30S-seq redefines the bacterial Ribosome Binding Site

The translation initiation step is rate limiting for the efficiency of gene expression in all organisms. However, the mechanism of ribosome recruitment to mRNA start sites strikingly differs between eukaryotes and prokaryotes. The eukaryotic small (40S) ribosomal subunit binds 5 end caps and scans for the start codon while the bacterial small (30S) subunit directly binds to the Shine-Dalgarno (SD) motif close to the initiation site. Pioneer studies have shown rare 30S loading events further upstream within 5 untranslated regions (5UTRs), at ribosome standby sites1-3. Together with the frequent occurrence of long bacterial mRNA 5UTRs and degenerated SD sequences, this indicates that the 30S subunit might bind upstream of the SD more commonly than currently thought. We therefore developed 30S-seq to map 30S-mRNA interactions in a bacterial transcriptome (Escherichia coli), inspired by translation complex profile sequencing (TCP-seq) previously used in eukaryotes4,5. Our results provide new and unsuspected insights into the behaviour of 30S and 70S complexes during the canonical translation initiation process. Notably, 30S subunits are recruited upstream of the start codon, primed to receive the SD released by the departing 70S ribosome. Remarkably, we also find hundreds of non-canonical 30S binding sites within mRNA 5UTRs, sometimes over 100 nucleotides upstream of the start region. We validated several of these upstream ribosome binding sites, and demonstrated their strong impact on gene expression. Thus, even in bacteria, ribosomes frequently bind mRNAs outside of the start region to initiate translation, challenging the classic ribosome binding site model.