The development of peptides as therapeutic agents has been hampered by their poor enzymatic stability and bioavailability. Many strategies, such as chemical modification, synthesis of peptidomimetics and formulation, have been employed to overcome these issues. For central nervous system (CNS) active peptides, the blood brain barrier is an added hurdle. Nasal delivery is believed to provide a direct access to the brain via the olfactory nerve, which would bypass the blood brain barrier. This route of administration, however, is dependant on the size and physico-chemical properties of the administered drug. For these reasons, three CNS active peptides were chosen as models. Leu-enkephalin, endomorphin-1 and a-conotoxin MII are three peptides that differ in their size, amino acid sequence and conformation. Using chemical modifications to improve their stability and ability to cross biological membranes, in vitro assessments of derivatives of these peptides were performed and in vivo nasal delivery was attempted on the most promising candidates. The chemical modifications consisted in the addition of lipids and/or sugars to the N- or C-terminus of the peptides. Assessment of the in vivo bioavailability after nasal administration, however, proved to be challenging. The initial method chosen for this purpose was the use of tritiated acetic anhydride which would radiolabel the peptide via acetylation at the N-terminus of the peptide derivatives. Consequently, in vitro stability and permeability of each acetylated derivatives was also studied. Acetylation of the lipidic derivatives, which formed an amide bond, proved to be beneficial for the stability of the lipidic peptides. In contrast, acetylation of the Nterminus sugar derivatives, which formed an ester bond at one or several positions of the sugar, was an unstable modification. Thus, an extraction method for the tested peptides from rat tissues was developed, and LC-MS/MS analyses were conducted to measure the level of peptide in the olfactory bulbs, brain and blood. Leu-enkephalin derivatives were all amide derivatives at the C-terminus of the peptide. The most successful Leu-enkephalinamide derivatives were C8-LeuEnk (2), C12- LeuEnk (3) and Lac-LeuEnk (8), which are the Leu-enphelinamide peptide modified with a C8 lipoamino acid, a C12 lipoamino acid and a lactose moiety respectively. They all exhibited improved permeability across Caco-2 monolayers and stability in Caco-2 cell homogenate and/or plasma. Problems of solubility encountered with C12-LeuEnk (3), however, hampered its testing in vivo after nasal administration. C8-LeuEnk (2) and Lac-LeuEnk (8) were administered intranasally to male Sprague-Dawley rats. Both peptides were found in the olfactory bulbs after 10 minutes administration (2: 49.2 ± 15.6 nM; 8: 40.6 ± 14.6 nM) while blood concentration remained low, showing that the peptide reached the olfactory bulbs directly from the nasal cavity via the olfactory nerve. Brain concentrations were 13.5 ± 10.1 nM for C8-LeuEnk (2) and 13.6 ± 6.9 nM for Lac-LeuEnk (8). These two peptides brain concentrations seemed to be high enough to exhibit analgesic effect when compared to their binding affinity in vitro. This was not statistically significant, however, due to the high standard deviations observed (Kiμ C8-LeuEnk (2) = 7.74 ± 1.15 nM; Kiμ Lac-LeuEnk (8) = 6.69 ± 1.81 nM). Endomorphin-1 was only modified at the N-terminus as previous results have shown that the activity of the peptide is strongly decreased by C-terminus modifications. The most successful modification, regarding permeability across Caco-2 monolayers and water solubility, was shown to be the addition of a lactose moiety to the N-terminus of the peptide. Lac-Endo1 (16) exhibited a permeability of 1.91 ± 0.76 x 10-6 cm/s and was soluble at the concentration used for in vivo nasal administration (2 mg/Kg, 50 μL administration). After 10 minutes administration, Lac-Endo1 (16) was found in the olfactory bulbs (418 ± 410 nM), in the brain (4.01 ± 4.61 nM) and in the blood (1.58 ± 1.85 nM). The large standard deviations observed reflect the difficulties encountered with the extraction process of this peptide. A direct transport for the nasal cavity to the olfactory bulb was observed as illustrated by the low blood concentrations. Brain concentrations, however, were too low to expect a strong analgesic effect from this compound after nasal administration (Kiμ Lac-Endo1 (16) = 11.3 ± 1.2 nM). a-Conotoxin MII is a 16 amino acid long peptide containing two disulfide bonds. The formation of these two disulfide bonds leads to low yields in the synthesis of