Search

Accurate RNA structure modeling is an important, incompletely solved, challenge. Single-nucleotide resolution SHAPE (selective 2'-hydroxyl acylation analyzed by primer extension) yields an experimental measurement of local nucleotide flexibility that can be incorporated as pseudo-free energy change constraints to direct secondary structure predictions. Prior work from our laboratory has emphasized both the overall accuracy of this approach and the need for nuanced interpretation of modeled structures. Recent studies by Das and colleagues [Kladwang, W., et al. (2011) Biochemistry 50, 8049; Nat. Chem. 3, 954], focused on analyzing six small RNAs, yielded poorer RNA secondary structure predictions than expected on the basis of prior benchmarking efforts. To understand the features that led to these divergent results, we re-examined four RNAs yielding the poorest results in this recent work: tRNA(Phe), the adenine and cyclic-di-GMP riboswitches, and 5S rRNA. Most of the errors reported by Das and colleagues reflected nonstandard experiment and data processing choices, and selective scoring rules. For two RNAs, tRNA(Phe) and the adenine riboswitch, secondary structure predictions are nearly perfect if no experimental information is included but were rendered inaccurate by the SHAPE data of Das and colleagues. When best practices were used, single-sequence SHAPE-directed secondary structure modeling recovered ~93% of individual base pairs and >90% of helices in the four RNAs, essentially indistinguishable from the results of the mutate-and-map approach with the exception of a single helix in the 5S rRNA. The field of experimentally directed RNA secondary structure prediction is entering a phase focused on the most difficult prediction challenges. We outline five constructive principles for guiding this field forward.

The ability to predict RNA secondary structure is fundamental for understanding and manipulating RNA function. The information obtained from selective 2'-hydroxyl acylation analyzed by primer extension (SHAPE) experiments greatly improves the accuracy of RNA secondary structure prediction. Recently, Das and colleagues [Kladwang, W., et al. (2011) Biochemistry 50, 8049-8056] proposed a "bootstrapping" approach for estimating the variance and helix-by-helix confidence levels of predicted secondary structures based on resampling (randomizing and summing) the measured SHAPE data. We show that the specific resampling approach described by Kladwang et al. introduces systematic errors and underestimates confidence in secondary structure prediction using SHAPE data. Instead, a leave-data-out jackknife approach better estimates the influence of a given experimental data set on SHAPE-directed secondary structure modeling. Even when 35% of the data were left out in the jackknife approach, the confidence levels of SHAPE-directed secondary structure prediction were significantly higher than those calculated by Das and colleagues using bootstrapping. Helix confidence levels were thus underestimated in the recent study, and the resampling approach implemented by Kladwang et al. is not an appropriate metric for evaluating SHAPE-directed secondary structure modeling.

Single-nucleotide-resolution chemical mapping for structured RNA is being rapidly advanced by new chemistries, faster readouts, and coupling to computational algorithms. Recent tests have shown that selective 2'-hydroxyl acylation by primer extension (SHAPE) can give near-zero error rates (0-2%) in modeling the helices of RNA secondary structure. Here, we benchmark the method using six molecules for which crystallographic data are available: tRNA(phe) and 5S rRNA from Escherichia coli, the P4-P6 domain of the Tetrahymena group I ribozyme, and ligand-bound domains from riboswitches for adenine, cyclic di-GMP, and glycine. SHAPE-directed modeling of these highly structured RNAs gave an overall false negative rate (FNR) of 17% and a false discovery rate (FDR) of 21%, with at least one helix prediction error in five of the six cases. Extensive variations of data processing, normalization, and modeling parameters did not significantly mitigate modeling errors. Only one varation, filtering out data collected with deoxyinosine triphosphate during primer extension, gave a modest improvement (FNR = 12%, and FDR = 14%). The residual structure modeling errors are explained by the insufficient information content of these RNAs' SHAPE data, as evaluated by a nonparametric bootstrapping analysis. Beyond these benchmark cases, bootstrapping suggests a low level of confidence (<50%) in the majority of helices in a previously proposed SHAPE-directed model for the HIV-1 RNA genome. Thus, SHAPE-directed RNA modeling is not always unambiguous, and helix-by-helix confidence estimates, as described herein, may be critical for interpreting results from this powerful methodology., (© 2011 American Chemical Society)

