Screening of BEST1 Gene in a Chinese Cohort With Best Vitelliform Macular Dystrophy or Autosomal Recessive Bestrophinopathy

Abstract

Purpose: Mutations in the BEST1 gene can cause Best vitelliform macular dystrophy (BVMD) and autosomal recessive bestrophinopathy (ARB). The aim of the current study was to establish the BEST1 mutation spectrum in Chinese patients with BVMD and ARB and to describe the phenotypic characteristics of patients carrying BEST1 mutations.

Methods: A total of 37 probands with a clinical diagnosis of BVMD (17 patients) or ARB (20 patients) were recruited for genetic analysis; of these, only 5 probands had a family history. All probands underwent detailed ophthalmic examinations. All coding exons and exon-intron boundaries of the BEST1 gene were screened by PCR-based DNA sequencing. In silico programs were used to analyze the pathogenicity of all the variants. Genomic DNA rearrangements of the BEST1 gene were identified by real-time quantitative PCR (RQ-PCR).

Results: For patients with BVMD, single heterozygous BEST1 mutations were identified in 13 patients and compound heterozygous mutations were found in 3 patients. For patients with ARB, biallelic mutations were found in 13 probands and single mutant alleles in six patients. Overall, 36 disease-causing variants (20 novel mutations) of the BEST1 gene were identified, including 28 (77.8%) missense, 3 (8.3%) nonsense, 4 (11.1%) splicing effect, and 1 (2.8%) frameshift small duplication mutations.

Conclusions: The mutation spectrum of the BEST1 gene in Chinese patients differed from those of Caucasian patients. Mutations that cause ARB differ from those that cause BVMD. BEST1 screening is important for precise diagnosis of BVMD or ARB.

Best vitelliform macular dystrophy (BVMD; OMIM 153700), also known as Best disease, is an early-onset inherited macular degeneration.¹ The characteristic sign of BVMD is the presence of bilateral yellowish yolk-like lesions in the macula. The lesions, which show considerable morphologic variability with time, evolve through the different stages of the disease: the previtelliform, vitelliform, vitelliruptive (scrambled egg), pseudohypopyon, atrophic, and cicatricial stages.² Electrooculography (EOG) is the most specific clinical diagnostic examination for BVMD, indicated by light-peak:dark-trough ratios (Arden ratios) usually less than 1.5.^1,2 Most BVMD is inherited in an autosomal dominant pattern with an incomplete penetrance and is caused by mutations of the BEST1 gene.^3,4 

Autosomal recessive bestrophinopathy (ARB; OMIM 611809), first detailed and described by Burgess and colleagues in 2008,⁵ is another disease caused by BEST1 mutations. In contrast to typical BVMD, the critical features of ARB are the presence of multifocal vitelliform lesions, with subretinal fluid or macular edema, scattered over the posterior pole of the retina. Some patients with ARB show hyperopia and shallow anterior chambers, with a corresponding high incidence of angle-closure glaucoma (ACG). Electrophysiological examinations may show a reduction in cone and rod responses in full-field ERG and an absent or markedly reduced light peak in EOG.⁵ 

The BEST1 gene, originally known as VMD2, consists of 11 exons that encode the bestrophin-1 protein (585 amino acids). The protein is predicted to have a short N terminus of 27 amino acids, 4 transmembrane domains, and a long C-terminal region of 294 amino acid residues localized to the cytoplasmic side of the membrane.⁶ This multifunctional protein is localized at the basolateral plasma membrane of RPE cells.^3,4,7 The exact functional role of bestrophin-1 is still not clear, but it is presumed to act as a Ca²⁺activated Cl⁻ channel (probably responsible for the reduced Arden ratio)⁸ and an inhibitor of intracellular voltage-dependent Ca²⁺ channels.^7,9,10 Thus far, more than 250 mutations have been implicated in a group of eye diseases collectively referred to as bestrophinopathies (http://www-huge.uni-regensburg.de/BEST1_database/). In addition to BVMD and ARB, the other two major phenotypes of bestrophinopathies are autosomal dominant vitreoretinochoroidopathy and microcornea, rod-cone dystrophy, cataract, and posterior staphyloma syndrome.^{1,3–5,11–21} In BVMD, most mutations are missense mutations and are nearly exclusively distributed within or close to transmembrane domains (in the first 310 residues).^{1,3,4,11–20} The BEST1 mutations identified in BVMD are extremely heterogeneous; most are rare and found uniquely in one family.^{1,3,4,11–20} Several mutations (p.T6P, p.R25W, p.R218C, p.Y227N, p.A243V, p.I295del, p.E300D, p.D301E, and p.D302N) have been found in more than three families.¹ Patients with the same mutation can show phenotypic variability; for example, patients with the mutation p.T6P, which has been frequently identified in the Dutch population, may have an abnormal EOG without ophthalmoscopic abnormalities or typical BVMD and multifocal vitelliform dystrophy.^3,19 Several notably frequent mutations observed in BVMD patients are usually ethnic specific, with allele frequencies between 20% and 45%. For example, in Danish BVMD patients, the most prevalent mutation is p.D302N, with an allele frequency of 44.4%.¹⁶ In the Italian population, the most common mutation p.R25W, has the highest allele frequency of 36.8%,¹⁷ followed by p.R218C with an allele frequency of 26.3%.¹⁷ The common mutation p.D302N in Danish patients may be due to a founder effect.¹⁶ 

