The aim of this dissertation is supporting theoretical transistor material design using a combination of electronic structure theories, transport simulations and local current density analysis. Our main efforts focus on using several carbon based materials including 1D defective zigzag graphene nanoribbons (ZGNRs), 2D fluorinated graphene and 1D fluorinated ZGNRs to design nanowires-based devices. The electronic properties of these materials are foremost studied within periodic density functional theory (DFT) framework. It is found that in defective ZGNRs the degree of defect dilution strongly influences the electronic structures of the system and a considerable bandgap can be introduced by engineering a regular C6 deficiency into the pristine ZGNRs structure. Combining Non-Equilibrium Green’s Function and the Landauer formalism, the global transport properties such as total current-bias voltage dependence of constructed transport models are evaluated. In order to illuminate the local origin of the variation observed in the conductivity, local cur- rent density maps of the devices are investigated via a procedure originally proposed by Evers and co-workers [Walz et al., Phys. Rev. Lett., 2014, 113, 136602.]. The presence of defects in the ZGNRs leads to a concentrated current flow in the middle region close to the defect edges. The degree of defect dilution as well as the width of the nanojunction have strong impact on the local current densities. Inspired by a recent experiment [Yang et al., Adv. Funct. Mater., 2017, 27, 1604096.] which proposed defective graphene materials as biosensor, we construct a transport model composed with a selected defective ZGNRs and a pyrene molecule serving both as adsorbent on the ZGNRs and as linker to the biomolecule. By modifying the distance between the pyrene molecule and the nanojunction plane, the presence of a heavy biomolecule is simulated. The transport calculations reveal a quantitative change in the total current-bias voltage dependence, which correlates to the experimental measurements. The numerical efficiency of the local current analysis procedure is improved by applying sparse matrix storage and spectral filtering techniques while maintaining the resolution standards. The resulting local current density maps qualitatively demonstrate the local variation of the interference between the linker molecule and the nanojunction plane, which elucidates the small quantitative changes found in the global transport properties. For 2D fluorinated graphene materials, on top of DFT, the G0W0 method as well as the G0W0Γ method within many-body Green’s function framework are employed. The results show that the fluorination degree has a determining impact on the bandgap value of the system, whereas the fluorination pattern strongly influences the characteristics of their band structures. Including the description of electron-hole interactions, the optical spectra based on Bethe-Salpeter equation (BSE) of various structures are calculated. Depending on the polarization of the applied electromagnetic field, the optical absorption spectra of the same structure could vary significantly. These interesting findings suggest potential optoelectronic applications of partially fluorinated graphene. The study of partially fluorinated ZGNRs is motivated by the tunability of the electronic properties found in 2D fluorinated graphene materials and an experimental publication [Withers et al., Nano letters, 2011, 11, 3912-3916.], where a controlled synthesis of fluorinated graphene via electron beam is reported. The correlation between the conductivity of resulting materials and the width of fluorinated area is revealed in the experimental measurement. In order to understand the detailed transport mechanism, different widths of edge fluorinated ZGNRs with various degrees of fluorination are considered. The total current-bias voltage dependence and the spin resolved local current density maps of selected systems suggest that specific fluorination pattern and fluorination degrees have determining impact on the conductivity and the electron flux migration pathway within the nanojunction. In addition, the well-known edge effect of the ZGNRs can be observed via spin-resolved local current analysis, which is especially promising for their potential spintronic application. Finally, we apply a similar procedure on a device model composed by the 1D triarylmethyls polymers. Depending on the orientation of the local magnetic moments on the central α carbons, either spin paired or antiferromagnetic electronic solution can be achieved. Different functional groups on the aryl rings can modify the conductivity of the system drastically. The local current analysis verifies the influence of the electronic solution of the system on its electron flux migration pathway.

Somatic mutations in cancer can result in neoantigens that serve as attractive targets for cancer immunotherapy. Most human cancers develop “spontaneously” and harbor fewer mutations than murine cancers induced with carcinogens. The murine fibrosarcoma Ag104A arose spontaneously in an aging C3H mouse and serves here as a model to analyze adoptive T cell therapy under conditions that closely resemble the clinical situation. A tumor-specific TCR (M2/3) was isolated from T cells of mice immunized with Ag104A cancer cells. To identify the target of M2/3, mutations in Ag104A were determined by whole exome sequencing and their expression was verified by RNA sequencing. The resulting 77 potential neoantigens were expressed as tandem minigenes and pulsed into target cells engineered with H-2k restriction elements. Subsequent analysis revealed that the specific M2/3 response was directed against an H-2Kk-presented peptide derived from mutant Trp53. The p53D253E mutation (mp53) was tested as target for neoantigen-specific TCR gene therapy. M2/3-engineered T cells were adoptively transferred into immunodeficient C3HxRag2-/- mice bearing established Ag104A tumors. The transferred T cells expanded in an antigen-specific manner and induced growth arrest of this otherwise aggressive and fast-progressing cancer followed by a phase of stable disease. However, all tumors relapsed 4-5 weeks after treatment. Analysis of the tumors that escaped mp53-specific TCR gene therapy showed that their populations were dominated by antigen-negative variants. Further analysis indicated that tumor escape was a result of low-frequency occurring mp53-negative variants due to intratumor heterogeneity. These findings suggest that neoantigen-specific TCR gene therapy is effective against a single target, since mp53-positive cancer cells were successfully eliminated in vivo. Interestingly, although the p53D253E mutation confers a growth advantage and higher tumorigenicity to Ag104A cancer cells and is supposed to be cancer-driving, it proved to be insufficient for eradication of established cancer when used alone as a target antigen of T cell therapy. Analysis of the Ag104A tumor model emphasizes intratumor heterogeneity as a critical parameter of neoantigen-specific TCR gene therapy that has to be addressed in clinical settings in order to make this form of immunotherapy successful for the treatment of human cancer.

This dissertation consists of six independent chapters contributing to the literature of labour and family economics. The main topic concerns how family policies impact on gender and socio-economic inequality in, at times, unintended ways.

Chapter 2 uses administrative linked employer-employee data to examine whether employers statistically discriminated against women of childbearing age (potential mothers) when they incurred direct costs associated with motherhood. Before 2006, large firms in Germany were obliged to pay for the generous maternity protection of female employees, such that firms’ expected wage costs depended on employees’ gender and age. From 2006 onward, all firms paid for maternity protection by contributing to the statutory health insurance system, where the contribution depends only on the number of employees and their wages and is thus independent of gender and age. We provide evidence that the reform was followed by an increase in female relative wages within large firms. This reform effect provides evidence for statistical employer discrimination in the pre-2006 setup.

Chapter 3 takes a firm-side perspective on parental leave. Motherhood and parental leave interrupt employment relationships, likely imposing costs on firms. We document that mothers who are difficult to replace internally take shorter leave and that their firms hire replacements more often. Introducing more generous parental leave benefits erases the link between mothers’ internal replaceability and their leave duration. In firms with few internal substitutes this reduces employment in the short-, but not longer-term. Firms respond by hiring fewer women of childbearing age into occupations where they are difficult to replace internally. Taken together, motherhood and generous parental leave policies burden firms that have few internal substitutes available.

Chapter 4 aims to improve the understanding of day care enrolment gaps by family background in a country with a universal day care system (Germany). Research demonstrates that children of parents with lower educational attainment and children of migrant parents may benefit the most from day care, making it important to understand why such enrolment gaps exist. We use a unique data set that records both parental demand for day care and actual usage to investigate determinants using complementary decomposition and quasi-experimental analyses. Our decomposition shows that (a) differences in demand are important but do not fully explain the enrolment gaps, (b) large shares of the gaps are unexplained, especially for migrant parents, and (c) the heterogeneous effects of access barriers (shortages and fees) may explain some of the remaining gaps. Our quasi-experimental design finds that reducing shortages significantly decreases the enrolment gap by parental education but not by parental migrant status. Similarly, using the synthetic control method we show that a reduction of fees reduces only the gap by parental education. We discuss implications for policy.

