Reduction of retinal ganglion cell death in mouse models of familial dysautonomia using AAV-mediated gene therapy and splicing modulators

**Background:** Familial dysautonomia (FD) is a rare recessive neurodevelopmental and neurodegenerative disease caused by a splicing mutation in the ELP1 gene (c.2204+6T>C), leading to tissue-specific skipping of exon 20 and reduced functional ELP1 protein. All FD patients develop progressive optic neuropathy due to retinal ganglion cell (RGC) death, resulting in visual impairment. Currently, no disease-modifying treatments exist. This study tested two therapeutic strategies: AAV-mediated gene replacement therapy and a small molecule splicing modulator (PTC680) to restore ELP1 levels and prevent RGC death in FD mouse models.

**Methods:** Two complementary FD mouse models were used. For gene therapy, Pax6cre;Elp1^loxp/loxp^ mice (retina-specific Elp1 knockout) received bilateral intravitreal injections at postnatal day 21 (P21) of AAV2 vectors expressing either mouse Elp1 (mElp1) or human ELP1 (hELP1) under the U1a promoter, with AAV2-eGFP as control. Vector doses ranged from 2.7×10^7 to 1.7×10^9 vg/eye. For splicing modulator studies, a new hybrid model (TgFD9;Pax6cre;Elp1^loxp/loxp^) carrying the human FD mutation was generated. Mice were fed chow containing 0.0008% PTC680 (1.2 mg/kg/day) or vehicle starting at embryonic day 18.5 (E18.5), P21, or P90, for 3 months. RGCs were quantified by Brn3a immunolabeling in retinal flat mounts, counting cells in four 1 mm^2 peripheral quadrants. ELP1 splicing was analyzed by RT-PCR.

**Key Results:** For gene therapy with mElp1, the high dose (1.6×10^9 vg/eye) increased RGC numbers by 137% (5696±252 vs. 2404±405 in untreated FD mice, p<0.0001), and the low dose (8.4×10^8 vg/eye) by 131% (5563±172, p<0.0001). Control eGFP virus also increased RGCs by 82% (4386±534), but mElp1 significantly outperformed eGFP by 30% (p=0.04). For hELP1, the highest dose (5.4×10^8 vg/eye) increased RGCs by 39% (4925±439 vs. untreated FD, p=0.02), with approximately 26% of RGCs expressing human ELP1 protein. For PTC680, treatment starting at E18.5 increased RGCs by up to 57% (p<0.001), at P21 by 57% (p<0.0001), and at P90 by 57% (p<0.0001) compared to vehicle-treated FD mice. PTC680 significantly increased exon 20 inclusion in human ELP1 transcripts (p<0.0001).

**Clinical Implications:** This study provides the first preclinical evidence that both gene replacement therapy and splicing modulation can effectively reduce RGC death in FD mouse models. The data support further optimization of AAV2 vectors for clinical translation, potentially using enhanced capsids or dual AAV systems to accommodate the large ELP1 transgene. The oral splicing modulator PTC680 shows efficacy even when started at 3 months of age, suggesting potential benefit for adolescent and adult FD patients who typically report vision loss in their teens. These findings lay the foundation for clinical trials aimed at preventing blindness in FD patients.