• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Jun 21, 2005; 102(25): 8933–8938.
Published online Jun 3, 2005. doi:  10.1073/pnas.0502350102
PMCID: PMC1157037
From the Cover

Structure, function, and evolution of the tRNA endonucleases of Archaea: An example of subfunctionalization


We have detected two paralogs of the tRNA endonuclease gene of Methanocaldococcus jannaschii in the genome of the crenarchaeote Sulfolobus solfataricus. This finding has led to the discovery of a previously unrecognized oligomeric form of the enzyme. The two genes code for two different subunits, both of which are required for cleavage of the pre-tRNA substrate. Thus, there are now three forms of tRNA endonuclease in the Archaea: a homotetramer in some Euryarchaea, a homodimer in other Euryarchaea, and a heterotetramer in the Crenarchaea and the Nanoarchaea. The last-named enzyme, arising most likely by gene duplication and subsequent “subfunctionalization,” requires the products of both genes to be active.

Keywords: molecular evolution, RNA–protein interactions, tRNA splicing

Gene duplication is the primary source of new genes. The “subfunctionalization” hypothesis argues that duplicate genes experience degenerate mutations that divide the activity encoded by a single ancestral gene among its descendants (1). Here, we report a striking example of subfunctionalization.

In Archaea, the tRNA endonuclease plays a key role in assuring the correct removal of the intron from pre-tRNAs and pre-rRNA (26), which constitute the core of the translation machinery. Crystal structures of the tRNA endonucleases from Methanocaldococcus jannaschii (METJA) and Archaeoglobus fulgidus (ARCFU), both belonging to the phylum Euryarchaeota, are available (7, 8). These structures differ in a remarkable way. The structure of the homotetrameric endonuclease from METJA reveals two different functional roles for the monomeric units. The METJA endonuclease is organized as a dimer of dimers, with one subunit from each dimer participating in the catalytic cleavage reaction (the catalytic subunit) and the other (structural) subunit acting to place the two catalytic subunits correctly in space.

The crystal structure of the ARCFU endonuclease, by contrast, shows it to be a homodimer. Each subunit contains two similar repeating domains that are homologous to the subunit structure of the homotetrameric enzyme from METJA; the C-terminal repeat (CR) is the active domain, and the N-terminal repeat (NR) acts to stabilize the dimer.

The overall shape and size of the homodimeric ARCFU endonuclease resembles that of the homotetrameric METJA enzyme.

Both METJA and ARCFU belong to the Euryarchaeota. Nothing is known about the tRNA endonuclease of Crenarchaeota, the other main family of Archaea. To determine the properties of a crenarchaeal endonuclease, we searched the genome sequence of Sulfolobus solfataricus (SULSO) for homologs of the METJA endonuclease and found two candidate sequences.

Characterization of the two gene products reveals that both are required for tRNA endonuclease activity, each presumably functioning like one-half of the ARCFU enzyme. Detailed analysis of the amino acid sequences of the two proteins supports the idea that they evolved by the process called subfunctionalization (1, 9, 10).

Materials and Methods

Identification of the Homologs. The genes coding for the two SULSO proteins were PCR-amplified from genomic DNA by using two primers designed to obtain an amplified fragment presenting an NdeI site upstream of the gene and a BamHI site downstream. The digested PCR fragments were cloned in both pET28 (Novagen) and in pCYCA-11b (11). These two plasmids harbor two origins of replication that are compatible with one another and also two genes coding for different antibiotic resistances (kanamycin and chloramphenicol). The same procedure was used to clone the gene encoding the tRNA splicing endonuclease from ARCFU genomic DNA. The two truncated forms of endonuclease, NR_ARCFU and CR_ARCFU, were generated by PCR using the cloned full-length gene as template. NR_ARCFU comprises residues 1–138, and CR_ARCFU comprises residues 139–326. The PCR products were digested and cloned as previously described in pET28b and pCYAC-11b, respectively. All sequences of the clones obtained were verified.