In contrast to the mutations found in BVMD, almost half of the mutations identified in ARB are located between residues 312 and 325, which are more C-terminal than most of the mutation sites associated with BVMD.⁵ The mutation p.R141H is the most common mutation for ARB, having been identified in several unrelated families of European ethnicity.^5,21 

At present, the BEST1-causing bestrophinopathies in the European population have been intensively investigated; however, the reports are limited with regard to the clinical features and the mutation spectrum of the BEST1 gene in Chinese patients with BVMD and ARB. Here, we performed a comprehensive molecular screening of 37 unrelated Chinese patients diagnosed with BVMD or ARB and described a somewhat different mutation spectrum, as well as 20 novel mutations, in the BEST1 gene. 

Materials and Methods

Patients

This study was approved by the ethics committee of Beijing Tongren Hospital and adhered to tenets of the Declaration of Helsinki. The protocol of this study was performed within the institutional guidelines of the Beijing Tongren Hospital Joint Committee on Clinical Investigation. Informed written consent was obtained from each enrolled patient (or guardians of the underaged patients) before participation in this study. In total, 37 unrelated patients were recruited at the Genetics Laboratory of Beijing Institute of Ophthalmology, Beijing Tongren Ophthalmic Center, during the period of 2011 to 2017. Of these, 17 patients were diagnosed with BVMD and 20 were diagnosed with ARB. In this study, only five probands had a family history and the remaining 32 unrelated patients were sporadic. Detailed ophthalmic examinations, including best-corrected visual acuity with E decimal charts, slit lamp biomicroscopy, color fundus photography, and optical coherence tomography (OCT), were conducted on all participants. Most patients also underwent fundus autofluorescence (FAF) and color discrimination tests (using the Lanthony Panel Hue15 or Farnsworth Panel D15 Test). Some patients underwent EOG and full-field ERG examinations. 

BVMD was diagnosed based on the following criteria: juvenile-to-adult onset vision loss or metamorphopsia; bilateral or unilateral macular lesions showing vitelliform, vitelliruptive, pseudohypopyon, or atrophic and cicatricial changes; and abnormal EOG (Arden ratio always below 1.5).²⁰ The clinical diagnosis criteria for ARB was as follows: juvenile-to-adult onset vision loss or metamorphopsia, bilateral multifocal vitelliform lesions with subretinal fluid or macular edema, abnormal EOG (Arden ratio always below 1.5), and an autosomal recessive or sporadic mode of inheritance.⁵ 

The patients diagnosed with BVMD were classified into five stages.² Patients in stage 0 exhibit a normal macula in the fundus photo, but an abnormal EOG. Patients in stage I showed a disturbance of the RPE of the macula. Patients in stage II had a typical vitelliform (egg yolk lesion) or breaking up of the vitelliform cyst (scrambled egg phase) in the macula. Patients in stage III (pseudohypopyon phase) had some yellow materials forming a horizontal level in the lesion. Patients in stage IV showed either atrophy of the RPE layer, or a white hypertrophic scar or neovascularization of fibrous tissue in the macula. 

PCR-Based Sequencing of the BEST1 Gene

Peripheral blood samples were collected from all participants for genetic analysis. The genomic DNA was then extracted using a genomic DNA extraction and purification kit (Vigorous Whole Blood Genomic DNA Extraction Kit; Vigorous, Beijing, China), following the manufacturer's protocol. All exons and flanking splicing sites of the BEST1 gene were amplified by the PCR. The primer sequences and related information are available on request. PCR assays were carried out using standard reaction mixtures and purified amplified fragments were sequenced using an ABI Prism 373A DNA sequencer (Applied Biosystems, Foster City, CA, USA). A published cDNA sequence of BEST1 (GenBank NM_004183) was used to compare with the sequencing results. For the BEST1 gene, cDNA numbering +1 corresponds to A in the ATG translation initiation codon. 

