Naturwissenschaftliche Fakultät
Refine
Year of publication
Document Type
- Article (1498)
- Doctoral Thesis (721)
- Preprint (4)
- Master's Thesis (3)
- Review (3)
- Bachelor Thesis (1)
- Book (1)
- Part of a Book (1)
Has Fulltext
- yes (2232)
Keywords
- - (427)
- climate change (17)
- graphene (15)
- Helicobacter pylori (14)
- Optics, quantum optics and lasers (14)
- quantum optics (11)
- Quantum information and quantum mechanics (10)
- catalysis (10)
- reactive oxygen species (10)
- remote sensing (10)
Institute
- Naturwissenschaftliche Fakultät (2232)
- Medizinische Fakultät (29)
- Technische Fakultät (3)
- Medizinische Klinik 5 - Hämatologie und Internistische Onkologie (2)
- Neurologische Klinik (2)
- Humangenetisches Institut (1)
- Institut für Biochemie (1)
- Institut für Physiologie und Pathophysiologie (1)
- Medizinische Fakultät -ohne weitere Spezifikation- (1)
- Plastisch- und Handchirurgische Klinik (1)
Rheumatoid arthritis (RA) is an inflammatory autoimmune disease that leads to joint inflammation and disability. Although research has advanced our understanding of RA, the exact cause is still unknown, and only a few patients achieve sustained remission. Autoantibodies, such as rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPAs), are present in the blood of around 60% of RA patients, often appearing years before disease onset. Individuals with detectable RF and/or ACPAs but no clinical symptoms have an increased risk of developing RA. It is speculated that autoantibody production, which targets the synovium, the affected tissue in RA, may be initiated outside the synovium. Lymph nodes (LN) play a crucial role in initiating adaptive immune responses and maintaining self-tolerance. As potential sites for autoantibody production, alterations within the LN may contribute to the loss of tolerance and the development of autoimmune diseases like RA. The objective of this study was to investigate the role of CCL19+ lymph node stromal cells (LNSC) in the development of arthritis. Spatiotemporal depletion of CCL19+ LNSCs in the popliteal lymph node (pLN) prior to the onset of collagen-induced arthritis (CIA) using Ccl19-Cre x iDTR mice significantly reduced disease severity. Spatiotemporal treatment with recombinant CCL19-IgG, injected locally in the footpad, confirmed the results and emphasized the importance of lymphocyte migration to and from the draining lymph nodes. mRNA sequencing analyses revealed that CCL19+ LNSCs downregulate the expression of tropomyosin receptor kinase (Trk) A just before disease onset. In in vitro co-culture assays, blocking TrkA in LNSCs resulted in increased T cell proliferation. Moreover, local treatment with a TrkA inhibitor exacerbated arthritis scores.
In conclusion, the study demonstrated the crucial involvement of CCL19+ LNSCs in the development of inflammatory arthritis. Targeting these cells via Trk could potentially serve as a therapeutic approach to prevent arthritis in at-risk patients. These findings shed light on the underlying mechanisms of arthritis onset and suggest novel avenues for intervention in the disease.
Long-time dynamics with ultrafast resolution in correlated electrons and electron-lattice systems
(2023)
The ultrafast time-evolution of the electronic structure in correlated
solids is often intertwined with the dynamics of the crystal lattice,
collective orders, or slow electronic variables such as non-thermal band
occupations, which might be orders of magnitude slower than the
intrinsic femtosecond electronic timescale. It is the interplay between
the electrons and the other degrees of freedom in solids, as well as the
interactions of the electrons among themselves, that leads to many
of the fascinating phenomena that are observed in condensed matter
using ultrafast time-resolved spectroscopies. Photoinduced insulator-to-
metal transitions in correlated Mott insulators, pump-induced melting
and recovery of charge density waves and non-thermal transitions to
inhomogeneously disordered and metastable states, represent just a few
examples. A better understanding of the non-trivial relaxation pathways
of electronic and structural degrees of freedom in these systems may
lead to new concepts for the description and manipulation of complex
materials. Here we report the microscopic theories and the numerical
methods that we developed to simulate the full thermalization dynamics
of strongly correlated electrons and coupled electron-lattice systems
that takes place on the slow timescales, while still taking into account
accurately the fast electronic degrees of freedom.
The interacting time-dependent Green’s function defined on the Keldysh
contour contains the full complexity of the nonequilibrium many-
body problem for the electronic degrees of freedom; its time evolution
is determined by the solution of the differo-integral Kadanoff-Baym
equations (KBEs). Through a controlled truncation of the memory
integrals in the KBEs, we can simulate both the fast prethermalization
and the slow relaxation dynamics in the paramagnetic and the symmetry-
broken antiferromagnetic Hubbard model. Afterwards, we discuss a
quantum Boltzmann equation (QBE) which only assumes a separation
of timescales, but is based on a non-perturbative scattering integral, and
makes no assumption on the spectral function such as the quasiparticle
approximation. In particular, a scattering integral corresponding to
nonequilibrium dynamical mean-field theory (DMFT) is evaluated in
terms of an Anderson impurity model in a nonequilibrium steady
state with prescribed distribution functions. With this, we successfully simulate the relaxation dynamics, following an interaction quench, of
a strongly correlated metal in the vicinity of the Mott transition in the
Hubbard model. In the second part of the dissertation, we discuss a
semiclassical approach to treat the coupled electron-phonon dynamics.
The effect of electronic fluctuations on the phonon is kept beyond
Ehrenfest dynamics, leading to a stochastic phonon evolution with
damping and noise terms that are self-consistently determined by the
electronic correlation functions in the fluctuating phonon field. We
benchmark results for the Anderson-Holstein model against numerically
exact quantum Monte Carlo data, and we find good agreement for the
phonon distribution function at temperatures comparable to the charge
ordering temperature. Together with a solution of the electronic model
based on the non-perturbative quantum Boltzmann equation, and with
a non-equilibrium generalization of statistical DMFT, we exploit the
semiclassical approach to address the coupled electron-lattice dynamics
in a photo-induced charge density wave transition in the Holstein model.
We simulate an ultrafast inhomogeneously disordered state where the
non-thermal order parameter on the lattice shows a spatial bimodal
distribution that cannot be described by the small Gaussian fluctuations
around the average taken into account in macroscopic theories like the
Ginzburg-Landau theory.
Quantum control has recently been a problem of great relevance and is gaining even more traction. In many fields, including quantum technologies, quantum metrology, and quantum computing, control of a system is essential.
In physics, quantum control has been studied both from the theoretical and the experimental point of view. Its tasks include purification of quantum states, optimal control, quantum feedback, and quantum state preparation. Each task is essential for studying and developing scalable and reliable quantum systems.
While quantum control was only approached through theoretical analytical solutions and gradient-based techniques in the past, in recent years, there has been considerable interest in applying machine learning (ML) techniques to such problems. One of the most promising branches of ML to do so is called Reinforcement Learning (RL).