Expression and Purification. Both proteins were overexpressed alone as hexahistidine-tagged protein (pET28) or coexpressed together with the untagged form (pCYCA-11b) in Escherichia coli BL21DE3 (Novagen). Cells were grown in a 1-liter culture of Luria–Bertani broth at 37°C in the presence of 30 μg/ml kanamycin (pET28) and adding, in the case of coexpression overproductions, 30 μg/ml chloramphenicol (pCYCA-11b). After cells reached an absorbance (at 600 nm) of 0.6, 1 mM IPTG was added to induce expression, and cells were grown for an additional 3 h. The bacterial pellets were resupended in 50 mM Tris·HCl, pH 8/500 mM NaCl/1 mM PMSF/0.1% Tween 20 and lysed by adding lysozyme to a final concentration of 100 μg/ml. The extract was then heated at 65°C for 20 min and centrifuged for 1 h at 4°C. The supernatant was loaded on a 2-ml TALON cobalt affinity column (BD Biosciences) and washed with lysis buffer, and retained protein was eluted by adding 100 mM imidazole to the lysis buffer. We recovered large quantities of homogeneous protein “suitable for crystallization trials.”

In Vitro Transcription. The DNA template was prepared and transcribed by T7 RNA polymerase (1214) using the T7-Megashort-script kit (Ambion, Austin, TX). Transcripts were internally labeled with [32P]UTP [800 Ci/mmol (1 Ci = 37 GBq), Amersham Pharmacia]. After phenol extraction and ethanol precipitation, transcripts were purified on a 10% denaturing polyacrylamide gel, eluted, and ethanol-precipitated.

In Vitro Splicing Assay. Labeled tRNA precursors (20 fmol) were incubated with purified splicing endonucleases in reaction mixtures containing 25 mM Tris·HCl, pH7.5/5 mM MgCl2/100 mM NaCl/10% glycerol at 65°C for 1 h. Cleavage products were analyzed after phenol extraction and ethanol precipitation by gel electrophoresis on 10% denaturing polyacrylamide gel. Image analysis was done by using a Molecular Dynamics Model Storm 860 PhosphorImager with imagequant 4 software.

Sequence Analysis. The gene coding for tRNA endonuclease subunits in METJA was used as the query sequence in blastp searches of the National Center for Biotechnology Information genomic database targeting the completed archaeal genomes including the species SULSO, Sulfolobus tokodaii (SULTO), Aeropyrum pernix (AERPE), Pyrobaculum aerophilum (PYRAE), ARCFU, Haloarcula marismortui (HALVO), Halobacterium sp. (HALN1), Methanothermobacter thermautotrophicus (METTH), METJA, Methanopyrus kandleri (METKA), Methanosarcina acetivorans (METAC), Methanosarcina barkeri (METBA), Methanosarcina mazei (METMA), Pyrococcus abyssi (PYRAB), Pyrococcus furiosus (PYRFU), Pyrococcus horikoshii (PYRHO), Ferroplasma acidarmanus (FERAC), Thermoplasma acidophilum (THEAC), Thermoplasma volcanium (THEVO), and Nanonarcheum equitans (NANEQ). The proteins detected were included in a multiple alignment built by using the program mafft (15). A refinement of the alignment was carried out by using the program rascal (16), and the quality was verified by normd (17).


Genes encoding the tRNA endonucleases were sought in 20 completed archaeal genomes by using the METJA subunit sequence as the probe in blastp searches. We found a single gene that codes for a polypeptide of ≈180 residues in the genomes of PYRAB, PYRFU, PYRHO, METJA, and METTH. Presumably, the enzymes from these organisms assemble as homotetramers (α4), like the METJA enzyme.

The single identified genes coding for an endonuclease in the genomes of METAC, METBA, METMA, ARCFU, HALVO, HALN1, FERAC, THEAC, and THEVO code for a subunit of ≈350 residues that presumably assembles into a homodimer, as does the ARCFU enzyme.

The search of the genomes of the Crenarchaeota SULSO, SULTO, AERPE, NANEQ, PYRAE, and METKA revealed two homologs. In the genome of SULSO, one gene codes for a protein, Q97ZY3, with an expectation value (E value) of 1e – 22, clearly indicating an ortholog. The size of this protein, 182 residues, suggests a possible assembly into a homotetrameric (α4) enzyme, as in the case of METJA. We named this protein α_SULSO. The second homolog, Q980L2, has an E value slightly above the twilight zone of 0.001 (0.004). The protein encoded by this gene contains 177 residues. We reasoned that the second gene could code for a second subunit of the endonuclease. The METJA enzyme is a homotetramer, whereas the ARCFU enzyme is assembled as a homodimer with the monomeric unit resulting from duplication and fusion of a METJA-like monomer (18). We named the second SULSO protein β_SULSO. The SULSO enzyme might thus assemble as a heterotetramer: α2β2.

