Van_vuure_w_2008_msc_thesis_mini_chemostat_for_bacillus_subtilus.

Transcript

OPTIMIZATION OF A MINI-SCALE CHEMOSTAT SYSTEM FOR THE ANALYSIS OF METABOLIC FLUX DISTRIBUTIONS IN STRESSED BACILLUS SUBTILIS - M.SC. RESEARCH PROJECT THESIS Wiebe van Vuure SUPERVISION ETH ZURICH Prof. Dr. Uwe Sauer Dr. Ir. Roelco Kleijn SUPERVISION TU DELFT Dr. Ir. Peter Verheijen Dr. Ir. Robbert Kleerebezem February – September 2008 Bacillus subtilis systems biology LST Research Project 2008 Abstract This thesis focuses on the metabolic aspects of cellular adaptation processes to oxidative and iron limiting stresses. These stress conditions are generally encountered by pathogenic species upon the invasion of host organisms. It has attracted the attention of researchers and resulted in the initiation of a European project on Bacillus subtilis systems biology. This project aims at elucidating the regulatory pathways of B. subtilis under oxidative and iron stress for the subsequent application to pathogens. The present study attempts to identify and quantify the adaptations in central carbon metabolism in response to imposed stresses. Therefore a recently developed system of mini-chemostats is used as a culturing system. This setup allows a quick and economic way of culturing under many different conditions. The main finding in oxidative stress adaptations is the re-routing of carbon through the pentose phosphate pathway in order to generate NADPH. NADPH is used by processes that relief the cells of oxidative stress. Next to the production of reduction equivalents, there was an increased amount of extracellular TCA cycle intermediates as fumarate and succinate. The fluxes through this pathway did not change. The iron limitation experiments yielded less clear results on the impact on metabolism. This might be due to the chosen method to deferrate the medium. Deferoxamine caused only minor iron stress and was not life threatening for the cells. That the cells were only slightly affected was shown by the increased production of siderophores, bacterial iron harvesting compounds. The cells were able to survive the situation by the production of siderophores. The results of this study aid the BaSysBio project in obtaining a better understanding of B. subtilis adaptation to a shift in environmental conditions. The generated data gives further insight into the adaptation processes and help model development by providing quantitative information on the stress responses. The results can be processed by BaSysBio and can eventually result in new strategies to combat pathogenic bacteria. M.Sc. Research Project i Bacillus subtilis systems biology LST Research Project 2008 Contents 1 Introduction........................................................................................... 4 1.1 Preface ........................................................................................... 4 1.2 Thesis project.................................................................................. 4 1.2.1 Bacillus Systems Biology Project ................................................. 4 1.2.2 Thesis project objectives............................................................ 6 1.3 Bacillus subtilis ................................................................................ 7 1.3.1 Metabolism ............................................................................... 9 1.3.2 Pathogenic properties.............................................................. 10 1.3.3 General stress response........................................................... 11 1.3.4 Oxidative stress ...................................................................... 13 1.3.5 Iron limitation ......................................................................... 17 1.4 Adaptation under stress conditions.................................................. 19 1.4.1 Oxidative stress ...................................................................... 20 1.4.2 Iron limitation ......................................................................... 21 1.4.3 Connection between iron and oxidative stress responses ............ 22 1.4.4 Carbon metabolism under stress .............................................. 23 2 Materials and methods.......................................................................... 27 2.1 Materials ....................................................................................... 27 2.1.1 Strains ................................................................................... 27 2.1.2 Growth Media ......................................................................... 27 Plate media ............................................................................ 29 2.1.3 2.1.4 Solutions ................................................................................ 29 2.2 Cultivations ................................................................................... 31 2.2.1 Fermentations......................................................................... 31 2.2.2 Precultures ............................................................................. 34 2.2.3 Shake flasks ........................................................................... 34 2.2.4 Microtiter 96-well plates........................................................... 34 2.2.5 Mini Chemostats ..................................................................... 35 2.3 Analytical procedures ..................................................................... 38 2.3.1 Optical density determination ................................................... 38 2.3.2 Cell dry weight determination................................................... 38 2.3.3 HPLC analysis ......................................................................... 38 2.3.4 GC-MS analysis ....................................................................... 39 3 Results ................................................................................................ 41 3.1 Growth medium optimization .......................................................... 41 3.2 Induction of stress responses ......................................................... 46 3.3 Stress dosage determination........................................................... 47 3.4 Continuous cultures ....................................................................... 52 3.5 Continuous cultures under limiting conditions................................... 59 4 Discussion ........................................................................................... 67 Optimized growth medium.............................................................. 67 4.1 4.2 Optimization of the chemostat setup ............................................... 68 M.Sc. Research Project ii Bacillus subtilis systems biology LST Research Project 2008 4.3 Oxidative stress ............................................................................. 68 4.4 Iron limitation ............................................................................... 71 5 Conclusions and recommendations ........................................................ 74 Acknowledgements ..................................................................................... 76 References ................................................................................................. 77 Appendices................................................................................................. 84 Appendix A Experimental data................................................................ 84 A.1 Chemostats................................................................................ 84 Appendix B Flux data............................................................................. 86 B.1 Flux ratios.................................................................................. 86 B.2 Net fluxes .................................................................................. 88 Appendix C Metabolic pathways.............................................................. 91 Appendix D Mini-scale chemostats - image overview ............................. 92 Appendix E Error Analysis ...................................................................... 95 Appendix F Figure and table list ............................................................. 96 M.Sc. Research Project iii Bacillus subtilis systems biology LST Research Project 2008 1 Introduction 1.1 Preface The genetic basis of cellular responses to stress conditions has been the subject of research in the last decades and many of the underlying principles have been elucidated. The traditional generation of mutants and the subsequent phenotypic observations have come a long way. The recent rise of computational power has made the development of complex predictive models feasible. This opened new doors for the analysis of host-pathogen interactions in many human diseases. An integrated model could be able to predict the effect of future therapeutics on known pathogens based on computer calculations. The understanding cellular processes in pathogens could aid in the development of modern pharmaceuticals. This study presents an examination of the adaptation processes in B. subtilis that allows survival under oxidative stress and iron limitation, two conditions often faced by closely related pathogens after host entrance. The study focuses on the adaptations in central carbon metabolism by means of metabolic flux analysis. 1.2 Thesis project This thesis project is carried out as part of the integrated European Bacillus Systems Biology Project (BaSysBio) (51). This chapter will give some insights about the BaSysBio project and how this thesis follows from it. The chosen approach is made clear and the final outcome is described. 1.2.1 Bacillus Systems Biology Project The BaSysBio project was launched on December 1st 2006 and involves fifteen European research institutions and an Australian university. The project aims at the development of so-called “systems biology” techniques. It enables the study of the global functioning of a model bacterium: Bacillus subtilis. The obtained knowledge will ultimately be extended to other pathogenic bacteria, thereby opening the way to applications in the fields of health and the environment. The project will receive a four-year, 12.1 million euro, funding from the Sixth European Union Framework Programme under the “life sciences, genomics and biotechnology for health” priority. Systems Biology The projects’ systems biology approach is part of an upcoming and promising field that focuses on the systematic study of complex interactions in biological systems. It integrates multiple scientific and technological fields and tools in order to obtain a holistic view. This is in contrast with traditional research which focuses on identifying individual genes, proteins and cells, and studying their 1 | Introduction 4 Bacillus subtilis systems biology LST Research Project 2008 specific functions. The traditional approach yields relatively limited insight as components rarely work independent and have many interactions with each other and with other compounds. Systems biology tries to integrate all data and determine the dynamic interactions of the systems’ compounds. The recent interest in this field of study can be ascribed to the advances in analytical techniques. The development of high throughput methods, chip technology and mass spectrometry gave rise to an array of research areas. These include fields as proteomics, metabolomics and transcriptomics. By bringing together teams in Europe that are specialized in different fields, the BaSysBio project will try to design and adapt high-throughput technologies. This should facilitate quantitative measurements and to subsequently develop predictive mathematical models that will make it possible to interpret the obtained experimental data. BaSysBio goals The BaSysBio project will use the model bacterium B. subtilis to gain insight into the global structure of the regulatory networks that control bacterial metabolism. This organism is non pathogenic, easy to handle, and has been a subject of research for many years now. There are two specific BaSysBio project goals: • Unravel the global regulatory structure of B. subtilis metabolism to understand how transcriptional regulation is integrated with other levels of control. • To achieve a quantitative understanding of the cellular transcriptional responses under conditions mimicking pathogenesis and to apply the acquired knowledge and the integrated modeling/experiments strategy developed in the model bacterium to understand the regulatory networks controlling pathogenesis in the closely related, disease-causing bacteria Bacillus anthracis and Staphylococcus aureus. The second point indicates the potential application of the results. B. anthracis is responsible for the well-known anthrax disease and S. aureus causes infections that are found after hospital treatment and which are secondary to the patient's original condition. Both species are exposed to a change in environmental conditions after the infection of a host organism. In order to maintain an optimal functioning (i.e. maximal cell growth, maximal ATP production) the cell must reprogram its metabolism. A quantitative description of these control processes would give an understanding of the adaptation and unravel general design principles in B. subtilis regulation. The models obtained could help understand the adaptation in other closely related pathogenic species. 1 | Introduction 5 Bacillus subtilis systems biology LST Research Project 2008 ETH project group At the ETH Zurich in Switzerland there are two institutes participating in BaSysBio; the Institute of Molecular Systems Biology (IMSB) and the Institute of Computational Science. The group of professor Sauer at the IMSB has been active in the metabolic engineering of B. subtilis for ten years and is a leader in 13 C-based experimental flux methods and software development. They have a quantitative metabolomics lab and use metabolic flux analysis and genome-scale modeling for phenotype prediction. The Sauer group will perform the flux experiments using the reference B. subtilis strain. In this way they will provide a data set that allows partners in the program to create models of the transcriptional control structure of metabolic networks. The experiments performed at the IMSB focus on the impact on metabolism resulting from growth on different substrates and under different stress conditions. 1.2.2 Thesis project objectives In order to answer BaSysBio’s central question of “How do B. subtilis cells adapt to a shift in environmental conditions?”, we are interested in the reprogramming that occurs in the cells metabolism. This thesis project aims at the generation of B. subtilis cultures under stress. B. subtilis cultures will be grown under specific conditions that are impacted by two different stressors. Considered are the iron limited growth and oxidative stress. These conditions mimic the environment that bacteria encounter when invading their hosts and are faced with the innate immune system. Subsequent experiments will identify the regulators of altered metabolism in different conditions by means of by model predictions and metabolomics analyses. To aid in the characterization of the transcriptome, tiling arrays will be generated and sent to collaborators in the project for further processing. In order to conduct the limitation and stress experiments needed for flux analysis, a simple and economical system of mini-scale chemostats will be applied to grow the cultures. Answers for BaSysBio This thesis project will contribute to the BaSysBio project by providing the following data • In obtaining quantitative information of the adaptation processes. • By determining the metabolic part of stress responses. • Generation of samples for tiling array analysis that aid in the discovery of novel transcripts as a function of stress. • On the determination of fluxes that can be used to: 1 | Introduction 6 Bacillus subtilis systems biology LST Research Project 2008 o Identify the main metabolic responses that combat stress. o Asses the redistribution of these metals in cytoplasmic proteins. o Determine whether the NADPH supply is the main metabolic burden of oxidative stress. State of the art It is widely accepted that iron limitation and oxidative stress play an important role in the defense against pathogens. The strategies employed by invading organisms to overcome these threats is therefore of major interest and has been the subject of many researchers to date. By now, many regulatory pathways have been elucidated through knockout and stress induction experiments. The pieces of the genetic response puzzle slowly start to come together. The introduction of transcriptomics and proteomics allowed the mapping of the full response but still fails at describing the system as a whole. In the last decade, the most detailed information was derived from yeast or plants. Less was done on B. subtilis, though more is being published (more details in chapter 1.4 and 1.4.4 of this thesis). The recent advent of metabolic flux analysis enables us to approach the problems from a different angle (22). This allows the observation of the central carbon metabolism fluxes, the end-result of all the genetic regulations. For B. subtilis there have been no combined investigations into the rearrangement of the central carbon metabolism for the wild-type organism exposed to oxidative stress and iron limited conditions. Research so far, indicates a probable switch in metabolism, but for B. subtilis, only mutants and closely related species were investigated what leaves a gap to be filled by this thesis project. 1.3 Bacillus subtilis The bacterium Bacillus subtilis can be found widespread in the upper soil layers. It is easily isolated from the rhizosphere of several plants. The cells are rod shaped, measure about 2 µm in length (figure 1-1) and are covered by flagella (figure 1-2). The gram-positive cell wall maintains an intracellular pressure of 20 atmosphere. B. subtilis has the ability to form protective endospores which allow the organism to tolerate extreme environmental conditions. 1 | Introduction 7 Bacillus subtilis systems biology Figure 1-1. Scanning Electron Micrograph of B. subtilis bacteria (60). LST Research Project 2008 Figure 1-2. Transmission Electron Micrograph of a B. subtilis cell in crosssection (74). The B. subtilis is a chemoorganoheterotroph1 and prefers glucose as a carbon and energy source. Taxonomic categorisation defines B. subtilis as a eubacterium. The phylogenetic relation to Staphylococcus, Listeria, Mycoplasma and mycobacteria make B. subtilis an interesting subject for molecular biologists and in pharmaceutical research. The ability to secrete extracellular enzymes is another reason of interest. B. subtilis is frequently applied in industry to produce enzymes for detergents and for the synthesis of riboflavin (vitamin B2). Because of its frequent use, and its role as a model organism for gram-positive organisms, the sequencing of the B. subtilis genome was started in 1990. The complete project took seven years and in 1997 the results showed that the genome consisted of 4,214,810 base pairs comprising 4,100 protein-coding genes. The GC content was 43.5% and that of AT 56.9% (42). In its natural environment, the upper soil layers, B. subtilis is confronted with a wide variety of biotic an abiotic factors. Alone or in combination, the bacterium encounters heat, hyper and hypo osmotic stress, heavy metals, pH changes, feed and energy scarceness, oxygen limitation, viral infections and chemical radicals and the resulting cell damage. In these environments the micro-organism is thus continually in a stress and starvation situation. It will have to adapt as good as possible to the ever changing environment to survive the competition for food and living space and to ensure the survival of its species. When encountering stresses, the cell has to adapt its gene expression as fast as possible, in the right way, as a response. The fine tuning of response by gene expression is made possible by organizing the genes in regulatory groups. One 1 Chemoorganoheterotrophic organisms require organic substrates to get carbon for growth and development and produce energy from oxido-reduction of an organic compound. 1 | Introduction 8 Bacillus subtilis systems biology LST Research Project 2008 can distinguish operons, regulons and stimulons. These are regulatory groups of genes under control of respectively, the same promoter, the same regulatory protein or an extracellular signal. One of the strongest properties of B. subtilis to survive is the ability to form endospores. The metabolic inactive, dormant endospore is resistant to dehydration, radiation, elevated temperatures and many other stressors. But before the bacterium decides to enter the sporulation phase it will apply an alternative strategy to overcome the stressful situation. It will try to employ the development of flagella and apply chemotaxis (migration response). Further reaction mechanisms consist of the production of the osmoprotectant glycine betaine as a reaction to osmotic stress and the regulation of C source metabolism when substrates become limited (65). Under glucose rich conditions the use of other C sources is repressed, this needs to be reversed in order to use available C sources. Other strategies include the production of competent cells, the production of extracellular enzymes that open up new food sources and the excretion of antibiotics to impede competitors. The unique properties, relations and applications of B. subtilis help explain why it is a commonly used lab strain. It is one of the best understood prokaryotes in terms of molecular biology and cell biology. It is extremely adaptable which allows the easy investigation of this bacterium. 1.3.1 Metabolism B. subtilis is a soil bacterium and this environment contains a wide variety of carbohydrates including a large number of polysaccharides derived from plants, animals and micro-organisms. B. subtilis has therefore developed a system with many polysaccharide-degrading enzymes. When secreted into the environment, they allow the breakdown and subsequent uptake of these polysaccharides to be metabolized. B. subtilis is able to utilize at least 18 different mono- or disaccharides as single-carbon and energy sources (65). These include: mono-, di- and oligosaccharides, amino sugars and their N-acetyl derivatives, sugarcontaining opines, glyconic, glycaric and glycuronic acids, and sugar derived polyalcohols (linear or cyclic). These compounds are all transported into the cells, phosphorylated and subsequently catabolized via glycolysis or the pentose phosphate pathway. Metabolic pathways The central pathways of carbon dissimilation, the glycolytic pathway and the citric acid cycle (TCA cycle), are conserved in B. subtilis and virtually all other 1 | Introduction 9 Bacillus subtilis systems biology LST Research Project 2008 organisms2. Apart from glucose, these central pathways can be fed by, hexoses, disaccharides, pentoses (through ribose 5-phosphate), fatty acids (through acetyl coenzyme A) and various amino acids (glutamate, glutamine, histidine, arginine and proline are converted to 2-ketoglutarate, alanine is converted to pyruvate and aspartate and asparagines are converted to fumarate) (66). Not all substrate molecules that enter the glycolytic pathway are fully catabolized to CO2 and H2O. Depending on the availability of nutrients, a significant fraction of the pyruvate or acetyl coenzyme A generated by catabolism may be converted to by-products that are excreted in the medium. The specific products are characteristics of the growth conditions and growth phase (17). Typically, the excreted products are acids (acetate or lactate) whose production lowers the pH of the medium. As cells use up available supplies of favoured substrates, such as glucose, the excreted acids are transported back into the cell (raising the pH of the medium) and further metabolized through the TCA cycle. Production and utilization of these by-products are tightly regulated. The utilization of acetate or fatty acids as sole carbon source is not possible in B. subtilis as it lacks a functional glyoxylate shunt (isocitrate to succinate and malatate via glyoxylate). Even though acetate is being transported and metabolised, it cannot grown on it. During the growth of B. subtilis, the TCA cycle peaks when cells enter the stationary growth phase. At the end of the exponential growth phase, when there is no glucose left, metabolites as acetoin and acetate are converted to acetyl coenzyme A and enter TCA cycle. 