Modelling the self-assembly of DNA multi-arm motifs

Lead Research Organisation: University of Oxford
Department Name: Oxford Chemistry


DNA is famous for its double helical structure, in which two strands wide around each other held together by the very specific interactions between the DNA bases that was first discovered by Watson and Crick, namely that adenine (A) bonds with thymine (T) and cytosine (C) bonds with guanine (G). The two strands are said to be complementary, because at each nucleotide the bases pair in the Watson-Crick fashion (A-T, and C-G). However, the specificity of the base pairing allows other structures other than one single double helix to be programmed into a set of DNA strands. For example, if you have 3 types of strands, say P, Q and R, where the first half of P is complement to half of Q, and the second half is complementary to half of R, and the remaining halves of Q and R are complementary, the system will want to form a structure with three double helical sections meeting at a junction. This basic approach can be scaled up to make a whole array of precisely-ordered structures on the nanometer length scale that assemble themselves driven by the desire of complementary stretches of DNA molecules to pair up. This field is called DNA nanotechnology, and in this proposal we wish to study nanostructures called DNA multi-arm motifs. The 3-arm example of this class of structures is somewhat similar to the junction mentioned above, except that each arm is made up of two parallel double helices. By designing these motifs so that there are short single-stranded overhangs at the each of arm that are complementary to the overhangs at the ends of other arms, these motifs can be made to form a whole variety of larger structures. For example, the 3-arm motifs can form tetrahedra, dodecahedra and buckyballs (all polyhedra where 3 edges meet at a vertex) depending on the precise design and the conditions under which they assemble, In experiments, the researchers are able to see if their designed sequences successfully assemble into the target structure, but cannot see how they manage to achieve this. This is where computer modelling and this proposal fits in. We want to use techniques that we recently developed to model DNA to visualize the mechanisms by which these remarkable molecules are able to so reliably self-assemble. The insights that we hope to obtain will be of great help to experimentalists, as it will give them a better idea of why some of their designs are successful whereas others fail to do what is expected. This in turn will help them to make the design procedure more rational and hence more likely to succeed. Hopefully, this will enable the experimentalists to significantly increase the types of structures they can produce.

Planned Impact

DNA nanotechnology is one of the most rapidly developing areas of nanotechnology, with the level of control and addressability one can achieve being unprecedented. Although, so far, activity in this area has almost completely been performed within academic institutions, this is likely to change significantly in the next decade. One of its particularly attractive features for commercial applications is the biocompatibility, and the way the structures can potentially respond to environmental cues.

For example, DNA polyhedra, such as those made from the DNA multi-arm motifs studied here, have obvious potential as capsules for drug delivery. The addressability of the structures allows one to attach almost whatever one would like (as long as it can be functionalized with a short single-stranded piece of DNA). Furthermore, the polyhedra could be designed to release or activate their cargo in response to a particular nucleic acid sequence. Specifically, this signal could be the presence of a particular messenger RNA transcript that could be indicative of a particular tissue-type or diseased state (e.g.\ cancer) of the cell. Although the current research is not directly exploring these potential applications, it is clear that an increased understanding of how to control the self-assembly and rationally design novel nanostructures, will significantly aid such a task.

Given the potential of the field of DNA nanotechnology, it is of strategic importance that the UK builds up expertise in this area. Together with the experimental group of Andrew Turberfield in Physics, Oxford is a world-leader in DNA nanotechnology, and the UK hub for this field. The proposed project will contribute to the building up of the theoretical side of this expertise, and help to make Oxford the world-leader in the simulation of DNA nanotechnology. Furthermore, this goal is very much in line with the EPSRC's strategic aims; it is notable that the ``control of self-assembly'' is one of the current topics on the EPSRC's physical sciences signposting list, and that assembly also features as one of the EPSRC's Chemical Sciences grand challenges. The strategic value of the proposal can also be appreciated by reference to the 2009 International Review of UK Chemistry Research, a key recommendation of which was for the `Integration of Computational Chemistry' and the `Need to enhance the participation of theory and computation, especially in areas that involve energy, materials and health applications'. This proposal is clearly participating in the development of novel materials, and we drew attention above to the possible health applications of this field.

The proposed project will contribute to our coarse-grained DNA simulation codes and utilities, which we plan to make publicly accessible through our web-site. As well as being useful to those who wish to model DNA nanotechnology, the model is also well-suited for studying certain structural changes in genomic DNA. To further enhance the impact of the model, we will also explore the possibility of incorporation of our model within a commercial molecular simulation code. In particular, the enhanced functionalities and the user-friendly interfaces interfaces within these packages may make the model more attractive to industrial and non-expert users.

DNA nanotechnology is an area of science that is ideally suited to public engagement, because it is very visually appealing and the principles underlying this technology are relatively easy to explain, relying less on detailed scientific knowledge and more on a natural sense of geometry. The PI is currently involved in local outreach events such as giving presentations on DNA nanotechnology at Oxfordshire schools. The PDRA would also be expected to be involved in such outreach, as well as events such as the Royal Society Summer exhibition.


10 25 50
Description The aims of the grant were to apply computer modelling to better understand DNA nanotechnology. DNA nanotechnology uses DNA as an engineering material to create structures and devices on the nanoscale. It is based on the selectivity of Watson-Crick base-pairing (a stretch of DNA will only bind to another piece of DNA if it has the correct sequence of matching complementary bases) and the ability to programme DNA (by choosing the order of the bases in a DNA strand) so that the DNA strands can self-assemble into the desired nanostructure.

