Genotype–Phenotype Analysis of Bietti’s Crystalline Dystrophy in Patients with CYP4V2 Mutations

Abstract

purpose. To evaluate the genotypic and phenotypic correlations of Bietti’s crystalline dystrophy (BCD) in patients with the CYP4V2 gene by mutation screening and clinical and electrophysiological assessment.

methods. Eighteen Chinese patients in 13 families with BCD were recruited for full ophthalmic examinations, optical coherence tomography (OCT), and visual electrophysiological tests, including electrooculography (EOG), full-field electroretinography (ERG), and multifocal electroretinography (mfERG). Peripheral venous blood was obtained from all index patients and their family members for genomic DNA extraction and CYP4V2 sequence screening by direct sequencing.

results. All 18 patients with BCD had mutations in the CYP4V2 gene: five were novel (Y219H, W244X, D324V, P396L, and R400C) and four had been reported. A common mutation occurred at the splice site IVS6-8del17bp/insGC of 12 patients, four being homozygous. OCT showed the presence of intraretinal crystals in all patients. Patients with more severe thinning of the retina had worse visual acuity, and there was moderate correlation between the OCT central foveal thickness and visual acuity (Spearman ρ = 0.46, P = 0.005). Patients with splice site mutations (i.e., homozygous IVS6-8del17bp/insGC or compound heterozygous IVS6-8del17bp/insGC and IVS8-2A>G) had lower EOG Arden index (P = 0.014) and were more likely to have a nonrecordable scotopic full-field ERG (P = 0.003) and nonrecordable 30-Hz flicker ERG (P = 0.043).

conclusions. BCD patients with homozygous IVS6-8del17bp/insGC or compound heterozygous IVS6-8del17bp/insGC and IVS8-2A>G mutations appeared to have more severe disease phenotype based on electrophysiological testing. The level of visual loss in BCD is related to the severity of retinal thinning.

Bietti crystalline dystrophy (BCD) is a progressive retinal degenerative disease first described by Bietti in 1937. ¹ It is characterized by progressive atrophy of the retinal pigment epithelium (RPE) and choriocapillaris, with the presence of limbal corneal and retinal crystals. ^{1 2} Although corneal crystals were described in the original cases, ¹ they were not present in many of the subsequently reported cases and therefore corneal crystals are not required for clinical diagnosis of BCD. ³ On the other hand, the pure retinal form of BCD is more common in Asians, in particular Chinese and Japanese, compared with Caucasians, ^{4 5 6 7} suggesting phenotypic variations of this disorder among different ethnic populations. ² The disease usually manifests between the second and fourth decades of life, and patients present with progressive night blindness, reduced vision, and visual field constriction. ⁴  

A causative gene for BCD, CYP4V2, has been identified, and it has been confirmed that BCD is a genetic disorder with autosomal recessive inheritance. ^{8 9} The CYP4V2 gene belongs to a member of the cytochrome P450 hemethiolate protein superfamily and is responsible for oxidizing various substrates in the metabolic pathway. Metabolic studies of cells cultured from patients with BCD were found to have abnormally high triglycerides and cholesterol storage and reduced metabolism of labeled fatty acid precursors into n-3 polyunsaturated fatty acids (n-3PUGA). ¹⁰ These findings suggest that BCD may result from a systemic abnormality of lipid metabolism. Though various mutations of the CYP4V2 gene have been reported in Japanese and Chinese populations, ^{6 11 12 13 14} specific genotype–phenotype correlations have not been defined. In this study, we sought to characterize the genotype–phenotype association in Chinese patients with BCD who had CYP4V2 mutations. We identified five novel changes in the CYP4V2 gene and showed that the homozygous IVS6-8del17bp/insGC or compound heterozygous IVS6-8del17bp/insGC and IVS8-2A>G mutations have a more severe disease phenotype. 

