Pedigree and haplotypes of a Minnesota6-generation family with X-linked high myopia and protanopia. Circles andsquares denote females and males, respectively; solid symbols denote affectedindividuals; and symbols with slashes denote deceased individuals. Obligatefemale carriers are denoted with a circle containing a dot. Unknown phenotypestatus is denoted with a circle or square containing a question mark. Eachindividual studied (plus sign) has alleles shown for X chromosome markersin descending marker order from the telomere of the p arm to the telomereof the q arm. Haplotypes were constructed on the basis of the minimum numberof recombinations between markers. The chromosome assumed to carry the diseaseallele is blackened. Only essential matings are shown; nonparticipating familymembers are not shown. Haplotype analysis indicates crossovers in individuals24 and 34, which defines the chromosomal region of interest. This region spans34.4 centimorgans from GATA31E08 to q-ter. Inferredand undetermined genotypes are denoted with parentheses and question marks,respectively.
Limited version of a branch ofthe original Bornholm eye disease pedigree with haplotypes of chromosome Xq27-q28markers. Symbols are explained in the legend to Figure 1. Individual 3 hasonly deuteranopia, which is unrelated to Bornholm eye disease (see the lastparagraph of the "Molecular Structure of Color Vision Genes" subsection ofthe "Results" section).
Comparative electroretinogramresponses between an affected Minnesota family participant (person 24) (someblink artifact), obligate carrier (person 33), and an unrelated normal malecontrol. A, Mixed rod and cone responses obtained with flashes of maximumintensity (0 dB) show a subnormal b-wave amplitude in the affected participant,but not in the carrier or control. B, Low-intensity (24-dB) dark-adapted scotopicrod responses showed no definite abnormality among the participants. C, Oscillatorypotentials, single white-flash, dark-adapted (scotopic rod) responses recordedwith the use of high-intensity 0-dB stimulation showed no definite abnormalityamong the participants. D, Maximum-flash (0-dB), photopic cone responses showmarked reduction of both a- and b-wave amplitudes in the affected participant,compared with normal amplitudes for the carrier and control recordings. E,The affected individual has, in contrast to the carrier and control, grosslydelayed and reduced amplitudes of his 30-Hz-flicker photopic cone electroretinogramresponses.
Schematic representation of theX chromosome and the Xq27/Xq28 region. Marker order and genetic distanceswere determined by reference to the National Center for Biotechnology Informationgenetic map, the Weber 9 Set, and the Généthon genetic map.The position of the markers relative to the chromosomal bands are approximate.Opsin 1 indicates red-green cone pigment gene array; cM, centimorgan.
A, Genomic structure of the normalhuman red and green pigment array. The red and green pigment genes span 15.2kilobase (kb) and 13.3 kb, respectively, with a 24.0-kb separation betweenthese 2 genes. Additional copies of the green pigment gene arranged in tandemat 24.0-kb intervals are found in 60% of the general white population. B,Relationship of the position of the 3′-red-green-5′ (R-G) hybridgene in the visual-pigment gene array found in the Minnesota (MN) family.Hybrids can occur because of the high degree of homology of these 2 genes(98%) (see the first paragraph of the "Molecular Structure of Color VisionGenes" subsection of the "Results" section). Individuals have a variable numberof green pigment genes. This family has 3 additional normal green pigmentgenes. C, Genomic structure of the human red and green pigment array fromDNA samples of individuals 26 and 28 with Bornholm eye disease (BED). Analysesof DNA samples showed a normal red pigment gene in the first position, followedby a green-red hybrid and 2 intact green pigment genes. The fusion point inthe hybrid gene was in intron 4 (G4R5). This gene array is typically foundamong subjects with deutan color vision defects. D, Their unaffected malecousin, BED-27, showed a normal red-green pigment array.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Young TL, Deeb SS, Ronan SM, et al. X-Linked High Myopia Associated With Cone Dysfunction. Arch Ophthalmol. 2004;122(6):897–908. doi:10.1001/archopht.122.6.897
Bornholm eye disease (BED) consists of X-linked high myopia, high cylinder,optic nerve hypoplasia, reduced electroretinographic flicker with abnormalphotopic responses, and deuteranopia. The disease maps to chromosome Xq28and is the first designated high-grade myopia locus (MYP1). We studied a second family from Minnesota with a similar X-linkedphenotype, also of Danish descent. All affected males had protanopia insteadof deuteranopia.
X chromosome genotyping, fine-point mapping, and haplotype analysisof the DNA from 22 Minnesota family individuals (8 affected males and 5 carrierfemales) and 6 members of the original family with BED were performed. Haplotypecomparisons and mutation screening of the red-green cone pigment gene arraywere performed on DNA from both kindreds.
