Genome-Wide Scan of Exfoliation Syndrome

Abstract

purpose. Exfoliation syndrome (XFS) is an age-related ocular condition that is characterized by the accumulation of fibrillogranular extracellular material in intra- and extraocular tissues. The purpose of the present study was to identify the genetic basis of XFS in a large Finnish family.

methods. A genome-wide scan with 1000 microsatellite markers was performed in an extended family from an island in the southwestern Finnish archipelago where XFS demonstrates an autosomal dominant mode of inheritance with incomplete penetrance. Two-point linkage analyses were performed with MLINK and multipoint linkage, using the Vitesse program.

results. Five chromosomal regions with markers showing two-point LOD scores more than 1.5 was identified by using a dominant mode of inheritance for the XFS trait. The most promising locus was assigned to 18q12.1-21.33 with a maximum two-point LOD score of 3.45 and a multipoint LOD score of 4.2. Some evidence of linkage was obtained at chromosomes 2q, 17p, and 19q, which were suggested in earlier reports to be possible regions of linkage to primary open-angle glaucoma (POAG).

conclusions. The study presented herein offers a starting point to unravel the molecular background of XFS.

Exfoliation syndrome (XFS), a risk factor for development and progression of glaucoma, ^{1 2 3} was first described in 1917 by the Finnish ophthalmologist Lindberg, ⁴ who noted the accumulation of fibrillogranular extracellular debris in ocular tissues in 50% of patients with open-angle glaucoma. Later on, similar material was also detected in many extraocular tissues. ⁵ XFS debris is suggested to be composed of a complex glycoprotein–proteoglycan structure bearing epitopes of the basement membrane, epitopes of elastic fiber system, and components of elastic microfibrils. ⁶ However, the exact composition of XFS material as well as the molecular mechanism responsible for its excessive production and progressive accumulation remains unknown. Pathogenetic factors suggested to influence the abnormal matrix metabolism in the ocular tissues include growth factors, proteolytic enzymes (matrix metalloproteinases [MMPs]) and their inhibitors (tissue inhibitors of metalloproteinases [TIMPs]), anterior chamber hypoxia, and increased oxidative stress. ⁶ In total, 23 genes with different expression patterns in the anterior segment tissues of eyes with XFS have recently been reported in a microarray study. ⁷  

The prevalence of XFS varies widely among populations, being highest in Scandinavian countries ^{8 9} and Greece, ¹⁰ but it is almost absent among the Inuits. ⁸ In a population-based study in Finland, the XFS was found in 22% of the population older than 70 years. ¹¹ The prevalence increases with age and is highest in the age group between 70 and 80 years. ⁸  

Familial aggregation and the increased frequency of XFS in relatives of affected subjects compared with relatives of unaffected subjects ^{12 13} suggest an underlying genetic component. ^{14 15} The main problems with studies on the genetic background of XFS have been the asymptomatic nature of XFS and late age of onset which make it difficult to collect multigeneration families with several affected individuals for linkage and association studies. A wide variety of inheritance models have been suggested depending on the study material, ^{15 16} and, of these, the autosomal dominant mode of inheritance with incomplete penetrance has received the most support (Sotirova et al. IOVS 1999;40:ARVO Abstract 512). ^{15 17} Most of these studies investigating XFS inheritance have been based on small pedigrees making hypotheses about the inheritance model uncertain. Thus, only a few molecular genetic studies of XFS have been performed. Loss of heterozygosity (LOH) has been reported on chromosome 13 on locus 13q12.11 and on chromosome 7 on loci 7p13, 7q21.3, ¹⁸ and 7q21.11, ¹⁹ in XFS specimens of the iris and anterior capsule. The genetics of primary open-angle glaucoma (POAG) has been studied in greater detail and at least 11 candidate loci (GLC1A-GLC1K) and three predisposing genes—myocilin (MYOC), optineurin (OPTN), and WD repeat-domain 36 (WDR36)— have been identified. ²⁰ Some of these POAG loci have been excluded for XFS patients. ^{21 22 23 24}  

The purpose of the present study was to identify susceptibility loci for XFS in a large Finnish family originating from a subisolate of the Finnish population in which inheritance of XFS resembles the autosomal dominant model. 