Before describing biochemical experiments that explore the activities of the tRNA endonuclease subunits, we present a bioinformatic analysis of the amino acid sequences found and the structures predicted for these sequences.

We first characterized all of the sequences at the level of their primary structure (Fig. 1). The homodimer subunits were divided into two monomers of roughly the same size as the tetrameric monomer on the basis of the information obtained from the crystallographic structure of the ARCFU endonuclease. The NR (NR_ARCFU) is composed of residues 1–138, and CR (CR_ARCFU) is composed of residues 139–326. All of the protein sequences were then aligned as described in Materials and Methods.

Fig. 1.
Primary sequences and secondary structures of tRNA endonuclease subunits. A sequence alignment of tRNA endonuclease subunits is presented as two separate groups as described in the text. Sequences belonging to the structural subunit class are colored ...

The sequences were divided into two groups in relation to the role played in the enzymatic complex according to the ARCFU structure. One set included the structural subunits: the putative β-subunits and the NRs. The second set included the catalytic subunits: the α-subunit and the CRs. The homotetrameric subunits were included in both groups, because they can act as catalytic or structural subunits, depending on how they assemble. For clarity, the numbering of the residues used throughout refers to the METJA sequence. The multiple alignments of the structural and catalytic subunits are shown in Fig. 1. The similarities and identities (red and blue letters in Fig. 1) were separately determined for each group. Subsequently, homology regions were highlighted by taking into consideration the similarities and identities of the two distinct groups at the same time.

When the two sets are compared, three groups of residues can be distinguished: residues conserved between the two classes (yellow boxes), residues conserved only in the catalytic class (green boxes), and residues conserved in the structural class (cyan boxes). The majority of residues conserved in common between the two classes of subunits (yellow boxes) are clustered in two distinct segments of 15 aa: one segment includes strand β4-helix α2-strand β5 and the second includes helix α4-loop-strand β6. The first pattern presents a motif [L-x-L-x-x-(L,I,F,V)-x-Y-L-x-x-(K,R)-G-x-(L,I)] that is localized in the N-terminal domain, and the other presents a motif [(K,R)-(Y,F)-xn-V-Y-x-D-L-(K,R)-x-(K,R)-G-(Y,F)-x-V] that is localized in the C-terminal domain. The first motif is not conserved in all of the structural subunits of crenoarcheote enzymes. In particular, in β-SULSO, this motif is highly divergent, and it has been deleted, along with the entire N-terminal domain, in β-SULTO and β-PYRAE. The second motif is universally conserved.

In the catalytic class (green boxes), the specific signature residues are clustered in loop L6, loop L7, and loop L9. The pattern of loop L6 is of six residues with a motif [(K,R)-(T,S)-(F,L,V,I)-K-(Y,F)-G]. Two of the three catalytic residues Y115 and H125 (residues in italics) are present in the conserved pattern of loop L7 [Y-xn-H-(A,S)-(D,E)-(Y,F,W)-(I,L,V)], although the size of this loop is variable. L9 presents an eight-residue motif [R-(V,L)-(A,S)-H-(G,S)-(V,T)-R-K-(K,R)-(L,M)], which includes the catalytic residue K156 (in italics).

The structural subunit class (cyan boxes) is characterized by a [V-D-(E,D)-(E,D)-x-(D,E)-x-T] conserved pattern in loop L10, and a conserved E at position 135.

The mutually exclusive sequence motifs of the two classes can be represented on a surface calculated from the structures of the monomer (Fig. 2A) and of the homotetramer of the METJA endonuclease (Fig. 2B). Extension of the information available from the crystallographic studies to the endonuclease alignments strongly supports the existence of a canonical structure. This structure is formed by two different domains (Fig. 1). The N-terminal domain is characterized by a five-strand antiparallel β-sheet (strands β1, β2, β3, β4, and β5) packed against helix α2 and α3, both of which lie perpendicular to α1. The C-terminal domain is characterized by a five-strand β-sheet (strands β6, β7, β8, β9, and β10) sandwiched between helix α4 and α5.

Fig. 2.
Structural analysis. (A) Results of the multiple sequence alignment mapped onto the three-dimensional structure of a METJA monomer (1A79). Residues conserved in the structural subunits are colored cyan; residues conserved in the catalytic subunit are ...