Chapter 5 estimates the effect of day care on parenting activities using time-diary and survey data for Germany. This is the first such study in economics to pay careful attention to issues of selection bias, and to provide a conceptual framework of underlying mechanisms. We find that while day care strongly reduces the amount of time parents spent with their child, parenting activities are only reduced by a few minutes per day. During non-center hours, parenting activities are not affected. An analysis of non-parenting activities reveals that day care is used to take up paid work, but also partly to ease time constraints. A reduction of leisure and sleep during non-center hours suggests that an increase in motivation may be responsible for keeping parenting activities constant during those times. Our findings represent novel evidence that activities in the home environment are a complement to day care, highlighting a credible alternative mechanism for child development effects of day care.

Chapter 6 uses novel time-use data from the GDR and reunified Germany and finds that women in the GDR, and later in East Germany, spend more time in paid work and less time doing housework, compared to West Germany. However, decomposing these gender housework gaps between the West and the East, we find that they are similar once individual time constraints are accounted for. Individual housework contributions are shown to be almost orthogonal to the partner’s labour supply. We discuss implications for the nature of gender norms, and effects of labour market policy targeted at gender gaps.

Chapter 7 examines how culture impacts within-couple gender inequality. Exploiting the setting of Germany’s division and reunification, I compare child penalties of couples socialised in a more gender-egalitarian culture (East Germany) to those in a gender-traditional culture (West Germany). Using a household panel, I show that the long-run child penalty on the female income share is 26.9 percentage points in West German couples, compared to 15.5 in East German couples. Additionally, the arrival of children leads to a stronger increase in the share of housework performed by West German women and they are responsible for a larger share of child care than those from the East. A battery of robustness checks confirms that differences between East and West socialised couples are not driven by current location, economic factors, day care availability or other smooth regional differences. I add to the main findings by using time-use diary data from the GDR and reunified Germany, comparing parents with childless couples of similar ages. This provides a rare insight into gender inequality in the GDR and allows to compare the effect of children in the GDR to the effects in East and West Germany after reunification. Lastly, I show that attitudes towards maternal employment are more egalitarian among East Germans, but that the arrival of children leads to more traditional attitudes for both East and West Germans. The findings confirm that socialisation has a strong impact on child penalties and thus on gender inequality as a whole.

One of the most pressing issues of our time is to move away from fossil fuels to stop or limit global warming. One important technology which is used to replace energy from fossil fuels are photovoltaic solar cells. This thesis deals with a material class which can be used as a solar cell absorber called kesterites.

All commercially relevant solar cell absorber materials have certain drawbacks like high material usage and utilizing scarce or poisonous elements. Kesterite solar cells rely on abundant and relatively environmentally friendly elements. Furthermore kesterites have a high absorption coefficient for visible light which allows to use them in so called thin-film solar cells which are very efficient in terms of material usage. Kesterite solar cells are not yet matured to a level where they can compete commercially. Their biggest issue is a low conversion efficiency of the incoming solar energy. It stagnated at 13 % for the last 7 years while beyond 20 % conversion efficiency are required to be successful in the commercial market. One of the causes of the low conversion efficiency is that kesterites can easily get disordered, in particular Cu and Zn atoms which share a plane in the unit cell can interchange easily. Partly or fully replacing Cu and/or Zn with another cation can reduce the tendency to disorder. In this thesis we investigate unsubstituted kesterite (Cu2ZnSnS4) as well as various substituted variants.

In thin-films compressive stress can occur for various reasons, e.g. if a reactant is sputtered in the synthesis. We use first principle methods to scan for high pressure phase transitions for ordered and disordered kesterite and the following substituted variants: Cu2FeSnS4, Cu2MnSnS4, Ag2ZnSnS4 and Ag2CdSnS4. We investigate how the electronic structure changes through the structural phase transition. Cu2CdSnS4 has a different crystal structure than Cu2ZnSnS4. Using first principle methods we determine the maximum amount of Cd which can be used in Cu2ZnSnS4 to substitute Zn without changing the crystal structure.

For all investigated materials we predict a high pressure transition to a compressed rocksalt structure. The electronic structure of this high pressure phase is in all cases metallic which renders the materials useless as a solar cell absorber. The transition pressure for Cu2ZnSnS4 is predicted at 16~GPa, in excellent agreement with experimental results. The transition pressure for the substituted variants Cu2FeSnS4, Cu2MnSnS4 and Ag2ZnSnS4 are lower but still well off the pressures we expect in thin-film synthesis. Only the transition pressure for Ag2CdSnS4 which is 4.7 GPa is close to pressures which can occur in thin-films. Therefore we advice to monitore the structure if this material is used as a solar cell absorber. For the solid solution series Cu2Cd(x)Zn(1-x)SnS4 we predict the kesterite-type structure to be most stable up to Cd(x)=0.51. Experimentally it is most stable up to Cd(x)=0.40. We show that the difference may be due to a small disorder in the experimental samples.

Diese Dissertation zeigt, wie die Depression in der stationären Therapie zweier psychiatrischer Kliniken in München und Berlin produziert und wie der Umgang mit ihr kulturell modelliert und sozial vermittelt wird. Dabei steht die kognitive Verhaltenstherapie im Mittelpunkt, die zum gegenwärtigen Zeitpunkt als die Mainstreamtherapie in Deutschland bezeichnet werden kann. Die unterschiedlichen Therapeuten versuchen, ihre Patienten zu einer anderen Wahrnehmung von sich selbst und ihrer Krankheit anzuleiten. Der Zweck all der stationären Interventionsformen zusammengenommen ist es, Patienten in Emotions- und Selbstmanager zu transformieren. Wie der therapeutische Prozess erfolgt und welche Widerstände und Probleme er auf Seiten der Patienten auslöst, wird in der Ethnografie deutlich. Im psychiatrischen Raum werden kulturelle Prägungen der Depressionstherapie sichtbar: So fordern die Therapeuten von den Patienten Individualisierung, Rationalität, Empowerment und Emotionsarbeit. Die Patienten können dies nur beschränkt umsetzen. Dies liegt nicht zuletzt an ebenfalls kulturell geprägten Paradoxien und Ambivalenzen in der Depressionstherapie, die in der vorliegenden Arbeit erläutert werden.

Neurulation is a complex morphogenetic process, requiring the integration of various molecular and cellular events in order to culminate in the development of sophisticated brain structures. Many mechanistic aspects of the regulatory cues, which pre-pattern the neural plate and further specification of the neuroepithelial stem cells, remain largely unknown. Nonetheless, understanding of the coordination between pivotal developmental processes is indispensable in successful prevention of congenital brain defects in human.

