.

Doctoral Thesis under a Cotutelle agreement between l’Université Pierre et Marie Curie and The University of New South Wales

French laboratory : Laboratoire d’Océanographie Biologique de Banyuls, Observatoire Océanologique. Université Pierre et Marie Curie. France. Ecole Doctorale B2M – Biochimie et Biologie Moléculaire Spécialité : Génome et Protéines. Australian laboratory: School of Biotechnology and Biomolecular Sciences. Faculty of Science. The University of New South Wales. Sydney, Australia.

PHYSIOLOGICAL AND MOLECULAR RESPONSES OF THE MARINE OLIGOTROPHIC ULTRAMICROBACTERIUM SPHINGOPYXIS ALASKENSIS RB2256 TO VISIBLE LIGHT AND ULTRAVIOLET RADIATION

Sabine MATALLANA SURGET

A thesis submitted in fulfilment of the requirements for the degrees of Doctor of Philosophy awarded by both UPMC and UNSW

May, 2009 CERTIFICATE OF ORIGINALITY

I hereby declare that this submission is my own work and that, to the best of my knowledge, it contains no material previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.

Sabine MATALLANA SURGET

I COPYRIGHT STATEMENT

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed Date : 14-09-09

AUTHENTICITY STATEMENT

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed Date : 14-09-09

II ABSTRACT

Ultraviolet radiation reaching the Earth’s surface (UVR, 280-400 nm) may penetrate deep into the clear oligotrophic waters influencing a large part of the euphotic layer. Marine heterotrophic bacteria at the surface of the oceans are especially sensitive to the damaging solar radiation due to their haploid genome with little or no functional redundancy and lack of protective pigmentation. In a context of climate change and ozone depletion, it is clearly important to understand the physiology and underlying molecular UVR responses of abundant marine bacteria species. We chose the marine ultramicrobacterium Sphingopyxis alaskensis as a reference species to study the impact of solar radiation due to its numerical abundance in oligotrophic waters and its photoresistance, previously reported. For this purpose, we focused on the formation of the two major UVB-induced DNA photoproducts (CPDs and 6-4PPs) as well as the differential protein expression under solar radiation.

We first demonstrated that the GC content of prokaryotic genome had a major effect on the formation of UVB-induced photoproducts, quantified by HPLC-MS/MS. Due to its high GC content, S. alaskensis presented a favoured formation of highly mutagenic cytosine-containing photoproducts and therefore would be more susceptible to UV- induced mutagenesis. By comparing S. alaskensis to another marine bacterium Photobacterium angustum, we observed for the latter strain a remarkable resistance to high UVB doses associated with a decrease in the rate of formation of CPDs explained by a non-conventional activity of photolyase. We also demonstrated that DNA damage in S. alaskensis was markedly modulated by growth temperature and time spent in stationary phase.

In order to assess the effects that environmental UV-R had on regulatory networks and pathways of S. alaskensis, and determine how the cell’s physiology was affected, a quantitative proteomics investigation was performed. Changes in proteome were analyzed, with the recent and powerful mass spectrometry based approach using iTRAQ methodology. Approximately, one third of the proteome of S. alaskensis was identified, with 119 statistically and significantly differentially abundant proteins. Cellular processes, pathways and interaction networks were determined and gave us unique insight into the biology of UV response and adaptation of S. alaskensis.

III ACKNOWLEDGEMENTS

I am first very grateful to Philippe Lebaron for welcoming me in his lab in Banyuls, when I first started my Master Degree and for encouraging me to do a PhD. Doing my PhD under a cotutelle agreement between France and Australia was a huge and wonderful positive experience. My PhD was supported by fellowship from the French Ministry, a CNRS-INSU through an ATIPE project and by the Australian Research Council. I have had the pleasure of working with and learning from many people. First and foremost, I have constantly felt throughout my PhD, that I was extremely lucky to have not one, but two very supportive and encouraging supervisors. Both Fabien Joux and Rick Cavicchioli have provided invaluable advice on all of the work I did in this thesis. They were always available throughout all stages of my PhD to meet and discuss results and ideas. Rick and Fabien, together your enthusiasm for combining the disciplines of microbial ecology and molecular approaches has created a stimulating environment that enabled our multidisciplinary work to be realised. Your trust in science, belief in me and the freedom you have allowed me have encouraged me to think creatively and take initiative in the lab. I thank you both warmly for a thoroughly enjoyable three and a half years.

Thank you very much to Mark Raftery for teaching me mass spectrometry and for his detailed assistance with the computers of BMSF. Thank you also for your patience and availability. I am also very touched that you have accepted our invitation to come in France, to be part of my thesis panel review. I am very grateful for this. Many thanks to Thierry Douki for collaborating on my thesis with the HPLC-MS/MS analysis of the DNA damage. I have learned a lot. It has been such a pleasure working with you and I am very glad to welcome you in Banyuls for my thesis defence. I am very grateful to Wade Jeffrey and Gérard Fonty for having accepted reviewing my manuscript and to be part of my thesis panel review.

I would also like to thank Jarah Meador for staying with us in Banyuls for few weeks and working with me on S. alaskensis. I keep excellent memories of working together. It was also very enjoyable working with Maher Abboudi on the optimization of ELISA methodology. Thank you Maher for your calmness. To Laurent Urios, your encouragement and support, especially in my early years as a scientist, was very much appreciated and will always be remembered.

Doing a PhD under a cotutelle agreement between two laboratories does not only allow to learn a lot on a scientific point of view but also to meet a lot of very nice people. Because I began my studies in Banyuls, first I would like to thank for their friendship and their help the following students that I met there, namely: Arturo Rodriguez Blanco, Elodie Peyric, Antoine Aze, Antoine Carlier, Olivier Zemb, Mickael Moulager, Raphaël Lami, Benoît Farinas, Marc Auffret, Robin Vuilleumier, Nathalie Parthuisot, Laure Bellec, Sarah Nahon, Charlotte Moritz, Mathieu Chatelain, Mélissa Laghdass and Caroline Sauret. Olivier it was funny to meet you in Banyuls and to become closer friend in Sydney, how small is the world! Thanks for all the good moments, we shared together. I especially thank Arturo, who has been down the long road with me. Thanks for your help and precious encouragement, our wonderful moments sharing together. Thanks for your love, support and the laughs. Thanks for too many things to mention here. To you I am forever grateful!

IV I am very grateful to all the members of the “microbes” team, for their invaluable help and friendship during my PhD: namely Laurent Intertaglia, Cécile Vilette, Audrey Calvez, Nicole Batailler, Philippe Catala, Muriel Bourrain, Julia Baudart, Delphine Guillebault, Jeff Ghiglione, Ingrid Obernosterer, Carmem Manes and Sébastien Peuchet. It was always really pleasant to work with enthusiastic people, who eagerly exchange ideas in a specific as well as broader context. Thanks to Gérard Peaucellier for our interesting chats on Proteomics.Thanks to Eliane Rubio and Gilles Vétion that were always on hand to help with the truly enthralling day-to-day activities involved in a PhD.

Thanks to Dominic Mooney, Kylie Jones, Claire Biron, Frederique Sanz and Zohra Boumédine for patiently dealing with all the administration that have occurred over the years of my PhD. Thanks also to Owen Parkes for proof-reading my introduction.

I would like to acknowledge sincerely all the members of Rick Cavicchioli’s team. It has been rewarding working with people from different scientific and cultural backgrounds and with whom I have shared a lot of laughs. Thanks to all the following friends that I have met through the UNSW and who have been a great source of support and are a reason for why I have absolutely enjoyed my PhD. Thanks to Dominic Burg for teaching me iTRAQ and showing me everything in the lab, when I first arrived. I am very grateful for all you have done, working with you on proteomics has certainly made it more enjoyable. Thanks for teaching me the real Aussie attitude. I keep excellent memories of all the “happy hours” spent together. Thanks for your friendship. I will always remember how much you and Iona were hospitable to me! Special thanks to Davide De Francisci, for the intellectual debate and your hypothesis-driven approach to science and life, the coffee breaks and the laughs! Thanks for all the wonderful week- ends (Shark Island was one of the best). Thanks to Charmaine for her kindness, for always being ready to help anyone of the team, and especially me. Thanks to Kevin for organizing Whales watching, a great experience. Thanks to Matt DeMaere, for being very welcoming and for sharing together good coffee breaks. I am also very grateful for the assistance of Lily Ting in teaching me manual annotation, but mostly for her friendship, and all of her “frenchy” interest. Thanks to Federico Lauro for computational analysis and to provide a list of parameters very useful for my proteomics analysis.

Thanks to Susan and Zoran’s family, who hosted me at their home, where they always made me feel welcome. Thank you so much for showing me around Sydney and for my birthday celebration! To my flatmates in Kingsford, Hiroko, Koji and especially Leo, who have given never-ending support and shared many a good time with me, I thank you.

To my brother Jean-Baptiste, your encouragement and friendship is truly valued. Thanks for visiting me in Sydney, our week-end spent in Melbourne have been some of the most relaxing times over the last few years! I keep wonderful memories of us down- under. To my sweet mum and dad, Sabine and Antonio, I am forever indebted. You have instilled in me the ethic of always completing what’s begun and give me the ability to overcome hardships and carry on when life gets tough. You have endured the ups and downs with me, always providing love, patience, care and support. This PhD is definitely a result of your many years of love and encouragement. Thank you for always being there for me!

