Direct RNA motif definition and identification from multiple sequence alignments using secondary structure profiles.

We present here a new approach to the problem of defining RNA signatures and finding their occurrences in sequence databases. The proposed method is based on "secondary structure profiles". An RNA sequence alignment with secondary structure information is used as an input. Two types of weight matrices/profiles are constructed from this alignment: single strands are represented by a classical lod-scores profile while helical regions are represented by an extended "helical profile" comprising 16 lod-scores per position, one for each of the 16 possible base-pairs. Database searches are then conducted using a simultaneous search for helical profiles and dynamic programming alignment of single strand profiles. The algorithm has been implemented into a new software, ERPIN, that performs both profile construction and database search. Applications are presented for several RNA motifs. The automated use of sequence information in both single-stranded and helical regions yields better sensitivity/specificity ratios than descriptor-based programs. Furthermore, since the translation of alignments into profiles is straightforward with ERPIN, iterative searches can easily be conducted to enrich collections of homologous RNAs.

RNAMotif, an RNA secondary structure definition and search algorithm.

RNA molecules fold into characteristic secondary and tertiary structures that account for their diverse functional activities. Many of these RNA structures are assembled from a collection of RNA structural motifs. These basic building blocks are used repeatedly, and in various combinations, to form different RNA types and define their unique structural and functional properties. Identification of recurring RNA structural motifs will therefore enhance our understanding of RNA structure and help associate elements of RNA structure with functional and regulatory elements. Our goal was to develop a computer program that can describe an RNA structural element of any complexity and then search any nucleotide sequence database, including the complete prokaryotic and eukaryotic genomes, for these structural elements. Here we describe in detail a new computational motif search algorithm, RNAMotif, and demonstrate its utility with some motif search examples. RNAMotif differs from other motif search tools in two important aspects: first, the structure definition language is more flexible and can specify any type of base-base interaction; second, RNAMotif provides a user controlled scoring section that can be used to add capabilities that patterns alone cannot provide.

Identification of alternate polyadenylation sites and analysis of their tissue distribution using EST data.

Alternate polyadenylation affects a large fraction of higher eucaryote mRNAs, producing mature transcripts with 3' ends of variable length. This variation is poorly represented in the current transcript catalogs derived from whole genome sequences, mostly because such posttranscriptional events are not detectable directly at the DNA level. Alternate polyadenylation of an mRNA is better understood by comparison to EST databases. Comparing ESTs to mRNAs, however, is a difficult task subjected to the pitfalls of internal priming, presence of intron sequences, repeated elements, chimerical ESTs or matches with EST from paralogous genes. We present here a computer program that addresses these problems and displays ESTs matches to a query mRNA sequence to predict alternate polyadenylation and to suggest library-specific forms. The output highlights effective polyadenylation signals, possible sources of artifacts such as A-rich stretches in the mRNA sequences, and allows for a direct visualization of EST libraries using color codes. Statistical biases in the distribution of alternative mRNA forms among EST libraries were systematically sought. About 1450 human and 200 mouse mRNAs displayed such biases, suggesting in each case a tissue- or disease-specific regulation of polyadenylation.

Predicting U-turns in ribosomal RNA with comparative sequence analysis.

The U-turn is a well-known RNA motif characterized by a sharp reversal of the RNA backbone following a single-stranded uridine base. In experimentally determined U-turn motifs, the nucleotides 3' to the turn are frequently involved in tertiary interactions, rendering this motif particularly attractive in RNA modeling and functional studies. The U-turn signature is composed of an UNR sequence pattern flanked by a Y:Y, Y:A (Y=pyrimidine) or G:A base juxtaposition. We have identified 33 potential UNR-type U-turns and 25 related GNRA-type U-turns in a large set of aligned 16 S and 23 S rRNA sequences. U-turn candidates occur in hairpin loops (34 times) as well as in internal and multi-stem loops (24 times). These are classified into ten families based on loop type, sequence pattern (UNR or GNRA) and the nature of the closing base juxtaposition. In 13 cases, the bases on the 3' side of the turn, or on the immediate 5' side, are involved in tertiary covariations, making these sites strong candidates for tertiary interactions.

Patterns of variant polyadenylation signal usage in human genes.