My Ph.D. work significantly contributed to understanding the mechanisms underlying the etiology of congenital brain disorders. I functionally characterized the potential genetic modifiers rendering mouse models with a higher predisposition to neural tube defects. Results from this work can further be extrapolated to human disease penetrance. One of the key signaling molecules at the onset of neurulation is the morphogen SHH, which is required for ventral midline specification. Mice deficient for the LRP2, a SHH co-receptor in the ventral forebrain, suffer from insufficient SHH uptake by the neuroepithelium and develop holoprosencephaly (HPE) on C57BL/6N background. Congenic Lrp2 mutants on FVB/N background, however, show normal SHH levels in the ventral forebrain and normal forebrain separation. We have identified new candidate modifier genes, Pttg1 and Ulk4, that influence the capacity of SHH signaling pathway. Higher expression levels of Pttg1 and Ulk4 in developing brain of FVB/N mice are likely to contribute to advantageous cellular environment, which maintains the necessary SHH morphogen gradient in the ventral forebrain, independent of the receptor-mediated uptake, hence preventing HPE. The rational is that these novel, disease relevant, positive modulators of the SHH pathway make the early forebrain developmental processes less susceptible to disturbances in the SHH pathway. Positive regulation of the SHH machinery is associated with efficient ciliogenesis. For the first time, I describe PTTG1 as a novel primary cilia component. I identified PTTG1 at the basal body – the primary cilia organizing center – and in the ciliary shaft. The variable localization of PTTG1 amongst different cilia suggested that it is not a structurally required component of the primary cilium, but may shuttle into the ciliary shaft in a regulated fashion to support ciliary function and/or axoneme assembly, ultimately enhancing SHH signaling capacity.

Besides the ventral forebrain defect, there is a SHH-independent, dorsal phenotype in Lrp2 mutant mice, reminiscent of neural tube closure defects (NTDs) affecting C57BL/6N and FVB/N mice. Combining mouse and Xenopus work from Kerstin Feistel’s lab, we unraveled a conserved, novel function of the receptor, beyond its highly anticipated signaling scope. Neural plate bending and closure is a dynamic process involving cytoskeletal remodeling and apical constriction. We characterized LRP2-dependent apical membrane remodeling necessary for efficient upfolding of the neural plate. We documented for the first time in neuroepithelium, functional interaction of LRP2 with PDZ-containing adaptor proteins, which serve as a bridge to link LRP2 to the intracellular scaffold of neuroepithelial cells. Endocytic activity of the receptor facilitates apical membrane remodeling during apical constriction and simultaneously contributes to the apicobasal distribution of VANGL2, indicating a close interface between cell shape control and maintenance of planar cell polarity. Furthermore, I report the function of LRP2 goes beyond the neuroepithelium, and affects cranial neural crest, thus placing LRP2 in the context of multiple stem cell niches in the developing brain.

To conclude, on one hand, my work describes new aspects relevant for understanding the genetic modulation of signaling pathways and thus genetic predispositions to neural tube disorders. On other hand, we provide a detailed functional characterization of cell-autonomous mechanism in NTD etiology. LRP2 serves as a hub, orchestrating the signaling and endocytic pathways with cytoskeletal remodeling, at the periciliary compartment of the neuroepithelial stem cell. By this mechanism, LRP2 ensures homeostasis of neural stem cell niches and, simultaneously, facilitates the biomechanics of neural tube formation.

Smart-wheelchair navigation is a daunting problem even when autonomous cars are slowly taking on the roads. Smart-wheelchairs ought to navigate in highly unstructured and diverse indoor and outdoor conditions that lack the conventions that simplify autonomous car driving. Furthermore, smart-wheelchairs have astringent limitations to the size, weight, energetic consumption, and cost of the components used for autonomous navigation. Despite the challenge, the development of smart assistive technology is necessary because the number of users is likely to rise due to demographic tendencies. This dissertation explores the use of the visual compass algorithm VCA for outdoor route-following with a smart-wheelchair. The VCA infers the travel direction by aligning the current-view with snapshots stored in a visual memory. Image matching is performed using simple IDFs on low-resolution panoramic images, as inspired by insect navigation findings. The VCA is an attractive alternative because it only requires an omnidirectional camera and modest computational resources. Nevertheless, the algorithm suffers from well-known limitations, and its performance for tasks like smart-wheelchair navigation is poorly understood because most studies evaluated have focused on highly controlled conditions. This investigation presents three extensions to the VCA. The first one is the use of texture-based representations of the panoramic images to reduce aliasing. Then, the TVCC, which enables the use of multiple representations and IDFs. The third extension introduces KT, i.e., using visual memories to navigate with a different setup than the one used to acquire them, which is an essential aspect of large-scale applications. Furthermore, this dissertation presents an algorithm to correct off-route translational error without resorting to multiscale analysis or probatory movements.

In der vorliegenden Arbeit wird die These aufgestellt, dass die heutige Landschaftsauffassung von dem neolithischen Fundort Göbekli Tepe (Südosttürkei) nicht auf Grundlage einer tatsächlichen Erforschung der Landschaft entstand. Stattdessen gehe ich davon aus, dass sich die Landschaftsauffassung durch eine semantische Übertragung der dem Platz zugeschriebenen Singularität auf die Landschaft herausbildete. Verschiedene Aspekte, die das heute gängige Narrativ zum Göbekli Tepe und insbesondere das zu seiner Landschaft prägen, werden als Spiegel rezenter Denkweisen und Wahrnehmungsformen angesehen. Ziel der Arbeit ist es einerseits diese These zu überprüfen und andererseits überprüfbare landschaftliche Analysen zu liefern, die als dringend notwendig erachtet werden, um zukünftig ein theoretisch begründetes und reflektiertes Narrativ zum Göbekli Tepe entwickeln zu können. Einem hermeneutischen Ansatz folgend, wird die Landschaft des Göbekli Tepe in ›zirkulär‹ aufeinander aufbauenden Teilen aus der Perspektive unterschiedlicher Landschafts bzw. Raumbegriffe betrachtet. Um sich der Landschaftswahrnehmung im Neolithikum nähern zu können, wird ein emphatischer Zugang gewählt, bei dem von den eigenen abweichende, intersubjektive, also nachvollziehbare Wahrnehmungsmuster betrachtet werden. Mit Hilfe dieser Vorgehensweise, die gewinnbringend archäologische und geographische Arbeitsweisen interdisziplinär miteinander verknüpft, erfolgt eine Neuinterpretation des Göbekli Tepe. Die Stätte wird hierbei nicht wie bisher meist angenommen als die Neolithisierung vorantreibend angesehen, sondern als rückwärtsgewandter Ort, an dem an der paläolithischen Lebensweise festgehalten wurde. Es wird gezeigt, dass die Standortwahl nicht aus den in der Literatur zumeist angeführten Gründen einer gewünschten Kontrolle, Effizienz oder Prominenz erfolgte. Stattdessen sind die Architektur und der Standort des Göbekli Tepe als Ausdruck der Unsicherheit zu interpretieren, die der Transformationsprozess von einer aneignenden zu einer produzierenden Lebensweise im frühen Neolithikum auslöste. Die These, dass das Narrativ zu der Landschaft des Göbekli Tepe stark auf heutigen Denkmustern aufbaut, wird als bestätigt angesehen.

In times of technological dynamics, knowledge intensification and increasing networking between business stakeholders, absorptive capacity has become one of the most prominent concepts in innovation and management research. However, despite the attention devoted to absorptive capacity in the academic community, there is still a limited understanding of how firms without institutionalised R&D departments and formal R&D expenditure can absorb and leverage external knowledge successfully. With its strong focus on R&D-intensive firms and high-tech industries as well as on R&D-related aspects, existing research has overlooked the variety in patterns of knowledge absorption and, due to its reliance on inappropriate measures, tends to underestimate absorptive capacity in non-R&D firms.

Therefore, the overall objective of this dissertation is to advance our understanding of absorptive capacity beyond the R&D-based context by providing novel empirical insights on how non-R&D-performing manufacturing firms build their AC to access and leverage external knowledge successfully in their innovation activities.

The dissertation consists of three distinct but complementary research papers. The empirical analyses in these papers are based on the qualitative data from multiple case studies of four non-R&D-performing German manufacturing firms, as well as the quantitative cross-sectional data from the German Manufacturing Survey 2015.