1.3.2 Pathogenic properties The property that makes B. subtilis so interesting is the fact that it is closely related to pathogenic species, but easy and safe to handle. It thus allows us a safe insight in the processes going on when a bacterium enters a host organism for infection. The first and immediate barrier is the innate immune system. This system is non specific and relies on the use of a variety of antimicrobial mechanisms to impose stress on the invading organism. The two mechanisms that are linked to this thesis, are the strategy of iron-withholding and the application of oxidative burst in phagocytic leukocytes. Oxidative burst Phagocytic leukocytes (neutrophils, eosinophils, monocytes/macrophages) play a critical role in innate immune responses against bacteria, fungi, and other pathogens. Neutrophil-mediated bacterial killing can use a killing mechanism that is initiated by the assembly of the NADPH oxidase complex at the phagosome 2 The pathways of the central carbon metabolism including the glycolysis, the citric acid cycle and the pentose phosphate pathway are depicted in Appendix C (page 91). 1 | Introduction 10 Bacillus subtilis systems biology LST Research Project 2008 membrane. The phagocyte NADPH oxidase is an enzyme with a redox center that transfers electrons from NADPH to molecular oxygen, thereby generating superoxide. This reaction by NADPH oxidase, the so-called oxidative burst, is shown in equation 1.1 (61). NADPH + 2O2 → NADP+ + 2O2− + H+ 1.1 However, superoxide has a low bactericidal potency, so within the phagosome, superoxide is converted into H2O2, which reacts with superoxide to generate hydroxyl radicals and singlet oxygen, both highly reactive and toxic compounds. Superoxide can also react with nitrogen oxide, generated by inducible NO synthase, to yield peroxynitrite, a very reactive nitrogen intermediate. The natural vulnerability of organisms of reactive oxygen species (ROS) is also exploited by plants and micro-organisms that wish to suppress the growth of their competitors. They secrete redox-cycling compounds that diffuse into nearby bacteria and generate oxidative stress. Plants also generate H2O2 by NADPH oxidases that are activated along plant wound sites, a natural habitat of B. subtilis. Iron-withholding An essential ability of pathogens is the ability to invade and multiply successfully within host tissues. Proliferation of a pathogen is critical for the establishment of an infection and it facilitates the pathogen to produce its full arsenal of virulence factors required for pathogenicity (53, 58, 73). In order to proliferate, the invading organism should multiply and synthesize biomass requiring iron. By keeping the iron levels low, the invading organisms face a problem with their iron acquisition. Freely available iron is virtually absent in normal human plasma. It contains an iron binding protein, transferrin, that ensures that the amount of free ferric iron remains about 10-18 M. To counteract the effects of the immune response and precautions of the host, organisms have developed many survival strategies. These will be explained in detail later in this chapter. 1.3.3 General stress response Bacteria like B. subtilis are regularly exposed to fluctuating environmental stresses. To ensure survival in these harsh conditions B. subtilis may employ its flaggellar machinery to find a more favorable location, or it tries to adapt to the changes in its vicinity by responding to the imposed stress. This is generally accomplished by changing the gene expression for the genes whose products are required to combat the stresses. The activation of specific transcription factors 1 | Introduction 11 Bacillus subtilis systems biology LST Research Project 2008 that interact with RNA polymerase, up-regulates the transcription of these stress response genes. An essential factor in stress resistance is the sigma factor (σ factor), a subunit of RNA polymerase. A sigma factor is a prokaryotic transcription initiation factor that enables specific binding of RNA polymerase to gene promoters. Different sigma factors are activated in response to different environmental conditions. Each of several sigma factors in the cell is required for the transcription of a specific subset of genes/operons within their regulon. The most important regulon in B. subtilis is the σB. The σB regulon comprises about 200 open reading frames (5% of it’s genome), the products of which aid in the general stress resistance of the cell (45). Every molecule of RNA polymerase contains exactly one sigma factor subunit. There are also anti-sigma factors that inhibit the function of sigma factors. The emerging view is that relatively few genes in the σB regulon serve to directly counter environmental stress. Instead it appears that the general stress response aligns metabolism and that creates a more passive stress resistance (57). An example is the induction of Thioredoxin A by σB in B. subtilis, which is thought to keep cytoplasmic proteins in a reduced state by reducing disulfide bonds generated by H2O2 (63, 67). The upstream component of the environmental stress signal transduction cascade is activated by a kinase that switches binding partners from a macromolecular complex, the stressosome. Details on this signaling cascade are shown in figure 1-3. 1 | Introduction 12 Bacillus subtilis systems biology LST Research Project 2008 Figure 1-3. Schematic of the environmental stress response. Enzymatic activities are italicized, and partner switching modules linked by dashed arrows (30). In a situation without stress, the RsbT kinases are sequestered by the stressosome. Under stressed conditions, the kinase activity of RsbT results in the phosphorylation of the stressosome, and RsbT is released. When RsbT interacts with RsbU it stimulates its phosphatase function towards RsbV-P. Dephosphorylated RsbV attacks the RsbW-σB complex and liberates σB to bind to the RNA polymerase to form the complex that transcribes the σB regulon. To reset the switch, the RsbX phosphatase dephosphorylates the stressosome, which can then re-capture RsbT, down-regulating RsbU phosphatase activity. The mechanisms by which stress is communicated are unknown. A drop in ATP is generally observed but it is not known whether the ATP levels themselves or other effectors trigger the induction (78). 1.3.4 Oxidative stress Life evolved on earth in an anaerobic way and when oxygen levels slowly rose, organisms had to develop mechanisms to defend themselves against the poisonous effects of oxygen. In the current age, oxygen is omnipresent and presents one of the most important stresses imposed on organisms on earth. Oxygen toxicity is caused by partially reduced oxygen species that are more reactive than is molecular oxygen itself (see figure 1-4). Such species are the inevitable by-products of aerobic metabolism where significant amounts are 1 | Introduction 13 Bacillus subtilis systems biology LST Research Project 2008 generated by enzymic misdirection of electron to O2 (46). It is assumed that flavin dependent transfer reactions of the respiratory chain are primarily responsible for the generation of ROS. In addition to the autoxidation of cellular components, exposure to ionizing radiation, redoxactive compounds or transition metals and depletion of antioxidants may contribute to the intracellular formation of ROS. If the amount of ROS increases to toxic levels, the generated reactive oxygen can result in the inactivation of enzymes (36), and the cells encounter oxidative stress. Figure 1-4. The redox states of oxygen with standard reduction potentials (36). ROS include singlet oxygen, O2- (superoxide), H2O2 (hydrogen peroxide), hydroxyl radicals, hypohalous acids and peroxynitrite. Oxygen crosses the membrane so freely that the intra- and extracellular levels are almost equal. The intracellular oxygen can take electrons from the exposed redox moieties of electron-transfer enzymes, thereby generating partially reduced oxygen species. Many organisms experience a steady flux of endogenously generated oxidants. There are several externally induced source of oxidative stress (37). An overview is shown in figure 1-5. First of all there are the plants and micro-organisms that secrete redox-cycling compounds in their environment to suppress the growth of their competitors. These compounds are toxic chemical agents that penetrate bacterial cells and catalyze electron transfer from redox enzymes to molecular oxygen, generating O2- and H2O2. These compounds are typically viologens, phenzines or quinones. 1 | Introduction 14 Bacillus subtilis systems biology LST Research Project 2008 Figure 1-5. Source of oxidative stress. Images shows (a) intracellular autoxidation, (b) environmental redox reactions, (c) H2O2 released by competing micro-organisms, (d) phagosomal NADPH oxidase, (e) redox-cycling antibiotics (36). Second you have the hydrogen peroxide which can pass the cell membrane as it is an uncharged molecule. H2O2 stress will therefore occur whenever H2O2 is available in the environment. Third is the generation of H2O2 by NADPH oxidases that are activated along plant wound sites and mammalian macrophages. The fourth source of oxidative stress is the presence of redox cycling drugs in the form of antibiotics. Mechanisms of damage Superoxide and hydrogen peroxide penetrate into the active site of enzymes with a solvent-exposed [4Fe-4S]2+ cluster. After penetration they bind the critical iron atom and oxidize the cluster to a redox state that is unstable. The iron atom is then released as ferric iron (Fe3+) and the cluster is left in an inactive [3Fe-4S]+ form. This and other mechanisms are depicted in figure 1-6. 1 | Introduction 15 Bacillus subtilis systems biology LST Research Project 2008 Figure 1-6. Mechanisms of damage by reactive oxygen species. (a) the Fentonmediated damage to proteins (rectangles) and DNA, (b) the oxidation of [4Fe-4S]2+ clusters, (c) inhibition of transketolase, and (d) disruption of the sulfur assimilatory pathway. The details of inactivation of transketolase and of sulfur metabolism remain unclear (36). Less known ROS activities are associated with the disruption of sulphur metabolism and the inhibition of transketolase (6). More is known on other damaging processes that results from intracellular presence of O2- and H2O2. The mutagenic properties of H2O2 result from its capacity to directly oxidize unincorporated intracellular ferrous iron (Fe2+) via the Fenton reaction (see equation 1.2 and 1.3). This form of iron is not incorporated into hemes, ironsulfur clusters of high affinity mononuclear sites within proteins. The ferrous iron that is involved in the reaction is known to be associated with DNA due to its Fe2+-chelating properties (34). The result of the Fenton reaction with DNA associated iron is the breakage of DNA strands (38). Fe2+ + H2 O2 → FeO2+ + H 2 O 1.2 FeO2+ + H+ → Fe3+ + HO• 1.3 Even though O2- is incapable of causing direct DNA damage, it participates in the release of iron from damaged dehydratase clusters. The released ferrous iron is known to bind to the surfaces of lipids and proteins who in turn suffer from the initial oxidative damage. This is the result of the reduction of ferric iron to form ferrous iron (equation 1.4) which can take part in the local generation of damaging hydroxyl radicals via the Fenton reaction (40). O2− + Fe3+ → O2 + Fe2+ 1.4 A selection of enzymes that is known to be vulnerable to oxidative stress is shown in table 1-1 (8). 1 | Introduction 16 Bacillus subtilis systems biology LST Research Project 2008 Table 1-1. Enzymes in B. subtilis known to be vulnerable to oxidative stress and their function (8). Enzyme Pathway Triacylglycerol lipase Glycerolipid metabolism Alpha-amylase Starch and sucrose metabolism Phosphoadenylyl-sulfate reductase Sulfur metabolism (thioredoxin) Adenylyl-sulfate reductase Selenoamino acid metabolism / Sulfur metabolism amidophosphoribosyltransferase Glutamate metabolism / Purine metabolism Oxalate decarboxylase dicarboxylate metabolism Glutamate-ammonia ligase Glutamate metabolism / Nitrogen metabolism / Peptidoglycan biosynthesis / Two-component system General 1.3.5 Iron limitation Iron exists in two redox states, the reduced Fe2+ ferrous form and the oxidized Fe3+ ferric form. Iron is an abundant metal in the Earth’s crust and was much incorporated in early life on this planet. Nowadays it is an absolute requirement for all life forms and is part of many biological processes such as respiration, N2 fixation, methanogenesis, H2 production and consumption, photosynthesis, the TCA cycle, oxygen transport, gene regulation and DNA biosynthesis. When oxygenic photosynthesis started to provide the atmosphere with molecular oxygen the much used iron showed to have extremely toxic side effects. In order to survive, organisms have to maintain a delicate iron homeostasis which secures adequate supplies and reduces the potentially harmful intracellular free iron levels. There are essentially five strategies applies to manage cellular iron which are applied depending on the environmental conditions (2). 1. High-affinity iron transport enabling iron to be scavenged, in various forms, from the surroundings. 2. Deposition of intracellular iron stores to provide a source of iron that can be drawn upon when external supplies are limited. 3. Employment of redox stress resistance systems (e.g. degradation of ironinduced reactive oxygen species and repair of redox stress-induced damage). 4. Control of iron consumption by down-regulating the expression of ironcontaining proteins under iron-restricted conditions. 5. An iron-responsive regulatory system that coordinates the expression of the above iron homeostatic machinery according to iron availability. Ferric iron (Fe3+) is almost insoluble under aerobic, aqueous and neutral pH conditions. After internalization, Fe3+ is reduced to Fe2+ by a process that involves hydrolysis and reductases, for subsequent incorporation into haem- or iron-containing proteins. For the acquisition of iron, B. subtilis produces high 1 | Introduction 17 Bacillus subtilis systems biology LST Research Project 2008 affinity ferric chelators called siderophores (iron carriers). About 500 siderophores have been characterised and classified. The molecules are low in molecular mass (< 1000 Da) and are highly specific towards ferric iron (Kaff > 1030). The siderophore in B. subtilis is the catecholic bacillibactin (BB) which is also found in its pathogenic family member B. anthracis. The process of bacillibactin mediated iron acquisition is still being researched and not completely clear (47). The most recent model (2006) is shown in figure 1-7. Figure 1-7. Most recent model of the bacillibactin-mediated iron acquisition pathway in B. subtilis. Pathway steps I, II, V and VI were functionally characterized. The broken arrows indicate putative pathway steps (47). Once internalized, metal ions are incorporated as cofactors into enzymes and metalloproteins, interact reversibly with metalloregulators and low molecular mass ligands, and can be incorporated into storage reservoirs or exist in a lowmolecular-weight pool. The disturbance of the iron homeostasis gives rise to numerous problems. The result of free Fe2+ and the generations of reactive oxygen species was already discussed in 1.3.4 but next to the disruption of iron-sulfur clusters, the lack of iron under iron limited growth condition could result in difficulties building enzymes with iron incorporated. Many enzymes function only with their proper iron sulphur center or iron cofactor. Effects of this are seen throughout the cell 1 | Introduction 18 Bacillus subtilis systems biology LST Research Project 2008 since these enzymes function in parts of the TCA cycle, respiration and the biosynthesis of amino acids. A selection of potentially vulnerable enzymes and their respective function is shown in table 1-2 (8). Inactivation of these enzymes may result in growth inhibition as has been shown for the biosynthesis of mutants defective in iron–sulfur clusters (18). Table 1-2. Selection of enzymes with iron atoms in B. subtilis and their function derived from (8). Enzyme Pathway Succinate dehydrogenase Benzoate degradation via CoA ligation Butanoate metabolism TCA cycle Oxidative phosphorylation Reductive carboxylate cycle (CO2 fixation) Two-component system - General Cysteine dioxygenase Cysteine metabolism Taurine and hypotaurine metabolism Unspecific monooxygenase Androgen and estrogen metabolism Arachidonic acid metabolism Caffeine metabolism Fatty acid metabolism gamma-Hexachlorocyclohexane degradation Linoleic acid metabolism Metabolism of xenobiotics by cytochrome P450 Tryptophan metabolism Levansucrase Starch and sucrose metabolism Two-component system - General Amidophosphoribosyltransferase Glutamate metabolism Purine metabolism Glycerol-3-phosphate cytidylyltransferase Glycerophospholipid metabolism Biotin synthase Biotin metabolism S-ribosylhomocysteine lyase Methionine metabolism Glutamate synthase Glutamate metabolism Nitrogen metabolism Phosphoadenylyl-sulfate reductase Sulfur metabolism (thioredoxin) Adenylyl-sulfate reductase Selenoamino acid metabolism Sulfur metabolism 3-deoxy-7-phosphoheptulonate synthase Phenylalanine, tyrosine and tryptophan biosynthesis Lysine 2,3-aminomutase Lysine degradation 1.4 Adaptation under stress conditions Because B. subtilis cannot exert much control over its environment, it has developed mechanisms that sharply adjust the synthesis of defensive proteins in response to stress. 1 | Introduction 19 Bacillus subtilis systems biology LST Research Project 2008 1.4.1 Oxidative stress For the response to oxidative stress, micro organisms use various sensing mechanisms. This is because superoxide and hydrogen peroxide stress do not always occur simultaneously and both differ in their biochemistry (36). The response to superoxide stress is regulated by two proteins: SoxR, a sensor protein that detects redox stress and SoxS, a secondary transcription factor that positively regulates a variety of genes. These genes are involved in cluster repair, drug efflux and resistance, or encode oxidant resistant dehydratase isozymes. Furthermore it induces the fur gene which is the iron-uptake regulatory protein. Hydrogen peroxide responses are mediated by the PerR regulator. This protein is B. subtilis’ equivalent for the more common OxyR system. PerR is a transcription factor and binds in the proximity of various H2O2 stress regulons. These in turn fulfil roles for H2O2 scavenging, heme synthesis, FeS cluster assembly, iron scavenging, iron-import control (Fur regulon), divalent cation import and the reduction of disulfide. The PerR protein contains a single ferrous iron atom in its center and in this form it functions as a transcriptional repressor. A H2O2 attack causes the oxidation of the iron center and the subsequent loss of DNA binding capacity. This allows the transcription of genes that are normally blocked by PerR. Cellular defenses against oxidative stress In reply to the ever present oxidative stress, B. subtilis induces proteins that collectively provide multiple levels of protection. The most obvious and important are the scavenging systems. There are two distinct groups of scavenging enzymes, those of superoxide and those of hydrogen peroxide. The first group consists of the protein superoxide dismutase (SOD). This enzyme catalyzes the dismutation of superoxide into oxygen and hydrogen peroxide at a rate that approaches catalytic perfection. The mechanism by which SOD performs the catalysis consists of two half reactions (see equation 1.5 and 1.6 for manganese cofactored SOD). Mn3+ - SOD + O2− → Mn2+ - SOD + O2 1.5 Mn2+ - SOD + O2− + 2H+ → M3+ - SOD + H2O2 1.6 The hydrogen peroxide is scavenged by specialised peroxidases (equation 1.7) and catalases (equation 1.8). RH2 + H2 O2 → R + 2H2 O H2 O2 + H2 O2 → O2 + 2H2 O 1 | Introduction 1.7 1.8 20 Bacillus subtilis systems biology LST Research Project 2008 The primary scavenger in B. subtilis is peroxidiredoxin AhpCF, a two component NAD(P)H peroxidase (49). AhpC is oxidized by H2O2 and the conformational change that results from this in turn attracts the AhpF, an NAD(P)H reducible flavoprotein, which aids in the further reduction of H2O2. The turnover rate of peroxidase is not very high but the rapid uptake of H2O2 into the first reaction half helps lowering the free H2O2 concentration. Other defenses include the exclusion and export of redox-cycling compounds, the protection of iron-sulfur clusters, the DNA repair processes, the protein repair processes and the control of unincorporated iron levels. The control of free iron in the cell is of major importance because ferrous iron can react with H2O2 and result in the formation of hydroxyl radicals. When the iron levels are kept low, the damage resulting from hydroxyl radicals is kept to a minimum. 1.4.2 Iron limitation The control of iron homeostasis in B. subtilis is mediated by the Fur protein (52). This transcription factor is activated by binding ferrous iron (a positive repressor). The activated dimeric Fur protein then binds a specific 19 bp DNA sequence (see figure 1-8), called the “Fur box”, within the promoter region of the regulated genes. When iron is scarce, the Fur protein is deactivated and B. subtilis expresses genes involved in the synthesis and uptake of the siderophore bacillibactin and uptake systems to pirate other microbial siderophores (2, 27). Fur also coordinates a response that spares iron by downregulating the expression of iron-containing proteins, including succinate dehydrogenase, aconitase, cytochromes, and biosynthetic enzymes for heme, cysteine, and branched chain amino acids (26). The exact affinity of Fur for Fe2+ is such that the right levels of iron are maintained for proper functioning of essential ironcontaining and iron-utilizing enzymes. When iron levels exceed those that are needed for metalloenzyme functioning, Fur gets reactivated and hinders the access of RNA polymerase resulting in the repression of downstream genes that function in iron uptake (44). Figure 1-8. Schematic representation of Fur-mediated gene repression (2). An analysis in B. subtilis on the role of fur showed that 46 genes were repressed by iron and Fur, all of which had defined or predicted Fur-binding sites (3). Of 1 | Introduction 21 Bacillus subtilis systems biology LST Research Project 2008 these genes there were 27 involved in iron uptake and the remainder encoded flavodoxins, flagella subunits, a cytochrome P450 monooxygenase homologue as well as genes involved in amino acid metabolism and nitrite reduction. Notably certain gram-positive bacteria, especially animal pathogens, can obtain iron from transferrins, ferritin, hemoglobin, and other iron-containing proteins of their host (58). Due to the iron withholding strategy pathogens have developed high affinity iron uptake systems that are critical for their survival. While some invading bacteria produce iron chelating siderophores that remove iron from the host sources, other bacteria rely on direct contact with host proteins. In the latter case iron can be abstracted from the protein surface or taking up heme into their cytoplasm. Heme can than be converted to biliverdin by the enzyme heme oxygenase with NADPH as the reducing agent. The resulting iron is released from the molecule as the ferric ion. 1.4.3 Connection between iron and oxidative stress responses From the previous paragraphs it becomes clear that there is an apparent link between metalloregulation and oxidative stress responses. Given the scenario of metalloproteins as the key targets for oxidative stress, a generalized downregulation of the enzymes involved in the metal-based detoxifying enzymes is an effective strategy to limit the availability of sensitive redox cofactors that would enhance the damage. Many enzymes that feel iron limitation are also influenced by O2- that attacks their iron centers. The key regulator for the inducible peroxide stress response, PerR, is a metal-binding repressor protein related to the ferric uptake regulator (Fur) family of proteins (10). The binding motifs of PerR and Fur differ only at one critical position. The result of this is that the cellular resistance to oxidative stress is linked, at the level of gene expression, to intracellular pools of iron as sensed by PerR. Another connection is the fact that the expression of the fur gene, encoding proteins involved in iron homeostasis, is induced by both PerR and SoxS, the regulatory proteins of the superoxide stress response (3, 25). This control seems reasonable since H2O2 is able to disrupt the ferrous Fur complex. In its inactivated form, Fur can no longer prohibit the generation of iron import proteins and the cell would suffer from over-imported iron levels. Transcriptome and proteome studies support these ideas and both members of the PerR and the Fur regulon are induced by peroxide and superoxide stress (49). On the other hand, it is known that iron starvation induces members of the PerR regulon in B. subtilis (3). Figure 1-9 shows a simplified overview of the components participating in stress responses. Many of them are related and effect each other. 