Complex intrachromosomal rearrangement in 1q leading to 1q32.2 microdeletion: a potential role of SRGAP2 in the gyrification of cerebral cortex

Background Van der Woude syndrome (MIM: 119300, VWS) is a dominantly inherited and the most common orofacial clefting syndrome; it accounts for ~2 % of all cleft lip and palate cases. Intellectual disability (ID) is characterized by significant limitations, both in intellectual functioning (cognitive deficit) and in adaptive behavior as expressed in conceptual, social and practical adaptive skills. Karyotyping has been the first standard test for the detection of genetic imbalance in patients with ID for more than 35 years. Advances in genetic diagnosis have laid chromosomal microarrays (CMA) as a new standard and first first-line test for diagnosis of patients with ID, as well as other conditions, such as autism spectrum disorders or multiple congenital anomalies. Case Presentation The present case was initially studied due to unexplained cognitive deficit. Physical examination at the age of 18 years revealed cleft palate, lower lip pits and hypodontia, accompanied with other dysmorphic features and absence of speech. Brain MRI uncovered significantly reduced overall volume of gray matter and cortical gyrification. Banding cytogenetics revealed an indistinct intrachromosomal rearrangement in the long arm of one chromosome 1, and subsequent microarray analyses identified a 5.56 Mb deletion in 1q32.1-1q32.3, encompassing 52 genes; included were the entire IRF6 gene (whose mutations/deletions underlay VWS) and SRGAP2, a gene with an important role in neuronal migration during development of cerebral cortex. Besides, a duplication in 3q26.32 (1.9 Mb in size) comprising TBL1XR1 gene was identified. Multicolor banding for chromosome 1 and molecular cytogenetics applying a battery of locus-specific probes covering 1q32.1 to 1q44 characterized a four breakpoint-insertional-rearrangement-event, resulting in 1q32.1-1q32.3 deletion. Conclusions Considering that the human-specific three-fold segmental duplication of SRGAP2 gene evolutionary corresponds to the beginning of neocortical expansion, we hypothesize that aberrations in SRGAP2 are strong candidates underlying specific brain abnormalities, namely reduced volume of grey matter and reduced gyrification. Electronic supplementary material The online version of this article (doi:10.1186/s13039-016-0221-4) contains supplementary material, which is available to authorized users.


Background
Van der Woude syndrome (MIM: 119300, VWS) is a dominantly inherited and the most common orofacial clefting syndrome; it accounts for~2 % of all cleft lip and palate cases [1]. A clinical synopsis for VWS predominantly includes mouth and teeth abnormalities (lower lip pits, cleft lip and/or palate and hypodontia).
In addition, abnormalities of limb, skin, nails and genital and/or hearing loss could be present [2]. 1q32-41 chromosomal location was mapped as being critical for VWS in 1987, when a patient with an interstitial deletion of chromosome 1q32-41 was reported [3]. Only in 2002, by direct sequencing, mutations in the IRF6 gene were detected and linked to VWS and popliteal pterygium syndrome [4]. About 70 % of VWS causal mutations occur in IRF6 gene [5], however, in less than 2 % of individuals with VWS entire IRF6 gene is deleted [6].
Nowadays it is worldwide recognized that IQ test score of 70-75 and lower, indicates a limitation in intellectual functioning or cognitive deficit. Intellectual disability (ID) is characterized by significant limitations, both in intellectual functioning (cognitive deficit) and in adaptive behavior as expressed in conceptual, social and practical adaptive skills [7]. ID is divided into five categories based on IQ (mild, moderate, severe, profound and unable to classify) [8]. In addition to that, ID can be grouped into syndromic (patients with one or multiple additional clinical features) and non-syndromic (i.e. ID as sole clinical feature). Karyotyping has been the first standard test for the detection of genetic imbalance in patients with ID for more than 35 years. Advances in genetic diagnosis have laid chromosomal microarrays (CMA) as a new standard and first first-line test for diagnosis of patients with ID. Cytogenetic and CMA as a "state-of-the-art" molecular-cytogenetic method are used in other conditions as well, such as autism spectrum disorders, multiple congenital anomalies and diverse brain diseases. Genetic findings in children with autistic disorders support conclusion that cytogenetic and molecularcytogenetic studies should be considered as compulsory in terms of detecting possible genetic causes of cognitive deficit and brain diseases [9,10].

