Position weight matrices (PWMs) have become a tool of choice for the identification of transcription factor binding sites in DNA sequences. PWMs are compiled from experimentally verified and aligned binding sequences. PWMs are then used to computationally discover novel putative binding sites for a given protein. DNA-binding proteins often show degeneracy in their binding requirement, the overall binding specificity of many proteins is unknown and remains an active area of research. Although PWMs are more reliable predictors than consensus string matching, they generally result in a high number of false positive hits. A previous study introduced a novel method to PWM training based on the known motifs to sample additional putative binding sites from a proximal promoter area. The core idea was further developed, implemented and tested in this thesis with a large scale application. Improved mono- and dinucleotide PWMs were computed for Drosophila melanogaster. The Matthews correlation coefficient was used as an optimization criterion in the PWM refinement algorithm. New PWMs keep an account of non-uniform background nucleotide distributions on the promoters and consider a larger number of new binding sites during the refinement steps. The optimization included the PWM motif length, the position on the promoter, the threshold value and the binding site location. The obtained predictions were compared for mono- and dinucleotide PWM versions with initial matrices and with conventional tools. The optimized PWMs predicted new binding sites with better accuracy than conventional PWMs.

Pantothenate synthetase (PS), which catalyzes the last step in the pantothenate (vitamin B5) biosynthesis, is a dimeric enzyme present in bacteria, fungi and plants. The enzymatic properties of PS from Escherichia Coli, Mycobacterium tuberculosi, Fusarium Oxysporum, Lotus japonicus, Oryza sativum, Brassica napus and Arabidopsis thaliana have been characterized. The chemical reaction and the proposed mechanism of reaction are identical for PS, irrespective of the source. However, from an enzyme mechanistic point of view, plant PS’s are dissimilar to their bacterial counterparts, in that they exhibit “allosteric behavior”, a property that has not been observed in the bacterial enzyme. The behavior is quite remarkable when one takes into consideration the fact that plant PS’s share a high degree of sequence identity (~ 40%) with the bacterial enzymes. Even more intriguing is the structural mechanism proposed to explain the observed differences in structure between the PS’s from E.Coli and M.tb, which share a 42% sequence identity. Till date there is no structural information available on the plant PS’s and of the substrate bound conformation of E.coli PS. This thesis aims to provide an understanding on some aspects of the structure – function relationship of this physiologically important enzyme. Specifically, the solution properties of E. coli PS have been examined using high-resolution multinuclear, multidimensional NMR methods. Given the large size of the full-length protein (~ 63 KDa), the structurally distinct N and C-terminal domains were cloned and expressed as individual proteins and their properties investigated. Towards this end, the tertiary fold of the 40 kDa dimeric amino-terminal domain of E. coli pantothenate synthetase has been determined using multidimensional multinuclear nuclear magnetic resonance (NMR) methods (PDB entry 2k6c). Sequence specific resonance assignments for backbone HN, 15N, 13Cα, 13C', sidechain 13Cβ and aliphatic 13CH3 (of isoleucine, leucine and valine residues) were obtained using perdeuterated ILV-methyl protonated samples (BMRB entry 6940). Secondary structure of nPS was determined from 13C secondary chemical shifts and from short and medium range NOEs. Global fold of the 40 kDa homo-dimeric nPS has been determined using a total of 1012 NOEs, 696 dihedral angles, 260 RDCs, 155 hydrogen bonds, radius of gyration potential, non-crystallographic symmetry potential and database derived potential based upon the Ramachandran map. The calculated structures, which show that the N-terminal domain forms a homo-dimer in solution, is of high stereochemical quality as judged by the Ramachandran statistics (70% of the residues have backbone dihedral angles in the allowed region, 25.5% in the additionally allowed region, 4.0% in generously allowed region, and only 0.5% in disallowed region). Dynamics of nPS, which has rotational correlation time τc of 17.3 ns, was investigated by 15N relaxometry measurements. Results of these studies indicate that the E. coli protein exhibits dynamic nature at the dimer interface. These structural and dynamic features of the protein were found to be of interest when correlated with NMR based substrate binding studies. Interaction of homo-dimeric amino-terminal domain (nPS) of E. coli pantothenate synthetase (PS; encoded by the gene panC; E.C. 6.3.2.1) with the substrates pantoate, β-alanine, ATP and the product pantothenate has been studied using isotopically edited solution NMR methods. Addition of pantoate prior to ATP has led to the interesting observation that pantoate binds each monomer of nPS at two sites. ATP displaces a molecule of pantoate from the ATP binding site. β-alanine and pantothenate do not bind the protein under the condition studied. Binding of pantoate and ATP also manifests as changes in residual dipolar couplings (RDCs) of backbone 1H-15N pairs in nPS when compared to the free form of the protein. Structures of homo-dimeric nPS bound to two molecules of pantoate (PDB entry 2k6e) as well as to pantoate + ATP (PDB entry 2k6f) were calculated by inclusion of hydrogen bonds between the ligands and backbone 1H-15N pairs of nPS in addition to other NMR derived restraints. The ligand bound structures have been compared to the similar forms of the M. tb PS. Structure of each monomer of nPS bound to pantoate and ATP shows the substrates in a favorable orientation for the intermediate pantoyl adenylate to form. Moreover, at all stages of substrate binding the symmetry of the dimer was preserved. A single set of resonances was observed for all protein-ligand complexes implying symmetric binding with full-occupancy of the ligands bound to the protein. In an effort to understand the structural basis of the observed enzymatic properties of plant PS’s, a structural model of the Arabidopsis PS was constructed. The results of these structural and substrate binding studies strongly suggest that 1 Substrate binding to PS occurs only at the active site. 2 There are no additional substrate binding sites which could potentially participate as regulatory sites. 3 Pantoate does not bind at the dimer interface to induce the observed homotropic effects. 4 The structural results presented on the substrate bound forms of nPS have direct implication for the development of novel antibacterial and herbicidal agents. Recently a great deal of interest has been evinced on the effects of molecular crowding on protein folding / unfolding pathways. Nuclear magnetic resonance is the only method by which high resolution structural information can be obtained on partially denatured states of a protein under equilibrium condition. Recent methodological advances have enabled the determination of high resolution structures using information embedded in the residual dipolar couplings. Molecular crowding using confinement may thus reveal important details about chaperone mediated protein folding. We have attempted to develop a protocol to study the effects of molecular confinement by sequestering proteins in poly-acrylamide gels and then subjecting these protein molecules to denaturation and then characterize these states by nuclear magnetic resonance. The preliminary results of these studies are described here.

