Molecular Vision 2006; 12:159-176 <http://www.molvis.org/molvis/v12/a18/> Received 13 October 2005 | Accepted 8 March 2006 | Published 10 March 2006

Download Reprint

CHST6 mutations in North American subjects with macular corneal dystrophy: a comprehensive molecular genetic review

Gordon K. Klintworth,^1,2 Clayton F. Smith,² Brandy L. Bowling²
 
 

Departments of ¹Pathology and ²Ophthalmology, Duke University Medical Center, Durham, NC

Correspondence to: Gordon K. Klintworth, MD, PhD, Duke University Medical Center, Box 3802, Durham, NC, 27710; Phone: (919) 684-3550; FAX: (919) 684-9225; email: Klint001@mc.duke.edu

Abstract

Purpose: To evaluate mutations in the carbohydrate sulfotransferase-6 (CHST6) gene in American subjects with macular corneal dystrophy (MCD).

Methods: We analyzed CHST6 in 57 patients from 31 families with MCD from the United States, 57 carriers (parents or children), and 27 unaffected blood relatives of affected subjects. We compared the observed nucleotide sequences with those found by numerous investigators in other populations with MCD and in controls.

Results: In 24 families, the corneal disorder could be explained by mutations in the coding region of CHST6 or in the region upstream of this gene in both the maternal and paternal chromosome. In most instances of MCD a homozygous or heterozygous missense mutation in exon 3 of CHST6 was found. Six cases resulted from a deletion upstream of CHST6.

Conclusions: Nucleotide changes within the coding region of CHST6 are predicted to alter the encoded protein significantly within evolutionary conserved parts of the encoded sulfotransferase. Our findings support the hypothesis that CHST6 mutations are cardinal to the pathogenesis of MCD. Moreover, the observation that some cases of MCD cannot be explained by mutations in CHST6 suggests that MCD may result from other subtle changes in CHST6 or from genetic heterogeneity.

Introduction

Macular corneal dystrophy (MCD; OMIM 217800), a rare inherited disorder first described in 1890 by Groenouw [1], has been identified throughout the world. MCD is most prevalent in India [2-4] and Saudi Arabia [5]. In some countries, this disease accounts for a high percentage (10-75%) of the corneal dystrophies requiring keratoplasty [6,7]. Clinically, MCD is characterized by a cloudiness of the cornea and irregularly shaped superficial opacities of both eyes that progressively extend through the entire thickness of the central and peripheral corneal stroma. The corneal stroma is thinner than normal [8-10]. In 1938, its autosomal recessive mode of inheritance became appreciated [11]. More than two decades later Jones and Zimmerman [12,13] differentiated MCD histopathologically from the other two major stromal corneal dystrophies known as granular and lattice corneal dystrophy. Histopathologically, MCD is typified by an intracellular storage of glycosaminoglycans (GAGs) within keratocytes and the corneal endothelium combined with an extracellular deposition of similar material in the corneal stroma and Descemet's membrane [5,14]. Guttae are common on Descemet's membrane [15]. In 1964, because of similarities to the systemic mucopolysaccharidoses, Klintworth and Vogel [16] suggested MCD might be a mucopolysaccharidosis localized to the cornea. A metabolic defect in keratan sulfate (KS), the major corneal GAG [17], was suspected because of the histochemical attributes of the corneal accumulations and that the cornea was the only site of overt abnormalities. Cell culture studies disclosed differences from the systemic mucopolysaccharidoses [18], and analyses of sulfated GAGs produced by organ cultures of corneas with MCD led to the discovery that corneal tissue with MCD does not synthesize KS [19] or normal KS containing proteoglycans (PGs) because of defective sulfation [20,21]. Immunochemical studies using an antibody that recognizes antigenic KS (AgKS) disclosed heterogeneity among cases of MCD based on the reactivity of corneal tissue with the antibody [22]. Subsequently this led to the recognition of three immunophenotypes that are clinically and histopathologically indistinguishable from each other [5,22-24]. Most often neither the serum nor the corneal tissue contain AgKS (MCD type I), but sometimes AgKS is absent in the corneal stroma and the serum but can be detected in the keratocytes (MCD type IA). A third immunophenotype (MCD type II) is characterized by the presence of AgKS in corneal tissue and detectable serum levels of AgKS that are often present in normal amounts. MCD type I corneas have been found to synthesize an abnormal KS-PG with lactosaminoglycan side chains that lack sulfate. In sharp contrast, an MCD type II cornea produced a normal ratio of KS-PG to dermatan sulfate-PG, but the net synthesis of PGs was below normal [25]. However, the chondroitin/dermatan sulfate side chains on decorin were sulfated. Subsequently, fluorochrome-assisted carbohydrate electrophoresis disclosed that the KS chain size within the cornea and cartilage in MCD type I was reduced and chain sulfation was absent [26]. In a cornea with MCD type II, the sulfation of N-acetylglucosamine and galactose was significantly reduced and the chain size was also reduced, but to a lesser degree than in MCD type I [26].

The absence or paucity of sulfate in KS and KS-containing PGs in corneas with MCD suggested that patients with MCD were deficient in a carbohydrate sulfotransferase that catalyzes the transfer of sulfate groups from 3'-phosphoadenosine 5'-phosphosulfate (PAPS) to KS in a manner comparable to which sulfate esters are put on other endogenous and exogenous substrates (PAPS + R-OH ->PAP + R-OSO₃). An analysis of the serum in patients with MCD type I disclosed normal levels of enzymatic activity for sulfating at least one of the two sugars present in KS and an enzyme deficient for sulfating N-acetylglucosamine (GlcNAc) was thought to be present [27]. Because the serum KS is considered to be derived from cartilage, the discovery of undetectable levels of AgKS in the serum of individuals with MCD type I suggested that the KS sulfotransferase deficiency was not restricted to the cornea. Direct evidence of cartilage involvement was obtained later in cartilage from the nose and ear in MCD type I [26,28]. Cartilage lacks abnormal accumulations comparable to those in the cornea, but the chondrocytes and extracellular matrix do not contain AgKS and the KS content of cartilage is at least 800 times lower than normal [28].

Using families from Iceland, where MCD was the most frequent indication for penetrating keratoplasty [7], the gene for MCD type I was mapped to chromosome 16 (16q22) and linkage data hinted that MCD type II was also at this locus [29]. Fine mapping refined the location of the MCD gene. Two candidate genes (TAT and LCAT) in that part of chromosome 16 were excluded [30-32].

