40639 research outputs found
Sort by
Das Immunrezeptorrepertoire in archivierten Lymphknoten von Melanompatienten
No full textDiese Dissertation ist gesperrt bis zum 31.07.2027!In dieser Arbeit wird zum ersten Mal die Zusammensetzung des Immunrezeptorreper-
toires aus einer größeren Menge an Sentinel-Lymphknoten von Patienten mit einem Ma-
lignen Melanom (UICC-Stadium III oder IV) sowohl im Bezug auf die T- als auch auf
die B-Zell-Rezeptorsequenzen untersucht. Dazu wurde die RNA aus insgesamt 118 LK
von 40 Patienten verwendet, welche in den Jahren 1995-2002 in der Universitätshautkli-
nik Tübingen operiert wurden.
Unter der Betrachtung verschiedener Diversitätsparameter wie Klonalität, Richness und
dem Chao1-Index betrachteten wir die Unterschiede im Repertoire im Bezug auf den vor-
liegenden Metastasierungsstatus. Dass diese Eigenschaften der Immunzellrepertoires
eine entscheidende Rolle bei der Immunantwort und Krebsentwicklung spielen, ist schon
aus früheren wissenschaftlichen Arbeiten bekannt.
Es ließ sich weiterhin ein signifikanter Zusammenhang zwischen einem breiteren Immun-
zellrepertoire vor allem in den B-Zell-Loci mit einem längeren Überleben feststellen. Ent-
sprechend korrelierte eine höhere Klonalität mit schlechterem Outcome. Diese Erkennt-
nis könnte theoretisch eine prognostische Anwendung haben. Eine höhere TZR-Diversi-
tät korreliert auch bei anderen Karzinomen oft mit einer besseren Prognose, bei Melano-
men ist die Rezeptordiversität im Blut bereits als prädiktiver Faktor für das Ansprechen
auf Immuntherapien bekannt.
Bei Betrachtung der LK-Repertoires im Hinblick auf Sequenzüberschneidungen intra-
und interindividuell ließen sich je nach Locus unterschiedliche Überschneidungsmuster
beobachten, deren Relevanz noch nicht abschließend geklärt ist.
Insgesamt bilden die erworbenen Erkenntnisse eine wichtige Ergänzung zu den vorhan-
denen Repertoireuntersuchungen im Bezug auf das Maligne Melanom. Die Einzigartig-
keit der Proben in Kombination mit der tiefgehenden Sequenzierung sowohl auf T- als
auch auf B-Zellebene bedingt eine große Menge an aufschlussreichen und spannenden
Daten, die zum besseren Verständnis der Immunantwort des Körpers auf Melanomenti-
täten in den Sentinel-Lymphknoten und der dahinterliegenden Prozesse eine tragende
Rolle spielen können
Size-Dependent Interactions Between Antimony and Iron(III) (Oxyhydr)oxide Nanoparticles: Implications for Adsorption, Sequestration, and Reduction
No full textNanoscale iron(III) (oxyhydr)oxide minerals, typically ranging from 1 to 100 nanometers in size, are common constituents of natural environments. Among these, goethite and hematite (hereafter referred to as iron(III) (oxyhydr)oxides) are especially abundant and geochemically significant in soils and sediments. These minerals occur across a wide range of particle sizes and are susceptible to microbial reduction, affecting the fate and mobility of trace elements, nutrients, and environmental contaminants. However, the size-dependent reduction behavior of iron(III) (oxyhydr)oxides in single and mixed mineral systems remains poorly understood.
Antimony (Sb) has been declared as a priority environmental pollutant, owing to its chronic toxicity and associated health risks. In natural waters, Sb predominantly exists in the +V and +III oxidation states, with Sb(V) prevailing under oxic to slightly reducing conditions, while Sb(III) is stable under reducing conditions. The mobility and fate of both species are tightly linked to sorption and coprecipitation reactions with iron(III) (oxyhydr)oxides. However, critical questions remain unclear, (i) the impact of the particle size and co-existing cation (e.g., Ni2+) on Sb(V) adsorption capacity and retention mechanisms (ii) how Sb(V)-coprecipitated iron(III) (oxyhydr)oxides of different particle sizes behave under microbial Fe(III) reduction, particularly in terms of Sb speciation and distribution.
