PMC:5985359 / 36245-44416 JSONTXT

{"project":"2_test","denotations":[{"id":"29526280-28405024-2044802","span":{"begin":155,"end":157},"obj":"28405024"},{"id":"29526280-28641106-2044803","span":{"begin":159,"end":161},"obj":"28641106"},{"id":"29526280-27112497-2044804","span":{"begin":163,"end":165},"obj":"27112497"},{"id":"29526280-22859208-2044805","span":{"begin":167,"end":169},"obj":"22859208"},{"id":"29526280-22726834-2044806","span":{"begin":171,"end":173},"obj":"22726834"},{"id":"29526280-17436254-2044807","span":{"begin":1430,"end":1432},"obj":"17436254"},{"id":"29526280-17436255-2044808","span":{"begin":1434,"end":1436},"obj":"17436255"},{"id":"29526280-28886202-2044809","span":{"begin":1547,"end":1549},"obj":"28886202"},{"id":"29526280-28562589-2044810","span":{"begin":1895,"end":1897},"obj":"28562589"},{"id":"29526280-28562589-2044811","span":{"begin":3057,"end":3059},"obj":"28562589"},{"id":"29526280-28562589-2044812","span":{"begin":3426,"end":3428},"obj":"28562589"},{"id":"29526280-25593321-2044813","span":{"begin":4542,"end":4544},"obj":"25593321"},{"id":"29526280-28118661-2044814","span":{"begin":4546,"end":4548},"obj":"28118661"},{"id":"29526280-11581212-2044815","span":{"begin":6543,"end":6545},"obj":"11581212"},{"id":"29526280-22859208-2044816","span":{"begin":7556,"end":7559},"obj":"22859208"},{"id":"29526280-26436889-2044817","span":{"begin":7794,"end":7796},"obj":"26436889"},{"id":"29526280-23380860-2044818","span":{"begin":7798,"end":7800},"obj":"23380860"},{"id":"29526280-27106101-2044819","span":{"begin":7802,"end":7804},"obj":"27106101"},{"id":"29526280-27151457-2044820","span":{"begin":7806,"end":7808},"obj":"27151457"}],"text":"Discussion\nASOs have recently emerged as a powerful therapeutic option for disease intervention, including those caused by trinucleotide repeat expansions.10, 39, 40, 41, 42 Here we show that a non-coding CTG trinucleotide repeat expansion in TCF4 (CTG18.1) confers greater than 76-fold risk for FECD in a large white British and Czech cohort. We demonstrate that primary CECs derived from FECD-affected subjects display the predicted hallmarks of primary and downstream repeat-expansion-associated pathology, and subsequently show that these changes are reversed by an ASO treatment specifically targeted at the CTG18.1 trinucleotide repeat expansion. An ASO-based treatment could therefore offer an innovative therapeutic approach that could benefit a substantial number of individuals affected by this common and sight-threatening condition.43\nThe data presented here suggest that the TCF4 repeat expansion leads to CEC-specific dysfunction, as unlike other trinucleotide expansion diseases, nuclear RNA foci are not observed in case-matched fibroblasts. These data, at least in part, explain the corneal-specific phenotype resulting from repeat expansions in this widely expressed gene and highlight the importance of investigating the trinucleotide expansion in primary human CECs. TCF4 haploinsufficiency causes the systemic condition Pitt-Hopkins syndrome, but the noncoding repeat expansion exclusively affects the cornea.44, 45 Interestingly, a recent study has reported that FECD is a common ocular finding in DM1-affected case subjects.46 Combined with our data, this suggests that the corneal endothelium is susceptible to toxicity induced by these genetically distinct repeat expansions, driven by RNA foci.\nIt is well established for a wide range of repeat expansion disorders that disease onset and incidence of RNA foci manifest only above a critical level of nucleotide repeats.47 A threshold for CTG18.1 repeat length and FECD association is yet to be fully defined. Performing FISH with 36 distinct CEC lines derived from FECD-affected subjects has enabled us to identify the threshold for the number of repeats required to produce nuclear RNA foci in our model (Figure 2B; Table S4). Nine CEC lines with CTG18.1 genotypes ranging from 12/12 to 18/31 were found to lack RNA foci. A further 27 lines with CTG18.1 genotypes ranging from 25/31 to 12/126 were all found to exhibit punctate nuclear RNA foci. These data allow us to correlate CTG18.1 genotype status with CUG RNA foci incidence and indicate that a repeat size of more than 31 trinucleotide repeats is sufficient to drive the accumulation of stable CUG RNA foci in primary CECs (Figure 2B; Table S4). This identified threshold also correlates notably with the bi-nominal distribution of CTG18.1 repeat length observed in our FECD cohort (Figure 1A), which is likely attributed to the instability of the repeat above approximately 30 copies. Interestingly, RNAs containing more than 30 CUG repeats have recently been demonstrated to undergo phase separation to form nuclear foci.47 The threshold for this repeat length-dependent process (30 CUG repeats) is remarkably similar to what we have observed with respect to CTG18.1-related foci occurrence in CECs (Figure 