Abstract
purpose. To develop and apply microarray-based resequencing technology to detect sequence alterations in multiple autosomal recessive retinal disease genes on a single high-throughput platform.
methods. Oligonucleotides corresponding to both strands of the target exons and the flanking intron sequences of 29,214 bp from 11 genes associated with autosomal recessive retinitis pigmentosa (arRP) were tiled on 20 × 25-μm microarrays (arRP-I arrays). A total of 155 exons were amplified from 35 arRP patient DNA samples, with each sample being sequenced on an arRP-I chip by hybridization.
results. With the arRP-I arrays, 97.6% of the tiled sequence were determined with more than 99% accuracy and reproducibility. Of the 2.4% unread sequence, 89.5% involved stretches of G or C. In analyzing the 903,140-bp sequence from the 35 patient samples, 506 sequence changes have been detected in which 386 are previously reported alterations, and 120 are novel. In addition to four known causative mutations, six novel sequence changes that are potentially pathogenic were observed. Additional analysis is needed to determine whether these changes are responsible for arRP in these patients.
conclusions. The use of microarray for sequencing is a novel approach, and the arRP-I chip is the first successful application of this technology for determining sequence alteration in multiple disease-related genes. These arrays can be used for high-throughput genotyping of patients with relevant retinal conditions. In addition, these arrays offer a unique opportunity to interrogate complex patterns of inheritance due to the involvement of more than one gene by screening multiple genes on a single platform.
Retinitis pigmentosa (RP) is a heterogeneous group of diseases characterized by progressive retinal degeneration leading to irreversible blindness or severe visual disability. It affects 1 in 3500 individuals worldwide. Clinically, it is characterized by an abnormal electroretinogram (ERG), loss of peripheral vision, and night blindness in the initial stage, which progresses to severe visual disability that can progress to complete vision loss.
1 RP can be inherited in an autosomal dominant or recessive or X-linked manner, as well as in digenic, mitochondrial, and simplex modes. Most patients with RP are simplex with no known affected relatives, although some of these may have autosomal or X-linked recessive RP or dominant RP with incomplete penetrance. Despite the nearly indistinguishable clinical characteristics of RP with different modes of inheritance, a broad genetic heterogeneity exists. At least 15 loci for autosomal dominant RP, 18 loci for autosomal recessive RP (arRP), and 5 loci for the X-linked form are known, and additional loci are yet to be mapped.
2 3 4 Phenotypic symptoms of RP can also overlap with various other retinal disease phenotypes.
1 5 In addition, several syndromes involving RP have been described (OMIM; Online Mendelian Inheritance in Man; http://www.ncbi.nlm.nih.gov/Omim/ provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). Clinical diagnosis is further complicated by the significant variation in the age of onset, severity, clinical presentation, and penetrance that is reported within and between families with retinal degenerations.
6 7 8 9 10 All these issues pose a challenge in providing a definite diagnosis to patients, particularly to those who have no clear family history.
Genes for approximately 110 retinal diseases have been cloned, and significant progress has been made in understanding the mechanism of these degenerations.
4 Various strategies to treat these conditions are being actively pursued. Results of studies of treatments for two different forms of early-onset retinal degeneration have been reported to be successful in animal models, and one can anticipate that some of these treatments may become available to patients in the future.
11 12 However, most of these treatments are specific to patients with particular genotypes, and thus, determination of the genotype of patients will become essential, for identifying individuals suitable for the available treatments.
The genetic heterogeneity, variable penetrance, and considerable task of analyzing all the known candidate genes temper one’s enthusiasm for genotyping patients with RP. So far, at least 32 genes have been identified for various forms of RP, and many more remain to be cloned. The sequence of all the known arRP genes alone adds up to more than 60 kilobases (kb). Current strategies for screening large numbers of bases for mutations are limited by the availability of high-throughput, sensitive, and rapid mutation-detection technologies.
Sequencing using microarrays, which is more typically referred to as resequencing, offers the promise of determining genotypes with a reasonable amount of effort. The past decade has seen the development of sequencing by hybridization to oligonucleotides on a microarray chip.
