Modelling aptamers with nucleic acid mimics (NAM): From sequence to three-dimensional docking

Abstract

Aptamers are single-stranded oligonucleotides, formerly evolved by Systematic Evolution of Ligands by EXponential enrichment (SELEX), that fold into functional three-dimensional structures. Such conformation is crucial for aptamers' ability to bind to a target with high affinity and specificity. Unnatural nucleotides have been used to develop nucleic acid mimic (NAM) aptamers with increased performance, such as biological stability. Prior knowledge of aptamer-target interactions is critical for applying post-SELEX modifications with unnatural nucleotides since it can affect aptamers' structure and performance. Here, we describe an easy-to-apply in silico workflow using free available software / web servers to predict the tertiary conformation of NAM, DNA and RNA aptamers, as well as the docking with the target molecule. Representative 2'-O-methyl (2'OMe), locked nucleic acid (LNA), DNA and RNA aptamers, with experimental data deposited in Protein Data Bank, were selected to validate the workflow. All aptamers' tertiary structure and docking models were successfully predicted with good structural similarity to the experimental data. Thus, this workflow will boost the development of aptamers, particularly NAM aptamers, by assisting in the rational modification of specific nucleotides and avoiding trial-and-error approaches.

Fig 1. Workflow to predict the tertiary structure of NAM aptamers (using the aptamers sequence as a starting point) and the docking model (including the identification of nucleotides that interact with the target molecule).

The yellow horizontal arrows are additional inputs necessary to run the workflow.

Fig 2. Aptamers’ secondary (A) and tertiary (B) structure obtained after executing the first five steps of the workflow (Mfold, 3dRNA, Biovia, PyMOL and QRNAS).

(A) The secondary structures obtained by Mfold and the Gibbs free energy (ΔG). (B) The overlap of the predicted tertiary structures (in red) and the corresponding experimentally resolved structures downloaded from the PDBe (in blue), and the RMSD values.

Fig 3. Comparison of the docking models deposited experimentally (in blue) in the PDBe database and the in silico docking models predicted through the described workflow (in red) for the 2ʹOMe (5D3G and 5HRT), DNA (5HRU) and RNA (6SY4) aptamers.

The target molecules (in green) were isolated from the aptamer-target complexes determined experimentally using the PyMOL software and used as a receptor in the docking prediction.

Fig 4. Identification of aptamer-target contact nucleotides and type of non-covalent interaction involved in the stabilization of the complexes for in silico (workflow) and experimental data (PDBe) docking.

Hydrophobic Interactions (orange), Hydrogen Bonds (blue), Salt Bridges (green), π-Stacking (brown) and π-Cation Interactions (yellow).