Materials and Methods

Recruitment of Subjects

This study was cross-sectional, and patients with BCD were recruited from Hong Kong Eye Hospital and Prince of Wales Hospital. All index patients with a diagnosis of BCD based on the typical fundus findings and clinical examinations were recalled for further assessment. Family members of the index patients were also invited for clinical examination and genetic assessment. Written informed consent was obtained from all patients and family members. The research protocol was approved by an ethics committee, and the study was performed in accordance with the tenets of the Declaration of Helsinki. 

Clinical Evaluations

All index patients and their family members were given a complete ophthalmic examination including visual acuity testing, noncontact tonometry, slit lamp examination and dilated fundus examination. Patients with BCD also underwent optical coherence tomography (OCT) assessment of the macula and electrophysiological tests including electrooculography (EOG), full-field electroretinogram (ERG), and multifocal electroretinogram (mfERG). OCT was performed with the StratusOCT (Carl Zeiss Meditec, Dublin, CA) in the line scan mode to obtain horizontal and vertical scans of 5-mm lengths, as well as the fast macular scan mode. The central foveal thickness (CFT) was measured with the retinal thickness mode measured from the inner surface of the RPE to the inner surface of the retina at the fovea. EOGs and full-field ERGs were recorded with a computerized recording unit (Nicolet Biomedicals, Madison, WI) according to ISCEV (International Society for Clinical Electrophysiology of Vision) standards. ^{15 16} mfERG recordings were performed (VERIS 4.8 system; ElectroDiagnostic Imaging, San Mateo, CA) with reference to the ISCEV guidelines. ¹⁷  

Mutation Analysis

Genomic DNA was extracted from peripheral venous blood leukocytes (QIAamp DNA Blood kit; Qiagen, Hilden, Germany). All 11 coding exons, including intron–exon boundaries, of the CYP4V2 gene (GenBank accession number NC_000004; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) were amplified by polymerase chain reaction (PCR) using primers published previously. ⁹ The amplified PCR products were purified, and directly sequenced using a dye terminator cycle sequencing kit (BigDye Terminator Cycle Sequencing Reaction Kit, ver. 3.1; Applied Biosystems, Inc. [ABI] Foster City, CA) on an automated capillary DNA sequencer (model 3130XL; ABI). The presence of IVS6-8del17bp/insGC, IVS8-2A>G, and H331P mutations was also investigated in 190 unrelated healthy Hong Kong Chinese control subjects by direct sequencing. Multiple sequence alignment of seven eukaryotic CYP4V2 homologues and CYP4 subfamily members was performed and analyzed with an online analysis tool (TCoffee, provided in the public domain by Institut de Biologie Structurale et Microbiologie, Marseille, France, at http://igs-server.cnrs-mrs.fr/Tcoffee/tcoffee_cgi/index.cgi). 

Statistical Analysis

The data were entered into statistical software (SPSS for Windows, ver. 11.0, SPSS Inc., Chicago, IL) for analysis. Analysis of categorical variables was performed with the Fisher exact test, and comparisons of continuous variables were performed with the nonparametric Mann-Whitney test. Correlation analyses between continuous variables were also conducted with bivariate Spearman correlation analysis, and the Spearman correlation coefficient ρ was calculated. P ≤ 0.05 was considered statistically significant. 

Results

Subject Demographics and Genotype Analysis

Eighteen patients with a clinical diagnosis of BCD in 13 unrelated families were recruited for the study (Table 1) . All patients were of Chinese ethnicity. The mean ± SD age was 50.5 ± 14.0 years (range, 30–75). The mean ± SD age at onset of the disease was 43.4 ± 13.5 years (range, 25–70). 