The formation of mature mRNAs in vertebrates involves the cleavage and polyadenylation of the pre-mRNA, 10-30 nt downstream of an AAUAAA or AUUAAA signal sequence. The extensive cDNA data now available shows that these hexamers are not strictly conserved. In order to identify variant polyadenylation signals on a large scale, we compared over 8700 human 3' untranslated sequences to 157,775 polyadenylated expressed sequence tags (ESTs), used as markers of actual mRNA 3' ends. About 5600 EST-supported putative mRNA 3' ends were collected and analyzed for significant hexameric sequences. Known polyadenylation signals were found in only 73% of the 3' fragments. Ten single-base variants of the AAUAAA sequence were identified with a highly significant occurrence rate, potentially representing 14.9% of the actual polyadenylation signals. Of the mRNAs, 28.6% displayed two or more polyadenylation sites. In these mRNAs, the poly(A) sites proximal to the coding sequence tend to use variant signals more often, while the 3'-most site tends to use a canonical signal. The average number of ESTs associated with each signal type suggests that variant signals (including the common AUUAAA) are processed less efficiently than the canonical signal and could therefore be selected for regulatory purposes. However, the position of the site in the untranslated region may also play a role in polyadenylation rate.

Determination of the rpoB gene sequences of Bartonella henselae and Bartonella quintana for phylogenic analysis.

Using the Genome Walker procedure, which allows PCR amplification of genomic DNA using a single gene-specific primer and direct automated sequencing methodology, we obtained the nucleotide sequence of the RNA polymerase beta subunit (rpoB) from Bartonella henselae and Bartonella quintana. A phylogenetic tree constructed from these data and other rpoB sequences available in GenBank is, in part, consistent with those previously derived from 16S rRNA gene sequences and confirms the position of Bartonella within the alpha subdivision of Proteobacteria. In fact, this analysis showed that rpoB data are similar to 16S rRNA data for the alpha, beta and gamma subdivisions of Proteobacteria. In contrast, concerning other bacteria included in our study, the topologies of phylogenetic trees were different. Based on the bootstrap values derived from rpoB phylogenic analysis, we believe that this molecule should contribute to better understanding the evolutionary process.

Novel selenoproteins identified in silico and in vivo by using a conserved RNA structural motif.

Selenocysteine is incorporated into selenoproteins by an in-frame UGA codon whose readthrough requires the selenocysteine insertion sequence (SECIS), a conserved hairpin in the 3'-untranslated region of eukaryotic selenoprotein mRNAs. To identify new selenoproteins, we developed a strategy that obviates the need for prior amino acid sequence information. A computational screen was used to scan nucleotide sequence data bases for sequences presenting a potential SECIS secondary structure. The computer-selected hairpins were then assayed in vivo for their functional capacities, and the cDNAs corresponding to the SECIS winners were identified. Four of them encoded novel selenoproteins as confirmed by in vivo experiments. Among these, SelZf1 and SelZf2 share a common domain with mitochondrial thioredoxin reductase-2. The three proteins, however, possess distinct N-terminal domains. We found that another protein, SelX, displays sequence similarity to a protein involved in bacterial pilus formation. For the first time, four novel selenoproteins were discovered based on a computational screen for the RNA hairpin directing selenocysteine incorporation.

Gene structure, expression pattern, and biological activity of mouse killer cell activating receptor-associated protein (KARAP)/DAP-12.

Natural killer cell and T cell subsets express at their cell surface a repertoire of receptors for MHC class I molecules, the natural killer cell receptors (NKRs). NKRs are characterized by the existence of inhibitory and activating isoforms, which are encoded by highly homologous but separate genes present in the same locus. Inhibitory isoforms express an intracytoplasmic immunoreceptor tyrosine-based inhibition motif, whereas activating isoforms lack any immunoreceptor tyrosine-based inhibition motif but harbor a charged amino acid residue in their transmembrane domain. We previously characterized KARAP (killer cell activating receptor-associated protein), a novel disulfide-linked tyrosine-phosphorylated dimer that selectively associates with the activating NKR isoforms. We report here the identification of the mouse KARAP gene, its localization on chromosome 7 and its genomic organization in five exons. Point mutation and transfection studies revealed that KARAP is a novel signaling transmembrane subunit whose transduction function depends on the integrity of an intracytoplasmic immunoreceptor tyrosine-based activation motif. In contrast to previous members of the immunoreceptor tyrosine-based activation motif polypeptide family, KARAP is ubiquitously expressed on hematopoietic and nonhematopoietic cells, suggesting its association with a broad range of activating receptors in a variety of tissues.

Quantitative analysis of the T cell repertoire selected by a single peptide-major histocompatibility complex.

The positive selection of CD4+ T cells requires the expression of major histocompatibility complex (MHC) class II molecules in the thymus, but the role of self-peptides complexed to class II molecules is still a matter of debate. Recently, it was observed that transgenic mice expressing a single peptide-MHC class II complex positively select significant numbers of diverse CD4+ T cells in the thymus. However, the number of selected T cell specificities has not been evaluated so far. Here, we have sequenced 700 junctional complementarity determining regions 3 (CDR3) from T cell receptors (TCRs) carrying Vbeta11-Jbeta1.1 or Vbeta12-Jbeta1.1 rearrangements. We found that a single peptide-MHC class II complex positively selects at least 10(5) different Vbeta rearrangements. Our data yield a first evaluation of the size of the T cell repertoire. In addition, they provide evidence that the single Ealpha52-68-I-Ab complex skews the amino acid frequency in the TCR CDR3 loop of positively selected T cells. A detailed analysis of CDR3 sequences indicates that a fraction of the beta chain repertoire bears the imprint of the selecting self-peptide.

