Authors :
Presenting Author: Dulcie Lai, PharmD, PhD – University of North Carolina
Meethila Gade, MPH – University of North Carolina; Brenda Porter, MD, PhD – Professor, Stanford University; Erin Heinzen, PharmD, PhD – Associate Professor, University of North Carolina
Rationale:
The SLC35A2 gene encodes a UDP-galactose transporter essential for cellular glycosylation. Germline genetic variants in SLC35A2 have been implicated in one type of congenital disorder of glycosylation associated with intractable seizures and a rare X-linked developmental and epileptic encephalopathy. Recently, somatic, loss-of-function SLC35A2 variants have been implicated in drug resistant neocortical epilepsy and in mild malformation of cortical development (MCD) with oligodendroglial hyperplasia in epilepsy (MOGHE); an emerging histopathological classification of MCD associated with drug resistant focal epilepsy. In this study, we leverage human brain tissue resected from drug resistant focal epilepsy harboring a pathogenic SLC35A2 variant in conjunction with SoMoSeq (Somatic Mosaic tissue Sequencing) to determine cell type specific variant burden.
Methods:
We obtained resected human brain tissue from a male with drug resistant focal epilepsy harboring a pathogenic SLC35A2 variant (c.C435A:p.Y145X) and an age-matched pediatric stroke case (control). Droplet digital PCR was used to validate and quantify variant allele fraction (VAF) in bulk tissue. Total nuclei were isolated using density gradient centrifugation and separated into neuronal (NeuN+) and non-neuronal (NeuN-) populations using fluorescence-activated nuclear sorting. Forty-six NeuN+ and 46 NeuN- single nuclei were sorted into each well of a 96-well plate in addition to two mini-bulk (n=50 nuclei) and two blank wells. A total of twenty four 96-well plates were collected (n=1104 NeuN+ nuclei and n=1104 NeuN- nuclei) and processed using SoMoSeq, a method optimized in our lab that enables the parallel separation of DNA and RNA for genotyping and transcriptomic analyses, respectively. All SLC35A2 variant harboring cells (n=91), matched variant negative cells (n=213), and stroke control (n=37) were selected for library preparation (Nextera XT DNA Library Kit) and sequenced on the Illumina NovaSeq6000 system to a targeted depth of 8-10 million reads per cell.Results:
The VAF in bulk tissue was determined as ~10% using ddPCR. This VAF was concordant with genotyping data in single nuclei with 110 (8%) variant positive nuclei of all genotyped wells (n=1351). Interestingly, variant positive nuclei are more prominent in NeuN- (13.5%) compared to NeuN+ (3%) populations suggesting that the variant is enriched in non-neuronal cell types.
Conclusions:
The application of SoMoSeq to pathogenic, somatic mosaic tissue provides a powerful tool to quantify cell type specific variant burden. Using brain tissue resected from a drug resistant focal epilepsy case harboring a pathogenic SLC35A2 variant, we have identified that eight percent of total nuclei harbor the SLC35A2 variant. Importantly, the variant is present in both neuronal and non-neuronal cell types, suggesting the variant arose from a common precursor that gives rise to both cell types. We are currently analyzing transcriptomic data to robustly classify the specific cell types harboring the SLC35A2 variants.
Funding: This work is funded by 1R01NS115017 and 1R01NS094596.