A sulfotransferase suspected of being defective in MCD would belong to the galactose/N-acetylgalactosamine/N-acetylglucosamine 6-O-sulfotransferase family of carbohydrate sulfotransferases. This family of enzymes catalyzes 6-O-sulfation on the 6-hydroxyl of GlcNAc, galactose (GAL), or N-acetylgalactosamine (GalNAc) [33]. All known sulfotransferases contain highly conserved regions [34], including a 5'-PSB (phosphosulfate-binding) and a 3'-PB (3'-phosphate binding domain) [35]. The carbohydrate sulfotransferases in different species possess marked sequence similarities at the amino acid level particularly in their catalytic domains and, with the exception of CHST3 (also known as GST-0), the coding sequences of the open reading frames (ORFs) for all CHST genes are within a single exon [36]. The galactose/N-acetylgalactosamine/N-acetylglucosamine 6-O-sulfotransferases have been designated as belonging to the GST family [36], but this term is not favored because they can be confused with glutathione transferases.

One potential candidate for the MCD disease gene was CHST1 (also known as GST-1) that encodes for KS 6-O-sulfotransferase, but it could be excluded because its gene had been mapped to human chromosome 11 (11p11.1-11.2) [37]. Further research on carbohydrate sulfotransferases resulted in finding a cluster of three carbohydrate sulfotransferases with apparent tissue restrictions on human chromosome 16 [33]. CHST4, which encodes high endothelial cell GlcNAc 6-O-sulfotransferase, was mapped to human chromosome 16 (16q23.1-23.2) [36]. Shortly thereafter, two more highly homologous carbohydrate sulfotransferases genes were identified in the same region by independent investigators [33,36]. These genes were CHST5 (also known as GST-4α and intestinal GST) [38] and CHST6 (also designated GST-4β) [36,39].

Because the gene responsible for MCD was suspected of encoding for a carbohydrate sulfotransferase and had been fine mapped to the same region as CHST5 and CHST6, these genes became logical suspects for the MCD disease gene. In 2000, Akama et al. [33] discovered mutations in CHST6 in MCD and also found insertional or deletional defects in the region between CHST5 and CHST6 in some cases. Subsequently, several laboratories confirmed these observations [2-4,39-52].

This report documents an analysis of CHST6 in 57 patients from 31 families from the United States with MCD, and reviews these findings with all previously documented mutations.

Methods

Subjects

We analyzed CHST6 in 141 individuals (57 affected patients, 57 carriers [parents or children], and 27 normal blood related family members) from 31 families with MCD. In all instances, the diagnosis of MCD was made on a combination of the characteristic clinical features together with the typical histopathologic findings in the corneal tissue obtained following a penetrating keratoplasty in one or both eyes. All patients were born in the United states and some ancestral lines could be traced to England, Ireland, Germany, Holland, Northern Ireland, Scotland, Switzerland, and Norway. Nonmolecular genetic studies have been previously reported on some of these cases [14,16,18,19,53-56]. This study was approved by the Internal Review Board of Duke University and conformed to the tenets of the declaration of Helsinki. Written informed consent was obtained from all participants. To evaluate the CHST6 gene in 11 deceased or unavailable patients from five families with MCD, we analyzed the DNA of surviving siblings, parents, and children to establish which mutations had been transmitted to the carriers. We also analyzed DNA from 50 control specimens from the United States and DNA from 10 normal spouses who married into these families.

Determination of MCD immunophenotypes

Prior to molecular genetic studies, the immunophenotypes had been determined in 32 of the 57 subjects with MCD by ascertaining the presence or absence of sulfated epitopes in KS in the corneal tissue and/or serum. The reactivity of formalin-fixed paraffin-embedded corneal tissue sections with the 5D4 anti-KS monoclonal antibody was evaluated immunohistochemically at a dilution of 1:5,000 using the avidin-biotin immunoperoxidase complex method [22]. For negative controls, diluted normal mouse serum was used. Serum AgKS levels were measured in 22 patients with MCD with an ELISA using an anti-KS monoclonal antibody (5D4; ICN Biomedical Inc., Costa Mesa, CA) directed against a highly sulfated epitope present on long KS chains [57]. These immunochemical studies were assessed in a masked fashion without clinical information or knowledge of the serum levels of AgKS.

DNA isolation

The coding region and upstream region of CHST6 in 57 patients, 57 carriers (parents or children), and 27 unaffected blood relatives from 31 families with MCD were screened for mutations. In 46 patients with MCD, CHST6 was analyzed on their own DNA, and it was done indirectly in 11 other affected subjects by studying the DNA of carriers related to them. Ten spouses of patients or their family members were also screened. DNA was isolated from blood relatives of 11 patients who were no longer available for study. DNA was extracted from peripheral blood leukocytes using the Puregene^TM Blood Kit (Gentra Systems, Minneapolis, MN) or from buccal samples using the Puregene^TM Buccal Cell Kit (Gentra Systems) as previously described by Liu et al. [40]. DNA was quantitated using the NanoDrop® ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies, Wilmington, DE).

Analyses with insufficient DNA

In cases when insufficient DNA was available for a complete analysis of CHST6, samples of genomic DNA were amplified using a protocol for the whole genome amplification (WGA) kit GenomiPhi (GE Healthcare, Piscataway, NJ, formerly Amersham Biosciences) [58-61].

Molecular analysis of the coding region of CHST6

The coding region of CHST6 was amplified in three overlapping fragments from genomic DNA by the polymerase chain reaction (PCR) using the same three pairs of primers as Akama et al. [33] (primers 17 and 19, 18 and 21, 20 and 22; Figure 1, Table 1). Each PCR was carried out in a 50 μl reaction mixture using the components of the Qiagen Taq DNA Polymerase Kit (Qiagen, Valencia, CA). Each reaction consisted of 1X PCR Buffer containing 1.5 mM MgCl₂, 1X Q solution, 0.2 mM dNTPs, 0.2 μM of each primer, 1 unit of Taq DNA polymerase (Qiagen), and 200-800 ng of genomic DNA. Amplification was performed in a PTC-225 PeltierThermal Cycler (MJ Research, Waltham, MA). PCR conditions for all three amplicons of the coding region were as follows: 2 min at 95 °C and 31 cycles of 1 min at 95 °C, 1 min at 60 °C, and 1 min at 72 °C, followed by a 4 °C hold cycle.

The resulting amplified PCR products were purified using QIAquick® PCR Purification Kit (Qiagen) and then sequenced on both strands using a fluorescent Big Dye Terminator Cycle Sequencing system (Applied Biosystems, Foster City, CA) combined with an ABI 377 PRISM DNA Sequencing instrument (Applied Biosystems). The sequences were then aligned to the CHST6 cDNA using the SeqWeb Sequence Analysis web-based program (Accelrys, San Diego, CA) to seek changes in the nucleotide and amino acid sequences of exon 3 in comparison to the published cDNA sequence of CHST6 (GenBank accession number NM_021615) and reported mutations. To help evaluate the significance of identified CHST6 nucleotide changes in patients with MCD, we compared them with the sequences found in other genes encoding the related family of carbohydrate sulfotransferases and with the protein sequences of conserved domains in the Conserved Domain Database (CDD) within the Entrez system NCBI Conserved Domain Database [62].

