Molecular Genetics of Primary Congenital Glaucoma in Brazil

Abstract

purpose. To determine the distribution of CYP1B1 gene mutations in Brazilian patients with primary congenital glaucoma (PCG).

methods. PCG diagnosis was established by presence of buphthalmos in at least one affected eye and associated high intraocular pressures before the age of 3 years. CYP1B1 mutation screening of 52 patients with PCG was performed by SSCP and direct sequencing of PCR fragments.

results. Eleven mutations, four of which are novel, were observed in 26 (50%) individuals. A new frameshift mutation (4340delG) was observed in 20.2% of all individuals screened. These individuals had early-onset, bilateral glaucoma that necessitated multiple surgical interventions. CYP1B1 mutations were twice as frequent in affected individuals of European descent as in individuals of African descent. Analysis of six intragenic single nucleotide polymorphisms (SNPs) established 5′-C-C-G-G-T-A-3′ as the most common haplotype among the affected Brazilian individuals. A nonsense mutation (W57X) previously reported in an individual with Peters anomaly (compound heterozygote) was also observed in two individuals with PCG but combined with different mutations. A newly developed SSCP assay enabled us to detect all DNA mutations and polymorphisms previously detected by direct sequencing.

conclusions. Our results indicate that CYP1B1 mutations may be responsible for half of cases of PCG in the Brazilian population. The SNP haplotype 5′-C-C-G-G-T-A-3′ was associated with the majority of CYP1B1 mutations. This haplotype harbors the high-activity V432 allele, which is emerging as a putative susceptibility factor in several cancers.

The severe primary congenital form of glaucoma (PCG; Online Mendelian Inheritance in Man [OMIM] 231300) has its onset in the neonatal or infantile period and is characterized by increased intraocular pressure, corneal edema, enlargement of the globe (buphthalmos), epiphora, photophobia, and blepharospasm. ¹ The pathologic increase of the intraocular pressure (IOP) is attributed to a developmental abnormality of the anterior chamber angle. ² Historically, Westerlund ³ attributed the first report of familial occurrence of congenital glaucoma to Junken, who in 1842 described a Swedish family in which apparently normal parents gave birth to two normal and seven affected children. According to the same author, in 1836 Grelios first acknowledged a higher incidence of the disease within an ethnic group—specifically, the Jewish population in Algiers. ³ This condition is now recognized worldwide, with a variable incidence ranging from 1:1,250 to 1:22,000. ^{3 4 5 6} For familial cases, the disease most frequently is transmitted as an autosomal-recessive trait with variable penetrance. ^{3 4 5 6 7} Parental consanguinity is frequently reported. There is a high rate of concordance among monozygotic twins and discordance among dizygotic ones. ^{2 3 4 5 6} The universal validity of the autosomal-recessive model has been challenged by reports of disease transmission in successive generations, unequal sex distribution among the affected individuals, and a lower-than-expected number of affected siblings in the familial cases. These observations have raised the possibility that PCG is a genetically heterogeneous disorder. ^{2 5 6 8 9 10} The first molecular evidence for this possibility came from the genetic linkage studies in our laboratory by identifying two separate genetic loci associated with the disease, thus confirming that PCG is indeed genetically heterogeneous. Most of our PCG families were linked to the GLC3A locus on 2p21. ¹¹ However, few other families were linked to the GLC3B locus on 1p36. ¹² The existence of linkage between the GLC3A locus and PCG phenotype was subsequently confirmed in PCG families from Saudi Arabia ⁷ and Romany Slovakians. ¹³  

In 1997, we reported that the cytochrome P4501B1 gene (CYP1B1; OMIM 601771) located within the GLC3A locus is mutated in individuals with PCG. ¹⁴ CYP1B1 is a monooxygenase that is capable of metabolizing various endogenous and exogenous substrates including steroids ¹⁵ and retinoids. ^{16 17} A specific CYP1B1 metabolite is most probably required for normal eye development, and its deficiency (or toxic accumulation) may result in PCG. ^{18 19 20} This initial report was followed by studies on the role of CYP1B1 in the etiology of PCG in various ethnic groups including Arabs, Turks, Romany, Japanese, and Amish. ^{7 21 22 23 24 25 26} These studies revealed extensive allelic heterogeneity by identifying more than 25 separate CYP1B1 mutations segregating with the disease phenotype. In the large family panels from Turkey and Saudi Arabia and in Romany Slovakians, CYP1B1 mutations were observed in 90% to 100% of the studied families. Reduced penetrance was reported in the Saudi Arabian families with PCG, ⁷ which was attributed to the existence of a dominant modifier locus that is located on 8p. ²²  

Herein, we describe the results of CYP1B1 screening of 52 Brazilian individuals with PCG. This report provides for the first time new information on the role of CYP1B1 in the pathogenesis of PCG in Latin America. 

