Our collaboration with Sam Schilit, Cynthia Morton, and the rest of the DGAP team on a case of congenital hearing loss with a de novo balanced translocation, dubbed DGAP242, was published this week in European Journal of Human Genetics. The article can be viewed here. Congrats to Sam & co. on the paper!
The genomics for DGAP242 took a slightly different route than most of the few hundred others we've done to date through DGAP. In a typical DGAP case, a patient is referred to our clinical coordinating center due a congenital anomaly and concomitant abnormal g-banded karyotype confirmed to be largely balanced by chromosmal microarray (usually aCGH). Since the resolution of microscopy-based cytogenetic methods like karyotyping are limited to ~3-5Mb, we almost always turn next to jumping library whole-genome sequencing to refine the karyotype to approximately kilobase resolution. At this scale, we can (a) clearly identify which gene--if any--was disrupted at a given breakpoint and (b) PCR & Sanger the breakpoint to achieve nucleotide resolution, both of which would be impossible from the karyotype alone. This approach has been fruitful not only in discovery of associations between diseases and genes but also clinical diagnostic applications.
For DGAP242, we did indeed take the tried-and-true DGAP jumping library approach, but we supplemented the jumping library with a newer method called targeted locus amplification (TLA). In TLA, DNA is crosslinked with formaldehyde, enzymatically digested, and religated, similar to how 4C and HiC chromatin capture libraries are prepared. However, unlike 4C/HiC, these proximity-ligated, circularized gDNA fragments are amplified specifically for a desired ("targeted") locus, resulting in an accumulation of reads specifically at the desired locus, or in this case the breakpoints of the reciprocal translocation. The resulting amplicons are then fragmented, barcoded, and passed along for standard NGS. TLA has some advantages over jumping libraries and traditional targeted capture resequencing by providing deep nucleotide coverage at modest sequencing cost without requiring intimate knowledge of the exact coordinates of the target locus, but conversely does not generally provide uniform coverage and offers no information beyond the target locus.
By this combination of genomics approaches, we narrowed the breakpoints to two genes, KIAA0825 and ESSRG. While little is known about the function and disease relevance of KIAA0825, estrogen receptor-related genes like ESSRG have previously been implicated in several types of hearing loss in humans and mouse models. Further, ESSRG is differentially expressed in the inner ear, suggesting a possibly relevant tissue-specific activity. With this evidence in hand, we felt confident concluding that ESSRG was the most likely driver of DGAP242's hearing loss phenotype, although roles of other genes (specifically KIAA0825) and the patient's overall genetic background could not be discounted. The diversity of DGAP242s phenotypes, including not only hearing loss but also hypotonia, developmental delay, and abnormal craniofacial features, hints at the possible contribution of genes beyond ESSRG alone, although ESSRG seemed to convincingly explain the referred primary phenotype in this case.
Give the paper a read and let us know if you have any questions or comments.
Citation: SLP Schilit et al., Estrogen-related receptor gamma implicated in a phenotype including hearing loss and mild developmental delay. Eur. J. Hum. Genet., online ahead of print (July 6 2016). DOI: 10.1038/ejhg.2016.64.