SUBJECTS

At the time of the original study in 1968,¹ the F family consisted of four generations of affected and unaffected persons. A fifth generation born since then includes an affected female and two affected males. A total of nine affected and 18 unaffected family members aged 14–79 provided blood samples for DNA analysis, representing four living generations of the five affected. Of these, 18 individuals were used for linkage; the pedigree is shown in fig 2. The phenotype of each individual was assessed and characterised on the basis of the previous report,¹ a clinical examination at the time of blood drawing, and by medical record review of all charts available on any person in the family. The results are summarised in table 1. Characteristic examples of the F syndrome limb phenotype are shown in fig 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Phenotype of F syndrome. (A) and (B) show the hand phenotype of a relatively mildly affected individual. Note the severe shortening of the thumb and partial syndactyly with the second digit. (C) shows the foot phenotype with fusion of the first and second toes. (D) is an X ray of (C), demonstrating a single first metatarsal and duplication of the first toe phalanges. The duplicated toe has, however, three phalanges (arrow, patient’s right foot) indicating that this is likely to be the second toe fused to the first. There is also duplication of distal phalanges in the other toes (arrow, left foot). In addition, there is fusion of metatarsals 3 and 4 at their proximal ends (*). (E) shows an affected hand with complete syndactyly between the first and second digits with the corresponding radiograph given in (F). (G) and (H) show a sternal deformity (arrow).

METHODS

Genomic DNA was extracted from peripheral blood lymphocytes by standard techniques. A panel of 395 microsatellite markers from the Généthon final linkage map⁵ with an average distance of 11 cM was used to perform a genome wide linkage analysis. PCR reactions were performed using the manufacturers’ protocols. Semi-automated genotyping was performed by a MegaBACE-1000 analysis system. Data were analysed by Genetic Profiler Software 1.5.

For fine mapping, a novel microsatellite marker M01HT1A located 704 kb q telomere from marker D2S434 and 477 kb centromere from marker D2S2250 using forward primer TTACAAggACAAgAAAAAggAAgg and reverse primer gggTgACAgggTgCgACTC was established. A two point lod score calculation was performed with LINKAGE v5.2 program package⁶ using an autosomal dominant model with 100% penetrance and a gene frequency for F syndrome at 0.00001. Multipoint lod score calculation was performed using Vitesse⁷ with the aid of Alan Young’s GAS2.0 interface. The most probable haplotypes were constructed with Simwalk2 v2.82.⁸

Candidate genes (IHH, WNT6a, WNT10a, PAX3, and STK36) were tested by amplifying all known exons of each gene (Ensemble database) using primers that were placed within the introns. PCR products were sequenced from both ends and compared to the wild-type sequence.

RESULTS

We performed a genome wide linkage search in 16 individuals, nine affected and seven unaffected relatives. The first evidence for linkage to chromosome 2q36 was indicated by the marker D2S377 with a lod score of Z_max = 1.505 at θ = 0 (table 2). Further fine mapping with an additional 10 markers confirmed the locus and documented critical recombination events between markers D2S434 and D2S2250 and markers D2S377 and PAX3.PCR1. Marker D2S2250 showed the highest two point lod score value of Z_max = 4.214 at θ = 0 (table 2). Genome wide multipoint analysis reached the same value indicating that full information from all the meioses had already been obtained using marker D2S2250 alone (not shown).

Haplotype analysis showed clear evidence that the mutant allele cosegregated in all the affected individuals and was absent from unaffected individuals (fig 2A). Individuals II.2 and IV.3 were identified as obligate recombinants determining the critical interval. To further narrow the 8.5 cM interval between markers D2S434 and D2S2228, novel microsatellite markers were developed based on the NCBI genomic sequence assembly data (build32; April 2003). Marker PAX3.PCR1 is located upstream of the PAX3 gene, some 530 kb towards the centromere from D2S360 (GDB:187424¹⁰; fig 3). Since this marker was still subject to recombination in individual II.2, the critical interval could be further reduced distally by about 1.53 cM. Unfortunately, the new marker M02HT1A identified in the interval between markers D2S434 and D2S2250 was uninformative for the relevant meiosis in individual III.7 (fig 2). Thus, the current critical interval is confined by the markers D2S434 and PAX3.PCR1 and spans a physical distance of approximately 4.6 Mb. This corresponds to a genetic distance of approximately 6.5 cM. For the Italian family, originally described by Camera and colleagues,⁴ haplotype analysis at chromosome 2q36 was consistent with linkage to this region (fig 2B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