To address these knowledge gaps, we synthesized [Sb(V)-containing] goethite and hematite of varying particle sizes, then investigated (i) reduction kinetics and extents of goethite and hematite using microbial and mediated electrochemical reduction, (ii) Sb(V) adsorption capacity and retention mechanisms through Sb(V) adsorption and Sb(V)–Ni2+ co-sorption experiments (iii) reductive dissolution behavior of Sb-coprecipitated iron(III) (oxyhydr)oxides and resulting Sb mobilization using Shewanella oneidensis MR-1.
We firstly found that small particles were preferentially reduced relative to their large counterparts in both single and mixed mineral systems, regardless of microbial or electrochemical treatments. The observed differences were attributed to the combined effect of higher thermodynamic favorability and greater surface availability. In mixed mineral systems, small particles were reduced slightly faster, whereas large particles were reduced notably slower and less extensively than solely predicted from single mineral systems. Specifically, when reduced alone, small particles showed Fe(III) reduction rate constants that were 1.5- to 3.6-fold higher than large particles, while when reduced together, the reduction rate constants for small particles were 6- to 21-fold higher than the rate constants for large particles.
Our findings also revealed that (i) Sb(V) adsorption capacities ranged from 2.5 to 55.7 mg/g, with higher affinity observed for smaller particles and for goethite relative to hematite, (ii) the presence of Ni2+ markedly enhanced Sb(V) adsorption onto both goethite and hematite, with adsorption capacities increasing by 16–89% as particle size decreased, (iii) the retention mechanism for Sb(V) adsorption changed from being dominated by edge-sharing inner-sphere complexes to enhanced electrostatic interactions and the likely formation of ternary surface complexes in the presence of Ni2+. These findings demonstrate that nanoparticulate goethite and hematite have an even greater potential for Sb(V) removal than previously expected, particularly in complex systems containing multiple coexisting ions.
Lastly, our findings revealed that (i) Sb(V) was incorporated into nanoscale iron(III) (oxyhydr)oxides to varying extents (Sb: Fe ratios ranging from 2:100 to 8:100), with no clear correlation between the extent of coprecipitation and mineral type or particle size. (ii) The presence of Sb(V) significantly inhibited both the extent and kinetics of microbial Fe(III) reduction across all samples of varying particle sizes, with suppression extent ranging from 25 to 80%. (iii) Sb release was limited (<10% of total Sb for goethite and <4% of total Sb for hematite), exhibited incongruence with the reductive dissolution of Fe, and showed only negligible Sb(V) reduction to Sb(III) across all sample conditions. These collective results indicate that the incorporation of Sb(V) into goethite and hematite may enhance the stability of these iron(III) (oxyhydr)oxides and serve as an effective sink for Sb immobilization.
Overall, this research advances the understanding of size-dependent reactivity of iron(III) (oxyhydr)oxides in processes of Fe(III) reduction, Sb(V) sorption, and Sb(V) coprecipitation. Collectively, these findings offer new insights into the role of nanoparticulate iron(III) (oxyhydr)oxides in mediating environmental redox reactions, enhancing antimony mitigation in multi-ion systems, and promoting the long-term sequestration of Sb under reducing conditions.Dissertation ist gesperrt bis 01. Oktober 2027
Novel perspectives on human nasal microbiome interactions
No full textDissertation gesperrt bis zum 15.09.2027Obwohl das nasale Mikrobiom eine Barriere gegen Krankheitserreger bildet, ist es im Vergleich zu anderen humanassoziierten Mikrobiomen bislang relativ wenig erforscht. In dieser Arbeit, die eine Literaturübersicht über das letzte Jahrzehnt bietet, beginnen wir im ersten Kapitel mit der Analyse der Faktoren, die das nasale Mikrobiom formen – darunter wirtsbezogene Faktoren, Umweltfaktoren sowie mikrobielle Interaktionen. Zudem werden die potenziellen Anwendungen des nasalen Mikrobioms und die Variabilität in der technischen Ausführung der Proben-gewinnung und Proben-Analyse beleuchtet. Im zweiten Kapitel wird anhand experimenteller Arbeiten und naturstoffchemischer Analysen ein neuartiges anti-Staphylococcus Bakteriozin namens Casipalin beschrieben, das aus dem humanen Kommensalen Staphylococcus capitis isoliert wurde. Dieses neuartige Lipopeptid