Un composé de type N-aryle pyridinium (PF) a montré un réel potentiel en tant qu’inducteur chimique de la maladie de Parkinson (MP). Son administration in vitro et in vivo a permis d’induire l’apparition des marqueurs cellulaires caractéristiques de la MP mais suivant une cinétique d’action beaucoup plus lente que les inducteurs parkinsoniens actuellement disponibles. Ces travaux préliminaires ont suggéré l’existence d’un lien entre le motif N-aryle pyridinium et la capacité pour la molécule qui le porte à induire chimiquement la MP. Ces résultats ont dès lors laissé envisager la préparation de composés structurellement inspirés, capables de mimer plus fidèlement les mécanismes biochimiques impliqués lors de l’apparition de la MP. Partant du même composé de départ – un dérivé de l’aniline – deux voies de synthèse ont été suivies et ont permis d’aboutir à des N-aryle pyridiniums diversement fonctionnalisés. La première voie d’accès, exploitant la réaction de Zincke, a permis d’aboutir à des composés ne portant pas de groupement hydroxyle sur le noyau pyridinium. La deuxième voie d’accès, comprenant une étape-clé d’activation anodique et inspirée de la préparation du PF, a permis d’aboutir à des N-aryle pyridiniums 3-hydroxylés. Pour ces deux voies de synthèse, la nature du composé de départ a contraint à la modification des protocoles opératoires décrits dans la littérature afin de pouvoir accéder aux molécules ciblées. L’évaluation toxicologique de ces N-aryle pyridiniums a donné lieu à une étude de relation structure/activité. Les résultats de cette étude ont permis d’identifier les éléments structuraux permettant à la fois l’inhibition du complexe I de la chaine respiratoire mitochondriale et l’induction d’une neurotoxicité in vitro. Il a également été possible d’identifier, pour la première fois, un paramètre permettant la modulation de la pénétration cellulaire dopaminergique et par suite, de l’activité de ces inducteurs chimiques de la MP / A N-aryl pyridinium derivative (PF) has shown promising results as a chemical inducer of Parkinson's disease (PD). Its in vitro and in vivo administration has induced the appearance of characteristical cellular markers of PD but following a much slower kinetics than currently available PD inducers. These preliminary studies have suggested the existence of a link between the N-aryl pyridinium moiety and the ability of the molecule bearing it to chemically induce PD. These results have therefore led to the preparation of structurally related compounds, able to mimic more closely the biochemical mechanisms involved in the onset of PD.Using the same starting material - an aniline derivative - two synthetic pathways have been followed and have led to diversely functionalized N-aryl pyridiniums. The first pathway, based on the Zincke reaction, afforded compounds without hydroxyl group on the pyridinium ring. The second pathway, including a key step of anodic activation and inspired the preparation of PF, led to 3-hydroxy-N-aryl pyridiniums. For both synthetic routes, the nature of the starting material has implied the modification of previously protocols described to gain access to the targeted molecules. The toxicological evaluation of these N-aryl pyridiniums gave rise to a structure / activity relationship study. The results of this study have led to the identification of structural elements allowing both the inhibition of complex I of the mitochondrial respiratory chain and the induction of in vitro neurotoxicity. It was also possible to identify for the first time, a parameter for the modulation of dopaminergic cell penetration and hence the activity of these chemical inducers of PD

L’objectif de cette thèse, réalisée en collaboration avec la société Mérial, concerne la santé animale et s’oriente vers la synthèse de molécules actives à visée préventive ou thérapeutique. Deux sujets distincts sont abordés. Dans un premier temps, le but recherché est la synthèse d’une série de molécules présentant une activité anti-coccidienne pour le traitement préventif des poulets. Quarante-trois composés originaux, issus d’une synthèse courte à deux ou trois étapes, ont été préparés par hémi-synthèse de la monensine. Quinze molécules ont été testées in vitro et trois d’entre elles montrent une activité importante vis-à-vis du parasite Eimeria tenella. Dans un second temps, nous nous sommes intéressés à la synthèse de nouveaux répulsifs pour un large panel d’insectes. Après avoir fait une étude bibliographique approfondie, nos recherches se sont concentrées sur la réactivité de la coumarine. Environ soixante-dix molécules ont été synthétisées par modifications fonctionnelles des 4-, 6- et 7- hydroxycoumarines et de la coumarine. La grande majorité a été testée en présence de drosophiles. Six molécules présentant une activité répulsive équivalente à celle du DEET, produit de référence, ont été identifiées / This thesis