Chemistry / Although molecules that bind DNA have the potential to modify gene expression, the reality of targeting DNA in a sequence-specific manner is still a problematic but worthwhile goal. The Iverson lab has been exploring DNA recognition through a motif known as threading polyintercalation based on connecting intercalating naphthalene diimide (NDI) units, which are molecules that insert themselves between DNA base pairs, together with peptide linkers. These polyintercalators interact with both DNA grooves by “threading” or winding through the DNA, like a snake might climb a ladder. Initially, two different bisintercalator modules with altered sequence specificities and different groove binding topologies were discovered and used to inspire the design of a hybrid NDI tetraintercalator. Surprisingly enough, this tetraintercalator bound sequence-specifically with a dissociation half-life of 16 days to its preferred 14 bp site, a record at the time it was reported for a synthetic DNA-binding molecule. The work reported here expands on the capabilities of this modular threading polyintercalation motif. Chapter 2 describes the ability of a new hybrid NDI tetraintercalator, where the bisintercalator modules are connected together in a different way compared to the previously studied tetraintercalator, to subtly discriminate between similar binding sites. Chapter 3 offers a structural understanding, through NMR analysis, for the sequence recognition abilities of this new tetraintercalator. Chapter 4 analyzes the binding abilities of an un-optimized NDI octaintercalator and proposes how to approach the second-generation design of longer polyintercalators. Chapter 5 describes the optimization of the originally designed NDI tetraintercalator by serially lengthening one of the linkers to produce a tetraintercalator with a 57 day dissociation half-life from its 14 bp sequence, a new record for a synthetic DNA-binding molecule. Using the optimized linker in the context of an NDI hexaintercalator allows for binding to a 22 bp designed site, a record for a synthetic non-nucleic acid molecule. Chapter 6 recounts a focused library screening to search for bisintercalators with new sequence specificities. These efforts have laid the groundwork to progress toward studies aimed at understanding how these molecules might function to prevent transcription in a sequence-dependent manner in vivo. / text

Biological transport is an essential phenomenon for the living systems. A mechanistic investigation of biological transport processes is highly important for the characterization of physiological and cellular events, the design and functioning of several biomedical devices and the development of new therapies. To investigate the physical-chemical details of this phenomenon, concerted efforts of both experiments and theory are necessary. Motor proteins constitute a major portion of the active transport in the living cell. However, the actual mechanism of how chemical energy is converted into their directed motion has still remained obscure. Recent experiments on motor proteins have been producing exciting results that have motivated theoretical studies. In order to provide deep insight onto motor protein's mechanochemical coupling we have used stochastic modeling based on discrete-state chemical kinetic model. Such models enable us to (1) resolve the contradiction between experimental observations on heterodimeric kinesins and highly popular hand-over-hand mechanism, (2) take into account the free energy landscape modification of individual motor domains due to interdomain interaction, (3) recognize the effect of spatial fluctuations on biochemical properties of molecular motors, and (4) calculate the dynamical properties such as velocities, dispersions of complex biochemical pathways. We have also initiated the investigation of the dynamics of coupled motor assemblies using stochastic modeling. Furthermore, an extensive Monte Carlo lattice simulation based study on facilitated search process of DNA-binding proteins is presented. This simulation shows that the accelerated search compared to pure Smoluchowski limit can be achieved even in the case where the one-dimensional diffusion is order of magnitude slower than the three-dimensional diffusion. We also show that facilitated search is not only the manifestation of dimensionality reduction but correlation times play a crucial role in the overall search times. Finally, a more general field of stochastic processes, namely first-passage time process is investigated. Explicit expressions of important properties, such as splitting probailities and mean first-passage times, that are relevant to (but not limited to) biological transport, are derived for several complex systems.

