Université de Mons-Hainaut Faculté des Sciences Service de Protéomie et Biochimie des Protéines

Centre d'étude de l'Energie Nucléaire Unité de Microbiologie

Molecular characterization of the life support bacterium S1H cultivated under space related environmental conditions

Dissertation originale présentée par

Felice Mastroleo en vue de l'obtention du grade de docteur en Sciences

Directeur de thèse: Prof. R. Wattiez

Co-directeur: Dr. Ir. R. Van Houdt (SCK•CEN)

Janvier 2009

"De la Terre à la Lune". Jules Verne. Paris, 1865.

REMERCIEMENTS

Mes premiers mots seront pour honorer la mémoire du Dr. Larissa Hendrickx, qui m'a fait confiance, voici bientôt 5 ans, et m'a donné la chance de commencer cette aventure.

Je tiens à remercier tout particulièrement les professeurs Ruddy Wattiez et Max Mergeay qui m'ont accueillit au sein de leur service respectif. Ils m'ont permis d'aborder un sujet captivant pour lequel mon intérêt ne s'est pas démenti, que du contraire, au cours de ces années. Leur emploi du temps surchargé ne les a pas empêché d'être là pour moi lorsque le besoin s'en faisait sentir, je leur en suis pour cela très reconnaissant.

J'exprime également toute ma gratitude envers les Drs. Natalie Leys et Rob Van Houdt qui ont su reprendre le flambeau de leur prédécesseur avec panache et enthousiasme, malgré un contexte difficile. Leur soutien m'a été d'un grand secours.

Je remercie également le SCK •CEN qui a permis le financement de cette thèse grâce à une bourse AWM.

Un grand merci également à tous les membres du service de Protéomie et Biochimie des Protéines de l'UMH, ainsi qu'à tout le groupe de Biologie Moléculaire et Cellulaire du SCK•CEN.

Je dédie ce travail à ma famille et à mes parents en particulier qui m'ont permis d'entreprendre et de mener à leur terme ces études universitaires.

CONTENT

LIST OF ABBREVIATIONS

RESUME

SUMMARY

1 – INTRODUCTION 1

2 – OBJECTIVES OF THE WORK 33

3 – WHOLE PROTEOME STUDY 35

4 – STUDY OF R. RUBRUM S1 H IN SPACE RELATED ENVIRONMENTAL CONDITIONS 75

A – THE MESSAGE 2 EXPERIMENT 75

B – THE BASE-A EXPERIMENT 101

C – R. RUBRUM S1H LIQUID CULTURE IN MODELED MICROGRAVITY 121

5 – GENERAL CONCLUSIONS AND FUTURE PERSPECTIVES 145

REFERENCES 151

ADDITIONAL DATA ARE PRESENTED IN THE ATTACHED CD-ROM

ABBREVIATIONS

2D-PAGE Bi-dimensional polyacrylamide gel electrophoresis aaRNA Amino allyl RNA ACLAME A classification of mobile genetic elements ACN Acetonitrile AHL N-acyl-L-homoserine lactone APS Ammonium persulfate aRNA Amplified RNA ATCC American type culture collection BASE-A Bacterial adaptation to the space environment – part A BCHL BLiSS Bioregenerative life support system CDS Coding sequence CELSS Closed ecological life support system CHAPS 3-[(3-Cholamidopropyl)dimethylammonio]-1-propanesulfonate CME Coronal mass ejection COG Cluster of orthologous group CRT DAE Dark aerobic DSB Double strand break DTE Dithioerythritol EDTA ethylenediaminetetraacetic acid emPAI Exponentially modified protein abundance index ERB Earth radiation belt ECF Extracytoplasmic function ESA European space agency FDR False discovery rate GCR Galactic cosmic rays Gy Gray H/L Heavy (H) to light (L) ICPL reagent form ratio HARV High aspect to ratio vessel HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HPLC High-performance liquid chromatography HZE high charge (Z) and high energy (E) particles ICC Ion charge current ICM Intracytoplasmic chromatophore membrane ICPL Isotope-coded protein label IEF Isoelectric focusing IPG Immobilized pH gradient IR Ionizing radiation ISS International space station LAN Light anaerobic LEO Low Earth orbit LET Linear energy transfer LSMMG Low shear modeled microgravity MaGe Magnifiying genome MALDI-ToF Matrix assisted laser desorption ionization-time of flight MELiSSA Micro-ecological life support system alternative MESSAGE 2 Microbial experiment in the space station about gene expression - part 2 MS Mass spectrometry MS/MS Tandem mass spectrometry MudPIT Multidimensional protein identification technology MW Molecular weight NASA National aeronautics and space administration NL Non linear OD optical density OSLD Optically stimulated luminescence detector PHB Polyhydroxybutyrate pI isoelectric point RBE Relative biological effectiveness RC Reaction center complex ROS Reactive oxygen species RP Reverse phase RPM Random positioning machine RT-qPCR Real time-quantitative polymerase chain reaction RuBisCO ribulose bisphosphate carboxylase RWV Rotating wall vessel SAA South atlantic anomaly SB3-10 Sulfobetaine-10 SCX Strong cation exchange SDS-PAGE Sodium dodecyl sulfate - polyacrylamide gel electrophoresis SILE Stable isotope labeling experiment SimRAD Simulation of ionizing radiation SPE Solar particle event SPY Sistrom peptone yeast Sv Sievert TBP Tributyl phosphine TED Track-etch detector TEMED N,N,N',N'-Tetramethylethylenediamine TFA Trifluoroacetic acid TGS Tris-glycine-SDS TLD Thermoluminescent detector

RESUME

Introduction . MELiSSA (acronyme pour Micro-Ecological Life Support System Alternative) est un système clos, développé par l'Agence Spatiale Européenne pour la régénération des consommables, et qui servira de support de vie pour les futurs voyages spatiaux de longue distance. Il est composé de processus interconnectés (bioréacteurs, compartiment pour plantes supérieures, unités de filtration, etc ) ayant pour objectif de recycler les déchets organiques en oxygène, en eau et en nourriture. Au sein de la boucle MELiSSA, l' α-protéobactérie pourpre non-sulfureuse R. rubrum S1H est utilisée pour convertir les acides gras volatiles, en provenance du digesteur situé en amont, en CO 2 et en biomasse. De plus, elle complète la + minéralisation des acides aminés en NH 4 , le tout étant ensuite transmis au compartiment nitrifiant situé en aval. Parmi les nombreux défis du projet, la stabilité fonctionnelle des bioréacteurs en conditions de vol spatial est d'une importance capitale pour assurer l'efficacité du système et par conséquent la sécurité de l'équipage. Lors de cette étude, nous avons caractérisé les changements induits chez R. rubrum S1H par les conditions environnementales rencontrées pendant un vol spatial. Matériel et méthodes . R. rubrum , inoculée sur milieu agar solide en condition d'obscurité et aérobie, a séjourné dans la Station Spatiale Internationale (ISS) en octobre 2003 (expérience MESSAGE 2) et en septembre 2006 (expérience BASE-A). De plus, dans un effort pour identifier une réponse spécifique de R. rubrum au vol spatial, nous avons réalisé des simulations terrestres des paramètres de radiation ionisante et de gravité rencontrés dans l'espace, utilisant des conditions de culture et de croissance identiques à celle rencontrées lors du vol spatial proprement dit. Ces cultures ont ensuite été analysées et comparées à leur contrôle respectif au niveau transcriptomique et protéomique. Pour se faire, nous avons développé et utilisé respectivement une puce à ADN de type oligonucléotidique, couvrant l'ensemble du génome de R. rubrum , et une approche protéomique différentielle à haut débit et sans gels couplée à un marquage isotopique de type ICPL (acronyme pour Isotope-Coded Protein Label). A côté de cela, l'entièreté du contenu en protéines de R. rubrum a été caractérisée à l'aide de la technologie MudPIT (acronyme pour Multidimensional Protein Identification Technology). Résultats . Nous avons pu mettre en évidence l'importance de la composition du milieu et du dispositif de culture dans la réponse de cette bactérie aux conditions environmentales liées au

vol spatial. De plus, nous avons montré pour la première fois qu'une faible dose de radiation ionisante (2 mGy) pouvait induire une réponse significative au niveau transcriptomique, mais sans altérer la viabilité de la cellule et en induisant un nombre restreint de protéines significativement régulées. A côté de cela, nous avons mis en évidence l'expression de gènes liés à la détection du quorum (quorum sensing) et à l'appareil photosynthétique de R. rubrum , au niveau transcriptomique et protéomique, lorsque la bactérie est cultivée dans l'obscurité et en condition aérobie, dans 2 dispositifs différents simulant la microgravité. Enfin, la technique MudPIT nous a permis d'identifier environ 25 % (c'est-à-dire 1.007 protéines) du protéome de R. rubrum . Associés à l'ensemble des données obtenues en étudiant une variété de conditions environmentales, ces derniers résultats seront particulièrement précieux pour l'annotation experte du génome de R. rubrum , qui compte toujours 25 % de gènes annotés comme codant pour des protéines hypothétiques. Conclusion . Notre approche, qui a intégré des données transcriptomiques et protéomiques, a permis d'évaluer la sensibilité de R. rubrum S1H à différentes conditions environmentales (vol spatial, radiation ionisante et microgravité simulée). De plus, dans la perspective d'une utilisation de cette bactérie dans un écosystème clos en conditions de lumière et anaérobie, une étude plus approfondie de la détection du quorum doit être envisagée. Finalement, des simulations terrestres impliquant la culture en bioréacteur nous apparaissent cruciales pour confirmer des changements métaboliques chez R. rubrum .

SUMMARY

Background . MELiSSA, which stands for 'Micro-Ecological Life Support System Alternative', is a closed regenerative life support system for future space flights under development by the European Space Agency. It consists of interconnected processes ( i.e. bioreactors, higher plant compartments, filtration units, etc.) targeting the total recycling of organic waste into oxygen, water and food. Within the MELiSSA loop, the purple non-sulfur α-proteobacterium R. rubrum S1H is used to convert fatty acids released from the upstream raw waste digesting reactor to CO 2 and biomass, and to complete the mineralization of amino + acids into NH 4 that will be forwarded to the nitrifying compartment. Among the numerous challenges of the project, the functional stability of the bioreactors under space flight conditions is of paramount importance for the efficiency of the life support system and consequently for the crew safety. In this study, changes induced by space flight related environmental conditions were investigated for R. rubrum S1H. Material and methods . R. rubrum S1H inoculated on solid agar medium in dark aerobic conditions was sent to the International space station (ISS) in October 2003 (MESSAGE 2 experiment) and September 2006 (BASE-A experiment). Moreover, in an effort to identify a specific response of R. rubrum to space flight, ground simulation of space ionizing radiation and space gravity were performed under identical culture setup and growth conditions encountered during the actual space flight experiment. The cultures were analysed and compared with their corresponding control at both the transcriptomic and proteomic level using respectively a newly developed R. rubrum whole genome oligonucleotide microarray and high throughput gel free proteomics using the Isotope-Coded Protein Label (ICPL) approach. In addition, R. rubrum whole protein content was further characterized using a multidimensional protein identification technology (MudPIT). Results . We could put forward the importance of the medium composition and the culture setup on the response of the bacterium to the space flight related environmental conditions. Moreover, we showed for the first time that a low dose of ionizing radiation (2 mGy) can induce a significant response at the transcriptomic level, although no change in cell viability and only a few significant differentially expressed proteins were observed. Besides, we detected the induction of quorum sensing and photosystem related genes , at both the

transcriptomic and proteomic level, when the bacterium was cultivated in liquid dark aerobic conditions in 2 different microgravity simulators. On the other hand, the MudPIT approach permitted us to identify about 25 % ( i.e . 1,007 proteins) of R. rubrum 's proteome. Together with the data gained from the variety of environmental conditions tested, these results will be particularly valuable for the expert genome annotation of the R. rubrum genome where 25 % of all genes are still annotated as coding for hypothetical proteins. Conclusion. Our approach integrating transcriptomic and proteomic data showed R. rubrum S1H susceptibility to various environmental conditions (space flight, ionizing radiation and low-shear modeled microgravity). Moreover, in the perspective of using R. rubrum in a closed-loop ecosystem under light anaerobic conditions, future studies of quorum sensing should be foreseen. Eventually, ground simulations involving culture in bioreactors are crucial to confirm change in R. rubrum S1H metabolism.

CHAPTER 1 - INTRODUCTION

1.1 SPACE EXPLORATION

Last year, with the birthday of Sputnik launched on October the 4 th 1957, we celebrated the fiftieth anniversary of space exploration. This major achievement seems only to have occurred as a side effect in a series of events. Indeed, without any doubt, the rocket originated from the war and in particular from the Nazi's huge means to bombard London and Paris using missiles. Depending on the point of view, space research originated from the Cold War and from the Soviet government desire to be able to launch a nuclear bomb to the United States, or it originated from the early work of Konstantin Eduardovich Tsiolkovsky, pioneer of astronautics (Encyclopædia Britannica, 2009). With regards to Sputnik, it came by chance from a reader of Jules Verne, Sergueï Pavlovitch Korolev, to reach an old dream: to travel in space (Ciel et Espace, 2007). This last event starts the Space Race where the Soviet space program achieved many of the first milestones (Figure 1.1): first animal within a satellite on the 3 rd of November 1957; first space ship to hit the Moon on the 12 th of September 1959; first living organism to go and come back from space on the 19 th of Augustus 1960; Yuri Gagarin is the first man to travel in space on the 12 th of April 1961; the first woman is placed in orbit on the 16 th of June 1963; the first moon landing on the 3 rd of February 1966; the first men on the Moon by the U.S.A on the 20 th of July 1969; the rendezvous on the 17 th of July 1975 of the Apollo and Soyuz spacecraft traditionally marks the end of the Space Race. After the first 20 years of exploration, focus shifted from one-off flights to renewable hardware such as the Space Shuttle program allowing large payloads to be carried to various orbits. With the manned orbital space stations, a new era in space exploration arose. Space stations are designed for living in orbit medium-term, for periods of weeks, months, or even years. Previous ones are the Almaz (Russia), Salyut series (Russia), Skylab (U.S.A.) and Mir (Russia). The International Space Station (ISS), a collaborative effort between the U.S.A., Russia, Japan, European Space Agency (ESA), Canada and Brazil is currently the only continuously manned station.

Chapter 1 – Introduction 2

Laïka, 11/03/1957 Valentina Sputnik, 10/04/1957 Iouri Gagarine, Terechkova, 1961 1963

The National Air and Space Museum display of the Apollo -Soyuz 'rendezvous' , 1975

Apollo Lunar Excursion Module on lunar surface, 1969

The launch of a Soyuz Soyuz spacecraft

Space Shuttle launch, TMA atop a Soyuz- TMA version , 2002 1981 FG rocket, 2001

Mir Space Station, 2000

International Space Station, june 2008

Figure 1.1 . Milestones in the space exploration program (Wikipedia and Zarya websites).

Chapter 1 – Introduction 3 Space stations are currently used to study the effects of long-term space flight on the human body as well as to provide platforms for a great number of scientific studies not possible on other space vehicles. In that respect one can imagine the final purpose of sending men to space. Unmanned missions with space probes, telescopes, and other instruments, resulted not only in revelations and sheer wonder but also produced transformational benefits to humanity in Earth observation, weather forecasting, navigation, and telecommunications. Unmanned missions don't need space suits, radiation shields, toilets, exercise bikes, a bail-out system during launch, or consumables other than energy. There's no argument when it comes to costs: Cassini, NASA's up-market robotic mission to Saturn, costed the U.S.A 3.4 billion dollars, and this investment studies the ringed planet and its moons since 2004 and will do so for at least another decade. By contrast, shuttle flights, which last around two weeks, costed about a billion dollar apiece in 2002, the last year before the Columbia tragedy, and costs have soared since then. «Why bother with human space flight, when robots do such a good job and do it so much more cheaply? Proponents of human space flight argue that only humans have the supple physical coordination and mental agility to get the most from an expedition. But the most compelling argument for human space flight may remain the one that worked at the beginning. Space exploration is ultimately about human dreams.» ( Space at 50 , National Geographic, October 2007). Moreover, manned space flight obeys political and strategic purposes. Space is a way to show technology superiority on the international scene. In that context, the European Space Agency's Aurora Program (ESA, 2001) and the National Aeronautics and Space Administration's Vision for Space Exploration (NASA, 2004) have as primary objective to create, and then implement, a long-term plan for the robotic and human exploration of the solar system, with Mars, the Moon and the asteroids as the most likely targets. A second objective is to search for life beyond Earth. Future missions under these programs will carry sophisticated exobiology payloads to investigate the possibility of life forms existing on other worlds within the solar system. It is clear from these objectives that the interdependence of exploration and technology is the basis of such programs. On the one hand the desire to explore provides the stimulus to develop new technology while on the other, the introduction of innovative technology will make exploration possible. Mars was chosen because it is the most Earth-like of the eight planets that make up our solar system and recent indications of the presence of water (Holt et al ., 2008) raise the likelihood to find traces of life. It represents, however, a major leap for humankind.

Chapter 1 – Introduction 4 To give some idea of how far it is: Mars is an estimated 400,000,000 km distant from the Earth at the farthest point of its orbit; the Moon is around 400,000 km, and the ISS is 'nearby' at approximately 400 km away from the Earth. With today’s technology it will take over two years to reach Mars and to return. Depending on the orbit strategy, it is possible to either shorten the travel time or to shorten the stay to up to two months. In the first case, the crew will be obliged to stay over a year and a half on Mars to await the next return opportunity, while in the second case, more than two years will be spent in traveling to and from Mars and the increase in velocity needed for the return trip will be much higher. To undertake such a mission will require tremendous efforts of organization, logistics and technological development. How will the astronauts survive for such a long period in an unfriendly environment? What will they eat, what will they drink and, not to forget, how will they deal with waste products? Finally, not the least of the problems will be learning to cope with the psychological pressure and stress of living in a confined space, for a long period of time, with a small number of colleagues. Research and simulation on the ground, as well as experience gained from working on orbital space station (Mir and ISS), will all help to meet and overcome these difficulties.

1.2 THE SPACE ENVIRONMENT

1.2.1 THE RADIATION ENVIRONMENT Astronauts, cosmonauts and taikonauts aboard a low Earth orbit (LEO) spacecraft such as NASA's Space Shuttle, ISS or on board spaceships travelling outside the Earth's magnetosphere (>70,000 km away from the Earth) are exposed to levels of radiation far higher than those encountered on Earth. Moreover these radiation conditions are dependent on a large number of parameters including the altitude and inclination of the spaceship's orbit, the orientation of the spaceship relative to the Earth and Sun and the particular phase of the 11-year solar cycle (Benton and Benton, 2001). The next section will focus on ionizing radiation (Table 1.1), since ultra-violet radiation is effectively shielded and only of concern outside a spaceship (and possibly through windows).

Chapter 1 – Introduction 5

1.2.1.1 IONIZING RADIATION ENVIRONMENT IN LOW EARTH ORBIT

The three principal sources of ionizing radiation in LEO are the following: (1) Galactic cosmic rays that are charged particles (protons, α-particles and, high charge Z and high energy E (HZE) particles) originating from beyond the solar system; (2) Energetic electrons and protons that are trapped in the geomagnetic field and make up the Earth's radiation belts (ERBs) and (3) Solar particle events that are high fluxes of charged particles (mostly low- energy protons and α-particles) encountered during rare but intense solar flares and Coronal mass ejections (CME) (Nicholson et al. , 2000; Benton and Benton, 2001). When passing through the skin and structure of a spaceship, primary ionizing particles can undergo interactions with nuclei that constitute the spaceship's mass, producing a wide variety of secondary particles: neutrons, protons, recoil nuclei, projectile fragments, γ-particles, etc. (Figure 1.2). While the number of different particle species is large and the energy spectrum they occupy quite broad, their fluxes are often low. But relatively rare events associated with solar flares and CMEs can produce sudden and dramatic increases in flux. Spaceship shielding is thus one of the most important factors in determining the characteristics of the ionizing space radiation inside a spaceship (Benton and Benton, 2001).

.

Figure 1.2 . Transport of primary radiation through the spacecraft structure and contents, and generation of secondaries. The most important secondaries in terms of radioprotection are neutrons and the high-LET target and projectile fragments (Benton and Benton, 2001). LET : linear energy transfer. SAA : South Atlantic Anomaly. GCR : galactic cosmic rays. SPE : solar particle events. HZE : high charge ( Z) and high energy ( E) particles.

Chapter 1 – Introduction 6 Early measurments on the Mir Orbital Station taken during a period of eleven years (1986 to 1997) showed dose rates ranging from 162 to 508 µGy/day (Benton and Benton, 2001). Monitoring radiation on board the International Space Station (ISS) in low Earth orbit ( ca . 400 km altitude) indicated an average dose of 180 µGy/day (Goossens et al ., 2006; Vanhavere et al ., 2008). While background radiation levels on Earth which come from a combination of terrestrial (from 40 K, 232 Th, 226 Ra, etc .) and cosmic radiations (photons, electrons, etc .), are fairly constant over the world, being 2-4 µGy/day (ISU, 2008). Hence, on board the ISS, organisms undergo radiation stress up to 40-fold higher than in their usual terrestrial environment.

1.2.1.2 IONIZING RADIATION ENVIRONMENT BEYOND THE MAGNETOSPHERE

The ionizing radiation environment outside the Earth's magnetosphere differs markedly from that encountered in LEO. Not only are the trapped radiation belts absent, but also interplanetary space lacks the protection afforded by the Earth's magnetic field and the associated geomagnetic cut-off. As an example, the Apollo measurements reflect the specific trajectory taken by the spacecraft through the Earth's radiation belts in transit to and from the Moon. The mean dose rate for the Apollo missions ranged from 220 to 1270 µGy/day (Benton and Benton, 2001). Thus, in deep space the ionizing radiation stress can be up to 300-fold higher than that encountered in terrestrial environments.

1.2.1.3 RADIATION ON MARS

As already mentioned, Mars is regarded as the most interesting planet in our solar system in a search for extraterrestrial life. This is mainly based on the fact that the early histories of Mars and Earth show similarities during the period when life emerged on Earth. Mars lacks a strong magnetic field to deflect lower energy particles. Thus, on the Martian surface, exposure to space radiation, while reduced compared to that encountered in deep space, is nonetheless omnipresent (Benton and Benton, 2001). Today, our knowledge about the radiation conditions on the surface of Mars and their biological effectiveness is still based on radiation measurements by instruments residing in Mars' orbit (Horneck et al ., 2001; Saganti et al ., 2002). However, the landing of the NASA's rovers 'Spirit' and 'Opportunity' and recently the Phoenix Lander will certainly extend our knowledge. Mars Odyssey Orbiter

Chapter 1 – Introduction 7 provided mean data (for the period from 03/13/2002 to 09/30/2003) ranging from 200-300 µGy/day to up to 20,000 µGy/day during solar particle events (NASA, 2005).

Table 1.1: Physical conditions prevailing on Earth, in low Earth orbit ( LEO ) ( ≈400 km altitude), in interplanetary space (>70,000 km) and on Mars ( ≈400,000,000 km) (Different sources see text above). Physical condition Earth LEO Interplanetary space Mars Ionizing (µGy/day) 2-4 160-500 a 220-1270 200-300 (max.20,000) Microgravity (g) 1 10 -3 – 10 -6 <10 -6 0.4 a: inside space station.

1.2.2 THE MICROGRAVITY ENVIRONMENT The space gravity environment is dramatically different from the terrestrial environment. -6 -3 -6 Gravity is comprised between 10 g and 10 g in low Earth orbit (LEO), and < 10 g in interplanetary space. This is to be compared to 1 g on Earth (Nicholson et al ., 2000) (Table 1.1). Life has evolved on Earth in a periodically changing environment. Climatic eras, natural cataclysms, cosmic radiation, and volcanism have influenced the development and adaptation of living organisms. However, one environmental factor has remained constant throughout the ages, namely the gravitational force. Therefore, one can expect that the transition from 1g to near 0g conditions may provoke unique effects on cells (Cogoli et al ., 1990).

1.3 LIFE SUPPORT FOR SPACE EXPLORATION

Redrafted after: Hendrickx L., De Wever H., Hermans V., Mastroleo F ., Morin N., Wilmotte A., Janssen P., Mergeay M. 2006. Microbial ecology of the closed artificial ecosystem MELiSSA (Micro-Ecological Life Support System Alternative): Reinventing and compartimentalizing the Earth's food and oxygen regeneration system for long- haul space exploration missions. Res. Microbiol. 157 (1):77-86.

1.3.1 LIFE SUPPORT SYSTEMS : CONCEPT AND PURPOSE From the Sputnik 2, with a hermetically sealed air conditioned compartment for the female Eskimo dog Laika 50 years ago, to the current 'luxurious' ISS habitat providing optimal comfort and safety to the crew of up to six people, it is clear that living organisms cannot enter the hostile environment of space without taking a little bit of 'Earth' along with them. The equipment that is needed to regulate the required environmental conditions is broadly defined as life support systems.

Chapter 1 – Introduction 8 Principally, life support systems can be divided in five main areas: (i) atmosphere management, (ii) water management, (iii) food production and -storage, (iv) waste management, and (v) crew safety (Eckart, 1996a). Especially the needs for water supplies in a closed artificial ecosystem immediately reflect upon the intensity of the challenge that life support systems have to cope with. For each day and each person, an amount of 3.56 kg of potable water and 26.0 kg of hygiene water is needed. Thanks to human ingenuity, 96.5 % of these water supplies can now be recuperated and recycled (Krump and Janik, 1992). Currently, water is purified and used again with the help of physico-chemical processes. Even water produced by the fuel cells of the space shuttle (e.g . during taxi flights) and urine is being recycled into potable water (NASA). However, once it is impossible to rely on constant re-supply of water, gas or food hauled from Earth ( i.e . by distance limitations), only bioregenerative life support systems capable to convert metabolic waste products into edible biomass will be able to support manned space exploration (Gustafson et al ., 1989). Several bioregenerative testbeds have been investigated since 1961. They consist of compartments allowing gas exchange between CO 2 and O 2 producing organisms (BIOS 1-2) (Eckart, 1996b), additional biological water purification and food production in BIOS 3 and in the US Closed Ecological Life Support System (CELSS), and the complete Earth-based ecological system for water recovery, plant growth, and domestic animal cultivation in the Biospheres and the Japanese Closed Ecology Experiment Facility (Eckart, 1996a). The European MELiSSA project (Micro Ecological Life Support System Alternative) was the first of its kind, offering an engineering approach in the development of a compact bioregenerative system for organic waste recycling with the use of microorganisms living in interconnected controllable bioreactors (Mergeay et al ., 1988).

1.3.2 DESIGNING THE MEL ISSA LOOP The driving elements for the design of an ideal bioregenerative life support system (BLiSS) are (i) the production of a highly nutritious biomass and O 2, (ii) with the direct use of light as a source of energy for microbiological biosynthesis, (iii) with limited O 2 consumption, (iv) in an easy to handle compact reactor setup adapted for space flight, (v) allowing an efficient and biosafe re-conversion of waste, CO 2, and minerals in simplified recycling steps, (vi) using known and already studied organisms that (vii) preferably possess a certain degree of 'space robustness'. A typical balanced diet contains carbohydrates, fat, and proteins as well as a number of vitamins and minerals. Hence, the most important criteria for the nature of the edible biomass

Chapter 1 – Introduction 9 are the high production rate and quality of its nutritious compounds. The organisms should equally possess a high rate of O 2 evolution and photosynthetic CO 2 fixation, combined with a limited volumic ratio. Cyanobacteria are the ideal candidate organisms as they generally fulfill the above stated conditions. They are a group of evolutionary ancient, morphologically diverse and ecologically important , which are able to carry out oxygenic , while using CO 2 as sole source of carbon (Garcia-Pichel, 2000). They made a significant contribution to the global primary production of the oceans and have become locally dominant primary producers in many extreme environments, like hot and cold deserts, hot springs, and hypersaline environments. Among the traditionally consumable cyanobacteria are Arthrospira , the terrestrial cyanobacterium Nostoc commune , and the symbiotic association Anabaena/Azolla . Arthrospira (cfr. Spirulina cakes) was grown and eaten in pre-Hispanic Mexico by the Aztecs as "tecuitlatl" or by the Kanembu tribes-women as "Dihé" cakes harvested from small soda lakes in the vicinity of Lake Chad, because of its extraordinary nutritional value (Garcia-Pichel, 2000; Baurain et al ., 2002). The main dietary features of spirulines are a very high content of proteins (up to 70 % dry weight) with a well balanced aminogram and a high content in vitamins and essential unsaturated fatty acids (Sautier and Tremolière, 1976). Culturing of Arthrospira sp . is easy and very productive as their growth rate tends to be very high. In addition, they are able to produce 20 times more proteins per hectare as compared to soja (Henrikson, 1997).

Hence, the easily cultivated food product Arthrospira sp. PCC8005 was selected as O 2 producer and essential food supplement in the conceptual design of the MELiSSA loop (Figure 1.3, table 1.2) (Mergeay et al ., 1988). In order to provide the consumers with all the requirements of a complete diet, a plant compartment, achieving the same goals as the Arthrospira sp . compartment, containing lettuce, red beet, and wheat was added to the loop (Figure 1.3).

With the choice of Arthrospira sp . and a plant compartment as sources of food and O 2 producing compartments, the first two conditions for the design of an ideal BLiSS were met.

In additional compartments it was necessary to focus on minimal O 2 consumption in simple and compact reactor processing systems enabling efficient and safe waste recycling. In order to fulfill these requirements we first should understand the nature of the waste that will be processed by the consumers. Unedible plant material as well as hygienic paper and fecal material will enter the MELiSSA loop. On Earth, similar waste streams such as sewage, organic slurries, manure and municipal solid wastes are typically treated by anaerobic

Chapter 1 – Introduction 10 digestion (Verstraete and Vandevivere, 1999). This process involves several subsequent steps performed by different groups of bacteria. Liquefaction or hydrolysis is the transformation of organic polymers by extracellular enzymes into monomers which can enter the bacterial cells.

(a)

(b)

Figure 1.3 . (a) Scheme of the MELiSSA loop (Lattenmayer, 2001). (b) Scheme presenting the main functional steps in the recycling of organic waste in a natural lake ecosystem. The MELiSSA system follows the four steps as indicated in the scheme in separate reactors using (1) an anaerobic thermophilic consortium, (2) the photoheterotrophic R. rubrum , (3) an Nitrosomonas europaea , Nitrobacter winogradskyi co-culture, and (4) the cyanobacterium Arthrospira species and a selection of plants, separately cultivated, to produce food, oxygen and H2O for (5) the human crew (Hendrickx and Mergeay, 2007).

During acidogenesis, the monomers are broken down into intermediate fermentation products (Table 1.2). The latter are mainly volatile fatty acids such as propionic acid, butyric,

Chapter 1 – Introduction 11 isobutyric, valeric and isovaleric acid. Acetogenesis and methanogenesis are the two final transformation steps which result in the production of CO 2 and energy-rich CH 4 (Smith et al ., 1980). However, due to its flammable nature, the production of methane is far from ideal in Life Support systems for closed environments. A partial anaerobic transformation, with a maximal production of volatile fatty acids is a far better option, because these can be used in downstream compartments for edible biomass production. Partial anaerobic digestion can either be achieved by inhibiting methanogenic activity - e.g. by reactor operation under slightly acidified conditions - or by making up consortia of selected non-methanogenic bacteria. In MELiSSA, preference was given to the first option.

Table 1.2. Chemical conversion describing the C, O, N and S cycles in the bacterial compartments ( BC ) of the life support system MELiSSA (Hendrickx et al., 2006).

Anaerobic processes can take place in mesophilic or thermophilic conditions. Metabolic pathways were found to be identical for both (Ahring et al ., 2001). However, digestion under thermophilic conditions has the advantage of higher metabolic rates, higher specific biogas production rates and a higher destruction of pathogens (Bendixen et al ., 1994; Burtscher et al., 1998; Burtscher et al ., 2003; Duarte et al ., 1992; Engeli et al ., 1993; Watanabe et al ., 1997).

Chapter 1 – Introduction 12 On the other hand, the thermophilic process is less stable than the mesophilic one and has higher energy requirements (Chachkhiani et al ., 2004; Elmashad et al ., 2001). However, thermophilic anaerobic digestion is now commonly applied in Europe for treatment of manure, mixtures of manure and organic wastes or the organic fraction of municipal solid waste, mainly for its better sanitizing effect (Hansen et al ., 1999). For the same reason, the MELiSSA substrate is treated in thermophilic conditions in the first compartment. An inoculum was cultivated by natural selection from fecal material, which is currently still being used. Liquefaction is hence achieved using the microbial consortium of the feed itself. In the second compartment of the MELiSSA loop, the low molecular carbon-compounds, primarily volatile fatty acids, as well as CO 2, H 2 and H 2S coming from the first compartment need to be treated (Figure 1.3, Table 1.2). This is achieved using the photoheterotrophic anoxygenic bacterium Rhodospirillum rubrum ATCC25903 (this strain is presented in detail in paragraph 1.4). Apart from Arthrospira sp ., the MELiSSA loop also contains a higher plant compartment for food production. Both the cyanobacteria and the plants preferentially take up as nitrate. Hence, the ammonium enriched waste of the MELiSSA consumer compartments needs to be nitrified to nitrate. Nitrification is generally performed by aerobic autotrophic micro-organisms using carbon dioxide from the atmosphere as the carbon source. Ammonium is oxidized first to nitrite by oxidizing micro-organisms and then nitrite is converted to nitrate by nitrite oxidizers. Both oxidation processes generate energy for the respective organisms (Table 2). Among the autotrophic ammonia oxidizers are the genera Nitrosomonas , Nitrosococcus , Nitrosospira , Nitrosolobus , and Nitrosovibrio while among the nitrite oxidizers are the genera Nitrobacter , Nitrococcus , Nitrospira , and Nitrospina . These groups of autotrophic nitrifiers consist solely of Gram negative bacteria (Prosser, 1989). In the first conceptual design of the MELiSSA loop a mixed culture was proposed with ammonia oxidizers belonging to the genera Nitrosomonas and nitrite oxidizers belonging to the genus Nitrobacter (Mergeay et al ., 1988). A mixture of nitrifiers would provide the system with stable reactor runs. For example, salt requirements, salt tolerance, and substrate affinity for ammonium differ significantly among ammonia oxidizers (Wagner et al. , 1995; Koops and Pommerenig-Röser, 2001). However, for simplifying the biological control of the reactor a co-culture of Nitrosomonas europaea ATCC19178 and Nitrobacter winogradskyi ATCC25391 was proposed for the third compartment. The latter microbiological concept of the nitrifying reactor using a synthetic medium for nitrification processes has already extensively been studied and oxygen supply was

Chapter 1 – Introduction 13 demonstrated to be critical (Gòdia et al ., 2002). Because human waste is fed to the MELiSSA loop, it is obvious that the composition of the feed in the third compartment will fluctuate in time. The controllable nature of the proposed co-culture can be questioned in light of discussions stating that more complex microbial communities show a higher functional stability (Fernández et al ., 1999). This is particularly true for nitrifyers in wastewater treatment systems (Sedlak, 1991). In addition, consortia involved in activated sludge processes may help to buffer inhibitory effects. It remains a challenge to obtain the ideal bacterial consortium providing stable treatment of effluents with high and fluctuating ammonium loads. Finally, the recycling of ammonium to nitrate, the main nitrogen precursor of Arthrospira sp. , allows for the closure of the MELiSSA loop (Figure 1.3) (Mergeay et al ., 1988).

1.3.3 STABILITY IN THE MEL ISSA LOOP Although the efficiency of recycling remains one of the most important aspects for the design of a functional life support system, it is important to note that several other factors will be important for the proper functional stability of the system. The general stability of the MELiSSA loop is dependent on the intrinsic character of the MELiSSA loop itself (design of the MELiSSA loop, reactor processing, efficiency of recycling, possibility for gene transfer, genetic evolution of the MELiSSA organisms, stress response, ... ) as well as on external influences like the physical conditions of the space environment and the composition of the consumer waste. The consumer waste containing fecal material may introduce micro-pollutants, such as heavy metals, hormones, or transformation products of pharmaceutical drugs. Insights into the behavior and effects of these compounds in natural ecosystems and closed loop systems are largely unknown. At present, such studies seem to be limited only to sewage treatment systems (Tabak et al ., 1970; Heberer et al ., 2002; Holbrook et al , 2002). The consumer waste is likewise a source of pathogenic organisms or their genes associated with virulence, and therefore a complete microbial separation between MELiSSA reactors and constant quality testing in all compartments of the MELiSSA loop is an absolute prerequisite to ensure the safety for consumption of the final food product. Especially the effects of microgravity, long time space radiation exposure, and changed electromagnetic fields are not fully understood.

Chapter 1 – Introduction 14

1.4 THE SECOND COMPARTMENT OF THE MEL ISSA LOOP

INHABITED BY RHODOSPIRILLUM RUBRUM S1H

1.4.1 STRAIN In light of designing a fully controllable and biosafe artificial bioregenerative ecosystem, ESA proposed as initial main constraint that the different bioreactors must be colonised in an axenic way only by known strains. Mergeay (1989) defined a list of potential microorganisms adapted for each compartment. The second compartment receives the low molecular carbon- compounds, primarily volatile fatty acids, as well as CO 2, H 2 and H 2S from the first (liquefying) compartment. The two main physiological features of the populating organisms must be: (1) Anaerobic metabolism that permits to limit the dependence to oxygen and; (2) Phototrophic metabolism that permits to use light as an energy source for the synthesis of complex nutritive molecules (proteins, fatty acids, carbohydrates,...). Moreover, in order to have a closed regenerative ecosystem, the biomass produced in that compartment must be edible by the crew members. It is thus desirable that the strain colonizing the bioreactor produces biomass with nutritional characteristic compatible with human diet. The microorganisms that possesses the above state requirements are the purple nonsulfur alpha- ( ), able to grow in the absence of oxygen and the presence of light (Mergeay et al ., 1988). More specifically, the strain Rhodospirillum rubrum was chosen in regard to its metabolic capabilities (diversity of assimilated substrates) (van Niel, 1944; Imhoff and Trüpfer, 1989; Madigan et al ., 2000). Assimilation of volatile fatty acids by that bacterium implies photoheterotrophic growth which is the most likely growth conditions in the second compartment of the MELiSSA loop and this particular metabolic strategy has been reviewed for Rhodospirillaceae by McEwan (1994). Furthermore, the strain is not toxic to humans and possesses highly nutritive characteristics (Kobayashi and Kurata, 1978; Vrati, 1984; Tranquille and Emeis, 1996a, b). Preliminary batch studies performed at the University of Gent (Belgium) by Christiaens (1990, 1991a,b) allowed to select for the specific strain R. rubrum ATCC25903 (designation S1H). The parent strain R. rubrum ATCC1170 (designation S1) is the type species for the Rhodospirillaceae and has been, and continues to be, the subject of a substantial amount of physiological and genetic analysis. Some of the areas of particular research interest are the photosystem (Vadeboncoeur et al ., 1979), (Lehman and Roberts, 1991), carbon monoxide

Chapter 1 – Introduction 15 oxidation (Kerby et al ., 1992), the ATP synthase (Andralojc and Harris, 1992) and more applied studies investigating bacterial production of biodegradable plastics (Brandt et al ., 1989) and hydrogen production (Najafpour G. et al., 2004). R. rubrum ATCC25903 differs from the parent strain ATCC1170 in that the former shows a greatly elevated threonine deaminase content, when grown on malate plus ammonium salt medium, as compared to the wild-type strain (approximatively 15 to 20-fold). As a consequence, growth of the strain ATCC25903 is much less susceptible to inhibition by the presence of exogenous L-threonine than the parent strain (Ning and Gest, 1966).

1.4.2 MORPHOLOGY Rhodospirillum spp . are gram-negative and motile. Cells are vibroid-shaped to spiral, 0.8-1.0 m wide; one complete turn of a spiral is 1.5-2.5 m wide and 7-10 m long (Imhoff and Trüpfer, 1989) (figure 1.4).

Figure 1.4 . Rhodospirillum sp. from DOE Joint Genome Institute.

1.4.3 PHOTOSYSTEM The purple phototrophic bacteria carry out anoxygenic photosynthesis; unlike cyanobacteria no O 2 is evolved. In fact, the bacteria are facultative aerobe and able to grow chemoheterotrophically under aerobic conditions and phototrophically under reduced-oxygen conditions in the presence of light. O 2 actually inhibits photosynthesis because it represses photopigment synthesis. During the transition from aerobic to reduced-oxygen conditions, photopigments and associated proteins are synthesized and inserted into a differentiated intracytoplasmic chromatophore membrane (ICM). Previous studies have demonstrated that the photosynthetic ICM is physically continuous with and formed by invagination of the cytoplasmic membrane (Crook et al ., 1986). Two protein pigment complexes, the P870 reaction center complex (RC) and the B880 light-harvesting I complex, are contained within the ICM of R. rubrum . The RC of R. rubrum consists of three polypeptide subunits, L, M, and H, which are present in a 1:1:1

Chapter 1 – Introduction 16 ratio. Bacteriochlorophyll (BCHL), bacteriopheophytin, the carotenoid spirilloxanthin (CRT), iron, and ubiquinone are associated with these polypeptides (Van Der Rest and Gingras, 1974; Vadeboncoeur et al ., 1979; Fotiadis et al ., 2004). Together, these pigments give their spectacular colors, usually purple, red or brown. R. rubrum has only one antenna complex, the B880 antenna complex, composed of α and β polypeptides, BCHL, CRT, diphosphatidylethanolamine, and diphosphatidylglycerol (Picorel et al ., 1983). These internal membranes allow purple bacteria to increase their specific pigment content and thus to better use the available light. Living cells of R. rubrum show absorption maxima at 375-377, 510-517, 546-550, 590-595, 807-808 and 881-885 nm (Imhoff and Trüpfer, 1989; Madigan et al ., 2000). The photosynthetic bacteria in general, and R. rubrum in particular, provide excellent model systems to study membrane assembly because the formation of the ICM can be induced by growth conditions and the RC is one of the best- understood bacterial membrane complexes (Hessner et al ., 1991).

1.4.4 METABOLISM These bacteria have been called "nonsulfur" because it was originally thought that they were unable to use sulfide as an electron donor for the reduction of CO 2 to cell material. However, sulfide can be used by most species, although the levels of sulfide well utilized by purple sulfur bacteria are toxic to most purple nonsulfur bacteria. A particular characteristic of this group is photoheterotrophy (where light is the energy source and an organic compound is the carbon source) that likely accounts for their competitive success in nature (Figure 1.5). They can be found in stagnant water bodies, lakes, waste-water ponds, sewage treatment plants, coastal lagoons, sediment, moist soil and paddy fields, growing best where there is a significant amount of soluble organic matter and therefore an obvious choice to treat the effluent coming from the first compartment. Purple nonsulfur bacteria are typically nutritionally diverse in this regard, using fatty, organic or amino acids; sugars; alcohol; and even aromatic compounds like benzoate as carbon source (van Niel, 1944; Imhoff and Trüpfer, 1989; Madigan et al ., 2000). Moreover, R. rubrum is one of the few bacteria able to metabolize carbon monoxide; it responds to CO by inducing at least two enzymes, CO dehydrogenase and a hydrogenase, that function to carry out the oxidation of CO with concomitant production of CO 2 and H 2 (Bonam et al ., 1989). Like all the Rhodospirillaceae , a specific feature of R. rubrum is the presence of the enzyme complex that allows fixation of atmospheric nitrogen (N 2) and growth in the

Chapter 1 – Introduction 17

+ - absence of ammonium (NH 4 ) and nitrate (NO 3 ), usually outcompeting other organisms (Madigan et al ., 2000).

Rhodospirillum rubrum

O2 O2

Chemotroph Phototroph (uses chemicals compounds (uses light as energy source) as energy source)

Chemolithotroph Chemoorganotroph (uses inorganic chemicals (Heterotroph) as electron donor) (uses organic chemicals as electron donor)

Chemolithoautotroph Photolithoautotroph Photoheterotroph Carbon source = CO 2 Carbon source = CO 2 Carbon source = organic Electron donor = H Electron donor = inorganic Electron donor = organic 2 Figure 15 . Overview of R. rubrum metabolism (after Madigan et al ., 2000).

Finally, R. rubrum grows at pH 6.0-8.5 (optimum 6.8-7.0) and at 30-35°C, requires biotin as growth factor (Imhoff and Trüpfer, 1989), and resists to the heavy metal oxide tellurite to a concentration of 20 µg/ml (0.1 mM) which constitutes up to 20-times the toxic concentration for many microorganisms (Moore and Kaplan, 1992).

1.4.5 GENETICS Since January 2005, the genome of the parent strain R. rubrum ATCC11170 (designation S1) is completely sequenced and comprises a chromosome (4.35 Mb; GenBank ID: CP000230) and a plasmid (53.7 kb; GenBank ID: CP000231). The GC content is quite high (65.4 %) which can limit efficiency of molecular methods to certain extend. Very little is known about the existence of restriction-modification systems in this bacterium. R. rubrum is very accessible to broad host range plasmids (Olsen and Shipley, 1973; Saegesser et al ., 1992) and to conjugative plasmids that were currently used to introduce transposons in this gender (Bao et al ., 1991; Jiang et al. , 1998). Within this context Hendrickx et al . (unpublished data) compared the frequency of transfer, mobilisation and retrotransfer in a mating involving R. rubrum with transfer, mobilisation and retrotransfer in a mating involving Cupriavidus (former Ralstonia ) metallidurans , an

Chapter 1 – Introduction 18 organism know to be very amenable for horizontal gene transfer by conjugation (Mergeay et al ., 1987; Springael et al ., 1994). In these experiment R. rubrum ATCC25903 was shown to have identical levels of conjugation frequencies and therefore this strain seems to be as amenable to gene transfer as most gram-negatives. As a result, among the MELiSSA strains, R. rubrum is certainly the strain that looks the most permeable to plasmid-mediated gene dissemination. This feature is also enhanced by the crucial position of R. rubrum in the second compartment, just downstream of the first compartment containing the yet unknown consortium. No phage was reported for R. rubrum , although a rhizobiophage may integrate his chromosome since it has been shown to do so in a tRNA gene of the closely-related 2.4.1. (Semsey et al ., 2002). No prophage was detected based on the ACLAME database ( http://aclame.ulb.ac.be/ ). Besides, the possibility to produce anti- viral or anti-microbial substances should be investigated which could be detrimental to optimal reactor conditions (Suwanto and Kaplan, 1991; Guest, 1974).

1.4.6 MUTAGENESIS Previous genetic analyses of bacterial photosynthesis have centered on relatively few species of closely related purple photosynthetic bacteria. Early genetic studies of photopigment biosynthetic pathways were undertaken primarily with Rhodobacter sphaeroides and Rhodospirillum rubrum (Sounder, 1978). The mutants provided a valuable source of intermediates in the tetrapyrrole and isoprenoid biosynthetic pathways, which has resulted in framing the basis of our current understanding of their biosynthetic schemes (Yildiz et al. , 1991). Since then, numerous mutants have been constructed providing a valuable basis for the study of photosystem structure and physiology (Hessner et al ., 1991; Ouchane et al ., 1997; Cheng et al ., 2000; Lupo and Ghosh, 2004, Fyfe et al. , 2004), nitrogen fixation (Lehman and Roberts, 1991; Zhang et al ., 1992; Zhang et al. , 2000) and carbon monoxide dehydrogenase characterization (Kerby et al ., 1997).

1.4.7 PREVIOUS EXPERIMENTS ON THE STRESS RESPONSE OF R. RUBRUM There are very few studies performed on the stress response of purple bacteria. Early reports of Clayton et al . (1958), Gunter and Kohn (1955) and Lambina (1961) studied the effects of ultraviolet and X-rays on the entire cell, and Bonam et al. (1989) exposed R. rubrum cells to CO. All studies are related to individual internal structures: differential radiation inactivation

Chapter 1 – Introduction 19 of the chromatophores to study the structure of the bacteriochlorophyll-protein complex (Picorel et al. , 1986; Gingras and Picorel, 1990; Wu et al. , 1991) and the structure of the ATPase (Chien et al ., 1993), and more recently L όpez-Marquès et al . (2004) studied the differential regulation of soluble and membrane-bound inorganic pyrophosphatases using pyrophosphate-based stress bioenergetics ( i.e. dark aerobic, light anaerobic or salt stress). The first proteomic study of R. rubrum was published this year by Selao et al . (2008). They used two-dimensional gel electrophoresis coupled to mass spectrometry identification and identified 44 proteins differentially expressed in R. rubrum grown anaerobic and photoheterotrophically, with different nitrogen sources.

1.5 THE STUDY OF BACTERIAL PHYSIOLOGY IN SPACE CONDITIONS Since the advent of space flight and the establishment of long-duration space stations in Earth's orbit, such as Skylab, Salyut, Mir and the ISS (see paragraph 1.1.), the upper boundary of our biosphere has extended into space. Such space missions expose humans and any other organism to living conditions not encountered on Earth. From the parameters related to the space environment described above (see paragraph 1.2.), ionizing radiation and microgravity are the main parameters that would affect living organisms.

1.5.1 BIOLOGICAL WEIGHTING OF RADIATION Ionizing radiation is radiation with sufficient energy to ionize molecules. There are two types of ionizing radiation, both produced by the decay of radioactive elements: electromagnetic (X- and γ-radiation that form part of the electromagnetic spectrum including visible light and radio waves) and particulate ( α- and β- and neutron-particles). Different types of ionizing radiation deposit energy in matter at different rates. α-, β-and neutron-particles produce ionization by collisions, depositing their energy within a short range after entering matter. γ- rays are photons that generate ions by several types of energy-absorption events (most commonly by the Compton effect, an increase in the wavelength of electromagnetic radiation when it collides with electrons in matter) and can penetrate deeply into a cell or tissue. Ion production is accompanied by the release of energetic electrons (Figure 1.6), and multiple ions and electrons can be generated in one event. The figure 1.6 shows the tracks of three different types of ionizing radiation. Small dots indicate energy deposition events. The inset depicts the ejection of an electron from an atom to generate an ion, mediated by an encounter with a γ-ray photon. The γ-ray transfers part of its energy to a valence electron, which is

Chapter 1 – Introduction 20 thereby ejected from the nucleus to create an ion. The scattered γ-ray can undergo additional Compton effects within the matter (Cox and Battista, 2005).

(>10 keV/µm)

Neutron,

Figure 1.6 . The figure shows the tracks of three different types of ionizing radiation. Small dots along the tracks indicate energy deposition events. The inset depicts the ejection of an electron from an atom to generate an ion, mediated by an encounter with a γ-ray photon. The γ-ray transfers part of its energy to a valance electron, which is thereby ejected from the nucleus to create an ion. (adapted from Cox and Battista, 2005).

