Open Access

Die Kenntnis über die im Gastrointestinaltrakt ablaufenden Prozesse spielt in der Entwicklung neuer Arzneiformen eine entscheidende Rolle. Besonders im Dickdarm ist dabei neben den physiologischen Bedingungen die bakterielle Besiedlung zu beachten, welche sowohl inter- als auch intraindividuell hoch variabel ist. Bislang gibt es keine einheitliche Methode zur Untersuchung des Einflusses der intestinalen Mikrobiota auf die Metabolisierung von Arzneistoffen. Diese Methoden sind jedoch entscheidend für das Verständnis des Einflusses der bakteriellen Metabolisierung auf die Pharmakokinetik und -dynamik der Arzneistoffe. Übergeordnetes Ziel dieser Arbeit war es, ein In vitro-Modell zu entwickeln und anzuwenden, welches die dynamischen Bedingungen im Colon ascendens, insbesondere im Hinblick auf die pH-Werte, Durchmischung und bakterielle Besiedlung, darstellt. Um dieses Ziel zu erreichen, wurde im Rahmen erster Versuche untersucht, wie es sowohl mit monographierten als auch biorelevanten Modellen möglich ist, die mechanische Belastung, die auf eine Arzneiform im GIT ausgeübt wird, darzustellen. Die Verwendung der SmartPill™ eröffnete die Möglichkeit, in den Apparaturen auftretende Drücke aufzuzeichnen. Außerdem konnten die gemessenen Drücke anschließend mit Daten aus In vivo-Studien verglichen werden. Die Untersuchungen ergaben, dass in den monographierten Apparaturen keine Drücke auftreten, die den während der Magen-Darm-Passage auftretenden Drücken entsprechen. Im Gegensatz dazu können im DOFTA gezielt Drücke und so auch vollständige Druckprofile simuliert werden. Im weiteren Verlauf der Arbeit waren die zuvor gewonnenen Erkenntnisse hilfreich für die Entwicklung des neuen Modells zur Darstellung des Colon ascendens. In das MimiCol wurden pH-Wert-Daten aus einer SmartPill™-Studie implementiert. Die Vorteile des neuartigen Bioreaktors MimiCol sind das kleinere Medienvolumen, das den In vivo-Bedingungen näherkommt, die Möglichkeit, Medienwechsel durchzuführen und dadurch Metabolite abzuführen und neue Nährstoffe hinzuzufügen sowie die genauere Simulation von In vivo-Durchmischungsmustern. Ziel der durchgeführten Untersuchung war der Vergleich der Metabolisierung des Modellarzneistoffs Sulfasalazin in dem neuartigen dynamischen Bioreaktor MimiCol und einem statischen Standard-Batch-Fermenter. Beide wurden mit der gleichen, kryokonservierten fäkalen Standardmikrobiota beimpft. Die Experimente zeigten, dass das MimiCol in der Lage ist, die dynamischen Bedingungen im aufsteigenden Dickdarm zu simulieren. Die dynamischen Bedingungen im MimiCol führten zu einer Verdopplung der Metabolisierungskonstanten im Vergleich zum statischen Batch-Fermenter. Das MimiCol ahmt, besonders in Bezug auf pH-Fluktuationen und Bakterienwachstum, die dynamischen Bedingungen im aufsteigenden Dickdarm nach und könnte sich in allen Phasen der Arzneimittel- und Formulierungsentwicklung als nützlich erweisen. Zur Erleichterung und Beschleunigung der Datengenerierung wurde im nächsten Schritt eine Erweiterung des Modells angestrebt. Hierbei war es die größte Herausforderung, die ursprünglichen Parameter auf ein erweitertes Modell mit einer anderen Steuerung und anderen Komponenten zu übertragen. Außerdem wurde in diesem Zuge die Charakterisierung komplexer Bakterienkulturen mittels 16S rRNA-Sequenzierung eingeführt. Bei der Erweiterung des Modells wurde besonderes Augenmerk auf die Einfachheit des Designs und die leichte Skalierbarkeit gelegt. Um zu beweisen, dass die Übertragung der Parameter erfolgreich war, wurde erneut der Abbau von Sulfasalazin untersucht und die bakterielle Zusammensetzung während des Experiments durch 16S rRNA-Sequenzierung analysiert. Die Übertragung der Versuchsbedingungen auf das neue Modell war erfolgreich. Kommerziell erhältliche Komponenten wurden in den Aufbau implementiert. Das Modell MimiCol³ repräsentierte das Colon ascendens in seinen Eigenschaften bezüglich des Volumens, pH-Werts und Redoxpotentials zufriedenstellend. Die 16S rRNA-Sequenzierung führte zu weiteren Erkenntnissen über die bakterielle Zusammensetzung in den drei Gefäßen. Der Abbau von Sulfasalazin stand in guter Übereinstimmung mit den In vivo-Daten und den im MimiCol gewonnenen Daten. Das neue Modell des Colon ascendens MimiCol³ ermöglichte es, zuverlässigere Daten zu sammeln, da drei Experimente gleichzeitig unter denselben Bedingungen durchgeführt wurden. Die durchgeführten Untersuchungen zeigen, dass ein wichtiges Instrument zur Untersuchung des Einflusses unseres Mikrobioms im Darm auf den Abbau von Arzneistoffen und Arzneiformen entwickelt wurde.

