Article Text

Clinical and genetic studies on 12 preaxial polydactyly families and refinement of the localisation of the gene responsible to a 1.9 cM region on chromosome 7q36


Polydactyly is the most frequently observed congenital hand malformation with a prevalence between 5 and 19 per 10 000 live births. It can occur as an isolated disorder, in association with other hand/foot malformations, or as a part of a syndrome, and is usually inherited as an autosomal dominant trait. According to its anatomical location, polydactyly can be generally subdivided into pre- and postaxial forms. Recently, a gene responsible for preaxial polydactyly types II and III, as well as complex polysyndactyly, has been localised to chromosome 7q36.

In order to facilitate the search for the underlying genetic defect, we ascertained 12 additional families of different ethnic origin affected with preaxial polydactyly. Eleven of the kindreds investigated could be linked to chromosome 7q36, enabling us to refine the critical region for the preaxial polydactyly gene to a region of 1.9 cM. Our findings also indicate that radial and tibial dysplasia/aplasia can be associated with preaxial polydactyly on chromosome 7q36.

Combining our results with other studies suggests that all non-syndromic preaxial polydactylies associated with triphalangism of the thumb are caused by a single genetic locus, but that there is genetic heterogeneity for preaxial polydactyly associated with duplications of biphalangeal thumbs. Comparison of the phenotypic and genetic findings of different forms of preaxial polydactyly is an important step in analysing and understanding the aetiology and pathogenesis of these limb malformations.

  • preaxial polydactyly
  • chromosome 7q36
  • localisation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The human limb bud starts to develop late in the fourth week of intrauterine life. Approximately four weeks later, an interplay of genes and molecular factors results in the development of a complete set of limbs with a well defined appearance, function, and a specific number of digits. Proper positional signalling within the three dimensional structure of the developing limb is of crucial importance for the future cell fate during embryogenesis. Disturbances in these signalling pathways can result in a large number of congenital limb deformities.

Limb malformations occur as isolated malformations of the hands or feet or as a part of a syndrome. In The Netherlands, approximately 16 per 10 000 children are born with congenital hand malformations.1-3 Of all congenital hand malformations, polydactyly is the most frequently observed and it is even one of the most frequent congenital disorders in general.4 The estimated prevalence of polydactyly, with or without an associated malformation, varies between 5 and 19 per 10 000 live births.1 2 4 5 Sporadic occurrence has been described, but the majority of cases show an autosomal dominant mode of inheritance. The large variety of recorded prevalences in different studies can partly be explained by the fact that clusters of affected families often inhabit specific geographical areas.

Our present knowledge on limb development and congenital limb disorders has mainly been gathered in two ways. Firstly, because of its accessibility and size, the vertebrate limb has become a model system for studying developmental mechanisms and pattern formation during embryogenesis of vertebrate embryos, and this has resulted in models that are now classical for vertebrate pattern formation.6Secondly, in recent years, congenital limb malformations have attracted enormous research attention. Finding the genes responsible and correlations of phenotypes with genotypes have made a large contribution to the understanding of human development.7However, the majority of genes involved in the aetiology of human limb malformations remain to be identified and in view of this most of the present classifications of congenital hand malformations are descriptive. The most frequently used are the ones by Swanson8 and Wassel9 based on anatomical findings, the classification by Temtamy and McKusick10widely used among geneticists, and the classification proposed by Winter and Tickle11 based on embryological findings. However, none of these classification systems has succeeded in clarifying the considerable overlap in phenotype between the defined subgroups.

Polydactyly can be defined as the duplication of a finger or a part of it. Isolated (non-syndromic) polydactyly can be generally subdivided into pre- and postaxial polydactyly. Preaxial polydactyly refers to an excess of parts on the radial side of the limb and it describes the so called duplicated thumbs, as well as the various forms of triphalangeal thumbs and index finger duplications. Temtamy and McKusick10 define the following four types of preaxial polydactyly: type I (PPD-I) or thumb polydactyly is the duplication of one or more of the skeletal components of a biphalangeal thumb; type II (PPD-II) or polydactyly of a triphalangeal thumb; type III (PPD-III) or polydactyly of an index finger; type IV (PPD-IV) or polysyndactyly. Both preaxial polydactyly and syndactyly are cardinal features of this phenotype, but syndactyly never occurs without polydactyly.