the derivatives of this peptide. Addition of a lipidic moiety to the peptide did not seem to improve its permeability through biological membranes. This modification resulted in highly lipophilic peptides with dissolution issues in water based media such as those used in the permeability experiments. The most successful a-conotoxin MII derivative was GS-Ctx (25) which exhibited a permeability of 4.22 ± 0.53 x 10-7 cm/s across Caco-2 monolayers. This permeability, however, was too low to consider in vivo administration. In conclusion, we successfully synthesised a series of derivatives of Leu-enkephalin, endomorphin-1 and a-conotoxin MII and screened them through Caco-2 monolayers for permeability and Caco-2 cell homogenates and human plasma for stability. Three derivatives (C8-LeuEnk (2), Lac-LeuEnk (8) and Lac-Endo1 (16)) were intranasally administered and found in the olfactory bulbs 10 minutes after administration. The low blood concentrations observed show that a direct transport from the nasal cavity to the brain occurs. Thus, nasal administration could be an option for delivering to the brain low molecular weight peptides exhibiting increased stability and permeability in vitro.

The development of peptides as therapeutic agents has been hampered by their poor enzymatic stability and bioavailability. Many strategies, such as chemical modification, synthesis of peptidomimetics and formulation, have been employed to overcome these issues. For central nervous system (CNS) active peptides, the blood brain barrier is an added hurdle. Nasal delivery is believed to provide a direct access to the brain via the olfactory nerve, which would bypass the blood brain barrier. This route of administration, however, is dependant on the size and physico-chemical properties of the administered drug. For these reasons, three CNS active peptides were chosen as models. Leu-enkephalin, endomorphin-1 and a-conotoxin MII are three peptides that differ in their size, amino acid sequence and conformation. Using chemical modifications to improve their stability and ability to cross biological membranes, in vitro assessments of derivatives of these peptides were performed and in vivo nasal delivery was attempted on the most promising candidates. The chemical modifications consisted in the addition of lipids and/or sugars to the N- or C-terminus of the peptides. Assessment of the in vivo bioavailability after nasal administration, however, proved to be challenging. The initial method chosen for this purpose was the use of tritiated acetic anhydride which would radiolabel the peptide via acetylation at the N-terminus of the peptide derivatives. Consequently, in vitro stability and permeability of each acetylated derivatives was also studied. Acetylation of the lipidic derivatives, which formed an amide bond, proved to be beneficial for the stability of the lipidic peptides. In contrast, acetylation of the Nterminus sugar derivatives, which formed an ester bond at one or several positions of the sugar, was an unstable modification. Thus, an extraction method for the tested peptides from rat tissues was developed, and LC-MS/MS analyses were conducted to measure the level of peptide in the olfactory bulbs, brain and blood. Leu-enkephalin derivatives were all amide derivatives at the C-terminus of the peptide. The most successful Leu-enkephalinamide derivatives were C8-LeuEnk (2), C12- LeuEnk (3) and Lac-LeuEnk (8), which are the Leu-enphelinamide peptide modified with a C8 lipoamino acid, a C12 lipoamino acid and a lactose moiety respectively. They all exhibited improved permeability across Caco-2 monolayers and stability in Caco-2 cell homogenate and/or plasma. Problems of solubility encountered with C12-LeuEnk (3), however, hampered its testing in vivo after nasal administration. C8-LeuEnk (2) and Lac-LeuEnk (8) were administered intranasally to male Sprague-Dawley rats. Both peptides were found in the olfactory bulbs after 10 minutes administration (2: 49.2 ± 15.6 nM; 8: 40.6 ± 14.6 nM) while blood concentration remained low, showing that the peptide reached the olfactory bulbs directly from the nasal cavity via the olfactory nerve. Brain concentrations were 13.5 ± 10.1 nM for C8-LeuEnk (2) and 13.6 ± 6.9 nM for Lac-LeuEnk (8). These two peptides brain concentrations seemed to be high enough to exhibit analgesic effect when compared to their binding affinity in vitro. This was not statistically significant, however, due to the high standard deviations observed (Kiμ C8-LeuEnk (2) = 7.74 ± 1.15 nM; Kiμ Lac-LeuEnk (8) = 6.69 ± 1.81 nM). Endomorphin-1 was only modified at the N-terminus as previous