Detailed analyses of mutations affecting structural formation of neuromuscular synapses in Drosophila melanogaster

The establishment of appropriate synapses between neurons and their target cells is an essential requirement for the formation of functional neuronal circuits. However, there is very little insight into the mechanisms underlying de novo formation of synapses and synaptic terminals. To identify novel genes involved in signalling or structural aspects of these processes I capitalised on possibilities provided by the model organism Drosophila. Thus, I contributed to a screen of a collection of third chromosomal mutations (Salzberg et al., 1997, Genetics 147, 1723ff.) selecting those mutant strains displaying structural defects of Drosophila neuromuscular junctions (NMJ). Carrying out genetic mapping experiments, I could assign 7 genes to interesting candidate mutations. All 7 mutations selected in this process cause size alterations of the embryonic NMJ, and one shows additional disturbances in the distribution of synaptic markers. 4 of these turned out to be transcription factors, not falling into the remit of this project. Only for one of these, the neuronal transcription factor Castor, I could show that its overgrown mutant NMJ phenotype is due to an increase in the number of motorneurons. The remaining genes encode a potential nitrophenylphosphatase, the translation initiation factor eIF4AIII, and a novel protein Waharan. Unfortunately, the nitophenylphosphatase gene was identified too late to carry out functional studies in the context of this project, but potential roles are discussed. eIF4AIII promotes NMJ size tempting to speculate that local translation at the NMJ is affected. I found that the synaptic scaffolding molecule Discs large (Dlg; orthologue of PSD95) is upregulated at eIF4AIII mutant NMJs. Targeted upregulation of Dlg can not mimic the eIF4AIII mutant phenotype, but dlg mutations suppress it. Therefore, Dlg function is required but not sufficient in this context. My findings are discussed in detail, pointing out future directions. The main focus of this work is the completely novel gene waharan (wah), an orthologue of the human gene KIAA1267 encoding a big brain protein of likewise unknown structure and function. My studies show that mutations or RNAi knock-down of wah cause NMJ overgrowth and reveal additional crucial roles in the patterning of wing imginal discs. RNAi studies suggest Wah to be required pre- and postsynaptically at NMJs and, consistently, wah is transcribed in the nervous system and muscles. Anti-Wah antisera were produced but could no longer be tested here, but preliminary studies with newly generated HA-targeted constructs suggest that Wah localises at NMJs and in neuronal nuclei. In silico analyses predict Wah to be structurally related to the Rad23-family of proteins, likely to target ubiquitinated proteins to the proteasome for degradation (Chen et al., 2002, Mol Cell Biol 22, 4902ff.) . In agreement with this prediction, poly-ubiquitinated proteins were found to accumulate in the absence of wah function, and wah-like mutant phenotypes were induced in NMJs and wing discs by knocking down proteasome function. My analysis further revealed that poly-ubiquitinated proteins are reduced in nuclei of wah mutant neurons and muscles, suggesting that Wah may play additional roles in ubiquitin-mediated nuclear import. Taken together, this study has uncovered a number of interesting candidate genes required for the de novo formation of Drosophila NMJs. 3 of these genes fell into the focus of this project. As discussed in detail, discovery of these genes and insights gained into their function have high potential to be translatable into vertebrate systems.

Gene expression analysis in ‘Candidatus Phytoplasma mali’-resistant and -susceptible Malus genotypes

Apple proliferation (AP) disease is the most important graft-transmissible and vector-borne disease of apple in Europe. ‘Candidatus Phytoplasma mali’ (Ca. P. mali) is the causal agent of AP. Apple (Malus x domestica) and other Malus species are the only known woody hosts. In European apple orchards, the cultivars are mainly grafted on one rootstock, M. x domestica cv. M9. M9 like all other M. x domestica cultivars is susceptible to ‘Ca. P. mali’. Resistance to AP was found in the wild genotype Malus sieboldii (MS) and in MS-derived hybrids but they were characterised by poor agronomic value. The breeding of a new rootstock carrying the resistant and the agronomic traits was the major aim of a project of which this work is a part. The objective was to shed light into the unknown resistance mechanism. The plant-phytoplasma interaction was studied by analysing differences between the ‘Ca. P. mali’-resistant and -susceptible genotypes related to constitutively expressed genes or to induced genes during infection. The cDNA-Amplified Fragment Length Polymorphism (cDNA-AFLP) technique was employed in both approaches. Differences related to constitutively expressed genes were identified between two ‘Ca. P. mali’-resistant hybrid genotypes (4551 and H0909) and the ‘Ca. P. mali’-susceptible M9. 232 cDNA-AFLP bands present in the two resistant genotypes but absent in the susceptible one were isolated but several different products associated to each band were found. Therefore, two different macroarray hybridisation experiments were performed with the cDNA-AFLP fragments yielding 40 sequences encoding for genes of unknown function or a wide array of functions including plant defence. In the second approach, individuation and analysis of the induced genes was carried out exploiting an in vitro system in which healthy and ‘Ca. P. mali’-infected micropropagated plants were maintained under controlled conditions. Infection trials using in vitro grafting of ‘Ca. P. mali’ showed that the resistance phenotype could be reproduced in this system. In addition, ex vitro plants were generated as an independent control of the genes differentially expressed in the in vitro plants. The cDNA-AFLP analysis in in vitro plants yielded 63 bands characterised by over-expression in the infected state of both the H0909 and MS genotypes. The major part (37 %) of the associated sequences showed homology with products of unknown function. The other genes were involved in plant defence, energy transport/oxidative stress response, protein metabolism and cellular growth. Real-time qPCR analysis was employed to validate the differential expression of the genes individuated in the cDNA-AFLP analysis. Since no internal controls were available for the study of the gene expression in Malus, an analysis on housekeeping genes was performed. The most stably expressed genes were the elongation factor-1 α (EF1) and the eukaryotic translation initiation factor 4-A (eIF4A). Twelve out of 20 genes investigated through qPCR were significantly differentially expressed in at least one genotype either in in vitro plants or in ex vitro plants. Overall, about 20% of the genes confirmed their cDNA-AFLP expression pattern in M. sieboldii or H0909. On the contrary, 30 % of the genes showed down-regulation or were not differentially expressed. For the remaining 50 % of the genes a contrasting behaviour was observed. The qPCR data could be interpreted as follows: the phytoplasma infection unbalance photosynthetic activity and photorespiration down-regulating genes involved in photosynthesis and in the electron transfer chain. As result, and in contrast to M. x domestica genotypes, an up-regulation of genes of the general response against pathogens was found in MS. These genes involved the pathway of H2O2 and the production of secondary metabolites leading to the hypothesis that a response based on the accumulation of H2O2 in MS would be at the base of its resistance. This resembles a phenomenon known as “recovery” where the spontaneous remission of the symptoms is observed in old susceptible plants but occurring in a stochastic way while the resistance in MS is an inducible but stable feature. As additional product of this work three cDNA-AFLP-derived markers were developed which showed independent distribution among the seedlings of two breeding progenies and were associated to a genomic region characteristic of MS. These markers will contribute to the development of molecular markers for the resistance as well as to map the resistance on the Malus genome.