The structure of the eukaryotic L5-5S rRNA complex was investigated in protection and interference experiments and is compared with the corresponding structure (L18-5S rRNA) in the Haloarcula marismortui 50S subunit. In close correspondence with the archaeal structure, the contact sites for the eukaryotic ribosomal protein are located primarily in helix III and loop C and secondarily in loop A and helix V. While the former is unique to L5, the latter is also a critical contact site for transcription factor IIIA (TFIIIA), accounting for the mutually exclusive binding of these two proteins to 5S RNA. The binding of L5 causes structural changes in loops B and C that expose nucleotides that contact the Xenopus L11 ortholog in H. marismortui. This induced change in the structure of the RNA reveals the origins of the cooperative binding to 5S rRNA that has been observed for the bacterial counterparts of these proteins. The native structure of helix IV and loop D antagonizes binding of L5, indicating that this region of the RNA is dynamic and also influenced by the protein. Examination of the crystal structures of Thermus thermophilus ribosomes in the pre- and post-translocation states identified changes in loop D and in the surrounding region of 23S rRNA that support the proposal that 5S rRNA acts to transmit information between different functional domains of the large subunit.

A class of RNA aptamers that demonstrates a high affinity for a large variety of C(2)H(2) zinc finger proteins was isolated from a library of random RNA sequences by the zinc finger protein TFIIIA. These aptamers have one or more copies of the consensus sequence GGGUGGG, which is part of a putative hairpin loop in the proposed structure of the most abundant aptamer, RNA1. Binding of zinc finger proteins to RNA1 relies upon zinc-dependent folding of the protein, the affinity of an individual protein for RNA1 being determined by the number of tandem zinc finger motifs. The properties of RNA1 were compared to the properties of two other aptamers from the same selection experiment: RNA21, which binds to some but not all zinc finger proteins tested, and RNA22, which binds only to the 5S rRNA binding zinc finger proteins TFIIIA and p43. The binding of three different zinc finger proteins to RNA1 was compared, and the results indicate that the RNA1-protein interaction occurs by several distinct mechanisms. Mutagenesis of RNA1 confirmed that the GGGUGGG consensus sequence presented in a hairpin conformation is required for high-affinity binding of zinc finger proteins.

A method to deduce RNA secondary structure on the basis of data from microarrays of 2'-O-methyl RNA 9-mers immobilized in agarose film on glass slides is tested with a 249 nucleotide RNA from the 3' end of the R2 retrotransposon from Bombyx mori. Various algorithms incorporating binding data and free-energy minimization calculations were compared for interpreting the data to provide possible secondary structures. Two different methods give structures with 100 and 87% of the base pairs determined by sequence comparison. In contrast, structures predicted by free-energy minimization alone by Mfold and RNAstructure contain 52 and 72% of the known base pairs, respectively. This combination of high throughput microarray techniques with algorithms using free-energy calculations has potential to allow for fast determination of RNA secondary structure. It should also facilitate the design of antisense and siRNA oligonucleotides.

Determining RNA secondary structure is important for understanding structure-function relationships and identifying potential drug targets. This paper reports the use of microarrays with heptamer 2'-O-methyl oligoribonucleotides to probe the secondary structure of an RNA and thereby improve the prediction of that secondary structure. When experimental constraints from hybridization results are added to a free-energy minimization algorithm, the prediction of the secondary structure of Escherichia coli 5S rRNA improves from 27 to 92% of the known canonical base pairs. Optimization of buffer conditions for hybridization and application of 2'-O-methyl-2-thiouridine to enhance binding and improve discrimination between AU and GU pairs are also described. The results suggest that probing RNA with oligonucleotide microarrays can facilitate determination of secondary structure.

