Protein Analysis of the TGFBIR124H Mouse Model Gives Insight into Phenotype Development of Granular Corneal Dystrophy

TGFBI R124H mutant allele silencing in granular corneal dystrophy type 2 using topical siRNA delivery

In recent years, success has been achieved in treating several eye conditions with oligonucleotide-based therapies. Herein, we outline the experimentation involved in progressing selection and development of a lead therapeutic siRNA for R124H mutation of TGFBI gene which causes Granular Corneal Dystrophy Type 2 (GCD2/Avellino CD). Firstly, a series of siRNA designs, generated by a gene walk across the R124H TGFBI mutation site, were tested and a lead siRNA identified. The lead siRNA was delivered into an immortalised human corneal epithelial cell line to assess on-target efficacy and off-target effects. The in vivo efficacy of the lead R124H TGFBI siRNA, complexed with Bio-Courier technology, silicon stabilized hybrid lipid nanoparticles (sshLNP), was assessed in a mouse model of GCD2 which expressed the human R124H TGFBI transgene. Following topical siRNA application for 5 consecutive days, expression of the R124H mutant TGFBI transgene was measured and shown to be reduced by 22.4 % (± 15.7 %, p < 0.05). We investigated gene expression in the mouse cornea and showed expression of murine Tgfbi was 20-fold lower than TGFBI in human cornea, and expression of the mutant TGFBI transgene was a further 3-fold lower. This estimated 60-fold lower mutant transgene expression may explain the low frequency of corneal deposits observed in this mouse model, limiting its usefulness to test whether siRNA silencing is capable of phenotypic improvement or regression of GCD2/Avellino corneal dystrophy. We assessed WT TGFBI silencing in human primary corneal epithelial cells (PCEC) derived from human corneal limbal biopsy material, which express TGFBI at a similar level to human corneal biopsy. We demonstrated that a single 100 nM siRNA treatment, delivered by the sshLNP to the primary human corneal epithelial cells, gave 26.6 % (± 6.6 %, p < 0.001) reduction in TGFBI mRNA and a 15.4 % (±10.5 %, p < 0.05 %) reduction in TGFBi protein after 48 h. In consideration of the mutant gene expression levels in existing models of GCD2 disease, an ex vivo model of mutation-expressing primary corneal epithelial cells generated from corneal limbal biopsies from GCD2 patients would be more suitable than existing transgenic mouse models for future pre-clinical work in the development of gene silencing therapies for corneal dystrophies.

Mimicking TGFBI Hot-Spot Mutation Did Not Result in Any Deposit Formation in the Zebrafish Cornea

PURPOSE

Mutations in transforming growth factor beta-induced (TGFBI) protein are associated with a group of corneal dystrophies (CDs), classified as TGFBI-associated CDs, characterized by deposits in the cornea. Mouse models were not proper in several aspects for modelling human disease. The goal of this study was to generate zebrafish mutants to investigate the corneal phenotype and to decide whether zebrafish could be a potential model for TGFBI-associated CDs.

METHODS

The conserved arginine residue, codon 117, in zebrafish tgfbi gene was targeted with Clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 method. Cas9 VQR variant was used with two target-specific sgRNAs to generate mutations. The presence of mutations was evaluated by T7 Endonuclease Enzyme (T7EI) assay and the type of the mutations were evaluated by Sanger sequencing. The mutant zebrafish at 3 months and 1 year of age were investigated under the microscope for corneal opacity and eye sections were evaluated histopathologically with hematoxylin-eosin, masson-trichrome and congo red stains for corneal deposits.

RESULTS

We achieved indel variation at the target sequence that resulted in p.Ser115_Arg117delinsLeu (c. 347_353delinsT) by nonhomology mediated repair in F1. This zebrafish mutation had the potential to mimic two disease-causing mutations reported in human cases previously: R124L and R124L + del125-126. Mutant zebrafish did not show any corneal opacity or corneal deposits at 3 months and 1 year of age.

CONCLUSION

This study generated the first zebrafish model mimicking the R124 hot spot mutation in TGFBI-associated CDs. However, evaluations even at 1 year of age did not reveal any deposits in the cornea histopathologically. This study increased the cautions for modelling TGFBI-associated CDs in zebrafish in addition to differences in the corneal structure between zebrafish and humans.

Mutation-induced dimerization of transforming growth factor-β-induced protein may drive protein aggregation in granular corneal dystrophy

Protein aggregation in the outermost layers of the cornea, which can lead to cloudy vision and in severe cases blindness, is linked to mutations in the extracellular matrix protein transforming growth factor-β-induced protein (TGFBIp). Among the most frequent pathogenic mutations are R124H and R555W, both associated with granular corneal dystrophy (GCD) characterized by the early-onset formation of amorphous aggregates. The molecular mechanisms of protein aggregation in GCD are largely unknown. In this study, we determined the crystal structures of R124H, R555W, and the lattice corneal dystrophy-associated A546T. Although there were no changes in the monomeric TGFBIp structure of any mutant that would explain their propensity to aggregate, R124H and R555W demonstrated a new dimer interface in the crystal packing, which is not present in wildtype TGFBIp or A546T. This interface, as seen in both the R124H and R555W structures, involves residue 124 of the first TGFBIp molecule and 555 in the second. The interface is not permitted by the Arg124 and Arg555 residues of wildtype TGFBIp and may play a central role in the aggregation exhibited by R124H and R555W in vivo. Using cross-linking mass spectrometry and in-line size exclusion chromatography-small-angle X-ray scattering, we characterized a dimer formed by wildtype and mutant TGFBIps in solution. Dimerization in solution also involves interactions between the N- and C-terminal domains of two TGFBIp molecules but was not identical to the crystal packing dimerization. TGFBIp-targeted interventions that disrupt the R124H/R555W crystal packing dimer interface might offer new therapeutic opportunities to treat patients with GCD.