1 | Introduction 22 Sensing Bacillus subtilis systems biology LST Research Project 2008 H2O2 Low iron O2- PerR regulon Fur regulon SoxR sigmaB Effector SoxS AhpCF Iron homeostasis NAD(P)H peroxidases SOD Superoxide dismutases General stress response Figure 1-9. interconnections between stress sensing and effector systems. Next to the individual pathways for oxidative and iron limitation there is the general stress response (σB). It defines a different and much larger set of proteins that have a similar function to those expressed by specific stress regulatory pathways. The non specific response to oxidative stress also includes catalases, DNA-protecting proteins, and proteins required for repairing the damage induced by oxidative stress. The main difference between both strategies is that the specific response is induced only by oxidative stress. The nonspecific response is co-induced, with other components of the general stress resistance machine, by all stimuli that activate σB (32). The general stress response will function when cell experience oxidative stress as they switch to a non-growing state. In this situation, unbalanced metabolism can result in significant oxidative damage. The general stress response therefore provides complementary protection under conditions in which the specific oxidative stress regulon controlled by PerR is inactive (10). In general, the non specific stress response is less sensitive to environmental stressors and the specific stress response mechanisms readily respond to lower degree of stress (33). 1.4.4 Carbon metabolism under stress As a result of the linkage between oxidative stress and iron limitation many of the problems that arise under these conditions impact carbon metabolism at the same location. Enzymes involved in metabolism often contain iron centers that are potential target sites for disruption by reactive oxygen species. This would render the enzymes dysfunctional and require alternative cellular strategies from the organism to survive. In a situation were iron is limited one can imagine that the synthesis of the iron centers in iron containing proteins is more difficult and their resulting intracellular concentrations will be lower. 1 | Introduction 23 Bacillus subtilis systems biology LST Research Project 2008 A closer look at the enzymes involved in central carbon metabolism under aerobic growth yields a list of possible targets of oxidative and iron limitation (41). Table 1-3 and table 1-4 show an overview of proteins that depend on an iron sulfur center or cofactor for proper functioning. Only the TCA cycle and the oxidative phosphorylation are involved while the glycolysis, the pentose phosphate pathway and the malic enzyme pathway contain no vulnerable enzymes (not more vulnerable than any other protein in the cell e.g. the oxidation of amino acids). Table 1-3. TCA cycle, possible target enzymes of oxidative stress and iron limitation Enzyme Property Reaction(s) EC number 4.2.1.3 aconitate hydratase Iron Sulfur cluster (1) citrate -> isocitrate (aconitase) (2) (1a) citrate -> cis-aconitate + H2O (3) (1b) cis-aconitate + H2O -> isocitrate succinate Iron Sulfur cofactor succinate + acceptor -> 1.3.99.1 dehydrogenase ironfumarate + reduced acceptor sulfur subunit Table 1-4. Oxidative phosphorylation, possible target enzymes of oxidative stress and iron limitation Enzyme Property Reaction(s) EC number NADH dehydrogenase Iron Sulfur cluster NADH + H+ + acceptor -> 1.6.99.3 NAD+ + reduced acceptor succinate Iron Sulfur cofactor succinate + acceptor -> 1.3.99.1 dehydrogenase fumarate + reduced acceptor A clear example is the enzyme aconitase in the TCA cycle that is known to switch function under iron limiting conditions (figure 1-10). Figure 1-10. The RNA-binding function of the aconitase depends on the iron status of the cell (12). When iron is limited, aconitase cannot function in the isomerization of citrate to isocitrate due to the absence of its iron sulfur cluster. In this situation, the enzyme gains the ability as an iron responsive element by binding specific mRNA sequences involved in iron homeostasis (12). 1 | Introduction 24 Bacillus subtilis systems biology LST Research Project 2008 In order to counteract the stress caused by reactive oxygen species, the cell needs vast amounts of reducing equivalents. Reducing molecules in the form of NADH and NADPH provide reducing equivalents to the redox enzymes that destroy the reactive oxygen species. One of the primary ROS scavengers induced by the PerR regulon, peroxidiredoxin AhpCF, is a NAD(P)H dependent peroxidase. NADPH is used as the main electron donor under oxidative stress and the NADPH/NADP+ ratio is usually kept high, in favor of the NADPH (56). There are multiple sites in metabolism that generate NADPH (75). An overview of NADPH generating enzymes and their role in oxidative stress is given in figure 1-11. NADPH G6PD 6PGD GDH ID ME ALD NADPH oxidase NADD NADP+ NADH Iron release ↑ ROS Reducing potential ↑ Thioredoxin PerR / Peroxidases NADK NAD+ ETC Reducing potential PerR / Peroxidases sigmaB ↓ ROS Figure 1-11. Pathways by which NAD and NADP are generated and affect cellular ROS generation in B. subtilis. Dashed lined indicate ROS formation, solid lines indicate ROS mitigation. Detailed description in text below. The putative transhydrogenase activity in B. subtilis (68) is omitted. It is seen that NADH contributes to cellular reducing potential, and can also increase ROS generation by the electron transport chain (ETC) or by inducing iron release from ferritin. NADPH plays a key role in cellular antioxidation capacity by delivering electron to peroxidases, and by contributing to thioredoxin reductase-mediated generation of thioredoxin. On the other hand, NADPH can also be used by NADPH oxidase to generate superoxide. NADPH is generated by enzymes from different pathways: • Glucose-6-phosphate dehydrogenase (G6PD – PP pathway) • 6-Phosphogluconate dehydrogenase (6PGD – PP pathway) • Isocitrate dehydrogenase (ID – from TCA cycle) • Malate dehydrogenase (Malic Enzyme, ME – from anaplerotic pathway) • Aldehyde dehydrogenase (ALD – various pathways) • Glutamate dehydrogenase (GDH – glutamate metabolism) These are all dehydrogenases that oxidize a substrate by transferring one or more protons and a pair of electrons to the acceptor NADP+. NADP+ can be 1 | Introduction 25 Bacillus subtilis systems biology LST Research Project 2008 derived from an internal pool or from NADH by NAD+ kinase and the subsequent action of NADH dehydrogenase. When not used for the relief of oxidative stress, NADPH is consumed in metabolic pathways involved in the biosynthesis of fatty acids, sterols, amino acids and purines. 1 | Introduction 26 Bacillus subtilis systems biology LST Research Project 2008 2 Materials and methods This chapter describes how the techniques, mentioned in the introduction, were applied in this research. Many methods on the cultivation of B. subtilis were inspired on the work of Harwood and Cutting (31). For all solutions were water is needed, ultrapure (dd)H2O with an electric resistance of more than 17 MΩ/cm (conductivity of less than 60 nS·cm) is used. This water is generated by a NANO pure ultrapure water system (SKAN AG, Basel, Switzerland). Remarks on used materials: • Where filter sterilizing is mentioned, larges volumes (> 50 mL) were processed with a 0.22 µm Steritop® filter (Millipore, MA, USA), and smaller volumes were run through a 0.22 µm Millex GP syringe tip filter (Millipore, MA, USA). • Solution were mixed under sterile conditions in a laminar flow cabinet (Lamin Air HB 2448, Heraeus Instruments, Hanau, Germany). • All compounds were weighed with either the Mettler AT200 or the PM4800 balance (both Mettler-Toledo, Greifensee, Switzerland). • Plasticware (eppies, needles, cuvettes, Petri dishes, 15 mL preculture tubes, other tubes?) from Greiner Bio-one (Frickenhausen, Germany) 2.1 Materials 2.1.1 Strains The BaSysBio Bacillus subtilis reference strain is used throughout the experiments. This B. subtilis 168 trp+ is stored, frozen in 15% (v/v) glycerol at -80 °C. Unlike the 168 parent strain, this strain is not auxotrophic for tryptophan due to the stable integration of the trpC gene into the genome. Table 2-1. Details of used strain. Species Description B. subtilis wild type 168 cured of trp auxotrophy Genotype trp+ Source Laboratoire de Génétique Microbienne, I.N.R.A, France 2.1.2 Growth Media The experiments required two different types of medium. The LB medium was used for initial precultures and M9 minimal medium for the final precultures and the actual experiments. 2 | Materials and methods 27 Bacillus subtilis systems biology LST Research Project 2008 LB medium Note: Miller formulation Total volume: 1 L 1 L ddH2O 10 g Bacto-Tryptone Becton Dickinson and Co 10 g NaCl Merck 5 g Yeast extract Becton Dickinson and Co Remarks: Medium can be autoclaved at 121 °C for 20 minutes and stored at RT. On the nomenclature of LB medium. The abbreviation LB was intended to stand for “Lysogeny Broth” and not the commonly seen, “Luria Bertani”. See the explanation in the postscript of (7). What is used here is LB medium, the Miller formulation (10 g/L NaCl). M9 medium Total volume: 1 L 781 mL ddH2O 1 mL CaCl2 solution 10 mL Trace elements 1 mL MgSO4 solution 1 mL Iron/Citrate solution 6 mL Glucose solution 200 mL M9 stock Remarks: Final glucose concentration in this (100 mM) (100x stock) (1 M) (50 mM / 100 mM) (50% w/v) (5x stock) medium will be 3 g/L. Contrary to what is described by Harwood and Cutting, MgSO4, CaCl2 are added as separate solutions to prevent precipitation problems. The FeCl3/citrate mix is also added as a separate solution as this compound is unstable. Store the solution in the dark at RT. The experiments with labeled substrate use the same M9 medium with an adjustment to glucose added. When [1-13C] experiments were performed, all the glucose was substituted by 1 gram of [1-13C]-labeled glucose (>99%, Spectra Stable Isotopes, MS, USA). The [U-13C] experiments required 0.2 gram of [U-13C]-labeled glucose (>99%, Spectra Stable Isotopes, MS, USA) and 1.6 mL of the glucose stock solution (50% w/v). M9 stock solution Note: 5x stock solution Total volume: 1 L 1 L ddH2O 42.5 g Na2HPO4·2H2O Merck Merck 15 g KH2PO4 Merck 5 g NH4Cl 2.5 g NaCl Merck Remarks: Adjust to pH 7.0 using 4M NaOH. Solution can be autoclaved at 121 °C for 20 minutes. Store at RT. 2 | Materials and methods 28 Bacillus subtilis systems biology LST Research Project 2008 The adjustment to pH 7.0 was done with a 691 pH meter (Metrohm, Zofingen, Switzerland). 2.1.3 Plate media LB agar medium Note: 1.5 % w/v agar Total volume: 1 L 1 L 10 g 10 g 5 g 15 g Remarks: Medium can at RT. ddH2O Bacto-Tryptone NaCl Yeast extract Agar be autoclaved at Becton Dickinson and Co Merck Becton Dickinson and Co Invitrogen 121 °C for 20 minutes and stored 2.1.4 Solutions Trace elements solution Note: 100x stock solution Total volume: 1 L 1 L ddH2O 100 mg MnCl2·4H2O Merck Sigma 170 mg ZnCl2 Merck 43 mg CuCl2·2H2O Merck 60 mg CoCl2·6H2O 60 mg Na2MoO4·2H2O Aldrich Remarks: Autoclavable and store in the dark at RT Calciumchloride solution Note: 100 mM Total volume: 100 mL 100 mL ddH2O 1.47 g CaCl2·2H2O Merck Remarks: Autoclavable and store at RT. Magnesiumsulfate solution Note: 1 M Total volume: 100 mL 100 mL ddH2O 24.6 g MgSO4·7H2O Merck Remarks: Autoclavable and store at RT. 2 | Materials and methods 29 Bacillus subtilis systems biology LST Research Project 2008 Glucose solution Note: 50% w/v Total volume: 1 L 1 L ddH2O 500 g C6H12O6 Merck Remarks: Use a warm water bath to speed up the dissolving. Solution is autoclavable at 121 °C for 20 minutes. Store at RT Deferoxamine mesylate solutions Note: 3.6 mM Total volume: 5 mL 5 mL ddH2O 11.8 mg Deferoxamine mesylate Remarks: Filter sterilize and store in dark. Note: 60 mM Total volume: 16.4 mL 16.4 mL ddH2O 0.6464 g Deferoxamine mesylate Remarks: Filter sterilize and store in dark. Sigma Aldrich Sigma Aldrich Paraquat solutions Note: 6 mM, Methyl viologen dichloride hydrate Total volume: 10 mL 10 mL ddH2O 15.43 mg Paraquat Sigma Aldrich Remarks: Filter sterilize and store in dark. Note: 0.3 M, Methyl viologen dichloride hydrate Total volume: 2 mL 2 mL ddH2O 0.1543 g Paraquat Sigma Aldrich Remarks: Filter sterilize and store in dark. 2 | Materials and methods 30 Bacillus subtilis systems biology LST Research Project 2008 Iron / Citrate solutions Note: 1.2 mM FeCl3 Total volume: 10 mL 10 mL ddH2O 0.0324 g FeCl3·6H2O Sigma Remarks: Filter sterilize and store in dark at 4 °C. Note: 50mM Citrate Total volume: 10 mL 10 mL ddH2O 1.0547 g Citric Acid Remarks: Filter sterilize and store in dark at 4 °C. Note: 50 mM FeCl3 / 100 mM citric acid Total volume: 100 mL 100 mL ddH2O 1.35 g FeCl3·6H2O Sigma Sigma 2.10 g C6H8O7·H2O Remarks: Filter sterilize and store in dark at 4 °C. Note: 50 mM FeCl3 Total volume: 100 mL 100 mL ddH2O 1.35 g FeCl3·6H2O Sigma Remarks: Filter sterilize and store in dark at 4 °C. Physiological Salt Note: 0.9% (w/v) isotonic buffer / wash buffer Total volume: 1 L 1 L ddH2O 9 g NaCl Sigma Remarks: Autoclavable and store at RT. 2.2 Cultivations 2.2.1 Fermentations This project relies on a number of different culturing methods that are derived from basic reactor design. This paragraph will give a short introduction on the differences and what calculations can be performed. Batch reactors Batch reactors are usually well sealed and properly mixed vessels (figure 2-1). This is the classical operation of the bioreactor that is used extensively in industrial processes. It is fed once at the start and then operated for several hours (figure 2-2). It can be run under sterile conditions and the risk of strain evolution is less then in a continuous reactor. Many industrial organisms start producing their valuable products when substrate gets limited, this reactors provides an ideal environment as no substrate enters the reactor. 2 | Materials and methods 31 Bacillus subtilis systems biology LST Research Project 2008 Time [h] Figure 2-1. Schematic view of a batch reactor. Figure 2-2. biomass (solid) and substrate (dashed) concentration during a batch culture. The disadvantage is that the experimental data produced, is hard to interpret because the environmental conditions experienced by the cells, vary with time. Continuous reactors Continuous reactors are fed and harvested at a constant rate throughout the process (figure 2-3). A typical operation of the continuous reactor is the chemostat. In this case the added medium is designed such that there is a single limiting substrate. This allows the controlled variation in the specific growth rate of the biomass. By varying the feed flow rate to the bioreactor, the environmental conditions can be varied and thereby valuable information concerning the influence of the environmental conditions on the physiology can be obtained. Time [h] Figure 2-3. Schematic view of a continuous reactor. Figure 2-4. biomass (X) and substrate (S) concentration during a continuous culture after a batch phase. After an initial start-up phase, the production of biomass in this reactor equals the amount of biomass pumped out of the system, a steady state (figure 2-4). The same holds for possible microbial products formed by the organisms. It is rarely used in industrial processes since it is sensitive to contaminations (via the 2 | Materials and methods 32 Bacillus subtilis systems biology LST Research Project 2008 feed stream), and to the appearance of spontaneously formed mutant that may out-compete the production strain. Determination of physiological parameters The use of bioreactors allows the calculation of many process parameters like substrate uptake rates, product formation rates and yields. In biological processes, the specific growth rate is depicted by µ. The doubling time of an organism (td) is calculated with equation 2.1. ln 2 2.1 td = µ The doubling time is equal to the generation time of the cell, i.e. the length of a cell cycle. The biomass yield on a certain substrate (Ysx) of the product yield (Ysp) can be calculated according to equation 2.2 and equation 2.3 respectively. They both require the substrate uptake rate (qs) and/or the production rate (qp). µ qS q = P qS YSX = 2.2 YSP 2.3 When an organism is cultured in a continuous reactor the dilution rate (D in equation 2.4) determines the growth rate (µ = D). D = F V 2.4 Thus, by varying the dilution rate (or the feed flow rate) in a continuous culture different specific growth rates can be obtained. This allows details physiological studies of the cells at a specific growth rate that corresponds to a certain environmental condition. The substrate mass balance allows us to calculate the substrate uptake rate (qs) as can be seen in equation 2.5. D C Sin − C S 2.5 qS = CX CSin is the concentration of substrate in the feed and CS is the substrate concentration in the reactor. With the definition of the yield coefficient ( equation 2.2) and knowledge that the growth rate equals the dilution rate we get an expression for the yields calculated from the concentration of substrates and biomass. ( YSX = C ) CX in S −CS 2.6 These calculations are valuable as the concentrations can be determined via HPLC measurements. 2 | Materials and methods 33 Bacillus subtilis systems biology LST Research Project 2008 2.2.2 Precultures The preculturing consisted of two steps, the LB preculture and the M9 preculture. The contents and the dilutions of the M9 preculture medium varied according to the experimental requirements. This relates to the iron content and the glucose content. Step 1 (LB preculture ) • Add 5 mL of LB medium to a 10 mL culture tube. • Inoculate the culture tubes from the B. subtilis -80 °C stock. • Incubate for 2-4 hours at 37°C, shaking at 200-300 rpm to an OD600 of 0.3-1.0. Step 2 (M9 preculture) • Prepare the appropriate amount and content of M9 medium based on experimental requirements. • 35 mL of M9 medium to a 500 mL shakeflask. • Prepare the dilutions of the LB preculture in M9 medium. Dilutions ratio depends on experimental setup. Generally three shakeflasks were used. • Incubate the shakeflasks overnight in the shaker at 37°C, 300 rpm. 2.2.3 Shake flasks • • • • • • • • Add 35 mL of M9 medium to each shakeflask Measure OD600 of the M9 precultures. Select the shakeflask with no apparent lysis and an OD up to OD600 = 1.0. Determine the amount of M9 preculture that has to be added. Starting OD600 of the shakeflask experiment should be around 0.05. Pelletize the cells in the required amount of preculture by centrifugation at room temperature. (3 minutes at 4000 rpm). Resuspend the pelletized cells in fresh M9 medium and inoculate the shakeflasks with the resuspended cells Incubate the shakeflasks at 37°C in the shaker (2.5 cm shaking diameter operated at 300 rpm) Check the growth of the culture by every hour sampling 1 mL of broth and measuring the OD600. Continue sampling until the OD600 reaches a plateau or starts to drop. At this point, the cells have reached the stationary phase and the glucose in the medium is depleted. 2.2.4 Microtiter 96-well plates The system allows the concurrent growth of 96 individual small-scale batch cultures (19). This type of plate (Megaplate deep-well plates, Polylabo, Geneva, Switzerland) contains 40-mm-deep square wells with a cross-section of 8.2 by 8.2 mm, a rounded bottom and a total filling volume of 2.4 mL. The minireactors are run with a culture volume of 1.2 mL. When placed in an orbital 2 | Materials and methods 34 Bacillus subtilis systems biology LST Research Project 2008 shaker at 300 rpm and a shaking diameter of 5 cm, the attained oxygen transfer rates are similar to shakeflasks. A schematic view is presented in figure 2-5. Figure 2-5. A square-well microtiter plate filled with liquid medium, covered as shown, and incubated on an orbital shaker. Adapted from (19). The deep well plates require the same preculture as the shakeflasks, except for the volumes of the M9 preculture that are kept to 5 mL per dilution. Conditions depend on the experimental requirements. This procedure describes a normal non stressed growth. • Fill all the 96-wells on the plate with 1,2 mL of M9 medium • Determine the amount of M9 preculture that has to be added to obtain a start OD600 of 0.05. Usually this is about 50 µL. • Place the plate into its special holder in the incubator and run it at 300 rpm at 37 °C. • Sampling is done with a multichannel pipet and the OD is determined in an xxx plate on a machine. • Maximum sampling rate was every 45 minutes and depends on the growth phase. 2.2.5 Mini Chemostats The mini-scale chemostat experiments are used in order to obtain biomass from a culture growing in a steady state. The use of a chemostat-type fermenter is a standard method to acquire such biomass from micro-organisms. The biomass can be used for further analysis. This research employs a system that was developed in 2005 by Annik Nanchen and Alexander Schicker from the Sauer group for the use with Escherichia coli (50, 64) but by now it has also been validated for B. subtilis (69). The setup basically consists of 8 glass tubes which are fed from a medium bottle via a feed2 | Materials and methods 35 Bacillus subtilis systems biology LST Research Project 2008 pump. An effluent pump controls the removal of effluent and, by doing so, also determines the aeration rate and the mixing of the broth. A drawing of a single reactor is shown in figure 2-6. Figure 2-6. Schematic drawing of the chemostat system (50). The simplicity of the system results from a smart use of pumping systems. The effluent is taken up by a needle just above the medium level in the reactor. The pump-outflow is much larger than the amount of medium being pumped in, so when the medium level drops, the effluent needle is able to draw in air. The air is withdrawn from a needle that is placed onto the bottom of the reactor and the incoming air results in a bubble-column that vigorously mixes the culture broth on its way up. The complete setup including all tubing, bottles and water bath is shown in figure 2-7. Figure 2-7. Actual setup of the mini-scale chemostats in the lab. The reactors are placed in the waterbath (center) and the feed enters from the right side. Effluent is pumped towards the left side. The system as used in this project differs in certain aspects from the original setup from 2005. This was mainly done in order to solve problems encountered by the researchers working with it since that date (chapter 3.4 will deal with this topic in detail). The main difference is the isolation of the different reactors. Where the air supply and the feed constructions were shared in previous 2 | Materials and methods 36 Bacillus subtilis systems biology LST Research Project 2008 incarnations of the device, this time, everything has been isolated. Currently, each reactor has a unique medium bottle, non-shared feed tubing and an individual aeration/moisturizing supply. A detailed illustration of the setup for future usage can be found in Appendix D on page 92. The experiments performed with the chemostats were used to validate the system and after approval, generate biomass under stress condition for subsequent flux analysis. A complete chemostat experiment requires three days when run at a growth-rate of µ = 0.3 h-1. Lower growth-rates extend the running time. Day 1 • Start LB preculture • Prepare M9 medium (enough for: n × 10 mL reactor volume + 3 × 35 mL precultures + n × 70 mL feed volume. (n = number of reactors)). • Start M9 preculture in proper dilutions Day 2 • Heat up the waterbath to 37 °C. • Measure OD600 of the M9 precultures. Select the shakeflask with no apparent lysis and an OD up to OD600 = 1.0. • Determine the amount of M9 preculture that has to be added. Starting OD600 of the 10 mL reactor should be around 1/20th of the final OD600 (i.e. OD600 = 1 needs 0.05). Pelletize the cells in the required amount of preculture by centrifugation at room temperature. (3 minutes at 4000 rpm). • Fill the reactors with 10 mL of M9 medium. • Resuspend the pelletized cells in fresh M9 medium and inoculate the shakeflasks with the resuspended cells • Attach the aeration tubing and the effluent pump tubing • Transfer the setup to the waterbath and connect to the effluent pump. Start aeration. This batch phase takes about 3.5 hours. • Fill the medium bottles with M9 medium • When the reactors reach an density of about OD600 = 0.4, one can connect the feed. • Connect the feed tubing to the medium bottles, start the pump and wait until the medium reaches the needles on the reactor side. • Connect the feed needles to the reactors and adjust the harvest needles to the proper volume level. • The chemostat is now running in continuous mode. Day 3 • When six volume changes have been pumped through the reactors you can disconnect the needles and remove the reactors. 