The formation of the Xenopus L5-5S rRNA complex depends on nonelectrostatic interactions. Fluorescence assays with 1-anilino-8-naphthalenesulfonate demonstrate that a hydrophobic region on L5 becomes exposed upon removal of bound 5S rRNA by treatment with ribonucleases. Several conserved aromatic amino acids, mostly tyrosines, were identified by comparative sequence analysis and changed individually to alanine. Substitution with alanine at any of three positions, Y86, Y99, or Y226, essentially abolishes RNA-binding activity, whereas those made at Y95 and Y207 have more modest effects. Replacement with phenylalanine at Y86 and Y226 does not change binding affinity, indicating that the aromatic ring of the side chain, not the hydroxyl group, is the critical functionality for binding. Alternatively, the phenolic hydroxyls at Y99 and Y207 do contribute to binding. The structural integrity of the mutant proteins was assessed using thermal denaturation and limited digestion with proteases. The T(m) of Y99A is 10 degrees C lower than that of the wild-type protein, and there are some differences in the protease digestion patterns that together indicate the structure of this mutant has been significantly perturbed. The structures of the other variants are not detectably different from the wild-type protein. These results provide evidence that intermolecular stacking interactions involving at least two tyrosine residues, Y86 and Y226, are necessary for formation of the L5-5S rRNA complex and can account, at least in part, for the contribution nonelectrostatic interactions make to the free energy of binding.

Diepoxyalkanes form interstrand cross-links in DNA oligomers preferentially at 5'-GNC sites. We have examined cross-linking by 1,2,3,4-diepoxybutane (DEB) and 1,2,7,8-diepoxyoctane (DEO) within a fragment of the 5S RNA gene of Xenopus borealis in both the free and nucleosomal states. Sites and efficiencies of interstrand cross-linking were probed through denaturing polyacrylamide gel electrophoresis and quantitative phosphorimagery. Both agents targeted 5'-GNC sites for cross-linking in the restriction fragment in its free state, and DEO also targeted 5'-GNNC sites. Monoalkylation occurred at all deoxyguanosines. The sites for both monoalkylation and interstrand cross-linking were similar in nucleosomal and free DNA, and cross-linked DNA was cleanly incorporated into the core particle structure. These findings suggest that the 5S core particle is able to tolerate any structural abnormalities induced by diepoxide cross-linking.

RNA multibranch loops (junctions) are loops from which three or more helices exit. They are nearly ubiquitous in RNA secondary structures determined by comparative sequence analysis. In this study, systems in which two strands combine to form three-way junctions were used to measure the stabilities of RNA multibranch loops by UV optical melting and isothermal titration calorimetry (ITC). These data were used to calculate the free energy increment for initiation of a three-way junction on the basis of a nearest neighbor model for secondary structure stability. Imino proton NMR spectra were also measured for two systems and are consistent with the hypothesized helical structures. Incorporation of the experimental data into the mfold and RNA structure computer programs has contributed to an improvement in prediction of RNA secondary structure from sequence.

Interstrand cross-linking studies with the antitumor drug cis-diamminedichloroplatinum(II) and its clinically inactive isomer, trans-diamminedichloroplatinum(II), were performed on a fragment of the 5S rRNA gene of Xenopus borealis in the free and nucleosomal state. 5S nucleosomes were formed via histone octamer exchange from chicken erythrocyte core particles. Native polyacrylamide gel electrophoresis was used to probe the ability of platinated DNA to reconstitute into core particles. Both isomers negatively impacted reconstitution when histones were present during incubation with the drug. When histones were not present during the drug treatment, platinated DNA was successfully reconstituted into core particles. These results suggest that platination of histones impedes reconstitution of free DNA. However, already-formed core particles were not disrupted upon platination. Sites of interstrand cross-linking were probed through denaturing polyacrylamide gel electrophoresis and quantitative phosphorimagery. We found both site-specific enhancement and depression of cis-diamminedichloroplatinum(II) cross-linking in the nucleosomal samples relative to free DNA at both drug concentrations that were tested (0.01 and 0.0025 mM). trans-Diamminedichloroplatinum(II) exhibited no detectable differences in the interstrand cross-linking of free and nucleosomal samples.