Bioinformatics Analysis

The potential functional impact of all the candidate variants was investigated using three programs, including PolyPhen2 (http://genetics.bwh.harvard.edu/pph/, in the public domain), Mutation Taster (http://www.mutationtaster.org/, in the public domain), and SIFT (http://sift.jcvi.org/, in the public domain). NetGene2 Server (http://www.cbs.dtu.dk/services/NetGene2/, in the public domain), Berkeley Drosophila Genome Project (http://www.fruitfly.org/seq_tools/splice.html, in the public domain), and Human Splicing Finder (http://www.umd.be/HSF3/, in the public domain) were used to assess the effects of mutations on splicing sites. Finally, we verified novel variations identified in the sequencing by conducting allele-specific PCR (AS-PCR) analysis in 100 healthy controls and family members. 

Real-Time Quantitative PCR and Multiplex Ligation-dependent Probe Amplification (MLPA) Analysis

Real-time quantitative PCR (RQ-PCR) and MLPA analysis were carried out to detect any genomic DNA rearrangements of the BEST1 gene in two patients with no mutations identified and in six ARB patients with only one mutation identified. The RQ-PCR reactions were performed on a Rotor-Gene 6000 instrument (Corbett Research, Mortlake, NSW, Australia) in a 10-μL final volume, containing 300-nM primers and 1 μL (100 ng) genomic DNA, using the Eva Green PCR Master Mix (Bio-Rad, Hercules, CA, USA), as we previously described.²² The primers information was summarized in Supplementary Table S1. Each assay was performed in triplicate. The human GAPDH (glyceraldehyde-3-phosphate dehydrogenase) gene was used as an internal control. The relative quantitation (RQ) of the target gene was accomplished using RQ manager software (Bio-Rad) and was calculated with the 2^−ΔΔCt method. The threshold value for normal was set at 0.8 to 1.3. The ranges of RQ values for deletions and duplications were defined as 0.45 to 0.74 and 1.6 to 1.8, respectively. MLPA assay was performed with a SALSA MLPA probemix P367-A2 BEST1-PRPH2a (Amsterdam, the Netherlands), following the manufacturer's protocols; this kit contains probes for each exon of the BEST1 gene. 

Results

BEST1 Mutations

The 17 unrelated patients with BVMD revealed one heterozygous BEST1 mutation in 13 patients, compound heterozygous mutations in 3 patients (010312, 010356, and 010424), and no mutation in 1 patient. The 20 patients with ARB revealed two mutant alleles in 13 patients, one mutant allele in 6 patients, and no mutation in 1 patient (Supplementary Table S2). One patient (010255) with ARB carried a complex mutation in which two variants were in cis positions on the same chromosome. In total, we identified 36 distinct disease-causing variants of the BEST1 gene in this cohort, which included missense (28/36, 77.8%), splicing effect (4/36, 11.1%), nonsense (3/36, 8.3%), and frameshift small duplication (1/36, 2.8%) mutations (Supplementary Table S3).^23–28 The RQ-PCR analysis revealed no large deletions or insertions of the BEST1 gene in eight patients. 

Of the 36 mutations, 14 were solely detected in patients with BVMD, 20 mutations were solely identified in patients with ARB, and only 2 mutations (p.R13H and p.A195V) were found in patients with both BVMD and ARB. All mutations identified in the patients with BVMD were missense mutations mostly located in the N-terminal half of the protein (Fig. 1). The most frequent mutation was the mutation p.R218C, with an allele frequency of 17.6% (3/17), followed by the mutation p.A195V found in two probands (12%, 2/17), and the remaining mutations were detected only once (Supplementary Tables S2, S3). The mutations identified in the patients with ARB included 14 missense, 3 nonsense, 4 splicing effect, and 1 frameshift small duplication mutations. Of these, the mutation p.R255W was the most frequent mutation, with an allele frequency of 15% (6/40), followed by the mutation p.A195V occurring three times (7.5%, 3/40), five mutations (p.R13H, p.R25W, p.Y44H, c.868–2A>G, and c.949–1G>A) occurring two times (5%, 2/40), and the remaining mutations were identified only once (Supplementary Tables S2, S3). 

Figure 1

 

Protein model of bestrophin-1 proposed by Milenkovic et al.⁶ Mutations detected in this study are indicated with colors and shapes. Circles indicate missense mutations, rhombuses indicate nonsense or frameshift mutations, star-marks indicate splice-site mutations; yellow indicates BVMD, red indicates ARB, blue indicates both BVMD and ARB.

