| Human Molecular Genetics | Pages |
Genetic and physical mapping of the McKusick-Kaufman syndrome
Introduction
Results
Pedigree analysis
Genetic analysis
Physical mapping and haplotype analysis
Discussion
Materials And Methods
Pedigree analysis
Genotyping and linkage analysis
Linkage analysis
Physical mapping and sequencing
Acknowledgements
References
Genetic and physical mapping of the McKusick-Kaufman syndrome
McKusick-Kaufman syndrome is a human developmental anomaly syndrome comprising mesoaxial or postaxial polydactyly, congenital heart disease and hydrometrocolpos. This syndrome is diagnosed most frequently in the Old Order Amish population and is inherited in an autosomal recessive pattern with reduced penetrance and variable expressivity. Homozygosity mapping and linkage analyses were conducted using two pedigrees derived from a larger pedigree published in 1978. The PedHunter software query system was used on the Amish Genealogy Database to correct the previous pedigree, derive a minimal pedigree connecting those affected sibships that are in the database and determine the most recent common ancestors of the affected persons. Whole genome short tandem repeat polymorphism (STRP) screening showed homozygosity in 20p12, between D20S162 and D20S894, an area that includes the Alagille syndrome critical region. The peak two-point LOD score was 3.33, and the peak three-point LOD score was 5.21. The physical map of this region has been defined, and additional polymorphic markers have been isolated. The region includes several genes and expressed sequence tags (ESTs), including the jagged1 gene that recently has been shown to be haploinsufficient in the Alagille syndrome. Sequencing of jagged1 in two unrelated individuals affected with McKusick-Kaufman syndrome has not revealed any disease-causing mutations.
INTRODUCTION
In 1964, McKusick et al. demonstrated that hydrometrocolpos, the result of a transverse vaginal membrane or vaginal atresia, is inherited in an autosomal recessive pattern in the Old Order Amish (1). In 1972, Kaufman et al. showed that polydactyly and congenital heart disease may accompany the finding of hydrometrocolpos in this syndrome (2). Hydrometrocolpos is the most distinctive finding in individuals affected with McKusick- Kaufman syndrome. The reproductive tract obstruction is due to a failure to canalize the junction between the inferior uterus and the vagina, structures derived from the paramesonephric ducts and urogenital sinus, respectively (3). The accumulation of secretions causes distension of the obstructed uterus in puberty or in infancy due to the effect of maternal hormones. The resulting mass increases abdominal pressure and can cause respiratory compromise, edema from venous compression or hydronephrosis from ureteral compression (4). Emergency surgery may be necessary, and the complications can be life threatening. Cardiac malformations also cause morbidity, and a number of individuals from Amish sibships with McKusick-Kaufman have died of congenital heart disease. Accurate genetic diagnosis of individuals and families at risk would be useful for anticipatory prenatal and pediatric care. Because the disorder occurs in a closed population with a founder effect, this provides an opportunity to map the locus for McKusick-Kaufman syndrome by homozygosity mapping.
RESULTS
Pedigree analysis
The starting points for this study were the affected persons identified in the expanded Old Order Amish McKusick-Kaufman pedigree (5). Field work confirmed the findings in the previously described affecteds and identified an additional sibship not previously described. The pedigree was analyzed with the PedHunter query software package applied to the Amish Genealogy Database (AGDB) (Agarwala et al., submitted). The AGDB is a computerized, updated and corrected version of the published Old Order Amish genealogy (6). Four of the affected sibships can be found in this genealogy. First, PedHunter and AGDB were used to identify two errors in the original pedigree (1) (Fig. 1). Second, PedHunter with the AGDB and manual searching of an earlier edition of the published genealogy (7) were used to connect the mother of the newly discovered sibship to the original pedigree (5) and to determine that the father of this sibship cannot be connected to any member of the published pedigree or their known antecedents through available genealogy data. The most simple explanations for our inability to connect the father of this sibship to the large pedigree are that the chosen common ancestor of the pedigree is erroneous or our genealogical data are incomplete. Alternatively, the father must carry a second disease allele, which is very unlikely for a rare disorder in a kindred with a high consanguinity rate. Third, PedHunter and AGDB were used to connect the four sibships present in the database into a pedigree with the minimum number of people required to establish the relationship among the sibships, as shown in Figure 1. This minimal pedigree defines the relationships that describe the simplest path of transmission of the disease gene. Because the father of the new sibship cannot be connected to the pedigree, and because three of the original sibships are not in the AGDB source data, the full pedigree was split into two parts that were analyzed separately (Fig. 1 and 2). PedHunter and AGDB defined a pedigree for sibships 1-4 that connected both parents of all sibships to a single founder couple. Some linkage analyses were performed on the full pedigree, with the father of the new sibship specified as unrelated, to compare the results with those of the split pedigree.