RL was applied successfully to solve complex problems that required finding control strategies in computer games and robotics. Reinforcement learning is a very general technique, and its framework can be applied with little effort to various tasks, so physicists saw it as a powerful tool for their quantum control tasks. Among the most notable applications of reinforcement learning in quantum control are state preparation, state transfer, parameter estimation, gate and circuit design, and quantum error correction.
This thesis will define the main concepts of machine learning and display the general framework of Reinforcement Learning and a few state-of-the-art algorithms. Several essential concepts of quantum control will also be presented, focusing on two RL applications in the field.
One of the applications is a deep reinforcement learning scheme used for preparing and stabilizing quantum Fock states and superpositions. The studied system is a cavity subject to quantum-non-demolition detection of photon number and controlled by only a simple linear drive, with RL reaching high fidelities in a multitude of tasks without any prior knowledge of the physical system. This work was inspired on an experimental application of Prof. Benjamin Huard’s experimental group at ENS Lyon, who also collaborated on the manuscript.
The second application showcases an extension of the popular GRAPE algorithm. GRAPE is a gradient-based technique primarily used to optimize quantum systems’ control sequences. Our implementation incorporates discrete and continuous strong stochastic measurements into an RL-inspired approach
This powerful technique is illustrated on a Jaynes-Cummings model with feedback. The strategies found by RL can prepare states and stabilize them in the presence of noise and are also human-interpretable.
All of the work in this thesis has been conducted under the supervision of Prof. Dr. Florian Marquardt at the Max Planck Institute for the Science of Light from January 2019 to March 2022.
The cranial neural crest plays a fundamental role in orofacial development and morphogenesis. As a pluripotent and dynamic cell population, the cranial neural crest is undergoing vast transcriptional alterations throughout embryogenesis and the formation of facial structures. Changes in expression patterns are brought about by several transcription factors and chromatin modifying complexes such as the Ep400/Tip60 complex. As combined chromatin remodeler and histone acetyltransferase the complex has two catalytically active subunits, Ep400 and Kat5. Previously it was noted that a neural crest specific deletion of either Ep400 or Kat5 caused orofacial deformations in mice and that heterozygous KAT5 mutations in humans are linked to the development of orofacial clefts. While the relevance of the Ep400/Tip60 complex for the cranial neural crest is thus established, the molecular causes and mechanisms were still unknown.
In this study we selectively inactivated Ep400 and Kat5 in the murine cranial neural crest cell line O9-1 and analyzed the transcriptomic changes to examine its role in neural crest biology. The inactivation of either subunit independently resulted in the reduced expression of genes linked to pathways involved in carbohydrate and amino acid metabolism as well as protein synthesis. Chromatin precipitation established the direct presence of Ep400/Tip60 in the promoter regions of several deregulated genes. In vitro assays confirmed the reductions in glycolytic ATP production and translation on a cellular level, as well as a significant reduction in proliferation rate. Following Ep400 and Kat5 inactivation in the mouse using floxed alleles and the neural crest-specific Wnt1-Cre, the same genes were found to be downregulated in the neural crest-derived early embryonic pharyngeal arch tissue. The tissue also showed reduced proliferation and increased apoptosis. In later embryonic stages, homozygous Kat5 inactivation resulted in almost complete ablation of all orofacial strictures, whereas heterozygous inactivation led to the development of a cleft palate and hypoplasia of the mandible.
In summary, this study demonstrated a direct regulation of energy metabolism and protein synthesis by Ep400/Tip60 in the cranial neural crest. Impaired activity of these processes by Ep400/Tip60 mutations led to reduced proliferation, increased apoptosis, and hypoplasia of the neural crest and its derivatives, which in the heterozygous state resulted in the formation of orofacial clefts. This study established the importance of chromatin remodeling as an important factor in early cranial neural crest development and orofacial cleft pathogenesis.
The simultaneous spectral and morphological analyses in complex regions in Imaging Atmospheric Cherenkov Telescopes data became possible due to the recently developed 3D template background model.
However, the capabilities of the model are partially restricted by systematic errors, particularly at low energies. Due to these uncertainties, the energy range of the analysis is usually shortened. A proper assessment of the model’s systematic errors can validate its accuracy. Nevertheless, to date there is no general method of estimating it.
This work presents a method for calculating the systematic errors of a 3D template background model, which also includes the error for any dataset and improvements in the analysis due to the convolution of the error in the maximum likelihood computation. The method generates lookup tables with the expected level of systematic error for the background model per run, in energy bins. By construction, the method is applicable to any version of a 3D template background model and any dataset and the produced lookup tables can be used regardless of the origin of the error for estimating the error of any dataset.
This work presents the improvement of incorporating systematic errors in analysis, through nuisance parameters and it is validated on a Crab Nebula dataset. This procedure widens the energy range towards low energies and provides a more accurate background estimation, and consequently a better characterization of the region in the analysis.
The origins of the highest levels of systematic errors presented in the method are explored. The first one, caused by the spatial miss modeling at low energies, is a generic issue arising from the model’s construction method and it is likely to appear in other versions. The second one, caused by the technique used for gamma-hadron separation intrinsic to this specific model version.
Finally, this framework was applied to spectral variability studies of the Galactic Center central source HESS J1745–290 in the TeV range, with H.E.S.S. data, including systematic errors in the 3D template background model. Motivated by the common analysis method, the corresponding source 4FGL J1745.6–285 observed by Fermi-LAT in the GeV range was included in this study. As a result of this investigation, no statistically significant spectral variability was observed nor was a correlation between the data from the different experiments. These results place constraints on two astrophysical models that predict spectrum variability due to flaring episodes from the central Super Massive Black Hole Sgr A* on a 15 year time scale.
Die Ausbildung von ektopisch lymphatischen Strukturen (ELS), die in ihrem Aufbau den sekundär lymphatischen Organen (SLO) ähneln, stellt ein Kennzeichen chronischer Entzündung dar. Das Auftreten von ELS ließ sich in den Meningen in einer Gruppe von Patienten mit sekundär progredienter Multipler Sklerose (SPMS) beobachten. Hierbei war eine Verbindung mit einer ausgeprägten kortikalen Pathologie und einem schwereren Krankheitsverlauf zu erkennen. Die in dieser Studie verwendete MP4-induzierte experimentelle autoimmune Enzephalomyelitis (EAE), ein chronisches Mausmodell der Multiplen Sklerose (MS), spiegelt durch die Entstehung von meningealen und perivaskulären B-Zell-Aggregaten die progrediente Phase der MS wider. Insgesamt zielte die hier durchgeführte Studie darauf ab, Initiatoren und Schlüsselmoleküle der B-Zell-Aggregatentstehung und -aufrechterhaltung zu identifizieren.