2 | Materials and methods 37 Bacillus subtilis systems biology • • • LST Research Project 2008 Measure the steady state OD of the reactors and centrifuge the broth (3’ at 10000 × g). Sample the supernatant and keep the cell pellet. Supernatant can be used for further HPLC measurements and is stored at -20 °C. The cell pellet is washed twice in isotonic wash buffer. (centrifuged at 15000 rpm) Rinse glasswork and tubing afterwards with ethanol (70%) and H2O. 2.3 Analytical procedures 2.3.1 Optical density determination Optical density of the cultures grown in the chemostat and shakeflasks was determined at a wavelength of 600 nm with a Pharmacia Novaspec® II VIS photospectrometer. The samples were diluted in physiological salt to reach a maximum linear absorbtion of 0.250. To asses the optical density of the 96-well plates, a light scatter device was used. Samples were diluted in physiological salt in a microtiter plate to obtain a maximal light scatter intensity of 0.300. The correlation between the lightscatter signal and the actual culture density was determined with a dilutions series of biomass from a mini-chemostat (resulted in equation 2.7). OD = 2.7402 ⋅ lightscatter - 0.0345 2.7 2.3.2 Cell dry weight determination Nitrocellulose filters with a pore size of 0.45 µm were used to determine the cell dry weight (Sartorius AG Germany, Cellulose Nitrate Filter 0.45 µm, type 1130647-N). The filters were tared after having been dried overnight in an oven at 85 °C and a subsequent cooling in a vacuum excicator for 30 minutes. The biomass from 9 ml of culture was collected on the filter and washed with 10 ml of demineralised water before being dried and tared as indicated above. Reference filters were rinsed with 9 mL of M9 medium instead of culture broth. 2.3.3 HPLC analysis A High Performance Liquid Chromatography RID/DAD system was used to determine the glucose and acetate levels in the samples derived from the chemostats. The machine was an Agilent/HP 1100 with an online degassing unit equipped with an AMINEX HPX-87H (BioRad, CA, USA) column and a column oven controlled by a µ oven controller model KU-OV with Eurotherm 2132 PID. The oven temperature was set to 60 °C. The mobile phase consisted of 5 mM sulfuric acid at a flowrate of 0.6 mL per minute. The operating pressure was about 51 bar. 2 | Materials and methods 38 Bacillus subtilis systems biology LST Research Project 2008 The samples from the -20 °C freezer were thawed out, and 200 µL was transferred into an HPLC vial. An auto-sampler injected a volume of 20 µL per sample into the column preceded by a needle wash with ddH2O. The column-flow was analyzed for acetate by a diode array detector at a wavelength of 210 nm. Glucose was detected by a refractive index module. An HPLC standard with succinate, fumarate, glucose, acetate and malate was used to validate the retention times. 2.3.4 GC-MS analysis Cell pellets from the chemostat experiments, stored at -20 °C in the freezer, were used for analysis by a Gas Chromatography / Mass Spectrometry combination. The method was based on the most recent description by members of the Sauer group (20). The setup consisted of an Agilent 6890 N Network gas chromatograph connected to an Agilent 5973 inert mass selective detector. The gas chromatograph was equipped with a Rtx-5Sil MS capillary column with a length of 30 m, an inner diameter of 0.25 mm and a film thickness of 0.25 µm (Restek, PA, USA). A mps 2 autosampler from Gerstel (Mühlheim an der Ruhr, Germany) handled the sample injection. GC-Method 1 µL of each sample was injected at a 1:10 split mode. The needle was washed three times in pyridine before and after each injection. The carrier gas helium was set at a column flow of 1.5 mL/min and the solvent delay was 2.7 minutes. The (glass wool) liner had a temperature of 230 °C and the transfer line was kept at a temperature of 280 °C. The temperature gradient in the GC was: 1 minute at 160 °C, ramp to 310 °C with 20 °C/min, hold 0.5 minutes at 310 °C. MS-Method The mass selective detection was done with a single quadrupole instrument with an EI ionization. The EI source temperature was 203 °C with an ionization energy of 70 eV. Detection was done in a full scan mode in the range of 180-550 m/z with 3-4 scans per minute. Sample preparation The biomass samples needed hydrolization and derivatization before GS-MS analysis. The procedure is described below: Hydrolysis • Thaw out the washed cell pellets from the -20 °C freezer • Resuspend cell pellet in 200 µL 6 M HCL and transfer the tubes to a sealable rack. 2 | Materials and methods 39 Bacillus subtilis systems biology • • LST Research Project 2008 Bake the sealed tubes overnight in an oven at 105 °C. Dry the hydrolyzate at 85 °C in a thermoblock in a fume hood. This will take approximately 5 hours. Derivatization • Add 30 µL of dimethylformamide (DMF, Merck Darmstadt, Germany) to the dried hydrolyzate • Roll the tube gently until the DMF gets a light brown color and transfer the solution to a new tube. • Add 30 µL of N-tertbutyldimethylsilyl-N-methyltrifluoroacetamide with 1% tertbutyldimethyl-chlorosilane (TBDMSTFA, Merck Darmstadt, Germany) and seal well • Incubate the tube, well sealed in a thermomixer at 85 °C for one hour. • Transfer the derivatized sample to a GC-MS vial and seal well. For validation purposes, the same procedure can be followed with 15 µL of an amino acid standard (1mM per amino acid, BioChemika , MO, USA). The sample is injected 5 times and the theoretical mass distribution is compared to measured mass distribution. The software FiatFlux includes a routine to perform batch-wise comparison on all amino acid fragments. 2 | Materials and methods 40 Bacillus subtilis systems biology LST Research Project 2008 3 Results 3.1 Growth medium optimization log(OD600) In order to perform the iron limited cultivations it had to be ensured that the cultures were not under iron stress already. When running a batch culture something appeared to be repressing the growth as in initial stages, the growth curve showed a notable speed-up effect during the experiments (figure 3-1). time Figure 3-1. Diagram of growth-rate lag issue in batch cultures. The solid line is the expected outcome; the dashed line shows the experimental outcome. The slope of a line represents the growth-rate. Next to this speeding-up the group of John Helman had discovered that the Fur regulon was activated in the initial growth phase (from personal communication). This indicated a response to low iron availability. Because the medium contains sufficient amounts of iron for B. subtilis to grow, there should be a different reason to account for the observed effect. It is known that B. subtilis is not able to transport the iron directly over the cell membrane and requires transporters to do so (see topic on siderophores in chapter 1.4.2). Apparently the iron supply was not optimal in these cases. It could well be that the production of siderophores takes time and the maximum growth rate is not reached immediately. The growth would speed-up after the iron carriers are generated and brought into action. If the growth medium already caused an iron stress it would be harder to distinguish between these conditions (as a control) and the future iron limitation experiments. To resolve the influence of the start-up phase of siderophore production an external chelating agent (iron binding compound) was added. External chelators can offer pre-bound compounds (like iron) to the cell that is suitable for immediate uptake. This would increase the intracellular iron availability and possibly solve the observed growth lag. A well known exogenous iron chelator, citric acid (citrate), was proposed as an addition to the medium (55). The addition of citrate is not a standard procedure for B. subtilis media 3 | Results 41 Bacillus subtilis systems biology LST Research Project 2008 according to (31). Figure 3-2 shows the molecular structure of citrate and the known iron-bound conformations. The exact configuration depends on the conditions, but its occurrence remains unpredictable. Figure 3-2. Citric acid (above) and the encountered complexes with iron (below). Adapted from (55). This hypothesis was tested in shakeflask experiments. This is a small scale batchreactor that facilitates a high sampling rate under optimal growth conditions. The detailed analysis of the growth-curves could aid in establishing a growth medium that is the most optimal for B. subtilis. To analyze the possible iron limitation during initial growth phase, the effect of different citrate concentrations on the growth was determined. Initial experiments were used to optimize the growth conditions. The adjustments included: • Premixing the iron and citrate before adding to the M9 medium in order to prevent the precipitation of salts. For the same reason, the FeCl3 was added to M9 done just before the medium was used. • Optimizing preculture dilution rates to obtain the most viable cells for the subsequent shakeflask experiments. A dilution series of 2500×, 3500× and 4500× was sure to produce a culture overnight in exponential growth. • Optimizing preculture transfer by minimizing the non aerated/heated/shaken periods. It was observed that the use of cell pellets that had been outside the medium for two hours reached a growth-rate of half the normal rate. • Acclimatizing the B. subtilis to the availability of citrate in the M9 precultures. This reduces the shock when transferred to the final shakeflasks. • Keeping the iron/citrate mix in the fridge and out of the light. This prevents oxidation. • Increasing the start OD to about 0.05 which allows the experiments to be performed in one day. Lower densities result in a longer start-up phase. 3 | Results 42 Bacillus subtilis systems biology LST Research Project 2008 After the determination of the appropriate experimental setup an evaluation was done with a range of citrate concentrations in the medium. This shows how B. subtilis performes under low and high citrate availability. The results of the citrate variation are shown in figure 3-3. 3 2.5 OD600 2 1.5 1 0 µM 50 µM 100 µM 50 µM + 250 µM 0.5 0 0 2 4 6 8 10 Time [h] Figure 3-3. Development of B. subtilis growth in shakeflasks with varying citrate concentrations. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. All media contained 50 µM Fe3+. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. One wants to find out if citrate is really enhancing the growth of B. subtilis. Figure 3-3 shows a distinction between the non-citrate grown cultures and the ones where citrate is added. All citrate cultures have a reduced lag phase and start their exponential growth-phase before the culture in regular medium (0 µM) does. The citrate-supplied cultures also reach a higher overall density. There is a small difference between the individual citrate media and their impact on growth. It appears that the higher the citrate concentration, the earlier the onset of exponential growth. A clear effect on the start-up phase. The impact on growth rate during the development of the experiment is shown in figure 3-4. 3 | Results 43 Bacillus subtilis systems biology LST Research Project 2008 1 -1 Growth rate [h ] 0.8 0.6 0 µM 50 µM 100 µM 50 µM + 250 µM 0.4 0.2 2 3 4 5 6 7 Time [h] Figure 3-4. Growth-rates of B. subtilis in shakeflasks experiments with varying citrate concentrations in the medium. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. If the citrate was really stimulating the onset of growth this would mean that the maximum growth rates would be obtained earlier in the experiments. Figure 3-4 shows that the citrate supplied cultures reach their maximum growth-rate earlier than the ones grown in regular medium without citrate. The maximum growthrates are not only reached earlier on but also maintained at a stable level until the end of the experiment where glucose is depleted. A determination of the maximum growth rates during the exponential phase requires a detailed look at the period between OD600 = 0.1 and OD600 = 1.5. This phase is shown in figure 3-5. 3 | Results 44 Bacillus subtilis systems biology LST Research Project 2008 1 0.5 0 log(OD600) 1 2 3 4 5 6 7 -0.5 -1 0 µM 50 µM 100 µM 50 µM + 250 µM -1.5 -2 -2.5 Time [h] Figure 3-5. Culture densities of B. subtilis grown under varying citrate conditions plotted on a logarithmic scale. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. All media contained 50 µM Fe3+. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. To study the maximum growth-rates and their onset one can plot these on a logarithmic scale. A clear exponential growth would show up as a straight line. Figure 3-5 illustrates the variety in onset of the exponential phase. The low citrate media enter this phase with a delay. The minor differences between the individual curves are hard to distinguish by eye but further analysis of figure 3-5 yielded the maximum growth rates. Results are shown in table 3-1. Table 3-1. Maximum growth rate of B. subtilis in shakeflasks under citrate varied conditions. Data derived from figure 3-5. Condition Growth-rate (h-1) Correlation 0 µM citrate 0.61 0.998 50 µM citrate 0.67 0.999 100 µM citrate 0.69 1.000 50 µM citrate + 250 µM citrate 0.66 0.999 The consequence of citrate addition on the maximum growth-rate as shown in table 3-1 indicates a positive effect of citrate in the medium. Higher citrate levels lead to a higher maximum growth rate. The supply of a significantly higher amount of citrate does not appear to affect the growth even more. Exploratory experiments already suggested that there is an optimum in the effect of citrate at a concentration of about 200 µM citrate (data not shown). 3 | Results 45 Bacillus subtilis systems biology LST Research Project 2008 These results lead to an alteration of the standard M9 minimal medium composition for future experiments. It was decided to use a 100 µM citrate concentration in the medium. Using a higher concentration would not improve the growth performance much, and a low citrate concentration would prevent the interference with the analyses on labeled substrates. This is important as citrate can be metabolized by B. subtilis and can cause unwanted labeling patterns in the amino acids. 100 µM of citrate would account for 1.6% of the total carbon in M9 medium. 3.2 Induction of stress responses To expose micro-organisms to an environmental stress in an experimental setup, one needs to use compounds that induce the required stress levels. For this research there are two situations of interest, the environment with oxidative stress and the one where low iron concentrations influence the growth of B. subtilis. Previous research showed that there are two commonly used methods to achieve these conditions. Oxidative stress In order to research the effects of oxidative stress many groups have been using the compound Paraquat (methyl viologen dichloride), a well-known herbicide. In target cells, Paraquat undergoes redox-cycling which may contribute to its toxic actions. Several NADPH oxidases have been identified as potential inducers of paraquat redox cycling including cytochrome P450 reductase and nitric-oxide synthase (29). These enzymes generate a reduced Paraquat radical that can act as an electron donor (see figure 3-6). Figure 3-6. Paraquat redox-cycling mechanism (29). Reacting rapidly with molecular oxygen, the Paraquat radical recycles back to Paraquat and in the process forms the highly toxic oxidants including the superoxide anion, hydrogen peroxide and hydroxyl radicals. These compounds can cause major cellular damage and as they are also generated at the expense of NADPH, the process creates an extra energy stress for the organism. 3 | Results 46 Bacillus subtilis systems biology LST Research Project 2008 Paraquat can only be degraded by a few micro-organisms in a photolytic process (9) and B. subtilis is not able to break it down. The ROS generating process will thus be recycled and remain active throughout the experiment. The application of Paraquat with B. subtilis is not new an has already been proven to increase the intracellular production of superoxide (11). Transcriptional profiling also indicated that the general stress response factor σB and the peroxide sensor PerR are induced after the administration of Paraquat (49). Iron limitation To obtain iron limited conditions in the chemostats all iron has to be removed from the medium. Solely omitting iron from the medium will not suffice, as iron is present as a contaminant in many chemicals. Furthermore, glassware is a notorious ion exchanger and can contain many iron molecules. The removal of iron can be achieved by the addition of an iron chelator, this is a compounds that binds iron. For this research, the iron chelator deferoxamine (DFO) is used (figure 3-7). This is virtually the only iron chelator currently in pharmacological use to overcome iron intoxication. Figure 3-7. The molecular structure of deferoxamine (5). Deferoxamine was originally isolated from Streptomyces pilosus and chelates iron by forming a stable complex with an iron atom. The complex prevents the iron from entering into further chemical reactions. Deferoxamine has a high affinity for ferric iron with Kaff = 1031. This is about the same as B. subtilis’ sideophore bacillibactin (Kaff = 1030). Deferoxamine extracts iron from ferritin but not readily from transferrin and not at all from cytochromes. Theoretically, deferoxamine is capable of binding about 8.5 parts by weight of ferric iron (14). 3.3 Stress dosage determination To make an educated guess on the precise dosage-effect of Paraquat and deferoxamine on B. subtilis growth in later chemostat runs, the effects were explored in 96-well microtiter plates. The results of the 96-well plates can be used to create a dose response curves which would allow the determination of useful growth condition in subsequent chemostat experiments. A probable growth curve could look like figure 3-8. 3 | Results 47 LST Research Project 2008 Maximum growth rate Bacillus subtilis systems biology Imposed stress Figure 3-8. Example of a dose response curve. The colored area indicates the soughtafter growth conditions. The area of interest is the one where a significant impact on growth is observed but where the B. subtilis still succeeds to survive. Growing the chemostat cultures under these conditions would probably mobilize the stress-response in a similar way. A literature study gave insights in the probable growth conditions of B. subtilis under Paraquat and deferoxamine stress. For example, 80 µM of Paraquat was known to result in the transcription of genes that contribute to stress resistance (71) and 800 µM of Paraquat resulted in growth defects (54). For iron deprivation, concentrations 5 µM to 30 µM deferoxamine had proven to be effective (1, 72). The conditions, tested for this research project, are compiled in table 3-2. 3 | Results 48 Bacillus subtilis systems biology LST Research Project 2008 Table 3-2. Concentrations of stressors tested to obtain a dose response curve. Values marked in bold-face are the indications derived from literature. Stressor Concentration (µM) Paraquat 0.5 1 5 10 20 50 100 150 250 500 1000 5000 10000 Deferoxamine 3 5 10 15 20 30 40 60 80 120 240 480 The medium used in the Paraquat tests was the exact same formulation as had been in use after the shakeflasks experiments. This contained 50 µM of FeCl3 and 100 µM of citrate. The deferoxamine experiments were performed in the same medium but without the addition of FeCl3. The biomass for the latter experiments was also obtained from M9 precultures without added iron. All controls were grown and precultured on the M9 medium established after the shakeflasks experiments (for composition see 2.1.2). The 96-well microtiter plates were run with varying amounts of deferoxamine and Paraquat to determine the required growth conditions for chemostat experiments. Initial attempts allowed the final fine-tuning of operational concentrations and quickly showed that B. subtilis is not able to grow in conditions where much Paraquat is present per cell. It was therefore not suitable to administer Paraquat at the start of the batch phase. Paraquat effects could only be determined when added after the onset of experimental growth when enough biomass has been generated. Another observation from the microtiter cultivations was the decreased maximum growth rate of the control cultures. The growth-rate did not exceed the µmax = 0.53 h-1. The fact that microtiter experiments yield lower maximum growth rates than shakeflasks is also a well known observation from other experimenters. 3 | Results 49 Bacillus subtilis systems biology LST Research Project 2008 Oxidative stress / Paraquat To explore the effect of Paraquat on B. subtilis growth, s wide range of Paraquat concentrations were tested in batch cultures. The resulting growth curves from these experiments are shown in figure 3-9. 4.5 4 50 µM 100 µM 150 µM 250 µM 500 µM 1 mM 5 mM 10 mM Control 3.5 OD600 3 2.5 2 1.5 1 0.5 0 0.0 2.0 4.0 6.0 8.0 Time [h] 10.0 12.0 14.0 Figure 3-9. The effect of varying Paraquat concentrations on the growth of B. subtilis in a batch culture. The dotted line indicates the moment of Paraquat administration. All curves are averaged from three independent wells. Glucose concentration in this experiment was 3 g/L. The growth-curves in figure 3-9 show a division from the moment of Paraquat addition. Only the 50 µM and the 100 µM regain growth (albeit slowly for the 100 µM) while the others show no increasing density and even lose density after a few hours. Notable is the maximum OD600 achieved by the control cultures, an OD600 normally approaches a value of 2.5 in this medium. With the results of this plate, subsequent experiments were planned to check the lower range of Paraquat concentrations (0 – 150 µM Paraquat). At these concentrations all cultures remained viable and resumed growth after the administration of Paraquat. After the determination of the maximum growth rate for each condition, these were plotted against the Paraquat concentration in figure 3-10. 3 | Results 50 Bacillus subtilis systems biology LST Research Project 2008 -1 Maximum growth rate [h ] 0.6 0.5 0.4 0.3 0.2 0.1 0 0 50 100 Paraquat [µM] 150 200 Figure 3-10. Dose response curve of B. subtilis. Maximum growth-rate as a function of the applied paraquat concentration. Dotted line represents the hand-drawn trend line. Data derived from at least 6 individual wells per data point. The plot of figure 3-10 clearly shows the relation between the added amounts of Paraquat and the resulting growth. Above 150 µM Paraquat there is no observed growth and at lower concentrations the growth-rates approach the µmax of the control cultures. It appears as if the growth-rates in the lower Paraquat range (0.5-10 µM) accumulate just above the 0.4 h-1 point. Iron limitation / deferoxamine The experiment for deferoxamine was set-up in a way that the plate contained cultures grown with 0 to 50 µM FeCl3 and cultures with from 10 to 80 µM deferoxamine. All of them were precultured in a non-iron containing M9 preculture. The control culture was grown under normal iron supplied conditions. The cell density during the experiment is shown in figure 3-11. 