Fakultät für Biologie - Digitale Hochschulschriften der LMU - Teil 05/06 Archiv

In the mammalian genome, cytosine methylation (5mC) plays a central role in the epigenetic regulation of gene expression and has been implicated in a variety of biological processes, including genome stability, imprinting or differentiation. Compared to other epigenetic marks, DNA methylation has been thought to be relatively stable. However, genome-wide loss of 5mC, or DNA demethylation, has been observed in specific developmental stages and in various types of cancer. The discovery of the TET family of enzymes in 2009 was a watershed moment in comprehending the mechanisms of DNA demethylation. TET proteins oxidize 5mC to 5- hydroxymethylcytosine (5hmC), 5-formlycytosine (5fC) and 5-carboxylcytosine (5caC), which not only serve as key intermediates in active DNA demethylation pathways, but can also act as independent epigenetic marks. In this study, various aspects of TET-mediated DNA demethylation have been intensively investigated. Using quantitative mass-spectrometry-based proteomics readers for the different cytosine derivatives in mouse embryonic stem cells (ESCs), neuronal progenitor cells, and adult mouse brain tissue were identified. Readers for these modifications are only partially overlapping and are dynamic during differentiation. Moreover, the oxidized derivatives of 5mC recruit distinct transcription regulators as well as a large number of DNA repair proteins, implicating DNA damage response as the main pathway contributing to active DNA demethylation. To identify additional non-canonical DNA bases, highly sensitive quantitative mass-spectrometry led to the discovery of 5-hydroxymethyluracil (5hmU) in ESCs. Genomic 5hmU is not generated via deamination of 5hmC, as widely suggested, but through direct oxidation of thymine by TET proteins. In addition, screening for specific 5hmU readers identified different transcriptional and epigenetic factors, implicating that this mark has a specific function in ESCs. So far, only little is known how TET enzymes are regulated and how they are modified by posttranslational modifications (PTMs). Mapping TET phosphorylation and glycosylation sites at amino acid resolution revealed that these PTMs are interdependent and mostly occur at regulatory protein regions. Finally, a reporter gene based assay could demonstrate that in vitro methylation causes gene silencing while subsequent oxidation, resulting in DNA demethylation, leads to gene reactivation in vivo. Different knockout and rescue experiments clearly show that oxidation of methylcytosine by TET proteins and subsequent removal by TDG or NEIL glycosylases and the base excision repair pathway results in reactivation of epigenetically silenced genes. In conclusion, this work provides new insights how TET proteins can set DNA modifications, how these oxidized bases are read by various factors and how TET proteins can be posttranslationally modified. Furthermore, removal of 5mC is achieved through TET-mediated oxidation and depends on the activity of specific glycosylases, which leads to gene reactivation.

Viele Vogelarten sind monogam, aber es kommt zugleich häufig zu Paarungen außerhalb des Paarbundes (Griffith et al. 2002). In einer großen Zahl von Studien wurde bereits der adaptive Wert der aus diesem Verhalten hervorgehenden „Fremdvaterschaften“, insbesondere für Weibchen, anhand einer Kosten-Nutzen-Analyse auf genetischer Ebene untersucht. Die Ergebnisse dieser Studien haben jedoch bisher nicht zu generalisierbaren Resultaten geführt (Akçay & Roughgarden 2007). Im Gegensatz dazu gibt es relativ wenig Informationen über verhaltensbegründete Kosten und Vorteile von Fremdvaterschaften, obwohl diese ein Verständnis der Mechanismen von Fremdverpaarungen erleichtern können. Die hier vorliegende Studie untersucht, wie zeitliche, räumliche, und soziale Faktoren beeinflussen wann, wo, und warum bestimmte Individuen Nachkommen mit einem anderen als ihrem sozialen Partner zeugen. Ich untersuche dies an der Blaumeise (Cyanistes caeruleus), einer Art, für die genetische Konsequenzen von Fremdvaterschaften bereits detailliert untersucht wurden, bei der die Kenntnisse im Bereich des außerpaarlichen Verhaltens selbst aber noch unvollständig sind. In den ersten beiden Studien stelle ich einen neuartigen Ansatz zur Untersuchung von Fremdvaterschaften vor und validiere