Introduction

Cataract, the cloudiness of the lens is the most frequent cause of blindness worldwide, representing almost half of all causes of blindness globally: the WHO estimates that 18 million people are bilaterally blind from cataract.1 Congenital cataracts are seen in 1–15 per 10 000 births in the United Kingdom and are a significant cause of childhood visual impairment.2, 3 Congenital cataract can occur in isolation or in association with other non-ocular manifestations and is a predominant feature in >200 genetic disorders. Congenital cataract may be familial and display considerable genotypic and phenotypic heterogeneity.4 Inheritance is most commonly autosomal dominant (AD), usually with complete penetrance but with highly variable expressivity. The phenotypic classification of the cataract depends on the position and type of the lens opacity such as: anterior polar, posterior polar, nuclear, lamellar, coralliform, blue-dot (cerulean), cortical, pulverulent, polymorphic, complete cataract, and posterior nuclear cataract.5, 6 Significant progress has been made in identifying the molecular genetic basis of human cataract. More than 40 genes have been recruited, including the large family of crystallins encoding transparent intracellular lens proteins, membrane gap junction proteins (connexins), water channel proteins (aquaporins), solute carrier protein, cytoskeletal proteins, transmembrane proteins, lens intrinsic membrane protein, chromatin-modifying protein-4B, receptor tyrosine kinase gene EPH receptor A2,7 an endoplasmic reticulum membrane-embedded protein, Wolframin,8 and transcription factor genes (eg, PAX6, FOXE3, EYA1, MAF, PITX3, and HSF4).7

Herein we report a novel mutation in the heat-shock transcription factor 4 (HSF4) gene causing an isolated autosomal-dominant lamellar cataract in a five-generation British family.

Methods

Phenotyping

In this study, the family was identified through the proband attending the Genetic Service at Moorfields Eye Hospital, London, UK. Local ethics committee approval was obtained and all of the participants gave written informed consent. All the family members underwent full ophthalmic examination, including slit lamp examination; all affected individuals were diagnosed as having isolated lamellar cataract.

Whole exome sequencing (WES) and bioinformatics analysis

Genomic DNA was extracted from EDTA sequestered blood samples taken with informed consent and local ethical approval using the Nucleon II DNA Extraction Kit (Scotlab Bioscience, Strathclyde, Scotland, UK). The DNA was sequenced at AROS Applied Biotechnology (Aarhus, Denmark). Exon capture and target enrichment was performed using the SureSelectXT Human All Exon V5 kit (Agilent, Santa Rosa, CA, USA). Paired-end sequencing was ran on an Illumina Hiseq 2500 high-throughput sequencer (Illumina, San Diego, CA, USA) generating mean exome coverage of 50 ×. Raw data in fastq format was analysed using the Phenopolis platform.9 The short read sequence data were aligned using novoalign (version 3.02.08). Variants and indels were called according to GATK best practices (joint variant calling followed by variant quality score recalibration).10 The variants were then annotated using the Variant Effect Predictor (VEP).11 Variants were then filtered to only contain novel variants not present in public control databases Kaviar12 and gnomAD (http://gnomad.broadinstitute.org/) and predicted to be moderately or highly damaging according to the VEP. Cosegregation of the filtered variants in both affected individuals was then performed. Finally, the list of variants was further screened using Phenopolis for genes associated with the Human Phenotype Ontology12, 13 term ‘lamellar cataract’ (HP:0007971) according to OMIM.14

Structural bioinformatics

The protein structure of HSF4 was analysed using SWISS-MODEL15 (https://swissmodel.expasy.org/repository/uniprot/Q9ULV5). The best PDB16 match, with a match of 75%, was the structure of the DNA-binding-domain (DBD) HSF1 (PDB ID 2LDU, solved with NMR by the Northeast Structural Genomics Consortium Target HR3023C 2011). An X-ray-solved structure of HSF1 DBD interacting with DNA was also found in PDB (PDB ID 5D5U17. All structures were downloaded in PDB format and analysed using Pymol (version 1.8) locally. The Pymol Mutagenesis tool was used to visualise the effect of the mutation on the HSF1–DNA complex.

Sanger sequencing

Bi-directional direct Sanger sequencing was performed to validate the variant identified by next-generation sequencing. Genomic DNA was amplified by PCR using GoTaq 2X master mix (AB gene; Thermo Scientific, Epsom, UK) and HSF4-specifc primers designed with Primer3 http://bioinfo.ut.ee/primer3-0.4.0/primer3/. PCR conditions were as follows: 94 °C for 10 min of initial denaturation followed by 30 cycles of amplification of 30 s at 94 °C, 30 s at 60 °C, and 45 s at 72 °C. After the PCR products were reacted with BigDye Terminator v3.1, they were run on ABI 3730 Genetic Analyzer (both from Applied Biosystems, Foster City, CA, USA) and analyzed using SeqMan Pro (version 8.0.2 from DNASTAR) sequence analysis. After validating the variant, family segregation was performed in all the individuals. Further, 96 unrelated patients from our autosomal-dominant congenital cataract (ADCC) panel were screened for this variant by bi-directional Sanger sequencing to look for any other occurrence of the mutation.

Results

A large five-generation pedigree comprising 16 members, including 9 affected individuals, 3 unaffected individuals, and 4 spouses, were examined and all affected members had evidence of lamellar cataract (Figure 1).

Figure 1
figure 1

Abridged pedigree of the British family with lamellar cataract. Squares and circles symbolize males and females, respectively. Open and filled symbols indicate unaffected and affected individuals, respectively.

