Family

The proband, born to non-consanguineous parents of Mexican origin, sought consultation at the age of 31 for whole-exome sequencing (WES)-based molecular diagnosis of ocular coloboma with microcornea at the Hospital de la Ceguera in Mexico City (APEC). The patient, her parents, brother, sister, and her daughter were included in the study for segregation analysis of the FZD5 c.800 C > T, p.Pro267Leu missense variant. All studies were conducted in accordance with the Helsinki Declaration of 1964. Prior to the molecular study, all participants or their legal guardians provided signed informed consent forms, including an iconographic informed consent form for the collection and use of clinical photographs.

Whole exome sequencing

Genomic DNA was extracted from peripheral venous blood using a QIAamp DNA kit (Qiagen, Victoria, Australia) according to the manufacturer’s protocol. Whole exome sequencing (WES) was performed at 3 billion, Inc (Seoul, South Korea) to capture all known coding regions of 19,433 known human genes using xGen Exome Research Panel v2 (Integrated DNA Technologies, Coralville, Iowa, USA). Sequencing was performed by Novaseq 6000 (Illumina, San Diego, CA, USA) as 150 bp paired-end reads. Sequencing data was analyzed using an internal bioinformatics pipeline ‘EVIDENCE’ (Seo et al. 2020). In brief, the FASTQ file was aligned to the Genome Reference Consortium Human Build 37 (GRCh37) and Revised Cambridge Reference Sequence (rCRS) of the mitochondrial genome using BWA-MEM, generating 109.83 mean depth-of-coverage within the 34,366,188 bases of the captured region. Approximately 98.7% of the targeted bases were covered to a depth of ≥ 20x. Variants were called using GATK v.3, annotated with Ensembl Variant Effect Predictor (VEP) and classified and filtered according to the American College of Medical Genetics and Genomics (ACMG) guideline (Richards et al. 2015). The final list of variants was manually reviewed by medical geneticists. Pathogenicity of each variant was further reviewed using VarSome (Kopanos et al. 2019).

Sanger validation and segregation analysis

The confirmation of the presence of the highest candidate variant identified by WES, FZD5 c.800C > T (p.Pro267Leu; NM_003468.4), in the proband’s DNA and its segregation analysis in the family were studied by PCR amplification and Sanger sequencing, using primers designed from intronic sequences flanking exon 2: Forward: 5’-GTCACACCCGCTCTACAACA-3’ and Reverse: 5’-AGGAAGACGATGGTGCACAG-3’ and the BigDye® Terminator v3.1 on an ABI 3500XL Genetic Analyzer (Applied Biosystems, Thermo Fisher Scientific, Courtaboeuf, France) as described in Gerber et al. 2021(Gerber et al. 2021). Data were analyzed using the ABI Sequencing Analysis 6 Software.

Clinical assessment of the Pro267Leu variant carriers

The proband, homozygous for the variant, underwent comprehensive ophthalmological and systemic clinical examinations. Additionally, family members carrying the variant in heterozygosity (proband’s father, mother, brother, sister, and daughter) underwent thorough ophthalmic evaluations.

Multiple sequence alignment of FZD5 across species and other members of the FZD family

Eighteen orthologs of human FZD5 were retrieved using the OrthoInspector website (https://lbgi.fr/orthoinspector/), (Linard et al. 2011; Nevers et al. 2019). The nine other Frizzled sequences (FZD1-FZD10) were aligned using PipeAlign2 (https://lbgi.fr/pipealign/), (Plewniak 2003). The percentage identity matrix of the ten FZD sequences was calculated using ClustalX2. Figure 1B was created using Jalview (Waterhouse et al. 2009).

Prediction of the protein stability change upon single variation

The cryoEM structure (PDB code 6WW2, chain R), the crystal structure of FZD4 (PDB code 6BD4, chain A), and the AlphaFold2 models of human FZD4 and FZD5, as well as zebrafish Fzd5, were submitted to the DynaMut2 (Rodrigues et al. 2021) and the PremPS web server (Chen et al. 2020). This analysis aimed to predict the protein stability change upon modification of proline to leucine and arginine (Pro267 in human FZD5, Pro251 in human FZD4, and Pro277 in zebrafish Fzd5) on the structure of FZD5.

Molecular dynamics web server-based simulation of mutant and wildtype FZD5

The AlphaFold2 model of human FZD5 was submitted to the CABS-flex 2.0 web server for rapid simulations to assess the flexibility of the structure, with a particular focus on predicting the flexibility level of the first intracellular loop (ICL1) within the context of the full-length structure. Default parameter values from the web server were utilized. Figures were created using PyMOL Molecular Graphics System, Version 3.0 Schrödinger, LLC.

