Insights into RNA-processing pathways and associated RNA-degrading enzymes in Archaea

Insights into RNA-processing pathways and associated RNA-degrading enzymes in Archaea Abstract RNA-processing pathways are at the centre of regulation of gene expression. All RNA transcripts undergo multiple maturation steps in addition to covalent chemical modifications to become functional in the cell. This includes destroying unnecessary or defective cellular RNAs. In Archaea, information on mechanisms by which RNA species reach their mature forms and associated RNA-modifying enzymes are still fragmentary. To date, most archaeal actors and pathways have been proposed in light of information gathered from Bacteria and Eukarya. In this context, this review provides a state of the art overview of archaeal endoribonucleases and exoribonucleases that cleave and trim RNA species and also of the key small archaeal proteins that bind RNAs. Furthermore, synthetic up-to-date views of processing and biogenesis pathways of archaeal transfer and ribosomal RNAs as well as of maturation of stable small non-coding RNAs such as CRISPR RNAs, small C/D and H/ACA box guide RNAs, and other emerging classes of small RNAs are described. Finally, prospective post-transcriptional mechanisms to control archaeal messenger RNA quality and quantity are discussed. Archaea, ribonucleases, RNA processing, RNA decay, RNA-binding proteins, CRISPR-Cas INTRODUCTION RNA biology plays a critical role in accurate gene expression, a prerequisite for all cellular processes from the simplest to most complex organisms. Expression of genetic information depends on messenger RNAs (mRNAs) to define the proteome and on ribosomal RNAs (rRNAs) and transfer RNAs (tRNAs) to decode mRNA sequence information as well as on additional regulatory non-coding RNAs (ncRNAs). During protein synthesis, bonds are formed in an active site of the ribosome residing on the large subunit, called the peptidyl transferase centre, which is devoid of any r-protein side chains, thus establishing the ribosome as a ribozyme (Nissen et al.2000). In this framework, RNA processing and decay paths stand at the centre of post-transcriptional regulation of gene expression and, therefore, are fundamental in defining the phenotypic characteristics of a cell. In addition, they permit a rapid adaptation to changes of environmental conditions. The continuing discovery of multiple mechanistically distinct pathways in Bacteria and Eukarya highlights the complexity of these processes. The quality and level of each cellular RNA species are tightly controlled and determined by the relative rates of transcription, maturation and decay (Durand et al.2015; Wagner and Romby 2015; Houri-Zeevi and Rechavi 2017). Based on evidence obtained from Eukarya and Bacteria, it is now established that an ordered action of a cell-specific set of RNA-modifying and RNA-processing enzymes is required from RNA synthesis to decay. Enzymes which act directly on RNA molecules are ribonucleases (RNases) that catalyse the exo- or endoribonucleolytic cleavage of phosphodiester bonds and act in concert with other RNA-related ancillary enzymes such as RNA helicases, poly(A) polymerases and pyrophosphohydrolases. RNA-processing enzymes are often found in RNA-degrading complexes and may be recruited to their respective targets by several means: cis-acting RNA sequence elements, trans-acting proteins or ncRNAs. RNA maturation or turnover is often initiated through internal cleavage by the action of endoribonucleases but also through 3΄-to-5΄ and 5΄-to-3΄ exoribonucleolytic decay by the action of the exoribonucleases. The tight teamwork between different RNA-degrading systems defines the outcome of each RNA species within the cell. Ribonucleases can be grouped according to their substrate specificity (single and/or double-stranded RNA or RNA/DNA hybrid), the nature of the product they release or their mechanism of action (processive or distributive). In-depth reviews covering enzymes involved in critical aspects of RNA processing in Bacteria and Eukarya have highlighted the considerable number of these processes from the nascent RNA to its degradation ((Durand et al.2015; Wagner and Romby 2015; Houri-Zeevi and Rechavi 2017) and chapters in Stoecklin and Mühlemann 2013). Archaeal microorganisms, omnipresent in Earth's ecosystems, constitute a unique domain of life with features at the frontier of Bacteria and Eukarya that had led to a revolution in the field of evolutionary biology. In this regard, elucidating fundamental cellular processes in Archaea is of a critical importance to understand life on earth. Moreover, the importance of studying archaeal microorganisms recently emerged from their identification in complex microbiomes of plants, animals and humans and opened the question of their impacts on the biosphere and health (Moissl-Eichinger et al.2018). Finally, the close phylogenetic relationship between Archaea and Eukarya increasingly highlights the advantage of using archaeal systems which use simpler proteins or ribonucleoprotein (RNP) complexes. Indeed, archaeal and eukaryotic cells share many key components of their processing systems, such as replication, transcription and translation machineries (Brochier-Armanet, Forterre and Gribaldo 2011; Adam et al.2017). This provides valuable insights in understanding associated enzymatic activities and molecular processes. Based on genomic data it has been accepted that archaeal phylogeny includes the TACK (Thaumarchaeota-Aigarchaeota-Crenarchaeota-Korarchaeota)/Proteoarchaeota group, the Euryarchaea and the DPANN (Diapherotrites, Parvarchaeota, Aenigmarchaeota, Nanoarchaeota and Nanohaloarchaeota) lineages (Fig. 1) (Guy and Ettema 2011). However, in view of the mosaic content of archaeal genomes, an enthusiastic debate is taking place around the topology of the archaeal phylogenetic tree and its connection to the eukaryal branch and questions the veracity of our present picture of the tree of life (Eme et al.2017; Levasseur et al.2017; Van der Gulik, Hoff and Speijer 2017). Figure 1. View largeDownload slide Taxonomic distribution of ribonuclease protein families across the archaeal phylogeny. The size of the filled circles is proportional to the percentage of members of ribonucleases in each family. Each family is represented by a single colour which is also used in Figs 2 and 4. Classification of taxonomic groups is according to the phylogenetic tree (Adam et al.2017), with the exception of the position of cluster I. The DPANN (Diapherotrites, Parvarchaeota, Aenigmarchaeota, Nanoarchaeota, Nanohaloarchaea) refers to the superphylum of Extremophile Archaea (Rinke et al.2013). The taxonomic status (Adam et al.2017) is shown in parentheses: C = Class; P = Phylum; SC = Super Class; SP = Super Phylum. To retrieve all members of each ribonuclease family, we collected an initial set of proteins using PFAM and/or COG profile annotations from complete archaeal genomes. Subsequently a graph partitioning method (MCL) was used to remove false positives. Hidden Markov model (hmm) profiles were built for each family and used with hmmsearch (HMMER package) to identify homologues in annotated archaeal genomes retrieved from NCBI (https://www.ncbi.nlm.nih.gov/genome). In addition, to identify potential unannotated genes, we performed tblastn searches. Finally, the percentage of occurrence of each ribonuclease family gene is computed for each taxonomic group (one strain per species) by taking into account the presence or absence of a candidate gene in a genome, i.e. paralogues are counted only once. For each taxonomic group, if complete annotated genomes (cg) were available, these were used in the analysis, otherwise incomplete genomes (ws) were used. For incomplete genomes (ws), the absence of a protein may be due to the fact that the genomes are partial. For references and further details, see section ‘Archaeal ribonuclease families’. Figure 1. View largeDownload slide Taxonomic distribution of ribonuclease protein families across the archaeal phylogeny. The size of the filled circles is proportional to the percentage of members of ribonucleases in each family. Each family is represented by a single colour which is also used in Figs 2 and 4. Classification of taxonomic groups is according to the phylogenetic tree (Adam et al.2017), with the exception of the position of cluster I. The DPANN (Diapherotrites, Parvarchaeota, Aenigmarchaeota, Nanoarchaeota, Nanohaloarchaea) refers to the superphylum of Extremophile Archaea (Rinke et al.2013). The taxonomic status (Adam et al.2017) is shown in parentheses: C = Class; P = Phylum; SC = Super Class; SP = Super Phylum. To retrieve all members of each ribonuclease family, we collected an initial set of proteins using PFAM and/or COG profile annotations from complete archaeal genomes. Subsequently a graph partitioning method (MCL) was used to remove false positives. Hidden Markov model (hmm) profiles were built for each family and used with hmmsearch (HMMER package) to identify homologues in annotated archaeal genomes retrieved from NCBI (https://www.ncbi.nlm.nih.gov/genome). In addition, to identify potential unannotated genes, we performed tblastn searches. Finally, the percentage of occurrence of each ribonuclease family gene is computed for each taxonomic group (one strain per species) by taking into account the presence or absence of a candidate gene in a genome, i.e. paralogues are counted only once. For each taxonomic group, if complete annotated genomes (cg) were available, these were used in the analysis, otherwise incomplete genomes (ws) were used. For incomplete genomes (ws), the absence of a protein may be due to the fact that the genomes are partial. For references and further details, see section ‘Archaeal ribonuclease families’. Archaeal RNA species and RNA biology-associated processes which finely regulate gene expression appear to be mosaic, with both eukaryal and bacterial features. A unique archaeal RNA polymerase (RNAP) related to the eukaryal RNA polymerase II (RNAPII) is in charge of transcribing all archaeal RNA classes of which mRNAs share bacterial features with no eukaryotic 5΄-end capping structure and generally no introns. To date, only a single example of an intron within a gene encoding a CBF5 pseudouridine synthetase has been identified in some Crenarchaeota (Yokobori et al.2009). In addition, archaeal pre-tRNAs contain introns that occur in eukaryal nuclear tRNAs, and several eukaryotic-like RNA-processing enzymes and RNA regulatory factors have been found to be encoded in archaeal genomes. This includes the 3΄-to-5΄ exosome machinery, the eukaryotic-like sets of ncRNAs namely C/D and H/ACA box small RNAs (sRNAs) guiding 2΄-O-ribose-methylation and pseudouridylation modifications and tRNA intron-processing enzymes. However, as in Bacteria, RNA transcription and translation are proposed to be coupled in Archaea. In this mosaic context, pathways dictating the biogenesis and the fate of archaeal RNAs wait for most of them to be precisely characterised. This review condenses the state of our current knowledge on associated players and pathways mediating maturation and degradation of archaeal RNA species. This focus opens with a comprehensive overview on the major protein families acknowledged to date which are capable of degrading or binding RNA molecules. This includes a comprehensive study of the taxonomic distribution of key ribonucleases setting the functional status of each enzyme across the archaeal phylogeny. Subsequently, maturation pathways leading to functional tRNAs, rRNAs, non-coding guide H/ACA and C/D box sRNAs, clustered regularly interspaced short palindromic repeat RNAs (CRISPR or crRNAs) and novel emerging classes of ncRNAs are described. Finally, a prospective on archaeal mRNA turnover and quality control pathways is discussed to propose mRNA-signalling pathways at regulating gene expression in Archaea. Archaeal ribonuclease families Our purpose is to provide a panoramic view of archaeal enzymes and machines that cleave and degrade the various RNA types in light of bacterial and eukaryotic systems to question their significance. Since 2001, studies on bacterial and eukaryal RNA-processing apparatuses have supplied the foundation for the discovery of RNA-degrading enzymes implicated in fundamental aspects of RNA processing in Archaea (see Koonin, Wolf and Aravind 2001; Klug et al.2007; Evguenieva-Hackenberg and Klug 2009; de Koning et al.2010; Clouet-d’Orval et al.2015; Phung, Bouvier and Clouet-d’Orval 2017 for review). The 10 archaeal RNase families highlighted in Tables 1A and 1B are defined based on their bacterial and eukaryal counterparts (Arraiano et al.2013; Stoecklin and Mühlemann 2013) (Table 2). We performed a phylogenomic search of members of each family within archaeal genomes to highlight the occurrence of these RNase families across the phylogeny of Archaea. This is shown and described in Fig. 1. Table 1A. Acknowledged archaeal endoribonuclease families. Activity Family Name Organism PDB accession Reference ENDO RNase P POP5/RPP30 P. horikoshii 2czv* Kawano et al. (2006) T. kodakarensis 3wz0 Suematsu et al. (2015) POP5 P. furiosus 2av5 Wilson et al. (2006) RPP21/RPP29 P. horikoshii 2zae* Honda et al. (2008) P. furiosus 2ki7 Xu et al. (2009) RPP21 P. furiosus 2k3r Amero et al. (2008) RPP29 M. thermautrophilus 1oqk Boomershine et al. (2003) A. fulgidus 1pc0, 1ts9 Sidote and Hoffman (2003); Sidote, Heideker and Hoffman (2004) RPP38/L7Ae M. jannaschii 1xbi Suryadi et al. (2005) P. abyssi 1pxw Charron et al. (2004a) A. pernix 2fc3 Bhuiya et al. (2013) P. horikoshii 2czw* Fukuhara et al. (2006) EndA/ Sen α4 M. jannaschii 1a79* Li, Trotta and Abelson (1998) α'2 A. fulgidus 1rlv*/ 2gjw/3p1y/ 1r0v Li and Abelson (2000); Xue, Calvin and Li (2006); Hirata, Kitajima and Hori (2011); Zhang and Li (2004) N. equitans 3iey Mitchell et al. (2009) T. acidophilum 2ohc Kim et al. (2007) α2β2 A. fulgidus 3p1z*/ 3ajv Hirata, Kitajima and Hori (2011); Okuda et al. (2011) P. aerophilum 2zyz Yoshinari et al. (2009) ε 2 ARMAN 4fz2* Hirata et al. (2012) rnz aRNase Z P. abyssi model* (39% Id. elac1) Pelota/ Dom34 aPelota T. acidophilum 2qi2* Lee et al. (2007) A. pernix 3wxm Kobayashi et al. (2010) A. fulgidus 3oby Lee et al. (2010) P. furiosus 3j15 Becker et al. (2012) PIN aNob1 P. horikoshii 2lcq* Veith et al. (2012) RNase H HII A. fulgidus 1i39*/ 3p83 Chapados et al. (2001); Bubeck et al. (2011) M. jannaschii 1eke Lai et al. (2000) Ferrodoxin like Cas6 P. furiosus 3pkm*/ 3ufc Wang et al. (2011); Park et al. (2012) S. solfataricus 3zfv Reeks et al. (2013) β-CASP aCPSF1 P. horikoshii 3af6*/ 3af5 Nishida et al. (2010) M. mazei 2xr1 Mir-montazeri et al. (2011) M. thermautotrophicus 2ycb Silva et al. (2011) Activity Family Name Organism PDB accession Reference ENDO RNase P POP5/RPP30 P. horikoshii 2czv* Kawano et al. (2006) T. kodakarensis 3wz0 Suematsu et al. (2015) POP5 P. furiosus 2av5 Wilson et al. (2006) RPP21/RPP29 P. horikoshii 2zae* Honda et al. (2008) P. furiosus 2ki7 Xu et al. (2009) RPP21 P. furiosus 2k3r Amero et al. (2008) RPP29 M. thermautrophilus 1oqk Boomershine et al. (2003) A. fulgidus 1pc0, 1ts9 Sidote and Hoffman (2003); Sidote, Heideker and Hoffman (2004) RPP38/L7Ae M. jannaschii 1xbi Suryadi et al. (2005) P. abyssi 1pxw Charron et al. (2004a) A. pernix 2fc3 Bhuiya et al. (2013) P. horikoshii 2czw* Fukuhara et al. (2006) EndA/ Sen α4 M. jannaschii 1a79* Li, Trotta and Abelson (1998) α'2 A. fulgidus 1rlv*/ 2gjw/3p1y/ 1r0v Li and Abelson (2000); Xue, Calvin and Li (2006); Hirata, Kitajima and Hori (2011); Zhang and Li (2004) N. equitans 3iey Mitchell et al. (2009) T. acidophilum 2ohc Kim et al. (2007) α2β2 A. fulgidus 3p1z*/ 3ajv Hirata, Kitajima and Hori (2011); Okuda et al. (2011) P. aerophilum 2zyz Yoshinari et al. (2009) ε 2 ARMAN 4fz2* Hirata et al. (2012) rnz aRNase Z P. abyssi model* (39% Id. elac1) Pelota/ Dom34 aPelota T. acidophilum 2qi2* Lee et al. (2007) A. pernix 3wxm Kobayashi et al. (2010) A. fulgidus 3oby Lee et al. (2010) P. furiosus 3j15 Becker et al. (2012) PIN aNob1 P. horikoshii 2lcq* Veith et al. (2012) RNase H HII A. fulgidus 1i39*/ 3p83 Chapados et al. (2001); Bubeck et al. (2011) M. jannaschii 1eke Lai et al. (2000) Ferrodoxin like Cas6 P. furiosus 3pkm*/ 3ufc Wang et al. (2011); Park et al. (2012) S. solfataricus 3zfv Reeks et al. (2013) β-CASP aCPSF1 P. horikoshii 3af6*/ 3af5 Nishida et al. (2010) M. mazei 2xr1 Mir-montazeri et al. (2011) M. thermautotrophicus 2ycb Silva et al. (2011) Structural models shown in Fig. 2 are indicated *. Note that aCPSF1 has been classified in here but carries also a 5΄-3΄ exoribonucleolytic activity (see section ‘Archaeal ribonuclease families’). View Large Table 1A. Acknowledged archaeal endoribonuclease families. Activity Family Name Organism PDB accession Reference ENDO RNase P POP5/RPP30 P. horikoshii 2czv* Kawano et al. (2006) T. kodakarensis 3wz0 Suematsu et al. (2015) POP5 P. furiosus 2av5 Wilson et al. (2006) RPP21/RPP29 P. horikoshii 2zae* Honda et al. (2008) P. furiosus 2ki7 Xu et al. (2009) RPP21 P. furiosus 2k3r Amero et al. (2008) RPP29 M. thermautrophilus 1oqk Boomershine et al. (2003) A. fulgidus 1pc0, 1ts9 Sidote and Hoffman (2003); Sidote, Heideker and Hoffman (2004) RPP38/L7Ae M. jannaschii 1xbi Suryadi et al. (2005) P. abyssi 1pxw Charron et al. (2004a) A. pernix 2fc3 Bhuiya et al. (2013) P. horikoshii 2czw* Fukuhara et al. (2006) EndA/ Sen α4 M. jannaschii 1a79* Li, Trotta and Abelson (1998) α'2 A. fulgidus 1rlv*/ 2gjw/3p1y/ 1r0v Li and Abelson (2000); Xue, Calvin and Li (2006); Hirata, Kitajima and Hori (2011); Zhang and Li (2004) N. equitans 3iey Mitchell et al. (2009) T. acidophilum 2ohc Kim et al. (2007) α2β2 A. fulgidus 3p1z*/ 3ajv Hirata, Kitajima and Hori (2011); Okuda et al. (2011) P. aerophilum 2zyz Yoshinari et al. (2009) ε 2 ARMAN 4fz2* Hirata et al. (2012) rnz aRNase Z P. abyssi model* (39% Id. elac1) Pelota/ Dom34 aPelota T. acidophilum 2qi2* Lee et al. (2007) A. pernix 3wxm Kobayashi et al. (2010) A. fulgidus 3oby Lee et al. (2010) P. furiosus 3j15 Becker et al. (2012) PIN aNob1 P. horikoshii 2lcq* Veith et al. (2012) RNase H HII A. fulgidus 1i39*/ 3p83 Chapados et al. (2001); Bubeck et al. (2011) M. jannaschii 1eke Lai et al. (2000) Ferrodoxin like Cas6 P. furiosus 3pkm*/ 3ufc Wang et al. (2011); Park et al. (2012) S. solfataricus 3zfv Reeks et al. (2013) β-CASP aCPSF1 P. horikoshii 3af6*/ 3af5 Nishida et al. (2010) M. mazei 2xr1 Mir-montazeri et al. (2011) M. thermautotrophicus 2ycb Silva et al. (2011) Activity Family Name Organism PDB accession Reference ENDO RNase P POP5/RPP30 P. horikoshii 2czv* Kawano et al. (2006) T. kodakarensis 3wz0 Suematsu et al. (2015) POP5 P. furiosus 2av5 Wilson et al. (2006) RPP21/RPP29 P. horikoshii 2zae* Honda et al. (2008) P. furiosus 2ki7 Xu et al. (2009) RPP21 P. furiosus 2k3r Amero et al. (2008) RPP29 M. thermautrophilus 1oqk Boomershine et al. (2003) A. fulgidus 1pc0, 1ts9 Sidote and Hoffman (2003); Sidote, Heideker and Hoffman (2004) RPP38/L7Ae M. jannaschii 1xbi Suryadi et al. (2005) P. abyssi 1pxw Charron et al. (2004a) A. pernix 2fc3 Bhuiya et al. (2013) P. horikoshii 2czw* Fukuhara et al. (2006) EndA/ Sen α4 M. jannaschii 1a79* Li, Trotta and Abelson (1998) α'2 A. fulgidus 1rlv*/ 2gjw/3p1y/ 1r0v Li and Abelson (2000); Xue, Calvin and Li (2006); Hirata, Kitajima and Hori (2011); Zhang and Li (2004) N. equitans 3iey Mitchell et al. (2009) T. acidophilum 2ohc Kim et al. (2007) α2β2 A. fulgidus 3p1z*/ 3ajv Hirata, Kitajima and Hori (2011); Okuda et al. (2011) P. aerophilum 2zyz Yoshinari et al. (2009) ε 2 ARMAN 4fz2* Hirata et al. (2012) rnz aRNase Z P. abyssi model* (39% Id. elac1) Pelota/ Dom34 aPelota T. acidophilum 2qi2* Lee et al. (2007) A. pernix 3wxm Kobayashi et al. (2010) A. fulgidus 3oby Lee et al. (2010) P. furiosus 3j15 Becker et al. (2012) PIN aNob1 P. horikoshii 2lcq* Veith et al. (2012) RNase H HII A. fulgidus 1i39*/ 3p83 Chapados et al. (2001); Bubeck et al. (2011) M. jannaschii 1eke Lai et al. (2000) Ferrodoxin like Cas6 P. furiosus 3pkm*/ 3ufc Wang et al. (2011); Park et al. (2012) S. solfataricus 3zfv Reeks et al. (2013) β-CASP aCPSF1 P. horikoshii 3af6*/ 3af5 Nishida et al. (2010) M. mazei 2xr1 Mir-montazeri et al. (2011) M. thermautotrophicus 2ycb Silva et al. (2011) Structural models shown in Fig. 2 are indicated *. Note that aCPSF1 has been classified in here but carries also a 5΄-3΄ exoribonucleolytic activity (see section ‘Archaeal ribonuclease families’). View Large Table 1B. Acknowledged archaeal exoribonuclease families. Activity Family Name Organism PDB accession Reference 3΄-5΄ EXO PDX/RNase PH Exosome (Csl4) A. fulgidus 3m7n*/3m85 Hartung et al. (2010) Exosome (Rrp4) A. fulgidus 2ba0*/2ba1 Büttner, Wenig and Hopfneret al. (2005) Exosome S. solfataricus 2je6/2jea/2jeb, 2br2, 3l7z, 4ba1/4ba2, 2c37/2c38/2c39 Lorentzen et al. (2005, 2007); Lorentzen and Conti (2012); Lu, Ding and Ke et al. (2010); Lorentzen and Conti (2005) P. abyssi 2po2/ 2pnz/2po0/ 2po1 Navarro et al. (2008) RNB/RNase II aRNase R H. volcanii model a* (28% Id. dis3l2) M. concili model b* (28% Id. dis3l2/OB-fold) M. conradii model c* (28% Id. dis3l2/ dsRBD) 5΄-3΄ EXO β-CASP aCPSF1b M. jannaschii model (32% Id. M. mazei aCPSF1) aCPSF2 S. acidocaldarius model* (33% Id. M. mazei aCPSF1) aRNase J M. psychrophilus 5haa/5hab* Zheng et al. (2017) Activity Family Name Organism PDB accession Reference 3΄-5΄ EXO PDX/RNase PH Exosome (Csl4) A. fulgidus 3m7n*/3m85 Hartung et al. (2010) Exosome (Rrp4) A. fulgidus 2ba0*/2ba1 Büttner, Wenig and Hopfneret al. (2005) Exosome S. solfataricus 2je6/2jea/2jeb, 2br2, 3l7z, 4ba1/4ba2, 2c37/2c38/2c39 Lorentzen et al. (2005, 2007); Lorentzen and Conti (2012); Lu, Ding and Ke et al. (2010); Lorentzen and Conti (2005) P. abyssi 2po2/ 2pnz/2po0/ 2po1 Navarro et al. (2008) RNB/RNase II aRNase R H. volcanii model a* (28% Id. dis3l2) M. concili model b* (28% Id. dis3l2/OB-fold) M. conradii model c* (28% Id. dis3l2/ dsRBD) 5΄-3΄ EXO β-CASP aCPSF1b M. jannaschii model (32% Id. M. mazei aCPSF1) aCPSF2 S. acidocaldarius model* (33% Id. M. mazei aCPSF1) aRNase J M. psychrophilus 5haa/5hab* Zheng et al. (2017) Structural models shown in Fig. 4 are indicated *. View Large Table 1B. Acknowledged archaeal exoribonuclease families. Activity Family Name Organism PDB accession Reference 3΄-5΄ EXO PDX/RNase PH Exosome (Csl4) A. fulgidus 3m7n*/3m85 Hartung et al. (2010) Exosome (Rrp4) A. fulgidus 2ba0*/2ba1 Büttner, Wenig and Hopfneret al. (2005) Exosome S. solfataricus 2je6/2jea/2jeb, 2br2, 3l7z, 4ba1/4ba2, 2c37/2c38/2c39 Lorentzen et al. (2005, 2007); Lorentzen and Conti (2012); Lu, Ding and Ke et al. (2010); Lorentzen and Conti (2005) P. abyssi 2po2/ 2pnz/2po0/ 2po1 Navarro et al. (2008) RNB/RNase II aRNase R H. volcanii model a* (28% Id. dis3l2) M. concili model b* (28% Id. dis3l2/OB-fold) M. conradii model c* (28% Id. dis3l2/ dsRBD) 5΄-3΄ EXO β-CASP aCPSF1b M. jannaschii model (32% Id. M. mazei aCPSF1) aCPSF2 S. acidocaldarius model* (33% Id. M. mazei aCPSF1) aRNase J M. psychrophilus 5haa/5hab* Zheng et al. (2017) Activity Family Name Organism PDB accession Reference 3΄-5΄ EXO PDX/RNase PH Exosome (Csl4) A. fulgidus 3m7n*/3m85 Hartung et al. (2010) Exosome (Rrp4) A. fulgidus 2ba0*/2ba1 Büttner, Wenig and Hopfneret al. (2005) Exosome S. solfataricus 2je6/2jea/2jeb, 2br2, 3l7z, 4ba1/4ba2, 2c37/2c38/2c39 Lorentzen et al. (2005, 2007); Lorentzen and Conti (2012); Lu, Ding and Ke et al. (2010); Lorentzen and Conti (2005) P. abyssi 2po2/ 2pnz/2po0/ 2po1 Navarro et al. (2008) RNB/RNase II aRNase R H. volcanii model a* (28% Id. dis3l2) M. concili model b* (28% Id. dis3l2/OB-fold) M. conradii model c* (28% Id. dis3l2/ dsRBD) 5΄-3΄ EXO β-CASP aCPSF1b M. jannaschii model (32% Id. M. mazei aCPSF1) aCPSF2 S. acidocaldarius model* (33% Id. M. mazei aCPSF1) aRNase J M. psychrophilus 5haa/5hab* Zheng et al. (2017) Structural models shown in Fig. 4 are indicated *. View Large Table 2. Distribution of ribonuclease enzymes and subunits across the three domains of life. Enzymes Bacteria Archaea Eukarya RNaseP/RPR Type A Type A/M/T H1 RNA RNaseP/RPP C5 protein/RnpA POP5 POP5 RPP21/29/30 RPP21/29/30 RPP38/L7Ae RPP38 RPP20,RPP25, RPP14,RPP40,POP1 EndA α / α'/ β/ ε Sen 2/15/34/54 RNase Z RNase Z aRNase Z RNase ZS/ RNase ZL Pelota/Dom34 aPelota Pelota/Dom34 Nob1 aNob1 Nob1 RNase H RNaseHI/HII aRNase HII RNase H1/H2 Cas6 Cas6 Cas6 CPSF-like Unch. CPSF* aCPSF1/1b/2 CPSF73/ Int11/RC-68 aCβx/aCβy/aCβz CSPF100Δ/Int9/RC-74Δ RNase J-like RNase J aRNase J Exosome/PNPase RNase PH Rrp41/42Δ Rrp41Δ/42 Δ Rrp4 Rrp4 Csl4 Csl4 Rrp44/Dis3 RNase R-like RNase R aRNase Ra/Rb/Rc Enzymes Bacteria Archaea Eukarya RNaseP/RPR Type A Type A/M/T H1 RNA RNaseP/RPP C5 protein/RnpA POP5 POP5 RPP21/29/30 RPP21/29/30 RPP38/L7Ae RPP38 RPP20,RPP25, RPP14,RPP40,POP1 EndA α / α'/ β/ ε Sen 2/15/34/54 RNase Z RNase Z aRNase Z RNase ZS/ RNase ZL Pelota/Dom34 aPelota Pelota/Dom34 Nob1 aNob1 Nob1 RNase H RNaseHI/HII aRNase HII RNase H1/H2 Cas6 Cas6 Cas6 CPSF-like Unch. CPSF* aCPSF1/1b/2 CPSF73/ Int11/RC-68 aCβx/aCβy/aCβz CSPF100Δ/Int9/RC-74Δ RNase J-like RNase J aRNase J Exosome/PNPase RNase PH Rrp41/42Δ Rrp41Δ/42 Δ Rrp4 Rrp4 Csl4 Csl4 Rrp44/Dis3 RNase R-like RNase R aRNase Ra/Rb/Rc CPSF73/CPSF100 heterodimer (cleavage and polyadenylation specific factor complex) involved in 3΄end processing of eukaryal canonical and histone pre-mRNA Int11/RC-68/ Int9/RC-74 (integrator complex) involved in 3΄end processing of eukaryal pre-snRNA (see Dominski, Carpousis and Clouet-d’Orval 2013 for review) * Uncharacterised CPSF-like β-CASP are detected in bacteria; Δinactive subunit. View Large Table 2. Distribution of ribonuclease enzymes and subunits across the three domains of life. Enzymes Bacteria Archaea Eukarya RNaseP/RPR Type A Type A/M/T H1 RNA RNaseP/RPP C5 protein/RnpA POP5 POP5 RPP21/29/30 RPP21/29/30 RPP38/L7Ae RPP38 RPP20,RPP25, RPP14,RPP40,POP1 EndA α / α'/ β/ ε Sen 2/15/34/54 RNase Z RNase Z aRNase Z RNase ZS/ RNase ZL Pelota/Dom34 aPelota Pelota/Dom34 Nob1 aNob1 Nob1 RNase H RNaseHI/HII aRNase HII RNase H1/H2 Cas6 Cas6 Cas6 CPSF-like Unch. CPSF* aCPSF1/1b/2 CPSF73/ Int11/RC-68 aCβx/aCβy/aCβz CSPF100Δ/Int9/RC-74Δ RNase J-like RNase J aRNase J Exosome/PNPase RNase PH Rrp41/42Δ Rrp41Δ/42 Δ Rrp4 Rrp4 Csl4 Csl4 Rrp44/Dis3 RNase R-like RNase R aRNase Ra/Rb/Rc Enzymes Bacteria Archaea Eukarya RNaseP/RPR Type A Type A/M/T H1 RNA RNaseP/RPP C5 protein/RnpA POP5 POP5 RPP21/29/30 RPP21/29/30 RPP38/L7Ae RPP38 