The present research was conducted to investigate the molecular etiology of non-syndromic CC in cases with genetic heterogeneity using WES analysis. In almost two thirds of these families, pathogenic and likely pathogenic variations explaining the etiology of CC were identified. Studies investigating non-syndromic CC cases in varying patient cohorts in different populations have reported varying rates of diagnosis. The diagnosis rate reported in patient cohorts investigated by a targeted next generation sequencing panel or WES analysis range between 42% and 75% (8, 9, 10, and 11). Although the sample size was small in the present research, the diagnosis rate was found to be 58%, and this rate was consistent with the literature. In their study investigating CC cases using whole genome sequencing (WGS), Ma et al. [12] concluded that the rate of diagnosis was 77%, which was 10% higher compared to the rate of diagnosis in WES. Another previous study stated that using WGS will likely increase the diagnosis rates to identify the molecular etiology of relevant cases, as WGS enables the detection of variants in GC-rich regions, indels in repetitive regions, and small copy number variations [12].
In the present research, we found variants in β-crystallin and γ-crystallin genes (CRYBB2, CRYGA, and CRYBB1) in 3 (F1, F2, F3) of the 12 families investigated. A de novo missense heterozygous c.355G>A (p.G119R) variant was detected in the CRYBB2 gene in a 6-month-old male patient (II-2) with bilateral membranous cataract in Family 1. Ma et al. [10] reported this variant for the first time in the literature and identified this variation as a de novo variant in a sporadic case. Later on, 3 individuals with nuclear cataract from the same family were found to have the same variant in a study conducted by Chen et al. in familial CC cases [13]. Finally, this variant was also reported by Bell et al. in 2021 as a de novo variant in a sporadic case [14]. This variant (c.355G>A; p.G119R) is located on the third Greek key domain of the CRYBB2 protein. Previous analyses predict that conversion of the hydrophobic glycine amino acid into a hydrophilic arginine amino acid is very likely to affect the three dimensional structure of this protein [10]. In these studies, it was seen that these cases that had the same variation in the same gene had different cataract types. This is supporting evidence for the clinical heterogeneity observed in CC cases. In the present research, we identified a missense heterozygous (c.239G>A; p.R80H) variant in the CRYGA gene in the index case in Family 2 and the mother of the case who had a similar phenotype. In the literature, this variant was identified in previous studies examining CC cases in families and the variation was reported in three individuals from the same family [11]. In the present research, a novel heterozygous in-frame deletion (c.126_128delAAC; p.T43del) in the CRYBB1 gene was identified in an index case in Family 3 with bilateral posterior polar cataract. Furthermore, it was found that the variant was also present in the mother of this case who had a similar phenotype. This variant is not present in the gnomAD population database and is considered a VUS variant according to ACMG criteria. In the present research, the variation was found to be co-segregated with the disease in the family. However, in order to clarify the pathogenicity of this variant, further experimental studies are needed.
In the present research, two different novel missense variations (c.144_145delGCinsAG; p.Q49E and c.65G>A; p.G22D) were identified in the gene encoding the gap junction alpha-3 protein (GJA3) in two different families (families F4 and F5). These variants were not previously reported in the literature or population databases. According to the ACMG criteria, these two variants are categorized as likely pathogenic. Previous studies in the literature report that the glycine amino acid located in codon 22 of the GJA3 protein is highly phylogenetically conserved among multiple species, and the change of glycine to serine in the same codon leads to the CC phenotype in these patients [15]. It is also reported in the literature that glutamine located in codon 49 is highly phylogenetically conserved among multiple species. These two variations (p.G22D, p.Q49E) are located on the N-terminus/transmembrane domain 1 boundary and first extracellular loop (EC1) domain of the GJA3 protein, respectively. These two domains of the GJA3 protein play a role in pore structure/gating and gap junction formation, respectively [16]. According to our results and supporting literature evidence, it can be speculated that both of these novel variants identified in the GJA3 gene lead to the CC phenotype.
In 2018, Ansar et al. [17] conducted a study on Pakistani families with CC and identified homozygous loss-of-function variations in the Dynamin-binding protein (DNMBP) in 3 different families, and this was the first study in the literature to associate DNMBP gene with the CC phenotype. In the present research, a first cousin marriage was observed between the parents in Family 6, and it was found that the index case received the diagnosis of bilateral lamellar cataract at the age of five months. In Family 6, a bi-allelic novel nonsense (c.1918C>T; p.R640*) variant was detected in the DNMBP gene in an index case. It was found that both parents in Family 6 were heterozygous for this variant. The present research is the second study in the literature that reports bi-allelic loss-of-function variants in the DNMBP gene. In the present study, a 6-month-old patient diagnosed with bilateral lamellar cataract in Family 7 was found to carry maternal c.395G>A (p.G132E) and paternal c.960G>T (p.Q320H) variations in the DNMBP gene, and furthermore, there was no consanguinity among the parents of this case. Although both variants were categorized under VUS variants according to ACMG criteria, there is a need for further studies to examine and clarify the pathogenicity of these variants.
