In the field of drug design and development, medicinal chemists use a variety of tools to quickly generate a series of hit compounds controlling a specific biological target and culminating in a lead compound. Process chemists seek efficient methods to synthesize the lead compounds provided by the medicinal chemists and using readily available and inexpensive starting materials, shortcuts and simpler routes. With this in mind, we aimed to design, develop and apply chemical tools to generate hit compounds, but also developing new simpler methods to make pharmaceutically relevant compounds. In this context, this thesis has two goals. In the first part, we focused on the enzyme prolyl oligopeptidase, reviewing its involvement in neurological disorders, such as Alzheimer's disease, and the efforts of several researchers to synthesize potent, selective inhibitors resembling the natural substrates of this enzyme. We proposed another method, using small pseudopeptidic and peptidomimetic inhibitors as chemical tools to better understand the shape, size and electronics of inhibitors and generate a more potent, selective prolyl oligopeptidase inhibitor. From our series of compounds, we discovered a few potent and highly selective, covalent inhibitors, one of them pseudo-peptidic (IC50 = 3-7 nM) and the other peptidomimetic (IC50 = 20-700 nM). In the second part of this thesis, with the goal of being able to exploit sugars in medicinal chemistry, we first reviewed the methods that exist to regioselectively functionalize the various hydroxyls of hexopyranosides. We compiled these methods into a table which chemists could consult when they are seeking to perform a specific reaction on a specific sugar. We then proposed to use a hydrogen bond accepting protecting group which can direct subsequent reactions to specific sites on sugars in an effort to reduce the number of protection and deprotection steps. We applied this protecting-directing group and developed methods to regio / Lors de la conception d'un nouveau principe actif, les chimistes médicinaux disposent d'un grand nombre d'outils leur permettant de générer rapidement une série de composés ayant une action spécifique. L'optimisation du procédé chimique requiert l'utilisation et le développement de méthodes toujours plus simples, rapides et peu couteuses. Ainsi, nous avons souhaité concevoir et utiliser quelques outils chimiques permettant de résoudre certains de ces problèmes. Cette thèse présente des outils spécifiques à la chimie médicinale et la chimie des procédés. Dans un premier temps, nous nous sommes intéressés à une enzyme : la prolyl oligopeptidase. Nous avons tout d'abord répertorié un grand nombre de données, quant à son implication dans certains troubles neurologiques (ex. : maladie d'Alzheimer), ainsi que de nombreux programmes de recherche visant le développement de nouveaux inhibiteurs conçus à partir de la structure du substrat naturel. Dans l'optique de mieux comprendre l'environnement chimique du site de liaison de l'enzyme et concevoir notre propre inhibiteur sélectif, nous avons synthétisé et testé une série de composés pseudo-peptidiques et peptidomimétiques. Lors de cette étude, nous avons identifié deux inhibiteurs covalents, actifs et sélectifs : l'un pseudo-peptidique (IC50 = 3-7 nM), l'autre peptidomimétique (IC50 = 20-700 nM). Dans la deuxième partie de cette thèse, nous avons rapporté les méthodes déjà existantes permettant une fonctionnalisation régioselective des différents groupements alcools des hexopyranosidases afin de rendre plus accessible l'utilisation des sucres en chimie médicinale. La compilation de cette recherche bibliographique dans une table permettra aux chimistes de facilement sélectionner la méthode la plus adéquate. Afin de réduire le nombre d'étapes (protections et déprotections) associées à la chimie des sucres, nous avons appliqué le groupement protecteur-directeur dév

