A Third Locus (GLC1D) for Adult-onsetPrimary Open-angle Glaucoma

Maps to the 8q23 Region

OVIDIU C. TRIFAN, MD, ELIAS I. TRABOULSI, MD, DILIANA STOILOVA, MD,IHUOMA ALOZIE, MD, RANDALL NGUYEN, MD,

SHARATH RAJA, MD, AND MANSOOR SARFARAZI, PHD

● PURPOSE: Two genes for adult-onset primaryopen-angle glaucoma have been mapped to chromo-somes 2cen-q13 and 3q21-q24. We studied a familywith adult-onset primary open-angle glaucoma inwhich the disease did not map to these two chromo-somal regions.● METHODS: We ascertained a four-generation fam-ily with adult-onset primary open-angle glaucoma inwhich the disease status of individuals was objec-tively assigned using defined criteria. Complete oph-thalmologic examinations, visual field testing, opticnerve head photographs, and venous blood sampleswere obtained. Family members were genotyped us-ing polymerase chain reaction amplification of mic-rosatellite polymorphic markers. Linkage analysiswas performed and lod scores were calculated. Hap-lotype transmission data were analyzed.● RESULTS: A total of 20 subjects in three successivegenerations agreed to participate in the study. This

included samples from eight affected subjects, oneglaucoma suspect, one normal individual, and twospouses in generations II and III, and an additionaleight individuals in generation IV. The phenotype inthis family appears to be variable, with onset of visualfield loss in middle age, followed by modest elevationof intraocular pressure and progression of the diseasein older individuals. Linkage was established with agroup of DNA markers located in the 8q23 region.A lod score value of 3.61 was obtained using markerD8S1471. Three other markers from the same re-gion gave lod score values of over 3.0. Haplotypetransmission data identified two recombinationevents that placed the gene in a 6.3-cM regionflanked by D8S1830 and D8S592. The disease-bearing haplotype was inherited by eight affectedsubjects and three glaucoma suspects.● CONCLUSION: We present evidence for a thirdadult-onset primary open-angle glaucoma locus(GLC1D) on chromosome 8q23. The genetic heter-ogeneity of adult-onset glaucoma is evident from themultiplicity of chromosomal loci associated with thisdisease. (Am J Ophthalmol 1998;126:17–28.© 1998 by Elsevier Science Inc. All rights reserved.)

G LAUCOMA IS THE SECOND MOST FREQUENT

cause of irreversible blindness in the UnitedStates, with over 1.5 million affected indi-

viduals.1 It is also one of the three major causes ofblindness worldwide.2 Glaucoma refers to a group ofdiseases characterized by irreversible loss of retinalganglion cells. It has been classified according to the

Accepted for publication Dec 5, 1997.From the Molecular Ophthalmic Genetics Laboratory, Surgical

Research Center, Department of Surgery, University of ConnecticutHealth Center, Farmington, Connecticut (Drs Trifan, Stoilova, Alozie,and Sarfarazi); The Johns Hopkins Center for Hereditary Eye Diseases,The Wilmer Eye Institute, Baltimore, Maryland (Drs Traboulsi,Nguyen, and Raja); and The Cleveland Clinic Eye Institute, Cleveland,Ohio (Dr Traboulsi). This work was supported by grants from theNational Eye Institute, Bethesda, Maryland (EY-09947, to Dr Sar-farazi); the International Glaucoma Association, London, UK (IGA-G249, to Dr Sarfarazi); and University of Connecticut General ClinicalResearch Center, Farmington, Connecticut (M01-RR-06192, to DrSarfarazi); and by a grant from the American Federation for AgingResearch, New York, New York (to Dr Traboulsi).

Reprint requests to Mansoor Sarfarazi, PhD, Surgical ResearchCenter, Department of Surgery, University of Connecticut HealthCenter, 263 Farmington Ave, Farmington, CT 06030-1110; fax: (860)679-7524; e-mail: msarfara@Cortex.UCHC.Edu

© 1998 BY ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED.0002-9394/98/$19.00 17PII S0002-9394(98)00073-7

anatomy of the anterior chamber (open angle,closed angle), age of onset (infantile, juvenile,adult) and etiology (primary, secondary).3 A defin-itive classification of glaucoma will be possible onlywhen all of the etiologic factors, including geneticloci and contributing risk factors, have been iden-tified.

Clinical findings common to the different formsof glaucoma include enlargement of the cup of theoptic nerve head secondary to loss of optic nerveaxons, characteristic visual field loss, and frequentelevation of the intraocular pressure. A large pro-portion of patients with primary open-angle glau-coma, however, have intraocular pressure within astatistically normal range.

