PMC:5985359 / 19667-36243 JSONTXT

{"project":"2_test","denotations":[{"id":"29526280-23185296-2044776","span":{"begin":797,"end":798},"obj":"23185296"},{"id":"29526280-26200491-2044777","span":{"begin":800,"end":802},"obj":"26200491"},{"id":"29526280-25298419-2044778","span":{"begin":804,"end":806},"obj":"25298419"},{"id":"29526280-24255041-2044779","span":{"begin":808,"end":810},"obj":"24255041"},{"id":"29526280-28358029-2044780","span":{"begin":1454,"end":1456},"obj":"28358029"},{"id":"29526280-27381832-2044781","span":{"begin":4035,"end":4040},"obj":"27381832"},{"id":"29526280-27196322-2044782","span":{"begin":4042,"end":4044},"obj":"27196322"},{"id":"29526280-25484056-2044783","span":{"begin":4046,"end":4048},"obj":"25484056"},{"id":"29526280-24268186-2044784","span":{"begin":4136,"end":4138},"obj":"24268186"},{"id":"29526280-23641909-2044785","span":{"begin":4140,"end":4142},"obj":"23641909"},{"id":"29526280-22194824-2044786","span":{"begin":4144,"end":4146},"obj":"22194824"},{"id":"29526280-25593321-2044787","span":{"begin":4454,"end":4456},"obj":"25593321"},{"id":"29526280-25722209-2044788","span":{"begin":4458,"end":4460},"obj":"25722209"},{"id":"29526280-25593321-2044789","span":{"begin":6606,"end":6608},"obj":"25593321"},{"id":"29526280-10970838-2044790","span":{"begin":7136,"end":7138},"obj":"10970838"},{"id":"29526280-19095965-2044791","span":{"begin":7140,"end":7142},"obj":"19095965"},{"id":"29526280-28442996-2044792","span":{"begin":7144,"end":7146},"obj":"28442996"},{"id":"29526280-22884328-2044793","span":{"begin":7148,"end":7150},"obj":"22884328"},{"id":"29526280-20098426-2044794","span":{"begin":7152,"end":7154},"obj":"20098426"},{"id":"29526280-28949831-2044795","span":{"begin":7156,"end":7158},"obj":"28949831"},{"id":"29526280-25593321-2044796","span":{"begin":8014,"end":8016},"obj":"25593321"},{"id":"29526280-28118661-2044797","span":{"begin":8018,"end":8020},"obj":"28118661"},{"id":"29526280-28118661-2044798","span":{"begin":8592,"end":8594},"obj":"28118661"},{"id":"29526280-19667189-2044799","span":{"begin":10362,"end":10364},"obj":"19667189"},{"id":"29526280-23511335-2044800","span":{"begin":10366,"end":10368},"obj":"23511335"},{"id":"29526280-26325627-2044801","span":{"begin":16410,"end":16412},"obj":"26325627"}],"text":"Results\n\nExpansion of the CTG18.1 Trinucleotide Repeat Confers Significant Risk for FECD\nA highly significant association between expansion of the CTG18.1 trinucleotide repeat (conservatively defined as ≥50 repeats) and FECD was identified (OR = 76.47; 95% CI: 47.45–123.2; p = 5.69 × 10−71) in the white European-only portion of the cohort (n = 392; Table 1). The distribution of the CTG18.1 expansion lengths among individuals affected by FECD and age-related macular degeneration (AMD), used as an ethnically matched control population for the purpose of this study, are summarized in Figures 1A and 1B and Table 1. For the AMD cohort, 4.2% (23/550) had one expanded copy (≥50 repeats) of the CTG18.1 allele, in line with reports from other unaffected populations screened for control purposes,9, 12, 13, 14 and none were found to have two expanded alleles. In contrast, 76.4% (344/450) of the FECD cohort had one or more expanded copies of the CTG18.1 allele, of which 4.0% (18/450) had bi-allelic expansions. Interestingly, male subjects had a higher incidence of expanded CTG18.1 alleles (81.6% versus 72.8% with at least one expanded allele; Table 1) and the FECD risk associated with repeat expansion at this locus was higher in males (OR = 95.04, 95% CI: 43.08–209.70, p = 1.62 × 10−29) than in females (OR = 66.78, 95% CI: 36.79–121.20, p = 2.06 × 10−43), supporting the hypothesis that interaction of this locus with gender could be important.29\nFigure 1 Expansion of CTG18.1 Is Associated with FECD in a British and Czech Cohort\n(A) Frequency histogram comparing relative distribution of CTG repeat length in Fuchs endothelial corneal dystrophy (FECD) and age-related macular degeneration (AMD) cohorts. The longest allele detected, per individual tested, is shown. In total the FECD (blue) and AMD (white) cohorts comprised 450 and 550 individuals, respectively.