Ionizing radiation is measured in the S.I. unit of absorbed dose per mass unit, the Gray (Gy), with 1 Gy equal to the net absorption of 1 Joule in 1 kilogram of material (water). However, radiation can have different degrees of effectiveness in producing effects in biological systems that largely depends on the local energy distribution, the Linear Energy Transfer (LET). Therefore, the same physical dose of radiation of different qualities can have different capacities to produce biological effects. This effect is described as the relative biological effectiveness (RBE) and is is the ratio of the physical dose of the test radiation and, for example gamma-rays, leading ot the same biological effect. The RBE value can be different for different biological systems, depending of their stage in the growth cycle and other environmental factors such as oxygen content. To assess the effectiveness posed by the radiation to humans and to the whole biosphere, estimates must be made of both the amount and type of radiation under consideration and the radiobiological effectiveness of the different components of the radiation. For this purpose, the quality factor Q has been introduced, which is the biological weighting function of ionizing radiation and has been obtained by averaging over a variety of RBE values for the same LET value. For gamma-rays, Q is equal to 1. For a given dose of high-LET radiation, the dose equivalent, H, is the product of Q and the absorbed dose:

H = QxD

Chapter 1 – Introduction 21 With H = the dose equivalent, Q = the quality factor and D = the absorbed dose. The S.I. unit for the dose equivalent is Sievert (Sv). For neutron-rays, Q can reach up to 20 depending on the neutron energy (ICRP, 1991). Q has originally been developed for radiation protection purposes. Therefore, it is mainly based on radiation risks for cancer induction in mammals.

1.5.2 BIOLOGICAL EFFECTS OF IONIZING RADIATION

1.5.2.1 THE IMPORTANCE OF THE REDOX BALANCE IN THE CELL The release of energy from growth substrates, and the capture of this energy for the biosynthesis of cell components and the processes that they support, involves redox reactions in which electrons and/or hydrogen atoms are transferred between donor and acceptor molecules. Bacteria lack the membrane-bound organelles of more complex organisms, so the opportunities to generate compartments with redox states that are optimized for different cellular processes are limited. However, sets of redox couples can be independent of each other if their activation energies are high and if there are no catalytic mechanisms to link them kinetically. Furthermore, interaction between different redox pathways might be limited by spatial constraints that are imposed by the cellular localization of the involved enzymes. For example, although the bacterial cytoplasm is generally a reducing environment, owing to the presence of high concentrations of reductants such as glutathione, during aerobiosis redox couples involved in respiration ( e.g . NADH/NAD+) persist in an oxidized steady state due to the rapid transfer of respiratory electrons to oxygen. Bacterial metabolism and cellular integrity are maintained by balancing the redox state of all the cellular components for optimal overall function. However, disturbance of this balance, notably an increase in oxidation reactions (oxidative stress), can damage essential cellular components (Green and Paget, 2004). Under normal growth conditions, there is a balance between pro-oxidants and antioxidants in the cell. Oxidative stress occurs as a consequence of an imbalance in favour of the pro- oxidants. It can be caused by exposure to increased levels of the reactive oxygen species – • (ROS), superoxide (O 2 ), hydrogen peroxide (H 2O2) and the hydroxyl radical (HO ), which are produced by the stepwise one-electron reduction of molecular oxygen (see equation 1):

•- • O2  O 2  H 2O2  HO  H 2O (1)

Superoxide and hydrogen peroxide are generated endogenously when bacteria grow under aerobic conditions by autooxidation of flavin cofactors of redox enzymes (Messler and Imlay,

Chapter 1 – Introduction 22 1999; Imlay, 2002). The highly reactive hydroxyl radical is generated when hydrogen peroxide reacts with Fe 2+ in the Fenton reaction (see equation 2), thereby linking cellular iron status to oxidative stress.

2+ + 3+ • Fe + H 2O2 + H  Fe + HO + H 2O (2)

So, exposure to oxidative stress is an unavoidable consequence of aerobic metabolism. However, ROS are also important elements of the host immune response, and many pathogens need to defend against oxidative stress in order to establish and maintain infections. ROS are useful components of host immune systems because they create lesions in DNA, damage the iron–sulphur clusters of key enzymes, oxidize protein thiols and cause peroxidation of lipids in the invading bacteria. The pervasive nature of the damage that is caused is illustrated by the reaction of superoxide with protein Fe–S clusters (see equations 3– 6). This reaction inactivates the protein and leads to the formation of hydrogen peroxide and Fe 2+ ,which in turn react to yield the hydroxyl radical. The hydroxyl radical is the strongest oxidant that can exist in aqueous environments. It reacts with bases and ribosyl moieties to generate DNA radicals that resolve into a range of sometimes lethal DNA lesions (Green and Paget, 2004).

2+ •– + 3+ [4Fe–4S] + O 2 + 2H  [4Fe–4S] + H 2O2 (3)

[4Fe–4S] 3+  [3Fe–4S] 1+ + Fe 2+ (4)

2+ + 3+ • Fe + H 2O2 + H  Fe + HO + H 2O (5)

• HO + DNA H 2O  DNA lesion (6)

1.5.2.2 ENVIRONMENTAL RADIATION AND LIFE

1.5.2.2.1 The dogma: DNA as the principal radiosensitive target Life on Earth, throughout its almost four billion years of history, has been shaped by interactions of organisms with their environment and by numerous adaptive responses to environmental stressors. Among these, radiation, both of terrestrial and of cosmic origin (see paragraph 1.2.1), is a persistent stress factor that life has to cope with. Radiation interacts with

Chapter 1 – Introduction 23 matter, primarily through the ionization and excitation of electrons in atoms and molecules. These matter-energy interactions have been decisively involved in the creation and maintenance of living systems on Earth. Because it is a strong mutagen, radiation is considered a powerful promoter of biological evolution on the one hand and an account of deleterious consequences to individual cells and organisms ( e.g. by causing inactivation or mutation induction) on the other. In response to harmful effects of environmental radiation, life has developed a variety of defense mechanisms, including the increase in the production of stress proteins, the activation of the immune defense system, and a variety of efficient repair systems for radiation-induced DNA injury. As a reactive chemical species, DNA is the target of numerous physical and chemical agents. These can induce a broad spectrum of DNA lesions, including damage to nucleotide bases, cross-linking, and DNA single- and double-strand breaks. Despite all of these lesions, the DNA is functionally more stable than the two other cellular macromolecules, RNA and proteins. This stability can be attributed to the following three factors: (i) the primary structure of DNA is all that is needed for transfer of information; (ii) because of the double- helical structure, DNA carries the information in duplicate; and (iii) there are molecular mechanisms of different complexity to undo the DNA damage, thus maintaining cellular survival and genetic integrity (Friedberg et al ., 1995; Vermeulen et al ., 1996). DNA repair encompasses the molecular reactions that eliminate damaged or mismatched nucleotide from DNA. A variety of repair mechanisms exist, each catalyzed by a different set of enzymes. Nearly all of these mechanisms depend on the existence of two copies of the genetic information, one in each strand of the DNA double helix: if the sequence in one strand is accidentally changed, information is not lost irretrievably because a complementary copy of the altered strand remains in the sequence of nucleotides in the other strand. Examples of different DNA repair mechanisms in microorganisms are: (i) direct repair including photoreactivation, akyltransfer and oxidative demethylation, (ii) excision repair including mismatch repair, base excision repair and nucleotide excision repair, (iii) recombinational repair, and other systems such as (vi) the SOS response.

1.5.2.2.2 Recent discoveries: extreme resistance in bacteria is dependent on protein protection Radiobiologists have long believed that ionizing radiation, like gamma rays, kills cells by shattering DNA. Recently, Daly et al. (2004), Daly et al . (2007) and Frederikson (2008)

Chapter 1 – Introduction 24 showed that proteins—not DNA—are the most sensitive targets, at least in some radiation- sensitive bacteria. The amount of DNA damage caused by a given dose of radiation is very similar for resistant and sensitive bacteria (Daly et al ., 2004; Gerard et al ., 2001). Yet, the range of resistance to ionizing radiation (IR) is large (Gerard et al ., 2001; Daly et al ., 2004; Cox and Battista, 2005), with a factor of 200 separating the most-resistant from the most-sensitive species (Daly et al ., 2004). For example, Deinococcus radiodurans can survive levels up to 10 kGy of IR that induced approximately 100 DNA double-strand breaks (DSBs) per genome, whereas Shewanella oneidensis is killed at 0.07 kGy, resulting in less than 1 DSB per genome (Daly et al ., 2004). Bioinformatic and experimental reports generally support that the genome configuration and copy number of D. radiodurans and its enzymatic protection and repair functions do not have unique properties that are essential or a prerequisite for this extreme- resistance phenotype (Cox and Battista, 2005; Qiu et al ., 2006). The observations of Daly et al . (2007) that IR-induced cellular protein damage, but not DNA damage (Gerard et al. 2001; Daly et al. 2004), is related to radioresistance and intracellular Mn/Fe concentration ratios, could help to explain why bacteria that encode a similar repertoire of DNA repair functions display such large differences in IR resistance (Daly et al . 2004; Cox and Battista, 2005; Qiu et al ., 2006). Specifically, Daly et al. (2007) proposed that in resistant bacteria proteins are protected from oxidation during irradiation by redox cycling of accumulated Mn(II), resulting in efficient survival and functioning of enzyme systems involved in recovery (Figure 1.7) (Daly et al ., 2004). This could explain why the polA gene of E. coli fully complements IR- and UV-sensitive D. radiodurans polA mutants (Gutman et al ., 2004). Similar, Daly et al . (2007) attributed the high level of radiation sensitivity of Fe-rich, Mn- poor bacteria to their susceptibility to global Fe-mediated oxidative protein damage during irradiation under aerobic or anaerobic conditions. Oxidative modification of proteins by IR could disrupt cellular functions involved in DNA repair either by loss of catalytic and structural integrity or by interruption of regulatory pathways, which in extremely radiation- sensitive cells might render protein damage lethal before significant DNA damage has accumulated (Qui et al ., 2006). In cells, oxidatively damaged DNA repair enzymes generated by sublethal IR doses would be expected to passively promote mutations by misrepair. Oxidized proteins, however, might also actively promote mutation by transmitting damage to other cellular constituents, including DNA (Du and Gebicki, 2004; Nauser et al ., 2005). Du and Gebicki (2004) even

Chapter 1 – Introduction 25 suggested that the well documented oxidation of lipids and DNA in cells by ROS appear to be secondary phenomena, subsequent to protein oxidation. In addition, because water molecules constitute the major component of an active organism, it will be the primary target of ionizing radiation leading to the production of ROS (figure 1.8). Concerning the bacterial membrane, studies with model systems indicate that the monounsaturated lipids found in most bacterial membrane are unreactive to oxidants (Bielski et al ., 1983). Therefore, only the polyunsaturated fatty acids within the membranes of photosynthetic bacteria, e.g . cyanobateria, are good candidates for peroxidation (Imlay, 2003).

Figure 1.7 . Model of ionizing radiation (IR) driven Mn and Fe redox cycling. Under IR, Fe(II,III) redox cycling ● ●- ●- is predicted to generate HO and O 2 , whereas Mn(II,III) redox cycling is predicted to favor O 2 scavenging without HO ● production (Daly et al ., 2007).

(a) indirect effect of IR IR H2O

(b) direct effect of IR

ROS

LIPIDS PROTEINS DNA

Lipid Pr ● Protein PrOOH Oxidized oxidation oxidation nucleic acids PrOO ● Membrane Protein Impaired damage dysfunction DNA repair

Figure 1.8 . Possible mechanisms of ionizing radiation ( IR ) induced oxidative stress and cellular damage. ROS : reactive oxygen species. Pr ●: carbon-centered radical. PrOO ●: peroxyl radical. PrOOH: protein hydroperoxide.

Chapter 1 – Introduction 26

1.5.3 BIOLOGICAL EFFECT OF CHANGE IN THE GRAVITY ENVIRONMENT

1.5.3.1 MECHANOSENSITIVE PROCESSES IN MICROORGANISMS Microbes have the ability to sense and respond to mechanical stimuli. The response of microbes to certain mechanical stimuli has profound effects on their physiology (Hamil and Martinac, 2001; Thomas et al ., 2002; Nickerson et al . 2003; Nickerson et al ., 2004; Vukanti et al ., 2008). The response of a cell to mechanical stimulation, such as stretch or shear force, is called mechanotransduction and is important for cell protection in both prokaryotes and eukaryotes (Ingber, 1997; Ingber, 1999; Hamil and Martinac, 2001). A great deal of progress has been made in understanding certain aspects of microbial mechanotransduction, for example, mechanisms used by bacteria to respond to changes in osmotic gradients (Blount and Moe, 1999; Hamil and Martinac, 2001; Pivetti et al ., 2003). Studies have also documented that microbes can sense and respond to changes in culture conditions when grown in the buoyant, low-fluid-shear environment of microgravity (Demain and Fang, 2001; Johanson et al . 2002; Klaus, 2002; Nickerson et al ., 2003; Nickerson et al., 2004; Vukanti et al ., 2008). It has been hypothesized that cells sense changes in mechanical forces, including shear and gravity, at their cell surface (Ingber, 1999; Thomas et al ., 2002).

1.5.3.2 MICROGRAVITY AND LOW FLUID SHEAR

Mechanical culture conditions in the quiescent microgravity environment of space flight are characterized by significant reductions in fluid shear (Hammond et al ., 2000; Klaus, 2002). This is because convection currents are essentially absent in microgravity (Klaus et al ., 1998; Klaus, 2002). Because actual space flight experiments are severely constrained by limited availability of electrical power, stowage space and crew time, and a general lack of sophisticated analytical equipment and expertise aboard the spacecraft, an effort has therefore gone into developing Earth-based systems that simulate microgravity (Table 3). Moreover, next to microgravity, other physical stresses exist in space (like change in ionizing radiations and magnetic field environment). Therefore, in most of the experiments conducted in space it is difficult to ascertain which factors accounted for the observed effects (Lynch and Matin, 2005). The most commonly used microgravity simulator is the Rotating Wall Vessel (RWV) culture apparatus (Synthecon, Texas, USA) developed by the NASA group at Johnson Space Centre in Texas.

Chapter 1 – Introduction 27

Table 3 . Synopsis of commonly used experimental approaches for achieving on-ground quasi weightlessness (Manti, 2006). Device Operating principle Major assets Major pitfalls Near quiescent fluid Limited gas exchange Gravity vector environment Bubble formation and Clinostat randomization by axial Reduced mass transport relatively high shear rotation conditions force

Gravity vector As above plus allowance Random positioning randomization coupled for medium exchange As above machine (RPM) to directional and larger specimens randomization

Very low shear and bubble formation Gravity vector Limited sustainable cell Rotating-wall vessel Constant free-fall randomization by solid- aggregate size; (RWV) condition body rotation No g-jitter Increased oxygen and nutrients availability

Short weightlessness Continuous free fall over Free fall machine Free fall conditions time limited time periods Mechanical stress

Partial gravity allowed Centrifuge free fall Free fall coupled with Short weightlessness (e.g . to simulated launch machine pulsed centrifugation time forces)

Small dependence on Bubble formation Dynamic neutralization Chemostat cell aggregate size Poor achievable flow of sedimentation velocity Good gas influx laminar conditions

Limited cell growth and Static neutralization of Neutral buoyancy As above metabolism sedimentation velocity No g-jitter

High magnetic field Magnetic levitation Weightless environment Use of strong magnetic balancing and opposing (Marco et al ., 2007) No bubble formation field (up to 17 Tesla) the force of gravity

This apparatus consists of a rotor, a cylindrical High Aspect to Ratio Vessel (HARV), and a platform on which the vessel is rotated (figure 1.9A). The HARV has separable front and back faces; the front face contains two sampling ports, and the back is provided with a semi- permeable membrane for aeration (figure 1.9B). The assembled vessel is filled to capacity (zero headspace) with medium and inoculum, and air bubbles are removed to eliminate turbulence and ensure a sustained low shear environment (< 0.1 N/m 2). It is then attached to the platform and oriented so that it is either rotated perpendicular to the gravitational vector, or parallel to this vector. Cells rotated in the former orientation experience normal gravitational forces and serve as a control (1 g) environment. In the vessel rotating around a horizontal axis, the liquid moves as a single body of fluid in which the gravitational vector is offset by hydrodynamic, centrifugal and Coriolis (circular

Chapter 1 – Introduction 28 movement) forces resulting in maintenance of cells in a continuous suspended orbit. In fact, this system "confuses" the biosystems ( e.g. cells growing culture) perception of gravity's direction. By placing cells along the axis of rotation and spinning them perpendicular to the gravity vector, they rotate through the vector. Because the cell spins at a constant rate and gravity remains constant, the gravity vector is nulled from the cell's perspective (Figure 1.9C) (Hammond and Hammond, 2001). Thus, the HARV bioreactor does not simulate microgravity as on the ISS, rather it randomizes gravity vectors and mimics the low turbulence of a space environment. Since the RWV apparatus provides a low-shear culture environment that simulates aspects of space (and therefore “models microgravity”), Nickerson et al. (2004) have adopted the terminology low-shear modeled microgravity (LSMMG) to refer to the RWV culture environment.

C

Figure 1.9 . Rotating Wall Vessel used to generate a simulated microgravity environment in ground-based investigations (in Lynch and Matin, 2005). HARV : High Aspect to Ratio Vessel.

Another microgravity simulator the Random Positioning Machine (RPM) or three- dimensional clinostat is a laboratory instrument to randomly change the position of an accommodated (biological) experiment in three-dimensional space, under the control of dedicated software running on a PC (Figure 1.10). The layout of the RPM consists of two frames and experiment platform. The frames are driven by means of belts and two

Chapter 1 – Introduction 29 electromotors. Both motors are controlled on the basis of feedback signals generated by encoders, mounted on the motor-axes, and by ‘‘null position’’ sensors on the frames. On the RPM, the samples are fixed as close as possible to the center of the inner rotating frame. This frame rotates within another rotating frame. While the RPM has been extensively use to study cytoskeleton structure and motility of human cells ( e.g . Meloni et al ., 2004) and plant gravitropism (Piconese et al ., 2003) only one work related to the study of bacterial behavior has been reported so far (de Vet and Rutgers, 2007).

1.5.3.3 MICROBIAL RESPONSES TO MICROGRAVITY AND OTHER LOW-SHEAR ENVIRONMENTS

The results from various studies regarding microbial responses to LSMMG are summarized in Table 1.4. Comparison of these results suggests that different organisms may dramatically differ in their responses to low-shear and space environments.

Figure 1.10 . Random Positioning Machine used to generate a simulated microgravity environment in ground- based investigations (Dutch space).

Albrecht-Buehler (1991) suggested that reduced gravity suppresses buoyancy-driven convection and thus limits the mechanism of mixing of fluid to diffusion. Along similar lines, McPherson (1993) suggested that the lack of convective mixing under reduced gravity conditions created a quiescent environment that resulted in a ‘depletion zone’ around a growing protein crystal, which favored the formation of a crystal with better quality. Based on these studies, Klaus et al . (1997) hypothesized that this same type of phenomenon might occur around a growing bacterial cell under reduced gravity conditions.

Chapter 1 – Introduction 30

Table 1.4 . Summary of LSMMG-induced effects on microbial physiology . Physiological effect induced by Bacterial species References LSMMG

Increased extracellular accumulation Micromicin B 17 Escherichia coli Fang et al . 1997 Rapamycin Streptomyces hygroscopicus Fang et al . 2000 Rhamnolipids Pseudomonas aeroginosa Crabbé et al ., 2008 (linked to cystic fibrosis lung pathology)

Increased virulence in murine model Salmonella enterica serovar Typhimurium Nickerson et al ., 2000 infection

Altered stress resistance Gao et al., 2001; Increased ethanol stress resistance Escherichia coli Lynch et al ., 2006 Increased acid, osmotic and Nickerson et al ., 2000; Salmonella enterica serovar Typhimurium thermal stress resistance Wilson et a l. 2006 Increased resistance to salt stress Escherichia coli Lynch et al ., 2006 and to penicillin and chloramphenicol Over-expression of genes associated with multiple stress (acidic Escherichia coli Vukanti et al ., 2008 stress, osmotic stress, oxidative stress, biofilm formation, lipid biosynthesis) Decreased oxidative stress (H O ) 2 2 Salmonella enterica serovar Typhimurium Wilson et al ., 2006 resistance

Increased survival within J774 Nickerson et al ., 2000; Salmonella enterica serovar Typhimurium macrophages Wilson et al ., 2006

Decreased generation time in M9 Salmonella enterica serovar Typhimurium Wilson et al ., 2006 minimal medium

Global alteration of gene expression Salmonella enterica serovar Typhimurium Wilson et al ., 2002 (LSMMG regulon)

Over-expression of starvation Escherichia coli Vukanti et al ., 2008 inducible genes

Similarly, a few studies have speculated that bacteria may indirectly respond to reduced gravity conditions because of changes in their immediate environment resulting either from changes in mass diffusion or other chemical alterations, such as accumulation of toxic by- products (Klaus et al . 1997), or limitations in the availability of nutrients (Baker et al. , 2004; Baker and Leff , 2004, 2006). Very recently, a gene expression study performed on Escherichia coli K12 support the hypothesis that bacteria under modeled reduced gravity conditions are actively responding to the changes in nutrient availability (Vukanti et al. , 2008). Creation of zones of nutrient depletion over-time in their immediate surroundings make these bacteria respond in a way that is similar to their entrance into stationary phase (starvation) which is generally characterized by expression of starvation inducible genes and genes associated with multiple stress responses.

Chapter 1 – Introduction 31

1.5.4 BIOLOGICAL EFFECT OF SPACE FLIGHT It is important to study the behavior of microbial populations in closed environments in order to estimate their real impact. In this respect, the spacecraft environment forms a special ecological niche where parameters such as acceleration, gravity, radiation, and electromagnetism are completely different than on Earth. As it is well known that environmental stresses induce and select for physiological, metabolic and/or genetic variations in microorganisms (Foster, 2007), it is envisioned that such adaptations will also likely occur under space flight conditions. The study of bacterial activity under space flight conditions is therefore highly important for the early detection of changes in bacterial communities and bacteria linked to medical, environmental, or life support issues (reviewed in Leys et al ., 2004). The results from various studies regarding microbial responses to space flight are summarized in Table 1.5.

Table 1.5 . Summary of space flight-induced effects on microbial physiology . Physiological effect induced by space flight Bacterial species References Effects on bacterial survival and growth characteristics No change in growth rate No induction of SOS response Escherichia coli Bouloc and D'Ari, 1991. Increased growth rate Escherichia coli Klaus et al ., 1997; Higher final cell concentration Bacillus subilis Kacena et al ., 1999. Survival rate promoted Rhodospeudomonas palustris Yang et al ., 1999. No increase in medicine resistance Rhodobacter sphaeroides Transcient increased in antibiotics resistance Escherichia coli Tixador et al ., 1985; Stapylococcus aureus Lapchine et al ., 1986. Normal DNA repair Escherichia coli Horneck et al ., 1996 No induction of SOS response Normal DNA repair Bacillus subtilis spore Horneck et al ., 1996 Kobayashi et al ., 1996; Enhanced recovery from radiation damage Deinococcus radiodurans Kobayashi et al ., 2000; Kobayashi et al ., 2004.

A signal transduction system involved in Escherichia coli Thévenet et al ., 1996. osmoregulation is functional

Enhanced plasmid exchange Bacillus thuringensis (various strains) De Boever et al ., 2007

Enhanced biofilm production Pseudomonas aeruginoasa in Pyle et al ., 2002

Enhanced virulence in a murine infection model Salmonella typhimurium Wilson et al ., 2007 Enhanced biofilm production

In agreement, with the effect of LSMMG, different organisms differ dramatically in their responses to space environments. However, one must be aware that the experiments presented here were subjected to the inconveniences of space biology. These include sample preparation long before the flight with prolonged storage, a strictly limited number of samples, strong acceleration during take-off, and a second storage period before recovery and analysis of the

Chapter 1 – Introduction 32 samples. These constraints impose a certain degree of caution in drawing conclusions (Thévenet et al ., 1996). In general one could conclude that microgravity has been shown not to hinder bacterial growth, on the contrary, it can enhance the growth of planktonic bacterial cultures, possibly through its influences on fluid dynamics (Klaus et al ., 1997). Biofilm formation can also occur in microgravity (Pyle et al ., 2002), an issue that for example must be addressed during the design of air and water recycling systems for long-term space flights. Various experiments suggested that antibiotics are less efficient in space flight conditions (Tixador et al ., 1985; Lapchine et al ., 1986). Bacteria exposed to space flight stresses may become more resistant to antibiotics over a short, introductory period, while losing most but not all of that resistance over the long term (Lapchine et al ., 1986). For the future, this possibly changing response of bacteria to antibiotics in space flight may imply that disinfection may be problematic. In addition, it may be difficult to treat an illness or injury with antibiotics on short term missions, due to the tendency of bacteria to resist them. On long-term missions, such as periods spent on space stations or trips to other planets, it may be difficult to predict the response of bacteria to a certain antibiotic. While most bacteria seem to become more susceptible to antibiotics after long-term space flight exposure (Lapchine et al ., 1986), a few may retain resistance, leading to potential hazard for the all crew. Recently, Wilson et al. (2007) provided the first direct evidence that growth during space flight can alter the virulence of a pathogen; in this study, Salmonella enterica serovar Typhimurium grown in space flight displayed increased virulence in a murine infection model compared with identical ground controls. Importantly, these results correlate with previous findings in which the same strain displayed increased virulence in the murine model after growth in the low-shear microgravity-like conditions of the RWV bioreactor (Nickerson et al ., 2000; Wilson et al ., 2002). In agreement with the increased virulence observed for the space flight samples, bacteria cultured in flight exhibited cellular aggregation and extracellular matrix formation consistent with biofilm production. Moreover, several Salmonella genes associated with biofilm formation changed expression in flight (Wilson et al ., 2007). Therefore, environmental conditions and crew member immune dysfunction associated with space flight may increase the risk of infectious disease during a long-duration mission (reviewed in Taylor et al ., 1997).

CHAPTER 2 - OBJECTIVES OF THE WORK

While foreseen to colonize the mineralizing compartment of the MELiSSA loop intended to insure human crew survival during long term space exploration missions, the susceptibility of Rhodospirillum rubrum S1H has not received much attention so far, as mentioned in chapter 1. However, the closing of the genome of its parent strain R. rubrum S1 in 2005 by the Joint Genome Institute opened the field of research by facilitating the study of that bacterium at the molecular level. After setting up the tools for the first large scale proteomic study, we will focus on the evaluation of R. rubrum S1H cultured on solid agar medium in space flight conditions. At this point, the proteomic approach will be completemented with a transcriptomic approach using for the first time a custom designed DNA microrray chip. On the other hand, in an effort to relate each single stress to its own bacterial response, ground simulation of ISS ionizing radiation and simulated microgravity experiments will be performed using each time the same culture setup as for the related space missions. To complete the present work, 2 microgravity simulators namely the Rotating Wall Vessel and the Random Positioning Machine will be compared to study the response of the bacterium to the change in gravity in liquid conditions, closer to the actual culture conditions in the MELiSSA loop. These data will give an overview of R. rubrum S1H response to a variety of environmental conditions based on the integration of transcriptomic and proteomic data and will permit to answer the following questions:

1. How far can transcriptomic and proteomic data be integrated? 2. How does R. rubrum S1H react to space flight related conditions? 3. What are the consequences of our results for future MELiSSA development?

CHAPTER 3 – WHOLE PROTEOME STUDY

3.1 INTRODUCTION Proteomics is the large-scale study of proteins, particularly their structures and functions. A cellular proteome is the collection of proteins found in a particular cell type under a particular set of environmental conditions. Proteomics provides a complementary understanding of an organism when coupled with a genomics approach. First, the level of gene transcription gives only a rough estimate of its level of expression into a protein. An mRNA produced in abundance may be degraded rapidly or translated inefficiently, resulting in a small amount of protein. Second, many proteins experience post- translational modifications that profoundly affect their activities. Finally, many proteins form complexes with other proteins or RNA molecules, and only function in the presence of these other molecules. The monitoring of gene expression using different technologies allows proteins of significance to be related to phenotypes associated with strain variability, environmental influences and the effects of genetic manipulation . As an example, gene- expression profiling and proteomic technologies have proven extremely useful to study the physiological response of bacterial cells to various environmental stress conditions ( e.g . Chhabra et al ., 2006; Wilson et al ., 2007; Pan et al ., 2008; Eraso et al ., 2008). Besides, annotation of protein-coding genes is a key goal of genome sequencing projects. In spite of tremendous recent advances in computational gene finding, comprehensive annotation remains a challenge. Proteomics based on mass spectrometry is a powerful tool for researching the dynamic proteome and suggests an attractive approach to discover and validate protein-coding genes namely the proteogenomic approach (Jaffe et al ., 2004; Fermin et al ., 2006; Ansong et al ., 2008). Other applications of bacterial proteomics were related to anti-bacterial drug discovery (Brötz-Oesterhelt et al ., 2005), vaccination research (Leroy et al ., 2007) and bacterial classification (Dworzanski et al., 2006). Today’s success of proteomics is based on the simultaneous advances in separation techniques, mainly bi-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and high- performance liquid chromatography (HPLC), and on mass spectrometry (MS) instrumentation

Chapter 3 – R. rubrum S1H – Whole proteome study 36 (Domon and Aebersold, 2006). During the last decade, mass spectrometry has become the core technology in proteomics. The application of mass spectrometry based techniques for the qualitative and quantitative analysis of global proteome samples derived from complex mixtures has had a big impact in the understanding of cellular function. First, strategies allowing high-throughput identification of proteins from highly complex mixtures include accurate mass measurement of peptides derived from total proteome digests and multidimensional peptide separations coupled with mass spectrometry . Second, recent developments in stable isotope labelling techniques and chemical tagging allow the mass spectrometry based differential display and quantitation of proteins. The third fundamental support for proteomics expansion is the huge increment in the entries in protein and nucleic acid databases, together with the development of bioinformatics tools (Chalkley et al ., 2005). Proteome analysis implies the ability to separate proteins with high resolution and reproducibility prior to characterization by mass spectrometry. Within that purpose, two approaches have been tested: (1) Sequential protein extraction analyzed by 2D-PAGE coupled to MALDI-ToF MS and (2) the shotgun proteomic approach also known as multidimensional protein identification technology (MudPIT) using electrospray ion-trap MS. For the latter technique, three prefraction methods were tested: on-line strong cation exchange, off-line isoelectric focusing and off-line SDS-PAGE (Figure 3.1). Except for the study of Selao et al . (2008) who used the 2D-gel approach and identified 44 differentially expressed proteins, no high-throughput proteomic studies have been carried out on R. rubrum S1H. Therefore, the use of shotgun proteomics, for the first time, will also help in the molecular characterization of this bacterium that occupies a key position (taking care of the carbon transformation) in the MELiSSA process.

3.2 MATERIAL AND METHODS Strain and media . The experiments reported here used R. rubrum strain S1H obtained from the American Type Culture Collection (ATCC25903). For the first generation proteomic experiments, R. rubrum S1H was cultivated in biological triplicates using 100 ml glass penicillin flasks filled with 40 ml of Sistrom-succinate medium (Sistrom, 1960) either wrapped with aluminum foil and vented through a 0.2 µm filter for dark aerobic growth or exposed to 20,000 lux and flushed with argon for light anaerobic growth in a growth cabinet CCL 300 BH (Angelantoni, Italy). Cells were incubated at 30°C on a rotary shaker (200 rpm) until end-exponential phase (OD 680nm ≈ 0.7). Thereafter, the 3 independent cultures were pooled, centrifuged at 7,000 rpm for 10 minutes at 4 °C and the resulting pellet was kept at -

Chapter 3 – R. rubrum S1H – Whole proteome study 37 80 °C until use. For the shotgun proteomic approach, R. rubrum S1H was also cultivated in biological triplicates but using 75 cm 2 vented cell culture flask (BD Falcon, USA) filled with 50 ml of Sistrom-succinate medium. Flasks were packed with aluminum foil and cells were kept rotating at 40 rpm and 30°C for dark aerobic growth until mid-exponential phase

(OD 680nm ≈ 0.3) was reached.

Extraction Digestion and separation Analysis

First generation proteomics 2D-PAGE Protein sequential extraction (Bio-Rad IEF-strip peptides tryspinolysis (a) reagents) from DAE MS (MALDI-ToF) or LAN grown 4-7 NL cultures

Second generation proteomics

Protein extraction peptides using guanidine (b) tryspinolysis SCX column RP column MS/MS chloride from DAE (Ion trap) grown cultures IEF-strip

Protein extraction peptides MS/MS using urea/thiourea tryspinolysis RP column (c) from DAE grown (Ion trap) cultures 3-10 L

Protein extraction peptides (d) using N-octyl- tryspinolysis RP column MS/MS glycoside (Ion trap)

3-10 L peptides Membrane protein extraction using (e) Laemmli buffer trypsinolysis RP column MS/MS (SDS, (Ion trap) β-mercaptoethanol) SDS-PAGE

Figure 3.1 . Schematic representation of the five sample workflows compared in this study. Details of each workflow are provided in the text. DAE : dark aerobic. LAN : light anaerobic. IEF : isoelectric focusing. NL : non linear. 2D-PAGE : bidimensional polyacrylamide gel electrophoresis. MS : mass spectrometry. MS/MS : tandem mass spectrometry. SCX : strong cation exchange. RP : reverse phase. SDS : sodium dodecyl sulfate.

3.2.1 FIRST GENERATION PROTEOMICS : 2D-PAGE APPROACH

3.2.1.1 PROTEIN EXTRACTION AND QUANTIFICATION

The ReadyPrep sequential extraction kit (Bio-Rad, USA) provides the reagents necessary to extract proteins of different solubility from cell lysates in a form suitable for 2D-PAGE. The extraction kit is based on the work of Molloy et al. (1998) and Herbert et al . (1998). This kit allows to solubilize different sets of proteins by using 3 solutions differing in their detergent

Chapter 3 – R. rubrum S1H – Whole proteome study 38 and chaotrope concentrations: Reagent 1 (40 mM Tris, 200 mM Tributylphospine (TBP), protease inhibitor cocktail (Roche, USA) extracted only the most soluble proteins, such as cytosolic proteins. Reagent 2 (8 M Urea, 4 % (w/v) CHAPS, 40 mM Tris, 0,2 % (w/v) Bio- Lyte 3/10, 200 mM TBP and protease inhibitor cocktail (Roche, USA) was used to extract proteins of intermediate solubility, while Reagent 3 (5 M Urea, 2 M Thiourea, 2 % (w/v) CHAPS, 2 % (w/v) SB3-10, 40 mM Tris, 0.2 % (w/v) Bio-Lyte 3/10, 200 mM TBP and protease inhibitor cocktail (Roche, USA) extracted proteins otherwise insoluble in Reagent 1 and Reagent 2, essentially membrane proteins. The following extraction protocol has been set up according to previous work done in the laboratory at UMH (Figure 3.2). Three extractions were performed for each reagent in order to optimize the extraction. The protein concentration of the supernatants was measured by the Bradford method (Bradford, 1976), according to the Bio-Rad Protein Assay kit, with bovine gamma-globulin as a protein standard. The Bio-Rad Protein Assay is based on the observation that the maximum absorbance for an acidic solution of Coomassie Brilliant Blue G-250 shifts from 465 nm to 595 nm upon binding to proteins.

Buffer 1 Buffer 1 Buffer 1

Bacterial pellet Pellet 1 Pellet 2

Treatment Treatment Treatment for cell lysis for cell lysis for cell lysis

Centrifugation Centrifugation Centrifugation

Pelle t 1 Supernatant Pellet 2 Supernatant Pellet 3 Supernatant 1

Buffer 2

Treatment for cell lysis

Centrifugation

Pellet 4 Supernatant

etc. Figure 3.2 . R. rubrum protein extraction protocol using the ReadyPrep sequential extraction kit (Bio-Rad). Treatment for cell lysis included (1) Sonication for 5 min and (2) Freezing at -80°C for 15 min. Centrifugation at 13,000 rpm and 4 °C for 15 min.

Chapter 3 – R. rubrum S1H – Whole proteome study 39

3.2.1.2 PROTEIN VISUALIZATION

(i) SDS-PAGE . This method is used to check the relevance of the differential extraction and the protein concentration estimation. Sodium Dodecyl Sulfate (SDS) polyacrylamide gels are the result of copolymerization of acrylamide monomers with a linker agent (N,N' –methylene bisacrylamide). The migration occurs first at 200 V and 15 mA/gel and then at 20 mA/gel as soon as the migration front reaches the resolving gel (see Table 3.1 for the composition of the gel).

Table 3.1 . Recipe for SDS-PAGE using the Mini-Protean II device (all components purchased from Bio- Rad except MilliQ 50 water from Millipore). Stacking gel Resolving gel Resolving gel Component (4 % acrylamide) (12 % acrylamide) (11 % acrylamide) Bis-Acrylamide 30 % 13.3 % (v/v) 40 % (v/v) 36.5 % (v/v) 1.5 M Tris-HCl pH 8.8 - 25 % (v/v) 25 % (v/v) 0.5 M Tris-HCl pH 6.8 25 % (v/v) - - MilliQ 50 water 60 + 1 % a (v/v) 30 + 1 a % (v/v) 37 + 1 a % (v/v) 10 % APS 0.5 % (v/v) 0.5 % (v/v) 0.5 % (v/v) TEMED 0.05 % (v/v) 0.05 % (v/v) 0.05 % (v/v) a: water replaces the SDS which is not added here because it is already present in the SDS-PAGE running buffer Tris-Glycine-SDS (TGS).

(ii) 2D-PAGE . Using this method, proteins are separated in a first dimension based on their isoelectric point (isoelectric focusing) and in a second dimension according to their molecular weight (SDS-PAGE). For isoelectric focusing, we used Immobiline Dry Strips non-linear 18 cm pH 4–7 (Amersham, USA). The strip was rehydrated overnight in rehydration solution (2 % (w/v) CHAPS, 8 M urea, 0.5 % (v/v) Pharmalyte 3–10, 13 mM dithioerythritol (DTE). After loading of 200 µg of protein, isoelectric focusing was performed on a Multiphor II system (Amersham, USA) equipped with a Pharmacia Biotech EPS3500 XL power supply using a 3-phase program (Table 3.2).

Table 3.2 . Step settings for the first dimension analysis of R. rubrum protein extract samples. Step number Time Power (W) Current (mA) Voltage (V) 1 1 min 5 2 500 2 1 h 30 5 2 3,500 3 14 h 20 5 2 3,500

After isoelectric focusing, the strips were equilibrated two times, 20 min each, first in equilibration solution (6 M urea, 30 % (v/v) glycerol, 2 % (w/v) SDS, 50 mM Tris- HCl pH 6.8) containing 65 mM DTE, and second, in equilibration solution containing 135 mM iodoacetamide. The purpose of this step is threefold: (1) to negatively charge the proteins using SDS, (2) to reduce the disulfide bridges in the proteins (with DTE) and (3) to covalently

Chapter 3 – R. rubrum S1H – Whole proteome study 40 block the free thiol groups with iodoacetamide. For the second dimension, the strips were placed on top of a 20 x 20.5 cm 11 % PAGE gel (see Table 2) in a 0.4 % (w/v) agarose gel solution made with SDS-PAGE running buffer containing 192 mM Glycine, 0.1 % (w/v) SDS and 25 mM Tris-HCl (pH 8.3). The Protean Plus electrophoresis 2D device (Bio-Rad, USA), which allowed high-throughput analysis since it can run up to 12 gels simultaneously, was used and gels were run for approximately 8 h at 200 V, 30 mA/gel at first and then 40 mA/gel as soon as the migration front reached the resolving gel. (iii) Gel staining. Coomassie Brilliant Blue G-250 (Bio-Rad, USA) allows detection of protein concentration up to 100 ng and is compatible with mass spectrometry analysis. The following protocol was used (all products are purchased from Merck, USA, except Coomassie Brilliant Blue G-250 from Bio-Rad, USA): 1. Fixation : 2 h (or overnight) - methanol 50 % (v/v) - phosphoric acid 3 % (v/v)

2. Wash : 3 x 30 min - MilliQ 50 water 1000 ml

3. Pre -incubation : 1 h - ammonium sulfate 17 % (v/v) - methanol 34 % (v/v) - phosphoric acid 3 % (v/v)

4. Incubation 1: 3 to 4 days. - ammonium sulfate 17 % (w/v) - methanol 34 % (v/v) - phosphoric acid 3 % (v/v) - Coomassie G-250 0.1 % (w/v)

5. Wash : 15 min - MilliQ 50 water 1000 ml

1 Incubation solution protocol: (1) Add methanol to G-250 and mix for 10 min.; (2) Add water and ammonium sulfate; (3) Add phosphoric acid; (4) Mix and then filter on 320 mm Ø filter (Shchleicher & Schuell).

Chapter 3 – R. rubrum S1H – Whole proteome study 41

3.2.1.3 PROTEIN IDENTIFICATION

(i) Sample preparation. Coomassie Brilliant Blue-stained proteins were excised from gel using a 1 mm sample corer (Fine Science Tools, USA). Excised gel pieces were placed in 1.5 ml polypropylene Eppendorf tubes and washed twice in 50 µl of 50 mM NH 4HCO 3. Next, the gel pieces were:

1. Incubated twice for 15 min in 40 µl NH 4HCO 3 (Thermomixer compact, Eppendorf);

2. Washed twice for 15 min in 40 µl 25 mM NH 4HCO 3 and 50 % (v/v) acetonitrile (BDH); 3. And finally the supernatant was decanted and gel pieces were dried in a centrifugal evaporator (Heto, Drywinner, Denmark) for 15 min; (ii) Trypsinolysis . Enzymatic digestion by trypsin resulted in a set of protein-specific peptides by cutting the protein at the carboxy-terminal domain of lysine and arginine. However, digestion is not achieved if one of these amino-acids is followed by a proline. Enzymatic digestion was performed by the addition of 10 µl of modified sequencing grade trypsin (0.02 mg/ml) (Promega, USA) in 25 mM NH 4HCO 3 to each gel piece. The samples were placed for 15 min at 4 °C and incubated overnight at 37 °C. The reaction was stopped with 1 µl 5 % (v/v) formic acid. At this point, the gel pieces were either stored at -20°C or directly used for downstream analysis. (iii) Matrix Assisted Laser Desorption Ionisation (MALDI) – Time of Flight (ToF) mass spectrometry (MS) analysis . For MALDI-MS, a 1.5 µl aliquot of the trypsin digested supernatant was spotted onto a sample plate with 1.5 µl of matrix and 1 µl of different internal standards. The matrix was prepared by mixing two different matrix solutions (500 µl of solution 1 + 500 µl of solution 2) as follows: - Solution 1 : α-cyano-4-hydroxycinnamic acid 1x10 -5 mM (w/v) acetonitrile 70 % (v/v) 5 % formic acid 30 % (v/v) - Solution 2 : 2.5-di-hydroxybenzoic acid 1.5x10 -5 mM (w/v) acetonitrile 70 % (v/v) 0.1 % tri-fluoro-acetic acid 30 % (v/v)

Internal standards were prepared as follows (all products purchased from Sigma, Germany): - 5 µl of bradykinin;

Chapter 3 – R. rubrum S1H – Whole proteome study 42 - 5 µl of adenocorticotropic hormone fragment 18-; - 5 µl angiotensin I; - 10 µl renin substrate tetradecapeptide; - 10 µl [GLu 1] – fibrinopeptide B.  + 35 µl of matrix.

MALDI-MS was performed using a Micromass M@ldi_spectrometer (Manchester, United Kingdom) equipped with a 337 nm nitrogen laser. The instrument was operated in the positive reflectron mode at 20 kV accelerating voltage with time-lag focusing. The resulting peptide masses were automatically searched against a local copy of the SWISS-PROT (Boeckmann et al ., 2003) R. rubrum ATCC11170 database using the ProteinLynx global server and the Protein Probe search engine (Micromass, USA). An initial mass tolerance of 50 ppm was used in all searches. Peptide modifications allowed during the search were carbamidomethylation of cysteins and oxidation of methionines. The maximum number of missed cleavages was set to 1.

3.2.2 SECOND GENERATION PROTEOMICS : MUD PIT APPROACH One alternative to 2D gel electrophoresis has been the development of whole proteome digestion and multidimensional separations (chromatographic and electrophoretic) coupled with tandem mass spectrometry (MS/MS) (Link et al ., 1999; Washburn et al. , 2001). This technique uses as first separating power (dimension) cation exchange (Peng et al ., 2003), capillary electrophoresis (Figeys et al., 1996; Tong et al ., 1999), capillary isoelectric focusing (Chen et al ., 2003), slab-gel isoelectric focusing (Cargile et al ., 2004ab) or immobilized pH gradient (IPG) (Gargile et al ., 2005), followed by reversed-phase liquid chromatography (RP- LC) as the second dimension, and finally MS/MS for subsequent peptide identification.

3.2.2.1 2D-LC MS/MS APPROACH

Our first MudPIT experimental setup included on-line strong cation exchange followed by reversed-phase chromatography and ion-trap MS/MS analysis (Figure 3.1b). (i) Protein extraction and quantification. Before protein extraction, the bacterial pellet was washed 3 times with phosphate buffered saline (Buffered saline pack, Pierce, USA). Samples were cleared by centrifugation at 13,500 rpm for 15 min at 4 °C. Protein samples were

Chapter 3 – R. rubrum S1H – Whole proteome study 43 obtained by high power sonication (U50 control, IKA labortechnik, Germany) of the washed bacterial pellet suspended in one pellet volume of 6 M guanidine chloride solution. Sonication was performed through 3 cycles of 10 seconds (40 % amplitude, cycle 1) followed by 1 min on ice. Samples were cleared by centrifugation at 13,500 rpm for 15 min at 4°C. Proteins in supernatant were reduced with tris(2-carboxyethyl)phosphine at 60 °C for 30 min and alkylated with iodoacetamide at 25 °C for 30 min. Proteins were recovered by acetone precipitation (1h) with an acetone/protein ratio of 4/1. After 15 min centrifugation at 13,500 rpm and acetone evaporation, the resulting pellet was dissolved in 25 mM Tris-HCl (pH 8), 2 M urea. Protein concentration was measured using the Non-Interfering Protein Assay TM Kit (Calbiochem, Germany) with bovine serum albumin as a protein standard. Overnight enzymatic digestion was carried out with modified sequencing grade trypsin (Promega, USA) at an enzyme/substrate ratio of 1/50 at 37 °C. (ii) Chromatographic separation. Eighteen micrograms of tryptic peptides obtained by enzymatic digestion were separated using an Ultimate 3000 chromatographic system (Dionex, USA). The chromatographic system was composed of two columns connected in series, a strong cation exchange (SCX) column (POROS 10S, 10 cm) and a reverse phase column (C18, 15 cm, ID 75 m) both from Dionex (USA). Peptides were injected in the SCX column and the flow through is directly loaded on a precolumn (C18 Trap, Dionex, USA). The peptides were then washed on the precolumn with the loading solvent (5 % (v/v) acetonitrile, 0.025 % (v/v) TFA) during 15 min. After washing, the acetonitrile (ACN) gradient was started and peptides were separated on the reverse column based on their hydrophobicity. The ACN gradient was 4 to 37 % of solvent B (solvent B: 80 % ACN, 0.08 % formic acid) in 100 min, 37 to 57 % of solvent B in 10 min and 57 to 90 % of solvent B in 10 min , 90 % of solvent B was maintained for 10 min and then reset to 4 %. Column equilibration at 4 % of solvent B was allowed for 10 min. After this first cycle, a first batch of peptides waq eluted from the SCX column by injection of 20 l loading solvent containing 1 mM NaCl. The eluted peptides were loaded on the precolumn and separated as described for the flow through. This sequence of elution from the SCX column with injection of a salt plug followed by separation on the reverse phase column was repeated 9 times with salt plug concentrations of 2, 5, 10, 25, 50, 100, 200, 500 and 1,000 mM NaCl. (iii) Mass spectrometry analysis . The separated peptides were analyzed on-line using an ion- trap mass spectrometer (HCT ultra PTM discovery, Bruker Daltonics, Germany). This mass spectrometer was used in the positive mode and spectra were acquired in a data dependent manner: MS scan range = 300-1,500 m/z, maximum accumulation time = 200 ms, Ion Charge

Chapter 3 – R. rubrum S1H – Whole proteome study 44 Current (ICC) target = 200,000. The top 4 most intense ions in the MS scan were selected for MS/MS in dynamic exclusion mode: mode = ultrascan, absolute threshold = 75,000, relative threshold = 1 %, excluded after spectrum count = 1, exclusion duration = 0.15 min, averaged spectra = 5, ICC target = 200,000. (iv) Data extraction and database search. Peptide peaks were detected and deconvoluted automatically using the Data Analysis 3.4 software (Bruker). The mass list was generated automatically in Mascot Generic Files format and searched against a local copy of the NCBInr 20080608 database using an in-house Mascot 2.2 server (Matrix Science) for protein identification (p < 0.05). The default search parameters used were: = Rhodospirillum rubrum ATCC11170, enzyme = trypsin, maximum missed cleavages = 1, fixed modification = Carbamidomethyl (C), variable modification = Oxidation (M), peptide mass tolerance ± 1.5 Da, fragment mass tolerance ± 0.5 Da, peptide charge =1+, 2+ and 3+; instrument = ESI-TRAP. Only sequences identified with a Mascot score 2 of at least 50 were considered. The false discovery rate was estimated using the Mascot 'decoy' option. If TP was true positive matches and FP was false positive matches, the number of matches in the target database was TP + FP and the number of matches in the decoy database was FP. The quantity that was reported was the False Discovery Rate (FDR) = FP / (FP + TP) (Elias et al ., 2005).

3.2.2.2 LC MS/MS APPROACH

Gan et al . (2005) showed that greater proteome coverage was achieved with protein fractionation prior to shotgun proteomics analysis. Indeed, the introduction of fractionation greatly reduced the complexity of the protein/peptides sample. Because the spectra were acquired in a data dependent manner, this improved the ability of the MS to detect low- abundance protein/peptides, since ion suppression effects caused by overlapping signals from high- to low-abundance ions were reduced (Gan et al ., 2005). Recently, an alternative to using strong cation exchange to separate peptides prior to LC- MS/MS was described where trypsin digested proteins were separated over an immobilized pH gradient (IPG) using isoelectric focusing (IEF), and this technique was termed peptide IPG-IEF (Cargile et al ., 2004ab). Analysis of tryptic peptides from proteins from Escherichia coli whole cell lysate using peptide IPG-IEF on pH 3-10 linear IPG strip allowed for the identification of 1,223 proteins

2 The Mascot score is given as S = -10*log 10 (P), where P is the probability that the observed match is a random event.

Chapter 3 – R. rubrum S1H – Whole proteome study 45 from LC-MS/MS analysis using an LCQ ion-trap MS (Cargile et al ., 2004b). Separating directly peptides on the IPG strip is more effective than separating proteins on gel and then performing an in-gel digestion. Indeed, using the latter method, some peptide bonds are not accessible to the enzyme due to the trapping of protein substrate in the gel and not all the peptides produced during the digestion can diffuse freely out of the gel. On the other hand, due to the high hydrophobicity of the membrane fraction, protein separation using mono- dimensional SDS-PAGE followed by in-gel digestion has been recommended (Ca ňas et al . 2007). Therefore, our experimental setup included off-line peptide IPG-IEF followed by reversed- phase (RP) chromatography and ion-trap MS/MS analysis (Figure 3.1c). In addition, in an effort to increase the proteome coverage, a second extraction was performed starting from the insoluble material from the first extraction (Figure 3.1d). Moreover, R. rubrum membrane samples were also characterized using protein SDS-PAGE prior to RP separation and MS/MS analysis (Figure 3.1e).