Methods

Subjects

The research followed the tenets of the Declaration of Helsinki. Individuals for this study were recruited among the subjects with PCG diagnosed in the hospitals at University of Campinas and University of Sao Paulo, Brazil. The Ethics Committee of both Universities approved the study. Fifty-two individuals with a diagnosis of PCG were analyzed. Each individual was examined by at least one of the authors. Buphthalmos was defined as horizontal corneal diameter higher than 12.5 mm. IOP measurements were obtained with Perkins or Goldmann tonometry in patients under general anesthesia. IOP higher than 19 mm Hg in the presence of buphthalmos was accepted as indicative of congenital glaucoma. Patients with biomicroscopic or gonioscopic changes inconsistent with the diagnosis of PCG were not included. Family history of PCG was established in 27 (51.9%) individuals. Twentyfour (46.2%) individuals had a negative family history; data were inconclusive for one additional individual. Parental consanguinity was reported in 14 (26.9%) individuals. Eight (15.4%) individuals had unilateral PCG, 21 (40.4%) identified themselves as being of African descent, and 30 (57.7%) indicated European descent (Portuguese/Italian). Data were not available for one individual. A panel of 50 DNA samples from unrelated Brazilian individuals with no reported family history of any hereditary eye disease was used as the normal control. 

Mutation Screening

Genomic DNA was prepared from peripheral blood. Three primer sets were used to amplify the coding regions of the CYP1B1 gene from genomic DNA: set 1 (1 forward [F], 5′[nt 3676]-tctccagagagtcagctccg-3′; and 1 reverse [R], 5′[nt 4461]-gggtcgtcgtggctgtag-3′ [786 bp]); set 2 (2F, 5′[nt 4199]-atggctttcggccactact-3′; and 2R, 5′[nt 4985]-gatcttggttttgaggggtg-3′ [787 bp]); set 3 (3F, 5′[nt 7740]-tcccagaaatattaatttagtcactg-3′; and 3R, 5′[nt 8624]-tatggagcacacctcacctg-3′ [885 bp]); PCR amplification was performed in a 50-μL volume consisting of 50 mM KCl, 10 mM Tris-HCl (pH 9.3), 1.5 mM MgCl₂, 0.2 mM of each dNTP, 1 U Taq DNA polymerase (Platinum Taq; GibcoBRL, Grand Island, NY), 10% dimethyl sulfoxide (DMSO; sets 1 and 2 only), 1.0 μL of a 20-μM stock solution of each primer oligonucleotide and 100 ng genomic DNA template. Thermocycling was performed (Gene Amp PCR Thermocycler; 9700; PE Applied Biosystems, Foster City, CA) under the following conditions: initial denaturation at 94°C for 2 minutes followed by 35 cycles each consisting of 30 seconds’ denaturation at 94°C, 30 seconds’ annealing at 55°C, and 1 minute’s extension at 72°C, followed by a final extension of 7 minutes at 72°C. Column purification of the PCR fragments was performed with a kit (QIAquick; Qiagen, Valencia, CA) according to the manufacturer’s protocol. Dye terminator sequencing (BigDye Terminator kit; PE Applied Biosystems) was performed according to the manufacturer’s protocol. Analysis of the sequencing reactions was performed on an automated DNA sequencer (ABI-377; PE Applied Biosystems). Comparative sequencing analysis between different samples was performed with the assistance of computer software (Sequencher, ver. 3.11; Gene Codes Corp., Ann Arbor, MI). 