(A) Pedigree of the F syndrome family. Haplotypes at 2q36.1 are shown for the individuals where DNA was available (digit numbers except II.1) or haplotypes could be inferred. Thirteen microsatellite markers are shown in order from centromere to q terminal. Segments of haplotypes which could not unambiguously be assigned to the paternal versus maternal haplotype are represented by a thin line. Inferred haplotypes are indicated in parentheses. Note that marker D2S434 is flanking the F syndrome locus on its centromeric borders, as defined through a recombination in individual II.1 (arrow), and that marker PAX3.PCR1 is flanking the F syndrome locus at its q terminal border as defined by a recombination in individual IV.3 (arrow). Flanking markers are underlined. (B) Pedigree of the Italian family showing the most probable haplotypes at 2q36.⁴ Four microsatellite markers are shown in order from centromere to q terminal. (A), (B) Generations are indicated by roman numerals on the left. Circles denote females, squares males. Filled symbols indicate diagnosis of F syndrome. A slash through a symbol marks a deceased individual. Haplotypes are presented as grey and black bars. The black haplotype cosegregates with the affected status.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Physical map of the linkage region on chromosome 2. Marker and gene positions are plotted in megabases according to the NCBI genomic sequence (build33; April-2003). Mapping intervals of the F family,¹ German syndactyly type 1 family¹¹ and Iranian syndactyly type 1 family¹² are shown as horizontal double arrows. Mapping interval of the mouse mutant doublefoot (Dbf¹³) is indicated as inverted double arrows. The genes Arpc2 and Slc4A3 represent approximate flanking locations from the region of conserved synteny on mouse chromosome 1. cen, centromeric orientation; q-ter, q-terminal orientation.

Searching for candidate genes using the NCBI MapViewer showed several interesting candidate genes, such as PAX3, WNT6, WNT10A, and IHH. We identified a 469C>G transition (based on accession number AB059570) resulting in substitution of proline to arginine (P155A) in exon 3 of the WNT6 gene. However, this change was also present in person III.1 who is unaffected (fig 2). This change does not correspond to a known SNP. The residue was not found to be conserved between mouse and human. Thus, sequencing of the exons and flanking splice sites of all these genes failed to detect mutations in the two F syndrome families, the original F syndrome family and the Italian family.⁴

DISCUSSION

F syndrome comprises carpal and tarsal synostoses, dysgenesis of the first and second finger with frequent syndactyly between them, fusion of first and second toes, syndactyly of toes and fusion of metatarsals, malformation of the sternum, and spina bifida occulta (fig 1, table 1). The presence of syndactyly of fingers and toes suggests that the causative gene is involved in early limb patterning at a stage when the number of digits is determined or when digits are separated from each other. Several pathways such as the hedgehog, fibroblast growth factor and bone morphogenetic protein pathways, are involved in these complex processes and an increasing number of phenotypes is being identified with defects in these regulatory pathways.¹⁴ It may be assumed that the identification of the F syndrome gene or mutation may give new insights into the pathogenesis of limb development and malformation syndromes with syndactyly or polydactyly.

We performed a genome wide linkage analysis of the F syndrome and were able to localise the underlying gene at chromosome 2q36 within a physical fragment of approximately 4.6 Mb correlating to a genetic distance of approximately 6.5 cM (fig 2A). PAX3 is located at the very telomeric end of the critical segment. Different mutations in PAX3 cause Waardenburg syndrome type 3 and craniofacial-deafness-hand syndrome.^15,16 Because affected individuals in Waardenburg syndrome type 3 have skeletal upper-limb hypoplasia,^16,17 we considered PAX3 a candidate gene for F syndrome. However, sequencing of PAX3 showed no mutation in our patients. WNT6 and WNT10A were considered interesting candidate genes. They are clustered in a head-to-head manner within an interval of less than 7 kb inside the critical segment.¹⁸Wnt6 is expressed in the developing murine limb.¹⁹ Similar results were obtained by Hayes et al¹³ who demonstrated expression of both Wnt6a and Wnt10a in the progress zone of the developing limb buds. However, no mutations in WNT6 and WNT10A could be demonstrated.

