Full-length RNA profiling reveals pervasive bidirectional transcription terminators in bacteria

The complexity and intricacy of prokaryotic transcriptomes—once deemed well understood and simple—are increasingly being discovered and appreciated. The ability to determine full-length sequences of all transcripts in the cell is essential for understanding the gene regulatory repertoire of a species. Standard RNA-seq requires fragmentation of RNA for short-read sequencing, which automatically decouples the 5' sequence of a RNA molecule from its 3' sequence. Intramolecular ligation of 5' and 3' ends represents a potential strategy for simultaneously capturing the sequences of both termini. However, counterintuitively, it is more challenging to circularize prokaryotic transcripts than eukaryotic ones due to a lack of poly(A) tails. Consequently, a method capable of comprehensively profiling full-length transcripts in prokaryotes is still urgently needed. In this work, we developed a workflow, termed SEnd-seq, that simultaneously reads both ends of all bacterial RNAs—including small non-coding RNAs—with single-nucleotide resolution. This method enabled us for the first time to determine the correlated occurrence of transcription start sites (TSS) and termination sites (TTS) across the whole transcriptome, and achieve an unprecedented map of its operon architecture. Using E. coli as a model system, we identified thousands of new TSS and TTS, many of which were found to be condition dependent. These results significantly expand our knowledge of the regulatory elements existing in the E. coli genome and improve our understanding of many aspects of bacterial RNA metabolism. We developed a high-throughput and unbiased method (SEnd-seq) to prepare the RNA-seq library, which could be used to analyze the E.coli transcriptome. We also used RNA-seq and ChIP-seq to verify some results.