RPP20,RPP25, RPP14,RPP40,POP1 EndA α / α'/ β/ ε Sen 2/15/34/54 RNase Z RNase Z aRNase Z RNase ZS/ RNase ZL Pelota/Dom34 aPelota Pelota/Dom34 Nob1 aNob1 Nob1 RNase H RNaseHI/HII aRNase HII RNase H1/H2 Cas6 Cas6 Cas6 CPSF-like Unch. CPSF* aCPSF1/1b/2 CPSF73/ Int11/RC-68 aCβx/aCβy/aCβz CSPF100Δ/Int9/RC-74Δ RNase J-like RNase J aRNase J Exosome/PNPase RNase PH Rrp41/42Δ Rrp41Δ/42 Δ Rrp4 Rrp4 Csl4 Csl4 Rrp44/Dis3 RNase R-like RNase R aRNase Ra/Rb/Rc CPSF73/CPSF100 heterodimer (cleavage and polyadenylation specific factor complex) involved in 3΄end processing of eukaryal canonical and histone pre-mRNA Int11/RC-68/ Int9/RC-74 (integrator complex) involved in 3΄end processing of eukaryal pre-snRNA (see Dominski, Carpousis and Clouet-d’Orval 2013 for review) * Uncharacterised CPSF-like β-CASP are detected in bacteria; Δinactive subunit. View Large Ubiquitous archaeal endoribonucleases To date, archaeal endoribonucleolytic activities were shown to be carried out by the universal RNase P and RNase Z families; the ubiquitous EndA, Nob1, aPelota and aCPSF1 families shared with Eukarya; and finally, the Cas6 family shared with Bacteria (Table 1A). Note that the aCPSF1 subfamily is described in the section ‘The universal β-CASP ribonuclease family’ together with other members of the β-CASP family. RNase P (EC 3.1.26.5) RNase P is a site-specific endoribonuclease which catalyses the hydrolysis of the phosphodiester bond at the junction between a 5΄-end single-stranded and a double-stranded region. This reaction requires Mg2+ ions for the attack of the phosphorous atom in the scissile bond (Guerrier-Takada et al.1983). In all domains of life, RNase P is best known for its role in removing the 5΄-leaders of tRNA precursors (pre-tRNAs) transcribed as mono- or polycistronic transcripts (Frank and Pace 1998; Evans, Marquez and Pace 2006; Esakova and Krasilnikov 2010; Jarrous and Gopalan 2010; Lai et al.2010b). However, RNase P is also involved in processing the precursor of 4.5S RNA, the transfer-messenger RNA (tmRNA) and riboswitches in bacteria, and of ncRNAs, such as intron-encoded box C/D small nucleolar RNAs, in Eukarya (Esakova and Krasilnikov 2010; Altman 2011). It has been proposed that this virtually ubiquitous enzyme has independently originated at least twice in evolution with different architectures (Lechner et al.2015). The most ancient type of RNase P holoenzyme is a ribozyme-based RNP complex assembled on a catalytic RNA subunit (RPR) (Guerrier-Takada et al.1983). This RNP complex adopts divergent scaffolds throughout the three domains of life (see Lai et al.2010b; Klemm et al.2016; Samanta et al.2016 for review). More recently, protein-only RNase P enzymes were discovered in human mitochondria (PRORP) and in the bacterium Aquifex aeolicus (HARP), highlighting a more intricate evolution of RNase P than previously thought (Hartmann and Hartmann 2003; Howard et al.2013; Lechner et al.2015; Klemm et al.2016; Nickel et al.2017). Using the HARP protein sequence from A. aeolicus, homologous proteins could also be identified in archaeal genomes (Nickel et al.2017). However tRNA-processing activity has not yet been shown for these archaeal HARP proteins. The diversity of RPRs and RPPs, composing RNase P in the three domains of life, has been extensively reviewed (Gopalan 2007; Altman 2011; Howard et al.2013; Klemm et al.2016; Samanta et al.2016) (Table 2). Briefly, RPRs are structurally related and classified according to specific structural features important for catalysis and substrate recognition (Gopalan 2007). In Archaea, three types of RPRs are now well defined: the archaeal A- and M-types characterised by two functional RNA domains, the substrate-binding domain (S-domain) and the catalytic domain (C-domain) (Fig. 2A); the newly identified T-type, a shorter RPR only retaining the conventional C-domain but lacking a recognisable S-domain, identified in Thermoproteaceae and the related Crenarchaea Caldivirga maquilingensis and Vulcanisaeta distributa (Lai et al.2010a). Figure 2. View largeDownload slide Domain organisation and structure of archaeal endoribonucleases. Structure (PDB accession numbers) and corresponding model organism references are reported in Table 1A and marked by an asterisk. When no structure was available (protein name tagged with an asterisk in figure), we used the Phyre2 server (http://www.sbg.bio.ic.ac.uk/phyre2) to generate a 3D protein model. Domains correspond to Pfam annotation (Pfam database http://pfam.xfam.org/). Catalytic sites are highlighted with yellow areas. (A) Archaeal RNase P is composed of a RNA moiety (ribonucleotide sequence shown in black), structured into two domains: a catalytic domain (C-domain) and a structural domain (S-domain). The secondary structure folding of P. horikoshii A-type is represented with typical helical structure (P) numbering, and three Kink-turn motifs (KT). Specific sets of heterodimer (RPP21•RPP29) and heterotetramer (POP5•RPP30) protein subunits bind to the S- and C- domains, respectively. In addition, the small RPP38/L7Ae protein binds to three Kink-turn motifs (KT1, KT2 and KT3) (see Fig. 3A). (B) Structure and domains of the four different forms of EndA found across the archaeal phylogeny are shown. The α4-type is a homotetramer containing four active catalytic sites. Each α subunit is composed of two Pfam domains (PF02778 ‘N-term’ and PF01974 ‘catalytic’, light and dark brown, respectively). The α2/β2-type heterotetramer is shown in dark pink. The α2-type is a heterodimer composed of two units of fused α-types leaving only two active catalytic sites. Finally, the ε2-type is a version with two active catalytic sites composed of a two units of fused α/β types. The archaeal EndAs recognise and cleave BHB motifs (see Fig. 3B). (C) Five other archaeal endoribonucleases, aPelota, RNase Z, Nob1, RNase HII, Cas6, acting as monomers or dimers are shown. The RNase Z which is characterised by a β-lactamase domain interrupted by a small insertion has no crystal structure reported to date. A Phyre2-built structural model of P. abyssi RNase Z shows that the flexible arm (light blue) protrudes from the classical β-lactamase domain. Note that Cas6 protein sequences show limited sequence conservation with only two common motifs: the ferredoxin fold or RRM (RNA recognition motifs) domains and a glycine-rich motif located at the C-terminus. The structure of P. abyssi Cas6 with an RNA fragment is shown. Figure 2. View largeDownload slide Domain organisation and structure of archaeal endoribonucleases. Structure (PDB accession numbers) and corresponding model organism references are reported in Table 1A and marked by an asterisk. When no structure was available (protein name tagged with an asterisk in figure), we used the Phyre2 server (http://www.sbg.bio.ic.ac.uk/phyre2) to generate a 3D protein model. Domains correspond to Pfam annotation (Pfam database http://pfam.xfam.org/). Catalytic sites are highlighted with yellow areas. (A) Archaeal RNase P is composed of a RNA moiety (ribonucleotide sequence shown in black), structured into two domains: a catalytic domain (C-domain) and a structural domain (S-domain). The secondary structure folding of P. horikoshii A-type is represented with typical helical structure (P) numbering, and three Kink-turn motifs (KT). Specific sets of heterodimer (RPP21•RPP29) and heterotetramer (POP5•RPP30) protein subunits bind to the S- and C- domains, respectively. In addition, the small RPP38/L7Ae protein binds to three Kink-turn motifs (KT1, KT2 and KT3) (see Fig. 3A). (B) Structure and domains of the four different forms of EndA found across the archaeal phylogeny are shown. The α4-type is a homotetramer containing four active catalytic sites. Each α subunit is composed of two Pfam domains (PF02778 ‘N-term’ and PF01974 ‘catalytic’, light and dark brown, respectively). The α2/β2-type heterotetramer is shown in dark pink. The α2-type is a heterodimer composed of two units of fused α-types leaving only two active catalytic sites. Finally, the ε2-type is a version with two active catalytic sites composed of a two units of fused α/β types. The archaeal EndAs recognise and cleave BHB motifs (see Fig. 3B). (C) Five other archaeal endoribonucleases, aPelota, RNase Z, Nob1, RNase HII, Cas6, acting as monomers or dimers are shown. The RNase Z which is characterised by a β-lactamase domain interrupted by a small insertion has no crystal structure reported to date. A Phyre2-built structural model of P. abyssi RNase Z shows that the flexible arm (light blue) protrudes from the classical β-lactamase domain. Note that Cas6 protein sequences show limited sequence conservation with only two common motifs: the ferredoxin fold or RRM (RNA recognition motifs) domains and a glycine-rich motif located at the C-terminus. The structure of P. abyssi Cas6 with an RNA fragment is shown. The A-type, related to the bacterial type but lacking the P13/P14 and P18 regions, is the most common in the archaeal phylogeny (Harris et al.2001) (Figs 1 and 2A). The M-type which is restricted to the Methanococci and Archaeoglobus resembles the A-type without the P8 and P16 regions, the P6 pseudoknot and a loop in P15 (Pannucci et al.1999; Gopalan 2007). While the archaeal A- and M-type RPRs contain all of the elements required for substrate recognition and catalysis, with biochemical properties similar to synthetic minimal bacterial RPRs, they are structurally defective in the absence of protein (Guerrier-Takada et al.1983; Kikovska, Svärd and Kirsebom 2007). Only A-type RPRs show residual in vitro endoribonucleolytic activity (Pannucci et al.1999; Gopalan 2007). Interestingly, the catalytic RPR is associated to a single protein in Bacteria, at least five in Archaea and up to 10 in eukaryal nuclei (Frank and Pace 1998; Evans, Marquez and Pace 2006; Ellis and Brown 2009; Hartmann et al.2009; Esakova and Krasilnikov 2010; Jarrous and Gopalan 2010; Lai et al.2010b; Samanta et al.2016) (Tables 1A and 2, Fig. 2A). Although archaeal RNase P RNAs are similar in both sequence and structure to those of Bacteria, all archaeal RNase P protein subunits are orthologous to their cousins in eukaryotic RNase P (Hall and Brown 2001, 2002; Samanta et al.2016) (Table 2). Until recently, the A- and M-types of archaeal RNase P were described as containing at least four proteins (POP5, RPP21, RPP29 and RPP30) (Table 1A). A fifth partner, RPP38, also known as 50S ribosomal protein L7Ae, was recently shown to be associated with the RNA component of both RNase P types (Fukuhara et al.2006; Cho et al.2010; Lai et al. 2014, 2017). Over the years, structural studies and biochemical reconstitutions of the RNase P subunits of Pyrococcus furiosus, P. horikoshii, Methanothermobacter thermoautotrophicum, Methanocaldococcus jannaschii and Methanococcus maripaludis have led to valuable insights into protein-aided RNA catalysis showing that protein content and divalent metal ions affect cleavage fidelity and catalytic efficiency (Kouzuma et al.2003; Tsai et al.2006; Pulukkunat and Gopalan 2008; Li, Willkomm and Hartmann 2009; Chen et al. 2010, 2011, 2012; Sinapah et al.2011; Gao et al.2017; Kimura 2017). The dominant structural feature of the 11-kDa RPP29 subunit is a sheet of six antiparallel β-strands, wrapped around a core of conserved hydrophobic amino acids (Fig. 2A) (Sidote and Hoffman 2003; Numata et al.2004; Sidote, Heideker and Hoffman 2004; Amero et al.2008). The RPP21 subunit comprises an N-terminal domain consisting of two long α-helices, and a central linker and C-terminal domain (CTD) which folds into a zinc ribbon domain forming an L-shaped structure (Fig. 2A) (Kakuta et al.2005; Amero et al.2008). These two subunits form a 30-kDa complex (RPP21•RPP29) (Honda et al.2008; Xu et al.2009) with enhanced substrate affinity (see Samanta et al.2016; Kimura 2017 for review). Despite weak sequence similarities of POP5 and RPP30, structural and footprinting studies revealed a heterotetrameric POP5•RPP30 complex, in which a homodimer of POP5 sits between two RPP30 molecules and contacts the C-domains of the RNA moiety by an hydrophobic interface (Kawano et al.2006; Tsai et al.2006; Honda et al.2010; Crowe et al.2011) (Fig. 2A). The POP5•RPP30 complex enhances the rate of RPR-catalysed self-processing by about 100-fold (Pulukkunat and Gopalan 2008). Finally, the RPP38/L7Ae protein family (see section‘Archaeal small RNA-binding proteins’) is described to specifically recognise a Kink-turn RNA structural motif (also named K-turn or KT), a key architectural motif in RNA structure, characterised by a three-nucleotide bulge flanked by a 5΄ canonical helix and a 3΄ non-canonical helix capped with two sheared non-canonical G•A base pairs in tandem that induces a tightly kinked structure with a twist (Huang and Lilley 2013, 2016) (see Lilley 2012; for review) (Fig. 3A). A novel integrated approach supports the formation of ‘double KT’ modules in A- and M-type archaeal RPR variants which are bound by the RPP38/L7Ae protein and are predicted to facilitate the functions of other RPPs (Lai et al.2017) (Fig. 2A). More specifically, it has been proposed that the formation of the K-turn in the C-domain of A-type RPR (KT1) could promote the interaction between the RPR and POP5•RPP30, while the double KT in the S-domain of both A- and M-types (KT2 and KT3) may facilitate a long-range tertiary interaction required for proper binding of the pre-tRNA substrate (Lai et al.2017). Figure 3. View largeDownload slide RNA structural motifs and small RNA binding proteins. (A) The structure of the Kink-turn motif with and without RPP38/L7Ae. Left panel: The Kink-turn (K-turn) RNA consists of a canonical (grey squares) and a non-canonical (unfilled squares) RNA stem separated by a short asymmetric loop (circles) with a sharp bend between the two stems. The internal loop contains a highly conserved flipped-out U (in blue circle) followed by a third non-Watson Crick base pair and the non-canonical stem usually contains two tandem sheared GA base pairs (non-Watson Crick base pairing is in blue triangle) (Lescoute et al.2005). Middle panel: The Kink-loop (K-loop) is an RNA structural motif related to the K-turn with a single stem closed by a terminal loop (Nolivos, Carpousis and Clouet-d’Orval 2005). Both RNA motifs are recognised by the L7Ae protein. Right panel: Ribbon structure representation of RPP8/L7Ae from P. horikoshii in complex with the K-turn of the SL12 RNA fragment (helix P12 of RPR) (PDB accession number:5DCV). The protein is represented by its ribbon diagram in purple and the RNA is drawn in black/grey stick form. The RNA–protein interaction occurs primarily through nucleotides in the non-canonical stem (sheared GA pairs) and at the 3-nucleotide bulge (G-G-U) (Oshima et al.2016). (B) Bulge-Helix-Bulge RNA (BHB) motifs. Left panel: The canonical BHB motif or hBHBh’ comprises two three-nucleotide bulges (B) enclosing a central main helix (H) of four base pairs. The bulges are flanked by shorter helices (h and h’) comprising at least two base pairs (Thompson and Daniels 1990; Marck and Grosjean 2003). Right panel: The relaxed versions of the BHB motif lack one of the flanking helices and one bulge and are accordingly referred to as hBH and hBh΄ (Marck and Grosjean 2003). Only the α2β2 and ε2 EndA forms can process relaxed BHB motifs (Tocchini-Valentini, Fruscoloni and Tocchini-Valentini 2005; Fujishima et al.2011). Grey squares are for Watson Crick base pairs and grey circles for single-stranded nucleotides. Scissors indicate the EndA cleavage sites within the bulges. (C) Structure of the AF-Sm1 heptamer Sm-like protein. The top view of the ribbon representation of the Archaeoglobulus fulgidus Sm1 (AF-Sm1) is shown (PDB accession number 1I5L, Törö et al.2001). For clarity, the monomers are drawn in a colour gradation from blue to violet. AF-Sm1 forms a seven-membered ring. It was shown that no conformational changes are induced by binding of RNA to the interior of the heptamer (Törö et al.2001). Figure 3. View largeDownload slide RNA structural motifs and small RNA binding proteins. (A) The structure of the Kink-turn motif with and without RPP38/L7Ae. Left panel: The Kink-turn (K-turn) RNA consists of a canonical (grey squares) and a non-canonical (unfilled squares) RNA stem separated by a short asymmetric loop (circles) with a sharp bend between the two stems. The internal loop contains a highly conserved flipped-out U (in blue circle) followed by a third non-Watson Crick base pair and the non-canonical stem usually contains two tandem sheared GA base pairs (non-Watson Crick base pairing is in blue triangle) (Lescoute et al.2005). Middle panel: The Kink-loop (K-loop) is an RNA structural motif related to the K-turn with a single stem closed by a terminal loop (Nolivos, Carpousis and Clouet-d’Orval 2005). Both RNA motifs are recognised by the L7Ae protein. Right panel: Ribbon structure representation of RPP8/L7Ae from P. horikoshii in complex with the K-turn of the SL12 RNA fragment (helix P12 of RPR) (PDB accession number:5DCV). The protein is represented by its ribbon diagram in purple and the RNA is drawn in black/grey stick form. The RNA–protein interaction occurs primarily through nucleotides in the non-canonical stem (sheared GA pairs) and at the 3-nucleotide bulge (G-G-U) (Oshima et al.2016). (B) Bulge-Helix-Bulge RNA (BHB) motifs. Left panel: The canonical BHB motif or hBHBh’ comprises two three-nucleotide bulges (B) enclosing a central main helix (H) of four base pairs. The bulges are flanked by shorter helices (h and h’) comprising at least two base pairs (Thompson and Daniels 1990; Marck and Grosjean 2003). Right panel: The relaxed versions of the BHB motif lack one of the flanking helices and one bulge and are accordingly referred to as hBH and hBh΄ (Marck and Grosjean 2003). Only the α2β2 and ε2 EndA forms can process relaxed BHB motifs (Tocchini-Valentini, Fruscoloni and Tocchini-Valentini 2005; Fujishima et al.2011). Grey squares are for Watson Crick base pairs and grey circles for single-stranded nucleotides. Scissors indicate the EndA cleavage sites within the bulges. (C) Structure of the AF-Sm1 heptamer Sm-like protein. The top view of the ribbon representation of the Archaeoglobulus fulgidus Sm1 (AF-Sm1) is shown (PDB accession number 1I5L, Törö et al.2001). For clarity, the monomers are drawn in a colour gradation from blue to violet. AF-Sm1 forms a seven-membered ring. It was shown that no conformational changes are induced by binding of RNA to the interior of the heptamer (Törö et al.2001). Even though there is no atomic resolution structure of full archaeal RNase P enzymes available to date, a three-dimensional (3D) model (not including RPP38/L7Ae) has been constructed based on crystal structures of the individual protein subunits and of the RPP21•RPP29 and POP5•RPP30 binary complexes (Kimura and Kakuta 2012). This proposed 3D model of P. horikoshii RNase P highlights that protein–RNA interactions in archaeal RNase P could replace RNA–RNA contacts in bacterial RNase P. This framework will assist further structural and functional studies on archaeal, as well as eukaryotic, RNase P enzymes. Interestingly, the taxonomic distribution of the RPP subunits and RPR types among the available archaeal sequenced genomes (Fig. 1) confirmed a previous analysis showing that genes coding for RPP21 are missing or are too divergent to be detected in genomes encoding T-type RPRs that lack an S-domain (Lai et al.2010a). Further investigations are needed to elucidate how RPP29 interacts with T-type RPRs in the absence of RPP21. Collectively, the plasticity in the subunit composition and of T-type RNase P in Thermoproteaceae highlights the rich evolutionary story of RNase P within the archaeal domain (Lai et al.2010b). Furthermore, it is interesting to note that the only known organism without RNase P is the archaeon Nanoarchaeum equitans, which belongs to the fast-evolving DPANN phylogenetic lineage with extremely reduced genome sizes (Brochier et al.2005; Rinke et al.2013), presumably because the pre-tRNAs in this organism are transcribed without leader sequences (Randau, Schröder and Söll 2008) (Fig. 1). Previous observations, in agreement with our phylogenomic analysis (Fig. 1), identified all five RPPs in one or more members of Euryarchaeota, Crenarchaeota, Thaumarchaeota, Nanoarchaeota (except for N. equitans), Bathyarchaeota, Nanohaloarchaeota, Diapherotrites and Lokiarchaeota (Samanta et al.2016). The RPP distribution in Thaumarchaeota is similar, with POP5 and RPP30 missing in some genomes whereas RPP21 homologues are absent in some Thermoproteales strains (Thermoproteus uzoniensis, Pyrobaculum arsenaticum and V. distribute) as previously reported by Lai et al. (2010a). In addition, RPP30 homologues are not found in genomes from the Diaforachaea group. However, the occurrence of the other partners suggests that these subunits may have diverged beyond the sensitivity of the similarity search methods. This is supported by the observation that RPP30 is less conserved than the other RNase P subunits (Samanta et al.2016). Last of all, it has also been proposed that these RPPs arose early in archaeal phylogeny and that putative ancestral archaeal RPP genomic loci include genes encoding other core cellular machineries including open reading frames for several ribosomal, exosomal and proteasomal proteins (Samanta et al.2016). The splicing endonuclease EndA (EC 3.1.27.9 transferred to EC 4.6.1.16) The archaeal RNA-splicing endonuclease, termed EndA, is a multimeric enzyme with variable subunit compositions (Fig. 2B, Table 1A) that recognises a specific RNA structural motif named Bulge-Helix-Bulge (BHB) which consists in its canonical form of a four-base-pair central helix flanked by two juxtaposed three-nucleotide bulges in a helical context with at least two adjacent base pairs (Fig. 3B) (Thompson and Daniels 1990; Marck and Grosjean 2003). Asymmetric endoribonucleolytic cleavages occur within the bulges producing two RNA molecules bearing a 3΄-end 2’, 3΄-cyclic phosphate and a 5΄-end hydroxyl group. EndA is strictly conserved among the archaeal phylogeny (Fig. 1) and catalyses critical steps of tRNA and rRNA processing (described in section ‘Processing pathways of archaeal ncRNAs’). Overall, archaeal EndA enzymes, classified upon their subunit composition, present the same structure (Fig. 2B), suggesting an accommodation over time to diverse canonical and non-canonical BHB splicing motif substrates (Fig. 3B) (Marck and Grosjean 2003). Their different quaternary structures that may coincide with the diversity of substrates observed within different archaeal groups could have arisen by gene duplication and subsequent subfunctionalisation (Tocchini-Valentini, Fruscoloni and Tocchini-Valentini 2005). The homotetrameric α4 form (Fig. 2B) with potentially four active catalytic sites is predominantly found in Euryarchaea (Fig. 1) but also, interestingly, in the recently discovered Asgard group described to be closely branched to Eukarya (Zaremba-Niedzwiedzka et al.2017). In contrast, the homodimeric (α2) form, predominantly found in Euryarchaeota, and the heterotetrameric (α2ß2) configuration, present mostly in Crenarchaeota and only exceptionally in Nanoarchaeota or the euryarchaeon Methanopyrus kandleri (Calvin and Li 2008), possess only two active catalytic sites (Fig. 2B). Briefly, the α2 homodimeric or α4 homotetrameric forms are only capable of processing canonical BHB motifs (Lykke-Andersen and Garrett 1997; Li, Trotta and Abelson 1998; Li and Abelson 2000) (Fig. 3B), while the α2ß2 heterotetrameric variant in addition can bind and process non-canonical BHB motifs (Fig. 3B). A fourth type of splicing endonuclease, the ε2 homodimer, with a similarly relaxed substrate recognition was described in the deeply branching Archaea ARMAN-1 and ARMAN-2 (Fujishima et al.2011; Hirata et al.2012) (Fig. 1). Finally, comparative sequence analysis of the subunits revealed strong similarity with two protein subunits, Sen34p and Sen2p, of the eukaryal heterotetrameric tRNA splicing endonuclease suggesting an ancient origin of EndA splicing enzymes before the emergence of Eukarya (Trotta et al.1997; Hirata et al.2012) (see Lopes et al.2015 for review) (Table 2). aRNase Z (EC 3.1.26.11) From respective sequence similarities, the RNase Z enzymes were classified as belonging to the family of metal-dependent β-lactamases (Aravind 1999) (Table 2). RNase Z homologues are present in organisms of all three domains of life (Späth et al.2008) and ubiquitously in archaeal genomes with the exception of Nanopusillus acidilobi 605887 (Fig. 1). RNase Z exists in two forms: a short form (250–350 aa) present in all domains of life and a long form (750–900 aa) restricted to Eukarya that probably arose from a gene duplication event (Vogel et al.2005; Ma et al.2017). Being a metallo-β-lactamase, RNase Z enzymes exhibit the characteristic αβ/βα fold of sandwiched β-sheets and in the catalytic centre the characteristic histidine motif (HxHxDH) and additional histidine and aspartate residues (Fig. 2C, Table 1A) (Aravind 1999; Ishii et al. 2005, 2007; Li de la Sierra-Gallay, Pellegrini and Condon 2005). In addition, RNase Z has a flexible arm (coined the exosite) that is inserted between the N- and C-terminal halves and protrudes from the protein body (Ishii et al.2007). Crystal structures of archaeal RNase Z have not yet been resolved; a model is shown in Fig. 2C. The short RNase Z dimer binds the pre-tRNA with one monomer by clamping the exosite onto the tRNA elbow by hydrogen bonds to the backbone of mainly the T-arm and acceptor stem (Ishii et al.2005; Li de la Sierra-Gallay et al.2006; Zhang and Ferré-D’Amaré 2016). Structures resembling these two stacked helices are in fact the minimal substrate of most RNase Z enzymes with the exception of the Haloferax volcanii RNase Z that also requires the D-arm (Schierling et al.2002; Schiffer, Rösch and Marchfelder 2002; Ishii et al.2005). Catalysis that depends on the coordination of Zn2+ ions generates a tRNA with a 3΄-end hydroxyl group (Späth, Canino and Marchfelder 2007). For the M. jannaschii, P. furiosus, H. volcanii, P. aerophilum and Thermoplasma acidophilum RNase Z enzymes, an in vitro tRNA-processing activity was sought (Schierling et al.2002; Schiffer, Rösch and Marchfelder 2003; Hölzle et al.2008; Späth et al.2008). These enzymes show only small differences in their biochemical