13 14 15 16 17 18 19 This technology has also been commercialized.
20 Recently, significant improvements have been made in this technology, and array platforms that can accommodate a larger number of bases have been developed.
21 Using this technology, Maitra et al.
22 have developed MitoChip for detection of human mitochondrial mutations, and Wong et al.
23 used this methodology to screen for mutations in the SARS (severe acute respiratory syndrome) virus. In both these cases, it was a haploid genome that was sequenced. Ours is the first report of the application of these resequencing platforms to analyze a diploid genome for mutations in multiple genes. We have screened 11 arRP genes for sequence changes in the coding regions and the flanking splice sites. Herein, we present the validation and potential application of these arRP-I chips to genotype patients with arRP.
The concentration of amplicons was determined by adding a fluorescent label (Picogreen; Molecular Probes, Eugene, OR) to the samples and measuring fluorescence (Spectro Fluor Plus; Tecan Austria Gesellschaft MBH, Salzburg, Austria), as described in the manufacturer’s protocol (Molecular Probes). Equimolar amounts of the 159 PCR products from each individual were pooled to ensure equivalent hybridization. Pooled samples were purified to remove the residual primers and nucleotides using a clean-up kit (QIA Quick PCR clean-up kit; Qiagen Inc., Valencia, CA).
Fragmentation of pooled samples was performed using fragmentation reagents (0.2 U DNAseI/μg DNA, from the GeneChip Resequencing Assay Kit; Affymetrix). A parallel fragmentation reaction was performed for each sample and analyzed on a 4% to 20% gradient acrylamide-TBE (Tris-boric acid-EDTA) gel followed by staining with a nucleic acid stain (SybrGold [1:50,000 in Tris-EDTA]; Molecular Probes) to evaluate the extent of fragmentation. Completely fragmented products were seen to run with 20- to 100-bp double-stranded (ds)DNA. Fragmented DNA was end labeled with biotin (GeneChip Reagent Kit; Affymetrix). Control plasmid DNA was fragmented and labeled separately.
Prehybridization, hybridization, washing, and scanning of the arRP-I chip were performed following the manufacturer’s protocol (GeneChip CustomSeq Resequencing Array Protocol, ver. 2; Affymetrix). A labeled plasmid control probe and gridding controls (Affymetrix) were included in the hybridization cocktail. Chip washing and scanning were then performed using manufacturer’s software (GeneChip Operating Software [GCOS], ver. 2.0; Affymetrix).
The 11 genes we selected carry at least 180 previously described mutations associated with early-onset retinal degenerations (HGMD, Human Gene Mutation Database; http://archive.uwcm.ac.uk/uwcm/mg/hgmd0.html/ provided in the public domain by the Institute of Medical Genetics, University of Wales College of Medicine, Cardiff, Wales, UK). These genes were arrayed on the 30-kb resequencing platform available through Affymetrix
(Table 2) . The array-based sequencing depends on differential hybridization of genomic fragments to short perfect-match and mismatch oligonucleotide probes. Exons with homologous sequences cross-hybridize on the chips and consequently are not suitable for sequencing with this technology. Sequences containing repeat regions are also not suitable. Analysis of the sequence on RepeatMasker revealed that the sequence of the coding region of
PDE6A and
PDE6B are highly homologous, except for exon 7. Thus, we selected the sequences of all exons of the gene
PDE6B, because it contains 20 known recessive RP mutations, whereas
PDE6A carries only three mutations, one each in exons 7, 13, and 14. Only the nonhomologous exon 7 of the
PDE6A gene was included on the chips. Exons 4 and 5 of
CNGA1 and exon 5 of
TULP1 contain repeat sequences and were not included. All the other exons in the 11 genes were found to be suitable for sequencing with the DNA sequencing arrays.