In the first paper, which relies on qualitative data, I explore how non-R&D firms organise their efforts to deploy absorptive capacity, and which personnel resources and organisational practices they rely on. The empirical results provide in-depth insights into the nature and specific character of the AC process in the non-R&D context, and highlight the focal role of a few key absorptive agents in the configuration of absorptive capacity. A crucial factor for success is the integration of these individuals into the process of absorptive capacity by means of aligned organisational practices.

Based on the quantitative cross-sectional data, the second paper compares non-R&D and R&D SMEs in terms of their openness to external knowledge and major external sources of innovation impulses. Further, it takes a deeper look at non-R&D SMEs and examines how the internal organisation of absorptive capacity in these firms varies depending on the type of external knowledge source targeted. The results indicate that both non-R&D and R&D SMEs are open to external knowledge sources to a similar degree. However, non-R&D SMEs are more likely to search for innovation knowledge from suppliers and less likely to search for science-based knowledge than R&D SMEs do. Moreover, the results show that different search patterns are associated with different modes of knowledge absorption.

The third and final paper explores the heterogeneity in patterns of absorptive capacity and explains how the configuration of absorptive capacity is shaped by a firm’s dominant innovation mode. To address its research objective, the paper develops an integrated conceptual framework of absorptive capacity that offers a more differentiated perspective of absorptive capacity and enables the investigation of heterogeneous configurations of absorptive capacity across different types of firms. The framework was illustrated and tested using the qualitative data from the case studies with non-R&D performing firms and was shown to be appropriate for the operationalisation of absorptive capacity configurations associated with the Doing, Using and Interacting (DUI) mode of innovation and practical, tacit knowledge. This makes this framework particularly relevant for non-R&D firms and SMEs.

This dissertation contributes to innovation management research and enriches the absorptive capacity literature in several important ways. First, by responding to the call to explore absorptive capacity beyond the R&D-based context, this dissertation provides novel empirical insights into the hitherto overlooked empirical context of non-R&D manufacturing firms. It brings more conceptual clarity and explains how non-R&D firms manage to build absorptive capacity directed towards strategically relevant knowledge. Second, this dissertation proposes a holistic conceptual framework that integrates fragmented findings from different research fields on the underlying building blocks of absorptive capacity, such as dimensions, constitutive elements and microfoundations. In doing so, it enriches our understanding of the internal configuration of a firm’s absorptive capacity. Third, this dissertation explores and explains inter-firm heterogeneity in patterns of absorptive capacity, which has received little attention in the previous literature. Thus, the dissertation provides a broader understanding of absorptive capacity beyond R&D and explains how firms can build and develop their absorptive capacity in a more differentiated manner. Finally, this dissertation, and in particular the developed framework, can serve as the cornerstone for the improved, more comprehensive operationalisation and measurement of absorptive capacity.

Timely and accurate genome duplication is essential to maintain genome integrity and cell survival. DNA replication-associated damage is one of the leading causes of genome instability and a precursor for carcinogenesis. The DNA replication fork (RF), the site for assembly of replication proteins, encounters a variety of obstacles, which slow or stall its progression, a process termed replication stress. Cells have evolved a number of mechanisms to stabilize stalled forks and to ensure replication restart and timely completion. However, during chronic stress, forks can no longer be stabilized and collapse, creating toxic DNA double-strand breaks (DSB). These DSBs, when left unrepaired, can lead to chromosomal rearrangements and promote genomic instability. RIF1, a multifunctional protein, is critical not only to promote fork stability and to ensure that replication is completed, but also to repair DSBs in the event of prolonged replication stress. While modulation of DSB repair pathways represents one of the resistance mechanisms to chemotherapeutic drugs, maintenance of fork stability is critical to prevent carcinogenesis from developing in the first place. Here, we have identified novel post translational modifications of RIF1 that are critical for its role in the maintenance of genome stability. Specifically, phosphorylation of a conserved cluster of SQ sites in RIF1 modulates its role in fork stabilization while being dispensable for its function in DSB repair.

Animals continuously sense the temperature in their environment, which is crucial for survival and for maintaining an optimal energy expenditure. Thermal perception is enabled by sensory afferent neurons that innervate the skin and express molecules that transduce thermal stimuli into electrical signals, which are later processed by the nervous system. In the recent years, studies in genetically modified mice have found sensory afferents and ion channels that transduce cooling in mammals, but the perceptual ability of mice to sense warmth and the underlying encoding mechanisms remain unknown.

In this work, I have investigated the neurobiological mechanisms that underlie the perception of warmth. To do so, I have employed the mouse (Mus musculus) as a model system due to both the phylogenetic proximity to humans and the availability of genetic tools for mechanistic studies.

Using a sensory detection behavior, I first show that mice perceive tiny (0.5oC) changes in temperature of the forepaw. Like humans, mice are able to discriminate warming from cooling, they are less sensitive to warmth and the baseline temperature strongly impacts the perceptual saliency. Together, these data indicate that mice and humans share many features of thermal perception, suggesting common sensory coding mechanisms.

Cooling perception in mice is known to require cool-activated sensory afferents that express the channel TRPM8, but the neurons encoding warmth are unknown. Here, warming recruited two polymodal afferent populations: one fired upon warming (warm-activated) and the other was both warm-silenced and cool-activated. To investigate their role in perception, I used gene knockouts and optogenetic afferent stimulation and found that mice sense and encode warming without the warm-activated ion channels TRPV1, TRPM3, TRPA1 and TRPM2. However, surprisingly, despite TRPM8+ afferent stimulation evoked a cooling percept, TRPM8-null mice cannot detect warming. Trpm8-/- mice possess warm-activated afferents but lack warm-silenced neurons, suggesting that cooling input from warm-inhibited fibers is required for warmth perception.

In preliminary work I have also investigated the role of primary somatosensory cortex in warm perception. Using intrinsic optical imaging I observed that cooling and touch, but not warming, elicits robust responses; but silencing of this brain region impaired the perception of warming stimuli.

Altogether, the data from my thesis suggest that warming perception is an integrative process and requires input from both warm- and cool-activated sensory pathways.

Anthropogenic activity causes excessive CO2 emission in the atmosphere. In the CO2 reduction reaction (CO2RR), CO2 is converted electrochemically into valuable products including methane, ethylene, and other hydrocarbons. Large-scale employment of efficient CO2RR could diminish the use of fossil-fuel resources and thereby counteract rising levels of atmospheric CO2. Although plenty of catalysts have been investigated, copper (Cu) is the only metal that can produce reasonable amount of hydrocarbon products with two or more carbon atoms (C2+ products). However, low selectivity and high overpotentials remain as two obstacles for practical application. The reactivity determinants and reaction mechanisms in CO2RR and their relation to formation of specific products are insufficiently understood only. Therefore, operando techniques that provide insight in reaction mechanism for the catalyst material during electrocatalytic operation are of great importance. In this thesis, operando Raman spectroscopy was adopted to investigate Cu-based foam-like materials for CO2RR catalysis. These Cu foams, which have been characterized regarding CO2RR activity previously, are shown to exhibit surface-enhanced Raman scattering (SERS) activity, thereby facilitating the detection of surface-bound molecular species. A normalization procedure was developed for meaningful analysis of the electric potential dependence of the detected species. A comprehensive overview of various potential dependencies of Raman spectra are reported. Special focus was put on the role of oxides and the bicarbonate electrolyte. Evidence for copper carbonate hydroxide (CuCarHyd), which resembles the mineral malachite, was obtained. Its carbonate ions can be directly converted to CO at low overpotentials. These and further experiments, including 13C isotope experiments and local pH detection, suggest a basic mode of CO2/carbonate reduction at Cu electrodes that contrasts previous mechanistic models: the starting point in carbon reduction is not CO2 but carbonate ions bound to the metallic Cu electrode in form of CuCarHyd structures. Cu oxides residues could enhance CO2RR indirectly by supporting formation of CuCarHyd motifs. The presence of CuCarHyd patches at catalytic potentials may result from alkalization in conjunction with local electrical potential gradients, enabling the formation of metastable CuCarHyd motifs over a large range of potentials. It is conceivable that the interaction of oxide and bicarbonate also may be the key for the enhanced C2+ production activity. Operando Raman spectroscopy was also used to investigate a bimetallic catalyst CuAg foam that shows a primary selectivity for CO. The Raman spectra on CuAg foam shows an earlier departure of CO from the surface. Further 13C isotope experiments probed a surface intermediate that can be assigned to the symmetrical stretching vibration of bidentate. Mass transport limitations of carbonate species may influence the CO2RR mechanism. Raman spectroscopy (without SERS) was used to detect the local pH at the Cu foam surface. Severe local alkalization was observed. The same process was analyzed in D2O electrolyte and using a flow-cell setup, persistent local alkalization was also observed. Further results of potential influences of local pH change on the CO2RR are presented.