Analysis of genomic DNA sequence upstream of the coding region of CHST6

Genomic DNA was screened for rearrangements upstream of the coding region of CHST6 according to the method of Akama et al. [33] using the same conditions as described. The primer pairs were the same except that primer number 2' was replaced by primer number 2, and primer number 6' was changed to primer number 6 (Figure 1, Table 1). HotStarTaq® DNA polymerase (QIAGEN) was used and the annealing temperatures were adjusted to 58 °C and 62 °C for the primer pair number 8 and number 14 and the primer pair number 2 and number 6, respectively. In our experience the changing of primers number 2' and number 6' to number 2 and number 6 yielded more consistent results, perhaps because the original primers were subject to potential mispairing due to the high degree of sequence similarity of CHST5 and CHST6 in this location. All PCR amplicons were electrophoresed on 2% agarose gels, and the gels were documented using the BioChemi Image Acquisition and Analysis Software (UVP BioImaging Systems, Upland, CA). All amplicons indicating upstream DNA rearrangements were sequenced to confirm their nucleotide order. Because two of the resulting amplicons with our substituted primer pairs extended into less homologous areas, their sequences could be more readily verified.

Evaluation of splice regions

In cases of MCD that could not be explained on the basis of mutations in either the coding region of CHST6 or upstream of CHST6, possible splice site mutations were evaluated by performing PCR. This involved using primer pairs that covered the exons and their splice sites (exon 1 [primer number 7 and 8], exon 2 [number 9 and 10], exon 3 [primers number 11 and 12, and number 13 and 14] and exon 4 [primers number 15 and 16]). The resulting amplicons were sequenced to determine if any nucleotide changes were present (Figure 1, Table 1).

Determination of chromosomal location of mutations

Two approaches were used to determine whether detected mutations were on the same or different chromosomes. When possible, genomic DNA from the parents, children, and siblings were analyzed. Parents, children, and siblings of those affected with a heterozygous CHST6 mutation were all presumed to be carriers of the MCD disease gene. When specimens of DNA were not available from appropriate family members, we determined whether multiple mutations were on the same or different chromosomes by digesting PCR products with Surveyor nuclease (Transgenomic, Inc., Omaha, NE), which cleaves double-stranded DNA at mismatched nucleotides [63]. Amplicons containing the mutations were heat denatured and mixed with amplicons from controls. After annealing, the heteroduplexes were digested with Surveyor nuclease, and the products were sized on an agarose gel.

Results

CHST6 mutations

The most frequent abnormalities found in this study were single base changes in the coding region of CHST6 that altered a coded amino acid. We identified 23 such examples in exon 3 of CHST6 (Table 2, Table 3).

Single nucleotide polymorphisms

We identified four single nucleotide polymorphisms (SNPs) that did not affect an amino acid (c.258A>C [Ala86Ala], c.294C>G [Ser98Ser], c.465G>A [Arg155Arg], and c.681C>T [Gly227Gly]) and which were presumably insignificant as known mutations could account for MCD in the affected individuals. In 14 individuals with MCD from five families (family 7, 13, 15, 22, and 26), c.484C>G (Arg162Gly) was associated with c.599T>G (Leu200Arg), but c.484C>G (Arg162Gly) was clearly not pathogenic. Arg162Gly was found in 11 blood relatives of persons with MCD and six of 50 American controls were heterozygous for this nucleotide change. Moreover, an unaffected 25-year-old son of a person with MCD in family 26 was homozygous for this change. He had no vision difficulties and a visual acuity of 20/15 when last examined at 23 years of age. The unaffected parent was heterozygous for Arg162Gly and has no other nucleotide changes in CHST6. Another unaffected 36-year-old sister of a patient with MCD in family 28 was also homozygous for this change. Her 61-year-old mother was heterozygous for Arg162Gly and carried Cys246Trp on the other chromosome, but had no corneal disease. An unaffected daughter of a patient in family 7 was homozygous for Arg162Gly and heterozygous for Leu200Arg. Her unaffected parent was heterozygous for Arg162Gly and has no other nucleotide changes in CHST6.

In the analysis of the splice sites, two previously reported SNPs were detected in the 3'-UTR region of exon 4 (c.3342A>T, rs424964 and c.3501G>A, rs10871313), which presumably are common variants in the general population. Some unaffected family members were homozygous for these SNPs. We have annotated nucleotide changes which are downstream from the coding region based on NM_021615 as it is considered the best estimate of the mRNA boundaries at this time (Personal communication, L. A. Pennacchio, DOE Joint Genome Institute, June, 2005).

Number of mutations in coding region

Although the coding region of CHST6 in MCD most often contained a single mutation, more than one nucleotide change on a single chromosome was detected in 11 families (families 6, 7, 9, 12, 13, 15, 22, 23, 26, 28, and 30). In family 20, some unaffected members contained two SNPs in the same chromosome, neither of which changed an amino acid. In three families (families 9, 12, and 30), the coding region of CHST6 contained two nucleotide changes within the same chromosome, both of which were consistent with disease-producing mutations (Ser131Leu plus Thr228Asp [family 30], Leu200Arg plus Arg334Cys [family 9], and His63Gln plus Arg114Cys [family 12]).

Insertions, deletions, frameshifts, and substitutions

Nucleotide deletions were found in the coding region of CHST6 in family 10 (c.51delG) and family 13 (c.740delG). An insertion was detected in family 23 (c.573_574insC), and family 3 had a complex mutation with a combination of a deletion plus an insertion at the same site (c.271_273delGCTinsA). Nucleotide insertions within exon 3 of CHST6 (c.573_574insC) caused a frameshift in family 23 (Ala192fs). Nucleotide deletions were found in the coding region of CHST6 in four families. Single nucleotide deletions resulted in frameshifts in family 10 (c.51delG, Gln18fs) and family 13 (c.740delG, Ala247fs). The entire coding region was deleted in family 5 and family 18. In family 5, exons 1 and 4 were amplifiable with PCR, but in family 18, none of the 4 exons were amplifiable. A deletion of 3 nucleotides and an insertion of 1 nucleotide (c.271_273delGCTinsA) in Family 3 resulted in a frameshift (Ala91fs). A nucleotide substitution (c.231G>A) in family 4 generated a heterozygous stop codon (Trp77X) which is predicted to encode a truncated sulfotransferase.

