In recent years, sRNAs (small non-coding RNAs) have been found to be abundant in eukaryotes and bacteria and have been recognized as a novel class of gene expression regulators. In contrast, much less is known about sRNAs in archaea, except for snoRNAs (small nucleolar RNAs) that are involved in the modification of bases in stable RNAs. Therefore bioinformatic and experimental RNomics approaches were undertaken to search for the presence of sRNAs in the model archaeon Haloferax volcanii, resulting in more than 150 putative sRNA genes being identified. Northern blot analyses were used to study (differential) expression of sRNA genes. Several chromosomal deletion mutants of sRNA genes were generated and compared with the wild-type. It turned out that two sRNAs are essential for growth at low salt concentrations and high temperatures respectively, and one is involved in the regulation of carbon metabolism. Taken together, it could be shown that sRNAs are as abundant in H. volcanii as they are in well-studied bacterial species and that they fulfil important biological roles under specific conditions.
sRNAs (small non-coding RNAs) in eukaryotes and bacteria
In the last few years, it has been revealed that the number of RNAs in all three domains of life is much higher than previously anticipated. In particular, sRNAs have been recognized as a novel class of gene expression regulators. In eukaryotes, miRNAs (microRNAs), siRNAs (small interfering RNAs) and piRNAs (piwi-interacting RNAs) represent three classes of very small RNAs (approx. 20–25 nt) that are involved in the regulation of translational efficiencies and biological half-lives of their target RNAs. Eukaryotes harbour several thousand sRNAs, and it has been estimated that up to 30% of eukaryotic transcripts are regulated by sRNAs. sRNAs play essential roles in development and differentiation, and an increasing number of human diseases are found to be related to the malfunction of sRNAs (reviewed in [1–5]). sRNAs in bacteria are somewhat larger and have sizes from approx. 50 nt to several hundred nucleotides. It has been estimated that Escherichia coli encodes approx. 200–300 sRNAs, corresponding to roughly 5% of all genes. However, the biological functions of only approx. 20 sRNAs have been characterized experimentally. Typically, bacterial sRNAs can hybridize to the 5′-region of their target transcripts, involving the Shine–Dalgarno sequence and/or the translation initiation codon. It has become clear that the functions of bacterial sRNAs can be diverse: they can down-regulate as well as up-regulate the translational efficiencies and they can interact with one specific target or act pleiotropically and regulate a variety of genes (reviewed in [6–8]). A database for non-coding RNAs summarizes the sRNAs known so far .
snoRNAs (small nucleolar RNAs) and other sRNAs in archaea
Much less is known about the existence and roles of sRNAs in the third domain of life, the archaea. Several years ago, it was discovered that Sulfolobus solfataricus harbours a specific class of sRNAs, i.e. the so-called snoRNAs that are involved in modification of rRNAs and tRNAs . Two classes of guide RNAs have been characterized that are involved in 2′-O-methylation and in pseudouridine formation, and in vitro systems for both activities have been established. The structures of several protein complexes with and without snoRNAs have been determined [11–16].
Using an RNomics approach, the existence of additional sRNAs has been shown for two archaeal species, i.e. S. solfataricus and Archaeglobus fulgidus [17,18]. However, the biological functions of these archaeal sRNAs remained unknown. Therefore we decided to use Haloferax volcanii to identify archaeal sRNAs and study their biological functions. H. volcanii is an archaeal model species because its genome is known, functional genomic techniques have been established, a sophisticated genetic system is available and many central biological functions are under active investigation [19,20]. Figure 1 gives an overview of the approaches used to study sRNAs of H. volcanii.
Overview of the approaches to identify sRNA genes in H. volcanii and to characterize their biological roles
Identification of sRNA genes in H. volcanii
Two different approaches were used for the identification of sRNAs, i.e. comparative computational genome analyses and experimental RNomics. The first approach makes use of the existence of the genome sequences of several haloarchaeal species of diverse genera. In silico libraries of intergenic regions of all haloarchaeal genomes were generated. They were searched for sequence similarities and putative RNA structures that were conserved between several species as well as for snoRNA motifs. The analyses led to the prediction that haloarchaea encode many sRNAs in the intergenic regions of their genomes. For H. volcanii, more than 70 sRNAs were predicted, including C/D box methylation guide snoRNAs. However, H/ACA snoRNAs involved in pseudouridine formation could not be detected.
In a second approach, RNomics was used to detect haloarchaeal sRNAs experimentally. Total RNA of H. volcanii was size-selected and used to generate a cDNA library representing sRNAs of sizes between 130 and 450 nt. After removal of clones with rRNA-derived inserts, approx. 200 clones were sequenced, leading to the identification of approx. 90 different sRNAs. Taken together, more than 160 sRNAs were discovered. The actual number in H. volcanii is most probably considerably larger, because (i) in the RNomics approach, most sRNAs were represented by a single clone, and (ii) the overlap of both approaches was very small. Nevertheless, it has become clear that haloarchaea encode many sRNAs, and the number is in the same range as in well-characterized bacterial species.
