Bioinformatic analysis for exploring relationships between genes and gene products

To carry out their specific roles in the cell, genes and gene products often work together in groups, forming many relationships among themselves and with other molecules. Such relationships include physical protein-protein interaction relationships, regulatory relationships, metabolic relationships, genetic relationships, and much more. With advances in science and technology, some high throughput technologies have been developed to simultaneously detect tens of thousands of pairwise protein-protein interactions and protein-DNA interactions. However, the data generated by high throughput methods are prone to noise. Furthermore, the technology itself has its limitations, and cannot detect all kinds of relationships between genes and their products. Thus there is a pressing need to investigate all kinds of relationships and their roles in a living system using bioinformatic approaches, and is a central challenge in Computational Biology and Systems Biology. This dissertation focuses on exploring relationships between genes and gene products using bioinformatic approaches. Specifically, we consider problems related to regulatory relationships, protein-protein interactions, and semantic relationships between genes. A regulatory element is an important pattern or "signal", often located in the promoter of a gene, which is used in the process of turning a gene "on" or "off". Predicting regulatory elements is a key step in exploring the regulatory relationships between genes and gene products. In this dissertation, we consider the problem of improving the prediction of regulatory elements by using comparative genomics data. With regard to protein-protein interactions, we have developed bioinformatics techniques to estimate support for the data on these interactions. While protein-protein interactions and regulatory relationships can be detected by high throughput biological techniques, there is another type of relationship called semantic relationship that cannot be detected by a single technique, but can be inferred using multiple sources of biological data. The contributions of this thesis involved the development and application of a set of bioinformatic approaches that address the challenges mentioned above. These included (i) an EM-based algorithm that improves the prediction of regulatory elements using comparative genomics data, (ii) an approach for estimating the support of protein-protein interaction data, with application to functional annotation of genes, (iii) a novel method for inferring functional network of genes, and (iv) techniques for clustering genes using multi-source data.

Genomic variation of HIV-1 and its effects on drug resistance and molecular epidemiology analyses

In the first part of this study human immunodeficiency virus type 1 (HIV-1) proviral DNA sequences derived from 201 clones of the C2-V3 env region and the first exon of the tat gene were obtained from six MV-1 infected heterosexual couples. These molecular data were used to confirm the epidemiological relationships. The ability of the molecular data to draw such conclusions was also tested with multiple phylogenetic analyses. The tat region was much more useful in establishing epidemiological relationships than the commonly used C2-V3.^ Subsequently, using nucleotide sequences from the first exon of the Tat gene, we tested the hypothesis that a Florida dentist (a common source) infected five of his patients in the course of dental procedures, against the null hypothesis that the dentist and each individual of the dental group independently acquired the virus within the local community. Multiple phylogenetic analyses demonstrated that the sequences of the five patients were significantly more related to each other than to sequences of the controls. Our results using Tat sequences, combined with envelope sequence data, strongly support a common phylogenetic epidemiological relationship among these five patients.^ A third study is presented, which deals with the effects of genomic variations in drug resistance. HIV-1 reverse transcriptase (RT) mutations were detected in DNA from peripheral blood mononuclear cells from 11 of 12 HIV-infected children after 11-20 months of zidovudine monotherapy. The codon 41/215 mutant combination was associated with general decline in health status. Patients developing the codon 70 mutation tended to have a better health status. ^

Binding of small molecules to DNA junctions

Previous results in our laboratory suggest that the (CG) 4 segments whether present in a right-handed or a left-handed conformation form distinctive junctions with adjacent random sequences. These junctions and their associated sequences have unique structural and thermodynamic properties that may be recognized by DNA-binding molecules. This study probes these sequences by using the following small ligands: actinomycin D, 1,4-bis(((di(aminoethyl)amino)ethyl)amino)anthracene-9,10-dione, ametantrone, and tris(phenanthroline)ruthenium (II). These ligands may recognize the distinctive features associated to the (CG)4 segment and its junctions and thus interact preferentially near these sequences. Restriction enzyme inhibition assays were used to determine whether or not binding interactions took place, and to approximate locations of these interactions. These binding studies are first carried out using two small synthetic oligomers BZ-III and BZ-IV. The (5meCG)4 segment present in BZ-III adopts the Z-conformation in the presence of 50 m M Co(NH3)63+. In BZ-IV, the unmethylated (CG)4 segment changes to a non-B conformation in the presence of 50 m M Co(NH3)63+. BZ-IV, containing the (CG)4 segment, was inserted into a clone plasmid then digested with the restriction enzyme Hinf I to produce a larger fragment that contains the (CG)4 segment. The results obtained on the small oligomers and on the larger fragment for restriction enzyme Mbo I indicate that 1,4-bis(((di(aminoethyl)amino)ethyl)amino)anthracene-9,10-dione binds more efficiently at or near the (CG)4 segment. Restriction enzymes EcoRV, Sac I and Not I with cleavage sites upstream and downstream of the (CG)4 insert were used to further localize binding interactions in the vicinity of the (CG)4 insert. RNA polymerase activity was studied in a plasmid which contained the (CG)4 insert downstream from the promoter sites of SP6 and T7 RNA polymerases. Activities of these two polymerases were studied in the presence of each one of the ligands used throughout the study. Only actinomycin D and spider, which bind at or near the (CG)4 segment, alter the activities of SP6 and T7 RNA polymerases. Surprisingly, enhancement of polymerase activity was observed in the presence of very low concentrations of actinomycin D. These results suggest that the conformational features of (CG) segments may serve in regulatory functions of DNA. ^