Mutations 5'-UTR upstream of CHST6

Four families (families 1, 11, 16, and 19) had deletions upstream of CHST6. In all cases these were 40 kb deletions from a region upstream of CHST5 to a homologous region upstream of CHST6 (Figure 2). In two of them, the deletions were heterozygous and each was associated with an additional mutation in the coding region of CHST6 on the other chromosome (Ser53Leu, Gln122Pro). One was found in a boy whose grandmother suffered from MCD due to homozygous Gln122 Pro mutation (family 1). A single copy of this mutation was inherited by the grandson, and the upstream deletion was inherited from his father. In family 16, there was a heterozygous upstream deletion and a heterozygous c.1A>T mutation (Met1?; Human Genome Variation Society discussions regarding the description of sequence variants), but it could not be determined whether or not they were on separate chromosomes. A homozygous upstream deletion was present in family 11.

Findings in splice regions

Splice site mutations were not found in the two cases of MCD from family 24 which could not be explained by mutations in the coding region of CHST6 or upstream of the gene.

Evaluation of maternal and paternal chromosomes

Because MCD is an autosomal recessive disorder, all affected individuals are expected to inherit a mutant gene from both the mother and the father. In 24 of the 31 families the MCD could be explained on the basis of two or more mutations in the CHST6 coding region and/or by an upstream deletion (Table 2, Table 3). Of the seven families that were not proven to have CHST6 mutations on both chromosomes, one family (family 23) was found to have two mutations. We were unable to establish whether they were or were not on the same chromosome because there was not enough DNA to amplify. In one other family (family 24), neither sequence analysis of the coding region of CHST6 nor a screening of the upstream region using the method of Akama et al. [33] disclosed mutations capable of explaining the genetic basis for MCD. In one other family (family 16), there was a heterozygous upstream deletion and a heterozygous coding region mutation, but it could not be determined if they were or were not on the same chromosome. In one family (family 27) there were no coding region mutations and there was insufficient DNA to determine whether or not upstream deletions or replacements were present. In three other families (families 29, 30, and 31), DNA could not be obtained from affected family members, and in each of these families, only one chromosome-carrying mutation could be detected among family members who were carriers.

Heterozygous mutations were detected in exon 3 of CHST6 in 18 families (families 1, 3, 4, 6, 7, 9, 10, 13, 15, 16, 17, 19, 20, 22, 23, 25, 26, and 28) and almost all were in association with another heterozygous mutation in the coding region of CHST6 on the other chromosome. The three exceptions had an upstream deletion (families 1, 16, and 19) on the other chromosome. In 14 instances, two or more independent heterozygous mutations were present. Aside from having two heterozygous nucleotide changes in CHST6 (c.484C>G, Arg162Gly and c.599T>G, Leu200Arg), family 9 had another homozygous missense mutation (c.1000C>T, Arg334Cys). Family 23 had two heterozygous mutations, but it was not possible to establish whether they were or were not on the same chromosome because of insufficient DNA. In family 1, one subject (case 1) had a homozygous c.365A>C (Gln122Pro) mutation, whereas an affected sibling (case 2) had a heterozygous c.365A>C(Gln12Pro) mutation together with a deletion upstream of CHST6 on the other chromosome 16.

Homozygous mutations were detected in the coding region of CHST6 in nine families (families 1, 2, 5, 8, 9, 12, 14, 18, and 21), and in two of them (families 5 and 18) a major portion of CHST6 that included the ORF was deleted. Another family (family 11) had a homozygous deletion upstream from CHST6. One family (family 9) had a homozygous c.1000C>T (Arg334Cys) mutation, but also two heterozygous mutations (c.484C>G, Arg162Gly; c.599T>G, Leu200Arg). In another family (family 1), one affected individual was homozygous for a single mutation (c. 365A>C, Gln122Pro), while another affected individual was heterozygous for this mutation, but also had a deletion upstream of CHST6 on the other chromosome. Whereas homozygous mutations are expected in the offspring of inbred matings, we were only able to establish consanguinity in two families with extensive genealogical analyses.

Results related to the MCD type

Most subjects studied had MCD type I (26 patients from 13 families). Six affected individuals with MCD type II were from four families (families 24, 27, 30, and 31). The c.231G>A mutation that generated a stop codon (Trp77X) was found in three subjects with MCD type 1. Affected subjects in two families (families 24 and 27) with MCD type II lacked mutations in the coding region of CHST6, and those in family 24 also had no deletion or insertion upstream of this gene. In family 27, there was insufficient DNA to determine whether or not there were upstream deletions or insertions. DNA was not available in two subjects with MCD type II but one allele in these cases could be determined by analyzing DNA from children of the subjects. The nucleotide changes in CHST6 in each of these carriers were c.392C>T (Ser131Leu), c681C>T (Gly227Gly), c.682A>G, 683C>A (Thr228Asp; family 30), and c.607G>A (Asp203Asn; family 31). In 14 families (19 affecteds and 23 carriers), the MCD immunophenotype was not determined.

Discussion

Eighteen of the mutations found in this investigation have apparently not been reported before (c.51delG [Gln18fs], c.137T>C [Leu46Pro], c.189C>G [His63Gln], c.217G>C [Ala73Pro], c.231G>A [Trp77X], c.274G>C [Val92Leu], c.340C>T [Arg114Cys], c.392C>T [Ser131Leu], c.529C>T [Arg177Cys], c.573_574insC [Ala192fs], c.607G>A [Asp203Asn], c.682A>G, 683C>A [Thr228Asp], c.738C>G [Cys246Trp], c.740delG [Arg247fs], c.744C>G [Ser248Arg], c.815G>A [Arg272His], c.1046G>A [Cys349Tyr], and c.1047C>G [Cys349Trp]), but 12 have been previously documented. They include c.1A>T (Met1?) [48], c.91C>T (Pro31Ser) [41] c.158C>T (Ser53Leu) [2,3], c.271_273delGCTinsA (Ala91fs) [42], c.277C>A (Arg93Ser) [43], c.363C>G (Phe121Leu) [3], c.365A>C (Gln122Pro) [43], c.599T>G (Leu200Arg) [41-43,51,52], c.827T>C (Leu276Pro) [43], c.1000C>T (Arg334Cys) [4], upstream deletions [33,52], and a deletion of the coding region [2,33]. Together with past reported genomic DNA analyses, this brings the number of identified CHST6 mutations in MCD to 124 (Table 4). This vast number underscores the marked allelic heterogeneity in CHST6 that has been previously documented in subjects with MCD in Britain [41], France [42], Iceland [40], India [2-4], Italy [52], Japan [33,47], Saudi Arabia [44], United States [43,46,50], and Vietnam [45,49]. The vast majority of subjects with MCD have missense and nonsense mutations in CHST6 that involve a single nucleotide change that is predicted to alter a conserved amino acid [2-4,33,39-45,47-52]. That these nucleotide changes are disease related is supported by the fact that they have not been detected in analyses of CHST6 in numerous control subjects in different parts of the world [2-4,33,40-45,47,49,52]. In an earlier study of Icelandic individuals, c.383C>T (Ala128Val) was detected in four of 50 healthy controls. However, it was most likely a missense mutation occurring in carriers, because it was associated with MCD when present in a homozygous state [40].