Der erste Teil der Studie fokussierte sich auf die Identifikation potenzieller Initiatoren der B-Zell-Aggregatbildung in der MP4-induzierten EAE. Zu diesem Zweck wurden Kleinhirne und Milzen MP4-immunisierter C57BL/6 (B6) Mäuse mittels Durchflusszytometrie (FACS) in der akuten und chronischen Krankheitsphase auf cluster of differentiation (CD)3-CD5-CD4+ RAR-bezogener Orphan-Rezeptor-γ (RORγt)+ Lymphgewebeinduktor (LTi)-Zellen und CD3+CD5+CD4+RORγt+ T-Helfer (TH)17-Zellen untersucht. Wegen der Abwesenheit von B-Zell-Aggregaten eigneten sich Myelin-Oligodendrozyten-Glykoprotein (MOG):35-55-immunisierte Mäuse als Kontrollmodell. Die Untersuchungsergebnisse zeigten weder im MP4- noch im MOG:35-55-Modell zu beiden Analysezeitpunkten LTi-Zellen. Jedoch konnte in beiden Mausmodellen eine Population an CD3-CD5-CD4-RORγt+ angeborenen lymphatischen Zellen (ILC) detektiert werden. Darüber hinaus überwogen TH17-Zellen in der chronischen Krankheitsstufe der MP4-immunisierten Mäuse.
Um weitere Schlüsselmoleküle in Bezug auf die B-Zell-Aggregatbildung und -aufrechterhaltung zu ermitteln, wurde im zweiten Teil der Studie überprüft, ob ein bestimmtes Zytokin- und Chemokinprofil mit der Schwere der B- und T-Zell-Pathologie korreliert. Hierfür wurden 34 Zytokine und Chemokine in Liquor cerebrospinalis (CSF)-Proben und den entsprechenden Serumproben von MP4-immunisierten Mäusen in der chronischen Krankheitsphase mittels MAGPIX® gemessen. Gleichzeitig wurde die B- und T-Zell-Pathologie im Kleinhirn dieser Versuchstiere mit Hilfe von Immunhistochemie (IHC) bewertet. Zusätzlich wurde die Expression entsprechender Chemokinrezeptoren im Kleinhirngewebe der MP4-immunisierten Mäuse untersucht. Die CSF-Analyse ergab eine signifikant erhöhte Konzentration an bestimmten Zytokinen und Chemokinen in MP4-immunisierten Mäusen im Vergleich zu Vehikel-immunisierten Mäusen. Die Unterschiede der meisten Zytokin- und Chemokinlevel waren unabhängig von der Schwere der B- und T Zell-Pathologie im Kleinhirn. Jedoch zeigte der C-C-Motiv-Chemokinligand (CCL)1 eine Assoziation mit hoher B- und T-Zell-Pathologie. Des Weiteren wurde eine signifikante Hochregulation von CCL1, CCL5, CCL7, CCL12, CCL22 und dem C-X-C-Motiv-Chemokinligand (CXCL)13 im CSF von MP4-immunisierten Mäusen mit einer hohen Anzahl an B-Zell-Aggregaten im Kleinhirn im Vergleich zu solchen ohne B-Zell-Aggregate detektiert. Zudem ergab sich eine signifikant erhöhte Expression von CXCL16 im Serum der Mäuse ohne B-Zell-Aggregate. Während der C-C-Motiv-Chemokinrezeptor (CCR)5 unabhängig von der Schwere der B- und T-Zell-Pathologie vorzufinden war, traten der C-X-C-Motiv-Chemokinrezeptor (CXCR)5 und CXCR6 vorwiegend im Zusammenhang mit hoher Pathologie auf.
Neben der Identifikation neuer Schlüsselmoleküle der B-Zell-Aggregatbildung und -aufrechterhaltung, fokussierte sich der dritte Teil der Studie auf das Enzym Matrixmetalloproteinase (MMP)-3. Eine Ribonukleinsäure (RNA)-Sequenzierung zeigte bereits, dass dieses Enzym in B-Zell-Aggregaten in der chronischen Phase von MP4-immunisierten Mäusen hochreguliert war. Um diese Ergebnisse auf die humane Erkrankung zu übertragen, wurden IHC-Färbungen von MMP-3 in Gehirnschnitten von MS-Patienten durchgeführt. Zudem wurde unter Verwendung von in-vitro-Analysen der Effekt von MMP-3 auf humane B-Zellen untersucht. B-Zellen wurden aus dem Blut von gesunden Probanden isoliert und vor der Zugabe von rekombinanter MMP-3 (rMMP-3) entweder mit R-848 und Interleukin (IL)-2 stimuliert oder unstimuliert weiterverwendet. Zur Überprüfung des Aktivierungsstatus der B-Zellen nach MMP-3-Behandlung wurde die Expression von Aktivierungsmarkern (CD69, CD80 und CD86) mit Hilfe von FACS bestimmt. Darüber hinaus wurde eine Analyse von B-Zell-bezogenen Genen unter Verwendung von RT2 profiler PCR arrays durchgeführt. Des Weiteren wurde zur Quantifizierung des IL-6-Levels der MMP-3-behandelten B-Zellen Dot Blot-Arrays und Enzyme-linked immunosorbent assays (ELISAs) verwendet. Die Untersuchung von MMP-3 ergab eine erhöhte Expression des Enzyms in Hirnschnitten von MS-Patienten, die B- und T-Zellen im Gewebe aufwiesen, im Vergleich zu solchen, die nur T-Zellen zeigten. Außerdem bewirkte MMP-3 sowohl eine erniedrigte Expression von CD69, CD80 und CD86 als auch eine Herunterregulation B-Zell-bezogener Gene. Ebenso wurde nach der Behandlung der B-Zellen mit MMP-3 eine erniedrigte IL-6-Konzentration detektiert.
Einerseits deutet diese Studie auf TH17-Zellen als potenziellen Initiator in der B-Zell-Aggregatentstehung in der MP4-induzierten EAE hin. Andererseits scheinen Untersuchungen von biochemischen Veränderungen im CSF eine geeignete Methode darzustellen, um Immunprozesse im CSF und ZNS der MP4-immunisierten Mäuse zu analysieren. Ein spezielles Profil an Chemokinen, die im CSF hochreguliert waren, und deren korrespondierende Rezeptoren scheinen in der Bildung und Aufrechterhaltung der B-Zell-Aggregate involviert zu sein. Des Weiteren wurde MMP-3 mit einer erhöhten B-Zell-Pathologie in Hirngewebe von MS-Patienten assoziiert und zeigte einen B-Zell-modulierenden Effekt durch Erniedrigung der B-Zell-Aktivierung und Zytokinproduktion.
Climatic impact on tree growth and seasonal wood formation along altitudinal gradients on Corsica
(2023)
The western Mediterranean is a hotspot for climate change. Trees in this area generally have to face a climate regime with a distinct dry period in summer, which is projected to intensify in the coming years. It is essential to understand how tree growth and seasonal wood formation are affected by climatic parameters in the past and today to anticipate how tree growth will change in the future.
The mountainous island of Corsica constitutes an excellent study area to research tree growth on an elevation gradient: Its ridges reach above the upper tree line on a rela-tively small area. Furthermore, extensive pine forests envelop large parts of the island. Pinus is the only tree genus that is distributed from the Mediterranean conditions at sea level up to subalpine climate conditions. The two native species Pinus pinaster (low elevation) and Pinus nigra subsp. laricio (high elevation) cover the island with an over-lapping area in mid-elevation. It is not yet fully understood how tree growth and its climatic control factors vary along an altitudinal gradient of this magnitude. Corsican pines allow for an investigation of a comprehensive climatological spectrum and facilitate growth comparisons within.