weist eine starke bakterizide Aktivität gegen Staphylococcus aureus auf, was auf sein Potenzial als therapeutischer Wirkstoff hinweist. Im dritten Kapitel werden bakterielle Nährstoffinteraktionen innerhalb des nasalen Mikrobioms beschrieben, wobei wir einen neuen taxonomischen Komplex innerhalb der Spezies Corynebacterium accolens identifizieren und zeigen, dass Mechanismen zur Eisenaufnahme ein zentrales Element für die Nischenanpassung dieses weit verbreiteten nasalen Kommensalen darstellen. Im vierten Kapitel analysieren wir die Verteilung biosynthetischer Gencluster in humanen, umwelt- und tierassoziierten Isolaten und heben die höhere Prävalenz bestimmter Bakteriozin-Klassen hervor. In den folgenden Kapiteln widmen wir uns den Interaktionen zwischen Wirt und Mikrobiom. Im fünften Kapitel konzentrieren wir uns auf das Zusammenspiel zwischen humanem IgA und dem nasalen Mikrobiom, einschließlich S. aureus. Im sechsten Kapitel werden wirtsassoziierte Fettsäuren untersucht, welche die Anpassung von S. aureus an Haut- und Nasenmikrobiome beeinflussen. Im abschließenden Kapitel präsentieren wir mithilfe von Metabolomik eine aktuelle Übersicht zur Zusammensetzung menschlicher nasaler Sekrete und zeigen interessante Muster von Wirts- und Bakterienmetaboliten in der nasalen Nische.Although the nasal microbiome serves as a barrier against pathogens, it remains relatively less studied compared to other human-associated microbiomes. In this work, reviewing the literature from the past decade, we start in the first chapter by dissecting the factors that shape the nasal microbiome, from host-related factors, environmental factors, and microbial interactions. Also, we shed light on the potential applications of the nasal microbiome and the technical variability associated with its analysis. In the second chapter, employing experimental work and natural products chemistry, we describe a novel anti‑staphylococcal bacteriocin, casipalin, isolated from the human commensal Staphylococcus capitis. This novel lipopeptide displays potent bactericidal activity against Staphylococcus aureus, suggesting its potential as a therapeutic agent. In the third chapter, we explore bacterial nutritional interactions within the nasal microbiome as we identify a new taxonomic complex within the Corynebacterium accolens species and reveal that iron acquisition mechanisms are a cornerstone for the niche adaptation of this highly prevalent nasal commensal. In the fourth chapter, we investigate the distribution of biosynthetic gene clusters across human, environmental, and animal-associated isolates, highlighting the higher prevalence of certain bacteriocin classes. In the subsequent chapters, we delve into host–microbe interactions. For example, in the fifth chapter, we focus on the interplay between human IgA and the nasal microbiome, including S. aureus, while in the sixth chapter, we focus on host-associated fatty acids that shape the adaptation of S. aureus in the human skin and nasal microbiome. In the final chapter, employing metabolomics, we offer an update on the human nasal secretion composition, revealing interesting patterns of host and bacterial metabolites in the nasal niche
Analyse von Polymorphismen als genetische Modifikatoren der Spinozerebellären Ataxie Typ 3 und Morbus Huntington
No full textDie Dissertation ist gesperrt bis zum 20. April 2027 !Die spinozerebelläre Ataxie Typ 3 und Morbus Huntington sind autosomal
dominant vererbte neurodegenerative Polyglutaminerkrankungen. Ursächlich für die Krankheitsentstehung sind expandierte Wiederholungen eines CAG Basentriplets in den krankheitsspezifischen Genen. Diese führen im translatierten Protein zu verlängerten Glutaminabschnitten. Es besteht eine negative Korrelation zwischen der Anzahl der CAG-Wiederholungen und dem Erkrankungsalter. Jedoch liegt eine große Variation des Erkrankungsalters auch bei gleicher Anzahl an CAG-Wiederholungen vor. Der Einfluss anderer Faktoren wird daher angenommen. Im Zuge der Pathogenese beider Erkrankungen kommt
es zu einer intranukleären Akkumulation der expandierten Proteine in den
betroffenen Neuronen. Ataxin-3, das translatierte Genprodukt bei der
spinozerebellären Ataxie Typ 3, befindet sich bei gesunden Personen im
Zytoplasma der Zellen. Bei Erkrankten wird es durch KPNA3, einem
Transportprotein, in den Zellkern transportiert und entfaltet dort zytotoxische Eigenschaften. Parkin ist ein Interaktionspartner von Ataxin-3 im Ubiquitinsystem, welches am Abbau fehlgefalteter Proteine beteiligt ist.