V PUBLICATIONS ARISING FROM THIS WORK

In Preparation

[1] Matallana-Surget S., T. Douki, J. Meador, R. Cavicchioli and F. Joux, Temperature-dependent formation of bipyrimidine photoproducts by UVB radiation in different conditions of starvation and light irradiation in the marine bacterium Sphingopyxis alaskensis, J. Photochem. Photobiol. B: Biol.

Published

[2] Matallana-Surget S., F. Joux, M. J. Raftery and R. Cavicchioli, The response of the marine bacterium Sphingopyxis alaskensis to solar radiation assessed by quantitative proteomics, Environ. Microbiol. 2009, 11, 2660–2675.

[3] Matallana-Surget S., T. Douki, R. Cavicchioli and F. Joux, Remarkable resistance to UVB of the marine bacterium Photobacterium angsutum explained by an unexpected role of photolyase, Photochem Photobiol Sci., 2009, 8, 1313-1320.

[4] Matallana-Surget S., J. A. Meador, F. Joux and T. Douki, Effect of the GC content of DNA on the distribution of UVB-induced bipyrimidine photoproducts, Photochem. Photobiol. Sci., 2008, 7, 794–801.

[5] Abboudi M., S. Matallana-Surget, J-F Rontani, R. Sempéré and F. Joux, Physiological alteration of the marine bacterium Vibrio angustum S14 exposed to simulated sunlight during growth, Curr. Microbiol., 2008, 57, 412–417.

[6] Matallana-Surget S., F. Joux, P. Lebaron and R. Cavicchioli, Isolation and characterization of oligotrophic marine bacteria. J. Soc. Biol., 2007, 201, 41–50.

VI CONFERENCE PROCEEDINGS

[1] Matallana-Surget S., F. Joux, M. J. Raftery and R. Cavicchioli, UV response of the marine oligotrophic ultramicrobacterium Sphingopyxis alaskensis, assessed by a quantitative proteomic approach using ITRAQ. (Oral communication-invited speaker) International conference on Proteomics in Plants, Microorganisms and Environment- Proteomlux 2008, organized by the Centre de Recherche Public Gabriel Lippmann, Luxembourg, October 22/25, 2008. [Awarded a complete grant].

[2] Douki T., S. Matallana-Surget, J. A. Meador and F. Joux. G+C content of genomes is a key parameter for the distribution of UVB-induced bipyrimidine DNA photoproducts. (Oral communication). 12th Congress of the American Society for Photobiology, University of Burlingame, CA. June 20-25, 2008.

[3] Matallana-Surget S., F. Joux, J. A. Meador and T. Douki. UVB induced DNA damage: high proportion of mutagenic photoproducts in GC-rich genomes. (Poster presentation). 12th Congress of the European Society for Photobiology, University of Bath, UK, September 1/6, 2007.[Awarded a Fellowship].

[4] Matallana-Surget S., F. Joux, R. Cavicchioli and T. Douki, How the DNA composition and the physiological state of marine bacteria may modify their DNA damages and their sensitivity to UV radiation. (Oral communication). 12thCongress of the European Society for Photobiology, University of Bath, UK, September 1/6, 2007. [Awarded a Fellowship].

[5] Matallana-Surget S., F. Joux and P. Lebaron. Importance of Oligotrophic Marine Bacteria. (Oral communication). Séance de la Société de Biologie. Océan et recherche biomédicale. Banyuls-sur-Mer, France, May 19, 2006.

VII TABLE OF CONTENTS

CERTIFICATE OF ORIGINALITY I COPYRIGHT / AUTHENTICITY STATEMENT II ABSTRACT III ACKNOWLEDGEMENTS IV PUBLICATIONS ARISING FROM THIS WORK VI CONFERENCE PROCEEDINGS VII LIST OF FIGURES XII LIST OF TABLES XV LIST OF APPENDICES XVI LIST OF ABBREVIATIONS XVII

CHAPTER 1 : INTRODUCTION 2

I. Heterotrophic marine bacteria 2 I.1 Role in the carbon fluxes and the food web 2 I.2 Abundance and diversity of bacteria 3

II. Oligotrophic bacteria 7 II.1 Concepts of oligotrophic and copiotrophic bacteria 7 II.2 Isolation of oligotrophic bacteria 11 II.3 Oligotrophic bacteria isolated by the dilution/extinction technique 15

III. Effects of solar radiation on marine microbial communities 18 III.1 Penetration of solar radiation in the atmosphere 18 III.2 Penetration of solar radiation into oceans 19 III.3 Direct effects of UVR on bacterial activity and diversity 21 III.4 Indirect effects of UVR on bacteria 23

IV. Affects of solar radiation on the DNA of microorganisms 24 IV.1 UV-induced DNA damage 24 IV.2 DNA repair mechanisms 34

V. Affects of solar radiation on gene expression in microorganisms 46 V.1 Transcriptome changes 46 V.2 Proteome changes 53 V.3 UV-induced protein damage 61

VI. Thesis outline 64 VI.1 Impact of GC content on DNA damage 64 VI.2 Factors modulating bacterial physiological states and responses to UV 65 VI.3 Role of spectral composition and UV intensity 67

VIII CHAPTER 2 : EFFECT OF THE GC CONTENT OF DNA ON THE DISTRIBUTION OF UVB-INDUCED BIPYRIMIDINE PHOTOPRODUCTS 70

I. Introduction 71

II. Materials and methods 74 II.1 Bacterial strains, media and culturing conditions 74 II.2 UVB irradiation of isolated DNA 74 II.3 UVB irradiation of bacterial cells and DNA extraction 75 II.4 HPLC-MS/MS analysis of bipyrimidine photoproducts 76 II.5 Dinucleotide frequencies 76

III. Results and discussion 78 III.1 Frequencies of bipyrimidine dinucleotides at different GC content differs from statistical distribution 78 III.2 GC content differently affects the yields of formation of the different DNA photoproducts 80 III.3 The proportion of cytosine-containing photoproducts correlates with that of the cytosine containing dinucleotides in irradiated isolated bacterial DNA 84 III.4 Favored formation of cytosine-containing photoproducts at high GC content in vivo 88 III.5 Increased UV-induced mutation rate in bacteria with high GC content? 91

IV. Conclusion 93

CHAPTER 3 : REMARKABLE RESISTANCE TO UVB OF THE MARINE BACTERIUM PHOTOBACTERIUM ANGUSTUM EXPLAINED BY AN UNEXPECTED ROLE OF PHOTOLYASE 95

I. Introduction 96

II. Materials and methods 100 II.1 Bacterial strains, media and culturing conditions 100 II.2 UVB exposure and repair conditions 100 II.3 Survival and DNA damage analysis 101

III. Results 103 III.1 Survival and DNA photoproducts following UVB radiation 103 III.2 Survival and DNA repair following photoreactivating light or darkness 105

IV. Discussion 109

V. Conclusion 114

CHAPTER 4 : TEMPERATURE DEPENDENT FORMATION OF BIPYRIMIDINE PHOTOPRODUCTS BY UVB RADIATION IN DIFFERENT CONDITIONS OF STARVATION AND LIGHT IRRADIATION IN THE MARINE BACTERIUM SPHINGOPYXIS ALASKENSIS 115

I. Introduction 116

IX II. Materials and methods 119 II.1 Bacterial strains, media and culturing conditions 119 II.2 UVB exposure and repair conditions 120 II.3 Survival and DNA damage analysis 120 II.4 Hierarchical clustering analysis 121

III. Results 122 III.1 Impact of temperature, starvation state and irradiation condition on viability 122 III.2 Quantitative changes in the yield of formation of the total UV-induced bipyrimidine photoproducts 124 III.3 Qualitative changes in the yield of formation of the different UV-induced bipyrimidine photoproducts 126

IV. Discussion 129

CHAPTER 5 : THE RESPONSE OF THE MARINE BACTERIUM SPHINGOPYXIS ALASKENSIS TO SOLAR RADIATION ASSESSED BY QUANTITATIVE PROTEOMICS 133

I. Introduction 134

II. Materials and Methods 137 II.1 Culture conditions and solar radiation treatment 137 II.2 Viability and protein synthesis activity 138 II.3 Protein Extraction, iTRAQ labeling and peptide purification 139 II.4 Mass spectrometry and protein identification 140 II.5 Protein quantification 142 II.6 Computational analysis 144

III. Results 146 III.1 Effect of light treatment on viability and rates of protein synthesis 146 III.2 Properties of proteins identified by LC-MS/MS 148 III.3 Differential protein abundance 149

IV. Discussion 161 IV.1 Overviewing the relationship between sunlight exposure and cell stress 161 IV.2 Describing the roles of specific proteins in the adaptive responses to sunlight. 163

V. Conclusion 171

CHAPTER 6 : GENERAL DISCUSSION 173

I. Summary of the findings 173

II. Comparative genome analysis: a tool to predict UV response 175

III. Distinct strategies of UV response 178 III.1 S. alaskensis 178 III.2 P. angustum 179

X IV. Impact of nutrient limitation on protein expression of S. alaskensis 181

V. Future prospects 182

VI. Concluding remarks 184

REFERENCES 185

APPENDICES 214

XI LIST OF FIGURES

Figure 1.1. Phylogeny of selected marine bacteria. Taxa with a gene size of 2 Mbp or less are boxed (Giovannoni and Stingl, 2007)...... 5

Figure 1.2. SeaWIFS map showing the average chl-a concentration over the word (http://oceancolor.gsfc.nasa.gov)...... 8

Figure 1.3. Steps in the dilution / extinction technique (Matallana-Surget et al., 2007)...... 14

Figure 1.4. Solar radiation fluxes outside the atmosphere and at the Earth’s surface. 19