Novel Mutations in the BEST1 Gene

The current study revealed 20 different novel mutations in the BEST1 gene. Of these, 14 (70%) were missense mutations, 2 (10%) were nonsense, 3 (15%) were splicing effects, and 1 (5%) was a frameshift small duplication mutation. Neither the novel missense mutations nor the synonymous variant p.R255R and the variant c.*24C>T were detected in the 100 healthy controls (200 alleles), as determined by AS-PCR analysis. Moreover, most of these mutations were not found in public databases, including the Exome Variant Server and1000 Genomes Database (Supplementary Table S3). Nine novel missense mutations were predicted to be disease-causing or probably damaging by three in silico analysis programs (Polyphen2, Mutation Taster, and SIFT). The remaining five mutations (p.R19C, p.L40P, p.Y44H, p.I73M, and p.R355H) were predicted to be probably damaging or disease-causing by one or two of these three programs. The synonymous variant c.763C>A (p.R255R) and the variant c.*24C>T were predicted to be disease-causing by Mutation Taster for the alteration of the downstream splice site. The remaining nonsense, splicing effect, and frameshift small duplication mutations were considered as obviously pathogenic mutations. 

Genotype-Phenotype Correlation

Best Vitelliform Macular Dystrophy.

Sixteen unrelated patients with BVMD harbored at least one disease-causing BEST1 mutation, and cosegregation analyses were performed in 15 probands (93.7%) (Supplementary Table S2). Of these, four probands had a family history, and only one pedigree (010378) showed a clearly autosomal dominant inherited pattern. No de novo mutation was identified in the remaining nine sporadic unrelated patients; therefore, the mutations detected in these probands had been transmitted from one of their asymptomatic parents. Asymptomatic parents from five families underwent clinical evaluation, including fundus photograph and OCT. They all had a normal fundus appearance; three had a mildly abnormal OCT performance, including a thicker and more reflective appearance of the interdigitation zone. Only one of them underwent an EOG examination, and her Arden ratio was 1.3 (Fig. 2B). All patients experienced different extents of defects in visual acuity (from finger counting to 0.9), and some of them also complained of metamorphopsia. The mean disease onset age was 21.7 years (range, 3–48 years) and one-third of these patients had visual symptoms in their first decade. Three patients (010150, 010324, and 010361) carrying the mutation p.R218C had a late-onset age (36–48 years). Two patients (010312 and 010356) harboring compound heterozygous mutations did not show an early-onset age; however, the macular lesion in the left eye of proband 010356 was larger than the typical vitelliform lesion of BVMD (Fig. 2D). Of the 32 eyes of the 16 patients with BVMD, 1 eye had a stage 0 lesion, 8 eyes had stage II lesions, 6 eyes had stage III lesions, and 17 had stage IV lesions. No stage I lesions were observed in the eyes of this cohort of patients. In seven patients, the stage of the vitelliform lesion differed between both eyes (Supplementary Table S2). 

Figure 2

 

Pedigrees, fundus appearances, and macular OCT images of four BVMD families. (A) Pedigrees and segregation analysis (heterozygous mutation p.Y97H in family 010388; compound heterozygous mutations p.R19C/p.A195V, p.S7N/ p.R13H, and p.A195V/p.V317M in families 010312, 010356, and 010424, respectively). Squares denote males; circles denote females; solid symbols indicate affected; open symbols indicate unaffected; open symbols with a spot indicate a carrier; slashed symbols indicate deceased; an arrow below the symbol indicates the proband; + indicates wild-type. (B) Color fundus (CF) photos show overall normal findings in the right eye and a vitelliruptive lesion in the left eye of the proband in family 010388. Macular OCT images show a thicker and more reflective appearance of the interdigitation zone in the right eye and subfoveal hyperreflective material located between the RPE and the neuroretina in the left eye. CF photos of the mother (010388-1) in this family show normal findings for the macula in both eyes; macular OCT images present a thicker and more reflective appearance of the interdigitation zone in both eyes. (C) CF photos of the proband in family 010312 show a pseudohypopyon lesion in the right eye and an atrophic lesion in the left eye; the OCT images show bilateral subfoveal clumping of hyperreflective material located between the RPE and the neuroretina. (D) Fundus photographs of the proband in family 010356 show bilateral atrophic lesions in the macular region; OCT images show subretinal detachment and clumping of hyperreflective material located at the RPE level, accompanied with disruption of the ellipsoid zone above the lesions in both eyes. (E) Fundus photographs of the proband in family 010424 show fibrous tissue in the macula of the right eye and a vitelliform cyst in the left eye; OCT images clearly demonstrate choroidal neovascularization in the right eye and subretinal detachment in the left eye.