Significant maximum logarithm of odds scores of 3.38 and 3.11 at thetas;= 0.0 were obtained with polymorphic microsatellite markers DXS8106 and DXYS154, respectively, in theMinnesota family. Haplotype analysis defined an interval of 34.4 cM at chromosomeXq27.3-Xq28. Affected males had a red-green pigment hybrid gene consistentwith protanopia. We genotyped Xq27-28 polymorphic markers of the family withBED, and narrowed the critical interval to 6.8 cM. The haplotypes of the affectedindividuals were different from those of the Minnesota pedigree. Bornholmeye disease–affected individuals showed the presence of a green-redhybrid gene consistent with deuteranopia.
Because of the close geographic origin of the 2 families, we expectedaffected individuals to have the same haplotype in the vicinity of the samemutation. Mapping studies, however, suggested independent mutations of thesame gene. The red-green and green-red hybrid genes are common X-linked colorvision defects, and thus are unrelated to the high myopia and other eye abnormalitiesin these 2 families.
X-linked high myopia with possible cone dysfunction has been mappedto chromosome Xq28 with intervals of 34.4 and 6.8 centimorgan for 2 familiesof Danish origin.
X-linked transmission of simple myopia has rarely been reported in theliterature. Brückner and Franceschetti1 describedan X-linked form of myopia up to 25 diopters (D) with macular and latticedegeneration, and retinal detachments as a common complication. Wold2 provided an incomplete description of families inwhich some members were highly myopic. Bartsocas and Kastrantas3 describeda family with X-linked high myopia in 3 generations. The obligate carriershad low degrees of myopia.
X-linked myopia transmission associated with other ocular findings,either as a possible secondary effect or as part of a syndrome, is more common.The association of X-linked transmission with congenital stationary nightblindness and retinitis pigmentosa has been well described.4,5 Forsiusand Eriksson6 described the Åland eyedisease, with X-linked myopia, albinism of the fundus, markedly impaired vision,abnormal dark adaptation, dyschromatopsia classified as protanomaly, fovealhypoplasia, and nystagmus. Francois et al7 describedan X-linked atypical achromatopsia combined with myopia, impaired vision,foveal aplasia, nystagmus, and photophobia. Mäntyjärvi et al8 described a family with high myopia and cone dysfunctionthat they mapped to Xp11.4-q13.1 and designated as progressive cone-rod dystrophylocus 3 (COD3), which may be the same X-linked progressivecone-rod dystrophy with myopia recently mapped by Jalkanen et al.9
The Bornholm eye disease (BED) was mapped to Xq28 by Schwartz et alin 199010 by linkage analysis using 3 restrictionfragment length polymorphic markers. No haplotyped interval was provided andno gene mutations have been reported for this disorder. This locus has beendesignated the MYP1 locus by the Human Gene NomenclatureCommittee (http://www.gene.ucl.ac.uk/nomenclature) (Online MendelianInheritance in Man entry 310460). The original 5-generation family studiedfrom the island of Bornholm, Denmark, was initially described by Haim et alin 1988.11 Affected males had an early-onset(1.5 to 5 years of age) X-linked form of myopia ranging from −6.75 to−11.25 D accompanied by several other ophthalmologic features. Six of8 examined affected males had astigmatism of 1 D or more. Their ophthalmoscopicexaminations showed temporal conus of the optic disc, as well as peripapillaryatrophy of the retinal pigment epithelium, giving the appearance of moderateoptic nerve hypoplasia. The choroidal vessels were described as visible inthe posterior pole, and no macular or lattice degenerations were noted. Nopatients sustained a retinal detachment. Electroretinogram (ERG) testing (performedin 4 of the 8 affected individuals) demonstrated subnormal flicker functionand normal results of scotopic testing. The refractive error, visual acuity,and results of ERG testing and color vision testing all remained stable inmost affected males after age 16 years. All affected males had a color visiondefect of deuteranopia.
We ascertained a large Minnesota family originating from the nearbyDanish islands of Møn and Zealand with X-linked high myopia and phenotypicfeatures similar to those described for BED. However, all affected individualsin the Minnesota family had a protanopic color vision defect. Thus, all affectedindividuals in the Minnesota and BED families had a color vision deficiency—albeitof different types—of protanopia and deuteranopia, respectively. Thisraised the question of whether color vision deficiency was part of the syndrome.We hypothesized that mutations in either the red or green pigment genes mighthave disrupted structure and function to cause cone dysfunction and the highmyopia associated with it. A few such mutations had previously been described,such as the C203R substitution in the green pigment gene.12-14 Weinvestigated the gross structure and sequence of the red and green color visiongenes in both the original BED family and the present Minnesota family. Weexamined the possibility that the phenotype of the 2 families mapped to thesame locus, and whether mutations in the visual pigment genes might be responsiblefor the ophthalmologic phenotype. Fine-point mapping and candidate regionmutation screening were performed in an effort to further refine the mappedinterval, and to uncover the genetic mechanisms associated with this disorder.