Nine mutations in the CYP4V2 gene were identified in our patients, including five novel (g.959T>C, p.Y219H; g.1037G>A, p.W244X; g.1276A>T, p.D324V; g.1492C>T, p.P396L; and g.1503C>T, p.R400C) and four previously reported mutations (Fig. 1 , Table 1 ). The autosomal recessive nature of BCD was confirmed in the listed families. The most common mutation observed was the IVS6-8delTCATACAGGTCATCGCG/insGC (IVS6-8del17bp/insGC) mutation found in 12 (66.7%) patients of 9 families, followed by H331P and IVS8-2A>G each in 4 (22.2%) patients of 2 and 3 families, respectively. The IVS6-8del17bp/insGC deleted the 3′ splicing acceptor site of intron 6 and caused skipping of exon 7. ^{6 9 18} Among the 190 unrelated controls, IVS6-8del17bp/insGC and H331P mutations were observed in one individual each (heterozygote carrier frequency = 1/190 = 0.005). None of the unrelated control subjects was found to have the IVS8-2A>G mutation. All the encoded amino acid residues involved in the six missense mutations were completely conserved among the CYP4V2 homologues and CYP4 subfamily members in the multiple sequence alignment (Fig. 2) , indicating a likely role of these residues in the functions or structural conformation of the CYP4V2 protein. 

Clinical Assessment

Seventeen (94.4%) of the 18 patients with BCD were symptomatic and 1 (5.6%) was asymptomatic, although there were signs of mild retinal changes. The commonest symptom on presentation was poor vision (10 patients, 55.6%), followed by night blindness (5 patients, 27.8%) and the presence of floaters (3 patients, 16.7%). The median visual acuity of the patients was 20/70 (range, hand motion to 20/25). The mean ± SD spherical equivalent refraction was −3.2 ± 3.2 D (range, +1.25 to −10.5) and the mean ± SD intraocular pressure was 13.5 mm Hg (range, 9.7–17.7). In some patients, the visual acuity was better in the fellow eye, mainly due to the less severe central macular involvement compared with the eye with worse visual acuity. 

Slit-lamp examination revealed corneal limbal crystals in both eyes of one patient (patient 17) and bilateral nuclear sclerotic cataract was found in five patients. Dilated fundus examination showed the presence of intraretinal crystalline deposits in the posterior pole of all 18 patients. There were some variations in the clinical phenotypes and the amount of crystalline deposits among the patients (Fig. 3) . In addition to retinal crystals, other retinal findings included bone spicule pigmentations (Figs. 3A 3B 3E 3G) , retinal vascular attenuation (Figs. 3C 3E) , diffuse choriocapillaris atrophy (Figs. 3A 3E 3H) , and retinal scarring (Fig. 3G) . Even in patients with the same genotype, the clinical phenotypes may be rather different. For example, in the three brothers with the homozygous H331P mutation (patients 13,14, and 15), more severe fundus changes with poorer visual acuity were found in the two older brothers, whereas the youngest brother showed mild retinal changes and was asymptomatic (Figs. 3F 3G) . Fundus appearance with OCT and electrophysiological findings of three selected patients are displayed in Figures 4to 6 . 

OCT Examinations

OCT examinations of the macula showed the presence of multiple intraretinal hyperrefractive lesions within the neurosensory retina in all patients (Figs. 4C 4D 5C 5D 6C 6D) . The mean ± SD CFT was 192.8 ± 48 μm (range, 118–300). Fifteen (41.7%) of the 36 eyes showed thinning of the retina at the central macular region outside the 5% limit of the normal population. The severity of retinal thinning was moderately correlated with the level of visual loss, with patients having more severe retinal thinning associated with worse visual acuity (Spearman ρ = 0.46, P = 0.005). 

Electrophysiology Findings

Table 2summarized the EOG, full-field ERG, and mfERG results in the 18 patients with BCD, all of whom had a reduced or borderline reduced EOG Arden index suggestive of diffuse RPE dysfunction. The mean ± SD EOG Arden index of all eyes was 1.38 (range, 1.0–1.8). No significant correlation was found between visual acuity and the Arden index (Spearman ρ = −0.18, P = 0.29). 

