Novel mutations identified in Chinese families with autosomal dominant congenital cataracts by targeted next-generation sequencing

Congenital cataract is a clinically and genetically heterogeneous lens disorder, characterized by opacification of crystalin lens at birth or during early childhood [1]. The prevalence of congenital cataracts varies from 1 to 6 per 10,000 live births [2]. Approximately one third of the cases have a family history [3]. The cataract may be an isolated anomaly, or part of a multisystem syndrome [4]. Both X-linked and autosomal recessive inheritance patterns have been reported for congenital cataract, however autosomal dominant trait is the most prevalent mode [5‐7]. Cataracts can be classified as sutural, whole lens, nuclear, lamellar, cortical, polar, cerulean, coralliform, and other subtypes, according to morphology of lens [8‐10].

To date, at least 30 pathogenic genes have been found to link to congenital cataracts. From the reported mutant genes in congenital cataract families, nearly half of the mutations associated with crystalin genes [11], including genes coding for crystalin families (CRYAA, OMIM 604219; CRYAB, OMIM 613763; CRYBA1, OMIM 600881; CRYBB1, OMIM 611544; CRYBB2, OMIM 601547; CRYBB3, OMIM 609741; CRYGC, OMIM 604307; CRYGD, OMIM 115700; CRYGS, OMIM 116100), gap junctional proteins (GJA3, OMIM 601885; GJA8, OMIM 116200), beaded filament structural proteins (BFSP1, OMIM 611391; BFSP2, OMIM 611597), and other functional genes (e.g., HSF4, OMIM 116800; MIP, OMIM 615274; PITX3, OMIM 610623; EPHA2, OMIM 116600) [7, 9, 12‐15].

Identification of accurate genetic cause of congenital cataract is essential for providing precise diagnosis and genetic counseling [8]. However, due to the high clinical and genetic heterogeneities, clinical and genetic diagnostic of congenital cataract, especially for nonsyndromic congenital cataracts, are limited by the traditional sequencing method, by which only few candidate genes can be sequenced at each time [16]. Recently, the next generation sequencing (NGS) combined with targeted genomic enrichment has proved to be an effective solution to the genetic test of genetically heterogeneous diseases and provides a new opportunity for genetic diagnostics of congenital cataracts [12, 17].

In this study, we collected information from five large Chinese families with congenital cataracts. Then we performed targeted enrichment and deep sequencing to detect the genetic mutations in these families. We identified five novel mutations in the GJA3 (S385Efs*83), GJA8 (F52 L), BFSP1 (G602R), EPHA2 (T511 M) and HSF4 (R119H) genes that potentially resulted in the development of congenital cataract. With Sanger sequencing, we confirmed that mutations were co-segregated with affected individuals in the five families, whereas mutations were not found in unaffected family members and normal controls. Bioinformatics analysis, conservative prediction and 3-D protein simulation indicated that the five mutations might be the pathogenic mutations for congenital cataract families. This study demonstrates that the targeted gene sequencing can be used as an effective tool for genetics diagnosis of congenital cataract.

We reported five novel mutations associated with the autosomal dominance cataract in five Chinese families respectively: c.154 T > C in GJA8, c.1152_1153insG in GJA3, c.1804G > C in BFSP1, c.1532C > T in EPHA2 and c.356G > A in HSF4. All of the five mutations were screened by targeted NGS for the 38 candidate genes of congenital cataracts, and verified through Sanger DNA sequencing. We confirmed that each mutation co-segregated with the disease phenotypes in the corresponding family and absent in all the unaffected individuals. Further, bioinformatics analysis, conservative prediction and 3-D protein simulation showed that these mutations might be deleterious. According to the ACMG criteria [20], c.1152_1153insG in GJA3 of Family 2 and c.356G > A in HSF4 of Family 5 are clearly pathogenic variants (class V); c.154 T > C in GJA8 of Family 1 is a likely pathogenic variant (class IV); c.1804G > C in BFSP1 of Family 3 and c.1532C > T in EPHA2 of Family 4 variants are unknown significance (class III) (Table 3). The unknown significance variants associated with congenital cataracts make them interesting candidates for further studies.

The lens has developed an extensive cell-cell interaction system using connexins to maintain its transparency. Three connexins are expressed in the lens: connexin 43 (Cx43), connexin 46 (Cx46), and connexin 50 (Cx50). Cx43 (GJA1) is expressed mainly in epithelial cells of lens, while Cx46 (GJA3) and Cx50 (GJA8) are expressed in lens fibre cells [21, 22]. GJA8 and GJA3 are the major connexin of the ocular lens, where gap junctions maintain ionic environment, water balance, transparency and optical properties of the lens [23]. To date, 65 variants in GJA8 and 43 variants in GJA3 have been reported in the HGMD (Professional 2019.1) to induce genetic cataracts, which account for about 1/4 of nonsyndromic familial cataract cases. The typical structure of connexin includes cytoplasmic NH2- and COOH- terminal domain, four transmembrane domains and two extracellular loops. The two extracellular loops mediate hemichannel docking between connexons and the E1 loop, which was also shown to be important for the voltage required for closure of gap junction pores [24]. In this study, we identified an amino acid change (F52 L) at the first external loop (E1) in GJA8 in family 1. The altered protein may disrupt normal interactions between the two connexins, which may reduce resistance of the intercellular channel and lead to the leakage of small ions. Moreover, F52 L is highly conserved among many species, so F52 L is very likely to cause disease. In Family 2, frameshift S385Efs*83 in GJA3 resulted from a guanine insertion that introduced a premature translation stop codon located in the COOH-terminus, which may interfere with the folding of the whole protein and resulted in cataract. This insertion mutation (c.1152_1153insG) is similar to the three mutations (c.1137dupC, c.1189dupG, c.1200dupC) reported previously [25‐27], thus providing further evidence that the GJA3 C-terminal domain plays an essential role in physiological function of the gene, and further expanding the mutation spectrum of GJA3 in association with congenital cataract.