3 | Results 51 Bacillus subtilis systems biology LST Research Project 2008 3 Control Fe 50 µM Fe 10 µM Fe 0 µM DFO 10 µM DFO 20 µM DFO 40 µM DFO 80 µM 2.5 OD600 2 1.5 1 0.5 0 0 2 4 6 Time [h] 8 10 12 Figure 3-11. Growth of B. subtilis in microtiter plates in media with varying iron or deferoxamine (DFO) concentration. All curves are averaged from at least eight independent wells. The growth curves of the cultures with altered conditions in figure 3-11 show an apparent contrast with the control situation. All of the tested conditions show a delayed onset of exponential growth. Another observation is that deferoxamine appears to stimulate the growth of B. subtilis. The cultures with no added iron (Fe 0 µM in the graph) are lagging behind the cultures with deferoxamine added. Deferoxamine even seems to improve the cultures when comparing to B. subtilis with a sufficient amount of iron available (Fe 50 µM). There are no major differences between the various amounts of deferoxamine added. All of them seem to follow the same curve. The lack of data points in this last experiment toughens the possible interpretation of the maximum growth rate. Results from an earlier deferoxamine microtiter experiment (data not included) showed that the growth-rates of cultures with a variety of deferoxamine concentrations were all close to each other at about µmax = 0.52 h-1. 3.4 Continuous cultures The final analysis of the adaptations to stress environments in B. subtilis cells was done with biomass from the mini-scale chemostats. These chemostats would run for at least 6 volume changes to obtain steady state and grow on a glucose concentration of 1 g/L. These conditions were verified with B. subtilis in previous research (50, 69). This research also showed that the cultures are not oxygen limited in this system at dilution rates ranging from 0.05 to 0.7 h-1. The expected growth in this system would result in an optical density of about 0.9 (69). The chemostats were initially tested in order to verify the accuracy of the system as it 3 | Results 52 Bacillus subtilis systems biology LST Research Project 2008 had fallen into disuse for some time. Initial runs were performed for a comparison of data with results from previous researchers. The first run was performed with the original material except for the feed pump-head tubing recently bought. The experiment used all eight reactors at four different dilution rates and was carried out for 70 hours after a batch phase of 6 hours. At this point all reactors would have been running for at least 6 volume changes. The substrate concentration in the chemostats was 1 gram glucose per liter. The density was measured to obtain an understanding how growth had developed inside the reactors. The results are plotted in figure 3-12. 1.4 1.2 1 0.114 OD600 0.114 0.8 0.198 0.198 0.294 0.6 0.294 0.450 0.4 0.450 0.2 0 0 20 40 60 80 Time [h] Figure 3-12. First chemostat run of B. subtilis with the original setup. All eight chemostats used the same medium and were inoculated from the same preculture. Four different dilution rates were tested in duplicate. Timepoint 0 denotes the feed start. The results show major fluctuations after the start of the feed. The intended moment of steady state, at 11 to 44 hours (depending on the dilution rate), displays a range of densities. These vary from 0.12 to 1.02. There appeared to be a correlation between the dilution rate and the end densities found in the reactors. Both cultures grown on D = 0.45 and 0.198 end up close to each other. This first experiment showed the formation of gas bubbles in the feed tubing. This caused a non-continuous feed and might have resulted in the varying densities observed in figure 3-12 above. The original setup used a type of silicone tubing with a relatively large inner diameter and a thin wall which could facilitate gas diffusion into the tubing. The tubing was therefore replaced by a different type with a thicker wall and a minimal inner diameter, a form of tubing regularly used in HPLC machines. A comparable experimental setup, with the flowrates in the same range as before resulted in figure 3-13. 3 | Results 53 Bacillus subtilis systems biology LST Research Project 2008 1 0.9 0.8 0.13 0.12 0.24 0.25 0.26 0.26 0.7 OD 600 0.6 0.5 0.4 0.3 0.46 0.50 0.2 0.1 0 0 10 20 30 40 50 Time [h] Figure 3-13. Second chemostat run with new feed tubing. B. subtilis grown in eight chemostats at different dilution rates. All eight chemostats used the same medium and were inoculated from the same preculture. Four different dilution rates were tested in duplicate. Timepoint 0 denotes the feed start. The observed stable densities in the experiments above might be ascribed to the more stable feeding in this experiment. The feed flow did not contain any gas bubbles anymore. This was a problem faced by many of my predecessors. The effect of the new tubing was that only two reactors appeared to perform well where the others drop in density to values in the lower regions of the graph. The expected growth would be at an end density of about 0.9, a value determined by the amount of glucose in the medium and the known biomass yield on glucose and already shown in literature (69). The reactors with dilutions rates of 0.24 and 0.46 appear to reach proper densities as could have been expected from earlier studies. It was clear that the chemostat setup was still not functioning as desired and the following aspects were systematically tested to optimize the system for the purpose of this study. Therefore twelve chemostat experiments were performed, a list of them is shown in table 3-3. Table 3-3. Overview of adaptations and checks performed in order to obtain a properly functioning mini-scale chemostat. System change / check Reason / result Fine-tuning of flowrates in feed More exact determination if the feed tubing flow aids in the result interpretation. The feed flow was weighed after a defined period of time. Determine if feed flow remains Check the application life of the feed constant over longer periods of time tubing. Could they be used over and 3 | Results 54 Bacillus subtilis systems biology Standard determination of tubing characteristics before and after the experiment Apply more sterile working methods Control of exact reactor volumes during the experiments Usage of a smaller range of dilution rates to pinpoint the problem Fine-tuning the batch phase by decreasing this period from 6 to about 3 hours Determining the exact volume inside the reactor. Create streak-out plates from different parts of the system. Needles, tubing, reactors. Determining the effluent flow (liquid) Determining the effluent flow (air) Check the pH of the reactors 3 | Results LST Research Project 2008 over? See of the flow per tubing remained constant between the individual experiments. General precaution to avoid infections. The setup was put together in a laminar flow cabinet. This aided in the sterile setup as connecting the tubing is quit tricky and prone to infections. This could now be done in a sterile environment. By exactly positioning the effluent needles the volume of the reactor remains the same. Attention should be paid to the bubble formation as they may rise up to a centimeter above the medium level. A decrease in volume results in a higher dilution rate and could result in the washout of a culture. By using only one dilution rate it was easier to tell whether or not the problems were due to a difference in feed-flows. The dilution rate was also chosen for the system to reach steady state within one day after feed start. (6 volume changes) Longer batch phase may yield cells in late exponential phase. This could influence the adaptation to a continuous feed-regime. General improvement of the setup. Allows more accurate result interpretation. Determination was done by measuring the weight of the reactor before and after the experiment. The volume at the start was known and final volume could be determined with the change in weight and the known culture density. See if there are any infections in the system. It might have been that infections could have taken place in the media of in the reactors. Check to see if the same volume of medium is pumped in as is being pumped out. In order to check for sufficient aeration, the air, pumped through the reactors was measured. This rate should be about 20 mL/min to obtain proper aeration and mixing of the broth. A change in pH can indicate whether or not the reactors are oxygen limited and 55 Bacillus subtilis systems biology Change of needles Removal of congestions in the tubing Addition of feed intake filters Recreating all stock solutions Reactor cleaning with soap, just water, or ethanol and water Cleaning of tubing in the system Refreshing the septum stoppers Glucose level check in all reactors and feeds with HPLC. LST Research Project 2008 the bacteria start to produce byproducts of the central carbon metabolism. Needles wear out after many sterilizations above the flame, this results in bad aeration by a deformation of the bubbles that exit the needle. At later stages this was omitted as no flame sterilization was needed. The later setups were all composed in the laminar flow cabinet. The tubing was often clogged by silicon or metal pieces from tubing and connectors. To overcome the clogging problems. In order to obtain fully sterile media after some problems with infections Remainders of soap in the glass tubes might influence the growth. Remainders of old chemostat run might influence future experiments Old punctured septi might not prohibit infections as well as new ones and the could leak are which deregulates the mixing of the reactor To see if there is a difference in glucose distribution in the various tubes. Of all these ideas, the following actions had a positive effect on the working of the chemostat and were applied in subsequent experiments. • Determination of the exact feed flowrates. Important for the final interpretation of the resultss and to see whether the tubing stays in a good condition throughout the experiments. • Connecting the system’s individual part in a laminar flow cabinet • Exact control of the reactor volume by correct placement of the effluent needle and correcting the reactor volume afterwards by measuring the weight of the hungate tube. • A short batch phase. Within three hours after the start of the batch phase the feed was started. This was mostly at OD600 of 0.3 to 0.4. This was to be sure that all cells were still in the exponential growth phase. • The cleaning of the system with, first ethanol and then with nanopure water. No soap was used. This holds for the tubing as well as the glassware. • Renewal of the septum stoppers of the hungate tubes and the medium bottles. After 4 runs these were generally completely punctures and led to aeration leaks. After the many checks and alterations, the results remained unusable. None of the attempts gave any insight into the actual solution of the problem. Figure 3 | Results 56 Bacillus subtilis systems biology LST Research Project 2008 3-14 shows an overview from the second to eleventh run and gives an impression of the variety of densities obtained in steady state per reactor. Each reactor was connected in the exact same way each time, grown on the same medium and at a dilution rate which should yield an end density of about 0.9. 1.4 1.2 OD600 1 0.8 0.6 0.4 0.2 0 0 0.1 0.2 0.3 0.4 0.5 0.6 -1 Dilution rate [h ] Figure 3-14. Overview of end densities as measured after six volume changes arranged at the various dilution rates used in the experiments. Data obtained from eight runs, each run is indicated by its own color. The reactors were grown at different dilution rates but within a range that should produce the same end density. The resulting densities show a major diversion between the experiments. The observed variation from the first experiments was still found in later runs and the adaptations and check did not yield a constant density after 6 volume changes. Figure 3-14 shows the apparent randomness in the outcome of the cultivations. A reason for this randomness might be the fact that during these runs, many factors were changed in order to obtain a functional setup. When in the last two runs of this result-set, only a minor issue in the batch phase was adjusted, the extracted data showed more similarities. The final densities of the reactors are shown in figure 3-15. 3 | Results 57 LST Research Project 2008 1.4 1.2 OD600 1.0 1.2 8 1.0 8 0.8 0.8 6 0.6 7 0.4 5 7 6 0.2 0.0 1 3 1.4 1 3 0 0.10 4 0.20 0.30 2 0.4 4 2 0.40 0.6 5 0.50 0.10 0.20 OD600 Bacillus subtilis systems biology 0.2 0.30 0.40 0.0 Dilution rate [h-1] Figure 3-15. The distribution of densities over the dilution rates after 6 volume changes in two successive runs. Numbers indicate the specific reactor tube. The apparent randomness is still noticeable when individual runs are inspected. If both runs are considered, there is a remarkable resemblance between the relative locations of the points. Although the dilution rate is slightly different in both cases, it is clear that reactors 5, 6 and 7 are always clustered together in the bottom left corner of the diagram. The same identification can be done for the other reactors. This shows that there is some sort of regularity between the runs. Another detail comes from the reactors 1, 3 and 8 in this setup that tend to grow well in both cases. Further observations indicate the specific distribution of the reactors. The reactors are always connected in pairs, 1 to 2, 3 to 4 and so on. The diagrams show that it is constantly one of this pair that reaches proper growth. The formation of pairs is a result of the coupling to the feed pump. A drawing of the connection of each reactor to the medium bottle is shown in figure 3-16. Medium Pump head 1 2 3 4 5 6 7 8 Figure 3-16. Split scheme of the feed tubing from the medium bottle through the pump to the individual reactors on the right. The exact pairs that can be seen in the density distributes are in fact coupled to each other in the final split of the feed tubing before the pump. It was known from previous experiments that a blockage in either of the channels could result in the withdrawal of medium from the malfunctioning tubing parts. If, during the chemostat run, the pressure inside the tubing would differ between the channels, due to local differences in the materials used, a possible variation in flow could be the result. 3 | Results 58 Bacillus subtilis systems biology LST Research Project 2008 All indications pointed to the feed system of the chemostat and in order to overcome the problem of varying end densities as quick as possible, the splittubing was replaced by a system of individual feed tubes connected to isolated medium bottles. After two comparative chemostat experiments, the system proved more constant. Growth in all reactors approached expected values and none of the cultures washed out. The result of these comparative chemostat experiments is shown in figure 3-17. 1.4 1.2 OD600 1 0.8 0.6 0.4 0.2 0 0.1 0.15 0.2 0.25 0.3 -1 Dilution rate [h ] Figure 3-17. Final optical densities after 6 volume changes of reactors connected to individual feed tubing. Figure based on two chemostats experiments. The system was ready for application in the stress imposed chemostat experiments. It is there that the new system would prove to be reliable and reach a 94.4% success rate. Control cultures reach a constant OD600 of about 1.05 ± 0.16 at a glucose concentration of 1 g/L. 3.5 Continuous cultures under limiting conditions The oxidative stress and iron limitation conditions in steady state growing cultures could be studies after the chemostat setup had been freed from its problems. The limitation experiments were performed by combining the M9 medium from the shakeflasks experiments and the dose response data from the microtiter plates. All stress experiments were executed at a dilution rate of 0.29 h-1 (± 0.02) and performed on [1-13C] and [U-13C] labeled glucose as substrate. This substrate is required for resolving the metabolic ratios from amino acid composition afterwards. The growth-rate was based on the data from the microtiter experiments. The dose response curve of these experiments indicated that B. subtilis would be able to survive at this growth-rate under Paraquat induced oxidative stress (dose response in figure 3-10 on page 51). The stress response experiments were performed on the same medium as before, a 1 g/L glucose medium and the measurements were performed after 6 volumes 3 | Results 59 Bacillus subtilis systems biology LST Research Project 2008 changes. After six volume changes the bacteria should be adapted to the continuous culturing (69). The oxidative stress experiments were performed by starting the reactors under normal conditions. A batch phase and a short continuous phase. After two volume changes, the paraquat was added to the medium and was slowly introduced to the culture. This was done to minimize the impact of the switch to a continuous growth regime resulting from two major changes. The biomass was harvested 6 volume changes after the addition of Paraquat. Paraquat was added in different concentrations. The endconcentrations in the reactors were 1, 5, 10, 15, 20 , 25 and 40 µM. Iron limitation was imposed by administration of a high dose of deferoxamine. The dose response curves had shown that even the highest tested levels of deferoxamine did not result in growth limitation. The chemostat experiments were therefore executed with the highest amount of deferoxamine available (3 mM per reactor) to be effective. The deferoxamine was added to the medium before the reactor start. Next to the iron withdrawal by deferoxamine, chemostats were also tested with an M9 medium lacking iron. All iron limitation cultures were precultered and grown in these ironless media. Every experiment also included two reactors grown under normal conditions (media with iron and no stress inducers) as a control. Integration of all chemostat results yields the physiological data as described in table 3-4. Table 3-4. Biomass yields and specific glucose uptake rates for B. subtilis grown under oxidative stress (Paraquat) and iron limitation (deferoxamine) at a growthrate of µ = 0.29 h-1 (± 0.02). Column ‘n’ denotes the number of reactors used in the calculation. Biomass yield Specific glucose uptake rate n (mmol · g-1 · h-1) (gcdw · g-1) Control 0.42 ± 0.08 5.49 ± 1.15 16 Paraquat (15 µM) 0.41 ± 0.08 5.57 ± 1.17 6 Deferoxamine (3 mM) 0.53 ± 0.04 4.15 ± 0.27 4 The results show virtually no difference in biomass yield between the controls and the Paraquat samples. The effect of deferoxamine is more evident. In that case, the biomass production per consumed glucose increases significantly. The uptake rates of Paraquat and the controls show the same pattern, both are alike and it is again deferoxamine that differs with a much lower glucose uptake rate. Oxidative stress The oxidative stress response was subject of seven chemostat experiments. Of the concentrations tested in these experiments (1, 5, 10, 15, 20, 25 and 40 µM) the 15 µM showed to yield the best cultures. Lower and higher concentrations 3 | Results 60 Bacillus subtilis systems biology LST Research Project 2008 have resulted in washouts of the system. These cultures did not achieve a steady state and were visually decreasing in density every hour. To check whether the cultures were truly growing under glucose limited conditions the reactor broth was analyzed for extracellular compounds via HPLC. When B. subtilis does not grow glucose limited, it will produce acetate and that would indicate a problem with the reactor setup. The results of this analysis showed that none of the reactors were producing acetate. Was found though, was the formation of fumarate. All reactors exposed to paraquat showed extracellular fumarate. When the paraquat concentration is increased, the amount of fumarate rises. This is shown in figure 3-18. 70 Fumarate [µM] 60 50 40 30 20 10 0 0 5 10 15 20 Paraquat [µM] Figure 3-18 Extracellular fumarate concentrations in B. subtilis Paraquat reactors. Data derived from single reactors. The investigation in the acetate levels yielded an observation of fumarate production. Another observation was the presence of additional byproducts at paraquat levels above 15 µM. In reactors with these higher concentrations the production of succinate, lactate and acetate was found. Next to the analysis with the HPLC, the retrieved biomass was put through GCMS analysis and subsequent calculations in FiatFlux (77). FiatFlux is a piece of Matlab code that is able to analyze the distribution of labeled carbon atoms, stemming from the fed glucose, to their final destination in the amino acids. This allows the determination of flux ratios in the metabolic network. All grown chemostat cultures were analyzed and the in vivo fluxes through major pathways of the central carbon metabolism assigned. Resulting data from the control cultures and the paraquat cultures is shown figure 3-19. 3 | Results 61 Bacillus subtilis systems biology LST Research Project 2008 0.8 Control PQ 10 uM PQ 15 uM PQ 20-25 uM 0.7 Relative flux ratios 0.6 0.5 0.4 0.3 0.2 0.1 0.0 PEP through PEP through PPP (ub) TK (ub) OAA from PYR PEP from OAA PYR from MAL (ub) PYR from MAL (lb) SER through glycolysis Pathways and reactions Figure 3-19. Flux ratios in B. subtilis under increasing oxidative stress. Fluxes are composed of data from different reactors under equal conditions. 10 µM: 5 reactors, 15 µM: 7 reactors, 20-25 µM: 2 reactors. The errors indicate deviations between various reactors. PQ: Paraquat. Flux analysis is a great tool to determine if there are any changes in the organization of the central carbon metabolism as a result of imposed stress conditions. The flux rations in figure 3-19 show that oxidative stress from Paraquat results in a down-regulated use of the glycolysis (serine though glycolysis) and an increase of the Pentose Phosphate pathway flux (phosphoenolpyruvate through pentose phosphate pathway). The flux into the TCA cycle coming from pyruvate appears to decrease a bit. Apparent is also the complete lack of malic enzyme activity. The knowledge of flux ratios from the experiments can be combined with a metabolic network model of B. subtilis to calculate the absolute fluxes throughout the network. The result of the calculations by FiatFlux, the relative distribution of absolute, is depicted in figure 3-20. 3 | Results 62 Bacillus subtilis systems biology LST Research Project 2008 Glucose 100 100 CO NADPH NADPH 31 46 Glc 6-P Rul 5-P 66 50 Frc 6-P Xul 5-P Rib 5-P Erythrose-4-P Frc 1,6-bP DihydroxyAcetone-P Glyceraldehyde 3-P NADH Sedoheptulose 7-P NADPH 1,3-BisP-glycerate Ser PGA Gly NADH 161 155 CO PEP 169 164 15 15 Pyruvate CO NADH 110 106 CO 41 35 CO NADH 3 1 Acetyl-CoA 3 -1 Acetate 95 94 Oxaloacetate 83 86 NADH CO NADPH Malate 2-Oxoglutarate 85 85 FADH Citrate Succinate CO NADH Figure 3-20. Relative distribution of absolute fluxes in B. subtilis under control conditions and under oxidative stress. Conditions: µ = 0.29 h-1 / qs,PQ= 4.3 mmol·g1 ·h-1 / qs.control = 4.4 mmol·g-1·h-1. Glucose concentration 1 g/L. Fluxes are normalized to glucose uptake rates. Upper number in the boxes are from the control culture, lower numbers are derived from a 15 µM paraquat reactor. 