Die Kv7-Kaliumkanalöffner Flupirtin und Retigabin waren wertvolle Alternativen bei der Pharmakotherapie von Schmerzen und Epilepsie. Beide Wirkstoffe werden aufgrund von unerwünschten Arzneimittelwirkungen derzeit jedoch nicht mehr eingesetzt. Die Flupirtin-induzierte Hepatotoxizität und die durch Retigabin hervorgerufenen Gewebeverfärbungen scheinen dabei auf den ersten Blick nicht zusammenzuhängen. Gleichwohl lassen sich wahrscheinlich beide Nebenwirkungen auf das gemeinsame oxidationsempfindliche Triaminoaryl-Grundgerüst zurückführen, welches zur Bildung von reaktiven Chinondiimin-Metaboliten neigt. Da hingegen der Wirkungsmechanismus, d. h. die Öffnung der Kv7-Kanäle, nicht an der Toxizität beteiligt zu sein scheint, hatte diese Arbeit zum Ziel, sicherere Alternativen für Flupirtin und Retigabin zu entwickeln. In einem Liganden-basierten Ansatz wurde eine Umgestaltung des Triaminoaryl-Kerns, den beide Wirkstoffe gemeinsam haben, vorgenommen, was zu Carba-Analoga führte, die durch eine erhöhte Oxidationsbeständigkeit sowie ein vernachlässigbares Risiko für die Bildung von chinoiden Metaboliten charakterisiert sind. Zusätzlich zu diesen verbesserten Sicherheitsmerkmalen offenbarten einige der neuartigen Derivate eine überlegene Kv7.2/3-Kanalöffnungsaktivität. Im Vergleich zu Flupirtin konnte die Potenz der Verbindungen um den Faktor 150 gesteigert werden, während die intrinsische Aktivität auf bis zu 176 % verbessert werden konnte, was die betreffenden Carba-Analoga zu vielversprechenden Kandidaten für eine weitergehende Entwicklung macht. Andererseits ermöglichten einige inaktive Verbindungen sowie die insgesamt deutlich abgestuften Kv7.2/3-Aktivitätsdaten die Etablierung von validen Struktur-Wirkungs-Beziehungen und Hypothesen zum Bindungsmodus, die mit Dockingergebnissen und Molekulardynamik-Simulationen korrelierten.