Fifteen-base pair flanking intron sequences on each side of each exon were included on the chips. Because the first and final 11 bp of each fragment cannot be read, we were able to analyze only 4 bp of the flanking intronic sequence from each splice junction. The total number of bases of double-stranded arRP gene sequence arrayed on the chips is 29,214. However, because 11 bp from the beginning and the end of each fragment tiled on arrays cannot be read, a total of 25,804 bp sequence of arRP genes, which includes sequence of all exons and 4-bp flanking intronic sequence, can be analyzed on each arRP-I chip. As a positive control, an 814-bp sequence from the plasmid TAG IQ-EX (Affymetrix) was tiled onto the arrays. The arRP-I chips were designed with 20 × 25-μm standard array features.
The arRP-I chip contains a total of 300,356 features arrayed on a 30-kb platform (1.28 × 1.28 cm). Each feature contains multiple copies of a specific 25-mer-oligonucleotide probe. Each nucleotide to be queried on the array platform is located at the 13th position of a 25-mer probe. At each base position, the gene chip tests for the wild-type nucleotide, for substitutions with the other three bases and for a single base deletion. The probes are tiled both for sense and antisense strands of the selected sequence. Therefore, a total of eight probes are tiled for each nucleotide position in addition to two ‘del’ cells corresponding to a single-base deletion on sense and antisense strands. During scanning, the software divides each probe cell into subunit squares or pixels (3 × 3 μm) and 56 raw pixel points are generated from each cell, of which 30 are used for base calling. The 26 border pixels are typically masked, and the mean and standard deviation of 30 observations are used for a base call. The differences in hybridization signal intensities between the sequences that bind to the specific probes and those that bind to the corresponding mismatched probes calculated using ABACUS determine the base at each sequence position
(Fig. 1) . Having 10 features of eight sets of probes for a single-base sequence, makes it possible to genotype ∼30 kb of DNA with more than 99% accuracy.
32
Genomic DNAs from 35 unrelated patients with recessive RP were amplified using primers described in the Methods section. The amplified products were sequenced on the custom designed arRP-I chips. The total arRP gene sequence that can be read on the chip is 903,140 bp (25,804 × 35). However, in our analysis of these 35 patients, amplification products of exons of some samples corresponding to 4,490 bp were not included in the hybridization cocktail. Hence, a total of 898,650 bp of dsDNA sequence from 35 patients was analyzed on arRP-I chips. The exons not included in the hybridization mix were sequenced by dideoxy sequencing (model 3100 sequencer; ABI) to complete the analysis for each patient. In total the DNA analysis software (GeneChip; Affymetrix) assigned 877,399 base calls out of the 898,650 bp analyzed, which ranged from 96% to 98.5% of the total on different chips. The called base sequence derived using these arrays corresponded to the reference sequence tiled on the chip with more than 99% accuracy. These results were obtained when the chips were used within 7 months of manufacturing; the actual shelf-life of these chips is yet to be determined.
In the 903,140 bp of sequence analyzed from the 35 samples, we identified a total of 506 sequence changes, of which 382 corresponded to previously reported single nucleotide polymorphisms (SNPs) and four changes represented previously reported mutations
(Tables 2 3) . An additional 113 changes corresponded to polymorphisms not previously reported, and the remaining seven nucleotide substitutions found were novel alterations
(Table 3) .
Analysis of Different Types of Known Sequence Changes.
Hybridization in the Absence of a Specific Sequence in the Sample.
Comparison with Sequence Generated Using the Dideoxy Chain Termination Method.
Microarray sequencing technology is a promising new diagnostic tool. Analysis of the arRP-I chips indicates that these sequencing arrays offer a rapid, cost-effective method for genotyping patients with the genetically heterogeneous disease arRP.
The arRP-I chips we designed produced 97.6% of the sequence analyzed with >99% accuracy and reproducibility. Similar resequencing arrays for the SARS coronavirus genome and the human mitochondrial genome have been described recently.
22 23 The sequence obtained using the SARS-CoV arrays was reported to be more than 99.99% accurate and reproducible,
23 and the MitoChip produced sequence with 96% accuracy and more than 99.99% reproducibility. Our arRP-I gene chips detected sequence changes involving any combination of nucleotide exchange in the homozygous and heterozygous state
(Table 4) . Overall performance of the custom-designed arRP-I sequencing arrays was comparable to the two previously reported custom-designed sequencing chips.
