Specific and non-specific purine trap in the T-loop of normal and suppressor tRNAs

To elucidate the general constraints imposed on the structure of the D and T-loops in functional tRNAs, active suppressor tRNAs were selected in vivo from a combinatorial tRNA gene library in which several nucleotide positions in these loops were randomized. Analysis of the nucleotide sequences of the selected clones demonstrates that most of them contain combination U54-A58 allowing the formation of the standard reverse-Hoogsteen base-pair 54-58 in the T-loop. With only one exception, all these clones fall into two groups, each characterized by a distinct sequence pattern. Analysis of these two groups has allowed us to suggest two different types of nucleotide arrangement in the DT region. The first type, the so-called specific purine trap, is found in 12 individual tRNA clones and represents a generalized version of the standard D-T loop interaction. It consists of purine 18 sandwiched between the reverse-Hoogsteen base-pair U54-A58 and purine 57. The identity of purine 18 is restricted by the specific base-pairing with nucleotide 55. Depending on whether nucleotide 55 is U or G, purine 18 should be, respectively, G or A. The second structural type, the so-called non-specific purine trap, corresponds to the nucleotide sequence pattern found in 16 individual tRNA clones and is described here for the first time. It consists of purine 18 sandwiched between two reverse-Hoogsteen base-pairs U54-A58 and A55-C57 and, unlike the specific purine trap, requires the T-loop to contain an extra eighth nucleotide. Since purine 18 does not form a base-pair in the non-specific purine trap, both purines, G18 and A18, fit to the structure equally well. The important role of both the specific and non-specific purine traps in the formation of the tRNA L-shape is discussed.     

Active suppressor tRNAs with a double helix between the D- and T-loops

One of the most conserved elements of the tRNA structure is the reverse-Hoogsteen base-pair T54--A58 in the T-loop, which plays a major role in the maintenance of the standard L-shape conformation. Here, we present the results of in vivo selection of 51 active suppressor tRNA clones, none of which contains base-pair T54--A58. In 49 clones, we found two regions in the D and T-loops that are complementary to each other. This finding suggests the existence of an inter-loop double helix consisting of three base-pairs, which could have the same role as base-pair T54--A58 in the fixation of the juxtaposition of the two helical domains within the L-shape. From this point of view, the appearance of the inter-loop double helix represents a compensatory effect for the absence of base-pair T54--A58. The results shed new light on the role of different elements of the tRNA structure in the formation of the standard L-shape conformation and on the possibility of synonymous replacements of one arrangement by another in functional RNA molecules.     

Structural compensation in an archaeal selenocysteine transfer RNA.

A new type of structural compensation between the lengths of two perpendicularly oriented RNA double helices was found in the archaeal selenocysteine tRNA from Methanococcus jannascii. This tRNA contains only four base-pairs in the T-stem, one base-pair less than in all other cytosolic tRNAs. Our analysis shows that such a T-stem in an otherwise normal tRNA cannot guarantee the formation of the normal interactions between the D and T-loops. The absence of these interactions would affect the juxtaposition of the two tRNA helical domains, potentially damaging the tRNA function. In addition to the short T-stem, this tRNA possesses another unprecedented feature, a very long D-stem consisting of seven base-pairs. Taken as such, a seven base-pair D-stem will also disrupt the normal interaction between the D and T-loops. On the other hand, the presence of the universal nucleotides in both the D and T-loops suggests that these loops probably interact with each other in the same way as in other tRNAs. Here, we demonstrate that the short T-stem and the long D-stem can naturally compensate each other, thus providing the normal D/T interactions. Molecular modeling has helped suggest a detailed scheme of mutual compensation between these two unique structural aspects of the archaeal selenocysteine tRNA. In the light of this analysis, other structural and functional characteristics of the selenocysteine tRNAs are discussed.     

Selenocysteine tRNA identification in the model organisms Dictyostelium discoideum and Tetrahymena thermophila

Characterizing Sec tRNAs that decode UGA provides one of the most direct and easiest means of determining whether an organism possesses the ability to insert selenocysteine (Sec) into protein. Herein, we used a combination of two techniques, computational to identify Sec tRNA genes and RT-PCR to sequence the gene products, to unequivocally demonstrate that two widely studied, model protozoans, Dictyostelium discoideum and Tetrahymena thermophila, encode Sec tRNA in their genomes. The advantage of using both procedures is that computationally we could easily detect potential Sec tRNA genes and then confirm by sequencing that the Sec tRNA was present in the tRNA population, and thus the identified gene was not a pseudogene. Sec tRNAs from both organisms decode UGA. T. thermophila Sec tRNA, like all other sequenced Sec tRNAs, is 90 nucleotides in length, while that from D. discoideum is 91 nucleotides long making it the longest eukaryotic sequenced to date. Evolutionary analyses of known Sec tRNAs reveal the two forms identified herein are the most divergent eukaryotic Sec tRNAs thus far sequenced.     

