The basic unit of eukaryotic chromatin is the nucleosome core, which contains 147 base pairs of DNA wrapped around an octamer of core histone proteins. Linker histones bind through their globular domain at the nucleosome dyad and to internucleosomal DNA through their C-terminal basic tail. The Saccharomyces cerevisiae linker histone homologue, Hho1p, contains two domains, GI and GII, that have sequence similarity to the globular domain of the canonical linker histone H1. The individual domains of Hho1p differ in their structural and functional properties, for example in 10 mM sodium phosphate GI is folded while GII exists as two species: folded and 'unfolded'. In Chapter 2 the structure of the second globular domain of Hho1p, GII, is further investigated. NMR studies indicate residual structure in the 'unfolded' form of GII, especially at the start of helices I and III. Chapter 3 considers the structural roles of Hho1p within chromatin. Semi-quantitative Western blotting is used to measure the abundance of Hho1p relative to nucleosomes in yeast. Analysis of reconstituted nucleosome arrays containing NGIL (Hho1p with the second globular domain removed) are indistinguishable from those containing full-length Hho1p, in gel-based assays and by analytical ultracentrifugation, suggesting the GII domain may not have a major role in chromatin compaction. Chapter 4 focuses on the interaction of Hho1p with chromatin proteins. Chemical cross-linking and gel filtration indicate that Hho1p does not interact significantly with the putative HMGB1 homologues Hmo1p and Nhp6ap in vitro. Hho1p and Htz1p, the yeast histone H2A.Z subtype, do not appear to interact directly in co-immunoprecipitation and chemical cross-linking assays, while chromatin immunoprecipitation studies show no evidence of colocalisation across the ADH2 and PHO5 genes. Hho1p and Sir2p cross-link in solution, but purification difficulties precluded further investigation. The effect of phosphorylation on the interaction of Hho1p and related truncation proteins with DNA and chromatin are investigated in Chapter 5. Phosphorylation reduces their affinity for linear DNA, but has different effects on the binding to four-way junction DNA for Hho1p and NGIL, compared with LGII (the linker region and GII domain of Hho1p). Phosphorylation has no obvious effect on the affinity of these proteins for chromatin in sucrose gradient centrifugation assays. NMR spectroscopy studies show that the linker region is mostly unstructured, with a short region showing some α-helical character. Phosphorylation of the linker domain changes its structural character.

Doctor of Philosophy / Department of Biochemistry and Molecular Biophysics / Jianhan Chen / Intrinsically disordered proteins (IDPs) are functional proteins that lack stable tertiary structures under physiological conditions. IDPs are key components of regulatory networks that dictate various aspects of cellular decision-making, and are over-represented in major disease pathways. For example, about 30% of eukaryotic proteins contain intrinsic disordered regions, and over 70% of cancer-associated proteins have been identified as IDPs. The highly heterogeneous nature of IDPs has presented significant challenge for experimental characterization using NMR, X-ray crystallography, or FRET. These challenges represent a unique opportunity for molecular mod- eling to make critical contributions. In this study, computer simulations at multiple scales were utilized to characterize the structural properties of unbound IDPs as well as to obtain a mechanistic understanding of IDP interactions. These studies of IDPs also reveal significant limitations in the current simulation methodology. In particular, successful simulations of biomolecules not only require accurate molecular models, but also depend on the ability to sufficiently sample the com- plex conformational space. By designing a realistic yet computationally tractable coarse-grained protein model, we demonstrated that the popular temperature replica exchange enhanced sampling is ineffective in driving faster reversible folding transitions for proteins. The second original contribution of this dissertation is the development of novel simulation methods for enhanced sampling of protein conformations, specifically, replica exchange with guided-annealing (RE-GA) method and multiscale enhanced sampling (MSES) method. We expect these methods to be highly useful in generating converged conformational ensembles.