22 23
The sequencing arrays are most appropriately used for patients with phenotypes that have been previously associated with the arrayed genes. Because the genes we selected for array can cause a spectrum of disease phenotypes, one may consider screening patients beyond the previously recognized phenotypes. These chips offer a fast and efficient method of screening for mutations in most recessive RP genes. Single large genes arrayed on these chips, such as
ABCA4, can also be analyzed for mutations in patients with other relevant phenotypes. Chips containing previously described mutations in the
ABCA4 gene have been designed and successfully used to detect these mutations in patients with Stargardt’s macular degeneration, cone–rod dystrophy, and arRP.
34 36 As observed in the patient samples analyzed using the arRP-I sequencing arrays, novel sequence alterations are likely to exist in patients with retinal conditions
(Table 4) . In contrast to the previously reported ABCA4 chips, the arRP-I chips can detect both novel and previously known single-nucleotide changes in the sequence.
36 Therefore, failure to detect mutations in a single gene or a group of genes using the arRP-I sequencing arrays will suggest exclusion of those genes for carrying mutations in the sequence tiled.
The arRP-I sequencing arrays offer for the first time an opportunity to screen simultaneously on a single platform for sequence alterations in multiple genes. This method of sequencing will allow us to address important questions about both simple Mendelian and complex forms of retinal degenerations involving sequence changes in more than one gene. A broad phenotypic variation within and between families with the same genotype has been described in several retinal diseases.
6 7 In addition, a digenic pattern of inheritance of RP,
37 and trialleleic involvement in Bardet-Biedl syndrome have been reported.
38 Our arRP-I chips may serve as valuable tools to evaluate the genetic basis of such complex phenotypes and to determine gene interactions.
Analysis of samples using the arRP-I sequencing arrays resulted in the identification of an average of 15 sequence changes in each sample. Some of the sequence changes detected are likely to be mutations associated with the phenotype. In addition, the disease-causing mutation may lie outside the putative coding region and the flanking regions tiled on the chips. A perceived problem of multigene sequencing is the interpretation of the results, as there is much intrinsic variation in the genome. The data generated by these high-throughput methodologies, in the long run, could be a valuable resource to study the phenotype–genotype associations. In addition, the polymorphisms detected may play a role as modifiers, either by directly influencing the function of the gene or by interacting with the environment. Careful evaluation of these sequence changes will provide an opportunity to ask questions about gene interactions and the role of these sequence alterations as modifiers of phenotype.
Generation of sequences by using the DNA array technology is simple, and the methodology is easy to establish. However, as is true of other sequencing methods, interpretation of results is time consuming. Although the data are generated in FASTA format, they do not directly interface with the sequence data available in public domains. Therefore, the interpretation of each sequence change must be performed manually. Some sequence stretches are not suitable for arrays, and these must be screened for mutations by an alternate method. Heterozygous deletions and insertions cannot be detected using resequencing arrays as well as changes in the other genomic regions (intergenic or intronic), which are not tiled on the chip. Despite these limitations, these arRP-I resequencing arrays offer an efficient and reliable method of high-throughput screening for mutations in genetically heterogeneous diseases such as arRP.
Comparison of the cost of sequencing using the arRP-I chips with the dideoxy sequencing revealed that the material cost of sequencing by arRP-I chips is 23.5% cheaper. However, resequencing using the arRP-I chips was found to be highly cost-effective when the labor and time required for the analysis were compared. With the arRP-I arrays, all 155 exons can be analyzed in a single assay, cutting down the time required for the analysis significantly.
Currently, the size of the platforms that are available limits our ability to array all known candidate genes for the arRP phenotype on a single matrix. However, larger platforms (300-kb) are likely to be available in the near future, which will solve this problem. Sequencing arrays are likely to play a major role in genotyping and molecular characterization of patients with hereditary diseases.
Supported by National Eye Institute Grant EY13198 (RA); the Foundation Fighting Blindness (RA, JRH); Research to Prevent Blindness Inc. (RA); National Eye Institute Core Grants EY07003 and EY07060 to the University of Michigan, Department of Ophthalmology and Visual Sciences; the Foundation Fighting Blindness, Canada; and the Fonds de la Recherche en Santé Quebec (FRSQ).