A positive family history of glaucoma is a majorrisk factor for developing the disorder.4–6 Glaucomamay be inherited in a X-linked,7 autosomal reces-sive,8 autosomal dominant,9–11 or multifactorial6,12

fashion. It appears, however, that primary open-angle glaucoma is caused predominantly by domi-nant genes in kindreds studied to date. In contrastto the less common juvenile-onset primary open-angle glaucoma, which usually manifests itself inlate childhood and early adulthood,13 with rapidprogression of visual field loss, poor response tomedical treatment,14,15 and moderately severe ele-vation of intraocular pressure, the adult-onset formof primary open-angle glaucoma (or chronic open-angle glaucoma) has a milder clinical course that isusually associated with mild to moderate elevationof intraocular pressure and a good response tomedical therapy.16,17 The locus for juvenile-onsetprimary open-angle glaucoma (GLC1A) maps tochromosome 1q21-q31,18 and the disorder is causedby mutations in the trabecular meshwork-inducedglucocorticoid response protein (TIGR) gene.19 Werecently mapped a chronic open-angle glaucomalocus (GLC1B) to the 2cen-2q13 region in a groupof families with low to moderately elevated intraoc-ular pressure.20 A second locus for chronic open-angle glaucoma (GLC1C) has been localized to the3q21-q24 region using linkage data from one NorthAmerican family.21

We present clinical, linkage, and haplotype trans-mission data supporting the assignment of a newgenetic locus, GLC1D, for chronic open-angle glau-coma to the 8q23 region.

METHODS

A LARGE KINDRED (REFERRED TO HERE AS FAMILY 1400)

has been ascertained through the registry of theJohn Hopkins Center for Hereditary Eye Diseases(Figure 1). The proband (III-1) in this pedigree wasdiagnosed with primary open-angle glaucoma at theage of 55 years, and a family history of glaucoma wasestablished. Consent was obtained to contact otherfamily members for their participation in the study.The study protocol and consent forms were ap-proved by the John Hopkins Institutional ReviewBoard and by the Joint Committee for ClinicalInvestigation. The available affected and unaffectedfamily members were invited to undergo a completeocular examination, visual field testing, and opticnerve head photography. Complete ophthalmologicexamination included applanation tonometry andrefraction. Visual field testing was done usingthreshold-related automated static perimetry (24-2program on the Humphrey Visual Field Analyzer).Slit-lamp examination was performed, and grossexternal, corneal, lenticular, and vitreal abnormal-ities were noted. The anterior chamber depth wasassessed and inspected for iridocorneal adhesions.Intraocular pressure measurements were taken twiceby two of us (E.I.T. and R.N., and/or E.I.T. andS.R.) and averaged. Direct and indirect ophthal-moscopy were performed, and optic cup/disk ratio,the presence of nerve fiber layer hemorrhages, andnotching of the optic nerve head neural rim wererecorded. Stereoscopic photographs of the opticnerve head were taken and evaluated for anychanges compatible with glaucoma.

Patients were diagnosed with chronic open-angleglaucoma when they fulfilled at least two of thefollowing criteria: (1) an intraocular pressure greaterthan or equal to 22 mm Hg in either eye; (2) visualfield defects in either eye that were compatible withthose characteristic of glaucoma as defined in the studyof Katz and associates22 (Figure 2); and (3) optic nervehead configuration compatible with glaucoma in ei-ther eye or significant asymmetry of cup/disk ratio(Figure 3). Subjects were considered chronic open-angle glaucoma suspects when they fulfilled only oneof the above criteria. All three parameters had to benormal for an individual to be considered unaffected atthe time of the study. In determining glaucomatous

AMERICAN JOURNAL OF OPHTHALMOLOGY18 JULY 1998

visual field loss, we examined the cross-meridional andglobal index analytic algorithms of each field. A visualfield was classified as abnormal when either the cor-rected pattern standard deviation was less than 2% orthe Glaucoma Hemifield Test was abnormal. Thisprovided a relatively high sensitivity and specificity fordetecting glaucomatous changes in all individuals.Borderline results on the Glaucoma Hemifield Testwere not considered abnormal.

Venous blood samples were collected from a totalof 20 subjects. DNA was extracted according to themethod previously described by Moore.23 Highly

polymorphic short tandem repeat polymorphismswith a firm location on genetic linkage maps wereamplified using multiplex polymerase chain reac-tion. The primer pairs were purchased from Re-search Genetics Inc, Huntsville, Alabama. Addi-tional information on bands, allele sizes, relativeorder, genetic distances between adjacent loci, andamplification conditions, when available, were ob-tained from the Genome Data Base (GDB) (http://gdbwww.gdb.org/), Genethon,25,26 Utah Marker De-velopment Group,24 Cooperative Human LinkageCenter27–29 (CHLC), and the Center for Med-

FIGURE 1. Pedigree of family 1400. Genotypic data for 10 short tandem repeat polymorphism markers from the8q23 region are shown below each family member. Solid bars represent the inherited affected chromosome; open barsrepresent the normal chromosome; the thin black line represents the uninformative portion of the chromosome. Notethat individual II-1 is recombinant for D8S592, whereas subject IV-1 is recombinant for D8S556 and D8S1830.Subjects II-3 and IV-6 were not available for this study. Solid squares and circles 5 affected, open symbols 5unaffected, checkered symbols 5 glaucoma-suspect family members.

GLC1D LOCUS ON 8q23Vol. 126, No. 1 19

ical Genetics at the Marshfield Medical Researchand Education Foundation (http://www.marshmed.org/genetics).