diesen. Fremdverpaarungen erfordern als Beteiligte sowohl ein Weibchen als auch ein Männchen und beide können beeinflussen, ob eine Paarung stattfindet. In dem von mir vorgeschlagenen Ansatz werden nicht nur Informationen über das Individuum, sondern auch über potentielle und realisierte außerpaarliche Partner mit eingebunden. Bei Arten wie der Blaumeise, deren Territorien nicht nur in einzelnen, sondern in allen Aspekten des Brutverhaltens (Nahrungssuche, Balz, etc.) genutzt werden, sind Interaktionen räumlich oft auf Individuen aus nahegelegenen Territorien beschränkt. Auch außerpaarliches Verhalten zeigt eine solche Beschränkung. Im vorliegenden Ansatz werden daher räumliche Verhaltensstrukturen explizit in das Modell mit einbezogen. Um solche räumliche Verhaltensinformationen für einen Langzeitdatensatz über Brutverhalten (die hier eingesetzen Daten von Blaumeisen umfassen 12 Jahre und 2 Populationen) zu erhalten, ist es notwendig, im Nachhinein die Position der Territorien der Brutpaare abzuschätzen. In der Ökologie werden zu diesem Zweck immer häufiger Thiessen-Polygone eingesetzt (z.B. Wilkin et al. 2006, Valcu & Kempenaers 2008). Im 1. Kapitel (Chapter 1) belege ich daher zunächst die Gültigkeit einer solchen Abschätzung von Territorien mit Hilfe von Thiessen-Polygonen, basierend of 14 publizierten Studien, für die detaillierten Informationen über Territorien vorliegen. Im 2. Kapitel (Chapter 2) stelle ich dann den neuartigen Ansatz im Detail vor. Anschließend wende ich diesen auf den genannten Langzeitdatensatz an, nutze dabei Thiessen-Polygone als Approximation für die räumliche Beschränkung von außerpaarlichem Verhalten und untersuche gleichzeitig die Korrelation relevanter Parameter mit außerpaarlichem Fortpflanzungserfolg. Meine Resultate bestätigen die Ergebnisse früherer Studien und zeigen, dass dieser Ansatz genutzt werden kann, um Hypothesen zu testen, die mit bisherigen Methoden nicht statistisch robust überprüft werden konnten. Fremdverpaarungen können früh morgens stattfinden und aus einer Studie über die nahe verwandte Kohlmeise (Parus major) geht hervor, dass Weibchen, die ihren Schlafplatz morgens früher verließen, mit größerer Wahrscheinlichkeit außerpaarliche Nachkommen hatten. Im 3. Kapitel (Chapter 3) untersuche ich mit Hilfe von korrelativen Daten über vier Jahre diesen Zusammenhang für Blaumeisen. Darüber hinaus wurden die Aufstehzeiten von Weibchen über zwei Jahre hinweg experimentell manipuliert, um einen direkten kausalen Zusammenhang zwischen Aufstehzeiten und außerpaarlichem Fortpflanzungserfolg zu ergründen. Ich konnte keinen solchen Zusammenhang feststellen, jedoch führte das Experiment interessanterweise in den zwei Jahren zu entgegengesetzten Ergebnissen. Dies könnte ein Hinweis darauf sein, dass das Verhalten von Weibchen am frühen Morgen durchaus Relevanz für Fremdverpaarungen hat, dass jedoch eine Interaktion mit Umwelteinflüssen diese Zusammenhänge beeinflussen kann. Im 4. Kapitel (Chapter 4) beschreibe ich, dass Blaumeisen vor oder während der Legephase der Weibchen immer wieder die Nester fremder Paare besuchen. Für Männchen korrelierte dieses Verhalten auch mit ihrem außerpaarlichen Fortpflanzungserfolg. Es erwies sich nämlich, dass ein Männchen, welches das Nest eines Weibchens besucht, mit deutlich größerer Wahrscheinlichkeit außerpaarliche Nachkommen mit diesem Weibchen zeugt. Im Gegensatz zu einer früheren Studie an Blaumeisen (Kempenaers et al. 1992) zeigt dieses Ergebnis, dass nicht nur das Verhalten des Weibchens, sondern auch das des Männchens einen wichtigen Einfluss auf die außerpaarliche Fortpflanzung haben kann. Im 5. Kapitel (Chapter 5) betrachte ich abschließend das Balzverhalten der Blaumeisen. Dies erfolgt mittels Audioaufnahmen von Rufen, die speziell im Balzkontext eingesetzt werden (Bijnens & Dhondt 1984). Zunächst beschreibe ich das zeitliche Vorkommen dieser Rufe im Laufe der Brutsaison und im Tagesverlauf. Im nächsten Schritt vergleiche ich diese Daten mit früheren Studien, um zu verifizieren, dass Balzrufe tatsächlich als Maß für das Balzverhalten verwendet werden können. Anschließend untersuche ich dann speziell Balzrufe von Männchen, die in einem Kontext auftreten, der nahelegt, dass die Rufe nicht an das soziale Weibchen gerichtet sind. Es ist wahrscheinlich, dass diese (zu großen Teilen) außerpaarliche Balz anzeigen. Es gab keine Korrelation der „außerpaarlichen Balzversuche“ von Männchen mit ihrem außerpaarlichen Fortpflanzungserfolg, jedoch verloren Männchen mit vielen außerpaarlichen Balzversuchen weniger Vaterschaft in ihrem eigenen Nest. Dies könnte darauf hindeuten, dass die Qualität von Blaumeisenmännchen auf Verhaltensebene relevant für den außerpaarlichen Fortpflanzungserfolg ist. In der allgemeinen Einleitung (General Introduction) beschreibe ich detailliert den biologischen Hintergrund dieser fünf Studien. Weiterhin diskutiere ich die Relevanz von Studien im Bereich der Forschung über Fremdvaterschaften, die anstelle der rein genetischen Betrachtung darüber hinaus auch das Verhalten untersuchen. In der allgemeinen Diskussion (General Discussion) erörtere ich die Implikationen dieser fünf Studien für außerpaarliches Verhalten: Das zeitliche Vorkommen von Balz und Paarungen im Tagesverlauf und über die Saison hinweg, das räumliche Vorkommen, die relative Wichtigkeit des Verhaltens von Männchen und Weibchen und welche Faktoren beeinflussen, dass es zu Fremdverpaarungen zwischen einem ganz bestimmten Männchen und Weibchen kommt.

This dissertation addresses the application of the statistical computing language R in the study of evolution and diversification of plants. The topics included range from the worldwide historical biogeography of the cucurbit family and the phylogenetic composition of the Mediterranean Oxalis flora in central Chile to the interplay between population genetics and climatic niche evolution in four Horde- um clades in the Americas. In these studies, I drew on existing methods in R and on java and C programs that could be easily integrated with R. Whenever necessary, I created additional software available in four new R packages. R's features, e.g., intersystem-interfaces, extensibility, reproducibility and advanced graphical capability, proved well suited for evolutionary and phylogenetic research. My coauthors and I addressed the history of Cucurbitaceae, one of the most economically important families of plants, using a multi-gene phylogeny for 114 of the 115 genera and 25 per cent of the 960 species. Worldwide sampling was achieved by using specimens from 30 herbaria. Results reveal an Asian origin of Cucurbitaceae in the Late Cretaceous, followed by the repeated spread of lineages into the African, American and Australian continents via transoceanic long-distance dispersal (LDD). North American cucurbits stem from at least seven range expansions of Central and South American lineages; Madagascar was colonized 13 times, always from Africa; Australia was reached 12 times, apparently always from Southeast Asia. Overall, Cucurbitaceae underwent at least 43 successful LDD events over the past 60 Myr, which would translate into an average of seven LDDs every 10 Myr. These and similar findings from other angiosperms stress the need for an increased tapping of museum collections to achieve extensive geographical sampling in plant phylogenetics. The second study focused on the interplay of population demography with the evolution of ecological niches during or after speciation in Hordeum. While large populations maintain a high level of standing genetic diversity, gene ow and recombination buffers against fast alterations in ecological adaptation. Small populations harbor lower allele diversity but can more easily shift to new niches if they initially survive under changed conditions. Thus, large populations should be more conservative regarding niche changes in comparison to small populations. My coauthors and I used environmental niche modeling together with phylogenetic, phylogeographic and population genetic analyses to infer the correlation of population demography with changes in ecological niche dimensions in 12 diploid Hordeum species from the New World, forming four monophyletic groups. Our analyses found both shifts and conservatism in certain niche dimensions within and among clades. Speciation due to vicariance resulted in three species with no pronounced climate niche differences, while species originating due to long-distance dispersals or otherwise encountering genetic bottlenecks mostly revealed climate niche shifts. Niche convergence among clades indicates a niche-filling pattern during the last 2 Myr in South American Hordeum. We provide evidence that species that did not encounter population reductions mainly show ecoclimatic niche conservatism, while major niche shifts