was completed in collaboration with the Animal Health Company Merial. The objective of the thesis is the synthesis of preventive and therapeutic bioactive molecules. Two separate subjects were treated. First, the aim was the synthesis of a series of molecules having an anticoccidial activity for the preventive treatment of chickens. Forty-three original compounds, issued from a short synthesis with two or three steps, were prepared from monensin by hemi-synthesis. Among them, fifteen were tested in vitro and three of them showed a significant activity against Eimeria tenella. Thereafter, we have worked on the synthesis of new repellents for a wide range of insects. After a comprehensive bibliographic study, our research was focused on the coumarin reactivity. More than seventy molecules were synthesized by structural modifications of 4-, 6- and 7-hydroxycoumarins and of coumarin. Almost all the compounds were tested in the presence of drosophila. Six molecules showing a repellent activity equivalent to DEET, used as a reference product, were identified

The main theme of this thesis is that there is a common structural basis for drugs acting on the central nervous system (CNS), and that this concept may be used to design new CNS-active drugs which have greater specificity and hence less side-effects. To develop these ideas, the biological basis of how drugs modify CMS neurotransmission is described, and illustrated using dopaminergic pathways. An account is then given of the use of physicochemical concepts in contemporary drug design. The complete conformational analysis of several antipsychotic drugs is used to illustrate some of these techniques in the development of a model for antipsychotic drug action. After reviewing current structure-activity studies in several classes of CNS drugs (antipsychotics, anti-depressants, stimulants, hal1ucinogens, anticonvulsants and analgesics), a hypothesis for a common structural basis of CNS drug action is proposed- This is based on a topographical comparison of the X-ray structures of eight representative CNS-active drugs, and consists of three parts: 1.there is a common structural basis for the activity of many different CNS-active drug classes; 2. an aromatic ring and a nitrogen atom are the primary binding groups whose topographical arrangement is fundamental to the activity of these drug classes; 3. the nature and placement of secondary binding determines different classes of CNS drug activity. A four-Point model for this common structural basis is then defined using 14- CNS-active drug structures that include the original eight used in proposing the hypothesis. The coordinates of this model are: R1 (0. 3.5, 0), R2 (0, -3.5, O), N (4.8. -0.3, 1.4), and R3 (6.3, 1.3, 0), where R1 and R2 represent the point locations of a hydrophobic interaction of the common aromatic ring with a receptor, and R3 locates the receptor point for a hydrogen bond involving the common nitrogen, N. Extended structures were used to define the receptor points R1, R2 and R3, and the complete conformational space of each of the 14 molecules was considered. It is then shoun that the model may be used to predict whether a given structure is likely to show CNS activity: a search over 1,000 entries in the current Merck Index shows a high probability (82%) of CNS activity in compounds fitting the structural model. Analysis of CNS neurotransmitters and neuropeptides shows that these fit the common model well. Based on the available evidence supporting chemical evolution, protein evolution, and the evolution of neurotransmitter functions, it is surmised that the aromatic ring/nitrogen atom pharmacophore proposed in the common model supports the idea of the evolution of CNS receptors and their neurotransmitters, possibly from an aromatic amine or acety1cho1ine acting as a primaeval communicating molecule. The third point in the hypothesis trilogy is then addressed. The extensive conformation-activity analyses that have resulted in well-defined models for five separate CNS drug classes are used to map out the locations of secondary binding groups relative to the common model for anti-psychotics, antidepressants, analgesics, anticholinergics, and anticonvulsants. With this information, and knowledge derived from receptor-binding data, it is postulated that drugs having specified activity could be designed. In order to generate novel structures having a high probability of CNS-activity, a process of drug design is described in which known CNS structures are superimposed topographically using the common model as a template. Atoms regarded as superfluous may be selectively deleted and the required secondary binding groups added in predicted locations to give novel structures. It is concluded that this process provides the basis for the rational design of new lead compounds which could further be optimized for potent and specific CNS activity.