This PhD thesis is part of the DFG-project bundle “CorsicArchive - Altitudinal Gradients and Forest Response: Climate, Hydrology and Isotope Variability of a Mediterranean Ecosystem”. The dendroecological subproject (summarised by this thesis) aims to find out how the growth of Corsican pines changed over the past centuries, how the climatic drivers vary across the elevation gradient, and how seasonal wood formation is affected by climate and altitude. A multiparameter approach allowed for the investigation of climate-growth relations on annual and intraannual resolution: Ring-width-based growth trajectories were established for 239 trees; electronical band dendrometers constantly monitored the stem circumference variations of 36 individuals during a remarkably dry 2017 and wet year (2018); and xylogenesis of 35 pines was analysed for a period of three years (2017-2019) and modelled.
The ring-width chronologies revealed generally higher long-term correlations of radial growth with precipitation in low elevations and with temperature in high elevations. Furthermore, the growth trajectories in different age classes indicated that predicted future climate conditions might shift the distribution ranges of both pine species further upward.
The temporal high-resolution measurements of the dendrometers revealed that the trees at high elevations sustained higher tree-water deficits than trees in low elevations in the exceptionally dry year 2017. The positive correlations between growth and precipitation and tree-water deficit and temperature imply that high evapotranspira-tion led to the interruption of growth and stem circumferential shrinking across the island. A novel integration of large-scale synoptic regimes showed that a majority of growth in the dry year occurred during relatively rare weather patterns.
The generalised additive models of tracheid cell differentiation phases unveiled that the growing season begins with a one-month offset between the high, middle, and low elevation bands. Temperature is driving the onset of growth in spring, but the fact that the trees cease growth simultaneously indicates that a combination of different climatic parameters is responsible for growth cessation. Under the same site conditions, P. nigra grows more during a shorter time than P. pinaster. The faster growth points towards a lower wood density, which would make the trees more vulnerable to mechanical stress. This result, combined with the finding that P. nigra is slowly superseded by P. pinaster in mid-elevations, suggests that the distribution of P. nigra will decline in the future.
The final section of this thesis combines and compares several methods for detecting seasonal wood formation. All methods for identifying intraannual growth detection have strengths and weaknesses; a complete picture is only given by a combination of methods.
This thesis improves the knowledge of inter- and intraannual growth of a widespread tree genus and how climatic impacts vary along an altitudinal gradient from Mediter-ranean to subalpine conditions. The findings can serve as a data basis for future growth models but can also support regional decision-makers and forest managers by revealing past and present climate-growth relations to prepare for future climate change.
The transition to CO2 neutral energy production relies on a secure supply of a diverse range of elements that are needed for green technologies. Tellurium (Te) is such an element, as it is critical for thin film photovoltaic (PV) production. However, the processes controlling the distribution and enrichment of Te in natural ore-forming systems are still poorly constrained.
Currently Te is almost exclusively a by-product of copper production, but rising Te demand is projected to outpace that for Cu in the near future, making additional Te sources necessary. High-K calc-alkaline to alkaline igneous hosted porphyry-epithermal systems, are exploited for Cu and Au but also contain appreciable amounts of Se, As, Ag and Te. Three high-K calc-alkaline to alkaline hosted porphyry-epithermal prospects on Limnos Island (Greece) and the alkaline-hosted world-class epithermal Au-Te deposit of Vatukoula (Fiji) were studied by in-situ trace element and S isotope analysis of hydrothermal pyrite. The results demonstrate that pyrite is a powerful tool to trace the fluid conditions which enhanced the deposition of Te and related element (e.g., As, Ag, Au and Se). A preferential fractionation of Te into the intermediate sulfidation epithermal pyrite, in comparison to the porphyry or high sulfidation epithermal pyrite is observed at Limnos. The highest Te concentrations at Limnos, which surprisingly were documented in sericitic alteration pyrite are linked to the successive retreat (telescoping) of the hydrothermal system. At Limnos unstable fluid conditions in relation to phase separation processes appear to hinder the concentration of Te, while stable fluid conditions are beneficial for the concentration of Te in pyrite. Nonetheless, the recovery of Te from pyrite at Limnos is projected to remain sub economical given the common median concentrations below 10 µg/g. The Vatukoula low sulfidation epithermal system by contrast features high levels of Te in tellurides and pyrite. Here an evolution from intermediate to low sulfidation fluid conditions has facilitated the incorporation of large quantities of Te into As rich pyrite. The ubiquity of pyrite at Vatukoula and its high Te contents of up to 1.4 wt. % further imply that pyrite could act as a future Te ore. However, current ore processing steps such as roasting strip Te from the pyrite which is processed for Au. Hence, to modify the processing for Te recovery from pyrite, a detailed mineralogical, chemical, and structural understanding of the incorporation is necessary. This thesis unveils, that there exists a positive correlation between As and the maximum Te content of pyrite at Vatukoula, up to an As threshold of up to 13 wt. %. Notably, pyrite with higher As contents between 23 and 43 wt. % only incorporates a maximum of 185 µg/g of Te, a previously unrecognized correlation. The extreme enrichment of Te in pyrite of Vatukoula are shown to consist of three separate mechanisms: (1) Concentration of Te in the crystal lattice, significantly surpassing the previously suggested solid solution limit in relation to As. (2) Nano-telluride inclusions, intra-grain
remobilized as melts in nano-cracks. (3) Crystal defects which exhibit notable enrichments of up to 1.6 wt. % Te though pipe diffusion. In conclusion, among the investigated porphyry and epithermal environments, alkaline igneous-rock hosted low sulfidation epithermal deposits hold the highest potential to be a future source of Te. Additionally, the consideration of pyrite is important for Te recovery as pyrite hosts up to 60 wt. % of the bulk Te in ores from Vatukoula. This thesis identified previously unrecognized Te incorporation mechanisms and revealed unknown relations of Te and As incorporation.
Summary
A fully functional neuritic network within the brain is essential for its integrity and a healthy life. Neurodegenerative diseases such as Parkinson’s disease (PD) and Huntington’s disease (HD) alter neuritic processes leading to psychiatric, motor, and cognitive impairments. Currently, no cure is available for either of these diseases. A better understanding of the underlying pathogenesis is there-fore crucial. Evaluating different disease models in vivo and in vitro mimicking important features of neurodegenerative diseases is a prerequisite to unravel processes interfering with the integrity of neuritic networks. In this thesis, the compensatory neuritogenesis within the striatum of a rat PD model and dysfunc-tional neuronal processes related to mutant huntingtin as well as a possible hun-tingtin lowering strategy were examined within primary HD-patient fibroblasts and patient-derived induced pluripotent stem cells (hiPSCs), neural progenitor cells (NPCs), and cortical neurons.