Full-field ERG of the patients showed various types of findings as summarized in Table 3 . For the scotopic ERG, seven (38.9%) patients showed a nonrecordable waveform, four (22.2%) had reduced amplitude with delayed implicit time, and seven (38.9%) had reduced amplitude with normal implicit time. For the maximal response ERG, four (22.2%) patients had nonrecordable waveform, nine (50.0%) had reduced amplitude with delayed implicit times, three (16.7%) had reduced amplitude with normal implicit times, and two (11.1%) had normal amplitude and implicit times. Photopic ERG showed nonrecordable ERG response in four (22.2%) patients, reduced amplitude with delayed implicit times in seven (38.9%), reduced amplitude with normal implicit times in five (27.8%), and normal amplitude and implicit times in two (11.1%). For the 30-Hz flicker ERG, 3 (16.7%) patients had nonrecordable waveforms, 13 (72.2%) had reduced amplitude, and 2 (11.1%) had normal amplitude. 

For the mfERG recordings, seven (38.9%) patients had nonrecordable mfERG recordings. Of the remaining 11 patients, 9 (50.0%) had severe reductions in the response amplitudes of both the central and peripheral macula, and 2 (11.1%) had a mild reduction of the central mfERG response amplitude with normal peripheral macular response amplitude. 

Comparisons of the Phenotypic Findings between Patients with Splice Site Mutations versus Those with Exon Mutations

To evaluate whether patients with splice site mutations causing skipped exons might have more severe disease phenotype, we grouped the patients with BCD into those with splice site mutations—that is, patients with splice site mutations on both alleles (homozygous IVS6to8del17bp/insGC or compound heterozygous IVS6to8del17bp/insGC and IVS8 to 2 A>G mutations) and patients without the two splice site mutations in both alleles. The phenotypic findings between the two groups were then compared. 

The mean ± SD age of patients with splice site mutations in both alleles was 48.3 ± 11.3 years, compared with 51.9 ± 15.8 years for those with exon mutations. The difference was not significant (Mann-Whitney test, P = 0.62). The mean ± SD age at onset of symptoms in patients with splice site mutations in both alleles was 40.6 ± 11.6 years, compared with 45.4 ± 15.0 years for patients with exon mutations (Mann-Whitney test, P = 0.43). The mean EOG Arden index in eyes with splice site mutations was 1.28, compared with 1.44 in eyes with exon mutations. The difference between the two groups was statistically significant (Mann-Whitney test, P = 0.014). Analysis of the ERG recordings showed that six (85.7%) of the seven patients with mutations in the introns had nonrecordable scotopic ERGs, whereas only 1 (9.1%) of the 11 patients with exon mutations had nonrecordable scotopic ERGs (Fisher exact test, P = 0.002). For the 30-Hz flicker ERG, a nonrecordable waveform was observed in 3 (42.9%) of the 7 patients with splice site mutations, compared with none in 11 patients with mutations within the exons (Fisher exact test, P = 0.043). No significant difference was observed in the proportion of patients with nonrecordable maximal response ERG, photopic ERG, or mfERG between patients with splice site or exon mutations. Also, the mean visual acuity and mean CFT between the two groups was similar (Mann-Whitney test, P = 0.43 and P = 0.94, respectively). 

