Experimental characterization of Cis-acting elements important for translation and transcription in Halophilic Archaea

Summary The basal transcription apparatus of archaea is well characterized. However, much less is known about the mechanisms of transcription termination and translation initation. Recently, experimental determination of the 5´-ends of ten transcripts from Pyrobaculum aerophilum revealed that these are devoid of a 5´-UTR. Bioinformatic analysis indicated that many transcripts of other archaeal species might also be leaderless. The´-ends and 3´-ends of 40 transcripts of two haloarchaeal species, Halobacterium salinarum and Haloferax volcanii, have been determined. They were used to characterize the lengths of 5´-UTRs and 3´-UTRs and to deduce consensus sequence-elements for transcription and translation. The experimental approach was complemented with a bioinformatics analysis of the H. salinarum genome sequence. Furthermore, the influence of selected 5´-UTRs and 3´-UTRs on transcript stability and translational efficiency in vivo was characterized using a newly established reporter gene system, gene fusions, and real-time PCR. Consensus sequences for basal promoter elements could be refined and a novel element was discovered. A consensus motif probably important for transcriptional termination was established. All 40 haloarchaeal transcripts analyzed had a 3´-UTR (average size 57 nt), and their 3´-ends were not posttranscriptionally modified. Experimental data and genome analyses revealed that the majority of haloarchaeal transcripts are leaderless, indicating that this is the predominant mode for translation initiation in haloarchaea. Surprisingly, the 5´-UTRs of most leadered transcripts did not contain a Shine-Dalgarno (SD) sequence. A genome analysis indicated that less than 10% of all genes are preceded by a SD sequence and even most proximal genes in operons lack a SD sequence. Seven different leadered transcripts devoid of a SD sequence were efficiently translated in vivo, including artificial 5´-UTRs of random sequences. Thus, an interaction of the 5´-UTRs of these leadered transcripts with the 16S rRNA could be excluded. Taken together, either a scanning mechanism similar to the mechanism of translation initiation operating in eukaryotes or a novel mechanism must operate on most leadered haloarchaeal transcripts. Author Summary Expression of the information encoded in the genome of an organism into its phenotype involves transcription of the DNA into messenger RNAs and translation of mRNAs into proteins. The textbook view is that an mRNA consists of an untranslated region (5´-UTR), an open reading frame encoding the protein, and another untranslated region (3´-UTR). We have determined the 5´-ends and the 3´-ends of 40 mRNAs of two haloarchaeal species and used this dataset to gain information about nucleotide elements important for transcription and translation. Two thirds of the mRNAs were devoid of a 5´-UTR, and therefore the major pathway for translation initiation in haloarchaea involves so-called leaderless transcripts. Very unexpectedly, most leadered mRNAs were found to be devoid of a sequence motif believed to be essential for translation initiation in bacteria and archaea (Shine-Dalgarno sequence). A bioinformatic genome analysis revealed that less than 10% of the genes contain a Shine-Dalgarno sequence. mRNAs lacking this motif were efficiently translated in vivo, including mRNAs with artificial 5´-UTRs of total random sequence. Thus, translation initiation on these mRNAs either involves a scanning mechanism similar to the mechanism operating in eukaryotes or a totally novel mechanism operating at least in haloarchaea