Les kallikréines (KLKs) tissulaires humaines sont des protéases à sérine « (chymo)trypsine-like » impliquées dans divers processus physiologiques. Parmi les 15 isoformes connues dans la littérature, la kallikréine 7 (KLK7) est particulièrement impliquée dans les processus de desquamation. La dérégulation de cette enzyme est associée à diverses atteintes dermatologiques, telles que le psoriasis ou la maladie de Netherton. Il est également admis que cette enzyme participe à l'invasion tumorale et à la progression du cancer de la prostate, des ovaires et du pancréas. L'utilisation d'inhibiteurs de la KLK7 pourrait donc constituer une approche prometteuse pour le traitement de certaines atteintes dermatologiques, et pour lutter contre la dissémination métastatique.Depuis deux décennies, plusieurs inhibiteurs synthétiques de cette isoforme ont été développés. Toutefois, la plupart de ces molécules présentent une sélectivité insuffisante et sont dépourvues de propriétés physicochimiques adaptées à une utilisation in vivo. Ce travail de thèse est consacré à la synthèse d'inhibiteurs réversibles et sélectifs de la KLK7. A ce titre, deux séries de composés ont été explorées. La première est issue d'un criblage de molécules pyrido-imidazodiazépinones, qui a permis d'identifier un inhibiteur réversible et sélectif de la KLK7, le JMV4967 (IC50 = 57,0 µM). Ce composé se caractérise par la présence, en position 2 du cycle diazépinique, d'un noyau phényle substitué en ortho par un groupement méthyle. Sur la base de ces résultats, une étude de relations structure-activité (SAR), a été initiée. Les meilleures inhibitions ont été obtenues avec les analogues du JMV4967 possédant en position 2 du cycle diazépinique, un groupement 3,4,5-triméthoxyphényle ou 3,4-diméthoxyphényle. Cette étude a permis l'identification du composé JMV5046 (IC50 = 33,5 µM). Une étude biochimique du JMV5046, a mis en évidence son caractère réversible et compétitif vis-à-vis du substrat dans le site actif de la KLK7, comme le hit initial. La seconde série a été développée à partir d'une quinazoline substituée par un amino-benzimidazole, et une étude de RSA a également été menée. Finalement, une étude de réactivité chimique, a été également entreprise afin d'accéder à deux nouvelles séries de composés diazépiniques fusionnés avec l'imidazo[1,2-a]pyridine ou l'indole. Cette étude a montré que la 2-amino-imidazopyridine pouvait subir une alkylation en position 3 du cycle, dans le cadre de la réaction de nitro-Michael. Les produits d'addition de Michael ainsi formés peuvent ensuite conduire, après réduction du groupement nitro, aux dérivés diazépiniques correspondants. Par ailleurs, nous avons pu montrer que l'acylation du 2-amino-indole, en présence d'un donneur d'acyle de type ester activé d'acide aminé, conduisait aux dérivés N-acylés correspondants. Ces derniers ont été utilisés par la suite, pour la synthèse de dérivés indolodiazépiniques via la réaction de cyclisation de Pictet-Spengler. La synthèse de ces composés, l'étude de leur activité inhibitrice sur la KLK7 sont décrites. Les expériences de modélisation moléculaire par docking in silico, permettant de préciser les bases structurales de l'inhibition de la KLK7 par le JMV5046, sont également présentées. / Human tissular kallikreins (KLKs) are members of (chymo)trypsin-like serine proteases, involved in various physiological pathways. Among the 15 known isoforms in the literature, kallikrein 7 (KLK7) plays a significant role in physiological and pathophysiological processes of the skin, like psoriasis and the Netherton syndrome. Several studies report also its implication in multiple processes leading to invasive and metastatic tumor growth, especially in prostatic, ovarian and pancreatic cancers. Strategies focused on KLK7 inhibition are a promising alternative for the treatment of such dermatological diseases, and to avoid metastatic dissemination. In the last two decades, many synthetic inhibitors of this KLK isoform have been developed. However, most of these molecules exhibit low selectivity and