Other MCD-causing mutations are nucleotide insertions or deletions in the coding region of CHST6 that cause frameshift changes and a few deletions or substitutions upstream of CHST6. The latter is a consequence of the marked sequence similarity of nucleotide sequences within the tandem CHST5 and CHST6 genes and the adjacent regions. It predisposes the region to anomalous chromosomal crossovers that lead to deletions and replacements within the involved chromosomes (Figure 2). Such defects in the promoter region were detected in six subjects with MCD in the present study and in 12 previously documented cases [33,52]. Fifty-five of the documented CHST6 mutations have been detected in more than one family with MCD; some of these families may be related to each other as a common ancestor has not been excluded yet.

Our detection of four SNPs in CHST6 that did not affect an amino acid (c.258A>C [Ala86Ala], c.294C>G [Ser98Ser], c.465G>A [Arg155Arg], and c.681C>T [Gly227Gly]) together with other reported SNPs brings the number of identified SNPs in CHST6 to 88. SNPs outside of the coding region could affect gene expression by altering or activating splice sites or by affecting the stability or translational efficiency of the transcript [64], but in individuals in whom we found these SNPs the MCD could be explained by other mutations. Although SNPs in the 3'-UTR have been reported to affect expression [65] and mRNA stability [66], the ones we detected (c.3342A>T and c.3501G>A) were also homozygous in some unaffected family members, indicating that they are not involved in the pathogenesis of MCD.

As illustrated in Figure 3, mutations found in MCD have been in highly conserved parts of carbohydrate 6-O-sulfotransferases, suggesting that the altered amino acids impair the normal functioning of the encoded enzyme and are pathogenic. Seventeen mutations involve the domains necessary for the interaction between the sulfotransferase and the sulfate donor PAPS [35]. Six of these mutations are situated within the 5'-PSB (5'-phosphosulfate-binding) domain that interacts with the 5'-phosphate group of PAPS [67], and 11 mutations involve the 3'-phosphate-binding domain that interacts with the 3'-phosphate of PAPS. There are also four mutations in the hydrophobic site, but none in any of the four glycosylation sites.

All SNPs in the coding region of CHST6 that alter a codon are presumed to be of pathogenic significance because a conserved amino acid is altered. However, at least one SNP (c.484C>G), which is predicted to change arginine to glycine at codon 162, does not seem to alter the encoded sulfotransferase significantly. This SNP has been previously reported in 6.2% of an American control sample [43] and in 9.4% of Italian controls [52]. We have also observed it in 12.0% of an American control sample and in 16.2% of an Icelandic control population (unpublished). Moreover, Abbruzzese et al. [52] suggested that this amino acid change does not affect enzyme activity since it is not conserved in mouse intestinal GlcNAc 6-O-sulfotransferase which has been shown to have the same enzymatic activity as human corneal GlcNAc 6-O-sulfotransferase [39].

Aside from the marked allelic heterogeneity of CHST6 in MCD, it is noteworthy that 10 families (families 6, 9, 12, 13, 15, 22, 23, 26, 28, and 30) with MCD in the present study had two SNPs in the coding region of CHST6 in a single chromosome. In addition to our study, other cis SNP combinations have included c.6G>A; 7C>A (Trp2X and Leu3Met) [4], c.166G>A; 167T>G; 500C>T (Val56Arg and Ser167Phe) [4], c.213G>T; 214C>T (Glu71Asp and Pro72Ser) [41], c.293C>G; 294C>G (Ser98Trp) [3], c.293C>T; 294C>G (Ser98Leu) [4], c.494G>C; 495C>T (Cys165Ser) [4,44], c.668G>A; 993G>T (Gly223Asp and Gln331His) [50], and c.484C>G; 599T>G (Arg162Gly and Leu200Arg) [43,52]. Together, these observations suggest that this part of human chromosome 16 may be prone to mutations that may not be readily corrected by the usual DNA repair mechanisms.

One case of MCD in this study could not be explained by mutations in the coding region of CHST6, by major deletions or insertions in the upstream region, or by splice site mutations which create or destroy signals for exon-intron splicing [68]. Others have also failed to detect mutations on one or both alleles of individuals with MCD [2-4,42,43]. A genetic explanation for these cases of MCD remains unknown, but several theoretical possibilities exist. It is conceivable that the relevant abnormality involves transcription factor binding sites (regulatory promoters) in the region immediately upstream of the minimal promoter that direct the expression of the gene or cis-acting distant genomic elements (enhancers, repressors, or insulators) located upstream or downstream of the transcription unit [69]. In this regard, Hemmerich et al. [36], suggested that the closely proximated CHST5 and CHST6 genes may be regulated by common promoters and/or enhancers, and they drew attention to a conspicuous triplet of binding sites for the zinc-dependent Sp1 transcription factor upstream of CHST6, which has three contiguous Zn (II) fingermotifs believed to be metalloprotein structures that interact with DNA. The Sp1 binding sites are present in the 5'-regulatory sequences of numerous genes that encode carbohydrate-modifying enzymes [70]. In other protein-coding genes, a 5'-promoter element contiguous with the transcription site (the minimal or core promoter) of the gene is necessary to splice the exons needed to produce the mRNA that encodes the appropriate protein [71]. This is presumably also true for CHST6. Other yet to be identified regulatory elements of CHST6 may also play a role, and genes on other chromosomes may affect the gene expression as well. A defect may reside in the 5'-untranslated region (5'-UTR) of CHST6, which has not been fully analyzed, because this part of other genes contributes to the specificity and overall efficiency of translation initiation [72]. Moreover, genetic heterogeneity remains to be excluded.

In their landmark publication, Akama et al. [33] provided evidence that MCD type II might be caused by genetic abnormalities upstream of CHST6. While a defect in the promoter region of CHST6 is expected to control corneal sulfotransferase activity and hence might cause the milder immunophenotype of MCD characterized by AgKS in the serum and cornea, other cases of MCD type II have not only had other mutations in the coding region of CHST6, but have also lacked evidence of a defect upstream of CHST6. A possible molecular explanation for the different immunophenotypes of MCD was provided by an unusual sibship that contained both MCD types I and II [30,50]. In that family, the sibling with the greater deficiency of AgKS (MCD type I) had homozygous c.418C>T (Arg140X) mutations in CHST6, which is predicted to generate a stop-codon and hence a truncated sulfotransferase. The individuals with the milder MCD type II in the sibship were heterozygous for c.418C>T and for cis c.993G>T and c.668G>A nucleotide changes in CHST6 and would presumably still have some residual enzymatic function albeit defective. It is noteworthy that the one family in the present study with a heterozygous stop codon in the coding region of CHST6 (family 4) had MCD type I. However, the molecular basis for the different immunophenotypes is clearly more complex as a molecular genetic study of MCD in Saudi Arabia found identical CHST6 mutations in families with MCD types I, IA, and II [44].

