Chromosomes 4q28.3 and 7q31.2 as New Susceptibility Loci for Comitant Strabismus

Abstract

purpose. This study was designed to localize chromosomal susceptibility loci for comitant strabismus among Japanese families by genome-wide linkage analyses.

methods. Fifty-five Japanese families, with at least two members with comitant strabismus (esotropia and/or exotropia), were subject to full ophthalmic examination, careful ocular history, and review of medical records. DNA was obtained and genotyping was performed with PCR amplification of 400 microsatellite markers. Parametric and nonparametric linkage (NPL) analyses scores were calculated. Linkage analysis was performed for the whole set of families (55 families), and then a second analysis was performed for two subgroups with the phenotypes, esotropia and exotropia.

results. A multipoint parametric heterogeneity logarithm of the odds (HLOD) score of 3.62 was obtained at marker D4S1575 under a dominant model, with a NPL score of 2.68 (P = 0.001). Testing under different penetrances and disease allele frequencies revealed two other susceptibility loci at 7q31.2 under a recessive model (HLOD scores = 3.93 and 4.40 at 125.2 cM and 107.28 cM, respectively). Analysis of the subgroups revealed new susceptibility loci for esotropia; one locus at 8q24.21 is worthy of further investigation.

conclusions. This study suggests multiple susceptibility loci for comitant strabismus. The loci at chromosomes 4q28.3 and 7q31.2 show a significant evidence of linkage.

Strabismus is the misalignment of the eyes and can be classified as comitant when a deviation is constant in all directions of gaze, or incomitant when a deviation varies in different directions of gaze. ¹ Ocular misalignment can lead to amblyopia, and reduction or loss of binocularity. ² Common clinical entities of comitant strabismus are infantile or accommodative esotropia, and constant or intermittent exotropia. Incomitant strabismus could be paralytic, either acquired such as cranial nerve palsies or congenital such as congenital superior oblique muscle palsy, or it could be restrictive, either acquired as in thyroid orbitopathy or congenital as in congenital fibrosis of the extraocular muscles (CFEOM). 

The prevalence of comitant strabismus in Japan is approximately 1% to 2% of the population ^{3 4 5} and approximately 2% to 4% among Caucasian populations. ⁶ Family studies of comitant strabismus show that >30% of probands have positive family history. ^{7 8 9 10} Twin studies have proven a higher concordance of strabismus in monozygotic than in dizygotic twins. ^{8 11} These lines of evidence suggest a strong hereditary background for comitant strabismus. 

To date, only two genome-wide linkage studies of comitant strabismus have been published. A susceptibility locus on chromosome 7p22.1 was reported in a genome-wide linkage study of seven families with comitant strabismus. Only one of those pedigrees under a model of recessive inheritance showed heterogeneity logarithm of the odds (HLOD) score of 4.51 at the proposed locus. ¹² The second study was our previous analysis of data from 30 families with comitant strabismus. ¹³ The number of families included in that study was too small to obtain statistically significant results and the study was considered a preliminary step along our ongoing search for chromosomal susceptibility loci for comitant strabismus. In the present study, we added 25 new families and then performed parametric and nonparametric linkage (NPL) analyses for the whole set of families (55 families). We reported our genome-wide scan of those families with comitant strabismus including both esotropia and exotropia, and revealed multiple susceptibility loci for comitant strabismus. 

Methods

Families’ Recruitment and Evaluation

Families participating in this study were enrolled from the Strabismus Clinic, Ophthalmology Department, Okayama University Hospital. Each family had at least two members assigned as affected with comitant strabismus. We studied in total 258 individuals in 55 Japanese families (data of the previously studied 30 families ¹³ and additional 25 new participating families) mainly from the Okayama area; 117 were males (45%) and 141 were females (55%). Ascertained as affected were 120 individuals: 73 females (61%) and 47 males (39%). 

This study was approved by the Ethics Committee of Okayama University Medical School and was performed in accordance with the Declaration of Helsinki. An informed consent was obtained from parents of affected children or from affected adults after thorough explanation of the nature and expected outcome of the study. 

