High-throughput detection of transcriptomic and epitranscriptomic variation and kinetics using MarathonRT

点击翻译按钮获取中文摘要

Project Summary The discovery and characterization of an efficient, ultraprocessive reverse transcriptase (MarathonRT) now makes it possible to develop high-throughput methods for accurate end-to- end sequencing of long RNA transcripts, thereby preserving information content on alternative splicing, editing and modification isoforms while conserving positional linkage information, thereby enabling one to distinguish RNA isoforms in complex mixtures without mapping to a reference genome. This type of technology is essential for deciphering the role of post- transcriptional RNA processing events during control of developmental stage, cell and tissue specificity and regulation of gene expression in higher organisms. It must be sufficiently efficient and accurate to power the long-read sequencing approaches that are used in single- cell RNAseq, particularly when transcript diversification is monitored as a function of time. The first two aims of the proposal are focused on high-throughput detection of RNA modifications (such as 2-O-methyl groups and N7-methyl guanosines). In the first aim, a unique MarathonRT primer extension protocol will be combined with a trained mutational profiling algorithm to recognize the positions and chemical identities of specific RNA modifications, reporting a modification signature that can be recognized at high throughput during long-read sequencing (MRT-ModSeq). In the second aim, MRT-ModSeq will be tested on unknown RNAs, where it will be used to predict sites of modifications on challenging long transcripts and robustness of the predictions will be directly evaluated using mass spectrometry. The second half of the proposal is focused on identification of linked alternative splicing and editing sites on long transcripts within complex cellular mixtures. In aim 3, MarathonRT will be incorporated into a workflow for accurately profiling the relative abundance and processing diversity of the highly complex paralytic (para) gene, which encodes more than 1 million possible processing variants, a subset of which are essential for the voltage-gating of a sodium channel. This sets the stage for Aim 4, in which sensitivity of the workflow must be further optimized and merged with data analysis strategies suitable for time-resolved single cell applications. The resulting method will be tested by monitoring full-length transcriptomic signatures induced by cell stress.