Volume 12 Supplement 13
Novel base triples in RNA structures revealed by graph theoretical searching methods
© Firdaus-Raih et al; licensee BioMed Central Ltd. 2011
Published: 30 November 2011
Highly hydrogen bonded base interactions play a major part in stabilizing the tertiary structures of complex RNA molecules, such as transfer-RNAs, ribozymes and ribosomal RNAs.
We describe the graph theoretical identification and searching of highly hydrogen bonded base triples, where each base is involved in at least two hydrogen bonds with the other bases. Our approach correlates theoretically predicted base triples with literature-based compilations and other actual occurrences in crystal structures. The use of ‘fuzzy’ search tolerances has enabled us to discover a number of triple interaction types that have not been previously recorded nor predicted theoretically.
Comparative analyses of different ribosomal RNA structures reveal several conserved base triple motifs in 50S rRNA structures, indicating an important role in structural stabilization and ultimately RNA function.
It is now clear that three-dimensional (3D) structure is as fundamental to the functionality of complex RNA molecules as it is for proteins. Awareness of the extent of the complexity and the diversity of RNA tertiary structures has expanded due to the availability of high resolution structures of large assemblies such as both ribosomal subunits [1–3], ribozymes [4, 5], and the P4-P6 domain of the group I intron , in addition to the early transfer-RNA structures [7, 8]. Base triples can provide key interactions in assembling a tertiary structure by docking a base-pair in a helical region, which may be either Watson-Crick or non-canonical, to a single-stranded nucleotide distant in the polynucleotide chain [9, 10]. In addition, neighboring base triples can form stacks or triple helices: the first such example in the RNA shallow groove was observed in the crystal structure of a frame-shifting pseudoknot . The occurrence of four base triples stacked together has also been observed in the structure of the Tetrahymena ribozyme .
As with base pairs, the key interactions that stabilize base triples are hydrogen bonds. Hydrogen bonds are highly directional and capable of defining specific interactions. Many previous reports discussing base triples consider a wide variety of possible interactions including single hydrogen bond interactions and interactions to non-base components of the nucleotides [11, 13]. However, it is clear that multiply hydrogen bonded triples, which consist of at least two hydrogen bonding interactions per base, will be especially stable and the structural conservation of such triples is therefore likely to be significant. They are thus expected to be influential constituents of RNA structure and are therefore of primary interest in this work. Nevertheless, RNA structures are highly dynamic and even triple interactions may be lost as a result of conformational changes. As the number of complex RNA structures increases, the ability to computationally detect and track such changes presents an interesting challenge. A library of 840 theoretically computed base triples has been compiled and is accessible as part of the NAIL (Nucleic Acid Interaction Library) database . Another resource, the NCIR database , consists of known non-canonical base interactions including base triples and is compiled through a literature search.
Here, we present a survey, that cross-references the patterns in the NAIL database , the Protein Data Bank (PDB)  and the records in NCIR , and is able to deliver a catalogue of base triples of a specific type in existing crystal structures. We use a graph theoretical approach that is capable of documenting occurrences of NAIL patterns in the PDB, and in doing so, find interactions that are predicted by NAIL but not recorded in NCIR. NCIR is a manual literature search and therefore, such a search is limited to what is reported in the available literature in addition to possibly incomplete coverage due to the manual and labor intensive nature of the compilation process. By employing high tolerances designed to give a ‘fuzzy’ search, our method was also able to retrieve previously unrecorded occurrences of triples contained within the PDB. Surprisingly, we also show that there are multiply hydrogen bonded base triples that occur in the PDB, but were not included in the NAIL dataset. Our investigation of these triples in the large ribosomal subunit structures also revealed conserved interactions that may be essential base interactions in the stabilization of rRNA structure.
RNA search database and query pattern library
The search database was compiled from PDB-sourced RNA and RNA-associated structures, solved by X-ray crystallography to a minimum resolution of 3Å. High resolution structures which became available at a later date, were also included as they became available. Several separate searches using structures solved at lower than 3Å resolution, such as the E. coli ribosome at 3.5Å , were also done. A library of 942 pattern matrices was generated as queries for the search engine, consisting of (i) The 840 base triple patterns from the Nucleic Acids Interaction Library  and (ii) 102 patterns generated by an alternative approach (Additional File 1 Figure S1).
Searching program and approach
Sequence alignments and structure superpositions
Source of sequences and structures used for sequence-structure comparisons.
