Discontinuous microduplications at chromosome 10q24.31 identified in a Chinese family with split hand and foot malformation

Background Split hand/foot malformation (SHFM) is a congenital disorder characterized by a cleft of the hands and/or feet due to dificiency of central rays. Genomic rearrangement at 10q24 has been found to cause nonsyndromic SHFM (SHFM3). Methods Four patients and fourteen unaffected individuals from a four-generation Chinese pedigree with typical SHFM3 phenotypes were recruited for this study. After informed consent was obtained, genome-wide copy number analysis was performed on all patients and two normal family members using the Affymetrix Cytogenetics Whole-Genome 2.7M Array. The results were then confirmed by real-time quantitative polymerase chain reaction in all available individuals of this pedigree. Candidate genes were further screened for mutation through sequence analyses. Results Copy number analysis showed a microduplication at chromosome 10q24.31-q24.32 co-segregating with the SHFM phenotype. Compared to other known genomic duplications for SHFM3, the duplication described here contains two discontinuous DNA fragments. The minimal centromeric duplicated segment of 259 kb involves LBX1, POLL and a disrupted BTRC. The minimal telomeric duplication of 114 kb encompasses DPCD and one part of FBXW4. No coding and splice-site mutations of candidate genes in the region were found. Conclusions Genomic duplications at chromosome 10q24.3, which were identified in the current study, provide further evidence for limb-specific cis-regulatory sequences in this region, highlighting the importance of chromosome 10q24.31-q24.32 in limb development and SHFM pathogenesis.


Background
Split hand/foot malformation (SHFM) is a congenital disorder characterized by variable degrees of median clefts of hands and/or feet due to absence of the central ray digits. It also presents syndactyly, and aplasia/hypoplasia of the phalanges, metacarpals and metatarsals. SHFM shows highly variability in phenotypes between and/or within families, even between limbs of a single patient, ranging from mild syndactyly to severe central clefting of the autopods, oligodactyly or monodactyly [1]. The condition can occur as an isolated entity or a part of other anomalies, with an estimated incidence of 1/8,500~1/25,000 [2].
The SHFM3 locus was initially mapped to chromosome 10q24 by linkage analysis [7][8][9]16], then narrowed to a 0.5-Mb region through the identified genomic duplications involving LBX1, BTRC and FBXW4 [17]. No intragenic mutations to date have been detected in candidate genes within or flanking duplicated regions. The relationship has been well established between SHFM3 and genomic duplications in 10q24 [17][18][19][20], a few were documented in Chinese patients [21] though a relative high birth prevalence rate of 0.64/10000 for nonsyndromic SHFM was reported [22]. In this study, we conducted an array-based genome-wide copy number variation (CNV) analysis on a Chinese family with SHFM and identified a discontinuous microduplication on 10q24. .32 as the likely cause.

Subjects
A SHFM family with four-generation-span was recruited after informed consent ( Figure 1). Four affected (II:5, III:9, III:10 and IV:3) and 14 unaffected family members received full clinical evaluation and blood samples were collected for further studies. Clinical records and radiographic images were published under the patients' written permission. This study was approved by the Research Ethics Committee of Sichuan University.

Mutation screening
Genomic DNA was isolated from blood samples using a commercially available kit (Bio Teke, Beijing, China). As TP63 is the only disease-causing gene for autosomal dominant SHFM, we first searched for possible mutations in the gene in family members as outlined previously [23]. We also screened candidate genes within the regions defined by array-based copy number variation analysis. Exons and their flanking intronic sequences of the LBX1, BTRC, DPCD and FBXW4 were amplified by polymerase chain reaction (PCR). All the resulting products were purified, and sequenced bi-directionally using an ABI 3730xl sequencer. All primer sequences are available on request.

Microarray analysis
Using the Affymetrix Cytogenetics Whole-Genome 2.7M array (Affymetrix Inc, USA), four affected and two unaffected individuals (II:6, III:8) in this family were selected to undertake CNV analysis. Experiments were carried out by Peking CapitalBio Company (Beijing, China) according to the manufacturer's protocol. The Affymetrix Chromosome Analysis Suite Software (ChAS) was used for genotype calling, quality control and CNV identification.

