A Nonsense Mutation in the Glucosaminyl (N-acetyl) Transferase 2 Gene (GCNT2): Association with Autosomal Recessive Congenital Cataracts

Abstract

purpose. To identify the genetic defect associated with autosomal recessive congenital cataract in four Arab families from Israel.

methods. Genotyping was performed using microsatellite markers spaced at approximately 10 cM intervals. Two-point lod scores were calculated using MLINK of the LINKAGE program package. Mutation analysis of the glucosaminyl (N-acetyl) transferase 2 gene (GCNT2) gene was performed by direct sequencing of PCR-amplified exons.

results. The cataract locus was mapped to a 13.0-cM interval between D6S470 and D6S289 on Chr. 6p24. A maximum two-point lod score of 8.75 at θ = 0.019 was obtained with marker D6S470. Sequencing exons of the GCNT2 gene, mutations of which have been associated with cataracts and the i blood group phenotype, revealed in these families a homozygous G→A substitution in base 58 of exon-2, resulting in the formation of premature stop codons W328X, W326X, and W328X, of the GCNT2A, -B, and -C isoforms, respectively. Subsequent blood typing of affected family members confirmed the possession of the rare adult i blood group phenotype.

conclusions. A nonsense mutation in the GCNT2 gene isoforms is associated with autosomal recessive congenital cataract in four distantly related Arab families from Israel. These findings provide further insight into the dual role of the I-branching GCNT2 gene in the lens and in reticulocytes.

Congenital cataracts are common disorders of the eye that often cause visual impairment or blindness in children. ^{1 2} At least one third of these cases are familial. ³ Identifying genes that cause inherited forms of cataract improves our understanding of mechanisms underlying cataract formation and provides further insight into normal lens development. The hypothesis that a significant portion of age-related cataracts could be influenced by mild mutations in genes that have already been implicated in families with congenital cataract ⁴ provides a tremendous incentive for studying these families. 

Although cataracts are most commonly inherited in an autosomal dominant manner (MIM 116600, 600897, 123680, 603212, 605749, 602669, 123590, 154050, 601885, 115650, 605728, 116800, 123740, 601202, 600881, 115660, 134790, 605387, 123580; and 601547; On-line Mendelian Inheritance in Man; http://www.ncbi.nlm.nih.gov/Omim/ provided in the public domain by the National Center for Biotechnology Information [NCBI], Bethesda, MD), recessive cataracts attract special interest because of a presumed influence of an autosomal recessive gene on nuclear sclerosing senile cataract pathogenesis. ^{5 6} To date, five loci for autosomal recessive cataracts have been described. These include two identified genes, CRYAA and LIM2, ^{7 8} and loci with as yet unknown genes (Gal A, et al. IOVS 2000;41:ARVO Abstract 1). ^{9 10} In addition, recent studies on the glucosaminyl (N-acetyl) transferase 2 (GCNT2) gene have begun to resolve the association between autosomal recessive congenital cataract and the rare adult i blood group phenotype. ^{11 12}  

I/i antigens are carbohydrate structures on glycoproteins and glycolipids on the cell surface of a variety of tissues and body fluids. ^{13 14} The i antigen epitope is a linear poly-N-acetyllactosamine chain that has Gal β1 to 4GlcNAc β1-3 unit repeats, and the I antigen structure is branched by the addition of an N-acetylglucosaminyl (GlcNAc) residue through β-1,6 linkage to a galactosyl residue. Conversion of the i antigen into an I structure first takes place in human red blood cells during the first 18 months after birth as a result of the expression of a specific transferase, I-branching GCNT2. Lack of this enzyme results in the adult i phenotype, a rare autosomal recessive condition, with only a few occurrences in thousands or tens of thousands. ¹⁵ The adult i phenotype was found to be highly associated with congenital cataract in Japanese ¹⁶ and Taiwanese ¹⁷ populations, but association with cataracts was not found to be as pronounced in the European population. ^{18 19} Tight linkage between the Ii blood group gene and a different cataract-related gene has been proposed, ^{19 20} but recent molecular genetic studies suggest that GCNT2 mutations in families from Japan and Taiwan are directly related to cataracts. Three GCNT2 splicing variants GCNT2A, -B, and -C, which differ at exon 1 but have identical exon 2 and 3 coding regions, are expressed differentially in specific tissues. Mutation events that occur in the specific exon 1 region of the GCNT2 gene may lead to a defect in one form of the GCNT2 enzyme and i phenotype in certain cell types, whereas those that occur in the common exon 2 to 3 region result in i phenotype as well as congenital cataract, because of the elimination of activity of all three forms of the GCNT2 enzymes. ^{11 12}  

In this study, we mapped the locus for autosomal recessive congenital cataract in four Arab families from Israel to a region of 6p24 spanning 13.0 cM including the I blood type locus. Sequencing exons of the GCNT2 gene showed a novel nonsense mutation in the common exon 2 of the gene. Blood typing of affected family members confirmed the possession of the adult i phenotype. 