Full size image

WES was undertaken in two affected individuals (IV-1, IV-6). Variant annotation and filtering was performed using the Phenopolis platform. From a total of 149 133 variants in the family, 43 441 were found to co-segregate in the two affected individuals. A filter for novel variants that were moderately or highly damaging was then applied and returned 79 variants. After gene panel screening, a rare heterozygous damaging variant, NM_001040667.2:c.190A>G, in HSF4 on chromosome 16q22.1 was identified. Direct sequencing confirmed that the missense mutation c.190A>G in exon 4 of HSF4 co-segregated with all the affected members of the family (Figure 2).

Figure 2
figure 2

Sequence analysis of HSF4. An unaffected individual (upper chromatogram illustrates a normal control and a missense mutation c.190A>G shown in few affected members of the family with lamellar cataract.

Full size image

This single base change is predicted to result in a lysine (K) to glutamic acid (E) amino-acid substitution (p.K64E) in the highly conserved DBD, as shown on the protein structure of HSF4 (Figure 3a). The structural change on HSF4 induced by the mutation was estimated to be negligible according to the difference in the PDB models returned by SWISS-MODEL (root mean square deviation of 0.009). However, the homologous amino acid has a hydrogen bond to the DNA molecule in the structure of DNA-binding complex of the homologous protein HSF1 (Figure 3b). The mutation from a lysine to a glutamic acid is likely to affect that bond, as the lysine is positively charged, whereas glutamic acid is negatively charged, which could have some effect on the binding affinity to the DNA molecule (Figure 3c). Further in silico analysis and experimental work is required to verify this claim.

Figure 3
figure 3

Structure of the DNA-binding domain of HSF4. (a) Known mutations and their sidechains are displayed in red. The K64E mutation is just downstream of an alpha helix and is a point of contact with the DNA molecule. (b and c) Point of contact of the amino acid (red) in the DBD interacting with the DNA molecule (orange and blue). (b) In the wild-type protein, the amino acid is lysine. (c) In the mutant protein, the amino acid is a glutamic acid.

Full size image

Discussion

Here we report a missense mutation NM_001040667.2:c.190 A>G in the HSF4 gene on chromosome 16q22.1 in a five-generation English pedigree with congenital lamellar cataract. All the affected family members had bilateral congenital cataract and age at onset varied from birth to age 5–7 years. One of the patients IV-6 also had glaucoma.

Lamellar cataract is also referred to as zonular cataract and is one of the commonest phenotypes of ADCC. The inner fetal nucleus is made up of a clear lens surrounded by an opacified shell that is, in turn, surrounded by clear cortex, which may contain opacities referred to as ‘riders’ or ‘cortical spokes’. Lamellar cataract represents a disturbance in the lens development at a particular time and the cataractous ‘shell’ varies in size according to the stage of fetal development at which the disturbance occurs.5 HSF4 is expressed in both cell types (epithelial cells and fibre cells) at an early stage of lens development and is required for normal cell growth and differentiation of these two cell types.18

HSF4 belongs to the family of heat-shock transcription factors that regulate the expression of heat-shock proteins in response to various stresses, such as high temperature, infection, and free radicals, and also in development. HSF4 expression is tissue specific and has two splice forms, (i) HSF4a, which inhibits transcription of other heat-shock factor genes by binding directly to the heat-shock element and (ii) HSF4b, which contains an additional 30 amino acids and acts as an activator of transcription. HSF4 comprises 15 exons and encodes a heat-shock transcriptional protein of 493 amino-acid residues with a DBD, an amino-terminal hydrophobic repeat (HR/A-B), an isoform-specific region and downstream of hydrophobic repeat (DHR).19, 20

As described in Table 1, 16 different mutations in HSF4 have been identified in different ethnicities for both for autosomal-dominant and -recessive congenital cataract, as well as age-related cataract.21, 22 Eight of these have been found in the Chinese population, three autosomal-recessive mutations in Pakistani families,23 and one splice site mutation in a large Tunisian family.24 Recently, Benham et al25 reported a novel homozygous mutation in an Iranian family causing congenital cataract. So far, only two mutations have been found in Europeans; in Danish families, p.R119C was a recurrent mutation; and a novel mutation p.H35Y in a British family with no description of the cataract phenotype.26 Each of these mutations were found in autosomal-dominant cataract, including our novel heterozygous mutation p.K64E in a large British pedigree with lamellar cataract.

Table 1 Published mutations in HSF4 that cause cataract
Full size table

Interestingly, all known autosomal-dominant mutations in HSF4 lie within the α-helical DBD, whereas the recessive mutations lie outside this highly conserved functional domain.27

These observations support that mutations in the DBD are dominant negative rather than loss of function. The binding affinity of HSF4 to the DNA might be perturbed that could interfere with the wild-type molecule. This could potentially disrupt the transcription factor ‘orchestra’ in eye development, leading to abnormal lens cells conducive to protein aggregates and hence the onset of cataract. It has also been confirmed that HSF4 regulates lens fibre cell differentiation by modulating the expression of certain lens structural proteins, such as lens-specific crystallins, beaded filament proteins, and fibroblast growth factors.18 Further studies have shown that HSF4 has a crucial part in the de-nucleation of lens fibre cells through regulating DNase 2β expression level and DNase activity.28 Merath et al29 have demonstrated that the functional loss associated with autosomal-recessive HSF4 mutations is due to a loss of regulatory domains situated at the C-terminal end, indicating that the transcriptional activation of HSF4 is mediated by interactions between activator and repressor domains within the C-terminal end. Furthermore, the Hsf4 knockout mice, which demonstrate a cataract phenotype, have abnormal lens development with nondegenerated nuclei in the secondary fibre cells, suggesting a critical role of HSF4 in the lens fibre cell differentiation.30