GENETIC TESTING AND MOLECULAR BIOMARKERS Volume 21, Number 6, 2017 ª Mary Ann Liebert, Inc. Pp. 357–362 DOI: 10.1089/gtmb.2016.0415

Molecular Genetic Characterization of a Chinese Family with Severe Split Hand/Foot Malformation

Lihua Cao,1 Wei Yang,2 Shusen Wang,1 Chen Chen,1 Xue Zhang,1,2 and Yang Luo1

Aims: Split hand/foot malformation (SHFM) is a congenital limb malformation characterized by underdevel- oped or absent central digital rays, clefts of the hands and feet, and variable syndactyly of the remaining digits. SHFM is a genetically heterogeneous disease; the aim of this study was to identify pathogenic variations in a Chinese family with SHFM. Materials and Methods: Haplotype analyses with microsatellite markers covering the five SHFM loci were performed to localize the causative locus. Real-time quantitative polymerase chain reaction (qPCR) assays and inverse PCR were performed to determine the copy number variations and to amplify junction breakpoints in affected individuals. Candidate genes were further screened for mutations through Sanger sequencing. Results: A potential haplotype in the SHFM3 locus was shared by all affected individuals. qPCR and inverse PCR showed a microduplication at chromosome 10q24 spanning 488,859 bp and encompassing five entire genes, LBX1, BTRC, POLL, DPCD, and FBXW4, that co-segregated with the SHFM phenotype. No coding or splice-site mutations of these genes were found. Conclusion: We determined the molecular basis of SHFM in a Chinese family by haplotype analysis, qPCR, inverse PCR, and Sanger sequencing. Our work extends the clinical spectrum of SHFM3; provides a fine-scale delineation of the chromosomal breakpoints helping to narrow the critical region of SHFM3; and facilitates an understanding of the mechanisms underlying abnormal limb development and extraskeletal anomalies.

Keywords: SHFM3, microduplication, mutation, breakpoint

Introduction spectively. The autosomal recessive SHFM6 has been found to be caused by homozygous WNT10B mutations. The ge- plit hand/foot malformation (SHFM) is one of the netic causes of SHFM2 and SHFM5 are still unknown Smost complex human congenital limb malformations. It (Crackower et al., 1996; Berdo´n-Zapata et al., 2004; Faiyaz- is characterized by variable degrees of median clefts of the Ul-Haque et al., 2005; Kano et al., 2005; Blattner et al., 2010; hands, feet, or both due to the absence of central ray digits. It Dai et al., 2013; Rattanasopha et al., 2014). also involves syndactyly as well as aplasia/hypoplasia of The relationship between SHFM3 and genomic duplica- phalanges, metacarpals, and metatarsals. SHFM occurs in tions at 10q24 has been well established (de Mollerat et al., 0.4–1.4 per 10,000 births and accounts for 8–17% of all limb 2003). The maximum duplicated segment contains at least six Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only. reduction defects and it is observed in 1.64 per 10,000 Chi- genes, KAZALD1, TLX1, LBX1, BTRC, POLL, and FBXW4, nese newborns (Duijf et al., 2003; Dai et al., 2010; Gurrieri and the minimal duplication includes only two genes, BTRC and Everman, 2013). Currently, six loci for SHFM have been and POLL (Lyle et al., 2006). identified, including SHFM1 (7q21), SHFM2 (Xq26), SHFM3 (10q24), SHFM4 (3q27), SHFM5 (2q31), and Materials and Methods SHFM6 (12q13). Of the six SHFMs, the autosomal dominant Subjects inherited SHFM1, three and four have been found to be as- sociated with chromosomal aberrations on 7q21 encom- Individuals of a Chinese family with SHFM were re- passing the DLX5 and DLX6 genes or their regulatory cruited, including three affected and two normal family elements, duplications at 10q24, and TP63 mutations, re- members; all underwent clinical evaluation and provided

1The Research Center for Medical Genomics, Key Laboratory of Cell Biology, Ministry of Public Health, Key Laboratory of Medical Cell Biology, Ministry of Education, China Medical University, Shenyang, China. 2McKusick-Zhang Center for Genetic Medicine, State Key Laboratory of Medical Molecular Biology, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China.