Discussion

In this study, we evaluated the genotypes and phenotypes of 18 patients with BCD in 13 unrelated families of Chinese ethnicity. This study is one of the largest series to date performed in the assessment of the genotype–phenotype association in BCD. Nine mutations, including two splice-site, one nonsense, and six missense mutations were identified in the CYP4V2 gene. The heterozygote carrier frequency for CYP4V2 mutation was determined to be 0.005 in our analysis of 190 unrelated healthy Hong Kong Chinese controls. This is consistent with the reported gene frequency in an epidemiologic survey of the general population in China. ⁵ Among the mutations, the IVS6-8del17bp/insGC is a common mutation in our BCD patients. It had been reported to be a frequent founder mutation in East Asian populations. ^{5 6 12} This 3′ splicing acceptor site change was expected to cause an in-frame deletion of 62 amino acid-encoding exon 7, which was confirmed by reverse transcriptase (RT)-PCR. ^{6 9 18} Another 3′ splicing acceptor site mutation, IVS8-2A>G, was predicted to disrupt the splicing of intron 8, resulting in an in-frame skipping of 45-amino-acid–encoding exon 9. ^{9 13} The novel W244X nonsense mutation should be a pathogenic null mutation, since the translated protein product was truncated early, with 282 amino acids shorter than wild-type and the deleted region included the heme coordinating I and L helices. ⁹  

All six amino acid substitutions identified in our study are completely conserved across species and CYP4 subfamilies in the multiple sequence alignment (Fig. 2) , indicating that they are essential for the structural and/or functional roles of the CYP4V2 protein. The positively charged R400, located in the β-strand region after the K helix, has been considered a mutation hotspot, as substitutions for C and H occur in this residue (patient 10). This mutation was predicted to alter heme coordination. The novel D324V substitution, located in the central I helix, would neutralize the negative charge in this residue and alter the local isoelectric point (pI) around the active site. The H331P (in the central I helix) and P396L (in the β-strand after the K helix) mutations result in insertion or deletion of a proline residue and may influence the stability of the protein heme coordination, depending on the secondary structures. ⁹ For the novel Y219H mutation, despite the similar aromaticity and steric hindrance between tyrosine and histidine, the local pI would be affected due to introduction of a charge from histidine into this position and therefore the interaction of the F helix with other secondary structures would be affected. 

The use of OCT in the assessment of BCD has been described in a case report. ¹⁹ OCT enables the in vivo evaluation of the location of retinal crystals in BCD as well as the measurement of retinal thickness. Our results showed that OCT CFT correlated moderately with the level of visual acuity of the patients. Retinal thinning with atrophy of the neurosensory retina may be due to the degeneration of photoreceptors as well as the inner retinal cells. Assessment of retinal thickness by OCT thus provided a useful parameter in determining the clinical severity of BCD. 

Various electrophysiological abnormalities in EOG and full-field ERG have been reported in patients with BCD. ^{6 20 21 22} These include a low EOG Arden index, the absence of an ERG waveform, subnormal scotopic and photopic ERG a- and b-wave amplitudes, decreased oscillatory potential (OP) amplitude, and reduced 30-Hz flicker ERG amplitude. ^{6 20 21 22} The variation in the results of electrophysiologic testing may be due to testing at different stages of the disease or to the heterogeneity of BCD. ^{6 20 21 22} It is known that EOG and ERG abnormalities may precede loss in visual acuity and visual field deterioration by many years. ²¹ It has also been suggested that the progression of BCD follows a rod–cone dystrophy pattern, and different stages of disease progression have been proposed. ²⁰ At the early stage of BCD, the RPE–choriocapillaris complex is mainly affected, and therefore reduction in EOG Arden index is one of the first abnormalities detected in BCD. This finding was confirmed in our study, as the EOG Arden index of all patients were found to be reduced. As the disease progresses, the number of photoreceptors may decrease with shortening of outer segments, and finally retinal phototransduction is affected in severe disease. ²¹ Our full-field ERG results suggest that cone functions are preserved in less severe form of BCD, as some patients were found to have a recordable photopic ERG despite having a nonrecordable scotopic ERG. 