The family studied in this investigation was a 6-generation family originatingfrom Denmark (Figure 1). Individual1 was from the island of Møn, and his wife, individual 2, was fromVortenberg on the island of Zealand. There was no known consanguinity. Eightaffected males and 5 female carriers were studied. In total, the DNA of 22people underwent genotyping with polymorphic microsatellite markers of theX chromosome (Figure 1).
All Minnesota family participants underwent extensive ophthalmologicevaluation, including Snellen visual acuity testing, slitlamp examination,intraocular pressure testing, cycloplegic refraction, and detailed funduscopy.In addition, fundus photography, ultrasound axial length, and keratometry(corneal curvature) measurements were performed on most participants.
Full-field, single-flash, and flicker ERG responses were recorded (UTAS-E2000 system; LKC Technologies Inc, Gaithersburg, Md) according to the recommendationsof the International Society for Clinical Electrophysiology of Vision.15 Burian-Allen bipolar electrodes were used. Mixedrod and cone responses were obtained by means of stimulation with flashesof maximum intensity (0 dB) in dark-adapted conditions. Dark-adapted scotopicrod responses were recorded by means of full-field white flashes of relativelylow intensity (24-dB attenuation). Single white-flash, dark-adapted oscillatorypotentials measuring scotopic rod responses were recorded with high-intensity0-dB stimulation. Photopic cone responses were elicited in light adaptationto white background (480 lumen/m2), and with maximum flash stimulation(0 dB). Flicker ERG responses (30 Hz) were recorded in light adaptation (480lumen/m2), using an averaging technique (n = 10) and stimulationwith maximum-intensity flashes.
Color vision was tested under standardized conditions, with pseudoisochromaticplates (Ishihara test for color blindness, 38-plate edition, 1988), the desaturatedpanel D-15 test, and the Farnsworth Munsell 100-hue test. The anomaloscopeNagel type II test was performed on 2 affected male subjects, 24 and 25.
Individuals with genomic DNA samples from the original BED family areshown in Figure 2. Individuals BED-16and BED-25 had deuteranopia only, and not the full clinical spectrum of eyefindings in the BED, such as myopia, photopic ERG abnormalities, and opticnerve hypoplasia. Individuals BED-26 and BED-28 were affected males, individualBED-22 was a carrier female, and individual BED-27 was an unaffected maleoffspring of BED-22. Individual BED-22 was also the monozygotic twin of themother of individual BED-26 (BED-20).
This research was performed according to the Declaration of Helsinkiand was approved by the local institutional review boards.
The DNA was isolated from peripheral-blood lymphocytes by standard techniques.The genome screen used polymorphic microsatellite markers from the Weber 4a,8a, and 9a sets (Research Genetics Inc, Huntsville, Ala).16 Forfine mapping, additional markers were selected from the Généthon(http://www.genethon.fr) and Marshfield (http://www.marshfieldclinic.org/research/genetics/) genetic map databases of chromosome Xq27 and Xq28. The 5′ markerof each primer set was modified with a special M13 sequence that allows forfluorescent detection.17
Polymerase chain reactions (PCRs) were prepared in 96-well plates with2.0mM magnesium chloride, 50mM potassium chloride, 10mM Tris hydrochloride(pH 9.0), 0.1% Triton X-100, 200µM of each dNTP, 1.0 pmol of each markerprimer, 0.08 pmol of M13 primer, 0.4 U of Taq DNApolymerase (AmpliTaq or AmpliTaq Gold; Perkin Elmer, Norwalk, Conn), and 20ng of DNA. Annealing temperatures were adjusted according to the specificcharacteristics of each marker. After amplification of 30 cycles, aliquotsof amplification products were mixed with 6X Ficoll buffer and separated byelectrophoresis through a preheated 23-cm, 6% polyacrylamide, 7M urea denaturinggel. A dual-dye infrared DNA analyzer was used (Li-Cor DNA 4200; Li-Cor Inc,Lincoln, Neb). Allele sizes were determined from computer images by meansof restriction fragment length polymorphism (RFLP) scan software (ScanalyticsInc, Fairfax, Va). Alleles were visualized as an autoradiogram-like imageon a computer, and size was determined with RFLP scan software (ScanalyticsInc). Allele sizes were then directly imported to a database (Filemaker ProInc, Santa Clara, Calif), which was then used for logarithm of odds (LOD)score determination.
Linkage analysis was performed with the C version of the LINKAGE package,FASTLINK18,19 and the utilityprograms Makeped, LINKAGE control program, and LINKAGE report program fromLINKAGE 5.1.19-22 Markerallele frequency estimates were based on the frequencies of alleles in marriedunrelated individuals in the families. Two-point LOD scores were also calculatedwith all alleles set at equal frequencies to control for allele frequencyeffects. Standard marker databases were used for intermarker recombinationfrequencies, marker order, and marker distances.