To assign a family member as affected, complete ophthalmic evaluations, including questions about family history, were performed on probands and all their available relatives by the examining ophthalmologists. Evaluations included visual acuity, angles of deviation at distance (5 m) and at near (0.3 m) with cover/uncover test, alternate prism and cover test (APCT) or Krimsky test (in young or uncooperative patients), accommodative convergence/accommodation ratio by lens gradient method, stereopsis tests, and cycloplegic refraction. The mean spherical equivalent in diopters (D) for myopic cases was −1.25 D (range from −0.25 D to −7.5 D), and the mean spherical equivalent for hyperopic cases was +2.7 D (range from +0.25 D to +8.0 D). For individuals whose examination was not possible, a history by interviews or analysis of medical records was used to confirm the affected status. 

According to the results of the examination, history, and medical records data, we included as affected any proband or available relative with constant or intermittent esotropic, exotropic, or hypertropic comitant misalignment, in the form of heterotropia (manifest misalignment) or heterophoria (latent misalignment). Any unavailable relative with a history of strabismus and/or strabismic amblyopia, strabismus surgery, or wearing spectacles to correct strabismus was considered as affected. 

Affected individuals were either esotropic (infantile, accommodative or partially accommodative esotropia or microtropia) or exotropic (intermittent or constant exotropia). Exclusion criteria were incomitant strabismus (paralytic or restrictive), strabismus associated with musculoskeletal or neurologic malformations, strabismus related to ocular trauma, surgery or diseases, and subjects with a history of low birth weight (<2500 g) or prematurity (birth before 37 gestational weeks). 

Forty-seven of the 55 studied families were two-generation families, four families were three-generation families and four were two-generation families with an affected aunt or uncle (some illustrative pedigrees are displayed in Fig. 1 ). The average number of participating individuals per family was 4.7 (range, 4–15), and the average number of affected individuals per family was 2.2 (range, 2–7). The summary of demographic data is presented in Table 1 . 

Genotyping

Genotyping was performed for 214 individuals. Genomic DNA was isolated from 10 mL peripheral blood obtained from the probands and their available relatives, as previously described. ¹³ Genome-wide search was performed with amplification by PCR of 400 markers in microsatellite regions (Linkage Mapping Set MD10; PE Applied Biosystems, Foster City, CA). The spacing between markers was on average 10 centimorgans (cM), and the average heterozygosity was 0.79. Each marker site was amplified by DNA polymerase (AmpliTaq Gold; Roche, Branchburg, NJ) at the manufacturer’s recommended conditions for the PCR system (GeneAmp 2400; Perkin-Elmer, Foster City, CA). DNA fragments were then mixed with size standards (GeneScan 400 HD [ROX]) in formamide, applied to a genetic analyzer (ABI PRISM 310; PE Applied Biosystems) and analyzed by analysis software (GeneScan; PE Applied Biosystems). 

Data Validation for Statistical Analyses

Determination of alleles was performed by two members of our research team, masked from pedigree structures or phenotypes. Any discrepancies were rechecked and resolved before proceeding. For Mendelian errors, data were checked with two analysis programs (PEDMANAGER [Whitehead Institute for Biomedical Research, MIT, Cambridge, MA] and PEDCHECK). ¹⁴ In case of any incompatibilities, the original data were reevaluated and the relationships of incompatible family members were checked. When Mendelian inconsistencies at multiple markers could not be resolved, data at these markers were considered missing. If missing data for one individual was >50%, that individual was eliminated. Marker allele frequencies were estimated from the whole dataset by a computer program (DOWNFREQ version 1.1; available at ftp://linkage.comc.columbia.edu/software/analyzenew [Joseph D. Terwilliger, 1995]). Because some of the founder individuals in the pedigrees were not typed, we chose to estimate marker allele frequencies from all individuals at all pedigrees as recommended by previous studies. ^{15 16}  

Linkage Analyses

Since the genetic components of comitant strabismus remain undetermined, we chose to approach our data by two types of analyses: model-dependent parametric (two-point and multipoint) analysis with an assumption of the mode of inheritance, and model-independent multipoint NPL analysis. We first performed these analyses for our entire set of families (55 families), and then we stratified the families into two subgroups according to phenotype: esotropia group (25 families) and exotropia group (26 families). Four families having both esotropic and exotropic patients were excluded from the stratification analysis. Parametric and nonparametric analyses were re-run for these subgroups separately. 