As main hypothesis, we investigated the following: In vivo and in vitro models of neuronal protein aggregation disorders present a powerful platform to reveal dis-ease specific compensatory mechanisms and to assess therapeutic targets. The aims were divided investigating PD and HD in an in vivo and in vitro model, re-spectively. Thus, the first aim of the present thesis is to examine whether the serotonergic input towards the dorsal striatum is altered due to the reduced do-paminergic input in a transgenic rat PD model. The second aim investigates whether mutant huntingtin causes dysfunctional neuronal processes and investi-gates a possible pharmacological intervention in primary HD-patient fibroblasts and patient-derived hiPSCs, NPCs, and cortical neurons.
Addressing the first aim, a transgenic rat PD model ubiquitously expressing hu-man α-synuclein using a bacterial artificial chromosome (BAC hα-syn tg rat) was used. The dopaminergic and serotonergic neuritic network was analyzed within the dorsal striatum, motor cortex M1/M2, raphe nuclei, and substantia nigra pars compacta of 12-month-old BAC hα-syn tg and non-tg rats. Structural analysis of the dopaminergic innervation showed a significant decrease in dopaminergic fi-ber density within the dorsal striatum of BAC hα-syn tg rats. Furthermore, the dopaminergic cell count within the substantia nigra was significantly reduced. In addition, a significant increase in serotonergic fiber density was detected by 100% within the dorsal striatum. In contrast, no structural changes of the dopaminergic and serotonergic innervation within the motor cortex M1/M2 were present. No changes in the number of serotonergic neurons and proximal fiber network within the dorsal and median raphe were observed. Immunofluorescence staining and imaging of transporter and enzymes important for L-DOPA uptake, dopamine synthesis, and dopamine release revealed the potential capacity of serotonergic neurons to synthesize dopamine from L-DOPA and its release. These findings indicate a site-specific compensatory neuritogenesis of serotonergic afferents within the dorsal striatum of BAC hα-syn tg rats. Together with the capacity of serotonergic neurons to synthesize and release dopamine, this might play a role in L-DOPA induced dyskinesia.
Addressing the second aim, an in vitro HD model, including HD-patient-derived fibroblasts, hiPSCs, NPCs, and cortical neurons, was established. HiPSCs showed a significant increase in NANOG expression while other pluripotency markers displayed no differences between HD and control indicating pathophysi-ological changes already at a stem cell level in cells derived from HD-patients. No differences in proliferation marker Ki67 and neural marker βIII-tubulin and CTIP2 expression were observed upon cortical neuron differentiation between HD-patients and controls. Furthermore, the expression of huntingtin was ana-lyzed in the in vitro HD model. No differences in total huntingtin levels were ob-served between HD-patient and control-derived fibroblasts, hiPSCs, NPCs, and cortical neurons. Mutant huntingtin was solely detected in HD-patient-derived cells. Treatment of HD-patient and control-derived fibroblasts and hiPSCs using Branaplam, a small molecule splice modulator, accomplished a profound reduc-tion in total and mutant huntingtin. This finding validates the in vitro HD model as a powerful screening platform for huntingtin lowering compounds and Branaplam as a potential compound to lower huntingtin levels in HD-patients.
These key findings demonstrate that in vivo and in vitro disease models are a powerful prerequisite for understanding underlying neurodegenerative processes in PD and HD. Thus, underlying pathogenic mechanisms are uncovered and new treatment strategies maybe tested in these models.
Decomposition Methods for Time-Dependent Mixed-Integer Nonlinear Optimization Problems on Graphs
(2023)
Decomposition can be the method of choice to deal with optimization problems that contain hard to solve model structures or that are of large scale.
The main idea is to decompose the problematic aspects of the problem into multiple smaller blocks that can be solved more easily.
Here, the challenge is to combine the single pieces to a solution that is not only feasible but maybe even optimal for the original problem.
In many cases, this can be done by introducing an iteration that eventually converges to a desired solution.
In this cumulative dissertation, we present several iterative decomposition methods that are tailored to different types of optimization models and use distinct approaches to split up the problems.
Our main motivation for this originates from the optimization of gas transport networks, where we encounter partial differential equations as well as discrete control decisions.
Additionally, we engage in the related field of district heating network optimization to study the challenges arising from large-scale and fully discretized systems as well as undesirable model features such as, \eg, complementarity constraints.
Here, we introduce two temperature mixing models that are well suited for optimization and a number of techniques to speed up the solution process, which are applied in numerical experiments.
As a next step, we develop an iterative time-domain decomposition method that is applied to optimal control problems subject to semilinear hyperbolic systems of partial differential equations.
For this, we derive first-order optimality conditions that are then split using a non-overlapping decomposition of the time horizon.
We exploit the fact that the resulting systems have a primal interpretation as so-called virtual control problems.
We prove the convergence of the iterative method and develop a posteriori error estimates.
Later, we extend the scheme to systems of ordinary differential equations with mixed-integer controls by using Pontryagin's maximum principle.
We again show the convergence and conduct a numerical case study.
Moreover, we use a consensus-based version of the classic penalty alternating direction method to solve tailored reformulations of transient gas network problems that allow us to minimize the number of coupling constraints between sub-problems.
Here, we utilize the quasi-separable structure of the network to decompose it into sub-networks with more desirable properties.
We also discuss different decomposition strategies and test them in a numerical case study.
Finally, we present a successive linear relaxation method for mixed-integer nonlinear problems with multivariate Lipschitz continuous nonlinearities.
The distinguishing feature of this algorithm is that it exploits no properties of the nonlinearities besides the Lipschitz constants.
Therefore, the method is applicable for problems with non-convex or even non-differentiable constraints.
The nonlinearities do not even need to be given in a closed form, which allows us to integrate black-box constraints into the model.
We prove that the algorithm converges to an approximate global optimum and we provide a worst-case estimate for the number of iterations.
The iterative method is applied to stationary gas transport problems, where implicitly given solutions of the differential equations are modeled via black-box constraints.
Professor Jens Kossmann was the Specialty Chief Editor of the Plant Biotechnology section at Frontiers in Plant Science from its inception until his death in March 2023 at the age of 59. He ran the section with great enthusiasm and was an important driver of its success. This obituary is designed to celebrate his contribution to plant science as well as his role as a mentor and friend.
A major part of this thesis covers the late-stage functionalization of pharmaceutical compounds meaning the selective modification of already existing active pharmaceutical ingredients (API).
Some APIs led to selectively chlorinated or fluorinated products after treatment with Selectfluor, whereas others formed a mixture of both. All halogenated products were then subjected to radioligand binding studies, where some showed increased affinity and selectivity.
In another part of this work, aluminum oxide nanoparticles were functionalized with selfassembled
monolayers in order to obtain tunable shell-by-shell systems with desired properties.
