Bei der adjuvanten Immuntherapie soll das Immunsystem von Tumorpatienten gezielt gegen Mikrometastasen aktiviert werden, die nach der operativen Entfernung des Primärtumors im Körper verbleiben. Tumor-assoziierte Antigene (TAA) spielen dabei die zentrale Rolle. Um der Heterogenität eines Tumors in der Expression einzelner TAAs Rechnung zu tragen, werden in modernen Vakzinierungsstrategien Pools von verschiedenen TAAs eingesetzt. Für das Mammakarzinom sind aber bislang nur wenige TAAs bekannt. Auf der Suche nach unbekannten Genen mit Mamma- bzw. Mammakarzinom-restringierter Expression in der Transkriptomdatenbank GeneExpress® wurde ein EST gefunden, das nur in 2 % der getesteten weibl. Normalgewebe ohne Mamma mit einer marginalen mittleren Expression von 16 FE (Fluoreszenzeinheit) detektiert wurde, aber in 63 % der Mammanormalgewebe (159 FE) und in 61 % der Mammakarzinome (339 FE) exprimiert wurde. Das korrespondierende UniGene-Cluster gab erste Hinweise auf die Zugehörigkeit dieses neuen Gens zu der Cytochrom P450 (CYP) Familie. Durch computergestützte Homologievergleiche mit verwandten Mitgliedern dieser Familie in Kombination mit RT-PCR konnte die cDNA-Sequenz dieses neuen CYP ermittelt werden: Durch Amplifikation der Gesamtlängen-cDNA wurden drei Transkripte in der Mammakarzinomzelllinie SK-BR-3 gefunden, die von zwei neuen CYP Genen stammen, CYP4Z1 und dem Pseudogen CYP4Z2P. Die cDNA von CYP4Z1 kodiert für ein 505 As großes Protein, das aufgrund von Homologien einer neuen Subfamilie innerhalb der CYP4 Familie zugeordnet werden kann. Sowohl die Sequenz als auch die vorhergesagte Sekundärstruktur zeigen alle charakteristischen Merkmale eines funktionellen Mitglieds dieser Familie. Aufgrund einer Nonsensemutation in Exon 8 kodiert die cDNA von CYP4Z2P (1436 bp) für ein verkürztes, nichtfunktionelles P450 von 340 As, das zu 96% identisch mit P450 4Z1 ist. Außerdem wurde in SK-BR-3 eine Spleißvariante von CYP4Z1 identifiziert. CYP4Z1 (50,8 kb) und CYP4Z2P (57,3 kb) liegen auf Chromosom 1p33-p34.1 und bestehen aus 12 Exons mit konservierten Exon-Intron-Grenzen. CYP4Z2P ist aus einer inversen Duplikation von CYP4Z1 hervorgegangen. Mittels Realtime RT-PCR mit cDNA von 17 Normalgeweben von gepoolten Spendern und von Mammakarzinomen konnte die auf Brustdrüsengewebe restringierte Expression beider Gene demonstriert werden: Die Expression von CYP4Z1 war in Mammakarzinomgewebe 3,6-mal höher als in Mammanormalgewebe, 60-mal höher als in Lunge und 84-mal höher als in Leber. Alle anderen getesteten weibl. Normalgewebe zeigten eine noch geringere Expression. Ein ähnlich stringentes Expressionsprofil ergab die Analyse von CYP4Z2P, allerdings mit einer deutlich niedrigeren Expressionsstärke. Das Mamma-restringierte Expressionsverhalten von CYP4Z1 wurde durch einen ?Cancer-Profiling-Array? (241 Tumor-/Normalgewebepaare von 13 Gewebetypen) bestätigt. Damit konnte gezeigt werden, dass CYP4Z1 bei 52 % der getesteten 50 Brustkrebspatientinnen im Tumor versus peritumoralem Normalgewebe überexprimiert war. Mit einem spezifischen Kaninchen-Antiserum konnte die Expression von P450 4Z1 Protein sowohl in CYP4Z1-transduzierten Zelllinien als auch in Mammagewebeproben mittels Western-Blot nachgewiesen werden. Konfokale Laser-Scanning Mikroskopie von MCF-7 Zellen, die das Fusionsprotein CYP4Z1-EGFP exprimierten, und die subzelluläre Fraktionierung der CYP4Z1-Transduktanten zeigten P450 4Z1 als integrales Membranprotein des endoplasmatischen Retikulums (ER). Für die Lokalisation und die Zurückhaltung im ER ohne Recycling aus dem Prä-Golgiapparat sind die ersten 32 As erforderlich, was Studien mit unterschiedlichen Deletionsmutanten aus der N-terminalen Sequenz von 4Z1 zeigten. Die Entdeckung eines neuen Mamma-spezifisch exprimierten P450 Enzyms eröffnet neue Möglichkeiten sowohl in Hinsicht auf eine Immuntherapie von Brustkrebs als auch für die Entwicklung neuer Chemotherapeutika, die spezifisch durch P450 4Z1 am Tumorort umgesetzt werden können.

