Tissue-specific TCF4 triplet repeat instability revealed by optical genome mapping

Abstract

Background: Fuchs endothelial corneal dystrophy (FECD) is the most common repeat-mediated disease in humans. It exclusively affects corneal endothelial cells (CECs), with ≤81% of cases associated with an intronic TCF4 triplet repeat (CTG18.1). Here, we utilise optical genome mapping (OGM) to investigate CTG18.1 tissue-specific instability to gain mechanistic insights.

Methods: We applied OGM to a diverse range of genomic DNAs (gDNAs) from patients with FECD and controls (n = 43); CECs, leukocytes and fibroblasts. A bioinformatics pipeline was developed to robustly interrogate CTG18.1-spanning DNA molecules. All results were compared with conventional polymerase chain reaction-based fragment analysis.

Findings: Analysis of bio-samples revealed that expanded CTG18.1 alleles behave dynamically, regardless of cell-type origin. However, clusters of CTG18.1 molecules, encompassing ∼1800-11,900 repeats, were exclusively detected in diseased CECs from expansion-positive cases. Additionally, both progenitor allele size and age were found to influence the level of leukocyte-specific CTG18.1 instability.

Interpretation: OGM is a powerful tool for analysing somatic instability of repeat loci and reveals here the extreme levels of CTG18.1 instability occurring within diseased CECs underpinning FECD pathophysiology, opening up new therapeutic avenues for FECD. Furthermore, these findings highlight the broader translational utility of FECD as a model for developing therapeutic strategies for rarer diseases similarly attributed to somatically unstable repeats.

Funding: UK Research and Innovation, Moorfields Eye Charity, Fight for Sight, Medical Research Council, NIHR BRC at Moorfields Eye Hospital and UCL Institute of Ophthalmology, Grantová Agentura České Republiky, Univerzita Karlova v Praze, the National Brain Appeal's Innovation Fund and Rosetrees Trust.

Keywords: Fuchs endothelial corneal dystrophy; Optical genome mapping; Somatic mosaicism; Tissue-specific repeat instability; Trinucleotide repeat expansion disease; Triplet repeat expansion-mediated disease.

Fig. 1

Optical genome mapping (OGM) effectively detects large expanded CTG18.1 alleles. (A) Detected traces after capillary electrophoresis of STR-PCR (i, iii, v) and TP-PCR (ii, iv, vi) products amplified from non-expanded whole-blood derived gDNA samples (i-ii), mono-allelic expanded whole-blood derived (iii-iv) and F35T cell-derived (v-vi) gDNA samples. Red boxes highlight the presence of expanded alleles as indicated by TP-PCR traces. (B) Schematic summary of OGM methodology; (1) extraction of ultra-high molecular weight (UHMW) gDNA that is (2) subsequently labelled via covalent modification at genome-wide CTTAAG hexamer motifs before (3) linearising and imaging the decorated molecules on nanochannels (image adapted from: https://bionanogenomics.com). (C) Histogram of OGM CTG18.1 molecule sizes (bp) observed in immortalised CEC line F35T. The red dotted line indicates alleles around the lowest detection threshold of the method, likely representing the non-expanded allele.

Fig. 2

Diseased corneal endothelial cells (CECs) display increased levels of CTG18.1 somatic instability compared to unaffected leukocytes. A series of peripheral blood leukocyte-derived (BL1-9 in blue) and corneal endothelial-derived (CEC1-9 in green) gDNA samples from nine unrelated FECD patients were analysed by OGM. Each grey box denotes samples from the same individual (subjects 1–9). In each instance, higher levels of somatic instability were detected in affected CECs compared to unaffected blood leukocyte-derived gDNA samples. The size (bp) of the CTG18.1 repeat-containing molecules is plotted (x-axis) against the total number of CTG18.1 molecules detected (y-axis). Red arrows depict the bin with most molecules detected above the 5439 bp threshold observed exclusively within the CEC-derived gDNA samples. Baseline CTG18.1 genotypes determined by STR-PCR analysis of leukocyte gDNA are shown in brackets, for each respective allele.

Fig. 3

Analysis of bi-allelic CTG18.1 corneal endothelial cells (CECs) indicates that molecules near the detection threshold of optical genome mapping (OGM) represent non-expanded alleles. CEC-derived gDNA samples from three unrelated FECD cases were analysed by OGM. Based on STR-analysis of leukocyte-derived DNA, we classified cases as CTG18.1 (A) expansion-negative (CEC-Ctr1), (B) mono-allelic expanded (CEC-9) or (C) bi-allelic expanded. Baseline STR genotyping results are shown in brackets for each respective allele. Red arrows depict the bin with most molecules detected >2000 bp threshold, exclusively observed in mono- (CEC-9; B) and bi-allelic (CEC-10; C) expanded CTG18.1 samples.

Fig. 4

Expanded CTG18.1 alleles behave dynamically in both peripheral blood leukocytes and corneal endothelial cells. Dot plots of molecules detected per individual capturing the distribution of all CTG18.1 alleles analysed in (A) peripheral blood leukocytes (BL; n = 24) and (C) corneal endothelial cell (CEC; n = 15) samples are presented. (A,C) Individuals with non-expanded CTG18.1 alleles (<50) are colour-coded in grey (Ctrl). Light blue and light green shades indicate FECD individuals with non-expanded CTG18.1 alleles in their peripheral blood leukocytes or CEC-derived gDNA, respectively. Lines represent the mean CTG18.1 expansion size in base pairs (bp) per sample. (B,D) Scatter plots of CTG18.1 mean molecule size (bp) against the largest progenitor allele is shown with polynomial regression for the expansion-positive dataset (red) and 95% confidence interval (CI) shown in blue for (B) peripheral blood leukocyte and (D) CEC sample series. All data series are arranged in order of the largest allele detected per individual according to baseline CTG18.1 genotype determined by the STR-PCR assay and indicated in brackets for each sample.