Abstracts

Rationale: There are various causes of epilepsy with mesial temporal lobe epilepsy (MTLE) being a category with diverse causes including focal cortical dysplasia, ischemic, tumors, gliosis, vascular lesions, with hippocampal sclerosis (HS) being the most common. In 2013, the International League Against Epilepsy (ILAE) defined HS subtypes based on patterns of neuronal cell loss within Ammon’s horn. Clinical outcomes have been correlated depending on the HS subtype classification. This also highlights the importance that vulnerable neurons can play in the disease biology. Bulk RNA sequencing has supported vulnerable neuron subtypes exist in epilepsy, however previous studies lack the single cell specificity to understand the effects on individual neuronal populations. Herein we examine new spatial transcriptional technologies in a MTLE-HS case with total sclerosis from a 58-year-old female with a 47-year history of epilepsy which included tonic-clonic seizures lasting 1-2 minutes. She had an amygdalo-hippocampectomy on 2/5/2013 at LAC-USC Medical center and has been seizure free since. Neuropathology diagnosed the case as ILAE type 1 (20x magnification H&E, figure 1B).


Methods: We utilized the 10X Visium spatial transcriptomic platform on FFPE fixed hippocampal tissue. After assuring the tissue had sufficient quality and quantity of RNA (e.g., DV200), the tissue was processed using standard methodologies for the 10X CytAssist, and was then sequenced using the Illumina NovaSeq 6000 platform at the University of Colorado’s Genomics Core. H&E images and raw sequences were co-processed using 10X SpaceRanger pipeline, and then statistically analyzed using the Seurat pipeline. Differential gene expression within tissue subset between a healthy control (HC) and the MTLE-HS case (EP) was analyzed for biological significance using Ingenuity Pathway Analysis. 10X Visium allows the assessment of all genes in a spatially resolved context, we can use gene-level biomarkers such as SNAP25 (Fig. 1C) and MBP (Fig 1D) to identify tissue substructure (Fig 1A), from which we can statistically analyze the effects of MTLE-HS in the various strata to understand how the disease state effects the tissue at histologic resolution.



Results: These data support the loss of neurons in CA2-CA4 and show that the effects of the disease are ubiquitous across the tissue, with the DG having the least effect (Fig 2). Pathway analysis of the differential genes between heathy control and MTLE-HS (Fig 2) illustrate significant changes within canonical pathways (A), disease and functions (B), and upstream regulators (C). The data support the hypothesis that a partial inhibition of MeCP2 activity appears to be the ultimate driver in this case. An ancillary analysis of MeCP2-related gene expression (Fig 3) illustrates that many genes related to MeCP2 are dysregulated and support the hypothesis that the gene responsible for Rett Syndrome, may also affect MTLE-HS


Conclusions: Spatial transcriptomics is a useful technique for neuropathologists to combine morphologic knowledge with gene transcription to identify key areas of dysregulation for disease.


Funding: University of Colorado Anschutz Medical Campus