Genomic imbalances in 17p13.3 are mainly associated with neuronal migration disorders. A ~ 1.3 Mb deletion within the 17p13.3 region extending from YWHAE to PAFAH1B1 is sufficient to cause MDS. In addition, recent studies have focused on a condition known as 17p13.3 microduplication syndrome [2, 3, 12]. Since individuals with either condition often exhibit poor prognosis, prenatal diagnosis of these genomic disorders is crucial. To investigate the clinical significance of CNVs involving 17p13.3 with varied sizes and gene content, we retrospectively analyzed the clinical data of eight cases. In addition, we discussed the potential implications of phenotype-associated genes located within these CNVs.
The CNVs involving 17p13.3 contained or overlapped with the MDS region in the eight cases. SNP array provided genetic diagnosis of MDS for cases 4, 6, and 7 and 17p13.3 duplication syndrome for case 1. Of these, case 7 displayed developmental delay (DD), congenital lissencephaly, and softening of the brain. In case 6, MDS may be the underlying cause of spontaneous abortion. In contrast, case 4 had no obvious clinical findings associated with MDS on the first-trimester ultrasound. MDS features, such as polyhydramnios, IUGR, ventriculomegaly, lissencephaly, and corpus callosum dysgenesis/agenesis, were often found in the second and third trimesters . Apart from this, case 1 had no abnormalities in brain structure and no IUGR, both of which were previously described in the 17p13.3 duplication syndrome .
Apart from the cases with definite syndromes mentioned above, we focused on the clinical significance of CNVs involving 17p13.3 that overlapped with the MDS region. Currently, MDS is regarded as a contiguous gene deletion syndrome, and PAFAH1B1, YWHAE, and CRK in this region are of prime interest. PAFAH1B1 is considered to cause an isolated lissencephaly sequence and contribute to MDS [3, 13]. YWHAE encodes 14-3-3ε, a phosphoserine/threonine-binding protein that plays a role in cortical development [14, 15]. In addition, YWHAE and TUSC5 appear to contribute to craniofacial dysmorphism , while CRK functions in cell proliferation, differentiation, migration, and axonal growth and is a typical candidate gene for growth restriction. Moreover, CRK appears to be related to limb abnormalities and craniofacial dysmorphism [2, 16, 17].
Case 5 harbored a de novo 2.1 Mb deletion containing YWHAE and CRK but not PAFAH1B1 and showed multiple abnormalities in the prenatal ultrasound. As previously reported, neurodevelopmental delay, growth retardation, craniofacial dysmorphisms, mild structural brain abnormalities, and seizures were observed in 17p13.3 deletions, including YWHAE and CRK but not PAFAH1B1 [14, 18, 19]. Similarly, case 8 carried a 1.6 Mb deletion in the 17p13.3 region encompassing YWHAE and CRK, along with a 4.0 Mb duplication in the 17q25.3 region. Previous studies have reported that 17q25.3 duplication was related to DD, growth retardation, and multiple congenital anomalies [20, 21]. Therefore, we hypothesized that deletions and duplications may contribute to the clinical phenotype of this patient. Overall, the 17p13.3 microdeletion including YWHAE and CRK but not PAFAH1B1 could be classified as pathogenic.
In case 2, the 17p13.3 duplication that included YWHAE and CRK was inherited from unaffected mother, and the ultrasound revealed a transient increase in NT. The fetus was then continued developing and had a good presentation at 15 months old. The duplication in case 2 can be identified as class I of 17p13.3 microduplication syndrome . The individuals in this category, including three patients from Bruno et al.  (cases 9, 11 and 12) and four from Bi et al.  (subjects 1–4), had autism manifestations, behavioral symptoms, learning disabilities, subtle dysmorphic facial features, subtle hand/foot malformations, and a tendency to postnatal overgrowth, among other disorders. Another study from Curry et al.  described eight patients in Group 1 17p13.3 microduplications who presented with developmental, behavioral and brain abnormalities, and rare variant phenotypes such as cleft palate and split hand/foot with long bone deficiency. Regarding inheritance of these 15 patients, six were de novo, six were inherited from an unaffected parent, and three were unknown. The duplication inherited from a normal parent may be owing to reduced penetrance and variable expressivity. In addition, we searched the DECIPHER database and found 12 duplications involving YWHAE and CRK, but not PAFAH1B1, which was approximately 300 kb. Eight out of 12 cases lacked parental analysis and showed a wide spectrum of phenotypes not characterized by autism. Therefore, the extent of contribution of the variants to their phenotypes cannot be ascertained.