Polymerase chain reaction was performed in 25 mlfinal reaction volume containing 100 ng genomic

DNA, 0.25 mM of each primer, 200 mM of eachdNTP, 50 mM KCl, 10 mM Tris (pH 8.4), 1.5 mMMgCl2, 0.01% gelatin and 0.5 U Taq polymerase(AmpliTaq; Perkin Elmer, Foster City, California).Amplifications were carried out in a GeneAmp 9600

FIGURE 2. (Top) Subject II-1. Goldmann visual fields. Only small central and peripheral islands of vision remain.(Bottom left) Subject II-8. Moderate visual field loss in both eyes. Her optic nerve heads photographs are shown in Figure3, top, left and right. (Bottom right) Subject III-5. Mild visual field loss in an arcuate pattern in the left eye. His optic nerveheads are shown in Figure 3, bottom, left and right. Note the deeper and slightly larger cup of the left nerve.

AMERICAN JOURNAL OF OPHTHALMOLOGY20 JULY 1998

thermocycler (Perkin Elmer) under the following con-ditions: initial denaturation for 3 minutes at 94 C, 30to 35 cycles of 10-second denaturation at 94 C, and30-second annealing at 55 C to 60 C, followed by afinal extension at 72 C for 3 minutes. The amplifiedproducts were separated by electrophoresis on 7%

denaturing polyacrylamide gels, silver stained,30 pho-tographed, and subsequently genotyped for each shorttandem repeat polymorphism.

Data obtained from the genotyping of short tandemrepeat polymorphisms were entered into a dedicatedcomputer program (Sarfarazi M, unpublished data,

FIGURE 3. (Top) Subject II-8. The right (left) and left (right) optic nerve heads. The cup/disk ratio of the right eyeis larger than that of the left eye. (Bottom) Subject III-5. The cup of the optic nerve head of the left eye (right) islarger and deeper than that of the right eye (left).

GLC1D LOCUS ON 8q23Vol. 126, No. 1 21

1986), checked for mendelian errors or inconsisten-cies, and subsequently prepared for the LINKAGEprogram.31,32 Two-point linkage was calculated withthe MLINK module of the LINKAGE package. Allcalculations were made under the assumption of anautosomal dominant mode of inheritance with anaffected gene frequency of 0.0001. No assumptionswere made for incomplete penetrance, mutation,and/or phenocopy. Because of a very limited numberof spouses that married into this family and lack of anyother reliable gene frequencies for the population inwhich this pedigree has been ascertained, equal allelefrequencies were used for all of the DNA markers.Because of the relatively late onset of the disease andlack of reliable penetrance ratios, and in order toeliminate the effect of possible incomplete penetrance,the lod score calculations were repeated for the af-fected meioses only. Haplotypes for the entire set ofmarkers were constructed and evaluated for genotypeinconsistency and inheritance from both parentalgenerations.

RESULTS

A SUMMARY OF CLINICAL INFORMATION ON AFFECTED

patients and glaucoma suspects in this family is

provided in Table 1. A total of 18 potentiallyinformative individuals agreed to participate in thestudy, of whom eight were firmly diagnosed ashaving chronic open-angle glaucoma. Two addi-tional members of this family did not agree toparticipate in the study. Family members IV-1, IV-2,IV-3, IV-4, and IV-5 were examined only once in1992 and refused follow-up examinations despiterepeated communication attempts and warningsthat they may have chronic open-angle glaucoma.Individual II-4 is a 76-year-old woman with intraoc-ular pressure readings in the low to mid-teens,normal optic nerve head configuration with a cup/disk ratio of 0.3 in each eye, mild cataracts, andabnormal visual fields. She was classified as a glau-coma suspect for the purpose of adhering to the setof criteria used in this study. We believe that shedoes not have chronic open-angle glaucoma. Herfield defects may be caused by cataracts. She wasconsidered as having an unknown disease status inthe linkage calculations. On the other hand, indi-vidual IV-1 has elevated intraocular pressure inboth eyes and slightly asymmetric cup/disk ratiowith a normal optic nerve head and normal visualfields. We believe that he is affected, but he wasclassified as a glaucoma suspect. His disease status

TABLE 1. Clinical Findings in Affected and Glaucoma Suspect Individuals

Subject,

Sex

Highest IOP

(Right/Left)

(mm Hg)

Optic

Nerve

Head

Humphrey 24-2

Visual Field Treatment

Affected

II-1, F 26/26 A A (advanced) Betaxolol, pilocarpine,

trabeculoplasty

II-6, M 21/22 A A None

II-8, F 26/24 A A Timolol maleate

II-9, M 20/23 A A None

III-1, F 20/21 A A Timolol maleate

III-3, F 22/22 A A Betaxolol hydrochloride

III-4, M 18/18 A A None

III-5, M 20/18 A A None

Glaucoma Suspects

II-4, F 13/13 N A None

IV-1, M 24/25 S/A N None

IV-2, M 18/19 S N None

IV-4, M 20/18 B/S N None

IOP 5 intraocular pressure; A 5 abnormal; B 5 borderline; N 5 normal; S 5 suspicious.