Autosomal Recessive Bestrophinopathy.

Nineteen patients with ARB in this cohort carried at least one BEST1 mutation. Except for proband 010286, whose parents' DNA was not available, the remaining 12 probands were determined to have compound heterozygous mutations on separate alleles, based on cosegregation analysis; this finding was consistent with autosomal recessive inheritance (Supplementary Table S2). The mean disease onset age was 20.9 years (range, 4–44 years), which is not significantly different from the onset age of BVMD. The patients had different extents of visual impairment (from no light perception to 0.8) and some had symptoms of metamorphopsia. Two unrelated patients (010289 and 010303) carrying different homozygous splicing effect mutations had an early-onset age (4 and 7 years), and the two probands harboring the same compound heterozygous mutations (p.A195V/p.R255W) showed a late-onset age (35 and 37 years). 

Clinical examination of the probands revealed a range of phenotypes. Most patients presented multifocal lesions in the maculae and around the optic disc. The typical lesions consisted of multifocal yellowish subretinal deposits, usually outlining the edges of the lesions, and subretinal fibrotic scars of variable sizes and numbers (Fig. 3B). Some patients also showed tiny yellow-white spots or pigment throughout the peripheral retina (Figs. 3C, 3D). The FAF examinations showed multifocal hyperfluorescent and hypofluorescent regions corresponding to the subretinal deposits and retinal scaring zone, respectively. Six probands (010387, 010392, 010408, 010413, 113360, and 113400) had a shallow anterior chamber depth (range, 1.94–2.44 mm; normal range, 2.5–3.0 mm), a short axial length (range, 21.21–23.03 mm; normal range, 24.00 mm), and a history of elevated IOP. They all underwent yttrium-aluminum-garnet laser peripheral iridotomy and trabeculoplasty or trabeculectomy due to ACG. Fundus examination showed glaucoma cupping and some irregular RPE atrophy or subretinal fibrosis in the maculae, with very tiny yellow or white spots (Fig. 4). All patients with ARB showed intraretinal cystoid fluid collection in or around the macula and most of them also displayed subretinal detachment in the macula by the OCT examination. 

Figure 3

 

Pedigree, fundus appearance, macular OCT images, and FAF images of families 010292, 010238, 010416, and 010397. CF photographs show multifocal vitelliform deposits throughout the post pole around the optic disc. The FAF images clearly show hyperfluorescent spots corresponding to the subretinal deposits depicted in the CF. The OCT images show serous neuroretinal detachment with intraretinal cystoid fluid collections and clumping of hyperreflective material located at RPE layer (B–D, E: 010397). (A) Pedigrees of families 010292, 010238, 010416, and 010397. Segregation analysis of the compound heterozygous mutations p.A195V/p.R255W, p.L109Y/p.T277M, and p.R25W/c.*24C>T and heterozygous mutation N99D in families 010292, 010238, 010397, and 010416, respectively. (B) CF photograph and the FAF image of the right eye of patient 010292 show a typical fundus appearance, as described above. (C) In addition to the multifocal vitelliform deposits in the posterior pole, the CF shows very tiny white spots throughout the retina, especially in the peripheral area. (The inset shows an enlarged view of the area, indicated by the black arrow). An FAF image shows hyperfluorescent spots corresponding to the subretinal deposits depicted in the CF. (D) A CF photo of the right eye of patient 010416 shows multifocal yellowish lesions in the posterior pole and around the superior temporal vascular arcade, and pigments dispersed in the midperipheral retina. The FAF image shows hyperfluorescent spots and hypofluorescent areas corresponding to the lesions and pigments shown in CF, respectively. The OCT image shows a mild cystoid edema in this patient. (E) The CF of the proband shows scarring in the macula and multifocal subretinal deposits near the superior temporal vascular arcade. The FAF image shows an area of hypofluorescence in the macula and patches of hypofluorescence. An OCT image shows subretinal detachment and subtle intraretinal cystoid fluid collections. The fundus, FAF, and OCT photos of the proband's parents (010397-1, 010397-2) are all normal, whereas the EOG of the mother shows a significantly decreased Arden ratio (1.3/1.2).

