In this work we report a new mechanism for ageing, where passive aggregation and stochastic segregation of protein aggregates can switch cells from a non-ageing to an ageing state. This switch is activated by the increase in the total amount of protein aggregates. We established a damage reporter system by labeling Hsp104, a chaperone which binds protein aggregates, with GFP. By observing that the accumulation of Hsp104 labeled aggregates correlated with the majority of cell death in the population, and that cells which are born with a high level of aggregates are more likely to die, we validated protein aggregation as being an ageing factor in S. pombe. To identify the mechanism of damage segregation, we monitored nucleation, fusion and partition of aggregates at division. We established that aggregates are present in the cytoplasm fraction which is not occupied by vacuoles and lipid vesicles, and that they are not actively transported by the cytoskeleton. Protein aggregates were not distributed in a biased manner at division. Their position in the cytoplasm dictates to which cell they will be partitioned at division, as confirmed by studies using an asymmetrically dividing mutant, pom1Δ, in which the larger cells inherited more aggregates. This shows that aggregate segregation in S.pombe is a stochastic process. Stochastic distribution contributes to a constant dilution of damage and the maintenance of a non-ageing division, where the total levels of damage in the individuals are on average maintained constant. Together with Steven Lade and Thilo Gross, from the MPI-PKS, we designed a model in which passive aggregation and stochastic segregation reproduced our experimental results. Surprisingly, when cells are exposed to heat stress and the total levels of protein aggregates increase, aggregates are unequally segregated, i.e., ageing is turned on. The switch in segregation results from increased fusion due to a higher number of aggregates, which generates large single aggregates, which are retained by one of the cells at division. Our model reproduced the heat stress condition, showing that fusion is an essential parameter to generate clean cells quickly after the levels of damage increase. Fission yeast cells can therefore switch between non-ageing and ageing like division depending on the total amount of damage at birth. To clarify if other cellular components could be ageing factors in S. pombe, we tested if the inheritance of the old cell wall at the cell pole was associated with an increase in division time, similarly to what occurs in E. coli. We also tested if the inheritance of the new centrosome-analog, the SPB, which is segregated to the ageing mother cell in S. cerevisiae, resulted in an increase in division time. We did not find evidence for ageing associated with these structures. Finally we determined that the feature of slow division was not a transmissible trait, i.e., daughters of slow diving cells divided faster than their mothers. Another ageing hallmark, the cumulative increase in division time with the total number of divisions before death, was not present in S.pombe. Our combined results from damage segregation and pedigree analysis show that stochastic segregation of damage is a viable strategy to avoid ageing. Passive aggregation in the presence of a high number of aggregates can switch on ageing, representing an alternative to active segregation mechanisms and to the existence of pre-defined ageing lineages, as shown for other organisms. Finally, our results show that ageing is not ubiquous to life, and that it can be a facultative strategy to cope with stress.

In the presented study, dimorphic yeasts were applied as model organisms to study the growth of fungal mycelia. When environmental conditions are chosen appropriately, yeast colonies are built up of well separated individual cells. Thus, in contrast to fungal mycelia the translocation of nutrients and information within the colony can be neglected. The study focuses on the question of how the growth behaviour of a population of single cells is regulated, and which differences can be expected when nutrient translocation actually occurs. To answer this question, at first, an effective method for the highly resolved estimation of biomass distributions inside the colonies was developed. This method facilitates a dynamic non-invasive monitoring of colony development. Furthermore, mathematical models were established which describe the development of the colonies based on the behaviour of discrete individual cells. Growth simulations allow a quantitative prediction, and, thereby, an in silico testing of hypothetic regulatory mechanisms. The growth behaviour of yeast colonies was investigated applying the model organisms Candida boidinii and Yarrowia lipolytica. The yeasts were cultivated on solid agar substrates at various degrees of carbon and nitrogen limitation, respectively. The highest gain of understanding was achieved for the growth of both yeasts on glucose as the limiting carbon source: Investigations showed that mycelial yeast colonies adapt to declining nutrient concentrations by decreasing the cell density