Figure 1.5. Z10% of UVB radiation in different oceanic areas...... 20

Figure 1.6. Sunlight composition penetrating into the water column...... 21

Figure 1.7. Direct and indirect DNA damage induced by UV and visible light...... 24

Figure 1.8. Chemical structure of thymine dimeric photoproducts (Douki et al., 2000)...... 25

Figure 1.9. DNA base products of interaction with reactive oxygen species (Cooke et al., 2003a)...... 27

Figure 1.10. Main monomeric DNA base photoproducts (Cadet et al., 2005)...... 27

Figure 1.11. Biological responses to DNA damage...... 31

Figure 1.12. Concept of retromutagenesis (Doetsch, 2002)...... 34

Figure 1.13. Diagram showing base excision repair (BER) (Augusto-Pinto et al., 2003) ...... 37

Figure 1.14. Diagram showing nucleotide excision repair (NER) (Augusto-Pinto et al., 2003)...... 39

Figure 1.15. Schematic representation of methyl directed DNA mismatch repair in E. coli (Joseph et al., 2006)...... 40

Figure 1.16. The steps of photoreactivation (Friedberg, 2003)...... 41

Figure 1.17. Homologous replication and non-homologous end-joining pathways of double-strand break (Shuman and Glickman, 2007)...... 45

Figure 1.18. DNA microarrays technology...... 47

Figure 1.19. The steps in the iTRAQ method...... 58

XII Figure 1.20. Formation of carbonylated proteins and their fate in the cell...... 63

Figure 2.1. Chemical structure of thymine–cytosine dimeric photoproducts...... 72

Figure 2.2. Frequencies of the bipyrimidine nucleotides in the genomes of 99 bacteria...... 79

Figure 2.3. Dose-course formation of the overall photoproducts within isolated DNA exposed to UVB radiation...... 82

Figure 2.4. Correlation between the proportion of cytosine-containing bipyrimidine dinucleotides and the proportion of cytosine-containing photoproducts in isolated DNA exposed to UVB...... 86

Figure 2.5. Level of bipyrimidine photoproducts within the DNA of S. alaskensis and P. angustum exposed to 20 min of UVB radiation...... 89

Figure 2.6. Proportion of the photoproducts of the four bipyrimidine dinucleotides within DNA of two bacteria: S. alaskensis or P. angustum ...... 90

Figure 3.1. The emission spectra for the lamps used to induce DNA damage (UVB) and to photoreactivate damage after UVB exposure (PER)...... 101

Figure 3.2. Changes in viability and quantity of total DNA lesions during UVB irradiation in P. angustum and S. alaskensis...... 104

Figure 3.3. Induction of the different CPD and 6-4 photoproducts in P. angustum and S. alaskensis...... 105

Figure 3.4. Changes in viability and repair of DNA lesions during photoreactivation and liquid holding in P. angustum and S. alaskensis...... 107

Figure 3.5. Distribution of bypyrimidine photoproducts in P. angustum and S. alaskensis after 3 h of PRL or darkness ...... 108

Figure 4.1. Growth curves of S. alaskensis at 12°C and 24°C...... 119

Figure 4.2. Changes in viability under Solar Simulator or UVB lamps according different temperatures (12°C and 24°C) and growth phases (SP and LSP)...... 123

Figure 4.3. Distribution of the yield of total UV-induced photoproducts according different temperatures (12°C and 24°C), growth phases (SP and LSP) and radiation sources (SOL and UVB)...... 125

Figure 4.4. Distribution of the yield of bipyrimidine photoproducts induced under Solar Simulator or UVB lamps according different temperatures (12°C and 24°C) and growth phases (SP and LSP) conditions...... 127

Figure 4.5. Dendrogram from hierarchical clustering analysis, based on the yield of individual bipyrimidine photoproducts induced...... 128

XIII Figure 5.1. Relative changes in viability and protein synthesis rates in S. alaskensis exposed to four different light treatments: Full sun, PAR+UVA, PAR and dark, for ML and SP cultures...... 147

Figure 5.2. Characteristics of all proteins identified (811) in all combined iTRAQ runs ...... 148

Figure 5.3. Total number of proteins with increased or decreased abundance caused by growth phase and irradiation treatment. The three different irradiation treatments (FS, PAR+UVA, and PAR compared to the dark control) are shown for ML8h, SP80min and SP8h...... 151

Figure 5.4. Venn diagrams showing the relationship of proteins from specific treatments...... 152

Figure 5.5. Quantitative proteomics grid of proteins with differential abundance. .. 156

Figure 5.6. HCA of quantitative proteomics data...... 160

Figure 5.7. Cartoon depicting important proteins and cellular processes that are important in S. alaskensis for its adaptive response to sunlight...... 165

Figure 6.1. Distribution of the total number (4903) of fully sequenced genomes accessible to date in the Genbank database...... 175

Figure 6.2. Extract of an “Abundance Profile Map” comparing the genomes of “Ca. Pelagibacter ubique”, Deinococcus radiodurans, Sphingopyxis alaskensis RB2256 and Photobacterium angustum S14...... 177

Figure 6.3. The dual role of photolyase in P. angustum S14...... 180

XIV LIST OF TABLES

Table 1.1. Main characteristics of oligotrophic and copiotrophic bacteria. (Matallana- Surget et al., 2007)...... 10

Table 1.2. A summary of the DNA lesions and the most relevant DNA repair mechanism for removing the lesions...... 35

Table 1.3. Amino acids residues of proteins that are oxidized and products formed. 61

Table 2.1. Effect of the variation in GC content on the yield of formation of bipyrimidine photoproducts in isolated DNA exposed to UVB radiation...... 81

Table 2.2. Ratio between the relative frequency of photoproducts and the frequency of bipyrimidine sites in genomes of different GC content...... 85

Table 2.3. Ratio between the yields of CPD and 6-4PP at TT and TC sites in DNA exhibiting various GC content...... 88

Table 3.1. Estimations of the rates of lesions repaired per 106 bases per hour in P. angustum during UVB radiation between 2 and 8.64 kJ m-2 (“UVB reapir”) and during the two first hours repair in the dark condition (“Dark repair”)...... 112

Table 5.1. Distribution of S. alaskensis proteins into COG categories ...... 150

XV LIST OF APPENDICES

Appendix A. Determination of the dinucleotide frequency for M. luteus (72% GC) in the high GC content region...... 215

Appendix B. Frequency of the four bipyrimidine dinucleotides in a series of 99 bacteria with increasing GC content...... 216

Appendix C. Effects of UVB radiation on the survival of V. natriegens, P. angustum and S. alaskensis...... 218

Appendix D. Evidence rating system for manual annotation of proteins...... 219

Appendix E. Cellular location predicted using PSORTb v.2.0 software for the total proteins observed in all iTRAQ runs...... 220

Appendix F. Unique proteins of mid-logarithmic phase identified by Scaffold...... 221

Appendix G. Unique proteins of stationary phase identified by Scaffold ...... 222

Appendix H. Representation of predicted protein interactions using BioLayout. .... 223

XVI LIST OF ABBREVIATIONS

2D-DIGE Two Dimensional Difference Gel Electrophoresis 2DE Two Dimensional gel Electrophoresis 6-4PP Pyrimidine (6-4) pyrimidone photoproduct. 8-HDF 8-Hydroxy-5-Deazaflavin 8-oxodG 8-oxo-2'deoxyguanosine A Adenine ACN Acetonitrile AP site Apurinic / Apyrimidic site ASW Artificial Sea Water ATP Adenosine Triphosphate BER Base Excision Repair BLAST Basic Local Alignment Search Tool C, Cyt Cytosine CE-SSCP Capillary Electrophoresis-Single strand conformation polymorphism CFC Chlorofluorocarbon CFU Colony-Forming Units chl-a chlorophyll-a Ci Curie COG Cluster of Orthologous Group CPD Cyclobutane pyrimidine dimer Cy Cyanine Da Dalton(s) DGGE Denaturation Gradient Gel Electrophoresis DNA Deoxyribonucleic Acid DNase Deoxyribonuclease DOM Dissolved Organic Matter dRpase Deoxyribophosphodiesterase DSB Double Strand Breaks EDTA Ethylene Diamine Tetra Acetic Acid EFM Epifluorescence Microscopy ELISA Enzyme-Linked Immunosorbent Assay EMBOSS European Molecular Biology Open Software Suite ER Evidence Rating ESC Ecological Species Concept FAD Flavine Adenine Dinucleotide FCM Flow Cytometry FISH Fluorescent In Situ Hybridization FS Full Sun (UVB+UVA+PAR) G Guanine Glufib Glufibrinopeptide GRAVY Grand Average of Hydropathicity HCA Hierarchical Clustering Analysis HFBA Heptafluorobutyric Acid HPLC High Performance Liquid Chromatography HR Homologous Recombination ICAT Isotope-Coded Affinity Tags