Figure 4

 

Fundus appearance of ARB patients with ACG. Three patients had diffuse yellow lesions in the macula, with or without yellow-white spots in the CF photographs. Two patients had subretinal multifocal deposits. All but one patient present typical glaucoma cups. FAF images show large hypofluorescent area in the macula and posterior pole. It is noteworthy that these four patients have lighter degree of subretinal detachment compared with those ARB patients without glaucoma. (A) The CF photo of patient 010387 (heterozygous mutation p.I123A fs15) shows a pale optic disc and defused yellowish lesion in the macula; the FAF image shows a large hypofluorescent area in the macula and posterior pole, the OCT image presents subtle cystoid edema and mild neurosensory retinal detachment. (B) The CF photo of patient 010392 (heterozygous mutation p.E35K) shows a diffuse yellowish lesion with tiny yellow-white spots scattered in the macula and near the inferior temporal vascular arcade. The C/D ratio is normal. The FAF image shows a large hypofluorescent area in the posterior pole and several hyperfluorescent spots located to the temporal side of the optic disc. The OCT image shows conspicuous cystoid macula edema, but subretinal detachment is mild. (C) A CF photo of patient 113360 (heterozygous mutation p.R255W) shows a diffuse yellowish lesion in the macula and the C/D ratio is 0.8. The FAF image shows a widespread area of hypofluorescence in the posterior pole and OCT shows cystoid macula edema with mild subretinal detachment. (D) A CF photo of patient 113400 (compound heterozygous mutation p.R25W/p.R255R) shows a typical glaucoma cup and multifocal yellowish deposits in the macula. OCT shows obvious cystoid macula edema, but mild subretinal detachment. (E) A CF photo of patient 010408 (heterozygous mutation p.L40P) shows a C/D ratio of 0.8 and central yellowish subretinal depositions. The FAF image shows hyperfluorescent spots in the macula and patchy hypofluorescence around the superior temporal vascular arcade. OCT shows central neuroretinal detachment together with cystoid macular edema.

Two patients of family 010413 are worth describing in detail. The 57-year-old proband was diagnosed with primary ACG and had undergone trabeculectomy 30 years previously. Since then, the IOP of her right eye had stayed within the normal limits, but the IOP of her left eye was still poorly controlled even after cyclocryotherapy, and its visual acuity had decreased to no light perception. In 2014, she was diagnosed with malignant glaucoma at our clinic, and she underwent combined phacoemulsification, intraocular lens implantation, goniosynechialysis, anterior vitrectomy, and peripheral iridotomy in her right eye. In a recent follow-up, she was diagnosed with suspected ARB, as an ophthalmologist observed some tiny yellowish lesions scattered over the posterior pole (Fig. 5B). The compound heterozygous mutations p.Q74X/ p.R255W were detected in this proband. During her family history review, she claimed that one of her brothers (II: 2), her sister (II: 4), and one nephew (010413–1, III: 4) had been diagnosed with macular degeneration, indicating a seemingly autosomal dominant pattern (Fig. 5A). Cosegregation analysis show that the affected brother (II: 2) carried the same mutations, whereas her nephew (010413–1, III: 4) carried a different compound heterozygous mutation (p.R255W/p.W287X) (Fig. 5A). Her brother (II: 2) refused to undergo ophthalmologic examinations, so we were unable to obtain his detailed clinical data. Her nephew (010413-1, III: 4), a 38-year-old male, had an analogic medical history. He had experienced traumatic retinal detachment in his right eye approximately 20 years previously and had not received any treatment. Five years previously, he was diagnosed with primary ACG and underwent trabeculectomy of the left eye. In 2014, he was diagnosed with malignant glaucoma (left eye) and underwent phacoemulsification, intraocular lens implantation, and anterior vitrectomy of the left eye in our hospital. The IOP was well controlled afterward. Similar to his aunt, he was diagnosed with suspected ARB in a recent follow-up because of the yellowish deposits in the macula and around the optic disc (Fig. 5C). 

Figure 5

 

Pedigree of family 010413, and fundus appearance, macular OCT images, and FAF images of the proband and the affected relative. (A) Pedigree of family 010413 and segregation analysis of the compound heterozygous mutations p.Q74X/p.R255W and p.R255W/p.W287X. (B) Fundus photograph of the right eye of patient 010413 shows a typical glaucoma cup and a diffuse yellowish lesion with multifocal subretinal depositions in the macula and posterior pole. An OCT photo shows hyperreflective material deposit in the RPE layer. (C) A fundus photograph of the left eye of patient 010413-1 shows a C/D ratio of 0.8 and tiny yellow or white spots in the posterior pole. The FAF image shows a large hypofluorescent area in the posterior pole and hyperfluorescent spots corresponding to the yellow spots in the fundus photograph. The OCT image shows intraretinal cystoid fluid collections in the macula and disruption of the ellipsoid zone.

