Refined Localization of the Gene Causing X-Linked Juvenile Retinoschisis TIINA ALITALO, **t TORBEN A. KRUSE,$ AND ALBERT DE LA CHAPEt.LE*-t *Department
of Medical
Genetics, University of Helsinki, Helsinki, Finland; tFolkh;ilsan Institute of Genetics, Helsinki, Finland; and +/nstitute of Human Genetics, University of Aarhus, Aarhus, Denmark Received
June
25,
1990,
revised
November
2. 1990
spokes of a wheel, which results from creasing of the superficial retinal layer, andperipheral schisis. Abnormalities of the retina have been noted in very young patients (Forsius et al., 1971) and are probably present in infancy. The phenotypic expression of the disease varies widely, and generally there is a very slow deterioration that occurs throughout life. The pathogenesis is unknown, although at least three different theories have been proposed (Yanoff et al., 1968; Kawano et al., 1981; Schepens, 1983; reviewed by Laatikainen et al., 1987). With the exception of one report (Arden et al., 1988), no abnormalities have been reported in heterozygous carriers. The RS gene is located in the distal short arm of the X chromosome (Wieacker et al., 1983; Alitalo et al., 1987, 1988; Dahl et al., 1988; Gellert et al., 1988). Attempts at determining the precise location of the RS locus have been limited by the lack of highly informative markers in Xp22.1-~22.2 (Mandel et al., 1989), which has also hampered the physical mapping of the gene. For carrier detection and prenatal diagnosis of RS a few flanking markers linked to the disease gene have been available (Alitalo et al., 1988). DXS43 and DXS207 are located about 2 CM distal and DXS41 6 CM proximal to the gene. In practice, the diagnostic value of these linked RFLPs is sometimes limited by the lack of demonstrable heterozygosity at the marker loci in female carriers. In this study we have performed linkage analyses with four additional polymorphic markers from region Xp22.2-~21.3 to localize the RS gene with respect to these markers as well as to the closest markers used earlier (Alitalo et al., 1988). The resulting map serves to improve the reliability of DNA linkage information when used in carrier detection and genetic counseling. We also report the allele frequencies of the markers in the Finnish population. To determine whether there are significant differences in haplotype distributions between two geographically
Previous linkage studies in X-linked juvenile retinoschisis (RS) placed the gene between the loci DXS43 and DXS41 in the region Xp22.2-~22.1. Here we have extended our earlier studies by analyzing 31 RS families with the markers DXSl6 (pSE3.2-L), DXS274, DXS92, and ZFX. Pairwise linkage analysis revealed significant linkage of the RS gene to all markers used; locus DXS274 (probe CRI-L1391) was tightly linked to the disorder, with a lod score of 9.02 at a recombination fraction of 0.05. The genetic map around the RS locus was refined by multilocus linkage studies in an expanded database including a large set of normal families (40 CEPH families). The results indicated that the RS gene locus lies between (DXS207, DXS43) and DXS274 with oddsof 1.8 X 104:1 favoring this most likely location over the second most likely location, i.e., distal to DXS43. Analysis by LINKMAP gave a maximum location score of 136.4 with the order XpterDXS16 - (DXS207, DXS43) - RS - DXS274 - (DXS41, DXS92)-Xcen. To assess the diagnostic value of the markers in Finnish patients, a total of 12 markers were tested for allele frequencies in 126 Finnish unrelated blood donors. With the exception of the markers DXS207, DXS43, and DXS92, allele frequencies did not show any significant deviation from the data published elsewhere. Haplotype analysis was performed with five DNA markers flanking the RS locus. Patients from southwest Finland had a haplotype association that differed from the haplotype association found in the patients from north central Finland, favoring the hypothesis that the mutations in the two groups arose independently. cr; issi Academic press. I~C.
INTRODUCTION
Juvenile retinoschisis (RS; MIM No. 31270, McKusick, 1988) was recognized as an X-linked disorder in 1913 (Pagenstecher, 1913). The most typical clinical feature is poor sight that cannot be corrected with glasses. In most cases the disorder is characterized by cystic degeneration in the fovea, recalling the 505
All
Copyright w 1991 rights of reproduction
hy Academic Press, Inc. in any form reserved.
506
ALITALO,
KRUSE.
AND
separated groups of patients as well as between patients and normal individuals from the same region, we analyzed haplotype associations with five markers. MATERIALS
AND
METHODS
Subjects
Thirty-one Finnish families (22 from southwest Finland, 8 from north central Finland, and 1 from southern Finland) with X-linked recessive retinoschisis were analyzed. A total of 232 individuals from these kindreds were available for study. The families, for which data on the segregation of DXS89, DXS85, DXS16 (pXUT23), DXSS, DXS207, DXS43, DXS197, DXS41, and DXS164 had been published earlier (Alitalo et al., 1988), were now further analyzed with the markers DXS16 (pSE3.2-L), DXS274, DXS92, and ZFX. Additional data for DXS197 were obtained from two families, previously untyped for this marker. Descriptions of the families have been published earlier (Alitalo et al., 1988). For the allele frequency studies and haplotype analysis, 5 ml blood in EDTA was obtained from normal blood donors from the Pori (southwest Finland; 68 individuals) and Oulu (north central Finland; 58 individuals) regions.