The mechanisms behind how a transcription factor elicits a given phenotype can be complex. The aim of the research presented herein was to provide experimental evidence to characterise the upstream and downstream regulation of the Arabidopsis thaliana R2R3-MYB transcription factor, AtMYB61. To address these aims, three separate experiments were undertaken. First, three direct downstream target genes of AtMYB61 were predicted based on a two-stage complete transcriptome analysis, using publicly available microarray datasets in combination with a custom microarray dataset comparing the transcriptomes of WT, atmyb61 and 35S::MYB61 plants. These candidate target genes encode the following proteins: a KNOTTED1-like transcription factor, a caffeoyl-CoA 3-O-methyltransferase and a pectin-methylesterase. AtMYB61 bound the 5’ non-coding regulatory regions of these target genes, as determined by electrophoretic mobility shift assay. Second, the preferred DNA-binding sites of recombinant AtMYB61 protein were assessed with a cyclic amplification and selection of targets (CASTing) assay. Key interactions between amino acids in the AtMYB61 DNA-binding site and nucleotides in the preferred DNA targets were predicted by molecular modeling. While recombinant AtMYB61 was sufficient to drive gene expression from CASTing-identified target DNA sequences in yeast, it did so in a manner that was not entirely consistent with predicted DNA-binding affinities determined by a nitrocellulose filter binding assay. Finally, the molecular components that function upstream to modulate AtMYB61 expression were determined. AtMYB61 was determined to be de-repressed by sucrose in a mechanism involving its second intron. An over-represented motif was conserved within the second intron of Brassicaceae AtMYB61 homologues and this motif functioned as a binding target for a putative sugar-mediated repressor, as determined by EMSA. Putative AtMYB61 repressor proteins that bound this motif in the absence of sucrose were affinity purified and characterised using LC-MS/MS, and the proteins identified based on their MS fingerprints.

The mechanisms behind how a transcription factor elicits a given phenotype can be complex. The aim of the research presented herein was to provide experimental evidence to characterise the upstream and downstream regulation of the Arabidopsis thaliana R2R3-MYB transcription factor, AtMYB61. To address these aims, three separate experiments were undertaken. First, three direct downstream target genes of AtMYB61 were predicted based on a two-stage complete transcriptome analysis, using publicly available microarray datasets in combination with a custom microarray dataset comparing the transcriptomes of WT, atmyb61 and 35S::MYB61 plants. These candidate target genes encode the following proteins: a KNOTTED1-like transcription factor, a caffeoyl-CoA 3-O-methyltransferase and a pectin-methylesterase. AtMYB61 bound the 5’ non-coding regulatory regions of these target genes, as determined by electrophoretic mobility shift assay. Second, the preferred DNA-binding sites of recombinant AtMYB61 protein were assessed with a cyclic amplification and selection of targets (CASTing) assay. Key interactions between amino acids in the AtMYB61 DNA-binding site and nucleotides in the preferred DNA targets were predicted by molecular modeling. While recombinant AtMYB61 was sufficient to drive gene expression from CASTing-identified target DNA sequences in yeast, it did so in a manner that was not entirely consistent with predicted DNA-binding affinities determined by a nitrocellulose filter binding assay. Finally, the molecular components that function upstream to modulate AtMYB61 expression were determined. AtMYB61 was determined to be de-repressed by sucrose in a mechanism involving its second intron. An over-represented motif was conserved within the second intron of Brassicaceae AtMYB61 homologues and this motif functioned as a binding target for a putative sugar-mediated repressor, as determined by EMSA. Putative AtMYB61 repressor proteins that bound this motif in the absence of sucrose were affinity purified and characterised using LC-MS/MS, and the proteins identified based on their MS fingerprints.

The human wildtype p53 tumor suppressor gene is found in two different forms, p53 Arginine and p53 Proline. This difference results in a substitution of a proline for an arginine at codon 72 producing the polymorphism. Knowing that apoptosis and cell cycle arrest are the two main functions of p53, the objective of this project was to determine the difference in the capacity of these allelic variants of p53 to induce apoptosis and/or cell cycle arrest in two different experimental model systems. The first experimental system was composed of non-transformed 10(1) cells and the second one was transformed Saos-2 cells. In the first experimental system, the two wildtype forms of p53 induced cell cycle arrest at the same level and did not induce apoptosis. On the other hand, in transformed cells, both p53Arg and p53Pro induced apoptosis at similar levels. No cell cycle arrest activity has been detected in Saos-2 cells. In conclusion, this study suggests that the induction of cell cycle arrest or apoptosis depends more on the cell type than on the type of the p53 protein. Also, the intensity of cell cycle arrest or apoptosis is independent of which allelic variant of p53 is present under the experimental conditions used in this study.