Oxide materials exhibit a wide variety of structures and properties. In particular, transition metal oxides tend to be highly stable while exhibiting a wide range of properties that can be used for numerous applications. This work focuses on investigating how the structures��� of 4d and 5d transition metal oxides influences their properties. Specifically oxides of Ru, Rh and Ir were investigated. A complete solid solution was found between isostructural Pb���Mn���O������ and Pb���Rh���O������. Pb���Rh���O������ shows a Verwey-type transition at 185 K. This transition remains with a 3 % substitution of Mn for Rh but disappears with a 4 % substitution of Mn for Rh. The structure was found to expand in the direction perpendicular to the layers of the structure, which is the c-axis, despite a contracting unit cell. Bi for Pb substitution in Pb���Mn���O������ was found to be limited as compared to in Pb���Rh���O������. Alkali metal substitution on the A-site of the orthorhombic perovskite SrRuO��� showed only low substitution levels were possible. Nonetheless, the substituted phases showed decreased ferromagnetic Curie temperatures, increased electrical resisitivity and relatively unchanged Seebeck coefficients. Thermoelectric studies of Sr[subscript 2-x]La[subscript x]CoRuO��� perovskite phases showed Sr���.���La���.���CoRuO��� with the best thermoelectric performance. This system showed possible correlations between cation ordering on the B-site and the charge carrier transport. A similar thermoelectric study of (RhV)[subscript 1+x]Ti[subscript 1-2x]O��� phases crystallizing in a disordered trirutile structure was done. Electron carriers were found to be dominant and dependent on Ti content. The electron carriers appear to become diminished at higher temperatures. Sr���IrO��� crystallizes in a K���NiF���-type structure. Effects of Ti, Fe and Co substitution for Ir were investigated. A complete Sr���Ir[subscript 1-x]Ti[subscript x]O��� solid solution was synthesized and characterized while limited solubility was found for Fe and Co substitutions. All substitutions showed a decrease in the c-cell parameter coupled with a decrease in octahedral tilting. All substitutions also showed a decrease in magnetic susceptibility and an increase in the paramagnetic effective moment was observed for Co and Fe doped samples. An incomplete solid solution was formed for Sr���Ti[subscript 1-x]Rh[subscript x]O��� phases; however effects of increased octahedral tilting with higher Rh content were observed. / Graduation date: 2013

Transition metal oxides exhibit potential in various application fields due to the special d-electrons. Solid state chemistry focuses on discovering the structure-property relationships. The work in this thesis mainly discusses compounds with hexagonal or brownmillerite-type structure and their practical properties. Hexagonal YIn[subscript 1-x]Fe[subscript x]O��� (x = 0-0.3, 0.7-1.0) phases have been prepared and characterized. All phases appear to have the ferroelectric structure known for YInO���. The color of the phases changes from yellow to orange to dark red with increasing Fe content. Magnetic measurements confirm high-spin Fe����� for all phases. Similarly, solid solution YAl[subscript 1-x]Fe[subscript x]O��� (x = 0-0.4, 0.7-1.0) phases were successfully synthesized through the sol-gel method. The Al-rich compounds present paraelectric YAlO��� structure while the Fe-rich side samples exhibit YFeO��� structure. The color of the compounds appear to be yellow with small Fe content and change to brown which has higher Fe content. Brownmillerite-type oxides Ba���In[subscript 2-x]Mn[subscript x]O[subscript 5+x] (x = 0.1-0.7) have been prepared and characterized. Magnetic measurements confirm that Mn in as prepared samples is substituting as Mn������ for all values of x with observed paramagnetic spin-only moments close to values expected for two unpaired electrons. Neutron diffraction structure refinements show Mn������ occupies tetrahedral sites for orthorhombic (x = 0.1) and tetragonal (x = 0.2) phases. For Mn ��� 0.3 samples, neutron refinements show the phases are cubic with disordered cations and oxygen vacancies. The colors of the phases change from light yellow (x = 0) to intense turquoise (x =0.1), to green (x = 0.2, 0.3) or dark green (x ��� 0.4). Solid solution Ba���In[subscript 2-x]Fe[subscript x]O[subscript 5+y] (x = 0.1-1.5) also exhibit brownmillerite-type structure. The color of the compounds appear to be green with small Fe content and change to black with higher Fe content (x ��� 0.3). Magnetic measurements and M��ssbauer spectroscopy conclude the mixed valence of Fe�����/Fe������ for all the phases. Nonstoichiometry compound YCu���.