SOD1: A Candidate Gene for Keratoconus

Abstract

purpose. To screen superoxide dismutase 1 (SOD1) on chromosome 21 as a possible candidate gene for familial keratoconus (KC).

methods. Total genomic DNA was extracted from the blood of 15 different KC families and 156 unaffected subjects. All five exons of the SOD1 gene were sequenced. For a rapid screening test, DNA was amplified by polymerase chain reaction (PCR), digested with HpyCH4 III or analyzed by radioactively end-labeled exon PCR. RNA was extracted from leukocytes and reverse transcribed to cDNA, and the PCR was amplified for splice variants. Some samples were cloned and sequenced.

results. A heterozygous genomic 7-base deletion in intron 2 of the SOD1 gene was identified in two KC families (pedigrees 1 and 6). The deletion segregated within pedigree 1 and was absent in 312 chromosomes from normal individuals. RNA from the proband of pedigree 1 showed that in addition to the wild-type transcript, two other transcripts were expressed for the CuZn SOD (SOD1) gene: lacking entire exon 2 (LE2) and lacking entire exon 2 and entire exon 3 (LE2E3).

conclusions. A unique genomic deletion within intron 2 close to the 5′ splice junction of the SOD1 gene was identified in three patients with KC. Moreover, mRNA from one affected individual also had two transcript splice variants (LE2 and LE2E3) that others have shown to code for proteins lacking the active site of the SOD1 enzyme. Further studies should be conducted to determine whether a causal relationship exists between these two events that may increase oxidative stress and be associated with KC.

The frequency of keratoconus (KC) is approximately 1 in 2000 individuals in the general population, and it is a primary indication for cornea transplantation in developed countries. ^{1 2 3} The major feature of KC is progressive corneal thinning that has been associated with increased degradative enzyme activities. ^{4 5 6 7 8} KC corneas also have an accumulation of cytotoxic by-products from the nitric oxide and lipid peroxidation pathways, ^{9 10} abnormal antioxidant enzymes, ^{11 12 13 14} and increased levels of mitochondrial DNA damage. ¹⁵ These findings suggest that ongoing oxidative stress contribute to KC but the underlying molecular and/or genetic abnormalities are unknown. 

The familial nature of KC has been reported in several studies. ^{16 17 18 19 20 21} Several modes of inheritance of KC have been suggested, including autosomal recessive and autosomal dominant. ^{2 22 23 24} There is a high concordance rate for KC in monozygotic twins. ^{16 17 18 19 20 21} It is associated with genetic disorders such as Leber’s congenital amaurosis, ^{25 26 27 28 29} trisomy 21, ^{1 30 31} and Turner’s syndrome. ^{32 33} Although recent genetic studies report linkage to at least seven different chromosomes (21, 20q12, 20 p11-q11, 17, 16q, 15q, 13, 5q14.3-q21.1, and 3p14-q13) within KC patients (Rabinowitz YS, et al. IOVS 1999;40:ARVO Abstract 2975), ^{28 34 35 36 37 38 39 40 41} to date, the underlying genetic defects for KC are not clear. 

There is a recognized association between KC and Down syndrome (trisomy 21). ^{1 30 31} Approximately 15% of Down syndrome patients have KC in the white population, although it is not common in Asian populations. ⁴² Increased oxidative damage and the superoxide disumutase-1 (SOD1) gene are also implicated in Down syndrome. Genetic studies show that KC may be linked to chromosome 21 (Rabinowitz YS, et al. IOVS 1999;40:ARVO Abstract 2975), the same chromosome on which SOD1 is located. Mutated SOD1 gene variants are associated with amyotrophic lateral sclerosis (ALS), ^{43 44} which is characterized by a rapid and progressive loss of motor function. These variants lead to structural defects in the SOD1 dimer that may contribute to nonspecific, neurotoxic effects on motor neurons. ⁴⁴ However, there is no known association between ALS and KC. 