The vitamin D receptor (VDR) and histone deacetylases (HDACs) are important chemotherapeutic targets. 1,25-dihydroxyvitamin D3 (calcitriol, or 1,25D) is the natural ligand of the VDR, stimulates immune responses, and inhibits cellular proliferation in a number of cancer cell lines. Small molecule HDAC inhibitors such as suberoylanilide hydroxamic acid (SAHA, Zolinza; Merck), block angiogenesis and promote cell apoptosis and differentiation and are being investigated as clinical treatments for a wide range of cancers. This thesis describes work on three projects focusing on the VDR and HDAC. In the first project, a hybrid molecule incorporating deacetylase activity into the structure of a 1,25D aromatic analog was identified in a virtual screen using FITTED, a docking program, synthesized, and tested. Surprisingly, it was found to be inactive as an agonist of the VDR. An in silico retrospective analysis of the novel hybrid compound and other non-steroidal VDR ligands revealed several general structural requirements for VDR activity. In a second project, computational modeling techniques including molecular dynamics simulations were employed to understand the differences between two ortho-aminoanilide hybrid compounds, which led one to act as a VDR agonist while the other acted as a VDR antagonist. Finally, as the mechanisms of anticancer activity of HDAC inhibitors are not well understood and current assays for HDAC inhibitors are cumbersome and not generally applicable to all HDAC isoforms, a fluorescence polarization assay for high-throughput screening of HDAC competitive inhibitors was developed. HDAC ligands combining the structures of SAHA and fluorescein were synthesized and determined to be well-suited as a fluorescence polarization assay probe. / Le récepteur de la vitamine D (RVD) et les histones déacétylases (HDAC) sont d'importantes cibles de chimiothérapie. La 1,25-dihydroxyvitamine D3 (calcitriol, ou 1,25D), le ligand naturel de haute affinité du RVD, stimule une réponse immunitaire et supprime la prolifération cellulaire chez des nombreuses lignées de cellules cancéreuses. De petites molécules inhibitrices des HDAC telle que l'acide subéroyl hydroxamique (SAHA, Zolinza; Merck) bloquent l'angiogénèse et provoque l'apoptose cellulaire et la différentiation et font présentement l'objet d'essais cliniques pour le traitement d'une variété de cancers. Cette thèse décrit les résultats de trois projets reliés au RVD et aux HDAC. Le premier projet concerne la création d'une molécule hybride incorporant une activité déacétylase à un analogue aromatique de la 1,25D. Cette hybride a été identifiée à l'aide d'un criblage virtuel utilisant le logiciel FITTED, créée puis testée en laboratoire. Elle a été trouvée sans activité en tant qu'agoniste du RVD. Une analyse in silico de cette molécule hybride et d'autres ligands non-stéroïdiens du RDV a révélé plusieurs des contraintes structurelles nécessaires à la possession d'une activité sur le RVD. Au cours du second projet, différentes techniques de modélisation par ordinateur ont été utilisées afin d'expliquer la différence entre deux molécules hybrides ortho-aminoanilide, l'une étant un agoniste du RVD et l'autre, un antagoniste. Finalement, puisque la compréhension des mécanismes de l'activité anticancer des inhibiteurs des HDAC est déficiente et puisque les tests biologiques des inhibiteurs des HDAC sont fastidieux et généralement inapplicables à tous les isoformes des HDAC, un test utilisant la polarisation de la fluorescence pour le criblage à haut rendement des inhibiteurs compétitifs des HDAC a été développé. Des ligands des HDAC combinant les structures de SAHA et de la fluorescéine ont été créés et il a été démontré qu'ils remplissaient convenablement le rôle de sonde dans des tests utilisant la polarisation de la fluorescence.