DE
LA
(:HAPELI,E
by multilocus likelihood calculations using the ILINK and LINKMAP programs. In ILINK calculations data from the CEPH database version 3 were included to improve estimates of genetic distances. Genetic distances were calculated using Haldane’s mapping function, which converts recombination fractions to genetic distances in centimorgans. The frequency of the RS gene in the general population was taken to be 0.0001, and the published RFLP allele frequencies were used (Table 1). Results with the two probes of the DXS16 locus were combined. Allele
Frequencies
Analysis
Methods for sample preparation have been documented elsewhere (Kunkel et al., 1977; Alitalo et al., 1987). DNA samples of 5 pugwere digested to completion using the appropriate restriction enzyme (Table 1). The fragments were separated by 0.7% agarose gel electrophoresis and transferred by the method of Southern (1975) to nitrocellulose membranes (Schleicher & Schuell, BA 85). For hybridization, the probes were radiolabeled by nick translation or random priming to a specific activity of approximately 5 X 10’ cpm/pg. Prehybridization and hybridization with [32P]dCTP-labeled probes were carried out according to standard procedures (Alitalo et al., 1987). Posthybridization washes were carried out at 65”C, twice at 3~ SSC and 0.1% SDS for 20 min and subsequently at increasing stringencies (1-0.1X SSC with 0.1% SDS), depending on the probe. The filters were exposed to Fuji X-ray film at -70°C for l-4 days with an intensifying screen. Linkage
Analysis
Genotypic data were obtained using the probes listed in Table 1. Pairwise and multilocus linkage analyses were performed using the LINKAGE 4.7 package of computer programs (Lathrop and Lalouel, 1984; Lathrop et al., 1984). Gene order was examined
Analysis
Comparisons of the allele frequencies were done by means of the x2 test. The marker genotypes and DXS207/DXS43/DXS197/DXS274/DXS41 haplotypes for RS chromosomes were derived from inspection of family data. The phase in one affected individual from each family was used in the analysis. Two families were excluded from the haplotype analysis; in one family the phase for the RFLP alleles could not be determined, and another was not known to originate either in the Pori or in the Oulu region. RESULTS Two-Point
DNA
and Haplotype
Linkage
Analysis
Results of the two-point linkage analysis are summarized in Table 2. The analysis revealed significant linkage of the RS gene to all markers used. Only three recombinations were found between RS and DXS274, giving a maximum lod score of 9.02 at a recombination fraction of 0.05. Due to a new probe pSE3.2-L, a total of 88 meioses informative for RS and DXS16 (pXUT23 and pSE3.2-L combined) were now observed, giving Z,,, = 12.60 at 8,,, = 0.06. A maximum lod score of Il.55 was obtained for DXS92 versus RS at a recombination fraction of 0.06, and a significant lod score was also obtained for ZFX versus RS, although with looser linkage. The new lod scores for RSDXS197 are higher than those previously reported (Alitalo et al., 1988), reflecting the addition to the analysis of a few previously untyped individuals of our original family panel. Multipoint
Linkage
Analysis
Multilocus linkage analyses were used to determine the most probable gene order and to evaluate the evidence against alternative orders. To determine the most likely position of ZFX, DXS274, DXS16, and DXS92 with respect to the five known markers around RS [DXS85-(DXS207,DXS43)-RS-DXS41DXS164; Alitalo et al., 19881, we conducted various seven-point linkage analyses using the genotypic data
REFINED
LOCALIZATION
OF
TABLE Allele
Frequencies
of Polymorphic
Locus
Probe
DXS89
pTAKlOa
MspI
DXS85
L782
EcoRI
DXS16
pXUT23
HglII
pSE.?.Z-L
Mspl
RC8
?‘nqI
Markers:
Enzyme
1
Comparison
Allele length (kb)
between
No. of S chromosomes scored
7.4 6.4 14.0 7.0 17.5
pPA4B pD2
PCUII
DXS197
pTS247
&lII
DXS’74
CKI-L1391
MspI
p99-6
PST1
DXS9‘7
pXG-16
Hind111
ZFX
pDP1039
MspI
DXS4
1
Nole. ’ For * P ** P *** Y
N.A., not available. references, see Kidd < 0.05. < 0.01. < 0.001.
108
116 “1 191 4 185
3.2 3.0 12.0 9.5, 2.5
DXS43
6.6 6.0 "5.0 20.0
40 69 154 38 188
11.0 85. 2.5 22.0 13.0 9.5
79 147 149 $3 137
5.5 5.3
89 140
5.1
86
Two-Point
D?s16 DXS197 DXS274 DXS9” ZFX
pSUT23 and pSE3.2.L pTS247 CRI-L1391 pXG-16 pDP1039
Reported allele frequency
0.19 0.81
0.19 0.81
0.64 0.36 0.78 0.22 0.48 0.52 0.10 0.88 0.02 0.82
0.60 0.40 0.84 0.16 0.53 0.47 0.7 0 0.84 0.06
0.61 0.39 0.4s o.s5 0.26
0.18 0.41 0.69 0.17 0.83 0.35 0.6:3 0.67 0.33 Cl.61 0.39 0.62 0.38
0.74
N.A. 0.56: 0.44 030 o.so 0.52 0.48
-
(DXSZO7,DXS43)-DXS274-(DXS4l,DXS92), we conducted a seven-point analysis, changing the order of the loci only by changing the position of RS. Recombination rates were estimated with the ILINK program by means of maximum likelihood under each order tested. The results are shown in Table 3. The analysis strongly suggested that the most likely location of the RS gene is between (DXS207,DXS43) and
TABLE
Probe
Frequencies”
et al. (1989).
from both RS and CEPH families (data not shown). The most likely gene order for the markers was used in furt,her analyses. Due to lack of recombinations, the location of DXS197 with respect to the other loci could not be determined on the basis of linkage methods. To determine the most likely position of RS with respect to six previously mapped markers DXSlG-
Locus
-
and Reported
Observed allele frequency
SO
7.0 5 .5 5.3
XbaI
Observed
43 183 133 74 176
12.5
DXS207
507
RETINOSCHISIS
Recombinants/ informative meioses
Linkage
Analysis
2
between
Lad score
RS and Five
at recombination
Marker
Loci
fractions
0.00
0.001
0.01
0.05
0.10
0.15
0.20
0.30
0.40
z,,,
8,,,
Confidence limits
5188
~ ,-x
5.68
10.40
12.55
12.25
11.20
9.80
6.37
2.67
12.60
0.06
0.02-0.13
O/22
4.82 --x ~ ,x,
4.81 5.16 4.66
4.72 7.93
4.33 9.02
3.82 8.61
3.30 7.78
2.75 6.75
1.64 4.39
0.58 1.88
4.82 9.02
0.00 0.05
0.00-0.10 0.01-0.14
9.36
11.50
11.23
10.26
8.96
5.76
2.32
11.55
U.06
3.97
5.09
5.16
4.76
3.26
1.37
5.20
0.1:i
0.02-0.14 0.06-0.24
3/48 4/67 B/50
- *,
+.41
-0.63
508
ALITALO,
TABLE Localization
’ Estimated 0.00.
0.02,
0.05,
recombination 0.03,
Frequencies
LA
CHAPELLE
Odds DXS92
1.1.1 1:7.% 1:1.8 1:l 1:5.6 1:9.6 1:5.2
DXS41mDXS92
-RS-DXS92
fractions
hetween
adjacent
loci:
x 106 I\ 106 x 10’ x 107 X lo* x 10” 4
0.02,
l~l~i’,‘,~l’,~,‘,‘,~,‘,~,~I 4 -4 -2 I) 2
0.00.
