RNA structure prediction from evolutionary patterns of nucleotide composition

Structural elements in RNA molecules have a distinct nucleotide composition, which changes gradually over evolutionary time. We discovered certain features of these compositional patterns that are shared between all RNA families. Based on this information, we developed a structure prediction method that evaluates candidate structures for a set of homologous RNAs on their ability to reproduce the patterns exhibited by biological structures. The method is named SPuNC for ‘Structure Prediction using Nucleotide Composition’. In a performance test on a diverse set of RNA families we demonstrate that the SPuNC algorithm succeeds in selecting the most realistic structures in an ensemble. The average accuracy of top-scoring structures is significantly higher than the average accuracy of all ensemble members (improvements of more than 20% observed). In addition, a consensus structure that includes the most reliable base pairs gleaned from a set of top-scoring structures is generally more accurate than a consensus derived from the full structural ensemble. Our method achieves better accuracy than existing methods on several RNA families, including novel riboswitches and ribozymes. The results clearly show that nucleotide composition can be used to reveal the quality of RNA structures and thus the presented technique should be added to the set of prediction tools.     

Different base per unit length ratios exist in single-stranded RNA and single-stranded DNA.

A significant difference was found to exist in the number of bases per unit length of single-stranded RNA as compared to single-stranded DNA when single-stranded RNA or DNA molecules of known nucleotide sequence were measured by electron microscopy using a cytochrome spreading technique. Using this technique, single-stranded RNA was found to have 17.5% more bases per unit of length than single-stranded DNA. These ratios were verified using four different denaturing conditions for the RNA: 80% formamide, 80% formamide plus glyoxal, 80% formamide/4M urea and 80% formamide/4M urea plus glyoxal. Molecules ranging in size from 1541 to 5386 nucleotides were examined and the number of bases per unit length was found to vary inversely with micrometer was consistent when RNA and DNA molecules of the same length were compared.     

Long-range translational coupling in single-stranded RNA bacteriophages: an evolutionary analysis.

In coliphage MS2 RNA a long-distance interaction (LDI) between an internal segment of the upstream coat gene and the start region of the replicase gene prevents initiation of replicase synthesis in the absence of coat gene translation. Elongating ribosomes break up the repressor LDI and thus activate the hidden initiation site. Expression studies on partial MS2 cDNA clones identified base pairing between 1427-1433 and 1738-1744, the so-called Min Jou (MJ) interaction, as the molecular basis for the long-range coupling mechanism. Here, we examine the biological significance of this interaction for the control of replicase gene translation. The LDI was disrupted by mutations in the 3'-side and the evolutionary adaptation was monitored upon phage passaging. Two categories of pseudorevertants emerged. The first type had restored the MJ interaction but not necessarily the native sequence. The pseudorevertants of the second type acquired a compensatory substitution some 80 nt downstream of the MJ interaction that stabilizes an adjacent LDI. In one examined case we confirmed that the second site mutations had restored coat-replicase translational coupling. Our results show the importance of translational control for fitness of the phage. They also reveal that the structure that buries the replicase start extends to structure elements bordering the MJ interaction.     

Sequence composition similarities with the 7SL RNA are highly predictive of functional genomic features

Transposable elements derived from the 7SL RNA gene, such as Alu elements in primates, have had remarkable success in several mammalian lineages. The results presented here show a broad spectrum of functions for genomic segments that display sequence composition similarities with the 7SL RNA gene. Using thoroughly documented loci, we report that DNaseI-hypersensitive sites can be singled out in large genomic sequences by an assessment of sequence composition similarities with the 7SL RNA gene. We apply a root word frequency approach to illustrate a distinctive relationship between the sequence of the 7SL RNA gene and several classes of functional genomic features that are not presumed to be of transposable origin. Transposable elements that show noticeable similarities with the 7SL sequence include Alu sequences, as expected, but also long terminal repeats and the 5′-untranslated regions of long interspersed repetitive elements. In sequences masked for repeated elements, we find, when using the 7SL RNA gene as query sequence, distinctive similarities with promoters, exons and distal gene regulatory regions. The latter being the most notoriously difficult to detect, this approach may be useful for finding genomic segments that have regulatory functions and that may have escaped detection by existing methods.     

