Last week, uniQure achieved a breakthrough in developing disease-modifying therapies for Huntington’s Disease by showing that AMT-130 slowed disease progression by 75%. AMT-130 is a DNA-directed RNAi construct targeting a sequence in the triplett-expanded exon 1 of the 67 exons-containing huntingtin gene. Since other approaches aimed at lowering full-length huntingtin, but leaving exon 1 mRNA untouched, have fallen short (e.g. tominersen by Ionis and Roche) this is very strong support that targeting the aberrant exon 1 transcript is essential for success.
The aberrant exon 1 transcript, and likely the protein produced from that, is loaded with polyglutamine-encoding CAG tripletts, has been shown in animal studies to be exceptionally toxic and is also produced in patients (see previous blog entry 'Huntington's Disease Therapeutics Finally Breaking Through!'), the amount of which is correlated with triplett expansion size based on comments by uniQure CMO Walid Abi-Saab at a conference earlier this year. Add to this the observation that a phase 3 trial with tominersen had to be stopped due to the oligo performing worse than control, maybe it is time for a heretical new idea:
accept that the old notion of Huntington’s Disease being an autosomal dominant disease caused by toxic full-length triplett-expanded huntingtin has been wrong, and full-length huntingtin indeed is an important protein to have. Instead, leave full-length huntingtin alone and exclusively target the aberrant exon 1 ‘fragment’. Hey, have you ever considered the possibility that the polyQ htt protein aggregates are not cleavage products from full-length Htt? Why always so complicated? Why consider both exon 1 and full-length Htt contributing meaningfully to disease? Shouldn't the properties of a short protein largely made up of polyglutamine be radically different from that of a much larger one containing the same polyglutamine stretch?
Sarepta and Arrowhead have shared plans to enter the clinic with a systemically administered RNAi conjugate for Huntington’s Disease. A systemically administered option would be a very valuable one since, especially in light of the one-time administered AMT-130, repeat intrathecal administration is less attractive. The Tfr1-targeted ARO-HTT being prepared for clinical trial application this year apparently does not affect exon 1 transcript. The excuse given is that it is difficult to find an effective RNAi trigger in the merely 102 nucleotide exon 1 target space.
The good
news: by focusing exlusively on aberrant exon 1 transcript, 7000+ nucleotides of
additional target space (from 5’ end of intron 1) opens up.
What good is it to have a trigger that is great at knocking down a transcript
when it is aimed at the wrong target? It
is not too late and can be explained to the market.
5 comments:
I thought that uniqure targets both the full length and the exon 1 fragment? So apparently it is not harmful to reduce the full length protein? Or is this too simplified and too early to tell
You are raising a good point. The lack of apparent toxicity of AMT-130 that is also targeting AMT-130 could be either due to the exquisite specificity of targeting the AAV to where it is needed most (tominersen e.g. preferentially engages target in the cortex, not deep nuclei) and/or the modest knockdown efficacy of AMT-130 (e.g. Sogorb-Gonzalez et al 2024). My main point though is that one should not limit onself to 102nt targeting space because of aiming to knock down full-length Htt as well, but embrace the 7000+ nt of 'intron 1' as additional space for finding highly active RNAi triggers.
I definitely agree, but in the theoretical case you target the intron 1 sequence, you throw in pretty much 100% with the exon 1a hypothesis. Whereas targeting exon 1 like uniqure leaves both options open (full length and shorter toxic fragments).
I'm not fully on board with exon 1a yet myself, cause if I remember correct, tominersen had pretty decent preclinical data and ability to improve the phenotype of mouse models. And like you say, tominersen probably is not hitting the key brain regions, but rather more cortical regions in patients. On top of that Tomi appears to have a class-effect dose limiting toxicity of the ASO. So in my mind, there hasn't been a fair head-to-head of targeting full length over just the 1a yet.
Will definitely be interesting to see how Alnylam progresses vs Arrowhead for instance! Exciting times and hope for HD patients
I hope you don't mind me spamming here, but I dug a bit deeper into the preclinical data of Tominersen because of your post. In one of the original tomi preclinical papers they test in a couple mouse models (YAC128, BACHD) and show HTT knockdown and improvement of behavior phenotype of the mice. I doublechecked and what is called huASO in the paper matches tominersen sequence exactly.
Now in a more recent paper with YAC128 mice, no effect on exon 1a is shown. https://pubmed.ncbi.nlm.nih.gov/35793238/
In this case I dont believe its tominersen because they speak of targeting exon 36. Still, whats interesting is lack of reduction in 1a when hitting full length htt in the YAC128 mouse cells. So this begs the question how did tominersen improve the mouse phenotype if indeed it wasn't hitting 1a? Is it a fluke or was there some minor reduction of 1a because of pre-mRNA reduction maybe? I have no idea at this moment.
Another interesting thing in the first paper is that Ionis apparently had an ASO called HuASOex1 that targets exon 1! Specifically tested in the R6/2 mouse (exon 1 model), but never progressed any further that I'm aware of.
I'm by no means an HD or ASO expert but it is still interesting to speculate as to why tominersen failed
here is the original tominersen paper I refer to btw: https://pmc.ncbi.nlm.nih.gov/articles/PMC3383626/
Post a Comment