Bacteroides fragilis NCTC 9343
Borrelia burgdorferi B31
Candidatus Protochlamydia amoebophila UWE25
Rhodopirellula baltica SH 1
Thiomicrospira denitrificans ATCC 33889
Acidobacteria bacterium Ellin345
Baumannia cicadellinicola str. Hc
Azoarcus sp. EbN1
Rhodobacter sphaeroides 2.4.1
Anaeromyxobacter dehalogenans 2CP-C
Proteobacteria; delta/epsilon subdivisions
Anabaena variabilis ATCC 29413
Dehalococcoides sp. CBDB1
Bacillus subtilis subsp. subtilis str. 168
Bifidobacterium longum NCC2705
PDB: 2j00 / 2j01
Thermotoga maritima MSB8
Aquifex aeolicus VF5
Sulfolobus solfataricus P2
Pyrobaculum aerophilum str. IM2
Nanoarchaeum equitans Kin4-M
Searching for structural patterns with graph theory
NASSAM , a graph theoretical search engine, was used to screen a nucleic acid database for matches to the 840 patterns derived from the NAIL database and a further 102 from our own procedure (see Additional File 1 Figure S1). The results from this primary screen were then filtered for specified hydrogen bonding interactions. The output from this secondary screen, which represent actual occurrences in the PDB, was used to compare and contrast the content of base triple interactions in the NAIL and NCIR databases. The probable importance of particular base triples as conserved motifs was able to be contextually presented by collectively comparing the structural conservation of the hits from our searches and their sequence conservation in a diverse set of sequences. Comparisons of the annotated triples for each structure were integrated to structure superpositions and sequence alignments information to investigate the conservation of triple hits in the prokaryotic ribosomal subunits (Additional File 1 Table S2).
Comparative overview of triples from the NASSAM search
Triple types found in high resolution RNA Structures that are unrecorded in NCIR
On examination of the new triple types in the 23S rRNA of H. marismortui and T. thermophilus structures, we find that many triples had already been annotated on the secondary structure diagram [1, 2]. However, one of the hits AGU1 (Figure 2A) appears to be completely unrecorded in this structure or any other, although it had been theoretically predicted by NAIL. This new triple is a GU (G924.U919) wobble pair in helix 37 in which the guanine is also participating in an AG N3-amino, amino-N1 interaction to A166 [PDB: 1ffk_0] which is situated in a hairpin loop in helix 11. Thus this triple links bases in domains I and II. Domain II accounts for much of the “back” of the ribosomal particle and thus this base triple is far from the functional sites of 23S rRNA. The A166 and G924 positions are conserved in alignment of 23 prokaryotic species sequences. However, in the sequence of Nanoarchaeum equitans Kin4-M, an archaeal species, the U919 position is replaced with a cytosine. The ACG1 triple (Figure 2A) was found in only in the two structures of the hepatitis delta virus (HDV) ribozyme [25, 26] using a NAIL input pattern which involves a protonated cytosine. Its position between sites for catalysis and protein interaction, may suggest that it contributes to stabilization of the helix backbone necessary for the correct folding of either or both these sites. Another protonated NAIL pattern that finds a match in the database is the ACC1 triple (Figure 2A). This triple, which involves an adenosine protonated at N1, is found in two different locations in the 23S rRNA structures of H. marismortui. One of these, the A2485.C2104.C2536 triple, was found to be conserved in our alignments of 23S rRNA sequences. It is within the peptidyl transferase region and is discussed below. The GGG2 triple (Figure 2A), though not listed in NCIR, has been previously identified in a synthetic RNA-DNA hybrid molecule , but we find no other examples in naturally occurring structures. NCIR is the result of a manual literature search and therefore, by implication, such a search is limited to what is reported in the available literature in addition to possibly incomplete coverage due to the manual and labor intensive nature of such a compilation process. Furthermore, to our knowledge, NCIR is also not automatically updated as newer structures become available.
Previously recorded triple types not predicted in NAIL
Seven triple types were found to be recorded in NCIR, but were not predicted in NAIL. This may be because of limitations already discussed by Walberer et al., namely: the fixed length of the initial hydrogen bond and problems associated with modeling of non-planar geometries. Thus in the AAG1, AAG2 and AGU2 triples (Figure 2B) a completely planar triple geometry could result in close approaches and possible atomic clashes. The ACG2 triple is found in the H. marismortui 23S ribosomal RNA structure and involves a rather non planar association of G2033 to the A1742.C2037 pair. This triple caps a short (4bp) double helix which sweeps up from the end of the double helix to complete the triple. The AGU2 triple (Figure 2B), found in the structure of the T. thermophilus 16S rRNA [PDB: 1fjg_A, 1n32_A], consists of an AU reverse Watson-Crick pair and an AG N7 amino, amino N3 pair in a situation where different edges of the shared adenine participates in different non-canonical pairings to form the triple. This triple (A55.G357.U368), is in fact part of a tetrad, listed in NCIR, as G357 also forms a Watson-Crick base pair with C54. In addition to the inter-base hydrogen bonds there is also a hydrogen bond with good geometry between the ribose O2* of G357 and the O2 acceptor of U368. This close approach may be the reason for its exclusion from the NAIL predictions. Another type of triple not predicted in NAIL is CCG1 which occurs in the 5S subunit (Figure 2B) and consists of a Watson-Crick G66.C15 base pair with a bridging C113 [PDB: 1ffk_9]. This interaction forms a junction between three double helices. The GGU1 triple was found to be present in aspartate-tRNA structures as well as 23S rRNA structures (Figure 2B). In the tRNA structures, this triple caps the anticodon stem, stacking with the A24.U11 base pair which in turn stacks on a group of three triples (U12.A23.A9;Ψ13.G22.A46; A14.A21.U8), the triple being augmented by a hydrogen bond between the O2* of residue 45 and the O3* of residue 9. The other occurrence of GGU1 is in the 23S ribosomal RNA structures [G2092.G2093.U2652, PDB: 1ffk_0] where the two guanosines are sequential and form a platform-like structure. This triple caps a stack of two other triples: G2094.A2649.C2651 and A2095.A2612.U2650.