���Ti���.���O[subscript 3-��] has been prepared and characterized. Structure study indicates that oxygen vacancy is favored under the synthesis condition. This change in oxygen content was further studied in the Mn-doped system. And the effect of stoichiometric difference in the Mn-doped samples was not as obvious as the initial compound. The disorder in the cation site enhanced the tolerance of the structure in the aspect of oxygen content. The hexagonal phases LnCu���.���Ti���.���O��� (Ln = Y, Tb-Lu) phases were prepared by the traditional solid state reactions. The prepared compounds were reduced at high temperature in the reduction atmosphere created by the H���/N��� gas mixture. Study on the structure and properties changes by reduction was conducted by X-ray diffraction, optical measurement, magnetic measurement and thermalgravimetric analysis. And we observed some evidence of the presence of Cu��� in the reduced phase by these characterization methods. Solid solution YMn[subscript x]Ti[subscript y]O[subscript 3-��] (y = 0.1-0.4) was successfully prepared through conventional solid state approach. All the samples showed hexagonal structure. But the structure transition from ferroelectric P6���cm to paraelectric P6���/mmc occurred when Ti amount is higher than 0.2. Based on the neutron diffraction refinement, the lattice expanded in the ab plane but contracted along the c axis direction. / Graduation date: 2013

This thesis develops an integrated methodology based on the desirability index and QSAR models to virtually optimise molecules. Statistical and algorithmic tools are proposed to search in huge collections of compounds obtained by combinatorial chemistry the most promising ones. First, once the drugability properties of interest have been precisely defined, QSAR models are developed to mimic the relationship between those optimised properties and chemical descriptors of molecules. The literature on QSAR models is reviewed and the statistical tools to validate the models, analyse their fit and their predictive power are detailed. Even if a QSAR model has been validated and sounds highly predictive, we emphasise the importance of measuring extrapolation by the definition of its applicability domain and quantifying the prediction error for a given molecule. Indeed, QSAR models are often massively applied to predict drugability properties for libraries of new compounds without taking care of the reliability of each individual prediction. Then, a desirability index measures the compromise between the multiple estimated drugability properties and allows to rank the molecules in the combinatorial library in preference order. The propagation of the models prediction error on the desirability index is quantified by a confidence interval that can be constructed under general conditions for linear regression, PLS regression or regression tree models. This fulfills an important lack of the desirability index literature that considers it as exact. Finally, a new efficient algorithm (WEALD) is proposed to virtually screen the combinatorial library and retain the molecule with the highest desirability indexes. For each explored molecule, it is checked if it belongs to the applicability domain of each QSAR models. In addition, the uncertainty of the desirability index of each explored molecule is taken into account by gathering molecules that can not be distinguished from the optimal one due to the propagation of QSAR models prediction error. Those molecules do not have a significantly smaller desirability than the optimal molecule found by WEALD. This constitutes another important improvement in the use of desirability index as a tool to compare solutions in a multicriteria optimisation problem. This integrated methodology has been developed in the context of lead optimisation and is illustrated on a real combinatorial library provided by Eli Lilly and Company. This is the main application of the thesis. Nevertheless, as the results on desirability index uncertainty are applicable under general conditions, they can be applied to any multicriteria optimisation problem, like it often occurs in industry.