The group of catalytic subunits shares a similar secondary structure (Fig. 1). The crenarchaeote proteins, in particular, contain an insertion in the loop between β3 and β4 rich in hydrophilic amino acids. CR_THEVO and CR_THEAC have lost helices α3 and α4. The size of loop L7 varies among all of the members of this group, but the two amino acids of the catalytic triads are always conserved. Variation in size and divergence in sequence are also present in loop L10.

The group of structural subunits shows, in the N-terminal domain, strong sequence divergence. In the case of β_SULTO and β_PYRAE, the entire N-terminal domain has been deleted. NR_HALN1 and NR_HALVO show a larger loop between β6 and β7. The sequence and size of loop L7 is extremely varied in all of the class members. NR_ARCFU, in particular, shows a major deletion at this location. NR_THEAC, NR_THEVO, NR_FERAC, NR_HALVO, and NR_HALN1 are deleted in the loop between α5 and β9. It is evident from the alignment that loop L10 and the residues in the β10 strand, known to play a structural role in the dimerization of the ARCFU enzyme and the tetramerization of the METJA enzyme (7, 8), are conserved.

It is evident from this analysis and the published work of others that the known tRNA endonucleases of Archaea can take one of two forms: (i) a homotetramer in which two of the subunits play a structural role and two contain the catalytic sites or (ii) a homodimer, each of whose subunits, twice the length of the METJA class, contains a structural part and a catalytic part (7, 8, 18, 19). The bioinformatic analysis suggests that the latter enzyme arose from a gene duplication/fusion event, followed by a few mutations that allow for subfunctionalization of the two regions of the protein. To illustrate this possibility, we reversed the putative evolutionary path by cutting the gene encoding the ARCFU tRNA endonuclease into two segments, expressed each polypeptide independently, and assayed their ability to function in vitro.

For these experiments, we used the substrate tRNA archaeuka (Fig. 3A), a synthetic molecule containing an intron recognized by either archaeal or eukaryotic endonucleases (12, 20). Two truncated forms of the ARCFU enzyme, NR_ARCFU and CR_ARCFU, were generated. NR_ARCFU was cloned in pET28 to express a protein with an N-terminal His tag (His-6 NR_ARCFU). CR_ARCFU was cloned in pCYAC-11b. His-6 NR_ARCFU was overexpressed alone or in the presence of CR_ARCFU. After lysis, the crude extracts were pulled down by using a cobalt TALON affinity resin and extensively washed to remove nonspecifically bound proteins. The bound proteins were eluted from the resin with imidazole and assayed. Fig. 3B (lanes 4 and 5) shows that the excision of the intron from pre-tRNA archaeuka occurs only when the two truncated proteins are coexpressed. The purified proteins were visualized by Coomassie blue staining on SDS polyacrylamide gels (data not shown). These results are consistent with the conclusion that NR_ARCFU is the structural subunit, acting to place CR_ARCFU, the catalytic subunit, in space correctly.

Fig. 3.
tRNA endonuclease activity assays. (A) Pre-tRNAarchaeuka consists of two regions derived from yeast pre-tRNAPhe (nucleotides 1–31 and 38–76) joined by a 25-nt insert that corresponds to the bulge–helix–bulge motif of archaeal ...

Having shown that the two halves of a homodimer endonuclease could be assembled into an active enzyme, we investigated the endonuclease predicted from the genome sequence of the crenarchaeote SULSO. Recall that the genome analysis reveals one ORF (α) with very high similarity to the homotetramer of METJA and another with significantly lower similarity (β).

The gene for α_SULSO and the gene coding for the METJA monomer were cloned in pET28. The constructs code for proteins with a modified N terminus presenting a His-6 tag. The tagged enzymes were purified by affinity chromatography as described in Materials and Methods. The formation of aggregates was prevented by using high salt. Purification resulted, in both cases, in homogeneous proteins, as judged from Coomassie blue staining on SDS/PAGE. To determine whether the α_SULSO enzyme has a dual functional role like the METJA monomer, we tested the purified α_SULSO protein for cleavage of pre-tRNA archaeuka. We already knew that the intron of this substrate could be precisely excised by using a partially purified extract from SULSO (12). Fig. 3C (lane 6) shows that the substrate is correctly cleaved by the purified endonuclease from METJA but not by the protein α_SULSO (lane1). Leaving other possibilities aside for the moment, we conclude that α_SULSO cannot tetramerize like METJA. This observation led us to consider the possibility that β_SULSO allows the α-subunit to assemble properly.