Methods

Patient Data

The patient data in the present study were from a population-based study conducted from 1960 to 1962 on an isolated island, Kökar, in the Åland Archipelago in the southwestern Finnish territory. This island has been permanently inhabited since the 14th century. Five hundred ninety-five subjects older than 10 years (85% of the population) participated in the study. During the subsequent decades, several follow-up studies of the same population were performed; the latest in 2001 and 2002 had 183 participants. Altogether, 530 subjects older than 50 years had undergone ophthalmic examination by the end of 2002. 

For all participants, both affected and unaffected, a comprehensive ophthalmic examination was performed including both slit-lamp biomicroscopy and funduscopy, before and after dilatation. XFS was recorded without grading when a grayish central disc, with or without a peripheral band on the anterior lens capsule, and/or fibrillary material on the pupillary ruff was detected (Fig. 1) . A subject was recorded as affected (XFS-positive), if XFS was detected in at least one eye. Suspect changes such as Krukenberg's spindle and pigmentation of cornea endothelium, diffuse haze on the anterior capsule, were noted as negative, and the subject was classified as unaffected. The recorded age of the subjects was the age either when XFS was observed for the first time, or if XFS was not present, the age when the last examination was performed. Diagnosis of glaucoma was based on structural changes in the optic disc (e.g., cup-to-disc ratio ≥ 0.7, localized thinning of the rim, asymmetry ≥ 2 D between the discs), corresponding visual field defect, open anterior chamber angles, and/or intraocular pressure more than 22 mm Hg. Glaucoma without XFS was recorded as POAG and that with XFS as exfoliation glaucoma (XFG). 

The study was approved by the Ethics Committee of the Department of Ophthalmology, Central Hospital of Åland, Mariehamn, Finland, and was conducted in accordance with the Declaration of Helsinki. The subjects gave informed consent after explanation of the nature and possible consequences of the study. 

Laboratory Procedures

EDTA blood samples were collected from 183 family members who participated in the latest follow-up study. Genomic DNA was extracted from peripheral blood with the standard phenol-chloroform procedure. For mutation screening, PCR and sequencing reactions of the OPTN and MYOC genes and subsequent analyses were performed as described elsewhere. ²¹ The genome-wide map of 1000 microsatellite markers spanning an intermarker interval of ∼3 cM was based on the deCode marker map. ²⁵ Genotypes were constructed with standard procedures. ²⁵ The Pedcheck program was used for genotyping error elimination. ²⁶  

Statistical Analyses

The two-point linkage analyses under both homogeneity and heterogeneity were performed using the MLINK program of the LINKAGE package. ^{27 28} The HOMOG 3.35 program ²⁹ was used for heterogeneity testing and to calculate the proportion of families showing linkage (α). The ANALYZE program was used to conduct these analyses. ³⁰ Analyses were performed using both dominant (dom) and recessive (rec) models with high penetrance and low phenocopy rate (dom 0.001, 0.999, 0.999; rec 0.001, 0.001, 0.999) and a rare disease allele frequency (0.0001). Since XFS was prevalent in 14.5% of the Kökar population and we hypothesize that it is a complex disorder, we tested the range of disease allele frequencies accommodating up to 0.14 (0.01, 0.1, 0.14; Supplementary Table S1). Clinically unaffected family members chosen for the genome scan were scored as unaffected because their mean age at the time of the last examination was 75.3 years (range, 51–94) and because using an affected-only method would have led to substantial loss of information. However, to compare the models affected-only analyses were performed (Supplementary Table S1). The marker allele frequencies were estimated using the DOWNFREQ 2.1 program. ³¹ Regions with the most promising two-point LOD score results (LOD score, >1.0) were further examined by multipoint analysis with the Vitesse program. ^{32 33}  

Results

In the cross-sectional part of the study from 1960 to 1962, the prevalence of XFS was 18.4% in 76 subjects older than 70 years. During the period 1960 to 2002, 530 subjects older than 50 years were examined at least once. Of these, 76 were affected with XFS, 23 men, mean age 71.9 years (range, 62–90) and 53 women, mean age 72.3 years (range, 53–88). XFG was found in 24 subjects (11 men, 13 women) and POAG in 12 subjects (7 men, 5 women). Moreover, in the whole population, two women younger than 50 years were affected (one with XFS, one with XFG). 