357 358 CAO ET AL.

peripheral venous blood samples (Fig. 1). All participants spanning the SHFM3 region at an average density of one provided informed consent in advance and the study was marker every 36 kb as described (Lyle et al., 2006). Seven- approved by the China Medical University Institutional Re- teen assays were performed according to the results of the view Board. The siblings were retrospectively diagnosed previous 16 assays to further refine the approximate break- with epileptic seizures by reviewing their medical records point locations. Quantification of target sequences was nor- and their mother’s recollections. All affected individuals malized to an assay from chromosome 21, and the relative received radiographic examination of all limbs and magnetic copy number (RCN) was determined on the basis of the resonance imaging (MRI) of the head. Clinical records and comparative DDCT method using a normal control DNA as radiographic images were published with the patients’ writ- the calibrator. A cutoff RCN of 1.3 was used for duplication. ten permission. All analyses were carried out in Excel 2007.

Haplotype analysis Amplification and analysis of breakpoints Twenty short tandem repeat microsatellite markers cov- Inverse PCR was performed to amplify junction break- ering dominant SHFM1–5 loci were selected for haplotype points in the affected individuals. When breakpoints mapped analysis. The design of the primer pairs was based on the to within 2–4 kb by qPCR, inverse PCR was performed using genomic sequences flanking the microsatellite repeats, a forward primer from the distal region and a reverse primer polymerase chain reaction (PCR) was performed under from the proximal region, both deduced to be within the standard conditions, PCR products were separated by elec- duplication, to generate a specific product in affected indi- trophoresis on 8% denaturing polyacrylamide gel, and allele viduals. Then, the specific band was subjected to DNA se- fragments were detected with routine silver staining. Hap- quencing. RepeatMasker was used to evaluate interspersed lotypes were determined based on each individual’s genotype repeat element content flanking the breakpoints, and the and kinship. BLAT program was used to determine the origin of inserted sequences at junctions. Then, the sequences were aligned Real-time quantitative PCR analysis using the nucleotide BLAST program to check for possible identical sequences between proximal and distal breakpoints. SYBR Green real-time quantitative PCR (qPCR) assays were designed using Primer Express software. The unique- Mutation screening ness of the amplicons was checked using UCSC in silico PCR. A total of 33 qPCR assays were designed, 16 assays To analyze the genetic causes of discordant phenotypes within the family, exons and their flanking intronic sequences

for the genes LBX1, BTRC, POLL, DPCD, and FBXW4 of the affected individuals were amplified and sequenced using an ABI 3730 sequencer.

Results Clinical reports Three affected individuals were available for this study (Fig. 1). The proband (III:2), a 15-year-old girl, had severe truncation defects of all limbs characterized by monodactyly of both hands and adactyly of both feet. All her long bones were normal except for a short and right diastrophic forearm (Fig. 2). Her first seizure took place in her sleep at 13 years of age. Since then, she had experienced two tonic–clonic sei- zures during her waking hours. Her electroencephalogram (EEG) was normal when recorded at intermission. Head MRI

Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only. showed that she had a cyst within the left lateral ventricle (Fig. 3). Her mother exhibited similar limb defects excepting that she had normal long bones (Fig. 2). Her older brother, 18 years old, had even more severe truncation defects of all limbs characterized by adactyly of both hands and feet (Fig. 2). He had short stature (144.5 cm) and intellectual disability. He experienced frequent seizures from the age of three, and epileptiform activity was recorded by EEG. Head MRI showed him to have regional cerebromalacia and cor- FIG. 1. Pedigree and disease haplotype segregation of a tical atrophy within the right parietal and occipital lobes and Chinese family with SHFM. Black shading on the left half an arachnoid cyst within the right temporal lobe (Fig. 3). represents SHFM, and gray shading on the right half repre- sents seizures. Open symbols represent family members with Haplotype analysis normal hands and feet and no seizures. Circles and squares indicate females and males, respectively. The disease hap- A haplotype was shared by all affected individuals at the lotype is boxed. SHFM, split hand/foot malformation. SHFM3 locus, and no shared haplotype was observed at the MOLECULAR GENETIC CHARACTERIZATION OF SHFM 359