The reactions of enolates are some of the most basic transformations in organic chemistry. While there are limited examples of catalytic asymmetric enolate alkylations, many are restricted to specific acidic substrates or require preactivation of the enolate. As a starting point for the development of a more general method, we have begun to study intramolecular alkylations of -bromo oxazolidinone imides as a prelude to developing the more difficult intermolecular process. We have found that intramolecular alkylation takes place with soft enolizing bases in the presence of catalytic amounts of magnesium perchlorate and DBU. Importantly, the reaction is subject to ligand acceleration, particularly by Pybox type ligands, with enantioselectivities of up to 63%. Under catalytic conditions there were difficulties in obtaining a consistent reaction in terms of conversion of starting material to product. Once these obstacles are addressed an intramolecular catalytic asymmetric alkylation reaction can be reported. / Les réactions des énolates font partie des transformations les plus fondamentales en chimie organique. Plusieurs exemples d'alkylations asymmétriques et catalytiques d'énolates existent; cependant ils sont limités à des cas où les substrats sont acides ou nécessitent une préactivation. L'alkylation intramoléculaire de -bromo oxazolidinone imides a été choisie comme point de départ pour le développement d'une nouvelle méthode générale. Nous avons découvert que l'alkylation intramoléculaire s'effectue en présence de quantité catalytique de perchlorate de magnésium et DBU (1,8-Diazabiciclo[5.4.0]undec-7-ene). Au cours de réactions, l'utilisation de ligands, particulièrement du type Pybox, induit une augmentation de la vitesse de réaction et permet d'atteindre une activité énantiosélective jusqu'à 63%. Certaines difficultés existent toujours : par exemple, l'obtention de résultats concluants en terme de conversion du substrat de départ en produit.

Ionic liquid-supported synthesis is a new concept in organic chemistry. This concept has been well demonstrated in this thesis including ionic-liquid-supported Swern oxidation, oligosaccharide synthesis, imidazolium oligomer synthesis and peptide synthesis. / Ionic liquid-supported Swern oxidation provided a new odorless approach to oxidize primary and secondary alcohols into aldehydes and ketones as compared to the conventional method where DMSO was used as an oxidant and the toxic and volatile compound dimethylsulfide was generated quantitatively. The new ionic liquid-supported organosulfur reagents were thermally stable, nonvolatile and odorless, which were recoverable and recyclable after Swern oxidation. / Chapter 3 of this thesis described the first example in the research field of ionic liquidsupported oligosaccharide synthesis. This new approach possesses the advantages of both conventional solution phase synthesis and solid phase-supported synthesis. It made possible the chemical assembling of glycals and the cleavage of oligosacharide from the support without need of chromatography to purify the products. / Ionic liquid-type oligomers were designed to solve the problems involved in the oneionic-unit ionic liquid-supported biopolymer synthesis because these compounds could have stronger ionic effects on support-bound molecules, which made easier the chemical assembling and purifications of large biopolymers than one-ionic-unit ionic liquidsupported organic synthesis. The thermogravimetric analysis (TGA) showed that these imidazolium oligomers have very good thermal stability. They also have great potential application in the electrochemistry. / Imidazolium oligomer-supported peptide synthesis provided a new approach to prepare peptides in a straightforward way. The amino acid coupling reactions were conducted in homogeneous solution phase, which did not require largely excess of reagents to push the reaction to completion and the purification could be performed in a simple way such as centrifugation, decantation and washing without need of special solvents. Moreover, the cleavage of the peptide from the imidazolium oligomer support was done by hydrolysis under basic conditions and the product was obtained in NMR and MS purity. Importantly, imidazolium oligomer-supported peptide block coupling proceeded very well, which provided an approach to convergently synthesize peptides. This approach is very useful for the peptide synthesis in industry especially for the synthesis of peptides with less than 20 amino acids.

