Editor—The van der Woude syndrome (VWS, OMIM 119300) is the most common form of syndromic cleft lip and palate, affecting 1/30 000 of the general population.1 The syndrome is inherited in an autosomal dominant fashion with almost complete penetrance.2 The clinical features of VWS include paramedial pits on the lower lip, cleft lip with or without cleft palate, cleft palate, and hypodontia. However, the presence and severity of the individual features is highly variable within and between families.3

A constitutional interstitial deletion of chromosome 1q32-q41 was first described by Bocian and Walker4 in a patient with lip pits. Linkage of VWS to this locus was then established by Murrayet al 5 and subsequently Schutteet al 6 refined the locus to a 1.6 cM region between D1S491 and D1S205. Recently, a microdeletion test using a novel polymorphic marker D1S3753 derived from the VWS critical region was also used.7 To date, 1q32-q41 is the only locus described for VWS and the VWS gene has yet to be identified. Recently, a large Brazilian VWS family with multiple cases affected with cleft palate was studied by Sertié et al.8 Linkage analysis was performed to search for a modifying locus for the cleft palate phenotype and suggestive linkage was shown in chromosome 17p11.2-p11.1 with a maximum lod score of 2.05 at D17S1824. The authors proposed the region as a modifying locus for cleft palate in VWS.

The homogeneous nature of the Finnish population is conducive for genetic studies especially for linkage disequilibrium analysis.9 The clinical features of Finnish VWS patients have previously been described,1 10 but no genetic studies have been reported. In this study, five Finnish VWS families were tested for linkage to the 1q32-q41 region and to the proposed modifying locus in chromosome 17p11.2-p11.1.

The VWS patients in this study were ascertained through hospital records at the National Cleft Centre in Helsinki, to which all patients with orofacial clefts in Finland are referred. The diagnosis of VWS in the participating patients was based on clinical examination, medical records, telephone interviews, or information from other family members. The pedigrees of the five families are shown in figs 1-3 and the clinical features of all affected cases are detailed in table 1. Within this group of patients the male to female ratio was 3:2 (18:12). Twenty six (87%) affected members had cleft palate or submucous cleft palate, while only three (10%) were affected with cleft lip and palate. Lip pits (including the microform of “wave-like” lower lip) were observed in nine subjects (30%). In family 60, no obvious VWS features were found in the five unaffected subjects who were shown to be disease haplotype carriers (IV.3, IV.5, III.2, III.7, and II.2, fig2). However, in two of these subjects (II.2 and II.4) speech problems were reported, which might imply the presence of submucous cleft palate.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Pedigree of family 60 which shows linkage to the 1q VWS region. The disease associated haplotype is represented by a hatched vertical bar. A meiotic recombination was detected between D1S2136 and D1S3753 in III.4 in family 60, placing the disease locus centromeric of D1S3753.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Pedigrees of families 59, 61, and 62.

The five families were analysed for the involvement of the VWS disease predisposition locus in 1q32-q41 as well as for the putative modifying loci in 17p11.2-p11.1 using linkage analysis. Microsatellite markers from the respective regions were amplified by PCR, pooled according to the emission spectra of the fluorescent dyes, and run on an ABI 377 laser fluorescent sequencing machine.

Nine polymorphic microsatellite markers covering the 1q32-q41 VWS region were genotyped and multipoint linkage was calculated as the logarithm of odds (lod) ratio using the GENEHUNTER program.12 The analysis was performed for individual families as well as the total group of families by considering genetic heterogeneity. VWS was modelled as an autosomal dominant trait with 95% penetrance.13 The disease allele frequency was taken as 1.5 × 10^-5 1 with a mutation rate of 1.8 × 10^-5.14 Representative multipoint lod scores for linkage between the VWS phenotype and microsatellite markers at 1q32-q41 are given for each family in table 2. Families 59, 60, and 61 displayed linkage to 1q32-q41 and their cumulative multipoint lod scores were higher than 3.8 across the VWS region. For family 57, haplotype analysis and multipoint lod scores below −10 in 1q32-q41 clearly excluded linkage (table 2, fig 1). Family 62 was uninformative for all nine VWS markers. Therefore, 25 additional microsatellite markers from chromosome 1 were typed. Markers adjacent to the VWS region included D1S1660 - 6 cM - D1S1678 - 8 cM - (nine VWS markers) - 2 cM - D1S2141 - 7 cM - D1S549 - 5 cM - D1S1656 - 2 cM - D1S3462. By observing the extended haplotypes (not shown) and multipoint lod scores of −1.9 along the VWS region (table 2), linkage could be excluded. For the whole group of Finnish VWS families, the maximum multipoint lod scores, under heterogeneity, was 2.53 from D1S245 to D1S205 with an alpha (proportion of linked families) of 53%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Exclusion of linkage between the VWS phenotype and markers within the VWS locus in 1q in family 57. Affected family members are indicated by filled symbols, unaffected members by open symbols, and obligate gene carriers with a dot. Subjects who were clinically re-examined in this study are marked with an asterisk. The genotypes for the markers in 1q32-q41 are shown under each subject, where “0” denotes the null allele in D1S3753. The best fitting disease associated haplotype is represented by a hatched vertical bar.

