Suzanne de Bruijn

180 Chapter 4 RP17-SVs create new topologically associating domains and ectopic enhancer-gene interactions Using the wildtype retinal organoid Hi-C map, we modelled the TAD boundaries, CTCF site orientation and retinal TF binding site positions for each unique RP17-SV ( Figure 4A , S9 ). In NL-SV1 the duplication contains part of the YPEL2 TAD, the boundary to the neighboring region and GDPD1 . This results in the creation of a neo-TAD which now contains the previously separated YPEL2 enhancers and the GDPD1 gene in one domain. To directly investigate the effect of the SVs in retinal cells, dermal fibroblasts from UK- SV2 were reprogrammed to iPSC and differentiated to 3D ROs thus creating an in vitro model ( Supplementary Results ). In this case, the duplicated regions are also inverted. Hi-C of RP17 ROs (UK-SV2) revealed the creation of two neo-TADs, compared to control ROs ( Figure 4B ). The rearrangement of CTCF sites caused by the SV creates boundaries for two novel domains (neo-TAD 1 and 2), where neo-TAD 2 contains a duplicated copy of GDPD1 and SMG8 , and the retinal enhancers, confirming the modelling for this SV ( Figure 4A ). Furthermore, based on our predictions, neo-TADs are created in each of the RP17 cases and GDPD1 is predicted to gain ectopic access to the retinal-specific enhancers ( Figure 4A , S9 ). Therefore, the potential convergent mechanism for retinal degeneration is transcriptional activation and expression of GDPD1 through juxtaposition of retinal TF binding sites within active compartments, bounded by CTCF sites. This model would also fit with a dominant gain- of-function mechanism of disease. Next, we assessed retinal enhancer expression in control and UK-SV2 ROs, by enhancer RNA qPCR ( Supplementary Results ). A significant increase of the retinal enhancer was detected in RP17 ROs ( Figure 4C ), demonstrating that this transcriptionally active retinal enhancer in the neo-TAD could drive retinal expression of GDPD1 . Differential expression of GDPD1 in RP17 iPSC derived photoreceptor precursors and 3D retinal organoids Our experimental data and modelling predict GDPD1 enters a neo-TAD with retinal enhancers in all RP17-SVs. An extra copy of YPEL2 enters the neo-TAD of NL-SV1, and SMG8 enters this domain in UK-SV2 ( Figure 5 , S9 ). To experimentally validate the consequence of RP17-SVs in genomic and cellular context, qPCR was performed to assess differential expression in PPCs (NL-SV1) and ROs (UK-SV2). The expression of GDPD1 , YPEL2 and SMG8 , was compared to controls ( Supplementary Results ).