FIG. 2. Photographs and radiographs of the affected individuals. (A–D) Individual II1, (E–H) individual III1, and (I–L) individual III2. (A, B, I, J) Two affected individuals present with typical monodactyly anomaly in both hands and (E, F) one patient shows adactyly deformity. All affected individuals display adactyly malformation in both feet (C, D, G, H, K, L). Color images available online at www.liebertpub.com/gtmb

SHFM1, SHFM2, SHFM4, or SHFM5 locus (Fig. 1). Fur- in unaffected individuals; in addition, the RCN of FGF8 was thermore, obviously biased silver density toward the possible about 1.0 in all family members, indicating that the region,

affected allele fragments was observed in the markers around including BTRC, was duplicated, and the FGF8 was located the LBX1 and FBXW4 genes (Fig. 4), so it was presumed here outside the duplicated region. that the affected individuals had duplications in this region. To determine the size of the duplicated area in this family, we performed 14 additional qPCR assays to cover a region of qPCR analysis 722 kb in one affected individual using a cutoff RCN of 1.3 for duplication and identified duplication of at least 446 kb. To confirm whether the SHFM phenotype is associated To further refine the extent of breakpoints, 17 extra qPCR with duplication in this region, two qPCR assays were per- assays flanking the breakpoints on the basis of the above 16 formed to determine the RCN of BTRC and FGF8. We ex- qPCR assay genomic coordinates were performed. A total of amined all family members and detected that the RCN of 33 qPCR assays were used in this study, and the duplicated BTRC was about 1.5 in all affected individuals and about 1.0 region encompassed five entire genes, LBX1, BTRC, POLL, DPCD, and FBXW4 (Fig. 5).

Characterization of breakpoints Using the inverse PCR assay, the corresponding PCR product was detected in all affected individuals and not Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only. in unaffected family members. Sequence analysis of the

FIG. 4. Two microsatellite markers at SHFM3 locus in FIG. 3. Magnetic resonance imaging of the epileptic in- individual II1 (heterozygote) showing obviously biased sil- dividuals. (A) Individual III1 presents regional cerebro- ver density toward the possible affected allele fragments. malacia and cortical atrophy within the right parietal lobe (A) A marker around the LBX1 gene. (B) A marker around and an arachnoid cyst within the right temporal lobe, and the FBXW4 gene. The arrows show the obviously biased (B) individual III2 shows a cyst within the left lateral silver density toward the possible affected allele fragments. ventricle. Color images available online at www.liebertpub.com/gtmb 360 CAO ET AL.

FIG. 5. Quantitative PCR showing the RCN changes of the affected individual at 10q24. (A) Scatter plot of RCN. Red dots indicate duplicated regions and green dots indicate normal regions. (B) Schematic of corresponding gene positions at 10q24. (C) Extent of the duplication and locations of breakpoints. qPCR, real-time quantitative polymerase chain reaction; RCN, relative copy number. Color images available online at www.liebertpub.com/gtmb

chimeric PCR fragments demonstrated that the duplication cyst within the right temporal lobe and the other had a cyst was tandem (head-to-tail), and the precise size was found to within the left lateral ventricle. Their mother had no history be 488,859 bp containing the entire LBX1, BTRC, POLL, of epilepsy or any structural brain abnormalities. The limb DPCD, and FBXW4; the centromeric and telomeric break- abnormalities also differed between the affected individu- points were located at chr10:102,965,036 bp and chr10:103, als, two patients presented with typical monodactyly in both 453,894 bp, respectively. There were no repetitive sequences such as LINE, LTR, SINE, or DNA transposons represented in the centromeric junction end; however, the telomeric breakpoint end was located in a simple CA repeat. There was only one base pair C microhomology between the centro- meric breakpoint and the telomeric breakpoint (Fig. 6).