Linkage analysis was performed with a myopia gene frequency of 0.01at 100% penetrance of the gene. For initial data analysis, high myopia wasassumed to have an X-linked inheritance mode, and the MLINK program18,22 was used to calculate 2-point LODscores between myopia and each marker. All affected individuals and informativespouses were included in the linkage analysis. The LODSCORE program was usedto calculate maximum LOD scores at the lowest recombination frequency.
Primers with a 19–base pair M13 tail were designed by using apublished DNA sequence of the locus control region (LCR) of the cone pigmentgenes on Xq28.23,24 The PCR reactionswere performed as previously described,23,24 andthe sequencing products were mixed with 6X Ficoll buffer and electrophoresedthrough a preheated 39-cm, 7% polyacrylamide, 7M urea denaturing gel in aDNA analyzer (Li-Cor Inc) as described in the preceding section. The basepair sequence was visualized on a computer with the use of Gene Imager software(Li-Cor Inc). The sequence was compared with that of unaffected controls andwith the published sequence.
The gross structure of the red and green visual pigment genes was determinedby means of quantitative PCR amplification followed by single-stranded conformationalpolymorphism analysis (SSCP) as described in detail elsewhere.25,26 Thisprocedure allowed determination of the ratio of red to green promoters, whichgives the total number of genes, as well as those of exons 2 to 5 (exons 1and 6 are identical between the red and green pigment genes). In additionto ratios, SSCP detects nucleotide variants.27
The primers used to cover the promoter, coding regions, and exon-intronjunctions are described in detail elsewhere.25,26 Thefollowing primers were used: 32 (antisense primer in exon 1) for the promoterregion, 130 (an antisense primer in intron 1) for exon 1, 174 (a sense primerin intron 1) for exon 2, 2A (a sense primer in exon 3) and 2Z (an antisenseprimer in exon 3) for exon 3, 30 (a sense primer in exon 4) and 2B (an antisenseprimer in exon 4) for exon 4, 3C (a sense primer in exon 5) and 79G (an antisenseprimer in exon 5) for exon 5, and 34 (an antisense primer in 3′-UTR)for exon 6. First, 12-kilobase (kb) segments extending from the promoter toexon 5 of the red and green pigment genes were amplified separately. The forwardprimer was either red promoter–specific (169RF) or green promoter–specific(171GF). The reverse primer in exon 5 (79G) was green-specific (5′),as the arrays were found not to contain red exon 5 (presence of red-greenhybrid gene in the first position).
The PCR reaction contained, in a total volume of 25 or 50 µL,0.25 or 0.5 µg of total genomic DNA, 0.2µM of each primer, 400µMof dNTPs, 1 × LA PCR buffer II (Applied Biosystems, Foster City, Calif)(magnesium chloride), and 1 or 2 U of Taq DNA polymerase(Takara LA; Takara Biotechnology, Shiga, Japan). The amplification conditionswere as follows: 1 minute of denaturation at 94°C was followed by 14 cyclesof 98°C for 10 seconds and 68°C for 12 minutes; 16 cycles of 98°Cfor 10 seconds and 68°C for 12 minutes plus 15 s/cycle; and 72°C for10 minutes. Second, a segment encompassing exon 5, intron 5, and exon 6 wasamplified with forward primers 3C and 34 (5′-GCAGTGAAAGCCTCTGTGACT-3′)primers.22 Third, the PCR products were gelpurified with a gel extraction kit (QIAquick; Qiagen, Valencia, Calif) andthe desired portions were directly sequenced by means of a kit (BigDye TerminatorCycle Sequencing kit; Perkin Elmer, Foster City, Calif) (total, 40 cycles)and a DNA sequencer (ABI Prism; Applied Biosystems).
Southern analysis was performed with a 1.3-kb complementary DNA (cDNA)clone of the red cone pigment gene using standard techniques.13,23,24 Thishs7 probe was cut from its pUC19 vector, gel purified, and labeled with α-32P-dCTP. Patient sample genomic DNA was digested with EcoRI and BamHI (Gibco BRL, Gaithersburg,Md). Each reaction contained 7 µg of genomic DNA, 25 U of each enzymein a total volume of 50 µL, and was incubated for 2 hours at 37°C.The digested products were separated by electrophoresis on a 0.7% SeaKem agarosegel (FMC; BioProducts, Rockland, Md). After separation, the cut DNA was capillaryblot transferred onto a membrane overnight and blot fixed by UV crosslinking.Prehybridization and hybridization steps were performed with the labeled cDNA.The hybridized bands were visualized after overnight exposure of the membraneto radiographic film at −80°C.