Despite a century of research, the evolutionary origin of angiosperms remains uncertain. Morphological studies have identified the gnetophytes as the sister group of angiosperms mainly because of the similar organization of their reproductive structures. Molecular studies have been ambiguous as to whether these two groups are closely related. Study of the development of seed plant reproductive structures can help to untangle this issue. Here, I report the cloning of five MARS-box floral homeotic gene homologues from the gnetophyte Ephedra andina. Three of these genes belong to AG, AGL6 and TM3 subfamilies. These monophyletic groups comprise angiosperm as well as conifer homologues. Phylogenetic analysis of the plant MADS-box gene family reveals that within subfamilies, Ephedra genes always form subclades with other gymnosperm genes to the exclusion of all angiosperm genes. These results suggest that gnetophytes are more closely related to conifers than to angiosperms.

No description available.

Knowing the sequence specificities of transcription factors allows us to surmise their functions and establish their regulatory roles in genomes. The most common DNA-binding domain among eukaryotic transcription factors is the Cys2His2 zinc finger domain; however, despite their prevalence, the specificities of the majority of Cys2His2 zinc finger proteins remain unknown due to the difficulty in assaying them. My objective was to develop a new phage displayed-based assay, in which individual Cys2His2 domains are displayed on phage in an otherwise constant three-finger protein scaffold. In Chapter 2, I discuss evidence for the modularity of the Cys2His2 domain, since my assay requires that zinc fingers be modular. In Chapter 3, I describe my results on the development of this phage display-based assay. This work provides support for a new strategy to determine the specificities of individual zinc fingers, which can be used to infer specificities for multi-finger Cys2His2 proteins.

Knowing the sequence specificities of transcription factors allows us to surmise their functions and establish their regulatory roles in genomes. The most common DNA-binding domain among eukaryotic transcription factors is the Cys2His2 zinc finger domain; however, despite their prevalence, the specificities of the majority of Cys2His2 zinc finger proteins remain unknown due to the difficulty in assaying them. My objective was to develop a new phage displayed-based assay, in which individual Cys2His2 domains are displayed on phage in an otherwise constant three-finger protein scaffold. In Chapter 2, I discuss evidence for the modularity of the Cys2His2 domain, since my assay requires that zinc fingers be modular. In Chapter 3, I describe my results on the development of this phage display-based assay. This work provides support for a new strategy to determine the specificities of individual zinc fingers, which can be used to infer specificities for multi-finger Cys2His2 proteins.

Doctor of Philosophy(PhD) / The present study deals with the synthesis and characterization of six tri-palladium complexes code named MH3, MH4, MH5, MH6, MH7 and MH8 that contained two planaramine ligands bound to the central or each of the terminal metal ions. The activity of the compounds against human cancer cell lines: A2780, A2780cisR and A2780ZD0473R, cell uptake, levels of DNA-binding and nature of interaction with salmon sperm and pBR322 plasmid DNA have also been determined. Whereas cisplatin binds with DNA forming mainly intrastrand GG adduct that causes local bending of a DNA strand, the tri-palladium complexes are expected to bind with DNA forming a number of long-range interstrand GG adducts that would cause a global change in DNA conformation. Among the designed complexes, MH6 that has two 2-hydroxypyridine ligands bound to each of the two terminal palladium ions is found to be most active. The compound also has the highest cell uptake and Pd-DNA binding levels. In contrast, MH8 which has two 4-hydroxypyridine ligands bound to each of the two terminal palladium ions is found to be least active. The results indicate that, as applied to the terminal metal centres, 2-hydroxypyridine would be more activating than 4-hydroxypyridine perhaps because of greater protection provided to the terminal centres from coming in contact with the solvent molecules. In contrast, when bound to the central metal centre, 4-hydroxypyridine appears to play a slightly greater activating role than 2-hydroxypyridine or 3-hydroxypyridine, suggesting that non-covalent interactions such as hydrogen bonding associated with the ligand rather than its steric effect may be a more important determinant of antitumour property. The results illustrate structure-activity relationships and suggest that the tri-palladium complex containing two 2-hydroxypyridine ligands bound to each of the three metal centres or the compound that contains two 2-hydroxypyridine ligands bound to each of the two terminal metal centres and two 4-hydroxypyridine ligands bound to the central metal centre, may be much more active than any of the designed complexes.