Single-Strand Conformation Polymorphism Analysis

Thirteen primer pairs (Table 1) covering the coding regions of the CYP1B1 gene were designed with the assistance of the PRIMER3 program (provided by the Massachusetts Institute of Technology, Cambridge, MA, and available at: http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). PCR amplification was performed in a 20-μL volume consisting of: 50 mM KCl, 10 mM Tris-HCl (pH 9.3), 1 to 3 mM MgCl₂, 0.1 mM from each dNTP, 0.1 U Taq DNA polymerase (AmpliTaq; Perkin Elmer, Norwalk, CT), 50 ng genomic DNA and 0.5 μL primer mix (20 μM stock) under the following thermocycling conditions: initial denaturation at 94°C for 2 minutes followed by 35 cycles each consisting of 30 seconds’ denaturation at 94°C, 30 seconds’ annealing at 55°C and 30 seconds’ extension at 72°C, followed by a final extension of 7 minutes at 72°C. The PCR samples were mixed 1:1 with formamide loading dye (0.05% bromphenol blue, 0.05% xylene cyanol, 2 mM EDTA in deionized formamide), incubated at 95°C for 5 minutes and snap frozen on ice. Approximately 8 μL from each sample were subjected to acrylamide gel electrophoresis (10% acrylamide/bis [37.5:1]) sequencing gel and 0.5× TBE). Electrophoresis was performed at room temperature on an air-cooled apparatus (model S2001; GibcoBRL) with 0.5× TBE running buffer at 7 W for 12 hours. Silver staining was performed as described previously. ¹¹  

Comparative Sequencing Alignment

The amino acid sequences of cytochrome P450 proteins were obtained from SwissProt and GenBank (Swiss Prot is provided by the Swiss Institute of Bioinformatics, Geneva, Switzerland, and is available at no fee to academic users at http://www.expasy.org; GenBank is provided by the National Center for Biotechnology Information, Bethesda, MD, and is available at http://www.ncbi.nlm.nih.gov/Genbank). Computer-assisted sequence alignment (Fig. 1) was performed with the pattern-induced multisequence alignment program (PIMA; provided by Baylor College of Medicine Search Launcher and available at: http://searchlauncher.bcm.tmc.edu). ²⁷ The missense mutations were mapped against a three-dimensional model of the conserved COOH terminus of CYP1B1 described earlier. ²¹  

Results

Sequence analysis of the coding regions of the CYP1B1 gene in 52 Brazilian PCG subjects identified 17 different alterations in the DNA sequence. Six of these alterations were observed in both normal and affected individuals: g.3793T→C, g.3947C→G (R48G), g.4160G→T (A119S), g.8131C→G (L432V), g.8184C→T (D449D), and g.8195A→G (N453S). This was in agreement with our earlier reports, ^{14 21} which recognized these changes as single-nucleotide polymorphisms (SNPs). The remaining 11 DNA alterations (Table 2) were not observed in the panel of 50 control samples and therefore could not be classified as polymorphisms. Seven of these changes were identified in earlier reports as disease-causing mutations. The remaining four DNA alterations (g.3860C→T, g.8165C→G, g.4340delG, g.8214_8215delAG) are reported herein for the first time. CYP1B1 mutations were observed in 26 (50%) individuals. An overall percentage of patients with CYP1B1 mutations in our sample is presented in Table 3 . A brief description of these mutations is presented in the following sections. Information on the clinical presentation and SNP haplotypes associated with each of these mutations is summarized in Table 4 . 

g.3860C→T (Q19X)

A thymine-to-cytosine transition affecting the first nucleotide of codon 19 was observed on one PCG chromosome. The observed haplotype for the six intragenic SNPs on this chromosome consists of: 5′-C-[Q19X]-C-G-C-C-A-3′. The termination codon of TAG created by this novel nonsense mutation is predicted to result in a premature translation termination, truncating 524 amino acids. 

g.4340delG

A new mutation, represented by a single-nucleotide (guanosine) deletion in exon 2 was documented for 21 (20.2%) PCG chromosomes, thus making it the most frequent mutation observed in our sample. Because both positions 4339 and 4340 are occupied by guanosine, it is not possible to determine exactly which of the positions is deleted. The frameshift introduced by this deletion creates a premature stop codon 51 base pairs downstream of the new reading frame. Because this mutation occurs upstream from the sequences encoding the conserved COOH terminus of CYP1B1, the resultant protein would not possess the conserved core structures (e.g., meander, hem-binding) required for the proper function of cytochrome P450. The PCG chromosomes carrying this mutation appear to share an identical haplotype for the six intragenic SNP markers flanking the mutated position: 5′-C-C-G-[4340delG]-G-T-A-3′. 

g.8165C→G (A443G)

A guanosine-to-cytosine transversion affecting the second nucleotide of codon 443 was observed on one PCG chromosome. This missense mutation substitutes glycine (aliphatic, polar-uncharged) for alanine (aliphatic, nonpolar, hydrophobic) within the conserved meander region of CYP1B1. 

g.8214_8215delAG

A 2-bp deletion (AG) affecting the third nucleotide of codon 459 and the first nucleotide of codon 460 was observed on one PCG chromosome and represents a new mutation. The frameshift produced by the deletion of two nucleotides alters the protein sequence and introduces a stop codon (TGA) 58 nucleotides into the new reading frame, thus eliminating the essential hem-binding region that is located downstream from this mutation. 