Pancreatitis is an inflammatory disorder of the pancreas with a mortality rate of 5% and severe negative effects on the quality of life. Of all non-malignant gastrointestinal diseases, it is the most common reason for hospitalization. Pancreatitis is a disease of multiple etiologies with different underlying pathomechanisms. Due to the diversity of mechanisms by which homeostasis within the exocrine pancreas can be disrupted, finding appropriate therapeutic approaches is challenging. Current treatment options are inadequate and are mostly limited to supportive treatment like fluid administration, bowel rest, antibiotics and pain control. Although significant advancements have been achieved in recent decades, the mortality rate for pancreatitis has not decreased. Furthermore, progress is slow due to limited patient sample availability and lack of an appropriate cell model. Taking samples from a human pancreas is typically avoided, because damaging the pancreatic tissue can itself induce pancreatitis. Additionally, while it is possible to keep individual acini in culture, it is not possible to grow pancreatic acinar cells. Thus, less appropriate cell models, often derived from pancreatic cancer samples, have to be used. The most common animal model for pancreatitis is mice, with caerulein administration being the most common method of inducing pancreatitis. However, the use of animal models has significant drawbacks, as they are time-consuming, costly, and pose ethical questions. Furthermore, exposing the pancreas to appropriate stimuli in animal models is difficult. For example, alcohol is the leading cause of pancreatitis in humans, but is typically avoided by animals. Thus, alcohol feeding methods had to be developed to overcome the natural aversion of rodents to alcohol. Results obtained from animal models are also often not transferable into clinical trials and outcomes in humans remain largely unpredictable. Due to the lack of experimental models, our understanding of this highly complex disease is still limited and significant progress is required for the development of effective therapy options. In this dissertation recombinantly expressed trypsin isoforms and variants of the serine protease inhibitor Kazal-type 1 (SPINK1) inhibitor are used to investigate mechanisms, by which tryptic activity is regulated in pancreatic acinar cells. With premature tryptic activity in the exocrine pancreas being the common focal point of most etiologies connected to pancreatitis, trypsin represents by far the most promising target for treating pancreatitis. Understanding the mechanisms by which the pancreas protects itself and rationalizing mutations that can undermine these protective mechanisms, are important steps towards developing effective therapies.

Universal products provide an axiomatic framework to study noncommutative independences general enough to include, besides the well known "single-faced" case (i.e., tensor, free, Boolean, monotone and antimonotone independence), also more recent "multi-faced" examples like bifree independence. Questions concerning classification have been fully answered in the single-faced case, but are in general still open in the multi-faced case. In this thesis we discuss how one can use insights in the relation between universal products and their associated moment-cumulant formula as a starting point towards a combinatorial approach to (multi-faced) universal products. We define certain classes of partitions and discuss why the defining axioms are sufficient to associate to each of them a multi-faced universal product. For the two-faced case we present our result that every positive and symmetric universal product can be produced in this fashion and we outline how these results might contribute to a classification of positive and symmetric universal products.

Infective endocarditis (IE) is a potentially life-threatening infection of the endocardial surfaces of the heart, most frequently the valves. It is typically caused by bacteria, less commonly by fungi. Over the past years, the morbidity and mortality of IE have gradually increased, and it is now the fourth most common life-threatening infection after sepsis, pneumonia, and intra-abdominal abscess. Despite advances in cardiac imaging and diagnostic techniques, the diagnosis of IE remains challenging. The lack of fast and reliable diagnosis of IE can lead to serious complications. Therefore, new diagnostic and therapeutic tools are urgently needed. This study had two main aims: (i) to investigate whether a pathogen-specific antibody response in IE patients is mounted against different IE pathogens and whether analysis of such a response might be useful for complementing the classical blood culture diagnosis, and (ii) generate and characterize neutralizing monoclonal antibodies (mAbs) against three virulence factors of Staphylococcus aureus (S. aureus), which is the most common etiological agent in IE. Our research group has recently established an xMAP® (Luminex®) technology-based serological assay that simultaneously quantifies the antibody response against 30 different pathogens. Within the research consortium Card-ii-Omics, we conducted a prospective, observational clinical discovery study involving 17 IE patients and 20 controls (i.e., patients with non-infectious heart-related conditions). Plasma samples were obtained on the day of IE diagnosis from all patients, while samples at later dates over the course of infection were available for only some patients. Invasive pathogens were identified by blood culture. The infection array revealed antibodies against a