A pedigree was constructed for all XFS-affected subjects. Seventy-five of the 78 who were XFS positive were linked in a large pedigree with 332 examined subjects, in whom the mode of inheritance seemed to resemble an autosomal dominant model. The three most recent generations had been examined several times, but the earlier ones were deceased or unable to participate when the population-based study was conducted. Within this large family, XFS was found in 22% of the examined relatives compared with a 14.5% prevalence in the whole population on the island, suggesting a genetic background. 

Among the 183 subjects in the latest follow-up study in 2001 and 2002, 28 had XFS and, of these, 9 had XFG, 7 had POAG, and 148 were unaffected family members. 

Mutation Screening

Molecular genetic analyses were started by sequencing the coding region and splice sites of the MYOC gene and the OPTN gene underlying POAG. ²⁰ The first four patients with glaucoma in the latest follow-up study, three with XFG and one with POAG, were chosen for mutation screening. Three patients with XFG had Arg76Lys change in the first exon of the MYOC gene (Fig. 2) . This alteration has been considered to be a non–disease-causing polymorphism. ³⁴ No sequence alterations in the MYOC gene were found in the patient with POAG. In the OPTN gene one of the patients with XFG had an Met98Lys alteration in the exon 5 of the gene (Fig. 2) . This alteration was considered to be a risk-associated factor or a dominant susceptibility allele in the original study. ³⁵ Due to this finding, we sequenced this exon in 70 family members, of whom 18 had XFS, 6 had XFG, 4 had POAG, and 42 were healthy relatives. Thirty-five (18 XFS, 6 XFG, and 11 healthy relatives) individuals of these 70 are also included in the present genome-wide scan. The Met98Lys change was found in one of the two sons and in two nephews of the index case, but not in other family members. None of the mutation carriers, except for the index case, had any symptoms of XFS/XFG or POAG at the time of the examination (age, 40–61 years). 

Genome-Wide Scan

From the 183 family members examined in 2001 and 2002, 64 were selected for a genome-wide scan, including all the 28 XFS affected (mean age 70 years; range 59–90), of whom 9 had XFG, and, as unaffected controls, the oldest possible relatives: 22 close relatives (siblings, uncles, aunts, cousins; mean age 77 years; range, 63–90), 10 distant relatives (mean age, 80 years; range, 70–94), and 4 offspring to 3 affected mothers included in the screening (mean age, 58 years; range, 51–65). None of the control subjects had POAG. The pedigree of the analyzed subjects (Fig. 2)is a part of the original large pedigree. The family was divided into five subfamilies for the linkage analysis. 

A genome-wide scan with 1000 microsatellite markers was performed. When the autosomal dominant mode of inheritance with rare disease allele frequency (0.0001) was assumed, seven markers at chromosomal regions 2q32.3, 5q33.3, 17p13.3, 18q12.1-21.33, and Xp22.2 suggested evidence for linkage (Z _max > 1.5). In addition, 17 markers at regions 1q, 4p, 4q, 5p, 7p, 7q, 10p, 10q, 13q, 15q, 16q, 19q, Xp, Xq exceeded a two-point LOD score of 1.0. The results for the regions with the highest LOD scores using rare disease allele frequency (0.0001; Table 1 ; Fig. 3 ) and regions of particular interest for XFS (Table 2)are presented in more detail. The effect of higher disease allele frequencies (0.01, 0.1, 0.14) and an affected-only approach to the LOD scores are demonstrated in Supplementary Table S1. 