Identification of the duplication by quantitative PCR
To confirm duplications at 10q24.31-q24.32 mapped by CNV analysis, six EvaGreen assays were designed to determine the relative copy number (RCN) for all family members. Primers for amplifying selected fragments in duplicated regions were described previously [20]. PCR conditions and locations of assays are presented in Additional file 1: Table S1. All reactions used 500nM of each primer, 10 ng of genomic DNA, and SsoFast EvaGreen Supermix quantitative PCR (qPCR) buffer (Bio-Rad, California, USA). PCRs were performed by using a Bio-Rad C1000™ Thermal Cycler (Bio-Rad, USA) in a 10-μl volume in 96-well plates with four replicates per sample. Reactions were run in a Bio-Rad CFX96™ Real-Time System with the following conditions: 98°C for 2 min, 98°C for 1 sec, annealing for 5 sec, and 40 cycles of 98°C 1 sec. The quantification of the target sequences was normalized to one assay from chromosome 11 [18], the RCN was determined on the basis of the comparative 2 -ΔΔCt method with a normal control DNA as the calibrator [24]. The experiments were repeated three times. A~1.5-fold RCN was used for duplication. Relative DNA copy number was obtained by pair-wise comparisons of test and control DNAs.

Clinical report
Five individuals in this family were affected by SHFM with autosomal dominant transmission, but only four of them were available for this study (Figure 1). The proband (IV:3), one three-year-old boy, was affected by severe distal deficiency affecting all 4 limbs. Other three patients exhibited similar phenotypes including ectrodactyly of central digits, bilateral median cleft in feet, 3/4 finger or toe syndactly, hypoplasia of metacarpals and metatarsals. Representative clinical observations from photographs and X-ray data are summarized in Figure 2 and Additional file 2: Table S2. Notably, phenotypic variations were observed between affected family members. Triphalangeal thumb was only identified in patient II:5, and complete 1/2 toe syndactyly in individual IV:3. Similar hypoplasia/agenesis of 1 st ray existed in III:9, III:10 and IV:3. No non-limb malformations were identified in any patient.

Mutation screening in candidate genes
Before genome-wide genotyping, we did not identify any mutation in exons of TP63 gene using PCR and direct sequencing analysis. Also we sequenced four known genes in the critical regions at 10q24, and found no pathogenic mutation in the coding region of these genes.

Copy number variation analysis
Using array-based genome-wide CNV, we identified similar duplications at 10q24.31-q24.32 in the four patients, but not in the two unaffected family members. The duplicated genomic contents contain two segments. The centromeric duplications, ranging from 247 kb to 260 kb, share a common overlapping segment containing LBX1, POLL and a disrupted BTRC. The telomeric duplications of 114-125 kb comprise DPCD and one part of FBXW4 gene (Additional file 3: Table S3, Additional file 4: Figure S1). With use of qPCR assay, we confirmed the duplication co-segregating with SHFM phenotypes in the family. As predicted by ENCODE (http://www. genome.gov/10005107), this genomic region contains a variety of regulatory elements including enhancers, promoters and multiple transcription factor-binding sites. Particularly, the centromeric duplications involve two negative enhancers, hs324 and hs939, while the region between two duplications harbors a human limb-development-related positive enhancer, hs326 (chr10:103,266,649-103,267,972, http://enhancer.lbl.gov).