requirements, in the cleavage site position and in substrate specificity. It is interesting to note that for RNase Z of H. volcanii a high-salt concentration inhibits its activity, contrasting to the organism's halophilic life style (Späth et al.2008). While the maturation state of the pre-tRNA influences the processing activity of RNase Z, the outcome is species and tRNA specific. Intron-containing precursors and substrates with a long 5΄-leader are not processed by H. volcanii RNase Z but can be cleaved by P. furiosus RNase Z (Späth et al.2008). The presence of a CCA in the 3΄-trailer inhibits cleavage by T. acidophilum RNase Z but does not influence P. aerophilum and M. jannaschii RNase Z activities (Schiffer, Rösch and Marchfelder 2003; Späth et al.2008). aNob1 (EC 3.1.) Several archaeal homologues of eukaryotic ribosome assembly factors have been identified and are valued as models for functional and structural studies (Veith et al.2012; Hellmich et al.2013). This is the case for the eukaryotic endoribonuclease Nob1 (Table 2). Nob1 of P. horikoshii (Pho) efficiently cleaves RNA substrates encompassing a D-site, in vitro in a manganese-dependent manner. In addition, the structure of Pho-Nob1, solved by nuclear magnetic resonance spectroscopy, revealed the presence of a PIN (PilT N-terminus) domain typical of many nucleases and of a zinc ribbon domain (Table 1A). These two domains are structurally connected by a flexible linker (Fig. 2C) (Veith et al.2012). However, Pho-Nob1 lacks a large insertion in the PIN domain and a long C-terminal tail that are both typical of eukaryotic Nob1 (Veith et al.2012). Further studies showed that the structure and function of Nob1 are conserved in all the major branches of Archaea with the exception of Korarchaea and some DPANN with incomplete genomes (Fig. 1). Nevertheless, it should be noted that shorter CTDs are observed in some Crenarchaeota such as certain Thermoproteales (13 out of 17 genomes), the Acidilobales (2/2), the Fervidicoccales (1/1) and some Desulfirococcales (3/15). RNase HII (EC 3.1.26.4) RNA-primed replication intermediates in archaeal cells are similar to those formed during eukaryotic DNA replication (Kochiwa, Tomita and Kanai 2007). RNase H enzymes play a crucial role in DNA replication by removing the RNA primer of Okazaki fragments, and in DNA repair by removing the occasional single ribonucleotides incorporated in the DNA (see Tadokoro and Kanaya 2009 for review). It was recently shown that initiation of the misincorporated ribonucleotide repair pathway in Archaea relies on RNase HII (Heider et al.2017). The RNase HII gene is conserved in all complete archaeal genomes with the exceptions of Nitrosopumilus sediminis and N. koreensis AR1 (Fig. 1, Table 1A). The RNase HI genes have a lower occurrence and a discontinuous distribution (Kochiwa, Tomita and Kanai 2007). This result is consistent with an essential and non-overlapping function of the RNase HII in archaeal genomes. RNase H enzymes which cleave the P-O 3΄-bond of the RNA strand embedded in RNA/DNA hybrid substrates with a two-metal-ion catalysis mechanism vary greatly in domain structures and substrate specificities (see Tadokoro and Kanaya 2009 for review) (Fig. 2C). RNase H from Bacteria and Archaea have been classified into type 1 RNase H (RNase HI) and type 2 RNase H (RNase HII) based on differences in their amino acid sequences (Kochiwa, Tomita and Kanai 2007) (Table 2). The structures of M. jannaschii (Lai et al.2000), Thermococcus kodakarensis (Muroya et al.2001) and Archaeoglobus fulgidus (Bubeck et al.2011) share a main chain fold, termed the RNase H-fold, and steric configurations of the four acidic active-site residues (Fig. 2C). In addition, it has been reported that the 3D structure of RNase H is similar to that of the PIWI domain of the P. furiosus argonaute-like (Ago) protein suggesting an evolutionary relantionship between the RNAse H and the Ago proteins in prokaryotes. Interestingly, in the presence of Mn2+ ions, Pf RNAse HII digests RNA–RNA duplexes (Kitamura et al.2010). aPelota (EC 3.1.) Drosophila melanogaster Pelota and Saccharomyces cerevisiae DOM34, initially identified as important proteins for meiotic cell division, were recently shown to be key players in the no-go decay (NGD) surveillance pathway by which mRNAs containing stalled ribosomes are degraded by endoribonucleolytic cleavage (see Simms, Thomas and Zaher 2017 for a review). Pelota-like proteins (or aPelota), homologues of eukaryal Pelota, were also identified in archaeal genomes (Table 1A). The protein aPelota from Sulfolobus solfataricus shares about 20% identity with its eukaryal counterparts (Ragan et al.1996) (Table 2) and is widely conserved in Archaea (Atkinson, Baldauf and Hauryliuk 2008) (Fig. 1). Recently, the crystal structures of archaeal Pelota proteins from T. acidophilum, A. fulgidus and S. solfataricus revealed three structural domains (Lee et al.2007, 2010). The central and CTDs (Fig. 2C) share similarities with the eukaryotic translational release factor eRF1, while the N-terminal domain has a weak similarity with RNA-binding Sm-fold proteins (described in section ‘Archaeal small RNA-binding proteins’) (Fig. 3C). In addition, these structural studies reported an interdomain structural plasticity of aPelota proteins suggesting that large conformational changes are essential for their functions (Lee et al.2010). It was shown that aPelota from T. acidophilum has a ribonucleolytic activity (Lee et al.2007). In addition, the archaeal translational elongation factor EF1α binds to archaeal RF1 and aPelota in a GTP-dependent manner (Kobayashi et al.2010; Kobayashi, Ishitani and Nureki 2013). It was proposed that a positively charged patch of aPelota domain 1 interacts with the decoding centre of the ribosome to specifically recognise the empty site A of the stalled ribosome and thereby induce the destabilisation of the ribosome (Kobayashi et al.2010; Becker et al.2012; Kobayashi, Ishitani and Nureki 2013). Cas6 The CRISPR-Cas defence mechanism involves repeat-associated mysterious proteins (RAMPs) that are classified into three large groups: Cas5, Cas6 and Cas7. These proteins form large RNP complexes together with CRISPR RNA (crRNA) dedicated to the destruction of invading DNA or RNA (Makarova et al. 2011, 2015). In type I and type III systems, the Cas6 protein processes the crRNA precursor into crRNAs of about 70–80 nucleotides length with a 5΄-hydroxyl and a 3΄- or cyclic 2΄-3΄- phosphate group (Hale et al.2008, 2009; Makarova et al.2015) (see Charpentier, van der Oost and White 2013 for review). Note that in type I-C systems, this reaction is catalysed by Cas5d (Garside et al.2012; Nam et al.2012). Cas6 protein sequences show limited conservation with only two common motifs: the ferrodoxin fold or RRM (RNA recognition motifs) domains and a glycine-rich motif located at the C-terminus (Table 1A, Fig. 2C). Cas6 is a metal-independent nuclease and the predicted active site of the enzyme shows similarities to the active site of the archaeal EndA endoribonuclease. Interestingly, the nature of amino acids located in the active site differs in the Cas6 proteins from the different subtypes. Some Cas6 proteins act as homodimers, whereas others are active as monomers. Several Cas6 crystal structures have been obtained and reveal conserved structural features (see Hochstrasser and Doudna 2015 for a review). Cas6 enzymes contain two RAMP domains that form RRM folds. The C-terminal RRM fold consists of a glycine-rich loop (G-loop), a groove-binding element and a β-hairpin. The G-loop binds and stabilises the RNA substrate, the groove-binding element interacts with major groove of the RNA structure and the β-hairpin is important for positioning the scissile phosphate. In the cleavage reaction, Cas6 interacts with the 5΄-region of the CRISPR repeat and cleaves at a specific site within the 3΄-region. Substrate recognition of some of the Cas6 proteins evolved together with the CRISPR-repeat sequence, and certain Cas6 proteins can only recognise specific repeat sequences. Depending on the ability of the repeat sequence to form stem-loops, Cas6 proteins either interact with a preformed stable stem-loop structure or fold an unstructured RNA into a stem-loop structure for their cleavage (see Charpentier, van der Oost and White 2013 for a review). Most prevalent archaeal 3΄-to-5΄ exoribonucleases 3΄-end-RNA trimming activities are critical in all three domains of life. In Archaea, this activity is carried out by two types of 3΄-to-5΄ exoribonucleases (Table 1B). The archaeal exosome machinery, homologous to the eukaryotic exosome, is commonly found in Archaea. In contrast, the archaeal RNase R-like enzyme, homologous to the bacterial RNase R, is only sporadically encountered (Fig. 1, Table 2). aExosome (EC 3.1.13.1) More than a decade ago, the detection of an operon encoding homologues of Rrp4, Rrp41 and Rrp42, eukaryal rRNA-processing proteins that form the core of the eukaryal exosome, in most archaeal genomes strongly suggested the existence of a conserved archaeal 3΄-end RNA-degrading machinery (Koonin, Wolf and Aravind 2001) (Table 2). The complete set of exosome protein subunits is conserved in most archaeal genomes with the exception of species from Methanococci, Methanomicrobiales, Haloferacales and Halobacteriales (Fig. 1) (Evguenieva-Hackenberg et al.2014). Many studies using different archaeal model organisms, in particular Sulfolobales and Pyrococcales, allowed a thorough characterisation of the archaeal exosome by determining its structures and in vitro enzymatic activities (see Hartung and Hopfner 2009; Chlebowski et al.2013; Evguenieva-Hackenberg et al.2014 for recent reviews). Briefly, the archaeal exosome (aExosome), commonly composed of a nine-subunit core, appears as heterogeneous complexes (Fig. 4A). Consistently, three Rrp41 and Rrp42 subunits, belonging to the PH RNase family, form a hexameric barrel-like structure with a narrow RNA entrance pore to which a variable heterotrimeric RNA-binding cap of Rrp4 and/or Csl4 subunits is tightly bound (Evguenieva-Hackenberg et al.2003; Ramos et al.2006; Witharana et al.2012) (Table 1B). In this context, the Rrp4 subunit confers strong poly(A) specificity to the exosome while Csl4 allows the binding of a aDnaG subunit to the exosome (Witharana et al.2012; Hou, Klug and Evguenieva-Hackenberg 2013). These cap RNA-binding subunits harbour different types of RNA-binding domains (S1-domain, KH-motif and Zn-ribbon) (Fig. 4A). Three active sites in the lumen of the barrel structure carried by the Rrp41 subunits are responsible for the phosphorolytic catalysis of single-stranded RNA from the 3΄-end and the release of nucleoside 5΄-diphosphates (NDPs) (Lorentzen et al. 2005, 2007; Hartung et al.2010). The aExosome is also able to perform the reverse reaction by using NDPs to synthetise heteropolymers, such as adenine-rich tails, at the 3΄-end of RNAs (Portnoy et al.2005). The 3΄-to-5΄ exoribonucleolytic and RNA-tailing activities can be modulated with Pi concentration and depend on the ATP:ADP ratio.(Evguenieva-Hackenberg et al.2008). Figure 4. View largeDownload slide Domain organisation and structure of archaeal exoribonuclease and β-CASP ribonuclease families. Structure (PDB accession numbers) and corresponding model organism references are reported in Table 1B and marked by an asterisk. Domains correspond to Pfam annotation (Pfam database http://pfam.xfam.org/). Catalytic sites are highlighted with yellow areas. (A) Structures and domains of the archaeal 3΄-to-5΄ exoribonucleases are shown. Upper panel: archaeal exosomes are usually formed by the association of three modules composed of two subunits Rrp41 and Rrp42 of the PDX protein (only Rrp41 is active). Two different archaeal exosome structures are shown, one with a cap formed by three Csl4 subunits and the other by three Rrp4 subunits. Lower panel: Phyre2-built structural models of the three different forms (annotated a to c) of archaeal RNase R are displayed. The a form (RNase Ra) is restricted to the RNB domain and the b and c forms (RNaseb and RNasec) display 3΄-extension RNA binding domains. (B) Structures and domains of the major types of β-CASP ribonucleases are displayed. Only the structures of aCPSF1 and aRNase J have been resolved. Phyre2-built structural model for S. acidocaldarius aCPSF2 is shown. aCPSF1 is a unique β-CASP ribonuclease carrying 5΄-to-3΄ exo- and endoribonucleolytic activities with a single catalytic site and composed of a β-lactamase/ β-CASP core and an N-terminal extension with two KH motifs. In contrast, aCPSF2 and aRNase J are restricted to the β-lactamase/ β-CASP core and carry only a 5΄-to-3΄ exoribonucleolytic activity. Figure 4. View largeDownload slide Domain organisation and structure of archaeal exoribonuclease and β-CASP ribonuclease families. Structure (PDB accession numbers) and corresponding model organism references are reported in Table 1B and marked by an asterisk. Domains correspond to Pfam annotation (Pfam database http://pfam.xfam.org/). Catalytic sites are highlighted with yellow areas. (A) Structures and domains of the archaeal 3΄-to-5΄ exoribonucleases are shown. Upper panel: archaeal exosomes are usually formed by the association of three modules composed of two subunits Rrp41 and Rrp42 of the PDX protein (only Rrp41 is active). Two different archaeal exosome structures are shown, one with a cap formed by three Csl4 subunits and the other by three Rrp4 subunits. Lower panel: Phyre2-built structural models of the three different forms (annotated a to c) of archaeal RNase R are displayed. The a form (RNase Ra) is restricted to the RNB domain and the b and c forms (RNaseb and RNasec) display 3΄-extension RNA binding domains. (B) Structures and domains of the major types of β-CASP ribonucleases are displayed. Only the structures of aCPSF1 and aRNase J have been resolved. Phyre2-built structural model for S. acidocaldarius aCPSF2 is shown. aCPSF1 is a unique β-CASP ribonuclease carrying 5΄-to-3΄ exo- and endoribonucleolytic activities with a single catalytic site and composed of a β-lactamase/ β-CASP core and an N-terminal extension with two KH motifs. In contrast, aCPSF2 and aRNase J are restricted to the β-lactamase/ β-CASP core and carry only a 5΄-to-3΄ exoribonucleolytic activity. Despite recent advances in dissecting the aExosome composition and activities through in vitro structural and biochemical studies, its contribution in specific biological pathways remains unknown, especially in RNA maturation and RNA decay mechanisms. First clues may come from observations showing an interaction with the splicing enzyme EndA in M. thermautotrophicus (Farhoud et al.2005), the S1-domain of the RNA-binding protein TK2227 in T. kodakarensis (Li et al.2010) and the translational elongation factor EF1α in S. solfataricus (Witharana et al.2012). Moreover, the heterogeneity of the exosomal complexes and its localisation at the cell periphery as observed in S. solfataricus may reflect a multilayered regulation of its 3΄-tailing or degradation activities upon substrate recognition (Witharana et al.2012). The physiological relevance of the composite RNA-binding cap in mechanisms of substrate selection is still lacking. Its understanding may come from the analysis of archaeal exosomes from species in which the set of subunits is incomplete. It will be interesting to understand how aDnaG is recruited in Diaforarchaea species in which the Cls4 subunit appear to be absent. Nevertheless, it should be noted that Csl4 is encoded in a separate operon to Rrp4, Rrp41 and Rrp42 which precludes the identification of potentially highly divergent copies in those genomes. In addition, monitoring how the absence of the Rrp4 subunit in Asgard genomes impacts aExosome activity may help identifying the specific functions of each cap subunit. Finally, the absence of the core subunit Rrp42 in some DPANN may suggest a different aExosome composition in these strains (Fig. 1). Interestingly, in species from the Methanomicrobiales, Haloferacales and Halobacteriales group, the absence of the exosome seems to be compensated by the presence of other 3΄-to-5΄ exoribonucleases, namely RNase R-like enzymes (see below, Fig. 1). However, in the case of the Methanococci group, no enzyme with an equivalent activity could be detected suggesting that in these strains, a 3΄-to-5΄ exoribonucleolytic activity may not exist and may be compensated by other endo- or 5΄-to-3΄ ribonucleolytic machineries (Clouet-d’Orval et al.2015). aRNase R (EC 3.1.13.1) As mentioned above, in some cases, the 3΄-to-5΄ exoribonucleolytic activity of an archaeal cell is not dictated by an exosome complex but instead by a bacterial-like RNase R enzyme (Portnoy and Schuster 2006) (Figs 1 and 4A, Table 2). This is the case for halophiles and methanococcales in which the operon encoding the exosomal subunits is missing (Fig. 1). Briefly, RNase R-like enzymes which are characterised by an RNB domain with a unique αβ-fold are part of the RNase II/RNB family present in the three domains of life (Fig. 4A, Table 2). It is now well established that bacterial RNase R encoded by the rnr gene and composed of the RNB domain flanked by RNA-binding domains is able to degrade structured RNAs from their 3΄-ends to their 5΄-ends to perform important functions in the cell from RNA quality control to virulence (for a review, see Arraiano et al.2013). Most knowledge on archaeal RNase R (aRNase R) comes from in vitro biochemical studies carried out with that from H. volcanii in which the rnr gene has been shown to be essential (Portnoy and Schuster 2006; Matos et al.2012). Interestingly, our searches for RNase R-like members in archaeal genomes identified at least three types of domain organisation (arbitrarily named a, b and c here) with none resembling the bacterial enzyme (Figs 1 and 4A and Table 1B). In the halophile group, aRNase Ra is restricted to the RNB domain with no extra RNA-binding domain. In contrast, two other forms, RNase Rb and RNase Rc, found in some methanococcales, contain an extra CTD with RNA-binding properties (Fig. 4A). These different forms of aRNase R have certainly resulted from horizontal gene transfer (HGT) from bacterial genomes. More specifically, we used the archaeal proteins annotated as RNase R as a query to search the NCBI non-redundant bacterial database with blastp. The 200 best hits were retrieved and aligned with the archaeal sequences (mafft program) and the sequence redundancies were removed with Jalview (98% identity threshold). A phylogenetic tree, computed with PhyML, revealed that archaeal sequences were dispersed as several clusters in the bacterial tree, suggesting multiple independent HGTs from bacterial to archaeal genomes. RNaseRa homologues are found in all analysed Haloferacale and Halobacteriale strains (Fig. 1). All sequences are clustered in a subtree connected to the bacterial sequences with a long branch. This pattern is in agreement with the presence of an RNaseRa homologue in the last common ancestor of Haloferacales and Halobacteriales. This ancestral gene was probably acquired by lateral gene transfer from Bacteria to Archaea. Furthermore,three RNase Rb homologue groups (Methanomicrobiales strains, Methanosaeta concilii GP6, Methanocella paludicola SANAE, Methanocella arvoryzae MRE50) are observed (Fig. 1). This is in agreement with three independent HGT events from bacterial strains. Finally, the RNase Rc homologues are found in one subtree that includes one sequence from Methanocellale strain (Methanocella conradii HZ254) and sequences from Methanosarcinale strains (Fig. 1). The location of the M. conradii sequence as an outgroup may indicate an initial HGT from Bacteria to Archaea followed by a second HGT between Archaea. However, more data are required to confirm the evolutionary origin of archaeal aRNase R homologues and to understand how they work in RNA processing and RNA decay, and whether they recruit other specific or common factors. The universal β-CASP ribonuclease family In the last few years, a prevalent family of enzymes, namely the β-CASP ribonucleases, has emerged as central in providing additional ribonucleolytic activities in archaeal cells (for a review, see Clouet-d’Orval et al.2015) (Tables 1A and 1B). Briefly, members of the β-CASP ribonuclease family are of major importance in RNA-processing and RNA decay pathways in Eukarya and Bacteria (for a review, see Dominski, Carpousis and Clouet-d’Orval 2013) (Table 2). Typically, β-CASP members harbour a β-lactamase domain, composed of a four-layered αβ/βα architecture, in which a so-called β-CASP domain is inserted (Fig. 4B). Several crystal structures of bacterial, eukaryal and archaeal members have reported an identical structural bipartite topology with a catalytic site formed at the interface of the two domains and involving conserved histidine and aspartic acid residues that coordinate two Zn2+ ions. Importantly, β-CASP ribonucleases have the propensity to carry dual, 5΄-to-3΄ exo- and endo-ribonucleolytic activities with a single active site (Li de la Sierra-Gallay et al.2008; Dorléans et al.2011; Newman et al.2011). Archaeal β-CASP members have been identified by in-depth phylogenomic analyses that revealed three major archaeal β-CASP clusters designated aCPSF1, aCPSF2 and aRNase J, based upon their close relationship to either bacterial RNase J or eukaryal CPSF73 (cleavage polyadenylation specific factor) (Dominski, Carpousis and Clouet-d’Orval 2013; Phung et al.2013; Clouet-d’Orval et al.2015) (Table 2 and Table 1A). Interestingly, the congruence of phylogenetic trees based on aCPSF1, aCPSF2 and aRNase J sequences and a set of concatenated conserved protein sequences reflecting the evolution of the Archaea showed that the three major archaeal β-CASP clusters have been inherited vertically (Clouet-d’Orval et al.2010; Phung et al.2013). This indicates that β-CASP ribonucleases are ancient enzymes that existed in the last universal common ancestor before the separation of the actual three domains of life. Features of each β-CASP ribonucleases clusters are briefly introduced below. aCPSF1 (CPSF73, EC 3.1.4.1) Members belonging to the aCPSF1 cluster typically contain, in addition to the β-CASP/β-lactamase core domains, a supplementary N-terminal domain composed of two consecutive tandem KH motifs. Three crystal structures of aCPSF1 from P. horikoshii, Methanosarcina mazei and M. thermautotrophicus show a tripartite architecture with the KH-domains link to the central core via a long helical linker and two molecules per asymmetric unit (Nishida et al.2010; Mir-Montazeri et al.2011; Silva et al.2011) (Fig. 4B, Table 1A). It was shown that P. abyssi aCPSF1 is dimeric in solution with a highly conserved short C-terminal segment serving as interface (Phung et al.2013) (see Dominski, Carpousis and Clouet-d’Orval 2013; Clouet-d’Orval et al.2015 for reviews). The in vitro enzymatic activity of thermococcales and methanococcales aCPSF1 revealed a dual activity with an endoribonucleolytic activity at single-stranded CA site and a 5΄-to-3΄ exoribonucleolytic activity restricted to 5΄-monophosphorylated RNA substrates (Levy et al.2011; Phung et al.2013). These activities are comparable to those displayed by the eukaryal CPSF73 (see Dominski, Carpousis and Clouet-d’Orval 2013 for review). Remarkably, an aCPSF1 member is encoded in each single archaeal genome (Fig. 1). Together, these observations suggest that the dual activities of aCPSF1 are at the centre of critical and conserved metabolic pathways important for the archaeal cell physiologies which remain to be determined. In addition, a small group of β-CASP proteins, aCPSF1b, closely related to aCPSF1 but lacking the N-terminal extension (Phung et al.2013), is restricted to the Methanococci group in which no 3΄-to-5΄ exoribonucleolytic machineries could be detected by homology searches (Fig. 1). A distantly related sequence is also present in Korarchaeum cryptofilum OPF8. aCPSF2 (CPSF73, EC 3.1.4.1) The aCPSF2 group was identified as a cluster of less conserved sequences compared to the aCPSF1 cluster (Phung et al.2013; Clouet-d’Orval et al.2015) (Table 1B, Fig. 4B). Members of this group, consistently found in the TACK/Proteaoarchaeota and more sporadically in Euryarchaeota, are architecturally restricted to the β-CASP/β-lactamase core domains except for the S. solfataricus member which possess an N-terminal extension (for reviews, see Dominski, Carpousis and Clouet-d’Orval 2013; Clouet-d’Orval et al.2015). To date, no aCPSF2 structures have been reported (a model is shown in Fig. 4B), and curiously enough, the 5΄-to-3΄ ribonucleolytic activity detected in vitro for Sulfolobales members requires Mg2+ ions in contrast to other β-CASP ribonucleases (Levy et al.2011). In S. acidocaldarius, the gene encoding aCPSF2 is not essential and its deletion is not associated with a modification of the growth rate (Märtens et al.2013), meaning that, in these conditions, the exoribonuclease aCPSF2 does not affect the overall cellular metabolism. aRNase J (EC 3.1.13.B1) More than a decade ago, homologues of bacterial RNase J were proposed to be encoded in archaeal genomes (Even et al.2005). It is now clear that archaeal RNase J (aRNase J) exists throughout most Euryarchaea with the exception of the Diafoarchaea and Archeoglobi groups (Fig. 1, Table 1B). In contrast to bacterial RNase J proteins which harbour a CTD, aRNase J enzymes are restricted to the β-CASP/β-lactamase core domains (Fig. 4B) (Clouet-d’Orval et al.2010). An aRNase J 5΄-to-3΄ exoribonucleolytic activity was first identified for recombinant proteins from Thermococci and Methanococci. In vitro, these enzymes show a high processivity, no detectable endoribonucleolytic activity under the conditions tested and a greatly reduced activity in degrading RNA substrates with 5΄-triphosphate ends (Clouet-d’Orval et al.2010; Levy et al.2011). Recently, the first RNA-free and RNA-bound structures of aRNase J from Methanolobus psychrophilus provided detailed mechanistic insights. Furthermore, the unique conserved archaeal loops I and II were found to be involved in the in vitro ribonucleolytic activity and in RNA interaction of aRNase J (Clouet-d’Orval et al.2010; Zheng et al.2017). Finally, distinct dimerisation/tetramerisation patterns were observed for archaeal and bacterial RNase Js (Zheng et