Submitted for publication January 4, 2005; revised March 16 and April 6, 2005; accepted April 14, 2005.
Disclosure:
M.N.A. Mandal, None;
J.R. Heckenlively, None;
T. Burch, None;
L. Chen, None;
V. Vasireddy, None;
R.K. Koenekoop, None;
P.A. Sieving, None;
R. Ayyagari, 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.
Corresponding author: Radha Ayyagari, Department of Ophthalmology and Visual Sciences, University of Michigan, 1000 Wall Street, Room 325, Ann Arbor, MI 48105;
[email protected].
Table 4. List of Known Sequence Changes Tested on the arRP-I Sequencing Arrays as the Positive Control
Table 4. List of Known Sequence Changes Tested on the arRP-I Sequencing Arrays as the Positive Control
| Gene | Nucleotide Change | Amino Acid Change |
| Sequence changes tested on the first set of chips | | |
| ABCA4 | C52T | R18W |
| A286C | N96H |
| G514A | G172S |
| G768T | V256V |
| G873C | P291P |
| T1622C | L541P |
| G1715A | R572Q |
| G1927A | V643M |
| G2552A | G851D |
| G2588C | G863A |
| G2828A | A943Q |
| T2966C | V989A |
| C3056T | T1019M |
| C3113T | A1038V |
| C3292T | R1098C |
| C3322T | R1108C |
| G3386T | R1129L |
| C389T | R1300X |
| C4139T | P1380L |
| C4203A | P1401P |
| G4463A | C1488Y |
| C4577T | T1526M |
| G4919A | R1640Q |
| C5381A | A1794D |
| A5844G | P1948P |
| C5908T | L1970F |
| C6079T | L2027F |
| T6285C | D2095D |
| G46320A | R2107H |
| G6529A | D2177N |
| Sequence changes tested on the second set of chips | | |
| ABCA4 | C1269T | H423H |
| C2041T | R681X |
| C3758T | T1253M |
| G5682C | L1894L |
| A5814G | L1938L |
| A5844G | P1948P |
| G5882A | G1961E |
| T6285C | D2095D |
| Rhodopsin | C173G | T58R |
| RPE65 | T1156C | Y368H |
| Sequence change included as a positive control along withtest samples | | |
| ABCA4 | C3758T | T1253M |
Table 1. arRP Genes Arrayed on the Chips and the Number of Missense Mutations Previously Reported in These Genes in Patients with arRP, LCA or CSNB
Table 1. arRP Genes Arrayed on the Chips and the Number of Missense Mutations Previously Reported in These Genes in Patients with arRP, LCA or CSNB
| Genes | Additional Phenotype | Number of Exons | Exons Tiled | Number of arRP Mutations Reported | Number of LCA/CSNB Mutations Reported |
| ABCA4 | | 50 | 50 | 9 | |
| CNGA1 | | 8 | 6 | 4 | |
| CRB1 | LCA, AD | 11 | 11 | 15 | 23 |
| MERTK | | 19 | 19 | 3 | |
| PDE6A | | 22 | 1 (exon 7) | 1 | |
| PDE6B | CSNB, AD | 22 | 22 | 21 | 23 |
| RGR | | 7 | 7 | 7 | |
| RHO | | 5 | 5 | 4 | |
| RLBP1 | | 7 | 7 | 3 | |
| RPE65 | LCA, AR | 14 | 14 | 15 | 47 |
| TULP1 | | 14 | 13 | 5 | |
| Totals | | 179 | 155 | 87 | 93 |
Table 2. Summary of Design of the arRP-I Chips and Results of Analysis
Table 2. Summary of Design of the arRP-I Chips and Results of Analysis
| Design of arrays | |
| Size of an array | 1.28 × 1.28 cm |
| Total number of features/cells in an array | 300,356 |
| Size of a feature/cell | 20 × 25 micron |
| Total double stranded arRP genes sequence arrayed per chip | 29,214 |
| Number of base pairs of arRP genes sequence that can be analyzed per chip | 25,804 |
| Number of base pairs of control plasmid DNA sequenced per chip | 814 |
| Sample analysis | |
| Total samples analyzed | 35 |
| Total base pairs of DNA sequenced on chip | 898,650 |
