The main objective of the work presented in this thesis was to study the behaviour of biopolymer films with respect to their mechanical and physicochemical properties and to test hypotheses as to their molecular origins. The thesis describes four studies. In the first study the granule, paste and film properties of common starches from six different botanical sources; i.e. cassava, corn, pea, potato, rice and wheat; were characterised with the aim of identifying the main factors which affect the properties of the starch films.

The 3D structure of the Na+-hydantoin transport protein, Mhp1, in different conformational states had already been established. To elucidate the molecular mechanisms of Mhp 1 these structural data needed support from functional data. The aim of this project was to characterise functionally Mhp 1 and to combine these data with structural information to illuminate the mechanisms of ligand binding and transport. This potentially gives insights into the molecular mechanisms of structurally similar transporters of therapeutic importance, which are reviewed. The wild-type Mhp1 protein, sixteen proteins mutated in the ligand binding site and six proteins mutated in the sodium binding site of Mhpl were expressed and purified prior to biochemical and biophysical analyses. The sodium-dependent binding of L-benzylhydantoin (L-BH) to the purified proteins was investigated using spectrophototluorimetry. The uptake of radiolabelled L-indolylmethylhydantoin (L-IMH) by the Mhpl proteins expressed in whole cells was also assessed. The mutation of Gln42, Trpl17, Gln121, Gly219, Trp220 or Asn318 in the ligand binding site impaired severely ligand binding and/or transport, revealing important interactions that these residues make with the bound ligand. The mutation of Ala317 in the ligand binding site or Ser312 or Thr313 in the sodium binding site removed the sodium-dependence of ligand binding and the Ala309Asn mutant exhibited a sodium-independent high ligand binding affinity. The mutation of Met39, Ile41, Gln153, Ala222 or Asn314 had significant but less dramatic impacts. Analyses of all of the mutant phenotypes lead to proposals of how the ligand and sodium binding sites might be linked. A library of seventy five compounds was designed to probe the recognition of ligands by Mhpl and these were screened for their ability to inhibit radiolabelled L-IMH uptake by the wild-type protein. The hydantoin moiety and a hydrophobic R group attached at the 5-postion of the hydantoin were shown to be crucial in generating compounds with a reasonable binding affinity. Five compounds, DIL naphthylmethylhydantoin (DIL-NMH), D-NMH, L-NMH, bromovinylhydantoin (BVH) and iodovinylhydantoin (IVH), were discovered that were more potent than the previously identified highest affinity ligand, L- IMH. The binding of L-NMH to the purified wild-type protein was investigated by spectrophotofluorimetry and near UV synchrotron radiation CD (SRCD). The lack of significant transport of radiolabelled L-NMH by cells expressing Mhpl suggested that L- NMH may act as a competitive inhibitor. A crystal structure ofD/L-NMH solved in complex with wild-type Mhpl, by collaborators, showed that the extracellular helix 10 'thin' gate did not occlude the ligand binding site fully, suggesting that closure of the extracellular 'thin' gate may be a prerequisite for larger conformational changes to convert Mhp1 from an outward- facing to an inward-facing structure.

A blue shift in the CH stretching vibration of formic acid cyclic dimer of 6.6 ern" (symmetric) and 3 ern" (antisymmetric) is observed by high resolution Raman (symmetric) and infrared (antisymmetric) spectroscopy. This is corroborated by theoretical ab initio calculations where blue shifts in the CH stretching vibration of 12.79 ern" (symmetric) and 10.26 ern" (antisymmetric) are calculated (CP corrected MP2/6-311++G(d,p) level of theory). This is unusual due to the CH bond not playing a direct part in the bonding of the dimer. The electric dipole moment derivative curve with respect to bond length of the CH bond in formic acid is found to be unusual. The equilibrium bond length is on the negative gradient side of the maximum of the dipole, and this has been used to explain interesting behaviour observed, including the blue shift of the CH stretching vibration and how the contribution of electrostatics to the interaction energy can cause a blue shift of the stretching vibration in the spectrum. A mechanism is proposed where the electron density is transferred from the CH bond, through to the OH site where bonding does take place. This in turn causes the CH bond to have increased polarity, and therefore the bond contracts due to this interaction. Several chloroform complexes are investigated, which show either blue shifting or red shifting of the CH stretching vibration. Complexation with dimethyl ether shows an experimental red shift of the CH stretching vibration of chloroform of -1.5 ern", and a theoretical shift of -2.11 crn'. The complex of chloroform with trimethyl amine shows an experimental red shift of the CH stretching vibration of chloroform of -54 ern", and a theoretical prediction of -79.51 ern". Both of these complexes show a 1: 1 stoichiometric equation. The chloroform self dimer shows blue shifts in the CH stretching vibration, calculated to be 2.1 and 8.8 ern", experimental results are currently inconclusive. Morokuma Kitaura energy decomposition has been used to understand the energy contributions to intermolecular bonding. Electrostatic interaction and exchange repulsion have been shown to be the main contributions to bonding, but some unusual cases, for example the CH bond of formic acid cyclic dimer, have shown electrostatics to cause a blue shift. - 3 - A tuneable stimulated Raman photoacoustic spectroscopy (PARS) set up has been further applied to the trace detection of H2' and has achieved a detection limit of 6.69 ppm by volume. A non-dispersive Raman shifter method has also been investigated as a simpler alternative to the tuneable PARS set up and has achieved a less sensitive detection limit of 108 ppm by volume. Methane has also been detected qualitatively via this method, using the Raman shifter as a source of infrared light.