al.2017). In conclusion, the roles of archaeal β-CASP ribonucleases in specific biological mRNA decay and RNA quality control mechanisms remain to be elucidated. The challenge will be to identify how β-CASP ribonucleases are specifically recruited and if they are part of multiprotein complexes as are their bacterial and eukaryal counterparts. The expected outcome will clarify the roles in archaeal cell physiology of 5΄-to-3΄ exo- and endo-ribonucleolytic activities carried by this conserved enzyme family. Archaeal sRNA-binding proteins In RNA biology processes, the RNA-associated proteins of the Sm/Hfq and ribosomal L7Ae superfamilies, sometimes called RNA chaperones, have an important role in RNP complex assembly and have emerged as key players in myriad of RNA-processing pathways in Eukarya and Bacteria. Archaeal Sm-like proteins Members of the Sm-like/Hfq proteins form a superfamily of RNA-binding proteins of ancient related origin (Mura et al.2013). They contain a common bipartite sequence, known as the Sm-domain, consisting of two Sm motifs connected by a region of variable length and motif. And while the proteins do not show a high similarity at the primary structure level, they share a tertiary structure and a multimeric quaternary architecture that supports interactions with partner molecules. Another conserved characteristic of the Sm and Sm-like proteins is their preference for single-stranded uracil-rich RNAs. The ubiquitous eukaryal Sm proteins and their Sm-like paralogues (>18 paralogues) are molecular scaffolds for the assembly of RNP complexes. They were first identified in spliceosomal RNP complexes where they associate with uridine-rich small nuclear RNAs. Hfq, the bacterial branch of the Sm superfamily, was discovered 50 years ago as a host replication factor for the bacteriophage Qβ (Franze de Fernandez, Eoyang and August 1968). In many bacterial organisms, Hfq is perceived primarily as the core component of a global post-transcriptional network involving small regulatory RNAs (Nielsen et al.2007). It is now proposed that homologues exist in the three domains of life and several recent studies try to overcome the gaps in our knowledge on the relevance of this RNA-associated protein family in Archaea. Archaeal Sm-like/Hfq superfamily proteins were also detected by database searches (Salgado-Garrido et al.1999). While their biological functions remain to be discovered, they were fundamental in elucidating the biochemical and structural properties of their eukaryal counterparts. To unify the nomenclature, the archaeal homologues have recently been termed SmAPs (Mura et al.2013). These proteins share the Sm-fold, common to all members of the Sm superfamily even for highly divergent sequences, but in general do not contain any C-terminal extension as observed in some eukaryotic Sm-like and some bacterial Hfq (Mura et al.2013) (Fig. 3C). Archaea often possess two SmAP paralogues, SmAP1 and SmAP2. In the euryarchaeal Pyrococcales and Halophiles, only one Sm-like protein has been identified. This is orthologous to SmAP1. There is a high-sequence similarity between members of the SmAP1 family, with up to 60% sequence identity (Törö et al.2001). The identity level drops to 30% or less (Törö et al.2001) between SmAP1 and SmAP2 proteins from the same organism. Crenarcheota, like S. solfataricus (Sso), S. tokodaii (Sto) and P. aerophilum (Pae), can possess an additional Sm-like paralogue, SmAP3 (Mura et al.2003b, 2013). SmAP3 proteins are particular in possessing a longer CTD (Mura et al.2003b). Finally, Methanococcus jannaschii (Mja) contains an Hfq-like protein (Nielsen et al.2007) rather than a conventional Sm-like protein (Salgado-Garrido et al.1999). SmAP1 proteins The crystal structures of many archaeal SmAPs have been resolved since the beginning of the 2000s. In particular, the crystal structures of SmAP1 cyclic homoheptamers of crenarchaeota P. aerophilum (Pae-SmAP1/PDB code 1ISF) (Mura et al.2001), and of euryarchaeota A. fulgidus (Afu-SmAP1/PDB code 1I4K) (Törö et al.2001), P. abyssi (Pab-SmAP1/PDB code 1H64) (Törö et al.2002) and M. thermautitrophicum (Mth-SmAP1/PDB code 1IS1) (Collins et al.2001) (Fig. 3C) strengthen the proposed model for human Sm core domain (Kambach, Walke and Nagai 1999) by providing the first atomic resolution of Sm monomers in an intact ring. SmAP1 monomers adopt a barrel-type structure consisting of an N-terminal α-helix followed by a strongly bent five-stranded β-sheet (Collins et al.2001; Mura et al.2001; Törö et al.2001, 2002). The highly curved nature of the β-sheet allows the formation of a compact hydrophobic core. The cyclic heptameric ring, composed of seven monomers (7-mer), involves interactions between the β4 and β5-strands of adjacent monomers to form a continuous β-sheet in the oligomer. The loop L4 that links the strands β3 and β4 and that define the distal face varies more in length and amino acid sequence than any other loops. Most SmAP1s exhibit a short loop L4 when compared to eukaryotic Sm-like proteins and do not include extended N- and C-termini, with the exception of Mth-SmAP1 that contains a slightly longer N-terminal sequence when compared to other archaeal proteins (Collins et al.2001). The cyclic heptameric complex is extremely stable and can resist denaturing conditions, even in the absence of RNA, as shown for eukaryotic Sm-like but not Sm proteins. The residues in loops L2, L3 and L5 lie towards the central cavity of the ring and residues located in loops L3 and L5 form a conserved uridine-binding pocket (Törö et al.2001), also called internal binding site (Thore et al.2003). This central cavity is surrounded by a positively charged region on the loop L4/distal face (Törö et al.2001; Thore et al.2003). Beyond their capacity to form cyclic heptameric rings, SmAP1 proteins can also assemble into two stacked heptameric rings and, like Pae- and Mth-SmAP1s, can also polymerise into fibrillary ultrastructures (Mura et al. 2003a, b). This oligomeric plasticity of SmAPs, which is observed across the entire protein family and may have unsuspected biological significance, will not be further developed in this review (see Mura et al.2013 for review). Both Afu- and Pab-SmAP1 proteins were shown to preferentially bind U-rich RNA molecules (Törö et al.2001; Thore et al.2003). Binding of U5 or U7 oligonucleotides by Afu-SmAP1 (PDB code 1I5L) (Törö et al.2001) or Pab-SmAP1 (PDB 1M8V) (Törö et al.2002; Thore et al.2003), respectively, does not induce major structural changes and involves the conserved His37, Asn39 and Arg63 residues of the internal binding site of the same monomer. A second binding site, the external binding site, was identified in the crystal structure of the Pab-SmAP1/U7 complex (Thore et al.2003). This site consists of the residues Arg4 from the N-terminal α-helix and Tyr34 from the β2-strand of two monomers of two cyclic heptamers in head-to-head orientation. The Tyr34 residue is absolutely conserved in all the archaeal SmAP1 proteins (Thore et al.2003). Its substitution by a valine does not affect Pba-SmAP1 binding to an oligo(U), but it strongly reduced binding to the Eukaryotic Sm consensus ‘AAUUUUUGG’ RNA. These results suggest an interplay between the two RNA-binding sites where the external binding site stabilises additional nucleotides after a specific recognition of the U-rich sequence by the internal binding site (Thore et al.2003). SmAP2 proteins Crystal structures of SmAP2 proteins from S. solfataricus ((Märtens et al.2017); PDB code 4XQ3) and A. fulgidus ((Törö et al.2002); PDB code 1LJO) have also been resolved. Afu-SmAP2 has essentially the same monomer fold as Afu-SmAP1 and harbours the conserved uridine-binding pocket (with the exception of His37 which is replaced by Tyr), suggesting a common RNA-binding mode for both proteins. Band-shift experiments confirmed that Afu-SmAP2 has a clear preference for U-rich sequence (Achsel, Stark and Lührmann 2001). However, conversely to Afu-SmAP1, the oligomerisation of Afu-SmAP2 is strongly dependent on the pH and the presence of RNA molecules. Indeed, while the electron micrographs of an Afu-SmAP2/U10 complex show that in the presence of an oligo(U), Afu-SmAP2 can assemble into a cyclic heptameric ring (Achsel, Stark and Lührmann 2001), at acidic pH and in the absence of RNA, Afu-SmAP2s form a cyclic hexameric (6-mer) structure in the crystals that is not stable (Törö et al.2002). The opposite charge distribution of the L4/distal face that is predominantly positive in Afu-SmAP1 and negative in Afu-SmAP2 creates differences in the ionic interactions between monomers that could explain the different oligomerisation behaviour of Afu-SmAP1 and 2 (Törö et al.2002). Nonetheless, while Sso-SmAP1 and -SmAP2 also exhibit similar opposite charge distribution, Sso-SmAP2 proteins were shown to assemble into cyclic heptameric rings in the crystals (Märtens et al.2015). Finally, another clear difference between free Afu-SmAP1 heptamers and Afu-SmAP2 hexamers is the diameter of the central cavity of the Afu-SmAP2 (≈8Å) that is smaller than that of SmAP1 proteins (10–15Å) and that will not allow RNA binding without significant structural rearrangement (Törö et al.2002). As suggested by the electron micrographs of Afu-SmAP2/U10 complexes, it was proposed that assembly around a U-rich RNA sequence may lead to the formation of an Afu-Sm2 heptamer (Törö et al.2002). SmAP3 proteins SmAP3 from P. aerophilum (Pae-SmAP3) contains the conserved Sm-domain linked to a novel CTD of nearly 70 residues. The crystal structure of Pae-SmAP3 shows that these additional residues form a compact CTD with a novel mixed α/β fold (Mura et al.2003b). The strong sequence conservation of Pae-, Sto- and Sso-SmAP3 CTDs suggests that they will have a similar structure (Mura et al.2003b). A comparative structural analysis found weak structural similarity between the CTD of Pae-SmAP3 and a CTD of yeast TATA-box binding protein (Mura et al.2003b). Although it is unclear how the CTD of SmAP3 would modify its function (biochemical or structural), it suggests that SmAP3s could have novel nucleic acid binding properties and may be used for physiological functions entirely unrelated to RNA metabolism (Mura et al.2003b). Moreover, Pae-SmAP3 monomers assemble into an unexpected 14-mer structure composed of two heptamers with different conformations (apical or equatorial) that is perforated by a cylindrical pore and is bound to 14 cadmium ions (Mura et al.2003b). Interestingly, it was shown that Sso-SmAP1, Sso-SmAP2 and Sso-SmAP3 can co-precipitate, suggesting that in vivo, the Sso-SmAPs could assemble into hetero-oligomers (Märtens et al.2017). Similar behaviour was also previously observed for Afu-SmAP1 and Afu-SmAP2 (Törö et al.2001). While hetero-oligomeric SmAP assemblies remain to be characterised in Archaea, a differential composition could impact their biological functions as shown for eukaryal Sm-like hetero-heptamers (Mura et al.2013). Hfq-like proteins A homologue of bacterial Hfq was identified in the genome of M. jannashii which does not contain an open reading frame encoding a SmAP protein (Salgado-Garrido et al.1999; Nielsen et al.2007; Mura et al.2013). Like bacterial Hfq, Mja-Hfq contains the conserved Hfq Sm2 motif (YKHAI consensus) but possesses a shorter C-terminal tail and N-terminal α-helix compared to Hfq homologues of Escherichia coli (Eco) and Staphylococcus aureus (Nielsen et al.2007; Mura et al.2013). The charge distribution on loop L4 of Eco-Hfq and Mja-Hfq is also different as Mja-Hfq L4 is predominantly negatively charged (Nielsen et al.2007; Mura et al.2013). The resolution of Mja-Hfq crystal structure (PDB code 2QTX) identified a cyclic hexameric complex also characteristic of Hfq proteins (Nielsen et al.2007). Mja-Hfq preserves the conserved uridine-binding pocket (also named distal binding site) but the distal adenine-binding site and lateral RNA-binding site show considerable structural changes compared to bacterial Hfq (Nikulin et al.2017). Despite these differences, Mja-Hfq is able to partially complement the pleiotropic phenotypes of Hfq knockout mutants in E. coli (Sittka et al.2009) and Salmonella enterica (Nielsen et al.2007). Moreover, Mja-Hfq forms a ternary complex with the Spot42 sRNA and its sucC mRNA target in vitro (Nielsen et al.2007). Finally, Hfq homologues were also discovered on four Thermococcales plasmids and three unrelated Methanococcales plasmids (Krupovic et al.2013). These archaeal Hfq proteins are particular and contain an additional N-terminal C2H2-type zinc finger domain suggesting an additional potential role in DNA or RNA binding (Krupovic et al.2013). Archaeal L7Ae ribosomal protein L7Ae is a member of the L7Ae/L30 protein family found in both Archaea and Eukarya. The members of the L7Ae/L30 protein family share the capacity to bind K-turn RNA motifs (Mao, White and Williamson 1999; Ban et al.2000; Kuhn, Tran and Maxwell 2002) (Fig. 3A). Archaeal L7Ae is homologous to the human 15.5kD protein and yeast Snu13 small nucleolar RNP core proteins and shares their capacity to bind the K-turn of C/D and H/ACA-box sRNAs (Kuhn, Tran and Maxwell 2002; Rozhdestvensky et al.2003). In contrast to 15.5kD/Snu13 proteins, L7Ae also has the ability to recognise Kink-loop (K-loop) motifs (Vidovic et al.2000; Hamma and Ferré-D’Amaré 2004; Nolivos, Carpousis and Clouet-d’Orval 2005; Li and Ye 2006; Soss and Flynn 2007; Gagnon et al.2010). An alignment of several archaeal L7Ae and eukaryotic 15.5 kD proteins uncovered a highly conserved RNA-binding region containing amino acid residues important for the binding of K-turn motifs (Gagnon et al.2010). The archaeal L7Ae protein is a component of the large ribosomal subunit of Haloarcula marismortui (Ban et al.2000) and serves as a core protein of the C/D and H/ACA box sRNP nucleotide modification complexes (Kuhn, Tran and Maxwell 2002; Tang et al.2002; Rozhdestvensky et al.2003; Charron et al.2004a; Baker et al.2005; Gagnon et al.2010). It is also part of the archaeal RNase P complex (Kouzuma et al.2003; Fukuhara et al.2006; Lai et al.2014; Oshima et al.2016) (see section ‘Ubiquitous endoribonucleases’). Moreover, recent evidence also suggests that L7Ae can regulate translation by direct binding to the 5΄-untranslated regions of some messengers, including its own transcript (Daume et al.2017). Finally, an in silico approach in P. abyssi identified three unknown small ncRNAs containing K-turn motifs as interacting partners of L7Ae (Phok et al.2011). Together, the presence of L7Ae in RNase P, C/D and H/ACA box sRNPs and ribosome in Archaea offer fascinating prospects for future integrative studies in understanding coordination in the regulation of macromolecular machineries involved in specific facets of translation. Processing pathways of archaeal ncRNAs Maturation of tRNAs Maturation of tRNAs is a highly conserved process in all three domains of life. The tRNA sequence is transcribed as a longer precursor (pre-tRNA) containing a 5΄-leader and a 3΄-end trailer sequence (Fig. 5A). Thus release of the mature tRNA requires processing at both transcript ends, and often post-transcriptional addition of a 3΄-CCA-end. Furthermore, extensive nucleoside modifications and, in some cases, removal of introns are needed. Figure 5. View largeDownload slide Processing pathways of transfer RNA precursors. (A) Maturation of a tRNA precursor in Archaea. The tRNA precursor is processed at the 5΄-end by RNase P, while the 3΄-end is generated by RNase Z cleavage. If not encoded by the tDNA, the CCA triplet is added by the tRNA nucleotidyl-transferase. If an intron (brown) is present within the tRNA, the Bulge-Helix-Bulge RNA (BHB) motif formed at the exon-intron junctions is recognised and cleaved by EndA. Subsequently, the tRNA-halves are joined by the RtcB RNA ligase and the intron is circularised. tRNA maturation is finalised by nucleoside modifications at various positions (black and grey dots). The order of processing events given here is deduced from work in H. volcanii, different processing orders have been observed as explained in the text. For references and further details, see section ‘Maturation of tRNAs’. (B) Intron positions in archaeal tRNAs. The secondary structure of a tRNA is shown with the tRNA universal numbering (Sprinzl et al.1987). Filled circles indicate the nucleotide located 5΄ to the insertion site (including tRNA interruptions found in split-tRNAs). Positions conserved across phyla are indicated in grey, and sites specific for Thermoproteales are given in blue. Larger circles indicate positions with high frequency (more than 10) according to Sugahara et al. (2008), Fujishima et al. (2009, 2011). Figure 5. View largeDownload slide Processing pathways of transfer RNA precursors. (A) Maturation of a tRNA precursor in Archaea. The tRNA precursor is processed at the 5΄-end by RNase P, while the 3΄-end is generated by RNase Z cleavage. If not encoded by the tDNA, the CCA triplet is added by the tRNA nucleotidyl-transferase. If an intron (brown) is present within the tRNA, the Bulge-Helix-Bulge RNA (BHB) motif formed at the exon-intron junctions is recognised and cleaved by EndA. Subsequently, the tRNA-halves are joined by the RtcB RNA ligase and the intron is circularised. tRNA maturation is finalised by nucleoside modifications at various positions (black and grey dots). The order of processing events given here is deduced from work in H. volcanii, different processing orders have been observed as explained in the text. For references and further details, see section ‘Maturation of tRNAs’. (B) Intron positions in archaeal tRNAs. The secondary structure of a tRNA is shown with the tRNA universal numbering (Sprinzl et al.1987). Filled circles indicate the nucleotide located 5΄ to the insertion site (including tRNA interruptions found in split-tRNAs). Positions conserved across phyla are indicated in grey, and sites specific for Thermoproteales are given in blue. Larger circles indicate positions with high frequency (more than 10) according to Sugahara et al. (2008), Fujishima et al. (2009, 2011). The number of tRNA genes in archaeal genomes ranges from 36 in M. smithii to 61 in M. bakeri (Chan and Lowe 2009, http://gtrnadb.ucsc.edu/). The archaeal tRNA genes are either encoded as single transcriptional units in operons together with other tRNAs or in polycistronic units in combination with other transcripts such as ribosomal and messenger RNAs. A recent analysis of transcriptional start sites in H. volcanii identified individual sites for 39 of 52 tRNAs (Babski et al.2016), whereas in M. thermoautotrophicus 10 of 39 tRNAs are apparently single-gene transcriptional units and two operons exist with five transcriptionally linked tRNAs (Smith et al.1997). Remarkably, archaeal tDNA can adopt a variety of non-canonical arrangements. In N. equitans six tRNA genes are split with their 5΄-halves containing a sequence stretch at their 3΄ end that can base pair with the sequence preceding the cognate 3΄-halves to form a non-canonical BHB motif (Fig. 3B). After recognition by EndA, both halves are joined in a trans-splicing reaction that releases a mature tRNA (Randau et al.2005). Disrupted tRNAs are not only found in reduced-genome archaea but also exist in free-living Crenarchaeota such as Caldivirga maquilingensis, Aeropyrum pernix, Staphylothermus marinus, Staphylothermus hellenicus and Thermosphera aggregans (Chan, Cozen and Lowe 2011). Interestingly, the location of the disruption corresponds in most cases to the canonical site of tRNA introns between positon 37 and 38 (Fig. 5B) although it can also be positioned between nucleotides 30/31 or 59/60 in some Staphylothermus species, N. equitans and Thermofilum pendens, respectively (Chan, Cozen and Lowe 2011). In C. maquilingensis, tRNA genes are even split into three segments (Fujishima et al.2009). Finally, as an extreme case of fragmented tDNAs, permutated tRNA genes have been found where both tRNA-halves are part of one operon but the 3΄-half precedes the 5΄ portion (Chan, Cozen and Lowe 2011). Generation of the full-length tRNA depends on a BHB motif encoded by the termini of the pre-tRNA. Upon splicing, the two halves are joined into circular intermediates that are subsequently opened by the action of RNase P and RNase Z to form a mature tRNA (Chan, Cozen and Lowe 2011). Altogether, the archaeal tDNA landscape codes for a multitude of tRNA-precursor variants, each requiring specific maturation pathways (Fujishima and Kanai 2014). 5΄-end tRNA processing The 5΄-leader sequence of a pre-tRNA is removed in all organisms by the universal RNase P (Figs 1 and 5A). RNase P substrate recognition relies on distinct structural characteristics of the conserved tRNA 3D structure. Especially critical is the so called ‘elbow’ where D- and T-loop nucleotides interact via base-stacking to form a hydrophobic surface (Zhang and Ferré-D’Amaré 2016). As shown for bacterial RNase P, the distance along the T-arm-acceptor-stem-axis correctly positions the cleavage site (Reiter et al.2010). Other determinants are the +1 nucleotide of mature tRNA, the discriminator nucleotide at position 73 and the pre-tRNA-leader (Kirsebom 2007 and references therein). 3΄-end tRNA processing Processing of the pre-tRNA 3΄ end is catalysed by the universal endonuclease RNase Z (Figs 1 and 5A). RNase Z recognises key substructures of the tRNA molecule, especially the T-arm and acceptor stem and cleaves directly after the discriminator base in most cases (see section‘Ubiquitous endoribonucleases’). In the majority of Archaea (e.g. H. volcanii, T. acidophilum or N. equitans), the CCA-end essential for amino acid attachment, accurate positioning for peptide-bond formation and tRNA quality control (Wilusz et al.2011) is not encoded in the tDNA (Hartmann et al.2009). Therefore, the CCA-motif is added by the nucleotidyl-transferase after RNase Z cleavage. Consequently, tRNA molecules with post-transcriptionally added CCA-ends should not be substrates for RNase Z since this would result in an endless cycle of CCA-addition and cleavage (Minagawa et al.2004). Indeed, the presence of a CCA in the 3΄-trailer inhibits cleavage by T. acidophilum RNase Z, whereas RNase Z enzymes from P. aerophilum and M. jannaschii process CCA-containing trailers in vitro (Schiffer, Rösch and Marchfelder 2003; Minagawa et al.2004; Späth et al.2008). Both species are examples for the uncommon case of archaeal genomes encoding a mixed population of tDNAs with varying portions of encoded and non-encoded CCA-ends (Marck and Grosjean 2002). Whether CCA removal is prevented by a protein factor or fast amino acetylation in vivo waits to be determined. Currently, no information is available on the mode of 3΄-processing in the rare cases of archaeal species such as A. pernix or P. furious that encode only CCA-containing tRNA genes (Marck and Grosjean 2002). Pyrococcus furiosus RNase Z has only been studied in vitro and in combination with non-CCA substrates (Hölzle et al.2008; Späth et al.2008). However, the bacterial RNase Z in Thermotoga maritima, where almost all tDNAs encode the CCA triplet, does not process at the discriminator base but after the CCA due to differences in substrate positioning by this bacterial RNase Z (Minagawa et al.2004; Ishii et al.2005). In bacterial systems, with encoded CCA-ends, a common 3΄-end maturation pathway is an exonucleolytic processing (Li and Deutscher 1996). But what holds true for archaeal systems with CCA-encoding tDNAs still awaits to be determined. The post-transcriptional addition of the CCA-end is achieved by the enzyme ATP(CTP):tRNA nucleotidyltransferase (EC 2.7.7.25) also termed the CCA-adding-enzyme or tRNA terminal transferase (Deutscher 1990) (Fig. 5A). It catalyses the template and primer-independent addition of nucleotides and is either essential if CCA is not encoded in the tDNA or required for maintenance of CCA integrity (Zhu and Deutscher 1987; Aebi et al.1990). The terminal transferase is conserved in all domains of life but two classes distinct in sequence and especially structure exist: class I enzymes are present in Archaea and class II enzymes found in Eukarya and Bacteria (Yue, Maizels and Weiner 1996). Both classes share only a homologous N-terminal catalytic core, whereas the other domains (neck, body and tail) differ structurally as well as in their arrangement resulting in a distinct overall architecture: U shaped in the class I of A. fulgidus opposed to the sea-horse shape of class II enzymes (Okabe et al.2003). Co-crystallisation of the A. fulgidus class I terminal transferase dimer with different substrates shows that the tRNA T-arm-acceptor-stem-half is recognised by a complementary shaped cleft in the body and tail domain (Tomita et al.2006). The tRNA elbow is contacted by the tail domain placing the discriminator base between head and neck domain fixing the 3΄-end within the active site (Zhang and Ferré-D’Amaré 2016). All interactions except for the base at position 72 are to the phosphate-sugar-backbone of tRNA allowing for sequence-independent recognition (Xiong and Steitz 2004). The specificity of the nucleotide-binding pocket of the A. fulgidus terminal transferase is not a function of the protein residues alone but co-catalytically adapted by influence of the growing 3΄-terminus during CCA-addition resulting in switching from CTP to ATP specificity for the final nucleotide transfer (Tomita et al.2006; Pan, Xiong and Steitz 2010). A growing body of evidence suggests that the terminal transferase can add more than one CCA to a tRNA or tRNA-like 3΄-end acting as a degradation signal in the course of RNA surveillance (Wilusz et al.2011). Indeed, as shown for the A. fulgidus terminal transferase in vitro, the tRNA is twisted and compressed by movement of the head domain after the first CCA-addition and if the tRNA structure is destabilised a bulge will form upon torsion and a second CCA cycle could ensue (Kuhn et al.2015). tRNA intron splicing On average, 15% of archaeal tRNAs contain introns ranging in size from 11 to 129 nucleotides (Yoshihisa 2014). This ranges from 48% intron-containing tRNAs in Crenarchaeota and almost 70% in Thermoproteales. Multiple introns are encoded in 17% of all intron-containing tRNAs as for example in T. pendens or P. calidifontis tRNAs (Marck and Grosjean 2003; Sugahara et al.2008; Yoshihisa 2014). However, tRNA introns are present across all