In an effort to understand the possible involvement of the SOD1 gene in KC, we performed mutation analysis of the entire coding sequence using a set of 15 unrelated individuals, each with a family history of KC. We found an IVS2 +50del7 change within intron 2 in two families. This 7-base deletion segregated with the KC subjects in pedigree 1. The analysis of the SOD1 mRNA of one of these subjects showed in addition to the wild-type transcript, the presence of LE2 (lacking entire exon 2) and LE2E3 (lacking entire exon 2 and entire exon 3) variants, which others have shown lack the active site of the enzyme encoded in exons 2 and 3. ⁴⁵  

Materials and Methods

Patient Selection

Patients were identified from a private practice setting by a corneal specialist (ABN). The diagnosis of KC in all patients was based on more than one of the following criteria: Munson’s sign; Rizzuti phenomenon; slit-lamp findings of stromal thinning, Vogt’s striae, Fleischer ring, or scarring (epithelial or subepithelial); retroillumination signs: scissoring on retinoscopy, oil-droplet sign (Charleux); photokeratoscopy signs: compression of mires inferotemporally, inferiorly or centrally; and videokeratography signs: localized increased surface power and/or inferior superior dioptric asymmetry. Two cases were unilateral and all others were bilateral (Table 1) . All the patients had at least one other family member with documented KC. Four patients had undergone bilateral corneal transplantation for KC. Control individuals (n = 156) were over the age of 65, had undergone a thorough ophthalmic examination, and did not have any signs or family history of KC. None of the subjects showed any signs or symptoms of Down syndrome or ALS. Patient consent was obtained according to the Cedars-Sinai Medical Center Institutional Review Board no. 2247 and according to the guidelines of the Declaration of Helsinki. 

DNA Extraction

The peripheral blood (5 mL) was collected in tubes containing 10 mM EDTA. Genomic DNA from white blood cells was isolated with a kit (Puregene DNA; Gentra, Minneapolis, MN). Samples were aliquoted and stored until further use. 

DNA Sequencing

Primers were designed flanking all exons of the gene (F, forward; R, reverse): exon 1F: GAT TGG TTT GGG GCC AGA GTG, exon 1R: GAC CCG CTC CTA GCA AAG GTG; exon 2F: ACT CCC AAG TCT GGC TGC TTT TT. exon 2R: GGG GTT TTA ACG TTT AGG GGC TA; exon 3F: ATG CAG GTC AGC ACT TTC TCC AT, exon 3R: GAA CTC CAG AAA GCT ATC GCC ATT; exon 4F: CCT TGA TGT TTA GTG GCA TCA GC, exon 4R: TCT GGA TCT TTA GAA ACC GCG ACT; exon 5F: TTT GGG TAT TGT TGG GAG GAG GT, exon 5R: AAA TCT GTT CCA CTG AAG CTG TT. They were localized within the intron to detect any variations near the splice site. After polymerase chain reaction (PCR), the samples were electrophoresed on a 2% agarose gel. The bands were cut from the gel and purified using columns (Qiagen, Valencia, CA). DNA sequencing was performed with a terminator sequencing kit (Thermosequenase; USB, Cleveland, OH), as recommended by the manufacturer, and products were separated on a sequencing gel as described by Small et al. ⁴⁶ The SOD1 RNA sequence (GenBank ID accession NM_000454; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information [NCBI], Bethesda, MD) was used as a reference for numbering the base changes. The NCBI sequence of SOD1: version NM_000454.4 was used as the reference sequence for RNA, and AY835629 was used for genomic DNA sequence. 

Radioactively End-Labeled Exon PCR

The primer designated exon 2F was end labeled as described by Udar et al. ⁴⁷ After PCR with labeled primer exon 2F and primer exon 2R, the products were run on a sequencing gel. ⁴⁷ Statistical analysis was performed with the Fisher exact two-tailed test. 

SOD1 Exon 2 Digestion with the HpyCH4 III Restriction Enzyme

The PCR product (10 μL) for exon 2 was digested with HpyCH4 III (New England Biolabs, Ipswich, MA) overnight at 37°C. The digested products were run on a 2% agarose gel and visualized with a phosphorescence imager (FMBIOIII; Hitachi, Inc., San Francisco, CA). 

RNA Extraction

The peripheral blood (8 mL) was collected in tubes containing 10 mM EDTA from two normal control subjects (non-KC) and the proband of pedigree 1. The white blood cells were isolated with a ficoll gradient column. RNA was extracted from the cells (RNeasy Mini kit; Qiagen) and quantified using the (2100 Bioanalyzer nano-RNA protocol; Agilent, Palo Alto, CA). 