Unlike missense mutations, the exon-skipping mutations are expected to cause a gross conformational change in the CYP4V2 protein structure. ⁹ Therefore, patients with homozygous IVS6-8del17bp/insGC or compound heterozygous IVS6-8del17bp/insGC and IVS8-2A>G may exhibit a more severe disease phenotype than those with homozygous or heterozygous amino acid substitution mutations. Our electrophysiological findings supported this proposition as our patients with splice site mutations had significantly lower EOG Arden index compared with patients with exon mutations. Patients with splice site mutations were also more likely to have nonrecordable scotopic ERGs and 30-Hz flicker ERGs. However, despite the more severe ERG changes and EOG reduction, no significant difference in visual acuity was observed between the two groups of patients. This result may be due to the preservation of the central macular function in the patients, as the mfERG findings were not significantly different between those with spice site and exon mutations. 

Our study showed that genetic assessment of patients with BCD may provide some clues into the severity of retinal diseases. Knowledge of alterations in amino acid sequence and levels and stability of protein conformation attributed to exon skipping would be useful in predicting disease progression. Among the three affected brothers in family 10 with the homozygous H331P mutation, the clinical phenotype appeared different, and the youngest brother was still asymptomatic visually at the time of the study. This individual should be observed for development of BCD later in life, and a regular follow-up has been arranged for him. It is also noted that there may be considerable variability in the amplitude of the maximal response ERG amplitude among patients with homozygous IVS6-8del17bp/insGC mutations. ^{6 11 12} For example, in the studies by Lin et al. ⁶ and Wada et al., ¹¹ some patients with homozygous IVS6-8del17bp/insGC mutations still had relatively preserved full-field ERG responses. In another study by Lee et al., ¹² a patient with a compound heterozygous mutation carrying one splice site mutation was also found to have nonrecordable full-field ERGs with more severe phenotype than did some patients carrying homozygous splice site mutations. Moreover, patients with missense mutations are expected to have more severe phenotypes as was observed in a patient with a homozygous T340X mutation in the study by Wada et al. ¹¹ However, clinical assessment in our patient with the W224X mutation (patient 4), as well as in a patient with the Q450X mutation reported by Lin et al., ⁶ demonstrated that the phenotype was less severe than that of other patients. Therefore, other environmental or modifying genetic factors may exist in influencing the severity in the deterioration of retinal function in BCD. Further research into the structure and function of the CYP4V2 protein as well as into the biochemical profiles of the patients would be beneficial in improving the understanding in the natural course of BCD and provide information to translate into the development of potential therapy for this progressive retinal disorder. 

In conclusion, this study of Chinese patients with BCD provides new information on phenotype–genotype correlation with regard to the CYP4V2 sequence alterations. Homozygous IVS6-8del17bp/insGC or compound heterozygous IVS6-8del17bp/insGC and IVS8-2A>G mutations appear to lead to a more severe disease phenotype. 

 Submitted for publication June 4, 2007; revised July 18, 2007; accepted September 5, 2007.

 Disclosure: T.Y.Y. Lai, None; T.K. Ng, None; P.O.S. Tam, None; G.H.F. Yam, None; J.W.S. Ngai, None; W.-M. Chan, None; D.T.L. Liu, None; D.S.C. Lam, None; C.P. Pang, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Demographics and Genotypes of 18 Chinese Patients with BCD

