A Familial Cri-du-Chat/5p Deletion Syndrome Resulted from Rare Maternal Complex Chromosomal Rearrangements (CCRs) and/or Possible Chromosome 5p Chromothripsis

Cri-du-Chat syndrome (MIM 123450) is a chromosomal syndrome characterized by the characteristic features, including cat-like cry and chromosome 5p deletions. We report a family with five individuals showing chromosomal rearrangements involving 5p, resulting from rare maternal complex chromosomal rearrangements (CCRs), diagnosed post- and pre-natally by comprehensive molecular and cytogenetic analyses. Two probands, including a 4½-year-old brother and his 2½-year- old sister, showed no diagnostic cat cry during infancy, but presented with developmental delay, dysmorphic and autistic features. Both patients had an interstitial deletion del(5)(p13.3p15.33) spanning ∼26.22 Mb. The phenotypically normal mother had de novo CCRs involving 11 breakpoints and three chromosomes: ins(11;5) (q23;p14.1p15.31),ins(21;5)(q21;p13.3p14.1),ins(21;5)(q21;p15.31p15.33),inv(7)(p22q32)dn. In addition to these two children, she had three first-trimester miscarriages, two terminations due to the identification of the 5p deletion and one delivery of a phenotypically normal daughter. The unaffected daughter had the maternal ins(11;5) identified prenatally and an identical maternal allele haplotype of 5p. Array CGH did not detect any copy number changes in the mother, and revealed three interstitial deletions within 5p15.33-p13.3, in the unaffected daughter, likely products of the maternal insertions ins(21;5). Chromothripsis has been recently reported as a mechanism drives germline CCRs in pediatric patients with congenital defects. We postulate that the unique CCRs in the phenotypically normal mother could resulted from chromosome 5p chromothripsis, that further resulted in the interstitial 5p deletions in the unaffected daughter. Further high resolution sequencing based analysis is needed to determine whether chromothripsis is also present as a germline structural variation in phenotypically normal individuals in this family.


Introduction
Genomic imbalances such as chromosomal aberrations have been long recognized to be a major cause for genetic disorders, resulting in miscarriages, neonatal birth defects, postnatal developmental delay, autistic spectrum disorders and intellectual disability [1]. The phenotypes seen in each affected individual with chromosomal syndromes clearly depend on the particular chromosomal region with a variability in the clinical presentations [2]. Chromosome 5p deletion or Cri-du-chat syndrome (CDCs, MIM 123450) was first described by Lejeune in 1963 [3] and it is the one of most common chromosomal deletion syndrome in humans [4]. The incidence of CDCs is between 1:50,000 to 1:37000 live births [5]. The hallmark clinical features of CDCs include high-pitched cat-like monochromatic cry, microcephaly, a round face, hypertelorism, micrognathia, epicanthal folds, hypotonia, prominent nasal bridge, and severe psychomotor and intellectual disability [6]. Recurrent respiratory infections are also frequently observed in CDCs and pneumonia is the major cause of neonatal or infantile death [4,7].
Approximately 80% of the CDCs patients carry a de novo 5p terminal or interstitial deletion and the majority of these deletions are paternal origin [6]. Less than 5% of the patients have de novo translocations or other rare chromosomal aberrations such as complex chromosomal rearrangements (CCRs) [8,9]. About 10%-15% of the 5p deletions result from unbalanced segregation of a parental balanced rearrangement such as translocation or inversion [10], but very rarely from a balanced parental insertion [11] or CCRs [12].
CCRs are constitutional structural rearrangements involving more than three chromosomes with more than two breakpoints [13]. Typically, CCRs are three-way translocations with one breakpoint in each chromosome; however, CCRs with up to fifteen breakpoints have been reported [14]. Individuals with de novo CCRs, resulting in genomic imbalances at the chromosome breakpoints are frequently reported to have a high incidence of abnormal phenotypes and developmental delay/intellectual disability [15]. Heterozygous carriers of balanced CCRs are usually phenotypically normal, but have a significant risk [16] to have multiple spontaneous abortions [13] or chromosomally abnormal offspring [17].
Chromothripsis is a phenomenon in which tens to hundreds of genomic rearrangements occur in a one-off cellular crisis. Originally, it was observed in 2-3% of different cancer types [18]. Recent studies have shown that pediatric patients with constitutional abnormalities and de novo CCRs or complex genomic rearrangements (CGRs) also harbor chromothripsis [19,20]. Using mate-paired sequencing along with molecular cytogenetic analyses in a family trio, including a proband with constitutional defects and CCRs of t(1;10;4)(p32.2;q21.1;q23) [19], a direct evidence of chromothripsis was found in two of the three chromosomes involved in the CCRs, and therefore, this catastrophic event most likely also has driven the de novo CCRs or CGRs in germline of the patients. However, such event has never been observed in phenotypically normal individuals.
Recent advances in clinical genetic settings have enabled integrated molecular and cytogenetic testing. The traditional chromosome analysis, Fluorescence in situ Hybridization (FISH), multiplex ligation-dependent probe amplification (MLPA), quantitative polymerase-chain-reaction (q-PCR) and short tandem repeats (STR) assay, and microarray based testing etc. are clinically utilized and have shown significant impact clinical human genetics, especially in diagnosing and counseling familial syndromes e.g. familial CCRs [21,22] or familial 5p2/CDCs syndrome [23,24].
Here, we report a 5p2/CDCs syndrome family with very rare and unique maternal CCRs, diagnosed by integrated molecular and cytogenetic analyses.

