High-resolution multimodal profiling of human epileptic brain activity via explanted depth electrodes
Anuj Kumar Dwivedi, Arun Mahesh, Albert Sanfeliu, Julian Larkin, Rebecca A. Siwicki, Kieron J. Sweeney, Donncha F. O’Brien, Peter Widdess-Walsh, Simone Picelli, David C. Henshall, Vijay K. Tiwari
Anuj Kumar Dwivedi, Arun Mahesh, Albert Sanfeliu, Julian Larkin, Rebecca A. Siwicki, Kieron J. Sweeney, Donncha F. O’Brien, Peter Widdess-Walsh, Simone Picelli, David C. Henshall, Vijay K. Tiwari
View: Text | PDF
Research Article Neuroscience

High-resolution multimodal profiling of human epileptic brain activity via explanted depth electrodes

Abstract

The availability and integration of electrophysiological and molecular data from the living brain is critical in understanding and diagnosing complex human disease. Intracranial stereo electroencephalography (SEEG) electrodes used for identifying the seizure focus in patients with epilepsy could enable the integration of such multimodal data. Here, we report multimodal profiling of epileptic brain activity via explanted depth electrodes (MoPEDE), a method that recovers extensive protein-coding transcripts, including cell type markers, DNA methylation, and short variant profiles from explanted SEEG electrodes matched with electrophysiological and radiological data allowing for high-resolution reconstructions of brain structure and function. We found gene expression gradients that corresponded with the neurophysiology-assigned epileptogenicity index but also outlier molecular fingerprints in some electrodes, potentially indicating seizure generation or propagation zones not detected during electroclinical assessments. Additionally, we identified DNA methylation profiles indicative of transcriptionally permissive or restrictive chromatin states and SEEG-adherent differentially expressed and methylated genes not previously associated with epilepsy. Together, these findings validate that RNA profiles and genome-wide epigenetic data from explanted SEEG electrodes offer high-resolution surrogate molecular landscapes of brain activity. The MoPEDE approach has the potential to enhance diagnostic decisions and deepen our understanding of epileptogenic network processes in the human brain.

Authors

Anuj Kumar Dwivedi, Arun Mahesh, Albert Sanfeliu, Julian Larkin, Rebecca A. Siwicki, Kieron J. Sweeney, Donncha F. O’Brien, Peter Widdess-Walsh, Simone Picelli, David C. Henshall, Vijay K. Tiwari

×

Figure 3

Single-source high throughput multi-omic profiling of EEG electrodes in epilepsy.

Options: View larger image (or click on image) Download as PowerPoint
Single-source high throughput multi-omic profiling of EEG electrodes in ...
(A) Selection of EEG electrodes: EEG electrodes were selected from SOZ, PZ, and NIZ. The attached nucleic acid material was extracted from each electrode, and whole transcriptome and methylome data were generated. (B) Seven electrodes explanted from the brain of a patient with FCD were used. On average, 100 million and 85 million reads were sequenced for the methylome and transcriptome, respectively. Notably, 91 million and 74 million methylome and transcriptome reads, respectively, mapped to the human genome. Similarly, 9 electrodes explanted from the patient with TLE were used. An average of 100 million and 95 million reads were sequenced for the methylome and transcriptome, respectively. Prominently, an average of 95 million and 87 million methylome and transcriptome reads, respectively, mapped to the human genome. Additionally, 8 electrodes explanted from the patient with RE were used. On average, 117 million and 101 million reads were sequenced for the methylome and transcriptome, respectively. Importantly, an average of 105 million and 92 million methylome and transcriptome reads, respectively, mapped to the human genome. (C) Genome-wide transcriptome and methylome density. The transcriptome and methylome density of each electrode from all 3 patients were plotted across gene body and flanking regions (3 kb). A consistent negative correlation between methylation and transcription was observed across all electrodes.