Methods

Families and DNA Specimens

Four Arab families were recruited at the Sapir Medical Center, Kfar Saba, and the Ha-Emek Medical Center, Afula, Israel (Fig. 1) . The families are highly consanguineous, including five first cousin marriages, and 13 affected offspring. Families 56005, 56007, and 56009 live in an Arab village in central Israel, and families 56003a and -b that were described previously ¹⁰ live in another Arab village 40 miles away. During the past century, ancestors of the families migrated between these two villages. A detailed family history was obtained from older family members to establish a connection between the different families. A single common ancestor for all four families has not been identified; however, a connection between families 56003a and -b has been established, and it is suspected that a single founder mutation is responsible for the disease in all these pedigrees. 

Participants gave informed consent to the study protocol, which was approved by the Sheba Medical center and National Eye Institute institutional review boards, and conformed to the tenets of the Declaration of Helsinki. Bilateral leukocoria (white pupil) was noticed usually by parents within the first month after birth. The opaque lens interfered with ophthalmoscope retinal examination which was accomplished only after lens extraction. Surgery was performed during the first year. No photography or pathologic findings of the lens were available. Ophthalmic examinations of the parents and unaffected siblings did not reveal any ocular abnormalities in any family. Heparinized blood was obtained from each participant for genomic DNA isolation and blood typing. 

Genotyping

A genome scan was performed with samples from family 56007. For markers showing lod scores greater than +1.0, we performed fine mapping, using all four identified families. The genome scan was performed using the microsatellite markers in a commercial mapping system (Prism Linkage Mapping Set MD-10; Applied Biosystems, Inc. [ABI], Foster City, CA). Multiplex polymerase chain reaction (PCR) was performed as described. ²¹ Briefly, each reaction was performed in a 5-μL mixture containing 40 ng genomic DNA, various combinations of 10 μM fluorescent-dye-labeled primer pairs, 0.5 μL 10× PCR buffer (Buffer II; Gene Amp; ABI), 250 dNTP mix (Gene Amp; ABI), 2.5 mM MgCl₂, and 0.2 U Taq DNA polymerase (AmpliTaq Gold Enzyme; ABI). Amplification was performed in a thermocycler workstation (Prism 9700; ABI). Initial denaturation was performed for 12 minutes at 95°C, followed by 10 cycles of 15 seconds at 94°C, 15 seconds at 55°C, and 30 seconds at 72°C, and then 20 cycles of 15 seconds at 89°C, 15 seconds at 55°C, and 30 seconds at 72°C, finishing with a 20-minute extension cycle at 72°C and a final hold at 4°C. PCR products from each DNA sample were pooled and mixed with a loading cocktail (ABI) and loading dye and separated on a 5% denaturing polyacrylamide gel in a sequencer (model 377; ABI). The alleles were analyzed on computer (Genscan 3.1 and Genotyper 2.1 software; ABI). 

Linkage Analysis

Two-point linkage analysis was performed by using the FASTLINK version ²² of MLINK from the LINKAGE program package, ²³ and maximum lod scores were calculated by using ILINK, assuming an autosomal recessive model of inheritance and 100% penetrance in both sexes. Gene frequency of 0.004 was chosen, considering an estimate for disease prevalence in the population. ²⁴ The marker order and distances in Figure 1 and Table 1 were obtained from the Gènèthon database (http://www.genethon.fr/ provided in the public domain by the French Association against Myopathies, Evry, France) and the NCBI chromosome 6 sequence map (http://www.ncbi.nlm.nih.gov/mapview/). Equal allele frequencies were initially assumed for the genome scan. Allele frequencies for markers used in final linkage analysis (Table 1) were estimated from an analysis of more than 50 unrelated and unaffected individuals of Israeli Arab ethnicity. Haplotypes were constructed so as to minimize recombinants. 

