Hepatotoxic Potential of Therapeutic Oligonucleotides Can Be Predicted from Their Sequence and Modification Pattern

Abstract

Antisense oligonucleotides that recruit RNase H and thereby cleave complementary messenger RNAs are being developed as therapeutics. Dose-dependent hepatic changes associated with hepatocyte necrosis and increases in serum alanine-aminotransferase levels have been observed after treatment with certain oligonucleotides. Although general mechanisms for drug-induced hepatic injury are known, the characteristics of oligonucleotides that determine their hepatotoxic potential are not well understood. Here, we present a comprehensive analysis of the hepatotoxic potential of locked nucleic acid-modified oligonucleotides in mice. We developed a random forests classifier, in which oligonucleotides are regarded as being composed of dinucleotide units, which distinguished between 206 oligonucleotides with high and low hepatotoxic potential with 80% accuracy as estimated by out-of-bag validation. In a validation set, 17 out of 23 oligonucleotides were correctly predicted (74% accuracy). In isolation, some dinucleotide units increase, and others decrease, the hepatotoxic potential of the oligonucleotides within which they are found. However, a complex interplay between all parts of an oligonucleotide can influence the hepatotoxic potential. Using the classifier, we demonstrate how an oligonucleotide with otherwise high hepatotoxic potential can be efficiently redesigned to abate hepatotoxic potential. These insights establish analysis of sequence and modification patterns as a powerful tool in the preclinical discovery process for oligonucleotide-based medicines.

Antisense Oligonucleotide Therapies

“…In other studies, the CTGT motif was found to increase the likelihood of hepatotoxicity with 2 0 -MOE PS ASOs (A. Levin: personal communication, DIA oligonucleotide conference, 2013). Very recently, a bioinformatic analytic program called a ''random forests classifier set'' was developed, which predicted with 80% accuracy, the lowor high hepatotoxic potential of LNAs based on basepair sequence and backbone modifications (Hagedorn et al 2013). These authors effectively demonstrated how to redesign a potentially hepatotoxic ASO to mitigate hepatic effects.…”

Quantum Mechanical Studies of DNA and LNA

“…As an example, a single nucleotide modification (e.g., perfect match vs. one base pair mismatch) may lead to major, or ''all or none,'' phenotypic changes. The QM modeling data explains and illustrates reasons underlying some of the unexpected observations relating to major phenotypic changes observed for otherwise ''closely related'' molecules (vide infra) (Seth et al, 2009;Straarup et al, 2010;Laxton et al, 2011;Stanton et al, 2012;Hagedorn et al, 2013).…”

“…Another example was recently reported where a specific LNA 13-mer was hepatotoxic. However, adding just a single additional LNA nt rendered the resulting 14-mer non-hepatotoxic (Hagedorn et al, 2013). This demonstrates that single-nucleotide substitutions may induce major phenotypic changes-changes that are not ''inherited'' from the single nucleotide per se, but result more likely from the nucleotide insertion/substitution in a specific sequence/compound context.…”

“…We are currently establishing such large databases of thousands of oligonucleotides tested both in vitro and in vivo. A simplified decomposition of oligonucleotides has been developed, combined with a random forest classification that with *80% significance can predict the hepatotoxic potential of LNA/DNA-gapmer oligonucleotides (Hagedorn et al, 2013). Predicted experimental outcome by in silico algorithms is linked with uncertainty.…”

Engineered microRNA therapeutics

“…A series of algorithms has recently been described that can help predict those chemically modified oligonucleotides with increased potential to induce liver injury. 5 A class associated off-target effect of oligonucleotides is to stimulate an immune response that may represent activation of membrane-associated and/or cytosolic nucleic acid sensors such as the toll-like receptors and retinoic acid-inducible gene 1, respectively. Binding of oligonucleotides to these sensors leads to the stimulation of an interferon response.…”