IHH appeared to be a further candidate gene located within the F syndrome critical region. Ihh is known to mediate condensation, growth, and differentiation of cartilage.²⁰ Mutations in IHH cause brachydactyly type A1²¹ and acrocapitofemoral dysplasia,²² two conditions involving the growth and differentiation of limb skeletal elements. The fused gene (FU, or STK36) encodes a serine/threonine kinase positively acting in the hedgehog pathway. Given the previous finding that many polydactyly conditions are caused by mutations in genes involved in the hedgehog pathway, we considered fused as a further candidate for F syndrome. However, no mutations were identified in IHH or STK36.²³

Other conditions with limb malformations and syndactyly/polydactyly are linked to this region of 2q36. Syndactyly type 1 (SD1) also referred as “type Lueken syndactyly” was mapped to a region overlapping with the F syndrome defining a 9.4 cM region between markers D2S2319 and D2S344,¹¹ (fig 3). The locus for SD1 could be confirmed in a second six generation Iranian family.¹² Syndactyly type 1 is characterised by variable syndactyly between the second and third toes, extending to complete syndactyly between the second and the fifth finger and the first to fifth toes. The involvement of the second finger appears to be a rather rare event in syndactyly type 1, raising the possibility that this family may be unique, and has therefore been classified as syndactyly type Lueken.²⁴ In contrast to the F syndrome phenotype, polydactyly and involvement of the thumbs is not observed. In particular, the characteristic syndactyly of the first and second fingers and the fusions of carpal and tarsal bones are not part of syndactyly type 1. However, both conditions affect the patterning of early digit formation and in particular the separation of digits, a process known to involve extensive apoptosis partly controlled through the bone morphogenetic protein pathway. Hence, it is conceivable that the F syndrome and syndactyly may be caused by different mutations in the same gene.

The mouse mutant doublefoot (Dbf) was mapped to a 0.4 cM region on mouse chromosome 1.^13,25 This region is syntenic to human 2q36 and contains the F syndrome critical interval including the genes Ihh, Wnt6, Wnt10a, and fused (fig 3). Dbf is characterised by preaxial (anterior) duplication of digits resulting in one to four additional digits and syndactyly. Ectopic expression of Ihh throughout the distal limb bud mesenchyme appears to be a main causative event²⁶ resulting in two hedgehog signals, one posterior by regular Shh expression and one anterior by ectopic Ihh expression. No mutation has been detected for Dbf to date but regulatory mutations affecting Ihh have been proposed. The phenotypic overlap between F syndrome, SD1 and Dbf raises the question of whether IHH misexpression could be involved in the pathogenesis of the human condition as well. IHH is causally involved in another condition already mentioned above, brachydactyly type A1. This condition is characterised by hypoplasia or aplasia of the middle phalanges, a phenotype not present in F syndrome or syndactyly type 1.

Dundar described a novel syndrome in a large Turkish family that maps to 7q36.² The authors describe overlap with F syndrome and concluded that the two conditions are related but distinct. Both conditions show syndactyly and involvement of the sternum but neither the characteristic syndactyly of fingers 1 and 2 nor the tarsal or metatarsal synostoses were observed in the Turkish family. Our present results clearly show that F syndrome and the previously published acropectoral syndrome² can be considered to be clinically and genetically separate conditions. Interestingly, the conditions described by Dundar et al map in close vicinity to the sonic hedgehog (SHH) gene which shows a high degree of homology to IHH. Like the locus at 2q36, the locus at 7q36 harbours several limb malformation syndromes such as preaxial polydactyly (PPD) types 2 and 3, triphalangeal thumb-polydactyly syndrome (TPT), and Haas type syndactyly (syndactyly type IV).²⁷

Recently this form of TPT/PPD has been shown to be caused by mutations in a Shh long range cis-regulatory element, located within intron 5 of the adjacent C7orf2 gene.²³ In addition, the syntenic region in the mouse is disrupted in the mouse mutant Sasquatch (Ssq) exhibiting preaxial polydactyly.²³ Acheiropodia is a recessive condition, characterised by a phenotype resembling the murine Shh null phenotype with hemimelia of both upper limbs and a distal third of both lower limbs. Acheiropodia is caused by homozygous deletion of exon 3 of C7orf2 and parts of the flanking introns.²⁸ The mutation has also been interpreted as loss of a cisregulatory putative Shh regulatory element, located in an C7orf2 intron, approximately 1 Mb away from Shh.²³ If the clustering of limb phenotypes around the IHH locus has pathogenetic similarities to those conditions clustered around the SHH a regulatory mutation leading to disruption of the fine tuned expression of IHH may also be considered as a possible cause for F syndrome.