In dieser Arbeit wurde untersucht, ob Polymorphismen in den für KPNA3 und Parkin codierenden Genen KPNA3 und PRKN einen Einfluss auf das
Erkrankungsalter der beiden Erkrankungen haben. Für jeden Polymorphismus wurde ein auf beide Erkrankungen anwendbarer Genotypisierungsassay mittels High Resolution Melting etabliert. Statistisch analysiert wurden die Ergebnisse mit Hypothesentests sowie linearer Regressionsanalyse. In die Auswertung wurden mehr als 1.500 Proben aus Europa, Kanada und Südamerika miteinbezogen. Für beide Polymorphismen zeigte sich ein signifikanter Einfluss auf das Erkrankungsalter der spinozerebellären Ataxie Typ 3. Dabei unterschied
sich das Erkrankungsalter bei KPNA3 je nach Genotyp um etwa 4 Jahre und bei PRKN um etwa 3 Jahre. Unter Berücksichtigung dieses Einflusses wurde für beide Polymorphismen je eine Formel erstellt, mit der das Erkrankungsalter für die spinozerebelläre Ataxie Typ 3 genauer vorhergesagt werden kann. Weiterhin eröffnen sich durch die Ergebnisse neue Behandlungsansätze für eine zukünftige Therapie der Erkrankung
Reciprocal Effects Between Teachers and Their Student Groups
No full textTeachers play a central role in shaping student learning, and much educational research has focused on how teacher characteristics and practices affect students. However, some research-ers have highlighted the importance of also considering how students may influence teachers. Reciprocal models propose that teacher-student interactions are dynamic and bidirectional, with students potentially affecting teachers’ emotions, motivation, and instructional quality. Recogniz-ing these reciprocal influences is important, as assuming only one-way effects may lead to mis-interpretations of observed relationships between attributes of teaching and student outcomes. This dissertation addresses these gaps using a triangulation approach across four studies, each exploring how student groups influence teachers and teaching. Study 1 surveyed 108 teachers about their perceptions of student group effects on their motivation, emotions, and be-havior. Teachers reported that student engagement, classroom climate, and misbehavior signif-icantly influenced their motivation and instruction—often in non-linear ways. Both highly moti-vated and highly challenging groups were perceived to improve instructional quality. Study 2 analyzed self-ratings from 162 teachers and 4,059 students across three time points using 522 random-intercept cross-lagged panel models. The study found as much evi-dence for effects of students on teachers as of teachers on students: teacher variables like the facilitation of group work and work avoidance and student characteristics like intrinsic motiva-tion, self-concept, and learning strategies mutually reinforced each other. A specifically robust bidirectional link was identified between teacher work avoidance and student interest. Study 3 tested a key theoretical model—Frenzel et al.’s (2018) proposition of bidirectional emotional transmission between students’ enjoyment and teachers’ enthusiasm—on three sep-arate samples as well as on a pooled sample including ratings from over 6,401 students and 308 teachers. Cross-lagged analyses showed significant reciprocal effects between teacher enthusiasm and student interest and enjoyment in mathematics, supporting the tested theory. Study 4 used a quasi-experimental design to assess whether changes in student groups in-fluence perceptions of teaching quality. 146 teachers who taught either the same or different student groups over two years were compared. Results showed student ratings of teaching quality varied significantly depending on the group, while teacher ratings were less affected but still influenced by changes in the student group, indicating that teaching quality depends also on the taught students. Together, these studies provide strong evidence that student groups shape teacher behav-iors, emotions, and instructional quality. Student factors like engagement, motivation, misbehav-ior, and classroom climate interact with teachers’ motivation and teaching style. The findings support refining reciprocal models of classroom dynamics and suggest that teachers may either reduce their efforts due to socio-emotional strain or adapt their instruction in response to chal-lenging student groups. Future research should explore what conditions trigger these different responses, and whether institutional support or teacher training can improve adaptive teacher responses. Intensive longitudinal data derived from automated AI supported assessment of teaching situations may allow further insights in processes of teacher-student interactions.Die Dissertation ist gesperrt bis zum 14. April 2027
Advancing Microbial and Metagenomic Analysis through Computational Tools and Integrated Data Systems
No full textStudying microbes is essential as they influence nearly every aspect of our planet’s ecosystems and living organisms. Microbiomes exist across diverse environments, including marine, soil, the human body, and even the International Space Station (ISS), with the human gut microbiome alone containing approximately ≈ 22 million genes, reflecting its immense complexity and functional potential. Traditional culture-based techniques are often ineffective since most microbes cannot be cultured. Advancements in sequencing technologies have enabled faster and more cost-effective analysis of microbial genetic information. Approaches such as amplicon sequencing provide taxonomic insights, while whole-genome shotgun sequencing (metagenomics) allows simultaneous analysis of taxonomic composition and functional diversity.