Reverse Transcription–Polymerase Chain Reaction

RNA (250 ng) was reverse transcribed into cDNA with an oligo (dT) primer (1 μM; BD Biosciences, Mountain View, CA) and 200 U of reverse transcriptase (SuperScriptII; Invitrogen, Carlsbad, CA). The RNA and primer were incubated initially for 3 minutes at 70°C, 2 minutes at 50°C, and held at 42°C, while 1 mM deoxyribonucleoside triphosphate, 1× first-strand buffer, 0.1 M dithiothreitol (DTT) and reverse transcriptase were added (Invitrogen). The reaction proceeded for 1 hour at 42°C and was terminated at 70°C for 15 minutes. cDNA was stored at 20°C. 

RT-PCR was performed with 1 μL of cDNA, Taq polymerase buffer, 200 μM deoxyribonucleoside triphosphates, 1.25 U Taq polymerase (Promega Biotech, Madison, WI) and 250 nM of primers either from exon 1F: AGT GCA GGG CAT CAT CAA TTT CGA GCA G, exon 4R: GAT GCA ATG GTC TCC TGA GAG TGA GAT C, or exon 1F; exon 5R: GTA CTT TCT TCA TTT CCA CCT TTG C in a total volume of 50 μL. Amplified products were resolved by electrophoresis in 2% agarose gels and visualized under ultraviolet light after they were stained with ethidium bromide. Routine PCR controls without cDNA (water control) were negative. Gels were scanned using the phosphorescence imager (Hitachi). 

Cloning and Sequencing of the Splice Variants

The amplified cDNA PCR products were purified (NucleoSpin columns; BD Biosciences Clontech) and were then subcloned into the pGEM-T easy vector (Promega, Madison, WI). Nine of the clones were sequenced and analyzed (BioEdit 7.0.5 sequence alignment editor; Tom Hall, Ibis Therapeutics, Carlsbad, CA). 

Results

All five SOD1 exons and flanking intronic regions were sequenced in 15 individuals with KC (Table 1) , representing unrelated families. We found a 7-base deletion in two patients with KC, the probands of pedigrees 1 and 6, which represents 2 of 30 chromosomes (Table 1 ; Figs. 1 2 ). The nucleotide change IVS2+50del7 segregated in one family (see pedigree 1, Fig. 2A ). Segregation studies in the other family were not possible due to lack of available samples. The DNA from 156 normal unaffected individuals (control) was also analyzed for the presence of this deletion by using the radioactive PCR method. The change was absent in these 312 control chromosomes (P < 0.008). We also developed a method that allowed rapid screening for the IVS2+50del7 mutation. The HpyCH4 III restriction enzyme digested the samples with the 7-base deletion into two distinct bands (255 and 223 bp), whereas the wild-type had a single 283 bp band (Fig. 2C) . 

To understand the significance of the 7-base intronic deletion close to the splice site that is present in three affected individuals representing two families and is absent in control subjects, we analyzed the SOD1 transcripts from the RNA extracted from the blood of one of the affected individuals (proband, pedigree 1) and two normal control subjects. RT-PCR analysis using primers located within exon 1 and either exon 4 or exon 5 demonstrated in the KC sample the presence of two transcripts in addition to the wild-type product encompassing the entire gene (Fig. 3) . The PCR products were cloned, sequenced, and aligned with the wild-type SOD1 sequence (NM_000454). The results demonstrated that there were two splice variants present (Fig. 3 , Table 2 ) LE2 and LE2E3. The normal individuals showed the wild-type transcript. Figure 4shows the sequenced clones from the proband of pedigree 1 with the LE2 and LE2E3 variants. 