Completely novel base triple interactions
A further two triple types, CGU1 and GGG1 (Figure 2C) were neither predicted in the NAIL nor listed in the NCIR databases. The CGU1 triple consists of a reverse Watson-Crick GC base pair G515.C548 with a bridging O4 from U519. The triple caps a succession of base triples formed by residues 512-515 and 521 to 523 in T. thermophilus tRNA-Gln in complex with its cognate tRNA synthetase  [PDB: 1g59_B: C512.C509.G523; U513.A546.G542; A514.U508.A521]. The second novel triple type, GGG1 (Figure 2C), was found in two structures of H. marismortui 23S ribosomal subunits [PDB: 1k9m, 1kd1 , 1m90 ] but not in the other H. marismortui high resolution rRNA structures searched. The GGG1 triple type consists of a G.G N3-amino symmetric base pair between G512 and G487. The O6 atom of G512 accepts two hydrogen bonds from N1 and N2 of G504 to complete the triple. This triple can be further extended into a quadruple interaction, which is unlisted in NCIR, via a Watson-Crick pairing between G487 and C515. In some other 23S structures, G512 and G487 are further apart and do not hydrogen bond, although in all the structures, hydrogen bonding occurs from G512 N2 to the O4* in the ribose of G487. The interaction is sandwiched between another triple U488.G503.A513 and a tetrad A485.A509.C505.U481.
Base triples as constituents of interactions between RNA secondary structures
Many of the base triples found contain a Watson-Crick pair. Thus, approximately 55% of the triples found in H. marismortui 23S rRNA, contain a Watson-Crick pair. Triples were observed as interactions involved in RNA helix packing, or at interfaces of RNA secondary or tertiary structure interactions. A listing of interactions between RNA secondary structure elements highlights the frequency of interactions involving at least one loop structure of any type interacting with either another loop or a helix (Additional File 1 Table S2). These interactions outnumber helix-helix and intra helix interactions. We further observed that the same triple types occur in different, larger structural motifs involving a variety of RNA secondary structures. For example, there are six AGC amino-N3, N1-amino; Watson-Crick triples in H. marismortui 23S rRNA, which are variously involved in helix to internal loop interactions, hairpin loop – hairpin loop – hairpin loop interactions, helix to hairpin loop interactions and helix to multi-branched loop interactions (Additional File 1 Table S2). Furthermore, three of these six occurrences are also involved in inter-domain interactions. One of these six AGC triples interfaces domain IV and domain VI and is conserved in the sequence alignments, while the other five show varying degrees of conservation.
Two occurrences of the common AGC amino-N3, N1-amino; Watson-Crick triples were also observed, in T. thermophilus 16S rRNA, one of which is also involved in an inter-domain interaction. Of the two AGC amino-N3, N1-amino; Watson-Crick triples, one involves bases in helices H8 and H14 (G347.C342.A160) and the other bases in helices H12 and H21. The helices for the first triple end in a GAAA (GNRA) tetraloop and a UACG (UNCG) tetraloop, respectively. The interaction between these loops has been previously noted to be an unusual packing arrangement. The second triple is between the second A (A608) of the GAAA loop and the C.G (C308.G292) base pair, which closes the UACG loop. Interestingly, this second example in 16S RNA, the adenine base, although not part of a GNRA tetraloop, is in a GAAAG internal loop sequence where the GAAA has a very GNRA like conformation. The significance of this may purely be that this base in the GNRA conformation is accessible and has both faces available for tertiary binding. This position may be particularly important for the interaction of tetraloop structures with other features.