O carvedilol é um anti-hipertensivo disponível na clínica como mistura racêmica, sendo o (S)-(-)-carvedilol um bloqueador ? e ?1-adrenérgico e o (R)-(+)-carvedilol apenas ?1-adrenérgico. O carvedilol é metabolizado principalmente por glicuronidação e pelo CYP2D6 a hidroxifenilcarvedilol e pelo CYP2C9 a Odesmetilcarvedilol. O presente estudo avalia a disposição cinética e o metabolismo dos enantiômeros do carvedilol, hidroxifenilcarvedilol e O-desmetilcarvedilol em voluntários sadios não comedicados (n=13) ou comedicados com dose única oral de glibenclamida (5 mg) e metformina (500 mg) (n=13) e em pacientes portadores de diabetes mellitus tipo 2 (com bom controle glicêmico e em tratamento com glibenclamida 5 mg/8h e metformina 500 mg/8h; n=14), fenotipados como metabolizadores rápidos (n=26) ou lentos (n=1). Os voluntários e pacientes receberam dose única oral de 25 mg de carvedilol racêmico e amostras seriadas de sangue foram coletadas até 24h após a administração. A frequência cardíaca foi avaliada na situação de exercício isométrico com o handgrip durante 2 min a 30% da contratilidade voluntária máxima e durante o repouso. Os enantiômeros do carvedilol e metabólitos foram analisados em plasma por LC-MS/MS empregando coluna Chirobiotic® V. O método foi linear no intervalo de 0,05 a 100; 0,05 a 10 e 0,02 a 10 ng/mL para os enantiômeros do carvedilol, hidroxifenilcarvedilol e O-desmetilcarvedilol, os desvios do estudo de precisão e exatidão foram inferiores a 15% e não foi observado efeito matriz. A farmacocinética avaliada por modelo não compartimental mostra acúmulo plasmático dos enantiômeros (R)-(+)-carvedilol, (R)-(+)-O-desmetilcarvedilol e (R)-(+)- hidroxifenilcarvedilol. A disposição cinética e o metabolismo dos enantiômeros do carvedilol não diferem entre os grupos de voluntários não comedicados e comedicados com dose única oral de glibenclamida e metformina. No entanto, os valores de AUC de ambos os enantiômeros do metabólito O-desmetilcarvedilol [(R)- (+): 6,92 vs 10,40 vs 11,91 ng.h/mL e (S)-(-): 2,36 vs 4,26 vs 3,98 ng/h/mL] são menores no grupo de pacientes portadores de diabetes mellitus tipo 2 quando comparados ao grupo de voluntários sadios não comedicados ou comedicados. Em compensação, os valores de AUC de ambos os enantiômeros do metabólito hidroxifenilcarvedilol [(R)-(+): 13,89 vs 6,60 vs 4,88 ng.h/mL e (S)-(-): 7,21 vs 1,50 vs 1,45 ng/h/mL] são maiores no grupo de pacientes portadores de diabetes mellitus tipo 2. Os parâmetros farmacocinéticos de um paciente metabolizador lento do CYP2D6, portador de diabetes mellitus tipo 2 e em tratamento com glibenclamida e metformina, permite inferir redução na formação de ambos os enantiômeros do metabólito hidroxifenilcarvedilol compensada pelo aumento na formação de ambos os enantiômeros do metabólito O-desmetilcarvedilol. Logo, a disposição cinética de ambos os enantiômeros do carvedilol sob a forma inalterada não difere entre metabolizadores rápidos e lentos do CYP2D6. O modelo não linear de efeitos mistos para a análise da disposição cinética e metabolismo populacional dos enantiômeros do carvedilol foi desenvolvido no NONMEM v.7.2 é preciso e possui capacidade preditiva adequada avaliada por métodos visuais do ajuste do modelo aos dados e ii bootstrap. Os valores de biodisponibilidade estimados pelo modelo para os enantiômeros (S)-(-) e (R)-(+)-carvedilol, respectivamente 16,43 e 25,4%, não diferem entre voluntários sadios e pacientes portadores de diabetes mellitus tipo 2 em tratamento com glibenclamida e metformina. Os valores de clearance pelo CYP2D6 estimados para o (S)-(-)-carvedilol foram de 1,65 vs 7,28 L/h, respectivamente, para os voluntários sadios e para os pacientes diabéticos, enquanto os estimados para o enantiômero (R)-(+)-carvedilol foram de 2,69 vs 13,7 L/h. Em relação ao clearance pelo CYP2C9, os valores estimados para o (S)-(-)-carvedilol foram de 16,2 vs 7,71 L/h, respectivamente, para os voluntários sadios e para os pacientes diabéticos, enquanto os estimados para o enantiômero (R)-(+)-carvedilol foram de 25,6 vs 10,5 L/h. Os valores de clearance por outras vias metabólicas são maiores para o (S)-(-)- carvedilol do que para o (R)-(+)-carvedilol (28,2 vs 4,86 L/h) e não diferem entre voluntários sadios e pacientes. Os valores de clearance total de ambos os enantiômeros do carvedilol não diferem entre os voluntários sadios e os pacientes portadores de diabetes mellitus tipo 2 [(S)-(-): 46,05 vs 43,19 L/h e (R)-(+): 33,15 vs 29,06 L/h], considerando que os menores clearances do CYP2C9 são compensados por maiores clearances do CYP2D6. A variação da frequência cardíaca induzida pelo exercício isométrico com o handgrip após a administração de dose única oral de 25 mg de carvedilol racêmico não mostra relação com as concentrações plasmáticas do (S)-(-)-carvedilol. / Carvedilol is an antihypertensive available as racemic mixture, the (S)-(-)- carvedilol is a ??and ?1 adrenergic blocker and (R)-(+)-carvedilol is only na ?1- adrenergic blocker. Carvedilol is metabolized primarily by glucuronidation and by CYP2D6 to hidroxifenilcarvedilol and CYP2C9 to O-desmetilcarvedilol. This study evaluates the disposition and metabolism of carvedilol, hidroxyphenilcarvedilol and Odesmethylcarvedilol enantiomers in health (n=13) and type II diabetes subjects treated with glibenclamide (5 mg/8h) and in a good glycemic control (n=13) and in a CYP2D6 poor metabolizer diabetes subject (n=1). The subjects received a single racemic carvedilol dose of 25 mg. blood samples wore collected until 24h. The heart rate was evaluated durig isometric handgrip exercise. Carvedilol and metabolites enantiomers wore evaluated in plasma sampels by LC-MS/MS. The pharmacokinetics was evaluate by