AMERICAN JOURNAL OF OPHTHALMOLOGY22 JULY 1998

was also considered unknown in the linkage analy-sis.

We first tested for the linkage of this family to thethree previously identified loci of GLC1A on 1q,GLC1B on 2q and GLC1C on 3q. Substantialevidence for exclusion (that is, lod score values of22.00 or less) were obtained for all DNA markersstudied from these three regions (data not shown).Consequently, a random genome search was initi-ated to identify the chromosomal location of theputative gene responsible for glaucoma in this fam-ily. After genotyping a large number of short tan-dem repeat polymorphism markers, alleles ofD8S1471 were found to segregate with the diseasephenotype in this family. A total of 30 additionalDNA markers from the same chromosomal regionwere studied, of which 25 provided positive lodscore values. As shown in Table 2, if individual II-4is considered unaffected, lod scores (abbreviated asZ) of over 3.0 were obtained at u 5 0.00 for fiveshort tandem repeat polymorphism markers ofD8S257 (Z 5 3.01), D8S1132 (Z 5 3.31), D8S1830(Z 5 3.01), D8S1471 (Z 5 3.61), and D8S588 (Z 53.14), with confidence intervals that varied between0.00 to 0.16 cM. Eight additional short tandemrepeat polymorphism markers also provided lodscore values between Z 5 2.11 to Z 5 2.75. Fourshort tandem repeat polymorphism markers showed

lod score values between 1.5 and 2.0, and five othersprovided Z values between 1.0 and 1.5. A summaryof lod score values for the most informative group ofDNA markers in this family is provided in Table 2.Lod score calculations were repeated assuming anunknown status for II-4 and either an affected orunknown status for IV-1. The values remainedsignificantly elevated for D8S1471 (Z 5 3.61 if II-4is unknown and IV-1 affected, and Z 5 3.31 if II-4is unknown and IV-1 unknown). Table 3 gives theremainder of the Z values for all markers under thislatter set of assumptions. Lod score calculationswere repeated using the affected meioses only inorder to eliminate possible bias because of incom-plete penetrance and/or to the contribution ofnormal subjects who may be asymptomatic genecarriers. As expected, the lod score values werereduced but still remained significantly positive.

In order to identify a flanking region that wouldharbor this newly identified locus, we constructed acomplete haplotype for all of the short tandemrepeat polymorphism markers in this region. Weidentified a telomeric recombination betweenD8S592 and all other markers distal to it in aglaucomatous subject carrying the affected haplo-type (II-1; Figure 1). Another centromeric recom-bination was also detected in subject IV-1 betweenD8S1830 and all other markers proximal to it.

TABLE 2. Two-point LOD Scores Between the GLC1D Locus and 14 DNA Markers on 8q23*

Marker

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45

D8S257 3.01 2.74 2.46 2.16 1.85 1.52 1.16 0.81 0.46 0.18

D8S556 2.11 1.91 1.71 1.5 1.27 1.05 0.81 0.58 0.35 0.16

D8S1132 3.31 3.02 2.72 2.39 2.05 1.69 1.31 0.91 0.53 0.22

D8S1830 3.01 2.74 2.46 2.16 1.85 1.52 1.17 0.82 0.48 0.21

D8S1122 2.41 2.19 1.95 1.7 1.44 1.16 0.87 0.56 0.27 0.07

AFM200VC7 2.75 2.47 2.17 1.86 1.53 1.19 0.84 0.51 0.23 0.06

D8S1471 3.61 3.3 2.97 2.62 2.26 1.86 1.45 1.02 0.59 0.23

D8S1139 2.41 2.19 1.95 1.7 1.44 1.16 0.87 0.56 0.27 0.07

GATA133C08 2.41 2.19 1.95 1.7 1.44 1.16 0.87 0.56 0.27 0.07

D8S1465 2.41 2.19 1.95 1.7 1.44 1.16 0.87 0.56 0.27 0.07

D8S547 2.11 1.91 1.7 1.47 1.24 0.99 0.72 0.46 0.22 0.05

D8S1779 2.11 1.91 1.7 1.49 1.27 1.04 0.82 0.6 0.38 0.18

D8S588 3.14 2.83 2.5 2.16 1.8 1.42 1.03 0.64 0.3 0.07

D8S592 2` 0.03 0.27 0.34 0.34 0.3 0.22 0.14 0.06 0.02

*The data in this table assume that the disease status of subjects II-4 and of IV-1 is unknown.

GLC1D LOCUS ON 8q23Vol. 126, No. 1 23

Although subject IV-1 has elevated intraocularpressure and large and slightly asymmetric cup/diskratio of RE, 0.6 and LE, 0.7, he has no apparentvisual field loss. We have not been able to reassesshis current status because he has consistently beenunavailable for further clinical and ophthalmologicexamination. If he is affected, and we believe thathe is, then this centromeric recombination could beused to narrow the region of interest. The precisecytogenetic map of the short tandem repeat poly-morphism markers used in this study is not avail-able. However, both flanking markers D8S556 andD8S588 have been mapped to the 8q23 region(Figure 4). Therefore, it is likely that this newchronic open-angle glaucoma locus is on the long

arm of chromosome 8 and is located within thisregion of 8q23.