XVII IDA Iodoacetamide IEF Isoelectric Focusing IR Infrared (>700 nm) OR Ionizing Radiation iTRAQ isobaric Tags for Relative and Absolute Quantification J Joules LC Liquid Chromatography LSP Late Stationary Phase ML Mid-Logarithmic MMR Mismatch Repair MS Mass Spectrometry MTHF Methenyltetrahydrofolylpolyglutamate MW Molecular Weight NCBI National Center for Biotechnology Information NER Nucleotide Excision Repair NHEJ Non-Homologous End Joining OD Optical Density PAR Photosynthetic Active Radiation, visible light (400-700 nm) bp base pair(s) PCR Polymerase Chain Reaction PER Photoenzymatic Repair pI Isoelectric Point POM Particular Organic Matter ppm parts per million PRL Photoreactivating Light (300-500 nm) RNA Ribonucleic Acid ROS Reactive Oxygen Species RP Reverse Phase SCX Strong Cation Exchange SDS-PAGE Sodium Dodecyl Sulphate-Polyacrylamide Gel Electrophoresis SP Stationary Phase sp. species SSB Single Strand Breaks STRING Search Tool for the Retrieval of Interacting Genes/Proteins T, Thy Thymine TCA Trichloroacetic Acid TCA cycle Tricarboxylic Acid cycle TRCF Transcription Repair Coupling Factor U Uracil UVA Ultraviolet A (320- 400 nm) UVB Ultraviolet B (280-320 nm) UVC Ultraviolet C (100-280 nm) UVR Ultraviolet Radiation (280-400 nm) VNSS Vaatanen Nine Salt Solution

XVIII CHAPTER 1

Introduction Chapter 1

CHAPTER 1 Introduction

I. Heterotrophic marine bacteria

I.1 Role in the carbon fluxes and the food web

Oceans contribute to about half of the global annual primary production, a process by which phytoplankton or photosynthetic bacterioplankton convert carbon dioxide

(CO2) into particular organic carbon (POM). A large fraction of organic matter produced by phytoplankton is released as dissolved organic matter, (DOM) in the food web and is then taken up almost exclusively by heterotrophic bacteria and archaea (Azam et al., 1983). According to the ‘microbial loop hypothesis’ described by Azam et al., in 1983, bacterioplankton are no longer regarded solely as final decomposers of organic matter but are placed in the centre of the food web, where they take up DOM, so remineralizing the carbon, and significantly contribute to the transfer of biomass to higher organisms, like heterotrophic flagellates and ciliates

(Azam et al., 1983).

The planktonic food web is composed of producers of organic matter, autotrophs, and consumers, heterotrophs. Bacteria are fundamental components of the organic carbon cycle in aquatic systems, responsible for up to 50% of primary production in the euphotic zone per day (Ducklow, 2000). The balance between these two trophic modes determines the net metabolism of the community, i.e., whether autotrophic production exceeds respiration, or whether heterotrophic processes prevail over autotrophic ones. Oligotrophic systems (nutrient-limited environments) are generally considered to be slightly net heterotrophic over an annual cycle (Duarte and Agustí,

2 Chapter 1

1998) because the respiration of bacteria can be higher than primary production at certain times.

I.2 Abundance and diversity of bacteria

The oceans are estimated to contain more than 1029 bacteria (Whitman et al., 1998).

Despite the low levels of nutrients, the concentration of bacteria in the first 200 meters of the ocean (i.e., the area photique) is about 105 - 106 cells/ml. This is the region of most biological activity. The bacteria can be counted directly in a sample of sea water using traditional techniques of epifluorescence microscopy (EFM) or flow cytometry (FCM). Those methods are based on the fluorescent labeling of DNA using a fluorochrome (e.g., DAPI, SYBR-Green) depending on the source of excitation light used (Porter and Feig, 1980; Gasol et al., 1999).

The cell volume of marine bacteria observed by EFM or FCM is very small ranging from 0.02 to 0.12 )m3 (Bjørnsen, 1986; Lee et al., 1987; Børsheim et al. 1990;

Nagata and Watanabe, 1990), which is about one order of magnitude smaller than commonly studied bacteria like Escherichia coli. They have a low apparent DNA content of between 1 and 2.5 fg/cell (Button et al., 2001; Schut et al., 1993). Despite their small biovolume, these bacteria represent a significant (and sometimes dominant) biomass and account for up to 90% of cellular DNA in many ocean systems (Paul et al., 1984; Fuhrman et al., 1989).

The study of bacterial diversity was revolutionized in the 90s by the introduction of molecular biology techniques (Giovannoni et al., 1990). The small subunit of the 16S rRNA gene that is highly conserved in bacteria has become a universal marker for

3 Chapter 1

phylogenetic analysis and is the main criterion by which microbial groups are identified and named (Amann and Ludwig, 2000; Giovannoni and Rappé, 2000). The gene that codes for 16S rRNA is ~1550 bp long, and contains both highly conserved and variable regions. The conserved regions of the gene can be used to design universal primers for PCR amplification and the variable regions can be used to distinguish one species from another. The bacterial DNA is extracted from a given complex sample and the 16S rRNA gene amplified by PCR. Today there are many molecular tools based on amplification of the 16S rRNA. They include 16S rRNA clone libraries, Denaturation Gradient Gel Electrophoresis (DGGE), Capillary electrophoresis-Single strand conformation polymorphism (CE-SSCP) and

Fluorescence In Situ Hybridization (FISH) (Amann and Ludwig, 2000).

- For a clone library, 16S rRNA amplicons are cloned and expressed in a host such as

Escherichia coli, followed by plasmid preparation of individual clones, reamplification of the 16S rRNA gene and sequencing (Amann et al., 1995).

- Denaturation Gradient Gel Electrophoresis (DGGE) and Capillary electrophoresis-

Single strand conformation polymorphism (CE-SSCP) (Muyzer, 1993; Larsen et al.,

2003) are techniques of structural analysis in which single-stranded 16S rRNA amplicons migrate differently through a gel or capillary according subtle variations in their sequence.

- Fluorescence In Situ Hybridization (FISH) was first introduced into bacteriology by

Giovannoni et al. (1988). This technique uses fluorescently labelled oligonucleotides as probes to hybridize specifically targeted regions on bacterial 16S rRNA, allowing direct counting of the specific, living cells in a given environment (Moter and Göbel,

2000).

4 Chapter 1

Those molecular methods have helped to reveal the incredible diversity of bacteria in the ocean. They are now commonly used in studies of microbial ecology to analyze the structures of communities and identify species (Muyzer, 1998). The phylogenetic tree in Fig. 1.1, based on 16S rRNA sequence information, shows the major bacterioplankton groups and provides the names of current cultivated strains and the sizes of their complete genomes as well as abundant uncultured groups (marked by an asterisk).

Figure 1.1. Phylogeny of selected marine bacteria. Taxa with a gene size of 2 Mbp or less are boxed (Giovannoni and Stingl, 2007).

5 Chapter 1

We do not yet know the total number of existing bacterial species, mainly because of the difficulty of choosing objective criteria to differentiate between species.

Hybridization DNA/DNA > 70% (Wayne et al., 1987) and the similarity of genes from the 16S rRNA > 97% (Hagström et al., 2000) have been suggested as indicating that two marine bacteria belong to the same species.

All the sequences of the 16S ribosomal DNA deposited in databases since 1990

(Genbank) have been analyzed to assess the number of species of marine bacteria on the surface of the ocean (Hagström et al., 2002). Based on 97% sequence similarity of the 16S rRNA gene, 609 of a total of 1 117 unique ribotypes came from uncultured environmental clones and 508 from cultured bacteria (Hagström et al.,

2002). The diversity distribution into the different clades obviously varies according to their culturability. Indeed, the major difference between the two distributions is the presence of the bacterial groups Planctomycetales and Verrumicrobiales in the uncultured bacterioplankton.

The recent large metagenomics or the culture-independent genomic analysis of an assemblage of microorganisms has the potential to answer fundamental questions in microbial ecology (Riesenfeld et al., 2004). One approach that has provided very useful information about the genome is the massive marine prokaryotic metagenome- sequencing project, which is providing a tremendous database for discovering metabolic capabilities and new ways to conceptualize and study prokaryotic biodiversity (Falkowski and de Vargas, 2004). However, organisms that are identical or cluster tightly by molecular criteria do not necessarily share all or, in some cases, essential physiological features. Thus, the molecular definition of species is not

6 Chapter 1

always adequate for assessing the functional diversity of prokaryotic communities.

The discovery of ecologically important differences in temperature optima attributed to hot spring microbes with less than 1% 16S rRNA sequence divergence led Ward in 1998 to advocate the more ‘‘natural’’ or ecological species concept (ESC) of

Simpson (Simpson, 1961).

In order to improve our understanding of marine microbial communities, we need to combine the molecular information with the physiological parameters of abundant, ubiquitous and cultivatable bacteria.

II. Oligotrophic bacteria

II.1 Concepts of oligotrophic and copiotrophic bacteria

The primary production of oceans is assured by phytoplankton living in the water column. The overall phytoplankton biomass in surface waters, often quantified by the concentration of chlorophyll a (chl-a), varies widely from 0.1 to 100 µg l-1 (Fig. 1.2).

Eutrophic waters such as estuaries or coastal waters are nutrient-rich environments with high chl-a contents and high primary production. In contrast, open oceans

(between 50 and 60%) have very low chl-a concentrations (<0.5 µg l-1), low primary production (<1 mg carbon per liter per day) (Schut et al., 1997) and the nutrient content (e.g., nitrate, phophate) is often at the limit of detection for conventional analytical techniques (Cole et al., 1998).

7 Chapter 1

Figure 1.2. SeaWIFS map showing the average chl-a concentration over the word (http://oceancolor.gsfc.nasa.gov).

The primary production of these so-called oligotrophic ocean areas was found to be mainly the result of active phytoplankton cells smaller than 2 µm across (Veldhuis et al., 2005). Another characteristic of these environments is their relatively high primary productivity (production/biomass), suggesting that the primary production and remineralization of organic matter are coupled by marine heterotrophic bacteria, in the absence of any significant external mineral elements. Moreover, it was recently suggested that global warming could increase the representativeness of the oligotrophic waters by increasing ocean stratification leading to reduced nutrient supply from deep waters (Cox et al., 2000; Bopp et al., 2001).