results have shown that the activity of the peptide is strongly decreased by C-terminus modifications. The most successful modification, regarding permeability across Caco-2 monolayers and water solubility, was shown to be the addition of a lactose moiety to the N-terminus of the peptide. Lac-Endo1 (16) exhibited a permeability of 1.91 ± 0.76 x 10-6 cm/s and was soluble at the concentration used for in vivo nasal administration (2 mg/Kg, 50 μL administration). After 10 minutes administration, Lac-Endo1 (16) was found in the olfactory bulbs (418 ± 410 nM), in the brain (4.01 ± 4.61 nM) and in the blood (1.58 ± 1.85 nM). The large standard deviations observed reflect the difficulties encountered with the extraction process of this peptide. A direct transport for the nasal cavity to the olfactory bulb was observed as illustrated by the low blood concentrations. Brain concentrations, however, were too low to expect a strong analgesic effect from this compound after nasal administration (Kiμ Lac-Endo1 (16) = 11.3 ± 1.2 nM). a-Conotoxin MII is a 16 amino acid long peptide containing two disulfide bonds. The formation of these two disulfide bonds leads to low yields in the synthesis of the derivatives of this peptide. Addition of a lipidic moiety to the peptide did not seem to improve its permeability through biological membranes. This modification resulted in highly lipophilic peptides with dissolution issues in water based media such as those used in the permeability experiments. The most successful a-conotoxin MII derivative was GS-Ctx (25) which exhibited a permeability of 4.22 ± 0.53 x 10-7 cm/s across Caco-2 monolayers. This permeability, however, was too low to consider in vivo administration. In conclusion, we successfully synthesised a series of derivatives of Leu-enkephalin, endomorphin-1 and a-conotoxin MII and screened them through Caco-2 monolayers for permeability and Caco-2 cell homogenates and human plasma for stability. Three derivatives (C8-LeuEnk (2), Lac-LeuEnk (8) and Lac-Endo1 (16)) were intranasally administered and found in the olfactory bulbs 10 minutes after administration. The low blood concentrations observed show that a direct transport from the nasal cavity to the brain occurs. Thus, nasal administration could be an option for delivering to the brain low molecular weight peptides exhibiting increased stability and permeability in vitro.

Despite the unfavorable pharmacokinetic properties associated with peptides, they are still of great interest in drug development due to a multitude of interesting biological functions. The development of peptidomimetics strives to maintain or improve the biological activity of a peptide concurrently with removing the unwanted properties. This thesis describes two synthetic approaches to peptidomimetics with particular emphasis on secondary structure mimetics. First the design, synthesis and evaluation of two beta-turn mimetics incorporated in the endorphin Leu-enkephalin is presented. The beta-turn mimetics were stabilized by replacement of the intramolecular hydrogen bond with an ethylene bridge, and the amide bond between Tyr and Gly was replaced with an ether linkage. Linear analogues of the two mimetics were also synthesized. The peptidomimetics and their linear analogues were evaluated in a competitive binding assay at two opiate receptors, my and delta. One of the cyclized beta-turn mimetics was found to be a delta receptor antagonist with an IC50 value of 160 nM. Second a synthetic strategy to a beta-strand mimetic using 2-fluoro-4-iodopyridine as scaffold is described. The synthesis involved a Grignard exchange reaction on the pyridine scaffold using an amino acid derivative as electrophile followed by an SNAr reaction using an amine as nucleophile. The synthesis of a tripeptidomimetic of Leu-Gly-Gly and attempts to introduce chiral building blocks at the C-terminal, as well as studies towards elongated mimetics are presented. Two additional studies deal with the synthesis of two classes of potential thrombin inhibitors based on the pyridine scaffold. The first class contain pyridine as central fragment (P2 residue) substituted with a para-amidinobenzylamine group as P1 residue and various benzoyl groups as P3 residues. Three potential thrombin inhibitors were synthesized and found to be microM inhibitors in an enzymatic assay. In the second class, the pyridine ring serves as P3 residue. This class also lacks a strongly basic group in the P1 position. A small library of eight compounds were synthesized and evaluated in the enzymatic assay. Unfortunately, these compounds lacked inhibitory activity.