Case presentation
At the age of 18 years the female patient was studied due to unexplained cognitive deficit. She was born after an uneventful pregnancy after 32 weeks of gestation (birth weight: 2400 g, birth length: 40 cm) as a child of closely related parents (3rd degree). According to scant data on her early development, she started to walk at the age of 2 years and her language development was severely delayed. The girl was born with palathoshisis which was surgically corrected at the age of 3 years, when also the encephalopathy, vesicoureteral reflux, expressive (motor) dysphasia and developmental delay were diagnosed. At the age of 9 years she still could not control the sphincters and was diagnosed with moderate to severe cognitive deficit and undeveloped speech (due to articulation disorder), and was moved from socially deprived environment in foster care homes. In addition to cleft palate as the primary clinical feature, she had thin lips with bilateral pits, indistinct and short philtrum, flat midface, narrow palpebral fissures, hypodontia (missing upper incisor) with irregular dental growth, acne-prone skin, hirsutism, broad thumbs, and short and tapering fingers. By the age of 18 years she developed some limited communication skills, but due to articulation irregularities the speech was rather incomprehensible and reduced to short sentences. Graphomotor development was delayed; no writing and reading skills were developed. She was deprived in spatial orientation and was not able to understand the concept of time.

Cytogenetics and molecular cytogenetics
This patient was studied within the project approved by the ethics committee of the School of Medical, University of Zagreb and the institution from which the patient comes. Peripheral blood was taken with a written informed consent by the institution from which the patient comes. Banding cytogenetics from peripheral blood lymphocytes was done according to the standard procedure. Fluorescence in situ hybridization (FISH) was done according to standard protocols using multicolorbanding probe-sets (MCB) and bacterial artificial chromosome (BACs) probes [9,10]. All BACs used in this study are listed in Additional file 1. Genomic DNA was extracted from whole blood by Puregene DNA Purification Kit (Gentra Systems, Minneapolis, MN, USA) following manufactureinstructions. Multiplex ligation dependent probe amplification (MLPA) analysis was performed according to the manufacturer's instructions (MRC-Holland, the Netherlands) using ABI-PRISM 3130XL Genetic Analyzer (Applied Biosystems, Foster City, USA). MLPA data analysis was done by use of Gen-eMarker software package (SoftGenetics, USA). Subtelomeric MLPA (P036-E2, P070-B2) and microdeletion/ microduplication (P245-B1, P297-B2) probe sets were used. Arraycomparative genomic hybridization (aCGH) was performed on Agilent oligonucleotide SurePrint G3 Human Genome microarray 4X180 K (Agilent Technologies, Santa Clara, CA, USA). The array was processed following the manufacturer's recommended protocol, and a sex-matched non-disease control sample (Promega, Mannheim, Germany) was used as a reference. Array results were analyzed using Cytogenomics software setting ADM2 aberration algorithm. Publicly available data for spatio-temporal expression profiles in brain [11] was screened by specially designed algorithm for all detected genes with aberrant copy number as previously described [12].

Brain MRI
Magnetic resonance imaging was performed on a 3T MR device (Magnetom TrioTim, Siemens; Germany), with 32-channel head coil, using standard set of sequences that included sagittal 3D magnetizationprepared rapid acquisition gradient echo (MPRAGE) sequence with voxel size 1 mm x 1 mm x 1 mm (TR = 2300 ms; TE = 3 ms; flip angle = 9 degrees; matrix : 256 x 256). Volumetric processing of MR images was conducted by using the CIVET (version 1.1.11) pipeline developed at Montreal Neurological Institute, Brain Imaging Centre [13]. This automated pipeline provided tools for observer independent corticometric analysis [14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. The calculated measurements included volumes of the grey and white matter tissue, of cerebrospinal fluid, and gyrification index for both hemispheres. Quantification of regional cortical thicknessaverages where lobe borders were determined by sulcal landmarks and detected by the CIVET pipeline, and regional/lobar cortical volume and surface area estimates were made for parietal, occipital, frontal, and temporal lobes, and for isthmus of cingulate gyrus, parahippocampal and cingulate gyrus, and insula. The gyrification index is a metric that quantifies the amount of cortex buried within the sulci compared to the amount of surface visible cortex. The gyrification index was computed by dividing the total cortical surface area with the area of the convex hull.