Discussion

SOD1 encodes a 645-bp transcript with a genomic length of 9320-bp and a 154-amino-acid (AA) protein product. In addition to the wild-type transcript consisting of five exons, distinct splice variants have been identified by RT-PCR. ^{45 48} When the SOD1 splice donor site sequences were matched with the corresponding consensus sequences, it was found that intron 1 had the lowest degree of match (77.2), and this may reduce the splice donor site usage and lead to alternative splicing. ⁴⁵ In addition, intron 2 had low splice acceptor site scores that may lead to reduced usage of the splice site and the lack of partial or entire exon 2 and/or exon 3. ⁴⁵  

Although mutations in SOD1 are known to be associated with familial ALS, ^{49 50 51 52} we did not find this 7-base deletion in the ALS database either as a mutation or polymorphism. This intronic deletion may influence the usage of the splice site such as that reported for exon 5 with resultant truncation or complete deletion of the protein. ⁵³ Intronic deletions have been reported previously as close to the splice junction or as far as 189 bp from the splice site. ^{53 54 55} Different SOD1 isoforms have various levels of carbonylation and susceptibility to oxidative modification, ⁵⁶ thereby leading to different functions for the isoforms. Moreover, the splice variants that are missing exon 2 or exons 2 and 3, yield proteins that lack the active site of the CuZn (SOD1) enzyme and are likely to have modified SOD1 function. ⁴⁵ In a cell-free translation system, the exon 2-skipping and exon 2- and 3-skipping SOD1 showed weak to no protein expression. ⁴⁸ In addition, transfection studies using 293T cells showed no expression for either exon-skipping SOD1. ⁴⁸ These findings imply that elimination of exon 2 and exons 2 and 3 has significant effects on SOD1 levels and functions. 

Our analysis of RNA transcripts showed that the KC proband from family 1 had LE2 and LE2E3 variants in addition to the wild-type transcript. Previous studies have demonstrated that the LE2 and LE2E3 variants code for inactive proteins and changed the translation reading frames. Based on these findings, we hypothesize that there is a relationship between the genomic change and the RNA variants and there is a relationship between the presence of the inactive variants and KC. 

The protein encoded by this SOD1 gene binds copper and zinc ions (CuZn-SOD) and is one of two isozymes responsible for destroying free superoxide radicals in the body. The encoded isozyme is a soluble cytoplasmic protein, acting as a homodimer to convert naturally occurring but harmful superoxide radicals to molecular oxygen and hydrogen peroxide. Mutations in this gene have been associated with familial ALS and rare transcript variants have been reported for the gene. ^{43 44} Some patients with familial ALS have mutated SOD1 gene variants, altered SOD1 crystallographic structures, and decreased SOD activity. ⁴⁴ By as yet unknown mechanisms, the mutated SOD1 variants may cause neurotoxic effects that damage the motor neurons. ^{43 44}  

The CuZn-SOD is an essential antioxidant enzyme found within the human cornea. We hypothesize that the predominance of a variant SOD1 transcript with exon 2-skipping or exon 2- and 3-skipping deletions may lead to a decrease in enzyme activities in KC corneas. This would lead to an inability of the corneal tissue to detoxify harmful superoxide radicals, thereby making them available for interaction with nitric oxide to form peroxynitrite, which is found excessively in KC corneas. ¹⁰ Of note, normal CuZn-SOD activities have been reported in sporadic KC ^{13 57} but have not been studied in familial KC. It is likely that there are multiple pathogenetic factors involved in both familial and sporadic KC so it would not be unexpected to have varied responses. 

Associations between KC and other candidate genes have been examined. The transforming growth factor-β induced gene (TGF-βI; BIGH3), which is associated with various corneal dystrophies, ^{58 59 60 61} was found not to be associated with KC. ⁶² VSX1 has been identified in a patient who has both KC and posterior polymorphism dystrophy. ³⁹ Moreover, novel mutations have been reported in the VSX1 gene in a series of individual patients with KC, ⁴⁰ but a biological mechanism connecting this gene to KC is not clear. Keratoconus is more common in patients with Down syndrome than in normal individuals. A defect of the SOD1 gene, with its location on chromosome 21 and association with Down syndrome, could play a role in the increased oxidative damage found in KC corneas. Moreover, recent studies show that oxidative stress elements, such as peroxynitrite and nitric oxide, can increase degradative enzyme activities, which could play a role in the stromal thinning characteristically found in KC. ⁶³ Future studies should include screening of both additional families with KC and sporadic KC patients. In addition, the possible relationships between the IVS2+50del7 genomic deletion and expression of the SOD1 variants LE2 or LE2LE3 should be explored in vitro. 