Position weight matrices (PWMs) have become a tool of choice for the identification of transcription factor binding sites in DNA sequences. PWMs are compiled from experimentally verified and aligned binding sequences. PWMs are then used to computationally discover novel putative binding sites for a given protein. DNA-binding proteins often show degeneracy in their binding requirement, the overall binding specificity of many proteins is unknown and remains an active area of research. Although PWMs are more reliable predictors than consensus string matching, they generally result in a high number of false positive hits. A previous study introduced a novel method to PWM training based on the known motifs to sample additional putative binding sites from a proximal promoter area. The core idea was further developed, implemented and tested in this thesis with a large scale application. Improved mono- and dinucleotide PWMs were computed for Drosophila melanogaster. The Matthews correlation coefficient was used as an optimization criterion in the PWM refinement algorithm. New PWMs keep an account of non-uniform background nucleotide distributions on the promoters and consider a larger number of new binding sites during the refinement steps. The optimization included the PWM motif length, the position on the promoter, the threshold value and the binding site location. The obtained predictions were compared for mono- and dinucleotide PWM versions with initial matrices and with conventional tools. The optimized PWMs predicted new binding sites with better accuracy than conventional PWMs.

Molecular dynamics (MD) simulations were used to investigate the structural basis for the functions of three proteins: Fc(gamma) receptor III (CD16), von Willebrand factor (VWF), and integrin. CD16, a heavily glycosylated protein expressed on human immune cells, plays a crucial role in immune defense by linking antibody-antigen complexes with cellular effector functions. Glycosylation of CD16 decreases its affinity for IgG. MD simulations were run for CD16-IgG Fc complexes with or without an N-glycan on CD16. The two simulated complexes show different conformations. Molecular Mechanics-Poisson Boltzmann Surface Area (MM-PBSA) approach was used to calculate the binding free energy of the CD16-IgG Fc complexes. The calculated binding free energy helped to identify critical residues. VWF, a multimeric multidomain glycoprotein, initiates platelet adhesion at the sites of vascular injury. A specific VWF metalloprotease, A Disintegrin And Metalloprotease with ThromboSpondin motifs member 13 (ADAMTS-13), cleaves the Tyr1605-Met1606 bond in the VWF A2 domain to generate the full spectrum of plasma VWF species. Shear stress or denaturants assist VWF cleavage by ADAMTS-13 due to the unfolding of A2. MD was used to simulate the unfolding processes of A2 under force or high temperature. The beta-strands of A2 were pulled out sequentially by force, during which the cleavage site changed in steps from the fully buried state to the fully exposed state. Thermal unfolding follows a very different pathway. Integrins are adhesion molecules mediating cell-cell, cell-extracellular matrix, and cell-pathogen interactions. Experiments suggest that integrins can undergo a large-scale change from a bent to an extended conformation, associating with a transition from low to high affinity states, i.e., integrin activation. Steered MD was utilized to simulate the bent-to-extended conformational transition in time of aVb3 integrin. The integrin was observed to change smoothly from the bent to the extended conformation. One major energy barrier was overcome, corresponding to the disruption of the interactions at Hybrid/EGF4/bTD interfaces. A partially extended conformation tends to bend back while a fully extended conformation is stabilized by the coordination of Asp457 with Ca2+ at alpha-genu. Unbending with separated legs overcomes more energy barriers.