CYP1B1 SNP Haplotypes

The SNP haplotypes associated with different CYP1B1 mutations are presented in Table 4 . The most frequent haplotype among affected individuals with CYP1B1 mutations appears to be 5′-C-C-G-G-T-A-3′. At least seven different mutations are associated with this haplotype. Complete haplotypes were also obtained in 47 of 50 normal individuals in the control group (data not shown). Of these, seven were homozygotes for the 5′-C-C-G-G-T-A-3′ haplotype. Assuming that the control Brazilian population is in Hardy-Weinberg equilibrium, the inferred haplotype frequency would be 0.39. Table 5 compares the percentage of homozygous and heterozygous individuals (Table 5A) and the allelic frequencies (Table 5B) for five SNPs affecting the amino acid sequence of CYP1B1 among three groups of individuals: Brazilians with PCG, Brazilian control subjects, and control subjects from populations in the United Kingdom and Turkey reported earlier. ²¹  

SSCP Analysis of CYP1B1

SSCP analysis was tested to provide a rapid and cost-effective alternative to direct DNA sequencing (Fig. 1) . Using uniform conditions for PCR amplification, gel electrophoresis, and detection method, we were able to detect all DNA alterations previously observed during our direct-sequencing analysis. 

Discussion

In 1997, we first reported CYP1B1 mutations in Turkish individuals with congenital glaucoma. ¹⁴ This initial report was followed by a number of studies investigating the role of CYP1B1 in the etiology of PCG in various ethnic groups. The first large panels of families to be studied (Turk, Arab, and Romany individuals) all historically attracted the attention of investigators, because of the high incidence of PCG, autosomal-recessive mode of inheritance, and relatively uniform clinical presentation. However, heterogeneity has been the hallmark of PCG at all levels of investigation, and therefore extending the CYP1B1 screening to other populations is essential for a better evaluation of its role in the pathogenesis of PCG worldwide. In this study, we sought to achieve this goal by reporting the results of CYP1B1 mutation screening in a group of 52 Brazilian individuals with PCG. 

Sequence analysis of the coding regions of the CYP1B1 gene identified 11 DNA mutations in 26 (50%) of 52 Brazilian probands. This is lower than the 90% to 100% reported in Turk, Arab, and Romany individuals ^{7 21 22 23} and higher than the 20% reported in Japanese individuals. ²⁸ However, direct comparison between these studies is difficult, because of differences in sample size, the composition of the samples (familial/sporadic, unilateral/bilateral) and the screening method used. In this Brazilian sample, CYP1B1 mutations were observed in 55.6% of the familial cases versus 41.7% of the sporadic ones. Mutations were also found in 55.8% of the individuals with bilateral disease compared with 12.5% of the individuals with unilateral disease. CYP1B1 mutations were much more frequent among the affected individuals of European descent (60%) than in those of African descent (33.3%). Of the 26 individuals with CYP1B1 mutations, 15 were homozygotes, 7 were compound heterozygotes, and 4 were heterozygotes (only one mutant allele was detected). Patients with PCG that had a heterozygous CYP1B1 mutation were also reported by Mashima et al. ²⁹ Because both screenings were limited to the coding regions of CYP1B1, it is reasonable to expect that these individuals may have mutations affecting the promoter region of the gene. ³⁰  

A single nucleotide deletion (4340delG), was observed on 21 (20.2%) PCG chromosomes in this sample of Brazilian individuals. Clinical evaluation of 12 individuals carrying this mutation (10 homozygotes and 2 compound heterozygotes) revealed a severe PCG phenotype characterized by early onset, aggressive course, and poor response to surgical treatment (Table 4) . In all but one individual, the disease onset was in the first month of life. Both eyes were affected in all cases. Increased intraocular pressure was observed in all but one individual (maximum IOP, 25–55 mm Hg). Each individual had undergone multiple (up to six) surgical interventions. The phenotype associated with this mutation appears to be more severe than the one in individuals without this mutation. The average number of surgeries, for example, was 2.6 in the left eye and 2.8 in the right eye (1.5 and 1.4, respectively, in individuals without this mutation). The uniformly severe phenotype associated with the 4340delG mutation raised the question of developing a rapid and cost-effective screening assay. In the current study, we performed a comprehensive SSCP assay covering the coding regions of the CYP1B1 gene. A total of 13 PCR fragments no longer than 250 nucleotides with 50-bp overlap were designed. SSCP analysis and silver staining enabled us to detect all DNA mutations and polymorphisms previously detected by DNA sequencing. Hence, this assay could be used for either complete or focused screening of the CYP1B1 gene. 