Patients and Family History
This study was approved by medical ethics and institutional review board at Zhongshan School of Medicine, Sun Yat-sen University and, a consent signed by both parents was obtained. Probands were a four and a half-year-old male and a two and a half-year-old female born to healthy and unrelated parents. Both pregnancies were uneventful with no history of drug or alcohol usage during the pregnancies. The birth weight, length and head circumference on both probands were all within normal range in Chinese neonatal population. There were no cat-like cries observed at the birth in both probands. The male proband has incompletion cleft-palate, congenital hypertrophic pyloric stenosis, developmental delay and dysmorphic features, including round face, micrognathia and hypertelorism, external canthus upslope, plagiocephaly as well as some midfacial hypoplasia, short philtrum, large nasal bridge, and low-set ears ( Figure 1). Simian creases were observed in both hands. In addition, he developed seizures at 8 months of age, and had neurodevelopmental abnormalities showing apraxia, callosum and alba dysplasia, cerebellum dysplasia, and ventriculomegaly. He raised his head at 9 months, crawled at 19 months and walked at 42 months and, had severe language delay with no speech till 38 months of age. Comparing with WHO 2006 reference, his circumference was 46.1 cm (23SD) at 4 years and 6 months and 46.5 cm (23SD) at age of 11 years and 9 months. The female proband had similar phenotypes including developmental delay, intellectual disability, as well as dysmorphic features. She died of pneumonia six months after the first genetic clinical visit.
The mother had a total of eight pregnancies ( Figure 2).Notably, in between two probands, she also had a first trimester miscarriage. Genetic counseling was provided to the family after the first genetic visit of the probands. She subsequently had five additional pregnancies including two spontaneous abortions, two terminated pregnancies due to abnormal prenatal testing results and one delivery to a phenotypically normal daughter. This phenotypically normal daughter was followed up and evaluated for several years. At age of seven years, her IQ was 105 by WAIS intelligence test for children, which is average by comparing with the same age pediatric population in China.

Conventional Cytogenetic Analysis
Chromosome analysis on peripheral blood (on I1, I 2, II1, II 2, III 1, III 3 and III 6), prenatal cord blood (on III5) and amniocytes (on III6, III 8) ( Figure 2) were performed following the standard clinical cytogenetics laboratory protocols [25,26]. The banding resolutions were ,550 and ,400 for blood samples and amniocytes, respectively. At least twenty metaphases were analyzed and chromosomal results were interpreted and reported according to the international system for human cytogenetic (ISCN 2005 and 2009) nomenclature.