The most promising chromosomal region was 18q12.1-21.33, where the highest two-point LOD score of 3.45 (θ = 0.04) was obtained, with marker D18S468 assuming the dominant mode of inheritance. Four markers surrounding the best marker exceeded a pair-wise LOD score of 1.0; D18S1135 (Z _max dom = 1.39, θ = 0.14, α = 1.00), D18S450 (Z _max dom = 1.49, θ = 0.08, α = 1.00), D18S64 (Z _max dom = 1.70, θ = 0.08, α = 0.84), and D18S1147 (Z _max dom = 1.68, θ = 0.12, α = 0.83) spanning an area of ∼30 cM. One allele on the best marker D18S468 was more often present (37%) in affected than in unaffected (26%) family members, and another allele was found more often in unaffected (34%) than in affected (25%) subjects. Three-point analysis in the best region using adjacent markers D18S1102–D18S468, D18S468–D18S1143, and D18S1143–D18S450 produced maximum multipoint LOD scores of 1.85 (θ = 0.03; 0.10), 2.28 (θ = 0.08; 0.05), and 1.73 (θ =0.15; 0.03), respectively. When we combined the information from markers D181135 and D18S468, located 5 cM apart, the maximum three-point LOD score was 2.80 (θ = 0.05; 0.10), and using markers D18S468 and D18S450, located 8 cM apart, the highest three-point LOD score of 4.33 (θ = 0.05; 0.08) was obtained. The complete results for three-point analysis over the region are shown in Table 3 . 

Two markers, D4S2936 and D4S394, situated 14 cM apart at region 4p16.1-16.3, resulted in slightly positive pair-wise LOD scores (Z _max > 1.0) (Table 1)and with adjacent markers D4S2936–D4S412 and D4S2935–D4S394 multipoint LOD scores of 1.07 (θ = 0,15; 0,04) and 1.51 (θ = 0,10; 0,019) were obtained, respectively. The singleton marker D10S602 at 10p15.3 produced a slightly positive two-point LOD score (Z _max > 1.0; Table 1 ) and adjacent markers D10S602–D10S1218 yielded a maximum three-point LOD score of 1.22 (θ = 0.10; 0.05). It is noteworthy that marker D10S602 resides next to the GLC1E locus (10p15-p14), underlying POAG. ²⁰  

The highest pair-wise LOD score on chromosome 2 was obtained with marker D2S117 (Z _max dom = 1.73, θ = 0.06, α = 1.00) at region 2q32.3. Multipoint analysis with the adjacent markers D2S117 and D2S311 produced a three-point LOD score of 1.21 (θ = 0.10; 0.03). Suggestive evidence of linkage was also observed on chromosome 17 with marker D17S849 (Z _max dom = 1.91, θ = 0.04, α = 1.00), on chromosome 5 with marker D5S2049 (Z _max dom = 1.61, θ = 0.00, α = 1.00), and on chromosome 19 with marker D19S932 (Z _max dom = 1.09, θ = 0.00, α = 0.92). 

A nominal linkage (two-point LOD score, >1.0) was observed at regions 7q21.3 with marker D7S554 and 7p12.3 with marker D7S2506, assuming an autosomal recessive mode of inheritance (Table 1) . Marker D15S1032 produced an individual positive LOD score (Z _max > 1.0) at location 15q21.2 (Table 1) . 

At chromosome region 16q12.2, marker D16S3034 (Z _max dom = 1.35, θ = 0.12, α = 1.00) exceeded the threshold LOD score of 1.0 and, together with adjacent marker D16S415, a three-point LOD score of 1.33 (θ = 0.20; 0.01) was produced. On chromosome X the highest pair-wise LOD scores were observed at DXS7108 (Z _max dom = 1.52, θ = 0.12, α = 1.00) and DXS6810 (Z _max dom = 1.18, θ = 0.10, α = 1.00) lying ∼50 cM apart (Table 1) . 

Discussion

A genome-wide scan was performed in 64 subjects from one large pedigree; of these, 28 were XFS/XFG affected and 36 were unaffected. The most promising locus was at 18q12.1-21.33, where several markers, using both rare and common disease allele frequencies, gave evidence of linkage (Supplementary Table S1). The other interesting loci were 2q32.3, 7p12.3, 7q21.3, 15q21.2, 16q12.2, 17p13.3, 19q12, and Xp11.3. 

The locus, 18q12.1-21.33, on chromosome 18 is considerable in size (∼30 cM). The annotated human sequence makes it possible to compile a comprehensive list of candidate genes in the region. This locus includes numerous genes with both known and unknown functions. Most of these genes are also expressed in ocular tissues at some level. Chromosome 18 has not been reported to be associated with either XFS or POAG and thus the finding at 18q12.1-21.33 provides evidence of a novel XFS susceptibility locus on this chromosome. Though so far XFS is the most common identifiable cause of POAG, ⁶ we were also interested in determining whether the loci identified for XFS are situated in regions that overlap the loci earlier associated with POAG. 