A recent report of CYP1B1 mutations in an individual with Peters anomaly (OMIM 604229) raised the question of a possible phenotypic heterogeneity of CYP1B1. ³¹ Peters anomaly is an anterior segment malformation consisting of a central corneal leukoma, absence of the posterior corneal stroma and Descemet membrane, and a variable degree of iris and lenticular attachments to the central aspect of the posterior cornea. ³² The subject (a male of Native Indian-French Canadian background) was found to be a compound heterozygote for two CYP1B1 mutations: M1T and W57X. The investigators hypothesized that the severity of the Peters phenotype correlates with the extent of predicted protein truncation. We detected the W57X mutation in two Brazilian individuals (patients 11 and 14, Table 4 ). Both individuals were females of European descent without a family history of PCG. Both of them were compound heterozygotes. In these two cases, however, W57X was combined with different mutations: 7901_7913delGAGTGCAGGCAGA and R368H. We have also observed a case of PCG from Israel, which was homozygous for the M1T mutation (Stoilov I, Sarfarazi M, unpublished data, 1998). Our observation suggests that presence of a particular mutation may not be sufficient by itself to determine the exact phenotypic outcome (i.e., PCG or Peters anomaly). 

Is it possible to attribute the phenotypic variability to the different combinations of CYP1B1 mutations observed in these three individuals? For such a possibility to be considered, it is essential to demonstrate that the mutant forms of CYP1B1 differ in their molecular and enzymatic properties. In fact, this has already been demonstrated by the functional analysis of two CYP1B1 mutations, G61E and R469W, conducted in our laboratory. ³³ Although both mutations impaired the normal function of CYP1B1 hemoprotein, the mechanisms of their action appeared to be quite different. The stability of the hemoprotein complex was reduced by G61E but increased by R469W. As far as the enzymatic activity of these mutants are concerned, major interference in A-ring 4-hydroxylation of 17β-estradiol was observed with G61E with only minor reduction of metabolism at other positions. We were surprised to note that despite its having a greater structural stability, the ability of the R469W allelic variant to metabolize steroids was greatly impaired at all positions. This data indicates that the CYP1B1 mutations associated with PCG affect the structure and function of the cytochrome P450 molecule in different ways. Therefore, individuals who are compound heterozygotes for such mutations may exhibit complex biochemical phenotypes, which could account for the phenotypic heterogeneity of CYP1B1. The deciphering of the phenotypic variability of CYP1B1 would then require meticulous comparative analysis of the genotype, biochemical phenotype, and clinical presentation of the relevant CYP1B1 mutations. 

In our earlier reports, we described six intragenic CYP1B1 SNPs. ^{14 21} The SNP haplotypes associated with the DNA mutations observed in the Brazilian population are presented in Table 4 . Table 5 provides a comparison between Brazilians with PCG, Brazilian control subjects, and control subjects from the United Kingdom and Turkey published previously. ²¹ The most common haplotype among the affected Brazilian individuals was 5′-C-C-G-G-T-A-3′. This haplotype was associated with at least seven separate mutations and appears to be shared by the individuals carrying the most frequent mutation, 4340delG. This particular haplotype was also observed on 94.7% of the Saudi Arabian PCG chromosomes. ²² In addition, its frequency in the general population appears to be similar between Saudi Arabians (0.35) ²² and Brazilians (0.39). These similarities are quite striking, because the two populations under discussion are geographically separated and are ethnically and culturally distinct from each other. On the protein level, the 5′-C-C-G-G-T-A-3′ haplotype would translate into: NH2-R48-A119-V432-D449-N453-COOH. 