noncompartimental model and higher levels of (R)-(+)-carvedilol, (R)-(+)-Odesmethylcarvedilol e (R)-(+)-hidroxiphenilcarvedilol levels are noticed. The carvedilol pharmacokinetics does not change between healthy and type II diabetes subjects. However the AUC values of both O-desmethylcarvedilol enantiomers are lower [(R)-(+): 6,92 vs 10,40 vs 11,91 ng.h/mL e (S)-(-): 2,36 vs 4,26 vs 3,98 ng/h/mL]in the diabetes subjects and the AUC values of both hidroxyphenilcarvedilol enantiomers are higher [(R)-(+): 13,89 vs 6,60 vs 4,88 ng.h/mL e (S)-(-): 7,21 vs 1,50 vs 1,45 ng/h/mL]. It is noticed in CYP2D6 poor metabolizer diabetes subject lower levels of hidroxyphenilcarvedilol but higher of O-desmetilcarvedilol and carvedilol disposition is not changed. A Non-linear mixed effects modelling was performed in NONMEM v.7.2 the model was validated by visual methods and bootstrap. The bioavailability of (S)-(-) and (R)-(+)-carvedilol was 16,43 e 25,4% and no covariate effect was noticed. The CYP2D6 clearance values were 1,65 vs 7,28 L/h to healthy and diabetes subjects, inasmuch (R)-(+)-carvedilol ones were 25,6 vs 10,5 L/h. The CYP2C9 clearance of (S)-(-)-carvedilol were 16,2 vs 7,71 L/h for healthy and diabetes subjects, while (R)- (+)-carvedilol ones were 25,6 vs 10,5 L/h. The (S)-(-)-carvedilol clearance by other metabolic routes are higher (28,2 vs 4,86 L/h) and does not change between healthy and diabetes subjects. Carvedilol total clearance also does not differ between healthy and diabetes subjects the because the lower CYP2C9 clearance are balanced by the higher CYP2D6 clearance. The cardiac frequency change induced by handgrip isometric exercise is not related with the (S)-(-)-carvedilol plasma levels.

La bioéquivalence, une mesure de substitution de l'innocuité et de l'efficacité à différents stades du processus de développement des médicaments, est tout particulièrement importante lors du développement d'un médicament générique. Entre autres critères, la bioéquivalence garantit que les médicaments génériques sont équivalents aux produits innovateurs ou de références approuvés en termes d’efficacité clinique et d’innocuité tout en contournant le long cours et le coût élevé des essais chez les animaux et des essais cliniques chez les patients exigés pour les médicaments innovants. Malgré les avancées dans le développement d'approches robustes au cours des dernières décennies, la pratique actuelle de la bioéquivalence fait toujours l'objet de controverses. Le but de cette thèse est d'explorer certaines de ces controverses et de les aborder en proposant des approches nouvelles et alternatives. L'une des questions les plus controversées dans la pratique actuelle de la bioéquivalence est l'extrapolation des résultats d'études de bioéquivalence d'une population à une autre. La majorité des études de bioéquivalence portant sur des formes pharmaceutiques orales efficaces par voie systémique reposent sur les critères de pharmacocinétique obtenus chez des sujets sains, alors que la population cible est constituée de patients. Ceci est basé sur l'hypothèse que si deux produits sont bioéquivalents dans une population, ils devraient l'être dans une autre. L'extrapolation des résultats des études de bioéquivalence ne se limite pas à celle des sujets sains aux patients. Depuis 2007, une proportion croissante d'études de bioéquivalence pharmacocinétique portant sur des soumissions génériques nord-américaines ou européennes a été réalisée auprès de populations géographiques/ethniques autres que celles visées, en raison du coût moins élevé de ces études en dehors de l'Amérique du Nord et de l'Europe. Dans le premier volet de cette thèse, nous avons examiné si les résultats de la bioéquivalence obtenus dans une population géographique ou ethnique pouvaient être extrapolés à une autre. À cette fin, nous avons extrait les résultats des études de bioéquivalence pharmacocinétique disponibles publiquement et provenant de soumissions génériques à Santé Canada et à la Food and Drug Administration des États-Unis. Pour dix médicaments différents, nous avons calculé l'effet d’un repas normalisé sur le produit de référence et comparé les résultats obtenus chez deux populations ethniques, les indiens et les nord-américains. Cette approche novatrice est basée sur le raisonnement suivant: si l'effet d’un repas sur le produit de référence est le même chez les populations indienne et nord-américaine, le produit générique et sa référence qui se sont révélés bioéquivalents dans la population indienne devraient également l'être dans la population nord-américaine. Pour 90% des médicaments à l'étude, une différence statistiquement significative a été détectée entre les deux populations après un repas. Pour 30% de ces médicaments, la différence s'est révélée d'une pertinence clinique possible. Les résultats de cette étude ont mis en évidence que l’extrapolation des résultats de bioéquivalence d’une population à l’autre devrait possiblement être reconsidérée pour certains médicaments. Les défis dans le contexte de la bioéquivalence ne se limitent pas toujours aux études pivots où la performance d’un produit générique est comparée à celle de la référence. En effet, une étude pilote peut être menée afin d’établir un protocole d’étude approprié pour cette étude pivot. Par conséquent, les résultats inexacts provenant d'une étude pilote, tels qu'une estimation imprécise du moment ou de la durée d’administration optimale de la dose lors de la comparaison du produit testé par rapport à la référence, pourront affecter négativement les résultats de l’étude de bioéquivalence. Ceci est particulièrement crucial pour les produits indiqués pour un usage topique dermatologique dont les corticostéroïdes