Alignment-based methods, which align reads to reference databases and bin them into taxonomic and functional categories, are powerful tools. However, these methods face significant computational challenges due to the exponential growth of sequencing data and reference databases. Additionally, they can generate terabyte-scale files for large projects, underscoring the need for efficient data management and accessibility. To study microbiomes comprehensively, it is necessary to move beyond sequencing and adopt multi-omics approaches, such as metabolomics, which provide a deeper understanding of microbiome functionality. Integrating such diverse datasets remains challenging and requires methods that are both efficient and accessible to researchers in a streamlined and less cumbersome manner. The same holds true for other bioinformatics methods beyond microbiome research, where simplifying complex workflows is equally important.
Addressing these challenges, this dissertation focuses on the key question: How can we optimize methods to enable users to easily analyze metagenomic data, seamlessly integrate different types of data, and ensure accessibility for analysis, even with limited computational resources or expertise?
This question is further divided into four aims. For Aim I, we focus on optimizing databases for alignment-based approaches, explored in Chapter I. As the NCBI-nr protein database (which contains protein sequences from all domains of life) continues to grow (reaching ≈ 812 million proteins as of October 2024), it has become a significant computational bottleneck for alignment-based methods. We explored this challenge in the context of the DIAMOND+MEGAN approach, which helps determine the taxonomic composition and functional potential of microbial communities while allowing exploratory analysis of results using the MEGAN GUI. We addressed this challenge in two scenarios: (1) when researchers are interested only in the prokaryotic content of a metagenomic sample, and (2) when exploration extends beyond the prokaryotic fraction.
In the first case, we investigated AnnoTree, a protein database based on GTDB, and compared it with the prokaryotic portion of NCBI-nr. Our results showed that AnnoTree maintained similar alignment and assignment rates, while outperforming NCBI-nr by assigning more reads to KEGG functional categories. AnnoTree demonstrated
comparable specificity, with minor trade-offs, and was twice as fast when tested on metagenomic datasets.
In the second case, we evaluated UniRef clustered versions (100, 90, 50) and NCBI-nr clustered versions (90, 50), which include protein sequences from all domains of life, for broader exploration beyond prokaryotic content. Our results showed that higher-resolution clustered databases exhibited alignment and assignment rates similar
to the full NCBI-nr database, with slightly improved taxonomic assignment rates while maintaining agreement and specificity, albeit with trade-offs. Furthermore, all clustered versions demonstrated better assignment rates for functional categories, with a minor exception. Importantly, these optimized databases were significantly smaller
in size and provided substantial speedup—NCBI-nr50, for instance, achieved a 17-fold speedup—thus greatly reducing computational costs.
For Aim II, we focused on providing novel methods and tools for efficient metagenomic data management and accessibility, explored in Chapter II. Alignment-based files, such as DIAMOND Alignment Archives (DAAs), can reach terabyte-scale for large projects due to the nature of alignment-based approaches, where a single read may generate
multiple hits in a database. This file size limits accessibility when users need to view alignments or perform further analyses. To address this challenge, particularly in the context of the DIAMOND+MEGAN approach, we developed MeganServer, a solution that eliminates the need for local file access.
MeganServer serves these files using the REST-API approach, allowing users to access and analyze data directly from the server where it was processed, without the need for local download. Users can interact with the data through any programming language, a web browser, or—most efficiently—via the MEGAN GUI, which acts as a client to the
server. By leveraging the MEGAN GUI, users can seamlessly perform comprehensive analyses and explore large metagenomic datasets without requiring programming expertise. Our results demonstrate that this approach is highly effective, significantly improving data accessibility and streamlining metagenomic analysis workflows.