(ienetic
DXS274, with odds of 1.8 X 104:1 favoring this location over the second best location. These results clearly excluded the location of RS proximal to DXS274, with a relative likelihood value of 1:5.6 x 107. An additional test with the same markers was performed with the LINKMAP program, which calculates the likelihood for the position of the disease locus with respect to a fixed map of markers. Although it was impossible to resolve definitely the order of the markers DXS207 and DXS43 and of DXS41 and DXS92, because no recombinations were observed between the loci in each of the two pairs, we found it reasonable to assume that those loci are not identical but are separated by small but unknown distances. Based on this assumption we conducted seven-point analysis with RS and the markers DXS16-DXS207DXS43-DXS274-DXS41-DXS92, in that order and with recombination fractions of 0.02, 0.01, 0.07, 0.03, and 0.01. The analysis gave a peak location score of 136.4 for the RS locus being located between DXS43 and DXS274 (Fig. 1). Allele
DE
Linkage
order
RS-DXSlG~DXS207FDXS43~DXS274DSS41 DXSl6-RS-DXS207~DXS43-DXS274-DXS41-DXSS’L DXS16-DXS207~RS~DXS43~DXS274~ DXSlG-DXS207-DXS43-RS-DXS274-DXS41-DXS92 DXSlG-DXS207-DXS43-DXS274-RS-DXS4l~DXS92 DXS16-DXS207TDXS43mDXS274-DXS41 DXS16-DXS207mDXS43mDXS274-DXS4l-DXS92RS
AND
3
of RS Using Seven-Point Analysis (ILINK) Locus
KRUSE,
and Haplotype
6
x
10
distance
II
14 16 18
(CM)
FIG. 1. LINKMAP analysis of the location of the RS locus. The location scores were calculated for different locations of the RS gene relative to a fixed map of six markers. The locus DXS207 was arbitrarily set at 0.00, and the other loci were positioned from it according to the genetic distances as described in the text.
genotype. The most frequently observed haplotype in normal chromosomes was the haplotype D (28%) while only 5% of RS chromosomes showed this pattern. In the eight RS families from Oulu, none of the RS chromosomes carried the A haplotype. Instead the majority (75%) of RS chromosomes carried the haplotype F, whereas only 16% of the normal chromosomes from Oulu showed this genotype. Again, the most frequently observed haplotype in normal chromosomes was the haplotype D (43%) while only 12.5% of RS chromosomes showed this pattern.
TABLE
4
Haplotype Analysis of 29 Retinoschisis (RS) and 76 Normal Chromosomes Using the Markers DXS207 (Aa), DXS43 (Dd), DXS197 (Ss), DXS274 (Ii), and DXS41 (Nn)”
Analysis
The allele frequencies for the 12 RFLPs in the Finnish population are shown in Table 1. The results for 9 markers are in good agreement with those of other studies, while 3 markers showed statistically significant differences from those published elsewhere. When the chromosomes were subdivided into Pori and Oulu categories, and the allele frequencies were compared with each other, significant differences emerged only in the allele frequencies of DXS197 (P -C0.05; data not shown). The result of the haplotype analysis (Table 4) showed that 76% of RS chromosomes in patients from 21 families in Pori shared a common DNA marker haplotype (the A haplotype), whereas only 5% of the normal chromosomes from Pori showed this
4
Number
of chromosomes
Retinoschisis Pori 21
Haplotype
AdSiN ADsIN AdsiN AdsIN Adsin ADsin adsin
16 3 1 1
(76%) (14%) (5%) (5%) -
Other u The capital letter refers smaller letter to the smaller
(‘%) Normal
Oulu 8
1 (12.5%;) 1 (1‘2.5%) 6 (75%)
2 1 2 11 4 2 5 3
Pori
Oulu
39
37
(5%) (3%) (5%) (28%)
9
to the larger fragment fragment size.
2 1 :I 16 (43%) 4 (11%) 6 (16%) 2
:1
size, and the
REFINED
LOCALIZATION
DISCUSSION
The results reported here allow us to map unambiguously the locus for X-linked retinoschisis between (DXS207,DXS43) and DXS274. Locus DXS274 represents a new marker closely linked to RS. It is the closest proximal marker currently available and thus of importance for carrier detection and prenatal diagnosis. DXS92 was also found to be located proximal to RS, close to DXS41, thus providing another valuable marker for diagnostic purposes. With the aid of CEPH data and new probes, our multilocus analysis resulted in a more refined estimate of the distance between flanking loci around RS. With the exception of Gellert et al. (1988), all of the few linkage studies between RS and anonymous RFLPs published to date have disclosed close linkage to DXS43 and DXS41 (Alitalo et al., 1988; Dahl et al., 1988; van Schooneveld et al., 1989; Sieving et al., 1989). Thus, current results are consistent with a homogeneous etiology of RS. Whether the apparent genetic homogeneity is contradicted by a marked variability in the phenotypic expression remains an open question. Different mutations of a single gene might give rise to widely different clinical features, such as in abnormalities of the dystrophin gene (Koenig et al., 1989). However, we note that there is great variability in RS phenotypes even within the Finnish families from the Pori region, where many of the families belong to one superpedigree and thus may have the same mutation (Vainio-Mattila et al., 1969). This may point to reasons other than heterogeneity for the clinical variation, such asmodifying genes or environmental effects on the retina at organogenesis or later. Based on the increasing recognition that deletions are responsible for genetic disorders, we assume that retinoschisis may also be caused by deletions. Alternatively or in addition, point mutations may occur such as the one in the rhodopsin gene that leads to autosoma1 dominant retinitis pigmentosa (Dryja et al., 1990). Which strategies should be used to approach the RS gene by physical mapping techniques is not entirely clear. We currently estimate the genetic distance between the closest. markers flanking the gene at some 7 CM, which may mean that RS is probably not more than 2-4 Mb away from the closest one. As stated in our previous study (Alitalo et al., 1988) there is one phase-unknown recombination between RS and (DXS207,DXS43) and another between RS and DXS207. Of the three recombinations found between RS and DXS274, two are phase-known. The fact that we now have found five recombination breakpoints in the region bracketed by DXS207 on the one side and DXS274 on the other will allow us to map RS more precisely as soon as more informative markers are isolated from that. interval.