Les neuropeptides endogènes chez l’homme, tels que les enképhalines et endomorphines, ont un potentiel thérapeutique considérable dans le traitement de la douleur. Ils agissent en activant les récepteurs opioïdes qui sont très largement distribués dans le système nerveux central ainsi que dans plusieurs tissus périphériques. Ces neuropeptides présentent, cependant, un certain nombre d’inconvénients qui limitent de manière importante leur efficacité thérapeutique. Tout d’abord, en raison de leur hydrophilie, ils ne passent pas la barrière sang/système nerveux, ce qui limite leur accès aux récepteurs opioïdes. De plus, ils présentent un temps de demi-vie plasmatique relativement court du fait d’une métabolisation rapide. Enfin, pour être efficaces, ces neuropeptides devraient résister à la protéolyse dans le système circulatoire et être suffisamment hydrophobes pour traverser ces barrières hémato-nerveuses.Le but de la thèse a consisté à créer de nouveaux nanomédicaments à base d’endorphines pour vectoriser et combattre la douleur de manière efficace.. Dans ce but, a été établi une liaison chimique covalente, enzymatiquement clivable (ester ou amide), entre le squalène (SQ, un lipide naturel et biocompatible) et le neuropeptide. Ce couplage donne lieu à des prodrogues qui ont la capacité de s'auto-assembler en nanoparticules (NPs) dans l'eau sans l’aide d'un agent tensio-actif. D’une manière générale, cette technologie présente de nombreux avantages tels qu’un taux de charge élevé en principe actif, une protection efficace de celui-ci vis-à-vis de la métabolisation et l’absence de phénomène de « burst release ».Durant ce travail de thèse, nous avons pu montrer pour la première fois que les Leu-enképhalines (LENK) pouvaient devenir efficaces pharmacologiquement une fois couplées au squalène, via une liaison amide (Am), ou via un bras espaceur, tel que le dioxycarbonyl (Diox), ou le diglycolate (Dig). Les prodrogues résultant de ce couplage ont toutes montré des propriétés d’auto-assemblage en milieu aqueux. Cette nanoformulation à base de squalène a permis, d’une part de protéger la LENK de la métabolisation rapide dans le plasma et d’autre part, de lui conférer un effet anti-hyperalgésique significatif dans un modèle de douleur inflammatoire induite chez le rat par la carragénine (test de Hargreaves). Il est important de souligner que cet effet anti-hyperalgésique a duré 3 fois plus longtemps qu’avec la morphine. Un prétraitement avec un antagoniste des récepteurs opioїdes imperméable à la BHE, comme la méthylnaloxone a complètement antagonisé l’effet anti-hyperalgésique des nanoparticules de LENK-SQ, démontrant ainsi que celles-ci agissent via les récepteurs opioïdes périphériques. De plus, l’étude de biodistribution de NPs LENK-SQ fluorescentes a montré une forte accumulation des celles-ci au niveau de la patte œdémateuse, mais aussi dans le foie, la rate et les poumons alors qu’aucun signal n’a pu être détecté au niveau cérébral, ce qui confirme bien l’effet périphérique de ces nanoparticules. Enfin, des études toxicologiques ont montré que malgré l’accumulation des NPs dans le foie, les taux d’aspartate transaminase (AST) et alanine transaminase (ALT) n’ont pas augmenté garantissant ainsi l’innocuité des NPs LENK-SQ après leur injection i.v. Cette étude représente une approche innovante et prometteuse permettant une distribution ciblée du neuropeptide endogène LENK dans les tissus œdémateux pour soulager efficacement la douleur inflammatoire. / Enkephalin is an endogenous pentapeptide producing potent analgesia by activating opioid receptors located on central and peripheral neuronal cell membranes. However, its clinical use has historically been limited due to pharmacokinetic issues, including restricted plasma stability and blood brain barrier impermeability. The aim of this project is to create a new enkephalin-based nanomedicine targeting pain, using biocompatible and biodegradable materials for drug delivery and targeting purposes, such as squalene (squalenoylation nanotechnology). This nanotechnology presents