Biogenesis of sRNAs in H. volcanii
sRNAs can be generated either as primary transcripts or by processing from larger precursor RNAs. In the former case, the sRNA genes should be preceded by the basal promoter elements BRE (transcription factor B-responsive element), TATA box and −10 element [21–23]. An in silico analysis revealed that only 12 sRNA genes were preceded by basal promoter elements, indicating that either most sRNAs are cleaved from precursor RNAs or that the promoters of sRNA genes are unusual and do not fit the consensus sequence of protein-encoding genes. The differentiation between these two possibilities by cloning upstream regions of sRNAs in front of a reporter gene and localizing promoter sequences in vivo is underway. Northern blot analyses with all experimentally identified intergenic sRNA genes have been performed, and the expression of the majority of genes could be verified. In several cases, larger bands were detected in addition to the sRNAs, indicating that precursors to these sRNAs exist. The precursor RNAs will be characterized by identification of their 5′- and 3′-ends using an established method . This will enable us to study processing biochemically in vitro using methods that have been established for the analysis of tRNA and 5S rRNA processing [24,25].
Generation and characterization of sRNA-deletion mutants
Haloarchaea were the first archaeal species for which a transformation system could be established . Strains and methods have been developed that enable the construction of chromosomal deletion mutants of protein-encoding genes using two consecutive selection steps [27–29]. The method has been applied to generate deletion mutants of sRNA genes. Five mutants have already been constructed and more are underway. All five mutants were viable, showing that these sRNAs are not essential during exponential growth in complex medium. H. volcanii can be grown in 96-well microtitre plates . This enables the easy phenotypic comparison of mutants and wild-type under many different conditions (‘phenotyping’). In the three cases investigated to date, mutant and wild-type grew indistinguishably under nearly all conditions, but a very specific phenotype of the mutant could be detected. Two of the sRNAs were found to be essential for the adaptation of H. volcanii to extreme conditions, either to high temperature or to low salt concentrations. In both cases, the mutants exhibited a very severe growth defect, in contrast with the wild-type. The proteomes of wild-type and mutants were compared by two-dimensional gel electrophoresis and, in both cases, several protein spots were identified that were nearly absent from the mutants (J. Straub, M. Brenneis, A. Jellen-Ritter, R. Heyer, J. Soppa and A. Marchfelder, unpublished work). After protein identification, we will test whether the respective transcripts are direct or indirect targets of the two sRNAs. The phenotype of the third mutant revealed that sRNAs in H. volcanii are involved not only in the adaptation to extreme conditions, but also in the regulation of carbon metabolism. The mutant analysis has led to the identification of the first biological functions of sRNAs in archaea(except for snoRNAs).
The LSm (Sm and like-Sm) protein of H. volcanii
The family of LSm proteins is present in archaea, bacteria and eukaryotes and contains RNA-binding proteins, including proteins essential for sRNA function (e.g. the Hfq protein of E. coli). The genome of H. volcanii encodes one LSm protein. It was shown to be constitutively expressed under various conditions at both the mRNA and the protein levels. The protein has been produced heterologously and is currently used to establish an in vitro assay for the characterization of sRNA binding. Deletion of the lsm gene (underway) or conditional silencing techniques will be used to study the biological function of the H. volcanii LSm protein in vivo.
Taken together, proof-of-principle experiments for most approaches listed in Figure 1 have been performed, which together are used to characterize the biological functions of sRNAs in H. volcanii. A higher number of sRNA genes could be identified than for any other archaeal species. Using Northern blot analyses, it could be shown that the sRNA genes are differentially expressed, and a general analysis of sRNA gene expression using DNA microarrays under various conditions is underway. The construction and characterization of deletion mutants has led to the discovery of the first biological functions of archaeal sRNAs and has identified sRNAs that are essential for growth at the temperature and osmolarity limits of H. volcanii. The collective effort of several research groups will lead to an overview of different biological functions of archaeal sRNAs in the near future and underscore that H. volcanii is an excellent model in which to study archaeal biology.
This work was supported by the Deutsche Forschungsgemeinschaft priority programme ‘Sensory and regulatory RNAs in Prokaryotes’ (SPP1258).
Molecular Biology of Archaea: Biochemical Society Focused Meeting held at University of St Andrews, U.K., 19–21 August 2008. Organized and Edited by Stephen Bell (Oxford, U.K.) and Malcolm White (St Andrews, U.K.).