Building on previous work, here Asami, Nagata, Yoshioka and colleagues from the Tokyo Medical and Dental University and Ionis Pharmaceuticals investigate new heteroduplex oligonucleotide (HDO) designs for antisense oligonucleotides.
For more information on antisense oligonucleotides please see this webinar by Dr. Robert MacLeod (Ionis Pharmaceuticals)
In the original design (top row in figure), a complementary RNA (coRNA) conjugated to α-tocopherol was hybridized to the DNA antisense oligonucleotide to form a heteroduplex. This HDO showed enhanced therapeutic potency and productive delivery to the liver due to increased association with serum lipoproteins and increased serum stability.
Both figures are from the article featured and available under Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) conditions.
The authors had reasoned that the RNA strand in such a heteroduplex design would be cleaved by endogenous RNase H1 but the cleavage pattern was incongruent with RNase H1. This suggested that other endogenous nucleases could cleave these HDOs, so the authors investigated if the RNA strand of the heteroduplex could be replaced with a DNA strand.
Indeed, tocopherol-conjugated HDOs containing coDNA strands were more efficient at degrading their target mRNAs in mouse liver than their 13 and 16-mer parental ASOs. However, sibling HDOs without tocopherol conjugation did not show such increases.
Tocopherol-conjugated HDOs with cEt replacing the LNA modifications in the gapmer strand also induced more efficacious knockdown compared to the parental ASO without tocopherol conjugation in vivo. As should be expected, a fully 2´-fluoro modified RNA complementary strand in the HDO led to significantly diminished activity.
Notably, sibling HDOs without tocopherol conjugation showed similar activity to those with the tocopherol conjugation in free uptake studies in primary hepatocytes in contrast what was seen in the in vivo studies. This suggests that tocopherol increases distribution to the liver in vivo. Further optimization of the design determined that coDNAs with three 2´OMe and phosphorothioate modifications at each end were most effective. Although degradation of coDNA strands was significantly faster than coRNA strands, target downregulation remained the same for both designs.
The use of a second strand to protect the gapmer during in vivo delivery is an interesting concept.