PPD-I, or thumb polydactyly, has further been subdivided into six subtypes by Wassel9 according to the level of duplication in bone anatomy. This type of preaxial polydactyly is usually sporadic, often unilateral, and less frequently associated with thenar hypoplasia than the other three types mentioned above.

In 1994, two independent studies reported linkage of two different phenotypes of preaxial polydactyly to chromosome 7q36. Two kindreds were affected with PPD-II/III12 13 and the other kindred was affected with complex polysyndactyly.14 Subsequently in two other studies, linkage of two families with PPD-II/III to the same chromosomal locus was reported.15 16

Our group is presently working on the identification of the gene responsible for PPD-II/III on chromosome 7q36. As part of these efforts, we collected a large number of families affected with different forms of poly- and syndactyly. In this paper we will address the question of which forms of polydactyly are linked to chromosome 7q36. We report the phenotypic and genetic findings of 12 families with different ethnic backgrounds that are affected with various forms of preaxial polydactyly.

All families with PPD-II/III could be linked to chromosome 7q36. Within the linked families a large variation in the phenotypes was observed, including PPD-I, postaxial polydactyly, and associated radial and tibial dysplasia/aplasia. Linkage to chromosome 7q36 of a family with PPD-I but not PPD-II/III could be excluded, indicating that there is genetic heterogeneity for different forms of preaxial polydactyly. A comparison of all phenotypes that have been linked to the genetic locus on chromosome 7q36 is an important step in analysing their aetiology and understanding their overlapping phenotypes.

Subjects and methods


After our initial studies, the two Dutch families originally described in 199412 13 were expanded by another seven families originating from the same small geographical region in The Netherlands. Clinical and genealogical investigation was performed and peripheral blood samples were obtained for DNA analysis. A genealogical search for the common ancestor of all affected families from this region was performed by using the population and census records and civil registration in the municipal archives. All participating family members were personally examined by one of the authors with consent.

In addition, through a review of medical records of the Department of Plastic and Reconstructive Surgery of the Sophia Children’s Hospital in Rotterdam, we ascertained a proband with PPD-I of the hands and feet. The proband had a positive family history for the same disorder. This family was not related to the above mentioned kindreds. Clinical investigation and peripheral blood sampling was performed by one of the authors with consent.


Through a review of medical records of the Department of Plastic, Hand and Reconstructive Surgery of St James’s University Hospital in Leeds, 10 probands with preaxial polydactyly were ascertained which were referred to the department for primary or secondary surgical correction of their hands. Each of these probands had a positive family history for the same limb malformation. Most of the probands and their relatives originated from the same small village in the Lake District in the United Kingdom. Clinical and genealogical investigation and peripheral blood sampling was performed by one of the authors with consent.


Through a review of medical records of the Department of Orthopaedic and Hand Surgery of the Social Security Hospital in Ankara, Turkey, four probands with PPD-II were ascertained, all of whom had been referred for hand surgery. All four patients were members of the same family in which PPD-II was inherited as an autosomal dominant trait with complete penetrance. Clinical investigation and peripheral blood sampling was performed by two of the authors with consent.


A case of a girl with bilateral tibial aplasia and preaxial polydactyly born into a family in which PPD-II was inherited as an autosomal dominant trait has previously been reported by one of us.17 This family will be referred to as Cuban family A. In addition to the already performed clinical and genealogical examination, peripheral blood samples were obtained for DNA analysis from all informative family members with each subject’s consent.

Through a review of medical records of the Department of Clinical Genetics of the National Center for Medical Genetics in Havana, Cuba, another family affected with PPD-II/III was ascertained. This family will be referred to as Cuban family B. Clinical and genealogical examination and sampling of the peripheral blood was performed by one of the authors with consent.


