Our previous discovery of the turnip crinkle virus tRNA-like translational enhancer (TCV TSS) motivated us to search for similar elements in other viruses. As a result, the middle of the UTR of the pea enation mosaic virus (PEMV) was found, by our modeling, to have a comparable T-shaped 3D motif (kl-TSS), a CITE, that is similar in both structure and function to the TCV element. This structure, however, does not contain pseudoknots like the TCV element. It was shown that a hairpin loop in the T-shaped PEMV engages in a long-distance kissing loop interaction with a hairpin stem loop just down-stream of the start codon. It was also shown that ribosomes bind to the kissing complex as well as to the T-shaped element alone. Analysis indicated that mutations in the element that reduced translation also reduced ribosome binding or the RNA-RNA kissing loop kl-TSS interaction. This element is also presumed to be important not only for delivery of ribosomes to the 5'end of the virus, but also in associating the eukaryotic initiation factor eIF4E with the 5'end due the binding of eIF4E with a nearby downstream element the PTE. We developed the programs CovaRna and CovStat to explore long-range co-varying RNA interaction networks using whole genome alignments. This new methodology was applied to Drosophila genomes. A parallel version of the program was devised to speed-up processing and the algorithms also rely on fast indexing schemes and conservative statistical methods to determine highly significant interactions. The methodology found interesting interactions that appear to be related to endogenous siRNAs, gene transport, and to synchronized mutations that may include common interaction partners. These results indicate the potential importance of such networks. The World Health Organization estimates there may be as many as 50-100 million cases of dengue fever per year. We characterized the untranslated regions of the virus focusing on the 3'UTR and its effects on replication and translation. We used MPGAfold and StructureLab to analyze the folding characteristics of a subgenomic (minigenome) RNA from dengue type 2 as well as mutated variants. We previously showed that MPGAfold is capable of predicting the secondary structure and folding dynamics including co-transcriptional folding of various RNAs. Our computational results were correlated with experiments on a 719 nt minigenome consisting mostly of the 5'and 3'UTRs and a dengue type 2 replicon containing a Renilla luciferase reporter. Attention was focused on the formation of two dumbbell-like structures found to form in the 3'UTR with the putative formation of pseudoknots involving regions downstream from the dumbbells. The computational results showed a significant preference for the formation of the 3'dumbbell over the 5'dumbbell. Mutational studies indicated that the motifs involved in dumbbell formation including the putative pseudoknots were important for replication but had less significant varying effects on translation. In a later related study Selective 2'-hydroxyl Acylation Analyzed by Primer Extension (SHAPE) analysis was done on the minigenome and the mutational variants. Most of the MPGAfold predicted structures were found to be present by SHAPE, including the 5'-3'interactions and the dumbbell structures. SHAPE found evidence for the formation of the 5'pseudoknot structure, but the formation of the 3'pseudoknot appeared to be complicated by the presence of a portion of the 3'pseudoknot sequence being involved in the cyclization motive involving a 5'sequence. Understanding sequence related conformational characteristics of kissing loop formation is important for understanding the formation of these motifs in RNA viruses such as HIV and is also important for understanding their formation for use in RNA-based nanodesign. We explored two HIV-1 kissing loop monomers (from subtype-A and subtype B). Our hypothesis in this study was that the structure of the monomer loop was important for the assembly of the dimer. It was experimentally shown that subtype-B monomers dimerize in high salt concentrations (NaCl) or in the presence of magnesium, while subtype-A requires coordination of magnesium ions with phosphate groups associated with bases in the loop irrespective of the salt concentration. We performed molecular dynamics simulations on the monomers with varying salt concentrations and with and without magnesium. We found that at low salt concentrations that the structures of the monomers were significantly deformed and not conducive for dimerization. At high salt concentrations, the simulations showed that the subtype-B monomer maintained its shape and was conducive for dimerization, while the subtype-A loop was significantly deformed. It was also found that two flanking bases in the loops were important for maintaining the structure of the loop. Subtype-A's loop was stabilized in the presence of magnesium coordinating with the two flanking bases. Restraints were also applied to atoms near the bottom of the loops with no salt or magnesium present. It was shown that the restraints maintained the shape of the monomer, but when released the monomers became distorted. Thus, the MD simulations strongly suggest that salt and magnesium play important but somewhat different roles in maintaining the proper conformations of the HIV monomers, which in turn may be indicative of proper dimer formation. MicroRNA 137 has been implicated with other microRNAs to be associated with neurogenesis and brain tumor development. Specifically, its up-regulation has been shown to be important for neuro cell differentiation and down-regulation has been implicated in disease processes. A genome-wide target mapping was performed on glioblastoma cells. 1468 genes were found to be negatively impacted by miR-137. Computational analysis also revealed that many of the genes had a highly likely interaction with miR-137 seed region. Several of the targeted genes were related to neurogenesis and oncogenic proteins. Other miRNAs, namely miR-124, 128 and 7 also regulate targets that are regulated by miR-137 and are associated with neurogenesis and tumorgenesis, thus their increase or decrease can have significant implications for the genetic control of neurogenesis or tumorgenesis. Over 150 million people are infected with hepatitis C virus (HCV). We investigated the synergistic effects that infection by the virus has on the course of the virus. The elevation of the expression of IFNL3 mRNA is known to help in clearance of the virus. In particular, a new means was discovered whereby HCV affects this antiviral response. This mechanism is dependent on a mutation found in the 3'UTR of IFNL3 mRNA that controls transcript stability. The mutation had an effect on AU-rich dependent decay of the transcript and the binding of a microRNA induced by HCV infection. RNA structural analysis showed that the indicated mutation in the 3'UTR of IFNL3 had the potential to alter the structure and thereby prevent AU rich element (ARE) binding proteins from degrading the RNA. In addition, it was found that the mutation prevents the binding of microRNAs that cause down-regulation of this gene. MicroRNAs, which arise from precursor transcripts, play an important role in cellular gene regulation. An algorithm was developed, MirID, to distinguish pre-miRNAs stem-loop structures from pseudo pre-miRNAs. The algorithm accepts an RNA sequence as input and classifies the sequence as either containing a pre-miRNA or a pseudo pre-miRNA. 74 different features were tested in various combinations to find the best models for classifying the pre-miRNAs. A support vector machine was used to construct a classifier ensemble. The accuracy of the method is further enhanced by the use of an AdaBoost algorithm. Results show high accuracy for classification.
Showing the most recent 10 out of 32 publications