Patient (Family) No.	Gender/Age (y)	Age at Onset	Visual Acuity			Fundus Findings	Mutation
								RE	LE
1 (1)*	F 70	60	HM	HM	CD, BSP, RVA, CA	Homozygous IVS6-8del17bp/insGC
2 (1)*	F 44	35	20/40	20/200	CD, BSP	IVS6-8del17bp/insGC + IVS8-2A>G
3 (1)*	M 42	35	20/70	20/70	CD, BSP	IVS6-8del17bp/insGC + IVS8-2A>G
4 (2)	M 59	55	20/40	20/30	CD, CA	IVS6-8del17bp/insGC + W244X, †
5 (3)	M 53	45	20/25	FC	CD, BSP, CA	Homozygous IVS6-8del17bp/insGC
6 (3)	F 52	49	20/50	20/70	CD, BSP, CA	Homozygous IVS6-8del17bp/insGC
7 (4)	F 68	51	20/50	20/400	CD, BSP, CA	Homozygous H331P
8 (5)	M 75	70	20/70	20/50	CD, CA	IVS6-8del17bp/insGC + P396L, †
9 (6)	F 30	29	20/30	20/30	CD, BSP, CA	IVS8-2A>G + D324V, †
10 (7)	M 63	54	FC	20/400	CD, BSP, RVA, CA	R400H + R400C, †
11 (8)	F 70	62	20/70	20/70	CD, BSP, CA	IVS6-8del17bp/insGC + Y219H, †
12 (9)	F 31	28	20/30	20/40	CD, BSP, CA	IVS6-8del17bp/insGC + H331P
13 (10)	M 47	31	HM	20/40	Few CD, BSP, RVA, CA	Homozygous H331P
14 (10)	M 45	37	20/70	20/40	Few CD, BSP, RVA, CA	Homozygous H331P
15 (10)	M 42	N/A	20/30	20/30	Few CD	Homozygous H331P
16 (11)	F 41	36	20/40	20/25	CD, BSP, CA, retinal scarring	IVS6-8del17bp/insGC + P396L, †
17 (12)	F 40	25	20/200	HM	CD, BSP, CA	Homozygous IVS6-8del17bp/insGC
18 (13)	F 37	35	20/25	20/30	CD, BSP	IVS6-8del17bp/insGC + IVS8-2A>G

 CD, crystalline deposits; BSP, bone spicule pigmentation; CA, choriocapillaris atrophy; RVA, retinal vascular attenuation; HM, hand motion; FC, finger count.

*  In family 1, patient 1 is the mother of patients 2 and 3. The mutate allele with the IVS8-2A>G mutation in patients 2 and 3 came from their father, and therefore the mutations in the two generations are different.

†  Novel mutation.

Figure 1.

 

Five novel mutations in the CYP4V2 gene. (A) Heterozygous Y219H, (B) heterozygous W244X, (C) heterozygous D324V, (D) heterozygous P396L, and (E) compound heterozygous R400C and R400H.

Figure 2.

 

Multiple sequence alignment of CYP4V2 protein with its homologues and CYP4 subfamily members. Arrows: positions of the amino acid substitutions. (A) Y219H, (B) D324V and H331P, and (C) P396L and R400C/R400H. hs, Homo sapiens; dm, Drosophila melanogaster; mm, Mus musculus; pp, Pongo pygmaeus.

Figure 3.

 

Fundus photographs of selected patients with BCD with mutations in CYP4V2 showing various clinical appearances of BCD. (A) Right eye of patient 1; (B) right eye of patient 2 who is a daughter of patient 1; (C) left eye of patient 4; (D) left eye of patient 10; (E) right eye of patient 13; (F) right eye of patient 15, who is the youngest brother of patient 13; (G) left eye of patient 16; and (H) right eye of patient 18.

Figure 4.

 

Clinical features of patient 6 with homozygous IVS6-8del17bp/insGC mutation. (A) Fundus photograph of the right eye and (B) left eye showing the presence of multiple crystalline deposits, diffuse choriocapillaris atrophy, and retinal bone spicule pigmentations. (C) OCT scan of the right and (D) left macula demonstrating fine intraretinal lesions with increased signal intensity (red arrows) consistent with intraretinal crystals. (E) Full-field ERG of the eye showing that all responses were nonrecordable.

Figure 5.

 

Clinical features of patient 8 with heterozygous IVS6-8del17bp/insGC and the novel Pro396Leu mutation. (A) Fundus photograph of the right eye and (B) left eye showing diffuse choriocapillaris atrophy with a few retinal crystals. OCT scan of the (C) right and (D) left macula demonstrating small intraretinal hyperintense lesions (red arrows) consistent with intraretinal crystals. (E) Full-field ERG of the patient showing reduced amplitude with normal implicit times for scotopic, scotopic maximum response, and photopic ERGs. The 30-Hz flicker response was normal.