have occurred in species that have undergone population bottlenecks. Our analyses allow the conclusion that population demography influences adaptation and niche shifts or conservatism in South American Hordeum species. Finally, I studied the phylogenetic composition of Oxalis flora of Mediterranean zone of Chile by asking whether in such a species-rich clade xerophytic adaptations arose in parallel, at different times, or simultaneously. Answering this type of question has been a major concern of evolutionary biology over the past few years, with a growing consensus that lineages tend to be conservative in their vegetative traits and niche requirements. Combined nuclear and chloroplast DNA sequences for 112 species of Oxalidales (4900 aligned nucleotides) were used for a fossil-calibrated phylogeny that includes 43 of the 54 species of Chilean Oxalis, and species distribution models (SDMs) incorporating precipitation, temperature, and fog, and the phylogeny were used to reconstruct ancestral habitat preferences, relying on likelihood and Bayesian techniques. Since uneven collecting can reduce the power of SDMs, we compared 3 strategies to correct for collecting effort. Unexpectedly, the Oxalis flora of Chile consists of 7 distant lineages that originated at different times prior to the last Andean uplift pulse; some had features preadapting them to seasonally arid or xeric conditions. Models that incorporated fog and a `collecting activity surface' performed best and identified the Mediterranean zone as a hotspot of Oxalis species as well as lineage diversity because it harbors a mix of ancient and young groups, including insuficiently arid-adapted species. There is no evidence of rapid adaptive radiation.

Postural control during standing and walking is an inherently unstable task requiring the interaction of various biomechanical, sensory, and neurophysiological mechanisms to shape stable patterns of whole-body coordination that are able to counteract postural disequilibrium. This thesis focused on the examination of central aspects of the functional roles of these mechanisms and the modes of interaction between them. A further aim was to determine the conditions of dynamic stability for healthy standing and walking performance as well as for certain balance and gait disorders. By studying movement fluctuations in the walking pattern it could be demonstrated that dynamic stability during walking depends on gait speed and is differentially regulated for the medio-lateral and the fore-aft walking planes. Stability control in the fore-aft walking plane exhibits attractor dynamics typical for a dynamical system. Accordingly, the most stable pattern of movement coordination in terms of minimal fluctuations in the order parameter (i.e., the relative phase between the two oscillating legs) can be observed at the attractor of self-paced walking. Critical fluctuations occur at increasingly non-preferred speeds, indicating a loss of dynamic gait stability close to the speed boundaries of the walking mode. Moreover, stability control during slow walking is critically dependent on sensory feedback control, whereas dynamic stability during fast walking relies mainly on the smooth operation of cerebellar pacemaker regions. Disturbances of sensory and cerebellar locomotor control in certain gait disorders could be further linked to a loss of dynamic gait stability, in particular an increased risk of falls. Furthermore, this thesis examined alterations in the sensorimotor postural control scheme that may trigger the experience of subjective imbalance and vertigo in the conditions of phobic postural vertigo and visual height intolerance. Both conditions are characterized by an inadequate mode of balance regulation featuring increased levels of open-loop balance control and a precipitate integration of closed-loop sensory feedback into the postural control scheme. This inadequate balance control strategy is accompanied by a stiffening of the anti-gravity musculature and is elicited by specific influences of attention and sensory feedback control. The findings of this thesis contribute to the understanding of central sensorimotor mechanisms involved in the control of dynamic postural stability during standing and walking. They further provide relevant information for the differential diagnosis and fall risk estimation of certain balance and gait disorders.