It has been demonstrated that human CYP1B1 can catalyze the hydroxylation of 17β-estradiol at positions C-4 and C-2. ¹⁵ Allelic variants containing valine at position 432 are characterized by a higher ratio of 4-hydroxyestradiol to 2-hydroxyestradiol production. ³⁴ Because 4-hydroxyestradiol is a known carcinogen, ³⁵ the possible association of the V432 allele with various cancers is currently under investigation by different research groups. So far, the V432 allele has been found to be a putative susceptibility factor in ovarian and prostate cancers and for the smoking-related head and neck squamous cell cancer. ^{36 37 38} According to our data, 42% of the Brazilians with PCG (Table 5A) were homozygous for the V432 polymorphism compared with 23% and 11% in the control Brazilian and English/Turkish groups, respectively. When the allelic frequencies are compared (Table 5B) L432V is the only polymorphism for which there is a reversal in the allele frequency between PCG and control groups, with V432 having the highest frequency in the PCG group. 

Does this mean that, later in life, individuals with PCG may also have a higher risk of development of cancer for which V432 was identified as a putative risk factor? In contrast to the modulating effects of polymorphic alleles such as V432, the mutations identified in individuals with PCG completely eliminate or severely diminish the enzymatic activities of CYP1B1 because of alterations in the conserved core structures and/or truncations of the protein. ^{21 33} Therefore, CYP1B1 mutations causing PCG effectively eliminate the high-activity V432 allele with which they appear to be associated in the Brazilian and Saudi Arabian populations. This may provide a biological advantage to heterozygous carriers of CYP1B1 mutations by reducing their susceptibility to various cancers. This possibility lies beyond the scope of this study but deserves investigation. 

 Supported by Grants EY-11095 from The National Eye Institute, Grant MO1RR-06192 from The Mary Jane & Peter DaPuzzo Research Fund of the Glaucoma Foundation and UCHC-GCRC (MS), and Fundação de Amparro a Pesquisa do Estado de São Paulo (VPC).

 Submitted for publication September 5, 2001; revised December 31, 2001; accepted January 16, 2002.

 Commercial relationships policy: N.

 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.

 

SSCP Primer Pairs

ID	Position*	Tm (°C)	Sequence (5′→3′)	Size (bp)
CYP 1F	3760	60.32	GTC-TCT-GCA-CCC-CTG-AGT-GT	225
CYP 2R	3984	60.70	ATC-AGT-GGC-CAC-GCA-AAC
CYP 3F	3864	59.63	AGA-CCA-CGC-TCC-TGC-TAC-TC	240
CYP 4R	4103	59.69	CGC-CAT-TCA-GCA-CCA-CTA-T
CYP 5F	4045	61.74	CTA-CGG-CGA-CGT-TTT-CCA-G	238
CYP 6R	4282	60.14	CTG-GCG-CGT-GAA-GAA-GTT
CYP 7F	4211	59.03	CAC-TAC-TCG-GAG-CAC-TGG-AA	226
CYP 8R	4436	60.74	CGA-AAC-ACA-CGG-CAC-TCA-T
CYP 9F	4295	61.82	CAA-GTC-CTC-GAG-GGC-CAC	259
CYP 10R	4553	60.84	AGT-ACT-GCA-GCC-AGG-GCA-T
CYP 11F	4477	60.60	GCT-CAG-CCA-CAA-CGA-AGA-GT	232
CYP 12R	4708	59.14	TTC-CGC-AGA-GAG-GAT-AAA-GG
CYP 13F	4598	59.92	AAC-CGC-AAC-TTC-AGC-AAC-TT	253
CYP 14R	4850	60.79	ACC-TGG-TGA-AGA-GGA-GGA-GC
CYP 15F	4760	59.98	AAC-GTA-CCG-GCC-ACT-ATC-AC	247
CYP 16R	5006	61.08	TGT-GAG-TCC-CTT-TAC-CGA-CG
CYP 17F	7838	58.91	TTT-GCT-CAC-TTG-CTT-TTC-TCT-C	238
CYP 18R	8075	59.72	TGG-TAG-CCC-AAG-ACA-GAG-GT
CYP 19F	7977	60.07	TGT-CCT-GGC-CTT-CTT-TTA-TG	245
CYP 20R	8221	59.98	TCA-TCA-CTC-TGC-TGG-TCA-GG
CYP 21F	8125	60.11	GAC-CCA-CTG-AAG-TGG-CCT-AA	225
CYP 22R	8349	59.67	ATT-CAT-TTT-CGC-AGG-CTC-AT
CYP 23F	8283	59.76	CTT-CAT-CTC-CAT-CCT-GGC-TC	205
CYP 24R	8487	59.90	CAG-CTT-GCC-TCT-TGC-TTC-TT
CYP 25F	8414	60.54	TGG-AGC-TCC-TTG-ATA-GTG-CTG	250
CYP 26F	8663	60.00	AGC-TCC-TGC-ATA-GCC-CAC-TA

*  GenBank U56438.