Despite overwhelming evidence implicating CHST6 in the pathogenesis of MCD, an understanding of the molecular events that lead to the characteristic lesions of MCD remain incomplete. The intracellular storage of GAGs may reflect a failure of transport of PGs with insufficiently sulfated lactosaminoglycans from the Golgi apparatus, but the characteristic extracellular deposits in Descemet's membrane remains to be explained. Also, a knockout of CHST6 might be expected to cause MCD, but mice and other nonprimates lack CHST6 and can be viewed as being equivalent to animals in which CHST6 has been deleted by genetic engineering yet they do not manifest MCD. However, the KS in corneas of mice are predominantly undersulfated and express a low level of reactivity with the 5D4 antibody used to differentiate MCD types I and II. The epitope of 5D4 is linear pentasulfated sequences of N-acetyl lactosamine disaccharides of KS PGs in which the GalNAC and Gal are sulfated. Hence, the 5D4 antibody reactivity in the mouse cornea is absent, but low sulfated KS is detectable with antibodies that react with lesser sulfated sequences of N-acetyl lactosamine disaccharides of KS (IB4 and 4D1) [73], and there is evidence that intestinal GlcNAc 6-O-sulfotransferase encoded by CHST5 in the mouse produces some sulfation of KS in the murine cornea [39].

Acknowledgements

We would like to thank the affected and unaffected individuals for participating in this study. Dr. Eugene J. Thonar kindly determined the serum antigenic keratan sulfate levels in 22 of the patients with MCD. This study was supported by a research grant from the National Eye Institute (R01-EY08249).

References

1. Groenouw A. Knötchenförmige Hornhauttrübungen (Noduli corneae). Archiv für Augenheilkunde 1890; 21:281-9.

2. Warren JF, Aldave AJ, Srinivasan M, Thonar EJ, Kumar AB, Cevallos V, Whitcher JP, Margolis TP. Novel mutations in the CHST6 gene associated with macular corneal dystrophy in southern India. Arch Ophthalmol 2003; 121:1608-12.

3. Sultana A, Sridhar MS, Jagannathan A, Balasubramanian D, Kannabiran C, Klintworth GK. Novel mutations of the carbohydrate sulfotransferase-6 (CHST6) gene causing macular corneal dystrophy in India. Mol Vis 2003; 9:730-4 <http://www.molvis.org/molvis/v9/a87/>.

4. Sultana A, Sridhar MS, Klintworth GK, Balasubramanian D, Kannabiran C. Allelic heterogeneity of the carbohydrate sulfotransferase-6 gene in patients with macular corneal dystrophy. Clin Genet 2005; 68:454-60.

5. Klintworth GK, Oshima E, al-Rajhi A, al-Saif A, Thonar EJ, Karcioglu ZA. Macular corneal dystrophy in Saudi Arabia: a study of 56 cases and recognition of a new immunophenotype. Am J Ophthalmol 1997; 124:9-18.

6. Santo RM, Yamaguchi T, Kanai A, Okisaka S, Nakajima A. Clinical and histopathologic features of corneal dystrophies in Japan. Ophthalmology 1995; 102:557-67.

7. Jonasson F, Johannsson JH, Garner A, Rice NS. Macular corneal dystrophy in Iceland. Eye 1989; 3:446-54.

8. Ehlers N, Bramsen T. Central thickness in corneal disorders. Acta Ophthalmol (Copenh) 1978; 56:412-6.

9. Donnenfeld ED, Cohen EJ, Ingraham HJ, Poleski SA, Goldsmith E, Laibson PR. Corneal thinning in macular corneal dystrophy. Am J Ophthalmol 1986; 101:112-3.

10. Quantock AJ, Meek KM, Ridgway AE, Bron AJ, Thonar EJ. Macular corneal dystrophy: reduction in both corneal thickness and collagen interfibrillar spacing. Curr Eye Res 1990; 9:393-8.

11. Bücklers M. Die erblichen Hornhautdystrophien: Dystrophiae corneae hereditariae. Stuttgart, Ferdinand Enke; 1938.

12. Jones ST, Zimmerman LE. Macular dystrophy of the cornea (Groenouw type II); clinicopathologic report of two cases with comments concerning its differential diagnosis from lattice dystrophy (Biber-Haab-Dimmer). Am J Ophthalmol 1959; 47:1-16.

13. Jones ST, Zimmerman LE. Histopathologic differentiation of granular, macular and lattice dystrophies of the cornea. Am J Ophthalmol 1961; 51:394-410.

14. Klintworth GK, Meyer R, Dennis R, Hewitt AT, Stock EL, Lenz ME, Hassell JR, Stark WJ Jr, Kuettner KE, Thonar EJ. Macular corneal dystrophy. Lack of keratan sulfate in serum and cornea. Ophthalmic Paediatr Genet 1986; 7:139-43.

15. Francois J. Heredo-familial corneal dystrophies. Trans Ophthalmol Soc U K 1966; 86:367-416.

16. Klintworth GK, Vogel FS. Macular corneal dystrophy. An inherited acid mucopolysaccharide storage disease of the corneal fibroblast. Am J Pathol 1964; 45:565-86.

17. Funderburgh JL. Keratan sulfate: structure, biosynthesis, and function. Glycobiology 2000; 10:951-8.

18. Klintworth GK, Hawkins HK, Smith CF. Acridine orange particles in cultured fibroblasts. A comparative study of macular corneal dystrophy, systemic mucopolysaccharidoses types I-H and II, and normal controls. Arch Pathol Lab Med 1979; 103:297-9.

19. Klintworth GK, Smith CF. Macular corneal dystrophy. Studies of sulfated glycosaminoglycans in corneal explant and confluent stromal cell cultures. Am J Pathol 1977; 89:167-82.

20. Hassell JR, Newsome DA, Krachmer JH, Rodrigues MM. Macular corneal dystrophy: failure to synthesize a mature keratan sulfate proteoglycan. Proc Natl Acad Sci U S A 1980; 77:3705-9.

21. Nakazawa K, Hassell JR, Hascall VC, Lohmander LS, Newsome DA, Krachmer J. Defective processing of keratan sulfate in macular corneal dystrophy. J Biol Chem 1984; 259:13751-7.

22. Yang CJ, SundarRaj N, Thonar EJ, Klintworth GK. Immunohistochemical evidence of heterogeneity in macular corneal dystrophy. Am J Ophthalmol 1988; 106:65-71.

23. Jonasson F, Oshima E, Thonar EJ, Smith CF, Johannsson JH, Klintworth GK. Macular corneal dystrophy in Iceland. A clinical, genealogic, and immunohistochemical study of 28 patients. Ophthalmology 1996; 103:1111-7.