Molecular basis of adaptive convergence in experimental populations of RNA viruses

Characterizing the molecular basis of adaptation is one of the most important goals in modern evolutionary genetics. Here, we report a full-genome sequence analysis of 21 independent populations of vesicular stomatitis ribovirus evolved on the same cell type but under different demographic regimes. Each demographic regime differed in the effective viral population size. Evolutionary convergences are widespread both at synonymous and nonsynonymous replacements as well as in an intergenic region. We also found evidence for epistasis among sites of the same and different loci. We explain convergences as the consequence of four factors: (1) environmental homogeneity that supposes an identical challenge for each population, (2) structural constraints within the genome, (3) epistatic interactions among sites that create the observed pattern of covariation, and (4) the phenomenon of clonal interference among competing genotypes carrying different beneficial mutations. Using these convergences, we have been able to estimate the fitness contribution of the identified mutations and epistatic groups. Keeping in mind statistical uncertainties, these estimates suggest that along with several beneficial mutations of major effect, many other mutations got fixed as part of a group of epistatic mutations.     

Comparison of proteins bound to the different functional classes of messenger RNA.

Proteins from nuclear ribonucleoproteins, informosomes, polysomal messenger ribonucleoproteins and cytoplasmic "binding factor" are characterized. 1. Nuclear ribonucleoproteins are purified from nuclei disrupted by ultrasonication. Possible contamination by nucleoplasm, histones or remaining cytoplasmic structures is controlled. 2. Informosomal proteins are obtained by mild RNAase degradation. This method gives informosomal proteins without appreciable contamination. 3. Polysomal messenger ribonucleoproteins are obtained from cells where the initiation of protein synthesis is arrested in order to release the messenger ribonucleoproteins from the polysomes. Their proteins are obtained like the informosomal proteins by mild RNAase digestion. No contamination by informosomes could be detected by sodium dodecyl sulfate gel electrophoresis. 4. Cytoplasmic "binding factor" proteins are purified by affinity chromatography. 5. The four sets of proteins are analysed by sodium dodecylsulfate acrylamide gel electrophoresis. In spite of the fact that some proteins from one or another kind of messenger ribonucleoprotein, have apparently the same molecular weight, the majority of proteins differ.     

High evolutionary conservation of the secondary structure and of certain nucleotide sequences of U5 RNA.

The nucleotide sequence of chicken, pheasant, duck and Tetrahymena pyriformis U5 RNAs as well as that of new mammalian variant U5 RNAs was determined and compared to that of rat and HeLa cells U5 RNAs. Primary structure conservation is about 95% between rat and human cells, 82% between mammals and birds and 57% between the Protozoan and mammals. The same model of secondary structure, a free single-stranded region flanked by two hairpins can be constructed from all RNAs and is identical to the model previously proposed for mammalian U5 RNA on an experimental basis (1). Thus, this model is confirmed and is likely to be that of an ancestor U5 RNA. The 3' region of the U5 RNA molecule constitutes domain A, and is common to U1, U2, U4 and U5 RNAs (2). The characteristic nucleotide sequences of domain A are highly conserved throughout the phylogenetic evolution of U5 RNA suggesting that they are important elements in the function of the four small RNAs. Another region of high evolutionary conservation is the top part of the 5' side hairpin whose conserved sequence is specific to U5 RNA. It might participate in the particular function of U5 RNA.     

Structural and functional accommodation of nucleotide variations at a conserved tRNA tertiary base pair.

The U8:A14 tertiary base pair of transfer RNAs (tRNAs) stabilizes the sharp turn from the acceptor stem to the dihydrouridine stem. This tertiary base pair is important for the overall L-shaped tRNA structure. Inspection of tRNA sequences shows that U8:A14 is highly conserved. However, variations of U8:A14 are found in natural sequences. This raises the question of whether all 16 permutations of U8:A14 can be accommodated by a single tRNA sequence framework and by the bacterial translational apparatus. Here we expressed the wild type and 15 variants of U8:A14 of an alanine tRNA amber suppressor in Escherichia coli and tested the ability of each to suppress an amber mutation. We showed that 12 of the 15 variants are functional suppressors (sup+) and 3 are nonfunctional (sup-). Of the 12 functional suppressors, the G8:G14 variant is the most efficient suppressor, whose suppression efficiency is indistinguishable from that of the wild type. Analysis of tRNA structure with chemical probes and the lead-cleavage reaction, however, showed a distinct difference between the G8:G14 variant and the wild type. Thus, two different structures of E. coli tRNAAla/CUA share an identical functional phenotype in protein synthesis. The remaining 11 sup+ variants with reduced suppression efficiencies are likely to have other structural variations. We suggest that the variations of these sup+ mutants are structurally and functionally accommodated by the bacterial translational apparatus. In contrast, the three sup- mutants harbor variations that alter the backbone structure in the corner of the L. These variations are likely to reduce the stability of the tRNA inside the cell or, among others, to interfere with the ability of the tRNA to functionally interact with elongation factor Tu and with the ribosome.     