BFSP1 (filesin) and BFSP2 (phakinin) are major components of the beaded filament, which are unique cytoskeletal lens structures. The biological functions of filesin and phakinin are still not clear, but some evidences indicate they play an important role in maintaining lens transparency and homeostasis during fetal development and fiber cell differentiation [28]. A novel mutation c.1804G > C(p.G602R) in BFSP1 was detected in Family 3. Alignment of the BFSP1 protein sequence among different species revealed that the Gly residue at position 602 was less conservative. MutationTaster and PROVEAN prediction tools showed the pathogenicity of G602R was neutral. However, M-CAP, SIFT and PolyPhen 2 analysis indicated that G602R was possibly damaging. Further, mutation was co-segregated with phenotypes in the Family 3 including five affected and four unaffected individuals and that variant frequency was 0.000066 in the ExAC browser, indicating that this variant was rare event in the human genome. Up to now, only six BFSP1 mutations have been reported and four BFSP1 mutations were involved in autosomal recessive cataract families [11, 29, 30]. And two mutations were found in autosomal dominant cataract families. In 2013, Wang et al. first found a heterozygous variant c.1042G > A(p.D348N) in BFSP1 in a 5-generation Chinese family in which 15 members had autosomal dominant nuclear cataract [31]. In 2017, Zhai et al. identified heterozygosity for a splice site mutation (c.625 + 3A > G) in BFSP1 in a 4-generation family co-segregating progressive punctate lamellar cataract [32]. The mutation (G602R) highlighted in this study is localized at the tail region of filesin, has an important effect on beaded filament formation as mutation D348N [31]. Taken together with previous research, the results of the Family 3 enriched the suspected pathogenicity of the BFSP1 mutation in human autosomal dominant congenital cataract.

The protein encoded by EPHA2, Ephrin Receptor EphA2, is spatially and temporally regulated in the cortical lens fiber cells, while its expression is lower in anterior epithelial cells, and absent in the nuclei of lens [33]. So far, 22 mutations of EPHA2 have been reported in the patients with congenital cataract, and most of them are in the SAM domain. After identification of p.P584L by Dave et al., we reported the second autosomal dominant mutation p.T511 M in the juxta membrane domain of the protein [34]. The pathogenicity of this mutation was proved in the following three aspects: (1) Protein sequence among different species revealed that the Thr residue at position 511 was highly conserved; (2) Bioinformatics analysis using five prediction tools indicated that T511 M was a pathogenic change. (3) 3-D protein simulation model predicted that amino acids change of M511 T would destroy H-bonding between T511 interacted via with residues N435 and Q515 of Ephrin type-A receptor 2. Furthermore, this mutation was co-segregated with phenotypes in Family 4. The mutation T511 M identified in EPHA2 gene is a known polymorphism (rs55747232), which raises doubt about its pathogenicity. In conclusion, we believe that M511 T in EPHA2 is a potential variation associated with congenital cataract.

HSF4 belongs to the family of heat-shock transcription factors that bind heat shock elements and activate downstream heat-shock response genes under conditions of stress [35]. It has been reported that HSF4 gene is responsible for both autosomal dominant and autosomal recessive cataracts [36]. We had screened the affected individuals in Family 5 and identified a missense mutation c.356G > A(p.R119H) in HSF4, and this mutation was co-segregated with the disease in all the affected individuals, but not observed in all the unaffected individuals. Protein sequence among different species revealed that the Arginine(R) residue at position 119 is high conserved, and five prediction tools showed p.R119H is pathogenic. 3-D protein simulation predicted that substitution of H119 destroyed the H-bonding, with which wild-type R119 interacted with residues L124 of Heat shock factor protein 4. Above all provided a persuasive evidence to its pathogenicity of p.R119H in HSF4 of Family 5.

In summary, we performed genetic analysis in five Chinese families with congenital dominant cataracts and identified five novel mutations, including an insertion mutation encoding p.S385Efs*83 in GJA3 and four missense mutations: p.F52 L in GJA8, p.G602R in BFSP1, p.T511 M in EPHA2 and p.R119H in HSF4. This work extended the mutation spectrum of congenital cataracts, and would provide more evidences for the precise diagnosis of the disease.

We firstly reported five novel mutations associated with autosomal dominant cataracts: c.154 T > C in GJA8, c.1152_1153insG in GJA3, c.1804G > C in BFSP1, c.1532C > T in EPHA2, c.356G > A in HSF4. This study expands the mutation spectrum of congenital cataracts, and provide solid evidence for genetic counseling and prenatal gene diagnosis of the cataract families.