3 | Results 63 Bacillus subtilis systems biology LST Research Project 2008 To find out if the effects of paraquat induced oxidative stress can be seen throughout the whole metabolic network, the analysis of flux ratios is a helpful tool. The resulting figure 3-20 shows a major adjustment in the split of the glycolysis and the pentose phosphate pathway. The flux into the pentose phosphate pathway increases with the introduction of oxidative stress at the expense of the glycolysis. The TCA cycle shows no change in fluxes. Resolving oxidative stress is coupled to the usage of NADPH. Analysis of flux ratios also yields the amounts of NADPH produced in the network and the amount of biomass NADPH consumption. NADPH production stems from the pentose phosphate pathway, the isocitrate dehydrogenase reaction and the malic enzyme reaction. NADPH consumption is based on the known biochemical requirements of NADPH for growth dependant macromolecule biosynthesis (15). The production and consumption can be compared and checked for imbalances. The results of this NADPH balance for a control culture and a 15 µM paraquat reactor are shown in figure 3-21. Relative NADPH flux [%] 200 PP Pathway Isocitrate Malic enzyme 150 100 50 0 Biomass NADPH consumption NADPH formation Control Biomass NADPH consumption NADPH formation Paraquat 15 µM Figure 3-21. Comparison of NADPH production and consumption rates in B. subtilis grown with and without 15 µM Paraquat. Bars represent the estimated formation and consumption from net analysis. Both samples were grown at µ = 0.29. The glucose uptake rate of the control was 4.4 mmol · g-1 · h-1 and of the Paraquat reactor 4.3 mmol · g-1 · h-1. This figure shows that the formation of NADPH under oxidative stress increases significantly. The generated NADPH is a result of an increased pentose phosphate pathway. The fluxes through isocitrate dehydrogenase remain the same. Iron limitation 3 | Results 64 Bacillus subtilis systems biology LST Research Project 2008 Iron experiments were performed with a deferoxamine concentration of 3 mM per reactors. This is about 0.16 gram deferoxamine being flushed through a reactor per experiment and would be able to bind 1.36 grams of iron. This amount of deferoxamine was employed in an attempt to bind all available iron. Lower concentrations had proven unsuccessful (see dose response results). Next to the cultures with deferoxamine there were cultures where neither the precultures nor the reactor media were supplied with iron. This could show the capacity of B. subtilis to harvest iron from contaminated ion-exchange surfaces as glass or silicone tubing. All iron deprivation experiments reached culture densities that were constant in time and did not washout. The biomass samples of these cultures was also analyzed with GC-MS and FiatFlux. The results from the flux analysis are shown in figure 3-22. 1.200 Control DFO 1.000 Relative flux ratios Ironless 0.800 0.600 0.400 0.200 0.000 PEP PEP through through PPP (ub) TK (ub) SER GLY from OAA E4P through from GLY SER from PYR TK PEP from OAA PYR from PYR from Labeled MAL (ub) MAL (lb) CO2 Pathway and reactions Figure 3-22 Flux ratios in B. subtilis under iron limited conditions. Fluxes are composed of data from different reactors under equal conditions. DFO: 4 reactors, Ironless: 4 reactors. The errors indicate deviations between various reactors. DFO: deferoxamine. Iron limitation could also have an effect on the metabolic network resulting from problems with iron containing proteins that would not function. These proteins are found in the TCA cycle and in parts of the respiratory chain. Flux ratios from the iron limiting experiments in figure 3-22 show little variation. Especially the controls and the ironless situations hardly differ. The notable alterations are in the amino-acid biosynthesis pathway of serine from glycine. Where the pathway in the direction of serine from glycine shows stable behavior, the exchange flux in the direction of serine is more than halved under deferoxamine growth. Other shifts are too small to consider, they do not exceed the measurement errors. 3 | Results 65 Bacillus subtilis systems biology LST Research Project 2008 Analysis of the extracellular compounds revealed the absence of the formation of acetate, succinate, fumarate or lactate during iron limited growth in chemostats. 3 | Results 66 Bacillus subtilis systems biology LST Research Project 2008 4 Discussion 4.1 Optimized growth medium Iron limitation can only be investigated when a reference cultures is sure to be grown without the stress of obtaining iron. Citrate was evaluated to see if it could resolve the iron stress observed in batch reactors. One of the properties of citrate is its ability to function as an exogenous iron chelator for B. subtilis (52). Higher citrate concentrations should allow the increased transport of iron into the cell but apparently the dosage of 100 µM suffices. This is in agreement with personal experience that states 10 µM of iron as the bare minimum for growth (J.D. Helmann, personal communication). This is easily accommodated for by citrate as it is able to bind three molecules of iron. That 100 µM citrate resolves B. subtilis apparent iron stress sufficiently can also be seen from the fact that offering more iron (via citrate) does not promote cell growth any further. This indicated that the iron stress has been resolved in this case. The addition of citrate is a practical solution to overcome the period in which B. subtilis is adapting to the new media conditions after inoculation. It is unknown if the observed up-regulation of Fur during B. subtilis growth has been reduced by the addition of citrate, but the results in this study show a decreased lag phase and an improved maximum growth-rate. The increased growth-rate could also indicate that citrate is being used in carbon metabolism by replenishing the TCA cycle. B. subtilis is known to be able to grow on glucose co-fed with citrate (16). This co-feeding is only possible under glucose limited conditions as the repressing mechanism for the use of other substrates is relieved. In the shakeflasks where glucose is abundant, B. subtilis will repress the uptake of other carbon sources and use glucose as the preferred carbon source (66). This is supported by the results in this report which shows that everincreasing amounts of citrate do not improve growth-rate. The shakeflasks do show however, that the maximum obtained culture density is higher with citrate added. It is doubtful whether this is the result of citrate in the medium as it accounts for only 1.6% of the total available carbon. The effect of citrate in the chemostat system has not been investigated here but under glucose limiting conditions, citrate will be taken up for metabolic purposes. Studies showed that the biomass yield increased when citrate was co-fed in glucose limited systems (43). Other studies showed that the co-fed citrate was mainly used for the generation of excess ATP used in maintenance metabolism. The partitioning of glucose-6-phosphate over the glycolysis and the pentose phosphate pathway remained unchanged (16). This indicated that the obtained flux ratios are not negatively influenced by the fact that the biomass is grown in the presence of citrate. The analysis of the flux ratios in FiatFlux showed that the summed fractional labeling of amino acids was close to 0.2. This is an indication 4 | Discussion 67 Bacillus subtilis systems biology LST Research Project 2008 of the incorporated labeled substrate. The observed values do not indicate a high usage of citrate. 4.2 Optimization of the chemostat setup The original chemostat setup that was available in the lab did not immediately yielded proper steady state cultures. After many chemostat optimization experiments, the setup appeared to be operating non-randomly. This pointed to a systematic error in the system and as most of the parts had proven not to affect the growth inside the chemostat. The solution to the problem was an adjustment of the feed tubing. By replacing the old tubing system that used many splits before arriving at the reactors was replaced by eight individual feed tubes. The real background of the problem remains unclear. The hypothesis of malfunctioning feed had already been tested by comparing the liquid flow in the effluent pump system with the feed flow-rates. At that moment both measurements did not indicate a difference between the incoming and outgoing volume. Furthermore, a decrease in feed-flow would not directly lower the final density of a reactor; the cultures are able to grow continuously at much lower feed rates. A possibility could be that the determination of the feed-flow was not correct because the measurements were done outside the system. Under normal operation the feed flow would be influenced by the underpressure in the reactor resulting from the effluent pump. The results pointed to a distribution of substrate in the system. In that case, some reactors would receive more glucose than others, and therefore achieve higher densities. What exactly happened inside the tubing splits is unsure, but they must have attributed to the problem as the isolation of reactor-channels solved the problem. In fact, the final adaptation is a more solid solution. By isolating each reactor vessel, they are in no way influenced by any of the processes in other reactors. Occlusions in one part would not affect other parts of the system any more. Moreover, the system features extra cultivation options in the new setup. The individual media bottles allow a per reactor control of what is being flushed through. 4.3 Oxidative stress The generation of oxidative stress conditions in B. subtilis cultures proved to be quite straightforward. The dose response of B. subtilis to Paraquat indicated that the intensity of the oxidative stress response remained constant at low concentrations of Paraquat. A binary effect on growth is observed after the administration of more than 20 µM Paraquat, from that point the growth shows a different response. The initial response at low Paraquat levels is apparently enough to overcome the effects of low oxidative stress. Even though Paraquat concentrations found in literature did not translate directly to a functional setup (see chapter 3.3), they provided an excellent starting point. 4 | Discussion 68 Bacillus subtilis systems biology LST Research Project 2008 The chemostat experiments were based on the dose response curve but still not all concentrations yielded a good steady state after 6 volume changes. An initial experiment with 40 µM Paraquat immediately washed out. This value lies close to the drawn trend line of the dose response curve. An unexpected result was the varying success with lower-than-lethal concentrations of Paraquat. Every concentration of paraquat administred, has shown a washout at least once. An adjustment of the feeding procedure where Paraquat was administered after two volume changes, did increase the success rate. After all experiments the 15 µM Paraquat concentration was the most successful. This concentration yielded more viable cultures than 20 µM and even more than the lower 10 µM. This effect might be due to an adaptation problem as increasing concentrations of Paraquat enter the reactor after feed-start. The washout of higher concentrations can be interpreted as an effect of high dosage. When the feed is started the concentration of paraquat slowly rises in the culture. It will reach the final concentration after a many hours. This period would allow the growth of B. subtilis but as the concentration increases, the environment might get too harsh to survive. This could also explain that some cultures washed out only a few hours before the harvesting moment. The steady state data from higher Paraquat concentration should therefore be interpreted with care. The biomass might have been collected at a point were the culture was already losing viability. The chemostats provided numerous biomass samples for further analysis. The 15 µM samples were the most abundant but also the successful 10 µM and 25 µM have been collected. When considering the biomass yields one might have expected a lower yield compared to the control cultures. A reasonable explanation would be the increased energy demand stemming from the formation of NADPH for the formation Paraquat mediated radicals and the subsequent sequestering of the reactive oxygen species. The glucose uptake rate is also slightly increased which could indicate an increase in central carbon metabolism for the generation of reduction equivalents. The variations are however very small and no final verdict can be given based on this data. An indepth analysis of the changes in central carbon metabolism allows a more detailed investigation. Research in yeast and in plants has shown that oxidative stress impacts the central metabolic pathways profoundly. These studies used transcriptomics, 13C labeling experiments and protein gels to analyze cellular adaptations in varying experimental setups. Their results showed that the TCA cycle is inhibited and carbon is re-routed into the oxidative part of the pentose phosphate pathway (4, 28, 70). By diverting the carbon away from the respiratory pathways and into the pentose phosphate pathway it generates reductants for antioxidant mechanisms. Rather than attempting to replace the inhibited enzymes, the cells mount an antioxidant response system and try to bypass damaged steps. Producing proteins that are immediately oxidatively damaged would be a waste of energy. 4 | Discussion 69 Bacillus subtilis systems biology LST Research Project 2008 Finally, there was an observed induction of transcripts of catabolic pathways that led to the mobilization of alternative internal carbon sources and their subsequent breakdown in order to generate ATP. The adaptations in B. subtilis during oxidative stress are less clear. Transcriptome analysis revealed that under superoxide en peroxide stress conditions, several genes which encode enzymes of glycolysis and the TCA cycle are downregulated (49). It was also shown that the zwf gene which encodes the enzymes catalyzing the first step of the gluconate bypass (figure 4-1) were upregulated under oxidative stress (59). The product of this reaction, gluconate, can be fed into the pentose phosphate pathway to yield NADPH. Figure 4-1. Diagram of the oxidative pentose phosphate pathway and related reactions in B. subtilis. The gluconate bypass is depicted as a dotted line and follows from glucose via gluconate to 6-phospho-gluconate (76). Other research has also indicated that the use of the pentose phosphate pathway depends on the [NADP+]/[NADPH]-ratio (16, 48). An increased demand for NADPH would direct more glucose into the pentose phosphate pathway. Flux ratios from the control cultures and the Paraquat cultures show a major difference in the flux being channelled into the pentose phosphate pathway. Higher concentrations of oxidative stress result in more than 20% increase of the pentose pathway flux. This could well indicate an extra demand for NADPH reductants. NADPH can be consumed in different processes: (i) in the reduction of reactive oxygen species, (ii) by the Paraquat induced radical generation or (iii) the conversion into NADH by transhydrogenase-like activity (69). Subsequent net analysis allows the calculation of an NADPH balance and shows that Bacillus has an excess formation of NADPH. This is according to what was found in literature under comparable conditions (62). The imposition of oxidative stress leads to an increased imbalance resulting from extra NADPH production but no higher NADPH consumption from biomass. Under oxidative stress, the glucose-6phosphate is shown to be rerouted into the pentose phosphate pathway at the expense of the glycolysis. This shows that external stressors do influence the 4 | Discussion 70 Bacillus subtilis systems biology LST Research Project 2008 normally preferred split ratio at this branchpoint. In this case that would be the result of an increased NADPH demand. Analysis of the other NADPH producing reactions, isocitrate dehydrogenase and the malic enzyme, shows that only the first has an increased response to oxidative stress. The malic enzyme seems not to function at all when compared to other wild-type strains of B. subtilis (69). This effect could be the result of the available citrate in the medium. Co-feeding experiments have already revealed that the flux through the malic enzyme decreases under glucose limited growth with citrate available (16). The TCA cycle has shown to be deregulated in plants and yeasts under influence of oxidative stress (4, 28, 70). According to the analysis of the netto fluxes in this study, this effect does not hold true for B. subtilis under oxidative stress. The continued usage of the TCA cycle suggests that the enzymes in the TCA cycle with probable oxidative stress susceptibility, aconitase and succinate dehydrogenase, are not causing a blockade of the cycle. It is not certain whether this is the result of the anti-oxidant systems in B. subtilis that prevent these enzymes from fatal damage or that the transcription levels are elevated to maintain proper TCA cycle function. While the flux through the TCA cycle seems not to be affected, the analysis of extracellular compounds revealed an increasing concentration of fumarate being released. And with the highest concentrations of Paraquat also succinate, acetate and lactate are found. This might indicate some problems in the TCA cycle. Fumarate is a product of succinate hydrogenase, an enzyme supposed vulnerable to oxidative stress. Paraquat induction could well result in the upregulation of this protein or its malfunctioning. Succinate levels can increase when succinate dehydrogenase is affected by reactive oxygen species and acetate might accumulate when acetylCoA cannot enter the TCA cycle anymore as a result of malfunctioning aconitase. There are no clear flux ratio results in this extreme Paraquat regime to illustrate this phenomenon due to the fact that only few cultures achieved steady state at Paraquat concentrations above 15 µM. 4.4 Iron limitation The role of carbon metabolism in survival strategies under iron limited growth is not yet fully understood. The iron sparing response in B. subtilis would aim primarily at repressing iron-rich proteins. For another bacterium of interest to the BaSysBio project, S. aureus, it has been shown that under iron deprived conditions, the central carbon metabolism was redirected to increase iron availability. Experiments revealed that S. aureus upregulated its glycolysis and decreased the TCA cycle usage (24). The result was an overproduction of acidic end products (lactate) that lowered the external pH. This could result in the release of iron from iron carriers as transferring which are available in the pathogens host. 4 | Discussion 71 Bacillus subtilis systems biology LST Research Project 2008 For B. subtilis there are not flux data under iron limitation. What was shown, separately, was that fur mutants grow better in an iron deprived environment compared to strains with an intact fur but a defect in their iron acquisition (26) and that a mutation in fur results in a lower TCA cycle flux (-23%) (21). An effect of iron limitation in B. subtilis might thus be observable in the TCA cycle. In order to explore the effects of iron on B. subtilis the iron chelator deferoxamine was applied. This was based on findings in literature which had successfully used the compound to create an iron limiting environment for microorganisms. The application of deferoxamine in this research proved unsatisfactory. The determination of the right dosage already indicated that the growth of B. subtilis with deferoxamine showed improved growth characteristics compared to cultures grown on ironless media. No dosage was identified that could completely inhibit the growth. The chemostats were therefore run with a maximal amount of deferoxamine. The results show that B. subtilis must have been able to sequester iron somewhere because all experiments yielded proper growth. The amount of deferoxamine used for each reactor, about 0.16 gram, would theoretically be able to bind 1.36 gram of iron. This also excludes the iron contamination in the chemicals used as a source of interference. The maximum iron contamination in the used chemicals was 0.005%. Deferoxamine would be able to handle 20 kg of this chemical. That is far more than the combined weight of the used chemicals in M9 medium. All cells in the deferoxamine chemostats remained viable and the results even showed a significant increase in biomass yield. This would not be possible in an iron limiting environment. It might be that deferoxamine is used as a carbon source by B. subtilis. The concentration of DFO used in the chemostats is so high that the total carbon content in the reactor is increased by 220%. Even though Bacillus spp. were not reported to profit from the availability of deferoxamine it is known that some organisms do (23). The really high concentrations might increase the chance of deferoxamine being metabolized. The analysis of the flux ratios in FiatFlux showed that the summed fractional labeling of amino acids was close to 0.2. This is an indication of the incorporated labeled substrate. The observed values do not indicate a high usage of external carbon from deferoxamine. The second question is whether deferoxamine is truly capable of creating an iron limited environment remains. An option to consider would be the competition between deferoxamine and bacillibactin for iron. Kinetic data shows that the iron affinities of deferoxamine and bacillibactin are equal. It might have been that he iron bound by deferoxamine was captured directly by bacillibactin after B. subtilis initiated the production of these siderophores. The source of iron, if not being added to the medium, is also not hard to find. It is known that iron can bind glass surfaces and contaminate the media from there. It is therefore that many 4 | Discussion 72 Bacillus subtilis systems biology LST Research Project 2008 studies used extensive deferration methods to clean their glasswork or even reverted to polystyrene bottles in stead of glass. Other options used in iron limitation studies was the application of mutants that lack the possibility to produce siderophores (35). Even though the cultures achieve normal growth, there appears to be a minor effect when the flux ratios are observed. The only difference between B. subtilis grown with deferoxamine is the change of flux between serine and glycine. When grown with deferoxamine the flux from glycine to serine remains constant but the exchange flux of serine from glycine is decreased. This could be an indication of iron limitation as glycine is one of the two precursors of bacillibactin. In a situation where much bacillibactin is required, an increase on the precursor generation is expected and the conversion of glycine into serine would not favor the situation. The flux of serine from glycine is known to be connected to specific growthrates but as all reactors in this system are grown at the same rate the effect must be resulting from deferoxamine. An increased siderophore production would support the observation that the supernatant of the deferoxamine grown reactors showed a slight yellow color. It is known that high concentrations of B. subtilis’ siderophores yield a brown color. The coloring together with the flux shift indicates a response of B. subtilis to iron limitation. This is not found in control cultures or in cultures grown without any supplied iron. This is an effect of the added deferoxamine, which has apparently generated some level of iron limitation. This level of iron limitation was apparently survivable with ease and a major adaptation in central carbon metabolism was not observed. Other changes in metabolism are not observed and a decreased TCA cycle flux was not seen. Other indications as the production of lactate for iron aquisition in S. aureus were not observed in these B. subtilis experiments. 