A peripheral blood sample for chromosome analysis of Minnesota participant25 was processed according to standard techniques, and prometaphase preparationswere analyzed by means of 850 high-resolution G-banding.28
Three generations of affected males were clinically studied in the Minnesotakindred. The youngest affected male studied was 5 months of age (subject 37);the oldest was 37 years of age (subject 25). Myopia was diagnosed at 6 yearsof age or younger (13 months to 6 years) in the affected males. Correctedvisual acuity was excellent to good in all affected males (range of 20/20to 20/40). On review of patients' ophthalmologic records, the visual acuityand degree of myopia was stable after puberty (age 16-20 years). We followedup all participants for 5 years, and the clinical features have been stablefor all affected males.
No patient had photophobia, nystagmus, night blindness, or paradoxicalpupillary response. Affected adult individuals had high myopia (mean sphericalrefractive component of −13.18 D; range, −10.25 to −18.25D). The average adult keratometry reading for astigmatic cylinder of 43.78D (range, +41.50 to +44.62 D) was not significantly higher (P = .18, 2-tailed t test) than the publishedmean ± SD adult normal value of 43.1 ± 1.62 D.29 Theaverage adult axial length of 28.39 mm (range, 26.84-31.73 mm) was significantlygreater (P = .009) than the published mean ±SD adult normal value of 24.2 ± 0.85 mm.29 Intraocularpressures were normal in all tested participants.
All affected males had fundus findings of optic nerve temporal conusand posterior pole retinal pigment epithelial thinning with prominent choroidalpattern. Optic nerve head caliber was normal. The peripheral fundus had lessretinal pigment epithelial thinning with more pigmentation relative to theposterior pole. No peripheral retinal degenerative changes were noted. A tapetalreflex was not observed. No macular changes (bull's eye, granular, etc) werenoted for any members. The vitreous was normal in all examined participants.
The ERG testing of all affected males in the Minnesota family showednormal rod-dominated scotopic waveforms (representative rod response b wave,174/172 µV and 100/113.5 milliseconds; normal, >120 µV and <108.5milliseconds). The mixed rod-cone b-wave amplitude responses were normal insome and depressed in others, with normal implicit times (representative maximalrod and cone response b wave, 179/277 µV and 44.5/44 milliseconds; normal,>320 µV and <53.5 milliseconds). The cone responses showed subnormalb-wave amplitudes and normal implicit times (representative cone responseb wave, 40/29 µV and 32.5/35.5 milliseconds; normal, >65 µV and<32.5 milliseconds). All affected males had reduced amplitudes of the 30-Hzflicker ERG recordings with increased implicit times (representative flickerresponse b wave, 37/20 µV and 32.5/35 milliseconds; normal, >42 µVand <30 milliseconds). All carrier females and unaffected males or femaleshad normal ERG recordings. Comparative ERG findings between an affected subject,an obligate carrier, and an unaffected male control are shown in Figure 3. While these findings are suggestiveof a cone dysfunction, ERG amplitude may also be influenced by increasingaxial length.30-33 Westallet al33 reported significant differences inERG amplitude between subjects with high myopia and those with small refractiveerrors. Implicit times, the ratio of b to a wave, and semisaturation constantsshowed no significant differences. These authors and others found these resultsto apply to rod and cone responses. However, cone responses were more affected,presumably because of change in geometry, which could distort the axis ofphotoreceptor segments or reduce the number of cones per unit area.30-33 Inthe latter case, it would not matter if all photoreceptors were included inthe ERG measurements. In addition, subnormal flicker ERG findings noted inthe BED kindred may be suggestive of cone dysfunction but, in view of highmyopia, are not diagnostic.
All affected males of the Minnesota family failed the Ishihara color-platescreening test and had color vision testing results consistent with protanopia,unlike the BED phenotype, which had deuteranopia as a clinical feature. TheRayleigh match range determined by type II Nagel anomaloscope was 73, indicatingdichromatic color vision consistent with a severe protan defect. Automatedstatic perimetry using programs 30-2 and 30-1 (Humphrey Field Analyzer; HumphreyInstruments, Inc, San Leandro, Calif) of the central visual fields and formalGoldmann visual field testing in individuals 24 and 25 were normal. Pertinentclinical eye findings are listed in Table1.
All carrier females, noncarrier females, and unaffected males had normalrefractive errors, ophthalmologic evaluations, ERG results, and color testingresults.
The Minnesota pedigree was studied with 23 markers spanning the entireX chromosome (Table 2 and Figure 4). One or multiple recombinationevents occurred between each of the markers outside but not within the Xq27-q28region and the putative disease gene. Haplotypes displaying the segregationmarker alleles in the chromosome Xq27-28 region in the family are shown in Figure 1.