In Vibrio harveyi induzieren die drei Autoinduktoren (AIs) Z-3-Aminoundec-2-en-4-on CAI 1, N-(3-Hydroxybutyryl)-D-Homoserinlakton HAI 1 und der Furanosylboratdiester AI 2 die Quorum Sensing (QS) Regulationskaskade. So werden unter Anderem Biolumineszenz, Biofilmbildung und Exoproteaseaktivität beeinflusst. Um die Sensorik der QS Kaskade zu verstehen, wurde die Expression verschiedener QS-Zielgene im Wildtyp untersucht. Im Vergleich mit einer AI-negativen Mutante zeigte sich, dass nicht die Zelldichte sondern die Konzentration der verschiedenen AIs und deren Verhältnisse zueinander die QS-Signalkaskade steuern. Beides verändert sich in Abhängigkeit von der Medien-zusammensetzung und erlaubt so die Modulation z.B. der Biolumineszenz (Kapitel 2.1.). Die extrazelluläre AI-Konzentration im Medium hängt entscheidend von der Produktion sowie dem Transport ab. Während HAI-1 aufgrund seiner chemischen Eigenschaften frei aus der Zelle diffundiert, müssen CAI-1 und AI-2 aktiv über die Membran transportiert werden. In Escherichia coli beeinflusst die Deletion des möglichen Transporters YdgG die extrazelluläre AI-2 Konzentration. Im Rahmen dieser Arbeit konnte bioinformatisch mit YhhQ ein zusätzlicher potentieller Exporter identifiziert werden. Bindestudien an beiden Proteinen, zeigten eine mikromolare Affinität für YhhQ jedoch nicht für YdgG. Im Gegensatz dazu ist die extrazelluläre AI-2-Konzentration der ΔydgG/ΔyhhQ Doppelmutante im Vergleich zu den Einzeldeletionen um 4 µM reduziert. In vivo Parallelstudien am YhhQ Ortholog in V. harveyi allerdings, lassen redundante Transportmechanismen vermuten. In diesem Zusammenhang könnte z.B. auch eine durch Phagen vermittelte Zelllyse eine Rolle spielen (Kapitel 2.2.1.). Unabhängig von Produktion und Export kann die externe AI-Konzentration auch durch aktive Reduktion in Form von Import und intrazellulärem Abbau kontrolliert werden. In diesem Zusammenhang sagte die AG Bischofs einen AI-2-Importer vorher und dessen Vorhandensein wurde im Verlauf dieser Promotion experimentell validiert (Kapitel 2.2.2.). Eine weitere Möglichkeit des sogenannten Quorum Quenching ist der Abbau von Acylhomoserinlaktonen wie HAI-1 durch Acylasen und Laktonasen. Bioinformatische Analysen aus der AG Streit identifizierten die fünf potentiellen Laktonasen VIBHAR_02708, VIBHAR_03213, VIBHAR_06370, VIBHAR_06736 und VIBHAR_06900. Jedoch, liegt nach intensiven Untersuchungen am Wildtyp und den „Laktonase“-Mutanten der Schluss nahe, dass es in V. harveyi keinen aktiven HAI-1 Abbau gibt. Vielmehr handelt es sich bei VIBHAR_06370 um eine β-Laktamase, die eine natürliche Ampicillinresistenz vermittelt (Kapitel 2.3.). Im Gegensatz dazu, kodieren VIBHAR_02708 und VIBHAR_03213 die Typ II Glyoxalasen GloB und GloC und tragen damit zur Detoxifikation von Methylglyoxal nicht nur in V. harveyi sondern auch in E. coli bei (Kapitel 2.4.).