\n(B) Bar chart illustrating the relative frequency of individuals with both alleles non-expanded (NE/NE), one expanded allele (E/NE), or both alleles expanded (E/E) in both the FECD and AMD cohorts. Expanded alleles are defined as ≥50 CTG repeats.\n\nCEC Cultures as a Model of FECD\nTo investigate CTG18.1 expansion-associated pathology, the occurrence of stable sense-strand-derived CUG RNA foci in fibroblast primary cultures was investigated in six independent fibroblast lines (F#1–6) derived from FECD-affected subjects with expanded CTG18.1 genotypes (Table S3). For each line, FISH was performed using a Cy3-(CAG)7 probe to determine the incidence of RNA-specific foci (Figure S3). Despite identifying multiple bright nuclear foci in a fibroblast line derived from a DM1 subject (positive control), none were detected in any of the FECD fibroblast lines investigated (Figure 2A).\nFigure 2 CTG18.1-Associated RNA Foci Occur in a Tissue-Specific Manner\n(A) Fluorescence in situ hybridization (FISH) was used to detect CUG-specific RNA foci in fibroblast and corneal endothelial cells (CECs) lines derived from three individuals with FECD and expanded TCF4 alleles. Fibroblast line BJL, non-expanded (NE) FECD, and healthy CECs were used as negative controls. Myotonic dystrophy 1 (DM1) fibroblasts were used as a positive control for foci detection (arrowheads). Each image is presented in greyscale and foci are indicated with arrowheads. Color insets (zoom panels) are presented.\n(B) Representative images of foci incidence among CECs derived from FECD-affected subjects with increasing CTG18.1 repeat lengths. Nuclei are stained with DAPI (blue). Foci detection was performed using Cy3-(CAG)7 probe (red, arrowheads). Scale bars, 10 mm.\nOn this basis, we tested the potential of using primary CECs, derived from tissue excised during endothelial keratoplasty, to investigate CTG18.1 expansion-associated pathology. The native “endothelial-like” properties of the CECs were confirmed by ICC and a variety of endothelial markers including sodium-potassium transporting adenosine triphosphatase (ATP1A1), zonula occludens 1 (ZO-1), N-cadherin, N-CAM, and CD16630, 31, 32 (Figure S4). Furthermore, the cultured CECs displayed distinctive polygonal morphology.25, 33, 34 FISH was performed, using the same Cy3-(CAG)7 probe as described above, in three distinct primary FECD CEC lines and corresponding individual-matched fibroblast lines (Figure 2A). In each instance, bright nuclear foci were detected in the FECD CECs, similar to those previously identified in corneal tissue,15, 35 whereas the corresponding fibroblasts were foci negative, suggesting that the endothelial-specific context is important and that the cultured primary CECs represent an ideal ex vivo system to investigate CTG18.1 expansion-associated corneal endothelial pathology (Figure 2A).