Mutation screening in candidate genes Five genes, LBX1, BTRC, POLL, DPCD, and FBXW4,in the duplication regions at 10q24 were sequenced and no dif- ferences in the coding region or the flanking intronic sequences were found between the two siblings and their mother.

Discussion SHFM occurs either as an isolated finding (nonsyndromic SHFM) or as a part of several syndromes. To date, 25 familial and sporadic cases of SHFM3 caused by rearrangements at 10q24 have been identified. These include 17 individuals with an isolated form of SHFM and 8 with syndromic forms Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only. of SHFM associated with various malformations, such as medulloblastoma, cleft palate, micrognathia, hearing prob- lems, renal hypoplasia, myopia, congenital cataracts, intel- lectual disability, and congenital heart defects (Elliott et al., 2005; Dimitrov et al., 2010; Filho et al., 2011). In this study, a tandem 489 kb microduplication at 10q24 was detected in a family with severe SHFM (adactyly or monodactyly). Two affected siblings in this family suffered seizures, which are more commonly found in patients with SHFM5 and FIG. 6. Amplification and analysis of the breakpoints. (A) Amplification of junctions showing segregation of the du- very rare in those with SHFM3. Only one individual with plication with the ectrodactyly phenotype in the Chinese SHFM3 has been reported to have seizures (Elliott and family. (B) Sequencing chromatogram of the breakpoints. Evans, 2006). Head MRI showed that both affected indi- (C) Sequence analysis of the junctions. Cen-Seq, centro- viduals studied here had structural abnormalities of the meric sequence; Pat-Seq, patient sequence; Tel-Seq, telo- brain; one had regional cerebromalacia and cortical atrophy meric sequence. Microhomology is boxed. Color images within the right parietal and occipital lobes and an arachnoid available online at www.liebertpub.com/gtmb MOLECULAR GENETIC CHARACTERIZATION OF SHFM 361

hands, while one patient had no fingers on either hand In summary, this study demonstrates that a tandem mi- (adactyly). The discordant phenotypes within the family croduplication at 10q24 causes SHFM3 with seizures. The may be explained by modifying factors, epigenetic events, extent of duplication was determined and the junction char- or environmental factors. acterized at the molecular level. Further study with more Until now, there has been no reliable explanation as to how patients and a functional and developmental approach in- the 10q24 genomic aberration causes an abnormal limb de- volving animal models would help elucidate the mechanism. velopment or why some patients have extraskeletal anoma- lies (Elliott and Evans, 2006). The real contribution of the Acknowledgments genes such as KAZALD1, TLX1, LBX1, BTRC, POLL, FBXW4, SUFU, and FGF8 within or in the vicinity of the The authors are grateful to the SHFM family for their duplication and how they are associated with the observed participation in this study. This work was supported by the phenotypic spectrum still remain to be established. Of these, National Basic Research Program of China grant No. BTRC, FBXW4, SUFU, and FGF8 are expressed in the de- 2012CB944600 and the Natural Science Foundation of China veloping limbs (Duijf et al., 2003). FBXW4 is considered the grant No. 81000253 and No. 81502176. best candidate gene due to its role in the SHFM3-like dac- tylaplasia mouse model (Sidow et al., 1999; Friedli et al., Author Disclosure Statement 2008). BTRC is another promising candidate because it is involved in the canonical WNT/b-catenin and NKFa sig- No competing financial interests exist. naling pathways, which are essential to limb development (Maniatis, 1999). LBX1 has been identified as a candidate References gene for familial idiopathic scoliosis. One patient with an isolated LBX1 microduplication showed no limb malforma- Berdo´n-Zapata V, Granillo-Alvarez M, Valde´s-Flores M, et al. tions, and there have been no cases of SHFM3 associated (2004) p63 gene analysis in Mexican patients with syndromic with scoliosis, vertebral malformations, or myopathy and non-syndromic ectrodactyly. J Orthop Res 22:1–5. (Ferna´ndez-Jae´n et al., 2014). Furthermore, LBX1 may be Blattner A, Huber AR, Ro¨thlisberger B (2010) Homozygous responsible for intellectual disability in 15% of SHFM3 pa- nonsense mutation in WNT10B and sporadic split-hand/foot tients. However, no point mutations were reported in any malformation (SHFM) with autosomal recessive inheritance. genes within the duplicated region, the elevated expression of Am J Med Genet A 152A:2053–2056. BTRC and SUFU in lymphoblastoid cells detected in SHFM3 Crackower MA, Scherer SW, Rommens JM, et al. (1996) patients suggested that the gene dosage or long-range control Characterization of the split hand/split foot malformation locus SHFM1 at 7q21.3-q22.1 and analysis of a candidate

