You are here: vision-research.eu » Vision Research » Visionary of the Quarter » Elfride De Baere (Q01-2020)
Inherited retinal diseases (IRD) are a major cause of early-onset blindness, affecting over two million people worldwide. They are characterized by a tremendous clinical and genetic heterogeneity, with over 270 disease genes identified (https://sph.uth.edu/retnet).
The major topics of my research deal with precision medicine in IRD (Figure 1).
We have identified novel genes implicated in IRD using a variety of genomic strategies (linkage studies and Sanger sequencing, candidate gene testing, homozygosity mapping, whole exome sequencing [WES]). Some of these genes play a role in general processes such as transcription (NR2E3, RAX2) (1,2), splicing (SF3B2) (Van Cauwenbergh, Van de Sompele et al. in preparation) and ubiquitination (RCBTB1) (3). The identification of additional families has been facilitated by the European Retinal Disease Consortium (ERDC) (www.erdc.info). To shed light on the molecular pathogenesis we have generated knockdown (sf3b2) (Van Cauwenbergh, Van de Sompele et al. in preparation) and knockout (rcbtb1, abca4, ush2a) models in Xenopus tropicalis (Figure 2) (Carron, Naert, Ascari et al. in preparation). The latter is a suitable model system for IRD, having the major cell types of the human eye and other characteristics that make it amenable for genome editing with high efficiency.
Despite large-scale testing of known IRD genes, ~50% of tested cases remain unsolved. We and others hypothesize that a large part of the missing heritability resides in the non-coding part of the genome that is not covered by exomic approaches. We provided proof-of-concepts for missing heritability in IRD. Copy number variants (CNVs) of IRD genes represent 5-10% of the mutational spectrum. We explored the genomic landscape of IRD genes to prioritize those genes susceptible to CNV formation (4). We used a customized microarray, arrEYE, for high-resolution CNV analysis of known and candidate IRD genes and of retina-expressed noncoding RNAs (ncRNAs) (5). The concept of non-coding variation in IRD has been strengthened by autosomal recessive monogenic IRD in which only a single heterozygous mutation can be found in the coding region of the expected disease-causing gene, the so-called monoallelic cases. We showed that a deep-intronic splice mutation in CEP290 accounts for ~20% of congenital blindness in Belgium (6). We identified missing variation in monoallelic patients with ABCA4-associated disease using targeted testing and ABCA4 locus resequencing (7,8,9). Deep-intronic splice variants including a Belgian founder, a recurrent milder variant, CNVs and variants with a putative effect on regulation were identified, revealing a molecular diagnosis in 84% of the cohort. IRD-associated regulatory variants affecting transcription are less studied. We linked 5’ untranslated region (5’UTR) regulatory mutations in NMNAT1 with Leber Congenital Amaurosis for the first time (10). As the interpretation of cis-regulatory variants is challenging, integrated cis-regulatory maps based on epigenome and transcriptome profiling in human retinal tissue and cells are required to close the gap between non-coding variants and IRD pathogenesis (11,12) (Figure 3).
We were able to rescue the effect seven deep-intronic ABCA4 variants using antisense oligonucleotides (AONs) (8,9). We showed that two adjacent mutations, c.4539+1100A>G and c.4539+1106C>T, could be corrected by the same AON using patient-derived photoreceptor precursor cells (PPCs) (Naessens et al. in preparation). We targeted a recurrent, dominant negative allele (c.166G>A) in NR2E3 using AONs, aiming at a specific downregulation of the mutated allele. We identified a subtle, mutant-specific downregulation for AONs containing specialized modifications (13). Further research is ongoing to improve AON specificity however (Figure 4).
ORCID: 0000-0002-5609-6895
Center for Medical Genetics Ghent (CMGG)
Campus Ghent University Hospital, MRB1,
Corneel Heymanslaan 10
9000 Ghent, Belgium
Email: Elfride.DeBaere[at]UGent.be
Website: www.debaerelab.com | www.startn.eu
Twitter: @elfridedebaere