The hidden transcriptome in cerebrospinal fluid and an exploration of its role in parkinson’s disease
In cerebrospinal fluid (CSF) many RNA-seq reads do not map to the genome with traditional genome aligners, and are termed “dark RNA” due to their unknown nature. Many of these reads map to the genome multiple times in different locations (colloquially termed “multi-mappers”). Because transposable elements (TEs) tend to move around via duplication, an emerging hypothesis is that they explain some of these multi-mappers. Extracellular vesicles (EVs) carry various types of genetic cargo (e.g. genes, proteins, RNA, etc.) and, as it is theorized that TEs may also be cargo in EVs, they may be found in various biofluids, including CSF. Prior research has shown that some neurological diseases, such as ALS, lupus, multiple sclerosis, and autism have been associated with transposable elements, usually with HERV (human endogenous retrovirus) and LINE 1 (long interspersed element). However, to our knowledge, Parkinson’s disease (PD) has not been investigated in the context of TEs. Here, we show that transposable elements are highly expressed, especially in the CSF, and that TE expression differs by biological sex and age. For this analysis a new pipeline was developed which aligns samples using STAR; TElocal and TEtranscripts created counts of the number of TEs based on a reference database . Differential expression was run using DESeq2 in R. CSF has notoriously poor genome mapping, and we found that in CSF TEs made up a higher percentage of reads that clearly map to the transcriptome (20%) than genes (15%). Unlike prior comparisons of gene diversity in the CSF and plasma, TEs had a similar distribution by class across the biofluids, with the majority of reads mapping uniquely to LINE, SINE (short interspersed element), and LTR (long terminal repeat). In an analysis of differentially expressed TEs across PD diagnosis, no significant TEs were detected. However, human satellite I (HSATI) was found to be significantly different across sex (lower in females). Further, when only female samples were analyzed, both MER11B and MER101-int were found to be significantly lower in participants with PD diagnosis. Additionally, in an analysis by age, a repeat of (CATTC)n was found to be significantly lower in those under age 65. Overall, these results demonstrate that transposable elements make up a significant part of biofluid samples, and tend to differ by sex, age, and, in females, Parkinson’s disease status.