| Total arRP gene sequence assigned by GDAS | 877,399 |
| Total percentage bases called (%) | 97.60% |
| Number of previously reported SNPs detected | 382 |
| Number of new polymorphisms detected | 113 |
| Number of previously reported mutations detected | 4 |
| Number of novel and potentially pathogenic changes detected | 7 |
| Replicate experiments | |
| First set | |
| Total number of chips used for the analysis | 2 |
| Number of known mutations tested (listed in table 4 ) | 30 |
| Number of mutations detected on chip 1 | 30 |
| Number of mutations detected on chip 2 | 30 |
| Second set | |
| Total chips analyzed for replicate experiments | 3 |
| Total number of base pairs analyzed | 10,069 |
| Base pairs assigned on chip 1 | 9,816 (97.5%) |
| Base pairs assigned on chip 2 | 9,741 (96.7%) |
| Base pairs assigned on chip 3 | 9,845 (97.8%) |
| Number of known sequence changes tested in this set (listed in table 4 ) | 10 |
| Number of sequence changes detected on chip 1 | 10 |
| Number of sequence changes detected on chip 2 | 10 |
| Number of sequence changes detected on chip 3 | 10 |
Table 3. Summary of Novel Potentially Pathogenic Nucleotide Changes and Previously Reported Mutations Detected in Patient DNA, with the arRP-I Chips
Table 3. Summary of Novel Potentially Pathogenic Nucleotide Changes and Previously Reported Mutations Detected in Patient DNA, with the arRP-I Chips
| Patient | Gene | Nucleotide Change | Amino Acid Change | Genotype | Reference |
| Novel changes | | | | | |
| KE727 | RHO | C959A | Thr320Asn | Hetero | |
| KE1246 | CRB1 | G2473A | Glu825Lys | Hetero | |
| R165 | TULP1 | IVS2 + 3, A > G | | Homo | |
| R206 | ABCA4 | G1699A | Val567Met | Hetero | |
| KE869 | RGR | C734T | Ser245Phe | Hetero | |
| R353 | MERTK | G500A | Arg167His | Hetero | |
| R376 | ABCA4 | IVS23-2, A > T | | Hetero | |
| Previously reported mutations | | | | | |
| KE385 | RPE65 | T963G | Asn321Lys | Hetero | 28 |
| KE1246 | ABCA4 | T3602G | Leu1201Arg | Hetero | 39 |
| KE1246 | ABCA4 | G5077A | Val1693Ile | Hetero | 34 |
| R376 | ABCA4 | C5327T | Pro1776Leu | Hetero | 39 |
The authors thank patients and their families for participating in the study; Randall Wallach, Deborah Eadie, and Regina Belaney for assistance in preparing the manuscript; and Miriam Meisler for critically reading the manuscript.
HeckenlivelyJR. Retinitis Pigmentosa. 1988;JB Lippincott Company Philadelphia.
DaigerSP. Identifying retinal disease genes: how far have we come, how far do we have to go?.BockG ChaderG GoodeJ eds. Retinal Dystrophies: Functional Genomics to Gene Therapy. 2004;17–36.Novartis Foundation Symposium Chichester, UK.
PreisingMN, HeegardS. Recent advances in early-onset severe retinal degeneration: more than just basic research. Trends Mol Med. 2004;10:51–54.
[CrossRef] [PubMed]
RetNet. University of Texas, Houston, TX. ;Available at http://www.sph.uth.tmc.edu/RetNet/.
TraboulsiE. Genetic Diseases of the Eye. 1998;Oxford University Press New York.
WeleberRG, CarrRE, MurpheyWH, SheffieldVC, StoneEM. Phenotypic variation including retinitis pigmentosa, pattern dystrophy, and fundus flavimaculatus in a single family with a deletion of codon 153 or 154 of the peripherin/RDS gene. Arch Ophthalmol. 1993;111:1531–1542.