No description available.

Biology can achieve phenomenal information storage/processing capabilities from just four hydrogen bonding molecules. The orthogonal self-sorting of these four hydrogen bonding motifs is achieved alongside the hydrophobic collapse of the DNA backbone. Orthogonal self-sorting without the aid of a polymer backbone is dicult to achieve, as shown by the minimal examples of self-sorting hydrogen bonding motifs available in the literature. This thesis describes the design, synthesis and binding studies of molecules that are capable of orthogonal self-sorting, without a preorganising backbone and begins to use them in a signalling cascade. In Chapter 2 ureidoimidazole, a conformer independent triple hydrogen-bond array, is introduced. Studies were carried out to investigate what effect, if any, preorganisation using intramolecular hydrogen bonding had on the binding affnity of triply hydrogen bonded complexes. Chapter 3 investigates another factor that can effect the binding affnity of complexes,the remote substituent effect. Two series of complementary molecules were synthesised so that they contained a variety of electron donating/withdrawing groups and the effect that these had on the binding affnity of the complex was measured. Chapter 4 describes a novel quadruple hydrogen-bond array, which was designed to interact strongly with its complementary partner. It was found that a combination of effects (differences in geometry and undesired conformers being favoured) lead to a low binding affnity being observed. Chapter 5 begins to investigate non-linear arrays, however none of those proposed were able to form heterodimers. Therefore a self-sorting system was assembled using a triple and quadruple hydrogen-bonded array. High fidelity interactions were achieved, even though it was possible to form undesired complexes. These undesired molecular interactions were exploited in Chapter 5, where a signalling cascade is described. Careful planning of the order in which to add the molecules can give different routes in which to achieve the self-sorting system, each with it's own fidelity trace. A photolabile tag was introduced to one of the molecules so that a photosensitive system could be achieved.

No description available.