Thioredoxin and Jab1 control estrogen- and antiestrogen-mediated progression of the cell cycle through p27 interactions

A major problem with breast cancer treatment is the prevalence of antiestrogen resistance, be it de novo or acquired after continued use. Many of the underlying mechanisms of antiestrogen resistance are not clear, although estrogen receptor-mediated actions have been identified as a pathway that is blocked by antiestrogens. Selective estrogen receptor modulators (SERMs), such as tamoxifen, are capable of producing reactive oxygen species (ROS) through metabolic activation, and these ROS, at high levels, can induce irreversible growth arrest that is similar to the growth arrest incurred by SERMs. This suggests that SERM-mediated growth arrest may also be through ROS accumulation. Breast cancer receiving long-term antiestrogen treatment appears to adapt to this increased, persistent level of ROS. This, in turn, leads to the disruption of reversible redox signaling that involves redox-sensitive phosphatases and protein kinases and transcription factors. This has downstream consequences for apoptosis, cell cycle progression, and cell metabolism. For this dissertation, we explored if altering the ROS formed by tamoxifen also alters sensitivity of the drug in resistant cells. We explored an association with a thioredoxin/Jab1/p27 pathway, and a possible role of dysregulation of thioredoxin-mediated redox regulation contributing to the development of antiestrogen resistance in breast cancer. We used standard laboratory techniques to perform proteomic assays that showed cell proliferation, protein concentrations, redox states, and protein-protein interactions. We found that increasing thioredoxin reductase levels, and thus increasing the amount of reduced thioredoxin, increased tamoxifen sensitivity in previously resistant cells, as well as altered estrogen and tamoxifen-induced ROS. We also found that decreasing levels of Jab1 protein also increased tamoxifen sensitivity, and that the downstream effects showed a decrease p27 phosphorylation in both cases. We conclude that the chronic use of tamoxifen can lead to an increase in ROS that alters cell signaling and causing cell growth in the presence of tamoxifen, and that this resistant cell growth can be reversed with an alteration to the thioredoxin/Jab1 pathway.

Multimode Analysis of Nanoscale Biomolecular Interactions

Biomolecular interactions, including protein-protein, protein-DNA, and protein-ligand interactions, are of special importance in all biological systems. These interactions may occer during the loading of biomolecules to interfaces, the translocation of biomolecules through transmembrane protein pores, and the movement of biomolecules in a crowded intracellular environment. The molecular interaction of a protein with its binding partners is crucial in fundamental biological processes such as electron transfer, intracellular signal transmission and regulation, neuroprotective mechanisms, and regulation of DNA topology. In this dissertation, a customized surface plasmon resonance (SPR) has been optimized and new theoretical and label free experimental methods with related analytical calculations have been developed for the analysis of biomolecular interactions. Human neuroglobin (hNgb) and cytochrome c from equine heart (Cyt c) proteins have been used to optimize the customized SPR instrument. The obtained Kd value (~13 µM), from SPR results, for Cyt c-hNgb molecular interactions is in general agreement with a previously published result. The SPR results also confirmed no significant impact of the internal disulfide bridge between Cys 46 and Cys 55 on hNgb binding to Cyt c. Using SPR, E. coli topoisomerase I enzyme turnover during plasmid DNA relaxation was found to be enhanced in the presence of Mg2+. In addition, a new theoretical approach of analyzing biphasic SPR data has been introduced based on analytical solutions of the biphasic rate equations. In order to develop a new label free method to quantitatively study protein-protein interactions, quartz nanopipettes were chemically modified. The derived Kd (~20 µM) value for the Cyt c-hNgb complex formations matched very well with SPR measurements (Kd ~16 µM). The finite element numerical simulation results were similar to the nanopipette experimental results. These results demonstrate that nanopipettes can potentially be used as a new class of a label-free analytical method to quantitatively characterize protein-protein interactions in attoliter sensing volumes, based on a charge sensing mechanism. Moreover, the molecule-based selective nature of hydrophobic and nanometer sized carbon nanotube (CNT) pores was observed. This result might be helpful to understand the selective nature of cellular transport through transmembrane protein pores.

A Rapid and Sensitive High-Throughput Screening Method to Identify Compounds Targeting Protein–Nucleic Acids Interactions

DNA-binding and RNA-binding proteins are usually considered ‘undruggable’ partly due to the lack of an efficient method to identify inhibitors from existing small molecule repositories. Here we report a rapid and sensitive high-throughput screening approach to identify compounds targeting protein–nucleic acids interactions based on protein–DNA or protein–RNA interaction enzyme-linked immunosorbent assays (PDI-ELISA or PRI-ELISA). We validated the PDI-ELISA method using the mammalian highmobility- group protein AT-hook 2 (HMGA2) as the protein of interest and netropsin as the inhibitor of HMGA2–DNA interactions. With this method we successfully identified several inhibitors and an activator for HMGA2–DNA interactions from a collection of 29 DNA-binding compounds. Guided by this screening excise, we showed that netropsin, the specific inhibitor of HMGA2–DNA interactions, strongly inhibited the differentiation of the mouse pre-adipocyte 3T3-L1 cells into adipocytes, most likely through a mechanism by which the inhibition is through preventing the binding of HMGA2 to the target DNA sequences. This method should be broadly applicable to identify compounds or proteins modulating many DNA-binding or RNA-binding proteins.