Deciphering signaling pathways for LPAR-dependent gene transcription and HCC cell proliferation
(2023)
Human hepatocellular carcinoma (HCC) is one of the most lethal cancers, indicated by a five-year survival rate of less than 15%. However, the molecular mechanisms involved in the carcinogenesis or progression of HCC are not fully elucidated. With limited therapeutic options currently available, there is an urgent pressure to decipher tumorigenic pathways and discover new molecular targets. We recently revealed that Filamin A (FLNA), an actin-binding protein, interacts with Myocardin-related transcription factor A (MRTF-A), which is a co-activator of the Serum Response Factor (SRF). Furthermore, we proved that this interaction regulates the transcriptional activity of MRTF/SRF and drives migration in melanoma cells. Based on the importance of this interaction, we first deciphered the role of FLNA phosphorylation at serine 2152 on the MRTF-A-FLNA interaction. We showed that Lysophosphatidic acid (LPA) stimulation and subsequent protein kinase C (PKC) activation drives FLNA phosphorylation at this residue. Furthermore, we revealed using FRET and immunoprecipitation assays with a non-phosphorylatable FLNA mutant (S2152A) that this LPA-inducible FLNA phosphorylation at serine 2152 acts as a molecular switch that triggers the MRTF-A-FLNA interaction. In addition, we demonstrated that dephosphorylation of FLNA at position 2152 due to PKC inhibition impairs MRTF-A-FLNA complex formation and reduces HCC cell proliferation. Moreover, PKC inhibition prevented the LPA-induced transcriptional activation of MRTF/SRF, which was reflected by a significant reduction in MRTF/SRF-specific target gene expression of SM22 and SRF on mRNA and protein levels.
In search of molecular targets, we investigated the expression of LPA receptors (LPARs) one to six in different liver and melanoma cancer cell lines. Furthermore, we examined the effects of deletion of the most promising LPARs on proliferation and induction of cellular senescence in HCC cells. Next, we demonstrated that LPAR1 is not only responsible for FLNA phosphorylation at serine 2152 but also plays a crucial role in the MRTF/SRF axis. Transient knockdown of LPAR1 resulted in decreased actin stress fiber formation, impaired F-actin polymerization, and downregulation of the RhoA-associated kinase 2 (ROCK2), thereby indicating the inactivation of RhoA. In addition, predominant nuclear MRTF-A was redistributed into the cytoplasm, and MRTF-A-FLNA interactions were significantly decreased after the transient knockdown of LPAR1. Since MRTF/SRF target gene expression relies on both the nuclear accumulation of MRTF-A and the interaction between FLNA and MRTF-A, we analyzed the transcriptional activation of SRF. Consequently, we could demonstrate decreased MRTF/SRF transcriptional activity after LPAR1 depletion, verified by strong downregulation of SM22, SMA, and SRF expression at both mRNA and protein levels. Moreover, depletion of LPAR1 resulted in growth arrest upon senescence induction in HCC cells. We were able to identify oncogene-induced senescence (OIS) as the underlying mechanism by observing dephosphorylation of the Retinoblastoma protein (Rb), ERK phosphorylation, and increased accumulation of H3K9me3.
We established a hitherto unknown interaction between LPAR1 and both MRTF-A and FLNA using proximity ligation assays (PLAs) and immunoprecipitations in HCC and melanoma cells. Next, we investigated the binding domains between the interacting proteins by immunoprecipitation assays with FLNA fragments or MRTF-A deletion mutants. We revealed that FLNA displays three different interaction domains for LPAR1: The first one is located at the N-terminal ABD-domain, the second region includes the amino acids (a.a.) 571-866, and the third one was found at the C-terminal end of FLNA. The interaction site in MRTF-A is located within the a.a. 301-506, however, the exact interaction site remains to be investigated. We propose a model in which FLNA interacts with LPAR1 or MRTF-A in different subcellular compartments. These interactions will, in turn, be regulated by LPA stimulation and subsequent FLNA phosphorylation at serine 2152. Using MRTF-A mutants with predominant nuclear (S454A) or cytoplasmic (N250) localization in immunoprecipitations in combination with PLA and immunofluorescence analyses, we demonstrated that the interactions occur in both nuclear and cytoplasmic compartments. We could demonstrate the increased formation of FLNA CT fragments after FLNA dephosphorylation at serine 2152 mediated by LPAR1/3 inhibition. Remarkably, MRTF-A and LPAR1 share one crucial binding domain with FLNA (a.a. 571-866). The deletion of the prominent binding domain (a.a. 571-866) of FLNA impaired the interaction with both proteins. Furthermore, we could demonstrate that the non-phosphorylatable FLNA mutant S2152A showed strongly decreased LPAR1 and MRTF-A interactions compared to wild-type FLNA. Consequently, we used the FLNA mutant (S2152A), which was less effective in LPAR1 and MRTF-A binding and demonstrated that the mutant regulated the subcellular localization of MRTF-A and the transcriptional activity of MRTF/SRF. Reconstitution of the non-binding FLNA mutants (S2152A, Δ571-866) after FLNA depletion resulted in decreased stress fiber formation, shortened focal adhesions, and induction of cellular senescence. More precisely, impaired complex formation by reconstitution of FLNA S2152A or Δ571-866 resulted in increased PML accumulation within the nucleus, thereby implying the induction of OIS. We showed that the MRTF/SRF-specific target gene SMA, a marker for fibrosis, is LPA-inducible. Furthermore, we could demonstrate that LPAR1 depletion or decreased complex formation strongly downregulates the expression of SMA. We found that the hepatoblastoma HepG2 and Hep3B cells express significantly lower levels of LPAR1 compared to the HCC HuH7 and HuH6 cells, eventually leading to excessive FLNA phosphorylation at serine 2152 in the HCC cells. Importantly, we could provide evidence for the overexpression of LPAR1 in human HCC tissues compared to non-tumorous tissues. We could verify the existence of the complexes in 3D-organoid models mimicking a tumor-specific microenvironment. Finally, we proved that antagonizing LPAR1 via the LPAR1/3 antagonist Ki-16425 is sufficient to downregulate the transcriptional activity of MRTF/SRF and induce proliferation arrest in HCC cells. Taken together, we were able to identify LPAR1 as a promising new therapeutic target for personalized HCC treatment.
Due to the advancing miniaturization in modern technologies and the multitude of microscopy applications, the research field of nanooptics is more important today than ever before. When working with nanostructures, a scattering behavior that can be described purely in terms of dipoles is a common encounter. In such cases, if a detailed understanding of the dipole excitation and the associated light scattering is available, a wide range of possibilities opens up. These possibilities include, in particular, the deliberate control of a known system and the resulting dipole radiation, as well as the reconstruction of excited dipoles as a tool for the investigation of unknown systems. Within this dissertation entitled “Dipoles in Nanooptics: From Their Tailored Excitation to a Nanoscopic Measurement Tool,” the necessary understanding is built and put into practice. For this purpose, versatile methods have been developed to determine the excited dipole moments, characterize optical components, and study sub-diffraction-limit size nanostructures.
At the onset of the work presented here, a theoretical foundation is established by gathering formalisms and equations from existing literature. In this process, all required calculations are adapted to the individual criteria applicable here and combined into a unified theory. This theoretical construct is divided into the following integral parts. First, tightly focused light fields are calculated from arbitrary paraxial input fields using vectorial diffraction theory. Second, the dipole moments excited in spherical nanoparticles are investigated via Mie theory. Last, the electromagnetic fields emitted by the dipoles are calculated utilizing Green functions, whereby both the resulting interaction with potential surrounding nanostructures and the radiated far-fields are determined.