PCR Amplification of the GCNT2 Gene

Genomic DNA was amplified by PCR using specific primer pairs (designed using the Primer Select program; DNASTAR Inc., Madison, WI) for each of the specific first exons and the common second and third exons for GCNT2A, -B and -C (Table 2) . To provide bidirectional coverage of the full length of the exon 1A/1B/1C coding sequences, each of these exons was divided by internal primers into three overlapping amplicons of approximately 300 bp. 

Amplification was performed in 20 μL reactions containing 80 ng genomic DNA, 10 picomoles forward and reverse primers, 200 nM dNTP, 1× PCR buffer, 1.5 mM MgCl₂, and 0.2 U Taq DNA polymerase (AmpliTaq Gold Enzyme; ABI). PCR amplification consisted of a denaturation step at 95°C for 9 minutes, followed by 35 cycles, each consisting of a denaturation step at 94°C for 30 seconds, an optimal annealing step for 30 seconds with temperatures as specified in Table 2 , and an extension step at 72°C for 60 seconds, followed by a final extension at 72°C for 7 minutes. 

DNA Sequencing

PCR products were analyzed on 2% agarose gels and purified by gel extraction (QIAquick; Qiagen, Valencia, CA). The PCR primers listed in Table 2 were used for bidirectional sequencing. Five microliters of PCR product was sequenced in a 10-μL reaction volume containing 3.2 picomoles of primer and 4 μL of dye terminator chemistry reaction mix (BigDye Terminator Ready; ABI). Cycling conditions were 96°C for 2 minutes, 25 cycles at 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes. Sequencing products were purified by gel filtration (Edge Biosystems, Gaithersburg, MD), dried, resuspended in 10 μL of formamide (Hi-Mi-Formamid; ABI) and denatured for 5 minutes at 95°C. Sequencing was performed on an automated sequencer (Prism 3100; ABI). Sequencing results were assembled and analyzed on computer (Seqman program of DNAStar software; DNASTAR Inc, Madison, WI). Mutations were confirmed by analyzing DNA from all affected individuals and unaffected family members. Fifty unrelated population-matched control DNA samples were analyzed by direct sequencing as well. 

I/i Blood Type Phenotyping

I/i phenotyping was tested at the Israeli National Blood Group Reference Laboratory (NBGRL) at Magen David Adom Blood Services in Israel. Testing was performed by conventional (tube) methods, ²⁵ using anti-I and anti-i from Serum Cells and Rare Fluids (SCARF) and anti-i from our in-house anti-sera collection. Cord red blood cells were used as a positive control for i and negative control for I. Adult red blood cells were the positive control for I and negative control for i. 

Results

A genome wide screen was initiated with 11 genomic DNA samples obtained from four affected and seven unaffected members of family 56007. After evaluating 148 of 385 markers (38% of the genome), we detected four consecutive 6p markers spanning a 22-cM region from D6S1574 to D6S289, segregating with the disease in this family. Pair-wise lod scores showed a maximum of 2.6 at θ = 0.00 in family 56007 (data not shown). Genotyping and haplotype results for the other three families with these markers are as shown in Figure 1 . Summed lod scores from all four families for these markers and the GCNT2 mutation are listed in Table 1 . Due to consanguinity in these families, affected family members were expected to show homozygosity for the mutation and for polymorphic markers in the vicinity of the disease gene. Centromeric obligate recombination events have taken place in affected individual 8 of family 56003 and individual 6 of family 56005 between D6S470 and D6S289. Telomeric obligate recombination events have taken place in affected individual 4 of family 56003 between D6S470 and D6S289. Thus, the disease-causing gene locus is placed on chromosome 6 between D6S470 and D6S289. The D6S470– D6S289 interval, which is the critical region mapped for these families includes the GCNT2 gene. Mutations in this gene have been shown recently to be associated with autosomal recessive congenital cataracts in families from Japan and Taiwan (Fig. 2A) . ^{11 12} Sequencing of the five exons that compose the three different GCNT2 isoforms shows a homozygous G→A substitution at position 58 of exon 2, resulting in a change of a tryptophan to a stop codon (W328X, W326X, and W328X) in all three isoforms (GCNT2A, -B, and -C) of the GCNT2 transcript respectively (Fig. 3) . This mutation was present in all affected siblings of the four families. Parents of all affected and individuals 5 and 11 of family 56007, individuals 10 and 20 of family 56005, and individual 11 of family 56003, who as obligate carriers are all heterozygotes for the mutation (Fig. 1) . Fifty population-matched control subjects were screened by means of direct sequencing, but only the wild-type variant was found (data not shown). 

Phenotyping of the corresponding red blood cells confirmed the presence of the adult i blood group. 