Adoptive transfer of T cells, engineered with antigen-specific receptors, is clinically used against acute lymphoblastic leukemia and certain types of large-B-cell lymphomas. Reliable and efficient use of such T cells against solid tumors has not been established yet. However, adoptive transfer of autologous, tumor infiltrating T cells (TILs) can eradicate established solid tumors in some patients. A limitation with this form of therapy is that isolation and expansion of TILs fails in the majority of patients and success seems to depend on the presence of a mutant neoantigen-specific T cell population in the transferred TIL product. Mutant neoantigen-specific T cells can be found in almost all analyzed cancer types. Furthermore, TCR genes can be isolated from TILs and used for viral transduction of patient‑derived, autologous peripheral blood lymphocytes (PBLs). As of yet, it is unknown how many different TCRs from T cell clonotypes of the polyclonal TIL response are needed to achieve cure. Therefore, the main objective of this thesis was to determine the requirements for mutant neoantigen-specific TCR-transduced T cells to achieve eradication of established solid tumors by adoptive transfer. For this, unmanipulated murine tumors were targeted using TCRs that recognized autochthonous, naturally occurring mutant neoantigens. Using such animal models, it was possible to show that tumors can be eradicated with just two cancer‑specific TCR clonotypes. One TCR had to target the cancer cells directly by recognition of a cancer-specific MHC class I restricted antigen, while the second TCR had to target a MHC class II restricted mutant neoantigen presented by cells of the tumor stroma. It shows that both TCR clonotype are sufficient and essential to achieve tumor rejection because a combination of two TCRs targeting independent MHC class I restricted mutant neoantigens was not as efficient and led to tumor escape even though the tumor that had been targeted was developed from a cancer cell clone, seemingly lacking tumor heterogeneity. Targeting only the tumor stroma with a MHC class II restricted mutant neoantigen‑specific TCR led to tumor destruction followed by long‑term growth arrest but never eradicated the tumor. This effect was caused by regression of tumor vasculature and depended on tumor stroma recognition only. For tumor eradication, direct cancer cell recognition was essential because lack of surface expression of MHC class I molecules by cancer cells led to failure of tumor rejection despite the presence of both TCR clonotypes. These data created the hypothesis that a four cell-type interaction is required for effective elimination of solid tumors and gives insights that simplifications for translation into clinical applications may be possible without sacrificing efficacy for adoptive, mutant neoantigen-specific, TCR-redirected T cell transfer.

The presented thesis concentrates on the metabolism of three AAS in the human body and the synthesis of potential new long-term metabolites. It demonstrates an effective way of introducing new markers to anti-doping analysis and finally confirming the proposed structures. First, the substrate specificity of a human CYP (21A2) was elucidated to use its stereoselective hydroxylation capacity for the targeted synthesis of new metabolites of several AAS. Second, new metabolites of MD and MT were synthesized chemically, characterized, and detected in urine samples. Third, a controlled administration study with DHCMT was performed to reveal the metabolism of this doping substance and provide kinetic data of the metabolic products. Different synthesis methods were used depending on the structure: either a full chemical or a combined chemical and biotechnological approach. Several analysis techniques were used to provide sufficient information during the particular project steps. To verify intermediate products in the synthesis, gas chromatography coupled by electron ionization to single quadrupole-mass spectrometry (GC-EI-MS) was used. NMR experiments together with GC-EI-MS and gas chromatography coupled by electron ionization to quadrupole time-of-flight-mass spectrometry (GC-EI-QTOF-MS) were used for the structure elucidation of the synthesized reference material. Post administration samples were analyzed by GC-EI- QQQ-MS and LC-ESI-QTOF-MS. The results of this thesis show the ideal way of providing reference material to the anti-doping community or every other research field that is dealing with the metabolism of endogenous and exogenous substances in case that the concentrations are too low for other means of comprehensive identification. In a first step, the synthesis of reference material must be planned, and the used compounds, e.g., enzymes or reagents, must be evaluated for their potential use. Subsequently, the synthesis has to be performed, and the product needs to be fully characterized so that there is no doubt of its structure. Finally, an administration of the parent compound and subsequent analysis of the excreted metabolites confirm the previous work. In summary, this work shows how new ideas about metabolism should be dealt with. It helps to dispel the doubts that have existed for years about postulated results. At the same time, it shows that steroid metabolism still has various blind spots, whose elucidation must be the goal of further research. Although anti-doping was the all-dominant topic, the gained results are also relevant for fields in physiology and pharmacology.

Diese Dissertation besteht aus sechs Kapiteln, die sich mit der ökonomischen Analyse von Bildungs- und Haushaltsentscheidungen befassen. Diesen Kapiteln geht eine Einleitung voraus, in denen gemeinsame und komplementäre Beiträge der Forschungsarbeiten dieser Dissertation zur Literatur dargestellt und die einzelnen Forschungsfragen motiviert werden. Im Schlussteil werden Limitationen der Forschungsarbeiten, offene Forschungsfragen und Handlungsempfehlungen für die Politik diskutiert.

Kapitel 2 befasst sich mit den Effekten von Informationen auf die erwarteten monetäre und nicht-monetären Erträge eines Masterstudiums sowie der beabsichtigten und tatsächlichen Aufnahme eines Masterstudiums. Hierfür wurden im Rahmen eines Experiments Studierenden gezielt Online-Informationen bereitgestellt. Sechs Monate nach der Informationsintervention lassen sich kausale Effekte auf die erwarteten Erträge beobachten. Die Effekte auf erwartete Erträge sowie die Effekte auf die beabsichtigte und tatsächliche Aufnahme eines Masterstudiums unterscheiden sich nach Geschlecht und akademischem Hintergrund. Insbesondere Männer schätzen infolge der Informationsintervention ihre Perspektiven ohne Masterstudium besser ein als zuvor. Ein Masterstudium verliert für sie an Attraktivität. Mithilfe weiterer Erhebungen wird gezeigt wie sich die Intervention auf die tatsächliche Aufnahme eines Masterstudiums auswirkt. Zusammengefasst zeigt dieses Kapitel die Relevanz von Informationen zu monetären und nicht-monetären Erträgen eines Masterstudiums für die Studienentscheidung auf.