OF
509
RETINOSCHISIS
Many of our families from the Pori region (southwest Finland) belong to one superpedigree (VainioMattila et al., 1969), and the families from the Oulu region (north central Finland) can be traced to a few pedigrees. As no genealogical connections between these two groups can be established (H. Forsius, personal communication), the occurrence of two geographically widely separated agglomerations could be explained in two alternative ways. First, there might be two or more different mutations. Second, both groups of families could originate from the same mutation in an early founder, but the geographic separation took place so long ago that genealogical studies have been unable to demonstrate a common ancestry. As our haplotype analysis showed a very profound difference between t,he two groups, we tentatively favor the hypothesis that at least two independent mutations are responsible for RS in Finland. The analysis also supports the theory that the families from southwest Finland have a common founder and that the families in north central Finland may have another common founder. In support of the hypothesis that one mutation is responsible for all Oulu patients is t,he fact that all patient haplotypes differing from the most common patient haplotype of the region can be created from this most common haplotype by a single recombinational event. This is also the case for the Pori region (assuming that DXS197 is distal t,o RS). At least two recomhinational events would have had to take place between DXS43 and DXS41 to explain a common origin for the Pori and Oulu mutations, indicating that the two mutations are probably independent. However, haplotype analyses of this type have only limited power of resolution, and the interpretations should he made with caution. Because there are a few families in Finland that can be genealogically traced to neither the southwestern nor north central pedigrees, they might represent different mutations. We note with interest that the peculiar Finnish population structure, due to small numbers of founders, profound isolation, and genetic drift (Norio et al., 1973; Vogel and Motulsky, 1986), does not necessarily mean that rare genetic disorders that are enriched in Finland are all caused by the same mutation. For example, in another developmental eye disorder, gyrate atrophy of the choroid and retina, at least two different mutations occur in Finland (Mitchell et al., 1989). We conclude that it is only by identifying and isolating the RS gene, and by characterizing its mutations, that these quest,ions can he answered. ACKNOWLEDGMENTS We are grateful to Professors H. Forsius and A. W. Eriksson making this study possible by giving us access to RS families.
for We
510
ALITALO,
KRUSE,
AND
DE
LA
13.
LAATIKAINEN, I,., TARKKANEN, A., AND QAKSELA, HereditaryX~linkedretinoschisisandbilateralcongenital inal detachment. Retina 7: 24-27.
14.
LATHROP, G. M.. AND LALOUEL, J. M. (1984). Easy tions of lod score and genetic risks on small computers. J. Hum. (km%. 36: 460-465.
15.
I,A’I’HROP, G. M., I,ALOUEL, J. M., JULIER, C., AND OTT, J. (1984). Strategies for multipoint linkage analysis in humans. Proc. Natl. Acad. Sci. (ISA 81: 3443-3446.
16.
MANUEL, J-L., WILLARD, H. F., NUSSBA~IM, R. L., ROMEO, G., PUCK, d. M., AND DAVIES, K. E. (1989). Report of the committee on the genetic constitution of the X chromosome. QQ&wut C’f’ll Genet.51: 384-437.
17.
MCKUSICK. V. A. (1988). “Mendelian Inheritance in Man: Catalogs of Autosomal Dominant, Autosomal Recessive, and X-Linked Phenot,ypes,” 8th ed., p. 1375. Johns Hopkins liniv. Press. Baltimore. MD.
18.
MI’~CHELL, G. A., BHODY, L. C., SIPILA, I., LOONEY, J. E., WONG, C.. ENCELHARDT, .J. F., PATEL, A. S., STEEL, G.. OBIE, C., KAISER-KUPFER, M., AND VALLE, II. (1989). At least two mutant alleles of ornithine 6-aminotransferase cause gyrate atrophy of the choroid and retina in Finns. Proc. Natf. Acad. Sci. USA 86: 197-201.
19.
NORI~, R.. NEVANLINNA, t 1973). Hereditary diseases Ann. C’lin. Rcs. 5: 109-141.
FORSIUS, H., ERIKSSON, A., NUUTILA, A., VAINIO~MATTILA, B., AND KRAUSE, U. (1971). A genetic study of three rare retinal disorders: Dystrophia retinae dysacusis syndrome, Xchromosomal retinoschisis and grouped pigments of the retina. Birth Dejects 7(3): 83-98.
‘0.
PAGENSTECWER, H. E. t 1918). Iieber eine unter dem Bilde der Netzhautahlosung verlaufende, erbliche Erkrankung der Retina. Gmefe”s Arch. (‘lin. Exp. fphthalmol. 86: 457-462.
21.
S~:HEPENS, (7. 1,. (1983). “Retinal Detachment and Diseases.” Vol. 2, pp. 568-588. Saunders, Philadelphia.
FORSIUS, H., KRAUSE, U., HELVE, J., VUOPALA, V., MusTONEN, E.. VAINIO-MATTILA, B.. FELLMAN, .J.. ANO ERIKSSON, A. W. (1973). Visual acuity in 183 cases of X-chromosomal retinoschisis. Canad. J. CIphthalmol. 8: 385-393. GELLERT, G., PETERSEN, J.. KRAWCZAK, M., AND ZOLL, B. (1988). Linkage relationship between retinoschisis and four marker loci. Hum. &net. 79: 382-384.
“2.
SIEVING, P., BINGHAM, E.. ROTH, M., YOUNG, M.. BOEHNKF,, M.. Kuo, S., ANL) GINSBERG, D. (1989). Linkage analysis of iuvenile X-linked retinoschisis using four DNA markers. Amer. J. Hum. Cknr~t. 45: Al61.
23.
SOU.~HERN, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Aid. 98: 5OB 517.
24.
VAINIO-MATTILA, B., ERIKSSON. A. W., AND FORSIIJS, H. (1969). X-chromosomal recessive retinoschisis in the region of’ Pori. Actn Ophthalmol. 47: 1135-I 148.
25.
VAN SCHOONEVELD. M., HOGENKAMP, TH., ORTH, U., NEUGEBAUER, M.. LISCH, K., BLEEKER-WAGEMAKERS, E. M., ANI) GAL, A. (1989). Linkage studies in X-linked juvenile retinoschisis. C~vtogewf. (‘~11 I&net. 51: 1096.
26.
VOGEL, F., ANU MOTULSKY, Problems and Approaches,” Verlag. Berlin/Heidelberg.
27.
WIEACKER, I’.. WIENKER, T. F., DALLAPICCOLA, B., BENDER, K.. DAVIES, K. E., AND ROPERS, H. H. (1983). Linkage relationships between retinoschisis, Xg, and cloned DNA sequence from the distal short arm of the X chromosome. Hum. Cknet. 64: 14:1-14.5.
“8.