Figure
Figure
Genetic analysis
Genotyping of the core family (sibship 1, Fig. 1) showed that the three affected children were homozygous for markers at 11 loci (13 markers) in the whole genome scan (385 markers). By adding additional affected individuals and by typing additional markers flanking the homozygous markers, all regions other than 20p were eliminated due to heterozygosity in multiple affected individuals. All affected individuals were homozygous for D20S160 and D20S1154, although D20S160 was not highly informative in this pedigree. To study this area further, additional markers were isolated from bacterial artificial chromosomes (BACs) in the region, as described below. An additional marker, NM2, showed homozygosity in all affecteds. Markers were ordered on the basis of physical mapping data in this region, which is an extension of the Alagille syndrome region physical map (8). Heterozygous markers in the region included D20S162, D20S901, NM1, D20S894, AFM164TG5 and D20S189. The physical ordering of the markers on the BAC clone-based Alagille syndrome region gave the order: Tel-(D20S162-D20S901-AFM164TG5)- NM1-NM2-D20S160-D20S1154-Cen. D20S894 and D20S189 did not map to this contig but were known to be centromeric of D20S1154 (8-10). Two-point LOD scores of the disease phenotype and markers in the region are shown in Table 1. The peak two-point LOD score is 3.33 at [thetas] = 0 with marker D20S1154. The LOD scores reported in the table were obtained using the split pedigrees. The results of two three-point analyses with two markers and the disease are shown in Figure 3. The highest three-point LOD score in these runs was 5.21. Comparison analyses done with the full pedigree showed consistently lower LOD scores (e.g. 3.33 for the split pedigree and 2.90 for the full pedigree) (Fig. 4).
NM1 and NM2 are polymorphic microsatellite markers [(GGAA)n and (CA)n repeats, respectively] that can be assayed by PCR. The primer sequences for NM1 are GAGTGGTTGGTTAATCACTC and CACCTGACCAAATTGTCTCA. The primer sequences for NM2 are TTGCAACTGCTAATCCACTA and GAATAGTGGGCTTAGAATCA. NM1 has a heterozygosity score of 0.78 with 10 alleles and NM2 has a heterozygosity score of 0.58 with three alleles. The allele sizes of the markers in CEPH individuals 1331-1 are 228 and 232 for NM1 and 232 and 234 for NM2. See Figure 5 for physical mapping data of these markers.Table
Marker
Recombination fraction
0
0.01
0.05
0.10
0.20
0.4
D20S162
1.97
1.91
1.67
1.37
0.83
0.13
NM1
2.93
2.84
2.47
2.02
1.18
0.17
NM2
1.74
1.67
1.40
1.10
0.60
0.07
D20S160
1.67
1.61
1.36
1.07
0.59
0.08
D20S1154
3.33
3.22
2.77
2.21
1.26
0.18
Physical mapping and haplotype analysis
Figure
Figure
Figure
The region of homozygosity in McKusick-Kaufman affecteds includes the jagged1 gene, known to cause Alagille syndrome. Alagille syndrome is a human developmental disorder that also causes congenital heart disease and other anomalies, although there is little overall similarity in the manifestations of these disorders. Because jagged1 mutations can cause human developmental anomalies, it is considered a candidate gene for McKusick-Kaufman syndrome. Twenty six exons and exon- intron boundaries (8) were sequenced but no alterations were found that were absent in the general population.