Kapitel 3 zeigt mithilfe von Zeitbudgeterhebungen aus der DDR und der wiedervereinigten Bundesrepublik Deutschland auf, dass Frauen in der DDR, und später in den ostdeutschen Bundesländern, mehr Zeit im Arbeitsmarkt und weniger Zeit mit Hausarbeit als westdeutsche Frauen verbringen. Bei der differenzierten Betrachtung dieser Unterschiede wird jedoch deutlich, dass keine Differenzen in der Hausarbeit zwischen ost- und westdeutschen Frauen bestehen, wenn Frauen mit gleichem zeitlichen Beschäftigungsumfang im Arbeitsmarkt verglichen werden. Da Männer eine höhere Arbeitsbelastung der Frauen am Arbeitsmarkt nicht durch die Übernahme von mehr Hausarbeit kompensieren, sind berufstätige Frauen in der DDR und BRD doppelt belastet. Das Kapitel schließt mit einer Diskussion über die Folgen von Geschlechternormen und möglichen Nebeneffekten von politischen Maßnahmen zur Erhöhung der Frauenerwerbsquote ab.

Kapitel 4 untersucht die Effekte eines umfangreichen Programms zur Eröffnung neuer Vorschulen in Kambodscha. Der Bau neuer Vorschulen im Jahr 2016 wurde mit zwei nachfrageseitigen Interventionen zur Stimulierung der Nachfrage nach Vorschulen in einer randomisierten kontrollierten Studie kombiniert. Während Kinder durch den verbesserten Zugang zu Vorschulen häufiger eine Vorschule besuchen, hatten die nachfrageseitigen Interventionen keinen zusätzlichen Effekt. Ein Jahr nach Beginn der Intervention wurden kleine Effekte auf die kognitive Entwicklung (0.04 Standardabweichungen) und sozio-emotionale Entwicklung (0.09 Standardabweichungen) der Kinder gemessen. Mittelfristige Effekte werden für Kinder des oberen Wohlstandsquartils zwei Jahre nach Beginn der Intervention gemessen. Dieses Kapitel zeigt zudem auf, dass das untersuchte Programm zu großen Verbesserungen in der strukturellen Qualität der vorhandenen Vorschulen geführt hat, jedoch nur geringe Effekte auf deren Prozessqualität hatte. Die Ergebnisse legen nahe, dass in Ländern mit niedrigem Einkommen, Kinder aus verhältnismäßig gutsituierten Haushalten stärker von den Vorschulangeboten profitieren. Ein Grund sind komplementäre Effekte des Vorschulbesuchs und bessere kognitive Stimulation und emotionale Unterstützung der Eltern. Die Ergebnisse zeigen auch auf, dass weitere Verbesserungen in der Prozessqualität von Vorschulen notwendig sind um die Entwicklung benachteiligter Kinder zu fördern.

Kapitel 5 untersucht die Effekte des gleichen Vorschulprogramms in Kambodscha mit besonderem Fokus auf die Tatsache, dass die neu geschaffenen Vorschulen mit bestehenden Vorschulangeboten von geringerer Qualität sowie familiärer Betreuung konkurrieren. In diesem Kontext wird aufgezeigt, dass sich die gemessenen Effekte des Vorschulprogramms auf die frühkindliche Entwicklung danach unterscheiden, ob ein Kind ohne die neuen Vorschulen eine andere Vorschule besucht hätte oder in familiärer Betreuung geblieben wäre. Mithilfe der Daten der randomisierten kontrollierten Studie werden verschiedene empirische Schätzmethoden angewandt, um den Effekt auf Kinder, die ohne die neuen Vorschulen in elterlicher Betreuung geblieben wären zu identifizieren. Es wird diskutiert, weshalb die Messung dieses Effekts von zentraler Bedeutung für die Wirkungsmessung von Vorschulangeboten ist und besser als konventionelle Schätzer (z.B. reduced-form estimates) für Vergleiche zwischen verschiedenen Studien und Kontexten geeignet ist. Es wird zunächst gezeigt, dass der Effekt auf die kognitiven Fähigkeiten von Kindern in kontrafaktischer elterlicher Betreuung, unter realistischen Annahmen, auf zwischen 0.14 und 0.39 Standardabweichungen eingegrenzt werden kann. Anschließend wird eine Instrumentvariablenschätzung angewandt, um die Effektgröße näher einzugrenzen. Die Ergebnisse zeigen, dass der Effekt in etwa 0.2 Standardabweichungen beträgt. Der Effekt auf Kinder, die ohne das neue Programm eine andere Vorschule besucht hätten, ist klein und nicht statistisch signifikant.

Kapitel 6 untersucht den Zusammenhang verschiedener Qualitätsindikatoren kambodschanischer Vorschulen mit der Entwicklung frühkindlicher kognitiver und sozio-emotionaler Fähigkeiten auf Basis eines value-added Modells. Diese Studie basiert auf einer Stichprobe von 327 Vorschulen und etwa 3,000 Kindern in Kambodscha. Die Qualitätsindikatoren basieren auf detaillierten Unterrichtsbeobachtungen und beinhalten mehrere Dimensionen von Struktur- und Prozessqualität, wie zum Beispiel die pädagogischen Ansätze der Lehrkraft zur Vermittlung des Unterrichtsstoffes und die Berücksichtigung individueller Bedürfnisse von Kindern. Mithilfe einer Faktorenanalyse wird zunächst aufgezeigt, dass häufig verwendete Qualitätsindikatoren in weitere latente Variablen unterteilt werden sollten. Jedoch lässt sich auch mit detaillierten Qualitätsindikatoren kein eindeutiger Zusammenhang zwischen der Lehrqualität und frühkindlicher kognitiver Entwicklung nachweisen. Auch mit der sozio-emotionalen Entwicklung bestehen nur schwache Zusammenhänge. Jedoch zeigt sich deutlich, dass Kinder in Vorschulen mit besonders geringer Strukturqualität, aufgrund eines Mangels an grundlegender Ausstattung, geringere kognitive und sozio-emotionale Fähigkeiten haben. Die Ergebnisse legen nahe, dass im untersuchten Kontext zunächst universelle Qualitätsstandards für alle Vorschulen erreicht werden sollten, bevor eine gezielte Förderung einzelner Schulen angestrebt wird. Der fehlende Zusammenhang zwischen der beobachteten Prozessqualität und frühkindlicher Entwicklung kann erklären, weshalb die Effekte von Lehrerfortbildungen auf Lernerfolge oftmals geringfügig sind und zeigen die Schwierigkeiten bei der Erfassung effektiver Bildungs- und Betreuungsmethoden auf.

Kapitel 7 fasst zunächst anhand bestehender Literatur zusammen, wie die Ernährungssituation von Kindern und deren häusliches Umfeld, einschließlich des Ausmaßes, in dem sie kognitive Stimulation und emotionale Unterstützung von den Eltern erhalten, eine tiefgreifende Rolle für die frühkindlichen Entwicklung spielen. Dieses Kapitel ergänzt die bestehende Literatur mit Evidenz dazu, welche komplementären Einflüsse die frühkindliche Erziehung und Ernährungssituation auf die Entwicklung bei kambodschanischen Kindern im Vorschulalter haben. Hierzu werden Paneldaten aus den Jahren 2016-17 zu elterlicher Erziehung, dem Ernährungsstatus und der frühkindlichen Entwicklung (exekutive Funktionen, sprachliche Entwicklung, frühe Rechenfertigkeiten und sozio-emotionale Entwicklung) von 6,508 kambodschanischen Kindern im Alter von 3 bis 5 Jahren genutzt. Die Ergebnisse zeigen, dass Ungleichheiten in der frühkindlichen Entwicklung, die mit dem sozio-ökonomischen familiären Hintergrund verbunden sind, bereits im Alter von 3 Jahren deutlich zu beobachten sind und sich bei Kindern im Alter von 4 und 5 Jahren weiter verstärken. Mit Hilfe eines hierarchischen Regressionsmodells lässt sich ein signifikanter Anteil dieser Ungleichheiten durch Unterschiede in der elterlichen Erziehung und dem frühen Ernährungsstatus, gemessen durch Wachstumsstörungen, erklären. Für besser gebildete Eltern lässt sich mehr kognitive Stimulation und emotionale Unterstützung der Kinder beobachten. Allerdings ist die positive Assoziation zwischen elterlicher Erziehung und frühkindlichen kognitiven Fähigkeiten bei Kindern ohne Wachstumsstörungen um 35%-54% stärker und die elterlichen Aktivitäten erklären nur etwa 8%-14% des kognitiven Entwicklungsunterschieds zwischen dem untersten und obersten Wohlstandsquintil. Die Ergebnisse unterstreichen den Bedarf an zusätzlicher Forschung, die die Wechselwirkungen zwischen Einflussfaktoren, die den Wohlstand der Familie mit der Entwicklung des Kindes verbinden, aufzeigt.