24. Edward DP, Yue BY, Sugar J, Thonar EJ, SunderRaj N, Stock EL, Tso MO. Heterogeneity in macular corneal dystrophy. Arch Ophthalmol 1988; 106:1579-83.

25. Midura RJ, Hascall VC, MacCallum DK, Meyer RF, Thonar EJ, Hassell JR, Smith CF, Klintworth GK. Proteoglycan biosynthesis by human corneas from patients with types 1 and 2 macular corneal dystrophy. J Biol Chem 1990; 265:15947-55.

26. Plaas AH, West LA, Thonar EJ, Karcioglu ZA, Smith CJ, Klintworth GK, Hascall VC. Altered fine structures of corneal and skeletal keratan sulfate and chondroitin/dermatan sulfate in macular corneal dystrophy. J Biol Chem 2001; 276:39788-96.

27. Hassell JR, Klintworth GK. Serum sulfotransferase levels in patients with macular corneal dystrophy type I. Arch Ophthalmol 1997; 115:1419-21.

28. Edward DP, Thonar EJ, Srinivasan M, Yue BJ, Tso MO. Macular dystrophy of the cornea. A systemic disorder of keratan sulfate metabolism. Ophthalmology 1990; 97:1194-200.

29. Vance JM, Jonasson F, Lennon F, Sarrica J, Damji KF, Stauffer J, Pericak-Vance MA, Klintworth GK. Linkage of a gene for macular corneal dystrophy to chromosome 16. Am J Hum Genet 1996; 58:757-62.

30. Liu NP, Baldwin J, Lennon F, Stajich JM, Thonar EJ, Pericak-Vance MA, Klintworth GK, Vance JM. Coexistence of macular corneal dystrophy types I and II in a single sibship. Br J Ophthalmol 1998; 82:241-4.

31. Liu NP, Baldwin J, Jonasson F, Dew-Knight S, Stajich JM, Lennon F, Pericak-Vance MA, Klintworth GK, Vance JM. Haplotype analysis in Icelandic families defines a minimal interval for the macular corneal dystrophy type I gene. Am J Hum Genet 1998; 63:912-7.

32. Liu NP, Dew-Knight S, Jonasson F, Gilbert JR, Klintworth GK, Vance JM. Physical and genetic mapping of the macular corneal dystrophy locus on chromosome 16q and exclusion of TAT and LCAT as candidate genes. Mol Vis 2000; 6:95-100 <http://www.molvis.org/molvis/v6/a13/>.

33. Akama TO, Nishida K, Nakayama J, Watanabe H, Ozaki K, Nakamura T, Dota A, Kawasaki S, Inoue Y, Maeda N, Yamamoto S, Fujiwara T, Thonar EJ, Shimomura Y, Kinoshita S, Tanigami A, Fukuda MN. Macular corneal dystrophy type I and type II are caused by distinct mutations in a new sulphotransferase gene. Nat Genet 2000; 26:237-41.

34. Hemmerich S, Rosen SD. Carbohydrate sulfotransferases in lymphocyte homing. Glycobiology 2000; 10:849-56.

35. Kakuta Y, Pedersen LG, Pedersen LC, Negishi M. Conserved structural motifs in the sulfotransferase family. Trends Biochem Sci 1998; 23:129-30.

36. Hemmerich S, Lee JK, Bhakta S, Bistrup A, Ruddle NR, Rosen SD. Chromosomal localization and genomic organization for the galactose/N-acetylgalactosamine/N-acetylglucosamine 6-O-sulfotransferase gene family. Glycobiology 2001; 11:75-87.

37. Fukuta M, Inazawa J, Torii T, Tsuzuki K, Shimada E, Habuchi O. Molecular cloning and characterization of human keratan sulfate Gal-6-sulfotransferase. J Biol Chem 1997; 272:32321-8.

38. Lee JK, Bhakta S, Rosen SD, Hemmerich S. Cloning and characterization of a mammalian N-acetylglucosamine-6-sulfotransferase that is highly restricted to intestinal tissue. Biochem Biophys Res Commun 1999; 263:543-9.

39. Akama TO, Nakayama J, Nishida K, Hiraoka N, Suzuki M, McAuliffe J, Hindsgaul O, Fukuda M, Fukuda MN. Human corneal GlcNac 6-O-sulfotransferase and mouse intestinal GlcNac 6-O-sulfotransferase both produce keratan sulfate. J Biol Chem 2001; 276:16271-8.

40. Liu NP, Dew-Knight S, Rayner M, Jonasson F, Akama TO, Fukuda MN, Bao W, Gilbert JR, Vance JM, Klintworth GK. Mutations in corneal carbohydrate sulfotransferase 6 gene (CHST6) cause macular corneal dystrophy in Iceland. Mol Vis 2000; 6:261-4 <http://www.molvis.org/molvis/v6/a35/>.

41. El-Ashry MF, El-Aziz MM, Wilkins S, Cheetham ME, Wilkie SE, Hardcastle AJ, Halford S, Bayoumi AY, Ficker LA, Tuft S, Bhattacharya SS, Ebenezer ND. Identification of novel mutations in the carbohydrate sulfotransferase gene (CHST6) causing macular corneal dystrophy. Invest Ophthalmol Vis Sci 2002; 43:377-82.

42. Niel F, Ellies P, Dighiero P, Soria J, Sabbagh C, San C, Renard G, Delpech M, Valleix S. Truncating mutations in the carbohydrate sulfotransferase 6 gene (CHST6) result in macular corneal dystrophy. Invest Ophthalmol Vis Sci 2003; 44:2949-53.

43. Aldave AJ, Yellore VS, Thonar EJ, Udar N, Warren JF, Yoon MK, Cohen EJ, Rapuano CJ, Laibson PR, Margolis TP, Small K. Novel mutations in the carbohydrate sulfotransferase gene (CHST6) in American patients with macular corneal dystrophy. Am J Ophthalmol 2004; 137:465-73.

44. Bao W, Smith CF, al-Rajhi A, Chandler JW, Karcioglu ZA, Akama TO, Fukuda MN, Klintworth, GK. Novel mutations in the CHST6 gene in Saudi Arabic patients with macular corneal dystrophy. ARVO Annual Meeting; 2001 April 29-May 4; Fort Lauderdale, FL.

45. Ha NT, Chau HM, Cung le X, Thanh TK, Fujiki K, Murakami A, Hiratsuka Y, Hasegawa N, Kanai A. Identification of novel mutations of the CHST6 gene in Vietnamese families affected with macular corneal dystrophy in two generations. Cornea 2003; 22:508-11.

46. Klintworth GK. The molecular genetics of the corneal dystrophies--current status. Front Biosci 2003; 8:d687-713.

47. Iida-Hasegawa N, Furuhata A, Hayatsu H, Murakami A, Fujiki K, Nakayasu K, Kanai A. Mutations in the CHST6 gene in patients with macular corneal dystrophy: immunohistochemical evidence of heterogeneity. Invest Ophthalmol Vis Sci 2003; 44:3272-7.