3.2.2.2.1 Peptide IPG-IEF: First extraction (i) Sample preparation. R. rubrum S1H was cultivated in dark aerobic culture conditions in biological triplicates as described above. (ii) Protein extraction and quantification. Protein samples were obtained by high power sonication (U50 control, IKA labortechnik, Germany) of the bacterial pellets suspended in one pellet volume 6 M urea/2 M thiourea solution in 10 mM HEPES (pH 8.0). Sonication and sample clearing by centrifugation was performed as mentioned above. Pellets were kept for the second extraction while the supernatants were first reduced with 10 mM DL-Dithiothreitol and then alkylated with 55 mM iodoacetamide respectively. Protein concentration of the supernatants was measured by the Bradford method (Bradford, 1976), according to the Bio- Rad Protein Assay kit, with bovine gamma-globulin as a protein standard. One hundred µg from each biological replicate were pooled. Prior to overnight digestion by trypsin (Promega, USA) at an enzyme/substrate ratio of 1/50 at 37 °C, the urea/thiourea concentration was reduced to 2 M by dilution with 10 mM NH 4HCO 3. Three hundred µg of digested proteins were then desalted using HyperSep TM SpinTip C18 (Thermo-Fischer electron, USA) following the manufacturer's instructions. The sample was evaporated to dryness in a vacuum centrifuge (Heto, Drywinner, Denmark) and resuspended in 350 µl of 8 M urea supplemented with a trace of bromophenol blue. A second protein sample was prepared following the same procedure to increase the peptide coverage.

Chapter 3 – R. rubrum S1H – Whole proteome study 46 (iii) IEF separation. Digested proteins (300 µg) were used to passively rehydrate linear pH 3-10 18 cm IPG strips (GE Healthcare, USA) overnight. Isoelectric focusing was conducted on a Protean IEF cell (Bio-Rad, USA) with a current limit of 50 µA per strip at 20 °C with the following focusing program: 300 Volt for 1 h, a gradient to 1,000 V for 1 h, a gradient to 4,000 V for 3 h, a gradient to 8,000 V for 3 h and 8,000 V until 100,000 V/h was reached. This last step lasted approximately 9 h. The strips were then cut (with plastic backing still in place) with a scalpel blade in 30 equal length pieces ( i.e . 5 mm). Peptides were extracted from each fraction by incubation in 100 µl of 0.1 % (v/v) formic acid for 1 h at room temperature. The extraction was repeated twice and subsequently combined with the initial fraction. Combined peptide extracts from each fraction were concentrated in a vacuum centrifuge (Heto, Drywinner, Denmark) to approximately 25 µl. Each fraction was desalted using Hypersep TM ziptip C18 tip (Thermo-Fischer electron, USA) following the manufacturer's instructions. (iv) Liquid chromatography-mass spectrometry. The peptides were analyzed by nanoLC- MS/MS using an LC Ultimate 3000 system (Dionex, USA) coupled to an HCTultra plus mass spectrometer (Bruker Daltonics, Germany). Six l of each fraction were loaded onto a precolumn (C18 Trap, 300 m ID x 5 mm, Dionex, USA) using the Ultimate 3000 system, delivering a flow rate of 20 l/min of loading solvent. After a 10 min desalting step, the precolumn was switched online with the analytical column (75 m ID x 15 cm PepMap C18, Dionex, USA) equilibrated in 96 % solvent A (0.1 % (v/v) formic acid in HPLC-grade water) and 4 % solvent B (80 % acetonitrile, 0.1 % formic acid in HPLC-grade water). Peptides were eluted from the precolumn to the analytical column and then to the mass spectrometer with a gradient from 4 to 57 % solvent B for 50 min and 57 to 90 % solvent B for 10 min, at a flow rate of 0.3 µl/min delivered by the Ultimate pump. The separated peptides were then analyzed on-line using an ion-trap mass spectrometer. This mass spectrometer was used in the positive mode and spectra were acquired in a data- dependent manner: MS scan range = 300-1,500 m/z, maximum accumulation time = 200 ms, ICC target = 200,000. The top 5 most intense ions in the MS scan were selected for MS/MS in dynamic exclusion mode: mode = ultrascan, absolute threshold = 75,000, relative threshold = 1%, excluded after spectrum count = 1, exclusion duration = 0.3 min, averaged spectra = 5, ICC target = 200,000. Data extraction and database search were performed as described above.

Chapter 3 – R. rubrum S1H – Whole proteome study 47 3.2.2.2.2 Peptide IEF-IPG: Second extraction Protein extraction and quantification. Protein samples, obtained by high power sonication of the bacterial pellets from the first extraction, were resuspended in one pellet volume of 1 % N-octyl-glycoside/600 mM NaCl/125 U benzonase (Novagen, Germany) solution in 25 mM

KH 2PO 4/150 mM NaCl (pH 7.5). Samples were cleared by centrifugation as mentioned above. Proteins in the supernatant were methanol/chloroform precipitated (Wessel and Flugge, 1984) and resuspended in one pellet volume 6 M urea/2 M thiourea solution in 10 mM HEPES (pH 8.0) (v/v). Proteins were then reduced, alkylated and quantified as mentioned above. The 3 biological replicates (500 µg) were pooled and trypsin digestion was conducted in 10 mM NH 4HCO 3 (pH 7.8) containing 60 % (v/v) methanol. Trypsin digestion was carried out in two stages (Chick et al ., 2008): 5 µg of trypsin was added to the protein sample by a 20 min water bath sonication at 4 °C and then incubation at 37 °C for 1.5 h followed by a extra addition of 5 µg trypsin and subsequent incubation at 37 °C for a 1.5 h. Methanol was evaporated using vacuum centrifuge and the peptide sample was split in two parts and desalted using spin tip as mentioned earlier. Peptide IPG-IEF and liquid chromatography- mass spectrometry analysis was performed as mentioned above.

3.2.2.2.3 SDS-PAGE prior to LC-MS/MS (i) Sample preparation . Six independent cultures of R. rubrum S1H were grown in Sistrom- succinate dark aerobic conditions and then pooled as described above. (ii) Protein extraction and quantification. The bacterial pellet was resuspended in extraction buffer (150 mM Tris-HCl, pH 7.2/300 mM NaCl with a trace of Mini EDTA-free protease inhibitor cocktail) (Roche, Belgium). Cells were ruptured by three passages through a French Press (Thermo-Fischer electron, USA) at 500 psi. After centrifugation at 13,200 rpm at 4 °C for 15 min, the supernatant was removed and the pellet was washed three times with the extraction buffer. After centrifugation, the pellet was resuspended in Laemmli buffer (2 % SDS, 10 % glycerol, 5 % β-mercaptoethanol, 0.002 % bromophenol blue and 0.125 M Tris-HCl) and sonicated in water bath six times 1 min at room temperature. After 1 min incubation at 90 °C, the lysate was centrifuged at 14,000 rpm at room temperature for 15 min to pellet cell debris. The membrane containing supernatant was loaded on a Precast polyacrylamide mini-gel 4 – 20 % (Pierce, USA) and was run at 150 V and 75 mA/gel. Staining of the gel was achieved using Imperial Protein Stain (Thermo- Fischer electron, USA) according to the manufacturer's instructions. The gel lane containing

Chapter 3 – R. rubrum S1H – Whole proteome study 48 proteins was cut in 29 pieces of 1 mm each. In-gel trypsin digestion and LC-MS/MS analysis were performed as described respectively in the 2D-PAGE gel analysis and in IPG-IEF paragraphs.

3.2.3 PROTEIN SUBCELLULAR LOCALIZATION PREDICTION Protein subcellular localization prediction for R. rubrum S1 was achieved using the web- based tool, P-CLASSIFIER (Wang et al., 2005) based on the extraction of the features from protein sequences and available at http://protein.bii.a-star.edu.sg/localization/gram- negative/index.html . Protein sequence information in FASTA format was retrieved from the Joint Genome Institute, Department of Energy (USA) ( http://genome.jgi- psf.org/finished_microbes/rhoru/rhoru.home.html ).

3.2.4 EXPONENTIALLY MODIFIED PROTEIN ABUNDANCE INDEX (EM PAI) The emPAI is a convenient and easily obtained index that can be used to produce protein expression data from any LC-MS/MS runs. This quantification index is based on the number of sequenced peptides per protein. It has been shown to be directly proportional to protein content (Figure 3.3) and has been defined as follow (Ishihama et al ., 2005):

PAI = N obsd /N obsbl

Where N obsd and N obsbl are respectively the number of observed peptides per protein and the number of observable peptides per protein. The emPAI is defined as follows. emPAI = 10 PAI – 1

Ishihama et al . (2008) also showed that protein abundances in the E. coli cytosol as measured by the emPAI approach correlate well with protein copy numbers per cell measured independently by isotope dilution using spiked E. coli BW25113 cells containing 40 proteins with known amounts (Figure 3.4).

Chapter 3 – R. rubrum S1H – Whole proteome study 49

Figure 3.3 . Relationship between protein concentration and emPAI for 46 proteins in neuro2a cells (Ishihama et al ., 2005).

Figure 3.4 . Correlation between observed emPAI values and independently measured protein copy numbers per cell. The Pearson correlation coefficient is 0.84 with a p-value < 10 -10 (Ishihama et al ., 2008).

This index was directly calculated by the Mascot research server but had to be normalized from one experiment to another to avoid bias inherent to data based acquired spectra already mentioned above. This was achieved by calculating the following percentage (Ishihama et al ., 2005):

Normalized emPAI = (emPAI protein /emPAI total ) x 100

Where emPAI protein and emPAI total are respectively the emPAI for a given protein and the sum of the emPAI of all the identified protein within a given experiment.

Chapter 3 – R. rubrum S1H – Whole proteome study 50

3.3 RESULTS

3.3.1 WHOLE PROTEOME IN SILICO PREDICTION In silico prediction tools are powerful tools that permit to easily predict biological characteristics of proteins based on amino-acid sequence. These data can be used to set up protocols and to assess the relevance of experimental results ( e.g . Bjellqvist et al ., 1993; Wang et al ., 2005). As mentioned earlier, the R. rubrum S1 genome comprises a chromosome (4.35 Mb; GenBank ID: CP000230) and a plasmid (53.7 kb; GenBank ID: CP000231). To characterize R. rubrum S1 proteome, the protein molecular weight (MW) and isoelectric point (pI) were predicted using the "Compute pI/Mw tool" from Expasy (http://www.expasy.ch/tools/pi_tool.html). In silico predicted protein MW showed that about 90 % of R. rubrum S1 proteins are distributed from below 10 kDa to 100 kDa (Figure 3.5).

900

800

700

600

500

400

Numberof proteins 300

200

100

0

Da Da Da Da kDa kDa 10 kDa 20 k 30 k 50 kDa 70 10 kDa 20 kDa < - 100 1 1 140 k 140 k 10- 20- 30-40 kDa 40 50-60 kDa 60- 70-80 kDa 80-90 kDa > 90- 30- 100- 110- 120-130 kDa 1 Mass weight Figure 3.5 . In silico predicted protein mass weight of the R. rubrum S1 whole proteome (3,829 candidate proteins).

On the other hand, in silico predicted protein pI showed a classical bi-modal distribution (Schwartz et al ., 2001), as previously observed for other bacterial and archaeal proteomes (Figure 3.6 and figure 3.7). This distribution was highly similar to that corresponding to Synechocystis sp. proteome, another phototrophic bacterium (Figure 3.7).

Chapter 3 – R. rubrum S1H – Whole proteome study 51

Figure 3.6. Histograms of isoelectric ( pI ) values at 0.1 unit intervals for (A) Escherichia coli K12–,(B) Synechocystis sp . strain PCC 6803–, (C) Methanococcus jannaschii –, (D) Pyrococcus abyssi –, (E) Thermotoga maritima –, and (F) Helicobacter pylori –predicted ORFs (from Schwartz et.al ., 2001).

1400

1200

1000

800

600 Numberof proteins

400

200

0 <3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 pI Figure 3.7 . In silico predicted protein isoelectric point ( pI ) of the R. rubrum S1 whole proteome (3,829 candidate proteins).

The P-classifier web-tool was used to predict the localization of the 3829 candidate proteins included in the R. rubrum proteome (Figure 3.8). This analysis predicted 68 % cytoplasmic, 16 % inner membrane, 13 % periplasmic, 2 % outer membrane and 1 % extracellular proteins. The amount of predicted membrane proteins for R. rubrum lied at the lower limit of the range (18 to 29 %) reported by Kihara and Kanehisa (2000) who investigated 11 bacterial complete genomes.

Chapter 3 – R. rubrum S1H – Whole proteome study 52

Outer membrane Extracellular 2 % 1 % Periplasmic 13 %

Inner membrane 16 %

Cytoplasmic 68 %

Figure 3.8 . In silico protein localization prediction of R. rubrum whole proteome (3,829 candidate proteins).

3.3.2 FIRST GENERATION PROTEOMICS USING 2D-PAGE We used protein sequential extraction based Bio-Rad kit and 2D-PAGE approach coupled to MALDI-ToF mass spectrometry to identify proteins from R. rubrum grown in light anaerobic (Figure 3.9) or dark aerobic (Figure 3.10) culture conditions. First, 2D-PAGE analyses of the different samples confirmed a clear sequential protein extraction. The last extraction with buffer 3 (urea, thiourea, CHAPS, SB3-10, Tris and Bio- Lyte 3/10) allowed to specifically extract membrane proteins except one cytoplasmic protein in both case (Table 3.3, spots 82 to 100 for LAN and Table 3.4, spots 66 to 74 for DAE). The sequential extraction of DAE cultures appeared to be more difficult than for the LAN cultures since more overlap was detected between buffer extraction 1 and buffer extraction 2 (Figure 3.11). Therefore, differential comparaison between the different conditions was difficult to achieve. Second, this approach allowed us to identify 56 (out of 100 analyzed spots) and 45 (out of 78 analyzed spots) non-redundant proteins in R. rubrum S1H cultivated respectively in light anaerobic (LAN) and dark aerobic (DAE) culture conditions (Table 3.3 and Table 3.4). Besides, 26 and 15 proteins were specific respectively to LAN and DAE cultures while 30 proteins were common to both bacterial culture conditions (Figure 3.11, Table 3.3 and Table 3.4). For example, we could identify the ribulose bisphosphate carboxylase (RuBisCO) enzyme (RruA2400) related to the fixation of CO 2 via photosynthesis specifically in the LAN growth conditions (spot number 13 in Table 3.3).

Chapter 3 – R. rubrum S1H – Whole proteome study 53

4 pI 7 4 pI 7 (a) (b)

1

3 2 4 70 71 7 5 68 12 72 73 11 10 14 69 9 8 6

15 13 16 74 17 76 2019 21 18

25 24 30 31 32 75 22 33 38 45 29 MW 26 23 42 48 MW 27 28 36 35 34 37 54 77 50 51 44 40 39 46 47 52 81 43 41 49 79 55 53

57 56 78 80 59 62 58 65 64 60 61

63 67 66

4 pI 7 (c)

93 91

92

89 100 85 98 97 87 MW 82 90 99 88 83 86 94 95 96

84

Figure 3.9. R. rubrum cultured in light anaerobic conditions ( LAN ). 2D Coomassie G-250 stained proteomic gel electrophoresis (non linear pI 4-7 18 cm strip, PAGE 11 %) on total protein extract using extraction solution 1 (a) , extraction solution 2 (b) and extraction solution 3 (c) (Bio-Rad). pI : isoelectric point. Number 1 to 100 indicates identified proteins mentioned in Table 3.3.

Chapter 3 – R. rubrum S1H – Whole proteome study 54

Table 3.3. Rhodospirillum rubrum S1H proteins identified in light anaerobic ( LAN ) culture conditions using Bio-Rad sequential extraction kit and 2D-PAGE coupled to MALDI-ToF MS analysis. Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction 1 RruA3511* acnA Aconitate hydratase 1 C 103 95.73 periplasmic 2 RruA2691 fusA Elongation factor G J 133 76.00 cytoplasmic 3 RruA3555* dnaK Molecular chaperone - DnaK O 148 68.00 cytoplasmic 4 RruA3419 pckA Phosphoenolpyruvate carboxykinase C 149 69.00 periplasmic 60 kDa chaperonin (Protein Cpn60) 5 RruA0162* groEL O 167 57.00 cytoplasmic (groEL protein) 6 RruA0072* htpG Heat shock protein Hsp90 O 93 ND cytoplasmic ABC-type oligopeptide transport system, 7 RruA3166 - E 178 68.00 periplasmic periplasmic component Extracellular solute-binding protein family 8 RruA2356* ddpA E 215 58.00 periplasmic 5 precursor Extracellular solute-binding protein family 9 RruA2356* ddpA E 215 58.00 periplasmic 5 precursor Extracellular solute-binding protein family 10 RruA2356* ddpA E 215 58.00 periplasmic 5 precursor Extracellular solute-binding protein family 11 RruA2356* ddpA E 215 58.00 periplasmic 5 precursor ABC-type oligopeptide transport system, 12 RruA0592* oppA E 77 ND periplasmic periplasmic component 13 RruA2400 - Ribulose bisphosphate carboxylase G 65 ND cytoplasmic Extracellular solute-binding protein family 14 RruA1917* ddpA E 111 57.90 periplasmic 5 precursor 15 RruA2702* tufB Translation elongation factor Tu J 103 43.00 cytoplasmic 16 RruA0221* pgk Phosphoglycerate kinase G 79 41.00 cytoplasmic 17 RruA2171* - Extracellular ligand-binding receptor E 101 38.00 periplasmic ABC-type branched-chain amino acid 18 RruA1746* livK E 75 38.00 periplasmic transport systems, periplasmic component ABC-type branched-chain amino acid 19 RruA1746* livK E 75 38.00 periplasmic transport systems, periplasmic component ABC-type branched-chain amino acid 20 RruA1746* livK E 75 38.00 periplasmic transport systems, periplasmic component ABC-type branched-chain amino acid 21 RruA1746* livK E 75 38.00 periplasmic transport systems, periplasmic component ABC transporter, polyamine transport 22 RruA1018* - E 91 40.00 periplasmic protein, periplasmic protein 23 RruA2047 - ABC transporter substrate-binding protein E 73 38.00 periplasmic 24 RruA2047 - ABC transporter substrate-binding protein E 73 38.00 periplasmic 25 RruA2047 - ABC transporter substrate-binding protein E 73 38.00 periplasmic ABC-type amino acid transport/signal 26 RruA1004* hisJ transduction systems, periplasmic E 75 36.00 periplasmic component/domain Putative periplasmic binding protein with 27 RruA2302* rbsB G 101 37.00 periplasmic substrate ribose precursor Putative periplasmic binding protein with 28 RruA2302* rbsB G 101 37.00 periplasmic substrate ribose precursor ABC-type amino acid transport/signal 29 RruA1004* hisJ transduction systems, periplasmic E 75 36.00 periplasmic component/domain 30 RruA0598 - ABC-type phosphate transport system P 94 36.00 periplasmic 31 NA N/A mixture: idem 30 + idem 32 Un. 74 NA NA Sulfate ABC transporter, periplasmic 31 RruA3400 sbp P 97 38.00 periplasmic sulfate-binding protein precursor Sulfate ABC transporter, periplasmic 32 RruA3400 sbp P 97 38.00 periplasmic sulfate-binding protein precursor Uncharacterized ABC-type transport 33 RruA1902 - system, periplasmic component/surface R 173 38.00 periplasmic lipoprotein ABC transporter substrate-binding protein 34 RruA2784 med R 108 35.57 periplasmic precursor Uncharacterized ABC-type transport 35 RruA1902 - system, periplasmic component/surface R 173 37.75 periplasmic lipoprotein Uncharacterized ABC-type transport 36 RruA1902 - system, periplasmic component/surface R 173 37.75 periplasmic lipoprotein TRAP-type C4-dicarboxylate transport 37 RruA3678* dctP G 186 35.40 periplasmic system TRAP-type C4-dicarboxylate transport 38 RruA3678* dctP G 186 35.40 periplasmic system Putative C4-dicarboxylate-binding 39 RruA2239 - G 62 36.74 periplasmic periplasmic protein 40 RruA2239 - Putative C4-dicarboxylate-binding G 62 36.74 periplasmic

Chapter 3 – R. rubrum S1H – Whole proteome study 55

Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction periplasmic protein 41 RruA2106 - Putative uncharacterized protein precursor R 83 33.70 periplasmic Putative ABC transporter periplasmic 42 RruA1336* rbsB G 166 34.24 periplasmic sugar-binding protein 43 RruA2106 - Putative uncharacterized protein precursor R 83 33.70 periplasmic Putative ABC transporter periplasmic 44 RruA1336* rbsB G 166 34.24 periplasmic sugar-binding protein TRAP-type C4-dicarboxylate transport 45 RruA3678* dctP G 186 35.40 periplasmic system ABC transporter amino-acid-binding 46 RruA2087* hisJ E 119 32.55 periplasmic protein ABC-type multidrug transport system, 47 RruA2498 - V 173 37.24 cytoplasmic ATPase component ABC transporter amino-acid-binding 48 RruA2087* hisJ E 119 32.55 periplasmic protein Molybdate-binding periplasmic protein 49 RruA0704 - P 93 26.88 periplasmic precursor 50 RruA2251* - Putative uncharacterized protein E 86 28.42 periplasmic 51 RruA2251* - Putative uncharacterized protein E 86 28.42 periplasmic 52 RruA2251* - Putative uncharacterized protein E 86 28.42 periplasmic ABC amino acid transporter, periplasmic 53 RruA1295* hisJ E 169 30.16 periplasmic ligand binding protein 54 RruA2417* - Lipoprotein, YaeC family precursor P 151 28.17 periplasmic mixture: RruA0894 + RruA0332: Putative 55 NA NA Un. 180 NA NA glutathione transferase mixture: RruA0894 + RruA0332: Putative 56 NA NA Un. 180 NA NA glutathione transferase mixture: RruA0894 + RruA0332: Putative 57 NA NA Un. 180 NA NA glutathione transferase 58 RruA1418* ahpC Alkyl hydroperoxide reductase subunit C O 215 20.69 cytoplasmic 59 RruA3271 gst Glutathione S-transferase O 82 23.43 cytoplasmic Outer-membrane lipoprotein carrier 60 RruA3698 lolA M 69 24.27 outer membrane protein - LolA Magnetic particle membrane specific 61 RruA3283 - S 166 17.53 cytoplasmic GTPase P16 62 RruA1760* sodA Superoxide dismutase P 114 24.93 periplasmic 63 RruA2169 - Nucleoside diphosphate kinase F 93 15.34 cytoplasmic 64 RruA0893 terZ Stress protein homologs of TerZ T 76 20.46 cytoplasmic 65 RruA1742 ppiB Peptidyl-prolyl cis-trans isomerase O 127 18.32 cytoplasmic 66 RruA3216 - Putative uncharacterized protein S 120 ND cytoplasmic 67 RruA2466 - Lactoylglutathione lyase E 95 16.39 cytoplasmic FKBP-type peptidyl-prolyl cis-trans 68 RruA1549* tig O 113 49.21 cytoplasmic isomerase (Trigger factor) 69 RruA2135 mdoG Periplasmic glucans biosynthesis protein P 104 57.52 cytoplasmic 70 RruA1226* atpD ATP synthase beta C 87 50.82 cytoplasmic 60 kDa chaperonin (Protein Cpn60) 71 RruA0162* groEL O 167 57.07 cytoplasmic (groEL protein) 72 RruA3567* hemD Uroporphyrinogen-III synthase H 100 74.01 cytoplasmic 60 kDa chaperonin (Protein Cpn60) 73 RruA0162* groEL O 167 57.07 cytoplasmic (groEL protein) 74 RruA1226* atpD ATP synthase beta C 87 50.82 cytoplasmic 75 RruA2171* - Extracellular ligand-binding receptor E 101 38.63 periplasmic ABC-type branched-chain amino acid 76 RruA1746* livK E 75 38.72 periplasmic transport systems, periplasmic component 77 RruA3567* hemD Uroporphyrinogen-III synthase H 100 74.01 cytoplasmic 78 RruA0894* terD Tellurium resistance protein TerD T 134 20.29 outermembrane 79 RruA0894* terD Tellurium resistance protein TerD T 134 20.29 outermembrane 80 RruA0894* terD Tellurium resistance protein TerD T 134 20.29 outermembrane 81 RruA1218 pspA Putative phage shock protein A K 127 25.26 cytoplasmic 82 RruA2211* ompC Outer membrane protein (Porin) - OmpC M 84 36.62 outermembrane 83 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane Outer membrane protein and related 84 RruA3328* ompA M 74 28.58 periplasmic peptidoglycan-associated - OmpA 85 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane Outer membrane protein and related 86 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 81 34.00 outermembrane OmpA Outer membrane protein and related 87 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 81 34.00 outermembrane OmpA Outer membrane protein and related 88 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 81 34.00 outermembrane OmpA 89 RruA0437* ompA Outer membrane protein and related M 81 34.00 outermembrane

Chapter 3 – R. rubrum S1H – Whole proteome study 56

Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction peptidoglycan-associated (Lipo)proteins - OmpA Outer membrane protein and related 90 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 81 34.00 outermembrane OmpA 91 RruA0488 btuB TonB-dependent receptor precursor H 128 69.54 outermembrane 92 RruA1798 tolC Outer membrane protein - TolC M 108 50.85 outermembrane 93 RruA1511 - Glycosyl transferase, family 2 M 135 86.47 cytoplasmic 94 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane 95 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane 96 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane 97 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane 98 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane 99 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 84 36.27 outermembrane Outer membrane protein and related 100 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 81 34.00 outermembrane OmpA *: protein identified in both LAN and DAE culture conditions. MW: molecular weight. NA : not applicable. ND : not determined. Protein localization was predicted using P-CLASSIFIER. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

4 pI 4 pI 7

1

2 (a) 3 (b) 4 43 5 6 7 42 44 8 45 9 10 46 15

16 14 13 12 11 48 19 47 17 25 49 18 26 50 20 21 22 23 24 51

28 31 52 27 54 MW 53 36 32 29 30 MW 33 34 65 35 64

38 37

55 39 41 58 57 56

40 60 59

62

61 63

Figure 3.10 . R. rubrum cultured in dark aerobic conditions ( DAE ). 2D Coomassie G-250 stained proteomic gel electrophoresis (non linear pI 4-7 18 cm strip, PAGE 11 %) on total protein extract using extraction solution 1 (a) , extraction solution 2 (b) and extraction solution 3 (c) (Bio-Rad). pI : isoelectric point. Number 1 to 74 indicates identified proteins.

Chapter 3 – R. rubrum S1H – Whole proteome study 57

4 pI 7

(c)

74

66 68 70 72

69 73 71 MW 67

Figure 3.10 (continued) . R. rubrum cultured in dark aerobic conditions ( DAE ). 2D Coomassie G-250 stained proteomic gel electrophoresis (non linear pI 4-7 18 cm strip, PAGE 11 %) on total protein extract using extraction solution 1 (a) , extraction solution 2 (b) and extraction solution 3 (c) (Bio-Rad). pI : isoelectric point. Number 1 to 74 indicates identified proteins mentioned in Table 3.4.

Table 3.4 . Rhodospirillum rubrum S1H proteins identified in dark aerobic ( DAE ) culture conditions using Bio- Rad sequential extraction kit and 2D-PAGE coupled to MALDI-ToF MS analysis. Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction 1 RruA3511* acnA Aconitate hydratase 1 C 154 96.13 periplasmic 2 RruA3785 pnp Polyribonucleotide nucleotidyltransferase J 184 76.40 cytoplasmic 3 RruA3555* dnaK Molecular chaperone - DnaK O 154 68.81 cytoplasmic 4 RruA0293 rpsA Ribosomal protein S1 J 120 63.04 cytoplasmic 5 RruA0072* htpG Heat shock protein Hsp90 O 189 68.96 cytoplasmic 60 kDa chaperonin (Protein Cpn60) 6 RruA0587 groEL O 117 57.66 cytoplasmic (groEL protein) 60 kDa chaperonin (Protein Cpn60) 7 RruA0587 groEL O 117 57.66 cytoplasmic (groEL protein) ABC-type oligopeptide transport system, 8 RruA0592* oppA E 143 60.78 periplasmic periplasmic component Extracellular solute-binding protein family 9 RruA2356* ddpA E 200 58.74 periplasmic 5 precursor Extracellular solute-binding protein family 10 RruA2356* ddpA E 200 58.52 periplasmic 5 precursor 11 RruA0454 pepB Leucyl aminopeptidase E 115 52.36 cytoplasmic Extracellular solute-binding protein family 12 RruA1917* ddpA E 184 57.96 periplasmic 5 precursor 13 RruA2086 glnA synthetase E 133 52.50 cytoplasmic 14 RruA2086 glnA Glutamine synthetase E 133 52.50 cytoplasmic FKBP-type peptidyl-prolyl cis-trans 15 RruA1549* tig O 89 49.21 cytoplasmic isomerase (Trigger factor) 16 RruA2201 - Peptidase S1C, Do O 79 52.86 cytoplasmic 17 RruA1885 eno Enolase G 100 44.97 cytoplasmic 18 RruA0221* pgk Phosphoglycerate kinase G 160 41.69 cytoplasmic 19 RruA2702* tufB Translation elongation factor Tu J 161 43.45 cytoplasmic 20 RruA2171* - Extracellular ligand-binding receptor E 94 38.63 periplasmic ABC-type branched-chain amino acid 21 RruA1746* livK E 125 38.63 periplasmic transport systems, periplasmic component ABC-type branched-chain amino acid 22 RruA1746* livK E 125 38.63 periplasmic transport systems, periplasmic component ABC-type branched-chain amino acid 23 RruA1746* livK E 125 38.63 periplasmic transport systems, periplasmic component

Chapter 3 – R. rubrum S1H – Whole proteome study 58

Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction Extracellular ligand-binding receptor 24 RruA3728 livK E 106 40.23 periplasmic precursor ABC transporter, polyamine transport 25 RruA1018* - E 88 40.54 periplasmic protein, periplasmic protein ABC transporter, polyamine transport 26 RruA1018* - E 88 40.54 periplasmic protein, periplasmic protein ABC-type amino acid transport/signal 27 RruA1004* hisJ transduction systems, periplasmic E 176 36.62 periplasmic component/domain Putative periplasmic binding protein with 28 RruA2302* rbsB G 96 37.19 periplasmic substrate ribose precursor 29 RruA0395 surA Parvulin-like peptidyl-prolyl isomerase O 107 33.00 periplasmic Putative ABC transporter periplasmic 30 RruA1336* rbsB G 106 34.23 periplasmic sugar-binding protein 31 RruA1210 mdh Malate dehydrogenase C 76 33.63 cytoplasmic TRAP-type C4-dicarboxylate transport 32 RruA3678* dctP G 138 35.40 periplasmic system ABC transporter amino-acid-binding 33 RruA2087* hisJ E 96 32.55 periplasmic protein 34 RruA2417* - Lipoprotein, YaeC family precursor P 144 28.17 periplasmic ABC amino acid transporter, periplasmic 35 RruA1295* hisJ E 90 30.16 periplasmic ligand binding protein 36 RruA2251* - Putative uncharacterized protein E 77 28.42 periplasmic 37 RruA0894* terD Tellurium resistance protein TerD T 91 20.29 outermembrane 38 RruA0894* terD Tellurium resistance protein TerD T 91 20.29 outermembrane 39 RruA1418* ahpC Alkyl hydroperoxide reductase subunit C O 193 20.69 cytoplasmic Outer membrane protein and related 40 RruA3328* ompA M 95 28.58 periplasmic peptidoglycan-associated (Lipo)proteins 41 RruA1760* sodA Superoxide dismutase P 68 24.93 periplasmic 42 RruA1226* atpD F1-ATP synthase, beta subunit C 104 50.82 cytoplasmic 43 RruA3511* acnA Aconitate hydratase 1 C 154 96.13 periplasmic 44 RruA3785 pnp Polyribonucleotide nucleotidyltransferase J 184 76.40 cytoplasmic 45 RruA3555* dnaK Molecular chaperone - DnaK O 154 68.81 cytoplasmic 46 RruA0293 rpsA Ribosomal protein S1 J 120 63.04 cytoplasmic FKBP-type peptidyl-prolyl cis-trans 47 RruA1549* tig O 89 49.21 cytoplasmic isomerase (Trigger factor) 60 kDa chaperonin (Protein Cpn60) 48 RruA0587 groEL O 117 57.66 cytoplasmic (groEL protein) 48 RruA3567* hemD Uroporphyrinogen-III synthase H 126 74.01 cytoplasmic 60 kDa chaperonin (Protein Cpn60) 49 RruA0162* groEL O 98 57.07 cytoplasmic (groEL protein) 60 kDa chaperonin (Protein Cpn60) 49 RruA0587 groEL O 117 57.66 cytoplasmic (groEL protein) 50 RruA1226* atpD F1-ATP synthase, beta subunit C 104 50.82 cytoplasmic 51 RruA2690 tufB Translation elongation factor Tu J 132 43.40 cytoplasmic 51 RruA2702* tufB Translation elongation factor Tu J 161 43.45 cytoplasmic 52 RruA2171* - Extracellular ligand-binding receptor E 94 38.63 periplasmic DNA-directed RNA polymerase subunit 52 RruA2664 rpoA K 80 37.54 cytoplasmic alpha ABC-type branched-chain amino acid 53 RruA1746* livK E 125 38.72 periplasmic transport systems, periplasmic component ABC-type amino acid transport/signal 54 RruA1004* hisJ transduction systems, periplasmic E 176 36.62 periplasmic component/domain 55 RruA2251* - Putative uncharacterized protein E 77 28.42 periplasmic 56 RruA0894* terD Tellurium resistance protein TerD T 91 20.29 outermembrane 57 RruA0894* terD Tellurium resistance protein TerD T 91 20.29 outermembrane 58 RruA1418* ahpC Alkyl hydroperoxide reductase subunit C O 193 20.69 cytoplasmic 59 RruA1418* ahpC Alkyl hydroperoxide reductase subunit C O 193 20.69 cytoplasmic 60 RruA1418* ahpC Alkyl hydroperoxide reductase subunit C O 193 20.69 cytoplasmic 61 RruA1070 - Heat shock protein Hsp20 O 98 17.30 cytoplasmic 62 RruA1227 atpC ATP synthase F1, epsilon subunit C 91 14.18 cytoplasmic 63 RruA0379 - Putative flavodoxin C 86 17.02 extracellular 64 RruA2417* - Lipoprotein, YaeC family precursor P 144 28.17 periplasmic 65 RruA2251* - Putative uncharacterized protein E 77 ND periplasmic 66 RruA2211* ompC Outer membrane protein (Porin) - OmpC M 80 36.62 outermembrane 67 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 80 36.27 outermembrane 68 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 80 36.27 outermembrane 69 RruA2212* ompC Outer membrane protein (Porin) - OmpC M 80 36.27 outermembrane Outer membrane protein and related 70 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 84 34.00 outermembrane OpmA Outer membrane protein and related 71 RruA0437* ompA M 84 34.00 outermembrane peptidoglycan-associated (Lipo)proteins -

Chapter 3 – R. rubrum S1H – Whole proteome study 59

Spot Gene Gene Mascot MW Localization Protein name COG number number name score (kDa) prediction OpmA Outer membrane protein and related 72 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 84 34.00 outermembrane OpmA Outer membrane protein and related 73 RruA0437* ompA peptidoglycan-associated (Lipo)proteins - M 84 34.00 outermembrane OpmA 74 RruA0072* htpG Heat shock protein Hsp90 O 189 68.96 cytoplasmic *: protein identified in both DAE and LAN culture conditions. COG nomenclature has been presented in the additional table 3.1. MW: molecular weight. ND : not determined. Protein localization was predicted using P-CLASSIFIER. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

LAN.B1 LAN.B2 DAE.B1 DAE.B2

42 3 4 19 14 8

0 LAN DAE 0 0 0 1 0 26 30 15 7 3

LAN.B3 DAE.B3

Figure 3.11. Venn diagrams showing non-redundant proteins identified in light anaerobic ( LAN ) and dark aerobic ( DAE ) culture conditions using 2D-PAGE coupled to MALDI-ToF approach. B1 , B2 and B3 refer to buffer 1, 2 and 3 from the ReadyPrep sequential extraction kit (Bio-Rad).

Clearly, the 2D-PAGE approach permitted us to identify only few major proteins such as chaperones and porins and required a relatively high amount of protein starting material. Moreover, quantitative analysis by gel comparison was not performed due to gel variation and extraction problems for the bacterium growing in different culture conditions. Today, new gel free based approaches have been developed and allowed to increase remarkably the number of identified proteins (Lipton et al ., 2002; Ishihama et al ., 2008). During this study, 3 different shotgun proteomic approaches have been used (Figure 3.1).

Chapter 3 – R. rubrum S1H – Whole proteome study 60

3.3.3 SECOND GENERATION PROTEOMICS

3.3.3.1 2D-LC MS/MS APPROACH

Using the 2D-LC MS/MS approach and starting with 18 µg of R. rubrum digested proteins 231 proteins including 38 hypothetical proteins were identified with a Mascot score of at least 50 (Additional table 3.1). The computed FDR was 0.34 % estimated at the peptide level. Among the 231 proteins, 69 % were predicted to be cytoplasmic, 1 % to be inner membrane, 24 % to be periplasmic, 4% to be outer membrane and 2 % to be extracellular proteins (Figure 3.12a). Our results show a fairly good coverage for cytosolic, periplasmic and extracellular proteins. This coverage was highly biased towards the identification of cytosolic proteins, presumably due to a better solubility of cytosolic proteins compared to membrane proteins, especially in our experimental conditions. Indeed, the protein extraction was realized, without detergent, in presence of 6 M guanidine chloride. Moreover, just before the trypsin treatment, the proteins were precipitated in presence of acetone and were resolubilized in 2 M urea. During these steps, it might occur that some hydrophobic proteins do not precipitate or do not resolubilise in 2 M urea.

3.3.3.2 PEPTIDE IPG-IEF LC MS/MS APPROACH

Clearly, improved coverage of membrane proteins from inner and outer membrane can be achieved by coupling an IPG-IEF strategy with better membrane-solubilising strategies. In the first step, the proteins were extracted in 6 M urea, 2 M thiourea solution instead of in 6 M guanidine chloride. The remaining pellet was then treated with N-octyl-glucoside to increase the membrane protein solubilisation. The first extraction protocol (using the urea/thiourea extraction buffer) using twice 300 µg of R. rubrum digested proteins, resulted in the identification of 905 proteins including 113 hypothetical proteins with a Mascot score of at least 50 (Additional table 3.2). The distribution of the 905 proteins in regard to subcellular localization prediction showed 74 % cytoplasmic, 7 % inner membrane, 16 % periplasmic, 2 % outer membrane and 1 % extracellular proteins (Figure 3.12b).

Chapter 3 – R. rubrum S1H – Whole proteome study 61

Extracellular Extracellular Outer membrane 2 % 4 % Outer membrane 1 % 2 % Periplasmic Periplasmic 16 % 24 % (a) (b)

Inner membrane 7 %

Cytoplasmic Inner membrane 69 % 1 % Cytoplasmic 74 %

Extracellular Extracellular 1 % (c) 4 % Outer membrane (d) Outer membrane 6 % 10 %

Periplasmic 16 %

Periplasmic 26 %

Cytoplasmic 59 % Inner membrane 8 % Cytoplasmic 69 % Inner membrane 1 %

Figure 3.12 . In silico protein localization prediction of the R. rubrum proteins identified using the 2D-LC (a) , IPG-IEF LC using the first (b) and second (c) protocols and the SDS-PAGE (d) LC-MS/MS (see material and methods). Respectively, 231, 905, 78 and 345 proteins were identified with a Mascot score above 50.

The second extraction protocol with the N-octyl-glucoside extraction buffer on the bacterial pellet from the first extraction and using twice 250 µg of R. rubrum digested proteins resulted in the identification of 78 proteins including 17 hypothetical with a Mascot score of at least 50 (Additional table 3.3). The distribution of these 78 proteins within the cells predicted 59 % cytoplasmic, 1 % inner membrane, 26 % periplasmic, 10 outer membrane and 4 % extracellular proteins (Figure 3.12c). The computed FDR estimated at the peptide level were 0.52 % and 1.72 % respectively for the first extraction and the second extraction. Only 10 proteins including one predicted membrane protein RruA2448 (hypothetical) were found to be specific to the second extraction (Additional table 3.3).

3.3.3.3 SDS-PAGE LC MS/MS APPROACH

From the 29 analyzed SDS-PAGE fractions (Figure 3.1d), we could retrieve 345 non- redundant proteins including 63 hypothetical proteins identified by a Mascot score of at least

Chapter 3 – R. rubrum S1H – Whole proteome study 62 50 (Additional table 3.4). The computed FDR was 0.82 % estimated at the peptide level. Protein prediction model foreseen 69 % cytoplasmic, 8 % inner membrane, 16 % periplasmic, 6 % outer membrane and 1 % extracellular proteins (Figure 3.12d).

3.3.4 COMPARISON BETWEEN THE DIFFERENT EXPERIMENTAL APPROACHES

3.3.4.1 FOREWORD

The 2D-PAGE technique was not compared to the other approaches for the following reasons: (i) no specific protein was found for this approach, all occurrences have also been detected using the MudPIT analysis, (ii) no quantitative information was available and (iii) only a few non-redundant proteins were identified. Besides, because of the low contribution and the lack of functional specificity of the 10 proteins gained from the second extraction protocol for the IPG-IEF, the latter were not taken into account for the comparison mentioned below except for the identification comparison where all proteins from the IPG-IEF approach were pooled (i.e . 915 occurrences). On the other hand, except for the number of identified proteins and protein localization comparison, the two replicate injections of the first protocol IPG-IEF samples were treated separately in an effort to assess the reproducibility of the method. Therefore, in the following section, we compared 2D-LC to IPG-IEF LC and SDS-PAGE LC MS/MS. Eventually, as a reminder, one must be aware that in our MudPIT approach peptide spectra were acquired in a data dependent manner (see material and methods). One peptide undetected during one run did not necessarily mean the peptide was absent. Therefore, the word "specifically" used below should not be understood as exclusive.

3.3.4.2 ABSOLUTE NUMBER OF NON-REDUNDANT PROTEIN IDENTIFIED Combining the results from all the approaches, we end up with the identification of 1,007 R. rubrum non-redundant proteins. When we compared the 3 approaches mentioned above in term of identified proteins, we saw a relative high overlap between the techniques and therefore only a few proteins remained specific to one particular approach (except for the IPG-IEF approach obsviously) (Figure 3.13). The 19 proteins specific for the 2D-LC approach were all predicted to be either cytoplasmic or periplasmic and included 12 hypothetical proteins (Table 3.5). On the other hand, of the 66 proteins specific to the SDS- PAGE membrane enriched samples 13 were predicted membrane proteins (Table 3.6). In that

Chapter 3 – R. rubrum S1H – Whole proteome study 63 purpose, we can mention that RruA1266 annotated as photosystem assembly protein could be incorrect since expert annotation using the MaGe platform (Vallenet et al . 2006) revealed that annotation has no experimental evidence and that protein could be annotated as conserved hypothetical protein only.

2D -LC IPG -IEF

19 99 544 106 7 166

66

SDS -PAGE

Figure 3.13 . Venn diagrams showing non-redundant proteins identified in the 2D-LC, IPG-IEF LC (pooled results from first and second protocols) and SDS-PAGE LC MS/MS approaches.

Table 3.5 . 19 R. rubrum S1H proteins identified in dark aerobic ( DAE ) culture conditions using specifically the 2D-LC MS/MS approach. Gene Gene Mascot MW Localization Product name COG #peptides number name score (kDa) prediction RruA0089 - ATP-dependent DNA helicase RecQ L 133 14 68.89 cytoplasmic RruA0296 - hypothetical protein Rru_A0296 S 183 4 12.61 cytoplasmic RruA0422 - hypothetical protein Rru_A0422 Un. 217 4 13.81 periplasmic RruA0497 - ABC transporter component E 72 8 39.34 cytoplasmic RruA1033 - hypothetical protein Rru_A1033 S 263 1 18.62 cytoplasmic RruA1042 - hypothetical protein Rru_A1042 S 614 13 10.23 cytoplasmic RruA1175 - CsbD-like S 884 26 7.61 cytoplasmic RruA1353 - hypothetical protein Rru_A1353 S 332 29 18.19 periplasmic RruA1354 - hypothetical protein Rru_A1354 Un. 388 15 12.81 cytoplasmic RruA1446 - hypothetical protein Rru_A1446 S 123 3 18.33 cytoplasmic RruA1665 rpmF 50S ribosomal protein L32 J 287 7 7.03 cytoplasmic RruA1867 - hypothetical protein Rru_A1867 M 75 10 16.58 cytoplasmic RruA2061 - hypothetical protein Rru_A2061 Un. 2708 3 21.44 periplasmic RruA2062 - hypothetical protein Rru_A2062 Un. 258 51 12.07 periplasmic RruA2208 - hypothetical protein Rru_A2208 R 174 4 9.59 cytoplasmic RruA2723 - rubrerythrin S 127 4 19.25 cytoplasmic RruA3286 - hypothetical protein Rru_A3286 Un. 252 7 16.24 periplasmic RruA3521 - 50S ribosomal protein L35P J 60 1 7.41 periplasmic RruB0047 - NAD-dependent epimerase/dehydratase M 96 10 38.71 cytoplasmic 'RruA' and 'RruB' refers to gene located respectively on the chromosome or on the plasmid. COG nomenclature has been presented in the additional table 3.1. MW: molecular weight. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Chapter 3 – R. rubrum S1H – Whole proteome study 64

Table 3.6 . 66 R. rubrum S1H proteins identified in dark aerobic ( DAE ) culture conditions using specifically the SDS-PAGE LC MS/MS approach. Gene Gene Mascot MW Localization Product name COG #peptides number name score (kDa) prediction RruA0096 - secretion protein HlyD V 175 3 34.14 cytoplasmic RruA0122 - hypothetical protein Rru_A0122 Un. 215 9 12.73 periplasmic RruA0231 - hypothetical protein Rru_A0231 Un. 63 1 21.7 periplasmic RruA0291 - hypothetical protein Rru_A0291 Un. 130 2 12.36 innermembrane RruA0333 dps Ferritin and Dps P 103 1 17.65 cytoplasmic ubiquinol-cytochrome c reductase, iron- RruA0364 - C 286 22 19.79 periplasmic sulfur subunit organic solvent tolerance protein OstA- RruA0434 - M 57 2 94.76 outermembrane like cyclic nucleotide-binding domain- RruA0496 - T 159 2 13.8 cytoplasmic containing protein RruA0636 - hypothetical protein Rru_A0636 S 77 2 11.53 cytoplasmic RruA0637 - hypothetical protein Rru_A0637 S 520 11 12.85 periplasmic RruA0755 - chemotaxis sensory transducer N 1066 22 60.75 innermembrane RruA0782 - hypothetical protein Rru_A0782 Un. 101 1 12.48 cytoplasmic RruA0823 - cupin region S 74 2 34.05 cytoplasmic RruA1073 motB hypothetical protein Rru_A1073 N 327 8 41.3 outermembrane MotA/TolQ/ExbB proton channel RruA1074 - Un. 422 9 46.94 innermembrane family protein RruA1091 - MotA/TolQ/ExbB proton channel U 345 5 33.08 innermembrane RruA1093 - hypothetical protein Rru_A1093 S 61 1 37.04 periplasmic RruA1130 glnK nitrogen regulatory protein P-II E 57 1 12.28 cytoplasmic RruA1218 pspA phage shock protein A, PspA K 144 3 25.26 cytoplasmic RruA1266 - (photosystem I assembly BtpA) R 144 5 28.56 innermembrane RruA1402 - CheW protein N 93 1 19.26 cytoplasmic RruA1438 rpoE sigma-24 (FecI) K 59 2 19.29 cytoplasmic plasmid maintenance system antidote RruA1535 - R 60 1 13.65 periplasmic protein methyl-accepting chemotaxis sensory RruA1554 - N 155 5 52.27 cytoplasmic transducer RruA1563 - NADH dehydrogenase subunit I C 144 3 19.09 cytoplasmic RruA1668 - integration host factor, alpha subunit L 207 5 11.44 cytoplasmic RruA1733 - hypothetical protein Rru_A1733 S 54 1 40.27 periplasmic RruA1843 - OmpA/MotB N 62 2 37.09 periplasmic RruA1888 - protein translocase subunit secG U 492 10 14.27 periplasmic RruA1891 - anthranilate synthase component I E 80 1 55.01 cytoplasmic RruA1985 - hypothetical protein Rru_A1985 Un. 263 6 15.59 outermembrane RruA2082 - hemerythrin HHE cation binding region P 86 2 23.81 cytoplasmic biopolymer transport protein RruA2154 - U 86 2 16.11 cytoplasmic ExbD/TolR RruA2218 - short chain dehydrogenase Q 75 4 28.28 cytoplasmic RruA2375 - hypothetical protein Rru_A2375 Un. 239 5 9.19 periplasmic RruA2419 - ABC transporter component P 106 3 27.98 cytoplasmic RruA2424 - gamma-glutamyltranspeptidase E 165 7 40.85 periplasmic RruA2438 - hypothetical protein Rru_A2438 Q 73 2 15.93 periplasmic RruA2450 - peptidase M48, Ste24p R 61 2 27.86 periplasmic Alpha,alpha-trehalose-phosphate RruA2485 - G 54 2 52.32 cytoplasmic synthase (UDP-forming) RruA2605 - phosphoheptose isomerase G 88 2 19.76 cytoplasmic RruA2608 - hypothetical protein Rru_A2608 Un. 151 3 10.75 cytoplasmic RruA2653 - hypothetical protein Rru_A2653 Un. 87 3 24.18 cytoplasmic RruA2760 - hypothetical protein Rru_A2760 Un. 394 5 7.68 periplasmic RruA2774 tar chemotaxis sensory transducer N 1658 31 77.36 innermembrane RruA2839 - hypothetical protein Rru_A2839 Un. 54 2 27.14 cytoplasmic RruA2894 fepC ABC transporter component P 109 1 32.03 innermembrane RruA2910 - hypothetical protein Rru_A2910 Q 201 6 33.1 outermembrane ABC-type uncharacterized transport RruA2911 - R 350 8 21.95 cytoplasmic system auxiliary component-like RruA2968 - hypothetical protein Rru_A2968 Un. 116 1 20.08 periplasmic RruA3038 - hypothetical protein Rru_A3038 Un. 230 3 14.43 cytoplasmic RruA3117 - polysaccharide export protein M 72 1 22.51 periplasmic RruA3153 - virulence factor protein S 90 2 99.35 cytoplasmic RruA3154 - hypothetical protein Rru_A3154 S 266 8 111.29 cytoplasmic RruA3155 tonB hypothetical protein Rru_A3155 M 127 5 55.93 periplasmic RruA3193 - glutathione S-transferase-like protein O 67 1 24.62 innermembrane RruA3253 - hypothetical protein Rru_A3253 A 188 3 38.72 cytoplasmic RruA3299 - glycoside hydrolase family protein G 74 1 76.16 cytoplasmic RruA3409 - hypothetical protein Rru_A3409 Un. 82 2 31.63 cytoplasmic RruA3414 - ABC transporter component E 87 2 27 cytoplasmic RruA3437 - aminoglycoside phosphotransferase R 59 1 39.31 cytoplasmic RruA3474 bfr bacterioferritin P 77 2 18.57 cytoplasmic RruA3593 - phosphoribosyl-ATP pyrophosphatase E 71 3 12.42 cytoplasmic

Chapter 3 – R. rubrum S1H – Whole proteome study 65

Gene Gene Mascot MW Localization Product name COG #peptides number name score (kDa) prediction RruA3649 - 2',5' RNA ligase J 89 1 21.69 cytoplasmic RruA3699 acrB acriflavin resistance protein V 221 6 109.79 innermembrane RruB0016 - Phage integrase L 56 3 23.76 cytoplasmic 'RruA' and 'RruB' refers to gene located respectively on the chromosome or on the plasmid. MW: molecular weight. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified

3.3.4.3 MOLECULAR WEIGHT AND ISOELECTRIC POINT

In contrast with the other approaches, the overall trend of the molecular weight (MW) distribution (Figure 3.14a) for the proteins identified during the IEF experiments was very similar to the theoretical proteome, suggesting no real bias in our approach. By comparing MW profiles of the proteins identified in the SDS-PAGE and 2D-LC experiments, we could see that the 2D-LC and the SDS-PAGE approaches were more favourable to a range of MW of 10-20 kDa and 20-30 kDa, respectively. Besides, we could see similar isoelectric point (pI) distribution for the proteins identified in the IEF experiments compared to the theoretical distribution (Figure 3.14b), strengthening the hypothesis that the IEF experiment correctly reflected the proteome coverage.