The prokaryotic ribosomal subunits: occurrences and conservation of base triples
Geometric families for 23S rRNA base triples
The conservation of geometric orientation for a triple position to an extent enables the conservation of 3-D space for a triple and thus preserving the backbone conformation of an RNA molecule despite variations in the nucleotide sequence. An analysis of the interaction geometry for 23S rRNA triples was used to investigate occurrences of repeats for base geometry and the resulting effect such conservation may have on sequence variation while still preserving the general conformation of the sugar-phosphate backbone. 23S rRNA triples from H. marismortui and T. thermophilus 23S rRNA structures were described using the nomenclature proposed by Leontis and Westhof . Triples of equivalent positions in the E. coli and D. radiodurans structures identified via the structural alignment were also described using the same nomenclature. The sequence alignment data was then integrated to this information and the variability of base sequence or content for these triples could be observed.
All the triples observed in this survey contain at least one base which uses its Watson-Crick edge to interact with other bases. The most common triple in 23S rRNA is the AGC triple from the Cis S / WC – Cis WC / WC geometric family (Additional File 1 Table S3). In the 23S rRNA structure, this geometry appears to be exclusive to the AGC triple. However, the most common geometric family in the 23S subunit is the Cis WC / WC – Trans H / WC family which has seven examples. These seven examples occur in triples with three different base compositions, ACU, AUU and CGG (Additional File 1 Table S3). The majority of triples occur as composite base pairs where one base interacts via two of its edges to the other two bases of the triple. However, four cases were observed where two bases are interacting with two edges (ACG, CGG, GGU, GUU - Additional File 1 Table S3). All four of these cases are unique and are not repeated in the 23S rRNA structure.
3D space conservation in triples with unconserved base content
Sequence variation has also been observed at points of inter-domain interactions. These triples are quite possibly the result of opportunistic interactions resulting from the placement of the component nucleotides by the conservation of the sugar-phosphate backbone fold. These opportunistic interactions are therefore expected to vary between different species and therefore the geometric family of the interactions is also not expected to be conserved. One such example is the A2841.C2087.G2657 triple in H. marismortui (Figure 4A) which has equivalently placed bases in the E. coli, T. thermophilus and D. radiodurans structures that however differ in sequence. This is an interaction between domains V and VI where the geometry of interaction for positions equivalent to A2841 (DVI) are not conserved. As previously noted, the formation of triples that are unconserved in sequence can be seen as an opportunistic mechanism for effecting sequence diversity while still preserving the backbone conformation of the structure in general.
A stacked UAU Hoogsteen, Watson-Crick triple motif in prokaryotic 23S rRNA
One interesting highly conserved occurrence in prokaryotic 23S rRNA is the presence of two UAU Hoogsteen, Watson Crick interactions stacked on each other. These triples [U2116.A2118.U2276 and U2115.A2470.U2277; PDB: 1ffk_0] have been previously recorded in NCIR. However, when viewed together the adenines appear stacked in opposite positions to each other (Figure 3C). This same stacked motif was also found in our survey of the E. coli [PDB: 2awb], T. thermophilus [PDB: 2j01] and D. radiodurans [PDB: 1nkw] 23S rRNA structures. The superposition for the formations in all four structures shows good structural conservation. To our knowledge, there have been no previous discussions or hypotheses with regard to the possible functions of this novel stacked UAU Hoogsteen, Watson Crick triples motif. This motif is situated at a junction which joins the three subdomains of domain V (Figure 3C). One of these three subdomains form the binding site for protein L1, another forms the majority of the central protuberance region and the third extends to the direction of domain VI and the putative peptidyl transferase active site . It has been previously observed that the U2115.A2470.U2277 triad is stacked below a UGCAG pentad (U2278.G2471.C2114.A2633.G2630). Our results demonstrate that there are actually two structurally conserved UAU triples situated at a multi-loop junction stacked to the UGCAG pentad. At present, the only other occurrence of this kind of UAU Hoogsteen, Watson Crick triple in a non-23S rRNA structure is a lone triple in the structure of cysteinyl tRNA synthetase [PDB: 1u0b].