YANOFF, M.. Histopathology 79: 49-53.
thank Dr. P. Sistonen for providing t he blood samples for the study of allele frequencies. We thank Drs. K. Davies, H. Donis-Keller. P. Green, L. Kunkel, S. Latt, D. C. Page, P. L. Pearson, H. F. Willard, and S. Wood for probes. This study was supported by grants from the Sigrid .JusBlius Foundation, the Academy of Finland and the Emil Aaltonen Foundation.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
ALITALO, T., KAKNA, J., FOI~SIUS, H., AND DE LA CHAPELLE, A. (1987). X-linked retinoschisis is closely linked to DXS41 and DXS16 hut not DXS85. C/in. Cknet. 32: 192-195. ALITALO, T., FORSIUS, H., KAI
of Medical
Genetics, University of Helsinki, Helsinki, Finland; tFolkh;ilsan Institute of Genetics, Helsinki, Finland; and +/nstitute of Human Genetics, University of Aarhus, Aarhus, Denmark Received
June
25,
1990,
revised
November
2. 1990
spokes of a wheel, which results from creasing of the superficial retinal layer, andperipheral schisis. Abnormalities of the retina have been noted in very young patients (Forsius et al., 1971) and are probably present in infancy. The phenotypic expression of the disease varies widely, and generally there is a very slow deterioration that occurs throughout life. The pathogenesis is unknown, although at least three different theories have been proposed (Yanoff et al., 1968; Kawano et al., 1981; Schepens, 1983; reviewed by Laatikainen et al., 1987). With the exception of one report (Arden et al., 1988), no abnormalities have been reported in heterozygous carriers. The RS gene is located in the distal short arm of the X chromosome (Wieacker et al., 1983; Alitalo et al., 1987, 1988; Dahl et al., 1988; Gellert et al., 1988). Attempts at determining the precise location of the RS locus have been limited by the lack of highly informative markers in Xp22.1-~22.2 (Mandel et al., 1989), which has also hampered the physical mapping of the gene. For carrier detection and prenatal diagnosis of RS a few flanking markers linked to the disease gene have been available (Alitalo et al., 1988). DXS43 and DXS207 are located about 2 CM distal and DXS41 6 CM proximal to the gene. In practice, the diagnostic value of these linked RFLPs is sometimes limited by the lack of demonstrable heterozygosity at the marker loci in female carriers. In this study we have performed linkage analyses with four additional polymorphic markers from region Xp22.2-~21.3 to localize the RS gene with respect to these markers as well as to the closest markers used earlier (Alitalo et al., 1988). The resulting map serves to improve the reliability of DNA linkage information when used in carrier detection and genetic counseling. We also report the allele frequencies of the markers in the Finnish population. To determine whether there are significant differences in haplotype distributions between two geographically
Previous linkage studies in X-linked juvenile retinoschisis (RS) placed the gene between the loci DXS43 and DXS41 in the region Xp22.2-~22.1. Here we have extended our earlier studies by analyzing 31 RS families with the markers DXSl6 (pSE3.2-L), DXS274, DXS92, and ZFX. Pairwise linkage analysis revealed significant linkage of the RS gene to all markers used; locus DXS274 (probe CRI-L1391) was tightly linked to the disorder, with a lod score of 9.02 at a recombination fraction of 0.05. The genetic map around the RS locus was refined by multilocus linkage studies in an expanded database including a large set of normal families (40 CEPH families). The results indicated that the RS gene locus lies between (DXS207, DXS43) and DXS274 with oddsof 1.8 X 104:1 favoring this most likely location over the second most likely location, i.e., distal to DXS43. Analysis by LINKMAP gave a maximum location score of 136.4 with the order XpterDXS16 - (DXS207, DXS43) - RS - DXS274 - (DXS41, DXS92)-Xcen. To assess the diagnostic value of the markers in Finnish patients, a total of 12 markers were tested for allele frequencies in 126 Finnish unrelated blood donors. With the exception of the markers DXS207, DXS43, and DXS92, allele frequencies did not show any significant deviation from the data published elsewhere. Haplotype analysis was performed with five DNA markers flanking the RS locus. Patients from southwest Finland had a haplotype association that differed from the haplotype association found in the patients from north central Finland, favoring the hypothesis that the mutations in the two groups arose independently. cr; issi Academic press. I~C.
INTRODUCTION
Juvenile retinoschisis (RS; MIM No. 31270, McKusick, 1988) was recognized as an X-linked disorder in 1913 (Pagenstecher, 1913). The most typical clinical feature is poor sight that cannot be corrected with glasses. In most cases the disorder is characterized by cystic degeneration in the fovea, recalling the 505
All
Copyright w 1991 rights of reproduction
hy Academic Press, Inc. in any form reserved.
506
ALITALO,
KRUSE.
AND
separated groups of patients as well as between patients and normal individuals from the same region, we analyzed haplotype associations with five markers. MATERIALS
AND
METHODS
Subjects
Thirty-one Finnish families (22 from southwest Finland, 8 from north central Finland, and 1 from southern Finland) with X-linked recessive retinoschisis were analyzed. A total of 232 individuals from these kindreds were available for study. The families, for which data on the segregation of DXS89, DXS85, DXS16 (pXUT23), DXSS, DXS207, DXS43, DXS197, DXS41, and DXS164 had been published earlier (Alitalo et al., 1988), were now further analyzed with the markers DXS16 (pSE3.2-L), DXS274, DXS92, and ZFX. Additional data for DXS197 were obtained from two families, previously untyped for this marker. Descriptions of the families have been published earlier (Alitalo et al., 1988). For the allele frequency studies and haplotype analysis, 5 ml blood in EDTA was obtained from normal blood donors from the Pori (southwest Finland; 68 individuals) and Oulu (north central Finland; 58 individuals) regions.