Ribosomal protein L18 from Bacillus stearothermophilus (bL18) includes a previously unreported phosphoserine residue. The folded conformation of the protein is stabilized by the dianionic form of the phosphate group of that residue. In the absence of Mg2+, the pK(a) of the phosphate group is so high that the protein is not fully folded at pH 7. In the presence of Mg2+, its pK(a) drops significantly, and consequently the native conformation of bL18 becomes stable at pH 7 and the protein is able to bind to 5S rRNA. Dephosphorylated bL18 does not bind to 5S rRNA at neutral pH.

Ribosomal 5S RNA presents a particular challenge to structural investigations since this polynucleotide is too large for complete NMR characterization but lacks significant tertiary structure to modulate, for example, diagnostic alkylation of guanine N7 by dimethyl sulfate. Nickel- and cobalt-dependent reagents that are sensitive to the N7 and aromatic face of guanine have now been applied to 5S rRNA (Xenopus lavis) and provide structural information that was not previously available from traditional chemical or enzymatic probes. Although G75 had repeatedly demonstrated an average reactivity with dimethyl sulfate and minimal reactivity with RNase T1, this residue was the major target of both metal-dependent reagents. Such reactivity provides crucial support for a structural model of loop E identified by prior physical, but not chemical, methods. Similarly, the tetraloop structure of loop D was more accurately reflected by the reactivity of G87 and G89 in the presence of the nickel reagent rather than in the presence of RNase T1. In addition, nickel-dependent modification of guanine residues surrounding the three-helix junction of loop A suggests an organization that is less compact than previously considered.

The Rev protein of the human immunodeficiency virus type 1 (HIV-1) has been studied by time-resolved fluorescence spectroscopy. The single tryptophan residue of Rev, Trp45, located within the arginine-rich RNA-binding domain of the protein, was utilized as an intrinsic spectroscopic probe. In addition, five peptides spanning different lengths of the arginine-rich domain, each containing the tryptophan residue, and two C-terminal deletion mutants of Rev, Rev M9 delta 14 and Rev M11 delta 14, were examined. Rev M9 delta 14 lacks residues 68-112 whereas Rev M11 delta 14 is missing residues 92-112 of the C-terminus of Rev. The fluorescence decay of Trp45 in wild-type Rev was resolved into four discrete lifetime components, and decay-associated spectra (DAS) were obtained for each component. The fluorescence decays of all five peptides and Rev M9 delta 14 were resolved into three lifetime components. The fluorescence decay of Rev M11 delta 14 was resolved into four components similar to those found for wild-type Rev. These results indicate that the activation domain (residues 78-93), present in wild-type Rev and Rev M11 delta 14, induced a unique tryptophan environment, characterized by a short-lived, blue-shifted emission, attributed to higher order assembly of Rev. In addition, fluorescence anisotropy decay data obtained for wild-type Rev and the two C-terminal deletion mutants also indicate that the activation domain mediates self-association of Rev. Based on the anisotropy decay results for wild-type Rev, the distribution of oligomers is independent of salt concentration. The average fluorescence lifetime of Trp45 was reduced upon complexation of Rev with a 40-mer fragment of the Rev response element containing the minimal element for Rev binding (F8-RRE), and the emission was blue-shifted. In addition, the local rotation of the tryptophan side chain was blocked in the protein-RRE complex. These results indicate that Trp45 directly interacts with the RRE. Rev is also shown to bind to 5S RNA, resulting in very similar changes in the time-resolved tryptophan fluorescence to those observed upon complexation of Rev with F8-RRE.