unsuitable physicochemical properties for in vivo use. This thesis is devoted to the synthesis of selective and reversible inhibitors of KLK7. Two series of compounds have been explored. The first one, derived from a screening of pyrido-imidazodiazepinones compounds, which led to the discovery of a reversible and selective inhibitor of KLK7, JMV4967 (IC50 = 57.0 µM). This compound is characterized by the presence of an ortho methyl substituted phenyl group, at position 2 of the diazepine ring. Based on these results, a structure-activity relationships (SAR) study was initiated. The best inhibitions were obtained with JMV4967analogs bearing a 3,4,5- trimethoxyphenyl group or 3,4- dimethoxyphenyl group, at position 2 of the diazepine ring. This study led to the discovery of one compound, JMV5046 (IC50 = 33.5 µM). A biochemical study of JMV5046, highlighted its reversible and competitive behavior against the substrate in the KLK7 active site, as previously observed for the initial hit. The second series was developed from an amino-benzimidazole substituted quinazoline, and a SAR study was also carried out. A chemical reactivity study was also initiated to access two new series of imidazo[1,2-a]pyridine-fused or indole-fused diazepine compounds. This study showed that 2-amino-imidazopyridine could undergo alkylation at position 3 of the ring, in the nitro-Michael reaction context. The obtained Michael adducts could then give, after reduction of the nitro group, the corresponding diazepine derivatives. We also showed that, in the presence of an acyl donor (as an activated amino acid ester), 2-amino-indole was N-acylated, unlike the 2-amino-imidazopyridine which leads to an exclusive C-acylation in the same conditions. The N-acylated derivatives were then used for the indolodiazepines synthesis, using Pictet-Spengler cyclization reaction.The synthesis of these compounds and their inhibiting activity against KLK7 are described. Molecular modeling experiments by in silico docking, to determine the structural requirements for KLK7 inhibition by JMV5046, are also presented.

No capítulo 1, a modelagem HQSAR, a docagem e os estudos de ROCS foram construídos utilizando uma série de 57 inibidores de cruzaína. O melhor modelo HQSAR (q2 = 0,70, r2 = 0,95, r2test = 0,62, q2rand. = 0,09 and r2rand. = 0,26) foi utilizado para predizer a potência de 121 compostos extraídos da literatura (conjunto de dados V1), resultando em um valor de r2 satisfatório de 0,65 para essa validação externa. Uma validação externa adicional foi empregada utilizando uma série de 1223 compostos extraído dos bancos de dados ChEMBL e CDD (conjunto de dados V3); nessa validação externa o valor de AUC (área sob a curva) para a curva ROC foi de 0,70. Os mapas de contribuição, obtidos para o melhor modelo HQSAR 3.4, estão de acordo com as predições do modo de interação e com as bioatividades dos compostos estudados. Nos estudos de ROCS, a forma molecular utilizada como filtro, foi útil na rápida identificação de modificações moleculares promissoras para inibidores de cruzaína. O valor de AUC obtido com a curva ROC foi de 0,72, isso indica que o método foi muito eficiente na distinção entre inibidores ativos e inativos da enzima cruzaína. Em seguida, o melhor modelo HQSAR foi utilizado para predizer os valores de pIC50 para novos compostos. Alguns dos compostos identificados, utilizando esse método, demonstraram valores de potência calculada maior do que a série de treinamento em estudo. No capítulo 2, os efeitos sobre a potência na inibição da enzima cruzaína pela substituição de um grupo nitrila como warhead por outros grupos foi avaliada. Com a síntese de 20 compostos do tipo dipeptidil, avaliou-se a relação estrutura-atividade (SAR), baseado na troca do grupo warhead na porção P1\'. O grupo oxima foi mais potente que o grupo correspondente nitrila em 0,7 unidades logarítmicas. Os compostos do tipo dipeptidil aldeídos e azanitrila obtiveram potências mais elevadas do que o correspondente dipeptidil nitrila em duas de magnitude. Os compostos dipeptidil alfa-beta