40 2D-LC 35 IEF.I.1-LC IEF.I.2-LC SDS-PAGE-LC 30

25

(a) 20

15

10 Percentage of identified proteins identified Percentage of 5 0 a D Da kDa Da Da kDa Da Da Da 0 kDa 0 kDa 0 kDa kDa 10 k 30 k 40 70 k 80 k 90 00 10 k 20 k 30 k 40 kDa < - - - - - 1 1 1 1 1 0 0 0 0 0 - - - >140 kDa 10-2 2 3 40-5 50-6 6 7 8 90 10 20 100- 1 1 130- MW Figure 3.14 . Distribution of R. rubrum S1H identified proteins from the 2D-LC, IPG-IEF injection 1 ( IEF.I.1 ) and injection 2 ( IEF.I.2 ), and SDS-PAGE LC MS/MS based on molecular weight ( MW) (a) and isoelectric point ( pI ) (b) .

Chapter 3 – R. rubrum S1H – Whole proteome study 66

50 2D-LC 45 IEF.I.1-LC IEF.I.2-LC 40 SDS-PAGE-LC

35 30 25 (b) 20

15

proteins identified Percentage of 10 5

0

3 4 5 6 0 1 2 < -8 1 1 1 3- 4- 5- 6-7 7 8-9 9- 0- >13 1 11- 12-13 pI

Figure 3.14 (continued) . Distribution of R. rubrum S1H identified proteins from the 2D-LC, IPG-IEF injection 1 ( IEF.I.1 ) and injection 2 ( IEF.I.2), and SDS-PAGE LC MS/MS based on molecular weight ( MW) (a) and isoelectric point ( pI ) (b) .

3.3.4.4 NUMBER OF PEPTIDES PER PROTEINS

SDS-PAGE experiment significantly increased the number of peptides identified per protein thereby increasing the confidence level of the protein identification. In contrast with the MW and the pI distributions, a notable difference appeared between the first and the second injection of the IPG-IEF from the first protocol regarding the number of peptides per identified protein (Figure 3.15). Indeed, the first injection showed about half the percentage of identified proteins by more than 10 peptides in regard to the second injection. Moreover, the latter analysis detected fewer proteins with only 1 or 2 peptides. This could be related to the preparation of the samples for the IPG-IEF experiment. Indeed, for the first injection sample, trypsinolysis was stopped using the usual 5 % formic acid. However, this lead to the precipitation of peptides that could not be completely resolubilized using NaOH. Therefore, the inconsistency regarding the number of peptides per protein between the first and the second injection confirmed that a bunch of peptides were lost during the preparation of the sample for the first injection.

Chapter 3 – R. rubrum S1H – Whole proteome study 67

35 2D-LC IEF.I.1-LC IEF.I.2-LC 30 SDS-PAGE-LC

25

20

15

10

Percentage of Percentageidentified proteins

5

0

1 2 3 4 5 6 7 8 9 10>10 Number of peptides/proteins

Figure 3.15 . Distribution of R. rubrum S1H identified proteins from the 2D-LC, IPG-IEF injection 1 ( IEF.I.1 ) and injection 2 ( IEF.I.2 ), and SDS-PAGE LC MS/MS based on the number of peptides detected per protein.

3.3.4.5 EXPONENTIALLY MODIFIED PROTEIN ABUNDANCE INDEX (EM PAI)

One way to get an idea of the quantity of a protein in a complex mixture is to calculate its emPAI (Ishihama et al ., 2005), which basically represents the number of identified peptides divided by the number of theoretical tryptic peptides. The median raw emPAI values were respectively: 0.56, 0.27, 0.39 and 0.43 for the 2D-LC, IEF first injection, IEF second injection and SDS-PAGE approaches. The emPAI value confirmed the data from the number of peptides per protein since the emPAI was lower for the first injection where more proteins were identified with 1 or 2 peptides. Besides, the most abundant proteins such as the chaperonins CsbD-like (RruA1175) and

GroEL (RruA0587), the F 0F1 ATP synthase subunit beta AtpD (RruA1226), the elongation factors Tuf1 and Tuf2 (RruA2690 and RruA2702), the cytochrome C (RruA1020) and the ribosomal RpsS (RruA2684) were present in the cytoplasmic and periplasmic fractions (Figure 3.16 and Table 3.7). We could also mention the protease co-factor HflC (RruA2202) and the tellurium stress protein TerD (RruA0894) as the most abundant proteins in the inner membrane and outer membrane fraction respectively. It could be surprising to find the OmpA protein (RruA3328) predicted to be periplasmic but looking into detail for the localization score prediction (26.7 % cytoplasmic, 13.3 % inner membrane, 20.0 % outer membrane and 40.0 % periplasmic) indicated some membrane domains. On the other hand, protein annotated

Chapter 3 – R. rubrum S1H – Whole proteome study 68 with the term 'metabolism' like the thiamine biosynthesis (RruA2008), the pyruvate kinase (RruA2465) and the NADH peroxidase (RruA2309) related proteins represented less than 0.02 % of the total emPAI and therefore could be seen as low abundant.

8 RruA1175 2D-LC 7 IEF.I.1-LC IEF.I.2-LC SDS-PAGE-LC 6 RruA1226 RruA0587 5 RruA1020 RruA2690 RruA2684 4 RruA2702 3 RruA2202 RruA3328 RruA0894 2

1

Percentage of the total emPAI per experiment per the total of emPAI Percentage 0 Cytoplasmic Inner membrane Periplasmic Outer membrane Extracellular

Gene Gene emPAI emPAI emPAI emPAI Product name number name 2D-LC IEF.I.1-LC IEF.I.2-LC SDS-PAGE-LC RruA1175 - CsbD-like 7.49 ND ND ND RruA1226 atpD F0F1 ATP synthase subunit beta 0.38 1.73 1.25 6.14 RruA0587 groEL chaperonin GroEL 2.13 1.24 1.65 5.57 RruA2690 tuf1 elongation factor Tu 1 4.21 1.36 1.06 0.32 RruA2702 tuf2 elongation factor Tu 2 3.88 1.24 1.06 0.28 RruA1020 - cytochrome c, class I 0.96 2.82 0.69 ND RruA2684 rpsS SSU ribosomal protein S19P ND 2.75 1.39 0.2 RruA3328 ompA OmpA/MotB 0.35 0.52 0.31 2.61 RruA0894 - putative tellurium resistance protein TerD 2.39 0.92 0.57 0.22 RruA2202 - putative HflC, protease cofactor ND 0.15 0.21 2.34

Figure 3.16 and Table 3.7 . The figure shows emPAI distribution of R. rubrum S1H identified proteins from the 2D-LC, IPG-IEF injection 1 ( IEF.I.1 ) and injection 2 ( IEF.I.2 ), and SDS-PAGE LC MS/MS experiments in regard to protein localization prediction. Each marker on the plot represents one single protein. The table shows the 10 most abundant proteins using the different approaches (highest value in bold ). ND : not detected.

Considering emPAI related COG distribution from all data (Figure 3.17), the 'function unknown' (S) class contained the most abundant protein whatever the technique employed, namely RruA1175. The second most abundant protein was retrieved in the 'energy and metabolism' (C) class, namely RruA1226 while the third one was the GroEL chaperone (RruA0587) from the 'posttranslational modification, protein turnover, chaperones' (O) class. Then the top 5 was rounded out with RruA2690 and RruA2702, both being elongation factors from the 'translation, ribosomal structure and biogenesis' (J) class. In that purpose, the classes showing the most abundant proteins logically appeared to be the O and the J class containing respectively the chaperones and the ribosomal proteins. We could also mention the relative

Chapter 3 – R. rubrum S1H – Whole proteome study 69 high abundance of RruA3283, a granule associated protein also named phasin most commonly associated with polyhydroxybutyrate (PHB) inclusions. Finally, SDS-PAGE approach seemed to be the most appropriate approach to study membrane proteins since it gave the most abundant proteins in the 'cell envelope biogenesis, outer membrane' (M) class (Figure 3.17).

8 RruA1175 2D-LC IEF.I.1-LC 7 IEF.I.2-LC SDS-PAGE-LC RruA1226 6 RruA0587

5 RruA2690 4 RruA3283 RruA2702 3

2

1 Percentage of the total emPAI per experiment per thetotal of emPAI Percentage

0 EGMDCSRPUIFN H VL O A Q T K J Un. COG

Figure 3.17 . emPAI distribution of R. rubrum S1H identified proteins from the 2D-LC, IPG-IEF injection 1 (IEF.I.1 ) and injection 2 ( IEF.I.2 ), and SDS-PAGE LC MS/MS experiments in regard to the COG classification. See the text for gene name. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

3.3.5 IDENTIFICATION OF PROTEIN SPECIFIC FOR THE RHODOSPIRILLACEAE

FAMILY By means of phylogenomic, Gupta and Mok (2007) have described signature proteins for the alpha Proteobacteria and its main groups including the group. Within the latter group, they found 14 proteins unique for the Rhodospirillaceae family ( Rhodospirillum and Magnetosprillum ). During our experiments, we could identify 5 of these Rhodospirillaceae specific proteins: RruA1756, RruA2112, RruA2510, RruA3662 and RruA3739 (Table 3.7). All these 5 proteins were identified thanks to the first protocol of the

Chapter 3 – R. rubrum S1H – Whole proteome study 70 IPG-IEF MS/MS approach (Additional table 3.2) while RruA2112 and RruA2510 were also identified after SDS-PAGE LC MS/MS (Additional table 3.4). Beside, RruA3662 and RruA3739 were also detected with the 2D-LC and the second protocol of the IPG-IEF MS/MS approaches. Concerning their abundance based on emPAI measurement, only RruA1756 (0.48) and RruA3662 (0.61) were above the emPAI median value for the IEF first injection (0.39). While RruA2510 (0.35) was closed to the median value, RruA3739 (0.25) and especially RruA2112 (0.12) were below that value. These 5 proteins were previously all annotated as 'hypothetical protein'. From now on, they could be upgraded to 'protein of unknown function'. Only a putative secreted fate for RruA2112 and RruA3739 proteins could be deduced from the presence of predicted signal peptide cleavage sites in the amino acid sequences (SignalP 3.0 server: http://www.cbs.dtu.dk/services/SignalP/ ).

Table 3.7 . Rhodospirillaceae group specific proteins detected during this study. Gene Gene MW Product name Mascot score # peptides SC (%) number name (kDa) RruA1756 - hypothetical protein Rru_A1756 118.87 2 9.35 15.78 RruA2112 - hypothetical protein Rru_A2112 197.76 3 20.68 26.22 RruA2510 - hypothetical protein Rru_A2510 191.62 3 31.52 20.58 RruA3662 - hypothetical protein Rru_A3662 351.86 5 29.41 12.62 RruA3739 - hypothetical protein Rru_A3739 165.96 4 14.54 58.03 SC : sequence coverage. MW: molecular weight.

3.3.6 IDENTIFICATION OF NEW RHODOSPIRILLUM RUBRUM S1 PROTEINS Magnifying Genomes (MaGe) is a microbial genome annotation system based on a relational database containing information on bacterial genomes, as well as a web interface to achieve genome annotation projects (Vallenet et al ., 2006). Rhodospirillum rubrum S1 is part of the MaGe 'MagnetoScope' project (https://www.genoscope.cns.fr/agc/mage/wwwpkgdb/MageHome/index.php?webpage=mage ) The MaGe system used a CoDing Sequence (CDS) model algorithm which is different from NCBI and allowed them to identify 258 extra CDS (MaGe website). The present study permitted us to confirm the translation of two of these 'extra CDS', namely RHOP2K1139 and RHOP2K1150 both detected in the 2D-LC while RHOP2K1139 alone was detected in the first protocol of the IPG-IEF LC MS/MS approach. Reliable sequence coverage and fragmentation spectra lead to a Mascot score above 50 (Figure 3.17). The emPAI value for RHOP2K1139 was 1.39 which was almost triple the median in 2D-LC experiment (0.56). Besides, RHOP2K1150 emPAI value was exactly equal to the median for the same experiment. The relative abundance of RHOP2K1139 was even more pronounced in

Chapter 3 – R. rubrum S1H – Whole proteome study 71 the IPG-IEF experiments where the same 1.39 emPAI value was found for the first (median = 0.27) and second injection (median = 0.39). No known function had been found yet for these two proteins but they have putative orthologs in magneticum AMB-1 and Rhodobacter sphaeroides 2.4.1 (MaGe website) and then they were annotated as 'conserved hypothetical proteins'.

RHOP2K1139 RHOP2K1150

Sequence coverage: 45 % Sequence coverage: 84 % (protein score: 174) (protein score: min. 65)

1 MGQDAR KDAL NNRHAELEER LEHEATRPIP DTSVVTTLKR QKLRIKDEIH 1 MDTDDLDPR A KPVKPVSLDD MTVKDLEDYI SALEAEIARA RAQIAAKGGA 51 RMEAR 51 FNAAASVFK R

MS/MS Fragmentation of LEHEATRPIPDTSVVTTLKR MS/MS Fragmentation of DLEDYISALEAEIAR (ion score: 92) (ion score: 106) Intensity Intensity

m/z m/z

Figure 3.17 . Sequence coverage and peptide fragmentation spectra of proteins identified from the MaGe CDS models. Matched peptides are shown in bold in the sequence coverage. For RHOP2K1150, the 84 % protein sequence coverage resulted from 2 analyses.

3.4 DISCUSSION First generation vs second generation proteomics . Recently, Selao et al . (2008) published the first differential study of the R. rubrum S1 proteome using 2D-PAGE and MALDI-ToF. They have identified 44 differentially expressed proteins from gels containing more than 600 protein spots when R. rubrum S1 was grown anaerobic and photoheterotrophically, with different nitrogen sources. Similar 2D-maps were obtained in our study. However, the experimental approach did not allow us to perform differential proteomics. Indeed, overlap between the different protein fractions (especially in the DAE growth conditions) were observed indicating that the sequential extraction was not quantitative. As a consequence, a reliable differential study could not be achieved. Our study made clear that a second generation proteomic approach was much more efficient in terms of protein identification and was less time-consuming in detecting protein from complex samples. A high prefractionation appeared to be an essential step in reducing sample complexity and increasing depth of field to detect low-abundance proteins. Indeed, by design,

Chapter 3 – R. rubrum S1H – Whole proteome study 72 first and second generation proteomic techniques favor abundant proteins. Therefore, fractionation of the sample was essential for the detection of less abundant proteins. We showed that IEF fractionation before LC-MS/MS analysis resulted in a substantial increase in protein identification number compared to 2D-LC-MS/MS. On-line strong cation exchange prefractionation before LC-MS/MS analysis appeared to be slightly complementary to the IEF-LC MS/MS approach. However we should mention that the protocol for the 2D-LC approach was not optimal. Indeed, it constituted our first contact with the MudPIT approach and 2 major parameters have been improved since then, and were used for the IPG-IEF approach: (i) Peptides have to be desalted in order to clean them and to reduce unspecific interactions with the SCX, (ii) Multiple injections appeared to be crucial to increase the protein coverage. As mentioned above, spectra were acquired in a data-dependent manner and multiple injections increased the probability for a peptide to be selected for sequencing and therefore increased the proteome coverage. Moreover, our present capillary configuration for the 2D-LC analysis could be adapted to an off-line SCX to increase the sample loading capacity and to use as much as possible for the IPG-IEF approach (Gan et al ., 2005). However, to date, the best proteome coverage has been achieved using the peptide IPG-IEF prefractionation step before mass spectrometry (Graumann et al ., 2007). Based on the present results, we could propose a specific workflow based on peptides IPG- IEF and proteins SDS-PAGE prefractionation and multiple sample injection to insure maximal bacterial proteome coverage (figure 3.1c,d and e). Indeed, we showed that the IPG- IEF approach covered all classes of proteins (from cytoplasmic to extracellular) while SDS- PAGE appeared to be the best complement to the first approach. However, even without considering the improvement concerning the loading capacity mentioned above, the 2D-LC approach should not be neglected because it coupled low protein amount with low lab work and gives results in a few days in regard to the IPG-IEF approach that needs about 2 weeks. It could therefore easily be added to whole proteome studies and will always give 'a plus' to the proteome coverage. R. rubrum proteome status . The current best proteome coverage has been achieved by Rohmer et al . (2008) covering 63 % ( i.e . 1,083 proteins) of the γ-proteobacterium Francisella tularensis subsp.novicida U112 proteome. This has been reached by analyzing quadruplicate injections on a linear ion trap-Fourier transform-ion cyclotron resonance mass spectrometer that permits higher sensitivity and faster data acquisition than the ion-trap used here. Nevertheless, pooling the data from the different approaches, we could identify 25 % ( i.e .

Chapter 3 – R. rubrum S1H – Whole proteome study 73 1,007 proteins) from the R. rubrum S1H proteome cultivated in dark aerobic growth conditions which is comparable to recent studies using a similar mass spectrometer (Pan et al ., 2008; Ishihama et al ., 2008; Assiddiq et al ., 2008). Moreover, our results further supported the observation that the number of proteins expressed in proteobacteria in a stable condition is around 1,000 and this number remains similar independent of genome size (Kolker et al ., 2003; Corbin et al ., 2003; Jaffe et al ., 2004; Elias et al ., 2006; VerBerkmoes et al ., 2006; Zhang et al ., 2006; Ishihama et al ., 2008; Rohmer et al ., 2008). Therefore, our study constituted the first large scale proteomic study of R. rubrum S1H since Selao et al . (2008) identified 44 differentially expressed proteins of 600 protein spots using the 2D-PAGE approach. Furthermore, we could confirm the translation of 16 % of the total hypothetical proteins (154 items) including 5 signature proteins for the Rhodospirillaceae group (Gupta and Mok, 2007). Moreover, 2 newly predicted R. rubrum CDS from the MaGe gene annotation platform which were not included in the NCBI annotation were identified with high confidence. In that purpose, all the MS and MS/MS data acquired so far could be used in the future to look for new Open Reading Frames namely the proteogenomic approach, a cost-effective means to add value to genome annotation (Jaffe et al ., 2004; Fermin et al ., 2006; Ansong et al ., 2008). Exponentially modified Protein Abundance Index . The emPAI analysis permitted us to estimate the absolute protein amount of R. rubrum S1H grown in dark aerobic culture conditions and put forward common highly abundant proteins like those involved in protein synthesis and energy metabolism (Ishihama et al ., 2008; Assiddiq et al ., 2008). On the other hand, our study revealed the high abundance of chaperone proteins, already mentioned for Methylobacterium extorquens AM1, another α-proteobacterium (Laukel et al ., 2004, Bosch et al ., 2008). Because the emPAI index is easily calculated from the information output of a database search engine like Mascot, this information should be kept in the reporting of all proteomic results. Hence, the emPAI approach could be used to quantify protein abundance in cells from various growth conditions (Ishihama et al ., 2005). However, one must be aware that this approach is biased by the prefractionation method that could change the actual protein concentration due to sample preparation and peptides ionization. Moreover, Hendrickson et al. (2006) mentioned that the emPAI approach is not as sensitive as stable isotope labeling for detecting small changes ( i.e . in the order of two-fold) in protein abundance. As a consequence, it could not be considered as an exclusive alternative to the isotope-based ICPL approach (Smidt et al., 2005) that will be used to quantify proteins in the rest of our work.

CHAPTER 4 - STUDY OF R. RUBRUM S1H IN SPACE RELATED ENVIRONMENTAL CONDITIONS

CHAPTER 4A - THE MESSAGE 2 EXPERIMENT

4A.1 INTRODUCTION

Numerous in-flight studies have confirmed that space flight can have a profound effect on a variety of microbial parameters including changes in microbial proliferation rate, cell morphology, cell physiology, cell metabolism, genetic transfer among cells, and viral reactivation within the cells (reviewed in Nickerson et al , 2004; Leys et al ., 2004; Nicholson et al ., 2005; Klaus and Howard, 2006). For instance, Klaus et al . (1997) reported an increased final cell concentration of Escherichia coli and Bacillus subtilis grown in liquid culture in space flight conditions. However, E. coli and B. subtilis cultured on solid agar medium in the same hardware showed no increase in final biomass yield (Kacena et al., 1997). Their observations lead to the conclusion that the unique behavior of fluid movement (lack of sedimentation and natural convection) due to the quiescence of low gravity in space flight, as opposed to direct cellular effect of gravity, is responsible for the kinetic changes in flight (Kacena et al ., 1997). Experiments concerning α-proteobacteria indicated that the major morphological and physiological characteristics of two purple non-sulfur phototrophic bacteria, i.e. Rhodopseudomonas palustris H32 and Rhodobacter sphaeroides ZGY, were unchanged after space flight on a satellite for 15 days (Yang et al ., 1999). On the other hand, Magnetospirillum magneticum , a bacterium closely related to R. rubrum in terms of 16S RNA gene similarity, showed impaired magnetotaxis after culturing on board the Space Shuttle and the Space Station Mir (Urban, 2000). In the present study the changes induced by space flight were investigated for R. rubrum S1H. R. rubrum inoculated on rich agar medium was sent to the ISS during the Spanish Soyuz

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 76 mission CERVANTES (MESSAGE 2 experiment, October 2003). After 10 days flight in the ISS, R. rubrum cultures returned to Earth. These cultures were subjected to both transcriptomic and proteomic analysis and compared with the corresponding ground control. Whole genome oligonucleotide microarray and high throughput proteomics, which offer the possibility to survey respectively the global transcriptional and translational response of an organism, were used to test the effect of space flight. Moreover, in an effort to identify a specific stress response of R. rubrum to space flight, ground simulation of space ionizing radiation and space gravity were performed under identical culture setup and growth conditions encountered during the actual space flight experiment. This study is unique in combining the results from an actual space experiment with the corresponding space ionizing radiation and modeled microgravity ground simulations, which lead to a more solid dissection of the different factors acting in space flight conditions.

4A.2 MATERIAL AND METHODS

(i) Strain and media. Sistrom-succinate medium (Sistrom, 1960) was used for liquid cultures and was supplemented with 3% peptone, 1% yeast extract (BD Biosciences, USA) and 2% agar for solid cultures (Sistrom-peptone-yeast) of R. rubrum S1H (Saegesser et al ., 1992). The rich 869 medium [LB Broth Base (Lennox L Broth Base powder, Invitrogen, USA), glucose

(1 g/l) and CaCl 2.6H 2O (0.345 g/l)] supplemented with 2 mM H 2TeO 4 was used for post-flight analysis. (ii) Space flight experiment setup. Three independent cultures of R. rubrum S1H were grown to stationary phase in dark aerobic conditions and were resuspended in 0.85 % NaCl

(Saline tablets, BRS, Oxoid) to a final OD 680 of 0.600 and transported at room temperature from SCK•CEN (Mol, Belgium) to Baikonour (Kazakhstan) 8 days before launch. Inoculation was performed 15 hours before launch on Sistrom-peptone-yeast plates. Drops of 10 6 cells of R. rubrum S1H, as well as dilution series for viable count purpose, were spotted in biological triplicates spread over 2 agar plates (Figure 4a.1). After inoculation, Petri plates were first sealed with parafilm and Kapton tape next, individually packed in Ziplock bag and then jointly packed in a second Ziplock bag and placed in a sealed jar. Radiation dosimeters (Track-Etch Detectors TED's, Optically Stimulated Luminescence Detectors OSLD's and ThermoLuminescent Detectors TLD's, described in Goossens et al ., 2006) and temperature sensors were added in the jar. This jar was then wrapped in protective foam and NOMEX

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 77 fabric bags and stored at room temperature (Figure 4a.1). Finally, the exterior of the package was disinfected with 3 % hydrogen peroxide wipes according to Russian flight procedures and was placed into the Soyuz (TMA-3) 8 hours before launch. Upon arrival in the ISS (after 2 days of flight in the Soyuz), the containers were incubated for 8 days in the Russian Service Module of the ISS behind structural bars (Figure 4.1D), at a relative constant temperature of 21°C ± 2°C. After a total of 10 days flight (October, 18-28 2003) in the ISS, the samples returned back to Earth (Soyuz TMA-2). Prior to return the protective foam was removed to reduce the weight of return cargo and containers were stowed in the download Soyuz vehicle. After landing in the Kazakhstan desert, hardware was handed over to an ESA representative for transport, the jar containing the cultures in Petri dishes was transferred to an active cooling ground transportation box at 4°C and transported by helicopter and airplane to Moscow and Amsterdam (The Netherlands) without exposure to X-rays scanning. The container arrived in SCK•CEN (Mol, Belgium), 40 hours after the return of the Soyuz vehicle. In contrast, ground control samples returned immediately after preparation to SCK•CEN and were cultured in dark conditions under comparable temperature and time profile as the ISS samples. From both ground control and ISS exposed samples, colonies were harvested for both transcriptomic and proteomic analysis. After 10 days flight in the Soyuz and ISS, dosimeters gave a total dose of combined high linear energy transfer (LET) and low-LET particles of ca . 2 mGy (Goossens et al , 2006). (iii) Simulated space ionizing radiation experiment setup. R. rubrum S1H agar cultures prepared according to the space flight experimental setup described above were exposed to ionizing radiations in the calibration room facility at SCK•CEN (Belgium). In order to mimic the ISS ionizing radiation environment, pure 137 Cs gamma and 252 Cf neutron sources were used as representative of low-LET and high-LET particles respectively. The source reference values were traceable to primary standards (PTB for gamma and NPL for neutrons). During irradiation, samples were kept rotating at 2 rpm in between the two sources for 10 days at 21°C ± 2°C, being exposed to a total dose of ca . 2 mGy (Figure 4a.2). Actual absorbed dose was assessed in situ by using TLD's (LiF:Mg,Ti 'TLD-100' from Harshaw Bicron, USA and LiF:Mg,Cu,P 'MCP-N' from TLD-Poland, Poland) and OSLD's (Al2O3:C 'TLD-500' from Harshaw-Bicron) for gamma detection and bubble detectors (BD-PND-BDT from Bubble Technology Industry, Canada) for neutron detection. Non-irradiated control samples were kept outside the irradiation bunker at the same temperature for the same period of time. Cultures were again used for both transcriptomic and proteomic analysis.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 78

d

a c b

Figure 4a.1. MESSAGE 2 hardware's. a . Jar containing Petri dishes. b. Protective foam. c. Transportation bag. d. Container in the Service Module in the ISS.

Sample

a b

Figure 4a.2 . Ionizing radiation ( a) and RPM ( b) experimental setup.

(iv) Simulated microgravity experiment setup. R. rubrum S1H cultures, prepared according to the space flight experimental setup described above, were cultivated in the ESA random positioning machine (RPM) facility located at the University of Sassari (Italy). The RPM was operated as a random walk 3D-clinostat (basic mode) with an angular velocity of rotation of 60 deg/s for 10 days. The RPM was located in a room with an ambient temperature of ca . 22 °C. Cultures were used for both transcriptomic and proteomic analysis.

4A.2.1 TRANSCRIPTOMIC ANALYSIS (i) Microarray analysis. The R. rubrum S1 genome comprising a chromosome (4.35 Mb; GenBank ID: CP000230) and a plasmid (53.7 kb; GenBank ID: CP000231) was used to design 60-mer aminosilane-modified oligonucleotide probes corresponding to the 3,829 candidate protein-encoding genes. The oligonucleotides were synthesized and spotted in triplicate onto glass slides by Eurogentec S. A. (Liège, Belgium), thus providing three technical replicates for each of the 3

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 79 biological samples. Total RNA was extracted using a SV Total RNA Isolation System kit (Promega, USA) and stored at -80 °C. The quality/integrity of total RNA was checked using a BioAnalyzer 2100 (Agilent, USA) and total RNA quantity was measured using the NanoDrop ND-1000 spectrophotometer (NanoDrop technologies, USA). Ten micrograms of RNA was reverse transcribed by random priming according to the Pronto kit instructions (Corning-Promega, USA) and directly labeled by incorporation of Cy3-dCTP or Cy5-dCTP nucleotides (Amersham Bioscience, United Kingdom) for respectively control and experimental samples. The two differentially labeled cDNAs were resuspended in the universal hybridization buffer (Pronto kit; Corning-Promega, USA), mixed, and added jointly to the surface of the spotted microarray glass slide for overnight hybridization at 42 °C using a HS 4800 Pro automated hybridization station (TECAN, Switzerland). The slide was washed according to the Pronto kit protocol and then dried in the hybridization station. The array was scanned with Genepix 4100A a laser scanner (Axon Instruments, The Netherlands) at 532 nm to detect the green fluorescence emitted by Cy3 (control sample) and at 635 nm to detect the red fluorescence emitted by Cy5 (test sample). Spots were analyzed using GenePix Pro v.6.0.1 software and flagged according to their quality. Quality check was performed for each array using the ArrayQuality package from BioConductor (Gentleman et al ., 2004). Raw median intensity data were imported into R version 2.7.0 for statistical analysis using the LIMMA package version 2.14.1 (Smyth, 2005) as available from BioConductor. Raw data were background-corrected based on convolution of normal and exponential distributions with an offset of 50 (Ritchie et al ., 2007). Data were normalized within each array using the loess normalization algorithm (Smyth and Speed, 2003) and between each array using Aquantile normalization (Yang and Thorne, 2003). The in-slide replicate correlations were calculated using the Duplicate correlation function in the LIMMA package (Smyth et al ., 2005). The log expression values were fit to a linear model and moderated t- statistics were calculated using the empirical Bayes method (Smyth, 2004). P-values were corrected for multiple testing using the Benjamini and Hochberg’s method to control the false discovery rate (Benjamini and Hochberg, 1995). Only genes with a p-value below 0.05 and showing a fold change below 0.5 or above 2 fold were kept for further data interpretation. The full description of the array analysis platform and the complete array data have been deposited at the Gene Expression Omnibus website (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE14265. Circular plots were constructed using the Circos library v.0.37.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 80 (ii) Real time quantitative PCR (RT-qPCR) analysis. Reverse transcription of target total RNA from the space and corresponding ground sample (n = 2) was performed using the TaqMan kit in a GeneAmp PCR system 2700 (both from Applied Biosystems, USA) following the manufacturer's instructions. The quality of the resulting cDNA was assessed via PCR reaction of the 16S rRNA and subsequent agarose gel electrophoresis of the amplified cDNA fragments. The primers for RT-qPCR were designed with Primer Express 3.0 (Applied Biosystems, USA) using factory default parameters. All primers were checked for specificity within the genome of R. rubrum S1 (ATCC1170) using the Basic Local Alignment Research Tool of the Joint Genome Institute (URL: http://genome.jgi-psf.org/cgi- bin/runAlignment?db=rhoru&advanced=1 ) with a maximum "expect value" of 1 e -2. Designed primer specificity and efficiency were tested in vitro on R. rubrum S1H genomic DNA sample. 3 primer sets were designed (Table 4a.1).

Table 4a.1 . Sequences of gene specific primers used for RT-qPCR. Target gene Primer sequence Gene product name Fw-rRNA 23S 5'- AGGATAGGCGGGAGGCTTT-3' 23S ribosomal RNA Rv-rRNA 23S 5'-AGACGGAAGGGCGGTATTTC-3' Fw-RruA0894 5'- CGCCCGCAAGCAGAATT-3' stress protein TerD Rv-RruA0894 5'- GCCATCGGCCTGATTGAC-3' Fw-RruA1760 5'-ACCCTCAATTCGTTGATCAAAGA-3' superoxide dismutase Rv-RruA1760 5'-GCCACGGTGGTAATGATATCCT-3'

The primers were purchased from Eurogentec (Belgium). RT-qPCR was performed on the corresponding cDNA using a MESAGREEN qPCR MasterMix Plus for SYBR assay 1 low ROX (Eurogentec, Belgium). Five microlitres of an 1:20 cDNA dilution, 2.5 µl of forward and reverse primers (20 µM), 12.5 µl of MESAGREEN qPCR MasterMix and 5.0 µl of sterile

H2O were mixed in the reaction wells. A non-template control was included during each RT- qPCR experiment to check the purity of the reagents. Each reaction was performed in duplicate using a 7500 Fast Real-Time PCR system (Applied Biosystems, USA) following the manufacturer's instructions. The 23S rRNA was used as a normalization gene while the samples of interest for these reactions were diluted 1,000 times to avoid saturation of the fluorescence detector. The cycle threshold values retrieved from the real-time PCR instrumentation were imported into Microsoft Office Excel XP and the data were analyzed performing the 2 -DeltaDeltaCt calculation described by Livak and Schmittgen (2001).

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 81

4A.2.2 PROTEOMIC ANALYSIS

4A.2.2.1 2D GEL ELETROPHORESIS SEPARATION APPROACH

(i) Protein extraction and quantification . Protein extraction was performed using the buffer 1 (40mM Tris) from the ReadyPrep sequential extraction kit (Bio-Rad, USA). The Protein concentration of the supernatants was measured by the Bradford method (Bradford, 1976), according to the Bio-Rad Protein Assay kit, with bovine gamma-globulin as a protein standard. (ii) First dimension . Each sample was subjected to isoelectric focusing in non-linear Immobiline Dry Strips (18 cm) pH 4–7 (Amersham Pharmacia Biotech, USA). The strips were rehydrated overnight in rehydration solution (2 % w/v CHAPS, 8 M urea, 0.5 % v/v Pharmalyte 3–10, 13 mM dithioerythritol (DTE). Isoelectric focusing was performed on a Pharmacia Biotech Multiphor II system equipped with a Pharmacia Biotech EPS3500 XL power supply using a 3-phase program: 500 V for 1 min, 3,500 V for 1.5 h and 3,500 V for 14.5 h. After isoelectric focusing, the gels were equilibrated twice, 20 min each, in equilibration solution (6 M urea, 30 % v/v glycerol, 2 % w/v SDS, 50 mM Tris- HCl pH 6.8) containing 65 mM DTE for the reduction step, and containing 135 mM iodoacetamide for the alkylation step. (iii) Second dimension. Using the SE 600 device (Hoefer, USA), the strips were then placed on top of 12 % Poly-Acrylamide Gel Electrophoresis gel in a 0.4 % w/v agarose gel solution made with SDS-PAGE running buffer TGS (192 mM Glycine, 0.1 % w/v SDS, 25 mM Tris- HCl, pH 8.3). The second dimension was run for approximately 4 h at 250 V, 20 mA/gel at first and then 30 mA/gel. Gels were then silver stained using the protocol from Mortz et al . (2001). (iv) Protein identification . Protein patterns within the gels were analyzed as digitalized images using a high-resolution scanner (ScanMaker 9800 XL, Microtek) in combination with the analysis software PDQuest (Bio-Rad, USA). Spots of interest on the gel were excised using a 1 mm sample corer (Fine Science Tools Inc.), destained, washed and then digested with trypsin overnight at 37 °C. Matrix Assisted Laser Desorption Ionisation (MALDI) – Time of Flight (ToF) Mass Spectrometry (MS) analysis was performed using a Micromass M@ldi_spectrometer (Manchester, UK) equipped with a 337 nm nitrogen laser. The instrument was operated in the positive reflectron mode at 20 kV accelerating voltage with time-lag focusing. The resulting peptide masses were automatically searched for in a local

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 82 copy of the SWISS-PROT (Boeckmann et al ., 2003), R. rubrum ATCC11170 database using the ProteinLynx global server and the Protein Probe search engine (Micromass). An initial mass tolerance of 50 ppm was used in all searches. Peptide modifications allowed during the search were carbamidomethylation of cysteine and oxidation of methionine. The maximum number of missed cleavages was set to 1.

4A.2.2.2 MUDPIT APPROACH USING ISOTOPE -CODED PROTEIN LABEL (ICPL)

(i) Protein extraction and quantification . Bacterial cells were collected by centrifugation at 7,000 rpm for 10 minutes. Protein samples were obtained by high power sonication (U50 control, IKA labortechnik, Germany) of the bacterial pellet suspended in 6 M (v/v) guanidine chloride. Sonication was realised by 3 cycles of 10 seconds (40 % amplitude, cycle 1) followed by 1 min. cooling on ice. Samples were cleared by centrifugation at 13,500 rpm for 15 min at 4 °C. Protein concentration of the supernatants was measured by the Bradford method (Bradford, 1976), according to the Bio-Rad Protein Assay kit, with bovine g-globulin as a protein standard. Protein concentration of each sample was adjusted to 5 µg/µL as required for labeling. Before labeling, the three biological replicates were pooled. (ii) ICPL labeling . ICPL labeling of proteins using the ICPL-kit (SERVA Electrophoresis, Heildberg, Germany) was performed as described previously (Schmidt et al ., 2005). Briefly, two protein mixtures of 100 µg each, obtained from two different culture conditions were first individually reduced and alkylated to denature proteins and to ensure easier access to the free amino acid groups that are subsequently derivatised with deuterium free (light) or deuterium containing (heavy) form, respectively of the ICPL reagent. The heavy (H) form of the ICPL reagent was always used to label the experimental sample while the light (L) form was always used to label the control sample. Labelled samples were then combined in a H/L ratio of 1:1. After subsequent overnight digestions by trypsin (protein/enzyme ratio of 1/50 and incubation at 37 °C) and by endoproteinase Glu-C (protein/enzyme ratio of 1/30 and incubation at 25 °C), peptides were desalted using HyperSep TM SpinTip C18 (Thermo-Fischer electron, USA) and analyzed by Multi- Dimensional Protein Identification Technology (MuDPIT). Since differentially labeled protein samples derived from identical peptides differ in mass, they appear as doublets in the acquired mass spectra. From the ratio of the ion intensity of these sister peptides pairs, the relative abundance of their parent protein in the original samples was determined. An H/L ratio larger than 1 or lower than 1 indicates respectively over-expression or down-expression of the protein of interest in the experimental sample.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 83 (iii) Chromatographic separation . MudPIT, a non-gel approach for the identification of proteins from complex mixtures, consists of a 2-dimensional chromatography separation, prior to electrospray mass spectrometry. The peptides obtained by enzymatic digestion were separated using an Ultimate 3000 chromatographic system (Dionex, USA). The chromatographic system is composed of two columns connected in series, a strong cation exchange (SCX) column (POROS 10S, 10 cm) and a reverse phase column (C18, 15 cm, ID 75 m) both from Dionex. Peptides are injected in the SCX column and the flow through is directly loaded on a precolumn (C18 Trap, Dionex, USA). The 18 µg peptides sample was washed on the precolumn with the loading solvent (5 % acetonitril, 0.025 % trifluoroacetic acid) during 15 min. After washing, an acetonitril (ACN) gradient was applied to separate the peptides on the reverse column based on their hydrophobicity. The ACN gradient was 4 to 37 % of solvent B (solvent B: 80 % ACN, 0.08 % formic acid) in 100 min, 37 to 57 % of solvent B in 10 min and 57 to 90 % of solvent B in 10 min 90 % of solvent B is maintained for 10 min and then reset to 4 %. Column equilibration at 4 % of solvent B is allowed for 10 min . After this first cycle, a first lot of peptides were eluted of the SCX column by injection in this column of 20 l of loading solvent containing 1 mM of NaCl. The eluted peptides were loaded on the precolumn and analyzed as described for the flow through. This sequence of elution from the SCX column with injection of a salt plug followed by separation on the reverse phase column is repeated 7 times with salt plug concentration of 5, 10, 25, 50, 200, 500 and 1000 mM NaCl. (iv) Mass spectrometry analysis . Peptides separated in the reverse phase column were analyzed on line using an ion-trap mass spectrometer (HCT ultra PTM discovery, Bruker Daltonics, Germany). This mass spectrometer was used in the positive mode and spectra were acquired in a data dependent manner: MS scan range = 300-1,500 m/z, maximum accumulation time = 200 ms, ICC target = 200,000. In order to focus on the differentially expressed proteins, the following H/L ratio detection parameter was set up for the stable isotope labeling experiment (SILE): the H/L ratio must be smaller than 0.83 or larger than 1.2. The top 3 most intense ions in the MS scan were selected for MS/MS in dynamic exclusion mode: mode = ultrascan, absolute threshold = 75,000, relative threshold = 1 %, excluded after spectrum count = 1, exclusion duration = 0.3 min, averaged spectra = 5, ICC target = 200,000. Since the spectra were acquired in a data dependant manner and not based on specific ion selection, each ICPL sample has been run twice to increase the peptide detection number and the protein identification and quantification accuracy.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 84

m/z m/z m/z 871.60 871.60 870.38

870 870 870 869.84 869.84 +MS, 1212.8min #34873 +MS, 1213.4min1213.4min #34895 #34895 +MS, 1213.1min #34884

868.40 868.40 868.35 868 868 868 867.92 866.47 866 866 866 865.92 865.92 865.46 865.42

864.90 864.41 864.41 864 864 864 0.5 0.95 863.79 2.4 863.48 863.48 863.05 862 862 862 862.13 862.10 862.10 861.17 860.93 860.93 860.90 860 860 860 nd the end (bottom) of its elution time from the 859.92 end of this period thanks to the SILE detection 859.47 859.44 859.42 859.42 858.59 858.59 858.41 e averaged value of the H/L ratio was actually 1.05. 858 858 858 857.11 856.78 856.78 856 856 856 856.03 855.17 855.17 854 854 854 853.98 5 5 5 3 2 1 0 8 6 4 2 0 2.0 1.5 1.0 0.5 0.0 x10 x10 x10 x10 Intens. Intens. Intens.

m/z

870

peptide at the beginning (top), the middle (middle) a 868 +MS, 1212.8-1213.7min #(34873-34902) 866.92 sequence 866.42 (red diamond) for MS/MS only at the beginning and at the 866

d over the 3 periods of time mentioned above, showed th 865.44 GYLDK*LK*TTDVVR

864.94 864

1.05 97

862 Signal/noise 861.40

860 859.43 108

score Peptide

858

. the On right: representation of a RruA1224 related 856 Protein ID RruA1224 854 Figure 4a.3 reverse phase column. The peptide has been selected parameters. theOn left: the spectra, manually summe 5 5 4 3 2 1 0 x10 Intens.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 85 (v) Data extraction and database search . Data acquired from both runs were combined in a single file using AnalysisCombiner (Bruker Daltonics, Germany) and data were extracted using DataAnalysis 3.4 (Bruker Daltonics, Germany) generating a peak list in Mascot Generic File format. The peak list was searched against a local copy of the NCBInr database released in June 2008 (taxonomy = R. rubrum S1) using an in-house Mascot 2.2 server (Matrix Science) for protein identification (p < 0.05). Protein quantification was completed using WARP-LC (Bruker Daltonics) that detects isotopic pairs resulting from differentially labeled proteins and calculates intensity ratio of each peptide pair based on peak intensity. Protein identification was achieved using the Mascot search engine including the following parameters: database = NCBInr, organism = Rhodospirillum rubrum ATCC11170, fixed modification = carbamidomethyl cystein, variable modification = oxidation of methionine, ICPL light and heavy (on lysine-K- and N-terminal), mass tolerance in MS = 1.3 Da, mass tolerance in MS/MS = 0.5 Da. The false discovery rate (FDR) was estimated using the Mascot 'decoy' option. Only proteins identified with a Mascot score above 50 and quantified by at least 2 different peptides or by the same peptide detected at different times during the analysis were considered. In any case, proteins quantified with a mean calculated peptide ratio showing a standard deviation higher than 20 % were manually checked. Figure 4a.3 illustrated the importance of manual checking.

4A.3 RESULTS

4A.3.1 TRANSCRIPTOMIC ANALYSIS Using the whole genome DNA chip, respectively 4.94 % and 0.78 % of the genes (out of 3826 genes retained) were identified as significantly up- and down-regulated after a 10 days space flight compared to the ground control. Concerning the ionizing radiation simulation, 0.13 % and 0.10 % of the genes (out of 3,794 genes retained) were identified as respectively significantly up- and down-regulated after a 10 days ionizing radiation experiment compared to non-irradiated samples. Eventually for the modeled microgravity experiment, 3.88 % and 0.26 % of the genes (out of the 3,814 genes retained) were identified as respectively significantly up- and down-regulated after 10 days culturing using the random positioning machine compared to culturing in normal gravity. A circular plot of the genes differentially expressed in the chromosome of R. rubrum S1H showed that these genes were distributed globally throughout the chromosome for all culture conditions (Figure 4a.4).

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 86 When sorting the differentially expressed genes according to their functional category, the 'unclassified' (Un.) class was the most numerically abundant in the ionizing radiation and RPM experiments and the second most in the MESSAGE 2 experiment (Figure 4a.5ab). Moreover, of the 373 non-redundant significant genes, 30 % coded for hypothetical proteins (Additional table 4a.1).

(i) (ii)

(iii) (iv) (v) (vi)

Rhodospirillum

rubrum S1 Chromosome 4.35 Mb

Figure 4a.4 . Map of the 4.35-Mb circular R. rubrum S1 chromosome. From the outside to the inside: (i) genes on direct strand, (ii) genes on reverse strand, genes up-regulated above 2 fold (in red color) and down-regulated below 0.5 fold ( green color) in (iii) the MESSAGE 2 experiment, (iv) the simulation of ISS ionizing radiation and (v) the simulation of microgravity by cultivation on RPM. (vi) gene position (x 10,000 bp).

4A.3.1.1 THE MESSAGE 2 EXPERIMENT

One clear observation was the high number of up-regulated genes related to the functional category: 'translation, ribosomal and biogenesis' (J) (Figure 4a.5a). Indeed, 29 ribosomal protein encoding genes as well as 1 translation initiation (RruA3781) and 2 translation

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 87 elongation (RruA2690 and RruA2691) related genes were overexpressed only in the MESSAGE 2 flight experiment (Additional table 4a.1).

40

35

30

25 (a) MSG2 UP 20 simRAD UP RPM

Numberof genes 15

10

5

0 EGDMNBHVLCSRPU I FOQTKJUn. Functional category (COG)

40

35

30 (b) 25 MSG2 DOWN 20 simRAD DOWN RPM DOWN Number of genes of Number 15

10

5

0 EGDMNBHVLCSRPU I FOQTKJUn. Functional category (COG)

Figure 4a.5 . Functional classification of the significant genes up-regulated ( a) and down-regulated ( b) during the MESSAGE 2 experiment, the simulation of ISS ionizing radiation (simRAD) and the modeled microgravity experiment (RPM). COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 88 Within the 'energy production and conversion' (C) class, an operon coding for a ubiquinone oxidoreductase nuoABCDEFG (RruA1555 to RruA1561) was specifically up-regulated in space flight conditions with nuoA , nuoD and nuoE showing a fold changes between 1.86 and 1.96 (p-value between 1.17x10 -5 and 6.06x10 -9). The operon sdhCDAB (RruA1202 to RruA1205) coding for a succinate dehydrogenase was also up-regulated in the space samples with sdhA and sdhB being over-expressed 1.73 fold (p-value: 7.14x10 -6) and 1.92 fold (p- -8 value: 6.68x10 ) respectively. In addition, three F 0F1 ATP synthase subunits (RruA1223, RruA1225 and RruA3246) were over-expressed only in the space samples. Related to transcription (class K), an alternative sigma factor of extracytoplasmic function (ECF) (RruA3287) and six transcriptional regulators were specifically up-regulated in the space samples (Additional table 4a.1). Finally, we also found a flagellar protein encoding gene cluster RruA2532 to RruA2535 from the 'cell motility and secretion' (N) class up-regulated in the space samples. Over-expressed genes from the 'inorganic ion transport and metabolism' (P) class contained genes related to the redox balance of the cell like a ferredoxin (RruA0077), a Fe-S cluster- related gene (RruA1069), a ferritin and dps related gene (RruA1499), a superoxide dismutase (RruA1760), and a bacterioferritin RruA2195 ( bfr ) (Additional table 4a.1). The same trend in gene expression could be confirmed by real time PCR for sodA with a 2.30 ± 0.60 fold induction. In the same way, tellurium resistance related to terB (RruA0891) was specifically up-regulated in the space samples. It was jointly over-expressed with RruA0894 encoding for TerD, and for which expression was confirmed by real time PCR (1.65 ± 0.21). In addition, RruA0890 ( terC ), RruA0892 and RruA0893 ( terZ ) showed respectively a fold change of 1.45 (p-value: 8.69x10 -4), 1.83 (p-value: 1.44x10 -6) and 1.77 (p-value: 9.72x10 -8). The genomic basis of R. rubrum tellurium resistance has not yet been described but it apparently shares common feature with the operon terZABCDEF described for the plasmid determinant R478 (IncHl2) (Taylor, 1999) except that R. rubrum tellurium resistance genes are carried by the chromosome. This enhanced tellurium resistance for the space samples was confirmed at the phenotypic level by plating space and control colonies on 869 solid medium supplemented with 2 mM tellurate and subsequent cultivation in light anaerobic conditions. The space cultures showed full growth on 869 plus tellurate which was not the case for the corresponding ground samples (Figure 4a.6).

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 89

(a) (b)

Figure 4a.6 . R. rubrum on solid agar 869 + 2mM tellurate and light anaerobic culture condition for MESSAGE 2 ground (a) and space (b) samples.

Within the 'posttranslational modification, protein turnover, chaperones' (O) class, up- regulated (RruA1550-clpP - and RruA2203-hflK-) as well as down-regulated genes (RruA0586-groES -, RruA0587-groEL -, RruA0753-clpB - and RruA3555-dnaK -) were found to be specific for the space samples (Additional table 4a.1). For the 'Cell envelope biogenesis, outer membrane' (M) class, space specific up-regulation was found for: an ompA -like gene (RruA0269), murE (RruA0955), wcaA (RruA1005), a peptidase (RruA1232), a porin (RruA2211), mscS (RruA3072) and ompA (RruA3328). Within the 'Carbohydrate transport and metabolism' (G) class, a major facilitator transporter (RruA1114) and the glycogen metabolism related RruA2246 and RruA2576 were specifically up-regulated together while hppA (RruA1818) involved in UDP-glucose conversion (class C). In addition, RruA2485 coding for a key enzyme in the biosynthesis of the osmo- (and thermo) protectant trehalose was up-regulated together with a trehalose synthase related gene cluster (RruA1604 and RruA1605) (Additional table 4a.1). Related to osmolarity, the Kdp K +-uptake system including the 3 ATPase subunits kdpABC , the sensor kinase kdpD and the cytosolic response regulator kdpE (RruA1155 to RruA1159) were slightly down-regulated in the space samples. Within the latter gene cluster, the fold change (and p-value) were respectively 0.79 (2.38x10 -2), 0.84 (2.63x10 -1), 0.54 (1.99x10 -5), 0.79 (4.57x10 -2) and 0.73 (1.03x10 -2). While RruA3670 encoding for a potassium efflux protein showed an increased expression only in the space samples (fold change of 1.78 with a p-value of 1.97x10 -4). As mentioned above, genes from the 'unclassified' (Un.) and the 'function unknown' (S) class were very abundant in the differentially expressed genes. The former category comprised 28 genes coding for hypothetical proteins that were differentially expressed only in the space samples like the gene cluster RruA0422, RruA0423 and RruA0424. On the other hand, the S class contained 10 space specific up-regulated hypothetical protein encoding genes as well as the chaperone CsbD (RruA1175) and a phasin (RruA3283). Noteworthy, the top 3 of the most induced genes were all hypothetical protein encoding genes namely RruA3369 (13.99 fold), RruA2713 (8.48 fold) and RruA0160 (7.50 fold) (Additional table 4a.1).