A role for the bulged nucleotide 47 in the facilitation of tertiary interactions in the tRNA structure.

Based on computer modeling and with the use of energy minimisation procedure, we show that the bulged nucleotide 47 in the yeast tRNA(Phe) structure plays an important steric role, allowing the formation of canonical tertiary interactions 15-48 and 22-46 within the D-domain. The absence of nucleotide 47 can be compensated by the presence of a wobble pair U13-G22, whose unusual stereochemistry permits as well the formation of the canonical tertiary interactions. The tRNA database show that the vast majority of the cytosolic tRNAs have either a nucleotide at position 47 or a wobble pair U13-G22. On the contrary, many mitochondrial tRNAs, having a Watson-Crick pair 13-22, do not have nucleotide in position 47, which suggests that their tertiary interactions within the D-domain must differ from those in cytosolic tRNAs.     

Importance of conserved residues for the conformation of the T-loop in tRNAs

The conformation of the T-loop of yeast tRNA(Asp) was studied by structural mapping techniques using chemical and enzymatic probes and by three-dimensional graphics modeling with the known crystallographic structures of tRNAs as references. The structural importance of C61 (conserved in the T-stem of all tRNAs) for the loop conformation was directly checked by ethylnitrosourea phosphate alkylation, either on the 3'-half tRNAAsp molecule or on a variant in which C61 was replaced by U61. The reactivity of P60 against ethylnitrosourea alkylation in the variant emphasizes the role of the hydrogen bond between this phosphate and position N4 of C61 for stabilizing the conformation of the T-loop. Experiments on several tRNA variants, containing C61 but altered in the sequence or in the length of the T-loop, indicate that other structural features help to stabilize the hydrogen bond network around P60. Evidence is presented that the reverse Hoogsteen base pair T54-A58 contributes to this stabilization by maintaining the hydrogen bonding between the 2'OH of ribose 58 and P60. Using graphics modeling and based on the chemical data. T-loops of several variants were constructed. It appears that both the constant length of the T-loop and the presence of psi 55 are crucial for the correct interaction between the T- and D-loops. The conclusion of this study is that the T-loop in tRNA possesses an intrinsic conformation (mainly governed by the constant residues) existing primarily without the structural context of the entire tRNA molecule.     

Nucleotide sequences of animal mitochondrial tRNAs(Met) possibly recognizing both AUG and AUA codons.

To elucidate the role of modified nucleosides of tRNA in mitochondrial translation systems, especially with regard to their codon recognition, we purified mitochondrial tRNAs(Met) isolated from liver of frog, chicken and rat, and determined their nucleotide sequences. All of these tRNAs(Met) were found to possess 5-formylcytidine in the first letter of the anticodon, which is known to be prerequisite for bovine mt tRNA(Met) to decode AUA codon as well as AUG codon. These tRNA possesses two pseudeuridines in similar positions, and only chicken tRNA(Met) had ribothymidine at the first position of the T-loop, which is always found in the usual tRNAs. Considering that AUA codon is used as five times frequently as AUG codon in these animal mitochondrial genomes, it is deduced that 5-formylcytidine at the wobble position is essential for the recognition of both AUA and AUG codons.     

Importance of the reverse Hoogsteen base pair 54-58 for tRNA function

To elucidate the general constraints imposed on the structure of the D- and T-loops in functional tRNAs, active suppressor tRNAs were selected in vivo from a combinatorial tRNA gene library in which several nucleotide positions of these loops were randomized. Analysis of the nucleotide sequences of the selected clones demonstrates that among the randomized nucleotides, the most conservative are nucleotides 54 and 58 in the T-loop. In most cases, they make the combination U54-A58, which allows the formation of the normal reverse Hoogsteen base pair. Surprisingly, other clones have either the combination G54-A58 or G54-G58. However, molecular modeling shows that these purine–purine base pairs can very closely mimic the reverse Hoogsteen base pair U-A and thus can replace it in the T-loop of a functional tRNA. This places the reverse Hoogsteen base pair 54-58 as one of the most important structural aspects of tRNA functionality. We suggest that the major role of this base pair is to preserve the conformation of dinucleotide 59–60 and, through this, to maintain the general architecture of the tRNA L-form.     