Discussion

This study showed a nonsense mutation in exon 2 of the GCNT2 gene responsible for autosomal recessive congenital cataract and the adult i phenotype in four consanguineous Arab families from Israel. This documents a large series of patients (n = 13) and the first description of the adult i phenotype associated with cataracts in this ethnic group. 

The i/I antigens are specific sphingoglycolipids (GLCs) present on the membranes of most human cells and on soluble glycoproteins in various body fluids, including milk, ²⁶ saliva, ²⁶ plasma, ²⁷ and amniotic fluid. ¹⁵ They were first identified on red blood cells (RBCs), where their expression was found to be developmentally regulated. The change from the linear i carbohydrate to the branched I structure gradually takes place as GCNT2 branching enzyme begins to be expressed during the first 2 years of age. ²⁸ I expression gradually decreases during the development and differentiation of mouse embryos and of embryonal carcinoma cells, which resemble multipotential cells of early embryos, ²⁹ pointing to potential roles of i/I antigens in regulation of cell growth and differentiation of various tissues. Thus, the presence of sphingoglycolipid antigens within different layers of the mammalian and chick lens, ³⁰ along with the linkage between congenital cataract formation and loss of lens GCNT2 activity strongly suggests an essential role of I/i antigens in lens development. 

The I-branching GCNT2 gene locus was recently shown to have an unusual molecular genetic arrangement, consisting of three different transcript forms, designated GCNT2A, -B, and -C, each possessing an alternatively spliced exon 1 but identical exon 2 and 3 coding regions. ¹¹ This unusual genomic organization is present in the mouse genome as well, and suggests conservation of the I-gene locus during evolution. ³¹  

Expression studies of the three GCNT2 transcripts suggest that while GCNT2C is expressed in erythrocytes, only the GCNT2B transcript is expressed in the lens, which lacks the other two forms of the enzyme. ^{11 31} In agreement with this observation, the mutation described in the present study, which is predicted to truncate 75 amino acids from the carboxyl end of all three forms of the enzyme, resulted in both adult i phenotype and congenital cataract, whereas mutation events in exon 1C that were previously described in whites without congenital cataract are expressed solely in reticulocytes. ¹¹ These do not affect the GCNT2B transcript in the lens and are not associated with congenital cataracts. Other expression studies of the GCNT2 gene in Chinese hamster ovary cells have demonstrated the importance of the truncated carboxylic end of the enzyme, for the GlcNAc-transferring (branching) activity. ¹²  

Of interest, GCNT2 isoforms are abundantly expressed in various nonerythroid tissues but, so far, allelic variants of the gene have been related only to the adult i phenotype, with or without congenital cataract. Further molecular studies on recessive congenital cataracts may detect mutations in exon 1 of GCNT2B which would be predicted to inactivate selectively the transcript expressed in the lens and therefore would be expected to result in congenital cataract without the i blood phenotype. 

We have mapped three congenital cataract families to the short arm of chromosome 3. ¹⁰ After establishing suggestive linkage to the short arm of chromosome 6 in family 56007, we checked all our recessive congenital cataract families for linkage to 6p. Surprisingly, two of the three families (56003a and 56003b in this study) showed linkage to 6p and shared the same GCNT2 nonsense mutation. In retrospect, linkage to 3p in these two families was probably a false-positive result. In the third 3p family, linkage to 6p was excluded, GCNT2 was mutation free, and affected cataract family members had the normal I blood group. Another report of a Lebanese Arab family with autosomal recessive congenital cataract mapped to the same region on 3p with maximum lod scores between 4.41 and 6.64 at θ = 0.00 (Gal A, et al. IOVS 2000;41:ARVO Abstract 1). Thus, it is still likely that an autosomal recessive cataract locus is located on 3p21. 

Anti-I antibody is considered as a benign, naturally occurring cold reactive autoantibody observed when testing is conducted at room temperatures or below. However, strong anti-I which may cause severe hemolytic anemia after blood transfusion has been reported in patients with the adult i blood group. ³² Therefore, consideration could be given to examining patients with autosomal recessive and sporadic congenital cataracts who are to undergo transfusion for the presence of the adult i phenotype. 

 ErP is a recipient of an American Physicians Fellowship for Medicine in Israel.

 Submitted for publication October 9, 2003; revised December 23, 2003; accepted January 14, 2004.

 Disclosure: Er. Pras, None; J. Raz, None; V. Yahalom, None; M. Frydman, None; H.J. Garzozi, None; El. Pras, None; J.F. Hejtmancik, 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.

