Stabilization of the anticodon stem-loop of tRNA Lys,3 by an A + -C base-pair and by pseudouridine 1 1Edited by I. Tinoco

A conserved pseudouridine modification in eukaryotic U2 snRNA induces a change in branch-site architecture

“…Thermal denaturation studies of branch-site duplex constructs demonstrated added stability upon substitution of uridine with pseudouridine at the conserved position in the U2 sequence (Table 5)+ The decrease in ⌬G8 37 of 0+7 kcal/mol in cBP compared with its unmodified counterpart was comparable to free energy differences measured for tRNA anticodon stem loops with pseudouridine incorporation at the closing base pair of the loop (Durant & Davis, 1999;Yarian et al+, 1999)+ A similar decrease was noted by Hall and McLaughlin (1991) upon substitution of two side-by-side uridine residues with pseudouridine in a self-complementary 11-nt duplex+ Although the sequences varied in these experiments, the data suggest that increased stability is a common feature of pseudouridine, probably induced by the additional hydrogen bonding in each case+ Incorporation of the pseudouridine also resulted in a more cooperative melting transition than observed in the unmodified duplex, which, combined with the slower exchange of protons observed in the unpaired region of the cBP duplex, suggests that stacking and additional hydrogen bonding in the bulged region contributes to the stability of the duplex+ In contrast, incorporation of a deoxypseudouridine residue at the same location increased the free energy of formation (⌬G8 37 ) of the branch-site interaction by 0+7 kcal/mol relative to uBP (and by 1+4 kcal/mol with respect to cBP)+ This finding is in accord with experimental data of Bevilaqua and Turner (1991) that demonstrated unfavorable ⌬G8 37 contributions to RNA duplex formation by deoxyriboses+ Davis (1995) observed that the values of J H19-H29 couplings measured for riboses in a short single-stranded oligo decreased when a pseudouridine residue was substituted for uridine in a given sequence+ Applying the generally held view that the thermal stability of RNA helices is related to A-form helical parameters (Lee & Tinoco, 1977), Davis attributed the greater propensity toward C39-endo ribose conformation to increased base stacking+ This conclusion was further supported by circular dichroism experiments in which incorporation of a pseudouridine residue caused increased temperature dependence of ellipticity at 270 nm (Davis, 1995)+ In our experiments, however, J H19-H29 couplings indicative of non-C39-endo conformation were measured for the riboses of residues opposing the pseudouridine in cBP (U22, A23, and A24), and one-dimensional phosphorous studies showed a 30% broader chemical shift range for cBP compared with an analogous unbulged helix (ubBP)+ In contrast, all nonterminal riboses of uBP had very small couplings characteristic of a C39-endo pucker, and one-dimensional phosphorous NMR data suggest that the backbone conformation of uBP differs very little from an unbulged A-form helix+ We consider it very interesting, therefore, that the presence of the pseudouridine in cBP resulted in a deviation from helical parameters, but that this duplex also exhibits greater thermal stability than uBP, the duplex that appears to maintain A-form geometry+ Although the chemical shift of individual phosphorous resonances is not a singularly reliable predictor of specific perturbations in backbone geometry (Nikonowicz & Gorenstein, 1990), we speculate that the increased stability of cBP is the result of stabilizing interactions involving the pseudouridine NH1 and unusual backbone conformation in the bulged...…”