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 90

4A.3.1.2 THE SIMULATION OF ISS IONIZING RADIATION USING THE MESSAGE 2 SETUP The 2 mGy irradiation experiment elicited a moderate response with 5 up- and 4 down- regulated genes, for which 2 of the down-regulated genes were also down-regulated either in space or during microgravity simulation. Of these 9 differentially expressed genes, 6 coded for a hypothetical protein (Table 4a.2).

Table 4a.2 . List of significant genes specific to the simulation of ISS ionizing radiation ( simRAD ) compared to the MESSAGE 2 ( MSG2 ) and the modeled microgravity ( RPM ) experiment. Gene Gene FC p-value FC p-value FC p-value Product name COG number name MSG2 MSG2 simRAD simRAD RPM RPM RruA0197 - hypothetical protein Rru_A0197 Un. 0.59 8.11E-07 0.46 1.42E-10 0.49 3.86E-09 RruA0198 - hypothetical protein Rru_A0198 Un. 0.68 4.30E-03 0.50 1.42E-06 0.85 2.54E-01 RruA1244 - hypothetical protein Rru_A1244 Un. 0.48 3.02E-03 0.21 2.51E-08 1.08 7.73E-01 Sec-independent protein translocase RruA1771 - U 1.05 8.06E-01 2.20 2.03E-05 1.04 8.66E-01 TatC response regulator receiver domain- RruA2367 ompR K 0.94 6.62E-01 2.24 1.14E-08 0.86 2.25E-01 containing protein RruA2933 - hypothetical protein Rru_A2933 Un. 0.71 6.79E-03 0.47 1.03E-07 1.43 5.17E-03 RruA2994 - hypothetical protein Rru_A2994 D 1.24 3.03E-01 2.57 9.39E-06 0.88 5.81E-01 RruA3636 - hypothetical protein Rru_A3636 Un. 0.85 6.40E-02 2.22 2.30E-11 0.94 5.67E-01 RruA3719 - methyltransferase FkbM H 1.03 8.20E-01 2.26 1.42E-10 1.04 7.71E-01 FC : fold change. Red highlight : up-regulated genes. Green highlight : down-regulated genes. COG nomenclature: D - Cell division and chromosome partitioning, H - Coenzyme metabolism, U - Intracellular trafficking, secretion, and vesicular transport, Un . – Unclassified.

4A.3.1.3 THE MODELED MICROGRAVITY (RPM) EXPERIMENT USING THE MESSAGE 2

SETUP

158 significantly differentially expressed genes were found during the microgravity simulation, for which 37 % appeared to code for hypothetical proteins (Additional table 4a.1). Therefore, not surprisingly genes from the functional categories: 'unclassified' (Un.) and the 'function unknown' (S) were the most numerically abundant (Figure 4a.5). Among the 'general function prediction (R) category, we could mention the phage shock related pspB (RruA1217) and the dnaJ -like RruA3559. In addition, up-regulation an alternative sigma factor of ECF (RruA2397), a grpE chaperone (RruA3643) and the fur gene (RruA3767) were found only in the RPM samples. In the 'Cell envelope biogenesis, outer membrane' (M), the specific up-regulation of two glycosyl related genes (RruA0938 and RruA2938) as well as the plasmid-borne gene RruB0046 involved in lipopolysaccharide biosynthesis could be observed. Three genes from the 'Defense mechanisms' class were also up-regulated only in the RPM sample namely RruA1118, RruA2806 and RruB0040 coding respectively for a efflux pump, an hypothetical protein and for a secretion protein (Additional table 4a.1).

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 91

4A.3.1.4 OVERLAPPING GENES BETWEEN THE THREE CULTURE CONDITIONS TESTED

No gene at all was common to the three culture conditions (Figure 4a.7). One single hypothetical gene (RruA1244) was shared between MESSAGE 2 and the ionizing radiation experiment, while another hypothetical gene (RruA0197) was common to the irradiation and the RPM experiment. Within the 11 significant genes jointly expressed in the MESSAGE 2 and RPM experiments, 4 coded for hypothetical proteins (Table 4a.3), with RruA3369 the most induced gene during the MESSAGE 2 experiment (13.99 fold) as mentioned above.

MSG2 simRAD

207 1 7

0 11 1

146

RPM

Figure 4a.7 . Venn diagrams showing the relation between the 373 significant genes categorized from the MESSAGE 2 ( MSG2 ), the simulation of ISS ionizing radiation ( simRAD ) and the modeled microgravity experiments ( RPM ).

Table 4a.3 . List of the 13 significant genes shared by at least 2 cultures conditions within the MESSAGE 2 (MSG2 ), the simulation of ISS ionizing radiation ( simRAD ) and the modeled microgravity ( RPM ) experiments. Gene Gene FC p-value FC p.value FC p.value. Product name COG number name MSG2 MSG2 simRAD simRAD RPM RPM RruA0197 - hypothetical protein Rru_A0197 Un. 0.59 8.11E-07 0.46 1.42E-10 0.49 3.86E-09 RruA0637 - hypothetical protein Rru_A0637 S 3.18 1.06E-09 0.81 1.61E-01 3.51 5.88E-10 RruA1244 - hypothetical protein Rru_A1244 Un. 0.48 3.02E-03 0.21 2.51E-08 1.08 7.73E-01 RruA1537 - hypothetical protein Rru_A1537 Un. 2.70 5.43E-08 0.87 3.92E-01 2.88 3.01E-08 RruA2091 - hypothetical protein Rru_A2091 Un. 3.97 5.54E-12 0.95 7.33E-01 2.75 1.61E-08 flagellar hook-associated protein 2 (FliD, RruA2535 - N 2.81 1.42E-10 0.93 6.10E-01 2.61 2.56E-09 filament cap protein) RruA2666 rpsM 30S ribosomal protein S13 J 3.08 4.43E-12 0.89 3.35E-01 2.35 7.93E-09 RruA2721 porA 2-oxoglutarate synthase, alpha subunit C 2.50 1.89E-07 1.30 7.87E-02 2.24 2.71E-06 RruA3205 rpmG 50S ribosomal protein L33P J 3.08 4.45E-11 0.57 2.58E-05 2.58 6.27E-09 RruA3369 - hypothetical protein Rru_A3369 Un. 13.99 1.00E-17 0.82 2.65E-01 2.35 7.87E-06 RruA3681 - YceI S 2.27 1.00E-08 1.30 2.14E-02 2.31 1.61E-08 RruA3746 rplY 50S ribosomal protein L25P J 2.77 3.31E-10 1.00 9.84E-01 2.05 1.03E-06 RruA3775 - XRE family transcriptional regulator K 2.08 2.35E-06 1.06 6.98E-01 2.23 7.97E-07 Red highlight : up-regulated genes. Green highlight : down-regulated genes. FC : fold change. COG nomenclature: N - Cell motility and secretion, C - Energy production and conversion, S - Function unknown, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 92

4A.3.2 PROTEOMIC ANALYSIS

4A.3.2.1 THE MESSAGE 2 EXPERIMENT

2 dimensional (D) protein maps were made for the 3 independent cultures from MESSAGE 2 and the corresponding ground controls (Figure 4a.8). These 2D-maps indicated several changes in protein expression when comparing space vs ground samples: both up-regulation (13 spots) and down-regulation (8 spots) were detected in the space samples compared to the ground control samples. Unfortunately, it was not possible to identify the spots of interest due the low amount of starting material (about 40µg of proteins).

4A.3.2.2 THE SIMULATION OF ISS IONIZING RADIATION EXPERIMENT USING THE MESSAGE

2 SETUP

For the simulation of ISS ionizing radiation experiment, the computed FDR was 0.34 % estimated at the peptide level. From these peptides, 341 proteins were identified including 192 quantified with at least 2 peptides (Additional table 4a.2). This represents respectively 6.9 % and 5.0 % of the total candidate protein-encoding genes. The median value for the protein fold change distribution plot was 1.07 indicating a correct mixing of the control and the experimental samples during the differential proteomic protocol (Figure 4a.10). Considering the threshold mentioned in the material and methods part, only 3 proteins showed a significant up-regulation (Table 4a.4). However, we could underlign the slight up-regulation of the osmoprotectant glycine betaine transporter periplasmic subunit (ProX) (Table 4a.4 and figure 4a.9).

4a.3.2.2.1 Comparison with the corresponding transcriptomic data None of the 9 differentially expressed genes from the simulation of the ISS ionizing radiation using the MESSAGE 2 setup were identified or quantified using the differential proteomic approach.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 93

4 pI 7 Space

MW

Figure 4a.8 . 2D silver stained proteomic gel electrophoresis (non linear pI 4-7; PAGE 12 %) on total protein extract using Buffer 1 (Bio-Rad) from R. rubrum S1H grown on Sistrom-peptone-Yeast plates in DAE culture conditions in Space or on Ground. pI : isoelectric point. MW: molecular weight. Solid line : space Ground up-regulated proteins; Dashed line : space down- regulated protein

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 94

Figure 4a.4 . Significant differentially expressed proteins from the simulation of ISS ionizing radiation using the MESSAGE 2 setup. C Gene FC Mascot MW Gene name Product name O # (H/L) number simRAD Score (kDa) G RruA0930 eutG Iron-containing alcohol dehydrogenase C 2.08 ± 0.02 2 85.92 39.91 RruA0931 putA aldehyde dehydrogenase C 1.54 ± 0.07 8 534.37 55.73 glycine betaine transporter periplasmic RruA2477 proX E 1.58 ± 0.10 2 162.49 37.23 subunit 'RruA' refers to gene located on the chromosome. FC : fold change. MW: molecular weight. COG nomenclature: E - Amino acid transport and metabolism, C - Energy production and conversion.

Intens. K*TVTHQQGSYSALMADTITR +MS, 2143.9-2144.8min #(63305-63333) x10 4 5 779.1

4 1.72 3 777.1 2 795.4 1 786.4 766.7 770.9 773.1 774.8 784.2 790.8 793.7 0 765 770 775 780 785 790 795 m/z

Intens. IDQLK*DPK*IAALFDHDGDGK*ADLAGCTPGWGCE +MS, 2998.6-2999.0min #(88890-88903) x10 4 1302.5 6 1.44 1296.5 4 1307.5 1309.8 2 1286.6 1291.7 1294.5 0 1285.0 1287.5 1290.0 1292.5 1295.0 1297.5 1300.0 1302.5 1305.0 1307.5 1310.0 m/z

Figure 4a.9 . Spectra related to the two peptides ( sequences shown in bold ) used to quantify the glycine betaine transporter protein (RruA2477) using the ICPL approach. Heavy/Light forms ratio is shown in red . K* : ICPL labeled lysine. m/z : mass to charge ratio.

4A.3.2.3 THE MODELED MICROGRAVITY EXPERIMENT USING THE MESSAGE 2 SETUP

In the modeled microgravity experiment, the computed FDR was 0.59 % estimated at the peptide level. 406 proteins were identified including 224 quantified with at least 2 peptides (Additional table 4a.4). This constituted respectively 10.6 % and 5.9 % of the total candidate protein-encoding genes.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 95

2.5

2

1.5

Foldchange 1 0 20 40 60 80 100 120 140 160 180 200

0.5

ratio simRAD/control 0 Protein number

Figure 4a.10 . Fold change distribution of the protein quantified by the differential proteomic approach for the simulation of ISS ionizing radiation using the MESSAGE 2 setup ( simRAD ).

2.5

2

1.5

Foldchange 1 0 50 100 150 200 250

0.5

ratio RPM/control 0 Protein number

Figure 4a.11 Fold change distribution of the protein quantified by the differential proteomic approach for the modeled microgravity ( RPM ) experiment using the MESSAGE 2 setup.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 96 The protein fold change distribution plot showed that the fold change was centered again on 1 (median value = 1.04) indicating a correct mixing of the control and the experiment samples during the differential proteomic protocol (Figure 4a.11). Five of the quantified proteins were found to have a fold change above the fixed threshold (Table 4a.5).

Table 4a.5 . Significant proteins from the modeled microgravity ( RPM ) experiment using the MESSAGE 2 setup. Gene Gene Mascot MW Product name COG FC RPM # (H/L) number name Score (kDa) RruA0738 grxC glutaredoxin GrxC O 1.55 ± 0.13 5 271.50 9.63 RruA0893 terZ stress protein T 1.57 ± 0.05 19 1185.06 20.46 RruA0894 terD stress protein T 1.56 ± 0.08 6 851.47 20.29 RruA2083 creA CreA S 1.60 ± 0.11 5 189.51 17.12 RruA2161 ppk polyphosphate kinase P 2.04 ± 0.25 3 113.83 84.55 'RruA' refers to gene located on the chromosome. FC : fold change. Mean ± standard deviation. MW: molecular weight. COG nomenclature: S - Function unknown, P - Inorganic ion transport and metabolism, O - Posttranslational modification, protein turnover, T - Signal transduction mechanisms.

4a.3.2.3.1 Comparison with the corresponding transcriptomic data Of the 158 significant differentially expressed genes in modeled microgravity, 16 were confirmed to be present as protein in the same sample (Table 4a.6). However, none of the quantified proteins passed the threshold limit set for significance in proteomics.

Table 4a.6 . Transcriptomics versus differential proteomics approach for the RPM experiment. RPM Gene RPM Gene number Product name COG FC name FC protein mRNA RruA0414 rpsF SSU ribosomal protein S6P J 2.22 1.29 ± 0.13 RruA1072 rpmE 50S ribosomal protein L31 J 2.68 0.95 ± 0.05 RruA1527 uup putative ABC transporter ATP-binding protein R 2.89 IBNQ RruA1537 - hypothetical protein Rru_A1537 Un. 2.88 IBNQ RruA1799 pcm protein-L-isoaspartate(D-aspartate) O-methyltransferase O 2.38 0.79 ± 0.04 RruA2302 rbsB periplasmic binding protein/LacI transcriptional regulator G 2.23 0.87 ± 0.15 RruA2666 rpsM 30S ribosomal protein S13 J 2.35 1.27 ± 0.17 RruA2721 porA 2-oxoglutarate synthase, alpha subunit C 2.24 IBNQ RruA3205 rpmG 50S ribosomal protein L33P J 2.58 0.83 ± 0.07 RruA3249 dsbG DSBA oxidoreductase O 2.33 IBNQ RruA3282 - hypothetical protein Rru_A3282 S 2.01 0.9 ± 0.02 RruA3643 grpE GrpE protein O 2.74 1.34 ± 0.05 RruA3700 acrA secretion protein HlyD M 2.08 IBNQ RruA3744 - signal transduction protein T 3.00 IBNQ RruA3746 rplY 50S ribosomal protein L25P J 2.05 IBNQ RruB0025 - hypothetical protein Rru_B0025 Un. 2.78 IBNQ 'RruA' and 'RruB' refer to gene located respectively on the chromosome and on the plasmid. FC : fold change (p-value < 0.05). IBNQ : identified protein but not quantified by at least two H/L occurrences (see material and methods). Proteins are quantified by mean ± standard deviation. COG nomenclature: G - Carbohydrate transport and metabolism, M - Cell envelope biogenesis, outer membrane, C - Energy production and conversion, S - Function unknown, R - General function prediction only, O - Posttranslational modification, protein turnover, T - Signal transduction mechanisms, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 97

4A.4 DISCUSSION

Genes coding for hypothetical proteins . This group appeared to constitute a non-negligible part (30 %) of the differentially expressed genes detected in the 3 studied culture conditions and even occupied the top 3 up-regulated genes in the space flight experiment (RruA3369, RruA2713 and RruA0160). The amount of genes annotated as hypothetical proteins ( ca . 25 %) in the genome of R. rubrum indicates that a substantial part of the genetics of this organism is not well-known; therefore exploring new environmental and stress conditions could assign novel functions to these hypothetical proteins. In addition, considering the increasing number of sequenced genomes, genes still coding for (conserved) hypothetical protein could be related to functions virtually specific to the novel environmental stressor of interest. More genes coding for hypothetical proteins linked to the MESSAGE 2 related experiments will be discussed in chapter 4c. Possible osmotic stress. The transcriptomic results suggest a response to what is felt by the cell as a possible osmotic stress. Together with the genes coding for a key enzyme in the biosynthesis of trehalose, two membrane-based osmosensors, the mechanosensitive channel MscS and, to a lesser extend, the two-component sensor kinase KdpD were differentially expressed in MESSAGE 2 space samples (Wood, 1999; Ballal et al ., 2007; Hurst et al ., 2008; Gunasekera et al ., 2008). The transcriptional regulator KdpE represses the transcription of the kdpABC operon coding for a K +-dependent ATPase uptake pump. This system has been shown to be functional in R. rubrum S1 and produces a 70-75 kDa polypeptide, which showed immunological cross- reactivity to E. coli KdpB antiserum, when grown in conditions of limiting K + (Abee et al ., 1992). The complete stimulus that is sensed by KdpD still remains unclear. KdpD probably integrates various parameters like: changes in turgor, in phospholipid composition of the membrane, in extra and/or intracellular K + concentration, in medium osmolarity, in ionic strength of the cytoplasm and in internal ATP concentration (Zimmann et al ., 2007). Our data may support a role of KdpD in the response to the change in cell turgor. Indeed, we reported the induction in the space samples of mscS encoding a mechanosensitive channel that has a major role as emergency safety valve during hypoosmotic shock, i.e. sudden reduction of the external osmolarity (reviewed in Booth et al ., 2007). Levina et al . (1999) determined that the degree of hypoosmotic shock required to activate the E. coli mechanosensitive channel was a change of 150 mM NaCl ( ca . 2 atm), which is just below the lysis pressure (by bursting) of E. coli mutants lacking the MscS (and MscL) channel activities. Thus these data indicate that

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 98 usually mechanosensitive channels are not activated by small changes in osmolarity because they disrupt ionic homeostasis of the cell. A disruption of the cellular homeostasis may elicit the observed increase in expression for the 'energy production and conversion' class of proteins, in order to restore the homeostasis. Redox balance of the cell and oxidative stress . The release of energy from growth substrates, and the capture of this energy for the biosynthesis of cell components and the processes that they support, involves redox reactions in which electrons and/or hydrogen atoms are transferred between donor and acceptor molecules. Bacterial metabolism and cellular integrity are maintained by balancing the redox state of all the cellular components for optimal overall function. However, disturbance of this balance, notably an increase in oxidation reactions (oxidative stress), can damage essential cellular components. Obviously because water molecules constitute the major component of an active organism, it will be the primary target of ionizing radiation leading to the production of reactive oxygen species (ROS) which have the ability, either directly or indirectly, to damage all biomolecules, including proteins, lipids, DNA and carbohydrates. Moreover, the generation of ROS can be amplified by the presence of transition metal ions, such as Fe 2+ via the Fenton reaction (Green and Paget, 2004). The observed increase in the 'energy production and conversion' class of proteins together with the prevailing ionizing radiation may contribute to the upregulation of several genes related to the redox balance in space. However, simulation of ionizing radiation alone did not trigger this response. Therefore, in space, the effect of ionizing radiation was enhanced compared to that observed on ground facilities with single beams of ionizing radiation. Rea et al . (2008) recently underlined the difficulty to measure and above all, to reproduce different radiation components at the same time over a wide energy range. Besides, it has been suggested that tellurium toxicity stems from its strong oxidizing ability, which may interfere with many cellular enzyme processes (Summers and Jacoby, 1977). Therefore up-regulation of tellurium resistance genes in space flight samples may be elicited by the oxidative stress undergone by the cell. Genes related to DNA repair were not clearly induced. Although this bacterium can be 60 classified as radiosensitive since the dose that leaves 10% survival (D 10 value) after Co irradiation in minimal liquid medium was 7 times lower (data not shown) than the 700 Gy D 10 value for the reference strain Escherichia coli K12 (see for example, Qiu et al ., 2006), culturing in space flight on rich medium for 10 days does not seem to induce severe DNA damage in R. rubrum . Together, with a more pronounced response to simulated microgravity

Chapter 4a – R.rubrum S1H in Space – The MESSAGE 2 experiment 99 that slightly overlaps with the response to space flight, the impact of microgravity seems to be larger than that of ionizing radiation in space under these culturing conditions. Integrating transcriptomics and quantitative proteomics results . The specific search for differential expressed protein using the isotopic pairs detection parameter (SILE) obviously limited the number of detected proteins. Moreover, the relative poor correlation between transcriptomic and proteomic data has been reported before (Cox et al. , 2007; Giotis et al ., 2008). Julka and Renier (2004) concluded their research by mentioning that one must be cautious in concluding that the presence or absence of significant mRNA abundance change of a gene detected by transcriptomics necessarily corresponds to presence or absence of significant protein abundance change detected by proteomics. The discrepancy between the mRNA log-ratio and the protein log-ratio of a gene can stem from sustained protein presence from transient transcriptional induction, post-transcriptional regulation or possible measurement errors or any combinations of these causes. Therefore, while not differentially expressed at the transcriptomic level, the up-regulation of the osmoprotectant transporter ProX (among the very few up-regulated proteins) in the simulation of ISS ionizing radiation remains an interesting result that could be related to a probable oxidative stress too. Indeed, exposure of the cells to one type of stress can also condition them against other, seemingly unrelated, stresses. In that respect, it has been shown that when bacteria are challenged with high osmolarity, they can acquire increased resistance to high temperature and oxidative stress via an RpoS-dependent mechanism (Hengge-Aronis et al ., 1993; Smirnova et al ., 2000; Cánovas et al ., 2001). However, based on the genome annotation, R. rubrum does not appear to have a specific sigma factor involved in general stress or entry into stationary phase, e.g. RpoS. This has also been observed for the closely related α-proteobacterium Rhodobacter sphaeroides 2.4.1 (Mackenzie et al ., 2007). Thus, we can suspect these functions to be ascribed to another sigma factor or to non-sigma-factor-type regulator(s). Palma et al . (2004) reported the response of Pseudomonas aeruginosa to hydrogen peroxide to induce an up-regulation of extracellular protein such as ECF. Interestingly, we found the up-regulation of an alternative sigma factor of ECF in the MESSAGE 2 experiment that could be involved in the oxidative stress response of R. rubrum . Finally, the use of 2 ground simulation experiments compared to the actual space flight experiment permitted to put forward specific functional categories as well as cluster of genes with possible physiological link between each other and to take a first step in dissecting the factors influencing in-space culturing of R. rubrum .

CHAPTER 4B - THE BASE-A EXPERIMENT

4B.1 INTRODUCTION

Three years after the MESSAGE 2 experiment, SCK•CEN succeeded to participate in a second space flight journey in October 2006. The 'Bacterial Adaptation to the Space Environment' – part A (BASE-A) experiment was originally designed to validate the response of the bacterium Cupriavidus (former Ralstonia ) metallidurans CH34 to space flight that was hypothesized after its previous two missions to the ISS (MESSAGE 1 & MESSAGE 2 experiments). One of the purposes was to visually follow the growth of selected C. metallidurans mutant strains during space flight. Therefore, the cultures were allowed to grow in a 6 well plate placed inside custom designed transparent 'Biocontainers' (Figure 4b.2). In the BASE-A experiment, one biocontainer, able to contain two 6 well plates, was dedicated to R. rubrum cultures. Two different culture conditions were tested: the rich medium previously used in the MESSAGE 2 experiment and consisting of Sistrom-succinate supplemented with peptone and yeast extract (SPY), and the minimal medium Sistrom-succinate. Sample retrieval after a 12 day space flight directly gave an unfortunate result since the R. rubrum cultures placed on rich medium were dead. One could hypothesize that the R. rubrum culture grew too fast in the rich medium and consumed all the oxygen present (indicated by a methylene oxygen indicator strip placed inside the 6 well plate, figure 4b.1) ultimately leading to cell death. Indeed, R. rubrum is not able to grow in dark conditions on SPY in the absence of oxygen (personal observation). Moreover, there were no single (micro)colonies observed (Figure 4b.1) and the total RNA extracted from the bacterial mat was completely degraded as assessed by using the BioAnalyzer 2100 (data not shown). Therefore only the data extracted from the R. rubrum cultures grown in the 6 well plates with Sistrom-succinate medium either in space flight, in simulation of ISS ionizing radiation environment or in modeled microgravity using the random positioning machine (RPM) are discussed in this chapter.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 102

SPY- FD4 Sis - FD8

SPY- FD8 Sis - FD8

Figure 4b.1 . BASE-A space samples of R. rubrum Sistrom-Peptone-Yeast ( SPY ) and Sistrom-succinate ( Sis ) cultures. The reactive zone of the test strip ( circle ) contains methylene blue which adopts its blue oxidized form in an aerobic environment and its colourless reduced form in an anaerobic environment. Arrow : colony growth. FD : flight day number.

4B.2 MATERIAL AND METHODS

(i) Space flight experiment setup. Cultures of R. rubrum S1H grown to stationary phase in dark aerobic conditions were resuspended in 0.85 % NaCl (Saline tablets, BRS, Oxoid) to a final OD 680 of 0.600 and transported at room temperature from SCK•CEN (Mol, Belgium) to Baikonour (Kazakhstan) 10 days before launch. Three biological independent culture suspensions were deposited 24 hours before launch as 4 spots of 20 µl corresponding to 6x10 6 cells or as 1 mat of 100 µl corresponding to 3x10 7 cells per ca. 10 cm² surface on a 5 ml layer agar medium in 6 well culture plates (CellStar 6, Greiner Bio-One, Belgium) and kept at room temperature. An oxygen indicator strip (Anaerotest, Merck) was added at the bottom of the multiwell plate in between the wells, to visualise oxygen concentration in the gas phase in the multiwell plate during incubation pre-, in- and post-flight. Culture plates were sealed with 1 layer of Parafilm® and 1 layer of Scotch® tape. Radiation dosimeters (Track-Etch Detectors – TED's - and Optically Stimulated Luminescence Detectors – OSLD's described in Vanhavere et al ., 2008) and temperature sensors were added in the container (Figure 4b.2). Two culture plates were sealed hermetically in 1 polycarbonate 'Biocontainer' (PedeoTechniek, Belgium) (Figure

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 103 4b.2) and vacuum sealed in a highly transparent Minigrip®polyethylene bag of 60µm thick. During the 2-days trip in the Soyuz TMA-9 to the ISS, the temperature of the pouch was maintained at 22 ± 1 °C. After docking, all samples were stored at ambient temperature in the ISS Russian Zvezda service module (the exact location is unknown) and a temperature of 20 ± 2 °C was recorded. The samples returned back to Earth after 12 days in-flight (September, 17- 28 2006). After landing in the Kazakhstan desert, the biocontainer was transferred to a ground transportation box at 4°C and transported to Moscow and thereupon to Amsterdam (The Netherlands) without exposure to X-rays scanning. 24h after landing of the Soyuz sample analysis was started at SCK•CEN. The parallel ground control experiment was maintained at 22 °C during transport from Baikonur to Belgium in transportation box, next at 22.5-23.0 °C in an incubator in the lab, and simultaneously cooled down to 4°C after landing of the space samples. From both ground control and ISS exposed samples, colonies and cell lawns were harvested for transcriptomic and proteomic analysis. For this particular space journey, the dosimeters gave a total dose of combined high linear energy transfer (LET) and low-LET particles of ca . 2.6 mGy (Vanhavere et al , 2008). (ii) Simulated space ionizing radiation experiment setup. R. rubrum S1H agar cultures prepared according to the setup described above were exposed to ionizing radiations in the calibration room facility at SCK•CEN (Belgium) as for the simulation of MESSAGE 2 ionizing radiation experiment (see Chapter 4a). During irradiation, samples were kept rotating at 2 rpm in between the two sources ( 137 Cs and 252 Cf) for 10 days (maximum irradiation time allowance) at 20.0 °C ± 0.5 °C, being exposed to a total dose of 2.1 mGy. Actual absorbed dose was assessed in situ by using TLD's (LiF:Mg,Cu,P 'MCP-N' from TLD-Poland, Poland). Non-irradiated control samples were kept outside the irradiation bunker at the same temperature for the same period of time. Cultures were then used for transcriptomic and proteomic analysis. (iii) Simulated microgravity experiment setup. R. rubrum S1H cultures, prepared according to the setup described above, were cultivated in the ESA random positioning machine facility (Fokker Space, The Nederlands) located at the University of Sassari (Italy) as for the MESSAGE 2 RPM simulation (see Chapter 4a). The RPM was operated as a random walk 3D-clinostat (basic mode) with an angular velocity of rotation of 60 deg/s for 10 days. The RPM was located in a room with an ambient temperature of ca . 22°C. Cultures were again used for transcriptomic and proteomic analysis.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 104

Figure 4b.2 . Schematic view of the BASE-A biocontainer. In green, the 6 well culture plate (Vanhavere et al . 2008).

4B.2.1 TRANSCRIPTOMIC ANALYSIS The same whole genome oligonucleotide microarrays were used as described earlier (see Chapter 4a), except for the RPM experiment for which the slides were spotted in-house without the following probes: RruA0080, RruA0167, RruA0849, RruA1962 and RruA2278. Total RNA was extracted using a SV Total RNA Isolation System kit (Promega, United States) and stored at -80°C. RNA amplification was performed using consecutively both MessageAmp™ II-Bacteria Kit for prokaryotic RNA amplification producing the amplified RNA (aRNA) and Amino Allyl MessageAmp TM II aRNA Amplification kit (Ambion, United States) producing the Amino Allyl RNA (aaRNA) following the manufacturer’s instructions. The quality/integrity of total RNA and amplified RNA was checked using a BioAnalyzer 2100 (Agilent, United States). aaRNA was labeled using the ChipShot™ Indirect Labeling and Clean-Up System (Promega and Corning, United States) following the manufacturer’s

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 105 instructions. The C hipShot™ Indirect System provides reagents and enzymes for generating fluorescently labeled cDNA using a two-step labeling process that includes synthesis of aminoallyl-modified cDNA, followed by conjugation to the CyDye™ NHS ester. Labeled cDNA was then hybridized to the printed microarray and scanned as previously described for the MESSAGE 2 experiment (see Chapter 4a). Microarray statistical analysis was performed using BioConductor as described in Chapter 4a. The circular plots were constructed using the Circos library v.0.37. The full description of the array analysis platform has been deposited at the Gene Expression Omnibus website (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE14265.

4B.2.2 MUDPIT APPROACH USING ISOTOPE -CODED PROTEIN LABEL (ICPL) Protein extraction and quantification were performed as described for the MESSAGE 2 experiment. The same significance threshold has been set meaning a fold change lower than 0.7 and higher than 1.5.

4B.3 RESULTS

4B.3.1 TRANSCRIPTOMIC ANALYSIS Using the whole genome DNA chip, respectively 1.7 % and 0.4 % of the genes (out of 3,143 genes retained) were identified as significantly up- and down-regulated after a 12-day journey inside the ISS compared to the ground control. Concerning the ionizing radiation simulation, 4.5 % and 1.2 % of the genes (out of 3,723 genes retained) were identified as respectively significantly up- and down-regulated after a 10-day ionizing radiation experiment compared to non-irradiated samples. For the modeled microgravity experiment, 0.6 % and only one of the genes (out of the 3,797 genes retained) were identified as respectively significantly up- and down-regulated after 10 days culturing on the random positioning machine compared to the normal gravity control. A circular plot of the genes differentially expressed in the chromosome of R. rubrum S1H showed that theses genes were not distributed globally throughout the chromosome for the ionizing radiation experiment where the up-regulated genes seemed to be essentially clustered on the last third of the chromosome (Figure 4b.3).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 106

(i) (ii)

(iii) (iv) (v) (vi)

Rhodospirillum rubrum S1 Chromosome 4.35 Mb

Figure 4b.3 . Map of the 4.35-Mb circular R. rubrum S1 chromosome. From the outside to the inside: (i) genes on direct strand, (ii) genes on reverse strand, genes up-regulated above 2 fold (in red color) and down-regulated below 0.5 fold ( green color) in (iii) the BASE-A experiment, (iv) simulation of ISS ionizing radiation and (v) simulation of microgravity by culturing on RPM. (vi) gene position (x 10,000 bp).

Considering the 261 non-redundant significant differentially expressed genes from the 3 cultures conditions, 31 % appeared to code for hypothetical proteins (Additional table 4b.1). In addition, genes from the 'unclassified' functional category that included 44 hypothetical proteins appeared to be the most numerically abundant category whatever the culture conditions and the way of induction (Figure 4b.4ab).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 107

30

25

20

(a) BASE-A UP 15 simRAD UP RPM UP Number of genes

10

5

0 EGDMNBHVLCSRPU I FOQTKJUn. Functional category (COG)

30

25

20

BASE-A DOWN (b) simRAD DOWN 15 RPM DOWN

Number of genes of Number

10

5

0 EGDMNBHVLCSRPU I FOQTKJUn. Functional category (COG)

Figure 4b.4 . Functional classification of the significant genes up-regulated ( a) and down-regulated ( b) during the BASE-A experiment, the simulation of ISS ionizing radiation ( simRAD ) and the modeled microgravity experiment ( RPM ) (0.5 > fold change > 2 and p-value < 0.05).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 108

4B.3.1.1 THE BASE-A EXPERIMENT

As mentioned above, differentially expressed genes from the 'unclassified' functional category were the most numerically abundant including 8 up- and 7 down-regulated genes. These were all genes coding for hypothetical proteins including the up-regulated RruA2588, RruA2590 and RruA3310 (Additional table 4b.1). The latters have been annotated using the MaGe platform as 3 transposases. Six genes from the 'energy production and conversion' (C) class were up-regulated (Additional table 4b.1) and 5 from the 'Amino acid transport and metabolism' (E) (Additional table 4b.1). The most over-expressed gene was RruA3097 that is annotated as a putative protease. Finally, in the space samples slight up-regulation was observed for RruA0807 coding for a conjugal transfer protein, RruA1661 coding for an extensin-like protein, a DNA mismatch repair related gene RruA2946, and the beta- lactamase-like encoding gene RruA2629 (Additional table 4b.1).

4B.3.1.2 THE SIMULATION OF ISS IONIZING RADIATION USING THE BASE-A SETUP

This condition gave the highest number of significant differentially expressed genes with 169 up- and 43 down-regulated genes (Additional table 4b.1). While the fold change neither in the BASE-A nor in the RPM experiment exceeded 4, the ionizing radiation rendered 20 genes with a fold change above 5 (Table 4b.1).

Table 4b.1 . Distribution of the significant differentially espressed genes in regard to the fold induction in the simRAD experiment. SimRAD culture conditions Total number of genes Gene coding for hypothetical proteins fold change > 10 4 1 10 > fold change > 5 16 2 5 > fold change > 2 149 47 0.5 > fold change > 0.3 39 15 Fold change ≤ 0.3 5 2

Twenty-six genes from the 'unclassified' class were up-regulated including 23 genes coding for hypothetical proteins, one putative polyhydroxybutyrate depolymerase (RruA1585) as well as the alpha- (RruA2976) and beta-subunit (RruA2977) of the light-harvesting complex B-870. The 12 down-regulated genes were all annotated as hypothetical proteins (Additional table 4b.1). The 'amino acid transport and metabolism' (E) class counted 17 up- and 1 down-regulated genes including the highly over-expressed sulfate adenylylstransferase (RruA2289) and the binding-protein dependent transport system (RruA0780) showing respectively a fold change

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 109 of 14.65 and 8.76 (Additional table 4b.1). These genes were also over-expressed but to a lesser extend in the BASE-A experiment (see below). The 'general function prediction only' (R) class included 14 up- and 4 down-regulated genes with respectively 4 and 1 genes coding for hypothetical proteins. The most up-regulated gene from this class was RruB0019 encoding for an O-methyltransferase family protein and is located on the plasmid (Additional table 4b.1). In addition, a peptidase (RruA0902) and a major facilitator family protein (RruA3661) were up-regulated while a tellurium stress related protein (RruA0892) and the putative Fe-S clustered gene encoding protein RruA1069 were down-regulated (Additional table 4b.1). The 'function unknown' functional category contained 13 up-regulated and 7 down-regulated genes. The lowest down-regulated gene was RruA2723 coding for a rubrerythrin. The 'energy production and conversion' (C) and the 'signal transduction' (T) classes contained equal amounts of differentially expressed genes. While almost half of these genes from category C were shared with the BASE-A experiment (see below), the 12 up- and 2 down- regulated genes from category T were under-expressed specifically in the ionizing radiation experiment (Additional table 4b.1). Finally, 10 differentially expressed genes belonged to the 'cell envelope biogenesis, outer membrane' (M), for which 5 up-regulated genes encoded for glycosyl transferase proteins (RruA0579, RruA2494, RruA2737, RruA3090, RruA3375 and RruA3657), and 10 belonged to the 'transcription' (K) class that contained only 1 down-regulated gene namely RruA0603 encoding for a cold-shock protein. The 'inorganic ion transport and metabolism' (P) class contained 9 genes including the up-regulated ferric uptake regulator gene fur (RruA3788).

4B.3.1.3 THE MODELED MICROGRAVITY (RPM) EXPERIMENT USING THE BASE-A SETUP

With 23 up- and 1 down-regulated gene, the RPM experiment seemed to induce less differential expression of genes than the BASE-A experiment or the simulation of ISS ionizing radiation (Additional table 4b.1). The gene RruA3477 related to a major facilitator transporter as well as a phage integrase encoding gene (RruA0674) were up-regulated in the RPM samples. In addition, 4 genes coding for hypothetical proteins (RruA0086, RruA1537, RruA1758 and RruA2208) were specifically up-regulated after RPM culturing (Additional table 4b.1).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 110

4B.3.1.4 OVERLAPPING GENES BETWEEN THE THREE CULTURE CONDITIONS TESTED

Overlapping genes between the different culture conditions were only obvious for BASE-A and the simulation of ISS ionizing radiation (Figure 4b.5).

BASE-A simRAD

25 37 170

2 1 5

21

RPM

Figure 4b.5 . Venn diagrams showing the relation between the 253 significant genes categorized from the BASE- A, the simulation of ISS ionizing radiation ( SimRAD ) and the modeled microgravity experiments ( RPM ).

BASE-A and the simulation of ISS ionizing radiation shared 38 differentially expressed genes (same way), while the only gene (RruA0269) common to the BASE-A and the RPM samples was expressed in opposite direction (Additional figure 4b.1). Finally, the ionizing radiation and the RPM experiment have specifically 5 genes in common, all being differentially expressed in opposite direction except RruA1854 commonly up-regulated (Table 4b.2). Only the genes RruA0119, RruA0672 and RruA3286 encoding for hypothetical proteins were common to the three culture conditions, being down-regulated in BASE-A and the ionizing radiation experiment while being up-regulated in the RPM sample (Table 4b.2). The 38 genes shared between BASE-A and the irradiation experiment include over-expression of the chaperones RruA3332 and RruA3433 as well as the cell envelope related genes RruA1608 ( mscS ), RruA2737 ( wcaA ) and RruA3715. In addition, the hfq (RruA1684) gene encoding for a RNA-binding regulatory protein was down-regulated in both the BASE-A and the irradiation experiment. Interestingly, the gene RruA2016 encoding for a hypothetical protein has been found to be frequently associated with the chaperone dnaJ using the MaGe platform (Table 4b.2).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 111

Table 4b.2 . List of the 45 significant genes shared by at least 2 cultures conditions within the BASE-A, the simulation of ISS ionizing radiation ( simRAD ) and the RPM experiments. Genes Gene FC p-value FC p-value FC p-value Product name COG number name BASE-A BASE-A simRAD simRAD RPM RPM hypothetical protein RruA0119 - Un. 0.42 2.01E-02 0.37 1.03E-02 2.02 1.75E-02 Rru_A0119 preprotein translocase RruA0235 secA U 2.12 4.56E-03 9.21 1.98E-06 0.87 8.07E-01 subunit SecA hypothetical protein RruA0269 - M 0.50 3.83E-03 0.53 1.30E-03 2.20 2.12E-03 Rru_A0269 hydrogenase RruA0306 - formation HypD O 2.58 3.98E-03 5.05 2.23E-06 1.01 9.88E-01 protein cold-shock DNA- RruA0663 - binding protein family K 0.53 3.94E-03 0.41 4.93E-03 2.37 3.14E-03 protein hypothetical protein RruA0672 - Un. 0.50 5.39E-03 0.37 3.76E-05 1.85 1.95E-03 Rru_A0672 binding-protein dependent transport RruA0780 - E 2.12 2.69E-03 8.76 9.15E-07 0.96 9.50E-01 system inner membrane protein EmrB/QacA family RruA0858 emrB drug resistance G 2.65 3.83E-03 3.15 1.98E-05 1.27 3.43E-01 transporter RruA0920 - acetate kinase C 2.20 4.28E-03 4.83 3.45E-06 0.91 8.06E-01 RruA0927 - luciferase-like C 2.15 5.11E-03 2.35 6.57E-05 0.83 7.78E-01 Linocin_M18 RruA0974 - S 0.59 3.83E-03 0.37 8.52E-05 3.93 3.77E-04 bacteriocin protein hypothetical protein RruA1107 - K 2.27 1.41E-03 2.39 1.58E-05 0.92 8.46E-01 Rru_A1107 AsnC family RruA1172 - transcriptional K 2.64 4.28E-03 6.61 1.96E-05 0.85 6.07E-01 regulator RruA1236 - Iojap-related protein S 0.39 1.14E-02 0.50 2.14E-03 1.73 5.27E-02 RruA1315 - sterol-binding I 0.75 1.84E-02 0.46 7.14E-03 2.09 2.46E-04 butyryl-CoA:acetate RruA1382 - I 2.26 4.56E-03 3.04 4.28E-04 0.81 6.49E-01 CoA transferase hypothetical protein RruA1411 - S 2.26 7.33E-03 3.13 1.96E-05 0.92 8.81E-01 Rru_A1411 MscS RruA1608 mscS mechanosensitive ion M 2.21 9.39E-03 5.70 2.23E-06 0.96 9.52E-01 channel RNA-binding protein RruA1684 hfq R 0.45 4.06E-03 0.41 1.77E-04 1.44 2.11E-01 Hfq pyridoxal phosphate RruA1854 - biosynthetic protein H 1.07 6.07E-01 2.67 1.10E-04 2.07 8.09E-02 PdxJ pyruvate RruA1881 - dehydrogenase C 2.03 3.98E-03 3.26 7.01E-05 1.01 9.82E-01 (lipoamide) TetR family RruA1934 - transcriptional K 2.89 1.33E-03 5.21 1.98E-06 1.01 9.77E-01 regulator divalent cation RruA1995 - P 2.55 5.71E-03 2.58 6.32E-04 1.08 9.11E-01 transporter hypothetical protein RruA2016 - Un. 2.06 5.68E-03 2.11 1.10E-03 1.04 9.45E-01 Rru_A2016 nitrogenase MoFe RruA2286 nifE cofactor biosynthesis C 2.24 5.39E-03 5.34 5.64E-06 1.00 9.94E-01 protein NifE sulfate RruA2289 - adenylyltransferase E 2.67 3.83E-03 14.65 1.34E-06 0.91 8.33E-01 subunit 2 ABC transporter RruA2389 - P 2.35 4.03E-03 4.85 2.78E-04 0.95 9.43E-01 component hypothetical protein RruA2561 - Un. 2.71 3.83E-03 4.41 1.98E-05 0.98 9.66E-01 Rru_A2561 glycosyl transferase RruA2737 wcaA M 2.27 7.72E-03 2.29 1.15E-03 0.72 1.81E-01 family protein response regulator RruA2837 ompR receiver domain- K 2.16 3.98E-03 2.41 6.18E-04 1.05 9.52E-01 containing protein hypothetical protein RruA2850 cheL M 0.72 1.10E-02 0.33 2.39E-04 2.02 3.21E-04 Rru_A2850 RruA2895 - twin-arginine P 2.03 5.26E-03 8.30 9.67E-06 0.80 3.46E-01

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 112

Genes Gene FC p-value FC p-value FC p-value Product name COG number name BASE-A BASE-A simRAD simRAD RPM RPM translocation pathway signal hypothetical protein RruA3034 - Un. 2.02 1.68E-03 3.76 5.48E-06 0.95 9.12E-01 Rru_A3034 hypothetical protein RruA3181 - S 0.48 3.83E-03 0.47 1.43E-03 1.16 7.09E-01 Rru_A3181 acetylornithine RruA3277 - E 2.70 5.11E-03 2.33 1.53E-03 1.01 9.88E-01 aminotransferase hypothetical protein RruA3286 - Un. 0.40 1.31E-02 0.30 5.41E-03 3.21 8.40E-02 Rru_A3286 hypothetical protein RruA3317 mmeA Un. 2.01 7.72E-03 28.69 1.59E-07 0.78 4.42E-01 Rru_A3317 pantothenate RruA3320 - H 3.55 1.76E-04 4.05 3.15E-06 0.84 4.42E-01 synthetase cytochrome c oxidase RruA3332 - O 2.18 3.83E-03 6.02 5.31E-06 1.18 5.42E-01 cbb3-type, subunit I hypothetical protein RruA3394 - Un. 0.47 4.01E-03 0.35 1.08E-03 1.65 2.50E-01 Rru_A3394 formaldehyde RruA3405 adhC dehydrogenase C 2.09 3.83E-03 5.08 1.58E-05 1.04 9.52E-01 (glutathione) RruA3433 - thioredoxin O 2.52 3.28E-03 3.81 1.27E-05 1.31 1.84E-01 lipopolysaccharide RruA3715 - M 2.41 7.33E-03 4.88 1.56E-05 1.06 9.44E-01 biosynthesis RruA3733 - HNH endonuclease V 0.48 4.28E-03 0.41 8.52E-05 0.90 9.44E-01 RruA3746 rplY 50S ribosomal protein J 2.57 4.43E-03 4.43 9.67E-06 1.06 9.44E-01 Red highlight : up-regulated genes. Green highlight : down-regulated genes. FC : fold change. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un. – Unclassified.

4B.3.2 PROTEOMIC ANALYSIS

4B.3.2.1 THE BASE-A EXPERIMENT

For the BASE-A experiment, the computed FDR was 0.32 % estimated at the peptide level. From these peptides, 366 proteins were identified (Additional table 4b.2) including 248 quantified with at least 2 peptides. This represents respectively 9.5 % and 6.5 % of the total candidate protein-encoding genes. The median value for the protein fold change distribution plot was 0.91 (Figure 4b.6). Nine down-regulated proteins including 3 proteins annotated as hypothetical showed significant change in expression (Table 4b.3).

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 113

Table 4b.3 . Significant proteins from the BASE-A experiment. Gene Gene FC Mascot MW Product name COG # (H/L) number name BASE-A Score (kDa) RruA0148 typA GTP-binding protein TypA T 0.63 ± 0.05 7 156.56 67.32 RruA0221 pgk phosphoglycerate kinase G 0.69 ± 0.09 3 437.72 41.69 RruA0684 pncB hypothetical protein Rru_A0684 H 0.64 ± 0.05 2 124.32 45.36 RruA1262 ugpB extracellular solute-binding protein G 0.63 ± 0.05 3 84.47 46.98 RruA1656 - hypothetical protein Rru_A1656 S 0.67 ± 0.02 2 116.27 13.57 RruA1760 sodA superoxide dismutase P 0.63 ± 0.02 2 168.71 24.93 RruA2557 - hypothetical protein Rru_A2557 S 0.70 ± 0.06 13 596.75 35.38 RruA3572 - YciI-like protein S 0.66 ± 0.03 2 184.69 10.24 RruB0037 - hemolysin-type calcium-binding region Un. 0.70 ± 0.06 2 157.86 81.39 'RruA' refers to gene located on the chromosome while 'RruB' refers to gene located on the plasmid. FC : fold change. MW: molecular weight. COG nomenclature: G - Carbohydrate transport and metabolism, H - Coenzyme metabolism, S - Function unknown, P - Inorganic ion transport and metabolism, T - Signal transduction mechanisms, Un . – Unclassified.

2

1.8

1.6

1.4 1.2

1 0 50 100 150 200 250

Foldchange 0.8

0.6

0.4

0.2 ratio BASE-A/control 0 Protein number

Figure 4b.6 . Fold change distribution of the protein quantified by the differential proteomic approach for the BASE-A.

4b.3.2.1.1 Comparison with the corresponding transcriptomic data Of the 65 significant differentially expressed genes, 11 could be confirmed at the proteomic level (Table 4b.4). However, none of these proteins satisfied the threshold for significance.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 114

Table 4b.4 . Transcriptomics versus differential proteomics approach for the BASE-A experiment. BASE-A BASE-A Gene Gene BASE-A Product name COG FC mRNA number name FC protein mRNA p-value RruA0235 secA preprotein translocase subunit SecA U 2.12 4.56E-03 0.94 ± 0.06 RruA1566 - NADH dehydrogenase subunit L C 2.99 3.43E-03 1.01 ± 0.00 RruA1881 - pyruvate dehydrogenase (lipoamide) C 2.03 3.98E-03 1.39 ± 0.22 RruA2692 rpsG 30S ribosomal protein S7 J 2.70 1.15E-03 0.82 ± 0.03 RruA2837 ompR response regulator receiver domain-containing protein K 2.16 3.98E-03 0.85 ± 0.12 RruA3317 mmeA hypothetical protein Rru_A3317 Un. 2.01 7.72E-03 IBNQ RruA3320 - pantothenate synthetase H 3.55 1.76E-04 1.10 ± 0.09 RruA3332 - cytochrome c oxidase cbb3-type, subunit I O 2.18 3.83E-03 IBNQ RruA3405 adhC formaldehyde dehydrogenase (glutathione) C 2.09 3.83E-03 IBNQ RruA3433 - thioredoxin O 2.52 3.28E-03 0.96 ± 0.05 RruA3746 rplY 50S ribosomal protein L25P J 2.57 4.43E-03 IBNQ 'RruA' refers to gene located on the chromosome. FC : fold change. IBNQ : identified protein but not quantified by at least two H/L occurrences (see material and methods). Proteins are quantified by mean ± standard deviation. COG nomenclature: H - Coenzyme metabolism, C - Energy production and conversion, U - Intracellular trafficking, secretion, and vesicular transport, O - Posttranslational modification, protein turnover, chaperones, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

4B.3.2.2 THE SIMULATION OF ISS IONIZING RADIATION USING THE BASE-A SETUP

For this experiment, the computed FDR was 0.31 % estimated at the peptide level. 387 proteins were identified (Additional table 4b.3) including 282 quantified with at least 2 peptides. This corresponded respectively to 10 % and to 7 % of the total candidate protein- encoding genes. Plotting the fold change to the protein number also showed a correct median value (1.05) (Figure 4b.7). In this analysis, only 2 proteins were shown to be significantly down-regulated including 1 protein also found in the BASE-A sample (RruA0148) and 1 protein annotated as hypothetical (Table 4b.3 and Table 4b.5).