DE
LA
(:HAPELI,E
by multilocus likelihood calculations using the ILINK and LINKMAP programs. In ILINK calculations data from the CEPH database version 3 were included to improve estimates of genetic distances. Genetic distances were calculated using Haldane’s mapping function, which converts recombination fractions to genetic distances in centimorgans. The frequency of the RS gene in the general population was taken to be 0.0001, and the published RFLP allele frequencies were used (Table 1). Results with the two probes of the DXS16 locus were combined. Allele
Frequencies
Analysis
Methods for sample preparation have been documented elsewhere (Kunkel et al., 1977; Alitalo et al., 1987). DNA samples of 5 pugwere digested to completion using the appropriate restriction enzyme (Table 1). The fragments were separated by 0.7% agarose gel electrophoresis and transferred by the method of Southern (1975) to nitrocellulose membranes (Schleicher & Schuell, BA 85). For hybridization, the probes were radiolabeled by nick translation or random priming to a specific activity of approximately 5 X 10’ cpm/pg. Prehybridization and hybridization with [32P]dCTP-labeled probes were carried out according to standard procedures (Alitalo et al., 1987). Posthybridization washes were carried out at 65”C, twice at 3~ SSC and 0.1% SDS for 20 min and subsequently at increasing stringencies (1-0.1X SSC with 0.1% SDS), depending on the probe. The filters were exposed to Fuji X-ray film at -70°C for l-4 days with an intensifying screen. Linkage
Analysis
Genotypic data were obtained using the probes listed in Table 1. Pairwise and multilocus linkage analyses were performed using the LINKAGE 4.7 package of computer programs (Lathrop and Lalouel, 1984; Lathrop et al., 1984). Gene order was examined
Analysis
Comparisons of the allele frequencies were done by means of the x2 test. The marker genotypes and DXS207/DXS43/DXS197/DXS274/DXS41 haplotypes for RS chromosomes were derived from inspection of family data. The phase in one affected individual from each family was used in the analysis. Two families were excluded from the haplotype analysis; in one family the phase for the RFLP alleles could not be determined, and another was not known to originate either in the Pori or in the Oulu region. RESULTS Two-Point
DNA
and Haplotype
Linkage
Analysis
Results of the two-point linkage analysis are summarized in Table 2. The analysis revealed significant linkage of the RS gene to all markers used. Only three recombinations were found between RS and DXS274, giving a maximum lod score of 9.02 at a recombination fraction of 0.05. Due to a new probe pSE3.2-L, a total of 88 meioses informative for RS and DXS16 (pXUT23 and pSE3.2-L combined) were now observed, giving Z,,, = 12.60 at 8,,, = 0.06. A maximum lod score of Il.55 was obtained for DXS92 versus RS at a recombination fraction of 0.06, and a significant lod score was also obtained for ZFX versus RS, although with looser linkage. The new lod scores for RSDXS197 are higher than those previously reported (Alitalo et al., 1988), reflecting the addition to the analysis of a few previously untyped individuals of our original family panel. Multipoint
Linkage
Analysis
Multilocus linkage analyses were used to determine the most probable gene order and to evaluate the evidence against alternative orders. To determine the most likely position of ZFX, DXS274, DXS16, and DXS92 with respect to the five known markers around RS [DXS85-(DXS207,DXS43)-RS-DXS41DXS164; Alitalo et al., 19881, we conducted various seven-point linkage analyses using the genotypic data
REFINED
LOCALIZATION
OF
TABLE Allele
Frequencies
of Polymorphic
Locus
Probe
DXS89
pTAKlOa
MspI
DXS85
L782
EcoRI
DXS16
pXUT23
HglII
pSE.?.Z-L
Mspl
RC8
?‘nqI
Markers:
Enzyme
1
Comparison
Allele length (kb)
between
No. of S chromosomes scored
7.4 6.4 14.0 7.0 17.5
pPA4B pD2
PCUII
DXS197
pTS247
&lII
DXS’74
CKI-L1391
MspI
p99-6
PST1
DXS9‘7
pXG-16
Hind111
ZFX
pDP1039
MspI
DXS4
1
Nole. ’ For * P ** P *** Y
N.A., not available. references, see Kidd < 0.05. < 0.01. < 0.001.
108
116 “1 191 4 185
3.2 3.0 12.0 9.5, 2.5
DXS43
6.6 6.0 "5.0 20.0
40 69 154 38 188
11.0 85. 2.5 22.0 13.0 9.5
79 147 149 $3 137
5.5 5.3
89 140
5.1
86
Two-Point
D?s16 DXS197 DXS274 DXS9” ZFX
pSUT23 and pSE3.2.L pTS247 CRI-L1391 pXG-16 pDP1039
Reported allele frequency
0.19 0.81
0.19 0.81
0.64 0.36 0.78 0.22 0.48 0.52 0.10 0.88 0.02 0.82
0.60 0.40 0.84 0.16 0.53 0.47 0.7 0 0.84 0.06
0.61 0.39 0.4s o.s5 0.26
0.18 0.41 0.69 0.17 0.83 0.35 0.6:3 0.67 0.33 Cl.61 0.39 0.62 0.38
0.74
N.A. 0.56: 0.44 030 o.so 0.52 0.48
-
(DXSZO7,DXS43)-DXS274-(DXS4l,DXS92), we conducted a seven-point analysis, changing the order of the loci only by changing the position of RS. Recombination rates were estimated with the ILINK program by means of maximum likelihood under each order tested. The results are shown in Table 3. The analysis strongly suggested that the most likely location of the RS gene is between (DXS207,DXS43) and
TABLE
Probe
Frequencies”
et al. (1989).
from both RS and CEPH families (data not shown). The most likely gene order for the markers was used in furt,her analyses. Due to lack of recombinations, the location of DXS197 with respect to the other loci could not be determined on the basis of linkage methods. To determine the most likely position of RS with respect to six previously mapped markers DXSlG-
Locus
-
and Reported
Observed allele frequency
SO
7.0 5 .5 5.3
XbaI
Observed
43 183 133 74 176
12.5
DXS207
507
RETINOSCHISIS
Recombinants/ informative meioses
Linkage
Analysis
2
between
Lad score
RS and Five
at recombination
Marker
Loci
fractions
0.00
0.001
0.01
0.05
0.10
0.15
0.20
0.30
0.40
z,,,
8,,,
Confidence limits
5188
~ ,-x
5.68
10.40
12.55
12.25
11.20
9.80
6.37
2.67
12.60
0.06
0.02-0.13
O/22
4.82 --x ~ ,x,
4.81 5.16 4.66
4.72 7.93
4.33 9.02
3.82 8.61
3.30 7.78
2.75 6.75
1.64 4.39
0.58 1.88
4.82 9.02
0.00 0.05
0.00-0.10 0.01-0.14
9.36
11.50
11.23
10.26
8.96
5.76
2.32
11.55
U.06
3.97
5.09
5.16
4.76
3.26
1.37
5.20
0.1:i
0.02-0.14 0.06-0.24
3/48 4/67 B/50
- *,
+.41
-0.63
508
ALITALO,
TABLE Localization
’ Estimated 0.00.
0.02,
0.05,
recombination 0.03,
Frequencies
LA
CHAPELLE
Odds DXS92
1.1.1 1:7.% 1:1.8 1:l 1:5.6 1:9.6 1:5.2
DXS41mDXS92
-RS-DXS92
fractions
hetween
adjacent
loci:
x 106 I\ 106 x 10’ x 107 X lo* x 10” 4
0.02,
l~l~i’,‘,~l’,~,‘,‘,~,‘,~,~I 4 -4 -2 I) 2
0.00.