domains of life and their canonical position is between nucleotide 37 and 38 in the anticodon loop. In Archaea, they have been found at more than 30 non-canonical positions along the entire tRNA molecule (Marck and Grosjean 2003) (Fig. 5B). Introns are mostly located between the D-arm and the anticodon loop in archaeal species but also in the 3΄-half in Thermoproteales (Fujishima et al.2010). Archaea do not possess self-splicing group I introns but use an archaea-specific splicing mechanism that requires an enzyme with homology to parts of the eukaryotic tRNA splicing endonuclease (Table 2) (Lykke-Andersen et al.1997; Li, Trotta and Abelson 1998). Archaeal introns are characterised by BHB motifs recognised and cleaved by the EndA (Thompson and Daniels 1990) (described in section ‘Ubiquitous endoribonucleases’). The BHB motif or a variant thereof is the necessary and sufficient determinant for a splicing reaction in Archaea (Fig. 3B). The arbitrary location of archaeal introns and tRNA splits is only possible because of this ‘motif-only’ definition of splice sites. Non-canonical BHB motifs recognised by variant forms of EndA are often associated with tRNA introns at non-canonical positions or split-tRNAs (Marck and Grosjean 2003; Hirata et al.2012) (Fig. 3B). After cleavage by EndA, the exons are joined by the RNA ligase and the intron is released as a circularised form (Fig. 5A). BHB motifs formed by sequences adjacent to the junction in split-tRNAs trigger a trans splicing reaction used to combine the individually encoded tRNA-halves (see above, Randau et al.2005; Chan, Cozen and Lowe 2011). The 3΄-phosphate RNA splicing ligase was only recently purified from M. kandleri and assigned to the RtcB protein family (Englert et al.2011). A tRNA ligation activity was demonstrated for P. aerophilum RtcB in vitro and the P. horikoshii RtcB structure revealed a novel active centre composition and structural fold (Englert et al.2012). RtcB homologues are found across all domains even in bacteria although bacterial tRNAs contain self-splicing group I introns. Thus, the bacterial RtcB homologue cannot catalyse a splicing-associated ligation reaction but might be used to generate circular RNAs (see section ‘Small non-coding box C/D and box H/ACA guide RNA biogenesis’). Post-transcriptional tRNA chemical modifications Natural RNA molecules contain over 100 post-transcriptional modified nucleosides, needed for cellular functions of RNAs in all three domains of life (see Agris 2015; Agris et al.2017; Väre et al.2017 for a review). Chemical modifications of the canonical adenosine, guanosine, cytidine and uridine nucleosides are generated enzymatically during the maturation processes of major RNA species (tRNA, rRNA and mRNA). Briefly, modifications include pseudouridine formation, methylations, deamination, carbamoylation, acetylation, carboxylation and addition of amino acids, all of which can be found in two dedicated databases (RNAMDB, http://rna-mdb.cas.albany.edu/RNAmods/ (Cantara et al.2011) and Modomics http://modomics.genesilico.pl/ (Machnicka et al.2013)). RNA modification biosynthesis requires simple but also sequential and multiple enzymatic activities. With a large variety of post-transcriptionally modified nucleosides, tRNAs are the most extensively chemically modified RNA entity (17% of the nucleosides in average) with 22 modified nucleosides universally retrieved in all organisms (Motorin and Helm 2011; El Yacoubi, Bailly and de Crécy-Lagard 2012; Grosjean and Westhof 2016) (Fig. 5A). Some of these modifications have been shown to facilitate RNA folding, structure and conformational dynamics. More precisely, in Archaea, at least 47 different modifications have been identified with some of them, such as archaeosine or agmatidine, being specific to this domain of life (Grosjean et al.2008; Phillips and de Crécy-Lagard 2011). The numerous and diverse chemical moieties in tRNAs affect global 3D conformation by providing unique recognition determinants important for effective tRNAs decoding ability (see Väre et al.2017 for a review). Indeed, chemical moieties at position 34 (wobble base) and 37 (dangling base) in the tRNAs anticodon loop expand the genetic code by stabilising codon–anticodon pairs (Grosjean, de Crécy-Lagard and Marck 2010; Grosjean and Westhof 2016). As an example, the universal N6-threonylcarbamoyl adenosine (t6A37), which occurs exclusively at the position 31 of all tRNAs decoding ANN codons (N 1/4; A, C, U or G), facilitates tRNA binding to the ribosome and prevents frameshifting in the course of translation (see Pichard-Kostuch et al.2017 for a review). Order of tRNA-processing events? A recent dissection of M. kandleri RNA-seq data pin points the identification of tRNA precursors before and after the unique C8-to-U8 editing activity and enables the determination of the order of tRNA-processing events (Su, Tripp and Randau 2013). According to this, pre-tRNA is first processed at its 3΄-end followed by a cleavage at the 5΄-end, before nucleotide modifications are introduced and C-to-U editing occurs, followed by final removal of introns (Su, Tripp and Randau 2013). In contrast, in vitro studies show that a precursor containing a 5΄-leader sequence or an intron is only processed by the recombinant RNase Z from P. furiosus but not by the H. volcanii one (Späth et al.2008). Furthermore, a yeast tRNA precursor expressed in H. volcanii cells undergoes 5΄- before 3΄-processing (Palmer, Nieuwlandt and Daniels 1994). However, unpublished results from the Marchfelder lab demonstrate that upon depletion of RNase P in H. volcanii cells, the 5΄-end but not the 3΄end tRNA maturation is impaired. The order of the tRNA-processing events is therefore not clearly defined and a certain degree of flexibility exists. The expanding number of archaeal whole-transcriptome analyses will certainly facilitate our understanding of the degree of this plasticity. Ribosomal RNA biogenesis Ribosomes function in protein synthesis in all domains of life and are universally conserved macromolecular machineries composed of rRNAs and specific sets of ribosomal proteins. Eukaryal and bacterial ribosome biogenesis pathways are complex and involve RNA processing and ribosomal protein assembly that require a series of specific enzymes and assembly factors acting in coordinated and multilayered manner. Only a few steps in the archaeal pathways have been currently identified, but the larger part remains to be explored (for a review, see Ferreira-Cerca 2017) (Fig. 6). Despite the universal conserved status of the RNA moieties of the ribosome with a common core structure across all domains of life, domain-specific rRNA expansion segments mostly localised at the ribosome periphery are present in Archaea and Eukarya (Petrov et al.2014). Indeed analysis of 16S rRNA sequences led to the recognition of Archaea as a separate domain of life (Woese, Kandler and Wheelis 1990; Albers et al.2013; Gulen et al.2016). In addition, ribosomal proteins vary in size and number across the three domains with 32 universal- and more than 60 domain-specific ribosomal proteins (Márquez et al.2011) (for a review, see Ferreira-Cerca 2017). Typically archaeal ribosomes sediment at 70S and are composed of the 23S, 16S and 5S rRNAs. Figure 6. View largeDownload slide Processing pathways of ribosomal RNA precursors. (A) Structure of two archaeal rrn operons. The structure of the rrn operons of H. volcanii (Euryarchaeon) and S. acidocaldarius (a representative of the TACK superphylum) are given at the top and the bottom, respectively. In both structures, the 16S and 23S rRNAs are co-transcribed as large precursors. The rrn genes (black boxes) are flanked by external transcribed spacers (ETS) at the 5΄- and 3΄-end and separated by an internal transcribed spacer (ITS). While in genomes of the TACK phyla, the 5S rRNA is encoded separately, in euryarchaeal genomes, the 5S rRNA gene is located downstream of the 23S rRNA gene. In most Euryarchaea, the ITS1 encodes tRNAAla and in some cases the 3΄ETS codes for tRNACys (dark grey boxes). In H. volcanii, a tRNA-like structure (also in dark grey) is encoded immediately upstream of the 5S rRNA. Sequences flanking the 16S and 23S rRNA sequences (marked by light grey boxes) can fold into Bulge-Helix-Bulge RNA (BHB) motifs. (B) Archaeal rRNA processing steps. The rRNA precursor molecule (H. volcanii pre-rRNA is depicted as example) is processed by various nucleases. Some activities remain elusive (indicated by a question mark), while others have been characterised (indicated by scissors). The inverted repeats flanking the large rRNA genes fold into BHB motifs processed by EndA. After ligation by the RNA ligase, the circular 16S and 23S precursors are further matured by unidentified activities yielding the mature rRNAs. Based on in vitro experiments, it was proposed that the 16S rRNA 3΄-end is processed by Nob1. The tRNAs encoded in the ITS and 3΄-ETS are processed by RNase Z and RNase P. The tRNA-like structure at the 5΄-end of 5S rRNA is processed by RNase Z to release the mature 5S rRNA 5΄-end. However, the 5S rRNA 3΄-end maturation pathway is still unsolved. Cleavages have also been observed within tRNAs. For references and further details, see section ‘Ribosomal RNA biogenesis’. Figure 6. View largeDownload slide Processing pathways of ribosomal RNA precursors. (A) Structure of two archaeal rrn operons. The structure of the rrn operons of H. volcanii (Euryarchaeon) and S. acidocaldarius (a representative of the TACK superphylum) are given at the top and the bottom, respectively. In both structures, the 16S and 23S rRNAs are co-transcribed as large precursors. The rrn genes (black boxes) are flanked by external transcribed spacers (ETS) at the 5΄- and 3΄-end and separated by an internal transcribed spacer (ITS). While in genomes of the TACK phyla, the 5S rRNA is encoded separately, in euryarchaeal genomes, the 5S rRNA gene is located downstream of the 23S rRNA gene. In most Euryarchaea, the ITS1 encodes tRNAAla and in some cases the 3΄ETS codes for tRNACys (dark grey boxes). In H. volcanii, a tRNA-like structure (also in dark grey) is encoded immediately upstream of the 5S rRNA. Sequences flanking the 16S and 23S rRNA sequences (marked by light grey boxes) can fold into Bulge-Helix-Bulge RNA (BHB) motifs. (B) Archaeal rRNA processing steps. The rRNA precursor molecule (H. volcanii pre-rRNA is depicted as example) is processed by various nucleases. Some activities remain elusive (indicated by a question mark), while others have been characterised (indicated by scissors). The inverted repeats flanking the large rRNA genes fold into BHB motifs processed by EndA. After ligation by the RNA ligase, the circular 16S and 23S precursors are further matured by unidentified activities yielding the mature rRNAs. Based on in vitro experiments, it was proposed that the 16S rRNA 3΄-end is processed by Nob1. The tRNAs encoded in the ITS and 3΄-ETS are processed by RNase Z and RNase P. The tRNA-like structure at the 5΄-end of 5S rRNA is processed by RNase Z to release the mature 5S rRNA 5΄-end. However, the 5S rRNA 3΄-end maturation pathway is still unsolved. Cleavages have also been observed within tRNAs. For references and further details, see section ‘Ribosomal RNA biogenesis’. Diversity of rRNA operons in archaeal genomes Across all domains of life, rRNA encoding genes are organised in rrn operons (Lafontaine and Tollervey 2001). The majority of archaeal genomes possess only one rrn operon. However, up to four operons, scattered on the chromosome, are present in some euryarchaeal species such as M. vannielii or Halolamina sediminis (Acinas et al.2004; Stoddard et al.2015). The rrn operon organisation differs across archaeal lineages, and disrupted and imbalanced rDNA arrangements can be found. In general, euryarchaeal rDNA typically comprises the 16S and 23S rRNA genes linked by an internal transcribed spacer (ITS) and the 5S rRNA gene framed by 5΄- and 3΄-external transcribed spacers (ETS) (Klug et al.2007). The ITS typically encodes a tRNAAla as found in the model Euryarchaeota M. mazei or H. volcanii (Deppenmeier et al.2002; Hartman et al.2010) (Fig. 6A). The 3΄-flanking ETS occasionally encodes an additional tRNA as in H. volcanii and Haloquadratum walsbyi the tRNACys (Bolhuis et al.2006; Hartman et al.2010) and the 5΄-ETS of M. fervidus rrn operon contains the 7S RNA and tRNASer genes (Haas, Daniels and Reeve 1989). Often, crenarchaeal rrn operons do not encode tRNAs in their intergenic sequences and encode the 5S rRNA in a separate transcriptional unit elsewhere in the genome as in S. acidocaldarius or T. tenax (Durovic and Dennis 1994; She et al.2001; Siebers et al.2011). Analysis of available genomic data shows that this arrangement of coupled 16S-23S rRNA and separate 5S transcription without interspersed tRNAs is also common throughout the TACK superphylum in Candidatus Korarchaeum cryptofilum OPF8, Cenarchaeum symbiosum, Candidatus Caldiarchaeum subterraneum and P. furiosus (Hallam et al.2006; Elkins et al.2008; Nunoura et al.2011; Ikeda et al.2017). In some specific cases, archaeal genomes possess a completely different rDNA organisation as for instance in M. jannaschii where one rrn operon encodes all three rRNAs, while the other is missing the 5S rRNA (Bult et al.1996) or in N. equitans and T. acidophilum where all three rRNAs are transcribed separately (Ree and Zimmermann 1990; Waters et al.2003). Processing of primary rRNA into mature 16S and 23S rRNAs Ribosomal RNAs are part of a precursor transcript (pre-rRNA), which is matured into the functional rRNAs in a series of steps. In Bacteria and Eukarya, those pathways involve the coordinated action of endo- and exoribonucleases as well as RNA modification machineries which have been best characterised in E. coli and S. cerevisiae (Deutscher 2009; Henras et al.2015). In Archaea, these steps have been investigated by analyses of processing products and intermediates identified by nuclease protection assays, primer extensions and cDNA analyses and in vitro processing experiments (Chant and Dennis 1986; Durovic and Dennis 1994; Dennis, Ziesche and Mylvaganam 1998; Ciammaruconi and Londei 2001; Tang et al.2002). As a general feature in Bacteria, primary polycistronic rRNA precursors contain flanking inverted repeat regions at the 5΄-leader and 3΄-trailer sequences of the respective 16S and 23S rRNA precursors (16S and 23S pre-rRNAs, respectively) (Deutscher 2009). These repeats form extended helical structures which are cleaved by the bacterial RNase III within the double-stranded region to produce the 16S and 23S pre-rRNAs. In Archaea, these stems contain a BHB motif recognised by the splicing endonuclease EndA (Fig. 6B) as occurs in the tRNA splicing reaction (Chant and Dennis 1986; Thompson and Daniels 1990). Hence, the 16S and 23S pre-rRNAs are circularised and the ETS and ITS sequences adjacent to the BHB motif are ligated as shown in S. solfataricus, S. acidocaldarius, H. salinarum and A. fulgidus (Tang et al.2002; Danan et al.2012). The circular 16S and 23S pre-rRNAs are quickly opened by endoribonucleolytic cleavages and trimmed by unidentified endo- and exo-ribonucleases, respectively, to yield the mature transcripts (Chant and Dennis 1986; Tang et al.2002). The 5΄-end appears to be matured faster than the 3΄-end, which might reflect a sequential assembly of the ribosome or a sequential formation of structural elements needed for processing (Chant and Dennis 1986; Dennis, Ziesche and Mylvaganam 1998). In addition, a second pathway for the direct generation of the mature 5΄-end via an endoribonucleolytic activity was described in H. marismortui for both pre-rRNAs (Dennis, Ziesche and Mylvaganam 1998). This splicing-independent maturation of the 23S rRNA plays a minor role, whereas the release of the 16S rRNA by endoribonucleolytic cleavage is an obligatory outcome event in the context of the H. marismortui rRNA operon since EndA-processing stems are missing (Dennis, Ziesche and Mylvaganam 1998). With the exception of EndA activity, archaeal exo- and endo-ribonucleolytic activities which generate mature 16S and 23S rRNA remain to be discovered. However, more recently, conserved homologues of the eukaryal endonuclease Nob1 responsible for end cleavage in the yeast 18S rRNA maturation pathway were identified in most archaeal genomes (Fatica, Tollervey and Dlakić 2004) (Fig. 1, Table 2). Analysis of P. horikoshii Nob1 revealed not only a conserved domain structure but also a similar in vitro cleavage specificity against a pre-16S like substrate resulting in the exact mature 3΄-end of 16S rRNA (Veith et al.2012). Processing of resulting 5΄EST/ITS/3΄EST products The 5΄EST/ITS/3΄EST ligation by-product, containing tRNAs and generated after the processing by EndA, is further processed by the enzymes involved in tRNA processing such as RNase Z and RNase P at the tRNA or tRNA-like structures present (Chant and Dennis 1986; Tang et al.2002) (Fig. 6B). Processing by RNase P is a late event occurring subsequent to pre-16S rRNA excision and after the initiation of tRNA 3΄-processing (Chant and Dennis 1986; Dennis, Ziesche and Mylvaganam 1998). Moreover, the tRNA included in the 3΄-EST is also released by the tRNA-processing enzymes. In H. salinarum, the processing of tRNACys is unordered, but the processing of the tRNACys 5΄-end precedes 5S rRNA 3΄-end maturation (Chant and Dennis 1986). Finally, tRNAs of archaeal rRNA operons are sometimes subject to endoribonucleolytic cleavage within their body as suggested in H. salinarum and H. marismortui (Chant and Dennis 1986; Dennis, Ziesche and Mylvaganam 1998). Processing of mature 5S rRNAs Processing stems are not involved in 5S pre-rRNA release. In H. salinarum and H. marismortui, 5S pre-rRNA formation is rapidly followed by maturation into 5S rRNA (Chant and Dennis 1986; Dennis, Ziesche and Mylvaganam 1998). Moreover, analysis of 5S pre-rRNA-processing steps in H. volcanii cell extracts led to the identification of RNase Z as the endonuclease activity responsible for 5΄-end maturation (Hölzle et al.2008) (Fig. 6B). The region upstream of the 5S pre-rRNA transcript is able to fold into a tRNA-like structure comprising a D-replacement loop, a shortened acceptor stem as well as a shortened T-arm. In silico data show that the D- and T-arm-like regions can pair as in the canonical tRNA structure, giving rise to the tRNA elbow conformation recognised by RNase Z (Hölzle et al.2008; Zhang and Ferré-D’Amaré 2016). In vitro, cleavage of the 5S pre-rRNA tRNA-like structure at the 3΄-end releases the H. volcanii 5S rRNA mature 5΄-end as the cognate trailer (Hölzle et al.2008). Therefore, RNase Z is not only implicated in rRNA processing to generate mature tRNAs encoded within the ITS and ETS regions, but is also able to mature the 5΄-end of 5S rRNA. Finally, the recent circRNA-seq approach of Sorek and collaborators identified 5S rRNA circular intermediate forms in S. solfataricus (Danan et al.2012). However, whether the circular form is associated with the 5S rRNA-processing pathway has not been determined yet (Danan et al.2012). A recent study links the protein FAU-1, previously described as a RNA-binding protein, to pre-5S and 16S rRNA processing in P. furiosus and T. kodakarensis (Ikeda et al.2017). The FAU-1 protein appears also to be important in the quality control of the mature 23S and 16S transcripts (Mackie 2013; Sulthana, Basturea and Deutscher 2016). FAU-1 proteins of both Thermococcales were shown to process a 5S rRNA precursor in vitro, preferentially at UA sites (Ikeda et al.2017). Moreover, deletion of the gene in T. kodakarensis impaired cell growth and appeared to alter degradation patterns of the 23S and 16S rRNAs (Ikeda et al.2017). Homing introns within archaeal 16S and 23S pre-rRNAs In Crenarcheota and Aigarchaeota, 16S pre-rRNAs may contain introns (Nomura, Sako and Uchida 1998; Morinaga, Nomura and Sako 2002; Itoh, Nomura and Sako 2003; Nunoura et al.2011). This is also the case for the 23S pre-rRNA of Desulfurococcales and Thermoproteales (Itoh, Nomura and Sako 2003). The most numerous and diverse set of introns is found in the genera Thermoproteus, Pyrobaculum and Caldivirga. However, the intron distribution within species is very variable and exhibits a biogeographic rather than phylogenetic distribution (Jay and Inskeep 2015). The rRNA introns, like tRNA introns, contain BHB motifs that ensure their removal by the EndA splicing endoribonuclease during rRNA processing (Lykke-Andersen et al.1997). Despite some short and non-coding introns, archaeal pre-rRNA introns are generally rather long (often >700 nt) and mostly contain one or more genes that encode the homing endonuclease characterised by the LAGLI-DADG motif (Morinaga, Nomura and Sako 2002; Stoddard 2005). The sequence-specific homing endonuclease allows mobility of the intron by reintegration at compatible intron-less homing sites. The recognition sequences for archaeal homing endonucleases are pseudo-palindromic or asymmetric sites, mostly found in conserved regions of the rRNA (Stoddard 2005). The rRNA introns are stably maintained within the cells and have also been identified as circularised RNAs e.g. in Desulfurococcus mobilis and P. aerophilum (Kjems and Garrett 1988; Burggraf et al.1993). Post-transcriptional rRNA chemical modifications An increasing number of post-transcriptional chemical modifications are present on rRNAs from Bacteria to Eukarya. Indeed base modifications, such as adenosine methylation, 2’-hydroxyl group methylation and conversion of uridine residues to pseudouridine by base rotation, occur during ribosome biogenesis (for review, see Ferreira-Cerca 2017). These nucleoside chemical modifications are proposed to fine-tune the ribosome biogenesis and functions (Sloan et al.2017). In hyperthermophiles archaea, the observed number of modifications is close to those reported of eukaryotic rRNAs (Grosjean, de Crécy-Lagard and Marck 2010; Dennis et al.2015; Sloan et al.2017). In contrast, a reduced set is observed in halophilic archaea. As an example, in S. solfataricus, 67 2’-hydroxyl group ribose methylations were reported compared to only 4 in H. volcanii (Grosjean et al.2008; Dennis et al.2015). The functional significance of this disparity is not yet understood. The most ubiquitous modifications are pseudouridine (isomer of uridine) and 2’-O-ribose methylation. These facilitate RNA folding, structure and conformational dynamics (Motorin and Helm 2011). In Archaea and Eukarya, both modifications are governed by two independent mechanisms, an RNA-independent reaction and an RNA-dependent reaction (De Zoysa and Yu 2017). The first involves single protein enzymes called pseudouridine synthases and methyl-transferases that alone recognise the substrate and catalyse the isomerisation of uridine to pseudouridine and the transfer of a methyl group at the 2΄O-ribose, respectively. The second are RNA-guided pseudouridylation and methylation by box H/ACA and box C/D RNP complexes, respectively. Briefly the archaeal C/D box sRNA is characterised by short sequence motifs, two sets of box C and C’ (RUGAUGA) and box D and D’ (CUGA) sequences, which fold into the widespread K-turn or K-loop structural motifs that is recognised by L7Ae (Fig. 3A) (Klein et al.2001; Rozhdestvensky et al.2003; Nolivos, Carpousis and Clouet-d’Orval 2005). In addition, two other proteins, Nop5 and the methyltransferase fibrillarin are recruited to generate an active RNP complex (Omer et al.2002; Bortolin, Bachellerie and Clouet-d’Orval 2003). The archaeal H/ACA box sRNAs is composed of one, two or three H/ACA motifs. Each archaeal H/ACA stem-loop structure contains K-turn or K-loop structural motifs which binds L7Ae. The complete archaeal H/ACA sRNPs contain three other proteins: the pseudouridine synthetase aCbf5 and aNop10 and aGar1 required to reinforce aCBF5 activity (Charpentier, Muller and Branlant 2005; Muller et al.2008). The RNA component of the complex serves as a guide that base pairs with the RNA substrate and directs the enzyme (aCbf5 or aNop1) to carry out the pseudouridine formation or methyl transfer at a specific site. In Archaea, ribosomal and tRNA nucleosides can be modified by these RNA-guided pathways. Interestingly, in Archaea, the RNA-dependent reactions involve more than 100 small box guide RNAs that target specific sites of rRNAs (Muller et al.2008). In addition, these small-guide RNAs have been suggested to serve as ‘RNA chaperones’ in facilitating rRNA folding (Bachellerie et al.1995; Steitz and Tycowski 1995; Dennis et al.2015). However, in a subset of Archaea, the same modification can either be synthetised from protein-based or RNA-guide systems (Clouet-d’Orval, Gaspin and Mougin 2005). This is the case for the dimethylation of the two universally conserved adenosines located at the 3΄-end of the 16S rRNA, generally dependent on the protein-based enzyme, RsmA/KsgA/Dim1, but alternatively can use an sRNA-guide system as in N. equitans (Seistrup et al.2017). Moreover, the widespread methylation at position U2552 of the 23S rRNA is mostly performed by the bacterial-like methyltransferase RlmE/RrmJ in most Euryarchaea with the exception of Thermococcales and Crenarchaea, in which the C/D box sR25 is in charge of guiding methylation at this position in the absence of RlmE/RrmJ homologues (Dennis et al.2015). CRISPR RNA processing Another type of ncRNA has sparked great interest in recent years. Long thought of as non-coding junk DNA, the long repeat arrays found in most archaeal and bacterial genomes have now been identified as part of a specific and hereditary defence mechanism based on RNPs: the CRISPR-Cas system (for recent reviews, see van der Oost et al.2014; Makarova et al.2015; Koonin and Makarova 2017). Key players of this system are the small crRNAs and the Cas proteins (Makarova et al.2015) (Fig. 7A). CRISPR loci encode recurring sequence elements interspaced by unique spacer sequences matching foreign genetic elements that are transcribed into the crRNA (Bolotin et al.2005; Mojica et al.2009). The associated cas genes encode the proteins needed for the defence reaction (Makarova et al.2015), which commences in three functional stages: adaptation, processing, interference (see Rath et al.2015; Mohanraju et al.2016 for reviews). During adaptation parts of the nucleic acid of an invading genetic element are integrated