Lipids havepreviously been considered primarily as building blocks of the cell membrane, but are now also recognized as important cell signaling molecules. Lysophosphatidic acid (LPA) is a glycerophospholipid consisting of a phosphate head group, a linker region, and a lipophilic tail. LPA has earlier been shown to exert a diversity of cellular effects such as aggregation, apoptosis, contraction, migration, and proliferation. The effects of LPA are elicited by activation of its cognate G protein-coupled receptors LPA1, LPA2, and LPA3. In the present study we have used cultures of human smooth muscle cells (SMCs) and erythroleukemia cells (HEL), and isolated human platelets to characterize physiological effects of LPA compared with adrenaline and noradrenaline as well as structure-activity relationships of LPA. SMCs were isolated from biopsies of human myometrium obtained at cesarean sections. We show that cultured myometrial SMCs express multiple LPA and α2-adrenergic receptor subtypes. Treatment of SMCs with LPA and noradrenaline resulted in increases in proliferation. However, LPA elicits a much more pronounced stimulatory effect than noradrenaline. The ability to increase calcium might be one explanation why LPA is more effective. Further studies indicated that several pathways mediate the growth stimulatory effect of LPA where transactivation of epidermal growth factor receptors through matrix metalloproteinases as well as calcium/calmodulin-dependent protein kinases appears to be important. LPA enantiomers and LPA analogues were synthesized and characterized due to their capacity to increase calcium in HEL cells. Our study is the first to show that both natural (R) and unnatural (S) LPA enantiomers are capable of stimulating cells, suggesting LPA receptors are not stereoselective. Moreover, we have synthesized a LPA analogue with higher maximal effect than LPA by reducing the hydrocarbon chain length. In platelets we demonstrated that LPA is a weak calciumelevating compound which failed to stimulate aggregation. However, in combination with adrenaline, another weak platelet agonist, a complete aggregatory response was obtained in blood from some healthy individuals. These results are important since platelet activation is a key step in distinguishing normal from pathological hemostasis. Since LPA is present at high concentrations in atherosclerotic lesions, the synergistic effect of LPA and adrenaline might be a new risk factor for arterial thrombosis. / On the day of the public defence the status of the article IV was: Submitted for publication.

Infection by the hepatitis C virus (HCV) leads to inflammation of the liver, i.e. hepatitis. The acute infection often progresses to a chronic phase during which the liver function is gradually impaired. Approximately 20% of these chronic cases develop liver cirrhosis, with an ensuing increased risk of liver cancer. Global estimates of the total number of chronic cases range from 123–170 million. Yet, neither specific anti-HCV drugs nor vaccines are available. When drugs become available for daily clinical use, rapid development of drug-resistant strains is expected, making resistance an important issue. One of the most studied targets for specific anti-HCV drugs is the NS3 protease. The main objectives of the work presented in this thesis were to design and synthesise peptidomimetic inhibitors of this enzyme, and to establish the structure–activity relationships (SARs) regarding the inhibition of the wild type as well as of the known resistant variants A156T and D168V. Substituted prolines are common P2 residues in HCV NS3 protease inhibitors. To decrease the peptide character of the inhibitors, the non-coded phenylglycine was evaluated as a proline replacement in combination with known and novel P3 and P1 residues and P2 substituents. The results confirmed that phenylglycine is a promising P2 scaffold, with a possible π-stacking interaction with histidine 57 of the active site. However, to benefit from its full potential, additional optimisation is required. A 2(1H)-pyrazinone-based scaffold was introduced as P3 residue. Utilising the scope of the method developed for the pyrazinone scaffold synthesis, the phenylglycine side-chain was transferred to the scaffold. In combination with an aromatic P1 building-block, this design yielded achiral, peptidomimetic inhibitors, three times more potent than the tripeptide lead. The SARs for the inhibition of the resistant variants A156T and D168V were investigated for compounds based on either P2 proline or phenyl­glycine. It was concluded that the vulnerability of the inhibitors to alterations in the enzyme depends on the P2 and the P1 residue, not only on the P2 as previously suggested. These results provide important information for the design of a new generation of inhibitors with improved properties.