Signal recognition particle (SRP) is a ribonucleoprotein complex involved in the targeting of secretory proteins to the lipid bilayer of the endoplasmic reticulum. SRP contains the protein SRP19, which is an important structural and functional component, believed to promote the assembly of the particle. We have purified the human SRP19 protein to homogeneity from recombinant bacteria which overexpress the polypeptide, and have studied details of the binding to SRP RNA via gel mobility shift and RNase sensitivity assays. SRP19 interacts with two SRP RNA conformers with different affinities such that the more compact RNA species is bound more avidly. Furthermore, binding was found to be highly cooperative. Binding constants and Hill coefficients were determined for several mutant SRP RNAs in which individual RNA helices were deleted. These results confirmed that both SRP RNA helices 6 and 8 are important for SRP19 binding. Enzymatic RNA structure probing of a 150-nucleotide mutant SRP RNA fragment and of the corresponding RNA-SRP19 complex showed that cooperativity may be due to protein-induced conformational changes in the large domain of the SRP RNA. Finally, SRP19 bound specifically not only to SRP RNA but also to the A-form of Escherichia coli 5S ribosomal RNA, thereby indicating structural similarities between these two RNA molecules.

The effect of iron(II) bleomycin on a DNA-RNA heteroduplex was investigated using a substrate formed by reverse transcription of Escherichia coli 5S ribosomal RNA. Both strands of the heteroduplex were cleaved by FeII.BLM A2 at comparable concentrations; complete digestion of both strands was observed using 5 microM FeII.BLM A2. The DNA strand of the heteroduplex was cleaved predominantly at 5'-G-pyr-3' sites; the sites of cleavage of the DNA strand were a subset of those observed for the corresponding DNA strand of a DNA duplex of identical sequence. The sites of cleavage of the RNA strand of the heteroduplex involved both purines and pyrimidines and were found to be different than the sites of cleavage of the 5S rRNA alone, demonstrating that cleavage of the former must actually have involved heteroduplex recognition by FeII.BLM A2. Both the DNA and RNA strands of the heteroduplex were cleaved by FeII.BLM A2 in the presence of physiological concentrations of Mg2+, consistent with the possibility that DNA-RNA heteroduplexes may be therapeutically relevant targets for bleomycin.

We have used gel electrophoretic mobility measurements to investigate the conformation of the symmetric eubacterial loop E sequence of 5S rRNA (seven nucleotides in each strand). The loop strongly retarded the gel mobility of duplex RNAs containing it. In contrast, only asymmetric A5.An or U5.Un internal loops (n not equal to 5) strongly affected duplex RNA gel mobility. A phasing experiment, in which an A2 bulge and loop E were placed in the same duplex RNA and the number of base pairs between them varied, showed that loop E has a permanent bend and is torsionally stiff. A second phasing experiment substituting loop E for duplex sequences between two A2 bulges measured the helical twist associated with loop E; it is about 30 degrees (+/- 15 degrees) overwound compared to a duplex RNA of the same number of bases. Ribosomal protein L25 specifically recognizes loop E but had little or no effect on the twist of the loop. These results suggest that loop E adopts a specific, roughly helical structure.

The effects of a series of point mutations within the Xenopus borealis somatic-type 5S RNA gene on transcription factor IIIA (TFIIIA) binding affinity were quantified. These data define a critical sequence-dependent contact region within the classical box C promoter element from base pair 80 to 91. Substitution of GC base pairs at positions 81, 85, 86, 89, and 91 significantly reduce TFIIIA binding affinity. Base pairs located at other positions within the box C contact region provide a moderate contribution to TFIIIA-5S gene interaction. In contrast to the extensive set of sequence contacts within the box C element, TFIIIA interaction is localized primarily to two GC base pairs at positions 70 and 71 within the intermediate promoter element. A selected amplification and binding assay (SAAB) was performed with a synthetic internal control region (ICR) randomized from base pair 78 to 95 to identify box C promoter sequences bound with high affinity by TFIIIA. The wild-type 5S RNA gene sequence from 79 to 92 is strongly selected. These results are consistent with the critical role of the box C element in sequence-dependent promoter recognition by TFIIIA.

