Identification of a novel missense mutation of MIP in a Chinese family with congenital cataracts by target region capture sequencing

Congenital cataract is both clinically diverse and genetically heterogeneous. To investigate the underlying genetic defect in three-generations of a Chinese family with autosomal dominant congenital cataracts, we recruited family members who underwent comprehensive ophthalmic examinations. A heterozygous missense mutation c.634G > C (p.G212R) substitution was identified in the MIP gene through target region capture sequencing. The prediction results of PolyPhen-2 and SIFT indicated that this mutation was likely to damage the structure and function of MIP. Confocal microscopy images showed that the intensity of the green fluorescent signal revealed much weaker signal from the mutant compared to the wild-type MIP. The expressed G212R-MIP was diminished and almost exclusively cytoplasmic in the HeLa cells; whereas the WT-MIP was stable dispersed throughout the cytoplasm, and it appeared to be in the membrane structure. Western blot analysis indicated that the protein expression level of the mutant form of MIP was remarkably reduced compared with that of the wild type, however, the mRNA levels of the wild-type and mutant cells were comparable. In conclusion, our study presented genetic and functional evidence for a novel MIP mutation of G212R, which leads to congenital progressive cortical punctate with or without Y suture.


Results
Clinical features. We identified a Chinese family with three generations of individuals (eight affected individuals and seven unaffected individuals) with diagnoses of ADCC (Fig. 1). The proband (II:2) was a 60-year-old male with a complaint of blurred vision in both eyes who underwent bilateral cataract surgery in our hospital in January and April 2013. The other affected participants of the family were aged 20-77 years, and manifested bilateral fine punctate anterior and posterior cortical opacities, combined with Y-sutural cataracts. However, the proband had a different type of cataract with fine punctate cortical opacities. None of the younger affected patients (III:2, III:5, III:7) complained of significant visual deterioration, and they were unaware of their cataracts until the examinations. The punctate opacities increased in number and both the punctate and Y-sutural opacities became gradually denser with age (Fig. 2). There were no other ocular or systemic abnormalities or symptoms.
Capture panel sequencing results, variant analysis and validation. Using the capture panel described in the Methods, an average of 137× and 188× depth in the target region was achieved, and 96.49% and 97.08% of designed target regions were covered by at least 20× in the proband and his brother, respectively. This demonstrated that sufficient data quality was achieved to identify variants. In total, 723 and 744 variants were detected in the coding regions and adjacent intronic regions in the proband and his affected brother, 57 and 52 of which were rare (the variants were filtered out if their frequency was > 0.01 in 1000 the genome database, dbSNP, HapMap project or local database). Only four rare variants were found in the 42 known congenital cataract causing genes in both samples ( Table 1). Two of the four rare variants were in the LEPREL1 gene, which has been reported to be an autosomal recessive gene mainly cause myopia 11,12 . It seems not to have affected the patients as the disease in the inherited in a autosomal in this family exhibits an dominant pattern and does not show myopia. The heterozygous variant c.10-7C > G in the CRYGD gene is in a potential accept splice site, but was predicted as a polymorphism by Mutation Taster (http://www.mutationtaster.org/). Direct sequencing of the variant of CRYGD gene was done in samples of the family members, and the results indicated the heterozygous variant c.10-7C > G in the CRYGD gene only exhibited in the proband (II:2) and his affected brother (II:4). There was no cosegregation of the variant of CRYGD gene with affected members in this family. Therefore, it seems not to be a disease-causing mutation. The c.634G > C mutation of the MIP gene, which has been reported to cause autosomal dominant cataracts 13,14 , was a novel heterozygous missense mutation. This mutation changed Glycine  (GGG) to Arginine (CGG) at position 212 (p.G212R) (Fig. 3A). Segregation analysis was performed, and this mutation cosegregated well with all affected participants and was not found in unaffected members or the 100 unrelated normal controls. Bioinformatics analysis. The mutation was predicted to be probably damaging by PolyPhen-2 with a score of 0.958 (sensitivity: 0.63; specificity: 0.92), and to affect protein function by SIFT, with a score of 0.00 (media information conservation 3.01). Collectively, these results strongly indicated that the p.G212R mutation is likely to be deleterious to the protein and may therefore be responsible for the congenital cataracts. The result of a multiple sequence alignment showed that the Glycine at position 212 of MIP is highly conserved among various species (Fig. 3B).