Crystal structure of pseudouridine synthase RluA: indirect sequence readout through protein-induced RNA structure

RluA is a dual-specificity enzyme responsible for pseudouridylating 23S rRNA and several tRNAs. The 2.05 A resolution structure of RluA bound to a substrate RNA comprising the anticodon stem loop of tRNA(Phe) reveals that enzyme binding induces a dramatic reorganization of the RNA. Instead of adopting its canonical U turn conformation, the anticodon loop folds into a new structure with a reverse-Hoogsteen base pair and three flipped-out nucleotides. Sequence conservation, the cocrystal structure, and the results of structure-guided mutagenesis suggest that RluA recognizes its substrates indirectly by probing RNA loops for their ability to adopt the reorganized fold. The planar, cationic side chain of an arginine intercalates between the reverse-Hoogsteen base pair and the bottom pair of the anticodon stem, flipping the nucleotide to be modified into the active site of RluA. Sequence and structural comparisons suggest that pseudouridine synthases of the RluA, RsuA, and TruA families employ an equivalent arginine for base flipping.     

Comparison of the tertiary structure of yeast tRNA(Asp) and tRNA(Phe) in solution. Chemical modification study of the bases

A comparative study of the solution structures of yeast tRNA(Asp) and tRNA(Phe) was undertaken with chemical reagents as structural probes. The reactivity of N-7 positions in guanine and adenine residues was assayed with dimethylsulphate and diethyl-pyrocarbonate, respectively, and that of the N-3 position in cytosine residues with dimethylsulphate. Experiments involved statistical modifications of end-labelled tRNAs, followed by splitting at modified positions. The resulting end-labelled oligonucleotides were resolved on polyacrylamide sequencing gels and analysed by autoradiography. Three different experimental conditions were used to follow the progressive denaturation of the two tRNAs. Experiments were done in parallel on tRNA(Asp) and tRNA(Phe) to enable comparison between the two solution structures and to correlate the results with the crystalline conformations of both molecules. Structural differences were detected for G4, G45, G71 and A21: G4 and A21 are reactive in tRNA(Asp) and protected in tRNA(Phe), while G45 and G71 are protected in tRNA(Asp) and reactive in tRNA(Phe). For the N-7 atom of A21, the different reactivity is correlated with the variable variable loop structures in the two tRNAs; in the case of G45 the results are explained by a different stacking of A9 between G45 and residue 46. For G4 and G71, the differential reactivities are linked to a different stacking in both tRNAs. This observation is of general significance for helical stems. If the previous results could be fully explained by the crystal structures, unexpected similarities in solution were found for N-3 alkylation of C56 in the T-loop, which according to crystallography should be reactive in tRNA(Asp). The apparent discrepancy is due to conformational differences between crystalline and solution tRNA(Asp) at the level of the D and T-loop contacts, linked to long-distance effects induced by the quasi-self-complementary anticodon GUC, which favour duplex formation within the crystal, contrarily to solution conditions where the tRNA is essentially in its free state.     

Analysis of the stability and flexibility of RNA complexes containing bulge loops of different sizes

Molecular dynamics simulations of RNA molecules consisting of an antisense oligonucleotide forming a complex with a target strand thereby creating an internal bulge-loop with 3, 4, or 5 nucleotides have been performed with and without O2' methylation of the antisense strand. The methylation influcences minor groove hydration, in particular near guanines but also around the methylated O2', and it also reduces the flexibility of both RNA strands. A G.U wobble pair adjacent to the bulge-loop is also found to increase the flexibility of the bulge nucleotides, compared to the situation with a standard Watson-Crick G-C base-pair in the same position.     

Conserved nucleotides in the TAR RNA stem of human immunodeficiency virus type 1 are critical for Tat binding and trans activation: model for TAR RNA tertiary structure.