Despite the ecological importance of marine oligotrophic ecosystems, we still know very little about the diversity and adaptations of organisms living in these

8 Chapter 1

environments. This is particularly true for heterotrophic bacteria; their study is complicated by problems of observation, characterization and isolation. EFM and

FCM counting suggest that only 0.01 to 0.1% of all microbial cells from marine environments form colonies on the standard agar plates typically used for isolating bacteria (Amann et al., 1995; Porter and Feig, 1980; Kogure et al., 1979; Ferguson et al., 1984). Clearly, we need to isolate the uncultivated marine species using new techniques and characterize them by genomic studies. The abundance of these microorganisms suggests that they are important, and knowledge of their physiology may reveal their role in biogeochemical cycles.

Oligotrophs are defined as heterotrophic bacteria that can grow in minimal organic carbon substrate concentration (1 to 15 mg C/l), even though they may be able to grow in richer media (Morita, 1997). Many oligotrophic marine bacteria appear to be able to grow only at low nutrient concentrations (Morita, 1997). However, this second definition is arbitrary and it is difficult to test all possible growth conditions available for the organisms (Morita, 1997). The survival of bacteria in nutrient-poor environments is subject to certain adjustments (Van Guemerden and Kuenen, 1984).

The terms oligotrophic and copiotrophic bacteria were proposed by Poindexter et al.

(1981). Heterotrophic bacteria can be separated in two major groups according to their ability to take up organic matter. Species with low nutrient affinity and rapid growth (r-strategists) that depend on high carbon concentrations are called copiotrophic bacteria, while those with high nutrient affinity and slow growth (K- strategists) that are able to cope with low concentrations of carbon, are called oligotrophic bacteria (Andrews and Harris, 1986).

9 Chapter 1

Tab. 1.1 summarizes the main theoretical features of copiotrophic and oligotrophic bacteria, although it is sometimes difficult to categorize the two groups. Placement in a category may depend on the nature of the substrate considered. A switch from the oligotrophic state to copiotrophic over time is possible (adaptation to the environment). Because oceans contain microzones that are rich in organic matter

(e.g., around detritus particles), copiotrophic and oligotrophic bacteria can coexist

(Simu et al., 2005).

Table 1.1. Main characteristics of oligotrophic and copiotrophic bacteria. (Matallana-Surget et al., 2007).

Oligotrophic bacteria Copiotrophic bacteria Environment Low absolute concentration High level of substrates of nutrients and nutrients

Strategy K strategy r strategy

Max growth rate Low (0.1 h-1) High (> 1 h-1)

Cell volume 0.02 to 0.12 µm3 0.34 to 6.0 µm3 Ultramicrobacteria

Uptake system High affinity Poor affinity Poor specificity High specificity Simultaneous uptake of multiple substrates Ribosomal operon Simple copy Multicopy (Vibrio spp.:8 to 10)

Genome Small size Larger size (“Ca. Pelagibacter ubique”: 1.32 Mpb (V. angustum: 5.1 Mpb) C. oligotrophus: 2.9 Mpb)

10 Chapter 1

II.2 Isolation of oligotrophic bacteria

Initial attempts to isolate marine bacteria on nutrient rich agar plates revealed a discrepancy of up to three orders of magnitude between plate counts and the total number of cells that could be counted by microscopic examination (Jannasch and

Jones, 1959; Ferguson et al. 1984). The term “the great plate count anomaly” was suggested by Staley and Konopka in 1985 to describe this phenomenon. Different hypothesis have been proposed to explain this anomaly, including inactivation due to the close proximity of other cells, the antibiosis phenomenon that occurs in colonies on agar plates (Long and Azam, 2001), and the deleterious effects of lytic cycle induction in bacteria infected by temperate phages (Middelboe et al., 2001). Species which would otherwise be “culturable” may fail to grow because of their growth state in nature, such as dormancy. Bacteria suffering from a deficiency of pivotal nutrients would mostly be present in marine environments in an inactive form.

However, this may be challenged by two other observations. First, the identification of bacterial species in the marine environment using the techniques of molecular biology, indicates that uncultured strains are poorly represented in populations described to date (Eilers et al., 2000; Giovannoni and Rappé, 2000). This does not affect the importance of seeking new bacterial species using conventional agar plates, but raises the question of their ecological representativeness. Second, cell activity has been detected by microautoradiography, which suggests that a significant fraction of marine bacteria is active despite their small size (Joux and Lebaron,

2000).

Finally, “the great plate anomaly” could also be due to the substrate being suitable but present at too high a concentration (Jannash and Jones, 1959). The successful

11 Chapter 1

culture of many microorganisms seems to need oligotrophic conditions. This argument provides the basis for studying obligate oligotrophy. As demonstrated by

Button et al., (1993), the growth of marine bacteria in a liquid medium consisting solely of filtered autoclaved seawater is completely blocked by adding 5 mg C/l. The techniques of classical bacterial culture appear to select mostly copiotophic bacteria, thus introducing a bias in the representativeness of the isolated species. The study of the physiology and function of the overall marine bacterial communities therefore requires the development of alternative techniques of isolation. Efforts to optimize the composition of culture media have not provided striking results (Buck, 1974;

Goltekar et al., 2006).

New techniques have been developed for cultivating previously uncultured bacteria.

The extinction dilution approach has been the most successful isolation technique to date. Button and co-workers (1993) were the first to to describe the original procedure, in which seawater samples were diluted with filtered, autoclaved natural seawater without any additional nutrients until only a few organisms remained in each dilution tube (Button et al., 1993). The growth rate of initially unculturable cells under these circumstances was a doubling in from one day to one week and the stationary phase was reached at a concentration of 105-106 cells ml-1 (Button et al.,

1993; Schut et al., 1993). Long term incubation of these cultures (6-12 months) in the dark at 5°C initiated an unknown mechanism that enabled the cells to grow on a nutrient-rich medium, a transition from an obligate to a facultative oligotrophic state

(Schut et al., 1993). This could facilitate physiological studies of facultative oligotrophic bacteria in the laboratory. This extinction dilution method has recently been optimized for high-throughput culturing (HTC). The volumes of culture is

12 Chapter 1

reduced by using a microplate, which facilitates the observation of growth by epifluorescence microscopy (Connon and Giovanonni, 2002; Simu and Hagström,

2004) (Fig. 1.3). The HTC technique has been a great success in cultivating new strains. Two marine oligotrophic bacteria are used today to study designs in several laboratories. The application of the dilution / extinction technique has led to the cultivation of Sphingopyxis alaskensis and the first member of the SAR11 clade,

“Ca. Pelagibacter ubique” (Schut et al., 1993; Connon and Giovannoni, 2002). The two bacteria differ mainly in their ability to adapt to nutrient-rich environments. “Ca.

Pelagibacter ubique” is a strict oligotrophic bacterium, which cannot be cultivated at the organic carbon concentrations found in traditional growth media (Rappé et al.,

2002), while Sphingopyxis alaskensis is a facultative oligotrophic bacterium that can grow in standard and nutrient-rich microbiological media (Schut et al., 1993, 1995).

13 Chapter 1

Sample of sea water (inoculum)

Cells counts (EFM and FCM)

Dilution of sterile water Cells number / well 2048 1024 512 1/2 dilution 256 with sterile water 128 (200 µL/well) 64 32 16 8 4 2 1

Incubation (1-4 weeks) Cell growth control assessed by EFM or FCM

-Ln (1 -[Z/n]) % viable cells = X Z = number of positive wells n = number of replicates (wells) X = number of inoculated cells at the retained dilution

Positive cultures at the highest dilution

Isolation on Transfer in new agar plates medium of sea water

Molecular identification of purified cultures (sequencing of 16S rRNA gene)

Figure 1.3. Steps in the dilution / extinction technique (Matallana-Surget et al., 2007).

14 Chapter 1

II.3 Oligotrophic bacteria isolated by the dilution/extinction technique

II.3.1 “Candidatus Pelagibacter ubique”

“Ca. Pelagibacter ubique” is a strictly oligotrophic bacterium that was isolated by the dilution / extinction technique from seawater samples taken off the Oregon coast

(Rappé et al., 2002). It is a Gram negative bacterium, Phylum Proteobacteria, class alpha Proteobacteria and order Rickettsiales (Rappé et al., 2002). According to the sequencing of its 16S rRNA gene, “Ca. Pelagibacter ubique” is a member of the

SAR11 clade and could be one of the most abundant bacteria in the ocean surface waters (Rappé et al., 2002). Roughly a quarter of all rRNA gene sequences retrieved from marine environments belong to the SAR11 clade (Rappé et al., 2002) and

SAR11 bacteria often account for a large fraction (35%) of all prokaryotes in surface waters of the Sargasso Sea, and might exceed 50% of the cells found in some temperate ocean gyres (Morris et al., 2002). By extrapolation, there would be 2.4 x

1028 cells of the SAR11 clade in the oceans, half of which would be located in the euphotic zone (Morris et al., 2002). Although the important biogeochemical role of the SAR11 clade is not yet established, the study of this microbial group is undoubtedly important because of its predominance in the marine bacterial community.

“Ca. Pelagibacter ubique” is an ultramicrobacterium with a diameter of 0.12 to 0.2

µm and a cell volume of 0.01 )m3. This is the smallest bacterium isolated to date. Its very small genome (1.32 Mbp) has been completely sequenced and codes for 1393 proteins. The DNA occupies up to 30% of the cell volume (Rappé et al., 2002).