Generation of fzd5 missense and loss-of-function forms in zebrafish

The zebrafish c.830C > T (p.Pro277Leu; mi-fzd5) variant corresponding to the human c.800C > T (p.Pro267Leu) substitution and the c.39del (E13Efs*30) mutation modeling a complete FZD5 loss-of-function (lof-fzd5) were introduced into the zebrafish cDNA through site-directed mutagenesis by high-fidelity PCR amplification (Phusion polymerase) using the pCS2-fzd5 plasmid containing the wild-type (wt) zebrafish fzd5 cDNA (wt-fzd5) (Cavodeassi et al. 2005) as template and primers specific to the 830C > T and c.39del mutations (Forward: 5’- AGCGCTTCAAATATCtAGAGCGGCCGATTAT-3’ and Reverse: 5’- ATAATCGGCCGCTCTaGATATTTGAAGCGCT-3 and Forward: 5’- ACCATGGAACCTCAGGGATGCACCTGG-3’ and Reverse: 5’- CTGAGGTTCCATGGTGAAATGATGCTCG-3’respectively). Following amplification, template DNA was eliminated by digestion with the methylase-sensitive DpnI restriction enzyme for 1 h at 37 °C. XL1-Blue competent bacteria (Agilent Technologies) were transformed using the digested products and plated on LB Agar supplemented with ampicillin (100 µg/mL). Mutant variants from single colony minipreps were validated through Sanger sequencing using primers as described previously (Forward: 5’- CCTAACTGTGCACTGCCTTG-3’ and Reverse: 5’-ACCCCAGTGACACAAACAGA-3’for the mi-Fzd5 variant and Forward: 5’-ATTTAGGTGACACTATAG-3’ and Reverse: 5’-CTCTGGCCACTCAAACCCAT- for the lof-Fzd5 variant).

Zebrafish lines and husbandry

AB/tupl wildtype zebrafish strains were maintained and bred according to standard procedures (Aleström et al. 2020). All experiments conform to the guidelines from the European Community Directive and British legislation (Animal (Scientific Procedures) Act 1986) and were conducted under the authority of Project Licence PP5056153 granted to Florencia Cavodeassi by the British Home Office authorities.

Synthesis of mRNA, embryo microinjections and in situ hybridization

Capped wt-fzd5, mi-fzd5 and lof-fzd5 mRNAs were synthesized using SP6 mMessage Machine (Ambion), following manufacturer’s instructions. 20 and 40 picograms/embryo (pg/emb) were injected into one-cell stage fertilized embryos and allowed to develop at 30 °C until the end of gastrulation (10 h post-fertilisation, hpf), as previously described (Cavodeassi et al. 2005). Double axes were visualized using antisense RNA probes for mRNA detection of rx3 (eye field marker) and pax2.1 (midbrain marker), synthesized with RNA-polymerases (Promega) and DIG-labelled nucleotides (Roche) following manufacturer’s instructions. Whole-mount in situ hybridization was performed as previously described (Hernández-Bejarano et al. 2015, 2022). Embryos were mounted in glycerol and imaging was performed under a Leica stereomicroscope connected to an IDS digital camara operated by IDS Software Suite.

Analysis of protein localization

Fzd5 protein localization was examined in zebrafish embryos using a fzd5-RFP C-terminal fusion. A mi-fzd5-RFP fusion was generated by subcloning of the EcoRI-BspEI fragment of mi-fzd5 into EcoRI-BspEI-digested pCS2-wt-fzd5-RFP. pCS2-wt-fzd5-RFP and pCS2-mi-fzd5-RFP were used as templates to synthesize mRNA for injection as described above. 200 pg/emb were injected into one-cell stage fertilized embryos, allowed to develop at 30 °C until dome stage (4.5 hpf), and fixed overnight in 4% paraformaldehyde. Embryos were briefly washed in PBS + Triton 0.3% and incubated with phalloidin-FITC (at 0.5µM to detect subcortical actomyosin) and Hoechst (at 1 µg/ml, to detect DNA) in PBS + Triton 1%+DMSO 1% for 4 h at room temperature, briefly washed in PBS and mounted in 1% low-melt-point agarose for imaging.

Imaging was performed in a Nikon A1R inverted confocal microscope with a 40X dry lens and images were processed with Nikon NIS Elements C software.

TopFlash assays

pCS2-wt-fzd5-RFP and pCS2-mi-fzd5-RFP were used as templates to generate myc-tagged forms for transfection, using the the NEB Q5 site-directed mutagenesis kit (E0554; forward primer AGCGAAGAAGATCTGTAGAACTATAGTGAGTCGTATTAC; reverse primer AATCAGTTTCTGTTCGAGGACATGTGATGAG).

HEK293 cells were maintained in DMEM with 10% FCS at 37 °C in a humidified atmosphere of 5% CO2. Transfections were carried out using Polyethylenimine as described (Longo et al. 2013). Cells for TOPFlash assays were plated in quadruplicate on 24-well plates and transfected with M50 Super 8X TOPFlash along with pRLTK (Promega) for normalisation. Activation of the Wnt-pathway was achieved by co-transfection of lrp6 and fzd constructs as described (Hua et al. 2018). Cells were processed 48 h after transfection with the Dual-Luciferase® Reporter Assay System (Promega). Luminescence readings were made with a GloMax® Discover Microplate Reader and data analysed using Excel and JASP software (JASP Team, 2024) JASP (Version 0.17.3) on macOS 10.15.7.