emDNA – A Tool for Modeling Protein-decorated DNA Loops and Minicircles at the Base-pair Step Level

Understanding nucleic acid structure provides essential insight into functional roles fundamental to molecular biology. Unfortunately, experimentally determining DNA structure at high resolution is tedious, expensive and not always tractable, particularly for large complex systems, which can show significant molecular flexibility. There has been extensive research into developing computational tools for modeling and manipulating nucleic acid structure as a complementary approach to study nucleic acid structure in silico. It is well known from biophysical studies how proteins and other molecules interact with double-helical DNA in its immediate context, but less is known about the larger-scale structural interactions that occur in protein-decorated loops and minicircles. Indeed, minicircle DNAs are novel non-viral DNA vectors that have become promising options for gene therapy, DNA vaccines, or intermediates in cell-based therapies. Current computer models of DNA range from coarse-grained simulations that forgo fine-grained physical and chemical information to tackle larger-scale systems to precise all-atom molecular dynamics investigations that yield time and spatially dependent portrayals of small DNA fragments.

In a new study published in the Journal of Molecular Biology, investigators Robert Young, Nicolas Clauvelin and led by Professor Wilma Olson at Rutgers University developed a new tool for generating optimized structures of DNA at thermal equilibrium with built-in or user-generated elastic models. This new tool is called emDNA, a command-line software application that enables users to create spatially constrained, energy-optimized DNA models. The new tool enables users of any ability level to create and explore mesoscale models of their own design in conjunction with the case studies presented in the article. Applications of emDNA to a well-characterized minicircle with distinctive sequence-dependent properties were highlighted in the reported case studies. The emDNA software is available to the general public on GitHub (https://nicocvn.github.io/emDNA/).

In this case, treating DNA at the base-pair level with rigid-body characteristics makes it possible to create models with hundreds of base pairs of length from small, sequence-specific features discovered via experimentation. This framework is utilized by the emDNA software to create optimal DNA structures at thermal equilibrium with pre-built or user-generated elastic models. Researchers now have a new method for studying DNA functionality thanks to the emDNA program. During the energy minimization of polymeric structures, users can regulate the chain length, sequence composition, location of the fragment ends, and range of local structural deformations. The companion tool, emDNA_probind, offers the capability to optimize DNA structures containing one or more bound proteins through a dedicated minimization procedure. These emDNA features provide helpful insights into how local regions of interest interact with the overall structure of the DNA, such as how tailored ligands may modify looping propensities and related genetic activity. The overall DNA configuration can be changed by selecting intrinsic features at the base-pair step level. Although emDNA was designed to be user-friendly, there are some limitations to consider. Depending on the size of the DNA fragment and the quantity of ramped models, the optimization and protein-decoration ramping procedures may require higher computing resources.

For example, optimizing a 100-bp design may take less than two minutes, In contrast a 1000-bp configuration may take days, depending on the computational resources available. This can be a concern when utilizing the emDNA_probind tool since the uptake of certain proteins involves a considerable number of optimization steps. The authors are going to address broader functionality and improved usability in the following stages of emDNA software development. This will offer users the choice of working within the framework of the 256 possible tetramer steps or the 16 possible dimer steps by including an expanded base-pair step option. Their initial analyses of the high-resolution structural data demonstrate that nucleotide context, or the immediate neighbors of a dimer step, affects the typical step-level characteristics and modifies the overall structure of sample DNA chains.

In conclusion, researchers and students now can access cutting-edge DNA modeling technology with the emDNA program developed by Professor Wilma Olson and her research team. General users will also be able to manipulate and view higher-order conformations of DNA using techniques that were previously only available to experts. These new technologies’ increased understanding of mesoscale organization may open up new research directions, such as ligand design that increases the possibility of stable loop formation and subsequent control of genetic processes.