This thesis is an investigation on the functionalization of aryl C-H bonds in the presence of transition-metal catalysts and oxidants.In the first part of this thesis, an oxidative amidation of 2-arylpyridine derivatives and 1-methylindoles with a variety of amides is described. Copper(I) bromide is used as catalyst and tert-butyl peroxide (TBP) is employed as oxidant. High regioselectivity of this amidation process is achieved through chelation-assisted aryl C-H activation. In the second part of this thesis, a new concept is described for aryl-aryl coupling that involves oxidative decarbonylative coupling of aryl C-H bonds and readily available aldehydes and forms the aryl-aryl union with complete control of reaction sites. This process is catalyzed by (CO)2Rh(acac), along with TBP as an oxidant.The third and final part of this thesis describes a novel ruthenium- and copper-catalyzed domino reaction between alkynols and aldehydes, which affords 5-olefinated 3,4-dihydropyran derivatives efficiently with only water as the byproduct. / Ce manuscrit de thèse présente l'étude de la fonctionnalisation de liaisons C-H aryliques catalysée par des complexes de métaux de transition en présence d'un oxydant. Dans une première partie est décrite la réaction d'amidation oxydante de dérivés de 2 arylpyridine et de 1-méthylindole par une variété d'amides. Le bromure de cuivre(I) est utilisé comme catalyseur et le peroxyde de tert-butyle (TBP) comme oxydant. Dans ces conditions, une excellente régiosélectivité est obtenue pour cette transformation grâce à la présence d'un groupe directeur qui permet l'assistance par chélation de l'activation de la liaison C-H. Au cours de la deuxième partie est abordé un nouveau concept pour la synthèse de motifs aryl-aryl, impliquant une séquence décarbonylation/couplage oxydant. Cette réaction permet la formation d'un composé biaryle à partir d'arènes et d'aldehydes aromatiques aisément accessibles avec un contrôle total des sites réactionnels. Elle est catalysée par le complexe (CO)2Rh(acac) en présence de TBP comme oxydant. La troisième et dernière partie de cette thèse présente une nouvelle réaction domino catalysée par des complexes de ruthénium et de cuivre qui permet l'obtention de 3,4 dihydropyranes substitués en position 5 à partir d'aldéhydes et d'hydroxyalcynes avec la formation d'eau comme unique sous-produit.

The investigation and development of small molecule inhibitors for the estrogen receptor is presented. Successful interpretation of its key protein-protein interactions with several coactivators is described using computational methods. Further examination using a known set of small molecule ligands failed to predict affinity for the receptor. It was determined that a computational assessment was unfeasible for the design of new ligands. The design and synthesis of SERM-HDACi (selective estrogen receptor modulator-histone deacetylase inhibitor) multiple ligands is described in detail. Multiple ligands are a single entity created to target more than one site in the body, therefore a SERM-HDACi multiple ligand will potentially target the estrogen receptor and histone deacetylases. Docking studies were used to determine three potential SERM-HDACi hybrid candidates to be synthesized in the lab. Several synthetic routes were explored to achieve a stereoselective synthesis of the 4-hydroxytamoxifen tetrasubstituted olefin core, however all attempts were unsuccessful. Nevertheless, the syntheses of three multiple ligands were accomplished as an E/Z mixture. Separation of isomeric mixtures was achieved by reverse phase HPLC. Preliminary biological results show two of the three hybrids possess estrogen receptor antagonistic properties. Stereochemistry of one isomer from each hybrid, were determined through 1D and 2D spectroscopic methods and correlate well with the biological data as the more active isomers in ERα. / Le développement de nouveaux inhibiteurs du récepteur d'oestrogène est présenté. L'interprétation des principales interactions entre le récepteur et des coactivateurs à l'aide de méthodes computationnelle est décrite. Une évaluation plus approfondie de la surface du récepteur en utilisant un ensemble connu de petites molécules n'a pas mené à des résultats encourageants pour la création d'un nouvel ensemble de composés. Il a donc fallu recourir à une approche plus élémentaire. La conception et la synthèse de ligands hybrides (à la fois modulateurs sélectifs du récepteur des oestrogènes et inhibiteurs des histones déacétylases) (SERM-HDACi) sont décrites en détail. Des études d'affinité (docking) ont été utilisées pour déterminer trois nouveaux candidats potentiels SERM-HDACi à être synthétisé. Plusieurs voies de synthèse ont été explorées pour réaliser une synthèse stéréosélective d'un alcène tétrasubstituées, mais toutes les tentatives ont échoué. Néanmoins, les synthèses des trois composés ont été réalisées comme un mélange d'alcènes E/Z qui ont été ensuite séparés par HPLC en phase inverse. Les résultats biologiques préliminaires montrent que deux de ces molécules possèdent des propriétés d'antagonistes du récepteur d'oestrogène. La stéréochimie d'un isomère de chaque hybride a été déterminée grâce à des méthodes spectroscopiques 1D et 2D et une corrélation existe entre les données biologiques et un des isomères plus actif pour le récepteur d'oestrogène α.