constituent un cas d’espèce. En effet, leur bioéquivalence est démontrée par une mesure pharmacodynamique, le blanchiment cutané, à différents temps après application topique. L’intensité du blanchiment est comparée entre le produit générique et le produit de référence à une durée d’administration spécifique d’une dose donnée, la DD50, soit la durée associée à 50% de l’effet maximal observé. Par conséquent, cette durée d’administration de la dose doit d’abord être déterminée dans le cadre d’une étude pilote. L’agence réglementaire américaine recommande l’utilisation d’une approche populationnelle basée sur la modélisation non linéaire à effets mixtes pour l'estimation de la DD50 et ce, quelle que soit la méthode d'analyse. Étant donné qu’il existe différents types de méthodes d’analyse non linéaire à effets mixtes, chaque commanditaire peut en choisir une différente. Dans le deuxième volet de cette thèse, nous avons examiné si les mêmes estimations de DD50 pouvaient être obtenues en utilisant différentes méthodes non linéaires à effets mixtes. À cette fin, nous avons ajusté les données de blanchiment de la peau d’onze études avec deux méthodes non linéaires à effets mixtes différentes : le maximum de vraisemblance avec maximisation de l’espérance (MLEM) et l'estimation conditionnelle de premier ordre (FOCE). Les résultats ont favorisé MLEM, compte tenu d’une meilleure puissance discriminative pour l’estimation de la DD50 de population et d’une meilleure minimisation de la variabilité interindividuelle. Bien que l'approche de la bioéquivalence fondée sur la pharmacocinétique ait contribuée de manière significative au développement de versions génériques de haute qualité des formes pharmaceutiques orales indiquées pour un effet systémique, la disponibilité de versions génériques pour les produits dermatologiques topiques demeure limitée et ce, par manque de méthodes acceptées par les agences réglementaires pour l'évaluation de la bioéquivalence de ces produits. Dans le troisième volet de cette thèse, une nouvelle approche pour l’évaluation de la bioéquivalence de formulations de crème topique d’acyclovir a été développée en utilisant une analyse basée sur un modèle de données d’exposition locales récupérées à partir d’échantillons de peau abrasée prélevés à une seule durée d’administration de la dose, la DD50 à l’aide de bandes adhésives. Un seul échantillonnage de peau effectué à la DD50 a non seulement assuré que les données pharmacocinétiques étaient recueillies à la durée d’administration de la dose ayant le meilleur pouvoir discriminant pour détecter une différence au niveau des formulations, mais a également permis de diminuer considérablement le nombre d'échantillons à analyser. Et surtout, cette nouvelle approche a permis de générer un profil pharmacocinétique au niveau même de la peau. Ce faisant, nous avons pu utiliser l'analyse compartimentale populationnelle et contourner les nombreuses hypothèses et calculs sophistiqués requis par les méthodes précédentes. Notre approche a également permis de générer de nouveaux paramètres pharmacocinétiques permettant de décrire la vitesse et le degré d’exposition cutanée pour l'évaluation de la biodisponibilité et de la bioéquivalence topiques. Finalement, cette méthode a le potentiel de discerner une formulation bioéquivalente d’une autre qui ne l’est pas. / Bioequivalence is a surrogate measure of safety and efficacy in different stages of drug development process with the most pronounced significance in the development of generic drugs. Bioequivalence, among other standards, ensures that generic drugs are equivalent to their approved innovator or reference products in terms of clinical efficacy and safety while circumventing the lengthy-time course and high cost of animal and clinical trials in patients required for innovator drugs. Despite the advancements in development of robust bioequivalence approaches over the past decades, there are still controversies in the current practice of bioequivalence. The aim of this thesis is to explore some of these controversies and address them by putting forward new and alternative approaches. One of the most controversial issues in the current practice of bioequivalence is the extrapolation of bioequivalence study results from one population to another. The majority of bioequivalence studies for systemic effective oral dosage forms are conducted based on pharmacokinetic endpoints in healthy volunteers whilst the targeted population is patients. This is based on the assumption that if two products are bioequivalent in one population, they should be bioequivalent in another one. The extrapolation of bioequivalence study results is not limited to that from healthy volunteers to patients. Since 2007, an ever-increasing proportion of pharmacokinetic bioequivalence studies for North American or European generic submissions have been performed in geographical/ethnic populations other than the intended ones, due to the lower cost of these studies outside North America and Europe. In the first part of this thesis, we investigated whether the bioequivalence results obtained in one geographical or ethnic population can be extrapolated to another one. To this purpose, we extracted pharmacokinetic bioequivalence studies results from generic submissions to Health Canada and the US Food and Drug Administration. We calculated food effect for ten different reference drug products and compared the results for each product between two ethnic populations, Indians and North Americans. This is based on the reasoning that if food effect is found to be the same between the Indian and North American populations, then the generic product and its reference that were found to be