DISCUSSION

THE PHENOTYPE OF PRIMARY OPEN-ANGLE GLAU-

coma results from a genetically heterogeneous groupof ocular disorders. It is characterized by a slowlyprogressive optic neuropathy with visual field lossand, in more than two thirds of cases, elevation ofthe intraocular pressure. The severe and least com-mon juvenile-onset form has been mapped to the1q21-q31 region18 and subsequently confirmed byother groups.33–35 Genetic heterogeneity has also

TABLE 3. LOD Scores*

u

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45

D8S257 AFF 2` 1.46 1.51 1.41 1.25 1.04 0.81 0.55 0.31 0.12

UK 2.71 2.46 2.21 1.93 1.64 1.34 1.03 0.71 0.41 0.17

D8S556 Aff 2` 0.66 0.78 0.78 0.71 0.61 0.49 0.36 0.22 0.10

UK 1.83 1.66 1.48 1.30 1.11 0.91 0.71 0.51 0.32 0.15

D8S1132 Aff 2` 1.74 1.76 1.64 1.45 1.21 0.94 0.66 0.38 0.16

UK 3.01 2.74 2.46 2.16 1.85 1.51 1.17 0.81 0.48 0.20

D8S1830 Aff 2` 1.46 1.51 1.41 1.25 1.04 0.81 0.57 0.34 0.15

UK 2.71 2.46 2.21 1.93 1.65 1.35 1.04 0.73 0.44 0.19

D8S1122 Aff 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

UK 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

AFM200VC7 Aff 2.75 2.47 2.17 1.86 1.53 1.19 0.84 0.51 0.23 0.06

UK 2.45 2.20 1.92 1.63 1.33 1.02 0.71 0.42 0.19 0.05

D8S1471 Aff 3.61 3.30 2.97 2.62 2.26 1.87 1.45 1.03 0.61 0.26

UK 3.31 3.02 2.72 2.39 2.05 1.69 1.31 0.91 0.53 0.22

D8S1139 Aff 2.11 1.91 1.70 1.47 1.23 0.98 0.72 0.46 0.22 0.05

UK 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

GATA133C08 Aff 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

UK 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

D8S1465 Aff 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

UK 2.11 1.91 1.69 1.47 1.23 0.98 0.72 0.46 0.22 0.05

D8S547 Aff 1.81 1.63 1.44 1.24 1.03 0.81 0.58 0.36 0.16 0.04

UK 1.81 1.63 1.44 1.24 1.03 0.81 0.58 0.36 0.16 0.04

D8S1779 Aff 2.11 1.91 1.71 1.50 1.29 1.07 0.86 0.64 0.43 0.22

UK 1.81 1.63 1.45 1.27 1.08 0.90 0.71 0.53 0.35 0.17

D8S588 Aff 3.13 2.83 2.50 2.16 1.80 1.42 1.02 0.64 0.30 0.07

UK 2.83 2.55 2.25 1.93 1.60 1.25 0.90 0.54 0.25 0.06

D8S522 Aff 1.93 1.81 1.67 1.50 1.31 1.10 0.88 0.66 0.44 0.21

UK 1.63 1.54 1.41 1.27 1.10 0.92 0.74 0.55 0.36 0.17

*LOD scores when subject II-4 is coded as unknown and when subject IV-1 is coded as either affected (Aff; first line for each marker) or

unknown (UK; second line for each marker).

AMERICAN JOURNAL OF OPHTHALMOLOGY24 JULY 1998

been reported for this form of primary open-angleglaucoma.36–38 In a recently published paper, Stoneand associates19 reported three distinct mutations inthe TIGR gene that are responsible for this pheno-type. We have also identified a new mutation in theTIGR gene in one such family from Edinburgh.39

For the more frequent form of adult-onset primaryopen-angle glaucoma, the first locus (GLC1B) hasbeen mapped to the 2cen-q13 region20,38 and thesecond locus (GLC1C) to the 3q21-q24 region.21,38

After excluding genetic linkage in this NorthAmerican family to the loci for GLC1A, GLC1B,and GLC1C, and following a general positionalmapping strategy, we identified a new locus for

chronic open-angle glaucoma on the long arm ofchromosome 8. This mapping is supported by lodscore values of over 3.0 for five short tandem repeatpolymorphism markers located on the 8q23 region,and a maximum Z value of 3.61 with markerD8S1471. When we used only the affected meioses,a maximum Z value of 1.51 was obtained withmarkers D8S1122 and GATA133C08. We realizethe difficulties in assigning a disease status in mem-bers of families with chronic open-angle glaucoma;hence, we adopted an objective and simple set ofdiagnostic criteria. We adhered to these criteria inall individuals. This resulted in the assignment ofsuspect status to family member II-4, who most

FIGURE 4. Chromosome 8 map. The order of short tandem repeat polymorphisms from the q23 region ofchromosome 8. On the right side, the Genethon markers and their genetic distances are from Dib and associates26;on the left side are the Cooperative Human Linkage Center (CHLC) and Utah markers. All markers shown wereused in our study. On the left side of the figure are the observed recombinations in subjects IV-1 (1) and II-1 (2).The inherited affected region of the chromosome is represented by a solid black bar; the normal region of thechromosome is represented by an open bar; the uninformative portion is represented by a thin line.