In the case of Aim III, we focused on developing innovative approaches to integrate different types of data, a critical component of microbiome-based analysis, explored in Chapter III. To address this, we introduced the Microbiome Metabolome Integration Platform (MMIP), a user-friendly web resource designed to integrate datasets at different
levels. MMIP uses taxonomic profiles generated from amplicon sequencing data, which are further used to derive functional profiles and predict metabolic potential from them. It effectively integrates taxonomic, functional, and metabolic potential data. It enables users to correlate predicted metabolic profiles with real-time metabolomic data and identify both positively and negatively correlated ones.
MMIP leverages methods such as Community-wide Metabolic Potential (CMP) for metabolic potential generation and employs learning-based approaches to identify important features. The platform also supports a range of microbiome analyses, including taxonomic comparisons, α- and β -diversity assessments, functional evaluations, and feature identification. Our results on validation datasets demonstrated outcomes comparable to those reported in published studies. With its intuitive web interface, MMIP eliminates the need for programming or advanced integration expertise, making comprehensive microbiome data integration analysis accessible and straightforward for researchers.
Finally, Aim IV focuses on designing user-friendly web resources to support diverse bioinformatics analyses, explored in Chapter IV. I developed the web resources for PLaBAse (for plant growth promotion), DeepToA (for predicting metagenomic sample sources), and MuLan-Methyl (for methylation prediction). While the conceptualization, background methods, and analyses were performed by my colleagues Sascha Patz (PLaBAse) and Wenhuan Zeng (DeepToA, MuLan-Methyl), I implemented the entire web architecture and backend workflows/pipelines to ensure these tools are easily accessible to users.
These resources provide access to the novel Plant Growth-Promoting Traits (PGPT) ontology developed by Sascha Patz and deep learning- and language-model-based approaches implemented by Wenhuan Zeng. By enabling efficient and user-friendly bioinformatics analyses, these tools eliminate the need for high-end computational
resources or advanced programming skills, making complex analyses more accessible to a broader range of users.
Taken together, the aims addressed in this thesis provide optimized methods that enable users to easily analyze metagenomic data, seamlessly integrate different types of data, and ensure accessibility for analysis, even with limited computational resources or expertise. These contributions advance microbiome research and related fields by delivering freely available resources, databases, tools, integrated systems, and web platforms that streamline bioinformatics analyses. Researchers can leverage these resources directly or build upon them, deepening our understanding of microbial communities and other research areas while paving the way for future discoveries
Municipal Political Dynamics in Childcare Implementation: Conditions for Facilitating Women's Substantive Representation
No full textAbstract
This three-paper dissertation examines the political dynamics of female representatives shaping regional diversity in policy implementation within increasingly decentralized welfare states. This approach broadly contributes to the welfare state and policy implementation studies, and specifically advancing the concept of women’s substantive representation. This dissertation focuses on the variation in childcare supply expansion for children under three, addressing two key questions: (1) At the local level, which local-level women representatives substantially contribute to policy implementation? (2) What conditions facilitate/hinder the achievement of local-level women’s substantive representation? Childcare for under three is chosen because it is one of the new social services being expanded in developed countries; access to it is directly linked to “women’s interests,” and it reflects the defamilialization.
Paper 1 and Paper 2 examine western Germany despite the national policy guaranteeing the legal entitlement to obtain childcare placements after the age of one (KiföG; Kinderförderungsgesetz). While other OECD countries also implement childcare at the local level, western Germany was chosen because of the above research puzzle and because the history of the male-breadwinner model shows less path dependency. Paper 1 employs multilevel analysis to address the first question, while Paper 2 uses qualitative comparative analysis to address the second question. Papers 1 and 2 use a newly collected municipal-level dataset comprising variables related to political representation in politics and administration.
Paper 1 partly demonstrated that an increase in women holding political positions leads to the increased expansion of the childcare supply. While the increase in female council members contributed to the expansion of childcare regardless of partisanship, the role of female mayors was model-dependent. The results in Paper 2 identify two pathways for increased childcare supply, highlighting the critical role of female council members. The first path outlines the pathway for conservative parties to expand childcare, while the second one shows the path for liberal parties. In the second pathway, the alignment between the female deputy mayor and the female council member underscores the importance of a comprehensive analysis of representatives across political and administrative spheres.