mechanisms may be responsible for the SHFM phenotype gene for its expression during limb development. Hum Mol (Lyle et al., 2006). The current data indicate that there are no Genet 5:571–579. obvious correlations between the size of the 10q24 duplica- Dai L, Deng Y, Li N, et al. (2013) Discontinuous micro- tion and phenotype. The integrity of the extra copy of the duplications at chromosome 10q24.31 identified in a Chinese FBXW4 gene might contribute to the syndromic forms, and family with split hand and foot malformation. BMC Med syndromic patients usually have a rearrangement encom- Genet 14:45. passing the whole FBXW4 gene instead of the disrupted extra Dai L, Li YH, Deng Y, et al. (2010) Prevalence of congenital copy at the telomeric end found in nonsyndromic cases (El- split hand/split foot malformation in Chinese population. Si- liott et al., 2005; Elliott and Evans, 2006; Dimitrov et al., chuan Da Xue Xue Bao Yi Xue Ban 41:320–323. 2010). Fine delineation of the aberration breakpoints may de Mollerat XJ, Gurrieri F, Morgan CT, et al. (2003) A genomic help narrow the critical region of SHFM3 and explain to what rearrangement resulting in a tandem duplication is associated extent duplication could influence the observed phenotypic with split hand-split foot malformation 3 (SHFM3) at 10q24. differences. Among the 25 SHFM3 cases with duplication at Hum Mol Genet 12:1959–1971. 10q24, only 2 patients’ breakpoints have been cloned (Kano Dimitrov BI, de Ravel T, Van Driessche J, et al. (2010) Distal et al., 2005). This study established the genomic character- limb deficiencies, micrognathia syndrome, and syndromic istics of aberration breakpoints and determined the precise forms of split hand foot malformation (SHFM) are caused size of the duplications, which spanned 488,859 bp, including by chromosome 10q genomic rearrangements. J Med Genet Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only. five entire genes (LBX1, BTRC, POLL, DPCD, and FBXW4). 47:103–111. There were no highly homologous sequences in the break- Duijf PH, van Bokhoven H, Brunner HG (2003) Pathogenesis points, so the rearrangement mostly took place through either of split-hand/split-foot malformation. Hum Mol Genet 12: nonhomologous end-joining repair (NHEJ) or any alternative R51–R60. Elliott AM, Evans JA (2006) Genotype-phenotype correlations replication-based processes (Zhang et al., 2009; Mladenov in mapped split hand foot malformation (SHFM) Patients. and Iliakis, 2011). NHEJ is one of the prominent mechanisms Am J Med Genet A 140:1419–1427. underlying the repair of double-stranded DNA breaks. It in- Elliott AM, Reed MH, Roscioli T, et al. (2005) Discrepancies in volves joining two DNA ends in the absence of any sequence upper and lower limb patterning in split hand foot mal- homology and in a manner that tolerates nucleotide loss or formation. Clin Genet 68:408–423. addition at the junction site. The replication-based processes Faiyaz-Ul-Haque M, Zaidi SH, King LM, et al. (2005) Fine mainly include microhomology-mediated FoSTeS/MMBIR, mapping of the X-linked split-hand/split-foot malformation aberrant firing of replication origins, and replication timing- (SHFM2) locus to a 5.1-Mb region on Xq26.3 and analysis of associated mechanisms. These processes may occur multiple candidate genes. Clin Genet 67:93–97. times and may lead to complex rearrangements (Lee et al., Ferna´ndez-Jae´n A, Suela J, Ferna´ndez-Mayoralas DM, et al. 2007; Hansen et al., 2010). (2014) Microduplication 10q24.31 in a Spanish girl with 362 CAO ET AL.