Thesis (Ph. D.)--University of Virginia, 2008. / Title from title page. Includes bibliographical references. Also available online through Digital Dissertations.

The "RNA world" hypothesis makes two predictions that RNA should have been able both to catalyze RNA replication and to direct protein synthesis. The evolution of RNA-catalyzed protein synthesis should be critical in the transition from the RNA world to the modem biological systems. Peptide bond formation is a fundamental step in modem protein biosynthesis. Although many evidence suggests that the ribosome is a ribozyme, peptide bond formation has not been achieved with ribosomal RNAs only. The goal of this thesis is to investigate whether RNA could catalyze peptide bond formation and how RNA catalyzes peptide bond formation. Two systems have been employed to approach these questions, the ribozyme system and the ribosome system. Ribozymes have been isolated by in vitro selection that can catalyze peptide bond formation using the aminoacyl-adenylate as the substrate. The isolation of such peptide-synthesizing ribozymes suggests that RNA of antiquity might have directed protein synthesis and bolsters the "RNA world" hypothesis. In the other approach, a novel assay has been established to probe the ribosomal peptidyltransferase reaction in the presence of intact ribosome, ribosomal subunit, or ribosomal RNA alone. Several aspects of the peptidyltransfer reaction have been examined in both systems including metal ion requirement, pH dependence and substrate specificity. The coherence between the two systems is discussed and their potential applications are explored. Although the ribozyme system might not be a reminiscence of the ribosome catalysis, it is still unique in other studies. The newly established assay for ribosomal peptidyltransferase reaction provides a good system to investigate the mechanism of ribosomal reaction and may have potential application in drug screening to search for the specific peptidyltransferase inhibitors.