in their mycelium while the growth rate of the colony diameter remains constant. Under glucose limitation, the yeast C. boidinii grows diffusion-limited, i.e., the growth of the population is controlled by the amount of nutrient that diffuses towards the colony. The cessation of growth coincides with the depletion of the primary nutrient source glucose from the growth substrate. In contrast to these findings, it was shown that Y. lipolytica colonies continue to extend even after the complete consumption of glucose. In the absence of the primary nutrient source, the yeast assimilates biomass from the inner colony regions to facilitate the growth of the population. The suggested mechanism of coupled extension and decay processes was verified by a number of experiments. However, the mechanism which facilitates the transport of decay products to the growing colony boundary, i.e., the actual nature of the decay process, remains unclear. Mathematical simulations show that a continuous colony extension on the decay products of dying cells cannot be explained by the assumption that colonies are built up of uncoordinatedly growing single cells. Therefore, a hypothesis for the growth of Y. lipolytica colonies was derived which suggests that these populations are built up of tube-like hyphal cells. Accordingly, the measured drop of biomass density in the inner colony areas is the consequence of a cytoplasm transport towards the growing edge of the mycelium where it is assimilated as a secondary nutrient resource in the absence of glucose. It has to be emphasized that this hypothesis also provides a mechanistic explanation for the vacuolisation of hyphae in mycelia of higher fungi. / In der vorgestellten Arbeit wurden dimorphe Hefen als Modellorganismen für die Untersuchung des Wachstums von Pilzmyzelien eingesetzt. Bei geeigneter Wahl der Umgebungsbedingungen sind Hefekolonien aus Einzelzellen aufgebaut, wodurch im Gegensatz zu Myzelien höherer Pilze der Transport von Nährstoffen und Informationen innerhalb der Kolonie vernachlässigt werden kann. Im Mittelpunkt der Untersuchungen stand die Frage, wie das Wachstumsverhalten einer Population individueller Zellen reguliert ist, bzw. welche Unterschiede sich ergeben, wenn ein Nährstofftransport tatsächlich stattfindet. Um diese Fragestellungen bearbeiten zu können, wurde zunächst eine effektive Methode zur hoch ortsaufgelösten Bestimmung der Biomasseverteilung innerhalb der Kolonien entwickelt. Diese Methode ermöglicht ein dynamisches nichtinvasives Monitoring der Entwicklung einer Kolonie. Weiterhin wurden mathematische Modelle entwickelt, die das Wachstumsverhalteeiner Population auf der Grundlage des Verhaltens von diskreten Einzelzellen beschreibt. Die Wachstumssimulationen erlauben quantitative Vorhersagen und damit ein in silico Testen der Auswirkungen von hypothetischen Regulationsmechanismen. Das Wachstumsverhalten von Hefekolonien wurde anhand der Modellorganismen Candida boidinii und Yarrowia lipolytica untersucht. Die Hefen wurden auf festen Agar-Nährböden bei verschieden starker Kohlenstoff- und Stickstofflimitation kultviert. Der größte Erkenntnisgewinn wurde dabei für das Wachstum beider Hefen auf Glukose als limitierender Kohlenstoffquelle erzielt: Die Untersuchungen ergaben, dass myzelartig wachsende Hefekolonien bei sinkenden Nährstoffkonzentrationen eine geringere Zelldichte aber einen konstante Wachstumsgeschwindigkeit des Koloniedurchmessers aufweisen. Die Hefe C. boidinii wächst unter Glukoselimitation diffusionslimitiert, d.h. das Wachstum der Population wird durch die Menge der zur Kolonie diffundierenden Nährstoffe bestimmt. Der Abbruch des Koloniewachstums fällt mit dem Verbrauch der primären Nähstoffquelle Glukose zusammen. Im Gegensatz dazu konnte für das Wachstum von Y. lipolytica gezeigt werden, dass sich die Kolonien auch nach dem vollständigen Verbrauch von Glukose weiter ausdehnen. Im Abwesenheit der primären Nährstoffquelle nutzt die Hefe Zerfallsprodukte eigener Zellmasse aus dem Inneren der Kolonie als Nährstoff, um das weitere Wachstum der Population zu gewährleisten. Während der vorgeschlagene gekoppelte Ausdehnungs- und Zerfallprozess durch eine Reihe von Versuchen experimentell abgesichert wurde, bleibt der Mechanismus des Transports der Zerfallsprodukte zum Kolonierand, bzw. die eigentliche Natur des Zerfallsprozesses unklar. Simulationsrechnungen ergaben, dass eine kontinuierliche Ausdehnung der Kolonie auf Zellzerfallsprodukten sterbender Zellen nicht durch die Annahme erklärt werden kann, dass die Kolonien aus unkoordiniert wachsenden Einzelzellen aufgebaut sind. Aus diesem Grunde wurde für das Wachstum von Y. lipolytica die Hypothese abgeleitet, dass das Myzelium dieser Hefe aus schlauchartigen Hyphenzellen aufgebaut ist. Der gemessene Abfall der Biomassekonzentration im Kolonieinneren ist demnach die Konsequenz des Transports von Zytoplasma hin zum wachsenden Kolonierand, wo es in Abwesenheit von Glukose als sekundäre interne Nährstoffquelle assimiliert wird. Es ist zu beachten, dass diese Hypothese auch eine mechanistische Erklärung für die Ursachen der Vakuolisierung in Myzelien höherer filamentöser Pilze gibt.