bioequivalent in the Indian population should also be bioequivalent in North American population. For 90% of the study drugs, statistically significant difference was detected in the food effect between two populations. For 30% of these drugs, the difference was found to be of possible clinical relevance. The results of this study raised a flag for extrapolating the bioequivalence results from one population to another. Challenges in the context of bioequivalence are not always limited to the pivotal studies where the performance of a generic product is compared to that of Reference. Prior to pivotal bioequivalence studies, a pilot study may be conducted to establish an appropriate study design for the pivotal bioequivalence study. Therefore, inaccurate results from a pilot study, such as inaccurate estimation of time point or dose duration for comparison of test versus reference, can affect the bioequivalence outcomes adversely. An example to this case is the comparison of the extent of skin blanching, the pharmacological effect of generic versus reference products of topical dermatological corticosteroids at specific dose duration, DD50, where the effect is half maximal. This dose duration should initially be determined in a pilot study. The US FDA 1995 Guidance document recommends the use of non-linear mixed effect population modeling for the estimation of DD50, irrespective of the method of analysis. Given the availability of different types of non-linear mixed effect modeling methods, each sponsor could choose a different one. In the second part of this thesis we investigated whether the same DD50 estimates can be obtained when different non-linear mixed effect modeling methods are used. To this purpose, we fitted the skin blanching data from eleven studies with two different non-linear mixed effect modeling methods, the Maximum Likelihood Expectation Maximization (MLEM) and the First Order Conditional Estimation (FOCE). The results favored MLEM given its lower population DD50 estimates that would locate in a more discriminative portion of the Emax curve and better minimization of inter-individual variability. Although the pharmacokinetic-based bioequivalence approach has contributed significantly to the development of high-quality generic versions of systemic effective oral dosage form, the availability of generic versions of topical dermatological products remains constrained due to the limited methods accepted for bioequivalence evaluation of these products. In the third part of this thesis, a novel approach for the bioequivalence assessment of topical acyclovir cream formulations was developed based on the model-based analysis of local exposure data recovered from tape stripping of the skin at a single dose duration, DD50. Conducting the stripping procedure only at DD50 not only ensured that the PK data was collected at the dose duration that is most discriminative of formulation differences, but it also decreased the number of samples to be analyzed significantly. More importantly, our novel approach in generating the local PK profile in the skin (dermatopharmacokinetic profile) and the implementation of population compartmental analysis circumvented the numerous assumptions and sophisticated calculations that were inherent to previous methods, while yielding the PK parameters relevant for topical bioavailability and bioequivalence assessment (rate and extent of exposure to the skin). This method successfully concluded bioequivalence and its absence.

Bei der adjuvanten Immuntherapie soll das Immunsystem von Tumorpatienten gezielt gegen Mikrometastasen aktiviert werden, die nach der operativen Entfernung des Primärtumors im Körper verbleiben. Tumor-assoziierte Antigene (TAA) spielen dabei die zentrale Rolle. Um der Heterogenität eines Tumors in der Expression einzelner TAAs Rechnung zu tragen, werden in modernen Vakzinierungsstrategien Pools von verschiedenen TAAs eingesetzt. Für das Mammakarzinom sind aber bislang nur wenige TAAs bekannt. Auf der Suche nach unbekannten Genen mit Mamma- bzw. Mammakarzinom-restringierter Expression in der Transkriptomdatenbank GeneExpress® wurde ein EST gefunden, das nur in 2 % der getesteten weibl. Normalgewebe ohne Mamma mit einer marginalen mittleren Expression von 16 FE (Fluoreszenzeinheit) detektiert wurde, aber in 63 % der Mammanormalgewebe (159 FE) und in 61 % der Mammakarzinome (339 FE) exprimiert wurde. Das korrespondierende UniGene-Cluster gab erste Hinweise auf die Zugehörigkeit dieses neuen Gens zu der Cytochrom P450 (CYP) Familie. Durch computergestützte Homologievergleiche mit verwandten Mitgliedern dieser Familie in Kombination mit RT-PCR konnte die cDNA-Sequenz dieses neuen CYP ermittelt werden: Durch Amplifikation der Gesamtlängen-cDNA wurden drei Transkripte in der Mammakarzinomzelllinie SK-BR-3 gefunden, die von zwei neuen CYP Genen stammen, CYP4Z1 und dem Pseudogen CYP4Z2P. Die cDNA von CYP4Z1 kodiert für ein 505 As großes Protein, das aufgrund von Homologien einer neuen Subfamilie innerhalb der CYP4 Familie zugeordnet werden kann. Sowohl die Sequenz als auch die vorhergesagte Sekundärstruktur zeigen alle charakteristischen Merkmale eines funktionellen Mitglieds dieser Familie. Aufgrund einer Nonsensemutation in Exon 8 kodiert die cDNA von CYP4Z2P (1436 bp) für ein verkürztes, nichtfunktionelles P450 von 340 As, das zu 96% identisch mit P450 4Z1 ist. Außerdem wurde in SK-BR-3 eine Spleißvariante von CYP4Z1 identifiziert. CYP4Z1 (50,8 kb) und CYP4Z2P (57,3 kb) liegen auf Chromosom 1p33-p34.1 und bestehen aus 12 Exons mit konservierten Exon-Intron-Grenzen. CYP4Z2P ist aus einer inversen Duplikation von CYP4Z1 hervorgegangen. Mittels Realtime RT-PCR mit cDNA von 17 Normalgeweben von