\n\nRNA Foci Are a Biomarker of CTG18.1-Associated Pathology in CECs\nTo further explore the incidence of RNA foci, we investigated a total of 36 independent CEC lines derived from FECD-affected subjects by FISH (Table S4; Figure 2B). In summary, no foci were detected in 9 CEC lines derived from FECD-affected subjects with CTG18.1 genotype status ranging from 12/12 to 18/31, in addition to a further 4 control lines. Bright nuclear foci were clearly detected in 27 CEC lines derived from individuals with alleles ranging from 25/31 to 12/126. Interestingly, CTG18.1 allele length of 31 repeats appears to represent a critical threshold for foci occurrence in CECs as individuals with genotypes of 18/31 and 25/31 were foci negative and positive, respectively. No samples were available with an expansion in the range between 32 and 52 repeats (Figure 2B; Table S4). On this basis, we classified CEC lines selected for further experimental investigation as non-expanded (NE) if both CTG alleles contained \u003c31 repeats, whereas those with at least one allele ≥53 repeats were considered expanded (Tables S3–S9).\n\nRNA Splicing Factors MBNL1 and MBNL2 Are Sequestered by Nuclear RNA Foci\nTo determine whether RNA binding proteins were being sequestered by the RNA foci, in the cultured FECD CECs, we employed a dual FISH and ICC approach. The nuclear distribution of MBNL1 and MBNL2 was investigated in multiple CEC lines derived from individuals with and without expanded copies of the repeat (Table S5). MBNL1 displayed a diffuse nuclear localization in three non-expanded (NE) CEC lines and no foci were detected, as anticipated (Figure 3A). Striking co-localization was observed between MBNL1 and the CUG-specific RNA foci in all three independent CTG18.1 expansion-positive lines examined (Figure 3A), concordant with a previous observation made in FECD-diseased tissue.15 Similarly, the recruitment of MBNL2 to CUG-specific nuclear RNA foci was also detected in expansion-positive CECs (n = 3) and MBNL2 displayed diffuse nuclear localization in expansion-negative and control CEC lines (n = 4) (Figure 3B). These data demonstrate that RNA splicing factors are recruited to CUG RNA foci and suggest that this recruitment may induce a functional deficiency of proteins that would be predicted to alter alternative splicing events, in a similar downstream pathway to what has been established for DM1.18, 19, 20, 21, 22, 36\nFigure 3 MBNL1 and MBNL2 Are Sequestered to RNA Foci in Corneal Endothelial Cells (CECs) Derived from FECD-Affected Subjects with CTG18.1 Expansions\nRepresentative images of MBNL1 (A) and MBNL2 (B) protein nuclear localization in cell lines derived from expansion-positive FECD-affected subjects, expansion-negative FECD-affected subjects, and CECs derived from healthy individuals. RNA foci are labeled with a Cy3-(CAG)7 FISH probe and DAPI is used to stain nuclei. Co-localization of the MBNL proteins and RNA foci is represented in the bottom row of both panels. Scale bars, 10 μm.\n\nExpansion of CTG18.1 Is Associated with Altered mRNA Processing\nTranscriptomic analysis of endothelial tissue derived from individuals affected with FECD has previously suggested abnormal regulation of alternative pre-mRNA splicing in CTG18.1 expansion-positive tissue.15, 37 Therefore, we investigated whether signatures of differential splicing were present in cultured primary CECs. Total RNA was isolated from four FECD CTG18.1 expansion-positive and three FECD CTG18.1 expansion-negative CEC lines (CTG18.1 genotype listed in Table S6), in addition to four CEC lines derived from healthy control subjects. RT-PCR analysis was performed to investigate differential splicing of the three most robustly detected aberrant events observed in tissue derived from affected subjects: MBNL1 (MIM: 606516), MBNL2 (MIM: 607327), and NUMA1 (MIM: 164009).37 For each transcript analyzed, significantly different (p \u003c 0.001) patterns of splicing were observed only in the FECD expansion-positive lines compared to FECD expansion-negative and unaffected control lines, supporting the hypothesis that abnormal regulation of mRNA processing is specific to CTG18.1-related pathology (Figure 4).