Table 4b.5 . Significant proteins from the simulation of ISS ionizing radiation using the BASE-A setup. Gene Gene Mascot MW Product name COG FC simRAD # (H/L) number name Score (kDa) RruA0148 typA GTP-binding protein TypA T 0.63 ± 0.05 6 158.04 67.32 RruA0936 - hypothetical protein Rru_A0936 S 0.69 ± 0.03 2 67.95 41.66 'RruA' refers to gene located on the chromosome while 'RruB' refers to gene located on the plasmid. FC : fold change. MW: molecular weight. Mean ± standard deviation. COG nomenclature: S - Function unknown, T - Signal transduction mechanisms.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 115

2

1.8

1.6

1.4

1.2 1 0 50 100 150 200 250 300 Foldchange 0.8

0.6

0.4

0.2

ratio simRAD/control 0 Protein number Figure 4b.7 . Fold change distribution of the protein quantified by the differential proteomic approach for the simulation of ISS ionizing radiation ( simRAD ).

4b.3.2.2.1 Comparison with the corresponding transcriptomic data Among the 214 significant differentially expressed genes from the simulation of ISS ionizing radiation using the BASE-A setup, 25 corresponding proteins could be confirmed (Table 4b.6). However, none of these proteins had a fold change above 1.5 or below 0.7, even for the RruA1234 gene that showed a fold induction higher than 10 fold.

Table 4b.6 . Transcriptomics versus differential proteomics approach for the simulation of ISS ionizing radiation using the BASE-A setup. simRAD simRAD Gene Gene simRAD Product name COG FC mRNA number name FC protein mRNA p-value RruA0146 ppa inorganic diphosphatase C 7.17 2.13E-06 1.04 ± 0.14 RruA0210 rpmB 50S ribosomal protein L28 J 0.34 4.41E-06 1.07 ± 0.07 CheA Signal transduction histidine Kinases RruA0521 cheA N 0.40 4.64E-05 1.09 ± 0.10 (STHK) RruA0586 groES chaperonin Cpn10 O 0.47 3.85E-05 1.10 ± 0.04 response regulator receiver domain-containing RruA0665 - K 2.36 1.10E-04 1.18 ± 0.18 protein RruA0892 - stress protein R 0.49 2.78E-04 IBNQ RruA0972 tyrA prephenate dehydrogenase E 2.15 3.76E-05 IBNQ RruA0974 - Linocin_M18 bacteriocin protein S 0.37 8.52E-05 IBNQ RruA1043 rpsU 30S ribosomal protein S21 J 2.01 6.57E-05 1.09 ± 0.09 RruA1234 gpmI phosphoglyceromutase G 10.94 2.23E-06 1.08 ± 0.05 RruA1551 clpX ATP-dependent protease ATP-binding subunit O 0.46 5.41E-03 1.05 ± 0.06 RruA1881 - pyruvate dehydrogenase (lipoamide) C 3.26 7.01E-05 1.24 ± 0.09 RruA1917 ddpA extracellular solute-binding protein E 2.40 2.91E-02 1.10 ± 0.05 RruA2167 purM phosphoribosylformylglycinamidine cyclo-ligase F 2.49 4.52E-04 IBNQ RruA2687 rplD 50S ribosomal protein L4P J 14.25 1.79E-05 0.97 ± 0.07 response regulator receiver domain-containing RruA2837 ompR K 2.41 6.18E-04 1.41 ± 0.09 protein aspartyl/glutamyl-tRNA(Asn/Gln) RruA3177 gatC J 2.14 3.67E-03 1.08 ± 0.09 amidotransferase subunit C RruA3279 - hypothetical protein Rru_A3279 S 2.12 5.02E-04 IBNQ

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 116

simRAD simRAD Gene Gene simRAD Product name COG FC mRNA number name FC protein mRNA p-value RruA3320 - pantothenate synthetase H 4.05 3.15E-06 0.78 ± 0.11 RruA3405 adhC formaldehyde dehydrogenase (glutathione) C 5.08 1.58E-05 IBNQ RruA3433 - thioredoxin O 3.81 1.27E-05 1.05 ± 0.05 RruA3631 leuS leucyl-tRNA synthetase J 2.56 2.97E-04 1.01 ± 0.08 RruA3711 - thioredoxin-related O 2.10 2.93E-04 IBNQ RruA3746 rplY 50S ribosomal protein L25P J 4.43 9.67E-06 IBNQ RruA3785 pnp polynucleotide phosphorylase/polyadenylase J 2.22 8.52E-05 1.12 ± 0.07 'RruA' refers to gene located on the chromosome. FC : fold change. IBNQ : identified protein but not quantified by at least two H/L occurrences (see material and methods). Proteins are quantified by mean ± standard deviation. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, N - Cell motility and secretion, H - Coenzyme metabolism, C - Energy production and conversion, S - Function unknown, R - General function prediction only, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, K – Transcription, J - Translation, ribosomal structure and biogenesis.

4B.3.2.3 THE MODELED MICROGRAVITY EXPERIMENT USING THE BASE-A SETUP

The computed FDR was 0.30 % estimated at the peptide level. We could retrieve 429 identified proteins (Additional table 4b.4) among which 282 were quantified with at least 2 peptides. With a median value of 1.04, the protein fold change distribution showed again a correct mixing of the control and the experimental samples (Figure 4b.8). For this experiment, no proteins showed a fold change above 1.5 or below 0.7 fold.

4b.3.2.3.1 Comparison with the corresponding transcriptomic data Of the 29 significant genes identified at the transcriptomic level, 3 were also identified at the proteomic level (Table 4b.7) but none were found to be significantly differentially expressed.

Table 4b.7 . Transcriptomics versus differential proteomics approach for the modeled microgravity (RPM ) experiment using the BASE-A setup. RPM RPM RPM FC Gene number Gene name Product name COG FC mRNA protein mRNA p-value RruA0974 - Linocin_M18 bacteriocin protein S 3.93 3.77E-04 IBNQ RruA1072 rpmE 50S ribosomal protein L31 J 2.09 1.09E-01 1.04 ± 0.03 RruA1311 soxR MerR family transcriptional regulator K 0.33 4.29E-02 1.12 ± 0.06 'RruA' refers to gene located on the chromosome. FC : fold change. IBNQ : identified protein but not quantified by at least two H/L occurrences (see material and methods). Proteins are quantified by mean ± standard deviation. COG nomenclature: S - Function unknown, K – Transcription, J - Translation, ribosomal structure and biogenesis.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 117 2

1.8

1.6

1.4

1.2

1 0 50 100 150 200 250 300 Foldchange 0.8

0.6

0.4

0.2

ratio RPM/control 0 Protein number

Figure 4b.8 . Fold change distribution of the protein quantified by the differential proteomic approach for the modeled microgravity ( RPM ) experiment.

4B.3.3 COMPARISON WITH THE MESSAGE 2 RELATED EXPERIMENTS

4B.3.3.1 COMPARISON REGARDING THE TRANSCRIPTOMIC APPROACH

In the MESSAGE 2 related experiments, the space and the related RPM experiments elicited the highest number of differentially expressed genes. On the other hand, in the BASE-A related experiments, the highest number of significant genes were found in the space and in the related ionizing radiation experiments. A general comparison of the differentially expressed genes from the two space and their related simulation experiments indicated a low overlap (Figure 4b.9). MESSAGE 2 and BASE-A showed 5 genes in common including 3 genes encoding for hypothetical proteins (RruA0119, RruA0269, RruA3286), the NADH dehydrogenase subunit L encoding gene RruA1566 and the ribosomal protein encoding RruA3746. However, as mentioned above, about 600 genes from the BASE-A space experiment were not kept for statistical analysis after spot filtering quality control. As a direct consequence, of the 219 significant genes from the MESSAGE 2 experiment, 49 have no corresponding fold induction value in the BASE-A experiment. Only one hypothetical protein encoding gene (RruA2994) was jointly expressed in the ISS ionizing radiation simulations of MESSAGE 2 and BASE-A while 3 genes (including RruA1537 and RruA2850 coding for hypothetical proteins and the ribosomal protein encoding gene RruA1072) were common to the microgravity simulation experiments.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 118 However, analyzing differential gene expression of large sets of genes will give overlap in the studied conditions just by chance. Indeed, considering the length of the compared genes list and the total number of genes surveyed, we could expect an overlap of respectively 7, 2 and 3 genes (p-value < 0.05) for the space, the ionizing radiation and the modeled microgravity experiments taken 2 by 2. Therefore, when interpreting the present results one should consider that the observed overlap could be expected just by chance.

214 5 60 8 1 21 3 153 3 26

MSG2 BASE-A MSG2 BASE-A MSG2 BASE -A simRAD simRAD RPM RPM

Figure 4b.9 . Comparison of significant genes expressed in the MESSAGE 2 ( MSG2 ) and BASE-A related experiments. simRAD : simulation of ISS ionizing radiation. RPM : modeled microgravity experiment using the random positioning machine.

4B.3.3.2 COMPARISON REGARDING THE PROTEOMIC APPROACH

Within the differential proteomic approach, no overlap was found between the MESSAGE 2 and BASE-A related experiments ( i.e . ionizing radiation and RPM).

4B.4 DISCUSSION

Experimental setup and bacterial response . As already mentioned for the MESSAGE 2 related experiments, discrepancies between transcriptomic and proteomic data were also observed in the present study. Moreover, it was not possible to confirm or invalidate the transcriptomic data using quantitative PCR because the former were obtained from amplified mRNA using the Ambion kit which was unsuitable template material for PCR reaction (personal observation). However, sample retrieval after space flight indicated directly that the culturing conditions in BASE-A were not comparable to those in the MESSAGE 2 experiment since the cells inoculated on the rich SPY medium died. Therefore, although culturing on a rich or on a minimal mineral medium could affect the response to space flight (Wilson et al ., 2008) and other stresses, the differential response in BASE-A compared to MESSAGE 2 was more intricate than solely due to culture medium.

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 119 Interestingly, for the BASE-A setup, a more pronounced effect was observed for the simulation of ionizing irradiation than for microgravity simulation in contrast to the MESSAGE 2 experiment. In addition, R. rubrum 's response to the BASE-A experiment showed a larger overlap with its response to simulation of ionizing radiation than to simulated microgravity while the opposite was observed for the MESSAGE 2 experiment. One could therefore argue that the global experimental setup used in both experiments influenced the effect of space flight conditions including microgravity and ionizing irradiation on the R. rubrum culture, and consequently influenced the bacterium's response to it. Stress elicited by ionizing radiation solely has apparently a more pronounced effect in a minimal mineral medium compared to a rich medium where more anti-oxidant compounds are present and therefore less ROS are generated (Lee et al ., 2006) or produced ROS are more rapidly scavenged. In that purpose, we observed the significant up-regulation of the ferric uptake repressor fur in the simulation of ionizing radiation in the BASE-A setup. This made less iron available for the Fenton reaction, thereby linking cellular iron status to oxidative stress (Hantke, 2001; Imlay, 2003). Moreover, we observed the down-regulation of the hfq gene encoding for a RNA-binding regulatory protein in the simulation of ionizing radiation (and in the space experiment) using the BASE-A setup. This was notable since Vecerek et al. (2003) showed that hfq indeed negatively regulates the expression of fur in E. coli . For growth in (simulated) microgravity, fluidic and diffusion processes will be affected and therefore a different metabolism and higher growth rate in rich medium could perhaps result in either an inadequate replenishment of nutrients or in a build-up of secondary metabolites and waste products in the boundary layer surrounding the cell, eliciting a different but more profound response in rich medium compared to minimal mineral medium. Indeed, several studies already reported more apparent bacterial response when cultivation in space flight and space flight analogs were performed on rich medium (Baker and Leff, 2004; Baker and Leff, 2006; Benoit and Klaus, 2007; Wilson et al ., 2008). Genes coding for hypothetical proteins . We found about the same percentage (30 %) of differentially expressed genes coding for hypothetical proteins as we found in the MESSAGE 2 related experiments (chapter 4a). After manual annotation using the MaGe platform, 5 genes coding for hypothetical proteins were further explored: (i) RruA2588, RruA2590 and RruA3310, over-expressed in the space samples, appeared to belong to mobile genetic elements, i.e . insertion sequences (IS). The mobility of IS has been linked to various bacterial stress reponses (reviewed in Foster, 2007), including the well-known UV irradiation-induced induction of prophage; (ii) RruA2016, found in the space and the ionizing radiation

Chapter 4b – R. rubrum S1H in Space – The BASE-A experiment 120 experiments using the BASE-A setup, is a conserved protein frequently associated with DnaJ- like chaperones and therefore gave another argument in favor of a stress response; (iii) RruA1661 was annotated as coding for an extensin-like protein. The latter is typical for plants where a mechanical function is described. In bacteria the function remains unknown although these proteins are regularly found. More genes coding for hypothetical proteins linked to the BASE-A related experiments will be discussed in chapter 4c.

CHAPTER 4C - R. RUBRUM S1H LIQUID CULTURE IN MODELED MICROGRAVITY

4C.1 INTRODUCTION

In-flight research is associated with many practical (and other) limitations and therefore, the space experiments MESSAGE 2 and BASE-A were conducted only on solid agar medium. While giving us a first glimpse on how the bacteria can react to space flight conditions the culture conditions remained far from the actual liquid bioreactor condition foreseen in the MELiSSA loop. Therefore, we performed a series of experiments in liquid conditions using ground-based systems namely the Rotating Wall Vessel (RWV) able to create a low-shear modeled microgravity (LSMMG) environment and the Random Positioning Machine (RPM) technology. Both techniques have been extensively described in the chapter 1.

4C.2 MATERIAL AND METHODS

(i) Strain and medium. Three independent cultures of R. rubrum S1H grown to stationary phase in Sistrom-succinate medium and dark aerobic conditions were resuspended in 0.85 %

NaCl to a final OD 680 of ca . 0.600 to constitute stock cultures. (ii) Culture setup . All cultures were made at 21 °C. The RWV bioreactors (Cellon, Bereldange, Luxembourg) were filled completely with ca . 58 ml of culture medium containing 1 % inoculum from the stock culture. Air bubbles were carefully removed through the sampling ports (Figure 1.9b), using syringes (without needle), in order to avoid undesired shear stress. The vessels were mounted on separate RWV devices and were placed in a culture chamber, one in vertical position creating the low-shear modeled microgravity (LSMMG) environment while the control was placed in horizontal position. Gas exchange in the RWV bioreactors during growth was ensured by the gas-permeable silicone membrane present at the back of each RWV culture vessel (Figure 1.9b). Bacterial growth in RWV conditions was allowed at a rotational speed of 25 rpm. The same RWV vessels were also used to cultivate R. rubrum in liquid aerobic condition using the RPM. The

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 122 horizontal RWV incubation described above was also used as control for the RPM cultivation (Figure 4c.1). The RPM experiment was performed in the lab of Prof. P. Pippia at the University of Sassari (I) while RWV cultivation was performed in Mol at the SCK•CEN. The culture chamber in SCK•CEN permits to control the water saturation which is crucial for RWV cultivation to minimize evaporation of the culture medium through the gas-permeable silicone membrane. This was achieved by setting up the water saturation of the culture chamber to approximately 90 %. Nonetheless, no culture medium evaporation was observed in the RWV vessels during cultivation on the RPM device. (iii) Microarrays and RT-qPCR analysis . Microarray analysis was performed using biological and technical triplicates as previously described (chapter 4a). Genes were considered as significant when the fold change was higher than 2 or lower than 0.5 with a p- value lower than 0.05. For RT-qPCR analysis, the same biological triplicates were used and the following primer sets were designed (Table 4c.1):

Table 4c.1 . Sequences of gene specific primers used for RT-qPCR. Target gene Primer sequence Gene product name Fw-rRNA 16S 5'-GTGGAGCATGTGGTTTAATTCG-3' 16S ribosomal RNA Rv-rRNA 16S 5'-GGAAGTGTCACGGGATGTCAA-3' Fw-RruA0222 5'-GGTGGTGGCGATCAATGAC-3' glyceraldehyde-3-phosphate Rv-RruA0222 5'-CCGTGGACGGAATCGAACT-3' dehydrogenase Fw-RruA1518 5'- TGCAGGGCAACAGGATCTG-3' hypothetical protein Rru_A1518 Rv-RruA1518 5'-CACGTCGCGCAAAACATG-3' (putative transposase) Fw-RruA3396 5'-TTTTTGTTAATCGCCTGAAATGG-3' autoinducer synthesis protein Rv-RruA3396 5'-TCATCAAAACGGTCGATTTCC-3'

The primers were purchased from Eurogentec (Seraing, Belgium) and the RT-qPCR reaction was performed in duplicate as previously stated. The 16S rRNA was used as normalization for the RPM samples while the samples of interest for these reactions were diluted 1000 times to avoid saturation of the fluorescence detector. RruA0222 was the most stably expressed gene for the RWV samples and then it was used as normalization gene. (iv) Proteome characterization . Proteome analyses were conducted using MudPIT approach coupled to the ICPL technology as described above (in chapter 4a). Repeated sample injection of pooled biological quadruplicate was performed. Proteins were considered as significant when fold change was higher than 1.5 of lower than 0.7 fold as mentioned earlier. These thresholds were adapted when needed (Wang et al ., 2008) (see below).

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 123

Sassari, Italy Mol, Belgium

RPM modeled Control RWV vessel RWV modeled microgravity microgravity

Experimental + Control Control + Experimental samples samples

Figure 4c.1. Modeled microgravity experimental setup.

4C.3 RESULTS

4C.3.1 TRANSCRIPTOMIC ANALYSIS Using the whole genome DNA chip, 6.1 % of the genes (out of 3,819 genes retained) were identified as significantly up-regulated after 10 days culturing on the RPM in Sistrom- succinate medium compared to the 1 g gravity control. While, 0.3 % of the genes (out of 3,824 genes retained) were identified as significantly up-regulated after 10 days culturing in LSMMG compared to the 1 g gravity control. No genes were found to be significantly down- regulated neither during RPM nor during vertical RWV culturing. A circular plot of the genes differentially expressed in the chromosome of R. rubrum S1H showed that these genes were distributed globally throughout the chromosome (Figure 4c.2). All the 13 up-regulated genes that were found in the RWV culture were also found to be up- regulated in the RPM experiment (Table 4c.2). RruA3396 is the most induced gene (19.91 fold) in RPM and the second most induced gene for LSMMG (4.01 fold). RruA3397, transcriptionally coupled to RruA3396, was respectively the fourth and the third most up- regulated gene in RPM and LSMMG. While the high fold inductions were associated to excellent p-values, these induction trends could not be confirmed by RT-qPCR for RruA1518 (0.14 ± 0.00 for RPM and 1.02 ± 0.02 for LSMMG) and for RruA3396 (0.18 ± 0.00 for RPM and 1.29 ± 0.00 for LSMMG).

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 124

(i) (ii)

(iii) (iv) (v)

Rhodospirillum rubrum S1 Chromosome 4.35 Mb

Figure 4c.2 . Map of the 4.35-Mb circular R. rubrum S1 chromosome. From the outside to the inside: (i) genes on direct strand, (ii) genes on reverse strand, (iii) genes up-regulated (in red color) during RPM cultivation and (iv) during vertical RWV cultivation, (v) gene position (x 10,000 bp).

The RruA3396 gene encodes for a N-acyl-L-homoserine lactone synthase and catalyzes the synthesis of the N-acyl-L-homoserine lactone (AHL) signalling molecules. These low molecular-mass molecules are used in bacterial communication commonly referred to as 'quorum sensing'. This relies on the principle that when a single bacterium produces small signalling molecules, the extracellular concentration is below a certain threshold, while a critical concentration can be reached when the cell density increases, allowing the signalling molecule to be sensed and enabling the bacteria to respond. The classical AHL-dependent quorum sensing systems rely on two fundamental proteins, the I protein (RruA3396) homologous to LuxI and the R protein (RruA3395) homologous to LuxR of Vibrio fischeri, where the system was first described (Nealson, 1979). While the I protein catalyzes the synthesis of the AHLs, the R protein is a transcriptional regulator that triggers

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 125 the responses (Reading and Sperandio, 2006). RruA3395 was only significantly up-regulated during RPM cultivation (1.4 fold change, p-value of 8.76x10 -03 ). Production of AHLs by R. rubrum S1H cultured in Sistrom-succinate in light anaerobic conditions has been shown using thin layer chromatography analysis combined with the reporter strain Chromobacterium violaceum CV026. In addition, the produced AHLs were characterized in collaboration with Paul Williams (University of Nottingham, UK) and indicated the presence of 8 different AHL molecules namely: hexanoyl-, octanoyl-, and decanoyl-homoserine lactone, 3-oxo-octanoyl-homoserine lactone, 3-hydroxy-hexanoyl-, 3- hydroxy-octanoyl-, 3-hydroxy-decanoyl-, and 3-hydroxy-dodecanoyl-homoserine lactone (Van Houdt, unpublished results). Considering all the significantly up-regulated genes from the RPM experiment, almost 1/3 appeared to encode for hypothetical proteins (Table 4c.3, Additional table 4c.1). For both the RPM and RWV experiment, the 'unclassified' (Un.) COG category ranked as the most numerically abundant functional category of genes whose expression was significantly changed under simulated microgravity with respectively 58 genes (including 41 hypothetical) up-regulated in RPM and 4 hypothetical protein encoding genes up-regulated in RWV (Figure 4c.3).

Table 4c.2 . Common genes between the RPM and the LSMMG experiment. Gene Gene RPM LSMMG Product name COG number name FC FC ribosomal subunit interface protein, RruA0073 - J 5.68 2.11 putative RruA0160 - hypothetical protein Rru_A0160 E 2.22 2.36 RruA1072 rpmE 50S ribosomal protein L31 J 15.29 4.20 RruA1518 - hypothetical protein Rru_A1518 Un. 11.93 2.65 5,10-methylenetetrahydrofolate RruA1530 - E 12.38 2.71 reductase RruA1758 - hypothetical protein Rru_A1758 Un. 10.17 2.96 RruA1947 - nucleotidyltransferase-like R 14.17 2.41 RruA2850 cheL hypothetical protein Rru_A2850 M 10.21 2.11 RruA3394 - hypothetical protein Rru_A3394 Un. 8.63 2.18 RruA3396 - autoinducer synthesis protein Q 19.91 4.01 RruA3397 - hypothetical protein Rru_A3397 Un. 14.21 3.16 RruA3689 - MaoC-like dehydratase I 12.47 2.09 RruB0012 - DNA polymerase, beta-like region R 9.28 2.21 'RruA' refers to gene located on the chromosome while 'RruB' refers to gene located on the plasmid. FC : fold change (p-value < 0.05). COG nomenclature: E - Amino acid transport and metabolism, M - Cell envelope biogenesis, outer membrane, R - General function prediction only, I - Lipid metabolism, Q - Secondary metabolites biosynthesis, transport and catabolism, J - Translation, ribosomal structure and biogenesis, Un. – Unclassified.

Table 4c.3 . Significant gene distribution in regards to the fold induction rate in the RPM experiment. RPM culture condition Total number of genes Number of hypothetical genes 2 ≤ fold change ≤ 5 192 64 5 ≤ fold change ≤ 10 28 9 Fold change > 10 15 6

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 126 70

60

50

40 RPM liquid UP RWV liquid UP Number Number of genes 30

20

10

0 EGMNHVLCSRPU I FOQTKJUn. Functional category (COG)

Figure 4c.3 . Functional classification of the 235 and 13 significantly differentially expressed genes that respond respectively to RPM and vertical RWV culture conditions. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

When focusing on the RPM samples, 22 genes up-regulated during RPM incubation belonged to the functional category related to transcription (K), including 17 transcriptional regulators and the main sigma factor rpoD (Table 4c.4). Besides, the functional category (R) comprising the genes with 'general function prediction only' retrieved 20 up-regulated genes including the translational regulator hfq and 6 genes coding for hypothetical proteins (Table 4c.4). Next in order was class S with 18 genes with unkown function, for which 11 genes coding for hypothetical proteins with RruA0086 more than 10 fold over-expressed (Table 4c.4). Among the 7 genes left, we can cite a bacteriocin Linocin M18 (RruA0974) and a phasin (RruA3283) being slightly up-regulated. 'Amino acid transport and metabolism' was the fifth most represented functional category including 15 genes. Among these, the 5,10- methylenetetrahydrofolate reductase (RruA1530) was more than 10 fold over-expressed in the RPM culture condition while 3 hypothetical protein encoding genes were also up-regulated (Table 4c.4). Finaly, we also observed differential expression for the pucC gene being part of the puc operon coding for two light-harversting antenna subunits; for 4 genes coding for CRISPR-

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 127 associated (Cas) proteins related to phage defense system (RruA0170, RruA0178, RruA0179 and RruA0345) (Barrangou et al ., 2007; Brouns et al ., 2008); for a mechanosensitive ion channel related gene mscS (RruA1608); for the polyhydroxyalkanoate synthase (RruA1816); for the superoxide dismutase sodA (RruA1760), for a bacterioferritin bfr (RruA2195) and for the ferric uptake-related repressor fur (RruA3788) (Table 4c.5).

Table 4c.4 . Most numerically abundant categories of genes over-expressed in RPM samples. Genes Gene RPM liquid RPM liquid Product name COG number name FC p-value RruA0079 - CarD family transcriptional regulator K 3.09 1.96E-06 RruA0663 - cold-shock DNA-binding protein family protein K 8.33 1.09E-08 RruA0700 - transcription regulator K 2.15 1.63E-04 RruA1116 - TetR family transcriptional regulator K 3.95 1.44E-08 RruA1347 - GntR family transcriptional regulator K 4.03 5.22E-07 RruA1521 cspC cold-shock DNA-binding protein family protein K 2.69 6.00E-07 RruA1714 - LysR family transcriptional regulator K 3.89 1.09E-08 RruA1717 - ArsR family transcriptional regulator K 4.94 1.81E-09 RruA1829 - MucR family transcriptional regulator K 4.11 3.56E-09 RruA1914 cspC cold-shock DNA-binding protein family protein K 3.92 3.61E-09 RruA1951 - AraC family transcriptional regulator K 7.67 1.02E-08 RruA1961 - TetR family transcriptional regulator K 2.76 3.18E-05 RruA1975 - XRE family transcriptional regulator K 3.63 6.82E-09 RruA2233 - transcriptional regulatory protein K 2.70 2.23E-06 RruA2308 - ArsR family transcriptional regulator K 4.86 8.81E-10 RruA2574 - BadM/Rrf2 family transcriptional regulator K 2.29 4.23E-07 RruA2695 rpoB DNA-directed RNA polymerase subunit beta K 4.71 1.52E-07 response regulator receiver domain-containing RruA2837 ompR K 2.04 8.53E-04 protein RruA2882 rpoD RNA polymerase sigma factor RpoD K 2.78 1.82E-08 RruA2929 - two component transcriptional regulator K 2.83 9.15E-06 RruA3701 - TetR family transcriptional regulator K 2.12 1.50E-06 RruB0041 - XRE family transcriptional regulator K 3.24 3.15E-07

RruA0086 - hypothetical protein Rru_A0086 S 10.41 9.64E-12 RruA0637 - hypothetical protein Rru_A0637 S 3.17 9.20E-08 RruA0974 - Linocin_M18 bacteriocin protein S 2.63 5.35E-08 RruA1175 - CsbD-like S 2.91 1.29E-05 RruA1236 - Iojap-related protein S 5.08 1.09E-08 RruA1353 - hypothetical protein Rru_A1353 S 2.68 1.88E-06 RruA1355 - hypothetical protein Rru_A1355 S 2.94 5.66E-07 RruA1368 - hypothetical protein Rru_A1368 S 2.78 4.96E-06 RruA1648 - helix-turn-helix protein, CopG S 2.00 1.83E-05 RruA1663 - ubiquinol-cytochrome C chaperone S 2.50 9.04E-05 RruA1733 - hypothetical protein Rru_A1733 S 2.43 1.28E-06 RruA1991 - hypothetical protein Rru_A1991 S 2.80 3.41E-06 RruA2205 - hypothetical protein Rru_A2205 S 3.57 4.20E-08 RruA2399 - hypothetical protein Rru_A2399 S 2.08 3.14E-03 RruA2568 - HesB/YadR/YfhF S 6.48 4.89E-08 RruA2736 - hypothetical protein Rru_A2736 S 3.00 9.38E-08 RruA3279 - hypothetical protein Rru_A3279 S 2.76 8.10E-06 RruA3283 - phasin S 2.52 4.33E-06

RruA0256 - GCN5-related N-acetyltransferase R 2.80 5.51E-06 RruA0551 - TPR domain-containing protein R 4.50 2.26E-12 RruA0699 - plasmid stabilization system protein R 2.19 1.05E-04 RruA0729 - Short-chain dehydrogenase/reductase SDR R 10.28 5.83E-11 RruA0968 - ATPase R 2.86 4.43E-05 RruA1443 - hypothetical protein Rru_A1443 R 2.61 2.70E-06 RruA1501 - transport-associated protein R 2.28 9.82E-07 RruA1667 - 3-oxoacyl- R 3.31 2.73E-07 RruA1684 hfq RNA-binding protein Hfq R 2.05 3.31E-04 RruA1945 - Short-chain dehydrogenase/reductase SDR R 3.64 2.51E-06 RruA1947 - nucleotidyltransferase-like R 14.17 3.08E-10 RruA2193 aarF Abc1 protein R 2.33 1.44E-05 RruA2787 - inner-membrane translocator R 3.03 4.79E-09 RruA2972 - peptidase M16-like R 2.09 9.54E-04 RruA3354 - hypothetical protein Rru_A3354 R 2.23 1.11E-04 RruA3392 - hypothetical protein Rru_A3392 R 2.46 7.51E-06 RruA3520 - hypothetical protein Rru_A3520 R 4.09 3.82E-07 RruA3558 - hypothetical protein Rru_A3558 R 3.49 1.35E-07

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 128

Genes Gene RPM liquid RPM liquid Product name COG number name FC p-value RruA3772 - hypothetical protein Rru_A3772 R 2.05 2.94E-04 RruB0012 - DNA polymerase, beta-like region R 9.28 1.66E-09

RruA0160 - hypothetical protein Rru_A0160 E 2.22 2.87E-07 RruA0176 - acetylornithine aminotransferase E 7.40 1.27E-09 RruA0230 asnB asparagine synthase, glutamine-hydrolyzing E 7.82 1.76E-07 RruA0910 - ethanolamine utilization protein eutJ E 4.98 1.00E-06 binding-protein dependent transport system inner RruA1016 - E 3.62 2.94E-08 membrane protein RruA1132 - aminodeoxychorismate synthase, subunit I E 2.42 1.88E-06 RruA1530 - 5,10-methylenetetrahydrofolate reductase E 12.38 8.29E-13 RruA1546 - FAD dependent oxidoreductase E 2.06 5.40E-04 RruA2289 - sulfate adenylyltransferase subunit 2 E 3.13 7.51E-05 glycine betaine/L-proline transport ATP-binding RruA2475 - E 2.12 5.39E-05 subunit RruA2623 - chorismate synthase E 2.22 3.49E-04 RruA2753 - transglutaminase-like E 2.07 2.59E-03 RruA2777 - hypothetical protein Rru_A2777 E 4.81 1.35E-07 RruA3004 - aminotransferase, class I and II E 6.10 1.59E-07 RruA3160 rhaT hypothetical protein Rru_A3160 E 2.19 3.68E-07 'RruA' refers to gene located on the chromosome while 'RruB' refers to gene located on the plasmid. FC : fold change. COG nomenclature: E - Amino acid transport and metabolism, S - Function unknown, R - General function prediction only, K – Transcription.

Table 4c.5 . Selected significant genes from the RPM experiment. Gene Gene RPM RPM Product name COG number name FC p-value RruA0170 - CRISPR-associated Cse3 family protein Un. 2.31 7.89E-05 RruA0178 - CRISPR-associated Cas2 family protein L 4.47 5.79E-06 RruA0179 - CRISPR-associated Cas1/Cas4 family protein L 3.74 1.07E-06 RruA0345 - CRISPR-associated Cas5e family protein Un. 3.76 9.34E-07 RruA0618 pucC PucC protein G 2.65 1.29E-07 RruA1608 mscS MscS mechanosensitive ion channel M 4.34 1.40E-06 RruA1760 sodA superoxide dismutase P 4.27 9.48E-10 RruA1816 - Poly(R)-hydroxyalkanoic acid synthase, class I I 3.99 9.30E-07 RruA2195 bfr bacterioferritin P 3.34 3.48E-09 RruA3788 fur ferric uptake regulator family protein P 7.59 4.14E-08 'RruA' refers to gene located on the chromosome. FC : fold change. COG nomenclature: G - Carbohydrate transport and metabolism, M - Cell envelope biogenesis, outer membrane, L - DNA replication, recombination and repair, P - Inorganic ion transport and metabolism, I - Lipid metabolism, Un . – unclassified.

4C.3.2 PROTEOME ANALYSIS

4C.3.2.1 RANDOM POSITIONING MACHINE (RPM) EXPERIMENT

For the RPM experiment, the computed FDR was 0.43 % estimated at the peptide level. From these peptides, 451 proteins were identified including 301 proteins quantified with at least 2 peptides (Additional table 4c.2). This represents respectively 12 % and 8 % of the total candidate protein-encoding genes. In addition, the translation of 19 hypothetical proteins has been shown (Addition table 4c.2). The median value of the 301 quantified proteins was 0.91 (Figure 4c.4). This meant we did not have a H/L ratio of 1:1 (probably because of measurement errors in the protein quantification). Therefore the threshold of significance was adapted by multiplying it by the correction factor (0.91), giving 1.37 and 0.64, rendering

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 129 respectively 16 and 17 proteins that passed the significance threshold for up- and down- regulation. Among the identified proteins, 42 proteins, of which 31 were quantified, were also differentially expressed at transcriptional level as indicated by the microarray analysis (Table 4c.6).

4

3.5

3

2.5

2

foldchange 1.5

1 0 50 100 150 200 250 300 350

0.5

ratio RPM liquid/control 0 Protein number Figure 4c.4 . Fold change distribution of the proteins quantified by the differential proteomic approach in the RPM experiment.

4c.3.2.1.1 Correlation between microarray and 2D-LC-MS/MS analyses Considering the threshold of significance for the 31 quantified proteins, the up-regulated aldehyde dehydrogenase (RruA0931) and bacterioferritin (RruA2195) showed direct correlation with the microarray data while none of the related down-regulated proteins passed that threshold (Table 4c.6).

Table 4c.6 . Transcriptomics versus differential proteomics approach for the RPM experiment. RPM Gene Gene RPM Product name COG FC number name FC Protein mRNA RruA0073 - ribosomal subunit interface protein, putative J 5.68 0.78 ± 0.07 RruA0210 rpmB 50S ribosomal protein L28 J 2.19 0.97 ± 0.04 RruA0437 ompA OmpA family protein M 4.43 0.94 ± 0.07 RruA0464 - hypothetical protein Rru_A0464 Un. 3.91 IBNQ RruA0521 cheA CheA Signal transduction histidine Kinases (STHK) N 2.06 0.89 ± 0.09 RruA0596 fba fructose-1,6-bisphosphate aldolase G 2.20 1.00 ± 0.08 RruA0662 infA translation initiation factor 1 J 7.66 1.05 ± 0.09 RruA0707 - BolA-like protein T 2.46 0.84 ± 0.04 RruA0931 putA aldehyde dehydrogenase C 3.53 2.39 ± 0.33 RruA0974 - Linocin_M18 bacteriocin protein S 2.63 1.05 ± 0.10 RruA1072 rpmE 50S ribosomal protein L31 J 15.29 1.02 ± 0.05 RruA1211 sucC succinyl-CoA synthetase (ADP-forming) beta subunit C 2.04 1.05 ± 0.10

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 130

RPM Gene Gene RPM Product name COG FC number name FC Protein mRNA RruA1224 atpA ATP synthase F1, alpha subunit C 4.18 0.91 ± 0.05 RruA1340 dak1 Glycerone kinase G 2.55 0.86 ± 0.12 RruA1353 - hypothetical protein Rru_A1353 S 2.68 0.77 ± 0.06 RruA1365 rbsB ribose ABC transporter, periplasmic binding protein G 2.18 IBNQ RruA1521 cspC cold-shock DNA-binding protein family protein K 2.69 0.93 ± 0.10 RruA1537 - hypothetical protein Rru_A1537 Un. 15.13 IBNQ RruA1557 nuoC NADH dehydrogenase subunit C C 2.64 IBNQ RruA1717 - ArsR family transcriptional regulator K 4.94 IBNQ RruA1731 - hypothetical protein Rru_A1731 Un. 2.97 0.88 ± 0.05 RruA1760 sodA superoxide dismutase P 4.27 IBNQ RruA1829 - MucR family transcriptional regulator K 4.11 0.99 ± 0.13 RruA1914 cspC cold-shock DNA-binding protein family protein K 3.92 1.23 ± 0.01 RruA2193 aarF Abc1 protein R 2.33 IBNQ RruA2195 bfr bacterioferritin P 3.34 1.39 ± 0.12 RruA2666 rpsM 30S ribosomal protein S13 J 2.31 0.74 ± 0.07 RruA2678 rplN 50S ribosomal protein L14P J 2.54 0.91 ± 0.03 RruA2679 rpsQ SSU ribosomal protein S17P J 6.22 0.91 ± 0.04 RruA2687 rplD 50S ribosomal protein L4P J 10.33 0.89 ± 0.10 RruA2688 rplC 50S ribosomal protein L3 J 4.92 0.90 ± 0.08 RruA2695 rpoB DNA-directed RNA polymerase subunit beta K 4.71 0.84 ± 0.06 RruA2837 ompR response regulator receiver domain-containing protein K 2.04 IBNQ RruA2882 rpoD RNA polymerase sigma factor RpoD K 2.78 0.76 ± 0.11 RruA2936 - dihydroorotase F 2.09 IBNQ RruA3205 rpmG 50S ribosomal protein L33P J 4.10 0.96 ± 0.09 RruA3279 - hypothetical protein Rru_A3279 S 2.76 IBNQ RruA3283 - phasin S 2.52 0.70 ± 0.01 RruA3320 - pantothenate synthetase H 2.37 0.87 ± 0.06 RruA3419 pckA phosphoenolpyruvate carboxykinase C 2.28 0.91 ± 0.07 RruA3521 - 50S ribosomal protein L35P J 12.81 IBNQ RruA3755 - heat shock protein Hsp20 O 2.66 IBNQ 'RruA' refers to gene located on the chromosome. FC : fold change (p-value < 0.05). IBNQ : identified but not quantified. Proteins are quantified by mean ± standard deviation. COG nomenclature: G - Carbohydrate transport and metabolism, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

4c.3.2.1.2 Genes showing significant expression changes by proteomic analysis but not by microarray analysis Of the 301 quantified proteins, 270 showed no significant differences in mRNA levels by microarray hybridization. The 2 most significant over-expressed proteins (out of 16) were RruA0617 and RruA2974, related to the photosynthetic apparatus, with more than 3 fold change in expression while the cytochrome C protein (RruA1020) had a fold change above 2 (Table 4c.7). Within the 17 significant down-regulated proteins, we could mention the following chaperonins GroEL, ClpA and GrpE and the tellurium stress related protein TerZ (Table 4c.7).

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 131

Table 4c.7 . Significant proteins showing no differential expression at the transcriptomic level in the RPM liquid culture conditions. Gene Gene Mascot MW Product name COG FC RPM # (H/L) number name score (kDa) RruA0162 groEL chaperonin GroEL O 0.64 ± 0.04 11 727.98 57.07 RruA0271 folE GTP cyclohydrolase I H 0.57 ± 0.05 5 211.69 25.91 RruA0273 fabG 3-oxoacyl- Q 0.56 ± 0.05 3 168.30 25.88 RruA0414 rpsF SSU ribosomal protein S6P J 1.39 ± 0.17 5 484.44 18.18 NADH:flavin oxidoreductase/NADH RruA0475 - C 0.15 ± 0.05 4 476.40 38.68 oxidase RruA0500 - extracellular solute-binding protein P 1.39 ± 0.09 2 78.57 38.81 RruA0617 puhA photosynthetic reaction center, H-chain Un. 3.43 ± 0.23 9 866.16 27.92 RruA0893 terZ stress protein T 0.54 ± 0.02 19 1274.21 20.46 RruA0930 eutG Iron-containing alcohol dehydrogenase C 3.04 ± 0.47 7 345.62 39.91 lysine-arginine-ornithine-binding RruA1004 hisJ E 1.37 ± 0.09 54 4622.66 36.62 periplasmic protein RruA1020 - cytochrome c, class I C 2.25 ± 0.22 7 148.32 14.50 RruA1096 - hypothetical protein Rru_A1096 S 0.57 ± 0.08 2 97.71 29.28 RruA1101 smc hypothetical protein Rru_A1101 D 0.51 ± 0.05 5 548.39 72.50 RruA1262 ugpB extracellular solute-binding protein G 0.63 ± 0.10 11 307.46 46.98 trimethylamine-N-oxide reductase RruA1287 - C 1.91 ± 0.06 3 160.33 89.80 (cytochrome c) RruA1362 - xylose isomerase-like TIM barrel G 0.43 ± 0.04 3 291.78 33.69 RruA1589 frr ribosome recycling factor J 0.55 ± 0.07 4 178.54 21.14 RruA1746 livK extracellular ligand-binding receptor E 1.79 ± 0.16 10 411.89 38.72 RruA1827 glyA serine hydroxymethyltransferase E 0.56 ± 0.08 4 151.94 45.49 RruA1885 eno enolase G 1.46 ± 0.23 11 414.88 44.97 RruA1928 - 2-isopropylmalate synthase E 0.53 ± 0.03 3 132.75 61.45 RruA2014 ddpA twin-arginine translocation pathway signal E 1.44 ± 0.23 7 259.77 59.42 methylmalonate-semialdehyde RruA2071 putA C 1.40 ± 0.09 6 301.74 53.49 dehydrogenase RruA2083 creA CreA S 0.63 ± 0.09 6 174.37 17.12 ATP-dependent Clp protease ATP-binding RruA2109 clpA O 0.46 ± 0.08 2 108.70 86.55 subunit ClpA TRAP dicarboxylate transporter- DctP RruA2239 - G 1.44 ± 0.02 4 122.81 36.74 subunit RruA2974 pufM photosynthetic reaction centre M subunit Un. 3.34 ± 0.38 4 239.87 34.24 RruA3511 acnA aconitate hydratase C 1.42 ± 0.18 14 1154.30 96.13 ATP-dependent protease ATP-binding RruA3600 hslU O 0.62 ± 0.04 2 66.96 47.95 subunit RruA3643 grpE GrpE protein O 0.48 ± 0.04 5 138.37 23.82 2-octaprenyl-3-methyl-6-methoxy-1,4- RruA3707 ubiH benzoquinol hydroxylase / 2-octaprenyl-6- C 0.60 ± 0.06 2 67.07 45.82 methoxyphenol hydroxylase 'RruA' refers to gene located on the chromosome. FC : fold change. Proteins are quantified by mean ± standard deviation. MW: molecular weight. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, C - Energy production and conversion, S - Function unknown, P - Inorganic ion transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

4C.3.2.2 THE ROTATING WALL VESSEL (RWV) EXPERIMENT

For the RWV experiment, the computed FDR was 1.38 % estimated at the peptide level. From these peptides, 422 proteins were identified including 273 proteins quantified with at least 2 peptides (Additional table 4c.3). This represents respectively 11 % and 7 % of the total candidate protein-encoding genes. In addition, the translation of 41 hypothetical proteins has been shown (Additional table 4c.3). The median value of the fold change distribution was 0.77 indicating a notable bias in the protein mix (Figure 4c.5). Therefore, the significance

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 132 thresholds were adapted as described above for the RPM samples. The corrected thresholds were 1.16 and 0.54.

3.50

3.00

2.50

2.00

1.50 Foldchange

1.00 0 50 100 150 200 250 300

0.50

ratio RWV liquid/control 0.00 Protein number Figure 4c.5 . Fold change distribution of the proteins quantified by the differential proteomic approach in the RWV experiment.

Among the identified proteins, only 2 were also differentially expressed at transcriptional level as indicate by the microarray analysis and showed inverse correlation between proteome and transcriptome expresssion (Table 4c.8).

Table 4c.8. Transcriptomics versus differential proteomics approach for the RWV experiment. RWV Gene Gene RWV Product name COG FC number name FC Protein mRNA RruA0073 - ribosomal subunit interface protein, putative J 2.11 0.57 ± 0.08 RruA1072 rpmE 50S ribosomal protein L31 J 4.20 0.64 ± 0.07 'RruA' refers to gene located on the chromosome. FC : fold change (p-value < 0.05). Proteins are quantified by mean ± standard deviation. COG nomenclature: J - Translation, ribosomal structure and biogenesis.

4c.3.2.2.1 Genes showing significant expression changes by proteomic analysis but not by microarray analysis Of the 273 quantified proteins, 271 showed no significant differences in mRNA levels by microarray hybridization. Respectively, 27 and 43 proteins were significantly up- and down- regulated (Table 4c.10). The 2 most over-expressed proteins were RruA2974 and RruA0617, related to the photosynthetic apparatus, similar to what was observed for RPM cultivation. In addition, after 10 days of cultivation (in dark aerobic conditions) an intenser coloration of the

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 133 LSMMG cultures compared to the control cultures could be observed without notable changes in cell density and oxygenation (Figure 4c.6 and table 4c.9).

Low-shear LSMMG Control

Culture condition OD 600 viable count (CFU/ml) O2 mg/l Control low-shear 1.00 ± 0.04 2.16E+07 ± 1.12E+07 7.08 ± 0.20 LSMMG 0.93 ± 0.03 2.37E+07 ± 1.00E+07 6.97 ± 0.20

Figure 4c.6 and table 4c.9. Pellet and culture features after 10 days culture in control low-shear and low-shear modeled microgravity ( LSMMG ) using the Rotating Wall Vessel technology (mean value ± standard deviation of at least 3 biological replicates). OD : optical density. CFU : colony forming unit. O 2 saturation = 8.46 mg/l.

Table 4c.10 . Genes showing significant expression changes by proteomic analysis but not by microarray analysis in the RWV experiment. Gene Gene Mascot MW Product name COG FC RWV # (H/L) number name score (kDa) RruA0210* rpmB 50S ribosomal protein L28 J 0.53 ± 0.01 3 343.40 11.06 RruA0217* citE citrate lyase G 0.54 ± 0.05 4 514.04 35.09 periplasmic binding protein/LacI transcriptional RruA0247* rbsB G 0.54 ± 0.08 5 166.25 35.83 regulator RruA0332* gst glutathione S-transferase-like protein O 0.50 ± 0.04 4 359.65 25.43 bifunctional GMP synthase/glutamine RruA0355* guaA F 1.32 ± 0.05 2 68.03 57.34 amidotransferase protein RruA0437* ompA OmpA family protein M 2.69 ± 0.33 9 544.13 35.60 RruA0595* tktA transketolase G 1.35 ± 0.16 6 540.13 70.01 RruA0617 puhA photosynthetic reaction center, H-chain Un. 2.82 ± 0.10 6 367.29 27.92 RruA0662* infA translation initiation factor 1 J 0.53 ± 0.01 3 99.05 10.73 RruA0701* rpsD 30S ribosomal protein S4 J 0.45 ± 0.03 5 124.98 23.62 RruA0930 eutG Iron-containing alcohol dehydrogenase C 1.72 ± 0.18 2 180.81 39.91 RruA0931 putA aldehyde dehydrogenase C 1.88 ± 0.06 12 591.66 55.73 RruA1043* rpsU 30S ribosomal protein S21 J 0.41 ± 0.02 5 108.05 9.42 RruA1095* ompA OmpA/MotB M 2.05 ± 0.04 3 545.42 17.29 RruA1096 - hypothetical protein Rru_A1096 S 0.49 ± 0.04 4 149.08 29.28 RruA1185* rpsP 30S ribosomal protein S16 J 0.51 ± 0.08 9 323.97 13.90 RruA1203* sdhD succinate dehydrogenase subunit D C 1.19 ± 0.05 4 184.34 13.59 RruA1214* aceF 2-oxoglutarate dehydrogenase E2 component C 0.54 ± 0.03 6 522.26 45.30 RruA1218* pspA phage shock protein A, PspA K 1.38 ± 0.11 2 251.05 25.26 RruA1223* atpH F0F1 ATP synthase subunit delta C 0.35 ± 0.14 6 412.35 19.56 RruA1232* prc C-terminal processing peptidase S41A M 0.50 ± 0.08 6 206.48 47.87 RruA1309* fadB 3-hydroxyacyl-CoA dehydrogenase I 1.30 ± 0.09 12 1031.11 83.07 RruA1310* paaJ Acetyl-CoA C-acyltransferase I 1.44 ± 0.10 9 790.75 39.28 RruA1312* - AMP-dependent synthetase and ligase I 1.28 ± 0.05 3 279.15 61.88 RruA1313* fadL aromatic hydrocarbon degradation protein I 1.24 ± 0.06 14 748.31 45.34 RruA1316* caiC AMP-dependent synthetase and ligase I 1.16 ± 0.07 3 412.42 58.79 RruA1353* - hypothetical protein Rru_A1353 S 0.47 ± 0.07 5 129.49 18.19 RruA1362 - xylose isomerase-like TIM barrel G 0.39 ± 0.02 3 163.94 33.69

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 134

Gene Gene Mascot MW Product name COG FC RWV # (H/L) number name score (kDa) Alkyl hydroperoxide reductase/ Thiol specific RruA1418* ahpC O 0.54 ± 0.01 3 288.22 20.69 antioxidant/ Mal allergen RruA1589 frr ribosome recycling factor J 0.39 ± 0.06 5 149.48 21.14 RruA1595* hlpA Outer membrane chaperone Skp (OmpH) M 0.51 ± 0.02 5 263.24 21.72 RruA1665* rpmF 50S ribosomal protein L32 J 0.39 ± 0.05 3 91.77 7.03 RruA1760* sodA superoxide dismutase P 0.46 ± 0.05 4 192.10 24.93 RruA1765* yajC protein translocase subunit yajC U 1.42 ± 0.04 3 103.72 15.60 RruA1798* tolC Type I secretion outer membrane protein, TolC M 1.39 ± 0.06 2 146.30 59.16 RruA1829* - MucR family transcriptional regulator K 0.52 ± 0.03 3 186.36 15.71 RruA2041* bioA aminotransferase H 0.51 ± 0.02 3 240.83 49.34 RruA2087* hisJ extracellular solute-binding protein E 0.49 ± 0.04 4 235.92 32.55 RruA2152* rplM 50S ribosomal protein L13 J 0.54 ± 0.05 8 256.75 17.10 RruA2171* - extracellular ligand-binding receptor E 0.54 ± 0.05 5 422.88 38.63 RruA2193* aarF Abc1 protein R 1.19 ± 0.17 2 185.61 51.24 RruA2211* ompC porin M 1.64 ± 0.11 8 707.05 37.49 RruA2356* ddpA extracellular solute-binding protein E 0.50 ± 0.02 20 1511.78 59.24 RruA2575* - hypothetical protein Rru_A2575 S 0.52 ± 0.03 5 116.49 14.27 RruA2663* rplQ 50S ribosomal protein L17P J 0.54 ± 0.02 7 419.37 15.33 RruA2677* rplX 50S ribosomal protein L24P J 0.53 ± 0.04 8 368.91 11.29 RruA2681* rplP 50S ribosomal protein L16 J 0.50 ± 0.09 5 232.86 15.43 RruA2683* rplV 50S ribosomal protein L22 J 0.52 ± 0.04 12 427.70 14.14 RruA2689* rpsJ 30S ribosomal protein S10 J 0.52 ± 0.00 2 81.74 11.69 RruA2691* fusA translation elongation factor 2 (EF-2/EF-G) J 1.16 ± 0.08 7 513.97 76.45 RruA2694* rpoC DNA-directed RNA polymerase K 1.20 ± 0.07 15 641.63 155.85 RruA2723* - rubrerythrin S 0.39 ± 0.05 2 121.47 19.25 RruA2956* ppsA pyruvate phosphate dikinase G 1.18 ± 0.03 5 268.94 97.15 RruA2964* maoC MaoC-like dehydratase I 0.47 ± 0.08 4 195.34 15.40 RruA2974 pufM photosynthetic reaction centre M subunit Un. 3.00 ± 0.10 4 225.76 34.24 RruA3205* rpmG 50S ribosomal protein L33P J 0.49 ± 0.03 5 404.82 6.49 RruA3283* - phasin S 0.35 ± 0.01 2 69.99 17.53 RruA3328* ompA OmpA/MotB M 2.52 ± 0.13 5 322.98 28.87 RruA3373* - hypothetical protein Rru_A3373 - 0.52 ± 0.03 6 149.49 18.70 RruA3400* sbp thiosulphate-binding protein P 0.54 ± 0.01 3 210.14 38.52 putative branched-chain amino acid transport RruA3506* livK E 0.50 ± 0.00 2 130.80 44.51 system substrate-binding protein magnesium-protoporphyrin IX monomethyl ester RruA3548* - C 1.52 ± 0.11 8 215.04 61.72 anaerobic oxidative cyclase RruA3566* - HemY-like S 1.49 ± 0.04 2 141.93 46.63 RruA3662* - hypothetical protein Rru_A3662 Un. 0.50 ± 0.02 2 100.71 12.62 RruA3728* livK extracellular ligand-binding receptor E 0.41 ± 0.05 5 173.97 40.22 RruA3744* - signal transduction protein T 0.47 ± 0.05 4 291.53 16.39 RruA3782* rbfA ribosome-binding factor A J 0.44 ± 0.03 2 70.89 17.84 RruA3785* pnp polynucleotide phosphorylase/polyadenylase J 1.33 ± 0.07 7 316.05 76.40 RruA3794* - cysteine synthase E 1.40 ± 0.03 2 216.92 35.49 RruA3802* rpsT SSU ribosomal protein S20P J 0.49 ± 0.04 5 523.38 9.57 'RruA' refers to gene located on the chromosome. FC : fold change. Proteins are quantified by mean ± standard deviation. The " *" refers to proteins showing significant expression changes only in the RWV experiments and not in the RPM experiment (see below). COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Beside, we could also mention the up-regulation of 4 outer membrane proteins (RruA0437, RruA1095, RruA2211 and RruA3328) and the phage shock protein PspA (RruA1218) (Table 4c.10). The 'translation, ribosomal structure and biogenesis' (J) functional category grouped 18 of the 43 down-regulated proteins. This category contained a vast majority of ribosomal proteins (Table 4c.10).