insaturados foram menos potentes do que o correspondente dipeptidil nitrila. No capítulo 3, estratégias de química medicinal foram empregadas nas sínteses de 23 novos análogos, contendo o esqueleto básico de imidazopiridina. Sete e doze compostos sintetizados exibiram EC50 <= 1µM in vitro contra os parasitos Tripanosoma cruzi (T. cruzi) e brucei (T. brucei), respectivamente. Com os resultados promissores de atividade biológica in vitro, citotoxicidade, estabilidade metabólica, ligação proteica e propriedades farmacocinéticas, o composto 41 foi selecionado como candidato para os estudos de eficácia in vivo. Esse composto foi submetido em um modelo agudo da infecção com T. cruzi em ratos (cepa Tulahuen). Depois de estabelecida a infecção, os ratos foram dosados duas vezes ao dia, durante 5 dias; e monitorados por 6 semanas usando um sistema de imagem in vivo IVIS (do inglês, \"In Vivo Imaging System\"). O composto 41 demonstrou inibição parasitária comparável com o grupo de treinamento dosado com benzonidazol. O composto 41 representa um potencial líder para o desenvolvimento de novos fármacos para o tratamento de tripanossomíases. / In chapter 1, the HQSAR, molecular docking and ROCS were applied to a dataset of 57 cruzain inhibitors. The best HQSAR model (q2 = 0.70, r2 = 0.95, r2test = 0.62, q2rand. = 0.09 and r2rand. = 0.26) was then used to predict the potencies of 121 unknown compounds (the V1 database), giving rise to a satisfactory predictive r2 value of 0.65 (external validation). By employing an extra external dataset comprising 1223 compounds (the V3 database) either retrieved from the ChEMBL or CDD databases, an overall ROC AUC (area under the curve) score well over 0.70 was obtained. The contribution maps obtained with the best HQSAR model (model 3.4) are in agreement with the predicted binding mode and with the biological potencies of the studied compounds. We also screened these compounds using the ROCS method, a Gaussian-shape volume filter able to identify quickly the shapes that match a query molecule. The AUC obtained with the ROC curves (ROC AUC) was 0.72, indicating that the method was very efficient in distinguishing between active and inactive cruzain inhibitors. These set of information guided us to propose novel cruzain inhibitors to be synthesized. Then, the best HQSAR model obtained was used to predict the pIC50 values of these new compounds. Some compounds identified using this method has shown calculated potencies higher than those which have originated them. In chapter 2, the effects on potency of cruzain inhibition of replacing a nitrile group with alternative warheads were explored; with the syntheses of 20 dipeptidyl compounds, we explored the structure activity relationships (SAR) based on exchanging of the warhead portion (P1\'). The oxime was 0.7 units more potent than the corresponding nitrile. Dipeptide aldehydes and azadipeptide nitriles were found to be two orders of magnitude more potent than the corresponding dipeptide nitriles. The vinyl esters and amides were less potent than the corresponding nitrile by between one and two orders of magnitude. In chapter 3, we synthesized 23 new imidazopyridine analogues arising from medicinal chemistry optimization at different sites on the molecule. Seven and twelve compounds exhibited an in vitro EC50 <= 1µM against Trypanosoma cruzi (T. cruzi) and Trypanosoma brucei (T. brucei) parasites, respectively. Based on promising results of in vitro activity (EC50 &lt; 100 nM), cytotoxicity, metabolic stability, protein binding and pharmacokinetics (PK) properties, compound 41 was selected as a candidate for in vivo efficacy studies. This compound was screened in an acute mouse model against T.cruzi (Tulahuen strain). After established infection, mice were dosed twice a day for 5 days, and then monitored for 6 weeks using an in vivo imaging system (IVIS). Compound 41 demonstrated parasite inhibition comparable to the benznidazole treatment group. Compound 41 represents a potential lead for the development of drugs to treat trypanosomiasis.