Molecular Dynamics (MD) and Quartz Crystal Microbalance (QCM) techniques can provide unique insights into what drives protein-ligand association. The major urinary protein (MUP) binds small ligands in a deeply buried hydrophobic pocket. Detailed calorimetric studies have shown that ligand binding is driven by enthalpic effects, not entropic effects [1]. Previous studies have shown that this is due to 'dewetting' of the binding site cavity even in the absence of ligands, and have also characterised the complex changes in molecular flexibility that accompany ligand binding-features that may be correlated with NMR data [2]. Recent MD revealed the hydration effects of apo-MUP and also shown where certain regions of MUP become more flexible upon ligand binding. They have also shown a water molecule remains close to the tyrosine in the binding pocket [2]. In our current MD studies and OCM experiments we have used wild type and 2 different mutants of MUP to study the binding effects of the ligand IBM. The first mutant has an OH group removed from the binding site of MUP (i.e. tyrosine to phenylalanine (Y120F)). The second mutant has an extra OH group in the binding site (i.e. alanine to serine (A103S)). For all three systems the hydration and flexibility upon ligand binding has been analysed. The hydration analysis from MD reveal (from radial distribution curves and hydration density maps) there is a small density of water that remains even without the presence of the ligand for the WT MUP whereas a larger density of water remains in the binding cavity of the A103S hydrophilic MUP simulation. The results are based on the average structure generated from the 1 mus simulations. The Y120F MUP simulations reveal that there is no water molecules present in the binding cavity. However, as protein molecules are very dynamic in nature, water molecules are observed to hop in and out of the binding pockets for both mutant MUP (but not WT MUP) simulations over the 1 mus simulations. On the other hand the experimental QCM results reveal that on ligand binding no water loss is observed for Y120F mutant MUP whereas A103S and WT MUP have about 2 water molecules which are lost in the binding cavity. The flexibility results from the MD simulations reveal that WT MUP have some residues which increase in flexibility whilst other residues which decrease in flexibility on ligand binding. However, the Y120F hydrophobic MUP show an overall decrease in flexibility whereas the A103S MUP shows an overall increase in flexibility on ligand binding. In contrast the experimental OCM and AFM results reveal that there is an increase in flexibility on ligand binding to all 3 different types of MUP molecules. The experimental and the simulation data have shown a variation in results but it is to be noted that the results cannot be directly compared as the analytical experiments are a surface based techniques whereas the MD simulations do not involve a surface. However, the contrast observed between computer simulation and experiments has revealed important information on the ligand binding effects on MUP. [1] Bingham, R.J., J.B.C. Findlay, S.Y. Hsieh, A.P. Kalverda, A. Kjeliberg, C. Perazzolo, S.E.V. Phillips, K. Seshadri, C.H. Trinh, W. B. TurnbulI, G. Bodenhausen, and S.W. Homans. 2004. Thermodynamics of binding of 2-methoxy-3-lsopropylpyrazlne and 2- methoxy-3-lsobutylpyrazine to the major urinary protein. J. Am. Chem. Soc. 126:1675-1681. [2] Barratt, E., R.J. Bingham. D.J. Warner, C.A. Laughton, S.E.V. Phillips, and S.W. Homans. 2005. Van der Waals interactions dominate ligand-protein association in a protein binding site occluded from solvent water. J. Am. Chem. Soc. 127:11827-11834.

En chémoinformatique, les molécules peuvent être décrites en retranscrivant de façon chiffrée leurs caractéristiques. Ceci permet de pouvoir comparer in silico deux molécules pour trouver de nouveaux composés potentiellement intéressants ou pour prédire leur activité. Nous avons développé des descripteurs combinant les informations structurales des composés avec leurs propriétés chimiques. Les 2D-FPTs (triplets pharmacophoriques flous bidimensionnels) se basent sur l'énumération en trois points de caractéristiques pharmacophoriques, combinée à la prise en compte des distances topologiques entre chacune de ces caractéristiques. De plus, les 2D-FPTs introduisent 2 améliorations : La projection floue des triplets d'atomes sur les triplets pharmacophoriques de base (pour mimer la tolérance naturelle des récepteurs par rapport à leurs ligands), et le marquage par pharmacophores dépendants du pKa (qui prend en compte l'équilibre protéolytique). Une nouvelle formule de calcul de similarité est introduite, qui prend en compte l'absence simultanée d'un triplet comme moins contraignante et probante qu'une présence simultanée. Le développement des triplets est détaillé, puis plusieurs applications des triplets sont étudiées. L’échantillonneur SQS (Stochastic QSAR Sampler) est utilisé pour comparer la capacité des 2D-FPTs à construire des modèles de relation structure-activité (QSAR) par rapport à d’autres descripteurs. Ensuite, les 2D-FPTs sont comparés en profondeur avec d’autres descripteurs dans des études QSAR. Enfin, ils sont utilisés pour construire des cartes auto-organisatrices afin de les utiliser pour accélérer les recherches par similarité dans une base de données. / In chemoinformatics, compounds can be described by encoding their features as numeric data. This allows in silico comparisons of two molecules (ie calculating distance between two vectors) in order to discover new druglike compounds or to predict their activity. We developed our descriptors, combining structural information but also chemical properties and adding chemically-relevant improvements. The 2D-FPTs (bidimensional fuzzy tricentric pharmacophores) are based on the enumeration of 3 pharmacophoric features points, combined with the topological distances between each of these features. Furthermore, the 2D-FPTs introduce two improvements: the fuzzy mapping of molecular triplets on basis pharmacophoric triplets (this mimics the natural tolerance of receptors towards their ligands), and pKa-dependant pharmacophore flagging (which takes into account the proteolytic equilibrium). Moreover, a new similarity calculation formula is introduced, which accounts for the simultaneus absence of a triplet as a less-constraining indicator of similarity than its simultaneous presence. The fuzzy triplets development is detailed, then, several applications are studied. The SQS (Stochastic QSAR Sampler) is used to compare the ability of 2D-FPTs to build QSAR models to other descriptors. Then, the use of FPTs in QSAR studies was deeply examined and compared with existing descriptors. Finally, they are used to build self-organizing maps (SOM), in order to use the maps as an attempt to accelerate similarity searches in a database.