Mit pathogenen Mikroorganismen kontaminierte Lebensmittel können zu schwerwiegenden Erkrankungen führen und stellen ein Risiko für den Verbraucher dar. Dabei haben vor allem Lebensmittel tierischen Ursprungs ein hohes Übertragungspotential von Lebensmittelerregern auf den Menschen. Lebensmittelassoziierte Krankheitserreger wie Escherichia (E.) coli spielen eine bedeutende Rolle als Verursacher von Lebensmittelinfektionen beim Menschen. Durch den zunehmenden Einsatz von Antibiotika zur Bekämpfung bakterieller Infektionen wird die Entwicklung von Resistenzen gefördert. Bakteriozine stellen als antimikrobielle Proteine eine vielversprechende Alternative dar, weil sie nicht toxisch auf eukaryotische Zellen wirken und ubiquitär in der Umwelt und somit auch natürlicherweise im Lebensmittel vorkommen. Die von E. coli produzierten Colizine und Mikrozine gelten als große Hoffnung pathogene E. coli im Lebensmittel zu reduzieren. Ziel dieser Arbeit war es, E. coli-Isolate aus Lebensmittelproben auf das Vorhandensein von Bakteriozinen im Spot-Assay zu untersuchen und die Bakteriozinbildner anschließend weitestegehend zu charakterisieren. Dafür wurde mittels PCR zum einen auf das Vorhandensein bekannter Colizin- und Mikrozin-Gene und zum anderen die Einteilung der Isolate in bestimmte Phylogruppen untersucht. Darüber hinaus erfolgte eine Bestimmung des Molekulargewichtes der Bakteriozinextrakte mittels SDS-PAGE. Mittels Zellzahlreduktions-versuchen wurde der Einsatz von Colizinen im Nährmedium und im Lebensmittel untersucht. Aus verschiedenen Lebensmittelamatrices konnte bei 97,5 % der Proben E. coli isoliert und mittels MALDI-TOF MS-Analyse bestätigt werden. Dabei wurden 51,3 % der E. coli-Isolate aus Fleisch isoliert. Mindestens einen Referenzstamm lysierten 52,5 % der isolierten E. coli im Spot-Assay, wobei der Referenzstamm E. coli K-12 DH5α die höchste Sensitivität aufwies. Die Colizin-Gene B, E7 und Ia konnten am häufigsten bei den untersuchten lysierenden E. coli-Isolaten aus Lebensmitteln nachgewiesen werden. Die E. coli-Isolate wurden überwiegend den Phylogruppen A (37,5 %) und B1 (36,3 %) zugeordnet. Neun der 33 untersuchten Proteinextrakte zeigten eine Lyse-Bande zwischen 55 und 75 kDa auf dem Overlay-Agar des inkubierten SDS-Gels. Nach Detektion von Bakteriozinbildnern und Charakterisierung der Bakteriozine, wurde ein Bakteriozin-Konzentrat hergestellt. In Versuchen wurde anschließend ermittelt, ob dieses Konzentrat in der Lage ist, den Referenzstamm E. coli K-12 DH5α unter definierten Bedingungen zu reduzieren. Das Konzentrat wurde aus der Übernachtkultur des E. coli 2116-Isolates hergestellt. Das Isolat wies Gene für die Colizine E2 und K sowie für das Mikrozin B17 auf. Die E. coli 2116 Übernachtkultur wurde mittels Tangentialflussfiltration bei einem Cut Off von 100 kDa filtriert und bei einem Cut Off von 30 kDa aufkonzentriert. Nach unmittelbarer Zugabe des E. coli 2116-Konzentrats konnte die Zellzahl von E. coli K-12 DH5α von 107 KbE/ml sowohl bei 37°C als auch bei 4°C-Umgebungstemperatur um sechs Log-Stufen reduziert werden. Bei 37°C Umgebungstemperatur stieg die Zellzahl innerhalb von 24 h wieder auf 108 KbE/ml an. Bei 4°C Umgebungstemperatur blieb die Zellzahl über die Gesamtversuchsdauer konstant bei 101 KbE/ml. Der Einsatz von Bakteriozinen zur Reduktion von E. coli-Zellzahlen ist im Nährmedium bei einer Umgebungstemperatur von 4°C gut geeignet. Bei der Zugabe von E. coli 2116-Konzentrat zu Hackfleisch kam es abhängig vom eingesetzten Volumen des Konzentrates zu Reduktionen von einer Log-Stufe bis unterhalb der Nachweisgrenze. Je höher verdünnt das Konzentrat eingesetzt wird, desto weniger Bakteriozine sind im Ansatz vorhanden, weshalb die Reduktion schwacher ausfällt. Vermutlich kommt es zusätzlich zu Wechselwirkungen zwischen Bakteriozinen und Bestandteilen des Fleisches. Für zukünftige Versuche sollten deshalb verschiedene Bakteriozinkonzentrate auf unterschiedlichen Lebensmittelmatrices getestet werden. Außerdem sollten Versuche durchgeführt werde, welche die Mechanismen und Wechselwirkungen zwischen Bakteriozinen und verschiedenen Matrices genauer untersuchen.

Infrastructure providers rely on the execution of third-party applications to offer their platforms to customers and researchers. The execution of each application without a preceded security review poses a risk for the system and the overall IT infrastructure. However, if the unreviewed application is available for execution, it is also available for an analysis to generate appropriate countermeasures to prevent unwanted behaviour. This work investigates the capabilities of an automated secure environment. This environment is generated based on the analysis of applications that are only available in their machine code format. The analysis focuses on the interaction of the application with the operating system through the system call interface. Therefore this work describes required technologies and mechanisms to collect data, process it and generate rules for a secure environment to protect assets from attacks. This process and the result environment is tested with real-world applications and attacks to determine its effectiveness and overall costs. It is shown that the described solution is able to decrease the rate of successful attacks against the system from 83% to 9% in selected use-cases. This is achieved with an execution overhead of 823 ms average. These results demonstrate that it is possible to utilise automatic software analysis pipelines to build restricted execution environments for pre-compiled applications. It also highlight the advantages and limitations of the selected approach to focus the analysis on the system call interface.