For the parametric linkage analysis, we assumed two simple (recessive or dominant) models of inheritance, with a disease allele frequency (P_d) = 0.01, penetrance (ƒ) for gene carriers = 0.8, and a phenocopy rate for non–gene carriers = 0.01 (Table 2) . 

The two-point linkage analysis was performed (MLINK of LINKAGE 5.1, with the utility programs MAKEPED, PREPLINK, Linkage Control Program [LCP], and Linkage Report Program [LRP]). ^{17 18 19} The multipoint parametric linkage analysis was also performed (GENEHUNTER software version 2.0 beta). ²⁰ Logarithm of the odds (LOD) scores were calculated at every marker locus and at 4 intermarker locations along the genetic map of each chromosome. Since it is expected that multiple loci might be involved in susceptibility to comitant strabismus, we allowed the computation of LOD scores under heterogeneity, assuming that a proportion of the families, “α,” are linked at the locus of interest, and maximizing the HLOD at each position with respect to α. ^{21 22}  

The nonparametric model-free multipoint linkage analysis (NPL _all ) was also computed (GENEHUNTER software). The NPL score is a measurement of the degree of sharing of alleles among affected individuals, independent of the model proposed for the disease under study, and represents the deviation from normal Mendelian expectations. The all function examines all individuals simultaneously and assigns a higher score when more of them share the same allele by descent. ²⁰  

Map locations and intermarker distances from the p-terminal end of each chromosome were taken from the 1996 Genethon human linkage genetic map (Kosambi distance in cM). ²³ Based on the parametric and nonparametric analyses, our results were reported as LOD score and NPL score with its P value, respectively. We also determined the 1-LOD interval for regions with LOD scores reaching suggestive or significant linkage levels. ²⁴ The cut point for the significant levels was that recommended by Lander and Kruglyak. ²⁵  

For chromosomes with regions showing significant or suggestive evidence of linkage with our initial models of inheritance, LOD scores were recalculated on the assumption of different penetrances (0.9, 0.8, or 0.7) and different disease allele frequencies (0.1, 0.01, or 0.001) to study the effect of penetrance and disease allele frequency on the LOD scores. 

To determine whether the reported susceptibility locus for comitant strabismus at 7p22.1 ¹² also is significant in the linkage analysis of the Japanese population, we investigated that locus using our model in the whole set of families, and then using the model proposed in the aforementioned study. In a second step, we re-examined the locus in the esotropia subgroup separately since that locus was reported in a large family in which affected members were esotropic. 

Results

Two-Point Linkage Analyses

In the two-point parametric linkage analyses using the inheritance vectors for either a dominant or a recessive model, 19 markers showed a LOD score greater than 1.0 under the dominant model, and 19 markers under the recessive model. Suggestive evidence of linkage (LOD scores greater than 1.9) was obtained on chromosomes 1, 2, 4, 6, 7, 14, and 17. Some of the markers on chromosomes 1, 2, 6, and 7 showed a LOD score greater than 1.0 under both models (Table 3) . 

Multipoint Linkage Analyses

The results of the multipoint linkage analyses under the dominant model (using GENEHUNTER) in all 55 families are displayed in Figure 2 , and a summary of the markers showing HLOD score greater than 1.9 is shown in Table 4 . The strongest evidence of linkage was found at D4S1575 (multipoint HLOD = 3.62, α = 1.0) in the 4q28.3 region using the dominant model (model 1). The calculated 1-LOD support interval for this region, which has been suggested to contain the gene responsible for the linkage with up to 90% probability, ^{24 26} extended from 122.90 cM to 135.92 cM between markers D4S402 at 4q26 (HLOD = 2.71) and D4S424 at 4q28.3 (HLOD = 2.68). These three loci covered the total distance of 13.02 cM. The NPL score for D4S1575 was 2.68 (P = 0.001), and 2.49 (P = 0.003) and 2.01 (P = 0.01) for the markers, D4S402 and D4S424, respectively. 