Genomic DNA was isolated from peripheral blood as described by Miller et al. 18 Microsatellite markers were tested in multiplex reactions essentially as described by Weber and May19 using a Perkin-Elmer-Cetus 9600 Thermocycler. Initial denaturation was 10 minutes at 94°C followed by 25 cycles of 30 seconds denaturation at 94°C, 30 seconds annealing at 55°C, and 90 seconds extension at 72°C. After 25 cycles a final extension time of five minutes at 72°C was used. Gel electrophoresis on polyacrylamide gels was performed as described by Weber and May.19 Information for microsatellite markers D7S550, D7S559, D7S2423, and D7S2465 was obtained from the Genome DataBase (GDB).20 Marker order was also obtained from GDB. A microsatellite marker within the Sonic hedgehog gene (Shh) was described by Marigo et al.21


Pairwise lod scores were calculated for each family using the MLINK program of the LINKAGE package (version 5.1)22assuming polydactyly to be an autosomal dominant disease with a gene frequency of 0.001 and a conservative penetrance estimate of 95% (table 1). Mutation rate was set at zero and equal recombination rates between males and females were assumed. Marker allele frequencies were kept equal because population frequencies were not available and the families were too small to calculate reliable allele frequencies from people marrying into the polydactyly kindreds.

Table 1

Pairwise lod scores between chromosome 7q36 markers and PPD-II/III families at various recombination fractions



Seven investigated families were affected with a similar phenotype of PPD-II/III according to Temtamy and McKusick.10 A detailed description of the phenotype has been provided elsewhere.13 We had the opportunity to examine 298 subjects of whom 138 were affected, 94 were unaffected sibs, and 66 were healthy partners of affected subjects. By means of genealogical investigation, all affected families could be linked to a single common ancestral couple who lived approximately 200 years ago (fig 1).

Figure 1

Pedigree of the Dutch kindreds affected with PPD-II/III according to Temtamy and McKusick,9 and their common ancestral couple. Filled symbols represent affected subjects.

Given its size and structure, lod scores were not calculated for the complete family since strong evidence for linkage in this family has already been reported.12 Instead we performed haplotype analysis on the complete pedigree in order to detect recombination between the disease phenotype and genetic markers that would allow us to reduce the critical region for the genetic defect. In our original study we reported recombinations with marker D7S550 on the centromeric side of the critical region and with marker D7S794 on the telomeric side. In addition to what was reported in our original study, we tested four chromosome 7q36 markers (D7S550, D7S2465, D7S559, and D7S2423) on all available family members. All affected subjects share a common haplotype for the markers D7S2465 and D7S559 but several recombinants were observed with D7S550 on the centromeric side and with D7S2423 on the telomeric side, reducing the candidate region to a 1.9 cM interval between D7S550 and D7S2423 (table 2). The Sonic hedgehog (Shh) gene is localised within this 1.9 cM region close to D7S550. Shh is known to play an important role in limb development and specifically in the determination of the anterior-posterior pattern of the limb. Therefore, the gene can be considered as a good candidate for PPD. We tested a polymorphic marker within the Shh gene in the branch of the family that shows a recombination with D7S550. We detected a recombination between one affected subject and the Shh polymorphism, excluding Shh as a candidate gene for PPD (data not shown).

Table 2

Haplotype sharing between affected subjects in several branches of the Dutch PPD-II/III family

The “disease” haplotype was not found in unaffected subjects except for two brothers who shared the “disease” haplotype between D7S550 and D7S2423. These two brothers showed no (preaxial) polydactyly at first sight. On more detailed examination, one had a rudimentary postaxial polydactyly on his left hand in the form of a wart with a diameter of 3 mm on the lateral border of his middle phalanx.13 No other abnormalities were found. This subject had two children and one of them inherited the “disease” haplotype and showed full blown preaxial polydactyly. The second brother had no abnormal findings even on detailed examination. He had one unaffected child who did not inherit the “disease” haplotype. In total, 140 members of this family shared a common haplotype for D7S2465 and D7S559. Of these 140 subjects, 139 showed a polydactyly phenotype. The calculated penetrance is therefore almost complete.