\nFigure 4 Altered Pre-mRNA Splicing Events Are Specific to CTG18.1 Expanded Corneal Endothelial Cells (CECs) Derived from FECD-Affected Subjects\n(A) Reverse transcriptase (RT)-PCRs reactions are shown for three selected alternative splicing events investigated for the following transcripts: MBNL1, MBNL2, and NUMA1. Samples are grouped in the following categories; controls (lanes 1–4), FECD CTG18.1 expansion positive (lanes 5–8), and FECD CTG18.1 expansion negative (lanes 9–11).\n(B) Schematic representations of RT-PCR-generated amplicons are provided for each transcript-specific reaction. Primer locations are denoted with arrows. The respective sizes of all amplified products are given.\n(C) Percentage expression of amplicons of interest (A or B) relative to total amplified products, per reaction, are presented as a mean for each respective group (C, E, and NE). Error bars represent ±1 standard deviation. p values were calculated by one-way analysis of variance (ANOVA); ∗∗∗∗p \u003c 0.0001, ∗∗∗p \u003c 0.001, ns, non-significant.\n\nASO Treatment Reduces the Incidence of Nuclear RNA Foci\nA fully 2′-O-methyl-phosohorothioate(2′-O-Me-PS) modified (CAG)7 ASO complimentary to (CUG)n repeats has previously been shown to effectively silence DMPK1 trinucleotide repeat expansion transcripts in a DM1 humanized animal model and human DM1 cell system and to reduce the number of DM1-associated RNA foci in a repeat-length-dependent manner.23, 24 We therefore tested whether transfecting with 2′-O-Me-PS-(CAG)7 ASOs, complimentary to TCF4 CTG18.1-derived transcripts, could induce a similar reduction in CUG-specific RNA foci, a biomarker of CTG18.1-related pathology, in the CEC cultures.\nA series of six FECD CEC lines, with confirmed expanded CTG18.1 genotypes (Table S7), were selected for ASO treatment. Each independent line was transfected with 200 nM of either (CAG)7 or a control ASO of identical chemical structure and of comparable length but specific to a completely unrelated sequence. A final oligo concentration of 200 nM was selected for all experiments based on optimization data (Figure S1). All lines analyzed showed a striking reduction in foci number in response to (CAG)7 ASO treatment (Figures 5 and S5). Violin plots summarize the shift in distribution of RNA foci-positive nuclei observed comparing the control ASO-treated CECs versus (CAG)7 ASO treatment (n = 6) (chi square test, χ2 = 160.78, df = 4, p = 0.001; Figure 5B). The percentage of nuclei containing zero, one, two, three, and four or more foci was analyzed, and odds ratio (OR) test confirmed that the likelihood of finding zero foci was significantly increased in cells treated with the (CAG)7 ASO compared to cells treated with the control ASO (OR = 6.2024, 95% CI, p \u003c 0.0001).\nFigure 5 ASO-Mediated Treatment of Corneal Endothelial Cells (CECs) Significantly Reduces Foci Number and Rescues MBNL1 Nuclear Localization\n(A) Foci incidence for control and (CAG)7 antisense oligonucleotide (ASO)-treated FECD-affected subject-derived CECs. The graph shows percentages of nuclei that contain 0, 1, 2, 3, and 4 or more foci after treatment with the different ASOs. Mean ± SD are represented in each case (n = 6).\n(B) Violin plots representing the distribution of the frequencies of each group of nuclei (containing 0, 1, 2, 3, and 4 or more foci) in cells treated with control and (CAG)7 ASOs (n = 6).\n(C) Representative images of ASO treatment on foci incidence. Sense RNA foci detection using Cy3-(CAG)7 probe (red). Scale bars, 10 μm.\n(D) Number of MBNL1 puncta present per nucleus when cells were treated with either a control ASO or with the (CAG)7 ASO. The mean ± SD from 4 independent expansion-positive CEC lines, where a minimum of 95 nuclei were evaluated per line. p values were calculated using an unpaired two tailed t test; ∗∗∗∗p \u003c 0.0001.