48. Gruenauer-Kloevekorn C, Braeutigam S, Froster U, Duncker GIW. Molecular genetic findings and therapeutical options in a well examined german family with macular corneal dystrophy. ARVO Annual Meeting; 2005 May 1-5; Fort Lauderdale (FL).

49. Ha NT, Chau HM, Cung le X, Thanh TK, Fujiki K, Murakami A, Hiratsuka Y, Kanai A. Mutation analysis of the carbohydrate sulfotransferase gene in Vietnamese with macular corneal dystrophy. Invest Ophthalmol Vis Sci 2003; 44:3310-6.

50. Liu NP, Bao W, Smith CF, Vance JM, Klintworth GK. Different mutations in carbohydrate sulfotransferase 6 (CHST6) gene cause macular corneal dystrophy types I and II in a single sibship. Am J Ophthalmol 2005; 139:1118-20.

51. El-Ashry MF, Abd El-Aziz MM, Shalaby O, Wilkins S, Poopalasundaram S, Cheetham M, Tuft SJ, Hardcastle AJ, Bhattacharya SS, Ebenezer ND. Novel CHST6 nonsense and missense mutations responsible for macular corneal dystrophy. Am J Ophthalmol 2005; 139:192-3.

52. Abbruzzese C, Kuhn U, Molina F, Rama P, De Luca M. Novel mutations in the CHST6 gene causing macular corneal dystrophy. Clin Genet 2004; 65:120-5.

53. Klintworth GK, Smith CF. Abnormal product of corneal explants from patients with macular corneal dystrophy. Am J Pathol 1980; 101:143-58.

54. Klintworth GK. Research into the pathogenesis of macular corneal dystrophy. Trans Ophthalmol Soc U K 1980; 100:186-94.

55. Klintworth GK, Reed J, Stainer GA, Binder PS. Recurrence of macular corneal dystrophy within grafts. Am J Ophthalmol 1983; 95:60-72.

56. Klintworth GK, Smith CF. Abnormalities of proteoglycans and glycoproteins synthesized by corneal organ cultures derived from patients with macular corneal dystrophy. Lab Invest 1983; 48:603-12.

57. Thonar EJ, Meyer RF, Dennis RF, Lenz ME, Maldonado B, Hassell JR, Hewitt AT, Stark WJ Jr, Stock EL, Kuettner KE, Klintworth GK. Absence of normal keratan sulfate in the blood of patients with macular corneal dystrophy. Am J Ophthalmol 1986; 102:561-9.

58. Dean FB, Nelson JR, Giesler TL, Lasken RS. Rapid amplification of plasmid and phage DNA using Phi 29 DNA polymerase and multiply-primed rolling circle amplification. Genome Res 2001; 11:1095-9.

59. Nelson JR, Cai YC, Giesler TL, Farchaus JW, Sundaram ST, Ortiz-Rivera M, Hosta LP, Hewitt PL, Mamone JA, Palaniappan C, Fuller CW. TempliPhi, phi29 DNA polymerase based rolling circle amplification of templates for DNA sequencing. Biotechniques 2002; Suppl:44-7.

60. Lizardi PM, Huang X, Zhu Z, Bray-Ward P, Thomas DC, Ward DC. Mutation detection and single-molecule counting using isothermal rolling-circle amplification. Nat Genet 1998; 19:225-32.

61. Esteban JA, Salas M, Blanco L. Fidelity of phi 29 DNA polymerase. Comparison between protein-primed initiation and DNA polymerization. J Biol Chem 1993; 268:2719-26.

62. Marchler-Bauer A, Anderson JB, DeWeese-Scott C, Fedorova ND, Geer LY, He S, Hurwitz DI, Jackson JD, Jacobs AR, Lanczycki CJ, Liebert CA, Liu C, Madej T, Marchler GH, Mazumder R, Nikolskaya AN, Panchenko AR, Rao BS, Shoemaker BA, Simonyan V, Song JS, Thiessen PA, Vasudevan S, Wang Y, Yamashita RA, Yin JJ, Bryant SH. CDD: a curated Entrez database of conserved domain alignments. Nucleic Acids Res 2003; 31:383-7.

63. Oleykowski CA, Bronson Mullins CR, Godwin AK, Yeung AT. Mutation detection using a novel plant endonuclease. Nucleic Acids Res 1998; 26:4597-602.

64. Day DA, Tuite MF. Post-transcriptional gene regulatory mechanisms in eukaryotes: an overview. J Endocrinol 1998; 157:361-71.

65. Goto Y, Yue L, Yokoi A, Nishimura R, Uehara T, Koizumi S, Saikawa Y. A novel single-nucleotide polymorphism in the 3'-untranslated region of the human dihydrofolate reductase gene with enhanced expression. Clin Cancer Res 2001; 7:1952-6.

66. Frittitta L, Ercolino T, Bozzali M, Argiolas A, Graci S, Santagati MG, Spampinato D, Di Paola R, Cisternino C, Tassi V, Vigneri R, Pizzuti A, Trischitta V. A cluster of three single nucleotide polymorphisms in the 3'-untranslated region of human glycoprotein PC-1 gene stabilizes PC-1 mRNA and is associated with increased PC-1 protein content and insulin resistance-related abnormalities. Diabetes 2001; 50:1952-5.

67. Negishi M, Pedersen LG, Petrotchenko E, Shevtsov S, Gorokhov A, Kakuta Y, Pedersen LC. Structure and function of sulfotransferases. Arch Biochem Biophys 2001; 390:149-57.

68. Cooper TA, Mattox W. The regulation of splice-site selection, and its role in human disease. Am J Hum Genet 1997; 61:259-66.

69. Kleinjan DA, van Heyningen V. Long-range control of gene expression: emerging mechanisms and disruption in disease. Am J Hum Genet 2005; 76:8-32.

70. Kadonaga JT, Carner KR, Masiarz FR, Tjian R. Isolation of cDNA encoding transcription factor Sp1 and functional analysis of the DNA binding domain. Cell 1987; 51:1079-90.

71. Levine M, Tjian R. Transcription regulation and animal diversity. Nature 2003; 424:147-51.

72. Gray NK, Wickens M. Control of translation initiation in animals. Annu Rev Cell Dev Biol 1998; 14:399-458.

73. Young RD, Tudor D, Hayes AJ, Kerr B, Hayashida Y, Nishida K, Meek KM, Caterson B, Quantock AJ. Atypical composition and ultrastructure of proteoglycans in the mouse corneal stroma. Invest Ophthalmol Vis Sci 2005; 46:1973-8.

74. Antonarakis SE. Recommendations for a nomenclature system for human gene mutations. Nomenclature Working Group. Hum Mutat 1998; 11:1-3.