The experiments carried out in the framework of this thesis begin with the tailored excitation of extraordinary dipole moments and the investigation of thereby occurring scattering phenomena. For this purpose, specific paraxial light beams were tightly focused, and nanoparticles were selectively placed in the resulting focal fields. In these experiments, many possibilities were discovered that could be used for nanoscale manipulation of light in later applications. In addition, the fundamental understanding of the physical processes was refined, and the practical know-how for their measurement was obtained. These aspects include, in particular, the reconstruction of the dipole moments excited in the nanoparticles, which is essential for later experiments.
Subsequently, the expertise for measuring dipole moments was used in two projects for a practical application. First, an absolute characterization method was developed for high numerical aperture microscope objectives. Here the term “absolute” denotes that no calibrated reference object is needed for the measurement. This requirement for a calibration reference is a significant problem in many characterization methods for cutting-edge optical elements, which was circumvented here. The measurement's key ingredient was using the fields radiated by a dipole as a nearly perfect reference wave. After that, the second application deals with objects almost a million times smaller than the previously examined microscope objective. More specifically, it involves the study of clusters of spherical gold nanoparticles with a size below the diffraction limit of visible light. The investigated particles were scanned through a known focus field, and polarization-resolved far-field measurements were recorded. In the data analysis, the theoretical dipolar scattering of such particles is then used to implement an inverse modeling approach to retrieve unknown parameters of the sample. Although these particles are not optically resolvable with conventional microscopy, it was possible to identify the vast majority of particles and determine their positions and sizes down to a few nanometers.
In conclusion, it can be noted that throughout this dissertation, especially the interplay of theory and the designed experiments have proven to be very fruitful. Ultimately, the theory allowed most of the subsequent experimental results to be reproduced partially or as a whole. These calculations allowed for estimating the effects of many potential sources of error. Thereby, experimental limitations were often known in advance, and much time was saved in the implementation. Moreover, and most importantly, it enabled the extensive inverse reconstruction approaches to be implemented in the first place.
Algebra is a crucial component of mathematics education as it introduces learners to the mathematical world of modeling relationships and handling abstract quantities. The increasing volume of scholarly work in the field has been analyzed qualitatively in numerous systematic reviews—a quantitative breakdown of the field, however, remains a desideratum to date. With this study we contribute to closing this gap by reporting on the results of a bibliometric analysis. We retrieved data from Scopus and Web of Science databases and analyzed 1,825 articles published between 2003 and 2022. On the one hand, we provide insight into the current state of algebra education from primary up to tertiary education by describing the scientific production and its bibliographic topography. On the other hand, we analyzed the data to identify trends and future directions. The results of our study indicate, among other aspects, that APOS Theory and Realistic Mathematics Education are emerging themes in the field that have great potential to shape future research.
The main focus of this thesis is on the study of singular limits related to scalar conservation laws. These are first-order partial differential equations that describe how the amount of a physical quantity in a given region of space changes over time, solely determined by the flux of that quantity across the boundary of the region.
The first part of this manuscript deals with nonlocal regularizations of scalar conservation laws, where the flux function depends on the solution through the convolution with a given kernel. These models are widely used to describe vehicular traffic, where each car adjusts its velocity based on a weighted average of the traffic density ahead. First, we establish the existence, uniqueness, and maximum principle for solutions of the nonlocal problem under mild assumptions on the kernel and flux function. We then investigate the convergence of the solution to that of the corresponding local conservation law when the nonlocality is shrunk to a local evaluation (i.e., when the kernel tends to a Dirac delta distribution). For kernels of exponential type, we analyze this singular limit for initial data of bounded variation as well as for merely bounded ones, using Ole\u{\i}nik-type estimates. We also demonstrate how the techniques developed in this analysis can be used to study the long-time behavior of a nonlocal regularization of the Burgers equation and to show that the asymptotic profile is given by the $N$-wave entropy admissible solution. We also investigate the role played by artificial viscosity in the nonlocal--to--local singular limit process. Finally, we study the boundary controllability problem for nonlocal traffic models.
In the second part of this thesis, we address the controllability of scalar conservation laws on networks and its relationship to the vanishing viscosity singular limit. Our main analysis is carried out in the linear case: for a linear advection-diffusion equation, we show that the cost of controllability blows up exponentially as the viscosity parameter vanishes for small times and decays exponentially for a sufficiently long time-horizon. Finally, for nonlinear conservation laws, we prove a controllability result for entropy solutions using a Lyapunov approach and highlight the stability of this result when a small viscosity is added.
Investigating plant/root-soil interactions at different scales is crucial to advance the understanding of soil structure formation in the rhizosphere. To better comprehend the underlying interwoven processes an explicit, fully dynamic spatial and image-based modeling at the pore scale is a promising tool especially taking into account experimental limitations. We develop a modeling tool to investigate how soil aggregation, root growth and root exudates mutually interact with each other at the micro-scale. This allows the simultaneous simulation of the dynamic rearrangement of soil particles, the input and turnover of particulate organic matter, root growth and decay as well as the deposition, redistribution and decomposition of mucilage in the rhizosphere. The interactions are realized within a cellular automaton framework. The most stable configuration is determined by the amount and attractiveness of surface contacts between the particles, where organo-mineral associations preferably lead to the formation of soil aggregates. Their break-up can be induced by root growth or the degradation of gluing agents previously created after the decomposition of particulate organic matter and mucilage. We illustrate the capability of our model by simulating a full life cycle of a fine root in a two-dimensional, horizontal cross section through the soil. We evaluate various scenarios to identify the role of different drivers such as soil texture and mucilage. We quantify the displacement intensity of individual particles and the variations in local porosity due to the change in available pore space as influenced by the root growth and observe compaction, gap formation and a biopore evolution. The simulation results support that the deposition of mucilage is an important driver for structure formation in the rhizosphere. Although mucilage is degraded within a few days after exudation, it leads to a persistent stabilization of the aggregated structures for both textures in the vicinity of the root within a time frame of 1000 days. Local porosity changes are quantified for exudation periods of 1, 10 and 100 days and are already pronounced for short-term exudation of mucilage. This stabilization is significantly different from the structures encountered when only POM could trigger the evolution of gluing spots, and is still present after complete degradation of the root.
Standard subspaces are a well studied object in algebraic quantum field theory (AQFT). Given a standard subspace V of a Hilbert space H, one is interested in unitary one-parameter groups on H with U(t)V⊆V for every t∈R+. If (V,U) is a non-degenerate standard pair on H, i.e. the self-adjoint infinitesimal generator of U is a positive operator with trivial kernel, two classical results are given by Borchers’ Theorem, relating non-degenerate standard pairs to positive energy representations of the affine group Aff(R) and the Longo–Witten Theorem, stating the the semigroup of unitary endomorphisms of V can be identified with the semigroup of symmetric operator-valued inner functions on the upper half plane.
In this thesis we prove results similar to the theorems of Borchers and of Longo–Witten for a more general framework of unitary one-parameter groups without the assumption that their infinitesimal generator is positive. We replace this assumption by the weaker assumption that the triple (H,V,U) is a so called real regular one-parameter group.