Short Tandem Repeats (STR) Analysis
Genomic DNA from peripheral blood and amniocytes were extracted using a QIAamp DNA Blood Maxi Kits (QIAGEN, Germany) and QIAamp DNA Mini Kits (QIAGEN, Germany), respectively.
Polymerase-chain-reactions (PCR) were carried out on II 1, II 2 and III1 and prenatally on III-6, III-8 using 5 ng of DNA, 2.5 mM MgCl2, 10 mMTris-HCl pH 8.3, 50 mMKCl, 250 mM each dNTP, 0.625 pmol of each primer, 0.25 u HotstarTaq DNA polymerase (QIANGEN GmbH, Hilden, Germany) in a total reaction volume of 5 ul. All PCRs were performed on a GeneAmp 9700 Therm cycle (PE Applied Biosystem), an initial denaturation 96uC for 5 min was followed by 35 cycle of 30 s at 95uC, 30 s at a prime-specific annealing temperature and 40 s at 72uC. The final extension was at 72uC for 10 min. PCR products were analyzed on an ABI PRISM 3100 Genetic Analyzer (PE Applied Biosystems). Allele sizes and peak areas were determined using GeneScan version 3.1, Genotype version 2.1 and LINKAGE version 5.1 (Applied Biosystems, CA). The genotypes at each locus were examined in the tested subjects and compared with the parental genotypes to determine whether there was non-Mendelian segregation or an apparent deletion (i.e. absence of a maternal or paternal allele).

Oligonucleotide Array Comparative Genomic Hybridization (array CGH) and Data Analysis
Oligonucleotide array CGH was performed on the male proband (III 1 at 12 years and 9 months old) and his phenotypically normal younger sister (III 6 at 6 years and 1 month old), as well as, both parents (II 1, II 2) using oligonucleotide-based arrays containing 180,000 probes from Agilent Technologies (Santa Clara, CA) according to manufacturer's instruction, for a whole genome copy number analysis. Oligonucleotides probes on this 4x180K array were annotated against NCBI Build 37 (hg19). Array CGH image files were quantified using Agilent Feature Extraction software (version 10.5). Text file outputs containing quantitative data were imported into the Nexus copy number analysis software (Biodiscovery, Segundo, CA). Log2 ratio .0.2 was defined as gain and ,20.5 was defined as loss and all other ratios as normal.

Fluorescence in Situ Hybridization (FISH)
FISH analyses were performed according to published protocols [27] in order to confirm chromosome abnormalities, STR and array CGH results. A total of twenty-one bacterial artificial , and RP11-876N17) mapping to 5p13.3-p15.32 region and three BAC clones (RP11-1029D4, RP11-916H5, and RP11-945C11) mapping to chromosome 21q21.1-q22.11 were selected from the University of California-Santa Cruz (UCSC), with detailed corresponding linear map position available in genome (

Chromosome Analysis
Chromosome analysis of peripheral blood on the two probands (III 1, III 3) revealed an identical interstitial deletion of chromosome 5p13.3p15.3. A clinical diagnosis of CDCs or chromosome 5p deletion syndrome was made based on chromosomal analyses and clinical manifestations in both probands. Subsequently, this family was provided with genetic counseling. Chromosome studies on both parents were performed and revealed the phenotypically normal mother to carry an apparently complex karyotype of 46,XX,ins(11;5) (q23;p13.3p15.3),inv (7)(p22q32) showing an apparently one balanced insertion between chromosomes 5 and 11, and a pericentromeric inversion 7, whereas the father had a normal karyotype. Further chromosome analysis on maternal grandparents confirmed the abnormal finding in the mother to be a de novo event. Due to the 5p deletions in both probands, the miscarriage of III 2 and the complex karyotype in the mother, this couple agreed to perform prenatal diagnosis for all future pregnancies. Prenatal chromosome analysis was performed on III 5 using cord blood at 21 week gestational age, and on III 6 and III 8 using amniotic fluid at 19 week gestational age (Figure 2). III 5 and III 8 showed abnormal karyotypes of 46,XY,del(5)(p13.3p15.3)mat and 46,XX,del(5)(p13.3p15.3),inv(7)(p22q32)mat, respectively, and both pregnancies were elected for termination by parents after genetic counseling. III 6 showed an apparently balanced insertion of 46,XX,ins(11;5)(q23;p13.3p15.3)mat. Additional prenatal molecular based testing was recommended to confirm the chromosome result. Partial karyotypes of the male proband, the mother and the phenotypically normal sister are shown in Figure 3 A, B and C, respectively.

STR Results
STR testing was performed on both parents, the male proband III 1, as well as prenatally on III 6 and III 8 (Table 1 and Figure 4). In a total of 19 STR markers tested, the male proband (III 1) only obtained father's allele apparently at D5S676, D5S1953, D5S580, D5S1957, D5S2095, D5S630, D5S2004, D5S416, D5S2096, D5S2113, D5S385 and D5S2061 (Figure 4). Because three markers (D5S406, D5S635 and D5S2854) were not informative, the encompassed interstitial deletion was estimated to be approximately 22.5 Mb between bands 5p13.3 and 5p15.33 (chr5:7439488-29976793), which was consistent with the above chromosome finding. The same finding was observed in III 8 and confirmed 5p deletion by the prenatal chromosome analysis. STR analysis in III 6 showed a haplotype identical to the mother, which was consistent with the prenatal chromosome analysis of the apparently balanced insertion ins (11;5). This pregnancy was continued and a phenotypically normal baby girl was delivered.