 

Figure 1.

 

Family pedigrees and haplotypes. Four chromosome 6 short arm microsatellite markers allele readings, and the GCNT2 983 G→A change (allele 2) are shown. The marker order is shown to the left of each generation. *Family 56003a and -b were described previously corresponding to families III and II in Pras et al. ¹⁰

Table 1.

 

Two-Point Lod Scores between Autosomal Recessive Congenital Cataract Families, Four Chromosome 6p Markers and the Present GCNT2 983 G→A Mutation

Position (cM)	Mb	Marker Theta	0	0.01	0.05	0.1	0.2	0.3	0.4	Maximum Lod
8.7	5.9	D6S1574	−∞	2.22	3.35	3.56	2.94	1.87	0.76	3.57 at θ = 0.092
13.6	8.1	D6S309	−∞	1.64	2.68	2.73	2.02	1.05	0.2	2.77 at θ = 0.077
17.7	10.1	D6S470	−∞	8.67	8.46	7.61	5.54	3.34	1.33	8.75 at θ = 0.019
20	10.7	I-mutation	12.42	12.19	11.2	9.97	7.36	4.65	2.04	12.42 at θ = 0.00
30.7	15.3	D6S289	−∞	4.44	5.52	5.3	3.97	2.33	0.83	5.53 at θ = 0.058

Table 2.

 

DNA Primers and Their Position (Celera Database Clone hcg:1818200) and Annealing Temperatures Used for PCR Amplification of Human GCNT2 Exons

Exon	Annealing Temperature (°C)	Fw Primer	Rv Primer
1A	60	(11538) 5′-TGTAGACACAGGTTGCAGGTTAGCA-3′	5′-GCAGGTAGCTTCATCAAGGGTA-3′ (11760)
	55	(11672) 5′-TAGCAGAAGCCTGTCATCAG-3′	5′-CCTTCAGATACTGAACTATTTC-3′ (12163)
	55	(12103) 5′-AACACCTGCGGGCAAGACTT-3′	5′-CTTTTGTCCTGTGAACAGAGCGGTT-3′ (12463)
1B	55	(38989) 5′-AGACTTACAGATTTTGACGGT-3′	5′-TAGATATTTTGGGGCATGTA-3′ (39402)
	57	(39340) 5′-CCATCATCACTTTGACACCT-3′	5′-CTTATCACATAGGAAAGCTCT-3′ (39748)
	55	(39699) 5′-CTCATGCAATTGGACGGACT-3′	5′-GGGTGAGAACTATATATGTTCCAGTT-3′ (40078)
1C	55	(68534) 5′-GCAAATTCAACCTCTCACACCGATC-3′	5′-GGGGCATATAGATAGCCCTAA-3′ (68971)
	55	(68907) 5′-TGTCATGGTCATCCATAAGG-3′	5′-CTTGGTGGACATATTTAGTT-3′ (69313)
	55	(69234) 5′-AGGATTTAAAGGGAAAAATATC-3′	5′-TGAGTCAGTTCTCTAGGCGAGCAG-3′ (69607)
2	57	(103624) 5′-CTGAAGTGGAGAAACCCTGGCTTA-3′	5′-AACCCTGGATTCCACAGCTACCTT-3′ (104138)
3	55	(108936) 5′-AGTTGTAGTTAGTCGGAGAGTACCT-3′	5′-TATAATTACGTAGCCAGGTCCTGAA-3′ (109366)

Figure 2.

 

Genomic organization, positions of mutations (A) and transcript expression profile (B) of the GCNT2 gene. (A) The whole genomic sequence is contained within Celera sequence hcg:1818200 (myScience, http://myscience.appliedbiosystems.com/). Boxes: exons. Mutations identified previously, their respective citation references, and the novel W328X premature stop mutation are illustrated. The predictive amino acid alteration, its position in GCNT2A and -C, and corresponding adult i population type are shown. (B) Structures and expression profile of the three GCNT2 forms. Reference numbers of the quoted studies are shown in brackets. Exon 1A and 1C possess six more coding nucleotides than exon 1B. Thus, the GCNT2B protein is shorter by two amino acids than the GCNT2A and -C proteins. To arrive at the position of the same nucleotide change in GCNT2B, subtract 6 bp from the position indicated for GCNT2A and -C.

Figure 3.

 

Sequence chromatograms from a normal control subject (top) and a patient (bottom). A G→A change in the patient resulted in a premature stop codon (TAG).