On chromosome 2, the highest pair-wise LOD score was obtained with marker D2S117 at region 2q32.3. D2S117 is located on the previously suggested candidate region for POAG. This marker produced a two-point LOD score of 1.97 using the dominant inheritance model in the original genome-wide scan of POAG. ³⁹ Suggestive evidence of linkage was also observed on chromosome 17 with marker D17S849 and on chromosome 19 with marker D19S932. These two markers are situated at loci 17p13.3 and 19q12, which are two of the highest-scoring regions for adult onset POAG in Wiggs et al. ⁴⁰ In that study, marker D17S926, which overlaps marker D17S849, produced a maximum pair-wise LOD score of 2.23 and marker D19S414, residing 3 cM from D19S932, resulted in Z _max = 3.11, assuming an autosomal recessive mode of inheritance. These findings suggest the possibility of some shared factors behind XFS and POAG. It is noteworthy that the apolipoprotein E (ApoE) gene, which has been reported to be associated with the development of XFS ³⁷ and to participate in amyloid deposition and fibril formation, is located some distance away from D19S932, at 19q13.31. 

A nominal linkage was observed at regions 7p12.3 with marker D7S2506 and 7q21.3 with marker D7S554. These regions are the nearby regions 7p13, 7q21.3, and 7q21.11, where a loss of heterozygosity in XFS specimens has been reported with markers D7S478, D7S479, ¹⁸ and D7S820, ¹⁹ respectively. Chromosome 7 is also of interest because it carries gene sequences encoding elastin (at 7q11.23), laminin B1 (7q31.1), and collagen 1A2 (7q21.3), especially as XFS is commonly recognized as a type of elastosis. 

An individual positive LOD score was produced at region 15q21.2 with marker D15S1032, which is situated approximately 8 cM from the fibrillin-1 at locus 15q21.1. Fibrillin-1 is an intrinsic component of elastic microfibrils and XFS fibers ⁶ and has been observed to be upregulated in ciliary processes, iris tissue, and the lens epithelium of XFS-affected eyes. ⁷  

At chromosome location 16q12.2, marker D16S3034 exceeded the threshold LOD score of 1.0 and is situated at the same chromosome arm as MMP2 (at 16q13), a major MMP in human aqueous humor that cleaves type IV collagen. The slightly positive marker DXS6810 at chromosome region Xp11.3 lies next to TIMP1 (at Xp11.3-11.23), the endogenous inhibitor of MMP-1 and -9. Elevated TIMP-2 (tissue inhibitor of MMP-2) and low MMP-2 ³⁶ levels, as well as an excess of TIMP-1 over MMP-9, ³⁸ in the aqueous humor of XFS eyes and upregulation and increased production of TIMP-2 and a contrasting downregulation of TIMP-1 in anterior segment tissues of eyes with XFS ⁷ has been reported in previous studies. Excessive TIMPs have been suggested to prevent the destruction of newly synthesized matrix, leading to exfoliate accumulation. ³⁸ Further studies are needed to confirm the role of the MMP/TIMP pathway in the pathophysiology of XFS. 

No hint of linkage was obtained for markers at 1p13.2, near adenosine receptor A3 (ADORA3), which participates in aqueous humor secretion and IOP regulation, for example, and whose mRNA and protein have been found to be significantly upregulated in the nonpigmented ciliary epithelium of XFS/XFG eyes. ^{7 41} No evidence for linkage was identified either in loci 2p14-2cen or 2q35-q36, which have been suggested to be putative candidate loci for XFS (Sotirova et al. IOVS 1999;40:ARVO Abstract 512). 

The main problem with investigations of late-onset diseases is recruitment of affected individuals in several subsequent generations. In the family in the present study, detailed phenotypes are available from three generations. The mean age of unaffected family members included in the study at the time of the last examination was 75.3 years (range, 51–94). Approximately one fifth of examined family members are likely to be affected during their lifetimes; and, correspondingly, approximately 80% of the unaffected will stay unaffected. Therefore, we accepted a model in which the unaffected were scored as unaffected in linkage analyses. However, we also calculated two-point LOD score results obtained by marking all unaffected family members as unknown in the linkage analysis (Supplementary Table S1). Although there is some variation in LOD scores, as expected, depending on the model, the loci remain the same. 