\n(E) Representative example illustrating the reduction of MBNL1 puncta and the changes in MBNL1 localization when cells were treated with control and (CAG)7 ASOs. In all cases CECs were treated with 200 nM (CAG)7 or control ASO for 24 hr. Scale bars, 25 μm.\n\nASO Treatment Rescues MBNL1 Nuclear Localization\nTo determine the effect of (CAG)7 ASO treatment on MBNL1 distribution, four independent expansion-positive CEC lines (Table S8) were selected and ICC was performed, following treatment with either the (CAG)7 or control ASO. In all four lines treated with the (CAG)7 ASO, there was an obvious redistribution of MBNL1 and a significant reduction in MBNL1-positive puncta (p \u003c 0.0001), demonstrating that ASOs can rescue aberrant MBNL1 nuclear localization associated with the CTG18.1 expansion-related pathology (Figures 3A, 5D, 5E, and S2).\n\nASO Treatment Reduces Aberrant mRNA Processing\nWe investigated the potential of (CAG)7 ASO treatment to reverse the shift in alternative splicing events previously observed in expansion-positive CEC lines (Figure 4). Ten FECD expansion-positive CEC lines (Table S9) were treated with either the (CAG)7 or control ASO. Following treatment, RNA was extracted and RT-PCR was performed. For both MBNL1 and MBNL2 there was a highly significant (p \u003c 0.0001) shift in the relative proportions of alternatively spliced transcripts after (CAG)7 ASO treatment (n = 10) toward the control CEC spliceoform distribution (Figure 6). Furthermore, a significant (p ≤ 0.05) shift in the relative proportions of alternatively spliced transcripts was also demonstrated for NUMA1 (n = 10) (Figure 6). These data therefore demonstrate that the (CAG)7 ASO is effective at rescuing the differential splicing events associated with CTG18.1 expansion-related pathology (Figures 4 and 6).\nFigure 6 ASO Treatment Rescues Differential Splicing Events Underlying CTG18.1-Associated Pathology\n(A) Representative RT-PCR images for non-treated samples with either expanded (E) or control (C) (non-expanded CTG18.1 genotypes) are shown on the far left, for reference purposes. Representative RT-PCRs are shown for three independent ASO-treated CEC lines.\n(B) Schematic representations are provided for each respective reaction. Primer locations are denoted with arrows. The respective sizes of all amplified products are given.\n(C) Percentage expression of amplicons of interest (A or B) relative to total amplified products, per reaction, are presented as a mean for control-ASO versus (CAG)7 ASO-treated groups. The mean from 10 independent CEC lines ± SD. In all cases CECs were treated with 200 nM (CAG)7 or control ASO for 24 hr. p values were calculated using a paired two tailed t test for each event investigated; ∗∗∗∗p \u003c 0.0001, ∗p \u003c 0.05.\n\nIntraocular Injections Enable Effective In Vivo Delivery of ASOs to the Corneal Endothelium\nTo assess the likely accessibility of ASOs to the corneal endothelium, C57B16 mice were injected intravitreally with varying concentrations (0.025, 0.01, and 0.005 mg) of Cy-3-labeled 2′Ome-PS-(CAG)7. Confocal fluorescence microscopy revealed that the ASO was present in corneal endothelium, keratocytes, and stroma, specifically accumulating in both the nuclear and perinuclear region of both the endothelial and stromal cells. This localization increased in a dose-dependent manner, showing stronger accumulation in cell layers at 48 hr post-dosing (Figure S6). ASOs with identical chemistry, following intraocular injection, have previously been shown in vivo to display a similar peri-nuclear and nuclear localization in other cell types in conjunction with effective molecular activity.38 These data highlight the potential for the (CAG)7 ASO to target CECs in vivo, an essential prerequisite for the effective delivery of an ASO-mediated FECD therapy."}