4 | Discussion 73 Bacillus subtilis systems biology LST Research Project 2008 5 Conclusions and recommendations To date, no investigations have been done to elucidate the metabolic response of wild-type B. subtilis to oxidative and iron limiting conditions. This thesis tries to make a start in that direction. The results show that oxidative stress leads to a re-routing of carbon through the pentose phosphate pathway. The subsequent increase in NADPH production is not required for biomass NADPH consumption and it is therefore likely to play a role in the relief of oxidative stress. The study of iron limitation on B. subtilis did not yield a clear picture. It is unsure whether the conditions were truly iron limiting and there are no observed effects on the TCA cycle as had been indicated by other researchers. The use of deferoxamine for obtaining iron limited conditions appears unsuitable as the affinity for iron is equal to that of bacillibactin. The initial starting point of aiding the BaSysBio project in advancing towards a better understanding of B. subtilis adaptation to a shift in environmental conditions remained the main focus of this thesis. The generated results give further insight into the adaptation processes and help future researchers by providing new information on the stress responses. More specifically, BaSysBio is now provided with increased knowledge on the metabolic responses to combat stress, the NADPH supply has been proven to be a metabolic burden under oxidative stress and with insight in the fact that metal containing proteins appear to be functioning under oxidative stress. Future investigations could focus on obtaining truly iron limiting conditions for B. subtilis in steady state chemostat experiments. First of all it would be interesting to see how much iron is actually available in a culture. There are assays available that can indicate the amount of iron (13). Iron could possible originate from contaminants in the chemicals used. If no other chelator can be found that is able to bind iron strong enough not to pass it to bacillisbactin it could be considered to deferrate the medium with special ion exchange resins. These have been shown to generate practically iron-less environments. To asses whether or not B. subtilis is coping with iron stress an assay could be used that indicated the availability of siderophores. When deferrating the environment does not suffice it would be possible to revert to a different B. subtilis strain that lacks the ability to produce siderophores. The chemostat would be an ideal system to use for future experiments. The system is operational and can easily be extended from eight to twelve reactors. The prerequisites for oxidative stress conditions in the mini chemostat system have now been determined in this project. The chemostats can now be used without any problems to generate samples of cultures under oxidative stress for transcript analysis. These transcripts can aid in elucidating the stress response cascade and continue to improve genome-scale models. 5 | Conclusions and recommendations 74 Bacillus subtilis systems biology LST Research Project 2008 Another interesting point for BaSysBio could be the combination of stresses. Pathogens encounter a plethora of stress upon host-invasion. A combination of iron limitation and oxidative stress would form a starting point. 5 | Conclusions and recommendations 75 Bacillus subtilis systems biology LST Research Project 2008 Acknowledgements I wish to thank Prof. Dr. Uwe Sauer for the opportunity to carry out my research internship at the Institute of Molecular Systems Biology. His visionary approach and innovative ideas were an asset to the project. Additionally, I would like to thank him for the critical reading and correction of this manuscript. A second word of thanks goes out to Roelco Kleijn, without whom this project would not have been possible. I am thankful for the time he invested in assisting me on my experiments and thoroughly reading this document. I would like to thank Peter Verheijen from Delft for his words of wisdom and the continuous support throughout this project period. Finally I wish the Sauer group all the best and thank them for a wonderful time in Switzerland. Especially Daniel Heine, my fellow-Diplomand, who helped me taking the hurdle to start speak German again after twelve years. Acknowledgements 76 Bacillus subtilis systems biology LST Research Project 2008 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Allard, K. A., V. K. Viswanathan, and N. P. Cianciotto. 2006. lbtA and lbtB Are Required for Production of the Legionella pneumophila Siderophore Legiobactin. Journal of Bacteriology 188:1351-1363. Andrews, S. C., A. K. Robinson, and F. Rodriguez-Quinones. 2003. Bacterial iron homeostasis. Fems Microbiology Reviews 27:215-237. Baichoo, N., T. Wang, R. Ye, and J. D. Helmann. 2002. Global analysis of the Bacillus subtilis Fur regulon and the iron starvation stimulon. Molecular Microbiology 45:1613-1629. Baxter, C. J., H. Redestig, N. Schauer, D. Repsilber, K. R. Patil, J. Nielsen, J. Selbig, J. L. Liu, A. R. Fernie, and L. J. Sweetlove. 2007. The metabolic response of heterotrophic Arabidopsis cells to oxidative stress. Plant Physiology 143:312-325. Benjah 08.09.2008 2006, posting date. Deferoxamine - Wikipedia, the free encyclopedia. [Online.] Benov, L., and I. Fridovich. 1999. Why superoxide imposes an aromatic amino acid auxotrophy on Escherichia coli - The transketolase connection. Journal of Biological Chemistry 274:4202-4206. Bertani, G. 2004. Lysogeny at mid-twentieth century: P1, P2, and other experimental, systems. Journal of Bacteriology 186:595-600. Braunschweig, T. 2008, posting date. Braunschweig Enzyme Database: The Comprehensive Enzyme Information System. [Online.] Bromilow, R. H. 2004. Paraquat and sustainable agriculture. Pest Management Science 60:340-349. Bsat, N., A. Herbig, L. Casillas-Martinez, P. Setlow, and J. D. Helmann. 1998. Bacillus subtilis contains multiple Fur homologues: identification of the iron uptake (Fur) and peroxide regulon (PerR) repressors. Molecular Microbiology 29:189-198. Cao, M., C. M. Moore, and J. D. Helmann. 2005. Bacillus subtilis paraquat resistance is directed by sigma(M), an extracytoplasmic function sigma factor, and is conferred by YqjL and BcrC. Journal of Bacteriology 187:2948-2956. Commichau, F. M., and J. Stulke. 2008. Trigger enzymes: bifunctional proteins active in metabolism and in controlling gene expression. Molecular Microbiology 67:692-702. Cox, C. D. 1994. DEFERRATION OF LABORATORY MEDIA AND ASSAYS FOR FERRIC AND FERROUS-IONS, p. 315-329, Bacterial Pathogenesis, Pt A, vol. 235. Academic Press Inc, San Diego. Crosa, J. H., and O. Tualatin. 1995 Deferration using anguibactin siderophore Dauner, M., J. E. Bailey, and U. Sauer. 2001. Metabolic flux analysis with a comprehensive isotopomer model in Bacillus subtilis. Biotechnology and Bioengineering 76:144-156. References 77 Bacillus subtilis systems biology 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. LST Research Project 2008 Dauner, M., M. Sonderegger, M. Hochuli, T. Szyperski, K. Wuthrich, H. P. Hohmann, U. Sauer, and J. E. Bailey. 2002. Intracellular carbon fluxes in riboflavin-producing Bacillus subtilis during growth on two-carbon substrate mixtures. Applied and Environmental Microbiology 68:1760-1771. Dauner, M., T. Storni, and U. Sauer. 2001. Bacillus subtilis metabolism and energetics in carbon-limited and excess-carbon chemostat culture. Journal of Bacteriology 183:7308-7317. Dougherty, M. J., and D. M. Downs. 2006. A connection between ironsulfur cluster metabolism and the biosynthesis of 4-amino-5hydroxymethyl-2-methylpyrimidine pyrophosphate in Salmonella enterica. Microbiology-Sgm 152:2345-2353. Duetz, W. A., L. Ruedi, R. Hermann, K. O'Connor, J. Buchs, and B. Witholt. 2000. Methods for Intense Aeration, Growth, Storage, and Replication of Bacterial Strains in Microtiter Plates. Applied and Environmental Microbiology 66:2641-2646. Fendt, S. M., and M. Rühl. 2008. Quantitative measurement of in vivo fluxes with 13C metabolic flux analysis. Nature Methods Submitted paper. Fischer, E., and U. Sauer. 2005. Large-scale in vivo flux analysis shows rigidity and suboptimal performance of Bacillus subtilis metabolism. Nature Genetics 37:636-640. Fischer, E., and U. Sauer. 2003. Metabolic flux profiling of Escherichia coli mutants in central carbon metabolism using GC-MS. European Journal of Biochemistry 270:880-891. Flournoy, D. J. 1991. INVITRO ANTIMICROBIAL PROPERTIES OF DEFEROXAMINE MESYLATE. European Journal of Clinical Microbiology & Infectious Diseases 10:597-598. Friedman, D. B., D. L. Stauff, G. Pishchany, C. W. Whitwell, V. J. Torres, and E. P. Skaar. 2006. Staphylococcus aureus redirects central metabolism to increase iron availability. Plos Pathogens 2:777-789. Fuangthong, M., A. F. Herbig, N. Bsat, and J. D. Helmann. 2002. Regulation of the Bacillus subtilis fur and perR genes by PerR: Not all members of the PerR regulon are peroxide inducible. Journal of Bacteriology 184:3276-3286. Gaballa, A., H. Antelmann, C. Aguilar, S. K. Khakh, K.-B. Song, G. T. Smaldone, and J. D. Helmann. 2008. The Bacillus subtilis ironsparing response is mediated by a Fur-regulated small RNA and three small, basic proteins. Proceedings of the National Academy of Sciences 105:11927-11932. Gaballa, A., and J. D. Helmann. 2007. Substrate induction of siderophore transport in Bacillus subtilis mediated by a novel onecomponent regulator. Molecular Microbiology 66:164-173. References 78 Bacillus subtilis systems biology 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. LST Research Project 2008 Godon, C., G. Lagniel, J. Lee, J. M. Buhler, S. Kieffer, R. Perrot, H. Boucherie, M. B. Toledano, and J. Labarre. 1998. The H2O2 stimulon in Saccharomyces cerevisiae. Journal of Biological Chemistry 273:2248022489. Gray, J. P., D. E. Heck, V. Mishin, P. J. S. Smith, J. Y. Hong, M. Thiruchelvam, D. A. Cory-Slechta, D. L. Laskin, and J. D. Laskin. 2007. Paraquat increases cyanide-insensitive respiration in murine lung epithelial cells by activating an NAD(P)H : paraquat oxidoreductase Identification of the enzyme as thioredoxin reductase. Journal of Biological Chemistry 282:7939-7949. Hardwick, S. W., J. Pane-Farre, O. Delumeau, J. Marles-Wright, J. W. Murray, M. Hecker, and R. J. Lewis. 2007. Structural and functional characterization of partner switching regulating the environmental stress response in Bacillus subtilis. Journal of Biological Chemistry 282:11562-11572. Harwood, C. R., and S. M. Cutting. 1990. MODERN MICROBIOLOGICAL METHODS MOLECULAR BIOLOGICAL METHODS FOR BACILLUS. Hecker, M., J. Pane-Farre, and U. Volker. 2007. SigB-dependent general stress response in Bacillus subtilis and related gram-positive bacteria. Annual Review of Microbiology 61:215-236. Helmann, J. D., M. F. W. Wu, A. Gaballa, P. A. Kobel, M. M. Morshedi, P. Fawcett, and C. Paddon. 2003. The global transcriptional response of Bacillus subtilis to peroxide stress is coordinated by three transcription factors. Journal of Bacteriology 185:243-253. Henle, E. S., Z. X. Han, N. Tang, P. Rai, Y. Z. Luo, and S. Linn. 1999. Sequence-specific DNA cleavage by Fe2+-mediated fenton reactions has possible biological implications. Journal of Biological Chemistry 274:962971. Hoffmann, T., A. Schutz, M. Brosius, A. Volker, U. Volker, and E. Bremer. 2002. High-salinity-induced iron limitation in Bacillus subtilis. Journal of Bacteriology 184:718-727. Imlay, J. A. 2008. Cellular defenses against superoxide and hydrogen peroxide. Annual Review of Biochemistry 77:755-776. Imlay, J. A. 2003. Pathways of oxidative damage. Annual Review of Microbiology 57:395-418. Imlay, J. A., S. M. Chin, and S. Linn. 1988. Toxic DNA Damage by Hydrogen-Peroxide through the Fenton Reaction Invivo and Invitro. Science 240:640-642. Kern, A., E. Tilley, I. S. Hunter, M. Legisa, and A. Glieder. 2007. Engineering primary metabolic pathways of industrial micro-organisms. Journal of Biotechnology 129:6-29. References 79 Bacillus subtilis systems biology 40. 41. 42. 43. 44. 45. 46. 47. 48. LST Research Project 2008 Keyer, K., and J. A. Imlay. 1996. Superoxide accelerates DNA damage by elevating free-iron levels. Proceedings of the National Academy of Sciences of the United States of America 93:13635-13640. Kunst, F. 1997, posting date. KEGG: Kyoto Encyclopedia of Genes and Genomes. [Online.] Kunst, F., N. Ogasawara, I. Moszer, A. M. Albertini, G. Alloni, V. Azevedo, M. G. Bertero, P. Bessieres, A. Bolotin, S. Borchert, R. Borriss, L. Boursier, A. Brans, M. Braun, S. C. Brignell, S. Bron, S. Brouillet, C. V. Bruschi, B. Caldwell, V. Capuano, N. M. Carter, S. K. Choi, J. J. Codani, I. F. Connerton, N. J. Cummings, R. A. Daniel, F. Denizot, K. M. Devine, A. Dusterhoft, S. D. Ehrlich, P. T. Emmerson, K. D. Entian, J. Errington, C. Fabret, E. Ferrari, D. Foulger, C. Fritz, M. Fujita, Y. Fujita, S. Fuma, A. Galizzi, N. Galleron, S. Y. Ghim, P. Glaser, A. Goffeau, E. J. Golightly, G. Grandi, G. Guiseppi, B. J. Guy, K. Haga, J. Haiech, C. R. Harwood, A. Henaut, H. Hilbert, S. Holsappel, S. Hosono, M. F. Hullo, M. Itaya, L. Jones, B. Joris, D. Karamata, Y. Kasahara, M. KlaerrBlanchard, C. Klein, Y. Kobayashi, P. Koetter, G. Koningstein, S. Krogh, M. Kumano, K. Kurita, A. Lapidus, S. Lardinois, J. Lauber, V. Lazarevic, S. M. Lee, A. Levine, H. Liu, S. Masuda, C. Mauel, C. Medigue, N. Medina, R. P. Mellado, M. Mizuno, D. Moestl, S. Nakai, M. Noback, D. Noone, M. OReilly, K. Ogawa, A. Ogiwara, B. Oudega, S. H. Park, V. Parro, T. M. Pohl, D. Portetelle, S. Porwollik, A. M. Prescott, E. Presecan, P. Pujic, B. Purnelle, et al. 1997. The complete genome sequence of the Grampositive bacterium Bacillus subtilis. Nature 390:249-256. Lee, J., A. Goel, M. M. Ataai, and M. M. Domach. 1997. Supply-side analysis of growth of Bacillus subtilis on glucose-citrate medium: Feasible network alternatives and yield optimality. Applied and Environmental Microbiology 63:710-718. Lee, J. W., and J. D. Helmann. 2007. Functional specialization within the Fur family of metalloregulators. Biometals 20:485-499. Marles-Wright, J., and R. J. Lewis. 2007. Stress responses of bacteria. Current Opinion in Structural Biology 17:755-760. Messner, K. R., and J. A. Imlay. 1999. The identification of primary sites of superoxide and hydrogen peroxide formation in the aerobic respiratory chain and sulfite reductase complex of Escherichia coli. Journal of Biological Chemistry 274:10119-10128. Miethke, M., O. Klotz, U. Linne, J. J. May, C. L. Beckering, and M. A. Marahiel. 2006. Ferri-bacillibactin uptake and hydrolysis in Bacillus subtilis. Molecular Microbiology 61:1413-1427. Moritz, B., K. Striegel, A. A. de Graaf, and H. Sahm. 2000. Kinetic properties of the glucose-6-phosphate and 6-phosphogluconate dehydrogenases from Corynebacterium glutamicum and their application References 80 Bacillus subtilis systems biology 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. LST Research Project 2008 for predicting pentose phosphate pathway flux in vivo. European Journal of Biochemistry 267:3442-3452. Mostertz, J., C. Scharf, M. Hecker, and G. Homuth. 2004. Transcriptome and proteome analysis of Bacillus subtilis gene expression in response to superoxide and peroxide stress. Microbiology-Sgm 150:497-512. Nanchen, A., A. Schicker, and U. Sauer. 2006. Nonlinear dependency of intracellular fluxes on growth rate in miniaturized continuous cultures of Escherichia coli. Applied and Environmental Microbiology 72:1164-1172. Noirot, D. P. 2006, posting date. BaSysBio. [Online.] Ollinger, J., K. B. Song, H. Antelmann, M. Hecker, and J. D. Helmann. 2006. Role of the fur regulon in iron transport in Bacillus subtilis. Journal of Bacteriology 188:3664-3673. Ong, S. T., J. Z. S. Ho, B. Ho, and J. L. Ding. 2006. Iron-withholding strategy in innate immunity. Immunobiology 211:295-314. Passalacqua, K. D., N. H. Bergman, A. Herring-Palmer, and P. Hanna. 2006. The superoxide dismutases of Bacillus anthracis do not cooperatively protect against endogenous superoxide stress. Journal of Bacteriology 188:3837-3848. Pierre, J. L., and I. Gautier-Luneau. 2000. Iron and citric acid: A fuzzy chemistry of ubiquitous biological relevance. Biometals 13:91-96. Pollak, N., C. Dolle, and M. Ziegler. 2007. The power to reduce: pyridine nucleotides - small molecules with a multitude of functions. Biochemical Journal 402:205-218. Price, C. W., P. Fawcett, H. Ceremonie, N. Su, C. K. Murphy, and P. Youngman. 2001. Genome-wide analysis of the general stress response in Bacillus subtilis. Molecular Microbiology 41:757-774. Ratledge, C., and L. G. Dover. 2000. Iron metabolism in pathogenic bacteria. Annual Review of Microbiology 54:881-941. Reder, A., D. Hoper, C. Weinberg, U. Gerth, M. Fraunholz, and M. Hecker. 2008. The Spx paralogue MgsR (YqgZ) controls a subregulon within the general stress response of Bacillus subtilis. Molecular Microbiology 69:1104-1120. Roberts, M., and M. N. Birmele 20.08.2008 2008, posting date. Passive Observatories for Experimental Microbial Systems (POEMS). [Online.] Roos, D., R. van Bruggen, and C. Meischl. 2003. Oxidative killing of microbes by neutrophils. Microbes and Infection 5:1307-1315. Sauer, U., V. Hatzimanikatis, J. E. Bailey, M. Hochuli, T. Szyperski, and K. Wuthrich. 1997. Metabolic fluxes in riboflavin-producing Bacillus subtilis. Nature Biotechnology 15:448-452. Scharf, C., S. Riethdorf, H. Ernst, S. Engelmann, U. Volker, and M. Hecker. 1998. Thioredoxin is an essential protein induced by multiple stresses in Bacillus subtilis. Journal of Bacteriology 180:1869-1877. References 81 Bacillus subtilis systems biology 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. LST Research Project 2008 Schicker, A. 2005. Metabolic Flux responses to global regulator knockouts in Escherichia coli mini-scale chemostat cultures. ETH, Zürich. Sonenshein, A. L. 1993. Bacillus subtilis and other gram-positive bacteria: Biochemistry, physiology and molecular genetics. American Society for Microbiology (ASM). Sonenshein, A. L. 1993. Introduction to metabolic pathways, p. 127132, Bacillus subtilis and other gram-positive bacteria: Biochemistry, physiology and molecular genetics. American Society for Microbiology (ASM). Spector, A., G. Z. Yan, R. R. C. Huang, M. J. Mcdermott, P. R. C. Gascoyne, and V. Pigiet. 1988. The Effect of H2o2 Upon ThioredoxinEnriched Lens Epithelial-Cells. Journal of Biological Chemistry 263:49844990. Tännler, S. 2008. Novel metabolic engineering targets for riboflavin production and elucidation of maintenance and NADPH metabolism in Bacillus subtilis. Dissertation. Tännler, S., S. Decasper, and U. Sauer. 2008. Maintenance metabolism and carbon fluxes in Bacillus species. Tarrio, N., M. Becerra, M. E. Cerdan, and M. I. G. Siso. 2006. Reoxidation of cytosolic NADPH in Kluyveromyces lactis. Fems Yeast Research 6:371-380. Thackray, P. D., and A. Moir. 2003. SigM, an Extracytoplasmic Function Sigma Factor of Bacillus subtilis, Is Activated in Response to Cell Wall Antibiotics, Ethanol, Heat, Acid, and Superoxide Stress. Journal of Bacteriology 185:3491-3498. Tottey, S., S. A. M. Rondet, G. P. M. Borrelly, P. J. Robinson, P. R. Rich, and N. J. Robinson. 2002. A copper metallochaperone for photosynthesis and respiration reveals metal-specific targets, interaction with an importer, and alternative sites for copper acquisition. Journal of Biological Chemistry 277:5490-5497. Weinberg, E. D. 1996. Iron withholding: A defense against viral infections. Biometals 9:393-399. Weiner, A. 10.08.2008 2006, posting date. Bacterium Bacillus subtilis with a Tecnai T-12 TEM. The Weizmann Institute of Science, Rehovot, Israel. [Online.] Ying, W. H. 2008. NAD(+)/ NADH and NADP(+)/NADPH in cellular functions and cell death: Regulation and biological consequences. Antioxidants & Redox Signaling 10:179-206. Zamboni, N., E. Fischer, D. Laudert, S. Aymerich, H. P. Hohmann, and U. Sauer. 2004. The Bacillus subtilis yqiI gene encodes the NADP(+)-dependent 6-P-gluconate dehydrogenase in the pentose phosphate pathway. Journal of Bacteriology 186:4528-4534. References 82 Bacillus subtilis systems biology 77. 78. LST Research Project 2008 Zamboni, N., E. Fischer, and U. Sauer. 2005. FiatFlux - a software for metabolic flux analysis from C-13-glucose experiments. Bmc Bioinformatics 6:8. Zhang, S. Y., and W. G. Haldenwang. 2005. Contributions of ATP, GTP, and redox state to nutritional stress activation of the Bacillus subtilis sigma(B) transcription factor. Journal of Bacteriology 187:7554-7560. References 83 Bacillus subtilis systems biology LST Research Project 2008 Appendices Appendix A A.1 Experimental data Chemostats Overview of the reactors from the limitation experiments. Every experiment has two controls. Numbers mark Paraquat concentrations in Paraquat runs and iron concentration in iron runs. DFO: deferoxamine, plus or minus indicated the preculture, minus is from a preculture without iron. The gray numbers indicate washouts and have not been used for further calculations Reactor content 15 16 17 18 19 20 21 23 24 PQ PQ Fe Fe PQ PQ PQ PQ Fe / PQ 1 control control control control control control control control control 1 U 1 U 1 1 1 U U 2 control control control control control control control control control 3 10 10 0+ DFO 10 10 10 15 DFO 4 10 10 0+ DFO 10 10 10 15 DFO 5 (25) (25) 50025 15 15 1 0- 6 25 (25) 50025 15 15 5 0- 7 40 15 05015 25 (25) 10 15 PQ 8 40 15 05015 25 (25) 20 15 PQ Observed OD’s in the chemostats after 6 volume changes. OD 15 16 17 18 19 20 21 23 24 PQ PQ Fe Fe PQ PQ PQ PQ Fe / PQ 1 U 1 U 1 1 1 U U 1 1.17 1.166 0.185 0.175 0.68 1.1 0.89 1.025 1.19 2 1.115 1.0835 0.675 0.74 0.855 1.01 0.775 0.935 1.14 3 1.13 1.056 0.81 1.345 0.015 0 0 0 1.145 4 1.045 1.001 0.75 1.24 0.048 0 0 0 1.24 5 6 7 8 0.375 0.95 0 0 0.245 0.285 1.034 1.0395 0.98 0.995 0.98 0.915 0.93 1.075 0.99 1.01 0.011 0.007 0.213 0.71 0 0 0 0 0.75 0.995 0.21 0.32 1.005 1.0275 1.0475 1.0075 1.06 1.005 1.185 0 Biomass concentration per reactor (g/L) Biomass g/l 15 PQ 16 PQ 17 Fe 18 Fe 19 PQ 20 PQ 21 PQ 23 PQ 24 Fe / PQ Appendices 1 U 1 U 1 1 1 U U 1 0.501 0.499 0.291 0.471 0.381 0.439 0.509 2 0.477 0.464 0.289 0.317 0.366 0.432 0.332 0.400 0.488 3 0.484 0.452 0.347 0.576 0.490 4 0.447 0.428 0.321 0.531 5 6 0.407 7 8 0.419 0.398 0.426 0.460 0.443 0.419 0.424 0.445 0.392 0.432 0.304 0.531 0.321 0.430 0.454 0.426 0.440 0.430 0.448 0.507 0.431 84 Bacillus subtilis systems biology LST Research Project 2008 Specific substrate uptake rates per fermenter (mmol.g-1.h-1) Uptake rates rate 15 PQ 16 PQ 17 Fe 18 Fe 19 PQ 20 PQ 21 PQ 23 PQ 24 Fe / PQ Appendices 1 U 1 U 1 1 1 U U 1 4.393 4.408 7.559 4.673 5.775 5.015 4.319 2 4.610 4.744 7.615 6.946 6.012 5.089 6.632 5.498 4.509 3 4.549 4.868 6.346 3.822 4.489 4 4.919 5.135 6.854 4.145 5 6 5.411 7 8 5.245 5.527 5.166 4.782 4.971 5.245 5.192 4.945 5.618 5.089 7.240 4.145 6.854 5.115 4.849 5.166 5.003 5.115 4.907 4.338 5.102 85 Appendices SER through glycolysis PEP from OAA Run 20: PQ 4 Label: [1-13C] SER through glycolysis PEP from OAA Run 21: PQ 5 Label: [1-13C] b 0.271 0.108 0.307 0.185 1.011 0.360 0.090 -0.03 -0.02 0.122 2 ± a 0.067 0.236 0.027 0.094 0.024 0.286 0.049 0.165 0.038 0.998 0.024 0.467 0.013 0.085 0.026 0.003 0.016 0.002 0.021 0.147 control 2 ± a 0.067 0.259 0.027 0.104 0.024 0.348 0.053 0.223 0.037 0.999 0.026 0.366 0.013 0.064 0.025 0.029 0.016 0.018 0.023 0.132 control 34 ± 0.067 0.027 0.022 0.045 0.036 0.031 0.012 0.029 0.016 0.019 ± 0.067 0.027 0.024 0.046 0.037 0.025 0.013 0.025 0.016 0.022 b 0.240 0.096 0.300 0.192 0.973 0.452 0.072 0.019 0.011 0.149 b 0.259 0.104 0.358 0.166 1.000 0.401 0.059 0.041 0.025 0.131 3 a 0.247 0.099 0.305 0.106 1.007 0.398 0.107 -0.055 -0.033 0.131 3 ± a 0.067 0.027 0.022 0.042 0.035 0.030 0.012 0.028 0.015 0.019 ± 0.067 0.027 0.022 0.045 0.037 0.026 0.013 0.026 0.015 0.020 Reactor 1 2 3 a ± b ± a ± b ± a 0.549 0.011 0.558 0.011 0.547 0.011 0.560 0.011 0.021 0.048 0.006 0.047 -0.009 0.046 0.001 0.047 control ± ± 35 ± ± 0.067 0.027 0.023 0.045 0.037 0.024 0.012 0.027 0.016 0.019 b b b ± ± ± 10 uM PQ 4 a 10 uM PQ 4 a 10 uM PQ 4 a ± ± 36 ± 4 b ± a ± 0.271 0.067 0.257 0.067 0.109 0.027 0.103 0.027 0.311 0.023 0.338 0.024 0.064 0.047 0.079 0.054 1.004 0.037 1.014 0.037 0.375 0.024 0.387 0.024 0.082 0.012 0.059 0.013 0 0.025 0.031 0.025 0 0.016 0.019 0.016 0.129 0.020 0.146 0.021 0 mM Fe / DFO max 3 mM b b b b 0.252 0.101 0.35 0.085 1.006 0.398 0.059 0.036 0.022 0.137 ± ± ± ± 0.067 0.027 0.022 0.047 0.037 0.024 0.013 0.025 0.015 0.019 37 ± 0.067 0.027 0.022 0.046 0.037 0.027 0.013 0.023 0.015 0.022 ± 0.067 0.027 0.024 0.049 0.037 0.029 0.013 0.026 0.016 0.021 b 0.242 0.097 0.343 0.142 0.987 0.367 0.059 0.023 0.014 0.124 b 0.259 0.103 0.29 0.123 1.014 0.381 0.09 -0.01 -0.01 0.125 6 ± a 0.067 0.293 0.027 0.117 0.022 0.323 0.044 0.119 0.036 0.994 0.026 0.385 0.012 0.074 0.024 0.009 0.015 0.006 0.021 0.16 1/5 6 ± a 0.067 0.244 0.027 0.098 0.023 0.316 0.05 0.232 0.037 1.004 0.029 0.4 0.012 0.072 0.026 0.017 0.016 0.01 0.022 0.14 0 mM Fe 38 ± 0.067 0.027 0.024 0.051 0.037 0.027 0.013 0.026 0.016 0.023 ± 0.067 0.027 0.024 0.046 0.037 0.029 0.013 0.026 0.016 0.021 b 0.314 0.126 0.251 0.114 0.989 0.387 0.127 -0.09 -0.05 0.151 b 0.263 0.105 0.274 0.171 0.981 0.376 0.087 -0.01 -0.01 0.126 ± 0.067 0.027 0.023 0.045 0.037 0.027 0.012 0.027 0.017 0.022 ± 0.067 0.027 0.024 0.05 0.037 0.027 0.013 0.026 0.016 0.021 7 a 0.333 