The isolated island of Kökar offered a good starting point for genetic studies. Its small population made it possible to examine nearly all the inhabitants, and they were interested in the follow-up examinations. We hypothesize that widespread genetic and environmental homogeneity may make it easier to identify common predisposing alleles that are identical by descent (IBD). Because of the isolated location of the island, the founder effect cannot be excluded, although the prevalence of XFS is high in the elderly in the mainland of Finland (22%) ¹¹ and in Kökar (18.4%). Bearing in mind the high prevalence of XFS in Kökar and the possibility that rare alleles could have been enriched in this small, isolated population we examined the effect of disease allele frequency on LOD scores by using the range of disease allele frequencies accommodating up to 0.14 (0.0001, 0.01, 0.1, 0.14; Supplementary Table S1). We assume that XFS is caused by several predisposing genes, with an unknown allele frequency for each of them. Because of the complex nature of the disease, the true disease allele frequency is probably less than 0.14. Although, most of the subfamilies of the original family in the present study were linked to the same genetic loci (α = 1.0), there were also markers where only some of the families were linked (α < 1.0; Table 1 ). This evidence implies that other genetic loci and predisposing alleles are needed for the expression of the XFS phenotype in this subisolate. We assume that one major gene is responsible for XFS in our family, although we cannot exclude the impact of other susceptibility genes. This explanation may be due to high frequency and the multifactorial nature of the disease and the import of new alleles by married-ins in the pedigree. Heterogeneity demonstrated in the present study, as well as candidate genes and regions represented in previous investigations (Sotirova et al. IOVS 1999;40:ARVO Abstract 512), ^{7 18 19 36 37 38 41} support the idea that multiple genes underlie XFS. 

In conclusion, the most promising locus was assigned to chromosome 18q12.1-21.33. Markers on chromosomes 2, 17, and 19 producing positive signals overlap the previously suggested possible regions of POAG linkage. This finding leads to the question of whether there are possible common predisposing alleles. Further investigations of the regions in these two different phenotypes are warranted. 

 Supported by the Sigrid Jusélius Foundation, Helsinki, Finland; Eye and Tissue Bank of Finland; the Eye Foundation, Finland; the Glaucoma Foundation LUX, Helsinki, Finland; the Emil Aaltonen Foundation, Tampere, Finland; the Biomedicum Helsinki Foundation, Helsinki, Finland; the Medical Society of Finland, Helsinki, Finland; the Åland Cultural Foundation, Mariehamn, Finland; the Finnish Medical Association, Helsinki, Finland; and the Finnish Society of Sciences and Letters, Helsinki, Finland.

 Submitted for publication September 13, 2006; revised January 24 and May 15, 2007; accepted July 5, 2007.

 Disclosure: S. Lemmelä, None; E. Forsman, None; P. Sistonen, None; A. Eriksson, None; H. Forsius, None; I. Järvelä, 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.

 

Figure 1.

 

Exfoliation debris on the lens.

Figure 2.

 

The Finnish XFS family. Extended pedigree structure indicating the families included in the genome-wide scan. The three most recent generations were examined. *Individuals providing material for the genome-wide scan; arrows: sequence alterations in MYOC (Arg76Lys) (M) and OPTN (Met98Lys) (O) genes.

Table 1.

 