“…The structural features of the eukaryotic branch-site region that promote its unique recognition and chemical activity are of great interest+ Results of biochemical studies in which individual base functional groups were modified imply an extrahelical orientation of the adenine base (Query et al+, 1996), although solution NMR structures of other RNAs with unpaired adenosines show the base intercalated between adjacent base pairs (e+g+, Borer et al+, 1995;Smith & Nikonowicz, 1998;Varani et al+, 1999;see, however, Greenbaum et al+, 1996)+ Of particular interest is the RNA that forms the binding site for phage GA coat protein, because of the sequence similarity it shares with the U2 snRNA-intron branchsite pairing+ The solution structure of this RNA, characterized by an unpaired adenine that stacks within the helix (Smith & Nikonowicz, 1998), prompted us to ask whether the adenosine residue of the spliceosomal branch site is also stacked within the helix or adopts an extrahelical orientation+ Posttranscriptionally modified bases in snRNAs, including pseudouridines (c; a uridine base rotated to have a carbon-carbon glycosidic linkage and a second exposed imino group; Fig+ 1D), have long been noted+ Pseudouridines in tRNA molecules have been shown to stabilize existing helical structure by providing an additional hydrogen bond donor that participates in a water-mediated hydrogen bond with backbone phosphates (Arnez & Steitz, 1994;Davis, 1995)+ Molecular dynamics simulations support the premise that pseudouridines restrict the motion of neighboring bases (Auffinger & Westhof, 1998)+ Accordingly, the presence of pseudouridine residues has been associated with stabilization of short duplex RNAs and tRNA anticodon loops without perturbing local structure (Davis & Poulter, 1991;Hall & McLaughlin, 1991;Arnez & Steitz, 1994;Durant & Davis, 1999;Yarian et al+, 1999)+ Chemical mapping of U2 snRNA sequences of a number of species has documented many conserved pseudouridylation sites in U2 snRNA, particularly in the FIGURE 1. A: Commitment-complex snRNA-intron pairings+ The U2 snRNA pairs with the branch-site region of the intron, and the U1 snRNA forms an analogous pairing with the 59 splice site+ Sequences shown are from yeast+ c ϭ pseudouridine+ B: Yeast U2 snRNA-intron branch-site interaction shown in context with nearby U6 snRNA interactions observed upon complex B assembly (Madhani & Guthrie, 1994)+ Pairing of U2 snRNA with the branch-site sequence of the intron results in the bulging of the adenosine residue responsible for nucleophilic attack at the 59 splice site (shown by an arrow)+ Y ϭ any pyrimidine+ C: U2 snRNA-intron branch-site constructs used for NMR spectroscopy in these studies+ The unmodified duplex (uBP) and c-modified duplex (cBP) constructs are shown with open dots indicating base pairs expected from the predicted secondary structure+ So...…”

“…The propensity for extrahelical conformations of unpaired adenosines in crystallized molecules as a result of endto-end stacking and intermolecular interactions has been well documented (Hermann & Patel, 2000)+ Very few NMR-derived structures depict single-bulged adenosines in extrahelical base conformations+ In solution, it is simply less energetically favorable to expose a base to solvent than to stack one within a helix (Thiviyanathan et al+, 2000)+ It is therefore intriguing that a U r c modification in a base pair adjacent to the unpaired adenosine in the branch-site duplex is sufficient to shift the energetically favored conformation of the base from a stacked to an extrahelical position+ Pseudouridines in other contexts assume a stabilizing role in RNA structure+ Water-mediated hydrogen bonds between pseudouridine N1 imino protons and phosphate oxygen atoms, predicted by molecular dynamics simulations (Auffinger & Westhof, 1998), and observed in tRNA anticodon loop structures determined by NMR (Durant & Davis, 1999) and X-ray crystallographic methods (Arnez & Stietz, 1994), contribute to the stability of RNA molecules (Hall & McLaughlin, 1991Davis, 1995)+ Pseudouridines are also abundant in all of the snRNAs, although their specific roles are not known+ A study by Hall and McLaughlin (1991) of a duplex representing the pairing of human U1 snRNA and the intron 59 splice site suggests that two side-by-side pseudouridines in the U1 snRNA sequence (Fig+ 1A) do not alter helix structure, but together stabilize duplex formation by approximately 0+4-0+5 kcal/mol+ Although the duplex in this study contained two pseudouridines, there were no mismatched base pairs or bulged bases (the human intron sequence has two adenosines that oppose the pseudouridines in the U1 snRNA sequence)+ In contrast, the single pseudouridine residue in the U2 snRNA-intron duplex apparently forms a base pair with an adenosine that is immediately adjacent to an unpaired adenosine, and alone contributes approximately 0+7 kcal/mol to duplex stability+ If the pseudouridines in the U1 snRNA-intron duplex form water-mediated hydrogen bonds with phosphate oxygen atoms, as has been suggested (Hall & McLaughlin, 1992;Arnez & Steitz, 1994;Davis, 1995), we might expect the added stabilization to correlate with the number of pseudouridines+ The large decrease in ⌬G8 37 observed upon incorporation of pseudouridine at the conserved site in the U2 snRNA-intron construct implies that c35 may have a different structural role than the pseudouridines in the U1 snRNA-intron helix+ In the study by Hall and McLaughlin (1991), disappearance of pseudouridine NH1 resonances upon increasing temperature was concurrent with the disappearance of all other imino resonances in the duplex+ In contrast, the cNH1 resonance of cBP is the last imino resonance to disappear upon denaturation of the duplex at low pH+ These results indicate that cNH1, presumably located in the major groove of the helix, is more stable upon denaturation than the WatsonCrick bas...…”