Interactions between rRNA domains
More than 80% of the triples in both the large ribosomal subunit and in the 16S rRNA mediate intra-domain interactions. However, inter-domain interactions involving bases very distant to each other in the polynucleotide sequence were also observed. Two interactions that interface three of the six domains and three which interface two domains were observed in the large subunit structure (Figure 3A, 2B). In the first three-domain triple, a G.C Watson-Crick and A.G N3-amino, amino-N1 triple (G2449.C418.A1921) forms an interaction between domains I, IV and V (Additional File 1 Table S2). Domain IV constitutes much of the subunit interface in contact with the 30S particle and helices 67 to 71 form an area around the putative active site cleft on the subunit interface side of the 23S rRNA. The adenine in this triple is in helix 68 and appears to be located in a relatively variable region in our alignment. This is the only example we have observed which shows the interactions of three bases, each on a separate hairpin loop and on different domains. The second triple that interfaces three different domains is U1371.A2054.U2648. U1371 on a multi-branched loop in domain III hydrogen bonds with the A2054 in a multi-branched loop on domain IV, which in turn is hydrogen bonded to U2648 on a small internal loop in domain V . An interesting difference in the structural organization of the two ribosomal subunits is that while the domains of the 16S rRNA each form distinct components, the domains of 23S rRNA are much more intricately linked together, which has been suggested may reflect the lesser requirement for flexibility in this subunit. This is also reflected in our work, where multiply hydrogen bonded triples which are components of inter-domain interactions are more numerous in the 23S rRNA than the 16S rRNA subunit. We did not observe any triples that involved all three domains of the T. thermophilus 16S rRNA structure [PDB: 1fjg], although two triples were observed to interface two different domains. One example, A608.C308.G292, has been previously discussed. The other appears to be a U13.U20 in a hairpin loop interacting with A915, where the Watson-Crick face of U20 forms hydrogen bonds with both U13 and A915.
Base triples and links to ribozymic activity
Domain V of the H. marismortui 23S rRNA has been investigated for links to its ribozymic peptidyl transferase activity . There are 11 interactions that involve base triples in this domain. One cluster of interactions involves the double UAU Hoogsteen, Watson Crick discussed previously. Another set of base triples forms interactions on stems leading or exiting the multi-branched loop (helices 73, 89, 90) where the A-loop  is located. These include a triple [C2104.A2485.C2536; PDB: 1ffk]  that was also found to be structurally conserved in the structural alignments and is also conserved in all the sequences aligned. This ACC triple is adjacent to A2486, which was originally hypothesized as the peptidyl transferase acid-base catalytic residue by Nissen et al., although subsequent work has shown that peptide bond formation in the peptidyl transferase centre does not involve this acid-base catalysis mechanism[37–40]. The A2450.C2501 component of this triple (E. coli numbering used), is also involved in an A-minor interaction with A76 of P-site tRNA.
Triples in the vicinity of other ribozyme active sites have previously been observed; an example being the base triple sandwich in the structure of the Tetrahymena ribozyme, where the 3’-terminal guanosine (ωG), which serves as the attacking group for RNA cleavage, participates in a triple interaction with the G264.C311 base pair. This triple fits the multiply hydrogen bonded criteria of this survey and is in turn sandwiched by three other base triples. Of the 3 triples in the sandwich, the A263.C262.G312 interaction above the ωG triple, is also of the type covered in this survey. Both triples were found by a NASSAM search although they were not in the original search due to the resolution of the Tetrahymena ribozyme structure [3.8Å, PDB: 1x8w] being lower than our cut-off point. The inter-domain interactions and helix stabilization functions carried out by base triples may be important, albeit collective factors, in determining the correct folding of the RNA molecule to enable its catalytic functions.
The ribosomal polypeptide exit tunnel
Possible exclusivity of interactions to particular structures
Several types of triples appear to occur more than once within one structure, but are not presently seen in any other class of RNA structure. For example, the new ACC1 triple (Figure 2A) occurs at two different positions in the large ribosomal subunit but does not occur in any other RNA structure used in our search. The two CCG triples found, C37.G43.C46 and C15.G66.C113 [PDB: 1ffk_9], were found only in the 5S rRNA of the large ribosomal subunit. Several of the interactions found have only been observed at the same position in a particular structure, and do not occur anywhere else in that structure or any other structure. Moreover, the ribosomal structures are extremely large compared to the other RNA structures and are likely to contain a wider diversity of triples, and therefore these observations, although inconclusive, are noted for the record.
Future extensions to the NASSAM program - roles of RNA backbone components, water molecules and metal ions
At present the methodology is restricted to base-base interactions. Although this has revealed new types of interactions, it is clear that base to backbone interactions are also of great importance. While inclusion of backbone information will pose some challenges for the graph theoretical methodology as it greatly increases the number of nodes that can potentially be matched, this is nevertheless an important future enhancement that can be added. Another important extension that can be envisaged is the addition of information relating to water molecules and metal ions (especially Mg2+ ions) which play important roles in the stabilization of interbase interactions. Here there is another problem in that it is extremely difficult to accurately define waters and low atomic number metal ions (such as Mg2+) in crystal structures unless they are at least 2.0Å in resolution; very few RNA structures match this criterion. This problem has been discussed, for example, by Banatao et al (2003) , who observed that many metal sites are mislabelled or completely missing in RNA structures. Nevertheless, with ribosomal structures (such as PDB: 1vqs) already at 2.2 Å, this kind of analysis may soon be an accessible goal, and would clearly greatly increase the number of possible interaction patterns beyond those discussed here.