Alternate polyadenylation in human mRNAs: a large-scale analysis by EST clustering.

Alternate polyadenylation is an important post-transcriptional regulatory process now open to large-scale analysis by use of cDNA databases. We clustered 164,000 expressed sequence tags (ESTs) into approximately 15,000 groups and aligned each group to a putative mRNA 3' end. By use of stringent criteria to discard artifactual mRNA extremities, clear evidence for alternate polyadenylation was obtained in 189 of the 1000 EST clusters studied. A number of previously unreported polyadenylation sites were identified, together with possible instances of tissue-specific differential polyadenylation. This study demonstrates that, besides quantitative aspects of gene expression, the distribution of alternate mRNA forms can be analyzed through EST sampling.

No tRNA3Lys unwinding in a complex with HIV NCp7.

The nucleocapsid protein NCp7 of the human immunodeficiency virus (HIV) type 1 is important for the annealing of HIV RNA and tRNA3Lys, the tRNA acting as a primer during reverse transcription of HIV RNA. A wild type NCp7 and a Cys23 mutant having a disrupted zinc finger were analyzed with far UV circular dichroism (CD). CD data analysis revealed that NCp7 has a high content of extended structures in aqueous buffer, decreasing in Cys23 NCp7 and in NCp7 in the absence of zinc. An increase in beta-turn structures is observed in NCp7 bound to tRNA3Lys. Furthermore, CD data shows that Cys23 NCp7 binds tRNA3Lys. The CD spectrum of tRNA3Lys is typical of an A-form helix and retains this structure after binding of NCp7, which demonstrates that NCp7 does not induce tRNA3Lys unwinding. CD spectra of tRNA3Lys were measured from 5 to 80 degrees C to observe CD changes resulting from tRNA3Lys melting. Molecular modeling of the complex identifies two potential tRNA anticodon binding sites in the NCp7 N-terminal region and first zinc finger. In this model, both binding sites can interact with 12 nucleotides in the anticodon domain without requiring a base specificity.

Inferring the conformation of RNA base pairs and triples from patterns of sequence variation.

The success of comparative analysis in resolving RNA secondary structure and numerous tertiary interactions relies on the presence of base covariations. Although the majority of base covariations in aligned sequences is associated to Watson-Crick base pairs, many involve non-canonical or restricted base pair exchanges (e.g. only G:C/A:U), reflecting more specific structural constraints. We have developed a computer program that determines potential base pairing conformations for a given set of paired nucleotides in a sequence alignment. This program (ISOPAIR) assumes that the base pair conformation is maintained through sequence variation without significantly affecting the path of the sugar-phosphate backbone. ISOPAIR identifies such 'isomorphic' structures for any set of input base pair or base triple sequences. The program was applied to base pairs and triples with known structures and sequence exchanges. In several instances, isomorphic structures were correctly identified with ISOPAIR. Thus, ISOPAIR is useful when assessing non-canonical base pair conformations in comparative analysis. ISOPAIR applications are limited to those cases where unusual base pair exchanges indeed reflect a non-canonical conformation.

G.U base pairing motifs in ribosomal RNA.

An increasing number of recognition mechanisms in RNA are found to involve G.U base pairs. In order to detect new functional sites of this type, we exhaustively analyzed the sequence alignments and secondary structures of eubacterial and chloroplast 16S and 23S rRNA, seeking positions with high levels of G.U pairs. Approximately 120 such sites were identified and classified according to their secondary structure and sequence environment. Overall biases in the distribution of G.U pairs are consistent with previously proposed structural rules: the side of the wobble pair that is subject to a loss of stacking is preferentially exposed to a secondary structure loop, where stacking is not as essential as in helical regions. However, multiple sites violate these rules and display highly conserved G.U pairs in orientations that could cause severe stacking problems. In addition, three motifs displaying a conserved G.U pair in a specific sequence/structure environment occur at an unusually high frequency. These motifs, of which two had not been reported before, involve sequences 5'UG3' 3'GA5' and 5'UG3' 3'GU5', as well as G.U pairs flanked by a bulge loop 3' of U. The possible structures and functions of these recurrent motifs are discussed.

Identification of base-triples in RNA using comparative sequence analysis.

