Suzanne de Bruijn

160 Chapter 3.2 Essential questions that remain are how much reduction of mutant allele expression is needed to achieve a therapeutic effect and how much expression of the wildtype allele should be maintained for normal hair cell function. Here, even organoid models likely fall short, and the generation of a knockin animal model should be considered. An animal model could provide valuable insights in the feasibility and efficacy of the lead AON to relieve the burden of mutant RIPOR2 proteins in the auditory hair cells of the cochlea. These studies could also shed light on the balance between the knockdown of mutant RIPOR2 and maintained wildtype RIPOR2 synthesis that is needed for a clinically meaningful improvement. While the therapeutic efficacy of near complete and highly mutant allele-specific knockdown of RIPOR2 seems apparent, this does not necessarily reflect the minimal decrease in mutant RIPOR2 transcript levels that is required to lower the amount of mutant RIPOR2 protein to a level that can halt, or significantly delay, disease progression. With the current knowledge, it can only be speculated to which extent mutant allele knockdown would be required to achieve therapeutic potential. DFNA21 is an adult-onset, progressive disease, suggesting that mutant protein toxicity and burden is only slowly accumulating. Removing a small percentage of this burden could potentially already delay the onset of HL for several years, especially when treatment is initiated at an early age. Therefore, we speculate that a >50% reduction of mutant transcript levels, achieved upon delivery of 250 nM of the lead AON to patient- derived fibroblasts, might already be sufficient to provide a clinically meaningful outcome to DFNA21 patients especially when treatments are started at an early stage. Additionally, exon-skipping studies have indicated that already ~20% of wildtype allele expression can be enough for phenotypic rescue in recessively inherited hearing loss and retinal degeneration. 28,29 Hypothetically, an approximate 50% reduction in wildtype RIPOR2 transcripts could be tolerated in order to achieve a strong reduction in mutant RIPOR2 transcripts. The transient nature of gapmer-mediated gene knockdown is both an advantage and a potential drawback for future clinical applications. It lowers the risk of sustained adverse or off-target effects that could accompany genome editing techniques, but repeated delivery of gapmers is likely required to achieve maximum efficacy. However, a repeated surgical delivery directly into the cochlea is not feasible. The recent study by Lentz et al. indicated that, in a mouse model of USH1C, AONs can diffuse over the round window membrane to reach auditory hair cells of the adult cochlea. 15 Also, intratympanic injections of steroids such as dexamethasone, are routinely and repeatedly used to treat e.g. Meniere's disease or idiopathic sudden sensorineural hearing loss which suggests this method could be very promising for repeated AON delivery as well. 30,31 Recent studies on splice-modulation AONs in the mouse revealed a relatively stable effect on transcript level up to 200 days post-delivery. While the delivery method and