Paper 3 contributed to the first and second questions by examining Japan as the most likely case where an increase in women’s representation is expected to drive increased childcare expansion, similar to the results in Paper 1. A new dataset at the municipal level in Japan was created for Paper 3. However, Paper 3 finds that the increase in female councilors and administrative managers does not lead to the increased childcare expansion. Notably, regional sexism suppresses the expansion of childcare, implying that regional sexism might hinder the local women’s substantive representation. Comparisons with Germany suggested that the strength of the legal framework at the central level in favor of addressing women's issues creates favorable conditions to facilitate women’s substantive representation at the local level.Die Dissertation ist gesperrt bis zum 27. Mai 2027
Multiple Sklerose-assoziierte Uveitis – Klinischer Verlauf und Therapie
No full textDie Dissertation ist gesperrt bis zum 01. August 2027
Spatial Immuno-Oncological Profiling of Cancer Using Multiplexed Tissue Imaging
No full textDissertation gesperrt bis zum 27.11.2027!Tumorentwicklung und -progression werden nicht nur durch die intrinsischen Eigenschaften maligner Zellen bestimmt, sondern maßgeblich durch ihre komplexe, räumlich organisierte Interaktion mit dem umliegenden Tumormikromilieu moduliert. Im Rahmen dieser Arbeit habe ich eine hochdimensionale Bildgebungsplattform auf Basis der MACSima imaging cyclic staining (MICS) Technolgie etabliert und weiterentwickelt, die eine tiefgehende Immunphänotypisierung von Tumorgewebe auf Einzelzellebene ermöglicht. Durch die Entwicklung und Validierung eines Panels mit mehr als 121 Markern sowie die Optimierung von Analyseworkflows konnte die Plattform erfolgreich beim hepatozellulären Karzinom (HCC) eingesetzt werden, wodurch räumlich strukturierte immunsuppressive Nischen identifiziert wurden.
Aufbauend auf diesen Ergebnissen wurde die Plattform für die Analyse von Formalin-fixierten, Paraffin-eingebetteten (FFPE)-Proben des seltenen und aggressiven NUT-Karzinoms (NCs) adaptiert, welches durch Genfusionen des NUTM1-Gens charakterisiert ist. Aufgrund der Seltenheit des NCs ist das Probenmaterial begrenzt und meist auf FFPE-Gewebe beschränkt. Es wurde eine Kohorte mit 16 Tumorproben von 14 Patient*innen zusammengestellt und eine hochaufgelöste räumliche Einzelzellanalyse auf Proteinebene durchgeführt. Ergänzt durch Immunhistochemie (IHC), Sequenzierungsverfahren und Durchflusszytometrie wurde eine der bislang umfassendsten molekularen Charakterisierungen dieser seltenen Malignität auf DNA-, RNA- und Proteinebene erzielt. Die Ergebnisse zeigten eine ausgeprägte Heterogenität des Tumorimmunmikromilieus, einschließlich „heißer“ (immuninfiltrierter) und „kalter“ (immunexkludierter) Tumorphänotypen, sowie immunregulatorische Interaktionen zwischen myeloiden Suppressorzellen (MDSCs) und CD8+ T-Zellen. Darüber hinaus konnte mittels multi-omic Analysen das Immuncheckpoint-Molekül CD276 als konsistent überexprimiertes und therapeutisch relevantes Ziel identifiziert und funktionell in vitro und in Xenotransplantationsmodellen validiert werden.
Diese Arbeit zeigt die technische Realisierbarkeit und den biologischen Erkenntnisgewinn räumlicher Einzelzellanalysen in häufig auftretenden als auch in seltenen Tumoren auf und liefert neue Ansätze für das Verständnis und die Therapie des NUT-Karzinoms. Die hier entwickelte Plattform bietet eine skalierbare Grundlage für zukünftige Anwendungen raumbezogener Analysen in der Krebsforschung und soll einen Beitrag zur personalisierten Onkologie leisten.Tumor development and progression are shaped not only by the intrinsic properties of malignant cells but also by their complex and spatially organized interactions with the surrounding microenvironment. In this thesis, I established and advanced a high-dimensional imaging platform based on the MACSima imaging cyclic staining (MICS) technology for deep immune profiling of tumor tissue. By designing and validating a panel comprising over 121 markers and optimizing imaging and analysis workflows, we successfully applied the platform to hepatocellular carcinoma (HCC), uncovering spatially structured immunosuppressive niches.