Calcineurin (CaN) is a highly regulated Ser/Thr protein phosphatase that plays critical roles in learning and memory, cardiac development and function, and immune system activation. Alterations in CaN regulation contribute to multiple disease states such as Down syndrome, cardiac hypertrophy, Alzheimer’s disease, and autoimmune disease. In addition, CaN is the target of the immunosuppressant drugs FK506 and cyclosporin A. Despite its importance, CaN regulation is not well understood on a molecular level. Full CaN activation requires binding of calcium-loaded calmodulin (CaM), however little is known about how CaM binding releases CaN’s autoinhibitory domain from the active site. Previous work has demonstrated that the regulatory domain of CaN (RD) is disordered. The binding of CaM to CaN results in RD folding. Folding of the RD in turn causes the autoinhibitory domain (AID) located C-terminal to the RD to be ejected from CaN’s active site. This binding-induced disorder-to-order transition is responsible for the activation of CaN by CaM. In this work, we explore the nature of the disorder in the RD and its transition to an ordered state, demonstrating that the RD exists in a compact disordered state that undergoes further compaction upon CaM binding. We also demonstrate that a single CaM molecule is responsible for binding to and activating CaN. Finally, we determine that the CaM binding to CaN induces an amphipathic helix (the distal helix) C-terminal to the CaM binding region. The distal helix undergoes a hairpin-like chain reversal in order to interact with the surface of CaM, resulting in the removal of the AID from CaN’s active site. We employ site-directed mutagenesis, size-exclusion chromatography, protein crystallography, circular dichroism spectroscopy, fluorescence anisotropy and correlation spectroscopy, and phosphatase activity assays to investigate the ordering of CaN’s regulatory domain, the stoichiometry of CaN:CaM binding, and the impact of the distal helix on CaM activation of CaN.

The work presented in this thesis has contributed with knowledge to several aspects of protein-protein interaction involving PDZ domains. A substantial amount of our proteome contains regions that are intrinsically disordered but fold upon ligand interaction. The mechanism by which disordered regions bind to their ligands is one important piece of the puzzle to understand why disorder is beneficial. A region in the PDZ domain of nNOS undergoes such a disorder-to-order transition to form a b-sheet in the binding pocket of its partner. By studying the kinetics of interaction, in combination with mutations that modulate the stability of the aforementioned region, we demonstrate that the binding mechanism consists of multiple steps in which the native binding interactions of the b-sheet are formed cooperatively after the rate-limiting transition state. These mechanistic aspects may be general for the binding reactions of intrinsically disordered protein regions, at least upon formation of β-sheets.               The second part of this thesis deals with the engineering of proteins for increasing affinity in protein-protein interaction. Infection by high-risk human papillomavirus (hrHPV) can lead to cancer, and the viral E6 protein is an attractive drug target. E6 from hrHPV natively interacts with the well-characterized PDZ2 domain in SAP97, which we used as a scaffold to develop a high affinity bivalent binder of hrHPV E6. We initially increased PDZ2's affinity for E6 6-fold, but at the cost of decreased specificity. Attaching a helix that binds E6 at a distant site, increasing the affinity another14-fold, completed the design.             The final work of this thesis investigates if binding studies conducted with isolated PDZ domains is representative of the full-length proteins they belong to. It has been suggested that ligand binding in PDZ domains can be influenced by factors such as adjacent domains and interactions outside of the binding pocket. We studied these aspects for the three PDZ domains of PSD-95 and found that they on the whole function in an independent manner with short peptides as ligands, but that interactions outside of the PDZ binding-pocket may be present. The representative length of the PDZ interaction partner should therefore be considered.

No description available.

No description available.