GLC1D LOCUS ON 8q23Vol. 126, No. 1 25

likely did not have glaucoma, and to a youngerfamily member, subject IV-1, who was most likelyaffected, as evidenced by the elevation of intraocu-lar pressure in both eyes. Such rigid adherence tothe diagnostic criteria could possibly result in un-derdiagnosis or overdiagnosis of the disease in someindividuals, but it assures the objectivity that isessential for purposes of lod score calculations.When subject II-4 is considered normal and subjectIV-1 is coded as affected, the lod score increases to3.91 for marker D8S1471.

Construction, inspection, and transmission ofaffected-bearing haplotypes identified recombina-tion events that likely place this new locus in a6.3-cM region flanked by markers D8S1830 andD8S592. The cytogenetic localization of thesemarkers is probably in the 8q23 region, thus placingthis new locus for chronic open-angle glaucoma inthe same region of this chromosome.

The glaucoma phenotype in this family is char-acterized by an optic neuropathy and visual field lossthat appear to start in the third or fourth decade oflife. Patients have a mild to modest elevation ofintraocular pressure. The optic nerve head has a 0.3to 0.7 cup with some asymmetry in most patientsbetween the two eyes. The clinical course of one ofthe oldest patients (II-1) was that of relentlessprogression of the field loss and poor control of amildly elevated intraocular pressure with medica-tions and trabeculoplasty. Except for subject IV-1,younger family members have the lowest intraocularpressure measurements. Some older family members(III-4 and III-5) have abnormal visual fields andoptic nerve head configuration consistent with glau-coma but have normal intraocular pressures. Thesepatients were classified as affected, based on thedefining criteria, but were not started on medica-tions. All affected individuals in generation II hadelevated intraocular pressure. Hence, a rise in in-traocular pressure appears to be a late finding in thisfamily and appears to be preceded by the opticneuropathy and visual field loss.

The disease status of some of the individuals inthis kindred is uncertain for various reasons. Thefour members of the youngest generation haveinherited the affected chromosome but have onlybeen examined once, 4 to 5 years ago, and refuse tobe reevaluated. Individuals IV-1, IV-2, and IV-4

were considered glaucoma suspects based on abnor-mal visual field or optic nerve head configurationcompatible with glaucoma. Because these individu-als do not have any affected offspring and becausethere is no proof of incomplete penetrance in thiskindred, their clinical status as asymptomatic genecarriers needs to be evaluated when the causativegene in this family is identified.

A search for genes and/or expressed sequence tags(ESTs) physically mapped between the two markersD8S556-D8S592 reveals 41 unidentified transcripts,two mRNAs for an open reading frame (KIAA0003and KIAA0103), the precursor gene for Syndecan 2,and the early growth response gene–alpha (EGR-a).40 Of these, Syndecan 2 is a good candidate genefor this newly identified primary open-angle glau-coma locus. Syndecans are a family of integralmembrane proteoglycans that participate in cell-matrix interactions and the modulation of growthfactor response. Syndecan 2 (SYND2) precursor(also known as fibroglycan or heparin sulfate pro-teoglycan [HSPG] core protein) was first character-ized as the core protein of the cell surface-associatedheparin sulfate proteoglycans. SYND2 is a single48-kd gene that maps to chromosome 8q23.41

SYND2 is expressed in a variety of cells, includingvascular smooth muscle cells,42 tumor cells,43 andoligodendrocytes.44 Interestingly, in oligodendro-cytes (cells responsible for myelin sheath of theaxons), the expression of Syndecan 2 increases asthe cells differentiate from proliferative late progen-itors to postmitotic mature cells.44 Therefore, itcould be speculated that a mutation of the SYND2gene would alter the adherence properties of adultoligodendrocytes, leading to a demyelinization ofthe optic nerve fibers, rendering them more sensi-tive to other stress factors, such as elevated intraoc-ular pressure. It could further be hypothesized thatchanges in the surface adherence properties oftrabecular meshwork cells could over time lead toan increase in the intraocular pressure through an asyet unknown mechanism. Under this hypothesis,Syndecan 2 becomes an excellent candidate genefor chronic open-angle glaucoma phenotype in thefamily presented here.

Another positional candidate is EGR-a. Radia-tion hybrid (Genbridge 4) mapping at the White-head Institute (http://www-genome.wi.mit.edu/

AMERICAN JOURNAL OF OPHTHALMOLOGY26 JULY 1998

cgi-bin/contig/phys_map) placed this gene insidethe 6.3-cM region (EST119377) where the newlocus for chronic open-angle glaucoma is located.Early growth response gene-alpha belongs to thesubfamily of zinc finger proteins that bindguanosine-rich regions of DNA. Functional analysisof EGR-a demonstrates that it could bind to, andstimulate transcription from, a basic transcriptionelement consensus DNA sequence. Basic transcrip-tion element is a naturally occurring binding se-quence in the promoter region of some cytochromeP450 members and functions as a transcription mod-ulator. If EGR-a is the physiologic regulator of theCYP1B1 gene, which has recently been shown to beresponsible for primary congenital glaucoma,38,45

then it is possible that mutations in EGR-a may alsocause chronic open-angle glaucoma in the familystudied here.