(ienetic
DXS274, with odds of 1.8 X 104:1 favoring this location over the second best location. These results clearly excluded the location of RS proximal to DXS274, with a relative likelihood value of 1:5.6 x 107. An additional test with the same markers was performed with the LINKMAP program, which calculates the likelihood for the position of the disease locus with respect to a fixed map of markers. Although it was impossible to resolve definitely the order of the markers DXS207 and DXS43 and of DXS41 and DXS92, because no recombinations were observed between the loci in each of the two pairs, we found it reasonable to assume that those loci are not identical but are separated by small but unknown distances. Based on this assumption we conducted seven-point analysis with RS and the markers DXS16-DXS207DXS43-DXS274-DXS41-DXS92, in that order and with recombination fractions of 0.02, 0.01, 0.07, 0.03, and 0.01. The analysis gave a peak location score of 136.4 for the RS locus being located between DXS43 and DXS274 (Fig. 1). Allele
DE
Linkage
order
RS-DXSlG~DXS207FDXS43~DXS274DSS41 DXSl6-RS-DXS207~DXS43-DXS274-DXS41-DXSS’L DXS16-DXS207~RS~DXS43~DXS274~ DXSlG-DXS207-DXS43-RS-DXS274-DXS41-DXS92 DXSlG-DXS207-DXS43-DXS274-RS-DXS4l~DXS92 DXS16-DXS207TDXS43mDXS274-DXS41 DXS16-DXS207mDXS43mDXS274-DXS4l-DXS92RS
AND
3
of RS Using Seven-Point Analysis (ILINK) Locus
KRUSE,
and Haplotype
6
x
10
distance
II
14 16 18
(CM)
FIG. 1. LINKMAP analysis of the location of the RS locus. The location scores were calculated for different locations of the RS gene relative to a fixed map of six markers. The locus DXS207 was arbitrarily set at 0.00, and the other loci were positioned from it according to the genetic distances as described in the text.
genotype. The most frequently observed haplotype in normal chromosomes was the haplotype D (28%) while only 5% of RS chromosomes showed this pattern. In the eight RS families from Oulu, none of the RS chromosomes carried the A haplotype. Instead the majority (75%) of RS chromosomes carried the haplotype F, whereas only 16% of the normal chromosomes from Oulu showed this genotype. Again, the most frequently observed haplotype in normal chromosomes was the haplotype D (43%) while only 12.5% of RS chromosomes showed this pattern.
TABLE
4
Haplotype Analysis of 29 Retinoschisis (RS) and 76 Normal Chromosomes Using the Markers DXS207 (Aa), DXS43 (Dd), DXS197 (Ss), DXS274 (Ii), and DXS41 (Nn)”
Analysis
The allele frequencies for the 12 RFLPs in the Finnish population are shown in Table 1. The results for 9 markers are in good agreement with those of other studies, while 3 markers showed statistically significant differences from those published elsewhere. When the chromosomes were subdivided into Pori and Oulu categories, and the allele frequencies were compared with each other, significant differences emerged only in the allele frequencies of DXS197 (P -C0.05; data not shown). The result of the haplotype analysis (Table 4) showed that 76% of RS chromosomes in patients from 21 families in Pori shared a common DNA marker haplotype (the A haplotype), whereas only 5% of the normal chromosomes from Pori showed this
4
Number
of chromosomes
Retinoschisis Pori 21
Haplotype
AdSiN ADsIN AdsiN AdsIN Adsin ADsin adsin
16 3 1 1
(76%) (14%) (5%) (5%) -
Other u The capital letter refers smaller letter to the smaller
(‘%) Normal
Oulu 8
1 (12.5%;) 1 (1‘2.5%) 6 (75%)
2 1 2 11 4 2 5 3
Pori
Oulu
39
37
(5%) (3%) (5%) (28%)
9
to the larger fragment fragment size.
2 1 :I 16 (43%) 4 (11%) 6 (16%) 2
:1
size, and the
REFINED
LOCALIZATION
DISCUSSION
The results reported here allow us to map unambiguously the locus for X-linked retinoschisis between (DXS207,DXS43) and DXS274. Locus DXS274 represents a new marker closely linked to RS. It is the closest proximal marker currently available and thus of importance for carrier detection and prenatal diagnosis. DXS92 was also found to be located proximal to RS, close to DXS41, thus providing another valuable marker for diagnostic purposes. With the aid of CEPH data and new probes, our multilocus analysis resulted in a more refined estimate of the distance between flanking loci around RS. With the exception of Gellert et al. (1988), all of the few linkage studies between RS and anonymous RFLPs published to date have disclosed close linkage to DXS43 and DXS41 (Alitalo et al., 1988; Dahl et al., 1988; van Schooneveld et al., 1989; Sieving et al., 1989). Thus, current results are consistent with a homogeneous etiology of RS. Whether the apparent genetic homogeneity is contradicted by a marked variability in the phenotypic expression remains an open question. Different mutations of a single gene might give rise to widely different clinical features, such as in abnormalities of the dystrophin gene (Koenig et al., 1989). However, we note that there is great variability in RS phenotypes even within the Finnish families from the Pori region, where many of the families belong to one superpedigree and thus may have the same mutation (Vainio-Mattila et al., 1969). This may point to reasons other than heterogeneity for the clinical variation, such asmodifying genes or environmental effects on the retina at organogenesis or later. Based on the increasing recognition that deletions are responsible for genetic disorders, we assume that retinoschisis may also be caused by deletions. Alternatively or in addition, point mutations may occur such as the one in the rhodopsin gene that leads to autosoma1 dominant retinitis pigmentosa (Dryja et al., 1990). Which strategies should be used to approach the RS gene by physical mapping techniques is not entirely clear. We currently estimate the genetic distance between the closest. markers flanking the gene at some 7 CM, which may mean that RS is probably not more than 2-4 Mb away from the closest one. As stated in our previous study (Alitalo et al., 1988) there is one phase-unknown recombination between RS and (DXS207,DXS43) and another between RS and DXS207. Of the three recombinations found between RS and DXS274, two are phase-known. The fact that we now have found five recombination breakpoints in the region bracketed by DXS207 on the one side and DXS274 on the other will allow us to map RS more precisely as soon as more informative markers are isolated from that. interval.