a new concept with numerous advantages in comparison with the conventional nanocarriers, such as high drug loading and absence of “burst release”. Here, we show for the first time, that the rapidly metabolized Leu-enkephalin (LENK) neuropeptide may become pharmacologically efficient owing to its simple conjugation with the squalene (SQ) using three different chemical linkers, i.e., dioxycarbonyl (Diox), diglycolate (Dig), or amide bond (Am). The resulting prodrugs were able to self-assemble in nanoparticles in aqueous media. This new squalene-based nanoformulation prevented rapid plasma degradation of LENK and conferred to the released neuropeptide a significant anti-hyperalgesic effect in a carrageenan-induced paw edema model in rats (Hargreaves test). It should be stressed that this effect lasted 3 times longer than morphine. Pretreatment with brain impermeant opioid receptor antagonist naloxone methiodide (Nal-M) reversed the nanoparticles induced anti-hyperalgesia, indicating that LENK-SQ NPs acted through peripherally located opioid receptors. Moreover, the biodistribution of DiD-fluorescently labeled LENK-SQ NPs showed a strong accumulation of the fluorescence within the inflamed paw as well as in the liver, spleen, and lung, while no signal could be detected in the brain, confirming the peripheral effect of LENK-SQ NPs. Toxicological studies showed that despite nanoparticles accumulation in the liver, the levels of aspartate transaminase (AST) and alanine transaminase (ALT) were not increased after i.v. injection of LENK-SQ NPs, highlighting thus their safety. This study represents a novel drug targeting approach, allowing the specific delivery of LENK neuropeptide into inflamed tissues for pain alleviation.

In the last decades NMR spectroscopy has become an invaluable tool both in academic research and in the pharmaceutical industry. This thesis describes applications of NMR spectroscopy in biomedicinal research for structure elucidation of biologically active peptides and peptidomimetics as well as in studies of ligand-protein interactions. The first part of this thesis describes the theory and methodology of structure calculations of peptides using experimental restraints derived from NMR spectroscopy. This methodology has been applied to novel mimetics of the peptide hormones desmopressin and Leu-enkephalin. The results of these studies highlight the complicating issue of conformational exchange often encountered in structural determination of peptides and how careful analysis of experimental data as well as optimization of experimental conditions can enable structure determinations in such instances. Although the mimetics of both desmopressin and Leu-enkephalin were found to adopt the wanted conformations, they exhibited no or very poor biological activity. These results demonstrate the difficulties in designing peptidomimetics without detailed structural information of the receptors. A stereoselective synthetic route towards XxxΨ[CH2O]Ala pseudodipeptides is also presented. Such pseudodipeptides can be used as isosteric amide bond replacements in peptides in order to increase their resistance towards proteolytic degradation. The second part of this thesis describes the study of the interaction between compounds that inhibit pilius assembly, pilicides, and periplasmic chaperones from uropathogenic Escherichia coli. Periplasmic chaperones are key components in assembly of pili, i.e. hair-like protein complexes located on the surface of Escherichia coli that cause urinary tract infections. Detailed knowledge about this interaction is important in understanding how pilicides can inhibit pilus assembly by binding to chaperones. Relaxation-edited NMR experiments were used to confirm the affinity of the pilicides for the chaperones and chemical shift mapping was used to study the pilicide-chaperone interaction surface. These studies show that at least two interaction sites are present on the chaperone surface and consequently that two different mechanisms resulting in inhibition of pilus assembly may exist.