In conclusion, we have identified a new chromo-somal location in a large North American chronicopen-angle glaucoma family that maps to the 8q23region. We have further narrowed down the mini-mum critical region of this locus (that is, GLC1D)to about 6.3 cM and identified two potential can-didate genes that may prove to be involved in theetiology of the adult-onset primary open-angle glau-coma phenotype in this family.

ADDENDUM

While this paper was in press, we identified andpublished46 a new locus (GLC1E) for adult-onsetprimary open-angle glaucoma in the 10p15-p14region. This is the first locus for a typical “lowtension” form of primary open-angle glaucoma.

REFERENCES

1. Quigley HA. Open-angle glaucoma. N Engl J Med 1993;328:1097–1106.

2. Foster A, Johnson GJ. Magnitude and causes of blindness inthe developing world. Int Ophthalmol 1990;14:135–140.

3. Shields M, Ritch R, Krupin T. Classification of the glau-comas. In: Ritch R, Shields BM, Krupin T, editors. Theglaucomas. Volume 2. St Louis: Mosby, 1996:717–725.

4. Shin DH, Becker B, Kolker AE. Family history in primaryopen-angle glaucoma. Arch Ophthalmol 1977;95:598–600.

5. Miller R. Genetics of glaucoma and family studies. TransOphthalmol Soc UK 1978;98:290–292.

6. Lichter PR. Genetic clues to glaucoma’s secrets: Jackson

memorial lecture, part 2. Am J Ophthalmol 1994;117:706–727.

7. Netland PA, Wiggs JL, Dreyer EB. Inheritance of glaucomaand genetic counseling of glaucoma patients. Int Ophthal-mol Clin 1993;33:101–120.

8. Beiguelman B, Prado D. Recessive juvenile glaucoma.J Genet Hum 1963;12:53–54.

9. Crombie AL, Cullen JF. Hereditary glaucoma: occurrencein five generations in an Edinburgh family. Br J Ophthal-mol 1964;48:143–147.

10. Lee DA, Brubaker RF, Hrusca L. Hereditary glaucoma: areport of two pedigrees. Ann Ophthalmol 1985;17:739–741.

11. Bennett SR, Alward WL, Folberg R. An autosomal domi-nant form of low-tension glaucoma. Am J Ophthalmol1989;108:238–244.

12. Jay B, Peterson G. The genetics of simple glaucoma. TransOphthalmol Soc U K 1970;90:161–171.

13. Goldwyn R, Waltman SR, Becker B. Primary open-angleglaucoma in adolescent and young adults. Arch Ophthal-mol 1970;84:579–582.

14. Ellis OH. The etiology, symptomatology and treatment ofjuvenile glaucoma. Am J Ophthalmol 1948;31:1589–1596.

15. Johnson AT, Drack AV, Kwitek AL Cannon RL, StoneEM, Alward WLM. Clinical features and linkage analysis ofa family with autosomal dominant juvenile glaucoma.Ophthalmology 1993;100:524–529.

16. Vogel R, Crick RP, Newson RB, Shipley M, Blackmore H,Bulpitt CJ. Association between intraocular pressure andloss of visual field in chronic simple glaucoma. Br JOphthalmol 1990;74:3–6.

17. Jay JL, Murdoch JR. The rate of visual field loss inuntreated primary open-angle glaucoma. Br J Ophthalmol1993;77:176–178.

18. Sheffield VC, Stone EM, Alward WLM, et al. Geneticlinkage of familial open-angle glaucoma to chromosome1q21-q31. Nature Genet 1993;4:47–50.

19. Stone EM, Fingert JH, Alward WLM, et al. Identificationof a gene that causes primary open-angle glaucoma. Science1997;275:668–670.

20. Stoilova D, Child A, Trifan OC, Crick RP, Coakes RL,Sarfarazi M. Localization of a locus (GLC1B) for adult-onset primary open-angle glaucoma to the 2cen-ql3 region.Genomics 1996;36:142–150.

21. Wirtz MK, Samples JR, Kramer PL, et al. Mapping a genefor adult-onset primary open-angle glaucoma to chromo-some 3q. Am J Hum Genet 1997;60:296–304.

22. Katz J, Tielsch JM, Quigley HA, Sommer A. Automatedperimetry detects visual field loss before manual Goldmannperimetry. Ophthalmology 1995;102:21–26.

23. Moore D. Preparation of genomic DNA. In: Ausubel RH,Brent R, Kingston RE, et al, editors. Current protocols inmolecular biology. Sections 2.1.1–2.1.8. New York: Wiley,1994.

24. Utah Marker Development Group. A collection of orderedtetranucleotide-repeat markers from the human genome.Am J Hum Genet 1995;57:619–628.