The pull-down approach was then used to investigate the oligomerization of the two SULSO gene products. Two different constructs were produced for each gene to obtain an N-terminal-tagged version and an untagged version of the products. The tagged versions were overexpressed alone or in the presence of the untagged version of the other gene. After purification, the proteins were assayed by using pre-tRNA archaeuka as substrate. Fig. 3C (lanes 1 and 3) shows that neither β_SULSO nor α_SULSO alone cleaves the substrate; the activity is detectable only when the two proteins are coexpressed (Fig. 3C, lanes 2 and 4).

Thus, SULSO contains two genes, each encoding one subunit of a (presumably) heterotetrameric endonuclease, mimicking the specialization revealed in the assembly of the artificially created subunits of the ARCFU enzyme.


A Previously Unrecognized Form of Endonuclease in SULSO. We present here the finding of a third archaeal tRNA endonuclease complex in the crenoarchaeote SULSO that is the result of assembly of two different subunits homologous to METJA, each one presumably playing one of the two different roles in a heterotetramer (Fig. 4). To identify the SULSO enzyme, we searched the sequenced genome by using METJA as a probe and found two homologs of the gene. The proteins encoded by these genes were expressed alone or together and, after purification to homogeneity, were tested for specific cleavage activity.

Fig. 4.
Schematic representation of the phylogenetic tree of Archaea (adapted from ref. 21). Each rectangle surrounds species sharing the same endonuclease architecture. The subunits are named according to Fig. 5. The duplication events are indicated by an arrow. ...

The enzyme is functionally active only if both subunits are present. Analysis of all of the completed genomes of 20 Archaea helped to extend the existence of a second homolog of the METJA gene to all Creanoarcheota and the Nanoarcheota and to determine the phylogenetic distribution of the different endonuclease architectures.

Evolution of tRNA Endonucleases in Archaea. The distribution of subunits in the different archaeal phyla and comparison of the aligned proteins provide a unique picture of how these enzymes evolved (Fig. 4). The ancestor of the archaeal enzyme was probably a homotetramer that, after two gene duplication events (or horizontal gene transfer), gave rise to a homodimeric form and a heterotetrameric form. The duplication events were independent from one another. One event took place in the common ancestor of Crenoarcheota and Nanoarcheota and resulted in two genes coding for two subunits, whereas the other occurred in the ancestor of Archaeoglobales, Halobacteriales, and Methanosarcinales, resulting in an in-frame duplication and giving rise to a single gene coding for two fused subunits.

Gene Duplication in Archaeal tRNA Endonucleases: An Example of Subfunctionalization. Two pairs of subunits (A1 and A2, and B1 and B2) associate to form isologous dimers by means of extensive interactions between their β10 strands (Fig. 5A, arrow). The carboxyl half of β10 from one subunit forms main-chain hydrogen bonds with the symmetry-related residues of β10 from the other subunit (β10′), leading to a two-stranded β-sheet spanning the subunit boundary. The tetramer is formed by means of heterologous interactions between the two dimers. The main interaction between the two dimers is by means of the insertion of loop L10 (triangle in Fig. 5) from subunits A2 and B2 into a cleft in subunits B1 and A1 between the N- and C-terminal domains of each monomer. The interaction is primarily polar between acidic residues in loop L10 and basic residues in the cleft. This arrangement causes the two isologous dimers to be translated relative to each other by ≈20 Å, bringing subunits A1 and B1 much closer together than A2 and B2, which do not interact at all, and resulting in an arrangement of subunits in which only one symmetrically disposed pair of active sites can recognize the substrate. These weak interactions are likely to have been conserved in evolution because they lead to the required positioning of the two active sites.

Fig. 5.
Model of the tRNA splicing endonucleases of METJA, ARCFU, and SULSO. (A) model of the METJA homotetramer. Several important structural features discussed in the text are indicated: loop L10 (cyan triangle), the C-terminal β10 strand (arrow), and ...