Inhibition of the intrinsically disordered proteins (IDP) is a recognized issue in drug research. Standard approaches, based on key-lock model, cannot be used in the absence of rigid structure and defined active site. Here a basic helix-loop-helix leucine zipper (bHLHZip) domain of c-Myc was studied, which is intrinsically disordered and prone to aggregation. Chemical denaturation of proteins is a widely accepted technique to study protein folding, but here this methodology was applied to IDP, observing its effect on the structural ensemble of c-Myc by NMR spectroscopy. Nonlinear chemical shift changes indicated cooperative unfolding of the helical structure of the part of the leucine zipper domain in parallel with the melting of the N-terminal helix. Paramagnetic relaxation enhancement (PRE) was used to probe long-range structure and revealed presence of long-range contacts. The following search for inhibitors can be directed to the search for ligands, locking c-Myc in its more compact conformation. Protein self-association is a problem typical for IDPs and intrinsic process for all proteins at high concentrations. It leads to increased viscosity, gelation and possible precipitation, which cause problems in protein manufacturing, stability and delivery. If protein drugs require high dosing, special approaches are needed. At high concentrations proteins experience conditions close to the crystal state, therefore interactions in solution could potentially coincide with crystal lattice contacts. A range of diverse methods is used to study this process, but the complexity of the mechanism makes it hard to build a reliable model. Here, the self-association of streptococcal Protein G (PrtG) was studied using Nuclear Magnetic Resonance (NMR) spectroscopy in solution. The properties of protein-protein interactions at high concentration, up to ~ 160 mg/ml, were studied at residue-level resolution. Residue specific information on protein dynamics was obtained using 15N relaxation measurements. The experiments were carried out at multiple concentrations. Variation in the rotational correlation time over these concentrations showed changes in the protein dynamics, which indicated weak protein-protein interactions occurring in solution. Pulsed-field gradient NMR spectroscopy was used to monitor translational diffusion coefficients in order to estimate the degree of protein self-association. Oligomer formation was also monitored by looking at variations in 1H and 15N amide chemical shifts. Better understanding of protein self-association mechanisms under different conditions could assist in developing methods to reduce the level of reversible protein self-association in solution at high protein concentrations.

Intrinsically disordered proteins (IDPs) do not form stable tertiary structures like their ordered partners. They exist as heterogeneous ensembles that fluctuate over a time scale. Intrinsically disordered regions and proteins are found across different phyla and exert crucial biological functions. They exhibit transient secondary structures in their free state and become folded upon binding to their protein partners via a mechanism called coupled folding and binding. Some IDPs form alpha helices when bound to their protein partners. We observed a set of cancer associated IDPs where the helical binding segments of IDPs are flanked by prolines on both the sides. Helix-breaking prolines are frequently found in IDPs flanking the binding segment and are evolutionarily conserved across phyla. Two studies have shown that helix flanking prolines modulate the function of IDPs by regulating the levels of disorder in their free state and in turn regulating the binding affinities to their partners. We aimed to study if this is a common phenomenon in IDPs that exhibit similar pattern in the conservation of helix flanking prolines. We chose to test the hypothesis in c-Myb-KIX : IDP-target system in which the disordered protein exhibits high residual helicity levels in its free state. c-Myb is a hematopoietic regulator that plays a crucial role in cancer by binding to the KIX domain of CBP. Studying the functional regulation of c-Myb by modulating the disorder levels in c-Myb and in IDPs in general provides a better understanding of the way IDPs function and can be used in therapeutic strategies as IDPs are known to be involved in regulating various cellular processes and diseases. To study the effect of conserved helix flanking prolines on the residual helicity levels of c-Myb and its binding affinity to the KIX domain of CBP, we mutated the prolines to alanines. Mutating prolines to alanines increased the helicity levels of c-Myb in its free state. This small increase in the helicity levels of a highly helical c-Myb showed almost no effect on the binding affinity between cMyb and KIX. We hypothesized that there is a helical threshold for coupled folding and binding beyond which helicity levels of the free state IDP have no effect on its binding to their ordered protein partner. To test this hypothesis, we mutated solvent exposed amino acid residues in c-Myb that reduce its overall helicity and studied its effect on the binding affinity between c-Myb and KIX. Over a broad range of reduction in helicity levels of the free state did not show an effect on the binding affinity but beyond a certain level, decrease in helicity levels showed pronounced effects on the binding affinity between c-Myb and KIX.

