Abstract
For antisense applications, oligonucleotides must be chemically modified to be resistant to endogenous nucleases. Until now, antisense oligonucleotide (ASO) analogs have been synthesized and then tested for their ability to duplex with a target nucleic acid, usually RNA. In this work, using molecular dynamics calculations simulations, we systematically tested a series of chemically modified analogs in which the 2-deoxyribose was substituted for by one or two methylene groups on each side of the phosphate backbone, producing four compounds, of which three were previously unknown. We used a 9-mer sequence of which the solution structure has been determined by NMR spectroscopy and tested the ability to form stable duplexes of these acyclic analogs to both DNA and RNA. In only one case out of eight, we unexpectedly found the formation of a stable duplex with complementary RNA. We also applied limitations on end fraying because of the terminal AT base pairs, in order to eliminate this as a factor in the comparative results. We consider this a predictive method to potentially identify target ASO analogs for synthesis and testing for antisense drug development.