Visual evaluation of brain MRI exam at the age of 21 showed pineal cyst, 8 × 5 × 8 mm in size, without compressive effect on surrounding structures. Cavum septi pellucidi and cavum verge were present in medial line. There was reduced pneumatisation of mastoid cells on both sides. The rest of the findings were described as normal. Volumetric analysis revealed overall reduced volume of gray matter (−2.91 STDEV) and cortical gyrification (gray: −7.27 STDEV; white: −4.79 STDEV; mid: −8.80 STDEV) ( Table 2). Remaining findings of brain volumetric analysis are described in Additional file 3.

Clinical physical presentation -van der Woude syndrome
Here we report a 1q32.1-1q32.3 deletion of 5.56 Mb in a girl patient presenting a main features of VWS accompanied with severe cognitive deficit. Most prominent VWS phenotypic characteristics in the present case are well described in the literature and could be assigned to at least several genes (LAMB3, G0S2, IRF6, RD3, NEK2 and KCNH1) described in Table 2. Haploinsufficiency score 1 for IRF6 gene is extremely low (HI score 2.02 %), it is therefore most likely that in present case phenotype characteristics (cleft palate, lower bilateral pits and hypodontia) are consisted with the literature descriptions of VWS, and are repercussions of IRF6 haploinsufficiency. Besides, some cases with small deletions in IRF6 gene were described as well. Three cases with a microdeletion (0.1-1 Mb in size) were detected by CMA in a large cohort of patients with clinical diagnosis of VWS [29,30]. Mutation in G0S2 and NEK2 genes were associated with phenotype of VWS [31], but contribution of haploinsufficiency of these genes in our case is questionable due to two facts: I) high HI score for booth genes (G0S2 88,45 and NEK2 25,91), and II) NEK2 is only partly deleted. Mutations in LAMB3 gene causes autosomal recessive skin disorder with blisters on the skin -Junctional Epidermolysis (JEB) [31]. In presented case LAMB3 deletion might have unmasked the recessive allele, as the HI score is high (58.66). Therefore, the phenotype manifestations restricted to acne-prone skin could represent a very mild form of Non-Herlitz type of JEB (MIM 226650). Conformation of Non-Herlitz type of JEB could be done only by direct sequencing of affected LAMB3 gene. Mutations in KCNH1 gene are responsible for two severe  (Table 2.). The HI score for KCNH1 is relatively high (22.53) but in the present case distinct physical phenotype ((wide mouth, broad nasal bridge, hirsutism, broad thumbs and short and tapering fingers) could be due to KCNH1 haploinsufficiency.

Cognitive phenotype 2
Research of cognitive function has shown that group of patients with clinical VWS diagnosis have just a slightly wider distribution of full-scale IQ (84 to 118) compared to controls (96 to 123) [32]. The evaluation of brain morphology in patients with clinical diagnosis of nonsyndromic clefts of the lip and/or palate have pointed to some brain abnormalities. The most severely affected region was the left temporal lobe with decreased gray matter volume [33]. In presented case, brain MRI showed reduced temporal lobe in all volumetric measurements (Additional file 3), but this finding alone cannot sufficiently explain a cognitive deficit.
Five cases have been so far reported to have deletions which included additional genes outside of the IRF6 gene ( Fig. 6) [3,[34][35][36][37]. First case was microscopically detected 1q gross deletion (no high resolution mapping) that most likely extends beyond distal and proximal breakpoint of our case, so it is difficult to discuss molecular basis of cognitive phenotype. Additional two cases were assigned by STR mapping and showed overlapping genomic region (CAMK1G on the proximal end LAMB3, TRAF3IP3, IRG6 and part of SYT14 genes on the distal end) (Fig. 6). Cases studied with STR mapping provide only a limited insight into the genome; therefore the contribution of additional aberrations to the phenotype cannot be excluded.
