Open Access

Iron uptake is an essential process in all Gram-negative bacteria including cyanobacteria and therefore different transport systems evolved during evolution. In cyanobacteria, however, the iron demand is higher than in proteobacteria due to the function of iron as cofactor in e.g. photosynthesis and nitrogen fixation. Most of the transport systems depend on outer membrane localized TonB-dependent transporters (TBDTs), a periplasma-facing TonB protein and a plasma membrane localized machinery (ExbBD). So far, iron chelators (siderophores), oligosaccharides and polypeptides have been identified as substrates of TBDTs. However, in proteobacteria TonB-dependent outer membrane transporter represent a well-explored subject whereas for cyanobacteria almost nothing is known about possible TonB-dependent uptake systems for iron or other substrates. The heterocyst-forming filamentous cyanobacterium Anabaena sp. PCC 7120 is known to secrete the siderophore schizokinen, but its transport system has remained unidentified. For Anabaena sp. PCC 7120 22 genes were identified as putative TBDTs covering almost all known TBDT subclasses. This is a high number of TBDTs compared to other cyanobacteria. The expression of the 22 putative TBDTs individually depends on the presence of iron, copper or nitrogen. The atypical dependence of TBDT gene expression on different nutrition points to a yet unknown regulatory mechanism. In addition, the hypothesis of the absence of TonB in Anabaena sp. PCC 7120 was clarified by the identification of an according sequence, all5036. Inspection of the genome of Anabaena sp. PCC 7120 shows that only one gene encoding a putative TonB-dependent iron transporter, namely alr0397, is positioned close to genes encoding enzymes involved in the biosynthesis of a hydroxamate siderophore. The expression of alr0397 was elevated under iron-limited conditions. Inactivation of this gene caused a moderate phenotype of iron starvation in the mutant cells. The characterization of the mutant strain showed that Alr0397 is a TonB-dependent schizokinen transporter (SchT) of the outer membrane and that alr0397 expression and schizokinen production are regulated by the iron homeostasis of the cell. Additional two genes of Anabaena sp. PCC 7120 involved in this process were identified. SchE encoded by all4025 is a putative cytoplasmic membrane-localized transporter involved in TolC-dependent siderophore secretion. The mutation of schE resulted in an enhanced sensitivity to high metal concentrations and in drastically reduction of secretion of hydroxamate-type siderophores. IacT coded by all4026 is a predicted outer membrane-localized TonB-dependent iron transporter. Inactivation of iacT resulted in reduced sensitivity to elevated iron and copper levels, whereas decoupling the expression from putative regulation by exchange of the promoter resulted in sensitization against tested metals. Further analysis showed that iron and copper effects are synergistic because decrease of iron induced a significant decrease of copper levels in the iacT insertion mutant but an increase of those levels in Anabaena sp. PCC 7120 where expression of all4026 is under the trc-promoter. In consequence, the results unravel a link between iron and copper homeostasis.

The biogenesis and function of photosynthetically active chloroplasts relies on the import of thousands of nuclear encoded proteins via the coordinated actions of two multiprotein translocon machineries in the outer and inner envelope membrane. Trafficking of preproteins across the soluble compartment of InterMembrane Space (IMS) is currently envisioned to be facilitated by an IMS complex composed of outer envelope proteins Toc64 and Toc12, a soluble IMS component, Tic22 and an IMS-localized Hsp70. Among them, currently Tic22 is the only component that stands undisputed in terms of its existence. Having two closely related homologs in A. thaliana, their biochemical and functional characterization was still lacking. A critical analysis of Tic22 knockout mutants displayed growth phenotype reminiscent of ppi1, the mutant of Toc33. However, both the genes have similar expression patterns with no clear preference for photosynthetic or nonphotosynthetic tissues, which explained the absence of a detectable phenotype in single mutants. In addition, transgenic complementation study with either of the homolog affirmed the identical localization of both proteins in the IMS which characterizes the two homologs as functionally redundant. Based on the pale-yellow phenotype exhibited by the double mutant plants, an attempt to analyze the import capacity of a stromal substrate in the double mutant revealed threefold reduction when compared to wild-type acknowledging the essential role of Tic22 in the import mechanism. Initially, Tic22 was identified together with another protein, Tic20, which has been heavily discussed as a protein conducting channel in the inner membrane. Despite being characterized, in A. thaliana, two out of four homologs of Tic20 are differentially localized with one being additionally localized in mitochondria and the other, exclusively residing in the thylakoids. According to in silico analysis, for all the Tic20 proteins, a four-helix transmembrane topology was predicted. Accordingly, its topology was mapped by employing the recently established selfassembling GFP-based in vivo experiments. Astonishingly, the expression of one of the inner envelope localized Tic20 homolog enforces inner membrane proliferation affecting the shape and organization of the membrane. Therefore this study focuses on analyzing the effects of high envelope protein concentrations on membrane structures, which together with the existing results, an imbalance in the lipid to protein ratio and a possible role of signaling pathway regulating membrane biogenesis is discussed.