[CrossRef] [PubMed]CremersFP, van de PolDJ, van DrielM, et al. Autosomal recessive retinitis pigmentosa and cone-rod dystrophy caused by splice site mutations in the Stargardt’s disease gene ABCR. Hum Mol Genet. 1998;7:355–362.
[CrossRef] [PubMed]MorimuraH, FishmanGA, GroverSA, FultonAB, BersonEL, DryjaTP. Mutations in the RPE65 gene in patients with autosomal recessive retinitis pigmentosa or Leber congenital amaurosis. Proc Natl Acad Sci USA. 1998;95:3088–3093.
[CrossRef] [PubMed]Al-MaghthehM, VithanaE, TarttelinE, et al. Evidence for a major retinitis pigmentosa locus on 19q13.4 (RP11) and association with a unique bimodal expressivity phenotype. Am J Hum Genet. 1996;59:864–871.
[PubMed]WellsJ, WroblewskiJ, KeenJ, et al. Mutations in the human retinal degeneration slow (RDS) gene can cause either retinitis pigmentosa or macular dystrophy. Nat Genet. 1993;3:213–218.
[CrossRef] [PubMed]AclandGM, AguirreGD, RayJ, et al. Gene therapy restores vision in a canine model of childhood blindness. Nat Genet. 2001;28:92–95.
[PubMed]RaduRA, MataNL, NusinowitzS, LiuX, SievingPA, TravisGH. Treatment with isotretinoin inhibits lipofuscin accumulation in a mouse model of recessive Stargardt’s macular degeneration. Proc Natl Acad Sci USA. 2003;100:4742–4747.
[CrossRef] [PubMed]DrmanacR, DrmanacS, LabatI, CrkvenjakovR, VicenticA, GemmellA. Sequencing by hybridization: towards an automated sequencing of one million M13 clones arrayed on membranes. Electrophoresis. 1992;13:566–573.
[CrossRef] [PubMed]MaskosU, SouthernEM. Oligonucleotide hybridizations on glass supports: a novel linker for oligonucleotide synthesis and hybridization properties of oligonucleotides synthesised in situ. Nucleic Acids Res. 1992;20:1679–1684.
[CrossRef] [PubMed]SouthernEM, MaskosU, ElderJK. Analyzing and comparing nucleic acid sequences by hybridization to arrays of oligonucleotides: evaluation using experimental models. Genomics. 1992;13:1008–1017.
[CrossRef] [PubMed]SchenaM, ShalonD, DavisRW, BrownPO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science. 1995;270:467–470.
[CrossRef] [PubMed]HaciaJG, SunB, HuntN, et al. Strategies for mutational analysis of the large multiexon ATM gene using high-density oligonucleotide arrays. Genome Res. 1998;8:1245–1258.
[PubMed]HaciaJG, BrodyLC, CheeMS, FodorSP, CollinsFS. Detection of heterozygous mutations in BRCA1 using high density oligonucleotide arrays and two-colour fluorescence analysis. Nat Genet. 1996;14:441–447.
[CrossRef] [PubMed]YershovG, BarskyV, BelgovskiyA, et al. DNA analysis and diagnostics on oligonucleotide microchips. Proc Natl Acad Sci USA. 1996;93:4913–4918.
[CrossRef] [PubMed]PeaseAC, SolasD, SullivanEJ, CroninMT, HolmesCP, FodorSP. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc Natl Acad Sci USA. 1994;91:5022–5026.
[CrossRef] [PubMed]WarringtonJA, ShahNA, ChenX, et al. New developments in high-throughput resequencing and variation detection using high density microarrays. Hum Mutat. 2002;19:402–409.
[CrossRef] [PubMed]MaitraA, CohenY, GillespieSE, et al. The Human MitoChip: a high-throughput sequencing microarray for mitochondrial mutation detection. Genome Res. 2004;14:812–819.
[CrossRef] [PubMed]WongCW, AlbertTJ, VegaVB, et al. Tracking the evolution of the SARS coronavirus using high-throughput, high-density resequencing arrays. Genome Res. 2004;14:398–405.