gepoolten Spendern und von Mammakarzinomen konnte die auf Brustdrüsengewebe restringierte Expression beider Gene demonstriert werden: Die Expression von CYP4Z1 war in Mammakarzinomgewebe 3,6-mal höher als in Mammanormalgewebe, 60-mal höher als in Lunge und 84-mal höher als in Leber. Alle anderen getesteten weibl. Normalgewebe zeigten eine noch geringere Expression. Ein ähnlich stringentes Expressionsprofil ergab die Analyse von CYP4Z2P, allerdings mit einer deutlich niedrigeren Expressionsstärke. Das Mamma-restringierte Expressionsverhalten von CYP4Z1 wurde durch einen ?Cancer-Profiling-Array? (241 Tumor-/Normalgewebepaare von 13 Gewebetypen) bestätigt. Damit konnte gezeigt werden, dass CYP4Z1 bei 52 % der getesteten 50 Brustkrebspatientinnen im Tumor versus peritumoralem Normalgewebe überexprimiert war. Mit einem spezifischen Kaninchen-Antiserum konnte die Expression von P450 4Z1 Protein sowohl in CYP4Z1-transduzierten Zelllinien als auch in Mammagewebeproben mittels Western-Blot nachgewiesen werden. Konfokale Laser-Scanning Mikroskopie von MCF-7 Zellen, die das Fusionsprotein CYP4Z1-EGFP exprimierten, und die subzelluläre Fraktionierung der CYP4Z1-Transduktanten zeigten P450 4Z1 als integrales Membranprotein des endoplasmatischen Retikulums (ER). Für die Lokalisation und die Zurückhaltung im ER ohne Recycling aus dem Prä-Golgiapparat sind die ersten 32 As erforderlich, was Studien mit unterschiedlichen Deletionsmutanten aus der N-terminalen Sequenz von 4Z1 zeigten. Die Entdeckung eines neuen Mamma-spezifisch exprimierten P450 Enzyms eröffnet neue Möglichkeiten sowohl in Hinsicht auf eine Immuntherapie von Brustkrebs als auch für die Entwicklung neuer Chemotherapeutika, die spezifisch durch P450 4Z1 am Tumorort umgesetzt werden können.

Die Phänotypisierung von Cytochrom P450 (CYP) und Leberfunktionstests sind wichtige Methoden in der Klinik. Die Methoden nutzen die Pharmakokinetik (PK) von Testsubstanzen und ihren Metaboliten, um Einblicke in die Stoffwechselkapazität der Leber und in die Aktivität von Enzymen und Transportern zu gewinnen. Die Leberfunktionstests werden nicht nur von zahlreichen Proband:innenmerkmalen, sondern auch von den Besonderheiten der Untersuchung beeinflusst. Eine zentrale Herausforderung besteht darin, die verschiedenen Faktoren, die das Ergebnis der Messungen beeinflussen, voneinander zu trennen, um ihren jeweiligen Einfluss auf das Messergebniss zu untersuchen. In dieser Arbeit wurde die Herausforderung durch Metaanalysen und physiologisch basierte Pharmakokinetik Modellierung (PBPK) angegangen. Es wurde eine offene Pharmakokinetik-Datenbank (PK-DB) entwickelt und PK-Daten für ein breites Spektrum von Testsubstanzen kuratiert. Meines Wissens enthält PK-DB derzeit den größten offenen PK-Datensatz zu Testsubstanzen. Der Datensatz ermöglichte die Identifizierung und Quantifizierung von demografischen und rassischen Bias (Geschlecht, ethnische Zugehörigkeit, Alter, Gesundheitszustand), Meldefehlern und Unstimmigkeiten in der Literatur. Auf der Grundlage der Daten wurde eine Metaanalyse der PK von Koffein im Hinblick auf verschiedene Faktoren bzgl. Leberfunktion und CYP1A2-Aktivität durchgeführt. Insbesondere wurde das vorhandene Wissen über die Auswirkungen des Rauchens, der Einnahme oraler Verhütungsmittel, verschiedener Krankheiten und Begleitmedikationen auf die PK von Koffein durch Metaanalysen und Datenintegration konsolidiert. Ebenso wurde die Messgenauigkeit der Koffeinkonzentration in Bezug auf den Messprotokol analysiert. Darüber hinaus wurde der Einfluss des CYP2D6-Polymorphismus untersucht. Hierzu wurde ein PBPK-Modell für Dextromethorphan und seine Metaboliten Dextrorphan und Dextrorphan O-Glucuronid entwickelt und mit den PK-Daten kalibriert und validiert. / Cytochrome P450 (CYP) phenotyping and dynamic liver function testing are essential methods in clinical practice. These methods utilize the pharmacokinetics (PK) of test substances and their metabolites to gain insight into the liver's metabolic capacity and the activity of enzymes and transporters. Liver function tests are not only influenced by numerous characteristics of a studied subject but also by the specifics of individual study procedures. A key challenge is to disentangle the various factors which influence the outcome of the measurements from each other to study their influence on the dynamic liver function and CYP phenotype. In this work, the challenge was addressed through meta-analysis and physiologically based pharmacokinetic modeling. As a foundation, an open pharmacokinetics database was developed and pharmacokinetics data were curated for a wide range of test substances. To my knowledge, PK-DB currently contains the largest open pharmacokinetic dataset on substances used for phenotyping and dynamical liver function testing. The dataset allowed for identifying and quantifying demographic and racial bias (sex, ethnicity, age, health), reporting errors, and inconsistencies in pharmacokinetic literature. Based on the data, a caffeine pharmacokinetics meta-analysis was conducted concerning various factors affecting liver function and CYP1A2 activity. In particular, meta-analysis and data integration solidified existing knowledge on the effects of smoking, oral contraceptives, multiple diseases, and co-medications on caffeine pharmacokinetics. Similarly, the measurement accuracy of caffeine concentration was investigated with respect to various aspects of the measurement protocol. In addition, the impact of CYP2D6 polymorphism was investigated. Therefore, a PBPK model of dextromethorphan (DXM) and its metabolites dextrorphan (DXO) and dextrorphan O-glucuronide (DXO-Glu) was developed, and calibrated, and validated with pharmacokinetics data.