Discussion

In this study, we screened for BEST1 mutations in 37 unrelated Chinese patients with BVMD or ARB and presented the clinical characteristics of patients with BEST1 mutations. Our mutation-detection rate for the BEST1 gene was 94.1% in the patients with BMVD (16/17) and 95% in the probands with ARB (13/20 with two mutations and 6/20 with one mutation). This high mutation-detection rate is consistent with previous reports.^16,20,21 

More than 250 disease-causing mutations of the BEST1 gene have been reported, and yet 56% (20/36) of the mutations detected in this cohort of patients were novel, suggesting that the mutation spectrum of Chinese patients is somewhat different from those of patients of other ethnicity. In the patients with BVMD, 16 mutations were identified and all were missense mutations, which is consistent with previous observations^15,29 and further confirmed the hypothesis of a possible dominant negative effect of the abnormal protein in determining the BVMD phenotype.¹ Except for the mutation p.V317M, the remaining 15 mutations were exclusively identified in the coding exons 2 to 6, which encode the N-terminal half of the protein. This part of the protein shows the highest evolutionary conservation, implying the function importance.⁶ Most of these mutations are clustered in the N-terminal intracellular region and the second transmembrane domain, which are within the two hot regions (codons 6–30 and 80–104) for BVMD.^1,29 Interestingly, no mutation was identified in exon 8 (encoding amino acids 290 to 316), which is one of the mutation hot regions that harbors a disproportionate fraction (26%) of the known BVMD disease-causing mutations.^4,12,29,30 Eleven of the 16 mutations were transmitted in an autosomal dominant pattern. The mutation p.R218C was the most common mutation, identified in three unrelated patients. This mutation has been reported in several patients with BVMD of different ethnic backgrounds^3,4,11,17,29 and is frequently observed in Italian patients.¹⁷ Two previous studies reported this mutation in four unrelated Chinese patients with BVMD,^26,27 suggesting that it might also be a common mutation for Chinese patients. Three compound heterozygous mutations (p.R19C/p.A195V, p.S7N/p.R13H, and p.A195V/p.V317M) were detected in three patients with BVMD. Apart from p.R19C, the other four mutations have been reported previously, and the mutations p.A195V and p.R13H were identified both in the patients with BVMD and with ARB in the current study. The mutation p.S7N in a compound heterozygous state was first identified in a Japanese patient who also showed a typical BVMD phenotype.²³ A heterozygous p.R13H was first reported in a Caucasian patient with a BVMD phenotype; however, the patient had no well-described clinical features and no cosegregation analysis was conducted for this family.¹¹ In the current study, the patient harboring the compound heterozygous mutations p.S7N/p.R13H exhibited a BVMD phenotype (Fig. 2D), whereas the patient carrying the homozygous p.R13H showed a typical ARB phenotype. Their parents, who carried only one heterozygous mutation, showed no abnormal macular appearance, based on fundus photography and OCT examination. 

The patients with ARB revealed 22 distinct mutations, with 64% (14/22) being missense mutations. Unlike the BVMD mutations, more than one-third of the ARB mutations (9/22) were located in exons 7 to 11, which encode the C-terminal half of the protein.⁶ The remaining mutations were mainly clustered in the first transmembrane domain and the intracellular regions (Fig. 1). In the current study, the mutation p.R255W was the most frequent mutation for the patients with ARB, with an allele frequency of 15% (6/40), and all were in the heterozygous state. This mutation, in a heterozygous state, was first identified in a Chinese patient with BVMD,²⁷ and since then has been reported in two unrelated Chinese patients with BVMD in a compound heterozygous state.²⁶ Two recent studies reported several Japanese patients carrying the mutation p.R255W, in either the compound heterozygous or the homozygous state. These patients showed a phenotype of ARB,^31,32 whereas their heterozygous parents were entirely normal, both clinically and electrophysiologically.³² In the current study, seven patients with the mutation p.R255W (five in the compound heterozygous state and two in the heterozygous state) had a clearly ARB phenotype. Therefore, we speculated the mutation p.R255W might be a common recessive mutation for Asian patients. The mutation p.A195V is the second commonest mutation and was detected three times in the ARB patients, all were heterozygous. The mutation p.A195V is one of the most prevalent mutations among those patients with biallelic mutations,³² and the patients carrying the heterozygous mutation p.A195V compounded with another heterozygous mutation exhibited either a phenotype of BVMD or of ARB, whereas the individuals carrying the heterozygous mutation p.A195V alone did not shown any phenotype of BVMD or ARB.^15,23,32 In the current study, the mutation p.R25W was identified in two ARB patients in the compound heterozygous state (one coupled with a synonymous change c.763C>A and another one with a variant c.*24C>T). The father of proband 010397 carrying the mutation c.*24C>T was healthy, both clinically and electrophysiologically, whereas her mother, who harbored the mutation p.R25W, had an abnormal EOG (Fig. 3E). The mutation p.R25W has been reported many times and was highly prevalent among Italian patients with BVMD.^4,17,21 We speculated that the mutation p.R25W might be similar to the mutation p.D303G described in a Japanese family,³¹ which caused BVMD in a heterozygous state and ARB in a homozygous or compound heterozygous state. The finding of synonymous mutation as a disease-causing mutation is not rare. In 2010, Davidson et al.³³ reported that one synonymous mutation p.G34G caused ARB with another mutation in the compound heterozygous state by altering pre-mRNA splicing. The novel synonymous mutant allele p.R255R (c.763C>A) might cause the alteration of the downstream splice site predicted by Mutation Taster and Human Splicing Finder. The novel variant c.*24C>T was also predicted as deleterious by Mutation Taster. These two novel variants were not found in 100 healthy controls and cosegregated with the family; therefore, we could not exclude their pathogenicity. Additional functional analyses are required to determine their pathogenicity in the future. 