Introduction AP2α is a 52 kDa retinoic acid inducible and developmentally regulated activator of transcription, which binds to the DNA in a sequence-specific manner. Transcription factor AP-2α was isolated from HeLa cells by affinity chromatography using specific binding sites with in SV40 and human metallothionein promoters. Further screening of HeLa cDNA library with oligonucleotide probes predicted partial peptide sequence which led to the isolation of AP-2α cDNA and subsequently it was mapped to chromosome 6 near HLA locus. A differentially spliced version of AP-2α, which lacks most of the C-terminus, encodes a dominant negative protein (AP-2B). Subsequent studies led to the identification of four more isoforms: AP-2β, AP-2γ, AP-2δ and AP-2ε. AP-2 family members can form homo or hetero dimers among themselves through the unique C-terminal helix span helix motif and bind DNA through basic domain lies N-terminus of DNA binding domain. Several evidences suggest that AP-2α can act as a tumor suppressor gene. It has been shown that AP-2α can activate growth suppressor genes like p21WAF1/CIP1. Transforming viral oncogenes like adenovirus E1A and SV40 large T antigen have been shown to alter AP-2α function. In addition, reduced expression of AP-2α has been reported in human breast, ovary, colon, skin, brain and prostate cancers. Further, supporting evidences suggest that more invasiveness and tumorogenicity was observed when dominant negative mutant of AP-2α was expressed in melanoma cells. In this work, we have carried out a systematic study to find the various signal transduction pathways which regulate AP-2 activity as well as we attempted to demonstrate the importance of DNA binding domain in the growth inhibitory functions of AP-2α. HDAC inhibitors (HDIs) activate AP-2 activity through spleen tyrosine kinase (Syk) In the literature, ample evidences are available that genotoxic drugs such as adriamycin, induce tumor suppressors like p53 and p73. In this study, we have screened pharmacological drugs which damage DNA and specific inhibitors of various signal transduction pathways for their ability to activate AP-2 activity. AP-2 specific reporter, 3Χ-AP2-CAT was used in this study to measure the AP-2 activity. Of all the compounds studied, we found that Histone Deacetylase Inhibitors (HDIs) efficiently activated AP-2 activity and was found to be specific as they failed to activate 3X-AP2 mut CAT, which contains mutated AP-2 binding sites as well as pGL tk Luc, which contains thymidine kinase minimal promoter and no AP-2 binding sites. To understand the mechanism of HDI-mediated of AP-2 activation, AP-2 isoforms and its coactivators transcript and protein levels were analyzed. We found significant change in transcript levels of the some of the molecules tested. While the endogenous protein levels of various AP-2 isoforms were undetectable, we found stabilization of AP-2α protein expressed from exogenous source in cells treated with HDIs. HDI stabilized AP-2α was found to be functionally active as it showed increased sequence-specific DNA-binding as well as increased apoptosis. While HDIs known for their ability to modulate the gene activities by chromatin remodeling, it is also known that they alter various signal transduction pathways. In an effort to find pathway(s) by which HDIs activate AP-2 activity, we found that HDIs failed to activate AP-2 reporter in the presence of staurosporine suggesting the involvement a staurosporine sensitive pathway(s) in this process. Stauosporine is a non-specific kinase inhibitor of different signaling pathways. Further studies using different pathway specific inhibitors identified that spleen tyrosine kinase (Syk) is essential for HDIs mediated activation of AP-2 activity. Syk is a non receptor tyrosine kinase which is known to be activated in stress conditions. Syk is considered to be a tumor suppressor since Syk over expression leads to growth suppression of breast cancer cells and is also inactivated in a subset of breast cancers. These results suggest that HDI mediated activation of AP-2 involves AP-2α stabilization through Syk pathway. Regulation of AP-2 by MAP kinase pathway Cell growth, differentiation, and apoptosis are mediated by the activation of mitogenactivated protein kinase (MAPK) pathways. These kinases constitute MAP kinase cascades mainly regulated through phosphorylation status. In mammalian cells, at least four MAPKs, namely, extracellular signal-regulated kinases (ERKs), c-Jun N-terminal kinase/stress-activated protein kinases (JNK/SAPKs), p38 and ERK5/big MAP kinase have been identified. The ERKs are usually activated by mitogenic stimuli which in turn increase the proliferation and survival. Over expression of any activator of this signaling cascade lead to the unregulated proliferation of cells. In many cancers, ERK pathways are known to be up regulated. In this study, we found that MEK (MEK is the immediate upstream regulator of ERK) inhibitors - PD98059 and U0126 activate 3X-AP2-CAT suggesting that AP-2 activity is repressed by activated MAP kinase pathway. MEK inhibitor mediated activation was found to be specific because they failed to activate transcription from pGL tk Luc which contains thymidine kinase minimal promoter and no AP-2 binding sites. To understand the mechanism of MEK inhibitor-mediated of AP-2 activation, AP-2 isoforms and its coactivators transcript and protein levels were analyzed. We found significant change in transcript levels of the some of the molecules tested. The endogenous protein levels of various AP-2 isoforms were undetectable. When AP-2α was exogenously expressed, while no change in protein levels and DNA-binding ability was seen, we found evidence for appearance of post-ranslationally modified AP-2α protein in U0126 treated cells. We also found CITED2 (CBP/p300-interacting transactivator 2, co-activator of AP-2α) transcript levels were up regulated in UO126 treated cells. Post translational modifications of AP-2α and increased and increased CITED2 levels may be responsible for MEK inhibitor mediated AP-2 activation. Thus we conclude that ERK pathway, which is an oncogenic MAP kinase pathway, inhibits AP-2 activity thereby suggesting the importance of down regulation of AP-2 activity during transformation. Essential role of DNA-binding domain of AP-2α for its growth inhibitory functions Transcription factor AP-2α has three distinct domains, N-terminal transactivation domain (52-108 aa), C-terminal DNA binding domain (204-408 aa) and dimerization domain (277-395 aa) which lies within the DNA binding domain. AP-2α exerts its effects through binding to specific DNA sequence in the promoter of its target genes leading to either repression or activation. Recent evidences suggest that AP-2α represses many genes through its competitive binding to overlapping AP-2 and other transcription factor binding sites. This suggests an important role exclusively for the DNA binding domain in AP-2α mediated functions. To address the importance of DNA binding domain for AP-2α mediated apoptosis, we have tested the ability different deletion/point mutants of AP-2α with varying DNA binding and transactivation capability to perform growth suppressor function and ability to induce apoptosis. Replication-deficient recombinant adenoviruses expressing different mutants were used in this study. We found that an intact DNA-binding domain alone even in the absence of activation domain is sufficient for AP-2α to inhibit colony formation and to induce significant levels of apoptosis. These results suggest an important role for DNA binding domain growth inhibitory functions of AP-2α and thereby implying the importance of transcriptional repression in AP-2α functions.