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 135

4C.3.2.3 COMPARISON OF RPM AND RWV AT THE PROTEOMIC LEVEL

In contrast with the transcriptomic approach, not all significant differentially expressed proteins during RWV cultivation were also found during RPM cultivation (Figure 4c.7).

RPM RWV

26 7 63

Figure 4c.7 . Significant differentially expressed proteins in both the RPM and the RWV experiments.

4C.3.2.4 COMMON QUANTIFIED PROTEINS BETWEEN THE RPM AND THE RWV EXPERIMENT

All of the 7 common proteins showed same trend in fold change including the top over- expressed RruA0617 and RruA2974 as mentioned above (Table 4c.11). Among the down-regulated proteins, the hypothetical protein encoding gene RruA1096 posses a Tol-Pal system domain that has been shown to be necessary for maintaining the outer membrane integrity of Gram-negative bacteria (Dubuisson et al ., 2005). The down-regulation of the ribosome recycling factor encoding gene (RruA1589) could indicate slowdown in growth in regard to the control conditions.

Table 4c.11 . Common significant proteins between the RPM and the RWV experiment. Gene Gene Product name COG FC RPM FC RWV number name RruA0617 puhA photosynthetic reaction center, H-chain Un. 3.43 ± 0.23 2.82 ± 0.10 RruA0930 eutG Iron-containing alcohol dehydrogenase C 3.04 ± 0.47 1.72 ± 0.18 RruA0931 putA aldehyde dehydrogenase C 2.39 ± 0.33 1.88 ± 0.06 RruA1096 - hypothetical protein Rru_A1096 S 0.57 ± 0.08 0.49 ± 0.04 RruA1362 - xylose isomerase-like TIM barrel G 0.43 ± 0.04 0.39 ± 0.02 RruA1589 frr ribosome recycling factor J 0.55 ± 0.07 0.39 ± 0.06 RruA2974 pufM photosynthetic reaction centre M subunit Un. 3.34 ± 0.38 3.00 ± 0.10 'RruA' refers to gene located on the chromosome. FC : fold change. Proteins are quantified by mean ± standard deviation. COG nomenclature: G - Carbohydrate transport and metabolism, C - Energy production and conversion, S - Function unknown, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 136

4C.3.2.5 PROTEINS SHOWING SIGNIFICANT EXPRESSION CHANGES IN THE RPM EXPERIMENT

ONLY

For the 26 proteins only significant in the RPM experiment, 12 proteins were up-regulated including the cytochrome c protein (RruA1020) and a bacterioferritin (RruA2195) (Table 4c.12). While 14 proteins were down-regulated including chaperonines like RruA0162, RruA2109 and RruA3643 as well as a tellurium stress related protein RruA0893 (Table 4c.12).

Table 4c.12 . Proteins showing significant expression changes only in the RPM experiment. Gene Gene Product name COG FC RPM number name RruA0162 groEL chaperonin GroEL O 0.64 ± 0.04 RruA0271 folE GTP cyclohydrolase I H 0.57 ± 0.05 RruA0273 fabG 3-oxoacyl- Q 0.56 ± 0.05 RruA0414 rpsF SSU ribosomal protein S6P J 1.39 ± 0.17 RruA0475 - NADH:flavin oxidoreductase/NADH oxidase C 0.15 ± 0.05 RruA0500 - extracellular solute-binding protein P 1.39 ± 0.09 RruA0893 terZ stress protein T 0.54 ± 0.02 RruA1004 hisJ lysine-arginine-ornithine-binding periplasmic protein E 1.37 ± 0.09 RruA1020 - cytochrome c, class I C 2.25 ± 0.22 RruA1101 smc hypothetical protein Rru_A1101 D 0.51 ± 0.05 RruA1262 ugpB extracellular solute-binding protein G 0.63 ± 0.10 RruA1287 - trimethylamine-N-oxide reductase (cytochrome c) C 1.91 ± 0.06 RruA1746 livK extracellular ligand-binding receptor E 1.79 ± 0.16 RruA1827 glyA serine hydroxymethyltransferase E 0.56 ± 0.08 RruA1885 eno enolase G 1.46 ± 0.23 RruA1928 - 2-isopropylmalate synthase E 0.53 ± 0.03 RruA2014 ddpA twin-arginine translocation pathway signal E 1.44 ± 0.23 RruA2071 putA methylmalonate-semialdehyde dehydrogenase C 1.40 ± 0.09 RruA2083 creA CreA S 0.63 ± 0.09 RruA2109 clpA ATP-dependent Clp protease ATP-binding subunit ClpA O 0.46 ± 0.08 RruA2195 bfr bacterioferritin P 1.39 ± 0.12 RruA2239 - TRAP dicarboxylate transporter- DctP subunit G 1.44 ± 0.02 RruA3511 acnA aconitate hydratase C 1.42 ± 0.18 RruA3600 hslU ATP-dependent protease ATP-binding subunit O 0.62 ± 0.04 RruA3643 grpE GrpE protein O 0.48 ± 0.04 2-octaprenyl-3-methyl-6-methoxy-1,4-benzoquinol RruA3707 ubiH C 0.60 ± 0.06 hydroxylase / 2-octaprenyl-6-methoxyphenol hydroxylase 'RruA' refers to gene located on the chromosome. FC : fold change. Proteins are quantified by mean ± standard deviation. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, H - Coenzyme metabolism, C - Energy production and conversion, S - Function unknown, P - Inorganic ion transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, J - Translation, ribosomal structure and biogenesis.

4C.3.2.6 PROTEINS SHOWING SIGNIFICANT EXPRESSION CHANGES IN THE RWV EXPERIMENT ONLY

23 proteins were up-regulated in the group of proteins that were only significant in RWV including the top 4 membrane related proteins RruA0437, RruA1095, RruA2211 and RruA3328 as well as the phage shock protein A (RruA1218) (all included in the Table 4c.10, see "gene number*").

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 137 For the down-regulated proteins, the 'translation, ribosomal structure and biogenesis' (J) functional category was the most numerically important comprising 17 proteins, followed from far by the category of the 'amino acid transport and metabolism' (E) class with 5 proteins (Figure 4c.8).

18

16

14

12

10

RWV liquid 8 DOWN

Numberof proteins 6

4

2

0 E GMH C S P I O T K JUn.

Functional category

Figure 4c.8. Functional classification of the 42 significant proteins specifically down-regulated in the RWV culture experiment. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, N - Cell motility and secretion, H - Coenzyme metabolism, V - Defense mechanisms, L - DNA replication, recombination and repair, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, U - Intracellular trafficking, secretion, and vesicular transport, I - Lipid metabolism, F - Nucleotide transport and metabolism, O - Posttranslational modification, protein turnover, chaperones, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un . – Unclassified.

We could also put forward the down-regulation of the outer membrane chaperone RruA1595 and the superoxide dismutase RruA1760 ("gene number*" in the Table 4c.10). Besides, a Poly(R)-hydroxyalkanoic acid synthase (RruA2413) was down-regulated as well as a phasin protein (RruA3283). Eventually, the translation of 4 hypothetical proteins was confirmed specifically by the RWV experiment (see "gene number*" in Table 4c.10).

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 138

4C.3.3 COMPARISON WITH THE MESSAGE 2 AND THE BASE-A RELATED

EXPERIMENTS

4C.3.3.1 COMPARISON REGARDING THE TRANSCRIPTOMIC APPROACH

Low overlap could be put forward between the RPM or the RWV liquid growth conditions (in minimal medium) and the space related experiments (on agar rich and minimal medium) (Figure 4c.9).

RPM liq. MSG2 RPM liq. BASE-A

197 38 181 221 14 51

RPM liq. MSG2-RPM RPM liq. BASE-A-RPM

218 17 141 223 12 17

RWV liq. MSG2 RWV liq. BASE-A

10 3 216 11 2 63

RWV liq. MSG2-RPM RWV liq. BASE-A-RPM

11 2 156 10 3 26

Figure 4c.9 . Significant differentially genes in the RPM and the RWV liquid growth conditions compared to MESSAGE 2 ( MSG2 ), BASE-A and their related modeled microgravity ( RPM ) ground simulations.

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 139 Except for the RWV liquid and MESSAGE 2 related RPM experiment comparison, the number of overlaps between the 2 conditions of interest appeared to be significant (p < 0.05). The annotation for genes coding for hypothetical proteins was manually checked using the MaGe platform. RruA1537 was up-regulated in RPM liquid, MESSAGE 2, MESSAGE 2 RPM-related and BASE-A RPM-related experiments. RruA2850, a putative chemotaxis related gene was up- regulated in RPM and RWV liquid as well as in MESSAGE 2 RPM related and BASE-A RPM-related experiments. Besides, RruA0637, coding for a putative lipoprotein, and RruA3369 were up-regulated in both RPM liquid, MESSAGE 2 and MESSAGE 2 RPM- related experiments. RruA1608, related to a mechanosensitive ion channel protein was up- regulated in both the RPM liquid and the BASE-A RPM. For the hypothetical protein encoding gene RruA1758, joint up-regulation was found in RPM and RWV liquid and in BASE-A RPM related experiment. Eventually, RruA3286 coding for a putative transcriptional regulator was jointly up-regulated in RPM liquid, MESSAGE 2 and MESSAGE RPM related experiment (Table 4c.13).

Table 4c.13 . Significant differentially genes in the RPM and the RWV liquid growth conditions compared to MESSAGE 2 ( MSG2 ), BASE-A and their related modeled microgravity ( RPM ) ground simulations. FC FC FC FC Gene Gene FC FC Product name COG RPM RWV MSG2 BASE-A number name MSG2 BASE-A liq liq RPM RPM RruA1072 rpmE 50S ribosomal protein L31 J 15.29 4.20 1.12 2.68 0.65 2.09 RruA1537 - hypothetical protein Rru_A1537 Un. 15.13 1.78 2.70 2.88 0.59 2.51 RruA3397 - hypothetical protein Rru_A3397 Un. 14.21 3.16 1.66 1.18 0.38 1.00 RruA1947 - nucleotidyltransferase-like R 14.17 2.41 2.08 1.44 1.61 1.04 RruA0086 - hypothetical protein Rru_A0086 S 10.41 1.85 1.11 0.75 0.55 2.07 RruA2687 rplD 50S ribosomal protein L4P J 10.33 1.62 2.15 0.85 1.11 0.89 RruA2850 cheL hypothetical protein Rru_A2850 M 10.21 2.11 1.51 4.56 0.72 2.02 RruA1758 - hypothetical protein Rru_A1758 Un. 10.17 2.96 0.98 1.20 0.71 2.15 RruA3394 - hypothetical protein Rru_A3394 Un. 8.63 2.18 1.34 1.08 0.47 1.65 RruA1014 - hypothetical protein Rru_A1014 Un. 8.33 1.51 0.94 1.19 0.29 1.60 cold-shock DNA-binding protein RruA0663 - K 8.33 1.87 1.31 1.18 0.53 2.37 family protein asparagine synthase, glutamine- RruA0230 asnB E 7.82 1.27 2.09 1.67 0.72 1.02 hydrolyzing RruA1237 - cytidylyltransferase H 7.72 0.85 0.83 5.06 NA 1.47 RruA0176 - acetylornithine aminotransferase E 7.40 1.62 1.22 0.39 0.79 0.97 RruA2679 rpsQ SSU ribosomal protein S17P J 6.22 1.01 3.36 1.79 1.23 0.90 ribosomal subunit interface protein, RruA0073 - J 5.68 2.11 2.85 1.74 0.91 1.12 putative RruA1236 - Iojap-related protein S 5.08 1.15 0.68 0.79 0.39 1.73 RruA1717 - ArsR family transcriptional regulator K 4.94 1.31 3.24 0.75 0.69 1.20 RruA2688 rplC 50S ribosomal protein L3 J 4.92 1.17 2.18 1.12 1.68 1.08 RruA3286 - hypothetical protein Rru_A3286 Un. 4.36 0.93 4.66 0.79 0.40 3.21 RruA1608 mscS MscS mechanosensitive ion channel M 4.34 1.18 1.00 0.74 2.21 0.96 RruA1760 sodA superoxide dismutase P 4.27 1.48 4.01 0.93 0.85 1.23 RruA1922 - DSBA oxidoreductase Q 4.16 1.31 1.07 1.42 0.81 2.27 RruA3205 rpmG 50S ribosomal protein L33P J 4.10 1.09 3.08 2.58 NA 1.33 RruA1116 - TetR family transcriptional regulator K 3.95 1.62 1.39 1.06 0.80 2.24 RruA3002 - nicotinamidase Q 3.84 1.21 1.17 0.98 2.74 0.90 RruA0436 - hypothetical protein Rru_A0436 Un. 3.79 0.87 3.48 1.43 1.07 1.20 RruA0424 - hypothetical protein Rru_A0424 Un. 3.69 0.91 3.84 1.95 1.37 1.07 Short-chain dehydrogenase/reductase RruA1945 - R 3.64 1.17 1.27 2.15 1.14 0.94 SDR RruA3019 - hypothetical protein Rru_A3019 Un. 3.62 0.74 2.11 1.64 0.51 1.09 RruA1315 - sterol-binding I 3.53 1.58 1.12 1.22 0.75 2.09

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 140

FC FC FC FC Gene Gene FC FC Product name COG RPM RWV MSG2 BASE-A number name MSG2 BASE-A liq liq RPM RPM RruA2352 - hypothetical protein Rru_A2352 Un. 3.34 1.00 1.13 3.22 1.23 1.08 RruA2195 bfr bacterioferritin P 3.34 1.58 2.54 1.38 1.69 1.07 RruA2280 - hypothetical protein Rru_A2280 Un. 3.27 1.07 1.50 3.21 0.96 1.24 RruA0637 - hypothetical protein Rru_A0637 S 3.17 1.17 3.18 3.51 0.82 1.71 RruA0119 - hypothetical protein Rru_A0119 Un. 3.15 0.87 4.22 0.77 0.42 2.02 RruA2289 - sulfate adenylyltransferase subunit 2 E 3.13 0.83 0.84 1.74 2.67 0.91 RruA2787 - inner-membrane translocator R 3.03 1.06 1.47 2.55 1.27 0.66 RruA1731 - hypothetical protein Rru_A1731 Un. 2.97 1.99 0.40 0.91 1.05 1.35 RruA1175 - CsbD-like S 2.91 0.96 2.58 1.17 0.72 1.30 RruA3491 - hypothetical protein Rru_A3491 Un. 2.87 1.16 1.36 2.81 1.27 0.76 cold-shock DNA-binding protein RruA1521 cspC K 2.69 1.18 2.48 0.84 0.91 1.12 family protein RruA1353 - hypothetical protein Rru_A1353 S 2.68 1.18 2.46 1.01 0.55 1.21 RruA1286 - hypothetical protein Rru_A1286 Un. 2.67 1.14 1.41 0.46 1.00 1.04 RruA3369 - hypothetical protein Rru_A3369 Un. 2.66 1.14 13.99 2.35 1.31 1.43 RruA1557 nuoC NADH dehydrogenase subunit C C 2.64 1.35 2.11 0.70 0.94 1.04 RruA0974 - Linocin_M18 bacteriocin protein S 2.63 1.16 0.70 1.38 0.59 3.93 RruA2678 rplN 50S ribosomal protein L14P J 2.54 0.96 2.91 1.00 1.13 1.34 RruA3189 - hypothetical protein Rru_A3189 Un. 2.53 1.17 3.80 1.51 0.85 1.59 RruA3283 - phasin S 2.52 1.22 3.67 1.62 1.16 1.08 RruA3745 - hypothetical protein Rru_A3745 Un. 2.52 1.09 1.48 2.90 NA 1.09 RruA3392 - hypothetical protein Rru_A3392 R 2.46 1.17 1.73 1.23 2.40 0.91 RruA3733 - HNH endonuclease V 2.46 1.08 1.27 1.12 0.48 0.90 RruA2261 - hypothetical protein Rru_A2261 Un. 2.45 0.92 2.27 0.84 1.09 1.97 aminodeoxychorismate synthase, RruA1132 - E 2.42 0.89 1.05 2.48 0.82 0.98 subunit I RruA3320 - pantothenate synthetase H 2.37 1.06 1.30 0.89 3.55 0.84 RruA2666 rpsM 30S ribosomal protein S13 J 2.31 1.40 3.08 2.35 NA 1.37 RruA1501 - transport-associated protein R 2.28 1.20 4.80 1.04 0.89 1.04 RruA0999 - hypothetical protein Rru_A0999 Un. 2.24 1.21 2.69 0.78 1.57 1.14 RruA0160 - hypothetical protein Rru_A0160 E 2.22 2.36 7.50 1.11 0.94 1.42 RruA0210 rpmB 50S ribosomal protein L28 J 2.19 1.02 2.13 1.67 0.68 1.15 ribose ABC transporter, periplasmic RruA1365 rbsB G 2.18 1.37 0.30 0.86 0.84 1.21 binding protein response regulator receiver domain- RruA0522 - T 2.17 0.78 2.01 1.15 0.78 0.84 containing protein RruA3701 - TetR family transcriptional regulator K 2.12 1.15 2.30 1.83 1.23 1.02 RruA2399 - hypothetical protein Rru_A2399 S 2.08 1.33 2.45 1.01 0.52 1.58 CheA Signal transduction histidine RruA0521 cheA N 2.06 1.13 2.05 0.93 0.51 0.74 Kinases (STHK) RruA1684 hfq RNA-binding protein Hfq R 2.05 1.31 1.30 1.33 0.45 1.44 response regulator receiver domain- RruA2837 ompR K 2.04 1.02 1.26 0.73 2.16 1.05 containing protein RruA3469 - hypothetical protein Rru_A3469 Un. 2.01 0.87 2.12 1.11 1.47 1.15 NA : gene not kept after spot quality control. Red highlight : up-regulated gene. Green highlight : down- regulated gene. FC : fold change. COG nomenclature: E - Amino acid transport and metabolism, G - Carbohydrate transport and metabolism, D - Cell division and chromosome partitioning, M - Cell envelope biogenesis, outer membrane, H - Coenzyme metabolism, V - Defense mechanisms, C - Energy production and conversion, S - Function unknown, R - General function prediction only, P - Inorganic ion transport and metabolism, I - Lipid metabolism, Q - Secondary metabolites biosynthesis, transport and catabolism, T - Signal transduction mechanisms, K – Transcription, J - Translation, ribosomal structure and biogenesis, Un. – Unclassified.

4C.3.3.2 COMPARISON REGARDING THE PROTEOMIC APPROACH

As a consequence of the relative low number of significant regulated proteins, even less overlap was found within the proteomic approach (Figure 4c.10). Only the limited overlap between RPM liquid and the MESSAGE 2 RPM related experiment appeared to be significiant (p < 0.05) (Table 4c.14). In addition, the overlapping proteins showed inverted trend induction.

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 141

Table 4c.14 . Significant differentially expressed genes in the RPM compared to MESSAGE 2 ( MSG2 ) related modeled microgravity ( RPM ) ground simulations. Gene Gene Product name COG FC RPM liq #(H/L) FC MSG2-RPM #(H/L) number name RruA0893 terZ stress protein T 0.54 ± 0.02 19 1.57 ± 0.05 19 RruA2083 creA CreA S 0.63 ± 0.09 6 1.6 ± 0.11 5 FC : fold change. COG nomenclature: S - Function unknown, T - Signal transduction mechanisms.

RPM liq. MSG2-RPM RPM liq. BASE-A

31 2 8 32 1 9

RWV liq. MSG2-RPM RWV liq. BASE-A

69 1 9 69 1 9

Figure 4c.10 . Significant differentially proteins in the RPM and the RWV liquid growth conditions compared to MESSAGE RPM related, BASE-A and its related modeled microgravity experiment. As reminder: (i) there are no proteomic data about the MESSAGE 2 space experiment and (ii) No significant proteins were detected in the BASE-A RPM related experiment.

4C.4 DISCUSSION

Quorum sensing activation in modeled microgravity . The gene RruA3396 encoding for the N-acyl-L-homoserine lactone (AHL) synthase, a central player in the bacterial communication known as quorum sensing, was the most up-regulated gene in both modeled microgravity experiments. Although not yet demonstrated for R. rubrum , the AHL synthase is often regulated by AHLs in a positive feed-back loop (Reading and Sperandio, 2006), in this respect the efficient low-shear solid body rotation in the modeled microgravity systems could enhance expression of RruA3396 by accumulation of the AHLs in the bacterium surroundings. While the differential expression of rhlI, encoding for the AHL synthase that produces N-butanoyl-l-homoserine lactone, has already been shown in Pseudomonas aeruginosa cultured in RWV (Crabbé et al ., 2008), the present study is the first report of the induction of quorum sensing in a bacterium cultured in the RPM. Won et al. (2008) suggested recently that the biosynthesis of the photosynthetic apparatus of the closely related Rhodobacter sphaeroides 2.4.1 is regulated by a quorum sensing circuit

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 142 involving a 7,8-cis-N-tetradecenoyl-homoserine lactone. While that exact signal molecule has not been found in R. rubrum S1H, cells cultivated in Sistrom-succinate under dark aerobic conditions in modeled microgravity showed higher pigmentation than the normal gravity control, without change in cell density and oxygenation. At the transcriptomic level, we could indeed detect the up-regulation of pucC . Moreover, the top over-expressed proteins in both the RPM and the RWV samples were related to the photosynthetic apparatus, namely PuhA and PufM. In that purpose, PucC was probably not detected because the protein is predicted to have 12 transmembrane domains that render it hardly accessible to the proteomic approach used here. On the other hand, PuhA and PufM have respectively 1 and 5 predicted transmembrane domains (TMhmm website). Hwang et al. (1997) constructed Rhodobacter sphaeroides protein I mutants that appeared less pigmented than wild-type colonies on plates. Moreover, they showed that these mutants aggregated after reaching high densities in Sistrom-succinate cultures and formed a floc that settled to the bottom of the culture tube. The aggregation in broth was prevented by addition of purified 7,8-cis-N-tetradecenoyl-homoserine lactone (Hwang et al ., 1997). Therefore, one might imagine that in sufficiently large cell aggregates, light could become a limiting nutrient and escape from a community aggregate could be advantageous. In the perspective of using R. rubrum S1H in a closed loop ecosystem under light anaerobic conditions, future studies of quorum sensing in R. rubrum and its role in cell pigmentation and aggregation are important. Comparison of RPM and RWV results . To our knowledge, the study by Patel et al. (2007) on osteoblasts is the only one to date that has compared the RPM to the RWV technology. They concluded that similar results were obtained by the two different simulators. In contrast with their results, we found only a few significant genes from the RWV samples and all were included in the more pronounced response of R. rubrum to the RPM cultivation. Moreover, common genes showed much higher fold induction (at least 3 times) in the latter device. We observed that the RPM induced specifically the mechanosensitive mscS , the redox state of the cell related to sodA , bfr and fur as well as 4 genes related to phage defense system. On the other hand, this trend was not confirmed at the proteomic level for which only a few common proteins were found and for which the RWV appeared to induce higher number of significantly regulated proteins. In addition, a bias was introduced during the sample preparation from the RWV experiment and, without changing dramatically the present results, we can suspect to have missed a few significant regulated proteins. Therefore, one must be cautious in concluding one of the two simulators to induce a higher response of R. rubrum . Nevertheless, it appeared clearly that culturing in liquid conditions induced more genes with

Chapter 4c - R. rubrum S1H liquid culture in modeled microgravity 143 higher fold induction than culturing in solid agar medium (chapter 4a and 4b). A low overlap was observed between the RPM or the RWV liquid growth conditions and the space related experiments (on solid agar medium). Genes coding for hypothetical proteins . Beside the evident knowledge acquired by confronting R. rubrum to different environmental conditions, these multiple experiments also permitted us to have more insight into the unknown part of the R. rubrum genome. Hence, we found of particular interest the results for the hypothetical proteins encoding gene RruA1537 and RruA2850. Indeed, these 2 genes were the only two showing a similar trend in expression in the 4 conditions that included change in gravity. Studying the fate of a micropollutant that could interfere with the proper functioning of the MELiSSA system, Pycke (in preparation) recently reported the high over-expression of a cluster of genes coding for hypothetical proteins namely RruA0423-0424-0425-0426 in R. rubrum S1H cultivated in the presence of 25 ppb triclosan, a biocide commonly used in personal care products. These genes were named micropollutant up-regulated factor or ' muf '. In this study, we found mufM (RruA0424) over-expressed in the space and the microgravity simulation of MESSAGE 2 experiments (together with RruA0422 and RruA0423, see p. 91) as well as in the RPM liquid experiment. Taken together, these data suggested the possible induction of muf genes by chemical as well as mechanical stimuli. Because bacteria possess several universal stress protein e.g . 6 in E. coli (Kvint et al ., 2003) and 15 in Cupriavidus metallidurans CH34 (MaGe website, 'Cupriavidu2Scope Project'), we can speculate MufM to be part of a yet uncharacterized universal stress protein cluster in R. rubrum S1H, completing the arsenal of the 4 universal stress proteins that have been detected so far based on the genome annotation.

CHAPTER 5 – GENERAL CONCLUSIONS AND

FUTURE PERSPECTIVES

After 20 years of research and development on ground, the MELiSSA project entered a crucial phase by moving beyond the borders of the Earth. The present work has taken the α- proteobacterium Rhodospirillum rubrum S1H to new unexplored heights and revealed directly upon first analysis some interesting aspects. The body of this work included the optimization and implementation of high-throughput techniques to assess both the global transcriptional and translational response of this bacterium to conditions related to space flight. The final goal was to acquire an increased understanding of the physiology of R. rubrum in general and an improved knowledge for the study of the functionality and stability of R. rubrum in the MELiSSA loop in particular. Currently, 25 % of the 3,829 candidate protein-encoding genes are annotated as encoding for hypothetical proteins. In addition, manual re-annotation of certain genes and gene clusters showed that it's needed to check and if necessary to improve the current annotation. This limits the interpretation of transcriptomic and proteomic data to a certain extent but on the other hand it could possibly reveal new functional assignments to unknown genes and, in this way, improve the knowledge of R. rubrum S1H and related bacteria. More distressing are the possible implications for the MELiSSA loop in which R. rubrum S1H is a key organism in the second bioreactor and thus is vital for the closure and performance of the loop. An optimal reactor relies on a complete knowledge of the bacterium's metabolism in the reactor, its resilience towards any thinkable perturbation and its functional stability keeping in mind the continuous long-term operation mode of the reactors. Therefore, there's not only a need for studies addressing the impact of perturbations due to space flight related conditions such as the one conducted here, but also studies addressing other environmental and stress parameters as well as studies investigating the genetic stability of the bacterium and its susceptibility to lateral gene transfer. Now that this concern regarding the current molecular knowledge of R. rubrum S1H has been formulated, some main questions posed at the start or during the project will be addressed.

Chapter 5 – General conclusions and future perspectives 146 1. How far can transcriptomic and proteomic data be integrated? As already reported, the integration of transcriptomic and proteomic data appeared to be complicated mainly due to fundamental biological factors in the transcription and translation processes like differences in half-life in vivo (Gygi et al ., 1999; Nie et al ., 2006a). Moreover, Nie et al. (2006b) showed that mRNA abundance alone could explain only 20–28 % of the total variation of protein abundance, suggesting that the correlation between mRNA and protein levels cannot be determined by mRNA abundance alone. Therefore, integration of transcriptomic and proteomic data are of prime importance since it can lead to new insights into the physiology of an organism that would not have been apparent from differential analyses at either the mRNA or protein level alone (de Groot et al ., 2007). This was particularly obvious in our study of R. rubum in liquid modeled microgravity (chapter 4c). While being able to identify 25 % of R. rubrum S1H proteome, we could also confirm the translation of about 15 % of the hypothetical proteins. Therefore, for these gene models the annotation can already be upgraded from 'hypothetical protein' to 'protein of unknown function'. Although often ignored by authors because they are seen as dead-end in data interpretation, hypothetical proteins remain of interest, especially for R. rubrum for which they constitute about 25 % of the genome as mentioned above. As an example, by combining the transcriptomic data from all our experimental conditions, we could point out interest for hypothetical proteins such as RruA1537 and RruA2850 as proteins of potential new function adapted to the stress encountered in change in gravity conditions while RruA0424 could be part of a yet uncharacterized universal stress protein cluster in R. rubrum S1H. Future studies including the construction of specific mutants and further phenotypic, transcriptomic and proteomic characterization will be necessary to confirm or invalidate these observations.

2. How does R. rubrum S1H react to space flight related conditions? During this study R. rubrum S1H had the opportunity to fly twice to the ISS. Limited by the constraints inherent to space flight missions, 2 different solid medium based culture setups were studied at both the transcriptomic and the proteomic level using high-throughput techniques. To date, only the study by Wilson et al . (2007) used a similar approach using Salmonella enterica serovar Typhimurium and reported an increased virulence pattern for space grown cultures incriminating a major role to the global regulator Hfq, which is not the case in this study for R. rubrum S1H. These results indicate that either the role of Hfq is species- (or strain)-dependent or that perception of space flight differs from bacterium to

Chapter 5 – General conclusions and future perspectives 147 bacterium. This is not totally surprising since previous studies showed that different organisms may dramatically differ in their responses to the space environment as mentioned in chapter 1. Our study became original when, in an effort to study the main space-related stresses independently, ground-based simulations of ISS ionizing radiation and reduced gravity were performed using the same culture conditions and setup as for the actual space flights. This permitted us to put forward the importance of the medium composition and the culture setup on the response of the bacterium to the space flight related stress. Moreover, we showed for the first time that a low dose of ionizing radiation (2 mGy) can induce a significant response at the transcriptomic level, although no change in cell viability and only a few significant differentially expressed proteins were observed. This highlights the necessity to study the effect of low dose on microorganisms (especially the radiosensitive ones) because most of the previous studies focused on high doses and/or on the radioresistant bacterium Deinococcus radiodurans (Cox and Battista, 2005). In a more general way, ground simulations should be taken for what they are…simulations. Indeed, space flight conditions contain such a huge mix of perturbations starting from the launch (acceleration, vibration), passing through the ISS (large spectrum of ionizing radiation, coupled to microgravity and changes in the magnetic field) and return to Earth (deceleration, vibration). While our ground simulations did not give a large overlap with the actual space flight response, still we could retrieve information about the susceptibility of the bacterium to various environmental stress conditions (ionizing radiation and low-shear modeled microgravity cultures). As a consequence, these experiments did not permit us to say which stress contributes the most to the space flight response, ionizing radiation or the change in gravity. Although, dissection of such an intricate stress situation as space is not straightforward with possible (synergetic) interactions on multiple levels and other prevailing conditions (such as the experimental design, culture conditions, etc .) will probably also affect the contribution of ionizing radiation or reduced microgravity to the global space response. A step further in this study was to study R. rubrum S1H in liquid conditions, closer to the MELiSSA culture conditions, using modeling microgravity devices. First statement, the culture in liquid conditions appeared to induce more significant genes with higher fold induction than the culture in solid medium. Our results suggested that the response of the bacterium was at least partly related to a side-effect, accumulation of (by-) products around the cell environment, rather than a direct mechanical effect on the bacterial shape for instance, which is in accordance with results reported by previous studies (Tucker et al ., 2007; Vukanti

Chapter 5 – General conclusions and future perspectives 148 et al ., 2008). In that purpose, a clear difference in the expression of quorum sensing related genes was observed in liquid but not on solid medium. An observation worthwhile to be monitored in future space flight experiment. Indeed, the increased production of the N-acyl-L- homoserine lactones would probably activate the bacterium's quorum sensing regulon and for instance the observed increase in pigment production could be activated by quorum sensing (Puskas et al ., 1997; Hwang et al ., 2008).

2.1 Perspectives

Related to the MELiSSA loop, it's a necessity to further explore this unknown quorum sensing regulon in R. rubrum . Indeed, will a bioreactor in the MELiSSA loop probably run in a constant cell density mode and certain genes will be therefore activated or not depending on the threshold concentration of AHLs. Feeding the R. rubrum reactor with analogous signaling molecules present in the output of the first reactor could interact with the LuxR-like transcriptional regulator and inhibit or induce its quorum sensing regulon, putatively affecting the bioreactor operation. Furthermore, these analogous molecules together with the signaling molecules produced by the α-proteobacterium R. rubrum , suspected to carry aggregation suppression effect (Puskas et al. , 1997), could accidentally be fed to the next reactor in the loop, possibly acting on the biofilm formation and/or affecting the quorum sensing regulons in the strains occupying this third packed bed reactor, namely the α-proteobacterimu Nitrobacter europeae and the β-proteobacterium Nitrosomonas winogradskyi. Moreover, quorum sensing systems have already been shown not only to elicit a response inside a bacterial population but also to cross the inter-kingdom barrier (Shiner et al ., 2005; Hugues and Sperandio, 2006). Therefore, production of bacterial signaling molecules could not only influence the MELiSSA loop by affecting the bacterial population within a bioreactor but also between bioreactors and even the last compartment comprising the cyanobacterium Arthrospira sp . and higher plants. However, these molecules are not ubiquitously stable, and high pH (from 8 or higher) or elevated temperature (37 °C or higher) will result in lactonolysis (Yates et al ., 2002). The very recent (Oct. 2008) BASE-C space flight experiment involved R. rubrum S1H and other bacterial strains in liquid culture condition. This was a next step in filtering out 'side effects' due to, for instance, the launch, by activating, growing and fixing the cells in the ISS. Moreover the use of an in-flight centrifuge created an additional control eliminating the effect of microgravity. While this constitutes a huge improvement compared to the 'early' flight

Chapter 5 – General conclusions and future perspectives 149 experiments, some problems could remain since acceleration, vibration, etc . could cause damage to the 'inactive' cell that will be processed when the bacterium is activated (in-flight) and an extra response could be elicited rather than space flight alone. In a few words, no ground simulations would be able to reproduce all the stress factors related to a space flight mission (mentioned above). Still, they will always be more accessible than flight opportunities (but also in term of experiment size, weight, electric power requirements, etc .) and can give an indication of which particular physiological aspect to monitor during cultivation in actual space flight conditions.

3. What are the consequences of our results for future MELiSSA development? Seeing transcriptome or proteome effects without change in viability could raise the question of the actual effect at the phenotypic level. We saw phenotypic evidence of genes identified with the transcriptomic or the proteomic approach but, at this point no conclusion can be written concerning change in R. rubrum S1H metabolism. Nevertheless, we showed that R. rubrum S1H is able to sense a variety of environmental changes that are likely linked to space flight conditions (chapter 4a, 4b, 4c). Therefore, some subtle effects could accumulate on a continuous reactor culturing mode during an extended space exploration mission.

3.1 Perspectives

Ground simulations ( e.g. space microgravity and ionizing radiation) involving culture in bioreactors are crucial. Obviously culture in MELiSSA condition (light and anaerobic) has to be performed. This could also be coupled to samples received from the MELiSSA partner's bioreactors on regular basis to assess the effect of long-time culturing. Considering all these samples, a metabolomic approach should be foreseen to follow the behavior of key metabolites of interest. And, if bioreactors issues occurred, it will be possible to study the strain using the transcriptomic and proteomic tools we set up during this work. In that respect, efficient countermeasures against the effect of microgravity do not yet exist although a response elicited by the indirect effects of microgravity, such as lack of sedimentation and natural convection could be minimized with a "space-certified" hardware that ensures a completely stirred bioreactor.

Chapter 5 – General conclusions and future perspectives 150 On the other hand, a major issue for future manned space exploration is the potentially lethal damage to living organisms from exposure to radiation arising, from example, from the solar wind. In low Earth orbit, much of the particle flux from space is attenuated thanks to the protection created by the magnetosphere which is not the case in the interplanetary space (Benton and Benton, 2001). Conventional shielding like aluminum, polyethylene as well as hydrogen-rich polymers (Vana et al ., 2006) can not be considered due to weight and production of secondaries particles constraints. Since the 1960s there has been an interest in the viability of creating artificial magnetosphere structures to provide local protection of spacecraft instrumentation and astronauts (e.g . Levy and French, 1968), but these studies were criticized as requiring huge amount of power supply and coil assemblies to create the necessary field strength (Parker, 2006). Only very recently, Bamford et al . (2008) described a promising new experiment to test the shielding concept of a dipole-like magnetic field and plasma, surrounding a spacecraft forming a 'mini magnetosphere'. Nevertheless, ionizing radiation and its deleterious effects as well as many of the psychological and environmental issues associated with extended confinement and isolation remain crucial issues to consider for future manned space exploration.

In summary, we succeeded in (i) the set up of high-throughput molecular tools to characterize R. rubrum at both the transcriptomic and proteomic level. This permitted us to (ii) assess the susceptibility of this bacterium by integrating data from both space flight and ground simulations experiments and (iii) to propose critical parameters to monitor for future bioreactor studies. In this way, we achieved fundamental goals defined at the start of this research and paved the way for further investigations on R. rubrum S1H as well as on the other MELiSSA strains, with the final goal of developing a reliable and a sustainable life support system for manned extended space missions.

REFERENCES

Abee T., Knol J., Hellingwerf K.J., Bakker E.P., Siebers A., Konings W.N. 1992. A Kdp-like, high-affinity, K+-translocating ATPase is expressed during growth of Rhodobacter sphaeroides in low potassium media. Arch. Microbiol . 158 :374-380.

Albrecht-Buehler, G. 1991. Possible mechanisms of indirect gravity sensing by cells. ASGSB Bull. 4(2):25-34.

Ahring B.K., Ibrahim A.A., Mladenovska Z. 2001. Effect of temperature increase from 55 to 65°C on performance and microbial population dynamics of an anaerobic reactor treating cattle manure, Wat. Res. 35 :2446-2452.

Andralojc P.J., Harris D.A. 1992. Isolation and characterisation of a functional alpha beta heterodimer from the ATP synthase of Rhodospirillum rubrum. FEBS Lett . 310 (2):187-192.

Ansong C., Purvine S.O., Adkins J.N., Lipton M.S., Smith R.D. 2008. Proteogenomics: needs and roles to be filled by proteomics in genome annotation. Brief. Funct. Genomic Proteomic 7(1):50-62.

Assiddiq B.F., Snijders A.P., Chong P.K., Wright P.C., Dickman M.J. 2008. Identification and characterization of Sulfolobus solfataricus P2 proteome using multidimensional liquid phase protein separations . J. Proteome Res . 7:2253-2261.

Baker P.W., Leff L.G. 2004. The effect of simulated microgravity on bacteria from the Mir space station. Microgravity Sci. Technol . 15 , 35-41.

Baker P.W., Leff L.G. 2006. Mir space station bacteria responses to modeled reduced gravity under starvation conditions. Adv. Space Res . 38, 1152-1158.

Baker P.W., Meyer M.L., Leff L.G. 2004. Escherichia coli growth under modeled reduced gravity. Microgravity Sci. Technol. 15 , 39-44.

Ballal A., Basu B., Apte S.K. 2007. The Kdp-ATPase system and its regulation. J. Biosci . 32 (3):559-568.

Bamford R., Gibson K.J., Thornton A.J., Bradford J., Bingham R., Gargate L., Silva L.O., Fonseca R.A., Hapgood A., Norberg C., Todd T., Stamper R. 2008. The interaction of flowing plasma with a dipole magnetic field: measurements and modelling of a diamagnetic cavity relevant to spacecraft protection. Plasma Phys. Control. Fusion 50:124025(11pp).

Bao Y., Lies D.P., Fu H., Roberts G.P. 1991. An improved Tn7-based system for the single copy insertion of cloned genes into chromosomes of gram-negative bacteria. Gene 109 :167- 168.

References 152 Barrangou R., Fremaux C., Deveau H., Richards M., Boyaval P., Moineau S., Romero D.A., Horvath P. 2007. CRISPR provides acquired resistance against viruses in prokaryotes. Science 315 :1709-1712.

Baurain D., Renquin L., Grubisic S., Scheldeman P., Belay A., Wilmotte A. 2002. Remarkable conservation of internally transcribed spacer sequences of Arhtrospira ("Spirulina") (Cyanophyceae, Cyanobacteria) strains from four continents and of recent and 30-year-old dried samples from Africa. J. Phycol . 38 :384-393.

Bendixen H.J. 1994. Safeguard against pathogens in Danish biogas plants. Wat. Sci. Tech . 30 :171-180.

Benjamini, Y., Hochberg, Y. 1995. Controlling the False Discovery Rate - a Practical and Powerful Approach to Multiple Testing. J. Royal Stat. Soc. Series B-Meth . 57 :289-300.

Benoit M.R. Klaus D.M. 2007. Microgravity, bacteria and the influence of motility. Adv. Space Res . 39 :1225-1232.

Benton E.R., Benton E.V. 2001. Space radiation dosimetry in low Earth orbit and beyond. Nuclear Instruments and Methods in Physics Research B 184 :255-294.

- Bielski B.H.J., Arudi R.L., Sutherland M.W. 1983. A study of the reactivity of HO 2/O 2 with unsaturated fatty acids. J. Biol. Chem . 258 : 4759-4761.

Bjellqvist, B.,Hughes, G.J., Pasquali, Ch., Paquet, N., Ravier, F., Sanchez, J.-Ch., Frutiger, S. Hochstrasser, D.F. 1993. The focusing positions of polypeptides in immobilized pH gradients can be predicted from their amino acid sequences. Electrophoresis 14 :1023-1031.

Boeckmann B., Bairoch A., Apweiler R., Blatter M.C., Estreicher A., Gasteiger E., Martin M.J., Michoud K., O'Donovan C., Phan I., Pilbout S., Schneider M. 2003. The SWISS-PROT protein knowledgebase and its supplement TrEMBL in 2003. Nucleic Acids Res . 31 :365-370.

Blount P., Moe P.C. 1999. Bacterial mechanosensitive channels: integrating physiology, structure and function. Trends Microbiol . 7:420–424.

Bonam D., Lehman L., Roberts G.P., Ludden P.W. 1989. Regulation of carbon monoxide dehydrogenase and hydrogenase in Rhodospirillum rubrum: effects of CO and oxygen on synthesis and activity. J. Bacteriol . 171 (6):3102-3107.

Booth I.R., Edwards M.D., Black S., Schumann U., Miller S. 2007. Mechanosensitive channels in bacteria: signs of closure? Nat. Rev. Microbiol . 5(6):431-440.

Bosch G., Skovran E., Xia Q., Wang T., Taub F., Miller J.A., Lidstrom M.E., Hackett M. 2008. Comprehensive proteomics of Methylobacterium extorquens AM1 metabolism under single carbon and nonmethylotrophic conditions. Proteomics 8:3494-3505.

Bouloc P., D'Ari R. 1991. Escherichia coli metabolism in space. J. Gen. Microbiol . 1991 137 (12):2839-2843.

References 153 Bradford M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem . 72 :248- 254.

Brandt H., Knee Jr., Fuller R.C., Gross R.A., Lenz R.W. 1989. Ability of the phototrophic bacterium Rhodospirillum rubrum to produce various poly(hydroxyalkanoates): potential sources for biodegradable polyesters. Int. J. Biol. Macromol . 11 :49-55.

Brouns S.J., Jore M.M., Lundgren M., Westra E.R., Slijkhuis R.J., Snijders A.P., Dickman M.J., Makarova K.S., Koonin E.V., van der Oost J. Small CRISPR RNAs guide antiviral defense in prokaryotes. 2008. Science 321 (5891):960-964.

Brötz-Oesterhelt H., Bandow J.E., Labischinski H. 2005. Bacterial proteomics and its role in antibacterial drug discovery. Mass Spectrom. Rev . 24 (4):549-565.

Burtscher C., Fall P.A., Christ O., Wilderer P.A., Wuertz S. 1998. Detection and survival of pathogens during two-stage thermophilic/mesophilic treatment of suspended organic waste, Wat. Sci. Tech. 38 :123-126.

Burtscher C., Wuertz D. 2003. Evaluation of the use of PCR and reverse transcriptase PCR for detection of pathogenic bacteria in biosolids from anaerobic digestors and aerobic composters. Appl. Environ. Microbiol . 69 :4618-4627.

Cañas B., Piñeiro C., Calvo E., López-Ferrer D., Gallardo J.M. 2007. Trends in sample preparation for classical and second generation proteomics. J. Chromatogr. A . 1153 (1-2):235- 258.

Cánovas D., Fletcher S.A., Hayashi M., Csonka L.N. 2001. Role of trehalose in growth at high temperature of Salmonella enterica serovar Typhimurium. J. Bacteriol . 183(11):3365- 3371.

Cargile B.J., Bundy J.L., Freeman T.W., Stephenson J.L. Jr. 2004a. Gel based isoelectric focusing of peptides and the utility of isoelectric point in protein identification. J. Proteome Res . 3(1):112-119.

Cargile B.J., Sevinsky J.R., Essader A.S., Stephenson J.L. Jr., Bundy J.L. 2005. Immobilized pH gradient isoelectric focusing as a first-dimension separation in shotgun proteomics. J. Biomol. Tech. 16 (3):181-189.

Cargile B.J., Talley D.L., Stephenson J.L. Jr. 2004b. Immobilized pH gradients as a first dimension in shotgun proteomics and analysis of the accuracy of pI predictability of peptides. Electrophoresis 25 (6):936-945.

Chachkhiani M., Dabert P., Abzianidze T., Partskhaladze G., Tsiklauri L., Dudauri T., Godon J.J. 2004. 16S rDNA characterization of bacterial and archaeal communities during start-up of anaerobic thermophilic digestion of cattle manure. Bioresource Technol . 93 :227-232.

Chalkley R.J., Hansen K.C., Baldwin M.A. 2005. Bioinformatic methods to exploit mass spectrometric data for proteomic applications. Methods Enzymol . 402 :289-312.

References 154 Chen J., Balgley B.M., DeVoe D.L., Lee C.S. 2003. Capillary isoelectric focusing-based multidimensional concentration/separation platform for proteome analysis. Anal Chem . 75 (13):3145-3152.

Cheng Y.S., Brantner C.A., Tsapin A., Collins M.L. 2000. Role of the H protein in assembly of the photochemical reaction center and intracytoplasmic membrane in Rhodospirillum rubrum. J. Bacteriol . 182 (5):1200-1207.

Chhabra S.R., He Q., Huang K.H., Gaucher S.P., Alm E.J., He Z., Hadi M.Z., Hazen T.C., Wall J.D., Zhou J., Arkin A.P., Singh A.K. 2006. Global analysis of heat shock response in Desulfovibrio vulgaris Hildenborough. J. Bacteriol . 188 (5):1817-1828.

Chick J.M., Haynes P.A., Molloy M.P., Bjellqvist B., Baker M.S., Len A.C. 2008. Characterization of the rat liver membrane proteome using peptide immobilized pH gradient isoelectric focusing. J. Proteome Res. 7(3):1036-1045.

Chien L.F., Wu J.J., Tzeng C.M., Pan R.L. 1993. ATPase of Rhodospirillum rubrum requires three functional copies of beta subunit as determined by radiation inactivation analysis. Biochem. Mol. Biol. Int . 31 (1):13-18.

Christiaens H. 1990. Utilisation of various carbon and nitrogen sources by the phototrophs of compartment 2. Utilisation of an organic or inorganic nitrogen sources by phototrophs. Technical note 6.2., MELiSSA Contract 8125/NL/FG.

Christiaens H. 1991a. Photoautotrophic growth of Rhodospirillum rubrum and Rhodobacter capsulata . Technical note 1, MELiSSA contract 8125/NL/FG.

Christiaens H. 1991b. The growth of pure culture of Rhodospirillum rubrum and Rhodobacter capsulata on the supernatant of the liquefying compartment. Technical note 7.2, MELiSSA Contract 8125/NL/FG.

Ciel et Espace. "Spoutnik, 1957 – La véritable histoire de l'espace". Hors-Série, n°18 octobre 2007.

Clayton R.K., Bryan W.C., Frederick A.C. 1958. Some effects of ultraviolet on respiration in purple bacteria. Arch Mikrobiol . 29 (3):213-226.

Cogoli A., Bechler B., Lorenzi G. 1990. Response of cells to microgravity. in Fundamentals of Space Biology, Asahima M. and Malacinski G.M. editors. Japan Sci. So. Press, Tokyo/Springer-Verlag, Berlin:97-111.

Corbin R.W., Paliy O., Yang F., Shabanowitz J., Platt M., Lyons C.E. Jr, Root K., McAuliffe J., Jordan M.I., Kustu S., Soupene E., Hunt D.F. 2003. Toward a protein profile of Escherichia coli : comparison to its transcription profile. Proc. Natl. Acad. Sci. U S A . 100 :9232-9237.

Cox M.M., Battista J.R. 2005. Deinococcus radiodurans – the consummate survivor. Nat Rev Microbiol . 3(11):882-892.