The main aims of the grant were to investigate an archetypal class of polyhedral DNA nanostructures that are made from DNA tiles with multiple arms that can bind to each other, for example, four appropriately designed 3-arm tiles can assemble into a tetrahedron. One key property of these tiles that allows control of their behaviour is their flexibility. We have analysed the fundamental properties of the motif (a "bulge") that gives rise to this flexibility, the mechanism by which changes in this motif allows control of the self-assembly products (e.g. whether 3-arm tiles form tetrahedra or larger polyhedra) and how this influences the structure of the resulting polyhedra.

We have also collaborated with experimental groups. Firstly, to work out how a single-strand of DNA (rather than multiple strands in the examples above) can be designed to fold into a target structure (in this case, a square-based pyramid) with a robustness and speed that is similar to the folding of proteins into their native state. One particular interesting feature of the DNA structure is its highly-knotted nature. Secondly, we have been examining the behaviour of a DNA walker, and have shown the mechanism by which the step length of the walker is physically constrained.
Exploitation Route The basic motivation for the project is that if one has a better fundamental understanding of the processes relevant to DNA nanotechnology, it aids the rational design of more complex future DNA nanodevices. Thus, we hope our work will feed into future applications for DNA nanotechnology, such as for intelligent drug delivery vehicles or designable nanopores for interrogating biomolecules (e.g. sequencing DNA). The increasing interaction that we are having with experimental groups to work together on the latest developments in DNA nanotechnology is testament to the role that computer modelling can play in such tasks.

Also, the current project has contributed to the further development of our coarse-grained DNA model and the associated simulation code. These are available from the oxDNA web-site and are currently being used by a significant number of groups around the world, and we envisage that this will continue to grow.
Sectors Pharmaceuticals and Medical Biotechnology

Title oxDNA 
Description The simulation code for our coarse-grained model of DNA, oxDNA, is available online for anyone to download and use. This grant has contributed in part to the development of this model and the associated simulation code. In particular, an improved version of the model (termed oxDNA2) was released in 2015. 
Type Of Material Improvements to research infrastructure 
Year Produced 2013 
Provided To Others? Yes  
Impact This code has been downloaded and used by many research groups around the world. For example, there are now over 80 publications that have used the oxDNA model (an up-to-date list is on the oxDNA web-site), and these involve over twenty non-Oxford groups who use the code. 
Description Optimizing DNA walkers 
Organisation Case Western Reserve University
Department Department of Chemistry
Country United States 
Sector Academic/University 
PI Contribution We have performed computer modelling of the DNA systems being studied experimentally by the group of Eyal Nir. A DNA walker is a synthetic molecular model whose aim is to efficiently and accurately "walk" along a track. There have been a number of potential designs proposed (often in very high profile publications) but most have high error rates. Before any applications of such nanodevices can be considered, their efficiency and reliability needs to be optimized. The group of Eyal Nir are performing careful experimental studies to understand the competing processes that lead to errors, and how best to minimize them. In our simulations, we are able to compute the rates of the relevant processes and how these change as the design parameters are varied. The group of Eyal Nir are also performing fundamental studies on the dynamics of DNA hairpin opening and closing (a hairpin is a single-stranded that folds back on itself as it has self-complementary sections). We are performing simulations on these systems to better understand the fundamental reasons for the variation of the rates with the features of the hairpins (e.g. stem length) and external parameters, such as salt concentration.
Collaborator Contribution Eyal Nir's group is performing the experimental studies on the DNA walker and hairpin systems outlined above.
Impact Two joint papers: 1) J. Chem. Phys. 142, 234901 (2015); title: Introducing improved structural properties and salt dependence into a coarse-grained model of DNA. In this paper the Nir group performed experiments on hairpins that allowed us to better parameterize our model of DNA. 2) Nucl. Acids Res. 46, 1553-1561 (2018); title: DNA bipedal motor walking dynamics: An experimental and theoretical study of the dependency on step size. This paper presented a combined experimental (Nir group) and simulation (us) study on the detailed dynamics of a key process in the stepping of a DNA walker. We are currently preparing a third paper on the the kinetics of DNA hairpin formation .
Start Year 2013
Description Single-stranded DNA folding 
Organisation National Institute of Chemistry, Slovenia
Country Slovenia 
Sector Academic/University 
PI Contribution Our group performed computer modelling of the system that the group of Roman Jerala were studying experimentally. They were interested in investigating how best to design a single DNA strand to fold into a target structure (in this case a square-based pyramid). We simulated the kinetics and folding pathway of particular designs. What we learned from the modelling, helped to guide them towards their final optimal design which was able to fold rapidly and robustly. Note, the potential for less than optimal designs to fold incorrectly is very high due to the highly knotted nature of the target. The above project has now finished.
Collaborator Contribution The group of Roman Jerala performed the experiments on the systems detailed above.
Impact A joint paper that recently appeared in Nature Communications (Nat. Commun. 7, 10803 (2016); title: Design principles for rapid folding of knotted DNA nanostructures)
Start Year 2014
Title oxDNA 
Description The simulation code for our coarse-grained model of DNA, oxDNA, is available online for anyone to download and use. This grant has contributed in part to the development of this model and the associated simulation code. In particular, an improved version of the model (termed oxDNA2) was released in 2015. 
Type Of Technology Software 
Year Produced 2013 
Open Source License? Yes  
Impact This code has been downloaded and used by many research groups around the world. For example, there are now over 80 publications that have used the oxDNA model (an up-to-date list is on the oxDNA web-site), and these involve over twenty non-Oxford groups who use the code. 
Description Interview for "Theoretically Speaking" podcast 
Form Of Engagement Activity A broadcast e.g. TV/radio/film/podcast (other than news/press)
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact I was interviewed about my work using oxDNA for the Theoretically Speaking podcast (@TheoryPod) which currently has over 400 followers (the podcast started in November 2017).
Year(s) Of Engagement Activity 2018