We have studied the binding of a chemical carcinogen to DNA reconstituted with histone octamers to determine the effect that nucleosome structure has on covalent adduct formation. Reconstitution of a plasmid containing the somatic 5S rRNA gene from Xenopus borealis resulted in characteristic nucleosome structure, as determined by micrococcal nuclease digestion, shifted migration in agarose gels, and hydroxyl radical footprinting. Formation of covalent adducts by benzo[a]pyrenediol epoxide (BPDE) occurred initially at a slower rate in reconstituted DNA than in naked plasmid, but after 2 h the total adduction levels (adducts/plasmid) were equal in both samples. Analysis of adduction at the sequence level by primer extension indicated that, after a 2-h BPDE reaction, the degree of adduction within the 5S rRNA nucleosome was suppressed by approximately 50% compared to naked DNA. The rotational setting of the guanines on the helix did not explain the level of adduction observed, since guanines in close proximity to the histone core were equally susceptible to adduction as guanines on the outer nucleosome surface. At early reaction times with BPDE, however, sequences near the 5S nucleosome dyad, where known modulations in the minor groove width occur, were the least susceptible to adduction. These results indicate that the structural features of DNA assembled into nucleosomes contribute to the susceptibility of the DNA to modification by BPDE.

The solution structure of a 27-nucleotide duplex, including the internal loop E from Xenopus laevis 5S ribosomal RNA, has been studied by two-dimensional NMR spectroscopy, followed by restrained molecular dynamics. The highly conserved internal loop closes to form a G.A base pair and a reverse-Hoogsteen A.U base pair. Extensive interstrand stacking between these uncommon base pairs provides a structural explanation for an interstrand ultraviolet-induced cross-link. A guanosine residue is bulged into the major groove and may form a base-triple with the adjacent reverse-Hoogsteen A.U pair. The structure of the less highly-conserved portion of the loop is less well-defined by the NMR data. A single-nucleotide deletion mutant has a very different, open conformation without mismatched base pairs [Varani, G., Wimberly, B., & Tinoco, I. Jr. (1989) Biochemistry 28, 7760-7772]. The implications of the structure for binding of the transcription factor TFIIIA and the cytotoxin alpha-sarcin are discussed.

Nucleosomes were assembled onto either closed circular plasmids containing a single Xenopus 5S RNA gene or a linear tandemly repeated array of Lytechinus 5S RNA genes. Both chromatin templates were found to vary in their extent of compaction, depending upon the type and concentration of cation in solution. Compaction of these chromatin templates led to a significant inhibition of both transcription initiation and elongation by RNA polymerase III. Thus, the transcriptional repression observed after incorporation of genes into chromatin depends not only on occlusion of the promoter elements through direct contact with histones but also on compaction of nucleosomal arrays which occurs under the conditions of the transcription reactions.