Discussion
SHFM is a complicated limb malformation that shows highly phenotypic variability and genetic heterogeneity. Even though six SHFM loci have been mapped to date, causal genes for a large number of cases are still unknown. At the gene level, mutations in TP63 and WNT10B can lead to autosomal dominant SHFM4 and autosomal recessive SHFM6, respectively. Recently a point mutation of DLX5 gene was reported to be responsible for autosomal recessive SHFM [25]. Genomic rearrangements at 10q23-q24 have been identified in many familial or sporadic SHFM cases [17][18][19][20], the mechanisms remain to be elucidated. In current study, we identified a previously unreported copy number variation at 10q24.31-q24.32, segregating with SHFM in a four-generation Chinese family. The minimal duplicated genomic contents include a 247-kb fragment comprising LBX1 and part of BTRC, and another fragment of 114 Kb containing POLL, DPCD, and disrupted FBXW4. Our findings expand the knowledge of genomic rearrangements leading to SHFM, and highlight the important role of chromosomal 10q24 region in regulating limb development.
A genotype-phenotype study showed that preaxial involvement of the upper limbs can be locus discriminators for SHFM3 [26]. In the current study, various preaxial upper limb anomalies were found in the affected individuals including proximal placed thumb, triphalangeal thumb, preaxial polydactyly, and absence of paraxial rays. In addition, more severe affected feet than hands, and worsening phenotype in patients of younger generations, were observed in the family. As noted in literature [9,19], the observed phenotypes strongly suggested that current disorder should be mapped to SHFM3 locus, chromosomal 10q24.
SHFM is considered a failure to maintain medial apical ectodermal ridge (AER) [1], so genes expressed in AER are reasonable candidates for SHFM3, such as the three known SHFM genes TP63 ,WNT10B and DLX5. Microduplications involving LBX1, BTRC, POLL, DPCD and one part of FBXW4 (human dactylin gene) were shown for SHFM3 in two studies [17,19]. Another study reported that a 325-kb duplication containing BTRC and POLL leads to SHFM3, which is the smallest one identified to date [20]. A small number of genes at 10q24 like BTRC, FBXW4, SUFU and FGF8 are also expressed in the developing limbs [1]. Among them, FBXW4 has been considered as the best candidate gene due to its role in the SHFM3 'like' Dactylaplasia mouse model [27,28]. BTRC is another promising candidate, because its product gets involved in essential pathways for limb development, such as the canonical WNT/β-catenin and NKFα signaling pathways. Despite the frequent observations of genomic rearrangements at 10q24 in SHFM patients, no exon mutations were found in these candidate genes.
In this study, the duplicated segments contain three entire genes (LBX1, DPCD and POLL) and one part of two genes (BTRC and FBXW4). The duplicated portion of BTRC gene contains its promoter and exon 1, while the duplicated segment of FBXW4 includes 3 0 end of the gene. Since BTRC and FBXW4 are only partially duplicated in our SHFM cases, it is less likely that overexpression of these two gene causes SHFM. Comparing our findings with those of previous studies, BTRC is the most frequently identified gene involved in the genomic rearrangements. A duplication affecting BTRC and POLL only has been reported to be responsible for SHFM3 [20]. We speculate that limb-specific regulatory elements around or in BTRC gene may play an important role in SHFM pathogenesis, by controlling the expression of genes at 10q24 or the related downstream genes.
It is unclear how genomic rearrangement at this critical region causes the SHFM phenotype. The elevated expression of BTRC and SUFU in lymphoblastoid cells detected in SHFM3 patients suggested that the gene dosage (BTRC) or long-range control mechanisms (SUFU) may underlie overexpression [20]. In the Dactylaplasia mouse model, changes in the expression levels of Lbx1, Btrc, Poll, and Dpcd were not observed except for reduced levels of normal Fbxw4 transcript [19], which implies duplications may disrupt normal expression of related genes. An increasing number of studies have shown that copy number variations are a common cause of human genetic disorders [29]. Duplications of regulatory elements have been described for human limb malformations. Brachydactyly type A2 is caused by a tandem duplication of a 5.5 kb region 3' of BMP2 [30]. Triphalangeal thumb-polysyndactyly syndrome is associated with duplications of ZPA regulatory sequence, a long range cis-regulator for SSH gene [31]. Based on current understanding of the molecular mechanisms for limb anomalies associated with duplications, the duplications described here may alter the dosage of a regulatory element involved in limb development or disrupt the co-expression domains, consequently leading to SHFM phenotypes.

Conclusion
In summary, this report describes a discontinuous duplication at 10q24.3 responsible for a typical familial SHFM3. Our data suggest that chromosome 10q24.3 may contain key elements for regulating the coordinate expression of many genes for limb development. Future studies on the developmental effects of genomic rearrangements at this region in animal models would thus complement the human data presented here. Our findings could be of value to molecular diagnosis and deeper understanding of pathogenesis of this disorder.