This thesis focuses on the design and synthesis of inhibitors targeting Mycobacterium tuberculosis ribonucleotide reductase (RNR) and hepatitis C virus (HCV) NS3 protease; enzymes that have been identified as potential drug targets for the treatment of tuberculosis and hepatitis C, respectively. Small peptides have been recognized as inhibitors of these enzymes. However, the use of peptides as drugs is limited due to their unfavorable properties. These can be circumvented by the development of less peptidic molecules, often referred to as peptidomimetics. When this work was initiated, only a few inhibitors targeting M. tuberculosis RNR had been identified, whereas the HCV NS3 protease was an established drug target. Therefore, early peptidomimetic design strategies were applied to inhibitors of RNR while the NS3 protease inhibitors were subjected to modifications in a later stage of development. It has previously been shown that peptides derived from the C-terminus of the small subunit of M. tuberculosis RNR can compete for binding to the large subunit, and thus inhibit enzyme activity. To investigate the structural requirements of these inhibitors, different series of peptides were evaluated. First, peptides from an N-terminal truncation, an alanine scan and a designed library were synthesized and evaluated to examine the importance of the individual amino acid residues. Then, a set of N-terminally Fmoc-protected peptides was evaluated, and it was found that the N-terminal group improved the affinity of the peptides even when the length of the compounds was reduced. Furthermore, potential inhibitors of less peptidic character were generated by the introduction of a benzodiazepine-based scaffold. To further reduce the peptidic character and investigate the binding properties of HCV NS3 protease inhibitors, a series of tripeptides incorporating a β-amino acid was synthesized. Inhibition was evaluated and docking studies were performed to understand how the structural changes affected inhibitory potency. The results illustrated the importance of preserving the hydrogen bonding network and retaining electrostatic interactions in the oxyanion hole between inhibitor and protein.

β-lactam antibiotics have long been a reliable course of treatment for bacterial infections. However, with recent increases in resistance and rising populations of immunocompromised patients new β-lactams have been synthesized and tested. The Turos laboratory has recently discovered novel β-lactams that have a mode of action distinct from penicillin and other β-lactam antibiotics as cell lysis is not observed. In the current investigations, these compounds are shown to also have antifungal properties. The rising incidence and prevalence of invasive fungal infections has become an increasing concern. The most common fungal pathogens involved in these infections are species in the genus Candida. In this study antifungal activity is observed for a wide range of N-methylthio β-lactams against C. albicans, C. tropicalis, C. keyfr, C. glabrata, C. lusitinae, C. utilis, and C. parapsilosis. The structure-activity relationship based on studies of β−lactam derivatives leaving different substituents at various positions on the lactam ring are investigated, and the minimum inhibitory concentration values determined using standard methods. In studies towards understanding the mode of action, the products of the interaction between the drug and fungal cells in a suspension were investigated using nuclear magnetic resonance spectroscopy and transmission electron microscopy. The mode of action of these new lactams seems to be similar to that observed in bacteria, involving transfer of the methylthio group to a cellular thiol.