 Tm, melting temperature.

Figure 1.

 

Novel CYP1B1 mutations observed in Brazilian individuals with PCG. (A) Genomic structure of the CYP1B1 gene. (B) Electrophoretograms representing four novel CYP1B1 mutations observed in Brazilian individuals. (C) Sequence alignment of human (Q16678), mouse (Q64429), and rat (Q64678) CYP1B1 proteins. (D) SSCP detection assay.

Table 2.

 

CYP1B1 Mutations in 52 Brazilians with PCG

DNA	Protein	Location	Number of PCG Chromosomes* (%)	Origin
g.3860C→T	Q19X	Exon 2	1 (1)	Brazil
g.3976G→A	W57X	Exon 2, hinge region	2 (1.9)	Brazil, Canada 31
g.7940G→A	R368H	Exon 3, J helix	2 (1.9)	Brazil, Saudi Arabia 7 22
g.7996G→A	E387K	Exon 3, K helix	3 (2.9)	Brazil, US, Canada, Romany 21 23 25
g.8147C→T	P437L	Exon 3, Meander region	3 (2.9)	Brazil, Turkey 21
g.8165C→G	A443G	Exon 3, Meander region	1 (1)	Brazil
g.4340delG		Exon 2	21 (20.2)	Brazil
g.7901_7913delGAGTGCAGGCAGA		Exon 3	3 (2.9)	Brazil, Turkey 14
g.8037_8046dupTCATGCCACC		Exon 3	6 (5.8)	Brazil, US, UK, Turkey 21
g.8182delG		Exon 3	5 (4.8)	Brazil, US 21
g.8214_8215delAG		Exon 3	1 (1)	Brazil

 These sequence variations are described in relation to reference sequence of human CYP1B1 gene with Genbank accession number U56438

*   CYP1B1 mutations were detected in 26 individuals (52 PCG chromosomes). A total of 48 PCG chromosomes harboring CYP1B1 mutations are recorded in the table, because in four individuals only one mutation was detected in the coding regions of this gene. n = 104.

Table 3.

 

CYP1B1 Mutations Observed in Different Subgroups of Patients

Category	Total Individuals (%)	Number with Mutations (%)
Total Individuals	52 (100)	26 (50)
Familial cases	27 (51.9)	15 (55.6)
Sporadic cases	24 (46.2)	10 (41.7)
Unknown origin	1 (1.9)	1 (1.9)
Parental consanguinity	14 (26.9)	8 (57.1)
Unilateral disease	8 (15.4)	1 (12.5)
Bilateral disease	43 (82.7)	24 (55.8)
Unknown type	1 (1.92)	1 (1.9)
European descent	30 (57.7)	18 (60)
African descent	21 (40.4)	7 (33.3)
Unknown descent	1 (1.9)	1 (1.9)

Table 4.

 