Highest Two-Point LOD Scores

Marker	Location	Position (cM)	Model	Z max	θ	α
D1S245	1q32.2	217	Rec	1.098	0.00	1.00
D2S117	2q32.3	199	Dom	1.725	0.06	1.00
D4S2936	4p16.3	0.9	Dom	1.000	0.00	0.61
D4S394	4p16.1	15	Dom	1.328	0.10	1.00
DG4S26	4q35.2	202	Dom	1.253	0.14	1.00
D5S2031	5p14.3	42	Rec	1.183	0.00	1.00
D5S2049	5q33.3	166	Dom	1.605	0.00	1.00
D7S2506	7p12.3	72	Rec	1.253	0.00	1.00
D7S554	7q21.3	110	Rec	1.126	0.00	1.00
D10S602	10p15.3	6	Dom	1.284	0.08	1.00
D10S192	10q24.31	124	Dom	1.295	0.12	1.00
D10S1676	10q26.2	165	Dom	1.398	0.00	0.60
D13S293	13q34	132	Dom	1.062	0.00	1.00
D15S1032	15q21.2	52	Dom	1.186	0.10	1.00
D16S3034	16q12.2	68	Dom	1.354	0.12	1.00
D17S849	17p13.3	0.07	Dom	1.911	0.04	1.00
D18S1135	18q12.1	58	Dom	1.385	0.14	1.00
D18S468	18q12.3	63	Dom	3.446	0.04	1.00
D18S450	18q21.1	71	Dom	1.491	0.08	1.00
D18S64	18q21.32	84	Dom	1.695	0.08	0.84
D18S1147	18q21.33	89	Dom	1.682	0.12	0.83
D19S932	19q12	51	Dom	1.086	0.00	0.92
DXS7108	Xp22.2	20	Dom	1.519	0.12	1.00
DXS6810	Xp11.3	69	Dom	1.180	0.10	1.00
DXS8055	Xq23	118	Rec	1.086	0.00	1.00
DXS1047	Xq25	135	Rec	1.063	0.00	1.00

 Markers with Z _max > 1.0, using dominant (Dom: 0.001, 0.999, 0.999) or recessive (Rec: 0.001, 0.001, 0.999) mode of inheritance and a rare disease allele frequency (0.0001), are presented. The LOD scores are maximized over recombination fractions (θ) ranging from 0 to 0.5. α ranging from 0 to 1.0 indicates the proportion of families showing linkage. The markers with Z _max > 1.5 are shown in bold.

Figure 3.

 

Two-point LOD scores, using the dominant (Lod_Hom_dom) and recessive (Lod_Hom_rec) modes of inheritance under homogeneity, obtained from the whole genome-wide screen.

Table 2.

 

Overlapping Candidate Loci for XFS

Current Study			Suggested Loci/Gene for XFS
Locus	Marker	Locus	Gene	Reference
7q21.3	D7S554	7q21.3; 7q21.11	?	18 19
7p12.3	D7S2506	7p13	?	18
15q21.2	D15S1032	15q21.1	Fibrillin-1	7
16q12.2	D16S3034	16q13	MMP-2	7 36
19q12	D19S932	19q13.2	APOE	37
Xp11.3	DXS6810	Xp11.3–11.23	TIMP-1	7 38

 Markers with pair-wise LOD score >1.0 in the current study are presented.

Table 3.

 

LOD Scores on Chromosomal Region 18q12.1–21.33

Marker	Position (cM)	LODHom at θ									LODHet (Z max)	Multipoint LOD Score
				0.0	0.1	0.2	0.3	0.4
18S1135	58	−6.44	1.26	1.24	0.69	0.19	1.385
								1.42
D18S1102	60	−3.29	0.23	0.50	0.36	0.14	0.501
								1.85
D18S468	63	2.15	3.12	2.11	1.05	0.26	3.446
								2.28
D18S1143	68	−2.58	0.38	0.55	0.34	0.09	0.775
								1.73
D18S450	71	−0.16	1.48	1.12	0.61	0.21	1.491
								1.27
D18S474	75	−6.40	0.62	0.84	0.50	0.12	0.869
								—
D18S64	84	−2.41	1.67	1.36	0.75	0.23	1.695
								1.32
D18S1134	87	−4.58	0.25	0.60	0.45	0.18	0.597
								0.97
D18S1147	89	−4.06	1.55	1.56	1.00	0.38	1.682
								1.26
D18S68	91	−4.23	0.04	0.36	0.25	0.07	0.532

 Markers in a 33-cM region on chromosome 18 were analyzed using dominant mode of inheritance. The two-point homogeneity LOD scores (LODHom) are maximized over recombination fractions (θ) ranging from 0 to 0.5, whereas heterogeneity LOD scores (LODHet) yielding the highest LOD score (Z_max) are shown. The best multipoint LOD scores can also be found in the table.

Supplementary Materials

Supplementary Table S1 - (PDF) 

The authors thank the family members for participating and Helena Dahlbom for collecting the genealogical data.