Building on this foundation, the platform was adapted to study formalin-fixed paraffin-embedded (FFPE) samples of NUT carcinoma (NC), a rare and aggressive cancer driven by chromosomal rearrangements involving the NUTM1 gene. Due to the rarity of NC, tissue samples are scarce and often limited to FFPE material. A cohort comprising 16 tumor specimens from 14 patients was assembled, and a highly detailed spatial single-cell proteomics analysis of NC was performed. In addition, complementary technologies including immunohistochemistry (IHC), whole-genome sequencing (WGS) and RNA-sequencing (RNA-Seq), and flow cytometry were employed, enabling one of the most comprehensive molecular characterizations of this rare malignancy to date across the DNA, RNA, and protein levels. The data revealed significant heterogeneity within the tumor immune microenvironment (TIME), including “hot” immune-infiltrated and “cold” immune-excluded phenotypes, as well as immunoregulatory interactions between myeloid-derived suppressor cells (MDSCs) and CD8+ T cells. Furthermore, multi-omic profiling identified the immune checkpoint molecule CD276 as a consistently overexpressed and actionable therapeutic target in NUT carcinoma (NC), which was validated functionally, both in vitro and in xenograft models.
Overall, this work demonstrates the feasibility and value of spatial single-cell profiling in prevalent and rare cancers and provides novel biological insights and therapeutic implications for NC. This thesis presents a flexible and scalable framework for deep spatial tissue analysis, advancing the field of spatially resolved cancer research. By enabling comprehensive characterization of tumor ecosystems, it aims to support translational efforts and guide the development of future therapeutic strategies
Texture analysis of liver NET metastases on [68Ga]Ga-DOTATATE PET candidate to PRRT: Non-invasive histological and prognostic stratification
No full textDie Dissertation ist gesperrt bis zum 14. Oktober 2027 !Background and Objective: [177Lu]Lu-DOTATATE-PRRT works by binding to SSRs, expressed in high quantities by NET cells, and delivering a cytotoxic effect through a beta-particle emitting radioactive isotope. Clinical studies have demonstrated the efficacy of PRRT in extending progression-free survival and improving quality of life for patients with advanced, inoperable NETs. However, the duration of PRRT response is highly variable, ranging from a few months to years. Furthermore, the assessment of treatment response in neuroendocrine tumours poses unique challenges and traditional criteria, such as RECIST and PERCIST, have not been capable of adequately evaluating the efficacy of PRRT. Radiomics, defined as the high-throughput extraction of textural features from imaging, could fulfil this need. The aim of this study was to develop survival models that predict tumour progression in patients with liver-metastasized NETs, using clinical and textural data from PET/CT scans obtained before and after PRRT. Additionally, we investigated whether radiomic features correlated with histological characteristics to explore the viability of a "virtual biopsy" through texture analysis. Methodology: We retrospectively queried our institutional database for patients with NET LM who underwent treatment with PRRT. Then, using the software LifeX, we segmented VOIs on the pre-treatment [68Ga]Ga-DOTA-peptide PET/CT. The same lesions were then segmented on the post treatment PET/CT and radiomics features were extracted. We then constructed different Cox proportional hazard models to explore the relationship between progression free survival time and one or more predictor variables. The Cox models we developed had increasing complexity: starting with only clinical variables, we progressively added radiomic variables. The models began with clinical variables, followed by clinical + pre-treatment radiomics, then clinical + pre-treatment radiomics + post-treatment radiomics, and finally clinical + pre-treatment radiomics + post-treatment radiomics + delta radiomics. Key Findings: -1. Impact of Radiomics on Prognostic Models: PET radiomics demonstrated a significant role in enhancing model performance. Specifically, integrating PET radiomics (particularly post-treatments scans) improved the C-Index in test runs and reduced the standard deviation in models relying only on clinical data. -2. Virtual Biopsy Potential: The study supports the concept of "virtual biopsy," where radiomics can approximate histological insights. These findings highlight the utility of radiomic features in predicting tumour histotypes and grading, potentially offering a non-invasive alternative to conventional biopsy when a specimen cannot be obtained. Conclusion: This dissertation provides evidence that PET radiomic analysis is a valuable tool for the non-invasive prognostication and stratification of liver NET metastases. The integration of PET-derived texture features into predictive models can improve clinical decision-making for PRRT candidates, offering insights into tumour biology and enhancing individualised treatment planning