The results of our survey reveal that multiply hydrogen bonded base triples have a high degree of conservation between comparable ribosomal structures, therefore suggesting significant collective contributions towards the overall folding and stabilization of the RNA molecule. By annotating triple patterns and correlating theoretically predicted base triples from NAIL with the literature-based compilations in the NCIR database, we were able to discover a number of multiply hydrogen bonded base triple formations that had not been previously recorded and/or had not been predicted theoretically. The same annotation approach has enabled motif discovery by putting into context known patterns of interactions such as the stacked UAU Hoogsteen, Watson Crick motif. The value of expertly curated databases such as NCIR  and SCOR  are indisputable. Computationally generated libraries such as NAIL have proven to be an important resource towards the discovery of new base interactions. Computational structure annotation methods, which can quickly locate tertiary interactions of a given type, large or small, can be invaluable for the comparison of complex structures as well as for increasing the coverage and volume for manual curation. This capability provides the foundation for further experimental work in investigating the specific contributions from possibly essential base interactions which can in turn correlate RNA tertiary structure to their corresponding function. As an outcome of this work, we have made available the NASSAM program via a web enabled interface at http://mfrlab.org/grafss/nassam/.
Acknowledgements and funding
We thank the Biotechnology and Biological Sciences Research Council, the Royal Society and the Wolfson Foundation for funding and computing facilities. MFR was funded by the Ministry of Higher Education Malaysia FRGS grant UKM-ST-06-FRGS0006-2009 and the Universiti Kebangsaan Malaysia grant UKM-GGPM-KPB-101-2010. The authors gratefully acknowledge assistance from Ms. Hazrina Yusof Hamdani for the setting up of the web server for NASSAM.
This article has been published as part of BMC Bioinformatics Volume 12 Supplement 13, 2011: Tenth International Conference on Bioinformatics – First ISCB Asia Joint Conference 2011 (InCoB/ISCB-Asia 2011): Bioinformatics. The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2105/12?issue=S13.
- Ban N, Nissen P, Hansen J, Moore PB, Steitz TA: The complete atomic structure of the large ribosomal subunit at 2.4 Angstrom resolution. Science 2000, 289: 905–920. 10.1126/science.289.5481.905View ArticlePubMedGoogle Scholar
- Wimberly BT, Brodersen DE, Clemons WM, Morgan-Warren RJ, Carter AP, Vonrhein C, Hartsch T, Ramakrishnan V: Structure of the 30S ribosomal subunit. Nature 2000, 407: 327–339. 10.1038/35030006View ArticlePubMedGoogle Scholar
- Schluenzen F, Tocilj A, Zarivach R, Harms J, Gluehmann M, Janell D, Bashan A, Bartels H, Agmon I, Franceschi F, Yonath A: Structure of functionally activated small ribosomal subunit at 3.3 angstrom resolution. Cell 2000, 102: 615–623. 10.1016/S0092-8674(00)00084-2View ArticlePubMedGoogle Scholar
- Pley HW, Flaherty KM, McKay DB: 3-dimensional structure of a hammerhead ribozyme. Nature 1994, 372: 68–74. 10.1038/372068a0View ArticlePubMedGoogle Scholar
- Scott WG, Finch JT, Klug A: The crystal-structure of an all-RNA Hammerhead ribozyme - a proposed mechanism for RNA catalytic cleavage. Cell 1995, 81: 991–1002. 10.1016/S0092-8674(05)80004-2View ArticlePubMedGoogle Scholar
- Cate JH, Doudna JA: Metal-binding sites in the major groove of a large ribozyme domain. Structure 1996, 4: 1221–1229. 10.1016/S0969-2126(96)00129-3View ArticlePubMedGoogle Scholar
- Kim SH, Suddath FL, Quigley GJ, McPherso A, Sussman JL, Wang AHJ, Seeman NC, Rich A: 3-dimensional tertiary structure of yeast phenylalanine transfer-RNA. Science 1974, 185: 435–440. 10.1126/science.185.4149.435View ArticlePubMedGoogle Scholar