Unlike other organizations with a very small genome (Archaea bacteria and parasites), “Ca. Pelagibacter ubique” has complete biosynthetic pathways for each of

15 Chapter 1

the 20 amino acids and cofactors (Giovannoni et al., 2005a). Its genome contains no pseudogenes, introns, transposons, or extrachromosomal elements (Giovannoni et al.,

2005a). It has some paralogues and the smallest intergene regions observed to date

(Giovannoni et al., 2005a). “Ca. Pelagibacter ubique” can grow in a liquid medium consisting of sea water supplemented with inorganic nitrogen and phosphorus.

Adding a source of organic carbon at a very low concentration (peptone up to

0.001%) causes the leads culture to fail. It is thus a strict oligotrophic bacterium

(Rappé et al., 2002). A maximum concentration of 106 cells ml-1 can be reached after

15 days culture under appropriate growing conditions. This biomass, although small, may be sufficient for certain kinds of study. Proteorhodopsin (light-dependent proton pump) has been characterized in this strain by mass spectrometry, and verification that this protein is functional was produced after cloning into an expression vector, E. coli, by measuring the absorption at 530 nm (Giovannoni et al., 2005b). This protein occupies 20% of the surface of the plasma membrane internally and its synthesis is probably important in the ecology of the oceans by supplying energy for microbial metabolism. However the exact function of proteorhodopsin in “Ca. Pelagibacter ubique” is not yet clear, because this strain grows at the same rate under both light and darkness conditions (Giovannoni et al., 2005b). The function of proteorhodopsin in the field is probably more subtle.

II.3.2 Sphingopyxis alaskensis RB2256

Sphingopyxis alaskensis RB2256 is a facultative oligotrophic bacterium that was first isolated by the dilution / extinction technique from a fjord in Alaska (Schut et al.,

1993). S. alaskensis AFO1 strain was then isolated from the North Sea and the

Pacific Northwest (Eguchi et al., 2001). Prolonged (6-12 months) incubation of

16 Chapter 1

diluted cultures of S. alaskensis RB2256 in the dark at 5°C enabled the cells to adapt to a richer nutrient medium (Schut et al., 1993). This mechanism underlying transition from an obligate to a facultative oligotrophic state remains unclear but it may be due to some mutations (Schut et al. 1993). S. alaskensis is a gram negative bacterium, phylum Proteobacteria, class alpha Proteobacteria, order Rhodospirillales

(Eguchi et al., 2001). It is an ultramicrobacterium, with a cell volume of 0.05 )m3, diameter 0.05 to 0.09 )m3, according to the growth state (Schut et al., 1997). S. alaskensis remains ultramicro in size (<0.1 µm3), irrespective of whether it is growing or starved, so that it can avoid predation, and its high surface-to-volume ratio provides enhanced nutrient uptake. This last phenomenon is coupled with the ability to use low concentrations of nutrients due to its high affinity, broad specificity uptake systems (e.g., highest reported rates of alanine transport for any bacterium) and its ability to take up mixed substrates simultaneously. S. alaskensis as a model of facultative oligotrophic bacterium has a low growth rate (0.13 to 0.16 h-1) in organic carbon concentrations from 0.8 to 800 mg carbon / l (Schut et al., 1993; Eguchi et al.,

1996).

The genome of this bacterium has been completely sequenced. It is 3.2 Mbp with

3196 candidate protein-encoding genes (Cavicchioli et al., 2003). The physiology of

S. alaskensis differs in many ways from that of traditionally studied bacteria (E. coli,

Vibrio spp.). The genome of S. alaskensis has only 1 rRNA operon, while Vibrio spp has 8 to 11 in. The single rRNA operon can be considered as an adaptation to life in oligotrophic environment, where nutrient concentrations vary only slightly and rapid growth in response to a massive nutrient supply, as in coastal waters, is unnecessary

(Fegatella et al., 1998; Klappenbach et al., 2000). S. alaskensis also has a regrowth

17 Chapter 1

capacity without a lag time, even after seven days in the stationary phase (Fegatella et al., 1998). Bacteria in the stationary phase have many more ribosomes than are required for protein synthesis - an advantage when they encounter new carbon sources (Fegatella and Cavicchioli, 2000).

S. alaskensis is very resistant to stresses, such as high temperatures (up to 56°C), hydrogen peroxide (25 mM), ethanol (20%), and ultraviolet B light (Eguchi et al.,

1996; Joux et al., 1999). It is also resistant to these stresses in both the exponential and stationary phases, unlike other bacteria. This has been called "cross protection"

(increased cell resistance towards stresses at the beginning of stationary phase)

(Eguchi et al., 1996). This bacterium could have novel resistance mechanisms, not yet described in undifferentiated bacteria (Eguchi et al., 1996).

III. Effects of solar radiation on marine microbial communities

III.1 Penetration of solar radiation in the atmosphere

Sunlight produces a broad spectrum of radiation. The UVC (100-280 nm) fraction is efficiently absorbed by the atmosphere and does not reach the Earth’s surface. UVB

(280-320 nm) is strongly absorded by the ozone of the stratospheric layer, while

UVA (320-400 nm) and visible light, PAR (400-700 nm), reach the Earth’s surface with little attenuation. Finally the infrared, IR (>700 nm), reach the Earth’s surface almost without any attenuation (Fig. 1.4).

18 Chapter 1

Figure 1.4. Solar radiation fluxes outside the atmosphere and at the Earth’s surface (Thekaekara, 1976).

The introduction of chlorofluorocarbons (CFCs) due to human activity over the past

50 years, along with other chlorine- and bromine-containing compounds, has resulted in a steady increase in the concentration of CFCs in the athmosphere, with a corresponding decrease in stratospheric ozone. Molina and Rowland first proposed their role in the destruction of atmospheric ozone in 1974. The ozone hole over

Antarctica was first reported in 1985 (Farman et al., 1985). The damaging UVB radiation that reaches the Earth’s surface is one of the physical variables likely to be modified in the future due to interactions between global warming and ozone depletion.

III.2 Penetration of solar radiation into oceans

Solar ultraviolet radiation (UVR, 280-400 nm) has been shown to reach significant depths in many marine ecosystems, influencing a large part of the euphotic layer, where phytoplankton productivity takes place (Tedetti and Sempéré, 2006). UV

19 Chapter 1

radiation may penetrate deep into the clear oligotrophic waters, so that 10% of the irradiance at 340 nm (Z10%) reaches a depth of about 30 to 40 m, while radiation at

380 nm can reached a depth of 60 to 70 m (Obernosterer et al., 2001). Thus about half of the photic layer is affected by ultraviolet radiation in the most representative oligotrophic oceans (Fig. 1.5) (Obernosterer et al., 2001).

Figure 1.5. Z10% of UVB radiation in different oceanic areas.

Many factors influence the penetration of radiation into natural waters, including dissolved organic compounds whose concentration and chemical composition are likely to be influenced by future changes in climate and UV radiation.

The composition and intensity of sunlight varies according to the depth in the ocean.

Indeed, as illustrated in Fig. 1.6, organisms could be exposed at the surface to the full spectrum (UVB+UVA+PAR), UVA+PAR or to PAR only at the deepest meters of the euphotic layer.

20 Chapter 1

280 nm 320 nm 400 nm 700 nm UV-B UV-A PAR IR

UVB, UVA, PAR Euphotic layer Zone receiving sufficient visbile light UVA, PAR for photosynthesis (up to 1% of surface light intensity) PAR

DARK

Figure 1.6. Sunlight composition penetrating into the water column.

Both UVB and UVA can have important detrimental effects on bacterial activity,

phytoplankton photosynthesis and photochemical transformation of dissolved

organic matter.

III.3 Direct effects of UVR on bacterial activity and diversity

Bacteria are among the marine organisms most severely affected by UVR. It has

been hypothesized that this may be due to their high surface-to-volume ratio (small

cell size) (Karentz et al., 1991) and/or the absence of protective pigmentation

(Garcia-Pichel, 1994). Solar radiation may have detrimental effects on bacteria in

surface waters by reducing their metabolic and exoenzymatic activity, decreasing

their viability and inducing DNA damage (Herndl et al., 1993; Müller-Niklas et al.,

1995; Jeffrey et al., 2000). Sieracki and Sieburth (1986) first reported that significant

changes in solar UV radiation on aquatic ecosytsems may result in decreased

bacterial biomass. In both laboratory and in situ experiments, Kaiser et al. (1997)

21 Chapter 1

showed that DNA and protein synthesis may be inhibited by as much as 40% of the pre-exposure level. Some studies have demonstrated that bacterial production is more damaged by UVR than is phytoplankton production (Jeffrey et al., 1996; Bertoni and

Callieri, 1999; Sommaruga et al., 1999; Plante and Arts, 2000). However other studies have obtained the opposite result (Wickham and Carstens, 1998; Ferreyra et al., 2006). These inconsistent results should not be surprising considering the complex ways in which UVR impacts on whole microbial communities, and that these UVR-manipulation studies were conducted in different places and times of the year, with variable UV doses. On the other hand, Joux and colleagues in 1999 reported significant variations in the responses to UVB of several marine bacteria, suggesting the repercussion of solar radiation on the ecosystem can lead to altered community structure.