The cytoskeleton in most eukaryotes consists of actin filaments, intermediate filaments, microtubules and specific associated proteins. It determines the shape and the polarity of a cell and is inevitable for the coordination of cell movement. The regulation of this complex structure requires a highly organised and specialised signalling network. Ste20-like kinases and the regulator protein Mo25 (Morula protein 25) are part of this signalling network. The main objective of this work was the functional characterisation of the regulator protein Mo25 and the Ste20-like kinases Fray1, Fray2 (Frayed kinase 1/2) and DstC (Dictyostelium serine threonine kinase C) in the amoeba Dictyostelium discoideum (D. discoideum). An additional project was to map and characterise the actin and actin related genes in the genome of the fresh water foraminifer Reticulomyxa filosa (R. filosa). Mo25 is a highly conserved 40 kDa scaffolding protein with a 60% identity from amoeba to man. The disruption of the mo25 gene in D. discoideum results in very large, multinucleated cells which are unable to complete cytokinesis. Growth as well as development is severely delayed in the Mo25-minus strain. Furthermore, in phototaxis assays performed with multicellular aggregates (slugs), the Mo25-minus slugs were unable to migrate towards the light source. These findings imply that Mo25 plays an important role in cytokinesis, growth and cell polarity. We could link the Ste 20-like kinase SvkA (severin kinase), a homolog of the human Mst3, Mst4 (Mammalian Ste20-like kinase 3/4) and Ysk1/Sok1 (Yeast Sps1/Ste20-related kinase 1, Suppressor of Kinase 1) kinases to Mo25 as a binding partner. To further elucidate the interaction of Mo25 with SvkA as well as their role in cytokinesis or polarity signalling, we generated a series of GFP–Mo25 rescue constructs with distinct point mutations in protein-protein interaction surfaces and transformed these into the Mo25-minus background. The kinase domains of the Ste20-like kinases, Fray1 and Fray2 in D. discoideum are highly homologous to the catalytic domains of OSR1 (Oxidative stress response kinase 1) and SPAK (Ste20/SPS1-related proline-alanine-rich protein kinase) in humans and Frayed in fruit fly. Here, we generated the knockout clones Fray1-minus, Fray2-minus, and the double knockout Fray2Fray1-minus in D. discoideum. In developmental studies, Fray2-minus did not show an altered phenotype, whereas Fray1-minus and Fray2Fray1-minus developed slightly slower into fruiting bodies. When grown in shaking culture, Fray1-minus and Fray2Fray1-minus showed a reduced growth rate compared to Fray2-minus and the wild type. In addition, by using a GFP-Trap resin we identified a binding partner of Fray1, a yet unknown protein that we named FRIP (Fray Interacting Protein). FRIP mainly consists of a CBS (Cystathionine beta synthase) domain pair and is 30% identical to the gamma subunit of the AMPK (5‘ adenosine mono phosphate-activated protein kinase) complex in D. discoideum. The Ste20-like kinase DstC has been described to be a regulator of the actin driven process of phagocytosis. The catalytic domain of DstC is most similar to the mammalian kinase Mst2 (Mammalian Ste20-like kinase 2) and Hippo (“Hippopotamus”-like phenotype) of D. melanogaster. We could map the sorting signal that localises DstC to phagocytic cups and acidic vesicles to about 90 amino acids. Here we present an array of distinct point mutations for the identification of the exact localisation signal. R. filosa is a fresh water protist which belongs to the the group of Foraminifera within the Rhizaria. The R. filosa genome is the first foraminiferal and only the second rhizarian genome to be deciphered. In this bioinformatics project, we could identify, map and characterise four new actin genes in addition to the already known actin and about 40 genes that code for potential actin related proteins of seven different protein classes.

Muscle spindles are complex stretch-sensitive mechanoreceptors that consist of 4-12 specialized muscle fibers. These intrafusal muscle fibers are innervated in the central (equatorial) region by an afferent sensory axon and in both peripheral (polar) regions by efferent γ-motoneurons. Until now little is known about muscle spindle development at the molecular level, especially about the development of cholinergic specializations. My study shows that nicotinic acetylcholine receptors (AChR) are concentrated at the γ-motoneuron endplate as well as in the equatorial region. Moreover, enzymes required for the synthesis and removal of acetylcholine, including choline acetyltransferase (ChAT) and acetylcholinesterase (AChE), as well as vesicular acetylcholine transporter (VAChT) and the AChR-associated protein rapsyn are all concentrated at the polar γ-motoneuron endplate and (with the exception of AChE) also at the equatorial region. Finally, the presynaptic protein bassoon, involved in synaptic vesicle exocytosis, is also present at the γ-motoneuron endplate and at the