Only two reported cases were studied by high resolution array technique. Tan et al. [37] reported a patient with de novo 2.3 Mb 1q32.2 microdeletion, presenting orofacial VWS phenotype. He was otherwise healthy individual, followed up closely for the last 20 years and there was no evidence of cognitive deficit [37]. Gene content shared with our case included PLXNA2 and SYT14 (Fig. 6). Since there was no evidence of cognitive deficit in case of Tan et al. [37], we may only discuss the contribution of PLXNA2 and SYT14 genes to the cognitive phenotype in our case. PLXNA2 mediates signals from two semaphorins (SEMA3 and SEMA6) and plays a role in axon guidance (guiding axonal growth cones), invasive growth and cell migration. Although PLXNA2 HI score is 5.01, the lack of cognitive phenotype in case reported by Tan et al. could be due to incomplete penetrance, as knockout mice for SEMA3 and SEMA6 show the neural abnormalities to some extent. It is possible that plexin-semaphorn diversity and brain "plasticity" could prenatally compensate for PLXNA2 haploinsufficiency, meaning that a signal transduction of SEMA3 and SEMA6 could go via other receptors during neurodevelopment. SYT14 is expressed prenatally, has HI score of 18.19, and has a role in prenatal neurotransmission. In case of Tan et al. [37] distal breakpoint is disrupting SYT14, so the haploid state might possibly be sufficient for the normal development. The second case studied with CMA, was a 22-month-old baby presenting with a 2.98 Mb deletion in 1q32.2-q32.3 region, who in addition to VWS, displayed developmental delay and dysmorphism [36]. KCNH1 and RCOR3 genes were the only two deleted genes enriched in human brain and overlapped with present case (Fig. 6). Mutations in aforementioned KCNH1 gene are responsible for two syndromes, the Temple-Baraitser (MIM 611816) and Zimmermann-Laband (MIM 135500) syndrome 1, both being developmental disorders characterized by severe cognitive and physical abnormalities (Table 1.). RCOR3 gene important for restricting the neural features to neuron has HI score of 13.05 and is prenatally expressed. Taking into account the above mentioned facts, KCNH1 and ROCR3 are good candidate genes whose haploinsufficiency may be liable for severe cognitive deficit combined with dysmorphism.   Bold numbers and text are pointing to more than 3 standard deviation deference between control and patient measurements The main players for cognitive deficit (1q32.1 :SRGAP2 and CD55; 3q26.33 :TBL1XR1) In comparison to two cases studied by array, our case has a proximal breakpoint further extended in 1q32.1, and the deletion covers additional genes including SRGAP2 and CD55. SRGAP2 (SLIT-ROBO Rho GTPase activating protein 2, HGNC: 19751). There are four copies of SRGAP2 gene: the parental SRGAP2A and three duplicates (SRGAP2B, SRGAP2C and SRGAP2D) [38]. The first duplication event occurred~3.4 million years ago when only first 9 (out of the 22) exons of parental SRGAP2A were duplicated. This segmental duplication truncated SRGAP2B in its F-BAR domain, which led to the key functional consequences. Second segmental duplication occurred~2.4 million years ago copying SRGAP2B to SRGAP2C. The most recent segmental duplication happened~1 million years ago when segmental duplication of SRGAP2B gave rise to SRGAP2D (Fig. 7). Human-specific three-fold segmental duplication of SRGAP2 gene, evolutionary corresponds to the beginning of neocortical expansion in humans [38]. Studies on SRGAP2 function reveled that it regulates neural migration and neurite outgrowth and branching, and is expressed throughout the developing cortex in proliferative zones (ventricular zone VZ and subventricular zone SVZ) and in postmitotic zone (cortical plate CP) [39,40]. Description of SRGAP2 function and summary of research on loss/gain of function is presented in Table 3. SRGAP2C inhibits ancestral SRGAP2 function during cortical neural migration and displays spine morphology phenotype characteristics of SRGAP2 knockdown resulting in neoteny during dendritic spine maturation [41]. Overview of SRGAP2 paralogs is presented in Table 4. Haploid state of ancestral SRGAP2 and diploid state of regulatory SRGAP2C could potentially explain specific brain MRI findings in our case, that