 Supported by NIH Grant EY06807, the Iris and B. Gerald Cantor Foundation, the Schoellerman Charitable Foundation, the Discovery Eye Foundation, the Skirball Molecular Ophthalmology Program, Research to Prevent Blindness, the National Keratoconus Foundation, and the McCone Foundation.

 Submitted for publication November 23, 2005; revised February 10, 2006; accepted May 26, 2006.

 Disclosure: N. Udar, None; S.R. Atilano, None; D.J. Brown, None; B. Holguin, None; K. Small, None; A.B. Nesburn, None; M.C. Kenney, 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.

 

Patient Information and Results

Pedigree	Family	Sex	Age (y)	Bilateral	IVS2+50del7
1	KC-A	M	61	+	HT
2	KC-B	M	51	+	NL
3	KC-C	F	55	+	NL
4	KC-E	F	42	+	NL
5	KC-G	F	57	+	NL
6	KC-H	M	52	+	HT
7	KC-I	F	42	+	NL
8	KC-J	M	81	+	NL
9	KC-K	F	42	+	NL
10	KC-L	F	40	+	NL
11	KC-P	F	69	+	NL
12	KC-Q	M	54	OD	NL
13	KC-R	F	46	+	NL
14	KC-T	M	69	OS	NL
15	KC-S	F	43	+	NL

 HT, heterozygous; NL, normal.

Figure 1.

 

Sequencing identified a genomic 7-base deletion in intron 2 (IVS2+50 del7) of the SOD1 gene in two KC families. (A) Sequencing gel showing the 7-bp deletion in the proband from pedigree 1. A similar deletion was found in the affected offspring and in the proband from pedigree 6. (B) The location of the 7-base deletion from intron 2. The bold italic sequence indicates the primer location. The lowercase letters indicate the nucleotide position reported for SOD1 cDNA. The bottom number indicates the amino acid position. The boxed residues are deleted in the mutant allele.

Figure 2.

 

Pedigree 1 shows affected individuals 1 and 17, and pedigree 6 shows affected individual 11. (A) The IVS2+50del7 change (Del +) was found in these affected individuals. The unaffected individuals did not show the base change (Del−). ND, not determined. (B) Representative gel showing the IVS2+50del7 change in one allele of the proband from pedigree 1 (lane 1), the proband from pedigree 6 (lane 11), and the affected daughter of proband 1 (lane 17). The radioactively end labeled exon 2 PCR products were run on a 7 M urea acrylamide gel. The affected (Del+) showed heterozygous products with the 283- and 276-bp bands. The probands from the other pedigrees had only the homozygous 283 bp product. (C) Restriction enzyme digestion with HpyCH4 III shows 2 bands representing detection of IVS2+50del7 change.

Figure 3.

 

The proband from pedigree 1 (P1) showed wild-type variants but also had two major splice variant products. PCR amplification of the cDNA isolated from leukocytes was performed with either the primers 1F/4R (SOD1 1–4) or else the primers 1F/5R (SOD1 1–5). The P1 individual had wild-type products (primers 1F/4R, 298 bp; and primers 1F/5R, 366 bp) but also had two major splice variant products (primer 1F/4R, 201 and 131 bp; and primers 1F/5R 269 and 199 bp). The normal individuals (N1 and N2) showed the wild-type products with minimal, small-sized products. WT, wild-type; N1, normal 1; N2, normal 2; P1, proband of pedigree 1; −, water control.

Table 2.

 

Summary of SOD1 Variants with Different Exon Primer Pairs

	SOD1 1-4 Primers	SOD1 1-5 Primers
N1	Wild-type (298 bp)	Wild-type (366 bp)
N2	Wild-type (298 bp)	Wild-type (366 bp)
P1	Wild-type (298 bp)	Wild-type (366 bp)
	LE2 (201 bp)	LE2 (269 bp)
	LE2E3 (131 bp)	LE2E3 (199 bp)

Figure 4.

 

The results from sequencing clones from the proband of pedigree 1. The wild-type SOD1 (CuZn SOD) has all five exons. One variant is LE2 and the other is LE2E3.

The authors thank all participating patients and their families.