Phenotype–Genotype Correlation in Brazilian Individuals with CYP1B1 mutations

ID	Age at Diagnosis (mo)	U/B	IOP Max (mmHg)	Age at First Surgery (mo)	Visual Acuity OD (n Surgeries)	Visual Acuity OS (n Surgeries)	Genotype	SNP 1	SNP 2	SNP 3	SNP 4	SNP 5	SNP 6
2	—	B	—	—	— (—)	— (—)	[4340delG]+[4340delG]	C	C	G	G	T	A
3	<1	B	—	—	LP (—)	LP (—)	[4340delG]+[4340delG]	C	C	G	G	T	A
4	<1	B	48	—	LP (0)	CF (2)	[4340delG]+[4340delG]	C	C	G	G	T	A
29	<1	B	55	—	CF (4)	LP (6)	[4340delG]+[4340delG]	C	C	G	G	T	A
38	<1	B	42	11	LP (4)	HM (4)	[4340delG]+[4340delG]	—	C	G	G	T	A
46	<1	B	28	2	— (3)	— (3)	[4340delG]+[4340delG]	—	C	G	G	T	A
49	<1	B	36	2	— (2)	— (3)	[4340delG]+[4340delG]	—	C	G	G	T	A
54	<1	B	45	1	0.1–0.3 (2)	LP (2)	[4340delG]+[4340delG]	—	C	G	G	T	A
37	<1	B	14	—	LP (4)	CF (2)	[4340delG]+[4340delG]	—	C	G	G	T	A
7	<1	B	25	2	— (3)	— (2)	[4340delG]+[7901_7913delGAGTGCAGGCAGA]	—	—	G	G	T	A
40	1	B	—	2	LP (1)	>0.5 (2)	[4340delG]+[7901_7913delGAGTGCAGGCAGA]	C	C	G	G	T	A
26	1	B	35	2	LP (2)	CF (2)	[4340delG]+?	C	C	G	G	T	A
27	<1	B	26	1	CF (2)	LP (2)	[8037_8046dupTCATGCCACC]+[8037_8046dupTCATGCCACC]	C	C	G	G	T	A
39	<1	B	30	120	0.1–0.3 (3)	LP (2)	[8037_8046dupTCATGCCACC]+[8037_8046dupTCATGCCACC]	—	C	G	G	T	A
25	<1	B	30	1	0.1–0.3 (1)	>0.5 (1)	[8037_8046dupTCATGCCACC]+[R368H]	C	C	G	—	—	—
21	<1	B	34	—	LP (1)	LP (0)	[8037_8046dupTCATGCCACC]+[8182delG]	C	C	G	—	T/C	A/G
22	<1	B	25	2	LP (1)	0.1–0.3 (1)	[8182delG]+[8182delG]	C	C	G	C	C	G
52	2	B	30	9	LP (1)	HM (1)	[8182delG]+[8182delG]	—	C	G	C	C	G
31	30	B	50	204	LP (2)	LP (2)	[E387K]+[E387K]	T	G	T	C	C	A
32	<1	B	32	144	>0.5 (2)	LP (4)	[E387K]+[P437L]	—	C/G	G/T	G/C	T/C	A/G
10	—	B	—	—	—	—	[P437L]+[P437L]	C	C	G	G	T	A
42	—	B	26	—	LP (1)	>0.5 (1)	[8214_8215delAG]+?	C/T	C/G	G/T	G/C	T/C	A
11	12	B	34	—	0.1–0.3 (2)	LP (0)	[W57X]+[7901_7913delGAGTGCAGGCAGA]	C	C	G	G	T	A
14	—	B	—	—	—	—	[W57X]+[R368H]	C	C	G	G	T	A
43	<1	U	30	21	0.4–0.5 (2)	LP (2)	[A443G]+?	C	C/G	G/T	G	T	A
55	—	—	—	—	—	—	[Q19X]+?	C	C	G	C	C	A

 SNP 1–6 are: 1. 3793T→C; 2. 3947C→G (R48G); 3. 4160G→T (A119S); 4. 8131C→G (L432V); 5. 8184C→T (D449D); 6. 8195A→G (N435S). U/B, unilateral/bilateral; LP, light perception; CF, count fingers.

Table 5.

 

Distribution of the Polymorphic CYP1B1 Alleles

A. Genotypes
SNPs	g.3947C→G (R48G)					g.4160G→T (A119S)					g.8131C→G (L432V)					g.8184C→T (D449D)					g.8195A→G (N453S)
		C/C	C/G	G/G	G/G	G/T	T/T	C/C	C/G	G/G	C/C	C/T	T/T	A/A	A/G	G/G
		R	R/G	G	A	A/S	S	L	L/V	V	D	D/D	D	N	N/S	S
% Controls UK/TR 21 (n = 100)	51%	40%	9%	51%	40%	9%	54%	35%	11%	51%	39%	10%	59%	34%	7%
% Controls Brazil (n = 47)*	47%	47%	6%	51%	43%	6%	34%	43%	23%	30%	45%	25%	60%	36%	4%
% PCG Brazil (n = 45)*	64%	31%	5%	67%	29%	4%	20%	38%	42%	20%	38%	42%	78%	18%	4%

B. Allelic Frequencies
SNPs	R48G			A119S			L432V			N453S
		R	G	A	S	L	V	N	S
Controls UK/TR 21 (n = 100)	0.710	0.290	0.710	0.290	0.715	0.285	0.760	0.240
Controls Brazil (n = 47)*	0.705	0.295	0.725	0.275	0.555	0.445	0.780	0.220
PCG Brazil (n = 45)	0.795	0.205	0.815	0.185	0.390	0.610	0.870	0.130

 Genotypes are shown on the first row beneath the mutations in A. Amino acids are shown on the second row. In B, amino acids are shown below the alleles.

*  Only individuals for whom a complete haplotype could be constructed are included. TR, Turkey.

The authors thank the patients and their families who participated in this study.