The development of novel imidazopyridines, which includes both tetrahydroimidazo[1,5-a]pyridine (rIMP) and imidazo[1,5-a]pyridine (IMP) was investigated using conventional and microwave induced procedures that afforded compounds at high yield of 88-96%. rIMP was synthesized using a two-step procedure that involved the microwave synthesis of IMP, then the reduction of the pyridine moiety of the fused imidazopyridine rings using 10% Pd/C and hydrazine monohydrate. The microwave synthesis of imidazopyridines involved the one pot reaction of 2-benzoylpyridine, substituted benzaldehyde and ammonium formate in acetic acid under open vessel microwave conditions, which resulted in products within 40 minutes. Novel PEG-IMP development, involved the synthesis of ethylene glycol tethered benzaldehydes and IMPs using traditional Williamson etherification synthesis, which afforded products at a high yield of 92-95%. We have then shown IMP and rIMP roles in its antiproliferative property towards PCa cells, specificity in inhibiting PI3K isoforms, and structural motif’s interaction with different residues in the kinase binding domain of the class I PI3K isoforms. The antiproliferative property towards PC3 cells shows increased activity with compounds containing pyridyl group on carbon 3 of the imidazo[1,5-a]pyridine parent moiety with signs of toxicity to PC3 within 24 hours of incubation and at 1 μM of the parent compound. Furthermore, the IMPs were tested against five prostate cellular lines: PC3, RWPE1, D145, LNCaP and LNCaP C81. IMPs showed little activity towards RWPE1 and increased activity towards PC3 cells. We determined that functionalizing the phenyl group at position 1 increased the efficacy of rIMP compared to the IMP. After showing increased toxicity to PC3 cells, it was important to investigate the mechanism in which IMP pose toxicity towards PC3 cells. The biochemical assay showed that rIMP was more effective in inhibiting PI3Kα isoform compared to both pan inhibitor wortmannin and IMP. Both IMP and rIMP inhibited more than 60% of PI3Kγ isoform activity at nanomolar concentrations. After showing IMPs affinity to PI3K isoforms, we investigated the binding interactions rIMP and IMP towards the PI3K isoforms using MOE molecular modeling software.

No capítulo 1, a modelagem HQSAR, a docagem e os estudos de ROCS foram construídos utilizando uma série de 57 inibidores de cruzaína. O melhor modelo HQSAR (q2 = 0,70, r2 = 0,95, r2test = 0,62, q2rand. = 0,09 and r2rand. = 0,26) foi utilizado para predizer a potência de 121 compostos extraídos da literatura (conjunto de dados V1), resultando em um valor de r2 satisfatório de 0,65 para essa validação externa. Uma validação externa adicional foi empregada utilizando uma série de 1223 compostos extraído dos bancos de dados ChEMBL e CDD (conjunto de dados V3); nessa validação externa o valor de AUC (área sob a curva) para a curva ROC foi de 0,70. Os mapas de contribuição, obtidos para o melhor modelo HQSAR 3.4, estão de acordo com as predições do modo de interação e com as bioatividades dos compostos estudados. Nos estudos de ROCS, a forma molecular utilizada como filtro, foi útil na rápida identificação de modificações moleculares promissoras para inibidores de cruzaína. O valor de AUC obtido com a curva ROC foi de 0,72, isso indica que o método foi muito eficiente na distinção entre inibidores ativos e inativos da enzima cruzaína. Em seguida, o melhor modelo HQSAR foi utilizado para predizer os valores de pIC50 para novos compostos. Alguns dos compostos identificados, utilizando esse método, demonstraram valores de potência calculada maior do que a série de treinamento em estudo. No capítulo 2, os efeitos sobre a potência na inibição da enzima cruzaína pela substituição de um grupo nitrila como warhead por outros grupos foi avaliada. Com a síntese de 20 compostos do tipo dipeptidil, avaliou-se a relação estrutura-atividade (SAR), baseado na troca do grupo warhead na porção P1\'. O grupo oxima foi mais potente que o grupo correspondente nitrila em 0,7 unidades logarítmicas. Os compostos do tipo dipeptidil aldeídos e azanitrila obtiveram potências mais elevadas do