Ten probands were originally ascertained and appeared to be members of the same family. According to family history, 28 out of 81 subjects were affected. Twenty-five family members were personally examined. Fifteen of them were affected, five were unaffected relatives, and five were healthy partners of affected subjects who had affected children. The pattern of inheritance was autosomal dominant with complete penetrance and variable expression. All affected family members had hexadactyly of all four extremities, with opposable or non-opposable thumbs.

In the “opposable thumb phenotype” (fig 2), a fully developed thumb is placed in a normal “thumb” plane with an angle of 90° to the digital rays. The thenar muscles are well developed and opposition function is not significantly impaired. On xray, the thumb shows either a biphalangeal or triphalangeal appearance, but without noticeable increase in length. The supernumerary digits are always localised between the thumb and index finger, partly rotated towards the thumb plane.

Figure 2

The “opposable” phenotype in the British kindred.

In the “non-opposable phenotype” (fig 3), the thumbs are small and hypoplastic. The “second” digital rays have the appearance of a fully developed index finger. All digits are placed in the same plane. The thenar musculature is hypoplastic and the thumbs are only capable of “pseudo-opposition”.

Figure 3

The “non-opposable” phenotype in the British kindred. The photograph was taken postoperatively. Only amputation of the preaxial extra ray was performed. No further reconstructive surgery was performed.

All affected subjects showed hexadactyly of both feet. The supernumerary ray is localised between the big and the second toe. Onx ray, the supernumerary toe is biphalangeal and usually has the appearance of a somewhat hypoplastic (duplicated) hallux. The phenotype of this family can be classified as PPD-II/III.

All affected persons in this family shared a common haplotype for all four markers. However, this haplotype was different from the one in the other families in this study (fig 4A).

Figure 4

Pedigrees of PPD-II/III families with haplotype for markers D7S550, D7S2465, D7S559, and D7S2423 (from top to bottom). (A) UK family. (B) Turkish family. (C) Cuban family A. (D) Cuban family B. Filled symbols represent affected subjectss. The 0 next to IV.4 in (C) represents the additional radius and tibia dysplasia/aplasia. The + next to VI.1 in (D) represents an additional phenotype of scaphocephaly. The partly filled symbols in (D) with the top left quarter filled represent duplication of the thumbs, the top right quarter represents brachydactyly, and the bottom right quarter represents duplication of the big toes.


According to the family history, 21 out of 51 subjects were affected. Fifteen affected subjects, 12 unaffected relatives, and five healthy partners of affected subjects who had affected children were examined.

All affected family members had a strikingly similar phenotype of hexadactyly of both hands and feet (fig 5). The first digital ray was too long for a thumb and was rotated in the “thumb plane”. Onx ray, the thumb was triphalangeal, with a fully developed rectangular extra phalanx and a long, slender metacarpal bone. There were no sesamoid bones, which corresponded with the clinical finding of a moderately developed thenar eminence and moderate opposition impairment. The second digital ray had the appearance of a fully developed index finger. It was difficult to distinguish whether the duplicated digit had the “identity” of an index or a middle finger.

Figure 5

The “Turkish phenotype”.

On x ray of the feet, the supernumerary digit was located between the big and the second toe. Again, the identity of the duplicated ray caused difficulties in view of the fact that in some subjects it contained two and in other affected subjects three phalanges.

All affected subjects were symmetrically affected, with little intra- and interpersonal variation in phenotype. The phenotype of this family can be classified as PPD-II/III. Patients shared a common haplotype for all markers, but not with any of the other families in this study (fig4B).


Family A

Six out of 11 members of this family were affected with preaxial polydactyly of the hands and feet. All affected persons showed a very similar phenotype of bilateral, non-opposable, triphalangeal thumbs and preaxial polydactyly of both feet. No other malformations were recorded. The common phenotype of this family can be classified as PPD-II/III.

The previously described proband showed a much more severe phenotype.17 Both upper extremities had a mild degree of radial dysplasia, associated with hexadactyly of the left hand, and a non-opposable triphalangeal thumb on the pentadactylous right hand (fig6, IV.4 in fig 4C). All digital rays on the upper extremities had three phalanges and were placed in the same plane. Thenar hypoplasia was severe on both hands. In addition, there was cutaneous syndactyly between the first, second, and third digital ray of the left hand. Onx ray, the fourth metacarpal of the right hand showed a partial duplication in the form of a bifurcation.