Wolframin is a cellular protein located on the endoplasmic reticulum (ER) membrane. Wolframin plays a role in the regulation of membrane trafficking, secretion, and ER calcium homeostasis. Reports in the OMIM database show that WFS1 gene mutations are associated with Wolfram Syndrome, Wolfram-like syndrome, autosomal dominant deafness, and autosomal dominant cataract phenotypes. Previous studies have so far stated that there are two missense variations in the WFS1 gene associated with the non-syndromic CC phenotype, and both of these variations are registered in the Cat-Map database [18, 19]. In the present research, two variations were detected in families 8 and 9 and both were considered likely to be pathogenic according to ACMG criteria. Furthermore, these variations were also co-segregated with the disease in the family. In the present research, a missense heterozygous c.2603G>A (p.R868H) variant was detected in the WFS1 gene in Family 9. In previous studies, this variant was associated with the autosomal recessive hearing loss phenotype. However, in the present research, no additional anomaly was identified in the case except for the presence of cataract.
FYCO1 (FYVE and coiled-coil domain containing 1) protein plays a key role in the transport of microtubule-mediated autophagocytic vesicles in the cell [20]. In the literature, FYCO1 gene expression has been identified in the lens. Furthermore, literature evidence shows that the absence of this protein leads to the cataract phenotype as a result of elevated levels of reactive oxygen radicals [20]. In the present research, first cousin marriage between parents was identified in Family 10. In this family, it was found that the index case received a diagnosis of bilateral total white cataract at the age of six months. We found a homozygous missense c.265C>T (p.R89C) variant in the FYCO1 gene in the index case in Family 10 (F10, II-1). This variant reported as likely pathogenic without providing the relevant phenotype information in the ClinVar database. The fact that this variant is likely pathogenic according to ACMG criteria and it was reported to be co-segregated with the disease allows us to speculate that the variant can explain the etiology of the relevant case.
GCNT2 is an autosomal recessive gene that is associated with the CC phenotype and is expressed in erythrocytes and lens cells in the body. To date, studies in the literature have only reported limited numbers of missense, nonsense, and gross exonic deletions in the GCNT2 gene [21]. In the present research, Family 11 contained two siblings with CC cataract (F11, II-1, II-2) born from parents without any consanguinity. Therefore, possible autosomal recessive inheritance was considered, and the family was thus included in genetic analysis. A paternally inherited novel heterozygous missense c.58A>G (p.I20V) variant was detected in the GCNT2 gene in siblings with the CC phenotype. This variant was categorized as a VUS variant according to ACMG criteria. Literature evidence shows that bi-allelic GCNT2 gene variants can lead to the CC phenotype. We can talk about two possibilities in such a case: i) There may be a monoallelic gross exonic deletion or deep intronic variation in the GCNT2 gene that WES analysis fails to identify, or ii) the CC phenotype of the patients may be associated with another gene which is not revealed by WES analysis.
Literature evidences that there is LSS gene expression in the lens. Furthermore, LSS encodes lanosterol synthase, which plays a critical role as the rate-limiting enzyme in cholesterol synthesis. Previous reports in the literature show that LSS gene variations cause a decrease in cholesterol levels in the lens and lead to cataract development [22]. In the present research, a novel homozygous missense c.1673A>G (p.E558G) variant was detected in LSS gene in the proband in Family 12. The amino acid residue is located in a highly conserved region. Therefore, the variation may have an effect on the function of this protein. However, there is a need for further research to elucidate this issue.
There are certain limitations of the present research. i) The sample size was small. This may have had an effect on diagnosis rates and the protein groups in which variations were detected. Therefore, the results may not be generalizable to the entire population. ii) We could not perform a functional study for variants categorized as VUS variants according to ACMG criteria. iii) Due to the technical limitations of WES analysis, we could not investigate gross deletions/duplications, deep intronic variants, and variants in homopolymer regions in this study. In autosomal recessive genes with monoallelic variant, as in Family 11, the second variant should be studied in the relevant cases. Therefore, it would be beneficial to examine these cases using Multiplex Ligation-dependent Probe Amplification and WGS analysis.
To the best of our knowledge, the present research is one of the limited numbers of studies in the Turkish population in which genetically heterogeneous non-syndromic CC was investigated using WES analysis. Novel variants that we identified in DNMP, LSS, and WFS1 genes, which are rarely associated with the CC phenotype, have contributed to the mutation spectrum of this disease. Identifying the relevant molecular genetic etiology allows accurate genetic counseling to be provided to the families.