- Ladner JE, Jack A, Robertus JD, Brown RS, Rhodes D, Clark BFC, Klug A: Structure of yeast phenylalanine transfer-RNA at 2.5 A resolution. Proc Natl Acad Sci U S A 1975, 72: 4414–4418. 10.1073/pnas.72.11.4414PubMed CentralView ArticlePubMedGoogle Scholar
- Gautheret D, Major F, Cedergren R: Modeling the 3-dimensional structure of RNA using discrete nucleotide conformational sets. J Mol Biol 1993, 229: 1049–1064. 10.1006/jmbi.1993.1104View ArticlePubMedGoogle Scholar
- Moore PB: Structural motifs in RNA. Annu Rev Biochem 1999, 68: 287–300. 10.1146/annurev.biochem.68.1.287View ArticlePubMedGoogle Scholar
- Su L, Chen LQ, Egli M, Berger JM, Rich A: Minor groove RNA triplex in the crystal structure of a ribosomal frameshifting viral pseudoknot. Nat Struct Biol 1999, 6: 285–292. 10.1038/6722View ArticlePubMedGoogle Scholar
- Guo F, Gooding AR, Cech TR: Structure of the tetrahymena ribozyme: base triple sandwich and metal ion at the active site. Mol Cell 2004, 16: 351–362.PubMedGoogle Scholar
- Nagaswamy U, Larios-Sanz M, Hury J, Collins S, Zhang ZD, Zhao Q, Fox GE: NCIR: a database of non-canonical interactions in known RNA structures. Nucleic Acids Res 2002, 30: 395–397. 10.1093/nar/30.1.395PubMed CentralView ArticlePubMedGoogle Scholar
- Walberer BJ, Cheng AC, Frankel AD: Structural diversity and isomorphism of hydrogen-bonded base interactions in nucleic acids. J Mol Biol 2003, 327: 767–780. 10.1016/S0022-2836(03)00090-1View ArticlePubMedGoogle Scholar
- Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE: The Protein Data Bank. Nucleic Acids Res 2000, 28: 235–242. 10.1093/nar/28.1.235PubMed CentralView ArticlePubMedGoogle Scholar
- Schuwirth BS, Borovinskaya MA, Hau CW, Zhang W, Vila-Sanjurjo A, Holton JM, Cate JHD: Structures of the bacterial ribosome at 3.5 Angstrom resolution. Science 2005, 310: 827–834. 10.1126/science.1117230View ArticlePubMedGoogle Scholar
- Harrison AM, South DR, Willett P, Artymiuk PJ: Representation, searching and discovery of patterns of bases in complex RNA structures. J Comput-Aided Mol Des 2003, 17: 537–549.View ArticlePubMedGoogle Scholar
- Ullmann JR: Algorithm for subgraph isomorphism. J Acm 1976, 23: 31–42. 10.1145/321921.321925View ArticleGoogle Scholar
- McDonald IK, Thornton JM: Satisfying hydrogen-bonding potential in proteins. J Mol Biol 1994, 238: 777–793. 10.1006/jmbi.1994.1334View ArticlePubMedGoogle Scholar
- Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD: Multiple sequence alignment with the CLUSTAL series of programs. Nucleic Acids Res 2003, 31: 3497–3500. 10.1093/nar/gkg500PubMed CentralView ArticlePubMedGoogle Scholar
- Kabsch W: Solution for best rotation to relate 2 sets of vectors. Acta Crystallogr Sect A 1976, 32: 922–923. 10.1107/S0567739476001873View ArticleGoogle Scholar
- Bailey S: The CCP4 Suite - programs for protein crystallography. Acta Crystallogr Sect D-Biol Crystallogr 1994, 50: 760–763. 10.1107/S0907444994003112View ArticleGoogle Scholar
- Leontis NB, Westhof E: Geometric nomenclature and classification of RNA base pairs. RNA-Publ RNA Soc 2001, 7: 499–512.View ArticleGoogle Scholar
- Duarte CM, Pyle AM: Stepping through an RNA structure: a novel approach to conformational analysis. J Mol Biol 1998, 284: 1465–1478. 10.1006/jmbi.1998.2233View ArticlePubMedGoogle Scholar
- Ferre-D'Amare AR, Zhou K, Doudna JA: Crystal structure of a hepatitis delta virus ribozyme. Nature 1998, 395: 567–574. 10.1038/26912View ArticlePubMedGoogle Scholar
- Ke AL, Zhou KH, Ding F, Cate JHD, Doudna JA: A conformational switch controls hepatitis delta virus ribozyme catalysis. Nature 2004, 429: 201–205. 10.1038/nature02522View ArticlePubMedGoogle Scholar
- Ravelli RBG, Leiros HKS, Pan BC, Caffrey M, McSweeney S: Specific radiation damage can be used to solve macromolecular crystal structures. Structure 2003, 11: 217–224. 10.1016/S0969-2126(03)00006-6View ArticlePubMedGoogle Scholar