Minimizing anti-cancer drug toxicity is a major challenge for the pharmaceutical industry. Toxicity is most frequently due to either the direct interaction of the drug on previously unidentified targets or its conversion to metabolites by drug metabolizing enzymes (e.g. CYP450 enzymes) that cause cellular, tissue or organ damage. Pharmacoproteomics is beginning to take a central role in studying changes in protein expression corresponding to drug administration, the results of which, inform about the mode of action, toxicity, and resistance in pre-clinical and clinical stages of drug development. The main aim of this research is to apply comparative proteomics studies on livers from male and female mice xenograft models treated with major anti-cancer drugs (5-flourouracil, paclitaxel, cisplatin, and doxorubicin) and CYP inducer, TCPOBOP, to investigate their effect on protein expression profiles (proteome). Within this thesis, an attention is paid to optimise a highly validated proteomics workflow for biomarker identification. Proteins were extracted from liver microsomes of mice treated in two separate sets; Set A – male (5-fluoruracil, doxorubicin, cisplatin and untreated) or Set B – female (5-fluoruracil, paclitaxel, TCPOBOP and untreated) using cryo-pulverization and sonication method. The extracts were digested with trypsin ii and the resulting peptides labelled with 4-plex iTRAQ reagents. The labelled peptides were subjected for separation in two-dimensions by iso-electric focusing (IEF) and RP-HPLC techniques before analysis by mass spectrometry and database searching for protein identification. Set A and Set B resulted in identification and quantification of 1146 and 1743 proteins, respectively. Moreover, Set A and Set B recovered 26 and 34 cytochrome P450 isoforms, respectively. The microsomal changes after drug treatments were quite similar. However, more changes were observed in the male set. Up-regulation of MUPs showed the greatest distinction in the protein expression patterns in the treated samples comparing to the untreated controls. In Set A, 5-fluoruracil and cisplatin increased the expression of three isoforms (MUP1, 2, and 6), whereas doxorubicin has increased the expression of four isoforms (MUP1, 2, 3, and 6). On the other side, only TCPOBOP in Set B has increased the expression of two isoforms (MUP1 and 6). Our findings showed that the expression of MUP, normally involved in binding and excretion of pheromones, have drug- and sex-specific differences. The mechanism and significance of MUP up-regulation are ambiguous. Therefore, the impact of each therapeutic agent on MUP and xenobiotic enzymes will be discussed.

The DDT metabolite 3-methylsulfonyl-DDE (3-MeSO2-DDE) is bioactivated by cytochrome P450 11B1 (CYP11B1) in the adrenal cortex of mice and forms irreversibly bound protein adducts, reduces glucocorticoid secretion, and induces cell death selectively in cortisol-producing adrenocortical cells. 3-MeSO2-DDE has therefore been proposed as a lead compound for an improved adrenocortical carcinoma (ACC) therapy. The aims of this thesis were to (1) develop in vitro test systems based on murine and human adrenocortical cell lines and to (2) investigate the mechanisms behind 3-MeSO2-DDE toxicity in adrenocortical cells. The cytotoxic and endocrine-modulating effects of 3-MeSO2-DDE were compared to those of o,p′-DDD (mitotane), the current ACC therapy, and to those of several structurally analogous compounds in both murine and human cell lines. 3-MeSO2-DDE bioactivation and cytotoxicity proceeded in a similar manner in the murine adrenocortical Y-1 cell line as in mice in vivo. The effects were highly structure-specific. Moreover, 3-MeSO2-DDE formed irreversibly bound protein adducts and caused cell death also in the human H295R cell line, and was slightly more cytotoxic than o,p′-DDD. However, 3-MeSO2-DDE toxicity in human cells was not affected by the CYP11B1 inhibitor etomidate, suggesting that bioactivation in human cells is performed by additional/other enzyme(s) than CYP11B1. 3-MeSO2-DDE generated biphasic responses in cortisol and aldosterone secretion and in expression levels of the steroidogenic genes CYP11B1, CYP11B2, and StAR. Such hormesis-like responses were not seen for o,p′-DDD or the precursor DDT metabolite p,p′-DDE. In addition, the two o,p′-DDD enantiomers (R)-(+)-o,p′-DDD and (S)-(-)-o,p′-DDD exhibited slight differences in cytotoxic and endocrine-modulating activity in H295R cells. In conclusion, this thesis  provides  extended  knowledge  on  the  mechanisms  of  action  of 3-MeSO2-DDE and points out important differences in effects between murine and human cells. Lead optimisation studies of 3-MeSO2-DDE using the herein presented in vitro test systems are ongoing.