scoliosis and myopathy: the critical role of LBX. Am J Med Mladenov E, Iliakis G (2011) Induction and repair of DNA Genet A 164A:2074–2078. double strand breaks: the increasing spectrum of non- Filho AB, Souza J, Faucz FR, et al. (2011) Somatic/gonadal homologous end joining pathways. Mutat Res 711:61–72. mosaicism in a syndromic form of ectrodactyly, including eye Rattanasopha S, Tongkobpetch S, Srichomthong C, et al. (2014) abnormalities, documented through array-based comparative Absent expression of the osteoblast-specific maternally im- genomic hybridization. Am J Med Genet A 155A:1152–1156. printed genes, DLX5 and DLX6, causes split hand/split foot Friedli M, Nikolaev S, Lyle R, et al. (2008) Characterization of malformation type I. J Med Genet 51:817–823. mouse Dactylaplasia mutations: a model for human ec- Sidow A, Bulotsky MS, Kerrebrock AW, et al. (1999) A novel trodactyly SHFM3. Mamm Genome 19:272–278. member of the F-box/WD40 gene family, encoding dactylin, Gurrieri F, Everman DB (2013) Clinical, genetic, and molecular is disrupted in the mouse dactylaplasia mutant. Nat Genet 23: aspects of split-hand/foot malformation: an update. Am J Med 104–107. Genet A 161A:2860–2872. Zhang F, Khajavi M, Connolly AM, et al. (2009) The DNA Hansen RS, Thomas S, Sandstrom R, et al. (2010) Sequencing replication FoSTeS/MMBIR mechanism can generate geno- newly replicated DNA reveals widespread plasticity in human mic, genic and exonic complex rearrangements in humans. replication timing. Proc Natl Acad Sci U S A 107:139–144. Nat Genet 41:849–853. Kano H, Kurosawa K, Horii E, et al. (2005) Genomic re- arrangement at 10q24 in non-syndromic split-hand/split-foot malformation. Hum Genet 118:477–483. Lee JA, Carvalho CM, Lupski JR (2007) A DNA replication Address correspondence to: mechanism for generating nonrecurrent rearrangements as- Yang Luo, PhD sociated with genomic disorders. Cell 131:1235–1247. The Research Center for Medical Genomics Lyle R, Radhakrishna U, Blouin JL, et al. (2006) Split-hand/ China Medical University split-foot malformation 3 (SHFM3) at 10q24, development of 77 Puhe Road rapid diagnostic methods and gene expression from the re- Shenyang North New Area gion. Am J Med Genet A 140:1384–1395. Shenyang 110122 Maniatis T (1999) A ubiquitin ligase complex essential for the China NF-kappaB, Wnt/Wingless, and Hedgehog signaling pathways. Genes Dev 13:505–510. E-mail: [email protected]

Downloaded by 59.46.65.6 from online.liebertpub.com at 06/21/17. For personal use only.