Comparative sequence analysis has proven to be a very efficient tool for the determination of RNA secondary structure and certain tertiary interactions. However, base-triples, an important RNA structural element, cannot be predicted accurately from sequence data. We show here that the poor base correlations observed at base-triple positions are the result of two factors. (1) Base covariation is not as strictly required in triples as it is in Watson-Crick pairs. (2) Base-triple structures are less conserved among homologous molecules. A particularity of known triple-helical regions is the presence of multiple base correlations that do not reflect direct pairing. We suggest that natural mutations in base-triples create structural changes that require compensatory mutations in adjacent base-pairs and triples to maintain the triple-helix conformation. On the basis of these observations, we devised two new measures of association that significantly enhance the base-triple signal in correlation studies. We evaluated correlations between base-pairs and single stranded bases, and correlations between adjacent base-pairs. Positions that score well in both analyses are the best triple candidates. This procedure correctly identifies triples, or interactions very close to the proposed triples, in type I and type II tRNAs and in the group I intron.

A major family of motifs involving G.A mismatches in ribosomal RNA.

G.A oppositions and their flanking nucleotides in the internal loops of 16 S and 23 S rRNA were analyzed from a comparative structure perspective, resulting in an unexpectedly high incidence of the sequence motifs [formula: see text], [formula: see text] and [formula: see text]. The first two motifs can form similar three-dimensional structures containing sheared G.A or A.A pair conformations. Comparative sequence analysis revealed numerous sites in ribosomal RNAs with distinct combinations of nucleotides capable of forming this specific structure. In some cases, the sequence variations provide strong evidence for the sheared tandem structure occurring. Interestingly, the sequence changes would maintain a similar exposure of two adenines in the minor groove, suggesting the possibility that they are serving as a recognition or anchoring unit. These tandem conformations are related to that of the [formula: see text] tandem observed in 5 S and 28 S rRNA.

Fitting the structurally diverse animal mitochondrial tRNAs(Ser) to common three-dimensional constraints.

We propose three-dimensional models for animal mitochondrial (amt) tRNAs lacking the D-domain based on consideration of universal constraints on tRNA to maintain functionality. The available tRNA sequences are classified into two groups, and distinct models are proposed for both classes derived from common structural features. The distance between the anticodon and the acceptor stem is comparable in the models and corresponds to that observed in conventional tRNAs. This fact averts the problem of how a shorter mitochondrial tRNA could function within the context of a protein synthesis machinery suited to full-sized tRNAs. In the models, the angle which defines the relationship between the helical domains composed of the acceptor/T-stem and the anticodon/D-stem is greater than in conventional tRNAs. These structures resemble more a "boomerang" than an "L". However, even in the boomerang model, the inner surface of tRNA would be sufficiently uncluttered to avoid steric clashes when two tRNA molecules cohabit the ribosome.

Reproducing the three-dimensional structure of a tRNA molecule from structural constraints.

The three-dimensional structure of yeast tRNA(Phe) was reproduced at atomic resolution with the automated RNA modeling program MC-SYM, which is based on a constraint-satisfaction algorithm. Structural constraints used in the modeling were derived from the secondary structure, four tertiary base pairs, and other information available prior to the determination of the x-ray crystal structure of the tRNA. The program generated 26 solutions (models), all of which had the familiar "L" form of tRNA and root-mean-square deviations from the crystal structure in the range of 3.1-3.8 A. The interaction between uridine-8 and adenosine-14 was crucial in the modeling procedure, since only this among the tertiary pairs is necessary and sufficient to reproduce the L form of tRNA. Other tertiary interactions were critical in reducing the number of solutions proposed by the program.

Modeling the three-dimensional structure of RNA using discrete nucleotide conformational sets.

The flexibility about seven torsion angles in nucleotides constitutes a severe obstacle to computer modeling of RNA. The computational feasibility of RNA conformational searches can be enhanced by assigning to each nucleotide a set of discrete conformations. In this work, four types of discrete conformational sets for the atomic representation of nucleotide structures were defined and evaluated. These sets, comprising between 10 and 30 conformations, were tested for their ability to reproduce known RNA structures and to generate structures responding to new specifications. Conformational searches were performed with the MC-SYM program, which allows for the generation of all structures satisfying a predetermined set of three-dimensional constraints in a given discrete space. Results with known hairpin loop structures show that root-mean-square deviations of about 1.5 A for backbone atoms and about 2.0 A for all atoms between the modeled and X-ray crystal structures can be expected. The conformational set that gives the most faithful representation of test structures is based on the classification of nucleotide conformations derived from a structural database. Representative conformations are selected from each class that adequately sample variations in backbone direction, sugar pucker and base orientation. With this conformational set, most of the important features of test hairpin structures are reproduced with fidelity, indicating that biologically useful models can be constructed from the combination of discrete nucleotide conformations and an algorithm that rapidly and systematically scans the pre-defined conformational space.