In addition to the region on chromosome 4, seven regions on chromosomes 1, 7, 8, 11, and 20 showed HLOD scores greater than 1.9, suggestive evidence of linkage at the genome-wide level under either the dominant or the recessive model at 1p31.3, 1q31.1, 7p14.3, 7q31.2, 8q24.13, 11q24.2, and 20q11.23 (Table 4) . 

When the loci showing suggestive or significant evidence of linkage were retested using different combinations of penetrances and disease allele frequencies, the decrease of the penetrance to 0.7 yielded an increase in the HLOD scores at D4S1575 (HLOD = 3.76, P_d = 0.01), D11S1320 (HLOD = 2.17, P_d = 0.01), D20S195 (HLOD = 2.07, P_d = 0.001), and D1S207 (HLOD = 2.51, P_d = 0.1). An inheritance model of 90% penetrance and 0.01 disease allele frequency achieved an increase of the HLOD scores to 2.21 at D1S413, 2.88 at D8S284, and 2.57 at D7S484. 

Two findings were observed at chromosome 7, at the setting of 0.1 disease allele frequency and 70% penetrance under a recessive model. The first observation was that the HLOD score at D7S486 increased from 2.32 to 3.93 at α = 0.62 (NPL = 2.7, P = 0.001), which changed the findings at this locus from suggestive evidence to significant evidence of linkage (Table 5) . The second observation was a relocation of the point of maximum HLOD from marker D7S486 (at 125.2cM) to a region between markers D7S657 and D7S515 (at 107.28cM), which lay within the 1-LOD confidence interval of the marker D7S486. The HLOD score at this new locus increased to reach 4.40 at α = 0.69 (NPL = 2.45, P = 0.003), which supported a significant evidence of linkage (Fig. 3) . 

To correct type I errors resulting from maximizing LOD scores over a dominant model, a correction factor of ≈0.3 was considered. Another correction factor of ≈0.3 was added where maximization over penetrances was performed. Applying these corrections to our results led to HLOD = 3.32 at 4q28.3, and HLOD = 3.33 and 3.80 at 125.2 cM and 107.8 cM at 7q31.2, respectively, all of which still remained as significant evidence of linkage. ^{26 27}  

Stratification into Esotropia Subgroup and Exotropia Subgroup

Parametric and nonparametric multipoint analyses were retested for the esotropia and the exotropia subgroups separately. Five loci at 6q27, 8q24.21, 10q22.2, 15q22.2, and 17p13.1 obtained HLOD scores suggestive of linkage in the esotropia subgroup for a dominant model (P_d = 0.01, phenocopy rate = 0.01, and 80% penetrance) (Table 6) . A HLOD score of 2.79 (NPL = 2.25, P = 0.005) was found at 8q24.21 (139.24 cM) in the esotropia subgroup, which was 4.96 cM away from the locus showing suggestive evidence of linkage in the whole-family analysis at 134.28 cM (Fig. 4) . On the other hand, only one locus was found in the analyses of the exotropia subgroup, which reached a threshold suggestive of linkage at 14q21.3 region (HLOD = 1.91, α = 1, NPL = 1.85, P = 0.02) under a dominant model. 

Replication of the Locus at 7p22.1

The HLOD score at D7S513 using either the recessive or the dominant model: (penetrance = 80%, disease allele frequency = 0.01, and phenocopy rate = 0.01) failed to exceed 1.0 in the analysis of the whole set of families. Neither could we find evidence of linkage when we re-analyzed the data with the model proposed by Parikh et al., ¹² with higher penetrance and less disease allele frequency and phenocopy rate: (penetrance = 90%, disease allele frequency = 0.001 and phenocopy rate = 0.0001). In the analysis of the esotropia subgroup, using our recessive model, the HLOD score at D7S513 was 1.13. Using the recessive model proposed by Parikh et al., ¹² suggestive evidence of linkage could be obtained (HLOD score = 2.53, NPL = 2.58, P = 0.001) at 24.30 cM. 