The status of individual 37 was designated as unknown primarily becauseof his age. He was 5 months old at the time of his first examination and hada cycloplegic refractive error of +1.00 +1.00 at 90° OU. A repeat ophthalmologicexamination at 9 months of age showed a reduction of 1.00 D in hyperopic spherein both eyes, indicating a non–age-appropriate myopic shift. Resultsof fundus examination were normal. Haplotype analysis shows that he inheritedthe disorder (Figure 1). More clinicalinformation is needed and will be available when this subject is old enoughto undergo further testing.
Close linkage without recombination was found between the putative geneand markers DXS8106 at Xq27.3, DXS8103 at Xq28, and DXYS154 at Xq28, withmaximum LOD scores of 3.38, 3.01, and 3.11 at thetas; = 0.0, respectively.Analysis of the chromosome Xq27-28 haplotypes (Figure 1) suggests that the disease locus in the Minnesota familyis located in the chromosomal region between the marker loci GATA31E08 and the q-ter, due to informative recombination events betweenmarkers GATA31E08 and DXS8106 forindividuals 24 and 34. This region defines a 34.4-cM interval (Figure 4).
The BED family phenotype had a provisional assignment to the distalpart of the X chromosome at Xq28 with the use of a limited number of markersoriginally. No critical region was defined within Xq28. We performed genotypingof 6 members of the BED family with additional markers. We observed that thedisorders of the Minnesota and BED families of Danish origin map to the samelocus (Figure 2 and Table 3), suggesting that these families carry the same mutation.Since the 2 families originated from villages that are only 100 miles apart,we explored the degree of identity in genotypes and haplotypes at variousXq28 markers. The markers studied were DXS8106, DXS8028,DXS998, DXS8069, DXS8103, DXS8061, DXS8087, DXS1073, and DXYS154. Our haplotype analysis of the 6 participating members of theBED family with the use of chromosome Xq27.3-28 polymorphic markers showsthat the region of the X chromosome associated with high myopia and presumablecone dysfunction has different ancestral origins in the Minnesota and BEDfamilies (Table 3). Some markerscould not be typed, probably because of older DNA (DXS8028,DXS998, DXS8069, and DXS8103). We found evidenceof a recombination event that occurred in subject BED-20 (probable fraternaltwin), inferred from the haplotype of her affected son BED-26, between markers DXS8103 and DXS8061. This definesa critical interval of 6.8 cM for the site of the basic defect. The resultsare shown in Figure 2 and Table 3.
Figure 5A shows a diagramof a typical X-linked visual pigment gene array of a male with normal colorvision. Quantitative SSCP analysis showed that the ratio of green to red pigmentgene promoters for affected Minnesota subjects 29 and 35 was 3, indicatingthat each of their arrays contained 4 genes (Figure 5A). Sequence analysis of exons 2 to 5 showed that the arraysof both subjects were composed of 1 red-green hybrid gene with a fusion pointin intron 4 (R4G5) followed by 3 normal green pigment genes. Exon 3 of boththe hybrid and normal green pigment genes had alanine at the polymorphic aminoacid residue 180.24 Therefore, it is inferredthat the pigment encoded by this hybrid gene is identical in spectral propertiesto that encoded by the green pigment gene, and is consistent with protanopia.The SSCP showed no abnormal sequence variants in these 2 individuals. Thisgene array is typically found among subjects with severe protan color visiondefects and without high myopia.12,13,34,35
We also sequenced the 0.6-kb LCR of the visual pigment gene array located3.6 kb upstream from the first cone pigment gene (usually the red pigmentgene).24 This conserved 5′ LCR interactswith either the red or green pigment gene promoters and directs gene expression.We found no mutations in affected individuals, obligate carriers, and unaffectedindividuals in both kindreds.
Similar analyses of BED family DNA samples from 2 affected individuals(BED-26 and BED-28) showed a normal red pigment gene in the first position,followed by a green-red hybrid and 2 intact green pigment genes (Figure 5C). A fusion point in the hybridgene was in intron 4 (G4R5). This gene array is typically found among subjectswith deutan color vision defects.34,35 Theirunaffected male cousin, BED-27, showed a normal red-green pigment array (Figure 5D). It is noteworthy that 2 genearrays associated with deutan color vision segregate in this family. One isgenetically linked to the locus associated with the ophthalmic findings, andthe other—subject 3 in Figure 2—didnot have the ophthalmic syndrome and, on the basis of DNA marker analysis,had married into the BED family.10
To rule out the existence of rearrangements within the color visiongene array, we performed Southern blot analysis using EcoRI and BamHI restriction enzymes.13 Southern blot analysis using the cDNA probe for thered pigment gene did not show deletions or rearrangements. Five individualsfrom the Minnesota pedigree were included in this analysis: 2 affected (subjects29 and 35), 2 unaffected males (subjects 11 and 32 [an unrelated marry-in]),and 1 obligate carrier (subject 19). All of the individuals tested, affectedand unaffected, showed the same bands, consistent with the published resultsfor the wild-type sequence in previous studies using this probe.13 Thesedata rule out the presence of insertions, deletions, or inversions at thislocus (data not shown).
