Mapping of internal monophosphate 5′ ends of Bacillus subtilis messenger RNAs and ribosomal RNAs in wild-type and ribonuclease-mutant strains

The recent findings that the narrow-specificity endoribonuclease RNase III and the 5' exonuclease RNase J1 are not essential in the Gram-positive model organism,Bacillus subtilis, facilitated a global analysis of internal 5' ends that are generated or acted upon by these enzymes. An RNA-Seq protocol known as PARE (Parallel Analysis of RNA Ends) was used to capture 5' monophosphorylated RNA ends in ribonuclease wild-type and mutant strains. Comparison of PARE peaks in strains with RNase III present or absent showed that, in addition to its well-known role in ribosomal (rRNA) processing, many coding sequences and intergenic regions appeared to be direct targets of RNase III. These target sites were, in most cases, not associated with a known antisense RNA. The PARE analysis also revealed an accumulation of 3'-proximal peaks that correlated with the absence of RNase J1, confirming the importance of RNase J1 in degrading RNA fragments that contain the transcription terminator structure. A significant result from the PARE analysis was the discovery of an endonuclease cleavage just 2 nts downstream of the 16S rRNA 3' end. This latter observation begins to answer, at least for B. subtilis, a long-standing question on the exonucleolytic versus endonucleolytic nature of 16S rRNA maturation.