Global human land alteration poses an immense threat to natural ecosystems with consequences to wildlife populations on numerous scales. Land is altered for agricultural supply of the human population, the gain of resources, and industrial production as well as for providing living space. The sprawl of urban agglomerations around the world creates novel ecosystems for wildlife species that remained in natural remnants enclosed by urbanization processes or actively colonized urban areas. These novel habitats provide high food abundances and diverse breeding opportunities but are also characterized by a high degree of human-induced disturbance, land- and light pollution and habitat fragmentation. The capacity to adjust to these novel environmental conditions depends on the behavioural plasticity or ecological flexibility of species (and individuals), or on rapid evolutionary processes that provide the genetic base for adaptive trait changes. The red fox (Vulpes vulpes) as a generalist predator of medium size and a broad geographic distribution, managed to successfully inhabit cities around the globe. Due to their ubiquitous presence in human dominated landscapes, ranging from agricultural land to densely built-up areas, it is commonly assumed that red foxes cope well with human presence. Although the fox’s inherent behavioural plasticity obviously enables the species to populate those areas, living in close proximity to humans may come with some downsides too. High mortality rates, low average life spans and elusive behaviours indicate a trade-off for this naturally shy generalist that is poorly addressed. Do we thus draw wrong conclusions about the actual boundaries of the behavioural plasticity of red foxes, based on shallow observations? To address this issue we (i) looked at genetic patterns on the population level by analysing red fox samples across a rural to urban continuum. We investigated how the urban matrix affected gene flow in foxes and how the urban environment potentially shaped the red fox population genetics beyond the effects of single landscape elements. We then (ii) researched space use of foxes on an individual level by radio-collaring individuals across the Berlin area. We examined how foxes adjust their habitat use within the city depending on landscape - including manmade structures such as built-up areas and traffic infrastructure - as well as on human presence and activity. The results of the first chapter revealed that gene flow between urban and rural fox population of Berlin and Brandenburg was limited, resulting in two genetically separable populations. Landscape did effect gene flow through the urban matrix to a certain extent but seemed to play a minor role for fox dispersal. For instance, while built-up areas had only weak impeding effects on gene flow despite their high degree of urbanization, urban green spaces like city parks and forests did not serve as gene flow enhancement either. Foxes avoided crossing the city border and predominantly dispersed along urban transport infrastructure such as larger streets and railways, despite the inherent mortality risk. This indicates that also human-induced fear drives dispersal behaviour in the studied red fox population. The second chapter reports on movement and space use of the foxes based on the comparison of used to available habitat. The results show that foxes did not avoid built-up areas or high degrees of imperviousness (ground sealing), while high human population densities were avoided. The foxes further did not preferentially select green spaces like public parks or urban forests. Wasteland areas - including verges along railways - and gardens of residential houses were predominantly used by the studied individuals, providing sites inaccessible to humans or with low human presence. Finally avoidance of humans was more distinct during times of human activity. The results pinpoint that the foxes’ space use was partly driven by avoidance behaviour towards humans. Our study showed that although foxes cope well with the urban landscape as a species, human presence has consequences on a population level and on an individual scale. Human local and temporal activities pushed the foxes into an adjustment of movement patters and their use of the urban habitat. The results also revealed the limits of this adjustment even in a flexible species like the red fox. We hope that our findings enhance the consideration of multiple factors beyond landscape for future studies on the ecology of wildlife.

Additional works For studying space use behaviour, we used radio collars that include a tri-axial accelerometer that measures deflections of the unit within the three-dimensional space. As recorded acceleration data hold an understudied potential to analyse animal behaviour using remote tracking, we also included a methodological work into our project (third chapter). We radio collared captive foxes and documented the behaviours they displayed during measurement, to train different machine learning programs in the inference of behaviours from the acceleration data. We showed that neural networks may provide an improved ability for the classification of animal behaviours from acceleration data using machine learning compared to established approaches. The presence of foxes in urban areas also concerns people. In addition to the possible transmission of diseases, foxes cause damage on private property and in public spaces and induce disturbances (e.g. due to odours or noises). Short fleeing distances and unfamiliar approaches of the animals (in gardens and sometimes even houses) stir up fears in the population, but can also create annoyance and calls for control of the urban fox population. We therefore conducted a representative survey to look more closely into the factors affecting wildlife perception (fourth chapter). We found that attitude towards and risk perception of foxes mainly influenced the participants’ preferences on whether and how to deal with the fox population, while factual knowledge did not influence their positions. Risk perception and attitude mainly depended on education, age, gender and living environment of the participants.

Ethanol is the most frequently abused drug in the world with a worldwide mortality of three million deaths each year (5.3% in 2016). Excessive ethanol consumption can adversely affect several organs, where it can lead to cellular and mitochondrial toxicity. The brain is particularly susceptible to ethanol toxicity. The exact mechanism by which ethanol affects the brain is still unknown. Dopaminergic neurons are believed to play a role, especially in the rewarding and reinforcing (i.e. addictive) effects of alcohol. Thus, model systems enabling to detect early toxicity in human dopaminergic neurons may help to discover potential treatment strategies. In this study, I used human induced pluripotent stem cells (iPSCs) from control individuals and from patients with alcohol use disorder (AUD) to generate neural progenitor cells (NPCs) and neuronal cultures containing dopaminergic neurons. I developed a high-content analysis (HCA)-based assay, dubbed Mitochondrial Neuronal Health (MNH) assay, to quantitatively assess mitochondrial neuronal health in iPSCderived neurons following ethanol exposure. Acute and chronic exposure of ethanol on iPSC-derived neurons led to decreased mitochondrial membrane potential (MMP) and decreased neuronal branching complexity in a dose-dependent manner. Ethanol toxicity was similarly observed in control neurons and in neurons derived from individuals with AUD. Finally, I adapted the MNH assay to perform a proof-of-concept compound screening to identify modulators of ethanol-induced neurotoxicity. I found that the two drugs disulfiram and baclofen, which are used to treat AUD, and lithium caused neurotoxicity. In contrast, the spasmolytic drug flavoxate had a positive effect on mitochondrial neuronal health.

The properties of biological interfaces play a role in all processes where organisms interact with their environment. The mucus barrier encountered in higher organisms is a prominent example of such an interface and regulates passage of nutrients and pathogens. On a molecular level, cellular membranes composed of lipid bilayers represent another fundamental biological interface, which is almost always in contact with solutions containing ions or other solutes. This thesis studies the properties of these two interfaces and the interactions with their environment.

First, the barrier properties of an uncharged mucus analogous hydrogel are analyzed based on non-normalized experimental concentration profiles of penetrating tracer molecules. For this, a numerical model of the diffusion process is developed that allows for the extraction of diffusion constants of the tracer particles in the bulk solution and in the hydrogel, as well as free energy differences from which partition coefficients are computed. The computational extraction method is validated by comparison of the obtained diffusion constants with results from experiments and with scaling laws from polymer theory. Based on the extracted partition coefficients a free volume model is developed, which takes into account the tracer and hydrogel flexibility. The model suggests a broad pore size distribution of the unordered hydrogel, in which the larger pores are found to predominantly determine the partitioning process, a phenomenon which might be general to unordered biological hydrogels like mucus.

The second part of this thesis covers the interactions of lipid bilayers in contact with solutions containing different co-solutes or ions, which are analyzed using atomistic molecular dynamics simulations. The hydration repulsion of lipid bilayers, commonly observed for nanometer separations, is found to be universally increased by the presence of co-solutes. This effect is quantitatively reproduced from experiments, thus validating the modeling approach. The added repulsion is in a next step modeled as an osmotic pressure afforded by the co-solutes and further augmented by the incorporation of solute-solute and solute-lipid interactions. Finally, ionic adsorption to the lipid interface is investigated in detail by computation of the surface potential obtained from a combination of equilibrium and non-equilibrium simulations in the presence of an electric field. By developing an electrostatic continuum model, which additionally incorporates the presence of minute amounts of negatively charged surface active impurities, initially counterintuitive experimental data is unified for the first time with simulation results. The assumption of contaminations existing in experiments has previously explained a range of other puzzling surface properties and is found to also allow for in detail modeling of electrophoresis experiments on lipid bilayers in ionic solutions.