The translocation of nuclear-encoded precursor proteins into chloroplasts is a highly ordered process involving the action of several components to regulate this molecular ensemble. Not only GTP hydrolysis and GDP release but also the phosphorylation of TOC GTPases is a widely discussed mechanism to regulate protein import. The receptor component (Toc34) and its isoform of A. thaliana (atToc33) were found to be regulated by phosphorylation. Although the phosphorylation of Toc33 is already known for several years, several questions regarding the molecular components involved in the regulation of the phosphorylation process, precisely what is the protein kinase and where this kinase is initially localized, so far remained unclear. This thesis aimed at the defining of the phosphorylation status of TOC GTPases in monomeric and/or dimeric states, the identification of the nature of Toc33-PK (protein kinase), and in the same context it aimed at gaining first insights into the physiological significance of Toc33 phosphorylation. To this end, (I) An in vitro and in vivo system for investigating of TOC GTPases Phosphorylation (in monomeric or dimeric state) was developed. Since no information is available about the phosphorylation status of the Toc159 isoforms, the second receptor of the TOC complex, it was interesting to investigate whether these isoforms undergo phosphorylation or not. The results indicated that atToc159 isoforms are able to be phosphorylated by the kinase activity in purified outer envelope membranes (OEMs) of pea, but not atToc132. Moreover, an artificial dimer of psToc34 based on the interaction of a C-terminally fused leucine zipper was not phosphorylated. This result reflected the inability of the OEM kinase to phosphorylate the dimers of TOC GTPases. Also, In vivo labeling of atToc33 was developed and occurred in a dose-dependent manner. Therefore, this results evidenced that in vitro phosphorylation of atToc33 (both endogenous wild type and recombinant expressed proteins) is not artificial labeling but represents a physiological relevance. CD (circular dichroism) measurements revealed that recombinant GTPase domain of atToc33 is preferentially phosphorylated in its folded state. Therefore, it could be suggested that folding of atToc33rec is a prerequisite for its phosphorylation and the phosphorylation event occurs as a posttranslational modification most likely after insertion of Toc33 (Toc34) into the OE of chloroplasts. Secondly, (II) Isolation and identification of Toc33-PK from OEMs of chloroplasts was performed. Four independent strategies were developed to identify the Toc33-protein kinase: UV-induced and chemically-based crosslinking, different applied chromatographic techniques, identification of PK-Toc33 interaction by means of HDN-PAGE (histidine- and deoxycholate-based native PAGE), and finally mass spectrometric approaches were performed on fractions including the potential kinase activity. UV-induced crosslinking procedure was developed and resulted in covalent bonding of nine proteins to [a-32P] ATP, while chemically-based one was not significant. The applied chromatographic and HDN-PAGE approaches, including mass spectrometry, have revealed the identification of 13 protein kinases. Of these identified kinases, phototropin2 (Phot2, AT5G58140), leucine-rich repeat PK (LRR-PK, AT4G28650.1), and receptor-like transmembrane PK (RLK, AT5G56040.2) were selected as the most promising candidates (ca. kinase type and one transmembrane helix for membrane localization). (III) The physiological significance of Toc33 phosphoryation was shown to link this process with the environmental changes (especially, the light conditions). Identification of chloroplast OE-located PKs performed by nLC-MALDI-MS/MS resulted in the detection of Phot2. Furthermore, the subcellular localization of Phot2 in OEM of chloroplasts was confirmed by immunoblotting experiments using a-Phot2 antibody. The kinase activity of Phot2 towards TOC GTPases was characterized and revealed that fused GST-KD (kinase domain) protein able to specifically phosphorylate atToc33rec, but not atToc159rec. Also, endogenous atPhot2 was upregulated and heavily detected in the ppi1-S181A plant line (where serine to alanine exchange was performed to abolish the phosphorylation of atToc33). Hence, we suggested that certain signal cascades may directly or indirectly link Toc33 receptor phosphorylation, protein levels of Phot2 (as promising PK candidate), and irradiation conditions (as an inducing signal of the subsequent phosphorylation events). Light-dependent phosphorylation of Toc33 was shown either after de-etiolation conditions or after high light intensities of blue light was performed. Therefore, phosphorylation of Toc33 might be identified as an external regulatory signal to regulate preproteins import into chloroplasts in response to environmental conditions (e.g. light changes) or as a signal of chloroplast biogenesis.

