EPND-03. RNA-SEQ GENE EXPRESSION ANALYSIS OF 106 FORMALIN-FIXED PARAFFIN-EMBEDDED PAEDIATRIC EPENDYMOMAS

Autor: Timothy Ritzmann, Hazel Rogers, Andrew M Donson, Rebecca Chapman, Lisa Storer, Jennifer Ward, Alex Virasami, Thomas S Jacques, Simon ML Paine, Nicholas Foreman, Richard G Grundy
Rok vydání: 2017
Předmět:
Zdroj: Neuro-Oncology. 19:iv15-iv15
ISSN: 1523-5866
1522-8517
Popis: BACKGROUND: Although ependymoma is the second most common malignant childhood brain tumour, obtaining frozen samples from extended cohorts is challenging. A large proportion of research tissue is stored in formalin-fixed, paraffin-embedded (FFPE) blocks, some many decades old. Use of these for RNA sequencing would facilitate biological studies of this enigmatic tumour. AIMS: 106 historical ependymomas were analysed to investigate the feasibility and challenges of using FFPE samples for RNA-seq and whether previously described gene expression patterns from frozen material would be replicated. METHODS: Ethics approval and consent were obtained. Pathologist confirmed ependymoma samples were collected from UK centres. 14 cases had matched frozen samples. RNA was extracted using the Qiagen FFPE Allprep kit and 1000ng per sample underwent whole transcriptome library preparation. Sequencing had a target depth of 50 million paired-end reads. Following quality control, samples were aligned using TopHat2. Differential expression analysis (DEA) was performed using EdgeR. RESULTS: Libraries were produced from all samples. The proportion of FFPE reads aligning to the genome was less than for frozen equivalents (54.2%) and showed higher inter-sample variation. Interestingly, samples with lower alignment to the genome had more reads aligning to bacterial sequences, suggesting potential contamination of the blocks. FFPE had more adapter read through when compared to the frozen samples. Clustering demonstrated 4 groups; poor quality samples, supratentorial tumours and two posterior fossa (PF) tumour groups. DEA of the PF groups showed good concordance with PFA and PFB gene expression patterns. CONCLUSIONS: RNA sequencing of historical FFPE ependymomas replicates subgroups. The benefit of investigating a wider range of samples has to be weighed against some compromise on quality. Researchers must be aware of the potential for bacterial contamination. Greater sequencing depth and more samples in addition to care over storage may help ameliorate quality issues.
Databáze: OpenAIRE