1 2 aim

#apaperaday: Levels of Exon-Skipping Are Not Artificially Overestimated Because of the Increased Affinity of Tricyclo-DNA-Modified Antisense Oligonucleotides to the Target DMD Exon

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Levels of Exon-Skipping Are Not Artificially Overestimated Because of the Increased Affinity of Tricyclo-DNA-Modified Antisense Oligonucleotides to the Target DMD Exon

Today’s pick is from Oligonucleotide Therapeutics Society and Mary Ann Liebert journal Nucleic Acid Therapeutics by Doisy et al on a study to see how high affinity antisense oligonucleotides (ASOs) can lead to overestimation of exon skipping levels by RT-PCR DOI: 10.1089/nat.2024.0002

Exon skipping is an approved treatment for Duchenne. However, there is room for improvement, as current dystrophin restoration levels for approved ASOs are low. Researchers are exploring different ASO chemistry modifications to improve efficiency.

Authors refer to our recent work in collaboration with @DANorDNA @F_ANAtico where we showed that high-affinity ASOs can make it look like there is a lot of exon skipping, but this is actually an artefact.

Brief summary: High-affinity LNAs bind to the unskipped transcripts, interfering with cDNA synthesis and PCR amplification. Thus, the skipped transcripts are preferentially amplified, making it seem like there is a lot of exon skipping when there actually isn’t.

A countermeasure is heating RNA to 96°C before cDNA synthesis to melt off the high-affinity ASO. However, this affects transcript levels overall, as RNA is a fragile, sensitive molecule. This fragility affects both full-length & skipped transcripts, avoiding overestimation.

Authors here use tricycloDNA (tcDNA), another high-affinity ASO, albeit less so than LNAs. They wondered if their discrepancy between high exon skipping levels and low dystrophin restoration could be explained by this artefact, so they studied it.

First, they spiked tcDNA into the RT-PCR. This did not affect exon skipping levels, but overall PCR efficiency decreased, and at 20 ng spike-in, PCR failed completely. So, tcDNA negatively impacts cDNA synthesis and/or PCR amplification.

Authors checked if tcDNA was present in their RNA samples. Indeed it was, but levels were much lower than what was spiked in. PCR efficiency was OK, and there was no correlation between exon skipping levels and the amount of tcDNA in the RNA (R² was 0.04 ???? i.e., horizontal line).

Authors then did an extra purification step to remove tcDNA, which had no effect on exon skipping levels as expected. They also preheated samples to 96°C before amplification, which did not affect exon skipping levels.

It did impact the overall efficiency of the PCR, as expected. All work was done with mouse dystrophin exon 51-targeting tcDNA. However, for tcDNA targeting mouse exon 53, the authors observed something different, using a combination of 3 tcDNAs.

Authors found that tcDNA in the sample reduced PCR efficiency and showed lower levels of skipped transcripts for tcDNA, while a PMO control had unaffected skipping levels.

Authors conclude that the mix of tcDNA might have included more tcDNA, impacting cDNA and PCR efficiency. This serves as a warning for people using combinations of ASOs. Furthermore, this is likely a sequence-dependent effect.

With more GC in the ASO(s), the affinity will be higher, and the risk for high-affinity ASO chemistries interfering with cDNA synthesis is greater. I appreciate the authors following up on this and sharing their results!