<p>The cell cycle is one of the fundamental processes in all living organisms, and all cells arise from the division of existing cells. To better understand the regulation of the cell cycle, synchrony experiments are widely used to monitor cellular dynamics during this process. In such experiments, a large population of cells is generally arrested or selected at one stage of the cycle, and then released to progress through subsequent division stages. Measurements are then taken in this population at a variety of time points after release to provide insight into the dynamics of the cell cycle. However, due to cell-to-cell variability and asymmetric cell division, cells in a synchronized population lose synchrony over time. As a result, the time-series measurements from the synchronized cell populations do not accurately reflect the underlying dynamics of cell-cycle processes.</p><p>In this thesis, we introduce a deconvolution algorithm that learns a more accurate view of cell-cycle dynamics, free from the convolution effects associated with imperfect cell synchronization. Through wavelet-basis regularization, our method sharpens signal without sharpening noise, and can remarkably increase both the dynamic range and the temporal resolution of time-series data. Though it can be applied to any such data, we demonstrate the utility of our method by applying it to a recent cell-cycle transcription time course in the eukaryote <italic>Saccharomyces cerevisiae</italic>. We show that our method more sensitively detects cell-cycle-regulated transcription, and reveals subtle timing differences that are masked in the original population measurements. Our algorithm also explicitly learns distinct transcription programs for both mother and daughter cells, enabling us to identify 82 genes transcribed almost entirely in the early G1 in a daughter-specific manner.</p><p>In addition to the cell-cycle deconvolution algorithm, we introduce <italic>DOMAIN</italic>, a protein-protein interaction (PPI) network alignment method, which employs a novel <italic>direct-edge-alignment</italic> paradigm to detect conserved functional modules (e.g., protein complexes, molecular pathways) from pairwise PPI networks. By applying our approach to detect protein complexes conserved in yeast-fly and yeast-worm PPI networks, we show that our approach outperforms two widely used approaches in most alignment performance metrics. We also show that our approach enables us to identify conserved cell-cycle-related functional modules across yeast-fly PPI networks.</p> / Dissertation

in numerous disease states, including cancers and neurodegenerative diseases. All proteins are dynamic in nature, occupying a range of conformational flexibilities. This inherent flexibility is required for their function, with ordered proteins and IDPs representing the least flexible, and most flexible, respectively. As such IDPs possess little to no stable tertiary or secondary structure, they instead form broad ensembles of heterogeneous structures, which fluctuate over multiple time scales. Although IDPs often lack stable secondary structure they can assume a more stable structure in the presence of their binding partners in a coupled folding binding reaction. The phenomenon of the dynamic behavior of IDPs is believed to confer several functional advantages but remains poorly understood. To that end the dynamic and structural properties of a family of IDPs - p53 transactivation domains (TAD) was measured and compared with the sequence divergence. Interestingly we were able to find stronger correlations between the dynamic properties and the sequence divergence than between the structure and sequence, suggesting that the dynamic properties are the primary trait being xiii conserved by evolution. These correlations were strongest within clusters of the IDPs that correlated with known protein binding sites. Additionally, we show strong correlations between the several available disorder predictors and the backbone dynamics of this family of IDPs. This indicates the potential of predicting the dynamic behavior of proteins, which may be beneficial in future drug design. The limited number of atomic models currently determined for IDPs hampers understanding of how their amino acid sequences dictate the structural ensembles they adopt. The current dearth of atomic models for IDPs makes it difficult to test the following hypotheses: 1. The structural ensembles of IDPs are dictated by local interactions. 2. The structural ensembles of IDPs will be similar above a certain sequence identity threshold. Based on the premise that sequence determines structure, structural ensembles were determined and compared for a set of homologous IDPs. Utilizing orthologues allows for the identification of important structural features and behaviors by virtue of their conservation. A new methodology of creating ensembles was implemented that broadly samples conformational space. This allowed us to find recurring local structural features within the structural ensembles even between the more distantly related homologues that were processed. This method of ensemble creation is also the first method to show convergence of secondary structural characteristics between discrete ensembles.