2B). These phase separation data further support the use of agents that disrupt RNA-RNA base-pairing, such as ASOs, as viable treatment options for RNA foci-induced cellular toxicity.47 Taken together, these data suggest that a CTG18.1 length ≥32 should, in future, be considered as FECD risk associated. We repeated the tests for association with FECD defining the expanded repeats as ≥32, instead of the previously used more conservative threshold of ≥50. The association model became more significant (p = 3.79 × 10−74), although the disease risk conferred by this locus was lower (OR = 34.14; 95% CI: 23.35–49.91). Future analysis of CECs from individuals affected by FECD with repeat lengths in the unidentified range of 31–53 could further refine this important threshold.\nWe investigated the downstream consequences of RNA foci and have identified that sequestration of RNA splicing factors, MBNL1 and MBNL2 (Figure 3), in addition to abnormal patterns of alternative splicing were detectable in a repeat expansion-specific manner (Figure 4). These observations reinforce the notion that such RNA structures induce toxic gain-of-function effects that are likely to be disrupting overall cellular homeostasis, implicating aberrant RNA metabolism in the pathogenesis of CTG18.1-associated FECD.15, 37 Furthermore, these data demonstrate that these events are specific to CTG18.1-mediated FECD and are not a general downstream consequence of the disease, given that CECs derived from FECD-affected case subjects without expanded copies of the repeat did not display features of aberrant RNA metabolism.\nImportantly, we demonstrate here that an ASO targeted to the CTG trinucleotide TCF4 expansion can ameliorate disease-associated markers of RNA toxicity in CECs derived from FECD-affected subjects, specifically reducing RNA foci formation (Figures 5A–5C), prompting MBNL1 nuclear redistribution (Figures 5D and 5E) and partially suppressing differential splicing events (Figure 6). We additionally demonstrate that the ASO can access the corneal endothelium when injected intravitreally in mice (Figure S6). Entry of fluorescently labeled ASO to the corneal endothelial cells is an endogenous property of naked 2′Ome-PS-ASOs oligo and no other excipients are required to induce entry into these cells. Access to the CECs is also both dose and time dependent, suggesting that ASOs are rapidly taken up by cells after dosing. While it may be desirable to consider a topical eye drop therapy for FECD using an ASO approach, this is highly unlikely to be effective using naked ASOs given the structure of the corneal epithelium and the size and charge of ASOs; an adjunctive delivery technology would likely have to be considered in this case.\nFuture studies using CECs from affected individuals will be helpful in defining an effective ASO concentration, which can lead to a good clinical dose estimate when considering introduction to the anterior ocular chamber, which has a small fixed but rapidly exchanging volume. Detailed pharmacokinetic studies will be required to define the optimal dosing time interval, which is expected to be months, based on the rapid cellular uptake and the tissue/cellular longevity of similar 2′Ome-PS-ASOs in studies that have examined intraocular dosing of ASOs.48 It is important to note that in the final stages of FECD there is significant endothelial cell loss, and consequently ASO-specific treatment is intended to prevent further cell loss. The most likely group of affected individuals to benefit from such a therapeutic intervention will be those individuals who are in the early stages of disease and have not yet experienced significant endothelial cell loss. Importantly, these at-risk individuals can be effectively identified by a CTG18.1 genotyping test.\nIn summary, we demonstrate proof-of-concept data that a targeted (CAG)7 ASO treatment reduces gain-of-function RNA toxicity induced by TCF4 CTG18.1 expansion, in a cellular and human genomic context. With the absence of FECD animal models, human ex vivo models are vital, both to provide a validation of the therapeutic approach for FECD and to continue the translation of ASO therapies into the clinic. ASO therapies are already in clinical trials for a variety of repeat expansion disorders including DM141 and Huntington disease (MIM: 143100), and additionally, intraocular ASO therapies are proposed for retinitis pigmentosa (RP [MIM: 268000]), RP associated with Usher syndrome (MIM: 608400), and Leber congenital amaurosis (MIM: 610142).49, 50, 51, 52 We propose that our proof-of-concept study provides evidence to translate this therapeutic approach to FECD given the accessibility of the diseased tissue and the relatively delayed onset of disease, which provides a window of opportunity to identify at risk individuals with TCF4 repeat expansions and prevent disease progression in pre-symptomatic individuals."}