The determination of similarities in amino acid composition among proteins separated by two-dimensional gel electrophoresis

A simple and rapid procedure has been developed to determine similarities in amino acid composition among cellular proteins separated by two-dimensional gel electrophoresis. Cells in tissue culture are simultaneously labeled with two different amino acids each tagged with a different radioisotope. The proteins are then separated on two-dimensional gels and their location on the gels determined by Coomassie-blue staining or autoradiography. Elution of the protein from the appropriate region of the gel followed by liquid scintillation counting yields an isotope ratio which reflects the ratio of the two amino acids in the protein. Examples of the use of this technique in analyzing mutant proteins, proteins altered by carbamylation, and cell proteins with similar amino acid composition (e.g., actin and tubulin) are given.     

Heterogeneity in base sequence among different DNA clones containing equivalent sequences of rotavirus double-stranded RNA.

The nucleotide sequences for several complementary DNA clones of the rotavirus genome were determined. When the sequences obtained from different clones for the same regions (16,000 bases) were compared, differences in eight base positions were observed. These discrepancies, approximately 1 in 2,000 bases, may be due to differences in individual RNA genomes resulting from multiple passages; infidelity of DNA synthesis in the cloning procedure; or both factors. Whatever the cause, this frequency of base substitution found in sequences of complementary DNA obtained from the same isolate should be considered when comparing DNA sequences obtained from independent isolates. On the other hand, the frequency of base changes observed suggests that the rotavirus genome is very conserved since the virus used for cDNA synthesis has been continuously passaged for 6 years without plaque purification.     

Validation of skeletal muscle cis-regulatory module predictions reveals nucleotide composition bias in functional enhancers

We performed a genome-wide scan for muscle-specific cis-regulatory modules (CRMs) using three computational prediction programs. Based on the predictions, 339 candidate CRMs were tested in cell culture with NIH3T3 fibroblasts and C2C12 myoblasts for capacity to direct selective reporter gene expression to differentiated C2C12 myotubes. A subset of 19 CRMs validated as functional in the assay. The rate of predictive success reveals striking limitations of computational regulatory sequence analysis methods for CRM discovery. Motif-based methods performed no better than predictions based only on sequence conservation. Analysis of the properties of the functional sequences relative to inactive sequences identifies nucleotide sequence composition can be an important characteristic to incorporate in future methods for improved predictive specificity. Muscle-related TFBSs predicted within the functional sequences display greater sequence conservation than non-TFBS flanking regions. Comparison with recent MyoD and histone modification ChIP-Seq data supports the validity of the functional regions.     

Accessibility of U1 RNA to base pairing with a single-stranded DNA fragment mimicking the intron extremities at the splice junction.

A DNA fragment containing a 16 nucleotide sequence mimicking the intron extremities of premessenger RNA aligned as proposed previously (1,2) in a model of splicing mechanism was prepared and used as a probe for accessibility of the 5' extremity of U1 RNA. Hybridization of U1 RNA to the probe under non denaturing conditions and digestion of the hybrid with RNase H revealed that the sequence of U1 RNA which is complementary to the extremities of introns is accessible to hybridization and to enzymes. Therefore, the configuration of isolated U1 RNA satisfies the criteria required for the alignment of introns and further enzymatic reactions of splicing.     

Molecular classes of heterogeneous nuclear RNA in sea urchin embryos.

Nucleotide sequences within a phage genome can be detected in individual phage plaques by in situ hybridization with complementary RNA sequences. Results with phage A and a derivative having 10% of its DNA deleted indicate that sequences 500 to 1000 base-pairs long should be detectable with confidence.