- Sekine S, Nureki O, Shimada A, Vassylyev DG, Yokoyama S: Structural basis for anticodon recognition by discriminating glutamyl-tRNA synthetase. Nat Struct Biol 2001, 8: 203–206. 10.1038/84927View ArticlePubMedGoogle Scholar
- Hansen JL, Ippolito JA, Ban N, Nissen P, Moore PB, Steitz TA: The structures of four macrolide antibiotics bound to the large ribosomal subunit. Mol Cell 2002, 10: 117–128. 10.1016/S1097-2765(02)00570-1View ArticlePubMedGoogle Scholar
- Hansen JL, Schmeing TM, Moore PB, Steitz TA: Structural insights into peptide bond formation. Proc Natl Acad Sci U S A 2002, 99: 11670–11675. 10.1073/pnas.172404099PubMed CentralView ArticlePubMedGoogle Scholar
- Nissen P, Ippolito JA, Ban N, Moore PB, Steitz TA: RNA tertiary interactions in the large ribosomal subunit: The A-minor motif. Proc Natl Acad Sci U S A 2001, 98: 4899–4903. 10.1073/pnas.081082398PubMed CentralView ArticlePubMedGoogle Scholar
- Woese CR, Fox GE, Zablen L, Uchida T, Bonen L, Pechman K, Lewis BJ, Stahl D: Conservation of primary structure in 16S ribosomal-RNA. Nature 1975, 254: 83–86. 10.1038/254083a0View ArticlePubMedGoogle Scholar
- Lu XJ, Olson WK: 3DNA: a software package for the analysis, rebuilding and visualization of three-dimensional nucleic acid structures. Nucleic Acids Res 2003, 31: 5108–5121. 10.1093/nar/gkg680PubMed CentralView ArticlePubMedGoogle Scholar
- Hauenstein S, Zhang CM, Hou YM, Perona JJ: Shape-selective RNA recognition by cysteinyl-tRNA synthetase. Nat Struct Mol Biol 2004, 11: 1134–1141. 10.1038/nsmb849View ArticlePubMedGoogle Scholar
- Nissen P, Hansen J, Ban N, Moore PB, Steitz TA: The structural basis of ribosome activity in peptide bond synthesis. Science 2000, 289: 920–930. 10.1126/science.289.5481.920View ArticlePubMedGoogle Scholar
- Green R, Switzer C, Noller HF: Ribosome-catalyzed peptide-bond formation with an a-site substrate covalently linked to 23S ribosomal RNA. Science 1998, 280: 286–289. 10.1126/science.280.5361.286View ArticlePubMedGoogle Scholar
- Beringer M, Bruell C, Xiong LQ, Pfister P, Bieling P, Katunin VI, Mankin AS, Bottger EC, Rodnina MV: Essential mechanisms in the catalysis of peptide bond formation on the ribosome. J Biol Chem 2005, 280: 36065–36072. 10.1074/jbc.M507961200View ArticlePubMedGoogle Scholar
- Bieling P, Beringer M, Adio S, Rodnina MV: Peptide bond formation does not involve acid-base catalysis by ribosomal residues. Nat Struct Mol Biol 2006, 13: 423–428. 10.1038/nsmb1091View ArticlePubMedGoogle Scholar
- Schmeing TM, Huang KS, Kitchen DE, Strobel SA, Steitz TA: Structural insights into the roles of water and the 2 ' hydroxyl of the P site tRNA in the peptidyl transferase reaction. Mol Cell 2005, 20: 437–448. 10.1016/j.molcel.2005.09.006View ArticlePubMedGoogle Scholar
- Schmeing TM, Huang KS, Strobel SA, Steitz TA: An induced-fit mechanism to promote peptide bond formation and exclude hydrolysis of peptidyl-tRNA. Nature 2005, 438: 520–524. 10.1038/nature04152View ArticlePubMedGoogle Scholar
- Polacek N, Mankin AS: The ribosomal peptidyl transferase center: Structure, function, evolution, inhibition. Crit Rev Biochem Mol Biol 2005, 40: 285–311. 10.1080/10409230500326334View ArticlePubMedGoogle Scholar
- Moore PB, Steitz TA: The structural basis of large ribosomal subunit function. Annu Rev Biochem 2003, 72: 813–850. 10.1146/annurev.biochem.72.110601.135450View ArticlePubMedGoogle Scholar
- Xin Y, Olson WK: BPS: a database of RNA base-pair structures. Nucl Acids Res 2009, 37: D83–88. 10.1093/nar/gkn676PubMed CentralView ArticlePubMedGoogle Scholar
- Banatao DR, Altman RB, Klein TE: Microenvironment analysis and identification of magnesium binding sites in RNA. Nucl Acids Res 2003, 31: 4450–4460. 10.1093/nar/gkg471PubMed CentralView ArticlePubMedGoogle Scholar
- Klosterman PS, Tamura M, Holbrook SR, Brenner SE: SCOR: a structural classification of RNA database. Nucleic Acids Res 2002, 30: 392–394. 10.1093/nar/30.1.392PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.