Long non-coding RNA normalisers in human brain tissue.

Abstract

The family of long non-coding RNA (lncRNA) is of increasing scientific interest as there isemerging evidence, that lncRNAs are of essential importance for transcriptional and translational control, genomic imprinting and regulation of normal development as well as neuronal plasticity. As the generation of reliable expression profiles requires adequate normalisers, it is of fundamental importance to determine suitable references for lncRNA studies. However, to date no systematic analysis of potentiallncRNA normalisers has been performed on human postmortem brain tissuesamples. In this study, we investigated three different brain regions (cortex, white matter, and cerebellum) of human postmortem tissue and analysed the expression stability of 90 lncRNAs. Bioinformatical analysis was performed to identify stably expressed lncRNAs. Subsequently, lncRNAs were classified according to their stability values using the NormFinder algorithm. We identified 30 suitablenormalisers in cortex, 22 in white matter, and 41 in cerebellum. In addition, there were 13 suitable normalisers for studies comparing cortex and white matter, 25 for studiescomparing cortex and cerebellum and 7 for studiescomparing white matter and cerebellum. 5 lncRNAs (LUST, IGF2AS (family), 7SK, HOXA6as, NDM29) showed stable expression in all investigated brain regions. A subsequent analysis of the influence of postmortem intervals (PMI) on expression of lncRNAs revealed that expression levelsof the newly identified5 universal lncRNAnormalisers are stable within PMI of up to 27h. Thus, these 5lncRNAs may be applicable as references for accurate normalisation of lncRNA profiling in multiple brain regions during long PMI, enabling the generation ofhighly reproducible datasets in lncRNA studies of the human brain.