The intersubunit β10–β10′ interaction observed in the METJA endonuclease is now an intrasubunit interaction in the ARCFU endonuclease dimer. ARCFU endonuclease also uses a similar L10 interaction, but to form a dimer (Fig. 5B). The three residues at the previously identified active site in the METJA endonuclease (His-125, Lys-156, and Tyr-115; pentagon in Fig. 5) are superimposable with those in the ARCFU endonuclease (His-257, Lys-287, and Tyr-246). This result is consistent with the fact that both enzymes act on the same bulge–helix–bulge motif in their pre-tRNA substrates. The connection between the NRs and the CRs is a fully extended 10-aa residue segment that wraps around the C-terminal domain of the CR before joining the N-terminal domain (Fig. 5B). The two CRs, C1 and C2, occupy the same spatial location as subunits A1 and B1 in the METJA endonuclease (in Fig. 5B). The two NRs, N1 and N2, occupy the same spatial location as the A2 and B2 subunits in the METJA endonuclease and thus play the same structural roles as the A2 and B2 subunits. Significantly, the NR has retained the L10 structure-mediating dimerization of the ARCFU endonuclease, equivalent to its role in tetramerization of the METJA endonuclease (Fig. 5B). The L10 structure is absent from the CR, suggesting that L10 has a role only in dimerization.

We generated two truncated forms of the ARCFU enzyme, NR_ARCFU and CR_ARCFU. The two halves of the homodimer endonucleases can be assembled into an active heterotetramer; NR_ARCFU is the catalytic subunit and contains the conserved catalytic triad (Fig. 5B). SULSO contains two genes, each coding for one of the subunits of a heterotetrameric endonuclease, mimicking the specialization revealed in the assembly of the artificially created subunits of the ARCFU enzyme. The α-subunit of the of the SULSO enzyme contains the catalytic triad, and the β-sunbunit contains the loop L10.

The general model of the process of duplication–degeneration–complementation proposes that degenerative mutations facilitate the preservation of the duplicated functional gene (9, 10). The gene, once duplicated, can follow three possible alternative evolutionary fates. After mutations in the regulatory regions or a null mutation in the coding region, the activity of one copy can be lost (nonfunctionalization). One copy can acquire mutations that result in the gain of a new function (neofunctionalization). Finally, both copies can accumulate mutations that lead to specialization (e.g., catalytic or structural), such that each copy is necessary to function as the single-copy ancestor gene (subfunctionalization).

Sequence analysis based on the available structural data has revealed two sets of mutually exclusive and substantially conserved motifs, allowing assignment of the endonuclease subunits to two classes according to their functional role. The homotetramer presents both sets of motifs, which correspond to a subunit capable of performing either function, according to how it is assembled in the enzyme. The homodimer and heterotetramer present domains or subunits subspecialized in one function or the other, presenting, therefore, one set of motifs or the other. The biochemical data we have presented here clearly support qualitative subfunctionalization in the case of the homodimers and heterotetramers. Both domains or subunits are required for the correct assembly of the enzyme to process the pre-tRNA precursor. This subfunctionalization is even more evident in the case of an ARCFU heterotetramer reconstituted by expressing the homodimer gene as two truncated forms. The resulting heterotetramer is active.


Our results suggest that on an evolutionary scale, the common ancestor of all of the archaeal enzymes was a homotetramer with four identical subunits that could play both the catalytic role and the structural role interchangeably, as in the contemporary methanogens. The appearance of a second copy of the gene resulted in subfunctionalization of the two copies, resulting in the splitting of the two roles, as in the case of homodimers and heterotetramers. The accumulation of mutations compromised one function or the other, as is evident from the conservation of signature motifs.


We thank Mose Rossi [Istituto di Biochimica delle Proteine-Consiglio Nazionale delle Ricerche (CNR), Naples] for providing the genomic DNA of SULSO, METJA, and ARCFU; A. Ferrara for secretarial assistance; and G. Di Franco for technical assistance. This work was supported by the Programma Biomolecole per la Salute Umana Ministero dell'Universitá e della Ricerca Scientifica e Tecnologica (MURST) CNR L. 95/95 5%; Fondo Investimenti Ricerca di Base Ministero Instruzione Universitá Ricerca (MIUR) of Italy; Progetto Genomica Funzionale L. 449/97 MIUR-CNR; Progetto Strategico Tecnologie di Base della Postgenomica CNR; Progetto Strategico Biotecnologie MURST-CNR; Progetto Strategico Genetica Molecolare L. 449/97 MURST; and European Networks of Excellence (EUMORPHIA and MUGEN).