DISCUSSION
The positional cloning of genes that cause human developmental disorders is important for improved medical care and better understanding of the processes that underlie mammalian embryogenesis. In this study, the gene for a rare developmental anomaly, the McKusick-Kaufman syndrome, was mapped to 20p12. The study of a large inbred kindred allowed fine mapping of the gene to a <1 cM region that includes <500 kb of DNA according to the physical mapping data. The candidate region includes the jagged1 gene, and mutations in that gene were shown recently to cause the Alagille syndrome (8,10). Because this gene was known to cause another developmental disorder in humans when haploinsufficient, we hypothesized that amino acid substitutions in this gene might cause the McKusick-Kaufman syndrome. Sequencing of exons and exon-intron boundaries to search for mutations revealed no alterations in the coding region of the gene or the exon-intron boundaries. This sequencing does not entirely exclude this gene as a cause of McKusick-Kaufman syndrome as it is possible that there is a disease-causing mutation in the promoter. Another gene in the region, that for synaptosomal-associated protein (SNAP), is a candidate but appears to have brain-specific expression. A possible mouse model, ipv (imperforate vagina), has not been mapped but is inherited in an autosomal recessive pattern (11). Two other mouse mutants, dominant hemimelia (Dh) and loop tail (lp) include skeletal and urogenital anomalies (12,13). These two disorders map to mouse chromosome 1 (14,15), which is not syntenic to human chromosome 20. We conclude that the latter two disorders are not mouse models of McKusick-Kaufman syndrome because the regions are not syntenic.
The haplotype data show that three individuals (two females and one male) are homozygous for the disease-carrying haplotype but have no apparent manifestations of the disorder. A potential confounding factor in this determination is the degree of medical evaluation prevalent in the Amish community. Medical care is limited in this community and it is possible that subtle anomalies may not be diagnosed. One mother of an affected sibship and two other siblings of affected persons are homozygous in this region and apparently unaffected. It is possible that the homozygous, apparently unaffected, siblings may have a minor cardiac defect or, in the case of the female, an undiagnosed uterine anomaly. A review of 54 patients reported before 1987 showed that hydrometrocolpos was present in 95%, polydactyly in 93% and cardiovascular malformations in 9% (16). Clinical and pedigree studies of Amish families show that there are individuals who have all three anomalies, others with two anomalies in various pairwise combinations, and some individuals with only one of the anomalies of McKusick-Kaufman syndrome (5). For this study, the penetrance was modeled as three independent probabilities for each defect. Using these probabilities, we estimated that hydrometrocolpos affects 70% of females, polydactyly affects 60% of both sexes and congenital heart disease affects 15% of both sexes. These estimates predict that at least 9% of males and 3% of females would be expected to be non-penetrant for the disorder. Non-penetrance in autosomal recessive disorders is difficult to demonstrate in the absence of molecular genetic analysis. It may be impossible to prove non-penetrance for a pleiotropic disorder (even with molecular data) since subtle malformations may be difficult to diagnose. Although there are no clear human examples of this phenomenon (17), mice homozygous for the isoform IV mutation of the limb deformity gene (ld) show incomplete penetrance of renal agenesis or dysgenesis and non-penetrance for the limb malformations (18). The study described herein demonstrates that individuals homozygous for the Amish allele for McKusick-Kaufman syndrome may present with no apparent cardinal signs of the disorder.
