Synpolydactyly and HOXD13 polyalanine repeat: addition of 2 alanine residues is without clinical consequences

Type II syndactyly or synpolydactyly (SPD) is characterized by webbing between 3/4 fingers and between 4/5 toes with partial or complete digit duplication within the syndactylous web. Three loci have been identified at chromosomes 2q31, 22q13.31 and 14q11.2-q12, and have been designated as SPD1, SPD2 and SPD3, respectively [1‐3]. Unique expansion mutations in a poly alanine repeat (PolAR) of HOXD13 have been implicated in SPD1 families [4, 5]. Interestingly, all mutations discovered in at least 31 SPD1 families reported so far, were ≤ 7 alanine residue expansions. This observation has lead to a suggestion that expansion mutations ≤ 6 alanine residues do not cause the SPD1 malformation [4, 6]. However, no clinical evidence has emerged to support this hypothesis.

Within SPD1 families, there exists a broad spectrum of inter- and intra-familial phenotypic manifestations. Even though, for the phenotypes segregating in large families unique clinical categories have been proposed which remain mutually exclusive [7]. Here, we describe a large Pakistani and an Indian family with SPD1, identify four distinctive clinical categories of the phenotype in the affected subjects, and give evidence for a minor foot phenotype being a part of the SPD1 clinical spectrum. We observe identical 7 alanine expansion mutations in HOXD13 in the affected subjects in both families. We further show that two normal variants of PolAR are segregating in both families and that an expansion of two alanine residues in PolAR is not associated with the SPD1 malformation.

Genomic DNA was extracted according to standard methods. To test the candidate locus SPD1 at chromosome 2q31 for linkage with the phenotype segregating in the Pakistani family, microsatellite markers from this region were selected (D2S418, D2S1353, D2S2330, D2S2307, D2S2314, D2S138, and D2S1391). Genotyping and linkage analysis were performed as described before [9].

To screen HOXD13 for mutations, the coding sequences were PCR amplified with the primers described elsewhere [10], by employing the Advantage GC-Genomic PCR kit (CLONTECH, PaloAlto, CA, USA). The upstream and downstream regions flanking the first and last exon were amplified with the primer pairs: 5' → 3' (F1-agagggagagagggctagag, R1-ctacaacggcagaagaggac; F5-cttagaagccattcggttgt, R5-cagtaccttcagcctcctgt). Sequencing was performed on both strands with the BigDye Terminator Cycle Sequencing kit (Applied Biosystems, Foster City, CA, USA) with an automated ABI-377 genetic analyzer (Applied Biosystems). Trace chromatogram data were analyzed with Sequencher 4.7 software. For a robust screening of a polyalanine repeat expansion, one of the oligos from the primer pair exon1a [10], flanking the repeat, was fluorescently labelled (FAM), and the amplified PCR product was analyzed on a 6% denaturating polyacrylamide gel with the ABI-377 (TAMRA-500 as an internal control). Fragment analysis was performed using GeneScan (ver 3.1.2) and Genotyper (ver 2.0) softwares.

Among all non-syndromic syndactyly types SPD is clinically and genetically the most heterogeneous malformation. However, in large families it is possible to group the affected subjects on the basis of clinical variants. For instance, four distinct clinical categories have been appreciated in a large Turkish family showing SPD1 [7]. Typical SPD phenotype and various milder variants were also observed in a panel of SPD1 families reported by Goodman et al. [4]. Likewise, in the Asian families we have identified four distinctive clinical categories (Figure 3A, B, C, D). At present there is no specific molecular explanation for the occurrence of discrete clinical variants in SPD1 families. However, variations in the genetic backgrounds reflecting the presence of modifier loci in different families or in different loops of the same family could be involved. It might also well be that stochasticity in HOXD13 dosage during development could results in the phenotypic variability [11].

A milder foot phenotype (partially developed metatarsals), observed in two families with less explicit SPD features has been advocated as a separate entity associated with deletion mutations in HOXD13 outside the PolAR [12]. In contrast, we suggest that this clinical variant is a part of the SPD1 phenotypic spectrum. First, this milder foot phenotype was observed in a large Turkish SPD1 family [7]. Then, a Caucasian family with this phenotype was reported in which typical SPD clinical features were present in only few affected subjects [11]. Finally, in the present Pakistani family we observe this milder foot phenotype as well. The frequent recurrence of this clinical variant with SPD1 leads us to conclude that it is in fact a milder phenotypic manifestation of SPD1 and not a separate entity in itself. Additionally, the presence of this variant calls for a more thorough phenotypic assessment in SPD1 families involving close radiological evaluation.

The PolAR expansion mutations in HOXD13 are believed to confer gain-of-function [4]. However, this gain-of-function is confined to expansion with a threshold total length of 22 alanine residues in PolAR (i.e., expansion of ≥ 7 alanines). Therefore, it has been argued that the expansions of ≤ 6 alanine residues remain without clinical consequences [6]. Our findings in two Asian families provide for the first time molecular evidence to reinforce this hypothesis, as we show that there is repeat polymorphism in PolAR and that an addition of 2 alanine residues in the PolAR is not associated with the SPD1 phenotype. The presence of identical repeat polymorphisms and mutations in two geographically different families is a rare finding. The observation in the Pakistani family shows that the mutation as well as the normal repeat polymorphism in PolAR are inherited meiotically stable over ~6 generations. Nonetheless, our data does not indicate the 2 alanine repeat expansion in itself has a modifier effect on SPD1 phenotype. The phenotype in subjects harbouring both a 7 alanine expansion and a 2 alanine expansion in trans is not remarkably distinct from the typical SPD phenotype. Additionally, it is worth mentioning that polyalanine contractions of two and four alanine residues have been described previously without obvious pathogenic effect [13, 14], however, a 7 alanine residues contraction has been shown to be associated with brachydactyly/syndactyly phenotypes [15].

Since the proposal of purely clinical classification of non-syndromic syndactylies [16], the phenotypic spectrum of SPD has broadened extensively (reviewed in [17]). In certain instances, it overlaps with other well-characterized syndactyly types. For instance, isolated webbing of 3/4 fingers (MIM 185900) has been observed in the panel of SPD1 families reported by Goodman et al. [4], and is also present in the present two Asian families. Secondly, another separate entity, isolated 2/3 toes cutaneous syndactyly, has also been observed in SPD1 family subjects [4, 13]. The phenotypic overlap of SPD1 with clinically and genetically distinct syndactylies calls for further investigations. It is quite likely that there is a crosstalk between HOXD13 and the type I syndactyly locus (or loci) in defining the limb phenotype. We are embarking upon further studies in order to understand this clinical and genetic heterogeneity observed in synpolydactyly families.

Synpolydactyly is clinically and genetically very heterogeneous; however distinct non-overlapping clinical variants have been identified in families with synpolydactyly type I (SPD1). When a minor variant occurs as a predominant phenotype in a given family, it should not be considered as a separate entity. The expansion of two alanine residues in the polyalanine repeat of HOXD13 is not associated with an SPD phenotype.