Drug metabolism plays an important role in drug absorption and drug elimination. Therefore, it is crucial to understand the mechanism and kinetics of drug metabolism by various drug-metabolizing enzymes (DMEs). Cytochrome P450 enzymes (CYPs) are responsible for the metabolism of more than 60% of the top 200 prescribed drugs. X-ray and NMR data of CYP enzyme suggest that relatively large and flexible active sites are capable of multi-substrate binding. Due to the multiple substrate-binding, CYP reactions tend to show non-Michaelis Menten kinetics (atypical kinetics), multiple metabolite formation and sequential metabolism.To investigate the complexity of cytochrome P450 kinetics, saturation curves and intrinsic clearances (CLint) were simulated for single substrate and multi-substrate models using rate equations and numerical analysis. These models were combined with multiple product formation and sequential metabolism and simulations were performed with random error. All simulation and model fitting was performed using Mathematica. A concentration-dependent metabolite ratio plot can be observed from multi-substrate binding kinetics. Use of single substrate models to characterize multi-substrate data can result in inaccurate kinetic parameters and poor clearance predictions. It has been shown that use of different substrate concentrations may lead to highly variable in vitro CLint estimations when sigmoidal kinetics are observed. Comparing results for use of standard velocity equations with ordinary differential equations (ODEs) clearly shows that ODEs are more versatile and provide better parameter estimates. It would be difficult to derive concentration-velocity relationships for complex models, but these relationships can be easily modeled using numerical methods and ODEs. The model drug diazepam (DZP) was chosen as the probe substrate to demonstrate complex CYP kinetics with specific CYP enzyme sources, including rat liver microsome (RLM), human liver microsome (HLM), purified CYP enzyme isoforms and rat hepatocytes. All saturation curves display non-Michaelis-Menten kinetics, form multiple primary metabolites, and are sequentially metabolized to secondary metabolites. In addition, the sequential metabolism and disposition would be characterized in hepatocytes incubation under flow conditions. To provide in vivo evidence of the atypical kinetics and investigate CYP-mediated sequential metabolism, preliminary intravascular (IV) dosing PK studies with male rats was performed for DZP. In general, DZP and its metabolites were quantitated by LC/MS/MS. Numerical methods were used to solve ODEs and parameterize micro and macro rate constants for the models. It has been shown that more complex models that include explicit enzyme-product complexes can well characterize the datasets for diazepam sequential metabolism with CYP3A4. Uncommon DZP metabolite PK profiles are observed in rat PK studies. In summary, methods of in vitro data analysis are compared, new assays are developed, and new modeling approaches for complex drug and metabolite pharmacokinetics are being investigated. / Pharmaceutical Sciences

Minimizing anti-cancer drug toxicity is a major challenge for the pharmaceutical industry. Toxicity is most frequently due to either the direct interaction of the drug on previously unidentified targets or its conversion to metabolites by drug metabolizing enzymes (e.g. CYP450 enzymes) that cause cellular, tissue or organ damage. Pharmacoproteomics is beginning to take a central role in studying changes in protein expression corresponding to drug administration, the results of which, inform about the mode of action, toxicity, and resistance in pre-clinical and clinical stages of drug development. The main aim of this research is to apply comparative proteomics studies on livers from male and female mice xenograft models treated with major anti-cancer drugs (5-flourouracil, paclitaxel, cisplatin, and doxorubicin) and CYP inducer, TCPOBOP, to investigate their effect on protein expression profiles (proteome). Within this thesis, an attention is paid to optimise a highly validated proteomics workflow for biomarker identification. Proteins were extracted from liver microsomes of mice treated in two separate sets; Set A – male (5-fluoruracil, doxorubicin, cisplatin and untreated) or Set B – female (5-fluoruracil, paclitaxel, TCPOBOP and untreated) using cryo-pulverization and sonication method. The extracts were digested with trypsin ii and the resulting peptides labelled with 4-plex iTRAQ reagents. The labelled peptides were subjected for separation in two-dimensions by iso-electric focusing (IEF) and RP-HPLC techniques before analysis by mass spectrometry and database searching for protein identification. Set A and Set B resulted in identification and quantification of 1146 and 1743 proteins, respectively. Moreover, Set A and Set B recovered 26 and 34 cytochrome P450 isoforms, respectively. The microsomal changes after drug treatments were quite similar. However, more changes were observed in the male set. Up-regulation of MUPs showed the greatest distinction in the protein expression patterns in the treated samples comparing to the untreated controls. In Set A, 5-fluoruracil and cisplatin increased the expression of three isoforms (MUP1, 2, and 6), whereas doxorubicin has increased the expression of four isoforms (MUP1, 2, 3, and 6). On the other side, only TCPOBOP in Set B has increased the expression of two isoforms (MUP1 and 6). Our findings showed that the expression of MUP, normally involved in binding and excretion of pheromones, have drug- and sex-specific differences. The mechanism and significance of MUP up-regulation are ambiguous. Therefore, the impact of each therapeutic agent on MUP and xenobiotic enzymes will be discussed.

The full text will be available at the end of the embargo: 17th July 2024