Encapsulation is a wide spread industrial technique. The use of biopolymers to form capsules has an obvious advantage for food, biomedical or personal care applications, as they can be made of food grade and biodegradable materials. Polyelectrolyte complexes (PEC) have been widely used to form capsules by the multilayer technique, usually coating a solid core with a layer by layer deposition technique involving various steps of deposition of polyelectrolyte of opposite charge. In this thesis we focused on capsules made with a one step technique where a liquid drop containing a concentrated solution of the polycation is dropped into a dilute solution of the polyanionic solution. Chitosan was the natural polycation used, due to its interesting properties such as biodegradability, biocompatibility or muco-adhesion, and because it is one of the rare cationic polysaccharides which is safe to be used in food applications. The simple one step technique used to form these capsules brings an advantage in terms of manufacturing. The drop deformation due to the impact of droplets at the surface of a liquid bath was studied and the conditions for these droplets to detach from the surface and disperse into the solution were examined. It was found that a balance between the kinetic energy and the absorbed viscous energy of the drop during impact is necessary for the formation of capsules by this technique. Initial work was performed on capsules made with chitosan and polyanionic biopolymers such as alginate or pectins. However, these biopolymers are usually provided with a polydispersed molecular weight distribution and impurities. Polyacrylic acid was thus chosen to study the effect of molecular weight, charge density, and pH on the microstructure, mechanical, swelling and release behaviour of these capsules. The microstructure of the capsules was observed by cryo-scanning electron microscopy, and the effect of molecular weight charge density of the polyelectrolytes on the shell thickness were shown and related to the mechanical and release properties of the capsules. It was shown that the molecular weight has an important role as it determines the thickness of the shell. The charge density of the polyelectrolytes, which can be controlled by the pH and ionic strength of the solution, dictates the density of the shell. The study of the release properties of these capsules showed that they could be completely impermeable to molecules with a molecular weight higher than 2000 g/mol, and surprisingly, the capsules could withstand a wide range of pH extending beyond the pH where electrostatic interactions occur, as they only dissolved at extreme pHs of 2 and 12. Chitosan was also used to make swellable hydrogels microparticles, by cross-linking with glutaraldehyde. The swelling behaviour in gastro-intestinal environment was studied in vitro and in vivo using Magnetic Resonance Imaging techniques.

In all organisms studied so far, the recognition of a functional poly(A) site is essential for RNA polymerase II termination at the end of nearly all genes transcribed by this enzyme (Whitelaw and Proudfoot, 1986; Guo et al., 1995; Birse et al. 1997). A number of eukaryotes have some of their genes organised in polycistronic structures which resemble bacterial operons (Davis and Hodgson, 1997; Ganot et al., 2004; Spieth et al. 1993), and in C. elegans, approximately 20% of all genes are contained within these operon-like structures (Blumenthal et al., 2002). Here, functional poly(A) sites will be synthesised and recognised by RNA polymerase II at the end of each gene within the operon, however termination of the polymerase only occurs at the final gene of the polycistronic transcription unit In these studies, we analyse the halting of RN A polymerase II transcription at the end of monocistronic genes and furthermore observe how premature RNA polymerase II termination is prevented during polycistronic transcription in the nematode C. elegans. We predominantly make use of reverse transcriptase PCR-based techniques to examine these mechanisms. We show that a large increase in pre-mRNAs stretching into the 3' flank of genes can be detected in worms depleted of the riboexonuclease XRN-2, indicating that this enzyme may have a possible role in RNA pol II termination and 3' end formation in C. elegans. Furthermore, we provide evidence that the polymerase can read into telomeric structures in the nematode. Also, we demonstrate that an RNAi-mediated knockdown of the UI-70K subunit of the UI snRNP causes a drop in polycistronic transcripts, providing a link between cis- splicing and the prevention of premature RNA polymerase II termination at operon-internal poly(A) sites. Finally, we illustrate that operon-internal poly(A) sites are capable of directing efficient 3' end formation outside of a polycistronic background. Together, these findings provide valuable insights into the mechanisms involved in directing or preventing premature RNA polymerase II transcription termination at C. elegans poly(A) sites.