que o correspondente dipeptidil nitrila em duas de magnitude. Os compostos dipeptidil alfa-beta insaturados foram menos potentes do que o correspondente dipeptidil nitrila. No capítulo 3, estratégias de química medicinal foram empregadas nas sínteses de 23 novos análogos, contendo o esqueleto básico de imidazopiridina. Sete e doze compostos sintetizados exibiram EC50 <= 1µM in vitro contra os parasitos Tripanosoma cruzi (T. cruzi) e brucei (T. brucei), respectivamente. Com os resultados promissores de atividade biológica in vitro, citotoxicidade, estabilidade metabólica, ligação proteica e propriedades farmacocinéticas, o composto 41 foi selecionado como candidato para os estudos de eficácia in vivo. Esse composto foi submetido em um modelo agudo da infecção com T. cruzi em ratos (cepa Tulahuen). Depois de estabelecida a infecção, os ratos foram dosados duas vezes ao dia, durante 5 dias; e monitorados por 6 semanas usando um sistema de imagem in vivo IVIS (do inglês, \"In Vivo Imaging System\"). O composto 41 demonstrou inibição parasitária comparável com o grupo de treinamento dosado com benzonidazol. O composto 41 representa um potencial líder para o desenvolvimento de novos fármacos para o tratamento de tripanossomíases. / In chapter 1, the HQSAR, molecular docking and ROCS were applied to a dataset of 57 cruzain inhibitors. The best HQSAR model (q2 = 0.70, r2 = 0.95, r2test = 0.62, q2rand. = 0.09 and r2rand. = 0.26) was then used to predict the potencies of 121 unknown compounds (the V1 database), giving rise to a satisfactory predictive r2 value of 0.65 (external validation). By employing an extra external dataset comprising 1223 compounds (the V3 database) either retrieved from the ChEMBL or CDD databases, an overall ROC AUC (area under the curve) score well over 0.70 was obtained. The contribution maps obtained with the best HQSAR model (model 3.4) are in agreement with the predicted binding mode and with the biological potencies of the studied compounds. We also screened these compounds using the ROCS method, a Gaussian-shape volume filter able to identify quickly the shapes that match a query molecule. The AUC obtained with the ROC curves (ROC AUC) was 0.72, indicating that the method was very efficient in distinguishing between active and inactive cruzain inhibitors. These set of information guided us to propose novel cruzain inhibitors to be synthesized. Then, the best HQSAR model obtained was used to predict the pIC50 values of these new compounds. Some compounds identified using this method has shown calculated potencies higher than those which have originated them. In chapter 2, the effects on potency of cruzain inhibition of replacing a nitrile group with alternative warheads were explored; with the syntheses of 20 dipeptidyl compounds, we explored the structure activity relationships (SAR) based on exchanging of the warhead portion (P1\'). The oxime was 0.7 units more potent than the corresponding nitrile. Dipeptide aldehydes and azadipeptide nitriles were found to be two orders of magnitude more potent than the corresponding dipeptide nitriles. The vinyl esters and amides were less potent than the corresponding nitrile by between one and two orders of magnitude. In chapter 3, we synthesized 23 new imidazopyridine analogues arising from medicinal chemistry optimization at different sites on the molecule. Seven and twelve compounds exhibited an in vitro EC50 <= 1µM against Trypanosoma cruzi (T. cruzi) and Trypanosoma brucei (T. brucei) parasites, respectively. Based on promising results of in vitro activity (EC50 &lt; 100 nM), cytotoxicity, metabolic stability, protein binding and pharmacokinetics (PK) properties, compound 41 was selected as a candidate for in vivo efficacy studies. This compound was screened in an acute mouse model against T.cruzi (Tulahuen strain). After established infection, mice were dosed twice a day for 5 days, and then monitored for 6 weeks using an in vivo imaging system (IVIS). Compound 41 demonstrated parasite inhibition comparable to the benznidazole treatment group. Compound 41 represents a potential lead for the development of drugs to treat trypanosomiasis.