Discussion

This study is an extension of our earlier work aimed at localizing chromosomal susceptibility loci for comitant strabismus. ¹³ In addition to the previously-analyzed data of 30 families, we analyzed the data of 25 additional new families. At present, there is no clear evidence that the phenotypic variations of comitant strabismus reflect genetic heterogeneity. Furthermore, different entities of comitant strabismus might share common etiology such as binocular vision abnormalities or fusion deficits. ²⁸ We, therefore, chose to include in our primary analyses all the families with esotropia and/or exotropia. 

In our earlier study, we performed affected sib-pairs and nonparametric linkage analyses of multiple pedigrees for 30 families. ¹³ At that time, we failed to report any susceptibility loci in either analysis, which was attributed to the small sample size and to the choice of the nonparametric approach. The model-free approach has the advantage of not requiring the specification of a model of inheritance, which is legitimate in comitant strabismus wherein little is known about its mode of inheritance. ^{6 8 29 30} However, the fact that NPL analysis does not use information from unaffected members makes it rather conservative compared to the parametric model-dependent methods. ^{26 31} In this study, we chose to perform parametric linkage analysis under two simple dominant and recessive models, together with the NPL analysis. The model-dependent approach is expected to retain a high power to detect linkage, even when the disease is inherited in more complex way than these two simple models suggest. Another merit of choosing only two simple models for the analyses is the minimal cost of increasing type I errors when multiple models are assumed. ^{26 27 31} We presume that increasing the number of participating families and using parametric linkage analysis methods increased the power to detect linkage. 

The only susceptibility locus for comitant strabismus, as reported by Parikh et al., ¹² has been located at 7p22.1 (under the assumption of a recessive mode of inheritance in one large family). In the primary analyses of our present study, no evidence of linkage was found at that locus. A suggestive evidence of linkage at 7p22.1 was only obtained in our secondary analysis of the esotropia subgroup using the recessive model proposed by Parikh et al. ¹² Unlike that proposed model with high penetrance, we observed in our study that decreasing the penetrance was associated with increase of HLOD scores in most of the studied loci. Similar to the results reported by Parikh et al., ¹² our implicated loci showed linkage under either a recessive or a dominant model. This is anticipated in a complex disorder like comitant strabismus when the trait might be influenced by multiple loci. The inheritance at each of these loci would be expected to approximate a recessive or a dominant model of inheritance. ²⁶  

One limitation of the present study is the small size of the participating families–two-generation families in most cases. Multi-generational families are expected to give a stronger power to detect linkage. ^{1 25} On the other hand, these multiple small families are expected to segregate a smaller number of genes thus decreasing within-family heterogeneity; also, nonparametric statistics in those families tend to show normal distribution and to be robust. ^{21 25} Another limitation was using history, medical records, or interviews based on a defined questionnaire ⁷ to determine the affected status for unavailable members. Using history for unavailable family members could have easily missed cases of phoria or microtropia; which could have led to an underestimation of affected individuals, consequently decreasing the power to detect linkage. 

