Ferber, Matthew J.

AbstractBackgroundWe describe a patient presenting with pachygyria, epilepsy, developmental delay, short stature, failure to thrive, facial dysmorphisms, and multiple osteochondromas.MethodsThe patient underwent extensive genetic testing and analysis in an attempt to diagnose the cause of his condition. Clinical testing included metaphase karyotyping, array comparative genomic hybridization, direct sequencing and multiplex ligation‐dependent probe amplification and trio‐based exome sequencing. Subsequently, research‐based whole transcriptome sequencing was conducted to determine whether it might shed light on the undiagnosed phenotype.ResultsClinical exome sequencing of patient and parent samples revealed a maternally inherited splice‐site variant in the doublecortin (DCX) gene that was classified as likely pathogenic and diagnostic of the patient's neurological phenotype. Clinical array comparative genome hybridization analysis revealed a 16p13.3 deletion that could not be linked to the patient phenotype based on affected genes. Further clinical testing to determine the cause of the patient's multiple osteochondromas was unrevealing despite extensive profiling of the most likely causative genes, EXT1 and EXT2, including mutation screening by direct sequence analysis and multiplex ligation‐dependent probe amplification. Whole transcriptome sequencing identified a SAMD12‐EXT1 fusion transcript that could have resulted from a chromosomal deletion, leading to the loss of EXT1 function. Re‐review of the clinical array comparative genomic hybridization results indicated a possible unreported mosaic deletion affecting the SAMD12 and EXT1 genes that corresponded precisely to the introns predicted to be affected by a fusion‐causing deletion. The existence of the mosaic deletion was subsequently confirmed clinically by an increased density copy number array and orthogonal methodologiesConclusionsWhile mosaic mutations and deletions of EXT1 and EXT2 have been reported in the context of multiple osteochondromas, to our knowledge, this is the first time that transcriptomics technologies have been used to diagnose a patient via fusion transcript analysis in the congenital disease setting.

Wolf‐Hirschhorn syndrome (WHS) is a microdeletion syndrome characterized by distinctive facial features consisting of “Greek warrior helmet” appearance, prenatal and postnatal growth deficiency, developmental disability, and seizures. This disorder is caused by heterozygous deletions on chromosome 4p16.3 often identified by cytogenetic techniques. Many groups have attempted to identify the critical region within this deletion to establish which genes are responsible for WHS. Herein, clinical whole exome sequencing (WES) was performed on a child with developmental delays, mild facial dysmorphisms, short stature, failure to thrive, and microcephaly, and revealed a de novo frameshift variant, c.1676_1679del (p.Arg559Tfs*38), in WHSC1 (NSD2). While WHSC1 falls within the WHS critical region, individuals with only disruption of this gene have only recently been described in the literature. Loss‐of‐function de novo variations in WHSC1 were identified in large developmental delay, autism, diagnostic, and congenital cardiac cohorts, as well as recent case reports, suggesting that de novo loss‐of‐function WHSC1 variants may be related to disease. These findings, along with our patient suggest that loss‐of‐function variation in WHSC1 may lead to a mild form of Wolf‐Hirschhorn syndrome, and also may suggest that the developmental delays, facial dysmorphisms, and short stature seen in WHS may be due to disruption of WHSC1 gene.

Hartsfield syndrome has been recently reported to be associated with mutations in FGFR1 however, to this date; no familial cases have been reported. In this report, we describe two siblings with Hartsfield syndrome and a novel de novo FGFR1 mutation suggesting gonadal mosaicism. The proband presented at our institution at age 6 years with a clinical diagnosis of Hartsfield syndrome and requesting further genetic evaluation. Previous studies included a normal karyotype, oligonucleotide array, and single gene testing for nonsyndromic holoprosencephaly (SHH, SIX3, ZIC2, TGIF). At the age of 6 years, exome sequencing was performed and a de novo novel missense variant was identified in FGFR1 (coding for fibroblast growth factor‐1) on chromosome 8p12: c.1880G>C (p.R627T). Subsequently, a younger sibling was born with the same phenotype (holoprosencephaly, ectrodactyly of bilateral hands and feet and bilateral cleft lip and palate). Targeted sequencing of FGFR1 revealed the identical variant that was previously identified in the proband. To our knowledge this observation is the first documentation of familial recurrence of Hartsfield syndrome. As both parents were negative for the sequence variant in FGFR1 gene by testing peripheral blood samples, this suggests gonadal mosaicism. The frequency of gonadal mosaicism in Hartsfield syndrome is not known however given our case, this possibility should be taken in to consideration for recurrence risk estimation in children of clinically unaffected parents. © 2014 Wiley Periodicals, Inc.

BACKGROUND Next-generation DNA sequencing (NGS) techniques have the potential to revolutionize molecular diagnostics; however, a thorough evaluation of these technologies is necessary to ensure their performance meets or exceeds that of current clinical sequencing methods. METHODS We evaluated the NimbleGen Sequence Capture 385K Human Custom Arrays for enrichment of 22 genes. We sequenced each sample on both the Roche 454 Genome Sequencer FLX (GS-FLX) and the Illumina Genome Analyzer II (GAII) to compare platform performance. RESULTS Although the sequence capture method allowed us to rapidly develop a large number of sequencing assays, we encountered difficulty enriching G+C–rich regions. Although a high proportion of reads consistently mapped outside of the targeted regions, >80% of targeted bases for the GAII and >30% of bases for the GS-FLX were covered by a read depth of ≥20, and >90% of bases for the GAII and >80% of bases for the GS-FLX were covered by a read depth of ≥5. We observed discrepancies among sequence variants identified by the different platforms. CONCLUSIONS Although oligonucleotide arrays are quick and easy to develop, some problematic regions may evade capture, necessitating sequential redesigning for complete optimization. Neither sequencing technology was able to detect every variant identified by Sanger sequencing because of well-known drawbacks of the NGS technologies. The rapidly decreasing error rates and costs of these technologies, however, coupled with advancing bioinformatic capabilities, make them an attractive option for molecular diagnostics in the very near future.