Figure 6.

 

Clinical features of patient 9 with heterozygous IVS8-2A>G and the novel Asp324Val mutation. (A) Fundus photograph of the right eye and (B) left eye showing numerous crystalline deposits in the retina with a few bone spicule pigmentations. (C) OCT scan of the right and (D) left macula demonstrating intravitreal hyperintense lesions (red arrows) confirming the presence of intraretinal crystals. (E) Full-field ERG showing reduced amplitude with normal implicit times for the scotopic ERG. The maximum response scotopic and photopic ERGs and the 30-Hz flicker ERG were within normal limits.

Table 2.

 

Electrophysiological Findings in 18 Chinese Patients with BCD

Patient (Family)	Age (y)	EOG Arden Index (RE/LE)	Scotopic ERG	Maximal Response ERG	Photopic ERG	30-Hz Flicker ERG	mfERG
1 (1)	70	1.4/1.4	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable
2 (1)	44	1.1/1.2	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: reduced amp.	BE: central reduction in response amp.
3 (1)	42	1.2/1.2	BE: nonrecordable	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: nonrecordable
4 (2)	59	1.8/1.8	BE: reduced amp./normal i.t.	BE: low normal amp./normal i.t.	BE: low normal amp./normal i.t.	BE: reduced amp.	BE: central reduction in response amp.
5 (3)	53	1.3/1.3	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: central reduction in response amp.
6 (3)	52	1.4/1.2	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: central reduction in response amp.
7 (4)	68	1.5/1.4	BE: reduced amp./normal i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./normal i.t.	BE: reduced amp.	BE: central reduction in response amp.
8 (5)	75	1.5/1.4	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: normal	BE: mild reduction in central response amp.
9 (6)	30	1.2/1.1	BE: reduced amp./normal i.t.	BE: normal amp./normal i.t.	BE: normal amp./normal i.t.	BE: normal	BE: central reduction in response amp.
10 (7)	63	1.4/1.4	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./normal i.t.	BE: reduced amp.	BE: central reduction in response amp.
11 (8)	70	1.4/1.7	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: reduced amp./normal i.t.	BE: reduced amp.	BE: central reduction in response amp.
12 (9)	31	1.0/1.1	BE: nonrecordable	BE: nonrecordable	BE: nonrecordable	BE: reduced amp.	BE: nonrecordable
13 (10)	47	1.5/1.5	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: nonrecordable
14 (10)	45	1.6/1.5	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: nonrecordable
15 (10)	42	1.7/1.7	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: mild reduction in central response amp.
16 (11)	41	1.3/1.2	BE: reduced amp./normal i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: central reduction in response amp.
17 (12)	40	1.4/1.4	BE: nonrecordable	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: nonrecordable
18 (13)	37	1.1/1.3	BE: nonrecordable	BE: reduced amp./delayed i.t.	BE: reduced amp./delayed i.t.	BE: reduced amp.	BE: nonrecordable

 EOG, electrooculogram; ERG, full-field electroretinogram; mfERG, multifocal electroretinogram; BE, both eyes; RE, right eye; LE, left eye; amp., amplitude; i.t., implicit times.

Table 3.

 

Summary of Full-Field Electroretinography Findings in 18 Chinese Patients with Bietti’s Crystalline Dystrophy

Type of ERG	Nonrecordable	Reduced Amplitude with Delayed Implicit Time	Reduced Amplitude with Normal Implicit Time	Normal Amplitude and Normal Implicit Time
Scotopic ERG	7 (38.9)	4 (22.2)	7 (38.9)	0 (0.0)
Maximal response ERG	4 (22.2)	9 (50.0)	3 (16.7)	2 (11.1)
Photopic ERG	4 (22.2)	7 (38.9)	5 (27.8)	2 (11.1)

 Data are the number of patients (percentage of total).

The authors thank all the patients and their family members participated in the study.