The lacZ mRNA can be stabilised by the T7 late mRNA leader in E coli.

Previous work from this laboratory has shown that T7 RNA polymerase outpaces ribosomes in vivo, generating naked mRNA stretches which may be nuclease-sensitive. In particular, lacZ transcripts synthesised this way are highly unstable and yield little beta-galactosidase. We have argued that most of these transcripts are prematurely inactivated via an RNase E cleavage that occurs ahead of the leading ribosome, whereas a few escape this initial cleavage and are translated normally. Presumably, these rescued transcripts are later inactivated non-nucleolytically and subsequently scavenged by a process partially controlled by RNase E, as for the natural lacZ mRNA. In contrast, despite being synthesised by T7 RNA polymerase, T7 late transcripts are stable. The 5' regions of several of these transcripts, exemplified by the gene 10 mRNA, harbour hairpin structures which may act as barriers against RNase E action. To test whether these structures are indeed 5' stabilisers, we replaced the lacZ leader sequence by the corresponding region from T7 gene10. This replacement yielded a ca 2.9-fold increase in beta-galactosidase yield per transcript. This increase vanished in the presence of the rne-50 mutation which inactivates RNase E, and therefore it reflects a protection of the transcript against RNase E-dependent inactivation. Yet, the leader replacement did not stabilise the transcript chemically. We propose that this replacement inhibits the initial cleavage step but somehow facilitates the subsequent scavenging process.