The thesis is aimed at understanding of molecular mechanisms of autophagic degradation of peroxisomes (pexophagy) in the yeast Yarrowia lipolytica. This microorganism has been extensively used to explore peroxisome biogenesis (Titorenko and Rachubinski, 2000). Gunkel et al. (1999) intoduced Y. lypolitica into pexophagy studies. However, the field of pexophagic research on this yeast remains quite unexplored. This work involved following tasks: (1) the development and optimization of Y. lipolytica as a model system to study peroxisome degradation; (2) Y. lipolytica genes and proteins implicated in pexophagy should be found and characterized; (3) a proper easy-to-handle selection procedure to isolate novel peroxisome degradation-deficient(pdd) mutants of Y. lipolytica should be devised.

Over the past few years a large amount of work has provided growing insight into the molecular mechanisms that direct post-Golgi trafficking events in the budding yeast Saccharomyces cerevisae. However, a key event in this process, the formation of secretory vesicles at the Golgi and sorting of cargo into these transport carriers, remains poorly understood. It has been demonstrated that phosphatidylinositol 4-phosphate (PI(4)P) generated by the PI(4)-kinase Pik1p plays an essential role in maintenance of Golgi secretory function and morphology. Up to now relatively few targets of Pik1/PI(4)P signaling at the Golgi have been identified and it thus remains elusive how Pik1p mediates its essential function in Golgi secretion. During my thesis work, I used synthetic genetic array analysis (SGA) of a temperature-sensitive mutant allele of PIK1 (pik1-101) in order to gain better understanding of Pik1p function at the TGN and to isolate new regulators of post-Golgi transport in yeast. I identified a total of 85 genes, whose deletion resulted in a synthetic growth defect when combined with the pik1-101 mutation. 21 isolated deletion mutants were used for further analysis, several of which were found to share common trafficking phenotypes with the pik mutant. A striking result of the screen was the finding that Pik1p interacts genetically with several components of a potential post-translational modification pathway referred to as “urmylation pathway”. In addition, a novel, previously uncharacterized subunit of the Transport protein particle (TRAPP) complex was isolated as genetic interactor of Pik1p, suggesting a function for the TRAPP complex in a Pik1p dependent trafficking pathway. Using tandem affinity purification, I could also demonstrate that TRAPP shows previously unknown interactions with other regulators of post-Golgi transport. The second part of this thesis describes the development of a new visual screening approach. Recent work indicates that secretory cargo in yeast can be transported to the cell surface via at least two different exocytic branches. Upon block of one pathway cargo can be partially redistributed into the other pathway. This partial redundancy of exocytic pathways provides one explanation why genetic screens in the past were largely unsuccessful in identifying the molecular machinery that directs vesicle budding and cargo sorting at the TGN. I collaborated in the development of a novel screening method that was devised to circumvent this problem. The method took advantage of the systematic yeast knockout array and was based on the assumption that a defect in cargo sorting and cell surface transport could be detected as intracellular accumulation of a GFP-tagged model cargo. The suitability of our approach for identifying regulators of secretory transport has been demonstrated in a small-scale pilot study that will be presented in this thesis. The screening method proofed to be applicable on a genome-wide scale and can now be used for the screening of additional markers. This novel approach provides an entry point to the comprehensive study of TGN sorting.