In the current study, almost all the patients with BVMD presented a typical macular appearance and no clear genotype-phenotype correlation was observed; however, three patients carrying the heterozygous mutation p.R218C had a late-onset age, which suggested the existence of a genotype-phenotype association. Two of the three patients with BVMD who carried the biallelic BEST1 mutations (010312 and010356) did not show an early-onset age. Several articles reported that patients with BVMD may carry biallelic BEST1 mutations.^15,26,27,34 One recent Chinese study identified biallelic BEST1 mutations in 4 of 10 families with BVMD.²⁶ The other clinical feature in our patients with BVMD was the high percentage (69%, 9/13) of “sporadic patients.” The exact mechanisms of this reduced penetrance and variable expression are still unclear and may be related to other genetic and environmental modifiers. 

When compared with BVMD, ARB is a new phenotype with a considerable clinical variability, making it a more challenging diagnosis.^5,21,31 In this cohort, one patient, whose fundus showed many tiny yellow-white dots in the peripheral area, was misdiagnosed with chorioretinitis, and three patients were misdiagnosed with juvenile X-linked retinoschisis due to the obvious collection of intraretinal cystoid fluid in the macula. Two patients (010289 and 010303) carrying homozygous splicing mutations had an early-onset age, and two patients (010286 and 010292) with compound heterozygous mutations (p.A195V/p.R255W) had a late-onset age. Due to the small sample number, we were unable to draw the conclusion that patients harboring two deleterious mutation alleles might have an early-onset age. Angle-closed glaucoma was observed in 31% (6/20) of the ARB probands, which is lower than the rate (approximately 50%) found in the Caucasian ARB patients.^5,21,35 A recent Japanese study found no ACG in nine patients from seven unrelated families.³¹ Currently, only one study has reported four patients with a homozygous mutation p.C251Y in a Chinese family affected with ACG.³⁶ 

Although we analyzed all exons and flanking regions of the BEST1 gene, six of our patients diagnosed with ARB had only one disease-causing mutation identified. Monoallelic BEST1 mutation was identified in 6 of 20 probands with ARB (30%); the number is much higher than the rate 14% (1/7) reported by Boon and colleagues.²¹ All patients with only one mutation presented bilateral multifocal vitelliform lesions and four of them had ACG (Figs. 3D, 4A–C, 4E), so the other undetected mutation may be in the promoter region or in some deep intron regions. The patient (010360) with BVMD with no mutation identified was a 54-year-old male who showed a typical pseudohypopyon in the maculae in both eyes, but he did not undergo an EOG examination. Due to his late-onset age (53 years), his phenotype may be caused by mutation of other genes, as described previously.³⁷ 

In conclusion, our results revealed that the mutation spectrum of the BEST1 gene in patients of Chinese descent differs from those of Caucasian patients. Mutations that cause ARB differ from those that cause BVMD. Patients with biallelic BEST1 mutations may present either BVMD or ARB phenotypes. Screening of the BEST1 gene is essential for the correct diagnosis of BVMD and ARB because of considerable phenotypic variations. 

Acknowledgments

Supported by the High-level Talents training plan of the health system of Beijing (No. 2013-2-021). The funding organization had no role in the design or conduct of this research. 

Disclosure: L. Tian, None; T. Sun, None; K. Xu, None; X. Zhang, None; X. Peng, None; Y. Li, None