Heat stress transcription factors (Hsfs) have an essential role in heat stress response (HSR) and thermotolerance by controlling the expression of hundreds of genes including heat shock proteins (Hsps) with molecular chaperone functions. Hsf family in plants shows a striking multiplicity, with more than 20 members in many species. In Solanum lycopersicum HsfA1a was reported to act as the master regulator of the onset of HSR and therefore is essential for basal thermotolerance. Evidence for this was provided by the analysis of HsfA1a co-suppression (A1CS) transgenic plants, which exhibited hypersensitivity upon exposure to heat stress (HS) due to the inability of the plants to induce the expression of many HS-genes including HsfA2, HsfB1 and several Hsps. Completion of tomato genome sequencing allowed the completion of the Hsf inventory, which is consisted of 27 members, including another three HsfA1 genes, namely HsfA1b, HsfA1c and HsfA1e. Consequently, the suppression effect of the short interference RNA in A1CS lin e was re-evaluated for all HsfA1 genes. We found that expression of all HsfA1 proteins was suppressed in A1CS protoplasts. This result suggested that the model of single master regulator needs to be re-examined. Expression analysis revealed that HsfA1a is constitutively expressed in different tissues and in response to HS, while HsfA1c and HsfA1e are minimally expressed in general, and show an induction during fruit ripening and a weak upregulation in late HSR. Instead HsfA1b shows preferential expression in specific tissues and is strongly and rapidly induced in response to HS. At the protein level HsfA1b and HsfA1e are rapidly degraded while HsfA1a and HsfA1c show a higher stability. In addition, HsfA1a and HsfA1c show a nucleocytosolic distribution, while HsfA1b and HsfA1e a strong nuclear retention. A major property of a master regulator in HSR is thought to be its ability to cause a strong transactivation of a wide range of genes required for the initial activation of protective mechanisms. GUS reporter assays as well as analysis of transcript levels of several endogenous transcripts in protoplasts transiently expressing HsfA1 proteins revealed that HsfA1a can stimulate the transcription of many genes, while the other Hsfs have weaker activity and only on limited set of target genes. The low activity of HsfA1c and HsfA1e can be attributed to the lower DNA capacity of the two factors as judged by a GUS reporter repressor assay. HsfA1a has been shown to have synergistic activity with the stress induced HsfA2 and HsfB1. The formation of such complexes is considered as important for stimulation of transcription and long term stress adaptation. All HsfA1 members show synergistic activity with HsfA2, while only HsfA1a act as co-activator of HsfB1 and HsfA7. Interestingly, HsfA1b shows an exceptional synergistic activity with HsfA3, suggesting that different Hsf complexes might regulate different HS-related gene networks. Altogether these results suggest that HsfA1a has unique characteristics within HsfA1 subfamily. This result is interesting considering the very high sequencing similarity among HsfA1s, and particularly among HsfA1a and HsfA1c. To understand the molecular basis of this discrepancy, a series of domain swapping mutants between HsfA1a and HsfA1c were generated. Oligomerization domain and C-terminal swaps did not affect the basal activity or co-activity of the proteins. Remarkably, an HsfA1a mutant harbouring the N-terminus of HsfA1c shows reduced activity and co-activity, while the reciprocal HsfA1c with the N-terminus of HsfA1a cause a gain of activity and enhanced DNA binding capacity. Sequence analysis of the DBD of HsfA1 proteins revealed a divergence in the highly conserved C-terminus of the turn of β3-β4 