Figure 6

Proband from Cuban family A. Note hexadactyly of the left hand and a non-opposable triphalangeal thumb on the pentadactylous right hand; note also the severe bowing of the lower extremities with bilateral absence of the tibia and preaxial polydactyly of both feet.

Both lower extremities were short, with bilateral absence of the tibia, severe bowing of the lower legs, and preaxial polydactyly of both feet. The right foot had six digital rays, which corresponded with six metatarsal bones on x ray. The left foot had seven digital rays on x ray accompanied by only five metatarsal bones. The two “floating” supernumerary digits were located between the big and the second toes.

This family is too small to obtain a significant lod score of more than 3 by itself. However, all affected family members share a common haplotype, and none of the unaffected subjects share this haplotype resulting in positive lod scores for all four markers, indicating that this family is very likely to be linked to 7q36. IV.2 shows a recombination with marker D7S2423. The severely affected proband showed the same disease haplotype as all other affected family members (fig4C). The “disease” haplotype was different from that of the other families investigated.

Family B

We had the opportunity to examine 52 subjects of whom 32 were affected, 14 were unaffected sibs, and six were healthy partners of the affected subjects. The phenotype consisted of the opposable digit-like triphalangeal thumb in association with a preaxial extra ray which usually resembled a hypoplastic thumb. The common phenotype of this family can be classified as PPD-II/III.

Subject VI.1 showed the above described PPD phenotype on his left hand and foot. The right hand showed six well developed metacarpals with hypoplasia of the middle and distal phalanges. On the right foot, only the big and little toes were well developed, while the middle toes consisted of only the metatarsal bones and hypoplastic proximal phalanges. In addition, this subject also had mild scaphocephaly. No other relatives had craniosynostoses. Subject V.12 had brachydactyly type C of both feet and no polydactyly. Subject VI.5 had complete duplication of both thumbs or PPD-I (fig 7), but there was no triphalangism. The ulnar thumb showed a long, fully developed metacarpal with a well developed proximal and distal phalanx. The radial thumb showed a rather thumb-like, shorter metacarpal, with slightly hypoplastic proximal and distal phalanges. Subject IV.5 had PPD type I of his feet or duplication of the big toes. All subjects affected with polydactyly also had hexadactyly of their feet. Onx ray, the extra ray was localised between the big and the second toe and contained two phalanges.

Figure 7

X ray of V.15 affected with PPD type I from Cuban family B.

All affected subjects shared a common haplotype, different from the other families in this study, except V.12 affected with brachydactyly (fig 4D). The reason for the occurrence of additional congenital malformations in this family apart from polydactyly remained unclear.


Because of the finding of PPD-I in the Cuban family B, we included another family in which PPD-I occurred in our study. The Dutch proband with preaxial polydactyly type I came from a family affected with pre- and postaxial polydactyly. Twenty-three family members were personally examined. Sixteen subjects were unaffected relatives, two were healthy partners of affected subjects who had affected children, and five showed a polydactyly phenotype. The proband showed broad big toes, which on x ray appeared to be caused by complete duplication of the osseous elements. On both hands, this subject showed broad thumbs with ulnar clinodactyly in the interphalangeal joints. On x ray no duplication was seen, but there were broad, trapezoidal proximal and broad distal phalanges. The proband’s father and brother showed only duplications of their big toes. Two subjects from the pedigree showed only unilateral postaxial polydactyly type B of their right hands. No other malformations were found. This family was too small to obtain significant evidence for linkage, but we tested the chromosome 7q36 markers on the family branch affected with the PPD type I, excluding the subjects with the postaxial polydactyly, and found no segregation of the marker alleles with the disease (data not shown), indicating that PPD-I can be caused by more than one genetic locus.


