Non-coding RNA profiling by high throughput sequencing Expression profiling by high throughput sequencing
Transposable elements (TEs) are dynamically expressed at high levels in multiple human tissues, but the function of TE-derived transcripts remains largely unknown. In this study, we identify numerous TE-derived microRNAs (miRNAs) by conducting Argonaute2 RNA immunoprecipitation followed by small RNA sequencing (AGO2 RIP-seq) on human brain tissue. Many of these miRNAs originated from LINE-2 (L2) elements, which entered the human genome around 100-300 million years ago. We found that L2-miRNAs derive from the 3’ end of the L2 consensus sequence and thus share very similar sequences, indicating that L2-miRNAs could target transcripts with L2s in their 3’UTR. In line with this, we found that many protein-coding genes carry fragments of L2-derived sequences in their 3’UTR: these sequences serve as target sites for L2-miRNAs. L2-miRNAs and their targets were generally ubiquitously expressed at low levels in multiple human tissues, suggesting a role for this network in buffering transcriptional levels of housekeeping genes. Interestingly, we also found evidence that this network is perturbed in glioblastoma. In summary, our findings uncover a TE-based post-transcriptional network that shapes transcriptional regulation in human cells.
Small and total RNA sequencing data from AGO-RIPseq experiments on human glioblastoma and cortex tissue, mouse brain tissue and ESCs-derived neurons.