[CrossRef] [PubMed]KumarS, HandAT, ConnorJR, et al. Identification and cloning of a connective tissue growth factor-like cDNA from human osteoblasts encoding a novel regulator of osteoblast functions. J Biol Chem. 1999;274:17123–17131.
[CrossRef] [PubMed]DryjaTP, McGeeTL, HahnLB, et al. Mutations within the rhodopsin gene in patients with autosomal dominant retinitis pigmentosa. N Engl J Med. 1990;323:1302–1307.
[CrossRef] [PubMed]DryjaTP, FinnJT, PengYW, McGeeTL, BersonEL, YauKW. Mutations in the gene encoding the alpha subunit of the rod cGMP-gated channel in autosomal recessive retinitis pigmentosa. Proc Natl Acad Sci USA. 1995;92:10177–10181.
[CrossRef] [PubMed]HollanderA, BrinkJ, KokY, SoestS, BornLI, DrielM. Mutations in a human homologue of Drosophila crumbs cause retinitis pigmentosa (RP12). Nat Am. 1999;23:217–221.
ThompsonDA, GyurusP, FleischerLL, et al. Genetics and phenotypes of RPE65 mutations in inherited retinal degeneration. Invest Ophthalmol Vis Sci. 2000;41:4293–4299.
[PubMed]HuangSH, PittlerSJ, HuangX, OliveiraL, BersonEL, DryjaTP. Autosomal recessive retinitis pigmentosa caused by mutations in the alpha subunit of rod cGMP phosphodiesterase. Nat Genet. 1995;11:468–471.
[CrossRef] [PubMed]RiessO, NoerremoelleA, WeberB, MusarellaMA, HaydenMR. The search for mutations in the gene for the beta subunit of the cGMP phosphodiesterase (PDEB) in patients with autosomal recessive retinitis pigmentosa. Am J Hum Genet. 1992;51:755–762.
[PubMed]EichersER, GreenJS, StocktonDW, et al. Newfoundland rod-cone dystrophy, an early-onset retinal dystrophy, is caused by splice-junction mutations in RLBP1. Am J Hum Genet. 2002;70:955–964.
[CrossRef] [PubMed]CutlerDJ, ZwickME, CarrasquilloMM, et al. High-throughput variation detection and genotyping using microarrays. Genome Res. 2001;11:1913–1925.
[PubMed]KhaniSC, KaroukisAJ, YoungJE, et al. Late-onset autosomal dominant macular dystrophy with choroidal neovascularization and nonexudative maculopathy associated with mutation in the RDS gene. Invest Ophthalmol Vis Sci. 2003;44:3570–3577.
[CrossRef] [PubMed]StoneEM. Finding and interpreting genetic variations that are important to ophthalmologists. Trans Am Ophthalmol Soc. 2003;101:437–484.
[PubMed]JaaksonK, ZernantJ, KulmM, et al. Genotyping microarray (gene chip) for the ABCR (ABCA4) gene. Hum Mutat. 2003;22:395–403.
[CrossRef] [PubMed]KleveringBJ, YzerS, RohrschneiderK, et al. Microarray-based mutation analysis of the ABCA4 (ABCR) gene in autosomal recessive cone-rod dystrophy and retinitis pigmentosa. Eur J Hum Genet. 2004;12:1024–1032.
[CrossRef] [PubMed]KajiwaraK, BersonEL, DryjaTP. Digenic retinitis pigmentosa due to mutations at the unlinked peripherin/RDS and ROM1 loci. Science. 1994;264:1604–1608.
[CrossRef] [PubMed]BadanoJL, KimJC, HoskinsBE, et al. Heterozygous mutations in BBS1, BBS2 and BBS6 have a potential epistatic effect on Bardet-Biedl patients with two mutations at a second BBS locus. Hum Mol Genet. 2003;12:1651–1659.
[CrossRef] [PubMed]BriggsCE, RucinskiD, RosenfeldPJ, HiroseT, BersonEL, DryjaTP. Mutations in
ABCR (
ABCA4) in patients with Stargardt macular degeneration or cone–rod degeneration. Invest Ophthalmol Vis Sci. 2001;42:2229–2236.
[PubMed]