In this study, we report clinical as well as genetic findings in 12 families with preaxial polydactyly. In addition to the two Dutch families described in our original linkage study,12 we examined another seven families originating from the same geographical region. The phenotype of the affected subjects from these kindreds was very similar to that of those in our original study and could be classified as PPD-II/III according to Temtamy and McKusick.10 Extensive genealogical studies showed a single common ancestor couple for all nine families. This genetic relationship was confirmed by the finding that all affected subjects in the nine families shared a common haplotype for markers D7S559 and D7S2465. This haplotype was not observed in unaffected subjects apart from one, so the trait shows almost complete penetrance.

Subsequent haplotype analysis on all available family members showed recombination events with markers D7S550 and D7S2423 reducing the critical region for the PPD-II/III gene to a region of 1.9 cM. The British, Turkish, and Cuban PPD-II/III kindreds described in this study also showed strong evidence for linkage to chromosome 7q36. A recombination event in the Cuban family A confirmed D7S2423 as the flanking telomeric marker for the critical region.

Recent studies by Hing et al 15 and Radhakrishnaet al 16 described linkage of two families with PPD-II/III to chromosome 7q36; one family was of North American origin and the other of Indian origin, respectively. Until now, all families affected with PPD-II/III that have been investigated are linked to the same locus. Some families come from a completely different ethnic background and therefore the question arises whether this disorder is caused by a single, ancient mutation that has spread throughout the world population, or by several independent mutations in a single gene? We were not able to find a commonly shared allele for any of the tested markers between the Dutch, British, Turkish, and Cuban kindreds reported in the present study, suggesting that independent mutational events have occurred in the gene at chromosome 7q36.

Since one member of the Cuban family B showed a phenotype of PPD-I, and one subject from the kindred reported by Hing et al 15 showed a similar PPD-I phenotype, we wondered whether PPD-I is part of the phenotypic spectrum of the locus on chromosome 7q36 or if it can be caused by other genetic mutations. We ascertained a small family in which three subjects had a PPD-I phenotype and tested chromosome 7q36 markers. This Dutch kindred did not show linkage with the 7q36 locus, indicating that there is genetic heterogeneity for PPD-I. Considering the fact that in the Dutch PPD-I patients two have postaxial polydactyly type B (PAP-B), there is the possibility that the above described Dutch PPD-I kindred is affected with a distinct form of polydactyly affecting the preaxial as well as the postaxial borders of the upper and lower extremities. However, the classification of Temtamy and McKusick10 does not accomodate combined pre- and postaxial polydactyly as found in this family. Therefore the diagnosis in this family might be better described as crossed polydactyly as reviewed by Goldsteinet al,23 although the fact that the proband showed preaxial polydactyly on both hands and feet would fit best with PPD-I. In each case, linkage analysis in more families with PAP-B and PPD-I phenotypes are needed to study the molecular aetiology of these conditions.

In 1994, at the same time that we reported our original linkage findings, an independent study by Tsukurov et al 14 reported linkage of a kindred with complex polysyndactyly to chromosome 7q36. The family that participated in this study was of Dutch origin. It has been suggested that this family, as well as the kindred reported in our study,12 originated from the same common ancestor from Belgium.15 24 However, an extensive genealogical search, which identified the common ancestral couple of the Dutch PPD-II/III families in our study, excluded common ancestry for the complex polysyndactyly family in the period of the past 200 years. This finding suggested that the two phenotypes are either caused by different mutations in the same gene (allelic heterogeneity) or by mutations in different, but closely linked genes (locus heterogeneity).

Three patients from the kindreds reported here deserve particular attention because of their phenotype. Patient V.12 from Cuban family B is affected with brachydactyly. The patient does not share the “disease” haplotype that is found in other patients in this family, suggesting that the brachydactyly in this patient represents a separate finding unrelated to the gene defect on chromosome 7q36. Patient VI.1 from the same family has a more extensive limb phenotype than most other patients and mild scaphocephaly. This patient shares the “disease” haplotype with the other patients from the family. It is not clear whether the scaphocephaly phenotype may be connected with the gene defect on chromosome 7q36 or represents a separate entity.