The results of the secondary analyses, after phenotypic stratification into esotropia and exotropia subgroups, revealed new susceptibility loci. This could be attributed either to the loss of power resulting from a smaller sample size in the subgroups or to genetic heterogeneity, in which a modifier gene or a variably expressive gene plays a role in altering the results. ³²  

The Japanese population, in general, is of homogeneous nature, and the Okayama area has not experienced large immigrations, compared with larger cities such as Tokyo or Osaka. This might be the reason for a strong founder effect in our population sample. Hence, the number of genes underlying comitant strabismus in our families might be smaller compared to more heterogeneous populations. On the other hand, the use of samples derived from a rather isolated population could be the reason why the loci implicated in our families could be different from those found in different populations or ethnicities. ²¹  

A brief search in the OMIM (Online Mendelian Inheritance in Man) database for genes in our implicated loci, had reference to some genes which are expressed in the brain, such as neuronal cell adhesion molecule (NRCAM) at 7q31.1-q31.2, or mastermind-like 3 (MAML3) and Protocadherin 10 (PCDH10) at 4q28.3. Genes for comitant strabismus would be expected to be expressed in the brain, as they might be involved in binocular fusion and stereopsis. We have not thoroughly investigated the genes at our implicated loci, since they tended to extend over broad regions which are in need of fine mapping. 

Despite the efforts to understand the pathogenesis and the mode of inheritance of comitant strabismus, many questions remain unanswered. ^{2 6 33} Maumenee et al. ³⁴ concluded from their segregation analyses that etiologic heterogeneity might exist in families with comitant esotropia, including autosomal recessive cases, dominant cases, and aggregation of nongenetic cases. The relationship between the genetics of comitant strabismus and those of refractive errors has not yet been adequately investigated. Some studies succeeded in finding concordance between refractive errors (especially hyperopia) and some entities of comitant strabismus (comitant esotropia). The risk for developing comitant strabismus markedly increased whenever there was a family history of strabismus associated with hyperopia. ^{6 33} These observations might raise a question of whether the genes responsible for refractive errors might also influence the development of esotropia or exotropia. 

In summary, we report multiple susceptibility loci for comitant strabismus through genome wide linkage analyses of 55 Japanese families in the present study. The most significant of these loci were located at chromosomes 4q28.3 and 7q31.2. We believe that better understanding of the pathophysiological processes underlying comitant strabismus would be possible with more research in progress and that ultimately a preventive or curative treatment would be found. 

Web Resources

OMIM, Genethon map and linked regions were searched for positional candidate genes at: http://ncbi.nlm.nih.gov/. 

Software programs or their links are available at the following websites:

•  
  GENEHUNTER: http://www.broad.mit.edu/ftp/distribution/software/genehunter/
•  
  LINKAGE: ftp://linkage.rockefeller.edu/software/linkage
•  
  PEDCHECK: http://watson.hgen.pitt.edu/register/
•  
  DOWNFREQ: ftp://ftp.ebi.ac.uk/pub/software/linkage_and_mapping/linkage_cpmc_columbia/analyze/
•  
  PEDMANAGER: http://www-genome.wi.mit.edu/ftp/distribution/software/pedmanager/.

 Supported in part by a Grant-in-Aid (C19592023) for Scientific Research from the Japan Society for the Promotion of Science, and by NSFC Grants 30730057 (JO) and 30700442 (QZ).

 Submitted for publication June 16, 2008; revised August 7 and September 13, 2008; accepted December 29, 2008.

 Disclosure: S. Shaaban, None; T. Matsuo, None; H. Fujiwara, None; E. Itoshima, None; T. Furuse, None; S. Hasebe, None; Q. Zhang, None; J. Ott, None; H. Ohtsuki, 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.

 

Illustrative pedigrees for some of the families participating in the study. Closed symbols: individuals assigned as affected; open symbols: unaffected individuals; dots: individuals who were genotyped; squares: males, and circles: females. XT, constant exotropia; XpT, intermittent exotropia; MT, microtropia; Infantile ET, infantile esotropia.

Table 1.

 

Clinical Features of the 55 Families with Comitant Strabismus

Total number of families	55
Number of families with at least two members with esotropia	25
Number of families with at least two members with exotropia	26
Number of families with mixed phenotypes*	4
Total number of individuals, †	258
Males	117
Females	141
Total number of individuals genotyped	214
Males	88
Females	126
Total number of individuals affected	120
Males	47
Females	73
Number of individuals affected with esotropia	55
Accommodative esotropia	25
Infantile esotropia	11
Micro-esotropia	4
Unclassified esotropia	15
Number of individuals affected with exotropia	65
Intermittent exotropia	55
Constant exotropia	10
Average number of individuals per family (range)	4.7 (4–15)
Average number of affected individuals per family (range)	2.2 (2–7)

*  Mixed phenotypes indicate that both esotropic and exotropic patients are found in the same family.