UVR affects the growth and productivity by a number of mechanisms involving several molecular targets within the exposed cells. UVB and UVA (320-400 nm) cause distinct but overlapping damage. Exposure to UVB mainly has a direct effect on DNA by inducing dimerization of adjacent DNA bases, blocking DNA transcription and replication, while UVA and PAR can lead to detrimental oxidative stress by generating reactive oxygen species (ROS) that interact with DNA, proteins and lipids, or a beneficial action by activating a specific light-dependent repair pathway (Jeffrey et al., 2000; Häder et al., 2007). The net biological effect of UVR depends upon the balance between the rate of UVR-induced damage and the rate at which that damage is repaired. Therefore, the overall effect of solar UV on bacterioplankton depends on their position in the water column, as composition and intensity of sunlight changes according to the oceanic depth, as well as the time of

22 Chapter 1

exposure, as the organisms are passively moved in the mixing layer (Häder et al.,

2007).

III.4 Indirect effects of UVR on bacteria

The indirect effects of UVR on bacteria can be due to alterations of the chemical environment through photolysis of organic compounds (Anesio et al., 2005), production of toxic substances (Xenopoulos and Bird, 1997), changes in bacteriophage infection rates (Wilhelm et al., 2003), changes in grazer activity

(Vincent and Roy, 1993; Belzile et al., 2006) and possibly changes in competitive interactions (Plante and Arts, 2000).

The large reservoir of dissolved organic matter (DOM) includes both small, simple molecules (amino acids and carbohydrates) that have a fast turn-over of from a few minutes to a day (Keil and Kirchman, 1991) and macromolecular organic complexes

(humic substances) whose turn-over is very slow, from a century to millenia

(Williams and Druffel, 1987; Bauer et al., 1992). It has been shown that solar UVR can induce diverse phototransformation reactions depending on the nature of the organic matter considered (refractory or labile), resulting in either a decrease or an increase in carbon uptake, thus influencing the overall bacterial activity (Ogawa et al., 2001). UVR can transform labile, simple organic matter into refractory substances, but the process is not well understood (Ogawa et al., 2001). The opposite transformation is also conceivable, as humic substances strongly absorb UVR at the ocean surface, and solar radiation has been shown to photolytically cleave refractory compounds into labile ones that are easily taking up by heterotrophic bacteria

(Herndl, 1997).

23 Chapter 1

IV. Affects of solar radiation on the DNA of microorganisms

IV.1 UV-induced DNA damage

Nucleic acids are a key component of UV-induced cellular damage, with a maximum absorption in the 260 nm region and a rapid decline towards longer wavelengths.

DNA can be damaged by sunlight by direct absorption of UVB and/or by indirect damage due to oxidative stress that modifies the activity of molecules (Fig. 1.7).

Figure 1.7. Direct and indirect DNA damage induced by UV and visible light.

IV.1.1 Direct damage

The main direct effect of UVB on DNA is to cause dimerization of pyrimidine bases, leading to the formation of two major photoproducts: primarily cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone photoproducts (6-4PPs)

(Fig. 1.8). High UVB fluxes can convert 6-4PPs into Dewar valence isomers upon intramolecular cyclization of the pyrimidone ring (Matsunaga et al., 1993; Mitchell

24 Chapter 1

and Rosenstein, 1987). Consequently, UVB can directly induce a total of 12 photoproducts but they are not all produced with the same frequency.

Figure 1.8. Chemical structure of thymine dimeric photoproducts (Douki et al., 2000).

It is assumed that approximately 75% of the UVB-induced DNA damage consists of

CPDs and 25% of 6-4PPs, less than 1% is purine dimers and monobasic damage

(Friedberg et al, 1995; Pfeifer, 1997). Both classes of lesions distort the DNA helix.

CPDs induce a 7–9°bend in the B- form DNA and 6–4PPs induce a 44° bend (Kim et al., 1995; Wang and Taylor, 1991).

The frequency at which CPDs form between any two adjacent pyrimidines differs according to the following rank formation: (5’-3’ order) TT > TC > CT > CC

(Ravanat et al., 2001) and for 6-4PPs the rank was: TC > CC > TT, whereas 6-4PPs at CT sites cannot be formed. Indeed only CPDs are generated at CT sites in isolates exposed to UVB radiation (Ravanat et al., 2001). Even if the CC site is the least

25 Chapter 1

photoreactive of the four bipyrimidine sites, it has strong mutagenic properties, because the cytosine can undergo spontaneous deamination (Douki and Cadet, 1992), resulting in the formation of uracil-containg CPDs, that code for an adenine like a thymine (Peng and Shaw, 1996).

IV.1.2 Indirect damage

Both ultraviolet radiation (UVB, UVA) and visible light may cause oxidative stress by generating reactive oxygen species (ROS). ROS are produced from oxygen,

- giving rise to active molecules, such as O2 , and •OH. Nucleic acid can be oxidized by ROS irrespective of their origin, (Buechter, 1988). More specifically, since nucleic acids bind iron well DNA is an especially favoured target of •OH mediated

2+ 3+ − through a Fenton reaction (Fe + H2O2  Fe + •OH + OH ) (Rai et al., 2001). The resulting DNA radicals produce a broad spectrum of lesions, including base and sugar lesions, strand breaks, DNA-protein cross-links and base-free sites. Reaction of pyrimidines and purines in DNA with ROS result in multiple products as illustrated in Fig. 1.9.

Although more than 20 base lesions have been identified in DNA, the main monomeric DNA base photoproducts are the photohydration of cytosine and the 8- oxo-2'deoxyguanosine (8-oxodG) (Fig. 1.10) (Cadet and Douki, 2005). 8-oxodG had been the focus of intense research, since it is one of the most abundant lesions and because guanine residues have a lower oxidation potential than cytosine, thymine, or adenine bases (Gajewski et al., 1990). It is also the most detrimental lesion because of its mutagenic effect (Ames and Swirsky-Gold, 1991).

26 Chapter 1

Figure 1.9. DNA base products of interaction with reactive oxygen species (Cooke et al., 2003a).

Figure 1.10. Main monomeric DNA base photoproducts (Cadet et al., 2005).

27 Chapter 1

Another indirect damage occurring during oxidative stress is the addition of an aromatic amino acid to DNA, leading to DNA protein cross linking (Dizdaroglu,

1998). Finally, reaction of •OH (the highly reactive hydroxyl radical) with the sugar moiety of DNA by H abstraction leads to sugar modifications and strand breaks

(Sonntag, 1987).

Accurate measurements of both direct and indirect UV-induced lesions are essential for a better understanding of their induction and their biological effects.

IV.1.3 Detection of DNA damage

A number of methods are in use to detect DNA damage. DNA degradation was first analyzed using radioactive methods and associating the decrease in radioactivity with

DNA degradation (O’Brien and Houghton, 1982). Some methods involve the use of

DNA repair enzymes that are able to nick DNA at the site of the photodamage, thus enabling the quantification of the resulting strand breaks (Seawell et al., 1980). An alkaline agarose gel method for quantitating single-strand breaks in nanogram quantities of nonradioactive DNA was developed by Freeman et al. (1986). The idea of Singh and colleagues in 1988 was to combine DNA alkali gel electrophoresis with fluorescence microscopy to visualize the migration of DNA strands. The Single Cell

Gel Electrophoresis assay (also known as the comet assay) is a simple technique for quantifying low levels of DNA damage in individual eukaryotic cells; it is especially sensitive for detecting DNA double- and single-strand breaks, alkali-labile damage, and excision repair sites (Singh et al., 1988). If the negatively charged DNA contains breaks, the DNA supercoils are relaxed and the broken ends are able to migrate

28 Chapter 1

toward the anode during electrophoresis. If the DNA is undamaged, the fragment is too large to migrate.

Another method is the ligation-mediated polymerase chain reaction. This allows the mapping of dimeric pyrimidine photoproducts and particularly of CPDs in DNA, the latter lesions are revealed through the nicking activity of T4 endonuclease V (Drouin and Holmquist, 1996).

Most of the studies devoted to the measurement of bipyrimidine photoproducts done over the past 20 years have used an immunological approach (Mitchell, 1996; Cooke et al., 2003b). Until recently, the majority of research involving the measurement of

DNA photoproducts in marine microorganisms used techniques such as radioimmunoassays (Jeffrey et al., 1996; Meador et al., 2002; Wilhelm et al., 2002) and a modified immuno dot-blot assay (Boelen et al., 2002). For the radioimmunoassay, a known amount of a radio-labeled UV-irradiated single-stranded synthetic homopolymer is used as an antigen and mixed with a known amount of antibody to that antigen. As a result, the two chemically bind to one another. The sample containing an unknown quantity of photodamage is then added. The unlabeled (or "cold") antigen from the serum competes with the radiolabeled antigen for antibody binding sites. The bound and free antigens are then separated and the radioactivity of the free antigen remaining in the supernatant is measured. The exact amount of photoproduct can thus be determined. Nowadays this technique has been partly supplanted by the ELISA method, in which the antigen-antibody reaction is measured using colorimetric signals instead of a radioactive signal.

29 Chapter 1

All the above techniques suffer from the same drawbacks: they detect the overall amount of photoproduct but cannot simultaneously detect individual CPD and 6-4PP lesions. Those limitations are overcome by the specific, quantitative and sensistive method that uses HPLC-MS/MS. This method can be used to monitor the formation of 12 possible dimeric pyrimidine photoproducts and 8-oxodeoxyguanosine (Douki et al., 2000). These photoproducts include TT, TC, CT and CC cyclobutane dimers, together with the related (6-4) adducts and Dewar valence isomers (Ravanat et al.,

2001). The technique involves DNA digestion with appropriate enzymes: 5’, 3’- exonucleases and alkaline phosphatase leading to the release of damaged DNA as dinucleoside monpophosphate in contrast to the undamaged bases liberated as nucleosides (Douki et al., 1997). The reaction mixture is then resolved by HPLC-

MS/MS, providing a sensitive, highly specific quantitative, assay. Each photoproduct is identified on the basis of its retention time, molecular weight and fragmentation pattern (Douki et al., 2000). HPLC-MS/MS has been successfully used to measure photoproducts in the cellular DNA of marine plankton (Jeffrey et al., 2004).