Induktion von Pro-Autophagie-Signalen durch einen extra- oder intrazellulären Alpha-Toxin-Angriff

Autophagie ist ein konservierter, kataboler Mechanismus in allen eukaryoten Zellen. Unter anderem wird ihm eine wichtige Rolle als zellautonomer Abwehrmechanismus gegen Mikroorganismen zugeschrieben; von manchen Infektionserregern wird er jedoch unterlaufen oder sogar genutzt. Der stärkste Auslöser der Autophagie ist ein Mangel an Nährstoffen, insbesondere Aminosäuren. Über die Deaktivierung der Kinase mTORC1 und die Phosphorylierung des eukaryoten Translationsinitiationsfaktors eIF2α hemmt die Nährstoffknappheit die Proteinbiosynthese und aktiviert gleichzeitig Autophagie. Wie Mikroorganismen, insbesondere Bakterien, Autophagie auslösen oder manipulieren, ist derzeit Gegenstand intensiver Forschung. Modifikationen an Mikroben oder Phagosomen und Adapterproteine, die diese Veränderungen und Komponenten des Autophagieapparates erkennen, scheinen jedenfalls bei der selektiven Erkennung durch die Autophagie-Maschinerie wichtig zu sein. rnIn der vorliegenden Dissertationsarbeit wird die Rolle des membranporenbildenden α-Toxins von Staphylococcus aureus für die Induktion von Autophagie beleuchtet. Zum einen erwies sich die Akkumulation von (EGFP)-LC3(II), einem Marker der Autophagosomen, um intrazelluläre S. aureus als abhängig von α-Toxin. Zweitens, genügt extrazellulär appliziertes α-Toxin um (EGFP)-LC3(II)-positive Endosomen zu induzieren. Während der Angriff aus dem extrazellulären Raum jedoch binnen kurzer Zeit eine fokale Kumulation von phosphoryliertem eIF2α an der Plasmamembran induziert, die an der Internalisierung des Toxins beteiligt ist, findet sich am phagosomalen Kompartiment keine Toxin-abhängige Anhäufung von p-eIF2α oder proximalen Autophagieregulatoren. Dies impliziert, dass Toxin-Angriff auf die Plasmamembran, nicht aber auf das Phagosom, zu einer Reaktion führt, wie sie bei massivem Nährstoffmangel zu beobachten ist. Obwohl keine α-Toxin-abhängige Kumulation von p-eIF2α bei einem Angriff aus dem Phagosom erfolgt, findet sich um α-Toxin-produzierende Bakterien eine massive Kumulation von LC3 und Adapterprotein p62/Sequestosome1. Dies deutet daraufhin, dass der Ort des Angriffs - Plasmamembran oder Phagosom – für den Autophagie-induzierenden Mechanismus wichtig sein könnte. Der unterschiedliche Effekt auf die zellulären Ionenkonzentrationen, den ein Angriff auf die Plasmamembran oder auf ein Phagosom auslösen würde, bietet hierfür eine mögliche Erklärung. Die Aktivierung der Autophagie über Adapterproteine könnte dann als back-up Mechanismus fungieren, der auch dann greift, wenn eine Invasion ohne Schädigung der Plasmamembran erfolgt. Ein cross-talk der beiden Induktionswege ist angesichts der Bedeutung von p62 für die selektive und die Hunger-assoziierte Autophagie gut möglich; sezerniertes Toxin könnte durch die Aktivierung der basalen Autophagie Adapter-basierte Mechanismen verstärken.