Functional Anticodon Architecture of Human tRNALys3 Includes Disruption of Intraloop Hydrogen Bonding by the Naturally Occurring Amino Acid Modification, t6A

“…Structures of the unmodified ASL Lys3 UUU and the stem modified ASL Lys3 UUU -Ψ 39 have been reported (15), but neither bind the ribosome (14). We compare the structure of the functional ASL Lys3 UUU -t 6 A 37 to that of the nonfunctional, unmodified ASL Lys3 UUU .…”

Modification of the universally unmodified uridine-33 in a mitochondria-imported edited tRNA and the role of the anticodon arm structure on editing efficiency

“…Given the large extent of modification of the ASL of the imported mt tRNA Trp , we investigated the specific features of the tRNA Trp that the editing enzyme may utilize for recognition and discrimination in a cellular pool of different tRNAs+ In L. tarentolae, other mitochondrial tRNAs containing a C as the wobble base, such as tRNA Leu (CAG), tRNA Val (CAC), tRNA Glu (CTC), and tRNA Gln (CTG), are not subjected to editing (data not shown)+ We have studied the sequence requirements for tRNA Trp editing in vivo by transfecting cells with plasmids containing mutant tRNA Trp genes+ Previous work showed that L. tarentolae cells can express mutant tRNAs from transfected plasmids and that the tRNAs can be imported into the mitochondrion (Lima & Simpson, 1996;Sbicego et al+, 1998)+ To create specific hybridization sites for RT-PCR primers, five nucleotide changes were introduced in the anticodon arm of the wild-type tRNA Trp sequence, generating the MASL (for mutant of the anticodon stem loop) construct (Fig+ 6A)+ These nucleotide changes allowed amplification of transfected MASL tRNAs without interference from the endogenous wild-type tRNA Trp + This set of mutations was chosen using published sequence alignments of all tRNA Trp genes (Sprinzl et al+, 1998); only positions that covary in the sequence alignment and preserve base pairing were targeted (data not shown)+ L. tarentolae cells were transfected with the pLY15 plasmid (Sbicego et al+, 1998) containing the MASL tRNA gene, and the tagged tRNA Trp was amplified from mitochondrial RNA, and the resulting clones were scored for C34-to-U34 editing+ As shown in Figure 6B (lanes 1 and 3), these mutations did not affect transcription or mitochondrial import of the transfected tRNA variant+ The extent of editing of the MASL variant was similar to that of the wild-type tRNA Trp , as shown by the Hinf Iresistant RT-PCR product (Fig+ 6C, lane 8, and Fig+ 7, MASL 31A-39U)+ These data demonstrate that this MASL tRNA Trp variant is a suitable substrate for studying the effect of other anticodon arm mutations on tRNA editing in vivo+ A single base-pair reversal was introduced at the last position (or at the next to the last position) of the anticodon stem of the MASL tRNA Trp (MASL 31U-39A; Fig+ 7)+ These mutations had no effect on transcription of the gene or on import of the tRNA into the mitochondrion (data not shown), but completely inhibited C34 to U34 editing (Fig+ 7)+ Several testable hypotheses for the lack of editing in the MASL 31U-39A variant include: (1) a requirement for a specific base pair at the base of the anticodon stem, such as in the formation of ⌿ in the T⌿C loop of yeast tRNAs (Becker et al+, 1997); (2) a specific requirement for ⌿ at position 39, which was proposed to be necessary to ensure anticodon loop stability of tRNAs, either by hydrogen bonding with A31 or by the ability to stack with neighboring nucleotides (Durant & Davis, 1999); (3) a requirement for increased stability at the last position of the stem; (4) a requirement for interactions with neighboring nucleotides within the anticodon stem+…”