Result of Fluorescence Microscopy analysis. We investigated the expression of WT-MIP and
G212R-MIP after the transient transfection of HeLa cells. The green fluorescence was much weaker in HeLa cells transfected with the G212R construct than in those transfected with the wild type construct. The expressed G212R-MIP was diminished and almost exclusively cytoplasmic in the cells. In contrast, WT-MIP was stable dispersed throughout the cytoplasm, and it appeared to be in the plasma membrane, in the subcellular organelles and in the nuclear membrane (Fig. 4A).  Results of RT-PCR and Western blot analysis. RT-PCR was used to measure the RNA transcription level of WT-MIP and G212R-MIP after total RNA was extracted from the HeLa cells. The RT-PCR revealed a similar relative MIP mRNA expression levels in WT-MIP and G212R-MIP (Fig. 4B). However, Western blot analysis indicated that the G212R mutation significantly reduced protein expression levels of MIP, consistent with the green fluorescence signal intensity (Fig. 4C), indicating that the mutation decreased the protein production of MIP gene.

Discussion
In the present study, we investigated the genetic and functional defects of three generations of a Chinese family mainly affected with a typical Y-suture cataract combined with punctate cortical opacities. Target region capture sequencing revealed a novel missense mutation of the MIP gene in exon 4 at nucleotide 634, which caused a Glycine-to-Arginine substitution at 212 position (p.G212R).
To date, 16 mutations in the MIP gene have been reported to be associate with autosomal dominant cataracts, including 10 missense mutations 8,13,[15][16][17][18][19][20][21] ; one acceptor splice-site mutation 22 ; one donor splice-site mutation 23 ; one deletion that causes a frameshift at 638delG 14 ; one initiation codon mutation 24 ; and two nonsense mutations 25,26 . The cataract phenotypes are significantly different among the MIP mutation families. Phenotypically identical cataracts can result from mutations at different genetic loci and may have different inheritance patterns, while phenotypically variable cataracts can be found in a single large family 6 . In the family, the proband had a different type of cataract as compared with other affected family members although they had the same mutation. We compared the phenotype of our family with other reported families and found that the clinical features were very similar to those reported in other studies 14,25 : all manifested as fine punctate opacities in the cortex and Y suture. The cataract family reported by Yu et al. 25 was associated with a nonsense mutation (c.337C > T (p.R113X)) that produced a severely truncated protein. Geyer et al. 14 reported a single nucleotide deletion, that caused a frameshift and a premature stop codon that truncated six amino acids from the C-terminus of MIP. This family had different cataract phenotypes including both fine punctate opacities in the cortex and Y sutures, and fine white punctate opacities in the cortex, similar to our cataract family.
The MIP gene encodes a 28-kDa protein with 263 amino acids, and it is primarily and abundantly expressed in the lens. After Peter Agre's laboratory developed a functional assay for water channels, the MIP family became the aquaporin family and MIP became known as aquaporin 0 10 . Besides functioning as a water channel, MIP also plays a role in interfiber adhesion and intracellular interaction with lens fiber proteins, being required for lens transparency and accommodation. Mutations in the MIP gene in human and mice results in genetic cataracts; deletion of the MIP/AQP0 gene in mice results in dominant cataracts, decrease of water channel activity and a lack of suture formation required for maintenance of the lens fiber architecture 10,27 . As an intrinsic membrane protein, MIP inserts in the plasma membrane with six transmembrane bilayer-spanning domains (H1-H6), resulting in three extracellular loops (A, C and E), two intracellular loops (B and D), and the N-and C-terminal intracellular domains. It assembles as a tetramer with four water pores in the endoplasmic reticulum (ER) before being transported and inserted into the plasma membrane. Each monomer is a water channel and can function independently 28,29 . The first missense mutation in H6 of MIP (c.644G > A) was reported to be associated with punctate cataracst in 2014 8 ; this is the second substituation mutation in H6 that can cause ADCC.
In the present study, our data suggest that the c.634G > C, G212R substitution may contribute to ADCC pathogenesis in this family. First, we investigated the expression of WT and G212R -MIP proteins in transfected HeLa cells viewed by confocal microscopy. The remarkable difference in the green fluorescence signal intensity indicated that the production of the mutant protein decreased. The expressed WT-MIP was stable dispersed throughout the cytoplasm, and it appeared to be in membrane structure. In contrast, G212R-MIP was diminished and almost exclusively cytoplasmic in the cells. Second, immunoblot analysis indicated that the protein expression level of the mutant form of MIP was remarkably reduced compared with that of the wild type, however, the mRNA levels of the wild-type and mutant cells were comparable. These results were similar to previous p.G165D results 20 . All the above results indicated the mutation resulted in instability of the protein. The loss of the water channel function could reflect RNA instability, impaired translation, protein degradation, or differential post-translational modification as well as impaired trafficking. Several mutations have been functionally characterized in vitro. The p.E134G 13 , p.T138R 13 , p.G165D 20 , and p.G215D 8 mutations may result in the loss of water permeability due to impaired trafficking of mutant proteins to the plasma membrane. The 638G deletion resulted in the impairment of the cell membrane, localizing the mutant protein in the endoplasmic reticulum without trafficking it to the plasma membrane, and inducing cellular cytotoxicity 30 . Retention or accumulation of proteins in the ER can cause ER stress and an unfolded protein response, which have been implicated in the pathogenesis of several diseases 20 .
Target region capture sequencing is a combination of genomic region enrichment and next-generation sequencing technology. The compatible performance in capture coverage, as well as the lower cost and shorter time required, makes it a good choice for clinical diagnosis of a heterogeneous group of monogenic disorders 31 .
In summary, through target region capture sequencing, this study reported ADCC caused by a novel missense mutation c.634G > C in exon 4 of the MIP gene, resulting in a Glycine -to-Arginine (p.G212R) substitution in the H6 domain of MIP. Moreover, this study presented evidence that p.G212D caused the production of the mutant protein to be reduced and retained in the cytoplasm. Further studies are needed to elucidate the pathophysiologic changes caused by this mutation. This study expands the spectrum of mutations that cause congenital cataracts.