Furthermore, the likelihood of a single-gene mutation causing propositus manifestations cannot be ruled out. The two-hit model proposed by Girirajan et al.  suggests that a secondary disruptive event (another CNV, a point mutation or environment factors) could result in more severe clinical manifestations in neurodevelopmental diseases. Likewise, Tolezano et al.  investigated the genetic factors that contribute to variable expressivity of class I 17p13.3 microduplications, providing new evidence regarding the contribution of RORA and DIP2B to neurocognitive deficits such as autism and intellectual disability, respectively. Moreover, in group I 17p13.3 microduplication, Curry et al.  reported that disruption of ABR and duplication of BHLHA9 were associated with clefts and split hand/foot with long bone deficiency phenotypes, respectively. Capra et al.  reported that a boy carrying a maternally inherited 329.5-kb 17p13.3 duplication, including BHLHA9, YWHAE, and CRK, presented with mild dysmorphic phenotype, autism, and mental retardation, while his mother was affected by a bipolar and borderline disorder and was addicted to alcohol. It can be seen that phenotypic heterogeneity existed in the mother and her child. Another report  described two patients manifesting distinctive features (patient 1, primary hypothyroidism; patient 2, bilateral cryptorchidism) that were not previously described in the duplication 17p13.3 spectrum. Whether these rare manifestations observed in the two patients were caused by a two-hit event or not is not known. Overall, considering 17p13.3 microduplication showing reduced penetrance, variable expressivity, and lack of a clear pathogenic mechanism, the clinical significance of the microduplication encompassing only YWHAE and CRK, but not PAFAH1B1, requires further investigation.
Interestingly, case 3 also carried a 74.2 Mb mosaic duplication of approximately 3.5 on chromosome 17p13.2q25.3 and a 1.0 Mb deletion in the 17q terminus, in addition to deletion of the MDS region. The SNP data were consistent with that some cells have ring 17 while others have dicentric or interlock ring 17. Given the dosage sensitivity of genes and regions involved in the three CNVs, case 3 may show compound manifestations of these known genomic disorders, such as MDS, Potocki–Lupski syndrome (MIM:610883) , Charcot–Marie–Tooth disease, type 1A (CMT1A, MIM:118220) [28, 29], 17q11.2 duplication syndrome, 1.4-Mb (618874) [30, 31] and 17q12 duplication syndrome (MIM:614526) [32, 33]. Notably, the karyotype of case 3 is similar to previously reported “ring chromosome 17” syndrome , the manifestations of which include DD, seizures, short statures, microcephaly, and muscular hypotonia, among others. In contrast, ventricular septal defect and dysplasia of the corpus callosum were observed on ultrasound at 26+ weeks of gestation while other features could not be detected in the prenatal ultrasound.
A total of eight cases were detected with CNVs involving 17p13.3 in our report. However, the size and number of genes involved differed considerably, particularly when mixed deletion and duplication at chromosome 17 terminations and ring chromosome 17 were observed. It is estimated that approximately 80% of MDS cases are de novo, and approximately 20% of the conditions arise from balanced chromosomal rearrangement in parents 4. To our knowledge, no data have been presented from large-scale case studies to calculate the frequencies of 17p13.3 microduplication, mixed deletion/duplication on chromosome 17, and ring chromosome 17 to date, indicating the need for further study.
Our study has some limitations. First, its single-center nature and small number of patients resulted in fewer detectable 17p13.3 CNVs. Second, owing to the lack of functional experiments, whether the genes inside the duplication in case 2 are overexpressed is uncertain. Third, despite its notable advantages in CNV detection, SNP array cannot detect point mutations associated with neurodevelopmental disorders. Recently, with the development of next-generation sequencing, whole-exome or whole-genome sequencing may provide clinically relevant information in cases where SNP array fails to determine the underlying cause of a neurodevelopmental disorder.