No description available.

Position weight matrices (PWMs) have become a tool of choice for the identification of transcription factor binding sites in DNA sequences. PWMs are compiled from experimentally verified and aligned binding sequences. PWMs are then used to computationally discover novel putative binding sites for a given protein. DNA-binding proteins often show degeneracy in their binding requirement, the overall binding specificity of many proteins is unknown and remains an active area of research. Although PWMs are more reliable predictors than consensus string matching, they generally result in a high number of false positive hits. A previous study introduced a novel method to PWM training based on the known motifs to sample additional putative binding sites from a proximal promoter area. The core idea was further developed, implemented and tested in this thesis with a large scale application. Improved mono- and dinucleotide PWMs were computed for Drosophila melanogaster. The Matthews correlation coefficient was used as an optimization criterion in the PWM refinement algorithm. New PWMs keep an account of non-uniform background nucleotide distributions on the promoters and consider a larger number of new binding sites during the refinement steps. The optimization included the PWM motif length, the position on the promoter, the threshold value and the binding site location. The obtained predictions were compared for mono- and dinucleotide PWM versions with initial matrices and with conventional tools. The optimized PWMs predicted new binding sites with better accuracy than conventional PWMs.

Indiana University-Purdue University Indianapolis (IUPUI) / My thesis comprises of two individual projects which revolve around the importance of post-transcriptional regulation in liver. My first project is studying the integrated miRNA – mRNA network in NAFLD. For fulfillment of the study we conducted a genome-wide study to identify microRNAs (miRs) as well as the miR-mRNA regulatory network associated with hepatic fat and NAFLD. Hepatic fat content (HFC), miR and mRNA expression were assessed in 73 human liver samples. Liver histology of 49 samples was further characterized into normal (n=33) and NAFLD (n=16). Liver miRNome and transcriptome were significantly associated with HFC and utilized to (a) build miR-mRNA association networks in NAFLD and normal livers separately based on the potential miR-mRNA targeting and (b) conduct pathway enrichment analyses. We identified 62 miRs significantly correlated with HFC (p < 0.05 with q < 0.15), with miR-518b and miR-19b being most positively and negatively correlated with HFC, respectively (p < 0.008 for both). Integrated network analysis showed that six miRs (miRs-30b*, 612, 17*, 129-5p, 204 and 20a) controlled ~ 70% of 151 HFC-associated mRNAs (p < 0.001 with q < 0.005). Pathway analyses of these HFC-associated mRNA revealed their key effect (p<0.05) in inflammation pathways and lipid metabolism. Further, significant (p<2.47e-4, Wilcoxon test) reduction in degree of negative associations for HFC-associated miRs with HFC-associated mRNAs was observed in NAFLD as compared to normal livers, strongly suggesting highly dysfunctional miR-mRNA post-transcriptional regulatory network in NAFLD. Our study makes several novel observations which provide clues to better understand the pathogenesis and potential treatment targets of NAFLD. My second project is based on uncovering important players of post-transcriptional regulation (RBPs) and how they are associated with age and gender during healthy liver development. For this study, we performed an association analysis focusing on the expression changes of 1344 RNA Binding proteins (RBPs) as a function of age and gender in human liver. We identify 88 and 45 RBPs to be significantly associated with age and gender respectively. Experimental verification of several of the predicted associations in the mouse model confirmed our findings. Our results suggest that a small fraction of the gender-associated RBPs (~40%) are likely to be up-regulated in males. Altogether, these observations show that several of these RBPs are important developmentally conserved regulators. Further analysis of the protein interaction network of RBPs associated with age and gender based on the centrality measures like degree, betweenness and closeness revealed that several of these RBPs might be prominent players in liver development and impart gender specific alterations in gene expression via the formation of protein complexes. Indeed, both age and gender-associated RBPs in liver were found to show significantly higher clustering coefficients and network centrality measures compared to non-associated RBPs. The compendium of RBPs and this study will help us gain insight into the role of post-transcriptional regulatory molecules in aging and gender specific expression of genes.