sheet. As the vast majority of HsfA1 proteins, HsfA1a at this position comprises an Arg residue (R107), while HsfA1c a Leu and HsfA1e a Cys. An HsfA1a-R107L mutant has reduced DNA binding capacity and consequently activity. Therefore, the results presented here point to the essential function of this amino acid residue for DNA binding function. Interestingly, the mutation did not affect the activity of the protein on Hsp70-1, suggesting that the functionality of the DBD and consequently the transcription factor on different promoters with variable heat stress element number and architecture is dependent on structural peculiarities of the DBD. In conclusion, the unique properties including expression pattern, transcriptional activities, stability, DBD-peculiarities are likely responsible for the dominant function of HsfA1a as a master regulator of HSR in tomato. Instead, other HsfA1-members are only participating in HSR or developmental regulations by regulating a specific set of genes. Furthermore, HsfA1b and HsfA1e are likely function as stress primers in specific tissues while HsfA1c as a co-regulator in mild HSR. Thereby, tomato subclass A1 presents another example of function diversity not only within the Hsf family but also within the Hsf-subfamily of closely related members. The diversification based on DBD peculiarities is likely to occur in potato as well. Therefore this might have eliminated the functional redundancy observed in other species such as Arabidopsis thaliana but has probably allowed the more refined regulation of Hsf networks possibly under different stress regimes, tissues and cell types.

Die letzten Jahrzehnte brachten einen enormen Zuwachs des Wissens und Verständnisses über die molekularen Prozesse des Lebens.Möglich wurde dieser Zuwachs durch die Entwicklung diverser Methoden, mit denen beispielsweise gezielt die Konzentration einzelner Stoffe gemessen werden kann oder gar alle anwesenden Metaboliten eines biologischen Systems erfasst werden können. Die großflächige Anwendung dieser Methoden führte zur Ansammlung vieler unterschiedlicher -om-Daten, wie zum Beispiel Metabolom-, Proteom- oder Transkriptoms-Datensätzen. Die Systembiologie greift auf solche Daten zurück, um mathematische Modelle biologischer Systeme zu erstellen, und ermöglicht so ein Studium biologischer Systeme auch außerhalb des Labors. Für größere biologische Systeme stehen jedoch meistens nicht alle Informationen über Stoffkonzentrationen oder Reaktionsgeschwindigkeiten zur Verfügung, um eine quantitative Modellierung, also die Beschreibung von Änderungsraten kontinuierlicher Variablen, durchführen zu können. In einem solchen Fall wird auf Methoden der qualitativen Modellierung zurückgegriffen. Eine dieser Methoden sind die Petrinetze (PN), welche in den 1960er Jahren von Carl Adam Petri entwickelt wurden, um nebenläufige Prozesse im technischen Umfeld zu beschreiben. Seit Anfang der 1990er Jahre finden PN auch Anwendung in der Systembiologie, um zum Beispiel metabolische Systeme oder Signaltransduktionswege zu modellieren. Einer der Vorteile dieser Methode ist zudem, dass Modelle als qualitative Beschreibung des Systems begonnen werden können und im Laufe der Zeit um quantitative Beschreibungen ergänzt werden können. Zur Modellierung und Analyse von PN existieren bereits viele Anwendungen. Da das Konzept der PN jedoch ursprünglich nicht für die Systembiologie entwickelt wurde und meist im technischen Bereich verwendet wird, existierten kaum Anwendungen, die für den Einsatz in der Systembiologie entwickelt wurden. Daher ist auch die Durchführung der für die Systembiologie entwickelten Analysemethoden für PN nicht mit diesen Anwendungen möglich. Die Motivation des ersten Teiles dieser Arbeit war daher, eine Anwendung zu schaffen, die speziell für die