The unique architecture of branched oligonucleotides was exploited to study compounds that associate as two parallel duplexes with intercalating C•C+ base pairs (i-motif DNA). The formation of a branched cytosine tetrad was induced by joining the 5'-ends of a pair of pentadeoxycytidine strands with a branching riboadenosine (rA) linker. This arrangement causes the orientation of the dC strands to be parallel, and forces the formation of a C•C+ duplex that self associate into i-DNA. Presence of the i-motif in this structure is supported by thermal denaturation, native gel electrophoresis, CD, and NMR spectroscopy. / To further characterize the nature of the sugar-sugar interactions in the i-motif, branched oligonucleotides containing D-arabinocytidine and D-ribocytidine were synthesized and their association properties examined. The branched D-deoxycytidine analogue, rA(2',5'-dC 5)-3',5'-dC 5 (dV-5) served as model system. It is found that the arabinose substitution leads to hypochromic structures that are characteristic of four-stranded intercalated DNA and has little, if any, effect on the stability of the complex formed. Parallel experiments with the branched ribocytidine analogs gave very weak or no discernible UV transitions, consistent with no strand association in this case. These results are discussed in relation to expected steric interactions of oligocytidine strands within the 1-motif. / A divergent approach for the construction of branched nucleic acids was also studied. Chain assembly took place on the surface of controlled-pore glass solid support in the unconventional 5' to 3 ' direction. The branch junctures were introduced by use of N6-benzoyl-2',3'-O-bis(dimethoxytrityl)adenosine-5 '-O-(N,N-diisopropyl)-beta-cyanoethylphosphoramidite. Various dendrimers were prepared in comparable or better yields relative to the well-established convergent approach.

A general procedure for the solid-phase synthesis of branched oligonucleotide analogs, including some with similar base sequence to the yeast ( Saccharomyces cerevisiae) rp51 mRNA, is described using the automated solid-phase phosphite triester approach. Using readily available phosphoramidite reagents, branch-points were introduced by sequential removal of the phosphate (beta-cyanoethyl or methyl) and silyl protecting groups without detaching the nascent oligonucleotide from the solid-support. This then allowed the extension from the branch-point in the conventional 3',5 ', or the unconventional 5',3' directions, thus affording complete control over the base composition and direction of strand around the branch point. / The protocol was modified for the synthesis of msDNA, a DNA/RNA branched chimera. Using a combination of silyl and Fpmp groups for the protection of 2'-hydroxyl functions, a linear oligonucleotide was assembled. Branching was then introduced using the methodology developed for branched DNA. / Utilization of capillary electrophoresis (CE) in the analysis of branched nucleic acids was also touched upon. CE proved to be a very powerful technique in the separation and identification of a mixture of four branched oligonucleotides. This technique, CE, was also seen to be an attractive alternative for following the debranching reaction, thus eliminating the need for radioactively labeled substrates. / The unique structural properties of branched nucleic acids were exploited in investigation of the association properties of bNAs. By synthesizing branch molecules, capable of forming T:A/T triplexes in the parallel (Hoogsteen) or the antiparallel (reverse Hoogsteen) T:A/T motifs allowed, for the first time, a direct comparison of their relative stabilities. Most of the 0.25 kcal mol-1/TAT base triplet difference in stabilities was seen to come from differences in enthalpy, with the parallel (Hoogsteen) base triplet being more stable. / Studies to explore structural effects in the use of a branched adenosine as replacement for nucleotide loops in duplex DNA are also described. Branched oligonucleotides of the type rA[2',5'dC ndA10-5']3',5dC ndT10-3' and rA[2',5' dCn3',3' dA10-5']3', 5dCndT10-3' are shown to form hairpin duplexes with thermal stabilities comparable to or better than that of one with a natural deoxynucleotide loop. Molecular modeling was used to study the conformation of branched hairpin loops. In certain cases branched hairpins adopted conformations similar to the ones seen in naturally occurring super stable RNA hairpins. / The branched nucleic acids (bNAs) also proved to be invaluable tools for studying in vitro splicing and the substrate specificity of the yeast debranching enzyme.