Interaction between the human immunodeficiency virus type 1 (HIV-1) trans-activator Tat and its cis-acting responsive RNA element TAR is necessary for activation of HIV-1 gene expression. We investigated the hypothesis that the essential uridine residue at position 23 in the bulge of TAR RNA is involved in intramolecular hydrogen bonding to stabilize an unique RNA structure required for recognition by Tat. Nucleotide substitutions in the two base pairs of the TAR stem directly above the essential trinucleotide bulge that maintain base pairing but change sequence prevent complex formation with Tat in vitro. Corresponding mutations tested in a trans-activation assay strongly affect the biological activity of TAR in vivo, suggesting an important role for these nucleotides in the Tat-TAR interaction. On the basis of these data, a model is proposed which implicates uridine 23 in a stable tertiary interaction with the GC pair directly above the bulge. This interaction would cause widening of the major groove of the RNA, thereby exposing its hydrogen-bonding surfaces for possible interaction with Tat. The model also predicts a gap between uridine 23 and the first base pair in the stem above, which would require one or more unpaired nucleotides to close, but does not predict any other role for such nucleotides. In accordance with this prediction, synthetic propyl phosphate linkers of equivalent length to 1 or 2 nucleotides, were found to be fully acceptable substitutes in the bulge above uridine 23, demonstrating that neither the bases nor the ribose moieties at these positions are implicated in the recognition of TAR RNA by Tat.     

A primordial RNA modification enzyme: the case of tRNA (m1A) methyltransferase

The modified nucleoside 1-methyladenosine (m¹A) is found in the T-loop of many tRNAs from organisms belonging to the three domains of life (Eukaryota, Bacteria, Archaea). In the T-loop of eukaryotic and bacterial tRNAs, m¹A is present at position 58, whereas in archaeal tRNAs it is present at position(s) 58 and/or 57, m¹A57 being the obligatory intermediate in the biosynthesis of 1-methylinosine (m¹I57). In yeast, the formation of m¹A58 is catalysed by the essential tRNA (m¹A58) methyltransferase (MTase), a tetrameric enzyme that is composed of two types of subunits (Gcd14p and Gcd10p), whereas in the bacterium Thermus thermophilus the enzyme is a homotetramer of the TrmI polypeptide. Here, we report that the TrmI enzyme from the archaeon Pyrococcus abyssi is also a homotetramer. However, unlike the bacterial site-specific TrmI MTase, the P.abyssi enzyme is region-specific and catalyses the formation of m¹A at two adjacent positions (57 and 58) in the T-loop of certain tRNAs. The stabilisation of P.abyssi TrmI at extreme temperatures involves intersubunit disulphide bridges that reinforce the tetrameric oligomerisation, as revealed by biochemical and crystallographic evidences. The origin and evolution of m¹A MTases is discussed in the context of different hypotheses of the tree of life.     

Comparison of left-end DNA sequences of bacteriophages Mu and D108

The nucleotide sequences of the left ends of bacteriophage Mu DNA and that of its close relative D108 have been determined. The first 100 bp of phages Mu and D108 are substantially the same except for an octanucleotide change from bp 53 to 61 and other small interspersed base-pair changes from bp 61 to 200. The first five host nucleotides preceding the host-phage junction are generally, but not always, G + C-rich and these five nucleotides display no obvious consensus sequence. Both phages Mu and D108 share striking similarity in their end DNA sequences to the end sequences of the newly described Escherichia coli movable genetic element IS30.     

Transfer RNA genes from Dictyostelium discoideum are frequently associated with repetitive elements and contain consensus boxes in their 5' and 3'-flanking regions.

A total of 68 different tRNA genes from the cellular slime mold Dictyostelium discoideum have been isolated and characterized. Although these tRNA genes show features common to typical nuclear tRNA genes from other organisms, several unique characteristics are apparent: (1) the 5'-proximal flanking region is very similar for most of the tRNA genes; (2) more than 80% of the tRNA genes contain an "ex-B motif" within their 3'-flanking region, which strongly resembles characteristics of the consensus sequence of a T-stem/T-loop region (B-box) of a tRNA gene; (3) probably more than 50% of the tRNA genes in certain D. discoideum strains are associated with a retrotransposon, termed DRE (Dictyostelium repetitive element), or with a transposon, termed Tdd-3 (Transposon Dictyostelium discoideum). DRE always occurs 50 (+/- 3) nucleotides upstream and Tdd-3 always occurs 100 (+/- 20) nucleotides downstream from the tRNA gene. D. discoideum tRNA genes are organized in multicopy gene families consisting of 5 to 20 individual genes. Members of a particular gene family are identical within the mature tRNA coding region while flanking sequences are idiosyncratic.