High-resolution cytogenetic analysis was performed on DNA from affectedMinnesota male 24. No abnormalities were detected (karyotype not shown).
New findings in this study are the discovery of a second Danish familyin Minnesota with high-grade myopia that maps to the terminal arm of chromosomeXq28, and the narrowing of the critical interval from 34.4 to 6.8 centimorgan.The ethnic origin, many phenotypic aspects, and initial genetic mapping ofthe Minnesota pedigree potentially matched that of the 5-generation BED familywhose defect was designated the MYP1 or first myopialocus. However, affected members of the Minnesota family had a protan andnot a deutan color vision defect, as seen in affected members of the BED family.Apart from the presence of hybrid pigment genes (red-green [protan] in theMinnesota family and green-red [deutan] in the BED family) that were identicalto those usually found in individuals with these common X-linked color visiondefects, no inactivating missense mutations12,13 werefound at the color vision locus.
Red-green color vision defects are common and were among the first recognizedX-linked traits.35-38 Themolecular genetics of the visual pigments mediating normal and defective colorvision have been meticulously studied in recent years.12,24-26,34,35 Thered-green pigment gene complex maps to a subterminal site on the long armof the X chromosome. The red-green gene arrays are composed of a single redpigment gene (6 exons) and 1 or more green pigment genes (6 exons) locateddownstream (3′) of the red gene. The close homology of the red and greenopsin genes, including introns, makes them prone to unequal crossing overand accounts for the numerical polymorphism of the green genes, usually rangingbetween 1 and 3. However, only the single proximal red pigment gene and only1 proximal green pigment gene are expressed in the retina.35 Deuteranomalyis the most common defect in Caucasian populations (4% to 5% of males). Theother defects (protanopia, protanomaly, and deuteranopia) have frequenciesof approximately 1% each among males. Myopia and ERG changes are not associatedwith the common red-green color deficits.
The finding of an almost identical ophthalmologic phenotype mappingto the Xq27-28 region in 2 different families of Danish origin from islandsin proximity suggests the same mutation. Genotyping with DNA markers in thecritical X chromosome region, however, showed differences between affectedmembers of the 2 families, arguing against an identical mutation. Intuitively,the most likely model would be that of a single unequal crossover betweenthe red and green pigment gene in ancestral chromosomes that could have producedboth red-green (Minnesota) and green-red (BED) hybrid genes associated withdeutan and protan color deficiency, respectively, in affected members of the2 families. Such an unequal crossover event would have had to occur in a womanhomozygous for the underlying ophthalmologic defect but with normal colorvision genes. Furthermore, haplotypic differences between affected membersin the 2 families were inconsistent with this hypothesis.
A recent report describes a large 6-generation Asian Indian pedigreewith high-grade myopia (mean of −13.33 D) transmitted in an X-linkedfashion that maps to the pseudoautosomal region of chromosome Xq28.39 The authors obtained a maximum LOD score of 3.99at thetas; = 0 with marker DXYS154. Little informationwas provided regarding associated clinical features, however, making comparisonswith the Minnesota and BED pedigrees difficult. The affected individuals donot have a color vision deficiency, however (Uppala Radhakrishna, PhD, oralcommunication, November 18, 2002), which provides greater evidence that colorvision deficiency is not part of the phenotype. This report lends supportto an X-linked locus for myopia at chromosome Xq28.
The disease segregating in these pedigrees exhibits some clinical similaritiesto other reported types of X-linked cone dystrophy; however, the dissimilaritiespoint to a novel disorder. These types include COD1, which maps to Xp21.1-p11.240-42; COD2, which mapsto Xq2743; blue cone monochromacy, which mapsto Xq2723; and a cone dystrophy reported byReichel et al44, which maps to Xq28 and doesnot have an assigned gene name. The disease locus for our pedigrees mappedto chromosome Xq, rather than chromosome Xp, so it is unlikely that our pedigreesrepresent a COD1 disorder. Furthermore, with COD1 the affected males havereduced visual acuity, progressive macular atrophy, and progressive colordegeneration with ultimate achromatopsia. Our pedigrees also do not matchthe phenotypic and genetic characteristics described for COD2 (and blue conemonochromacy) because of the progressive nature of those disorders, includingreduced rod ERG in later stages and carrier female abnormalities.23,43 The phenotype of blue cone monochromacyincludes nystagmus, incomplete achromatopsia, and mutations in the LCR.23,43 The Xq28 mapped cone dystrophy describedby Reichel et al44 also does not fit with theMinnesota pedigree. Their affected participants exhibited reduced visual acuity,progressive macular atrophy, deteriorating color vision progressing to centralacquired achromatopsia, and a deletion in the red-cone pigment gene detectedby Southern blot analysis using the same hs7 cDNA probe. Table 4 shows comparative clinical features with the Minnesota pedigree,and those of others with X-linked cone dysfunction.10,23,40-47
A chromosome Xq28-linked cone dystrophy in 3 English families was recentlydescribed by Johnson et al.14 The affectedmembers had protanopia with an associated progressive cone dystrophy and werefound to have point mutations in the X-linked red pigment gene that encodesthe opsin gene and results in photoreceptor death, rather than the benignred-green hybrid gene found in the common protan defects.