A meiotic recombination between D1S2136 and D1S3753 was observed in an unaffected subject in family 60 (III.4, fig 2), which tentatively placed the VWS locus centromeric to D1S3753 (proposed region 1, fig 4). Furthermore, the affected subjects from family 60 and 61 shared a common haplotype from D1S1663 to D1S3754 (proposed region 2, fig 4). The allele frequencies for each marker making up the haplotype 3 - 5 - 3 - 5 - 3 were 0.53, 0.22, 0.34, 0.34, and 0.59, respectively; therefore, the random population frequency of this haplotype would be 0.8%. Further genealogical studies of the two families may show the contribution of this haplotype to VWS.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Mapping of the VWS locus within 1q32-q41 to a 130 kb interval, based on the combined results from this study and those previously reported. The physical distance is provided by the Sanger Centre and the VWS region is reported by Schutte et al.6The region of haplotype sharing between families 60 and 61 is indicated together with the critical meiotic recombinants. The proposed region 1 is based on the previously reported recombinants as well as on the recombination in III.4 in family 60 and region 2 is based on a common origin of the shared haplotype.

During the study, the inheritance of marker D1S3753 was found not to follow a Mendelian inheritance pattern, suggesting the presence of an undetected (null) allele or deletion (figs 1 and 2). The null allele was found in both affected and unaffected subjects, meaning that it may not be related to VWS. To verify the relationship of this allele to orofacial clefts, five control families with no history of cleft lip and palate were typed with this marker. The null allele of D1S3753 was found to be present in these families as well, therefore excluding its role in orofacial clefts.

For the linkage analysis of a modifying locus on chromosome 17, the five most informative microsatellite markers reported by Sertiéet al 8 were genotyped. The cleft palate phenotype in VWS was modelled as autosomal dominant with a penetrance of 70% and a disease gene frequency of 0.0018and only subjects with cleft palate were considered as affected, regardless of the presence of lip pits. Multipoint lod scores at 17p11.2-p11.1 were less than −2 in families 59 and 60 combined (linked to 1q32-q41) as well as in families 57 and 62 combined (unlinked to 1q32-q41) (table 3). Thus, linkage between the cleft palate phenotype and the 17p11.2-p11.1 region could be excluded in these families. Results from family 61 were inconclusive. When all five families were considered together, there was no evidence of heterogeneity and a 20 cM region in 17p11.2-p11.1 could be excluded with lod scores less than –2.

The Finnish population is useful for genetic studies owing to its homogeneous background. As a result of isolation, the Finnish population displays unique clinical and genetic profiles in a number of diseases.9 In this study, 90% of the Finnish VWS patients in this study had cleft palate in contrast with 30% of VWS patients in other countries.14 This predominance of cleft palate in our Finnish VWS families implies specific types of VWS gene mutation or the influence of genetic background (modifying genes). The clinical variability of cleft palate including submucous cleft also contributed to the low expressivity of VWS in the Finnish families, especially in family 60, in which five gene carriers did not have overt cleft palate.

To date, chromosome 1q32-q41 is the only reported VWS locus in different populations.6 8 15 16 In the present study, two Finnish VWS families (57 and 62) were unlinked to this locus, thus representing the first report of genetic heterogeneity in VWS. Both families had classical features of VWS including cleft lip and palate and lower lip pits. For family 57, detailed clinical re-examination was performed on most members to confirm the diagnosis of VWS. In addition, the genotyping was repeated and scored by independent investigators. In family 62, although II.1 was homozygous for all the markers from the VWS region, analysis with additional markers from chromosome 1 clearly excluded linkage.

One important goal of linkage analysis is to refine the genetic locus, typically by observing meiotic recombinational events in affected family members. In the three 1q linked Finnish VWS families, one critical recombination was found between D1S2136 and D1S3753 in an unaffected subject (III.4, family 60, fig 2). When combined with previously published linkage data this recombination would place the VWS disease locus between markers D1S491 and D1S3753, in a region of less than 200 kb. However, recombinations in unaffected subjects have to be interpreted with caution, especially when a low expressivity is observed, such as in this family.

Linkage disequilibrium is an efficient approach for mapping disease genes in isolated populations such as the Finnish.17 In our study, haplotyping of the VWS families showed a common shared region from D1S1663 to D1S3754 between affected members in families 60 and 61. This finding would tentatively refine the VWS locus to the D1S491-D1S2136 interval (proposed region 2, fig 4), the size of which has been estimated to be approximately 130 kb, as determined from genomic sequencing.18 Genealogical studies are now under way to identify the ancestral relationship between these families.

Typing of locus D1S3753 has been suggested as a standard test for microdeletion in VWS and other patients with orofacial clefts.7 Our results from the five VWS families and from five additional control families showed that this marker contains a commonly occurring null allele, implying that this is a polymorphism without any aetiological role in VWS. Recently, Watanabeet al 19 reported a commonly deleted 8 kb segment in the distal part of the VWS critical region in the general population. This, together with previous reports of frequent deletion in 1q32-q41, may represent an unstable region including D1S3753 within the VWS locus.

The Finnish VWS families in our study displayed predominantly cleft palate, which made them suitable for testing the modifying gene on chromosome 17p11.2-p11.1.8 Linkage was excluded in families either linked or unlinked to 1q32-q41, indicating that 17p11.2-p11.1 does not contain a major modifying gene for cleft palate in VWS.

In conclusion, we have provided the first evidence of genetic heterogeneity in VWS, tentatively refined the VWS locus to a 130 kb interval, and excluded 17p11.2-p11.1 as a major modifying locus for the cleft palate phenotype. The entire VWS critical region has now been sequenced and mutation analysis on both known and new genes within the region is being performed, but no pathogenetic VWS mutations have been reported so far.18 In the future, the identification of the VWS genes in 1q and elsewhere will shed light on the pathogenesis of VWS.