The coordination complex tris(4,7-diphenyl-1,10-phenanthroline)rhodium(III) [Rh(DIP)3(3+)], which promotes RNA cleavage upon photoactivation, has been shown to target specifically guanine-uracil (G-U) mismatches in double-helical regions of folded RNAs. Photoactivated cleavage by Rh(DIP)3(3+) has been examined on a series of RNAs that contain G-U mismatches, yeast tRNA(Phe) and yeast tRNA(Asp), as well as on 5S rRNAs from Xenopus oocytes and Escherichia coli. In addition, a "microhelix" was synthesized, which consists of seven base pairs of the acceptor stem of yeast tRNA(Phe) connected by a six-nucleotide loop and contains a mismatch involving residues G4 and U69. A U4.G69 variant of this sequence was also constructed, and cleavage by Rh(DIP)3(3+) was examined. In each of these cases, specific cleavage is observed at the residue which lies to the 3'-side of the wobble-paired U; some cleavage by the rhodium complex is also evident in several structured RNA loops. The remarkable site selectivity for G-U mismatches within double-helical regions is attributed to shape-selective binding by the rhodium complex. This binding furthermore depends upon the orientation of the G-U mismatch, which produces different stacking interactions between the G-U base pair with the Watson-Crick base pair following it on the 5'-side of U compared to the Watson-Crick pair preceding it on the 3'-side of U. Rh(DIP)3(3+) therefore serves as a unique probe of G-U mismatches and may be useful both as a model and in probing RNA-protein interactions as well as in identifying G-U mismatches within double-helical regions of folded RNAs.

The three-dimensional folding of Xenopus oocyte 5S rRNA has been examined using the coordination complex Rh(phen)2phi3+ (phen = phenanthroline; phi = phenanthrenequinone diimine) as a structural probe. Rh(phen)2phi3+ binds neither double-helical RNA nor unstructured single-stranded regions of RNA. Instead, the complex targets through photoactivated cleavage sites of tertiary interaction which are open in the major groove and accessible to stacking. The sites targeted by the rhodium complex have been mapped on the wild-type Xenopus oocyte RNA, on a truncated RNA representing the arm of the molecule comprised of helix IV-loop E-helix V, and on several single-nucleotide mutants of the 5S rRNA. On the wild-type 5S rRNA, strong cleavage is found at residues U73, A74, A101, and U102 in the E loop and U80 and G81 in helix IV; additional sites are evident at A22 and A56 in the B loop, C29 and A32 in helix III, and C34, C39, A42, and C44 in the C loop. Given the similarity observed in cleavage between the full 5S RNA and the truncated fragment as well as the absence of any long-range effects on cleavage in mutant RNAs, the results do not support models which involve long-range tertiary interactions. Cleavage results with Rh(phen)2phi3+ do, however, indicate that the apposition of several noncanonical bases as well as stem--loop junctions may result in intimately stacked structures with opened major grooves. In particular, on the basis of cleavage results on mutant RNAs, both loops C and E represent structures where the strands constituting each loop are not independent of one another but are intrinsically structured.(ABSTRACT TRUNCATED AT 250 WORDS)

The structure of helix I of the 5S rRNA from Escherichia coli has been determined using a nucleolytic digest fragment of the intact molecule. The fragment analyzed, which corresponds to bases (-1)-11 and 108-120 of intact 5S rRNA, contains a G-U pair and has unpaired bases at its termini. Its proton resonances were assigned by two-dimensional NMR methods, and both NOE distance and coupling constant information have been used to calculate structural models for it using the full relaxation matrix algorithm of the molecular dynamics program XPLOR. Helix I has A-type helical geometry, as expected. Its most striking departure from regular helical geometry occurs at its G-U, which stacks on the base pair to the 5' side of its G but not on the base pair to its 3' side. This stacking pattern maximizes interstrand guanine-guanine interactions and explains why the G-U in question fails to give imino proton NOE's to the base pair to 5' side of its G. These results are consistent with the crystal structures that have been obtained for wobble base pairs in tRNAPhe [Mizuno, H., & Sundaralingam, M. (1978) Nucleic Acids Res. 5, 4451-4461] and A-form DNA [Rabbinovich, D., Haran, T., Eisenstein, M., & Shakked, Z. (1988) J. Mol. Biol. 200, 151-161]. The conformations of the terminal residues of helix I, which corresponds to bases (-1)-11 and 108-120 of native 5S RNA, are less well-determined, and their sugar puckers are intermediate between C2' and C3'-endo, on average.