into the CRISPR loci as new spacers. In the processing stage, a precursor RNA transcribed from the CRISPR locus is generated and subsequently processed within the repeat sequences separating the individual spacers (Fig. 7A). The resulting crRNA is integrated into the effector complex and used during the third stage to sequence specifically base pair with the re-invading nucleic acid, which is subsequently cleaved and degraded. Figure 7. View largeDownload slide Processing pathways of CRISPR RNA and tRNA-derived fragment precursors. (A) crRNA maturation pathway. Components of the CRISPR-Cas system are encoded by the CRISPR-locus and the cas gene cassette. The CRISPR locus is composed of individual spacer sequences matching previously encountered invaders (blue boxes) interspaced by repeat sequences (black diamonds) and its transcription is governed by the promoter sequence contained in the leader region. The transcribed precursor, the pre-crRNA, may fold into secondary structures within the repeats. These motifs are recognised and processed by Cas6, releasing the mature crRNAs comprising an individual spacer sequence and repeat-derived 5΄-and 3΄-handles. Certain subtype-specific Cas proteins assemble together with the crRNA into cascade complexes to eliminate the invader DNA. The simplified overview depicted here is based on a type I system, typically found in haloarchaeal species. For references and further details, see section ‘CRISPR RNA processing’. (B) Generation of stable tRNA-derived fragments (tRFs). tRFs can be processed from mature tRNAs or precursor tRNAs. Three forms of tRFs have been identified in Archaea (Gebetsberger and Polacek 2013): the 5΄-tRF (red) stems from a cut in the D-loop of mature tRNA, whereas 3΄-CCA tRFs (orange) are released via processing within the T-loop. In contrast, the 3΄-U tRFs (blue) are generated from precursor tRNA processing and comprise the precursor 3΄-trailer. Figure 7. View largeDownload slide Processing pathways of CRISPR RNA and tRNA-derived fragment precursors. (A) crRNA maturation pathway. Components of the CRISPR-Cas system are encoded by the CRISPR-locus and the cas gene cassette. The CRISPR locus is composed of individual spacer sequences matching previously encountered invaders (blue boxes) interspaced by repeat sequences (black diamonds) and its transcription is governed by the promoter sequence contained in the leader region. The transcribed precursor, the pre-crRNA, may fold into secondary structures within the repeats. These motifs are recognised and processed by Cas6, releasing the mature crRNAs comprising an individual spacer sequence and repeat-derived 5΄-and 3΄-handles. Certain subtype-specific Cas proteins assemble together with the crRNA into cascade complexes to eliminate the invader DNA. The simplified overview depicted here is based on a type I system, typically found in haloarchaeal species. For references and further details, see section ‘CRISPR RNA processing’. (B) Generation of stable tRNA-derived fragments (tRFs). tRFs can be processed from mature tRNAs or precursor tRNAs. Three forms of tRFs have been identified in Archaea (Gebetsberger and Polacek 2013): the 5΄-tRF (red) stems from a cut in the D-loop of mature tRNA, whereas 3΄-CCA tRFs (orange) are released via processing within the T-loop. In contrast, the 3΄-U tRFs (blue) are generated from precursor tRNA processing and comprise the precursor 3΄-trailer. Despite their common functional principle, CRISPR-Cas systems are quite diverse. A plethora of individual types (termed I–VI) and an even broader variety of subtypes exhibiting specific differences both in composition and functional aspects exist (for an overview, see van der Oost et al.2014; Makarova et al.2015; Koonin and Makarova 2017). Currently, all CRISPR-Cas systems are divided into two classes, the main difference being that class 1 systems have a multiprotein effector complex (Cascade-, Csm-, or Cmr-complex) while the effector complex of class 2 systems consists of a single Cas protein (Cas9, Cas12 and Cas13). Archaea generally possess class 1 CRISPR systems, whereas Bacteria may also contain class 2 systems (Vestergaard, Garrett and Shah 2014; Makarova et al.2015). Recently, the first, and to date, only example of an archaeal class 2 system (type V) with the single effector Cas protein Cas12a has been described for Candidatus Methanomethylophilus alvus (Makarova et al.2015). Processing of the pre-crRNA precursor is quite uniform amongst different types of CRISPR-Cas systems and involves Cas proteins (see Charpentier et al.2015 for a review). Maturation of class 1 system crRNAs depends on the activity of the Cas6 endoribonuclease family (described in section‘Ubiquitous endoribonucleases’) (Charpentier et al.2015), with the exception of the type I-C system which depends on Cas5d (Garside et al.2012; Nam et al.2012). Cas6 binds pre-crRNA within the repeat region that often encodes palindromic motifs. If the repeats can stably fold into a hairpin, binding of Cas6 is governed by the structural element as seen in type I-D, I-E and I-F systems (Charpentier et al.2015). Unstructured repeats with an only weak folding potential as in type I-A, I-B and III rely on sequence elements for Cas6 binding that in turn may transiently stabilise the hairpin-fold (Wang et al.2011; Shao and Li 2013; Niewoehner, Jinek and Doudna 2014). The mature crRNA contains a single spacer sequence flanked by repeat-derived 5΄- and 3΄-handles. The 5΄-handle is commonly eight nucleotides long whereas the size of the 3΄-handle is quite variable as repeat length varies across species (Brouns et al.2008; Carte et al.2008; Lange et al.2013). After Cas6 processing, a second trimming step by a yet unidentified nuclease further reduces the 3΄-handle length via a ruler-dependent mechanism in type III systems (Hatoum-Aslan, Maniv and Marraffini 2011; Zhang et al.2012; Tamulaitis, Venclovas and Siksnys 2017). Similar trimming also occurs for type I-A, I-B and I-D systems in H. volcanii, Clostridium thermocellum, M. maripaludis, T. tenax and Synechocystis sp. PCC6803 (Richter et al.2012; Scholz et al.2013; Plagens et al.2014; Maier, Dyall-Smith and Marchfelder 2015). Interestingly, systems in which the crRNA seems to be further trimmed after processing by Cas6 possess repeat sequences for which no folding or only weak folding is predicted. Cas6 is only loosely bound to these handles and does not permanently (or not at all) associate with the effector complexes thereby rendering the 3΄-handle accessible (Charpentier et al.2015; Li 2015). Small non-coding box C/D and box H/ACA guide RNA biogenesis As mentioned above, 2’-O-ribose methylations are abundant in archaeal rRNAs with a prevalence in thermophilic species (Dennis et al.2015). In contrast, only few pseudouridinylations have been detected in archaeal RNAs so far. As a consequence, C/D box sRNAs outnumber H/ACA sRNAs in archaeal genomes (see Randau 2015 for a review). Indeed in silico analysis, immunoprecipitation experiments and more recently high-throughput RNA-seq datasets have revealed a plethora of C/D box sRNA sequences in various archaeal phyla like Ignicoccus hospitalis, M. maripaludis, M. kandleri, N. equitans, P. calidifontis, S. acidocaldarius and T. tenax (Gaspin et al.2000; Omer et al.2000; Phok et al.2011; Tripp et al.2017 and references therein). Comprehensive analyses revealed that only a fraction of archaeal C/D box sRNAs is flanked by an individual promoter while co-transcriptional arrangements with neighbouring genes are predominant (Tripp et al.2017). But co-transcription with other sequences requires subsequent processing events to generate mature guide RNAs. In the case of C/D box sRNAs co-transcribed with tRNAs as seen for the T. tenax tRNAPro or N. equitans tRNAVal, the release of C/D box sRNAs occurs by processing via the tRNA-processing enzymes (Randau 2012; Tripp et al.2017). Nevertheless, the general process releasing mature C/D box RNAs is still not understood. Although relaxed BHB motifs were predicted for some archaeal C/D box sRNAs, processing in S. acidocaldarius appeared independent of flanking sequences (Berkemer et al.2015; Tripp et al.2017). Maturation of archaeal small-guide RNAs could also involve other archaeal endo- and exoribonucleases such as the archaeal exosome and/or ribonucleases of the β-CASP family (see section ‘Archaeal ribonuclease families’), reminiscent of the eukaryotic guide sRNA maturation pathway (Massenet, Bertrand and Verheggen 2017). Finally, the most intriguing feature of archaeal small-guide RNA biogenesis is the discovery of significant amounts of circularised C/D box sRNAs in M. kandleri, N. equitans, P. furiosus and S. solfataricus (Starostina et al.2004; Danan et al.2012; Randau 2012; Su, Tripp and Randau 2013). It is proposed that circularisation stabilises RNAs in high-temperature conditions (Starostina et al.2004; Danan et al.2012). In P. furiosus, the circular C/D box RNAs are shown to be integrated into functional C/D box RNP complexes (Starostina et al.2004). C/D box sRNAs may also be embedded in tRNA intron sequences as seen in the intron of tRNATrp which is conserved across most Euryarchaeota (Clouet d’Orval et al.2001; Weisel, Wagner and Klug 2010). The C/D box sRNA is released by EndA and circularised by the RtcB ligase as seen in H. volcanii and H. salinarum (Clouet d’Orval et al.2001; Weisel, Wagner and Klug 2010). However, in the majority of cases, C/D box sRNAs are not intron-encoded and it remains to be determined how circularisation occurs. Emerging classes of ncRNAs Transfer RNA-derived fragments Fragments derived from tRNA sequences (tRNA-derived fragment or tRF), initially thought to be intermediates of tRNA degradation, are now identified as functional small ncRNAs (for reviews, see Keam and Hutvagner 2015; Kumar, Kuscu and Dutta 2016) (Fig. 7B). Their function was extensively studied in Eukarya but they have also been described in Bacteria and Archaea (Bernick et al.2012; Gebetsberger et al.2012; Heyer et al.2012; Kumar et al.2014). Transfer RNA-derived fragments can be generated from tRNA precursors or mature tRNAs. The amount of a given tRNA-derived fragment does not correlate with the amount of the tRNA progenitor implying that tRNA biogenesis and tRNA-derived fragment-processing pathways are independently regulated (Gebetsberger and Polacek 2013). The nature of the processing and regulatory factors generating tRNA-derived fragments is not known (Keam and Hutvagner 2015). Upon endoribonucleolytic cleavage, different classes of tRNA-derived fragments can emerge depending on the processing site and the nature of the precursor (Gebetsberger and Polacek 2013). Archaeal tRNA-derived fragments from several Pyrobaculum species, N. equitans and H. volcanii were found to encompass 5΄-tRFs (derived from a cleavage within the D-arm of mature tRNA), 3΄-CCA tRFs (derived from a cleavage within the T-loop of mature tRNA) and 3΄-U tRFs (remainder of 3΄-trailer of pre-tRNA) (Bernick et al.2012; Gebetsberger et al.2012; Heyer et al.2012; Randau 2012; Gebetsberger and Polacek 2013). A detailed analysis of tRFs is only available for H. volcanii (Gebetsberger et al.2012). Transfer RNA-derived fragments correspond to 11 of 51 tRNAs of H. volcanii with a majority derived from tRNA 5΄-ends in response to different stresses, as also seen in eukaryotes (Gebetsberger et al.2012; Heyer et al.2012). In an in-depth functional analysis of the alkaline-stress induced tRNAVal 5΄-tRF (tRF[Val]) (Gebetsberger et al. 2012, 2017), tRF[Val] was shown to co-purify with ribosomes and bind the small ribosomal subunits in vitro, inhibiting translation (Gebetsberger et al.2012). tRF[Val] was also recently demonstrated to bind the ribosome in proximity to the mRNA channel and to affect the overall protein biosynthesis in archaeal, bacterial and eukaryal ribosomes (Gebetsberger et al.2017). Similar translational control by tRFs occurs in eukaryotic HeLa cells (Sobala and Hutvagner 2013). tRFs are therefore potent regulators in all domains of life but their full gene regulatory potential has yet to be unravelled. Circular RNAs Another emerging class of small ncRNAs with an interesting biogenesis pathway are circular RNAs. They are rare but present in all domains of life. Most archaeal circular RNAs described are the result of the intron ligation step of the splicing reaction. Some are intermediates of rRNA-processing events, such as the circular precursors of 16S and 23S rRNAs (Tang et al.2002; Danan et al.2012) (section ‘Ribosomal RNA biogenesis’). But those transcripts are further processed and exert their biological function in the form of a linear end product. In contrast stable circular RNAs are generated by excision of the rRNA introns coding for homing endonucleases (Kjems and Garrett 1988; Burggraf et al.1993) (see section ‘Ribosomal RNA biogenesis’). In addition, tRNA precursor splicing events also produce stable circular RNA entities observed for the pre-tRNATrp of Euryarchaeota like H. volcanii (see section ‘Ribosomal RNA biogenesis’). A more systematic search for circular RNAs by the novel sequencing method, circ-RNA-seq, identified a plethora of circular RNAs in S. solfataricus, with a great diversity of sequences corresponding to rRNA intermediates, tRNA introns and C/D and H/ACA box RNAs, 7S SRP RNA and RNase P and open reading frame-derived circular RNAs from 37 different genes (Danan et al.2012). In addition, a circular form of the 7S SRP RNA has also recently been described for T. tenax where circularisation restores the permuted sequence by processing via the splicing machinery (Plagens et al.2015). Nevertheless, the molecular mechanisms that generate circular RNAs await discovery and the biological function and significance of these circular forms need to be experimentally tested. Circular RNAs have been proposed to represent either processing intermediates or to be a means of stabilisation and protection (Danan et al.2012). Messenger RNA turnover in Archaea—lessons from Eukarya and Bacteria For several decades, the post-transcriptional mechanisms by which the level, the stability and the quality of mRNAs are finely regulated have been an area of intense research interest and have been reported in comprehensive reviews (see chapters in Stoecklin and Mühlemann 2013). In essence, mRNA decay pathways are critical in defining timescales for cellular events by allowing transient adaptation relative to a growth rate and rapid responses to external stimuli to meet cellular requirements for homeostasis. The stability of an mRNA is determined by cell-specific set of ribonucleases, the sequence and structural features of the mRNA molecule itself and related proteins or regulatory RNAs. In this regard, small ncRNAs, such as sRNA, miRNA and siRNA, are critical and often mediate co-degradation, sequestration or stabilisation of target mRNAs in a regulatory hierarchy. In Archaea, in contrast to Bacteria and Eukarya, pathways controlling mRNA stability and quality are still poorly documented, even if several ribonuclease families carrying endo-, 5΄-to-3΄ and 3΄-to-5΄ exo-ribonucleolytic activities have been identified (Fig. 1) (Clouet-d’Orval et al.2015) (see section ‘Archaeal ribonuclease families’). Overview of mRNA decay pathways in Bacteria and Eukarya In Bacteria, most mRNAs are polycistronic with concurrent transcription and translation. A key event in their degradation is an initial cleavage by an endonuclease which is followed by the action of exoribonucleases (Fig. 8). This direct entry pathway controls the turnover of most mRNAs and recruits the hydrolytic endoribonucleases, RNase E in γ-proteobacteria and RNase Y in most Firmicutes as well as in some δ- and ε-proteobacteria (Laalami, Zig and Putzer 2014). Remarkably and despite their sequence disparity, both enzymes serve as assembly platforms for other enzymes, such as RNA helicases and exoribonucleases and are membrane bound (Bandyra et al.2013; Hui, Foley and Belasco 2014; Redder 2016). In addition, 5΄-end-dependent access pathways which sense the nature of the mRNA 5΄-terminus involve the recruitment of pyrophosphohydrolase enzymes of the Nudix family which trigger deprotection or ‘decapping’ (as a diphosphate is removed) and leave a 5΄-end monophosphate mRNA which is a subsequent substrate for 5΄-to-3΄exoribonucleases, such as RNase J in most Bacteria with the exception of the γ-proteobacteria (Bandyra et al.2013; Hui, Foley and Belasco 2014). Their association in multienzyme complexes or degradosome-like machineries and their subcellular localisation allow precise control of the mode of action of these potentially highly destructive enzymes (Kaberdin, Singh and Lin-Chao 2011; Deutscher 2015; Khemici et al.2015) (Fig. 8). In the case of mRNA lacking a termination codon, the stalled ribosomes are rescued by a mechanism known as trans-translation that involves a specialised bacterial RNA (tmRNA) with features of both tRNA and mRNA and an escort protein (SmpB) which has been shown to associate with bacterial RNase R in vitro (Inada 2013; Hui, Foley and Belasco 2014). More recently, in Firmicutes, a family of ribosome-dependent endoribonuclease Rae1/YacP has been proposed to play a role in a translation-related process with overlapping similarities with classical toxin–antitoxin systems involved in mediating bacterial stress responses (Leroy et al.2017). Figure 8. View largeDownload slide Schematic of the major mRNA quality control and decay pathways of the three domains of life. mRNA or precursor mRNA (pre-mRNA) (dark grey wavy line) is transcribed from genomic DNA (double light grey wavy lines) by the RNA polymerase (RNAP for Bacteria, aRNAP for Archaea and RNAPII (RNA polymerase II) for Eukarya). The 5΄cap structure (m7G) and poly-A tail (AAAAAA) of eukaryal mRNA are shown. The endo-, 5΄-to-3΄ exo- and 3΄-to-5΄ exoribonucleases are represented as red, blue and green ‘pacman’, respectively. In addition, the 3΄-to-5΄ exo machineries (archaeal and eukaryal exosome and bacterial PNPase) shown as green complexes are formed of multiple subunits. The ribonuclease tool kit boxes are specified at the bottom of each panel. Some of these machineries in bacterial cell and eukaryotic cytoplasm have been shown to be involved in the ribosome-rescue pathway. For clarity, mRNAs in polysomes are not represented. The mRNA processing and decay pathways (framed within each panel) are specified: for bacteria, it includes the direct entry and bypass decays, 5΄-end decay and sRNA mediated decay; for archaea, an non-sense mediated decay (NMD) has been proposed; for Eukarya, in the nucleus, it includes cleavage and polyadenylation, splicing and RNA surveillance whereas in the cytoplasm, it includes in addition to the classical decapping and deadenylation pathways, the micro-RNA and silencing RNA (miRNA/siRNA) decays, nonsense-mediated decay (NMD), no-go decay (NGD) and nonstop decay (NSD). Figure 8. View largeDownload slide Schematic of the major mRNA quality control and decay pathways of the three domains of life. mRNA or precursor mRNA (pre-mRNA) (dark grey wavy line) is transcribed from genomic DNA (double light grey wavy lines) by the RNA polymerase (RNAP for Bacteria, aRNAP for Archaea and RNAPII (RNA polymerase II) for Eukarya). The 5΄cap structure (m7G) and poly-A tail (AAAAAA) of eukaryal mRNA are shown. The endo-, 5΄-to-3΄ exo- and 3΄-to-5΄ exoribonucleases are represented as red, blue and green ‘pacman’, respectively. In addition, the 3΄-to-5΄ exo machineries (archaeal and eukaryal exosome and bacterial PNPase) shown as green complexes are formed of multiple subunits. The ribonuclease tool kit boxes are specified at the bottom of each panel. Some of these machineries in bacterial cell and eukaryotic cytoplasm have been shown to be involved in the ribosome-rescue pathway. For clarity, mRNAs in polysomes are not represented. The mRNA processing and decay pathways (framed within each panel) are specified: for bacteria, it includes the direct entry and bypass decays, 5΄-end decay and sRNA mediated decay; for archaea, an non-sense mediated decay (NMD) has been proposed; for Eukarya, in the nucleus, it includes cleavage and polyadenylation, splicing and RNA surveillance whereas in the cytoplasm, it includes in addition to the classical decapping and deadenylation pathways, the micro-RNA and silencing RNA (miRNA/siRNA) decays, nonsense-mediated decay (NMD), no-go decay (NGD) and nonstop decay (NSD). In contrast to bacterial mRNAs, eukaryal mRNAs are often monocistronic, must transit the nuclear membrane before being translated and have longer half-lives. In this setting, the maturation of eukaryal RNAPII-mRNA transcripts requires further intricate processes, such as 5΄-capping, 3΄-addition of extensive polyadenylation tracts, intron removal and chemical modifications. All of these sophisticated steps are orchestrated and controlled by RNA surveillance and decay pathways which recruit a number of multiprotein scaffolds (Chlebowski et al.2013; Nagarajan et al.2013; Porrua and Libri 2013; Schweingruber et al.2013; Inada 2017) (Fig. 8). Nuclear RNA (including cryptic stable transcript and snoRNA) control quality pathways and cytoplasmic mRNA decay pathways rely on the exosome/Rrp6/Dis3 machineries with dual, endo- and 3΄-to-5΄ exoribonucleolytic activities and on the ribonucleases from the XRN family with 5΄-to-3΄ exoribonucleolytic activity (Chlebowski et al.2013; Nagarajan et al.2013). These are associated with specific sets of co-factors to identify and degrade deadenylated and decapped transcripts (Stoecklin and Mühlemann 2013) (Fig. 8). The complexes are also involved in co-translational quality control mechanisms allowing elimination of a variety of aberrant mRNAs (errors in transcription, processing or induced by stresses). In addition, there are numerous quality control systems in place to differentiate between paused ribosomes and terminally stalled ribosomes that need to be rescued. In Eukarya, three cytoplasmic mRNA quality control systems monitor mRNAs for translational errors: nonsense-mediated decay (NMD), non-stop decay (NSD) and no-go decay (NGD) pathways (Inada 2013, 2017) (Fig. 8). Messenger RNA half-lives Earlier reviews specifically described half-lives of several archaeal mRNAs in different conditions (Evguenieva-Hackenberg and Klug 2009, 2011). In general, it appeared that some archaeal mRNAs have significantly longer average half-lives than most mRNAs in Bacteria. Briefly, half-lives of several transcripts varied from 7 to 57 min in M. vannielii (Hennigan and Reeve 1994), from 2 to 20 min in S. solfataricus and S. acidocaldarius (Bini et al.2002; Andersson et al.2006) and from 4 to 80 min in H. mediterranei (Jäger et al.2002). More recently, a first global picture of differential expression and half-lives for M. acetivorans provided evidence of drastic genome-wide shifts in RNA half-lives modulated by carbon source and led to the conclusion that more than half of metabolism genes are controlled by degradation (Peterson et al.2016). Nevertheless, even if a number of cases of mRNA cleavage have been documented in global analysis of mRNA decay at single-gene resolution, available data are still fragmentary. In M. jannaschii, evidence of directional 3΄-to-5΄ RNA degradation comes from the accumulation of transcript 5΄-ends from upstream of protein-coding genes (Zhang and Olsen 2009). However, the core subunits of the archaeal exosome in Methanoccocales are missing (Fig. 1). This is also the case in the halophilic euryarchaeon Halobacterium salinarum NRC-1 which instead possesses a bacterial-like RNase R (Fig. 1) with transcript stabilities ranging from 5 min to more than 18 min (Hundt et al.2007). However, to date no specific mRNA-endoribonucleolytic cleavage signal that could initiate RNA decay has been identified even though the aCPSF1 endoribonuclease is absolutely conserved in complete annotated archaeal genomes (Figs 1 and 8) (see Clouet-d’Orval et al.2015 for a review). Specific mRNA features Regardless of the similarities between archaeal and eukaryal transcription machineries, archaeal mRNAs show bacterial features with mostly no introns, no 5΄-methyl caps and no long 3΄-poly A tails and potentially polycistronic structures (Fig. 8). Nonetheless, in vitro and in vivo studies using the crenarchaeota S. solfataricus showed evidence for cap-like structure in which the archaeal translation initiation factor aIF2-γ subunit binds to the RNA 5΄-triphosphorylated ends and protects transcripts from a 5΄-end-dependent degradation pathway (Hasenöhrl et al.2008; Arkhipova et al.2015). In this instance, the aIF2-γ subunit resembles the eukaryotic cap complex strongly suggesting that a directional 5΄-to-3΄ degradation pathway may exist in archaeal cells. This is supported by the ubiquitous occurrence of a 5΄-to-3΄ exoribonucleolytic activity carried by members of the archaeal β-CASP family (Fig. 1). However, in Archaea, while 5΄-to-3΄ exoribonucleases, namely aRNase J and aCPSF2, have been identified to have a preference for monophosphorylated RNA substrates in Euryarchaea and Crenarchaea (Clouet-d’Orval et al.2010; Hasenöhrl, Konrat and Bläsi 2011), the molecular mechanism of 5΄-to-3΄ mRNA decay remains unknown and an enzyme capable of converting the 5΄-triphosphorylated ends of their mRNAs to monophosphates remains to be discovered. In eukaryotic cells, at least two mRNA decapping pyrophosphohydrolases of the Nudix superfamily, Dcp2 and Nudt16, remove the 5΄-cap structure (Arribas-Layton et al.2013). In Bacteria, the triphosphorylated 5΄-termini are converted to monophosphorylated ends by the Nudix superfamily RppH (Celesnik, Deana and Belasco 2007). Although Nudix hydrolases are widely distributed among the three domains of life, archaeal Nudix hydrolases with ‘decapping’ activity have not yet been identified. Two of these proteins from S. solfataricus were cloned and expressed in heterologous system but no pyrophosphohydrolase activity could be detected in in vitro assays (Hasenöhrl, Konrat and Bläsi 2011). It has been shown that large 5΄-UTRs contribute to the mRNA stability of methanogenesis genes in M. mazei, (Cao et al.2014). This is in contrast to other Euryarchaeota and Crenarchaeota in which leaderless transcripts are dominant (Babski et al.2016), and to eukaryal organisms in which mRNA stability is mainly regulated by elements embedded in the transcript 3΄-UTR (see Geissler and Grimson 2016 for a review). Finally, in many cases described above, mRNAs from genes known or predicted to be transcribed in operons exhibited similar half-lives. This was confirmed in a methanogenic archaeon by recent findings allowing the genome-wide mapping of the RNA-processing site (PSSs) obtained by the specific sequencing of the 5΄-monophosphate transcripts that arise from processing events (Qi et al.2017). This approach detected overrepresented 5΄-untranslated and intergenic regions in the polycistronic operons encoding ribosomal proteins signifying that mechanisms of ribosomal protein synthesis and stoichiometry may primarily consist of mRNA-processing–mediated post-transcriptional regulation (Qi et al.2017). Furthermore, archaeal regulatory sRNAs are undoubtedly involved in many biological processes, including metabolic regulation, adaptation to extreme conditions and stress response presumably by altering translation and stability of specific target mRNAs as reported in Bacteria and Eukarya (Marchfelder et al.2012; Babski et al.2014). In the last decade, three classes of small ncRNAs, namely intergenic, antisense and sense RNAs, that can target either the 5΄-UTRs or the 3΄-UTRs of their respective target mRNAs have been detected in several Crenarchaeota and Euryarchaeota (Babski et al.2014). These include A. fulgidus (Tang et al.2002), S. solfataricus (Tang et al.2005), Halophiles (Soppa et al.2009; Straub et al.2009; Fischer et al.2011), M. psychrophilus (Li et al.2015), T. kodakarensis (Jäger et al.2014), M. mazei (Jäger et al.2009) and P. abyssi (Phok et al.2011; Toffano-Nioche et al.2013). Some are differentially expressed with growth conditions (Straub et al.2009; Jaschinski et al.2014). Nevertheless, a mystery remains regarding mRNA decay directly mediated by sRNA since no RNase III-like double-stranded endoribonuclease activity has been detected so far in archaeal cells. It is conceivable that, in Archaea, the EndA activity cleaving within BHB motifs overcomes the lack of double strand-specific RNase III-like activity. Role of SmAPs in mRNA stability? The SmAPs (described in section ‘Archaeal Sm-like proteins’) could play important physiological functions in archaeal mRNA stability. Indeed, while in Eukaryotes, Sm proteins serve as molecular scaffold for RNP complex assembly, as exemplified by their role in rRNA processing by small nucleolar RNPs, in Bacteria, Hfq acts as an RNA chaperone, facilitating base-pairing interactions between small regulatory ncRNAs (sRNAs) and their mRNA targets, and directly influencing the structures of some RNAs (Mura et al.2013). However, the function of SmAP proteins in archaeal cells is currently mostly unknown but first details have been uncovered in H. volcanii and S. solfataricus. The smAP1 gene is very often encoded immediately upstream or even overlapping the gene for 50S ribosomal protein L37e, suggesting a conserved role in processing or stabilisation of rRNAs (Mura et al.2013). For the SmAP2 paralogue, in most Crenarchaeota and Thaumarchaeota, the gene lies directly upstream of, and on the same strand as, a methionine adenosyl transferase gene that could be involved in DNA or RNA methylation. Additionally, a role in tRNA processing or biogenesis was suggested since A. fulgidus Afu-SmAP1 and Afu-SmAP2 (Törö et al.2001) and Sso-SmAP1 and Sso-SmAP2 (Märtens et al.2015) associate with the RNA molecule of RNase P. Moreover, the interacting proteins and RNA substrates of Hvo-SmAP1 (Fischer et al.2010) and Sso-SmAP1 and Sso-SmAP2 (Märtens et al.2015, 2017) identified by co-immunoprecipitation experiments hint at possible SmAP biological functions. The candidate binding partners belong to similar functional classes (rRNA and tRNA modification and processing, RNA decay and translation) like the partners identified for bacterial Hfq and eukaryotic Sm-like proteins (Fischer et al.2010; Märtens et al.2017). While most interactions between SmAPs and candidate partners remain to be confirmed, a direct interaction between the Sso-SmAP1 and Sso-SmAP2 and the exosomal subunit aDnaG, required for efficient polyadenylation of RNA substrates and that might be required for membrane localisation of the Sso-aExosome, was confirmed by affinity purification (Märtens et al.2017). A link between the exosome and Sm-like proteins also exists in eukaryotes (Luz et al.2010). While intron-containing RNAs are enriched with Sso-SmAP1, suggesting possible functions in tRNA/rRNA processing, Sso-SmAP2 seems to bind predominantly to mRNAs and could therefore be involved in mRNA decay or stability (Märtens et al.2015). Finally, an H. volcanii strain deleted for the Sm1 motif of Hvo-SmAP1 (Δsm1) that does not influence L37e expression or affect cell growth was shown to exhibit enhanced swarming (Maier et al.2015). This observation is corroborated by differential RNA-seq gene expression analysis. Transcripts encoding proteins required for motility are upregulated in the Δsm1 mutant (Maier et al.2015). Whether Hvo-SmAP1 is directly involved in cell motility regulation or affects a downstream regulator like an sRNA remains to be determined (Maier et al.2015). Messenger RNA quality control pathway? Controlling mRNA quality is another important challenge for the cell to prevent the synthesis of harmful proteins derived from aberrant mRNAs. As mentioned earlier, in Eukarya, three different decay pathways stimulate the degradation of aberrant mRNAs and the dissociation of ribosomes by recruiting mRNA surveillance factors at the stalled ribosomes (Fig. 8). Recently, it emerged that the structural basis of these mRNA surveillance pathways could exist in Archaea. This is supported by the identification of an ubiquitous archaeal Pelota/Dom34-like protein (aPelota) (Ragan et al.1996) (described in section ‘Ubiquitous endoribonucleases’, Figs. 1 and 2C). Indeed, aPelota, in a complex with the archaeal translation elongator factor EF1α, could mimic the tRNA•EF-Tu complex bound to the A site of the ribosome similarly to proposed ribosome disassembly pathways in Eukarya (Kobayashi et al.2010; Saito et al.2010; Becker et al.2012). A structural basis and a universal mechanism for archaeal and eukaryal ribosome recycling were proposed from cryo-electron microscopy reconstructions of eukaryotic and archaeal ribosome recycling complexes containing the ATPase ABCE1 and the Pelota termination factor paralogue (Becker et al.2012). This is in agreement with early findings showing that the molecular mimicry of tRNA in the distorted ‘A/T state’ conformation by Pelota-like proteins enables the complex to efficiently detect and enter the empty A site of the stalled ribosome which involves the interaction of the positively charged patch of Pelota-like domain A with the decoding centre (Kobayashi et al.2010; Franckenberg, Becker and Beckmann 2012). Future work should determine if translation arrests induce endoribonucleolytic cleavages of the aberrant mRNA by aPelota bound to the A site of the ribosome and recruit other degradation factors, such as the exosome or β-CASP enzymes with the 3΄-to-5΄ and 5΄-to-3΄ exoribonucleolytic activities, respectively, to eliminate the aberrant mRNA fragments. A first clue towards the biological relevance of the 5΄-to-3΄ β-CASP ribonuclease of the aCPSF2 group comes from transcriptome analyses of wild-type and aCPSF2-deleted S. acidocaldarius strains which reveal a global effect with differential transcript abundance for 560 genes (Märtens et al.2013). Other indications may come from genomic context analysis where the gene encoding aCPSF1 in the majority of archaeal genomes is adjacent to the gene encoding a proteasome subunit suggesting interplay between RNA and protein-degrading machineries (Koonin, Wolf and Aravind 2001). In addition, a study in H. volcanii identified Hvo-aCPSF1 as intimately linked to partners of the formation of thiolated tRNA (Lys)UUU which includes homologues of ubiquitin-proteasome (Chavarria et al.2014). In conclusion, many open questions remain and coherent pictures of mRNA decay and RNA quality control pathways are still missing in Archaea (Fig. 8). Future work using improved genetic tools for established model organisms coupled with specific high-throughput RNA sequencing technologies, global measurement of mRNA half-lives and comparative genomics will certainly give an input to the field. The expected outcome will definitely illuminate the roles of the conserved 3΄-to-5΄ exosome machineries and β-CASP 5΄-to-3΄ exo/endoribonucleases in specific biological mRNA decay mechanisms. CONCLUDING REMARKS Despite the wealth of information gathered to date regarding RNA maturation, processing, function and life cycles within archaeal cells, many open questions remain, and RNA maturation and degradation pathways in Archaea are for the most part still a mystery. Several key steps of rRNA maturation, for instance, are still far from being understood, and a comprehensive overview of common principles and species-specific peculiarities is about to emerge. Future in-depth landscape exploration of these pathways in different genomic contexts will certainly provide general principles of ribosomal biogenesis and functional assembly in Archaea and will highlight processing and assembly steps shared between Archaea, Bacteria and Eukarya. In general, many ribonucleases and other RNA-modifying enzymes are still expected to be discovered. Challenges will be to identify the key events and to understand how the actors of RNA metabolism are specifically recruited in pathways controlling archaeal RNA life cycle. Furthermore, it should not be forgotten that archaeal species have very diverse ecological niches, and may have evolved specific and biochemically unique pathways, in addition to the conserved ones, to maintain the integrity of their RNAs in extreme conditions. Finally, deciphering RNA metabolism pathways at work in the recently discovered archaeal lineages will enlarge evolutionary perspectives in understanding the players acting in RNA biology throughout the three distinct phylogenetic domains, especially in the context of the recent lively debate that oppose the standard three-domain (Woese, Kandler and Wheelis 1990) to the two-domain (Eocyte) topology models of the tree of life (Koonin 2015; Raymann, Brochier-Armanet and Gribaldo 2015; Adam et al.2017; Da Cunha et al.2017; Eme et al.2017; Van der Gulik, Hoff and Speijer 2017) Acknowledgements We are indebted to Mike Chandler for critical reading of the manuscript and to Christine Gaspin for helpful discussions. FUNDING This work was supported by the Centre National pour la Recherche Scientifique (CNRS); the Idex-emergence program of the ‘Université de Toulouse’ (UPS) [RNarch to BCO]; the French ‘Agence Nationale pour la Recherche’ [ANR-16-CE12-0016-01 to BCO] and by funding of the DFG [MA1538/21-1 to AM]. Conflict of interest. None declared. REFERENCES Achsel T , Stark H , Lührmann R . The Sm domain is an ancient RNA-binding motif with oligo(U) specificity . Proc Natl Acad Sci USA 2001 ; 98 : 3685 – 9 . Google Scholar CrossRef Search ADS PubMed Acinas SG , Marcelino LA , Klepac-Ceraj V et al. Divergence and redundancy of 16S rRNA sequences in genomes with multiple rrn operons . J Bacteriol 2004 ; 186 : 2629 – 35 . Google Scholar CrossRef Search ADS PubMed Adam PS , Borrel G , Brochier-Armanet C et al. The growing tree of Archaea: new perspectives on their diversity, evolution and ecology . ISME J 2017 ; 11 : 2407 – 25 . Google Scholar CrossRef Search ADS PubMed Aebi M , Kirchner G , Chen JY et al. Isolation of a temperature-sensitive mutant with an altered tRNA nucleotidyltransferase and cloning of the gene encoding tRNA nucleotidyltransferase in the yeast Saccharomyces cerevisiae . J Biol Chem 1990 ; 265 : 16216 – 20 . Google Scholar PubMed Agris PF . The importance of being modified: an unrealized code to RNA structure and function . RNA 2015 ; 21 : 552 – 4 . Google Scholar CrossRef Search ADS PubMed Agris PF , Eruysal ER , Narendran A et al. Celebrating wobble decoding: half a century and still much is new . RNA Biol 2017 ; 1 – 17 , DOI: 10.1080/15476286.2017.1356562. Albers S-V , Forterre P , Prangishvili D et al. The legacy of Carl Woese and Wolfram Zillig: from phylogeny to landmark discoveries . Nat Rev Microbiol 2013 ; 11 : 713 – 9 . Google Scholar CrossRef Search ADS PubMed Altman S . Ribonuclease P . Philos T Roy Soc B 2011 ; 366 : 2936 – 41 . Amero CD , Boomershine WP , Xu Y et al. Solution structure of Pyrococcus furiosus RPP21, a component of the archaeal RNase P holoenzyme, and interactions with its RPP29 protein partner . Biochemistry 2008 ; 47 : 11704 – 10 . Google Scholar CrossRef Search ADS PubMed Andersson AF , Lundgren M , Eriksson S et al. Global analysis of mRNA stability in the archaeon Sulfolobus . Genome Biol 2006 ; 7 : R99 . Google Scholar CrossRef Search ADS PubMed Aravind L . An evolutionary classification of the metallo-beta-lactamase fold proteins . In Silico Biol (Gedrukt) 1999 ; 1 : 69 – 91 . Google Scholar PubMed Arkhipova V , Stolboushkina E , Kravchenko O et al. Binding of the 5΄-triphosphate end of mRNA to the γ-subunit of translation initiation factor 2 of the Crenarchaeon Sulfolobus solfataricus . J Mol Biol 2015 ; 427 : 3086 – 95 . Google Scholar CrossRef Search ADS PubMed Arraiano CM , Mauxion F , Viegas SC et al. Intracellular ribonucleases involved in transcript processing and decay: precision tools for RNA . Biochim Biophys Acta 2013 ; 1829 : 491 – 513 . Google Scholar CrossRef Search ADS PubMed Arribas-Layton M , Wu D , Lykke-Andersen J et al. Structural and functional control of the eukaryotic mRNA decapping machinery . Biochim Biophys Acta 2013 ; 1829 : 580 – 9 . Google Scholar CrossRef Search ADS PubMed Atkinson GC , Baldauf SL , Hauryliuk V . Evolution of nonstop, no-go and nonsense-mediated mRNA decay and their termination factor-derived components . BMC Evol Biol 2008 ; 8 : 290 . Google Scholar CrossRef Search ADS PubMed Babski J , Haas KA , Näther-Schindler D et al. Genome-wide identification of transcriptional start sites in the haloarchaeon Haloferax volcanii based on differential RNA-Seq (dRNA-Seq) . BMC Genomics 2016 ; 17 : 629 . Google Scholar CrossRef Search ADS PubMed Babski J , Maier L-K , Heyer R et al. Small regulatory RNAs in Archaea . RNA Biol 2014 ; 11 : 484 – 93 . Google Scholar CrossRef Search ADS PubMed Bachellerie JP , Michot B , Nicoloso M et al. Antisense snoRNAs: a family of nucleolar RNAs with long complementarities to rRNA . Trends Biochem Sci 1995 ; 20 : 261 – 4 . Google Scholar CrossRef Search ADS PubMed Baker DL , Youssef OA , Chastkofsky MIR et al. RNA-guided RNA modification: functional organization of the archaeal H/ACA RNP . Gene Dev 2005 ; 19 : 1238 – 48 . Google Scholar CrossRef Search ADS PubMed Ban N , Nissen P , Hansen J et al. The complete atomic structure of the large ribosomal subunit at 2.4 A resolution . Science 2000 ; 289 : 905 – 20 . Google Scholar CrossRef Search ADS PubMed Bandyra KJ , Bouvier M , Carpousis AJ et al. The social fabric of the RNA degradosome . Biochim Biophys Acta 2013 ; 1829 : 514 – 22 . Google Scholar CrossRef Search ADS PubMed Becker T , Franckenberg S , Wickles S et al. Structural basis of highly conserved ribosome recycling in eukaryotes and archaea . Nature 2012 ; 482 : 501 – 6 . Google Scholar CrossRef Search ADS PubMed Berkemer SJ , Höner ZU , Siederdissen C et al. Processed small rnas in archaea and BHB elements . Genomics Comput Biol 2015 ; 1 : 18 . Google Scholar CrossRef Search ADS Bernick DL , Dennis PP , Lui LM et al. Diversity of antisense and other non-coding RNAs in archaea revealed by comparative small RNA sequencing in four pyrobaculum species . Front Microbiol 2012 ; 3 : 231 . Google Scholar PubMed Bhuiya MW , Suryadi J , Zhou Z et al. Structure of the Aeropyrum pernix L7Ae multifunctional protein and insight into its extreme thermostability . Acta Crystallogr F Struct Biol Cryst Commun 2013 ; 69 : 979 – 88 . Google Scholar CrossRef Search ADS Bini E , Dikshit V , Dirksen K et al. Stability of mRNA in the hyperthermophilic archaeon Sulfolobus solfataricus . RNA 2002 ; 8 : 1129 – 36 . Google Scholar CrossRef Search ADS PubMed Bolhuis H , Palm P , Wende A et al. The genome of the square archaeon Haloquadratum walsbyi: life at the limits of water activity . BMC Genomics 2006 ; 7 : 169 . Google Scholar CrossRef Search ADS PubMed Bolotin A , Quinquis B , Sorokin A et al. Clustered regularly interspaced short palindrome repeats (CRISPRs) have spacers of extrachromosomal origin . Microbiology (Reading, Engl) 2005 ; 151 : 2551 – 61 . Google Scholar CrossRef Search ADS PubMed Boomershine W.P. , McElroy C.A. , Tsai H.-Y. , Wilson R.C. , Gopalan V. , Foster M.P. Structure of Mth11/Mth Rpp29, an essential protein subunit of archaeal and eukaryotic RNase P . Proceedings of the National Academy of Sciences of the United States of America 2003 ; 100 : 15398 – 403 . Google Scholar CrossRef Search ADS PubMed Bortolin M-L , Bachellerie J-P , Clouet-d’Orval B . In vitro RNP assembly and methylation guide activity of an unusual box C/D RNA, cis-acting archaeal pre-tRNATrp . Nucleic Acids Res 2003 ; 31 : 6524 – 35 . Google Scholar CrossRef Search ADS PubMed Brochier C , Gribaldo S , Zivanovic Y et al. Nanoarchaea: representatives of a novel archaeal phylum or a fast-evolving euryarchaeal lineage related to Thermococcales? Genome Biol 2005 ; 6 : R42 . Google Scholar CrossRef Search ADS PubMed Brochier-Armanet C , Forterre P , Gribaldo S . Phylogeny and evolution of the Archaea: one hundred genomes later . Curr Opin Microbiol 2011 ; 14 : 274 – 81 . Google Scholar CrossRef Search ADS PubMed Brouns SJJ , Jore MM , Lundgren M et al. Small CRISPR RNAs guide antiviral defense in prokaryotes . Science 2008 ; 321 : 960 – 4 . Google Scholar CrossRef Search ADS PubMed Bubeck D , Reijns MAM , Graham SC et al. PCNA directs type 2 RNase H activity on DNA replication and repair substrates . Nucleic Acids Res 2011 ; 39 : 3652 – 66 . Google Scholar CrossRef Search ADS PubMed Bult CJ , White O , Olsen GJ et al. Complete genome sequence of the methanogenic archaeon, Methanococcus jannaschii . Science 1996 ; 273 : 1058 – 73 . Google Scholar CrossRef Search ADS PubMed Burggraf S , Larsen N , Woese CR et al. An intron within the 16S ribosomal RNA gene of the archaeon Pyrobaculum aerophilum. Proc Natl Acad Sci USA 1993 ; 90 : 2547 – 50 . Google Scholar CrossRef Search ADS PubMed Büttner K , Wenig K , Hopfner K-P . Structural framework for the mechanism of archaeal exosomes in RNA processing . Mol Cell 2005 ; 20 : 461 – 71 . Google Scholar CrossRef Search ADS PubMed Calvin K , Li H . RNA-splicing endonuclease structure and function . Cell Mol Life Sci 2008 ; 65 : 1176 – 85 . Google Scholar CrossRef Search ADS PubMed Cantara WA , Crain PF , Rozenski J et al. The RNA modification database, RNAMDB: 2011 update . Nucleic Acids Res 2011 ; 39 : D195 – 201 . Google Scholar CrossRef Search ADS PubMed Cao Y , Li J , Jiang N et al. Mechanism for stabilizing mRNAs involved in methanol-dependent methanogenesis of cold-adaptive Methanosarcina mazei zm-15 . Appl Environ Microb 2014 ; 80 : 1291 – 8 . Google Scholar CrossRef Search ADS Carte J , Wang R , Li H et al. Cas6 is an endoribonuclease that generates guide RNAs for invader defense in prokaryotes . Gene Dev 2008 ; 22 : 3489 – 96 . Google Scholar CrossRef Search ADS PubMed Celesnik H , Deana A , Belasco JG . Initiation of RNA decay in Escherichia coli by 5΄ pyrophosphate removal . Mol Cell 2007 ; 27 : 79 – 90 . Google Scholar CrossRef Search ADS PubMed Chan PP , Cozen AE , Lowe TM . Discovery of permuted and recently split transfer RNAs in Archaea . Genome Biol 2011 ; 12 : R38 . Google Scholar CrossRef Search ADS PubMed Chan PP , Lowe TM . GtRNAdb: a database of transfer RNA genes detected in genomic sequence . Nucleic Acids Res 2009 ; 37 : D93 – 7 . Google Scholar CrossRef Search ADS PubMed Chant J , Dennis P . Archaebacteria: transcription and processing of ribosomal RNA sequences in Halobacterium cutirubrum . EMBO J 1986 ; 5 : 1091 – 7 . Google Scholar PubMed Chapados BR , Chai Q , Hosfield DJ et al. Structural biochemistry of a type 2 RNase H: RNA primer recognition and removal during DNA replication . J Mol Biol 2001 ; 307 : 541 – 56 . Google Scholar CrossRef Search ADS PubMed Charpentier B , Muller S , Branlant C . Reconstitution of archaeal H/ACA small ribonucleoprotein complexes active in pseudouridylation . Nucleic Acids Res 2005 ; 33 : 3133 – 44 . Google Scholar CrossRef Search ADS PubMed Charpentier E , Richter H , van der Oost J et al. Biogenesis pathways of RNA guides in archaeal and bacterial CRISPR-Cas adaptive immunity . FEMS Microbiol Rev 2015 ; 39 : 428 – 41 . Google Scholar CrossRef Search ADS PubMed Charpentier E , van der Oost J , White MF . crRNA Biogenesis . In: Barrangou R , van der Oost J (eds.) CRISPR-Cas Systems, Berlin, Heidelberg : Springer , 2013 , 115 – 44 . Charron C , Manival X , Cléry A et al. The archaeal sRNA binding protein L7Ae has a 3D structure very similar to that of its eukaryal counterpart while having a broader RNA-binding specificity . J Mol Biol 2004a ; 342 : 757 – 73 . Google Scholar CrossRef Search ADS Charron C. , Manival X. , Charpentier B. , Branlant C. , Aubry A. Purification, crystallization and preliminary X-ray diffraction data of L7Ae sRNP core protein from Pyrococcus abyssii . Acta Crystallographica. Sect. D, Biological Crystallography 2004b ; 60 : 122 – 4 . Google Scholar CrossRef Search ADS Chavarria NE , Hwang S , Cao S et al. Archaeal Tuc1/Ncs6 homolog required for wobble uridine tRNA thiolation is associated with ubiquitin-proteasome, translation, and RNA processing system homologs . PLoS One 2014 ; 9 : e99104 . Google Scholar CrossRef Search ADS PubMed Chen W-Y , Pulukkunat DK , Cho I-M et al. Dissecting functional cooperation among protein subunits in archaeal RNase P, a catalytic ribonucleoprotein complex . Nucleic Acids Res 2010 ; 38 : 8316 – 27 . Google Scholar CrossRef Search ADS PubMed Chen W-Y , Singh D , Lai LB et al. Fidelity of tRNA 5΄-maturation: a possible basis for the functional dependence of archaeal and eukaryal RNase P on multiple protein cofactors . Nucleic Acids Res 2012 ; 40 : 4666 – 80 . Google Scholar CrossRef Search ADS PubMed Chen W-Y , Xu Y , Cho I-M et al. Cooperative RNP assembly: complementary rescue of structural defects by protein and RNA subunits of archaeal RNase P . J Mol Biol 2011 ; 411 : 368 – 83 . Google Scholar CrossRef Search ADS PubMed Chlebowski A , Lubas M , Jensen TH et al. RNA decay machines: the exosome . Biochim Biophys Acta 2013 ; 1829 : 552 – 60 . Google Scholar CrossRef Search ADS PubMed Cho I-M , Lai LB , Susanti D et al. Ribosomal protein L7Ae is a subunit of archaeal RNase P . Proc Natl Acad Sci USA 2010 ; 107 : 14573 – 8 . Google Scholar CrossRef Search ADS PubMed Ciammaruconi A , Londei P . In vitro processing of the 16S rRNA of the thermophilic archaeon Sulfolobus solfataricus . J Bacteriol 2001 ; 183 : 3866 – 74 . Google Scholar CrossRef Search ADS PubMed Clouet d’Orval B , Bortolin ML , Gaspin C et al. Box C/D RNA guides for the ribose methylation of archaeal tRNAs. The tRNATrp intron guides the formation of two ribose-methylated nucleosides in the mature tRNATrp . Nucleic Acids Res 2001 ; 29 : 4518 – 29 . Google Scholar CrossRef Search ADS PubMed Clouet-d’Orval B , Gaspin C , Mougin A . Two different mechanisms for tRNA ribose methylation in Archaea: a short survey . Biochimie 2005 ; 87 : 889 – 95 . Google Scholar CrossRef Search ADS PubMed Clouet-d’Orval B , Phung DK , Langendijk-Genevaux PS et al. Universal RNA-degrading enzymes in Archaea: prevalence, activities and functions of β-CASP ribonucleases . Biochimie 2015 ; 118 : 278 – 85 . Google Scholar CrossRef Search ADS PubMed Clouet-d’Orval B , Rinaldi D , Quentin Y et al. Euryarchaeal beta-CASP proteins with homology to bacterial RNase J have 5΄- to 3΄-exoribonuclease activity . J Biol Chem 2010 ; 285 : 17574 – 83 . Google Scholar CrossRef Search ADS PubMed Collins BM , Harrop SJ , Kornfeld GD et al. Crystal structure of a heptameric Sm-like protein complex from archaea: implications for the structure and evolution of snRNPs . J Mol Biol 2001 ; 309 : 915 – 23 . Google Scholar CrossRef Search ADS PubMed Crowe BL , Bohlen CJ , Wilson RC et al. Assembly of the complex between archaeal RNase P proteins RPP30 and Pop5 . Archaea 2011 ; 2011 : 1 – 12 . Google Scholar CrossRef Search ADS Da Cunha V , Gaia M , Gadelle D et al. Lokiarchaea are close relatives of Euryarchaeota, not bridging the gap between prokaryotes and eukaryotes . PLoS Genet 2017 ; 13 : e1006810 . Google Scholar CrossRef Search ADS PubMed Danan M , Schwartz S , Edelheit S et al. Transcriptome-wide discovery of circular RNAs in Archaea . Nucleic Acids Res 2012 ; 40 : 3131 – 42 . Google Scholar CrossRef Search ADS PubMed Daume M , Uhl M , Backofen R et al. RIP-Seq suggests translational regulation by L7Ae in Archaea . MBio 2017 ; 8 : e00730-17 . Google Scholar CrossRef Search ADS PubMed De Koning B , Blombach F , Brouns SJJ et al. Fidelity in archaeal information processing . Archaea 2010 ; 2010 , DOI: https://doi.org/10.1155/2010/960298 . Dennis PP , Tripp V , Lui L et al. C/D box sRNA-guided 2’-O-methylation patterns of archaeal rRNA molecules . BMC Genomics 2015 ; 16 : 632 . Google Scholar CrossRef Search ADS PubMed Dennis PP , Ziesche S , Mylvaganam S . Transcription analysis of two disparate rRNA operons in the halophilic archaeon Haloarcula marismortui . J Bacteriol 1998 ; 180 : 4804 – 13 . Google Scholar PubMed Deppenmeier U , Johann A , Hartsch T et al. The genome of Methanosarcina mazei: evidence for lateral gene transfer between bacteria and archaea . J Mol Microbiol Biot 2002 ; 4 : 453 – 61 . Deutscher MP . Ribonucleases, tRNA nucleotidyltransferase, and the 3΄ processing of tRNA . Prog Nucleic Acid Re 1990 ; 39 : 209 – 40 . Google Scholar CrossRef Search ADS Deutscher MP . Chapter 9 maturation and degradation of ribosomal RNA in bacteria . In Condon C (ed). Molecular Biology of RNA Processing and Decay in Prokaryotes . Elsevier , 2009 , 369 – 91 . Deutscher MP . How bacterial cells keep ribonucleases under control . FEMS Microbiol Rev 2015 ; 39 : 350 – 61 . Google Scholar CrossRef Search ADS PubMed De Zoysa MD , Yu Y-T . Posttranscriptional RNA pseudouridylation . Enzymes 2017 ; 41 : 151 – 67 . Google Scholar CrossRef Search ADS PubMed Dominski Z , Carpousis AJ , Clouet-d’Orval B . Emergence of the β-CASP