25. Gyapay G, Morissette J, Vignal A, et al. The 1993–94Genethon human genetic linkage map. Nature Genet1994;7(suppl):246–339.

26. Dib C, Faure S, Fizames C, et al. The final version of theGenethon human genetic linkage map. Nature 1996;380:152–154.

27. Murray JC, Buetow KH, Weber JL, et al. A comprehensive

GLC1D LOCUS ON 8q23Vol. 126, No. 1 27

human linkage map with centimorgan density. Science1994;265:2049–2054.

28. Sheffield VC, Weber JL, Buetow KH, et al. A collection oftri- and tetranucleotide repeat markers used to generatehigh quality, high resolution human genomewide linkagemaps. Hum Mol Genet 1995;4:1837–1844.

29. Gastier JM, Pulido JC, Sunden S, et al. Survey of trinucle-otide repeats in the human genome: assessment of theirutility as genetic markers. Hum Mol Genet 1995;4:1829–1836.

30. Bassam BJ, Caetano-Annoles G. Silver staining of DNA inpolyacrylamide gels. Protocols Biotechnol 1993;42:181–188.

31. Lathrop GM, Lalouel JM. Easy calculations of LOD scoresand genetic risks on small computers. Am J Hum Genet1984;36:460–465.

32. Schaffer AA, Gupta SK, Shriram K, Cottingham RW Jr.Avoiding recomputation in genetic linkage analysis. HumHered 1994;44:225–237.

33. Richards JE, Lichter PR, Boehnke M, et al. Mapping of a genefor autosomal dominant juvenile-onset open-angle glaucomato chromosome 1q. Am J Hum Genet 1994;54:62–70.

34. Meyer A, Valtot F, Bechetoille A, et al. Linkage betweenjuvenile glaucoma and chromosome 1q in two Frenchfamilies. Comp Rend Acad Sci (Paris) 1994;317:565–570.

35. Wiggs JL, Haines JL, Paglinauan C, Fine A, Sporn C, LouD. Genetic linkage of autosomal dominant juvenile glau-coma to 1q21-q31 in three affected pedigrees. Genomics1994;21:299–303.

36. Graff C, Urbak SF, Jerndal T, Wadelius C. Confirmation oflinkage to 1q21-31 in a Danish autosomal dominantjuvenile-onset glaucoma family and evidence of geneticheterogeneity. Hum Genet 1995;96:285–289.

37. Wiggs JL, DelBono EA, Schuman JS, Hutchinson BT,Walton DS. Clinical features of five pedigrees geneticallylinked to the juvenile glaucoma locus on chromosome1q21-q31. Ophthalmology 1995;102:1782–1789.

38. Sarfarazi M. Recent advances in molecular genetics ofglaucomas. Hum Mol Genet 1997;6:1667–1677.

39. Stoilova D, Child A, Brice G, Crick RP, Fleck BW,Sarfarazi M. Identification of a new “TIGR” mutation in afamily with juvenile-onset primary open-angle glaucoma.Ophthalmol Genet 1997;18:109–118.

40. Schuler GD, Boguski MS, Stewart EA, et al. A gene map ofthe human genome. Science 1996;274:540–546.

41. Marynen P, Zhang J, Cassiman JJ, Van den Berghe H,David G. Partial primary structure of the 48- and 90-kilodalton core proteins of cell surface-associated heparinsulfate proteoglycans of lung fibroblasts: prediction of anintegral membrane domain and evidence for multipledistinct core proteins at the cell surface of human lungfibroblasts. J Biol Chem 1989;264:7017–7024.

42. Cizmeci-Smith G, Langan E, Youkey J, Showalter LJ, CareyDJ. Syndecan-4 is a primary-response gene induced by basicfibroblast growth factor and arterial injury in vascular smoothmuscle cells. Arterioscler Thromb Vasc Biol 1997;17:172–180.

43. Itano N, Oguri K, Nagayasu Y, et al. Phosphorylation of amembrane-intercalated proteoglycan, syndecan-2, ex-pressed in a stroma-inducing clone from a mouse Lewis lungcarcinoma. Biochem J 1996;315:925–930.

44. Bansal R, Kumar M, Murray K, Pfeiffer SE. Developmentaland FGF-2-mediated regulation of syndecans (1-4) and glypi-can in oligodendrocytes. Mol Cell Neurosci 1996;7:276–288.

45. Stoilov I, Akarsu AN, Sarfarazi M. Identification of threedifferent truncating mutations in cytochrome P4501B1(CYP1B1) as the principal cause of primary congenital glau-coma (Buphthalmos) in families linked to the GLC3A locuson chromosome 2p21. Hum Mol Genet 1997;6:641–647.

46. Sarfarazi M, Child A, Stoilova D, et al. Localization of thefourth locus (GLC1E) for adult-onset primary open-angleglaucoma to the 10p15-p14 region. Am J Hum Genet1998;62:641–652.

Authors InteractivetWe encourage questions and comments regarding this article via theInternet on Authors Interactivet at http://www.ajo.com/ Questions, com-ments, and author responses are posted.

AMERICAN JOURNAL OF OPHTHALMOLOGY28 JULY 1998