OF
509
RETINOSCHISIS
Many of our families from the Pori region (southwest Finland) belong to one superpedigree (VainioMattila et al., 1969), and the families from the Oulu region (north central Finland) can be traced to a few pedigrees. As no genealogical connections between these two groups can be established (H. Forsius, personal communication), the occurrence of two geographically widely separated agglomerations could be explained in two alternative ways. First, there might be two or more different mutations. Second, both groups of families could originate from the same mutation in an early founder, but the geographic separation took place so long ago that genealogical studies have been unable to demonstrate a common ancestry. As our haplotype analysis showed a very profound difference between t,he two groups, we tentatively favor the hypothesis that at least two independent mutations are responsible for RS in Finland. The analysis also supports the theory that the families from southwest Finland have a common founder and that the families in north central Finland may have another common founder. In support of the hypothesis that one mutation is responsible for all Oulu patients is t,he fact that all patient haplotypes differing from the most common patient haplotype of the region can be created from this most common haplotype by a single recombinational event. This is also the case for the Pori region (assuming that DXS197 is distal t,o RS). At least two recomhinational events would have had to take place between DXS43 and DXS41 to explain a common origin for the Pori and Oulu mutations, indicating that the two mutations are probably independent. However, haplotype analyses of this type have only limited power of resolution, and the interpretations should he made with caution. Because there are a few families in Finland that can be genealogically traced to neither the southwestern nor north central pedigrees, they might represent different mutations. We note with interest that the peculiar Finnish population structure, due to small numbers of founders, profound isolation, and genetic drift (Norio et al., 1973; Vogel and Motulsky, 1986), does not necessarily mean that rare genetic disorders that are enriched in Finland are all caused by the same mutation. For example, in another developmental eye disorder, gyrate atrophy of the choroid and retina, at least two different mutations occur in Finland (Mitchell et al., 1989). We conclude that it is only by identifying and isolating the RS gene, and by characterizing its mutations, that these quest,ions can he answered. ACKNOWLEDGMENTS We are grateful to Professors H. Forsius and A. W. Eriksson making this study possible by giving us access to RS families.
for We
510
ALITALO,
KRUSE,
AND
DE
LA
13.
LAATIKAINEN, I,., TARKKANEN, A., AND QAKSELA, HereditaryX~linkedretinoschisisandbilateralcongenital inal detachment. Retina 7: 24-27.
14.
LATHROP, G. M.. AND LALOUEL, J. M. (1984). Easy tions of lod score and genetic risks on small computers. J. Hum. (km%. 36: 460-465.
15.
I,A’I’HROP, G. M., I,ALOUEL, J. M., JULIER, C., AND OTT, J. (1984). Strategies for multipoint linkage analysis in humans. Proc. Natl. Acad. Sci. (ISA 81: 3443-3446.
16.
MANUEL, J-L., WILLARD, H. F., NUSSBA~IM, R. L., ROMEO, G., PUCK, d. M., AND DAVIES, K. E. (1989). Report of the committee on the genetic constitution of the X chromosome. QQ&wut C’f’ll Genet.51: 384-437.
17.
MCKUSICK. V. A. (1988). “Mendelian Inheritance in Man: Catalogs of Autosomal Dominant, Autosomal Recessive, and X-Linked Phenot,ypes,” 8th ed., p. 1375. Johns Hopkins liniv. Press. Baltimore. MD.
18.
MI’~CHELL, G. A., BHODY, L. C., SIPILA, I., LOONEY, J. E., WONG, C.. ENCELHARDT, .J. F., PATEL, A. S., STEEL, G.. OBIE, C., KAISER-KUPFER, M., AND VALLE, II. (1989). At least two mutant alleles of ornithine 6-aminotransferase cause gyrate atrophy of the choroid and retina in Finns. Proc. Natf. Acad. Sci. USA 86: 197-201.
19.
NORI~, R.. NEVANLINNA, t 1973). Hereditary diseases Ann. C’lin. Rcs. 5: 109-141.
FORSIUS, H., ERIKSSON, A., NUUTILA, A., VAINIO~MATTILA, B., AND KRAUSE, U. (1971). A genetic study of three rare retinal disorders: Dystrophia retinae dysacusis syndrome, Xchromosomal retinoschisis and grouped pigments of the retina. Birth Dejects 7(3): 83-98.
‘0.
PAGENSTECWER, H. E. t 1918). Iieber eine unter dem Bilde der Netzhautahlosung verlaufende, erbliche Erkrankung der Retina. Gmefe”s Arch. (‘lin. Exp. fphthalmol. 86: 457-462.
21.
S~:HEPENS, (7. 1,. (1983). “Retinal Detachment and Diseases.” Vol. 2, pp. 568-588. Saunders, Philadelphia.
FORSIUS, H., KRAUSE, U., HELVE, J., VUOPALA, V., MusTONEN, E.. VAINIO-MATTILA, B.. FELLMAN, .J.. ANO ERIKSSON, A. W. (1973). Visual acuity in 183 cases of X-chromosomal retinoschisis. Canad. J. CIphthalmol. 8: 385-393. GELLERT, G., PETERSEN, J.. KRAWCZAK, M., AND ZOLL, B. (1988). Linkage relationship between retinoschisis and four marker loci. Hum. &net. 79: 382-384.
“2.
SIEVING, P., BINGHAM, E.. ROTH, M., YOUNG, M.. BOEHNKF,, M.. Kuo, S., ANL) GINSBERG, D. (1989). Linkage analysis of iuvenile X-linked retinoschisis using four DNA markers. Amer. J. Hum. Cknr~t. 45: Al61.
23.
SOU.~HERN, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Aid. 98: 5OB 517.
24.
VAINIO-MATTILA, B., ERIKSSON. A. W., AND FORSIIJS, H. (1969). X-chromosomal recessive retinoschisis in the region of’ Pori. Actn Ophthalmol. 47: 1135-I 148.
25.
VAN SCHOONEVELD. M., HOGENKAMP, TH., ORTH, U., NEUGEBAUER, M.. LISCH, K., BLEEKER-WAGEMAKERS, E. M., ANI) GAL, A. (1989). Linkage studies in X-linked juvenile retinoschisis. C~vtogewf. (‘~11 I&net. 51: 1096.
26.
VOGEL, F., ANU MOTULSKY, Problems and Approaches,” Verlag. Berlin/Heidelberg.
27.
WIEACKER, I’.. WIENKER, T. F., DALLAPICCOLA, B., BENDER, K.. DAVIES, K. E., AND ROPERS, H. H. (1983). Linkage relationships between retinoschisis, Xg, and cloned DNA sequence from the distal short arm of the X chromosome. Hum. Cknet. 64: 14:1-14.5.
“8.
YANOFF, M.. Histopathology 79: 49-53.
thank Dr. P. Sistonen for providing t he blood samples for the study of allele frequencies. We thank Drs. K. Davies, H. Donis-Keller. P. Green, L. Kunkel, S. Latt, D. C. Page, P. L. Pearson, H. F. Willard, and S. Wood for probes. This study was supported by grants from the Sigrid .JusBlius Foundation, the Academy of Finland and the Emil Aaltonen Foundation.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
ALITALO, T., KAKNA, J., FOI~SIUS, H., AND DE LA CHAPELLE, A. (1987). X-linked retinoschisis is closely linked to DXS41 and DXS16 hut not DXS85. C/in. Cknet. 32: 192-195. ALITALO, T., FORSIUS, H., KAI