†  Total number of individuals participating in the study either undergoing examination, interviews, or review of their medical records.

Table 2.

 

The Inheritance Vectors for Dominant and Recessive Models

	Penetrances (f)
		f 0 (aff/++)	f 1 (aff/+d)	f 2 (aff/dd)
Genetic Model*
1	0.01	0.8	0.8
2	0.01	0.01	0.8

*  Models correspond to autosomal dominant (model 1) and recessive (model 2) patterns of inheritance, with penetrance value (f) = 80%, and disease allele frequency (P_d) = 0.01.

 +, wild-type allele at disease locus; d, disease allele at disease locus.

Table 3.

 

Two-Point Linkage Analysis in 55 Japanese Families with Comitant Strabismus

Marker	Assumed Model of Inheritance*	Cytogenetic Location	Location of Maximum LOD (cM)	LOD Score
D1S468	Dominant	1p36.32	6.2	1.11
D1S214	Dominant	1p36.32	8.9	1.29
D1S2667	Dominant	1p36.23	19.3	2.15, †
D1S413	Dominant	1q31.3	215.2	2.11, †
D2S337	Dominant	2p16.2	77.5	1.27
D2S125	Dominant	2q37.3	263.2	2.77, †
D4S1575	Dominant	4q28.3	130.6	2.67, †
D4S413	Dominant	4q32.1	157.8	1.93, †
D6S308	Dominant	6q23.2	135.8	1.99, †
D6S264	Dominant	6q26	169.4	1.84
D7S484	Dominant	7p14.3	50.8	2.62, †
D7S530	Dominant	7q32.1	135	1.97, †
D8S514	Dominant	8q24.12	126	1.42
D11S968	Dominant	11q24.3	140.4	1.70
D13S1265	Dominant	13q34	149	1.54
D14S292	Dominant	14q32.32	124.5	2.10, †
D17S799	Dominant	17p12	35.7	1.99, †
D19S420	Dominant	19q13.11	55.7	1.40
D20S196	Dominant	20q13.13	74.5	1.70
D1S2667	Recessive	1p36.23	19.3	1.16
D1S2841	Recessive	1p31.3	101	1.30
D1S206	Recessive	1p22.2	128.9	1.35
D2S2211	Recessive	2p25.2	8.3	1.33
D2S2259	Recessive	2p22.2	59.6	1.14
D2S125	Recessive	2q37.3	263.2	1.61
D3S2338	Recessive	3p25.1	31.3	1.11
D3S1300	Recessive	3p14.3	70.6	1.11
D4S419	Recessive	4p15.32	30.6	1.20
D4S392	Recessive	4q13.2	76.6	1.44
D4S402	Recessive	4q26	122.9	1.77
D4S1597	Recessive	4q33	169.7	1.23
D6S308	Recessive	6q23.2	135.8	1.95, †
D7S484	Recessive	7p14.3	50.8	2.14, †
D7S515	Recessive	7q22.1	112.8	1.25
D7S530	Recessive	7q32.1	135	1.79
D10S208	Recessive	10p11.22	61.9	1.70
D10S1686	Recessive	10q23.1	107.3	1.75
D21S1914	Recessive	21p13	8.6	1.66

*  Analysis under dominant or recessive models with disease allele frequency (P_d) = 0.01, penetrance (f) for gene carriers = 0.8, and a phenocopy rate for non-gene-carriers = 0.01.

†  Suggestive evidence of linkage (LOD scores greater than 1.9).

Figure 2.

 

HLOD scores (solid lines) for all chromosomes in parametric multipoint linkage analysis and NPL scores (broken lines) in NPL analysis of 55 Japanese families with comitant strabismus. HLOD scores are calculated on the assumption of disease allele frequency at 0.01, penetrance at 0.8 (80%), and a phenocopy rate for non–gene carriers at 0.01 under a dominant model of inheritance.