The linkage and pedigree analyses were facilitated by the use of the PedHunter software system on the Amish Genealogy Database (Agarwala et al., submitted). PedHunter addresses a long-standing problem in genetic analysis by rapidly and systematically specifying pedigrees to connect distantly related affected individuals. The PedHunter pedigrees are selected according to mathematically rigorous criteria, whereas manual searches and pedigree construction yields pedigrees of unknown accuracy. For the purposes of linkage analysis, it is reasonable to hypothesize that a recessive disease in a small closed population will have a single common ancestral source for all disease-carrying alleles. Nevertheless, it is problematic to construct large, unified pedigrees in this situation when the genealogy data are complex, incomplete or erroneous. It is well known that this can make the linkage analysis intractable, either in time for the Elston-Stewart algorithm or in memory for the Lander-Green algorithm. In the analyses reported here, forcing the two sub-pedigrees to be connected reduced the LOD score. This occurs because the father of sibship 7 is not included in the consanguinity loops and is treated as unrelated, which lowers the LOD score. PedHunter can be used rigorously to construct pedigrees for any population that has a genealogy, which should increase the already high genetic value of such populations.
The mesoaxial polydactyly typical of McKusick-Kaufman syndrome is commonly seen in Pallister-Hall syndrome (MIM 146510), the oral-facial-digital syndromes, Ellis van Creveld syndrome (MIM 225500), Holt-Oram syndrome (MIM 142900), Holzgreve syndrome (MIM 236110) and several other disorders as an uncommon malformation. Recently, it has been noted that children with Bardet-Biedl (MIM 209900, 209901, 600374 and 60015) syndrome may present with hydrometrocolpos and polydactyly, but these children are usually retarded and later develop obesity and retinal dystrophy (19). The data presented here show that McKusick-Kaufman syndrome is not allelic to Pallister-Hall syndrome (GLI3, 7p13) (20), Holt-Oram syndrome (TBX5 on 12q24) (21,22), Ellis van Creveld syndrome (4p16) (23) or Bardet-Biedl syndrome (16q21, 11q13, 15q22.3-q23, 3p13-p12) (24-27). These data, taken together, suggest that the gene that causes McKusick-Kaufman syndrome will be an important and previously unrecognized gene in the genetic pathway of limb and genital development.
MATERIALS AND METHODS
Pedigree analysis
This study was reviewed and approved by the National Cancer Institute Institutional Review Board. Affection status was assigned by a staged definition. A woman with hydrometrocolpos is defined as an affected. Siblings of women with hydrometrocolpos are affected if they have hydrometrocolpos, polydactyly or congenital heart disease alone or in any combination. Individuals with polydactyly or congenital heart disease who do not have a sibling affected by hydrometrocolpos would be designated as unaffected. The previously published pedigree (5) was evaluated using the PedHunter software package and the Amish Genealogy Database (AGDB). PedHunter facilitates the creation and verification of pedigrees within large genealogies. AGDB has been created specifically for the Old Order Amish community of Lancaster County. For our analyses, all affected sibships were used to formulate an `all shortest paths' query for which PedHunter uses Dijkstra's algorithm (28) to determine the most recent common ancestor and the pedigree(s) that connects this ancestor to the affected sibships while minimizing the inheritance path lengths. Multiple paths of equal length may exist between the common ancestor and each sibship. The set of affected sibships and the `all shortest paths' pedigree were then used to formulate a second `minimum pedigree' query. In a `minimum pedigree' query, PedHunter uses a branch and bound algorithm to determine a pedigree with the smallest number of meioses from the given `all shortest paths' pedigree. This is accomplished by maximizing the number of ancestors shared among branches of the pedigree.