broad range of pathogens in both controls and IE patients, suggesting a broad immune memory. Overall, antibody levels did not significantly differ between both groups, but we observed high antibody titers against those pathogens that were detected by blood culture. Whenever available (in the case of 13/17 IE patients), back-up and follow-up plasma samples (obtained before or after diagnosis, respectively) were included in the analyses that provided valuable information about the kinetics of antibody response during the course of infection. Notably, infection array data confirmed (and extended) the blood culture data in only 2/13 cases. In three cases, serology contradicted the microbiological diagnosis, and in three cases, the infection array was able to identify pathogens, while the microbiological diagnosis failed. In three cases, serology was negative while microbiological diagnosis was positive, and in two cases, both serology and microbiological diagnosis were negative. In 6 out of 8 cases with increases in antibody levels, this response was directed against gut microbes. This supports the leaky gut hypothesis, which assumes that breaching of the gut barrier causes translocation of gut microbes into the bloodstream, which then infect the heart valves. Moreover, we observed an increase in antibody titers in 4 patients against the yeast C. albicans, suggesting a secondary fungal infection. Finally, this study emphasized that the timing of plasma collection is crucial for studying antibody kinetics in IE. After demonstrating that pathogen-specific antibodies are generated during IE, we aimed to generate mAbs against the prime IE pathogen S. aureus and study their functions on a molecular level. Using the hybridoma technology, our research group has recently generated mAbs against two S. aureus surface proteins/adhesion factors (clumping factor A (ClfA) and fibronectin-binding protein A (FnBPA)), both involved in biofilm formation, as well as an extracellular enzyme, the staphylococcal serine protease–like protein B (SplB), a virulence factor. In this work, the sequences of the mAbs were determined from hybridoma RNA. Then those mAbs were produced at a larger scale in order to determine their binding and neutralizing capacities using in vitro assays such as ELISA, Western blot, Dot blot, microscale thermophoresis, and in a mouse model. The anti-SplB mAb specifically targeted SplB, with no cross-reactivity to other Spls or extracellular proteins (ECP) of S. aureus. Though anti-SplB mAb showed moderate binding to SplB with a Kd value of 2.54 μM and high sequence homology to the germline sequence, it neutralized the enzymatic activity of SplB up to 99% in 5-fold molar excess as showed in an in vitro substrate cleavage assay. Previous work showed that SplB facilitates the release of proinflammatory cytokines in endothelial cells and induces endothelial damage in mice. Here, we demonstrated that the anti-SplB mAb efficiently blocked the function of SplB in vivo, thus markedly reducing the damage to the endothelial barrier. In conclusion, we identified the strong neutralizing potential of a mAb against SplB, which merits further investigation as a candidate for the immunotherapy of SplB-induced S. aureus pathologies, including IE. High antibody titers against S. aureus adhesins, including ClfA and FnBPA, have been reported in IE patients. Besides, ClfA is involved in serious S. aureus bloodstream and biofilm-related infections. Similarly, FnBPA facilitates biofilm formation and inhibits macrophage invasion. These important properties make the two bacterial adhesins ideal candidates for a passive vaccination strategy. We generated two murine ClfA-mAbs, ClfA-002 and ClfA-004, which showed strong specificity to ClfA. However, ClfA-004 showed reduced binding strength compared to ClfA-002 due to a single non-synonymous nucleotide change (Phe Tyr) at the CDR3 region. While the ClfA-002 mAb reduced the binding of ClfA to fibrinogen by around 60%, the ClfA-004 had no inhibitory capacity. We also generated two murine and twelve humanized anti-FnBPA mAbs, which showed similar and moderate binding to FnBPA. One murine mAb (anti-FnBPA D4) partially inhibited the binding of FnBPA to fibronectin. FnBPA contains 11 tandem repeats that can all bind to fibronectin. This redundancy could be the reason for the lack of complete inhibition. Hence, in this work, we characterized the properties of neutralizing mAbs against two adhesins of S. aureus. These mAbs should be tested in the future, alone and in combination with other mAbs and antibiotics, for their ability to reduce staphylococcal biofilm formation. In conclusion, we showed that antibody profiling of IE patients can provide valuable insights into the causative agent(s), and can help in guiding the antibiotic therapy. However, sampling is crucial in IE, which often dwells for many weeks before being clinically diagnosed. Because of the severity of IE, which can be life-threatening, I suggest to establish biobanks to store patient samples upon hospital admission that will provide a baseline in case of a later microbial infection. Moreover, our results suggest that C. albicans plays an important and so far underestimated role in IE. In the second part of the thesis, we characterized several mAbs against an S. aureus protease and two adhesins. Of high interest is a neutralizing mAb against SplB, which shows promising results in vitro and in vivo. Further in vitro and in vivo tests need to be conducted to study the anti-biofilm activity of the anti-FnBPA- and anti-ClfA-mAbs and explore their utility as therapeutic agents.