ribonucleases: Highly conserved and ubiquitous metallo-enzymes involved in messenger RNA maturation and degradation . Biochim Biophys Acta 2013 ; 1829 : 532 – 51 . Google Scholar CrossRef Search ADS PubMed Dorléans A , Li de la Sierra-Gallay I , Piton J et al. Molecular basis for the recognition and cleavage of RNA by the bifunctional 5΄-3΄ Exo/Endoribonuclease RNase J . Structure 2011 ; 19 : 1252 – 61 . Google Scholar CrossRef Search ADS PubMed Durand S , Tomasini A , Braun F et al. sRNA and mRNA turnover in Gram-positive bacteria . FEMS Microbiol Rev 2015 ; 39 : 316 – 30 . Google Scholar CrossRef Search ADS PubMed Durovic P , Dennis PP . Separate pathways for excision and processing of 16S and 23S rRNA from the primary rRNA operon transcript from the hyperthermophilic archaebacterium Sulfolobus acldocaldarius: similarities to eukaryotic rRNA processing . Mol Microbiol 1994 ; 13 : 229 – 42 . Google Scholar CrossRef Search ADS El Yacoubi B , Bailly M , de Crécy-Lagard V . Biosynthesis and function of posttranscriptional modifications of transfer RNAs . Annu Rev Genet 2012 ; 46 : 69 – 95 . Google Scholar CrossRef Search ADS PubMed Elkins JG , Podar M , Graham DE et al. A korarchaeal genome reveals insights into the evolution of the Archaea . Proc Natl Acad Sci USA 2008 ; 105 : 8102 – 7 . Google Scholar CrossRef Search ADS PubMed Ellis JC , Brown JW . The RNase P family . RNA Biol 2009 ; 6 : 362 – 9 . Google Scholar CrossRef Search ADS PubMed Eme L , Spang A , Lombard J et al. Archaea and the origin of eukaryotes . Nat Rev Microbiol 2017 ; 15 : 711 – 23 . Google Scholar CrossRef Search ADS PubMed Englert M , Sheppard K , Aslanian A et al. Archaeal 3'-phosphate RNA splicing ligase characterization identifies the missing component in tRNA maturation . Proc Natl Acad Sci USA 2011 ; 108 : 1290 – 5 . Google Scholar CrossRef Search ADS PubMed Englert M , Xia S , Okada C et al. Structural and mechanistic insights into guanylylation of RNA-splicing ligase RtcB joining RNA between 3'-terminal phosphate and 5'-OH . Proc Natl Acad Sci 2012 ; 109 : 15235 – 40 . Google Scholar CrossRef Search ADS PubMed Esakova O , Krasilnikov AS . Of proteins and RNA: the RNase P/MRP family . RNA 2010 ; 16 : 1725 – 47 . Google Scholar CrossRef Search ADS PubMed Evans D , Marquez SM , Pace NR . RNase P: interface of the RNA and protein worlds . Trends Biochem Sci 2006 ; 31 : 333 – 41 . Google Scholar CrossRef Search ADS PubMed Even S , Pellegrini O , Zig L et al. Ribonucleases J1 and J2: two novel endoribonucleases in B. subtilis with functional homology to E.coli RNase E . Nucleic Acids Res 2005 ; 33 : 2141 – 52 . Google Scholar CrossRef Search ADS PubMed Evguenieva-Hackenberg E , Hou L , Glaeser S et al. Structure and function of the archaeal exosome . WIRES RNA 2014 ; 5 : 623 – 35 . Google Scholar CrossRef Search ADS PubMed Evguenieva-Hackenberg E , Klug G . RNA degradation in Archaea and Gram-negative bacteria different from Escherichia coli . Prog Mol Biol Transl Sci 2009 ; 85 : 275 – 317 . Google Scholar CrossRef Search ADS PubMed Evguenieva-Hackenberg E , Klug G . New aspects of RNA processing in prokaryotes . Curr Opin Microbiol 2011 ; 14 : 587 – 92 . Google Scholar CrossRef Search ADS PubMed Evguenieva-Hackenberg E , Roppelt V , Finsterseifer P et al. Rrp4 and Csl4 are needed for efficient degradation but not for polyadenylation of synthetic and natural RNA by the archaeal exosome . Biochemistry 2008 ; 47 : 13158 – 68 . Google Scholar CrossRef Search ADS PubMed Evguenieva-Hackenberg E , Walter P , Hochleitner E et al. An exosome-like complex in Sulfolobus solfataricus . EMBO Rep 2003 ; 4 : 889 – 93 . Google Scholar CrossRef Search ADS PubMed Farhoud MH , Wessels HJCT , Steenbakkers PJM et al. Protein complexes in the Archaeon Methanothermobacter thermautotrophicus analyzed by blue native/SDS-PAGE and mass spectrometry . Mol Cell Proteomics 2005 ; 4 : 1653 – 63 . Google Scholar CrossRef Search ADS PubMed Fatica A , Tollervey D , Dlakić M . PIN domain of Nob1p is required for D-site cleavage in 20S pre-rRNA . RNA 2004 ; 10 : 1698 – 701 . Google Scholar CrossRef Search ADS PubMed Ferreira-Cerca S . Life and death of ribosomes in archaea . In Clouet-d’Orval , B (ed.) RNA Metabolism and Gene Expression in Archaea . Cham : Springer International Publishing , 2017 , 129 – 58 . Fischer S , Benz J , Späth B et al. The archaeal Lsm protein binds to small RNAs . J Biol Chem 2010 ; 285 : 34429 – 38 . Google Scholar CrossRef Search ADS PubMed Fischer S , Benz J , Späth B et al. Regulatory RNAs in Haloferax volcanii . Biochem Soc Trans 2011 ; 39 : 159 – 62 . Google Scholar CrossRef Search ADS PubMed Franckenberg S , Becker T , Beckmann R . Structural view on recycling of archaeal and eukaryotic ribosomes after canonical termination and ribosome rescue . Curr Opin Struct Biol 2012 ; 22 : 786 – 96 . Google Scholar CrossRef Search ADS PubMed Frank DN , Pace NR . Ribonuclease P: unity and diversity in a tRNA processing ribozyme . Annu Rev Biochem 1998 ; 67 : 153 – 80 . Google Scholar CrossRef Search ADS PubMed Franze de Fernandez MT , Eoyang L , August JT . Factor fraction required for the synthesis of bacteriophage Qβ-RNA . Nature 1968 ; 219 : 588 – 90 . Google Scholar CrossRef Search ADS PubMed Fujishima K , Kanai A . tRNA gene diversity in the three domains of life . Front Genet 2014 ; 5 : 142 . Google Scholar CrossRef Search ADS PubMed Fujishima K , Sugahara J , Kikuta K et al. Tri-split tRNA is a transfer RNA made from 3 transcripts that provides insight into the evolution of fragmented tRNAs in archaea . Proc Natl Acad Sci USA 2009 ; 106 : 2683 – 7 . Google Scholar CrossRef Search ADS PubMed Fujishima K , Sugahara J , Miller CS et al. A novel three-unit tRNA splicing endonuclease found in ultrasmall Archaea possesses broad substrate specificity . Nucleic Acids Res 2011 ; 39 : 9695 – 704 . Google Scholar CrossRef Search ADS PubMed Fujishima K , Sugahara J , Tomita M et al. Large-scale tRNA intron transposition in the archaeal order Thermoproteales represents a novel mechanism of intron gain . Mol Biol Evol 2010 ; 27 : 2233 – 43 . Google Scholar CrossRef Search ADS PubMed Fukuhara H , Kifusa M , Watanabe M et al. A fifth protein subunit Ph1496p elevates the optimum temperature for the ribonuclease P activity from Pyrococcus horikoshii OT3 . Biochem Bioph Res Co 2006 ; 343 : 956 – 64 . Google Scholar CrossRef Search ADS Gagnon KT , Zhang X , Qu G et al. Signature amino acids enable the archaeal L7Ae box C/D RNP core protein to recognize and bind the K-loop RNA motif . RNA 2010 ; 16 : 79 – 90 . Google Scholar CrossRef Search ADS PubMed Gao X , Oshima K , Ueda T et al. A three-dimensional model of RNase P in the hyperthermophilic archaeon Pyrococcus horikoshii OT3 . Biochem Bioph Res Co 2017 ; 493 : 1063 - 68 . Google Scholar CrossRef Search ADS Garside EL , Schellenberg MJ , Gesner EM et al. Cas5d processes pre-crRNA and is a member of a larger family of CRISPR RNA endonucleases . RNA 2012 ; 18 : 2020 – 8 . Google Scholar CrossRef Search ADS PubMed Gaspin C , Cavaillé J , Erauso G et al. Archaeal homologs of eukaryotic methylation guide small nucleolar RNAs: lessons from the Pyrococcus genomes . J Mol Biol 2000 ; 297 : 895 – 906 . Google Scholar CrossRef Search ADS PubMed Gebetsberger J , Polacek N . Slicing tRNAs to boost functional ncRNA diversity . RNA Biol 2013 ; 10 : 1798 – 806 . Google Scholar CrossRef Search ADS PubMed Gebetsberger J , Wyss L , Mleczko AM et al. A tRNA-derived fragment competes with mRNA for ribosome binding and regulates translation during stress . RNA Biol 2017 ; 14 : 1364 – 73 . Google Scholar CrossRef Search ADS PubMed Gebetsberger J , Zywicki M , Künzi A et al. tRNA-Derived fragments target the ribosome and function as regulatory non-coding RNA in Haloferax volcanii . Archaea 2012 ; 2012 : 1 – 11 . Google Scholar CrossRef Search ADS Geissler R , Grimson A . A position-specific 3΄UTR sequence that accelerates mRNA decay . RNA Biol 2016 ; 13 : 1075 – 7 . Google Scholar CrossRef Search ADS PubMed Gopalan V . Uniformity amid diversity in RNase P . Proc Natl Acad Sci USA 2007 ; 104 : 2031 – 2 . Google Scholar CrossRef Search ADS PubMed Grosjean H , de Crécy-Lagard V , Marck C . Deciphering synonymous codons in the three domains of life: co-evolution with specific tRNA modification enzymes . FEBS Lett 2010 ; 584 : 252 – 64 . Google Scholar CrossRef Search ADS PubMed Grosjean H , Gaspin C , Marck C et al. RNomics and Modomics in the halophilic archaea Haloferax volcanii: identification of RNA modification genes . BMC Genomics 2008 ; 9 : 470 . Google Scholar CrossRef Search ADS PubMed Grosjean H , Westhof E . An integrated, structure- and energy-based view of the genetic code . Nucleic Acids Res 2016 ; 44 : 8020 – 40 . Google Scholar CrossRef Search ADS PubMed Guerrier-Takada C , Gardiner K , Marsh T et al. The RNA moiety of ribonuclease P is the catalytic subunit of the enzyme . Cell 1983 ; 35 : 849 – 57 . Google Scholar CrossRef Search ADS PubMed Gulen B , Petrov AS , Okafor CD et al. Ribosomal small subunit domains radiate from a central core . Sci Rep 2016 ; 6 : 20885 . Google Scholar CrossRef Search ADS PubMed Guy L , Ettema TJG . The archaeal ‘TACK’ superphylum and the origin of eukaryotes . Trends Microbiol 2011 ; 19 : 580 – 7 . Google Scholar CrossRef Search ADS PubMed Haas ES , Daniels CJ , Reeve JN . Genes encoding 5S rRNA and tRNAs in the extremely thermophilic archaebacterium Methanothermus fervidus . Gene 1989 ; 77 : 253 – 63 . Google Scholar CrossRef Search ADS PubMed Hale C , Kleppe K , Terns RM et al. Prokaryotic silencing (psi)RNAs in Pyrococcus furiosus . RNA 2008 ; 14 : 2572 – 9 . Google Scholar CrossRef Search ADS PubMed Hale CR , Zhao P , Olson S et al. RNA-guided RNA cleavage by a CRISPR RNA-Cas protein complex . Cell 2009 ; 139 : 945 – 56 . Google Scholar CrossRef Search ADS PubMed Hall TA , Brown JW . The ribonuclease P family . Method Enzymol 2001 ; 341 : 56 – 77 . Google Scholar CrossRef Search ADS Hall TA , Brown JW . Archaeal RNase P has multiple protein subunits homologous to eukaryotic nuclear RNase P proteins . RNA 2002 ; 8 : 296 – 306 . Google Scholar CrossRef Search ADS PubMed Hallam SJ , Konstantinidis KT , Putnam N et al. Genomic analysis of the uncultivated marine crenarchaeote Cenarchaeum symbiosum . Proc Natl Acad Sci USA 2006 ; 103 : 18296 – 301 . Google Scholar CrossRef Search ADS PubMed Hamma T , Ferré-D’Amaré AR . Structure of protein L7Ae bound to a K-Turn derived from an archaeal Box H/ACA sRNA at 1.8 Å resolution . Structure 2004 ; 12 : 893 – 903 . Google Scholar CrossRef Search ADS PubMed Harris JK , Haas ES , Williams D et al. New insight into RNase P RNA structure from comparative analysis of the archaeal RNA . RNA 2001 ; 7 : 220 – 32 . Google Scholar CrossRef Search ADS PubMed Hartman AL , Norais C , Badger JH et al. The complete genome sequence of Haloferax volcanii DS2, a model archaeon . PLoS One 2010 ; 5 : e9605 . Google Scholar CrossRef Search ADS PubMed Hartmann E , Hartmann RK . The enigma of ribonuclease P evolution . Trends Genet 2003 ; 19 : 561 – 9 . Google Scholar CrossRef Search ADS PubMed Hartmann RK , Gössringer M , Späth B et al. The making of tRNAs and more - RNase P and tRNase Z . Prog Mol Biol Transl Sci 2009 ; 85 : 319 – 68 . Google Scholar CrossRef Search ADS PubMed Hartung S , Hopfner K-P . Lessons from structural and biochemical studies on the archaeal exosome . Biochem Soc Trans 2009 ; 37 : 83 – 7 . Google Scholar CrossRef Search ADS PubMed Hartung S , Niederberger T , Hartung M et al. Quantitative analysis of processive RNA degradation by the archaeal RNA exosome . Nucleic Acids Res 2010 ; 38 : 5166 – 76 . Google Scholar CrossRef Search ADS PubMed Hasenöhrl D , Konrat R , Bläsi U Identification of an RNase J ortholog in sulfolobus solfataricus: implications for 5'-to-3' directional decay and 5'-end protection of mRNA in Crenarchaeota . RNA 2011 ; 17 : 99 – 107 . Google Scholar CrossRef Search ADS PubMed Hasenöhrl D , Lombo T , Kaberdin V et al. Translation initiation factor a/eIF2(-gamma) counteracts 5' to 3' mRNA decay in the archaeon Sulfolobus solfataricus . Proc Natl Acad Sci USA 2008 ; 105 : 2146 – 50 . Google Scholar CrossRef Search ADS PubMed Hatoum-Aslan A , Maniv I , Marraffini LA . Mature clustered, regularly interspaced, short palindromic repeats RNA (crRNA) length is measured by a ruler mechanism anchored at the precursor processing site . Proc Natl Acad Sci USA 2011 ; 108 : 21218 – 22 . Google Scholar CrossRef Search ADS PubMed Heider MR , Burkhart BW , Santangelo TJ et al. Defining the RNaseH2 enzyme-initiated ribonucleotide excision repair pathway in Archaea . J Biol Chem 2017 ; 292 : 8835 – 45 . Google Scholar CrossRef Search ADS PubMed Hellmich UA , Weis BL , Lioutikov A et al. Essential ribosome assembly factor Fap7 regulates a hierarchy of RNA-protein interactions during small ribosomal subunit biogenesis . Proc Natl Acad Sci USA 2013 ; 110 : 15253 – 8 . Google Scholar CrossRef Search ADS PubMed Hennigan AN , Reeve JN . mRNAs in the methanogenic archaeon Methanococcus vannielii: numbers, half-lives and processing . Mol Microbiol 1994 ; 11 : 655 – 70 . Google Scholar CrossRef Search ADS PubMed Henras AK , Plisson-Chastang C , O’Donohue M-F et al. An overview of pre-ribosomal RNA processing in eukaryotes . WIRES RNA 2015 ; 6 : 225 – 42 . Google Scholar CrossRef Search ADS PubMed Heyer R , Dörr M , Jellen-Ritter A et al. High throughput sequencing reveals a plethora of small RNAs including tRNA derived fragments in Haloferax volcanii . RNA Biol 2012 ; 9 : 1011 – 8 . Google Scholar CrossRef Search ADS PubMed Hirata A , Fujishima K , Yamagami R et al. X-ray structure of the fourth type of archaeal tRNA splicing endonuclease: insights into the evolution of a novel three-unit composition and a unique loop involved in broad substrate specificity . Nucleic Acids Res 2012 ; 40 : 10554 – 66 . Google Scholar CrossRef Search ADS PubMed Hirata A , Kitajima T , Hori H . Cleavage of intron from the standard or non-standard position of the precursor tRNA by the splicing endonuclease of Aeropyrum pernix, a hyper-thermophilic Crenarchaeon, involves a novel RNA recognition site in the Crenarchaea specific loop . Nucleic Acids Res 2011 ; 39 : 9376 – 89 . Google Scholar CrossRef Search ADS PubMed Hochstrasser ML , Doudna JA . Cutting it close: CRISPR-associated endoribonuclease structure and function . Trends Biochem Sci 2015 ; 40 : 58 – 66 . Google Scholar CrossRef Search ADS PubMed Hölzle A , Fischer S , Heyer R et al. Maturation of the 5S rRNA 5' end is catalyzed in vitro by the endonuclease tRNase Z in the archaeon H. volcanii . RNA 2008 ; 14 : 928 – 37 . Google Scholar CrossRef Search ADS PubMed Honda T , Hara T , Nan J et al. Archaeal homologs of human RNase P protein pairs Pop5 with Rpp30 and Rpp21 with Rpp29 work on distinct functional domains of the RNA subunit . Biosci Biotechnol Biochem 2010 ; 74 : 266 – 73 . Google Scholar CrossRef Search ADS PubMed Honda T , Kakuta Y , Kimura K et al. Structure of an archaeal homolog of the human protein complex Rpp21-Rpp29 that is a key core component for the assembly of active ribonuclease P . J Mol Biol 2008 ; 384 : 652 – 62 . Google Scholar CrossRef Search ADS PubMed Hou L , Klug G , Evguenieva-Hackenberg E . The archaeal DnaG protein needs Csl4 for binding to the exosome and enhances its interaction with adenine-rich RNAs . RNA Biol 2013 ; 10 : 415 – 24 . Google Scholar CrossRef Search ADS PubMed Houri-Zeevi L , Rechavi O A matter of time: small rnas regulate the duration of epigenetic inheritance . Trends Genet 2017 ; 33 : 46 – 57 . Google Scholar CrossRef Search ADS PubMed Howard MJ , Liu X , Lim WH et al. RNase P enzymes . RNA Biol 2013 ; 10 : 909 – 14 . Google Scholar CrossRef Search ADS PubMed Huang L , Lilley DMJ . The molecular recognition of kink-turn structure by the L7Ae class of proteins . RNA 2013 ; 19 : 1703 – 10 . Google Scholar CrossRef Search ADS PubMed Huang L , Lilley DMJ . The kink turn, a key architectural element in RNA structure . J Mol Biol 2016 ; 428 : 790 – 801 . Google Scholar CrossRef Search ADS PubMed Hui MP , Foley PL , Belasco JG . Messenger RNA degradation in bacterial cells . Annu Rev Genet 2014 ; 48 : 537 – 59 . Google Scholar CrossRef Search ADS PubMed Hundt S , Zaigler A , Lange C et al. Global analysis of mRNA decay in Halobacterium salinarum NRC-1 at single-gene resolution using DNA microarrays . J Bacteriol 2007 ; 189 : 6936 – 44 . Google Scholar CrossRef Search ADS PubMed Ikeda Y , Okada Y , Sato A et al. An archaeal RNA binding protein, FAU-1, is a novel ribonuclease related to rRNA stability in Pyrococcus and Thermococcus . Sci Rep 2017 ; 7 : 12674 . Google Scholar CrossRef Search ADS PubMed Inada T . Quality control systems for aberrant mRNAs induced by aberrant translation elongation and termination . Biochim Biophys Acta 2013 ; 1829 : 634 – 42 . Google Scholar CrossRef Search ADS PubMed Inada T . The ribosome as a platform for mRNA and nascent polypeptide quality control . Trends Biochem Sci 2017 ; 42 : 5 – 15 . Google Scholar CrossRef Search ADS PubMed Ishii R , Minagawa A , Takaku H et al. Crystal structure of the tRNA 3΄ processing endoribonuclease tRNase Z from Thermotoga maritima . J Biol Chem 2005 ; 280 : 14138 – 44 . Google Scholar CrossRef Search ADS PubMed Ishii R , Minagawa A , Takaku H et al. The structure of the flexible arm of Thermotoga maritima tRNase Z differs from those of homologous enzymes . Acta Crystallogr F 2007 ; 63 : 637 – 41 . Google Scholar CrossRef Search ADS Itoh T , Nomura N , Sako Y . Distribution of 16S rRNA introns among the family Thermoproteaceae and their evolutionary implications . Extremophiles 2003 ; 7 : 229 – 33 . Google Scholar CrossRef Search ADS PubMed Jäger D , Förstner KU , Sharma CM et al. Primary transcriptome map of the hyperthermophilic archaeon Thermococcus kodakarensis . BMC Genomics 2014 ; 15 : 684 . Google Scholar CrossRef Search ADS PubMed Jäger A , Samorski R , Pfeifer F et al. Individual gvp transcript segments in Haloferax mediterranei exhibit varying half-lives, which are differentially affected by salt concentration and growth phase . Nucleic Acids Res 2002 ; 30 : 5436 – 43 . Google Scholar CrossRef Search ADS PubMed Jäger D , Sharma CM , Thomsen J et al. Deep sequencing analysis of the Methanosarcina mazei Go1 transcriptome in response to nitrogen availability . Proc Natl Acad Sci USA 2009 ; 106 : 21878 – 82 . Google Scholar CrossRef Search ADS PubMed Jarrous N , Gopalan V . Archaeal/eukaryal RNase P: subunits, functions and RNA diversification . Nucleic Acids Res 2010 ; 38 : 7885 – 94 . Google Scholar CrossRef Search ADS PubMed Jaschinski K , Babski J , Lehr M et al. Generation and phenotyping of a collection of sRNA gene deletion mutants of the haloarchaeon Haloferax volcanii . PLoS One 2014 ; 9 : e90763 . Google Scholar CrossRef Search ADS PubMed Jay ZJ , Inskeep , WP . The distribution, diversity, and importance of 16S rRNA gene introns in the order Thermoproteales . Biol Direct 2015 ; 10 : 35 . Google Scholar CrossRef Search ADS PubMed Kaberdin VR , Singh D , Lin-Chao S . Composition and conservation of the mRNA-degrading machinery in bacteria . J Biomed Sci 2011 ; 18 : 23 . Google Scholar CrossRef Search ADS PubMed Kakuta Y , Ishimatsu I , Numata T et al. Crystal structure of a ribonuclease p protein Ph1601p from Pyrococcus horikoshii OT3: an archaeal homologue of human nuclear ribonuclease p protein Rpp21 . Biochemistry 2005 ; 44 : 12086 – 93 . Google Scholar CrossRef Search ADS PubMed Kambach C , Walke S , Nagai K . Structure and assembly of the spliceosomal small nuclear ribonucleoprotein particles . Curr Opin Struct Biol 1999 ; 9 : 222 – 30 . Google Scholar CrossRef Search ADS PubMed Kawano S , Nakashima T , Kakuta Y et al. Crystal structure of protein Ph1481p in complex with protein Ph1877p of archaeal RNase P from Pyrococcus horikoshii OT3: implication of dimer formation of the holoenzyme . J Mol Biol 2006 ; 357 : 583 – 91 . Google Scholar CrossRef Search ADS PubMed Keam SP , Hutvagner G . tRNA-derived fragments (tRFs): emerging new roles for an ancient RNA in the regulation of gene expression . Life (Basel) 2015 ; 5 : 1638 – 51 . Google Scholar PubMed Khemici V , Prados J , Linder P et al. Decay-initiating endoribonucleolytic cleavage by RNase Y is kept under tight control via sequence preference and sub-cellular localisation . PLoS Genet 2015 ; 11 : e1005577 . Google Scholar CrossRef Search ADS PubMed Kikovska E , Svärd SG , Kirsebom LA . Eukaryotic RNase P RNA mediates cleavage in the absence of protein . Proc Natl Acad Sci USA 2007 ; 104 : 2062 – 7 . Google Scholar CrossRef Search ADS PubMed Kim YK , Mizutani K , Rhee K-H et al. Structural and mutational analysis of tRNA intron-splicing endonuclease from Thermoplasma acidophilum DSM 1728: catalytic mechanism of tRNA intron-splicing endonucleases . J Bacteriol 2007 ; 189 : 8339 – 46 . Google Scholar CrossRef Search ADS PubMed Kimura M . Structural basis for activation of an archaeal ribonuclease P RNA by protein cofactors . Biosci Biotechnol Biochem 2017 ; 81 : 1670 – 80 . Google Scholar CrossRef Search ADS PubMed Kimura M , Kakuta Y . Structural biology of the ribonuclease P in the hyperthermophilic archaeon Pyrococcus horikoshii OT3 . In Prakash , A , Satyanarayana , T , Johri , BN (eds) Microorganisms in Sustainable Agriculture and Biotechnology . Dordrecht, Netherlands : Springer , 2012 , 487 – 508 . Kirsebom LA . RNase P RNA mediated cleavage: substrate recognition and catalysis . Biochimie 2007 ; 89 : 1183 – 94 . Google Scholar CrossRef Search ADS PubMed Kitamura S , Fujishima K , Sato A et al. Characterization of RNase HII substrate recognition using RNase HII-argonaute chimaeric enzymes from Pyrococcus furiosus . Biochem J 2010 ; 426 : 337 – 44 . Google Scholar CrossRef Search ADS PubMed Kjems J , Garrett RA . Novel splicing mechanism for the ribosomal RNA intron in the archaebacterium Desulfurococcus mobilis . Cell 1988 ; 54 : 693 – 703 . Google Scholar CrossRef Search ADS PubMed Klein DJ , Schmeing TM , Moore PB et al. The kink-turn: a new RNA secondary structure motif . EMBO J 2001 ; 20 : 4214 – 21 . Google Scholar CrossRef Search ADS PubMed Klemm BP , Wu N , Chen Y et al. The diversity of ribonuclease P: protein and RNA catalysts with analogous biological functions . Biomolecules 2016 ; 6 : 27 , https://doi.org/10.3390/biom6020027. Google Scholar CrossRef Search ADS Klug G , Evguenieva-Hackenberg E , Omer A et al. RNA processing . In: Cavicchioli , R (ed.) Archaea: Molecular and Cellular Biology . Washington DC : ASM Press , 2007 , 158 – 74 . Kobayashi K , Ishitani R , Nureki O . Recent structural studies on Dom34/aPelota and Hbs1/aEF1α: important factors for solving general problems of ribosomal stall in translation . Biophysics 2013 ; 9 : 131 – 40 . Google Scholar CrossRef Search ADS PubMed Kobayashi K , Kikuno I , Kuroha K et al. Structural basis for mRNA surveillance by archaeal Pelota and GTP-bound EF1α complex . Proc Natl Acad Sci USA 2010 ; 107 : 17575 – 9 . Google Scholar CrossRef Search ADS PubMed Kochiwa H , Tomita M , Kanai A . Evolution of ribonuclease H genes in prokaryotes to avoid inheritance of redundant genes . BMC Evol Biol 2007 ; 7 : 128 . Google Scholar CrossRef Search ADS PubMed Koonin EV . Origin of eukaryotes from within archaea, archaeal eukaryome and bursts of gene gain: eukaryogenesis just made easier? Philos T Roy Soc B 2015 ; 370 : 1 – 15 . Google Scholar CrossRef Search ADS Koonin EV , Makarova KS . Mobile genetic elements and evolution of CRISPR-Cas Systems: all the way there and back . Genome Biol Evol 2017 ; 9 : 2812 – 25 . Google Scholar CrossRef Search ADS PubMed Koonin EV , Wolf YI , Aravind L . Prediction of the archaeal exosome and its connections with the proteasome and the translation and transcription machineries by a comparative-genomic approach . Genome Res 2001 ; 11 : 240 – 52 . Google Scholar CrossRef Search ADS PubMed Kouzuma Y , Mizoguchi M , Takagi H et al. Reconstitution of archaeal ribonuclease P from RNA and four protein components . Biochem Bioph Res Co 2003 ; 306 : 666 – 73 . Google Scholar CrossRef Search ADS Krupovic M , Gonnet M , Hania WB et al. Insights into dynamics of mobile genetic elements in hyperthermophilic environments from five new Thermococcus plasmids . PLoS One 2013 ; 8 : e49044 . Google Scholar CrossRef Search ADS PubMed Kuhn C-D , Wilusz JE , Zheng Y et al. On-enzyme refolding permits small RNA and tRNA surveillance by the CCA-adding enzyme . Cell 2015 ; 160 : 644 – 58 . Google Scholar CrossRef Search ADS PubMed Kuhn JF , Tran EJ , Maxwell ES Archaeal ribosomal protein L7 is a functional homolog of the eukaryotic 15.5kD/Snu13p snoRNP core protein . Nucleic Acids Res 2002 ; 30 : 931 – 41 . Google Scholar CrossRef Search ADS PubMed Kumar P , Anaya J , Mudunuri SB et al. Meta-analysis of tRNA derived RNA fragments reveals that they are evolutionarily conserved and associate with AGO proteins to recognize specific RNA targets . BMC Biol 2014 ; 12 : 78 . Google Scholar CrossRef Search ADS PubMed Kumar P , Kuscu C , Dutta A Biogenesis and function of transfer RNA-related fragments (tRFs) . Trends Biochem Sci 2016 ; 41 : 679 – 89 . Google Scholar CrossRef Search ADS PubMed Li de la Sierra-Gallay I , Mathy N , Pellegrini O et al. Structure of the ubiquitous 3΄ processing enzyme RNase Z bound to transfer RNA . Nat Struct Mol Biol 2006 ; 13 : 376 – 7 . Google Scholar CrossRef Search ADS PubMed Li de la Sierra-Gallay I , Pellegrini O , Condon C . Structural basis for substrate binding, cleavage and allostery in the tRNA maturase RNase Z . Nature 2005 ; 433 : 657 – 61 . Google Scholar CrossRef Search ADS PubMed Li de la Sierra-Gallay I , Zig L , Jamalli A et al. Structural insights into the dual activity of RNase J . Nat Struct Mol Biol 2008 ; 15 : 206 – 12 . Google Scholar CrossRef Search ADS PubMed Laalami S , Zig L , Putzer H . Initiation of mRNA decay in bacteria . Cell Mol Life Sci 2014 ; 71 : 1799 – 828 . Google Scholar CrossRef Search ADS PubMed Lafontaine DL , Tollervey D . The function and synthesis of ribosomes . Nat Rev Mol Cell Biol 2001 ; 2 : 514 – 20 . Google Scholar CrossRef Search ADS PubMed Lai L , Yokota H , Hung LW et al. Crystal structure of archaeal RNase HII: a homologue of human major RNase H . Structure 2000 ; 8 : 897 – 904 . Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Microbiology Reviews Oxford University Press

Insights into RNA-processing pathways and associated RNA-degrading enzymes in Archaea