Methods
Clinical data and genomic DNA preparation. Three-generations of a Han Chinese family from Zhejiang Province with ADCC were recruited through the Department of ophthalmology, the first Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China. All study participants underwent detailed clinical and ophthalmological examinations. This study was approved by the Institutional Review Board of the First Affiliated Hospital, Zhejiang University School of Medicine, and adhered to the Tenets of the Declaration of Helsinki. Each participant was informed about the nature of the study, and written informed consent was obtained. Genomic DNA was extracted from the peripheral blood for PCR amplification using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's standard procedure. A total of 100 ethnically matched subjects without a family history of congenital cataracts were recruited as controls.

Target region capture sequencing and analysis
Library preparation and target region capture sequencing. A custom-made capture panel (BGI, Shenzhen China) was designed to capture 351 genetic eye disease genes (gene list not shown), which included 42 inheritable genetic congenital cataract-related genes (Supplemental Table 1). These genes were collected from OMIM (http://www.ncbi.nlm.nih.gov/omim/) and the published literatures. The genomic DNA of the proband (II:2) and his affected brother (II:4) was separated into approximately 200-300 base pair (bp) fragments and used to generate a paired-end library (Covaris S2, Woburn, MA, USA). The library capture was completed through BGI using the custom-made capture array and sequenced on an Illumina HiSeq2000 Analyzer (San Diego, CA). Image analysis and base calling were performed using the Illumina Pipeline to generate raw data.

Variant identification and validation.
To detect the variants in the patients, we applied filtering criteria to generate clean reads, and then aligned the clean reads to the human genome reference from the NCBI database (NCBI build 37.1) using the Burrows Wheeler Aligner (BWA) Multi-Vision software package. Single-nucleotide variants (SNVs) and insertions and deletions (InDels) were extracted by the Genome Analysis Tool Kit (GATK). All SNVs and InDels were annotated using the NCBI dbSNP, HapMap project, 1000 Genome Project and the BGI local database of a 1000 healthy adults. The candidate variations of known congenital cataract genes were validated by polymerase chain reaction (PCR) and Sanger sequencing. PCR primer sets were designed via Primer