PN-Modellierung und Analyse in der Systembiologie gedacht ist, also in ihren Analysemethoden und ihrer Terminologie sich an den Bedürfnissen der Systembiologie orientiert. Zudem sollte die Anwendung den Anwender bei der Auswertung der Resultate der Analysemethoden visuell unterstützen, indem diese direkt visuell im Kontext des PN gesetzt werden. Da bei komplexeren PN die Resultate der Analysemethoden in ihrer Zahl drastisch anwachsen, wird eine solche Auswertung dieser notwendig. Aus dieser Motivation heraus entstand die Anwendung MonaLisa, dessen Implementierung und Funktionen im ersten Teil der vorliegenden Arbeit beschrieben werden. Neben den klassischen Analysemethoden für PN, wie den Transitions- und Platz-Invarianten, mit denen grundlegende funktionale Module innerhalb eines PN gefunden werden können, wurden weitere, meist durch die Systembiologie entwickelte, Analysemethoden implementiert. Dazu zählen zum Beispiel die Minimal Cut Sets, die Maximal Common Transitions Sets oder Knock-out-Analysen. Mit MonaLisa ist aber auch die Simulation des dynamischen Verhaltens des modellierten biologischen Systems möglich. Hierzu stehen sowohl deterministische als auch stochastische Verfahren, beispielsweise der Algorithmus von Gillespie zur Simulation chemischer Systeme, zur Verfügung. Für alle zur Verfügung gestellten Analysemethoden wird ebenfalls eine visuelle Repräsentation ihrer Resultate bereitgestellt. Im Falle der Invarianten werden deren Elemente beispielsweise in der Visualisierung des PN eingefärbt. Die Resultate der Simulationen oder der topologischen Analyse können durch verschiedene Graphen ausgewertet werden. Um eine Schnittstelle zu anderen Anwendungen zu schaffen, wurde für MonaLisa eine Unterstützung einiger gängiger Dateiformate der Systembiologie geschaffen, so z.B. für SBML und KGML. Der zweite Teil der Arbeit beschäftigt sich mit der topologischen Analyse eines Datensatzes von 2641 Gesamtgenom Modellen aus der path2models-Datenbank. Diese Modelle wurden automatisiert aus dem vorhandenen Wissen der KEGG- und der MetaCyc-Datenbank erstellt. Die Analyse der topologischen Eigenschaften eines Graphen ermöglicht es, grundlegende Aussagen über die globalen Eigenschaften des modellierten Systems und dessen Entstehungsprozesses zu treffen. Daher ist eine solche Analyse oft der erste Schritt für das Verständnis eines komplexen biologischen Systems. Für die Analyse der Knotengrade aller Reaktionen und Metaboliten dieser Modelle wurden sie in einem ersten Schritt in PN transformiert. Die topologischen Eigenschaften von metabolischen Systemen werden in der Literatur schon sehr gut beschrieben, wobei die Untersuchungen meist auf einem Netzwerk der Metaboliten oder der Reaktionen basieren. Durch die Verwendung von PN wird es möglich, die topologischen Eigenschaften von Metaboliten und Reaktionen in einem gemeinsamen Netzwerk zu untersuchen. Die Motivation hinter diesen Untersuchungen war, zu überprüfen, ob die schon beschriebenen Eigenschaften auch für eine Darstellung als PN zutreffen und welche neuen Eigenschaften gefunden werden können. Untersucht wurden der Knotengrad und der Clusterkoeffizient der Modelle. Es wird gezeigt, dass einige wenige Metaboliten mit sehr hohem Knotengrad für eine ganze Reihe von Effekten verantwortlich sind, wie beispielsweise dass die Verteilung des Knotengrades und des Clusterkoeffizienten, im Bezug auf Metaboliten, skalenfrei sind und dass sie für die Vernetzung der Nachbarschaft von Reaktionen verantwortlich sind. Weiter wird gezeigt, dass die Größe eines Modelles Einfluss auf dessen topologische Eigenschaften hat. So steigt die Vernetzung der Nachbarschaft eines Metaboliten, je mehr Metaboliten in einem biologischen System vorhanden sind, gleiches gilt für den durchschnittlichen Knotengrad der Metaboliten.