It is very clear that chemistry including modified monomers play a key role in transforming nucleic acid technologies. In the context of drug development this has most clearly been shown in the case of GalNAc targeting ligands that have proven very important not only for siRNA therapeutics, but also for single stranded antisense therapeutics.
The GalNAc conjugated Inclisiran oligo was eventually acquired, along with The Medicines Company, by Novartis in late 2019 for about 10 billion USD, demonstrating how known chemistry, when combined in new contexts such as oligos, may become transformative in drug development. At the TIDES Europe Virtual 2020, Jesper Wengel, Professor, Center Director BioNEC, University of Southern Denmark, discussed chemically modified oligos that his lab has been involved with.
Developing LNA variants
Wengel’s lab has worked on large nucleic acids (LNA), and unlocked nucleic acids (UNA). They are proven to be mixable with other modifications and with DNA and RNA, which is very important for product tuning. This contrasts other modifications which are not completely mixable with existing monomers. That is a key advantage. A recent development in his lab is a 2’-Fluoro DNA modification.
His lab is continuously exploring variants of the LNAs towards addressing different challenges including how to possibly improve the pharmacokinetic properties of antisense oligonucleotides, biodistribution, duration of action, but also to possibly reduce off-target effects.
Wengel’s lab developed a lipophilic LNA derivative called Pal-LNA and demonstrated relatively efficient synthesis of this monomer. It has many steps but is a high yielding synthesis. Comparing the binding affinity of Pal-LNA monomer relative to the normal LNA, they show similar binding affinity, slightly lower for the palmitoylated version than the LNA. Therefore, even though a lipophilic group is pointing into the minor groove of duplexes with DNA and RNA, the binding affinity is still dominated by the high affinity induced by the conformationally locked LNA type sugar ring.
Pal constitution leads to albumin binding. In contrast, for a normal LNA gapmer, the binding to albumin is very weak at best, even for the all-phosphorothioate version. If two Pal-LNA monomers, one in each end, are introduced both for the all-phosphorothioate and for the all-phosphodiester, significant binding to albumin is achieved. The same binding to albumin is engineered by having two Pal-LNA monomers positioned towards one end together. The position at 3’ end or towards the 5’ end achieves the same result. However, one Pal is not enough, it does not lead to significant albumin binding shown by in vitro assays.
Wengel’s group demonstrated HER3 gene knockdown by gymnotic uptake (no transfection) of LNA phosphorothioate gapmer in vitro. With the diPal gapmer of phosphodiester, the same or even better potency was demonstrated. It could be of interest in light of the toxicity associated with phosphorothioate linkages.
Despite the fact that the phosphorothioates are known to bind proteins which increases plasma half-life, a much longer half-life in plasma was demonstrated by diPal addition. For this kind of antisense oligos, it seems that the presence diPal LNA monomer dictates the organ distribution of the molecule.
The lab has also developed a so-called mini-holliday junction-like construct (HJ-LNA) with four strands, enabling the attachment of a number of end-modified molecular units in a tetravalent fashion. HJ-LNA demonstrated increased half-life in plasma and a shift of organ distribution from kidney with the holliday junction to the liver, with the diPal version of the holliday junction. As a result, the organ distribution is again dictated by the presence of the two Pal-amino-LNA monomers.
Another modification, D-lyxo, is compatible with gene knockdown. With this new monomer, they have been able to construct a gap modification that is compatible with RNase H cleavage and gene knockdown and shows an interesting specificity, an increase relative to the all DNA version.
These results demonstrate that relatively simple chemistry, elaborating on known approaches can 1) increase the lipophilicity of the oligonucleotides and 2) engineer the conformational preference of a monomer through conformational restriction. It has been possible to modify and still see RNase H activity and to engineer some specificity into the cleavage of RNase H. In conclusion, there is, in Wengel’s opinion, a lot of chemistry to explore in order to optimize the properties of the antisense molecule not only to improve pKa, but also to reduce toxicity and off-target effects.