This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: METJA, Methanocaldococcus jannaschii; SULSO, Sulfolobus solfataricus; ARCFU, Archaeoglobus fulgidus; SULTO, Sulfolobus tokodaii; AERPE, Aeropyrum pernix; PYRAE, Pyrobaculum aerophilum; HALVO, Haloarcula marismortui; HALN1, Halobacterium sp.; METTH, Methanothermobacter thermautotrophicus; METKA, Methanopyrus kandleri; METAC, Methanosarcina acetivorans; METBA, Methanosarcina barkeri; METMA, Methanosarcina mazei; PYRAB, Pyrococcus abyssi; PYRFU, Pyrococcus furiosus; PYRHO, Pyrococcus horikoshii; FERAC, Ferroplasma acidarmanus; THEAC, Thermoplasma acidophilum; THEVO, Thermoplasma volcanium; NANEQ, Nanonarcheum equitans; CR, C-terminal repeat; NR, N-terminal repeat.

See Commentary on page 8791.


1. Kondrashov, F. A., Rogozin, I. B., Wolf, Y. I. & Koonin, E. V. (2002) Genome. Biol. 3, RESEARCH0008. [PMC free article] [PubMed]
2. Kjems, J. & Garrett, R. A. (1991) Proc. Natl. Acad. Sci. USA 88, 439–443. [PMC free article] [PubMed]
3. Thompson, L. D. & Daniels, C. J. (1988) J. Biol. Chem. 263, 17951–17959. [PubMed]
4. Abelson, J., Trotta, C. R. & Li, H. (1998) J. Biol. Chem. 273, 12685–12688. [PubMed]
5. Kjems, J. & Garrett, R. A. (1988) Cell 54, 693–703. [PubMed]
6. Kjems, J., Jensen, J., Olesen, T. & Garrett, R. A. (1989) Can. J. Microbiol. 35, 210–214. [PubMed]
7. Li, H., Trotta, C. R. & Abelson, J. (1998) Science 280, 279–284. [PubMed]
8. Li, H. & Abelson, J. (2000) J. Mol. Biol. 302, 639–648. [PubMed]
9. Force, A., Lynch, M., Pickett, F. B., Amores, A., Yan, Y. L. & Postlethwait, J. (1999) Genetics 151, 1531–1545. [PMC free article] [PubMed]
10. Lynch, M. & Conery, J. S. (2003) J. Struct. Funct. Genomics 3, 35–44. [PubMed]
11. Fribourg, S., Romier, C., Werten, S., Gangloff, Y. G., Poterszman, A. & Moras, D. (2001) J. Mol. Biol. 306, 363–373. [PubMed]
12. Fabbri, S., Fruscoloni, P., Bufardeci, E., Di Nicola Negri, E., Baldi, M. I., Attardi, D. G., Mattoccia, E. & Tocchini-Valentini, G. P. (1998) Science 280, 284–286. [PubMed]
13. Mattoccia, E., Baldi, I. M., Gandini-Attardi, D., Ciafre, S. & Tocchini-Valentini, G. P. (1988) Cell 55, 731–738. [PubMed]
14. Di Nicola Negri, E., Fabbri, S., Bufardeci, E., Baldi, M. I., Gandini Attardi, D., Mattoccia, E. & Tocchini-Valentini, G. P. (1997) Cell 89, 859–866. [PubMed]
15. Katoh, K., Kuma, K., Toh, H. & Miyata, T. (2005) Nucleic Acids Res. 33, 511–518. [PMC free article] [PubMed]
16. Thompson, J. D., Thierry, J. C. & Poch, O. (2003) Bioinformatics 19, 1155–1161. [PubMed]
17. Thompson, J. D., Plewniak, F., Ripp, R., Thierry, J. C. & Poch, O. (2001) J. Mol. Biol. 314, 937–951. [PubMed]
18. Kleman-Leyer, K., Armbruster, D. W. & Daniels, C. J. (1997) Cell 89, 839–847. [PubMed]
19. Lykke-Andersen, J. & Garrett, R. A. (1997) EMBO J. 16, 6290–6300. [PMC free article] [PubMed]
20. Fruscoloni, P., Baldi, M. I. & Tocchini-Valentini, G. P. (2001) EMBO Rep. 2, 217–221. [PMC free article] [PubMed]
21. Woese, C. R. (1996) Curr. Biol. 6, 1060–1063. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...