Array CGH and FISH Results
The male proband and his phenotypically normal sister were followed up extensively for clinical genetic evaluations. Eight years after the first genetic visit of the probands, the array CGH technology became available and it was introduced to this family. Because of the unique familial 5p-history and the rare CCRs in mother, this family agreed to have array CGH testing on both parents, the male proband III 1 (at 12 years and 9 months old) and his phenotypically normal sister III 6 (at 6 years and 1 months), for a whole genome copy number assessment. Array CGH revealed the minimal interstitial deletion in the male proband at chromosome bands 5p15.33-p13.3 to be 26.22 Mb, within the genomic interval 4200304-30493484 ( Figure 5A), which was slightly larger than the deletion detected by the STR testing although was consistent with the chromosome results. The discrepancy between array CGH and STR results are most likely due to the three non-informative STR markers tested fall in breakpoints regions (Table 1) Figure 5B).The proband and the normal sister shared the distal and proximal breakpoints in 5p15.33 and 5p13.3 respectively. There were no abnormal copy number aberrations identified in both parents.
To further investigate possible clinical consequences of the three deletions in 5p in the phenotypically normal sister III 6, we carefully analyzed genes in these genomic intervals. The first interstitial deletion in III 6 expanded approximately 2.89 Mb in 5p15.31p15.33, encoding 11 known genes including LOC340094, ADAMTS16, KIAA0947, FLJ33360, MED10, UBE2QL1, LOC255167, NSUN2, SRD5A1, PAPD7 and MIR4278. Of these, haploinsufficiency of SRD5A1 along with other genes in 5p15.1 region [28] are recently reported to be associated with hypospadias and cerebellar hypoplasia in a prenatal study. Only two genes LOC643401 and LSP1P3are located in the second deletion of 5p14.2 and both have not been characterized for gene function and disease association. No known genes were found in the third deletion region of 5p13.3p14.1.