A search for genes and/or expressed sequence tags physically mappedbetween marker GATA31E08 and q-ter shows 48 regulatoryor structural genes, 81 unidentified transcripts, 16 messenger RNAs, and 2open reading frames (Integrated X Chromosome Database [http://ixdb.molgen.mpg.de/maps.html]; National Center for Biotechnology Information database [http://www.ncbi.nlm.nih.gov/genemap/map.cgi]). Other close loci include those for blue cone monochromacy (mentionedherein), myotubular myopathy, adrenoleukodystrophy, dyskeratosis punctata,X-linked dominant chondrodysplasia punctata, glucose-6-phosphate dehydrogenase,cardiac valvular dysplasia 1, Emery-Dreifuss muscular dystrophy, factor VIII–associatedgene, hemophilia A, fragile X disorders type E and F sites, and biglycan.Biglycan, also termed proteoglycan I, is a small,leucine-rich proteoglycan expressed during chondrogenesis in cartilage andsclera.48 It functions in connective tissuemetabolism by binding to collagen fibrils and transforming growth factor βand may promote neuronal survival. Biglycan appeared to be a relevant candidategene and underwent mutational screening analysis. No mutations were found(data not shown). Previous mutational analysis of this 8-exon gene was performedby Das et al,49 who excluded it as a candidatefor X-linked dominant chondrodysplasia punctata, dyskeratosis congenita, andincontinentia pigmenti.
In conclusion, we report the molecular genetic findings of 2 familiesoriginating from Denmark with high-grade myopia and red-green color visiondefects. All affected males from the Minnesota family had an early onset andseemingly nondegenerative high-grade myopia, myopic fundus changes, severeprotanomaly, and reduced cone function on ERG testing. The cone abnormalitiessuggested by the ERG findings, however, might be secondary to or are accentuatedby the severe myopia. Two-point LOD score analysis confirms linkage of thephenotype to the telomeric end of the X chromosome. The haplotypes at theXq27-28 region in the Minnesota and BED families are different, suggestingthat independent mutational events led to the phenotype. In addition, thered and green pigment gene arrays of the affected individuals in the 2 familiesare different, one containing a red-green hybrid gene (Minnesota family) consistentwith the demonstrated protanopia, and the other containing a green-red hybridgene (BED family) consistent with the demonstrated deuteranopia. These findingsstrongly suggest that the color vision defects are unrelated to the underlyingdisease and do not play a role in causing cone dysfunction. Both phenotypesappear to be novel forms of X-linked myopia and nonprogressive cone dysfunction.Definitive identification of the DNA defect causing myopia and cone dysfunctionin these families should resolve the exact nature of the basic genetic defect(s).
Corresponding author and reprints: Terri L. Young, MD, Division ofOphthalmology, The Children's Hospital of Philadelphia, 34th and Civic CenterBoulevard, Philadelphia, PA 19104 (e-mail: firstname.lastname@example.org).
Submitted for publication February 11, 2003; final revision receivedJanuary 15, 2004; accepted January 15, 2004.
This research was funded by a Career Development Award from Researchto Prevent Blindness Inc, New York, NY (Dr Young); Vision Core grant 2P30EY001583-26(Dr Young) and grants 1K23EY00376 (Dr Young) and EY08395 (Drs Deeb and Motulsky)from the National Eye Institute, Bethesda, Md; and the Mable E. Leslie endowedchair of The Children's Hospital of Philadelphia, Philadelphia, Pa (Dr Young).
We thank the families for their participation in this study. The redcDNA probe designated hs7, containing exons 1 through 6 of the red pigmentgene, was kindly provided by Jeremy Nathans, PhD, The Johns Hopkins Schoolof Medicine, Baltimore, Md. We are grateful to Richard Purple, MD, the Universityof Minnesota Medical School, Minneapolis, for performing the extensive colorvision testing on the family participants. We also thank Janice Peterson,BS, and Sabitha Shriram, MS, for their technical assistance.
Create a personal account or sign in to: