Mutation update of transcription factor genes FOXE3, HSF4, MAF, and PITX3 causing cataracts and other developmental ocular defects

Abstract

Mutations in the transcription factor genes FOXE3, HSF4, MAF, and PITX3 cause congenital lens defects including cataracts that may be accompanied by defects in other components of the eye or in nonocular tissues. We comprehensively describe here all the variants in FOXE3, HSF4, MAF, and PITX3 genes linked to human developmental defects. A total of 52 variants for FOXE3, 18 variants for HSF4, 20 variants for MAF, and 19 variants for PITX3 identified so far in isolated cases or within families are documented. This effort reveals FOXE3, HSF4, MAF, and PITX3 to have 33, 16, 18, and 7 unique causal mutations, respectively. Loss-of-function mutant animals for these genes have served to model the pathobiology of the associated human defects, and we discuss the currently known molecular function of these genes, particularly with emphasis on their role in ocular development. Finally, we make the detailed FOXE3, HSF4, MAF, and PITX3 variant information available in the Leiden Online Variation Database (LOVD) platform at https://www.LOVD.nl/FOXE3, https://www.LOVD.nl/HSF4, https://www.LOVD.nl/MAF, and https://www.LOVD.nl/PITX3. Thus, this article informs on key variants in transcription factor genes linked to cataract, aphakia, corneal opacity, glaucoma, microcornea, microphthalmia, anterior segment mesenchymal dysgenesis, and Ayme-Gripp syndrome, and facilitates their access through Web-based databases.

Keywords: Ayme-Gripp syndrome; LOVD; anterior segment mesenchymal dysgenesis; aphakia; cataract; microcornea.

Figure 1. Schematic of FOXE3 gene, protein domains, and position of mutations

(A) The human FOXE3 gene (top) spans a single exon (E1) that encodes the FOXE3 protein (bottom), which contains the forkhead DNA-binding domain (pink). Nucleotide positions for cDNA and genomic DNA in the reference sequence (hg38) and the corresponding amino acid positions in the protein sequence are given. The UCSC genome browser (https://genome.ucsc.edu) was used to obtained cDNA, genomic and protein position information from FOXE3 human GRCh38/hg38 assembly (reference sequence NM_012186, uc001crk.3 at chr1: 47,416,072–47,418,052). Total 52 mutation studies, of which 33 represent unique mutations identified in independent studies are indicated as flags. These include 6 frameshift, 23 missense, 3 nonsense and 1 nonstop mutation. Nucleotide changes in DNA and amino acid changes in protein sequence are indicated. Changes in cDNA that are not documented in original article are represented as (*) or (**). Variants linked to eye defects are shown above the gene (and below the protein) and those associated with thoracic aortic aneurysms and acute aortic dissections are shown below the gene (and above the protein).

Figure 2. Multiple sequence alignment of FOXE3 protein sequences

Protein sequences for FOXE3 downloaded from the UCSC browser for Homo sapiens, Mus musculus, Gallus gallus, Xenopus tropicalis and Danio rerio shows amino acid conservation across different vertebrate species. Amino acid changes caused by specific human mutations are indicated by yellow arrowheads. The sequence consensus is presented for each amino-acid and the forkhead domain is indicated by green arrow.

Figure 3. Schematic of HSF4 gene, protein domains, and position of mutations

(A) The human HSF4 gene spans fifteen exons (E1-E15) that encode the HSF4 protein (bottom). While there are multiple HSF isoforms, the specific isoform (HSF4 isoform b) that has high expression in the lens is described here because of its relevance to the cataract defect. Exons 3 through 6 encode the DNA binding domain (pink), exons 6 through 8 encode an oligomerization domain termed hydrophobic repeat (HR- A/B) (blue) and exons 12 through 14 encode a region termed as downstream of hydrophobic repeat (DHR) (purple). HSF4b is derived by alternative mRNA splicing of exons 10 and 11 (exon 8 and 9 in the originally reported study). Nucleotide positions for cDNA and genomic DNA in the reference sequence (hg38) and the corresponding amino acid positions in the protein sequence are given. The UCSC genome browser (https://genome.ucsc.edu) was used to obtained cDNA, genomic and protein position information from HSF4 human GRCh38/hg38 assembly (reference sequence NM_001040667.2 at chr16:67163385–67169945). Total 18 mutations, of which 16 represent unique mutations identified in independent studies are indicated as flags. These include 2 frameshift, 13 missense and 1 nonsense mutation. Nucleotide changes in DNA and amino acid changes in protein sequence are indicated. In Table 2 * indicates nucleotide and amino acid positions as indicated in the original article

Figure 4. Multiple sequence alignment of HSF4 protein sequences

Protein sequences for HSF4 downloaded from UCSC browser for Homo sapiens, Mus musculus, Gallus gallus, Xenopus tropicalis and Danio rerio shows amino acid conservation across different vertebrate species. Amino acid changes caused by specific human mutations are indicated by yellow arrowheads. The sequence consensus is presented for each amino-acid. HSF-DNA binding domain and vertebrate heat shock transcription factor domains are indicated by green arrow.

Figure 5. Schematic of MAF gene, protein domains, and position of mutations

(A) The human MAF gene (top) spans two exons (E1-E2) that encode the HSF4 protein (bottom). Exon 1 encodes the transactivation domain (pink) and the bZIP domain (blue). Nucleotide positions for cDNA and genomic DNA in the reference sequence (hg38) and the corresponding amino acid positions in the protein sequence are given. The UCSC genome browser (https://genome.ucsc.edu) was used to obtained cDNA, genomic and protein position information from MAF human GRCh38/hg38 assembly (reference sequence NM_005360, uc002ffm.4 at chr16:79,593,838–79,600,714). Total 20 mutations, of which 18 represent unique mutations identified in independent studies are indicated as flags. These include 17 missense and 1 translocation mutation. Nucleotide changes in DNA and amino acid changes in protein sequence are indicated.

Figure 6. Multiple sequence alignment of MAF protein sequences

Protein sequences for MAF downloaded from UCSC browser for Homo sapiens, Mus musculus, Gallus gallus, and Xenopus tropicalis shows amino acid conservation across different vertebrate species. Amino acid changes caused by specific human mutations are indicated by yellow arrowheads. The sequence consensus is presented for each amino-acid. Transactivation domain and bZIP domain are indicated by green arrow.

Figure 7. Schematic of PITX3 gene, protein domains, and position of mutations

(A) The human PITX3 gene (top) consists of four exons (E1-E4), of which exons 2 through 4 encode the PITX3 protein. Exons 2 through 4 encode the homeodomain (pink) and exon 4 encodes the OAR domain (blue). Nucleotide positions for cDNA and genomic DNA in the reference sequence (hg38) and the corresponding amino acid positions in the protein sequence are given. The UCSC genome browser (https://genome.ucsc.edu) was used to obtained cDNA, genomic and protein position information from PITX3 human GRCh38/hg38 assembly (reference sequence NM_005029, uc001kuu.2 at chr10:102,230,186–102,241,474). Total 19 mutations, of which 7 represent unique mutations identified in independent studies are indicated as flags. These include 6 frameshift and 1 missense mutation.

Figure 8. Multiple sequence alignment of PITX3 protein sequences

Protein sequences for PITX3 downloaded from UCSC browser for Homo sapiens, Mus musculus, Gallus gallus, Xenopus tropicalis and Danio rerio shows amino acid conservation across different vertebrate species. Amino acid changes caused by specific human mutations are indicated by yellow arrowheads. The sequence consensus is presented for each amino-acid. Homeodomain and OAR domain are indicated by green arrow.