Abstracts

Rationale:
Tuberous sclerosis complex (TSC) is a neurodevelopmental disease characterised by the presence of tubers in many organs, including the brain. Patients with TSC often present with intractable focal epilepsy. In such patients, surgical resection may provide a good outcome. Although the causative genes have been identified as TSC1 and TSC2, ongoing work is still characterising the pathogenesis of TSC. In particular, recent works implicated the cellular and metabolic signalling, mammalian target of rapamycin (mTOR), in TSC. However, despite emerging evidence implicating dysfunction in metabolic signalling in TSC, there is a lack of understanding of what metabolic pathway is affected. To address this, we conducted an unbiased transcriptomics - metabolomics study to identify affected pathways.
Method:
Fresh-frozen surgically-resected epileptogenic tissue was collected from 9 paediatric patients with TSC. Control tissue was obtained from 10 samples of perilesional resected areas in 1 patient with Rasmussen’s encephalitis and 9 with focal cortical dysplasia. RNA was extracted and sequenced on Illumina HiSeq 2x150bp, single index platform with polyA selection library prep. Standard HISAT2- FeatureCounts – DESeq2 pipeline was conducted to generate differentially expressed genes (DEGs). DEGs were defined as genes having adjusted p-value< 0.05. For the untargeted metabolomics, frozen tissues were homogenised and extracted in 70% methanol solution. Metabolites were extracted and ran on a ZIC-pHILIC column on a Q-Exactive Orbitrap LC/MS. Metabolomics data were analysed using CompoundDiscoverer and TraceFinder to generate differentially expressed metabolites (DEMs). DEMs were defined as metabolites having adjusted p-value< 0.05 and a log-2fold change threshold of 1. Results7570 DEGs and 67 DEMs were identified from the transcriptomics and metabolomics analysis. Among the DEMs, lysine derivatives were significantly affected including increased amino-adipate and N-acetyl-L-2-amino adipic acid as well as decreased lysine and lysine-leucine dipeptide. Interestingly, changes in lysine metabolites were accompanied by associated changes in transcriptomics. Such changes included upregulation of AASS and DHTKD1; genes encoding enzymes involved in lysine degradation. Furthermore, we identified changes in genes downstream of lysine degradation such as upregulation of HSD17B10 and HADH; genes encoding enzymes linking lysine degradation with generation of aceto-acetyl-coA, a source of acetyl-coA and ketone bodies. Finally, genes involved in ketogenesis such as BDH2 and HMGCL were also upregulated.