Genotyping and linkage analysis
Marshfield screening set v6 (Research Genetics) (29) was used to genotype a sibship with three affected children. Markers homozygous in the three affected children were then tested on two affected cousins. Additional markers were then typed in the regions that showed consistent homozygosity. The original screening of the first sibship was done using fluorescent labeling of PCR products and computer readings of data as previously described (30). Subsequent marker typings were performed in 96-well polycarbonate plates (Corning Costar) using a PTC-100 thermocycler (MJ Research). DNA was isolated from whole blood samples by lysis and affinity chromatography (Qiagen). Each PCR amplification of a 20 ng sample of DNA was performed in a 15 µl reaction containing 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.01% gelatin, 0.2 mM dATP, 0.2 mM dGTP, 0.2 mM dTTP, 0.0024 mM dCTP, 1 µCi of [32P]dCTP (3000 Ci/mmol; Amersham), 0.45 U of Taq polymerase (Perkin Elmer), and 0.67 µM of each primer. After initial denaturation at 95°C for 3 min, amplification was performed for 30 cycles at 95°C (1 min), 55°C (1 min) and 72°C (1 min). A final extension step of 10 min at 72°C was also performed. Products were separated by electrophoresis at 80 W through gels containing 6% polyacrylamide and 8 M urea (Sequagel-6, National Diagnostics). The gels were exposed at -80°C to XAR film (Kodak).
Linkage analysis
The linkage analysis computations were done with FASTLINK 3.0P (31,32), which is a faster version of LINKAGE (33). The longer runs were done using parallel FASTLINK (34). Some analyses used the p4 multiprocessor system software (35) and SGI Challenge multiprocessor computers. Linkage analyses used a penetrance function of 0.001 for 1,1 and 1,2 genotypes and 0.8 for 2,2 genotypes.
Physical mapping and sequencing
Extension of the published BAC contig of this region was accomplished by sequencing both ends of the most centromeric BAC, 334G22. One primer pair was designed from the sequence from each end. The orientation of the ends was determined by amplification of both ends from 334G22 and the more telomeric BAC, 204H22. The primer pair that gave a product from 334G22 and not 204H22 was used to PCR screen a human BAC library (Genome Systems). PCR was used to map known microsatellite markers and sequence-tagged sites (STSs) to the new BAC. Isolation of new polymorphic markers was performed as described (36). The jagged1 exons and exon-intron boundaries were sequenced by standard techniques in an affected individual from this pedigree and from a sporadic case using the previously described primers (8).
ACKNOWLEDGEMENTS
The authors thank Victor McKusick, William Pavan, Francis Collins, Margaret Abbott, Marlene Dressman, Eric Green and Robert Nussbaum for providing clinical data, technical assistance, advice and reviewing previous drafts of this manuscript. R.A. is employed under contract to R.O.W. Sciences, Inc. The authors are grateful for the hospitality and participation of the families.
REFERENCES
*To whom correspondence should be addressed. Tel: +1 301 402 2041; Fax: +1 301 402 2170; Email: leslieb@helix.nih.gov
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Feb 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Z. A. Eldadah, A. Hamosh, N. J. Biery, R. A. Montgomery, M. Duke, R. Elkins, and H. C. Dietz Familial Tetralogy of Fallot caused by mutation in the jagged1 gene Hum. Mol. Genet., January 1, 2001; 10(2): 163 - 169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. David, P. Bitoun, D. Lacombe, J.-C. Lambert, A. Nivelon, J. Vigneron, and A. Verloes Hydrometrocolpos and polydactyly: a common neonatal presentation of Bardet-Biedl and McKusick-Kaufman syndromes J. Med. Genet., August 1, 1999; 36(8): 599 - 603. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. Agarwala, L. G. Biesecker, K. A. Hopkins, C. A. Francomano, and A. A. Schaffer Software for Constructing and Verifying Pedigrees within Large Genealogies and an Application to the Old Order Amish of Lancaster County Genome Res., March 1, 1998; 8(3): 211 - 221. [Abstract] [Full Text]
|
|
|
|
 |
|
 L. Vindevoghel, K.-Y. Chung, A. Davis, D. Kouba, S. Kivirikko, H. Alder, J. Uitto, and A. Mauviel A GT-rich Sequence Binding the Transcription Factor Sp1 Is Crucial for High Expression of the Human Type VII Collagen Gene (COL7A1) in Fibroblasts and Keratinocytes J. Biol. Chem., April 11, 1997; 272(15): 10196 - 10204. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||