Discussion
The majority of chromosome 5p deletions are associated with CDCs [6]. Traditionally, CDCs patients are diagnosed based upon the clinical manifestations, chromosome and/or FISH analyses, or molecular based testing such as MLPA or PCR. Although CDCs is a well-defined genomic disorder, individuals with this syndrome show phenotypic and cytogenetic variability. Several genotype-phenotype studies revealed that the size of chromosome 5p deletions could vary from a single chromosome cytogenetic band to the entire chromosome 5p, and the severity or spectrum of clinical phenotypes i.e. intellectual disability and microcephaly in CDCs or 5p syndromes are related to the size and the location of the deletions [6,29]. For example, the single chromosome band 5p15.2 deletion is reported to be responsible for dysmorphism and intellectual disability; the proximal region of 5p15.3 is associated for ''cat-like'' cry and speech delay [30]. These chromosomal bands are considered to be critic regions for CDCs. Genes in this region include the SEMAF, CTNND2 [5], which involve in brain development and function, the FLJ25076-UBC-E2 homologous gene which is highly expressed in thoracic and scalp tissues. Haploinsufficiency of these contiguous candidate genes are most likely the cause of classic spectrum in CDCs [30]. On the other hand, identical deletions of chromosome 5p14 region are reported in phenotypically normal parents and their affected children, indicating that 1) this 5p14 region is not as critic as the 5p15 region, or 2) this region encodes the possible recessive alleles resulting in reduced expressivity or variable clinical manifestations [6]. In our current study, three microdeletions resulting from the maternal CCRs were identified in the phenotypically normal sister (III 6) of the probands. Both second and third deletions are inside or overlapped with 5p14 region. Very few genes are encoded in these two deleted regions making these two genomic intervals less clinically relevant. Although there were 11 known genes (LOC340094, ADAMTS16, KIAA0947, FLJ33360, MED10, UBE2QL1, LOC255167, NSUN2, SRD5A1, PAPD7, and MIR4278) located in the first deletion in 5p15.33p15.31, only a few of them are fully characterized for gene functions and disease association. Recent whole exome sequencing study revealed that homozygous splicing mutation in NSUN2 to be a cause of Dubowitz-like syndrome, an autosomal recessive disorder characterized by the constellation of mild microcephaly, growth and mental retardation, eczema and peculiar facies [31]; The ADAMTS16 gene has been reported to play a role of the metalloproteinase in murine genitourinary development [32], and loss-of-function mutations of the MED10 gene have been previously linked with WNT/GSK3b/b-Catenin pathway function [33]. Only SRD5A1 haploinsufficiency has been speculated to be associated with cerebellar hypoplasia, hypospadias, and facial dysmorphisms in a prenatal study [28]. Our extensive clinical evaluation in the phenotypically normal sister III 6 showed no evidence of dysmorphism or intellectual disability or other apparent findings; the IQ test at age seven in this girl was 105. We postulate that these microdeletions in III 6 are either nonpathogenic benign copy number variant or only resulted in unnoticeable minor clinical manifestations. Certainly, this individual needed to be followed up and evaluated frequently in her future genetic counseling.
The majority of the CDCs patients are diagnosed in first month of life or within first year whereas the minority is diagnosed from 13 month to 47 years old [34]. Due to the limited clinical genetic testing in China, until recently, our patients were diagnosed at age 4 1/2 and 2 1/2 years old. Although our probands did not have the characteristic cat-like cry at birth or during their infancy (or the cat-like cry was not recognized), they did have other typical dysmorphic features as reported in CDCs and remarkable developmental delay and intellectual disability. The integrated molecule and cytogenetic testing was able to not only define the large interstitial deletion del(5)(p13.3p15.33) and confirm a clinical diagnosis in probands, but also help reveal the CCRs in the mother and therefore provide the appropriate genetic counseling to the family.
Microarray based testing with significantly improved overall resolution is a more accurate clinical testing for genome-wide copy number assessment in patients with congenital anomalies [35]. Our array CGH study was able to refine the deleted segment in the proband better than STR testing, because three STR markers D5S406, D5S635 and D5S2854 were all apparently not informa-tive. We believe that the power of microsatellite analysis is limited especially when it comes to the breakpoints involving complex rearrangements. In addition, array CGH did not detect apparent copy number aberrations in the mother, indicating that the three insertions and one pericentromeric inversion observed are all probably balanced at the array CGH testing level, and might also explain her normal phenotype. However, it did detect three microdeletions in the phenotypically normal daughter. These array findings have led to the discovery of two cryptic ins(21;5) in the mother by using additional targeted FISH analysis. Although we could not further investigate the second deletion (0.56 Mb) in III 6 by FISH due to lack of available BAC clones in 5p14.2 region, it is likely this deletion is also resulted from a maternal cryptic rearrangement. All these findings also raised questions - What had driven these germline copy number variations in the daughter? Given so many abnormal pregnancies produced, does the mother really carry all balanced rearrangements at the molecular level?
Recent studies have shown the correlations between CCRs and/or CGRs and chromothripsis in pediatric patients with congenital abnormalities [19,20]. Although the phenotypically normal mother did not have apparent copy number aberrations on 5p, this could be due to limited resolution (180 K total oligo coverage) of the oligo array CGH applied in this study. We believe that, the extremely complex CCRs in the mother might have resulted from chromosome 5p chromothripsis, and that was exactly what had driven so many affected individuals and pregnancies in this family. The microdeletions observed in the daughter are the direct evidence of such underline mechanism, although high resolution sequencing base testing is needed to confirm such speculation.
Without prenatal integrated molecular and cytogenetic testing, consequently, this woman would have produced four children and all with 5p deletion syndrome or CDCs (Figure 2). The clinical significances of the three microdeletions identified in the phenotypically normal sister were not clear, although they were most likely representing non-pathogenic based on current avail-able genomic variation databases; however, these findings will be valuable for her future genetic counseling.
In summary, we are reporting a very unique familial chromosome CDCs/5p deletion syndrome, resulted from unusual maternal de novo CCRs and/or chromosome 5p chromothripsis. The integrated molecular and cytogenetic testing not only fully characterized the maternal CCRs, but also enabled unveiling this familial event and providing accurate post-and pre-natal diagnosis as well as appropriate genetic counseling. Certainly, further sequencing based testing is needed to help determine whether chromothripsis also exists in phenotypically normal individuals in this family.