Oligonucleotides containing the modification 2-fluoro-2-deoxy-beta-D-arabinose (2'F-ANA) were synthesized via solid phase synthesis in order to investigate their ability to bind to nucleic acids (DNA and RNA). The ability of this modification to serve as a suitable antisense oligonucleotide construct was also evaluated. / All four nucleobase monomers (thymine, cytosine, adenine and guanine) were prepared for solid phase oligonucleotide synthesis, from which homopolymeric and heteropolymeric base sequences were assembled. These molecules bound with very good affinities to both DNA and RNA targets. Structural studies via NMR experiments demonstrated that in a 2'F-ANA/RNA duplex the 2'F-ANA residues adopt an O4' -endo conformation, similar to what has been proposed for the structure of DNA in a DNA/RNA hybrid. / A 2'F-ANA oligopyrimidylate formed a triple-helical complex with duplex DNA and hybrid DNA(Pu):RNA(Py) with an affinity higher than that of a corresponding DNA strand. Also, a cytosine-rich 2' F-ANA strand was found to form a complex at acidic pH which has properties similar to that of i-motif DNA. / Finally, 2'F-ANA strands when hybridized to RNA were found to activate RNase H, an enzyme that is involved in the mechanism of action of antisense drugs. 2'F-ANA is the first example of an antisense analogue that demonstrates improved binding to RNA relative to DNA and still retains the ability to elicit RNase H activity.

There is accumulating evidence that cysteine proteinase activity plays an important role in cancer cell invasion and metastasis. Extracellular cysteine proteases, in particular cathepsin B, have been implicated in a variety of pathological processes involving tissue remodeling states, such as inflammation, parasite infection, and tumor metastasis, by degradation of extracellular matrix components. As such, increasing efforts are directed toward the development of inhibitors that are specific to the enzyme cathepsin B in order to both dissect out the role of the enzyme in these pathological processes and as potential therapeutic agents. / Unlike other members of the papain superfamily of enzymes, cathepsin B has a unique 18-residue insertion loop termed the "occluding loop" which sits over the primed subsites of the enzyme. It has now been well established that this loop is responsible for cathepsin B's unique dipeptidylcarboxypeptidase activity due to the presence of two positively charged residues, His110 and His111, that anchor the C-terminus of peptide substrates and allow the enzyme to carry out its carboxypeptidase activity. It has been demonstrated that this loop is a flexible segment that can move up and out of the way in order to accommodate binding of larger molecules such as the propeptide. Close examination of the x-ray crystal structure of the mature enzyme indicates the presence of an electrostatic interaction between the side chain of the main chain residue Asp22 and the imidazole ring of the occluding loop residue His110. This salt bridge acts as latch between the loop "open" and loop "closed" conformations of the enzyme. / It has been postulated that the development of inhibitors which interact with the "occluding loop" might provide highly selective inhibitors for cathepsin B and in fact cathepsin B-specific epoxysuccinyl inhibitors have been made which exploit the interaction of between a negatively charged carboxylate at P2' and the two positively charged histidine residues of the occluding loop. Using site-directed mutagenesis, we have dissected out the individual contributions of the two occluding loop histidine residues (His110 and His111). The effect of pH on these interactions has also been evaluated and it has been shown that increasing pH results in increased loop flexibility and diminished inhibitor potency. / Further structure-activity relationships for epoxysuccinyl inhibitors were also established by varying our inhibitor design template both in the P1' and P2' positions using known substrate specificities in these positions. Extension of the inhibitor template into the unprimed subsite region yielded the most potent epoxysuccinyl inhibitor to cathepsin B reported to date (BzlNH-Phe-NH-(2 S,3S)-tEps-Leu-Pro-OH: k2/Ki 2 900 000 +/- 300 000 M-1s-1).