Table 4.

 

Multipoint Linkage Analysis in 55 Japanese Families with Comitant Strabismus

Chromosome	Adjacent Marker	Model of Inheritance	Cytogenetic Location	Position of Maximum LOD (cM)	1-LOD CI (cM)	HLOD Score	NPL Score	P *
1	D1S207	Recessive	1p31.3	113.58	46.78	2.07	1.96	0.01
1	D1S413	Dominant	1q31.1	206.40	27.74	2.10	2.19	0.008
4	D4S1575	Dominant	4q28.3	130.60	13.02	3.62, †	2.68	0.001
7	D7S484	Recessive	7p14.3	50.80	10.80	2.32	2.32	0.005
7	D7S486	Recessive	7q31.2	125.20	42.46	2.32	2.70	0.001
8	D8S284	Dominant	8q24.13	134.28	25.50	2.83	2.28	0.006
11	D11S1320	Dominant	11q24.2	135.54	16.20	1.97	1.87	0.02
20	D20S195	Dominant	20q11.23	45.82	16.32	2.01	2.39	0.004

*   P values are for NPL scores.

†  Significant evidence of linkage (LOD score greater than 3.3).

Table 5.

 

Heterogeneity Logarithm of the Odds (HLOD) Scores for the Marker D7S486, on the Assumption of Different Penetrances and Different Disease Allele Frequencies in Parametric Multipoint Linkage Analysis of 55 Japanese Families with Comitant Strabismus

	HLOD Score for D7S486 with Penetrance of*
		0.9	0.8	0.7
Pd , †
0.1	3.80	3.88	3.93
0.01	2.60	2.32	2.03
0.001	0.16	0.06	0.09

*  Penetrance of dd genotype (gene carriers) for a recessive model.

†  P_d = disease allele frequency at disease locus.

Figure 3.

 

HLOD scores (solid line) on chromosome 7 on the assumption of disease allele frequency at 0.1 and penetrance of 70% for a recessive model of inheritance. NPL scores (broken line) on chromosome 7 are also shown in parallel. Note that HLOD score increases to 3.93 (NPL = 2.45, P = 0.003) from 2.32 (NPL = 2.70, P = 0.001, see Tables 4 and 5 ) and that the point of maximum HLOD was relocated from marker D7S486 (at 125.2 cM) to a region between markers D7S657 and D7S515 (at 107.28 cM).

Table 6.

 

Parametric Multipoint Linkage Analysis and NPL Analysis in Either Esotropia Subgroup (25 Families) or Exotropia Subgroup (26 Families)

Subgroup	Chromosome	Adjacent Marker	Cytogenetic Location	Position of Maximum HLOD (cM)	HLOD Score	NPL Score	P *
Esotropia	6	D6S281	6q27	189.78	2.95	2.37	0.002
Esotropia	8	D8S284	8q24.21	139.24	2.79	2.25	0.005
Esotropia	10	D10S1652	10q22.2	82.4	2.36	2.94	0.0003
Esotropia	15	D15S153	15q22.2	57.4	2.02	2.33	0.003
Esotropia	17	D17S1852	17p13.1	20.48	2.02	1.79	0.01
Exotropia	14	D14S288	14q21.3	42.34	1.91	1.85	0.02

*   P values are for NPL scores.

Figure 4.

 

HLOD scores on chromosome 8 in parametric multipoint linkage analysis of all 55 families (solid line) and in the esotropia subgroup with 25 families (broken line). HLOD is calculated on the assumption of disease allele frequency at 0.01, penetrance of 80%, and a phenocopy rate of 0.01 under a dominant model of inheritance. Note the HLOD score of 2.79 (NPL score = 2.25, P = 0.005) on 8q24.21 (139.28 cM) in the esotropia subgroup, which is 4.96 cM away from the suggestive locus at 134.28 cM in the analyses of all families.