References 155 Cox B., Kislinger T., Wigle D.A., Kannan A., Brown K., Okubo T., Hogan B., Jurisica I., Frey B., Rossant J., Emili A. 2007. Integrated proteomic and transcriptomic profiling of mouse lung development and Nmyc target genes. Mol. Syst. Biol . 3:109.

Crabbé A., De Boever P., Van Houdt R., Moors H., Mergeay M., Cornelis P. 2008. Use of the rotating wall vessel technology to study the effect of shear stress on growth behaviour of Pseudomonas aeruginosa PA01. Environ. Microbiol . 10 (8):2098-2110.

Crook, S. M., S. B. Treml, and M. L. P. Collins. 1986. Immunocytochemical ultrastructural analysis of chromatophore membrane formation in Rhodospirillum rubrum . J. Bacteriol . 167 : 89-95.

Crump W. and Janik D., (Editor, Eds.), Book Introduction to Life Support Systems, Krieger Publishing, Malabar, Florida, 1992, pp. pages.

Daly M.J., Gaidamakova E.K., Matrosova V.Y., Vasilenko A., Zhai M., Venkateswaran A., Hess M., Omelchenko M.V., Kostandarithes H.M., Makarova K.S., Wackett L.P., Fredrickson J.K., Ghosal D. 2004. Accumulation of Mn(II) in Deinococcus radiodurans facilitates gamma-radiation resistance. Science 306 (5698):1025-1028.

Daly M.J., Gaidamakova E.K., Matrosova V.Y., Vasilenko A., Zhai M., Leapman R.D., Lai B., Ravel B., Li S.M., Kemner K.M., Fredrickson J.K. 2007. Protein oxidation implicated as the primary determinant of bacterial radioresistance. PLoS Biol . 5(4):e92.

De Boever P., Mergeay M., Ilyin V., Forget-Hanus D., Van der Auwera G., Mahillon J. 2007. Conjugation-mediated plasmid exchange between bacteria grown under space flight conditions. Microgravity Science and Technology 19 (5-6):138-144.

Demain, A. L., Fang A.. 2001. Secondary metabolism in simulated microgravity. Chem. Rec . 1:333–346. de Groot M.J., Daran-Lapujade P., van Breukelen B., Knijnenburg T.A., de Hulster E.A., Reinders M.J., Pronk J.T., Heck A.J., Slijper M. 2007. Quantitative proteomics and transcriptomics of anaerobic and aerobic yeast cultures reveal post-transcriptional regulation of key cellular processes. Microbiol . 153 :3864-3878. de Vet S.J., Rutgers R. 2007. From waste to energy: First experimental Bacterial Fuel Cells onboard the International Space Station. Microgravity sci. technol . XIX-5/6:225-9.

Domon B., R. Aebersold R. 2006. Mass spectrometry and protein analysis. Science 312 (5771):212-217.

Du J., Gebicki J.M. 2004. Proteins are major initial cell targets of hydroxyl free radicals. Int. J. Biochem. Cell. Biol . 36 :2334–2343.

Duarte E.A., Mendes N., Oliviera J.S. 1992. Removal of Salmonella , Streptococci , and coliforms in pig breeding effluent by anaerobic mesophilic digestion. Wat. Sci. Tech . 26:2169-2172 .

References 156 Dubuisson J.F., Vianney A., Hugouvieux-Cotte-Pattat N., Lazzaroni J.C. 2005. Tol-Pal proteins are critical cell envelope components of Erwinia chrysanthemi affecting cell morphology and virulence. Microbiol. 151 (10):3337-3347.

Dworzanski J.P., Deshpande S.V., Chen R., Jabbour R.E., Snyder A.P., Wick C.H., Li L. 2006. Mass spectrometry-based proteomics combined with bioinformatic tools for bacterial classification. J. Proteome Res . 5(1):76-87.

Eckart P., in: (L. Markley, W.L. Larson, R.E. Münch and J.F. Shea, Eds.), Space flight life support and biospherics, Bioregenerative life support concepts, Microcosm, Inc., Torrance, CA, 1996a, pp.249-364.

Eckart P., in: (L. Markley, W.L. Larson, R.E. Münch and J.F. Shea, Eds.), Space flight Life Support and Biospherics, Physico-chemical versus bioregenerative life support, Microcosm, Inc., Torrance, CA, 1996b, pp.125-174.

Elias J.E., Haas W., Faherty B.K., Gygi S.P. 2005. Comparative evaluation of mass spectrometry platforms used in large-scale proteomics investigations. Nat. Methods 2(9):667- 675.

Elias D.A., Monroe M.E., Smith R.D., Fredrickson J.K., Lipton M.S. 2006. Confirmation of the expression of a large set of conserved hypothetical proteins in Shewanella oneidensis MR- 1. J. Microbiol. Meth. 66 :223-233.

Elmashad H.M., Zeeman G., Lettinga G. 2001. Thermophilic anaerobic digestion of cow manure, proceedings at the 9th World Congress on Anaerobic Digestion, 505-508.

Encyclopædia Britannica. 2009. "Konstantin Eduardovich Tsiolkovsky". Encyclopædia Britannica Online. 09 Jan. 2009.

Engeli H., Edelmann W., Fuchs J., Rottermann K. 1993. Survival of plant pathogens and weeds during anaerobic digestion. Wat. Sci. Tech. 27 :69-76.

Eraso J.M., Roh J.H., Zeng X., Callister S.J., Lipton M.S., Kaplan S. 2008. Role of the global transcriptional regulator PrrA in Rhodobacter sphaeroides 2.4.1: combined transcriptome and proteome analysis. J. Bacteriol . 190 (14):4831-4848.

ESA's 2001 Aurora Program website (connection 01-12-2009): http://www.esa.int/SPECIALS/Aurora/SEMZOS39ZAD_0.html

Fang A., Pierson D., Mishra S., Demain A. 2000. Growth of Streptomyces hygroscopicus in rotating-wall bioreactor under simulated microgravity inhibits rapamycin production. Appl. Microbiol. Biotechnol . 54 :33–36.

Fang A., Pierson D., Mishra S., Koenig D., Demain A. 1997. Effect of simulated microgravity and shear stress on microcin B7 production by Escherichia coli and on its excretion into the medium. Appl. Environ. Microbiol . 63 :4090–4092.

References 157 Fermin D., Allen B.B., Blackwell T.W., Menon R., Adamski M., Xu Y., Ulintz P., Omenn G.S., States D.J. 2006. Novel gene and gene model detection using a whole genome open reading frame analysis in proteomics. Genome Biol . 7(4):R35.

Fernández A., Huang S., Seston S., Xing J., Hickey R., Criddle C., Tiedje J.. How stable is stable? Function versus community composition, Appl. Environ. Microbiol . 65 :3697-3704.

Figeys D., Ducret A., Yates J.R. 3rd, Aebersold R. 1996. Protein identification by solid phase microextraction-capillary zone electrophoresis-microelectrospray-tandem mass spectrometry. Nat. Biotechnol . 14 (11):1579-1583.

Foster P.L. 2007. Stress-induced mutagenesis in bacteria. Crit. Rev. Biochem. Mol. Biol . 42 (5):373-397.

Fotiadis D., Qian P., Philippsen A., Bullough P.A., Engel A., Hunter C.N. 2004. Structural analysis of the reaction centre light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J. Biol. Chem . 279 (3):2063-2068.

Fredrickson J.K., Li S.M., Gaidamakova E.K., Matrosova V.Y., Zhai M., Sulloway H.M., Scholten J.C., Brown M.G., Balkwill D.L., Daly M.J. 2008. Protein oxidation: key to bacterial desiccation resistance? ISME J . 2(4):393-403.

Friedberg E.C., Walker G.C., Siede W. DNA repair and mutagenesis, ASM press. Washington D.C., 1995.

Fyfe P.K., Isaacs N.W., Cogdell R.J., Jones M.R. 2004. Disruption of a specific molecular interaction with a bound lipid affects the thermal stability of the purple bacterial reaction centre. Biochim. Biophys. Acta. 1608 (1):11-22.

Gan C.S., Reardon K.F., Wright P.C. 2005. Comparison of protein and peptide prefractionation methods for the shotgun proteomic analysis of Synechocystis sp . PCC 6803. Proteomics 5(9):2468-2478.

Gao Q., Fang A., Pierson D.L., Mishra S.K., Demain A.L. 2001. Shear stress enhances microcin B17 production in a rotating wall bioreactor, but ethanol stress does not. Appl. Microbiol. Biotechnol . 56 :384–387.

Garcia-Pichel F., in: (J. Lederberg, Eds.), Encyclop. Microbiol., Cyanobacteria, Academic Press, 2000, pp. 907-929.

Gentleman, R. C., Carey, V. J., Bates, D. M., Bolstad, B., Dettling, M., Dudoit, S., Ellis, B., Gautier, L., Ge, Y., Gentry, J., Hornik, K., Hothorn, T., Huber, W., Iacus, S., Irizarry, R., Leisch, F., Li, C., Maechler, M., Rossini, A. J., Sawitzki, G., Smith, C., Smyth, G., Tierney, L., Yang, J. Y., and Zhang, J. 2004. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol . 5(10), R80.

Gerard E., Jolivet E., Prieur D., Forterre P. 2001. DNA protection mechanisms are not involved in the radioresistance of the hyperthermophilic archaea Pyrococcus abyssi and P. furiosus . Mol. Genet. Genomics 266:72–78.

References 158 Gingras G., Picorel R. 1990. Supramolecular arrangement of Rhodospirillum rubrum B880 holochrome as studied by radiation inactivation and electron paramagnetic resonance. Proc. Natl. Acad. Sci. USA 87 :3405-3409.

Giotis E.S, Muthaiyan A., Blair I.S., Wilkinson B.J., McDowell D.A. 2008. Genomic and proteomic analysis of the Alkali-Tolerance Response (AlTR) in Listeria monocytogenes 10403S. BMC Microbiol . 24 (8):102.

Gòdia F., Albiol L., Montesinos J.L., Pérez J., Creus N., Cabello F., Mengual X., Montras A., Lasseur C. 2002. MELiSSA: a loop of interconnected bioreactors to develop life support in Space. J. Biotechnol . 99 :319-330.

Goossens O., Vanhavere F., Leys N., De Boever P., O'Sullivan D., Zhou D., Spurny F., Yukihara E.G., Gaza R., McKeever S.W. 2006. Radiation dosimetry for microbial experiments in the International Space Station using different etched track and luminescent detectors. Radiat. Prot. Dosimetry 120 (1-4):433-437.

Graumann J., Hubner N.C., Kim J.B., Ko K., Moser M., Kumar C., Cox J., Schöler H., Mann M. 2007. Stable isotope labeling by amino acids in cell culture (SILAC) and proteome quantitation of mouse embryonic stem cells to a depth of 5,111 proteins. Mol. Cell. Proteomics 7(4):672-683.

Green J., Paget M.S. 2004. Bacterial redox sensors. Nat. Rev. Microbiol . 2(12):954-966.

Guest J.R. 1974. Bacteriocinogeny in the Athiorhodaceae. J. Gen. Microbiol ., 81 :513-515.

Gunasekera T.S., Csonka L.N., Paliy O. 2008. Genome-wide transcriptional responses of Escherichia coli K-12 to continuous osmotic and heat stresses. J. Bacteriol . 190 (10):3712- 3720.

Gunter S.E., Khon H.I. 1955. The effect of X-Rays on the survival of bacteria and yeast. I. A comparative study of the dose-survival curves of Azotobacter agile, Escherichia coli, Pseudomonas fluoresceno, Rhodopseudomonas spaeroides and Saccharomyces cerevisae irradiated in the resting state. J. Bacteriol . 71 (5):571-81.

Gupta R.S., Mok A. 2007. Phylogenomics and signature proteins for the alpha-proteobacteria and its main groups. BMC Microbiol . 7:106.

Gustafson K.R., Cardellina J.H., Fuller R.W., Weislow O.S., Kiser R.F., Snader K.M, Patterson G.M., Boyd M.R. 1989. AIDS-antiviral sulfolipids from cyanobacteria (blue-green algae), J. Natl. Cancer Inst. 81 :1254-1258.

Gutman P.D., Fuchs P., Minton K.W. 1994. Restoration of the DNA damage resistance of Deinococcus radiodurans DNA polymerase mutants by Escherichia coli DNA polymerase I and Klenow fragment. Mutat. Res . 314 :87–97.

Gygi S.P., Rochon Y., Franza B.R., Aebersold R. 1999. Correlation between protein and mRNA abundance in yeast. Mol. Cell. Biol. 19 :1720–1730.

Hamill O. P., B. Martinac. 2001. Molecular basis of mechanotransduction in living cells. Physiol. Rev . 81 :685–740.

References 159 Hammond T. G., Benes E., O’Reilly K.C.,Wolf D.A., Linnehan R.M., Taher A., Kaysen J.H., Allen P.L., Goodwin T.J. 2000. Mechanical culture conditions effect gene expression: gravity-induced changes on the space shuttle. Physiol. Genomics 3:163–173.

Hammond T. G., Hammond J.M. 2001. Optimized suspension culture: the rotating-wall vessel. Am. J. Physiol. Ser. F . 281 :F12–F25.

Hansen K.H., Angelidaki I., Ahring B.K. 1999. Improving thermophilic anaerobic digestion of swine manure. Wat. Res . 33 :1805-1810.

Hantke, K. 2001. Iron and metal regulation in bacteria. Curr. Opin. Microbiol . 4: 172–177.

Heberer T., Reddersen K., Mechlinski A. 2002. From municipal sewage to drinking water: fate and removal of pharmaceutical residues in the aquatic environment in urban areas, Wat. Sci. Tech . 46 :81-88.

Hendrickson E.L., Xia Q., Wang T., Leigh J.A., Hackett M. 2006. Comparison of spectral counting and metabolic stable isotope labeling for use with quantitative microbial proteomics. Analyst 131 :1335-1341.

Hendrickx L., De Wever H., Hermans V., Mastroleo F., Morin N., Wilmotte A., Janssen P., Mergeay M. 2006. Microbial ecology of the closed artificial ecosystem MELiSSA (Micro- Ecological Life Support System Alternative): Reinventing and compartimentalizing the Earth's food and oxygen regeneration system for long-haul space exploration missions. Res. Microbiol. 157 (1):77-86.

Hendrickx L., Mergeay M. 2007. From the deep sea to the stars: human life support through minimal communities. Curr. Opin. Microbiol . 10 (3):231-237.

Hengge-Aronis R., Lange R., Henneberg N., Fischer D. 1993. Osmotic regulation of rpoS- dependent genes in Escherichia coli . J. Bacteriol . 175 (1):259-265.

Henrikson R., (Editor, Eds.), Book Earth Food Spirulina: How this remarkable blue-green Algae can transform your health and our planet, Ronore Enterprises, Inc, PO Box 3937, San Rafael, CA 94912 USA, 1997, pp. Pages.

Herbert B.R, Molloy M.P., Gooley A.A., Walsh B.J., Bryson W.G., Williams K.L. 1998. Improved protein solubility in two-dimentional electrophoresis using tributyl phosphine as reducing agent. Electrophoresis 19 (5):845-851.

Hessner MJ, Wejksnora PJ, Collins ML. 1991. Construction, characterization, and complementation of Rhodospirillum rubrum puf region mutants. J. Bacteriol . 173 (18):5712- 5722.

Holbrook R.D., Novak J.T., Grizaard T.J., Love N.G. 2002. Estrogen receptor agonist fate during waste water and biosolids treatment processes: a mass balance analysis. Env. Sci. Tech . 36 :4533-4539.

Holt J.W., Safaeinili A., Plaut J.J., Head J.W., Phillips R.J., Seu R., Kempf S.D., Choudhary P., Young D.A., Putzig N.E., Biccari D., Gim Y. 2008. Radar sounding evidence for buried glaciers in the southern mid-latitudes of Mars. Science 322 (5905):1235-1238.

References 160 Horneck G., Baumstark-Khan C., Reitz G. 2002. Space microbiology: effects of ionizing radiation on microorganisms in space. In Encyclopedia of Environmental Microbiology. John Wiley & Sons, Inc. DOI: 10.1002/0471263397.env003.

Horneck G., Facius R., Reitz G., Rettberg P., Baumstark-Khan, Gerzer R. 2001. Critical issues in connection with human planetary missions: protection of and from the environment. Acta Astronautica 49 (3-10):279-288.

Horneck G., Rettberg P., Baumstark-Khan C., Rink H., Kozubek S., Schäfer M., Schmitz C. 1996. DNA repair in microgravity: studies on bacteria and mammalian cells in the experiments REPAIR and KINETICS. J Biotechnol . 47 (2-3):99-112.

Hughes D.T., Sperandio V. 2006. Inter-kingdom signalling: communication between bacteria and their hosts. Nat. Rev. Microbiol . 6(2):111-120.

Hurst A.C., Petrov E., Kloda A., Nguyen T., Hool L., Martinac B. 2008. MscS, the bacterial mechanosensitive channel of small conductance. Int. J. Biochem. Cell. Biol . 40 (4):581-585.

Hwang W., Lee K.E., Lee J.K., Park B.C, Kim K.S. 2008. Genes of Rhodobacter sphaeroides 2.4.1 regulated by innate quorum-sensing signal, 7,8-cis-N-(tetradecenoyl) homoserine lactone. J. Microbiol. Biotechnol . 18 (2):219-227.

ICRP, Recommendations of the International Commission on Radiological Protection, Report 60, Pergamon Press, Oxford, U.K., 1991.

Idaho State University (ISU). "Radioactivity in nature", ISU connection website (August 2008). http://www.physics.isu.edu/radinf/natural.htm#top.

Imhoff J.F., Trüpfer H.G. 1989. "The purple non sulfur bacteria" in : Bergey's Manual of systematic bacteriology, vol. 3, Staley J.T., Bryant M.P., Pfennig N., Holt J.C. (eds). Williams and Wilkinm: Baltimore. 1658-1666.

Imlay J. A. 2002. How oxygen damages microbes: oxygen tolerance and obligate anaerobiosis. Adv. Microbial Phys . 46 :111–153.

Imlay J.A. 2003. Pathways of oxidative damage. Annu. Rev. Microbiol . 57 :395-418.

Ingber D. E. 1997. Integrins, tensegrity, and mechanotransduction. Grav. Space Biol. Bull . 10 :49–55.

Ingber D. 1999. How cells (might) sense microgravity. FASEB J. 13 :S3–S15.

Ishihama Y., Oda Y., Tabata T., Sato T., Nagasu T., Rappsilber J., Mann M. 2005. Exponentially modified protein abundance index (emPAI) for estimation of absolute protein amount in proteomics by the number of sequenced peptides per protein. Mol. Cell. Proteomics 4(9):1265-1272.

Ishihama Y., Schmidt T., Rappsilber J., Mann M., Hartl F.U., Kerner M.J., Frishman D. 2008. Protein abundance profiling of the Escherichia coli cytosol. BMC Genomics 9:102.

References 161 Jaffe J.D., Berg H.C., Church G.M. 2004. Proteogenomic mapping as a complementary method to perform genome annotation. Proteomics 4:59-77.

Jiang Z.Y., Rushing B.G., Bai Y., Gest H., Bauer C.E. 1998. Isolation of Rhodospirillum centeneum mutants defective in phototactic colony motility by transposon mutagenesis . J. Bacteriol. 180 :1248-1255.

Johanson K., Allen P.L., Lewis F., Cubano L.A., Hyman L.E., Hammond T.G. 2002. Saccharomyces cerevisiae gene expression changes during rotating wall vessel suspension culture. J. Appl. Physiol . 93 :2171–2180.

Julka S., Regnier F. 2004. Quantification in proteomics through stable isotope coding: a review. J. Proteome Res . 3(3):350-363.

Kacena M.A., Leonard P.E., Todd P., Luttges M.W. 1997. Low gravity and inertial effects on the growth of E. coli and B. subtilis in semi-solid media. Aviat. Space Environ. Med . 68 (12):1104-1108.

Kacena M.A., Merrell G.A., Manfredi B., Smith E.E., Klaus D.M., Todd P. 1999. Bacterial growth in space flight: Logistic growth curve parameters for Escherichia coli and Bacillus subtilis . Appl. Microbiol. Biotechnol . 51 : 229-234.

Kerby R.L., Ludden P.W., Roberts G.P. 1997. In vivo nickel insertion into the carbon monoxide dehydrogenase of Rhodospirillum rubrum : molecular and physiological characterization of cooCTJ. J. Bacteriol. 179 (7):2259-2266.

Khemiri A., Galland A., Vaudry D., Chan Tchi Song P., Vaudry H., Jouenne T., Cosette P. 2008. Outer-membrane proteomic maps and surface-exposed proteins of Legionella pneumophila using cellular fractionation and fluorescent labelling. Anal Bioanal Chem . 390 (7):1861-1871.

Kihara D., Kanehisa M. 2000. Tandem clusters of membrane proteins in complete genome sequences. Genome Res . 10 (6):731-743.

Klaus D.M. 2001. Clinostats and bioreactors . Gravit. and Space Biol. Bull . 14 : 55-64.

Klaus D. M. 2002. Space microbiology: microgravity and microorganisms, p. 2996–3004. In G. Bitton (ed.), Encyclopedia of environmental microbiology. John Wiley & Sons, Inc., New York, N.Y.

Klaus D.M., Howard H.N. 2006. Antibiotic efficacy and microbial virulence during space flight. Trends Biotechnol. 24 (3):131-136.

Klaus D., Simske S., Todd P., Stodieck L. 1997. Investigation of space flight effects on Escherichia coli and a proposed model of underlying physical mechanisms. Microbiology 143 :449-455.

Klaus D. M., Todd P., Schatz A. 1998. Functional weightlessness during clinorotation of cell suspensions. Adv. Space Res . 21 :1315–1318.

References 162 Kobayashi Y., Kikuchi M., Nagaoka S., Watanabe H. 1996. Recovery of Deinococcus radiodurans from radiation damage was enhanced under microgravity. Biol. Sci. Space 10 (2):97-101.

Kobayashi M., Kurata S. 1978. The mass culture and cell utilisation of photosynthetic bacteria. Process Biochem . 13 :27-30.

Kobayashi Y., Narumi I., Satoh K., Funayama T., Kikuchi M., Kitayama S., Watanabe H. 2004. Radiation response mechanisms of the extremely radioresistant bacterium Deinococcus radiodurans . Biol. Sci. Space 18 (3):134-135.

Kobayashi Y., Watanabe H., Kikuchi M., Narumi I. 2000. Effect of the space environment on the induction of DNA-repair related proteins and recovery from radiation damage. Adv. Space Res . 25 (10):2103-2106.

Kolker E., Purvine S., Galperin M.Y., Stolyar S., Goodlett D.R., Nesvizhskii A.I., Keller A., Xie T., Eng J.K., Yi E., Hood L., Picone A.F., Cherny T., Tjaden B.C., Siegel A.F., Reilly T.J., Makarova K.S., Palsson B.O., Smith AL. 2003 . Initial proteome analysis of model microorganism Haemophilus influenzae strain Rd KW20. J. Bacteriol . 185 :4593-4602.

Koops H.-P., Pommerening-Röser A. 2001. Distribution and ecophysiology of the nitrifying bacteria emphasizing cultured species. FEMS Microbiol. Ecol . 37 :1-9.

Krijgsveld J., Gauci S., Dormeyer W., Heck A.J. 2006. In-gel isoelectric focusing of peptides as a tool for improved protein identification. J. Proteome Res . 5(7):1721-1730.

Kvint K., Nachin L., Diez A., Nyström T. 2003. The bacterial universal stress protein: function and regulation. Curr. Opin. Microbiol . 6(2):140-145.

Lambina V.A. 1961. The effect of X-rays upon photosynthetizing purple sulphur bacteria. Mikrobiologiia 30 :601-606. [Article in Russian]

Lapchine L., Moatti N., Gasset G., Richoilley G., Templier J., Tixador R. 1986. Antibiotic activity in space. Drugs Exp. Clin. Res . 12 (12):933-938.

Laukel M., Rossignol M., Borderies G., Völker U., Vorholt J.A. 2004. Comparison of the proteome of Methylobacterium extorquens AM1 grown under methylotrophic and nonmethylotrophic conditions. Proteomics 4:1247-1264.

Lee Y., Pena-Llopis S., Kang Y-S., Shin H-D, Demple B., Madsen E.L., Jeon C.O., Park W. 2006. Expression analysis of the fpr (ferredoxin-NADP + reductase) gene in Pseudomonas putida KT2440. Biochem. Biophys. Research Comm . 339 (4):1246-1254.

Lehman L.J., Roberts G.P. 1991. Identification of an alternative nitrogenase system in Rhodospirillum rubrum . J. Bacteriol. 173 (17):5705-5711.

Leroy B., Roupie V., Noël-Georis I., Rosseels V., Walravens K., Govaerts M., Huygen K., Wattiez R. 2007. Antigen discovery: a postgenomic approach to paratuberculosis diagnosis. Proteomics 7(7):1164-1176.

References 163 Levina N., Tötemeyer S., Stokes N.R., Louis P., Jones M.A., Booth I.R. 1999. Protection of Escherichia coli cells against extreme turgor by activation of MscS and MscL mechanosensitive channels: identification of genes required for MscS activity. EMBO J . 18 (7):1730-1737.

Levy R. H., French F. W. 1968. Plasma radiation shield - Concept and applications to space vehicles. J. Spacecraft Rockets 5(5):570-577.

Leys N.M., Hendrickx L., De Boever P., Baatout S., Mergeay M. 2004. Space flight effects on bacterial physiology. J. Biol. Regul. Homeost. Agents 18 (2):193-199.

Link A.J., Eng J., Schieltz D.M., Carmack E., Mize G.J., Morris D.R., Garvik B.M., Yates J.R. 3rd. Direct analysis of protein complexes using mass spectrometry. Nat. Biotechnol . 17 (7):676-682.

Lipton M.S., Pasa-Tolic' L., Anderson G.A., Anderson D.J., Auberry D.L., Battista J.R., Daly M.J., Fredrickson J., Hixson K.K., Kostandarithes H., Masselon C., Markillie L.M., Moore R.J., Romine M.F., Shen Y., Stritmatter E., Tolic' N., Udseth H.R., Venkateswaran A., Wong K.K., Zhao R., Smith R.D. 2002. Global analysis of the Deinococcus radiodurans proteome by using accurate mass tags. Proc. Natl. Acad. Sci. U S A 99 (17):11049-11054.

Lopez-Marquez R.L., Pérez-Castiñeira J.R., Losada M., Serrano A. 2004. Differential regulation of soluble and membrane-bound inorganic pyrophosphatases in the photosynthetic bacterium Rhodospirillum rubrum provides insights into pyrophosphatebased stress bioenergetics. J. Bacter. 186 (16):5418-5426.

Lupo D., Ghosh R. 2004. The reaction center H subunit is not required for high levels of light- harvesting complex 1 in Rhodospirillum rubrum mutants. J. Bacteriol. 186 (17):5585-5595.

Lynch S.V., Matin A. Travails of microgravity: man and microbes in space. Biologist , 52 (2): may 2005.

Madigan M.T., Martinko J.M., Parker J. 2000. Brock Biology of Microorganisms. Ninth edition. Prentice Hall: Upper Saddle River, New Jersey. 991pp.

Manti L. 2006. Does reduced gravity alter cellular response to ionizing radiation. Radiat. Environ. Biophys . 45 :1-8.

Marco R., Lavàn D.A., van Loon J.J.W.A., Leandro L.J., Larkin O.J., Dijkstra C.D., Anthony P., Villa A., Davey M.R., Lowe K.C., Power J.B., Medina F.J. Drosophila melanogaster, a model system for comparative studies on the responses to real and simulated microgravity. J. Gravit. Physiol . 14 (1):P125-6.

McPherson A. 1993. Effects of a microgravity environment on the crystallization of biological macromolecules. Microgravity Sci Technol . 6(2):101-109.

Meloni M.A., Galleri G., Camboni M.G., Pippia P., Cogoli A., Cogoli-Greuter M. 2004. Modeled microgravity affects motility and cytoskeletal structures. J. Gravit. Physiol . 11(2):197-198.

References 164 Mergeay M. 1989. Bibliographic review of microbial growth and survival in space conditions, list of strains for use in MELiSSA. Technical note 1, ESA/ESTEC 8125/NL/FG.

Mergeay M., Lejeune P., Sadouk A., Gerits J., Fabry L. 1987. Shuttle transfer (or retrotransfer) of chromosomal markers mediated by plasmid pULB113. Mol. Gen. Genet ., 209 (1):61-70.

Mergeay M., Verstraete W., Dubertret G., Lefort-Tran M., Chipaux C., Binot R.A. 'MELiSSA' - A micro-organisms-based model for 'CELSS' development, proceedings at the 3rd European Symposium on Space Thermal Control & Life Support Systems, Noordwijk, The Netherlands 1988,pp.65-68.

Messner K. R., Imlay, J. A. 1999. The identification of primary sites of superoxide and hydrogen peroxide formation in the aerobic respiratory chain and sulfite reductase complex of Escherichia coli . J. Biol. Chem . 274 :10119–10128.

Mizuno T., Kageyama M. 1978. Separation and characterization of the outer membrane of Pseudomonas aeruginosa . J. Biochem . 84 (1):179-191.

Molloy M.P., Herbert B.R., Walsh B.J., Tyler M.I., Traini M., Sanchez J.C., Hochstrasser D.F., Williams K.L., Gooley A.A. 1998. Extraction of membrane proteins by differential solubilization for separation using two-dimensional gel electrophoresis. Electrophoresis , 19 (5): 837-844.

Moore M.D., Kaplan S. 1992. Identification of intrinsic high-level resistance to rare-earth oxides and oxyanions in members of the class proteobacteria: characterization of tellurite, selenite, and rhodium sesquioxide reduction in Rhodobacter sphaeroides. J. Bact . 174 (5):1505-14.

Mortz E., Krogh T.N., Vorum H., Görg A. 2001. Improved silver staining protocols for high sensitivity protein identification using matrix-assisted laser desorption/ionization-time of flight analysis. Proteomics 1(11):1359-1363.

NASA's 2004 Vision for Space Exploration website ( connection 01-12-2009): http://www.nasa.gov/missions/solarsystem/explore_main.html

NASA connection website (July 2005): http://marie.jsc.nasa.gov/Data/Index.cfm.

NASA, http://advlifesupport.jsc.nasa.gov/

Nauser T., Koppenol W.H., Gebicki JM. 2005. The kinetics of oxidation of GSH by protein radicals. Biochem J . 392 (Pt 3):693-701.

Nealson K.H. 1979. Bioluminescence. Science 205 (4409):893.

Nicholson W.L., Munakata N., Horneck G., Melosh H.J., Setlow P. 2000. Resistance of Bacillus endospores to extreme terrestrial and extraterrestrial environments. Microbiology and Molecular Biology Rev . 64 (3):548-572.

References 165 Nicholson W.L., Schuerger A.C., Setlow P. 2005. The solar UV environment and bacterial spore UV resistance: considerations for Earth-to-Mars transport by natural processes and human space flight. Mutat Res. 571 (1-2):249-264.

Nickerson C. A., Ott C.M., Mister S.J., Morrow B.J., Burns-Keliher L., Pierson D.L.. 2000. Microgravity as a novel environmental signal affecting Salmonella enterica serovar Typhimurium virulence. Infect. Immun . 68 :3147–3152.

Nickerson C. A., Ott C.M., Wilson J.R., Ramamurthy R., LeBlanc C.L., K. Honer zu Bentrup, Hammond T., Pierson D.L. 2003. Low-shear modeled microgravity: a global environmental regulatory signal affecting bacterial gene expression, physiology, and pathogenesis. J . Microbiol. Methods 54 :1–11.

Nickerson C.A., Ott C.M., Wilson J.W., Ramamurthy R., Pierson D.L. 2004. Microbial Responses to Microgravity and Other Low-Shear Environments. Microbiology and molecular biology reviews 68 (2):345–361.

Nie L., Wu G., Zhang W. 2006a. Correlation of mRNA expression and protein abundance affected by multiple sequence features related to translational efficiency in Desulfovibrio vulgaris : a quantitative analysis. Genetics 174 (4):2229-2243.

Nie L., Wu G., Zhang W. 2006b. Correlation between mRNA and protein abundance in Desulfovibrio vulgaris : a multiple regression to identify sources of variations. Biochem. Biophys. Res. Commun . 339 :603–610.

Ning C., Gest H. 1966. Regulation of L-isoleucine biosynthesis in the photosynthetic bacterium Rhodospirillum rubrum. Proc. Natl. Acad. Sci. USA 56 :1823-1827.

Olsen R.H., and Shipley P. 1973. Host range and properties of the Pseudomonas aeruginosa R factor R1822. J. Bacteriol . 113 :772-80.

Ouchane S., Picaud M. Vernotte C, Astier C. 1997. Photooxidative stress stimulates illegitimate recombination and mutability in carotenoid-less mutants of Rubrivivax gelatinosus . The EMBO journal , 16 (15):4777-87.

Pan C., Oda Y., Lankford P.K., Zhang B., Samatova N.F., Pelletier D.A., Harwood C.S., Hettich R.L. 2008. Characterization of anaerobic catabolism of p-coumarate in Rhodopseudomonas palustris by integrating transcriptomics and quantitative proteomics. Mol. Cell. Proteomics 7(5):938-948.

Parker E.N. 2006. Shielding space travelers. Sci. Am . 294 (3):40-47.

Patel M.J., Liu W., Sykes M.C., Ward N.E., Risin S.A., Risin D, Jo H. 2007. Identification of mechanosensitive genes in osteoblasts by comparative microarray studies using the rotating wall vessel and the random positioning machine. J. Cell. Biochem . 101 (3):587-599.

Peng J., Elias J.E., Thoreen C.C., Licklider L.J., Gygi S.P. 2003. Evaluation of multidimensional chromatography coupled with tandem mass spectrometry (LC/LC-MS/MS) for large-scale protein analysis: the yeast proteome. J. Proteome Res . 2(1):43-50.

References 166 Piconese S., Tronelli G., Pippia P., Migliaccio F. 2003. Chiral and non-chiral nutations in Arabidopsis roots grown on the random positioning machine. J. Exp. Bot . 54 (389):1909-1918.

Picorel, R., G. Belanger, and G. Gingras. 1983. Antenna holochrome B880 of Rhodospirillum rubrum S1. Pigment, phospholipid composition, and polypeptide composition. Biochemistry 22 :2491-2497.

Picorel R., L'Ecuyer A., Potier M., Gingras G. 1986. Structure of the B880 holochrome of Rhodospirillum rubrum as studied by radiation inactivation method. J. Biol. Chem. 261 (5):3020-3024.

Pivetti C. D., Yen M.R., Miller S., Busch W., Tseng Y.H., Booth I.R., Saier M.H. Jr. 2003. Two families of mechanosensitive channel proteins. Microbiol. Mol. Biol. Rev . 67 :66–85.

Prosser J.I., in: (A.H. Rose and D.W. Tempest, Eds.), Advances in Microbial Physiology, Autotrophic nitrification in bacteria, Academic Press Inc., New York, 1989, pp.125-181.

Puskas A., Greenberg E.P., Kaplan S., Schaefer A.L. 1997. A quorum-sensing system in the free-living photosynthetic bacterium Rhodobacter sphaeroides . J. Bacteriol . 179 (23):7530- 7537.

Pyle B., Vasques M., Aquilina R. The effect of microgravity on the smallest space travelers. Bacterial physiology and virulence on Earth an in microgravity. National Aeronautics and Space Administration (NASA) 2002.

Qiu X., Daly M.J., Vasilenko A., Omelchenko M.V., Gaidamakova E.K., Wu L., Zhou J., Sundin G.W., Tiedje J.M. 2006. Transcriptome analysis applied to survival of Shewanella oneidensis MR-1 exposed to ionizing radiation. J. Bacteriol 188 (3):1199-1204.

Rea G., Esposito D., Damasso M., Serafini A., Margonelli A., Faraloni C., Torzillo G., Zanini A., Bertalan I., Johanningmeier U., Giardi M.T. 2008. Ionizing radiation impacts photochemical quantum yield and oxygen evolution activity of Photosystem II in photosynthetic microorganisms. Int. J. Radiat. Biol . 84 (11):867-877.

Reading N.C., Sperandio V. 2006. Quorum sensing: the many languages of bacteria. FEMS Microbiol. Lett . 254 (1):1-11.

Ritchie M. E., Silver J., Oshlack A., Holmes M., Diyagama D., Holloway A., Smyth G. K. 2007. A comparison of background correction methods for two-colour microarrays. Bioinformatics 23 (20):2700-2707.

Rohmer L., Guina T., Chen J., Gallis B., Taylor G.K., Shaffer S.A., Miller S.I., Brittnacher M.J., Goodlett D.R. 2008. Determination and comparison of the Francisella tularensis subsp.novicida U112 proteome to other bacterial proteomes. J. Proteome Res . 7(5):2016- 2024.

Shiner E.K., Rumbaugh K.P., Williams S.C. 2005. Inter-kingdom signaling: deciphering the language of acyl homoserine lactones. FEMS Microbiol. Rev . 29 (5):935-947.

References 167 Qiu X., Daly M.J., Vasilenko A., Omelchenko M.V., Gaidamakova E.K., Wu L., Zhou J., Sundin G.W., Tiedje J.M. 2006. Transcriptome analysis applied to survival of Shewanella oneidensis MR-1 exposed to ionizing radiation. J Bacteriol. 188 :1199–1204.

Saegesser S., Ghosh R., Bachofen R.. 1992. Stability of broad host range cloning vectors in the phototrophic bacterium Rhodospirillum rubrum. FEMS Microbiol. Lett . 95 (1):7-12.

Saganti P.B., Cucinotta F.A., Wilson J.W., Simonsen L.C., Zeitlin C. 2002. "Radiation climate map for analyzing risks to astronauts on the mars surface from galactic cosmic rays." NASA Johnson Space Center, Martian Radiation Environment Experiment: MARIE website connection (july 2005): http://marie.jsc.nasa.gov/.

Sautier C. and Tremolieres J. 1976. Valeurs alimentaires des algues spirulines chez l'homme. Ann. Nutr. Alim . 30 :517-533.

Schmidt A., Kellermann J., Lottspeich F. 2005. A novel strategy for quantitative proteomics using isotope-coded protein labels. Proteomics 5(1):4-15.

Sedlak R. (Eds.), Phosphorus and nitrogen removal from municipal wastewater: Principles and practice, Lewis Publishers, New York, 1991.

Selao T.T., Nordlund S., Norén A. 2008. Comparative Proteomic Studies in Rhodospirillum rubrum Grown under Different Nitrogen Conditions. J. Proteome Res . 7(8):3267-3275.

Semsey S., Blaha B., Köles K., Orosz L., and Papp P.P. 2002. Site-specific integrative elements of rhizobiophage 16-3 can integrate into proline tRNA (CGG) genes in different bacterial genera. J. Bacteriol . 184 :177-182.

Sistrom W. R. 1960. A requirement for sodium in the growth medium of Rhodopseudomonas sphaeroides . J. Gen. Microbiol . 22 :77-85.

Schwartz R., Ting C.S., King J. 2001. Whole proteome pI values correlate with subcellular localizations of proteins for organisms within the three domains of life. Genome Res . 11 (5):703-709.

Schmidt A., Kellermann J., Lottspeich F. 2005. A novel strategy for quantitative proteomics using isotope-coded protein labels. Proteomics 5(1):4-15.

Smirnova G.V., Muzyka N.G., Oktyabrsky O.N. 2000. The role of antioxidant enzymes in response of Escherichia coli to osmotic upshift. FEMS Microbiol. Lett . 186 (2):209-213.

Smith R.D., Anderson G.A., Lipton M.S., Pasa-Tolic L., Shen Y., Conrads T.P., Veenstra T.D., Udseth H.R. 2002. An accurate mass tag strategy for quantitative and high-throughput proteome measurements. Proteomics 2(5):513-523.

Smith M.R., Zinder S.H., Mah R.A. 1980. Microbial methanogenesis from acetate. Process. Biochem . 15 :34-39.

Smyth G. K. 2004. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol . 3:Article3.

References 168 Smyth, G. K. 2005. Limma: linear models for microarray data, pp. 397-420. In R. Gentleman, V. J. Carey, S. Dudoit, R. A. Irizarry, and R. Huber (Eds): Bioinformatics and Computational Biology Solutions using R and Bioconductor , Springer, New York.

Smyth G. K., Michaud J., Scott H. S. 2005. Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics 21 (9), 2067-2075.

Smyth G. K., Speed T. 2003. Normalization of cDNA microarray data. Methods 31 (4):265- 273.

Sounders V.A. 1978. Genetics of Rhodospirillaceae. Microbiol. Rev . 42 :357-384.

Springael A., Diels L., Mergeay L. 1994. Transfer and expression of PCE-degradative genes into heavy metal resistant Alcaligenes eutrophus strain. Biodegradation , 5(3-4), 343-357.

Summers A.O., Jacoby G.A. 1977. Plasmid-determined resistance to tellurium compounds. J. Bacteriol. 129 (1):276-281.

Tabak H.H., Bunch R.L. 1970. Steroid hormones as water pollutants. I. Metabolism of natural and synthetic ovulation-inhibiting hormones by micro-organisms of activated sludge and primary settled sewage. Dev. Ind. Microbiol . 11 :367-376.

Taylor D.E. 1999. Bacterial tellurite resistance. Trends Microbiol . 7(3):111-115.

Taylor G.R., Konstantinova I., Sonnenfeld G., Jennings R. 1997. Changes in the immune system during and after space flight. Adv. Space Biol . Med. 6:1-32.

Thévenet D., D'Ari R., Bouloc P. 1996. The SIGNAL experiment in BIORACK: Escherichia coli in microgravity. J. Biotechnol . 47 (2-3):89-97.

Thomas W. E., Trintchina E., Forero M., Vogel V., Sokurenko E.V. 2002. Bacterial adhesion to target cells enhanced by shear force. Cell 109: 913–923.

Tixador R., Richoilley G., Gasset G., Templier J., Bes J.C., Moatti N., Lapchine L. 1985. Study of minimal inhibitory concentration of antibiotics on bacteria cultivated in vitro in space (Cytos 2 experiment). Aviat Space Environ Med . 56 (8):748-51.

Tong W., Link A., Eng J.K., Yates J.R. 3rd. 1999. Identification of proteins in complexes by solid-phase microextraction/multistep elution/capillary electrophoresis/tandem mass spectrometry. Anal Chem . 71 (13):2270-2278.

Tranquille N., Emeis J.J. 1996a. R. rubrum food acceptability study for MELiSSA. Technical note 30.1., PRF 151712.

Tranquille N., Emeis J.J. 1996b. R. rubrum food acceptability study for MELiSSA. Technical note 30.2., PRF 151712.

Tucker D.L., Ott C.M., Huff S., Fofanov Y., Pierson D.L., Willson R.C., Fox G.E. 2007. Characterization of Escherichia coli MG1655 grown in a low-shear modeled microgravity environment. BMC Microbiol . 7:7-15.

References 169 Urban J.E. 2000. Adverse effects of microgravity on the magnetotactic bacterium Magnetospirillum magnetotacticum . Acta Astronaut. 47 (10):775-780.

Vadeboncoeur, C., H. Noel, L. Poirier, Y. Cloutier, and G. Gingras. 1979. Photoreaction center of photosynthetic bacteria. 1. Further chemical characterization of the photoreaction center from Rhodospirillum rubrum . Biochemistry 18 :4301-4308.

Vallenet D., Labarre L., Rouy Z., Barbe V., Bocs S., Cruveiller S., Lajus A., Pascal G., Scarpelli C., Médigue C. 2006. MaGe: a microbial genome annotation system supported by synteny results. Nucleic Acids Res . 34 (1):53-65.

Vana N., Hajek M., Berger T., Fugger M., Hofmann P. 2006. Novel shielding materials for space and air travel. Radiat. Prot. Dosimetry 120 (1-4):405-409.

Van Der Rest, M., and G. Gingras. 1974. The pigment complement of the photosynthetic reaction center isolated from Rhodospirillum rubrum. J. Biol. Chem . 249 :6446-6453.

Vanhavere F., Genicot J.L., O'Sullivan D., Zhou D., Spurny F., Jadrnickova I., Sawakuchi G.O., Yukihara, E.G. 2008. DOsimetry of BIological EXperiments in SPace (DOBIES) with luminescence (OSL and TL) and track etch detectors. Radiation Measurements 43 (2):694- 697.

Van Niel C.B. 1944. The culture, general physiology, morphology, and classification of the non-sulfur purple and brown bacteria. Bact. Rev. :1-118.

Vecerek B., Moll I., Afonyushkin T., Kaberdin V., Bläsi U. 2003. Interaction of the RNA chaperone Hfq with mRNAs: direct and indirect roles of Hfq in iron metabolism of Escherichia coli. Mol. Microbiol . 50 (3):897-909.

VerBerkmoes N.C., Shah M.B., Lankford P.K., Pelletier D.A., Strader M.B., Tabb D.L., McDonald W.H., Barton J.W., Hurst G.B., Hauser L., Davison B.H., Beatty J.T., Harwood C.S., Tabita F.R., Hettich R.L., Larimer F.W. 2006. Determination and comparison of the baseline proteomes of the versatile microbe Rhodopseudomonas palustris under its major metabolic states. J. Proteome Res . 5, 287-298.

Vermeulen W., Hoeijmakers J., Boostma D. 1996. HELIX, Amgens Magazine Biotechnol. 5:22-29.

Verstraete W., Vandevivere P. 1999. New and broader applications of anaerobic digestion. Crit. Rev. Env. Sci. Tech . 28 :151-173.

Vrati S. 1984. Single cell protein production by photosynthetic bacteria grown on the clarified effluents of a biogas plant. Appl. Microbiol. Biotechnol. 19 :199–202.

Vukanti R., Mintz E., Leff L. 2008. Changes in Gene Expression of E. coli under Conditions of Modeled Reduced Gravity. Microgravity Sci. Technol. 20 :41–57.

Wagner M., Rath G., Amann R., Koops H-P., Schleiffer K.H. 1995. In situ dentification of ammonia-oxidizing bacteria, System . Appl. Microbiol . 18 :251-264.

References 170 Wang Y., Mulligan C., Denyer G., Delom F., Dagna-Bricarelli F., Tybulewicz V.L., Fisher E.M., Griffiths W.J., Nizetic D., Groet J. 2008. Quantitative proteomic characterisation of a mouse embryonic stem cell model of Down syndrome. Mol. Cell. Proteomics . In press.

Wang J., Sung W.K., Krishnan, A., Li K.B. 2005. Protein subcellular localization prediction for Gram-negative bacteria using amino acid subalphabets and a combination of multiple support vector machines. BMC Bioinformatics 6:174.

Washburn M.P., Wolters D., Yates J.R. 3rd. 2001. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol . 19 (3):242-247.

Watanabe K., Kitamura T., Ochi S., Ozaki M. 1997. Inactivation of pathogenic bacteria under mesophilic and thermophilic conditions. Wat. Sci. Tech . 36 :25-32.

Wessel D., Flugge, U. I. 1984. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Analytical biochemistry 138 : 141-143.

Wilson J.W., Ott C.M., Höner zu Bentrup K., Ramamurthy R., Quick L., Porwollik S., Cheng P., McClelland M., Tsaprailis G., Radabaugh T., Hunt A., Fernandez D., Richter E., Shah M., Kilcoyne M., Joshi L., Nelman-Gonzalez M., Hing S., Parra M., Dumars P., Norwood K., Bober R., Devich J., Ruggles A., Goulart C., Rupert M., Stodieck L., Stafford P., Catella L., Schurr M.J., Buchanan K., Morici L., McCracken J., Allen P., Baker-Coleman C., Hammond T., Vogel J., Nelson R., Pierson D.L., Stefanyshyn-Piper H.M., Nickerson C.A. 2007. Space flight alters bacterial gene expression and virulence and reveals a role for global regulator Hfq. Proc. Natl. Acad. Sci . U S A 104 (41):16299-304.

Wilson J.W., Ott C.M., Quick L., Davis R., Höner zu Bentrup K., Crabbe A. et al . 2008. Media ion composition controls regulatory and virulence response of Salmonella in space flight. PLoS ONE , 3(12):e3923.

Wilson J.W., Ott C.M., Ramamurthy R., Porwollik S., McClelland M., Pierson D.L., Nickerson C.A. 2002. Low-shear modeled microgravity alters the Salmonella enterica serovar Typhimurium stress response in an RpoS- independent manner. Appl. Environ. Microbiol. 68 :5408–5416.

Wilson J.W., Ramamurthy R., Porwollik S., McClelland M., Hammond T., Allen P., Ott C.M., Pierson D.L., Nickerson C.A. 2002. Microarray analysis identifies Salmonella genes belonging to the low-shear modeled microgravity regulon. Proc. Natl. Acad. Sci. USA 99 :13807–13812.

Wood J.M. 1999. Osmosensing by bacteria: signals and membrane-based sensors. Microbiol. Mol. Biol. Rev. 63 (1):230-262.

Wu J.J., Ma J.T., Pan R.L. 1991. Functional size analysis of pyrophosphatase from Rhodospirillum rubrum determined by radiation inactivation. FEBS letter 283 (1):57-60.

Yang H.F., Huang Z.Y., Zhang D.M., Liu Z.H. 1999. [Biological effects of space flight on purple nonsulfur photosynthetic bacteria]. [Article in Chinese]. Space Med Med Eng (Beijing) 12 (1):46-50.

References 171 Yang Y. H., Thorne N. P. 2003. Normalization for two-color cDNA microarray data. In: D. R. Goldstein (ed.), Science and Statistics: A Festschrift for Terry Speed, IMS Lecture Notes - Monograph Series, Volume 40, pp. 403-418.

Yates E.A., Philipp B., Buckley C., Atkinson S., Chhabra S.R., Sockett R.E., Goldner M., Dessaux Y., Cámara M., Smith H., Williams P. 2002. N-acylhomoserine lactones undergo lactonolysis in a pH-, temperature-, and acyl chain length-dependent manner during growth of Yersinia pseudotuberculosis and Pseudomonas aeruginosa . Infect. Immun . 70 (10):5635-5646.

Yildiz F.H., Gest H., Bauer C.E. 1991. Genetic analysis of photosynthesis in Rhodospirillum rubrum. J. Bacteriol. 173 (13):4163-4170.

Zhang Y., Burris R.H., Roberts G.P. 1992. Cloning, sequencing, mutagenesis, and functional characterization of draT and draG genes from brasilense. J. Bacteriol . 174 (10):3364-3369.

Zhang W., Gritsenko M.A., Moore R.J., Culley D.E., Nie L., Petritis K., Strittmatter E.F. Camp D.G. 2 nd , Smith R.D., Brockman F.J. 2006. A proteomic view of Desulfovibrio vulgaris metabolism as determined by liquid chromatography coupled with tandem mass spectrometry. Proteomics 6:4286-4299.

Zhang Y., Pohlmann E.L., Ludden P.W., Roberts G.P. 2000. Mutagenesis and functional characterization of the glnB, glnA, and nifA genes from the photosynthetic bacterium Rhodospirillum rubrum. J. Bacteriol. 182 (4):983-992.

Zimmann P., Steinbrügge A., Schniederberend M., Jung K., Altendorf K. 2007. The extension of the fourth transmembrane helix of the sensor kinase KdpD of Escherichia coli is involved in sensing . J. Bacteriol . 189 (20):7326-7334.