After providing some basic theory for regular one-parameter groups, we investigate regular one-parameter groups that are infinitesimally generated by operator-valued Pick functions. We provide criteria for a Pick function to generate a regular one-parameter group and in this case we provide a formula for the multiplicity space of the regular one-parameter group and show that the dimension of the multiplicity space is compatible with composition of Pick functions.
We also investigate reflection positive regular one-parameter groups and provide a normal form for them. We do this by linking them to reflection positive Hilbert spaces of the form (L²(R,K),H²(C+,K),θ(h)) with some complex Hilbert space K and some function h∈L∞(R,U(K)), where the involution θ(h) on L²(R,K) is given by (θ(h)f)(x)=h(x)·f(−x), x∈R. In the multiplicity free case K=R we give a full classification of all functions h∈L∞(R,T) for which the triple (L²(R,C),H²(C+),θ(h)) is a maximal reflection positive Hilbert space. Also we give an explicit description of the Osterwalder–Schrader transform of these reflection positive Hilbert spaces.
We apply these results to prove that every positive contractive Hankel operator on H²(C+) can be extended to an involution θ(h) on L²(R,C) for which the triple (L²(R,C),H²(C+),θ(h)) is a maximal reflection positive Hilbert space.
Finally, we apply our results about (reflection positive) regular one-parameter groups in the context of standard subspaces and use them to provide analogues of Borchers’ Theorem and the Longo–Witten Theorem to pairs (V,U) for which the triple (H,V,U) is a real regular one-parameter group or for which the quadruple (H,V,U,J(V)) is a real reflection positive regular one-parameter group respectively.
The CD83 molecule is a powerful immunomodulatory protein which contributes to the resolution of inflammation. Expression of CD83 is a hallmark of mature dendritic cells (DCs), but this molecule is also expressed on other activated immune cells including B cells, T cells, regulatory T cells (Tregs), macrophages (Mφ) and on thymic epithelial cells (TECs). Two naturally occurring isoforms of this protein exist: a membrane-bound form (mCD83) and a soluble form (sCD83), which is the extracellular domain of the mCD83 protein. The soluble CD83 protein has profound pro-resolving functions, since administration of sCD83 in specific autoimmune disease and transplantation models, leads to reduction of disease severity on the one hand and on the other hand sCD83 application in transplantation significantly prolonged allograft survival by Treg induction. Conclusively, sCD83 is a potential candidate for new therapeutical treatment options for future clinical trials in the field of autoimmunity and transplantation.
Recent studies elucidated the biological role of the mCD83 protein on specific cell types, including DCs and Tregs, using conditional KO mice (cKO). Mice with CD83-deficiency on Tregs as well as DCs developed severe autoimmune reactions in the animal model for multiple sclerosis – i.e. the experimental autoimmune encephalomyelitis (EAE) model - compared to control animals. Therefore, the function of CD83 on Tregs as well as DCs indicates a similar pro-resolving function, as reported for the soluble CD83 protein (sCD83).
The first aim within the present study was to establish a new therapeutic concept of transplant-mediated tolerance induction by the sCD83 protein, within the high-risk cornea transplantation model, in cooperation with the Cornea Lab, Department of Ophthalmology, University Hospital of Cologne (PD Dr. rer. nat. Felix Bock and M.Sc. Alfrun Schönberg). Importantly, we show that sCD83 preincubation of donor-corneal allografts induces tolerogenic mediators within the graft recipients and prolongs graft survival. Preincubation of donor corneal tissue with sCD83 led to induction of Tregs, regulatory DCs as well as a shift from pro-inflammatory, classically activated Mφ (CAM) towards pro-resolving, alternatively-activated Mφ (AAM), in eye draining lymph nodes of graft recipients. Therefore, we further analyzed the impact of the sCD83 protein on Mφ and DC differentiation. Although previous studies reported an immunoregulatory function of sCD83 on the transition from iDCs to mDCs, there are no studies investigating whether sCD83 affects the monocyte to DC differentiation processes. Intriguingly, sCD83 induced a tolerogenic DC phenotype, which was characterized by low expression of costimulatory molecules, e.g. CD80 and CD86, whilst pro-resolving molecules including CD200R and Msr-1 were upregulated. MLR assays revealed a less stimulatory capacity of allogeneic T cells in the presence of sCD83-differentiated DCs. The phenotype of regulatory DCs was also confirmed by our collaborators, using the corneal transplantation model in vivo. Importantly, next to DCs, also Mφ play an important role in the modulation of the alloimmune response and there are no studies regarding the effect of sCD83 on Mφ. Therefore, we assessed the effect of sCD83 on Mφ and revealed that sCD83 induces alternatively activated Mφ (AAM)-like phenotypic and functional changes, which are able to induce Tregs.
Within the second aim of this study, the biological role of the mCD83 protein on Mφ was analyzed using conditional KO mice. First we assessed the expression of CD83 on these cells, using different pro- and anti-inflammatory stimuli, to generate two distinct Mφ populations: the pro-inflammatory, classically-activated Mφ (CAM) and pro-resolving, alternatively-activated Mφ (AAM), which have different phenotypes and functions. CD83 was stably and significantly upregulated on IL-4-stimulated pro-revolving AAM, while upon stimulation with pro-inflammatory stimuli such as LPS, Zymosan, TNF-α and IFN-ɣ, CD83 was only transiently upregulated. Furthermore, we found out, that CD83 is upregulated during the resolution of inflammation, indicating a pro-resolving function of CD83 expressed by Mφ.
To further characterize the functional relevance of CD83 on Mφ, we crossed CD83-floxed mice with CX3CR1-cre mice, to generate the conditional KO of CD83 on CX3CR1+ myeloid cells. Interestingly, deletion of CD83 results in a downregulation of MHCII and CD86 on Mφ, which is in line with previous studies investigating B cells as well as DCs. Moreover, we show that CD83 deletion results in a defect in the AAM phenotype and function. IL-4 stimulated CD83-deficient Mφ show an overactivated phenotype, which is characterized by reduced CD200R as well as Msr-1 expression levels and in line with that, we observed increased proinflammatory cytokine secretion in CD83-deficient AAM, including TNF-α, IL-6, CXCL1 and G-CSF. When we analyzed the phagocytic activity – a hallmark of AAM - CD83-deficient AAM showed a striking decrease in the engulfment of E.coli bacteria, which correlates with the observed reduced Msr-1 expression. When we analyzed the ability to kill E.coli, by unstimulated CD83-deficient Mφ, we observed an increase in their killing capacity, which was attributed to an increased ROS production. Moreover, when we cocultured CD83-deficient Mφ with allogeneic splenocytes in mixed lymphocyte reaction (MLR) assays, T cells proliferated stronger and proinflammatory cytokine expression levels were increased, whilst frequencies of Tregs were significantly reduced.
Taken together, within the present study, we established a new therapeutic approach in transplant mediated tolerance induction, using the high-risk cornea transplantation model, by preincubation of the donor tissue with sCD83. In addition, sCD83 modulates Mφ towards a pro-resolving AAM-like phenotype, whilst mCD83 deletion modulates Mφ towards a proinflammatory phenotype.