0.133 0.302 0.154 0.996 0.369 0.087 -0.03 -0.02 0.151 7 a 0.346 0.138 0.303 0.236 0.979 0.297 0.085 -0.011 -0.008 0.115 39 ± 0.067 0.027 0.024 0.051 0.037 0.026 0.013 0.026 0.016 0.023 ± 0.067 0.027 0.024 0.051 0.037 0.025 0.013 0.023 0.016 0.026 b 0.321 0.128 0.304 0.149 1.015 0.31 0.082 -0.007 -0.005 0.134 b 0.335 0.134 0.327 0.266 0.987 0.297 0.075 0.014 0.01 0.132 8 ± a 0.067 0.343 0.027 0.137 0.024 0.285 0.05 0.288 0.039 0.991 0.025 0.392 0.013 0.111 0.023 -0.05 0.016 -0.03 0.026 0.149 10/20 8 ± a 0.067 0.027 0.024 0.046 0.038 0.025 0.013 0.023 0.016 0.027 15 uM PQ 40 ± 0.067 0.027 0.023 0.045 0.038 0.027 0.012 0.027 0.016 0.022 ± b 0.349 0.140 0.324 0.344 0.983 0.360 0.074 0.025 0.016 0.153 b ± 0.067 0.027 0.022 0.040 0.037 0.026 0.012 0.024 0.015 0.024 ± 5 a ± b ± 15 uM PQ 6 a ± b ± 7 a ± b ± 25 uM PQ 8 a ± b ± 5 6 7 8 a ± b ± a ± b ± a ± b ± a ± b ± 0.598 0.011 0.576 0.011 0.559 0.011 0.576 0.011 0.43 0.011 0.427 0.011 0.420 0.011 0.421 0.011 -0.015 0.050 -0.04 0.052 -0 0.042 -0.01 0.042 -0.03 0.062 0.020 0.060 -0.031 0.061 -0.034 0.060 15 uM PQ 25 uM PQ 5 a 0.298 0.119 0.372 0.126 1.022 0.373 0.033 0.077 0.048 0.144 5 a 0.269 0.107 0.305 0.187 0.989 0.383 0.085 -0.006 -0.004 0.119 B.1 PEP through PPP (ub) PEP through TK (ub) E4P through TK SER from GLY GLY from SER OAA from PYR PEP from OAA PYR from MAL (ub) PYR from MAL (lb) Labeled CO2 b 0.302 0.121 0.303 0.157 1.019 0.358 0.080 -0.01 -0.01 0.137 Appendix B Reactor 1 2 3 a ± b ± a ± b ± a 0.694 0.011 0.699 0.011 0.715 0.011 0.723 0.011 0.013 0.045 0.018 0.043 0.024 0.042 0.014 0.044 control Reactor 1 a ± 0.280 0.067 0.112 0.027 0.331 0.024 0.138 0.052 1.022 0.038 0.392 0.028 0.063 0.013 0.027 0.026 0.016 0.016 0.153 0.023 Run 23: PQ 5 Label: [u-13C] PEP through PPP (ub) PEP through TK (ub) E4P through TK SER from GLY GLY from SER OAA from PYR PEP from OAA PYR from MAL (ub) PYR from MAL (lb) Labeled CO2 Reactor 1 a ± 0.303 0.067 0.121 0.027 0.314 0.024 0.289 0.044 1.005 0.038 0.452 0.027 0.095 0.013 -0.04 0.030 -0.02 0.017 0.141 0.019 Run 24: DFO 3 / PQ 6 Label: [u-13C] Bacillus subtilis systems biology LST Research Project 2008 Flux data Flux ratios 86 Appendices SER through glycolysis PEP from OAA Run 15: PQ1 Label: [1-13C] PEP through PPP (ub) PEP through TK (ub) E4P through TK SER from GLY GLY from SER OAA from PYR PEP from OAA PYR from MAL (ub) PYR from MAL (lb) Labeled CO2 Run 16: PQ 2 Label: [u-13C] SER through glycolysis PEP from OAA Run 17: DFO1 Label: [1-13C] PEP through PPP (ub) PEP through TK (ub) E4P through TK SER from GLY GLY from SER OAA from PYR PEP from OAA PYR from MAL (ub) PYR from MAL (lb) Labeled CO2 Run 18: DFO 2 Label: [u-13C] SER through glycolysis PEP from OAA Run 19: PQ 3 Label: [1-13C] b 0.245 0.098 0.312 0.085 1.000 0.352 0.099 -0.02 -0.02 0.115 b b 0.75 -0.02 control 2 a 2 a 0.232 0.093 0.266 0.207 0.987 0.501 0.074 -0.01 -0.01 0.149 control 2 ± a 0.067 0.238 0.027 0.095 0.023 0.334 0.049 0.222 0.037 0.976 0.022 0.313 0.013 0.071 0.025 0.011 0.016 0.008 0.020 0.130 control ± ± Reactor 1 2 12 a ± b ± a 0.749 0.011 0.736 0.011 0.770 0.006 0.036 0.004 0.037 -0.009 control Reactor 1 18 a ± 0.258 0.067 0.103 0.027 0.336 0.023 0.148 0.046 1.001 0.036 0.367 0.026 0.075 0.013 -0.01 0.025 -0.01 0.016 0.144 0.023 Reactor 1 a ± Reactor 1 a ± b 0.232 0.093 0.265 0.173 0.992 0.489 0.080 -0.02 -0.01 0.145 ± 0.067 0.027 0.023 0.044 0.037 0.025 0.012 0.030 0.016 0.016 3 a 0.297 0.119 0.354 0.084 1.026 0.319 0.075 0.016 0.011 0.107 35 ± 0.067 0.027 0.022 0.048 0.040 0.032 0.013 0.023 0.016 0.026 ± ± 10 uM PQ 4 a ± 4 36 b ± a ± 0.313 0.067 0.319 0.067 0.125 0.027 0.128 0.027 0.320 0.023 0.315 0.023 0.068 0.049 0.037 0.052 1.037 0.040 1.055 0.041 0.340 0.033 0.294 0.032 0.092 0.013 0.098 0.013 -0.01 0.024 -0.02 0.023 -0.01 0.016 -0.02 0.016 0.134 0.027 0.106 0.028 0 mM Fe / DFO max 3 mM b b 0.318 0.127 0.319 0.072 1.053 0.362 0.099 -0.01 -0 0.06 b ± 0.067 0.027 0.023 0.049 0.041 0.029 0.013 0.025 0.016 0.018 ± 5 a 0.291 0.117 0.293 0.195 1.070 0.441 0.086 0.023 0.013 0.104 5 a 37 ± 0.067 0.027 0.023 0.045 0.041 0.025 0.012 0.028 0.016 0.016 ± b 0.278 0.111 0.289 0.165 1.022 0.451 0.091 -0.02 -0.01 0.155 b 25 uM PQ 6 a 6 ± a 0.067 0.255 0.027 0.102 0.023 0.295 0.046 0.17 0.039 1.037 0.027 0.302 0.012 0.094 0.029 -0.01 0.016 -0.01 0.019 0.154 0 mM Fe ± 38 ± 0.067 0.027 0.023 0.046 0.04 0.027 0.012 0.023 0.016 0.029 ± b 0.279 0.112 0.314 0.167 1.053 0.323 0.083 0.011 0.008 0.127 b ± 0.067 0.027 0.022 0.043 0.041 0.026 0.012 0.023 0.016 0.025 ± 7 a 0.293 0.117 0.316 0.158 1.071 0.371 0.1 -0.02 -0.01 0.141 39 ± 0.067 0.027 0.023 0.047 0.042 0.024 0.013 0.026 0.016 0.021 b 0.293 0.117 0.312 0.143 1.043 0.386 0.1 0.003 0.002 0.111 8 ± a 0.067 0.261 0.027 0.104 0.023 0.325 0.047 0.112 0.04 1.047 0.023 0.39 0.013 0.086 0.026 0.008 0.016 0.005 0.018 0.151 50 mM Fe 40 ± 0.067 0.027 0.023 0.045 0.041 0.024 0.012 0.026 0.016 0.021 b 0.273 0.109 0.307 0.075 1.061 0.424 0.105 -0.02 -0.01 0.104 ± 0.067 0.027 0.023 0.048 0.041 0.023 0.012 0.028 0.016 0.016 7 8 a ± b ± a ± b ± 0.551 0.011 0.555 0.011 0.576 0.011 0.559 0.011 -0.1 0.057 -0.079 0.054 -0.023 0.053 -0.008 0.052 15 uM PQ b 0.244 0.098 0.332 0.156 0.988 0.343 0.078 0.012 0.008 0.150 ± 0.067 0.027 0.023 0.046 0.036 0.023 0.013 0.024 0.016 0.023 13 ± b ± 0.011 0.758 0.011 0.035 -0.004 0.037 19 ± 0.067 0.027 0.022 0.040 0.036 0.024 0.013 0.023 0.016 0.025 3 a 0.605 -0.009 3 a 0.303 0.121 0.319 0.181 0.991 0.338 0.083 0.006 0.004 0.149 b 0.334 0.134 0.306 0.137 1.006 0.367 0.101 -0.01 -0.01 0.127 4 ± a 0.067 0.342 0.027 0.137 0.023 0.326 0.046 0.135 0.038 0.986 0.024 0.37 0.013 0.09 0.026 0.012 0.016 0.007 0.021 0.139 10 uM PQ 4 14 ± b ± a 0.011 0.615 0.011 0.593 0.049 -0.003 0.048 0.009 10 uM PQ 20 ± 0.067 0.027 0.023 0.043 0.038 0.024 0.013 0.024 0.016 0.024 b 0.329 0.132 0.321 0.146 1.02 0.346 0.1 0 0 0.103 ± 0.067 0.027 0.023 0.048 0.039 0.023 0.013 0.025 0.016 0.02 15 ± b ± 0.011 0.604 0.011 0.041 -0.026 0.050 21 ± 0.067 0.027 0.023 0.048 0.037 0.024 0.013 0.026 0.016 0.021 5 a 0.405 -0.042 5 a 0.447 0.179 0.223 0.455 1.021 0.452 0.130 -0.086 -0.047 0.140 b 0.456 0.182 0.241 0.386 0.998 0.464 0.139 -0.1 -0.05 0.141 6 ± a 0.067 0.423 0.027 0.169 0.024 0.285 0.042 0.414 0.042 1.014 0.027 0.448 0.013 0.079 0.034 0.041 0.018 0.023 0.018 0.141 25 uM PQ 6 16 ± b ± a 0.011 0.408 0.011 0.564 0.064 -0.03 0.063 0.018 25 uM PQ 22 ± 0.067 0.027 0.024 0.040 0.041 0.027 0.012 0.031 0.017 0.018 b 0.465 0.186 0.232 0.381 0.969 0.428 0.113 -0.03 -0.02 0.131 ± 0.067 0.027 0.024 0.041 0.04 0.026 0.012 0.029 0.017 0.018 7 a 0.335 0.134 0.341 0.308 1.026 0.315 0.083 0.017 0.012 0.149 7 17 ± b ± a 0.011 0.567 0.011 0.048 -0 0.047 23 ± 0.067 0.027 0.023 0.038 0.041 0.027 0.012 0.028 0.016 0.018 ± 24 ± 0.067 0.027 0.023 0.042 0.04 0.024 0.013 0.023 0.016 0.027 b b 0.327 0.131 0.327 0.298 1.001 0.356 0.085 0.026 0.017 0.104 ± 40 uM PQ 8 a 8 ± a 0.067 0.322 0.027 0.129 0.023 0.302 0.043 0.132 0.039 0.943 0.022 0.328 0.013 0.099 0.025 -0 0.016 -0 0.019 0.105 15 uM PQ ± 25 ± 0.067 0.027 0.023 0.049 0.035 0.024 0.013 0.024 0.016 0.023 b b 0.344 0.138 0.301 0.283 1.025 0.373 0.103 -0.01 -0 0.131 ± ± 0.067 0.027 0.023 0.044 0.04 0.023 0.013 0.026 0.016 0.02 3 4 5 6 7 8 26 27 28 29 30 31 32 ± b ± a ± b ± a ± b ± a ± b ± a ± b ± a ± b ± a ± b ± 0.01 0.75 0.01 0.72 0.01 0.73 0.01 0.71 0.01 0.72 0.01 0.72 0.01 0.72 0.01 0.77 0.01 0.76 0.01 0.7 0.01 0.7 0.01 0.73 0.01 0.72 0.01 0.05 0 0.04 -0.02 0.04 -0.01 0.04 0.02 0.04 0 0.05 0 0.04 -0.01 0.04 -0.02 0.04 -0.03 0.04 0 0.04 -0.03 0.05 0.01 0.05 -0.04 0.05 0 mM Fe (pre +) 0 mM Fe (pre -) 50 mM Fe 34 ± 0.067 0.027 0.022 0.042 0.036 0.026 0.012 0.031 0.015 0.016 Reactor 1 2 3 a ± b ± a ± b ± a 0.708 0.011 0.693 0.011 0.715 0.011 0.724 0.011 0.027 0.060 0.032 0.059 0.018 0.045 0.029 0.046 control Bacillus subtilis systems biology LST Research Project 2008 87 Bacillus subtilis systems biology B.2 LST Research Project 2008 Net fluxes Net fluxes used for NADPH balancing REACTIONS glucose in zwf yqjI gntZ pgi rpe ywlF pfk-fbaA TK TA TK gap-pgk PGA -> PEP Pyruvate kinase PYR-AcCoA TCA OGA-MAL mdh mae pck anaplerosis acetate TH Respiration acetoin out acetoin in ribo pant growth Reactor 16-1 Control flux error 100.0 30.6 30.6 0.0 65.6 17.0 13.6 82.6 9.9 7.0 9.9 170.1 160.8 169.3 109.9 94.7 85.5 82.9 2.5 14.9 40.5 3.1 0.0 277.8 0.0 0.0 0.0 0.0 6.6 0.0 1.0 1.0 0.0 1.0 0.6 0.6 0.7 0.4 0.4 0.4 1.5 2.0 3.3 3.8 5.7 5.9 5.3 3.2 2.3 3.4 4.4 0.0 9.9 0.0 0.0 0.0 0.0 0.0 Reactor 24-7 15 uM Paraquat flux error 100.0 0.0 45.7 1.0 45.7 1.0 0.0 0.0 50.4 1.0 27.0 0.6 18.7 0.6 77.5 0.7 15.0 0.4 12.1 0.4 15.0 0.4 164.8 1.5 155.4 2.0 163.7 3.3 106.3 3.9 93.7 5.9 84.9 6.1 86.1 5.6 -1.2 2.9 14.7 2.3 35.3 3.0 0.5 4.5 0.0 0.0 274.6 10.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.7 0.0 reactions used for analysis of netto fluxes glucose in: GLC + ATP > Glucose-6-P gluconate bypass: Glucose-6-P + ATP > 6PG + NADPH zwf: Glucose-6-P > 6PG + NADPH yqjI: 6PG > Ru5P + CO2 + NADPH gntZ: 6PG > Ru5P + CO2 + NADH pgi: Glucose-6-P = F6P rpe: Ru5P = X5P ywlF: Ru5P = R5P pfk-fbaA: F6P + ATP > 2*T3P TK1: X5P + R5P = S7P + T3P TA: X5P + E4P = F6P + T3P Appendices 88 Bacillus subtilis systems biology LST Research Project 2008 TK2: S7P + T3P = E4P + F6P gap-pgk: T3P = PGA + ATP + NADH PGA > PEP:PGA > PEP Pyk:PEP > PYR + ATP PYR-AcCoA:PYR > AcCoA + CO2 + NADH TCA: OAA + AcCoA > OGA + CO2 + NADPH OGA-MAL: OGA = MAL + CO2 + 0.5*ATP + 2*NADH mdh: MAL = OAA + NADH mae-NADPH: MAL > PYR + CO2 + NADPH mae-NADH: MAL > PYR + CO2 + NADH pck: OAA + ATP > PEP + CO2 pycA: PYR + ATP + CO2 > OAA acetate: AcCoA > Acetate + ATP TH: NADPH = NADH Respiration: O2 + 2*NADH > 2*PO*ATP acetoin out: 2*PYR > Acetoin + 2*CO2 acetoin in: Acetoin > 2*NADH + 2*AcCoA riboflavin: 2*Ru5P + R5P + 1*PGA + 2*NADPH + 15*ATP > Ribo + 3*NADH ratios Ser from glycolysis = (2*[pfk-fbaA]-2*[TK2]-2*[TA])/(2*[pfk-fbaA]+[TA]+[TK1]) Gnd - NADPH = [yqjI]/([gntZ]+[yqjI]) PEP through PPP (ub) > ([TK1]+2*[TK2]+3*[TA])/(2*[pfk-fbaA]+[TA]+[TK1]) PYR from MAL (ub) > ([mae-NADH]+[mae-NADPH])/([Pyk]+[mae-NADH]+[maeNADPH]) PYR from MAL (lb) < ([mae-NADH]+[mae-NADPH])/([Pyk]+[mae-NADH]+[maeNADPH]) Mae - NADPH = ([mae-NADPH])/([mae-NADH]+[mae-NADPH]) OAA from PYR = [pycA]/([pycA]+[mdh]) PEP from OAA = [pck]/([pck]+[PGA > PEP]) biomass DNA: (0.2573*mu+0.5027)*(289*[R5P]111*[T3P]+747*[E4P]+1992*[PGA]+1493*[PEP]+4482*[PYR]+759*[AcCoA]+1 891*[OGA]+2684*[OAA]+22678*[NADPH]-3772*[NADH]+53248*[ATP]3847*[CO2]) RNA: (0.1714*mu + 0.0484)*(3077*[R5P]+1800*[PGA]+1277*[OAA]+3884*[NADPH]10277*[NADH]+37631*[ATP]+1800*[CO2]) 0.026*(3260*[R5P]+1630*[PGA]+1630*[OAA]+5820*[NADPH]10710*[NADH]+37967*[ATP]+1630*[CO2]) (0.0852*mu+0.0842)*(1929*[T3P]+263*[PGA]+3105*[PYR]+18007*[AcCoA]+1 561*[OAA]+42970*[NADPH]-1292*[NADH]+25383*[ATP]-5552*[CO2]) Appendices 89 Bacillus subtilis systems biology LST Research Project 2008 (-0.0331*mu + 0.0326)*(1603*[G6P]+3624*[T3P]+1683*[PYR]+2364*[AcCoA]+144*[OAA]+54 14*[NADPH]+3344*[NADH]+18843*[ATP]-496*[CO2]) (0.2523*mu+0.2488)*(3011*[G6P]+1076*[T3P]+500*[PEP]+1501*[PYR]+1258* [AcCoA]+500*[OGA]+500*[OAA]+4004*[NADPH]+1076*[NADH]+12521*[ATP]) (-0.0579*mu+0.0571)*(2379*[OGA]+2379*[NADPH]+9517*[ATP]) Appendices 90 Bacillus subtilis systems biology Appendix C LST Research Project 2008 Metabolic pathways Overview of metabolic pathways from (39) Appendices 91 Bacillus subtilis systems biology Appendix D LST Research Project 2008 Mini-scale chemostats - image overview Effluent pumphead connection Feed pumphead connection Appendices 92 Bacillus subtilis systems biology LST Research Project 2008 Top view of reactor setup Top view of reactor setup Appendices 93 Bacillus subtilis systems biology LST Research Project 2008 Overview of the whole chemostat setup with isolated tubing. Appendices 94 Bacillus subtilis systems biology Appendix E LST Research Project 2008 Error Analysis Example: determination cell dry weight ----OD MEASUREMENT Photospectrometer guestimate error: 0.003 endOD = x*ODread x is dilution with pipetting error 1 mL +/- 0.001 0.1 mL +/- 0.001 0.9 mL +/- 0.001 dilution = 0.1/(0.1+0.9) error in dilution = ((0.001^2)*3)^1/2 = 0.002 Metingen 0.113 -> 10*0.113 -> (10*(0.002)^2+0.03^2)^1/2 -> 1.13 +/- 0.03 0.112 -> 1.12 +/- 0.03 averageOD = (OD1 + OD2)/2 averageErr = (0.03^2+0.03^2)^(1/2) = 0.04 AverageOD = 1.13 +/- 0.06 ----CDW determination guestimate error scale 1 gram +/- 0.000001 weight filter: 0.0779 +/- 0.000001 weight filter + biomass: 0.0838 +/- 0.000001 biomass is filter2-filter1 error: (0.000001^2+0.000001^2)^(1/2) = 1.414*10^-6 biomass: 0.0059 +/- 1.414*10^-6 Volume of measured medium: 10 mL pipet with error 0.05 Appendices 95 Bacillus subtilis systems biology LST Research Project 2008 final measured medium volume : 9 mL +/- 0.05 medium density: biomass/volume error: ((1.414*10^-6)^2+0.05^2)^(1/2) = 0.05 medium density (per L): 1000*(0.0059/9) = 0.6556 g/L +/- 0.05 CDW: density/OD error: (0.06^2+0.05^2)^(1/2) = 0.08 CDW: 0.58 g/OD.L +/- 0.08 Appendix F Figure and table list Figures Figure 1-1. Scanning Electron Micrograph of B. subtilis bacteria (60)................. 8 Figure 1-2. Transmission Electron Micrograph of a B. subtilis cell in cross-section (74). ..................................................................................................... 8 Figure 1-3. Schematic of the environmental stress response. Enzymatic activities are italicized, and partner switching modules linked by dashed arrows (30). .......................................................................................................... 13 Figure 1-4. The redox states of oxygen with standard reduction potentials (36). .......................................................................................................... 14 Figure 1-5. Source of oxidative stress. Images shows (a) intracellular autoxidation, (b) environmental redox reactions, (c) H2O2 released by competing micro-organisms, (d) phagosomal NADPH oxidase, (e) redoxcycling antibiotics (36). ......................................................................... 15 Figure 1-6. Mechanisms of damage by reactive oxygen species. (a) the Fentonmediated damage to proteins (rectangles) and DNA, (b) the oxidation of [4Fe-4S]2+ clusters, (c) inhibition of transketolase, and (d) disruption of the sulfur assimilatory pathway. The details of inactivation of transketolase and of sulfur metabolism remain unclear (36). .............................................. 16 Figure 1-7. Most recent model of the bacillibactin-mediated iron acquisition pathway in B. subtilis. Pathway steps I, II, V and VI were functionally characterized. The broken arrows indicate putative pathway steps (47). ... 18 Figure 1-8. Schematic representation of Fur-mediated gene repression (2). ..... 21 Figure 1-9. interconnections between stress sensing and effector systems....... 23 Figure 1-10. The RNA-binding function of the aconitase depends on the iron status of the cell (12). .......................................................................... 24 Figure 1-11. Pathways by which NAD and NADP are generated and affect cellular ROS generation in B. subtilis. Dashed lined indicate ROS formation, solid lines indicate ROS mitigation. Detailed description in text below. The putative transhydrogenase activity in B. subtilis (68) is omitted................ 25 Figure 2-1. Schematic view of a batch reactor. .............................................. 32 Appendices 96 Bacillus subtilis systems biology LST Research Project 2008 Figure 2-2. biomass (solid) and substrate (dashed) concentration during a batch culture. ............................................................................................... 32 Figure 2-3. Schematic view of a continuous reactor. ...................................... 32 Figure 2-4. biomass (X) and substrate (S) concentration during a continuous culture after a batch phase. .................................................................. 32 Figure 2-5. A square-well microtiter plate filled with liquid medium, covered as shown, and incubated on an orbital shaker. Adapted from (19)................ 35 Figure 2-6. Schematic drawing of the chemostat system (50). ........................ 36 Figure 2-7. Actual setup of the mini-scale chemostats in the lab. The reactors are placed in the waterbath (center) and the feed enters from the right side. Effluent is pumped towards the left side................................................. 36 Figure 3-1. Diagram of growth-rate lag issue in batch cultures. The solid line is the expected outcome; the dashed line shows the experimental outcome. The slope of a line represents the growth-rate........................................ 41 Figure 3-2. Citric acid (above) and the encountered complexes with iron (below). Adapted from (55)................................................................................ 42 Figure 3-3. Development of B. subtilis growth in shakeflasks with varying citrate concentrations. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. All media contained 50 µM Fe3+. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. .......................................................................... 43 Figure 3-4. Growth-rates of B. subtilis in shakeflasks experiments with varying citrate concentrations in the medium. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. .......................................................................... 44 Figure 3-5. Culture densities of B. subtilis grown under varying citrate conditions plotted on a logarithmic scale. Curves are averages from two cultures. Values denote the added amount of citrate to the medium. All media contained 50 µM Fe3+. The sample with an extra 250 µM citrate was performed to check the effects of a major increase in citrate concentration. Glucose concentration in this experiment was 3 g/L. ............................... 45 Figure 3-6. Paraquat redox-cycling mechanism (29)....................................... 46 Figure 3-7. The molecular structure of deferoxamine (5)................................ 47 Figure 3-8. Example of a dose response curve. The colored area indicates the sought-after growth conditions.............................................................. 48 Figure 3-9. The effect of varying Paraquat concentrations on the growth of B. subtilis in a batch culture. The dotted line indicates the moment of Paraquat administration. All curves are averaged from three independent wells. Glucose concentration in this experiment was 3 g/L. ............................... 50 Figure 3-10. Dose response curve of B. subtilis. Maximum growth-rate as a function of the applied paraquat concentration. Dotted line represents the Appendices 97 Bacillus subtilis systems biology LST Research Project 2008 hand-drawn trend line. Data derived from at least 6 individual wells per data point. .................................................................................................. 51 Figure 3-11. Growth of B. subtilis in microtiter plates in media with varying iron or deferoxamine (DFO) concentration. All curves are averaged from at least eight independent wells. ....................................................................... 52 Figure 3-12. First chemostat run of B. subtilis with the original setup. All eight chemostats used the same medium and were inoculated from the same preculture. Four different dilution rates were tested in duplicate. Timepoint 0 denotes the feed start. ......................................................................... 53 Figure 3-13. Second chemostat run with new feed tubing. B. subtilis grown in eight chemostats at different dilution rates. All eight chemostats used the same medium and were inoculated from the same preculture. Four different dilution rates were tested in duplicate. Timepoint 0 denotes the feed start.54 Figure 3-14. Overview of end densities as measured after six volume changes arranged at the various dilution rates used in the experiments. Data obtained from eight runs, each run is indicated by its own color. The reactors were grown at different dilution rates but within a range that should produce the same end density. The resulting densities show a major diversion between the experiments........................................................ 57 Figure 3-15. The distribution of densities over the dilution rates after 6 volume changes in two successive runs. Numbers indicate the specific reactor tube. .......................................................................................................... 58 Figure 3-16. Split scheme of the feed tubing from the medium bottle through the pump to the individual reactors on the right. .......................................... 58 Figure 3-17. Final optical densities after 6 volume changes of reactors connected to individual feed tubing. Figure based on two chemostats experiments.... 59 Figure 3-18 Extracellular fumarate concentrations in B. subtilis Paraquat reactors. Data derived from single reactors. ......................................................... 61 Figure 3-19. Flux ratios in B. subtilis under increasing oxidative stress. Fluxes are composed of data from different reactors under equal conditions. 10 µM: 5 reactors, 15 µM: 7 reactors, 20-25 µM: 2 reactors. The errors indicate deviations between various reactors. PQ: Paraquat. ................................ 62 Figure 3-20. Relative distribution of absolute fluxes in B. subtilis under control conditions and under oxidative stress. Conditions: µ = 0.29 h-1 / qs,PQ= 4.3 mmol·g-1·h-1 / qs.control = 4.4 mmol·g-1·h-1. Glucose concentration 1 g/L. Fluxes are normalized to glucose uptake rates. Upper number in the boxes are from the control culture, lower numbers are derived from a 15 µM paraquat reactor. ................................................................................. 63 Figure 3-21. Comparison of NADPH production and consumption rates in B. subtilis grown with and without 15 µM Paraquat. Bars represent the estimated formation and consumption from net analysis. Both samples were grown at µ = 0.29. The glucose uptake rate of the control was 4.4 mmol · g1 · h-1 and of the Paraquat reactor 4.3 mmol · g-1 · h-1. ........................... 64 Appendices 98 Bacillus subtilis systems biology LST Research Project 2008 Figure 3-22 Flux ratios in B. subtilis under iron limited conditions. Fluxes are composed of data from different reactors under equal conditions. DFO: 4 reactors, Ironless: 4 reactors. The errors indicate deviations between various reactors. DFO: deferoxamine................................................................. 65 Figure 4-1. Diagram of the oxidative pentose phosphate pathway and related reactions in B. subtilis. The gluconate bypass is depicted as a dotted line and follows from glucose via gluconate to 6-phospho-gluconate (76). ............. 70 Tables Table 1-1. Enzymes in B. subtilis known to be vulnerable to oxidative stress and their function (8).................................................................................. 17 Table 1-2. Selection of enzymes with iron atoms in B. subtilis and their function derived from (8)................................................................................... 19 Table 1-3. TCA cycle, possible target enzymes of oxidative stress and iron limitation ............................................................................................. 24 Table 1-4. Oxidative phosphorylation, possible target enzymes of oxidative stress and iron limitation ................................................................................ 24 Table 2-1. Details of used strain................................................................... 27 Table 3-1. Maximum growth rate of B. subtilis in shakeflasks under citrate varied conditions. Data derived from figure 3-5. ............................................... 45 Table 3-2. Concentrations of stressors tested to obtain a dose response curve. Values marked in bold-face are the indications derived from literature. ..... 49 Table 3-3. Overview of adaptations and checks performed in order to obtain a properly functioning mini-scale chemostat. ............................................. 54 Table 3-4. Biomass yields and specific glucose uptake rates for B. subtilis grown under oxidative stress (Paraquat) and iron limitation (deferoxamine) at a growth-rate of µ = 0.29 h-1 (± 0.02). Column ‘n’ denotes the number of reactors used in the calculation. ............................................................ 60 Appendices 99