow of ions has been recently achieved using a different strategy. In this case an outer hydrophobic belt comprised of small chemical alkyl groups which mask the negatively charged oligonucleotide backbone has been introduced. This modification overcomes the otherwise inherent energetic mismatch to the hydrophobic environment of the membrane ( howorka Small 2013). These two works are the first examples showing the ability of DNA to assemble in a hydrophobic environment and open up the design of entirely new molecular devices for a broad range of applications including sensing, electric circuits, catalysis, and research into nanofluidics and controlled transmembrane transport. A common limit of the up to now created DNA nanostructures is that their 3D description can be obtained by low-resolution techniques such as AFM, Cryo-EM and small angle x-ray spectroscopy (SAXS) ( oliveira 2010 Simmel 2012). Despite this limitation such analysis have shown the presence of topologies that are not observed in short DNA strands indicating that the assembly in nanostructure give rise to curvutures and motif not observed in short classical DNA strands ( Simmel PNAS 2012). Similarly the usual biological information that comes from the study of the problem concerning the protein-DNA interaction is usually limited to the description of the static picture coming from the crystal structure of a protein in presence of a short DNA segment but DNA in a cell is not an inert like linear B-form. Instead DNA in cells has topology and topology affects curvature, twist, kinking, base flipping, denaturation adding further complexity to the system ( Fogg Jm 2012 Q. Rev. Biophys.).
An help in the comprehension of such a large degree of complexity may come from MD simulation that due to the strong improvement in the core computational system and on the force-field quality has permitted from one side to simulate large DNA nanostructure and from the other to carry out very long simulation on DNA helices or DNA motifs of limited size. Comparison of the simulated and experimentally determined structures has brought to the development of the parmbsc0 force field, a refinement of the AMBER parm99 force field. This force field has been derived by fitting to high-level quantum mechanical data and verified by a very extensive comparison between simulations and experimental data (Perez et al., 2007a) and has been also used for microsecond molecular dynamics simulations of a DNA duplex, where many relevant local transitions and sequence-dependent transitions coupled to partial and total base pair openings have been observed, demonstrating how long simulation can highlight features otherwise not detectable (Perez et al., 2007b). The importance of the sequence in inducing bending has been also shown on a series of simulations at various temperature and salt conditions of DNA A-tracts, The simulations indicate that the characteristic local conformational features and the global bending magnitudes observed in the simulation agree with the electrophoresis data, but not completely with NMR data (Lankas et al., 2010). The large DNA flexibility has been also evidenced by massive DNA simulations combined to X-ray analysis providing evidence of the existence of binormality in many dinucleotide pair steps, and of more intriguing bimodality in the distribution of a small numbers of helical parameters at given dinucleotide pair steps (Dans et al, 2012). The importance of long MD simulations has been also exploited to investigate the interactions of the TTA loop with the G-quartets an important DNA structural motif showing that structural dynamics on a microsecond timescale differs dramatically from previous shorter simulations. The analysis provides a detailed atomistic account of the dynamic nature of the TTA loops, highlighting their interactions with the G-quartets including formation of an A:A base pair, triad, pentad and hexad (Islam et al., 2013).
At the same time atomistic simulations of natural supercoiled DNA, that represent an usual topology for DNA in cells has shown that a wealth of non-canonical DNA structures such as kinks, denaturation bubbles and wrinkled conformations form in response to DNA bending and torsional stress (Mitchell et al., 2011, Liverpool tb 2008). Similarly simulation of an artificial DNA nanostructure has permitted to show unusual curvature and stacking that are not detected in the simulation of short DNA strand ( Falconi 2008).
All these data indicate that MD simulation is strongly contributing in describing the structural-dynamical features of “unusual” DNA structures but that a common effort with experimentalists must be done to optimize the approach and to define the systems that must be taken into account.