is mainly significantly reduced gyrification by two scenarios: I) SRGAP2C in normal diploid state could hypothetically inhibit all ancestral SRGAP2 expressions with functional consequence, so our case could present a complete knockout phenotype. A knockout mouse for ancestral SRGAP2 shows no abnormalities in cortical lamination, but neuronal migration is extremely fast. In our case the neurons are probably spreading toward their destination either too fast and/or too early or even could be misplaced due to their inability to establish the right speed of neuronal migration (regulated by SRGAP2A-2C interaction). In either case the neurons might fail to form stable or functional synapses leading to reduced gyrification; II) It is also possible that specific brain MR findings presented in our case are due to haploinsufficiency of only ancestral SRGAP2 (HI score 8.32), regardless of SRGAP2A-2C interaction. Mouse knockdown model for ancestral SRGAP2 displays significant decrease in both axonal and dendritic branching with accelerated radial migrationthe characteristics that could potentially explain reduced gyrification. However, only the postmortem brain analyses could verify these hypotheses.
CD55 (CD55 molecule, decay accelerating factor for complement, HGNC: 2665, 1q32) gene encodes glycoprotein that has a physiological role to inhibit the complement cascade, protecting autologous cells and tissues from complement-mediated damage. It was recently demonstrated that certain number of genes belonging to macrophages/immune system (including    paralogs have some additional variants CD55) show differential expression between ages of 3 to 6 months after birth (time of intense overshoottype synaptic formationnumber of synapses reach the peak, and pruning taking place after this peak) [42]. The process of synaptic phagocytosis by microglia which is occurring at the time of the overshoottype synaptic formations could cause damage to normal tissue and mitochondria. The fact that CD55 shows the higher expression at 6 months than at 3 months, may suggests that normal brain tissue is more protected at 6 M [43]. The deletion of CD55 gene in our case could have potential role in the "fine tuning" of synaptic pruning in a negative way. Reduced level of the glycoprotein, due to a deletion, could hypothetically make normal brain tissue more prone to negative side effects of phagocytosis during pruning period. Further functional analysis could shed more light on this process and reveal the neuro-protective genes. TBL1XR1 (transducin (beta)-like 1 X-linked receptor 1, HGNC:29529, 3q26.33) is essential in mediating transcription silencing by unliganded nuclear receptors (NRs) and other regulated transcription factors (TFs) [44][45][46]. Through the recruitment of the specific proteasome complex, TBL1XR1 can act as a transcription activator which mediates the exchange of corepressors for coactivators [47,48]. TBL1XR1 haploinsufficiency was so far described in only three patients, all characterized by facial dysmorphism, speech delay, mild to moderate cognitive deficit, and lack of autistic behaviors [49,50]. Interestingly, mutations in TBL1XR1 gene were recently identified in three patients displaying ASD and severe ID, but without any obvious dysmorphism or recurrent comorbidities [51,52]. HI score for TBL1XR1 is 6.81.

Conclusion
The deletion in our patient appeared to be a very rare event and provides a valuable insight for genotype-phenotype correlations in genetic disorders. The phenotype in presented case points on the importance of looking for the recessive disorders when deletion is detected. The deletion might in turn unmask recessive disorder; even when the HI score is low. Only by combining the GTG-banding, FISH and CMA analyses we were able to fully characterize the structural rearrangement within 1q. Brain MRI is crucial for the patients with cognitive deficit, as it allows better genotype-phenotype correlations and understanding of gene aberration consequences. Finally, the aberrations in SRGAP2 are strong candidates underlying specific brain abnormalities, namely reduced volume of grey matter and reduced gyrification. Further analyses of the cases with SRGAP2 aberrations should be highly informative in this respect.