annulospiral sensory nerve ending. During postnatal development, the AChR subunit composition at the γ-motoneuron endplate changes from the γ-subunit containing fetal AChR to the ε-subunit containing adult AChR. This is similar to the postnatal change at the neuromuscular junction. In the equatorial region the ε-subunit expression starts around postnatal week two; however the γ-subunit persists in the central region despite the onset of the ε-subunit expression. Therefore, the γ- and ε-subunits are simultaneously present in the equatorial region. This result was confirmed using a mouse line in which the AChR γ-subunit was genetically labelled by green fluorescence protein (GFP). In this mouse, the GFP-labelled AChR γ-subunits are concentrated at the contact site of the intrafusal fiber with the sensory nerve ending. This result indicates different AChR maturation occurs within two areas of the same intrafusal fiber. I also show that agrin and the agrin receptor complex (consisting of LRP4 and MuSK) are present in muscle spindles in the region of the sensory and motor innervation. Moreover, agrin, MuSK, and LRP4 are expressed by proprioceptive neurons in dorsal root ganglia but only agrin and LRP4 were detected in the cell body of γ-motoneurons in the spinal cord. In mice with a targeted deletion of agrin, AChR aggregates are absent from the polar region and γ-motoneuron endplates do not form. By contrast, AChR aggregates remain detectable in the central part of intrafusal fibers. Moreover, muscle-specific re-expression of mini-agrin is sufficient to restore the formation of synaptic specializations at γ-motoneuron endplates. These results show an unusual AChR maturation at the annulospiral endings and confirm that agrin is a major determinant for the formation of γ-motoneuron endplates. Agrin on the other hand appears dispensable for the aggregation of AChRs in the central region of intrafusal fibers.

The main function of glucocorticoids in corticotroph cells is to suppress proopiomelanocortin, the precursor of the stress hormone adrenocorticotropin (ACTH). Cushing’s disease is a rare but severe neuroendocrine condition caused by partially glucocorticoid resistant corticotroph adenomas, which consequently secrete excessive amounts of ACTH in an uncontrolled fashion. The patients suffer from chronic hypercortisolism due to excessive stimulation of the adrenal glands by ACTH to produce glucocorticoids. Impairing mutations of the glucocorticoid receptor (GR) only sporadically explain the reduced glucocorticoid sensitivity in the adenomas – the molecular mechanism behind the partial resistance is poorly understood. The function of GR depends on direct interactions with the molecular chaperone Hsp90. Both the reduction and overexpression of Hsp90 impedes GR activity in different experimental settings. Therefore, the expression of the inducible Hsp90α isoform was determined in biopsy specimens of corticotroph pituitary adenomas from patients with Cushing’s disease. Its strong overexpression compared to normal human pituitary cells paved the way to study its role in the function of corticotroph adenomas using small molecules which target Hsp90. The three distinct Hsp90 inhibitors 17–AAG, Novobiocin and Silibinin showed antiproliferative effects in AtT–20 cells through the degradation of the oncogenic client kinase Cdc2, a hallmark of pharmacologic inhibition of Hsp90. Surprisingly, only the N–terminal Hsp90 inhibitor 17–AAG caused the degradation of GR, as was reported also for other Geldanamycin–based Hsp90 inhibitors. Neither Silibinin nor the C–terminal Hsp90 inhibitor Novobiocin affected GR protein levels. These converging effects led to the assumption that both compounds bind to the same domain in Hsp90. It was shown here that Novobiocin displaces Silibinin from the C–terminal domain of Hsp90, and that these compounds dissociate mature GR from Hsp90 at the biochemical level. As a result, increased levels of mature receptor were present in the cell able to bind glucocorticoids with high affinity. This novel molecular mechanism proved to potentiate GR transcriptional activity in AtT–20 cells. The potentiation in GR activity also led to enhanced suppression of ACTH elicited by low concentrations of Dexamethasone in AtT–20 cells and in primary cultures of human corticotroph adenomas from patients with Cushing’s disease. In contrast, Silibinin did not show effects on rat normal pituitary cells. Finally, Silibinin reduced tumor growth, partially reverted hormonal alterations, and alleviated symptoms in a mouse allograft model for Cushing’s disease. These results suggest that the regulation of GR sensitivity by overexpressed Hsp90 may represent a pharmacologically reversible mechanism in the pathogenesis of this disease. Together, a proof of principle is provided that the clinically safe Hsp90 inhibitor Silibinin potentially restores glucocorticoid sensitivity in corticotroph adenomas in vitro and in vivo, and that it might be used to treat Cushing’s patients in the future.