The proband from Cuban family A showed a phenotype of bilateral tibial aplasia with polydactyly (MIM 188770) which has not previously been associated with the locus on chromosome 7q36.24 In addition, this patient had a mild degree of radial dysplasia on both upper extremities. Interestingly, in the kindred reported by Hinget al,15 one subject is also described with long biphalangeal thumbs and radiographical evidence of distal radial hypoplasia. This suggests that radial and tibial dysplasia/aplasia can be variant expressions of the PPD-II/III phenotype. Radial aplasia or hypoplasia usually occurs sporadically, which limits the possibilities for genetic research. In contrast, tibial aplasia or hypoplasia without associated duplication of the fibula, but with preaxial polydactyly of the toes and fingers, is known as a rare autosomal dominant trait with variable penetrance and expression.24 It would be of interest to test these families for linkage to chromosome 7q36.

More evidence for the suggestion that radial and tibial dysplasia/aplasia are variant expressions of the PPD-II/III phenotype comes from two mouse mutants strains, Hemimelic extra-toes (Hx) and Hammertoe (Hm).25 26 It has already been suggested that these mouse strains are excellent models for studying the PPD-II/III and complex polysyndactyly phenotypes linked to 7q36, respectively.12 14 Both mutations are localised very close together on mouse chromosome 5, in a region that is syntenic to human chromosome 7q36. The Hx mouse shows tibial aplasia and various degrees of radial dysplasia and extra metacarpals, metatarsals, and digits, all of which are located on the preaxial side. The striking resemblance of theHx mouse phenotype to the phenotype of the proband of Cuban family A strongly suggests that radial dysplasia and tibial hypoplasia are probably part of the phenotypic spectrum of the 7q36 locus.

Could the radial and tibial dysplasia/aplasia be the result of homozygosity for a mutation in the gene on chromosome 7q36, or is the variation the result of a modifier gene or an environmental factor? The possibility that the patients have a homozygous phenotype seems unlikely because both patients that have been described with radial/tibial dysplasia/aplasia in addition to PPD are heterozygous for the “disease” haplotype in the respective families. Furthermore, in both cases only one of the parents has the disease phenotype (fig4C).15 The answer to the question whether the variation in phenotype is caused by a modifier gene or by environmental factors could come from studies on the mouse model.

The Hx mutation arose in a non-inbred strain27 and shows great phenotypic variability, similar to that seen in the human families. The variation in phenotype is likely to have a genetic background since environmental factors for mice are strictly controlled. By breeding theHx mutation mice until the strain is completely isocongenic, the phenotype should become stable. These experiments are currently being performed by our group.

Several genes have been localised to chromosome 7q36 within the critical region. Two genes seem to be excellent candidates for the poly-/complex polysyndactyly phenotypes; the Sonic Hedgehog (Shh) gene and a homeobox gene HB9.21 28 The mouse homologue of Shh is known to be expressed in the developing limb and is an important signalling molecule for the anterior-posterior patterning of the early limb bud. In humans, mutations in this gene can cause holoprosencephaly. Marigo et al 21 cloned the human homologue of the gene and performed recombination analysis in the family with complex polysyndactyly. In an unaffected subject from this family the authors reported a recombination event, making it unlikely that the Shh gene is responsible for the complex polysyndactyly phenotype. We performed haplotype analysis with the same polymorphic marker in the branch of the Dutch pedigree that showed the recombination with D7S550. We detected a recombination event in an affected subject, thereby clearly excluding Shh as a candidate gene for PPD II/III. Additional evidence comes from experiments performed in the mouse mutant where several recombination events were observed between theHx locus andShh. 14

The homeobox gene HB9 is distantly related to theDrosophila melanogaster proboscipedia gene.28 The gene is mainly expressed in pancreas, small intestine, and colon. Whether the gene is expressed in the developing limb bud and is responsible for the polydactyly and complex polysyndactyly phenotypes is currently under investigation by our group.

The identification of the PPD gene(s) and the subsequent functional analysis will show how this gene is involved in the molecular pathways controlling the patterning of the preaxial portion of the limb, and in particular of the thumb.


The authors wish to thank all the participating families. Their contribution has been crucial for this work. We also wish to thank Professor Galjaard for his continuous support.