IV.1.4 Consequences of DNA damage

DNA damage has both cytotoxic and genotoxic effects. Both types of DNA lesions,

CPDs and 6–4PPs, block DNA/RNA polymerases. This may inhibit the polymerase action during DNA replication and transcription at or near the site of damage or result in lesion bypass with misincorporation that could eventually lead to mutations in either DNA or RNA (Sinha and Häder, 2002).

The CPDs are the most abundant and probably most cytotoxic lesions but the 6–4PPs may have more serious, potentially lethal, mutagenic effects (Sinha and Häder,

30 Chapter 1

2002). The primary response of the cell to UV-induced 6-4PP lesions is to trigger cell death (or suicide), whereas the main response of the cell to CPD lesions appears to involve cell cycle arrest. These findings suggest that CPD and 6-4PP have different biological effects in the UV-damaged cell. The "type" of DNA lesion is as important as the total "amount" of damaged DNA in the balance of survival and death (Lo et al., 2005).

A general response to DNA damage is a delay or arrest of the cell cycle (Fig. 1.11) to provide more time for DNA repair. Checkpoints were actually first described in bacteria, although the same terminology was not used when they were first proposed

(Cooper, 2006).

Cell Cycle Arrest Checkpoint

Stress Repaiir Response Pathways

DNA DAMAGE Regulation of transcription, Excision Repair translation and turnover of -Base Excision Repair (BER) multiple genes with various -Nucleotide Excision Repair (NER) responses -Mismatch Repair (MMR) Photoreactivation (PER) Strand Breaks Repair (RH / NHEJ)

Damage Cell Death Tolerance or Suicide

DNA/RNA polymerases bypass -Error free -Error prone (MUTAGENESIS)

Figure 1.11. Biological responses to DNA damage.

31 Chapter 1

For instance, the bacterial “replication checkpoint” was well described by

Helmstetter and Pierucci in 1968, showing the inhibition of the cell cycle progression when DNA synthesis is not complete. The cell cycle arrest may also enable the tight regulation of the large number of genes that somehow contribute to cell survival

(Fig. 1.11).

In some situations DNA or RNA polymerase can bypass the lesions, so giving rise to

DNA or RNA mutations. One such damage-tolerance mechanism, called translesion

DNA synthesis, involves the replication machinery bypassing sites of base damage, allowing normal DNA replication and gene expression to proceed downstream of the

(unrepaired) damage. Several new DNA polymerases that appear to be important in lesion bypassing have been discovered recently (Woodgate, 1999). These polymerases insert one or a few bases at the site of lesions. They then dissociate and are replaced by replicative polymerases (Tang et al., 1999). These novel DNA polymerases have very low fidelity, which makes it likely that they are responsible for DNA damage induced mutations (Woodgate, 1999). In fact, their low fidelity suggests that their synthesis and catalytic activity must be tightly regulated.

Bulky lesions such as CPD and 6-4PP photoproducts and breaks in the transcribed strand of the DNA template strongly interfere with the progression of E. coli RNA polymerases (Tornaletti and Hanawalt, 1999). A single unrepaired CPD is sufficient to completely block expression of a transcriptional unit (Protic-Sablic and Kreamer,

1985). DNA lesions such as 8-oxodG and abasic sites that are efficiently bypassed by bacterial RNA polymerase undergo base substitutions in the resulting transcripts.

32 Chapter 1

This so-called “transcriptional mutagenesis” could lead to the production of mutant proteins, with possible detrimental effects on cell physiology (Viswanathan and

Doetsch, 1998). Indeed, mutant proteins have the potential to alter nearly every aspect of cell behaviour and properties. Changes in protein structure and function due to amino acid substitutions or deletions are responsible for a multitude of biological responses, ranging from conferring selective growth advantages to cell death. Little is known about the nature and consequences of lesion bypass by RNA polymerase.

Interestingly, one potential consequence of transcriptional bypass is a switch of bacterial cells from a non-growth to a growth state. As illustrated in Fig. 1.12, cells in a non-growth state (part A) transcribe a miscoding DNA lesion (X) giving rise at the end of translation to mutant proteins. If the mutant proteins alter the cell’s physiology, they can induce de novo DNA synthesis (part B), so that the lesion (if still not repaired) becomes “fixed” as a permanent mutation in one of the two daughter DNA duplex. The other duplex produced from the undamaged template strand contains an undamaged gene that restors a non-growth state cell. A subsequent round of replication with or without lesion repair will lead to the generation of a dividing population of cells that has adapted to and could eventually escape from any original growth restrictive conditions (Doetsch, 2002). It should also be remembered that the capacity to continue replication in the presence of DNA damage and its potential mutation induction could be beneficial to unicellular organisms in evolutionary terms.

33 Chapter 1

Figure 1.12. Concept of retromutagenesis (Doetsch, 2002).

Experimental tools for evaluating the consequences of all DNA damage in the cell are becoming available. For example, Pandya et al. (2000) recently described an assay to determine the recombinogenic properties of single DNA adducts carried on plasmids. On a more global level, it would be interesting to test the role of DNA photoproducts in triggering changes in gene expression and protein modification.

Microarray and proteomic techniques can be used to determine the quantitative and qualitative changes induced by UV radiation.

IV.2 DNA repair mechanisms

To defend themselves all microrganisms have a wide variety of strategies to reverse, excise, or tolerate DNA damage (Tab. 1.2). Both bacterioplankton and phytoplankton can repair CPDs and 6-4PPs using light-dependent photoenzymatic

34 Chapter 1

repair (PER) or nucleotide excision repair (NER). Oxidative damage is repaired by base excision (Friedberg et al., 1995). The overall efficiency and relative contribution of each repair system to photoproduct removal is species-dependent and may be affected by synergistic interactions between the two communities (Mitchell,

1995; Joux et al., 1999; Meador et al., 2002).

Table 1.2. A summary of the DNA lesions and the most relevant DNA repair mechanism for removing the lesions.

Damage Direct (UVC, UVB) Indirect (UVB, UVA, PAR)

CPD 6-4PP Base modification DNA strand (oxidation, abasic site) breaks Repair NER NER BER RH PER (CPD photolyase) PER (6-4 photolyase) NER NHEJ MMR

IV.2.1 Excision repair

Dark repair is the light-independent repair of UV-induced damage and is believed to function more or less constantly. All dark repair mechanisms function by excision repair i.e., they cut out the damaged DNA and fill it with new material. There are two major excision repair pathways: BER and NER.

IV.2.1.1 Base excision repair (BER)

BER repairs damage to single nucleotides caused by oxidation, alkylation, hydrolysis or deamination. The pivotal enzymes in BER are DNA glycosylases. They remove modified or damaged bases by cleavage of the N-glycoside bond between the base and the 2-deoxyribose moieties of the nucleotide residues. Different DNA glycosylases remove different kinds of damage, and the specificity of the repair

35 Chapter 1

pathway is determined by the type of glycosylase involved (Seedberg et al., 1995).

Once the base is removed, the apurinic/apyrimidinic (AP) site is removed by an AP endonuclease or an AP lyase, which nicks the DNA strand 5’ or 3’ to the AP site, respectively and the -ose phosphate backbone is removed from the DNA by the action of deoxyribophosphodiesterase (dRpase). Finally the resulting single nucleotide gap is filled by the action of DNA polymerase, and a ligase attaches the repaired strand (Fig. 1.13). This repair pathway is less important in UVB-damaged

DNA since single base damage is only a minor part of UVB-induced DNA lesions.

36 Chapter 1

Figure 1.13. Diagram showing base excision repair (BER) (Augusto-Pinto et al., 2003).

IV.2.1.2 Nucleotide excision repair (NER)

NER removes a wide range of DNA distorting lesions, including CPDs, 6-4PPs and single-strand breaks. UvrABC endonuclease is a multienzyme complex that is involved in NER (Fig. 1.14). First, the UvrA dimer binds to UvrB to form a trimer

37 Chapter 1

which is able to detect DNA damage. The UvrB part attaches to the double helix at the damaged site. The UvrA dimer then leaves and a UvrC protein binds to the UvrB monomer resulting in the formation of a new UvrBC dimer. This dimer is responsible for cleaving the nucleotides either side of the DNA damage. UvrB cleaves phosphodiester bonds five nucleotides downstream of the DNA damage, and the UvrC cleaves another phosphodiester bond eight nucleotides upstream of the lesion, to create a twelve nucleotide excised segment that is removed by the DNA helicase II. Finally, DNA polymerase I fills in the correct nucleotides sequence and the last phosphodiester bond is completed by DNA ligase (Augusto-Pinto et al.,

2003).

6-4PP lesions are more efficiently removed than CPDs, because, as previously mentioned, 6-4PP lesions induce much more distortion of the DNA structure than

CPDs (Kim et al., 1995). Studies on human cells recently found that the nature of the pyrimidine involved in the dimeric lesion strongly affects the rate of CPD removal by the NER pathway, with the following decreasing order of efficiency CT > CC >

TC > TT (Mouret et al., 2008).

The coupling of DNA repair and transcription in bacteria is mediated by transcription-repair coupling factor (TRCF), which removes transcription elongation complexes stalled at DNA lesions and recruits the NER machinery to the site (Selby and Sancar, 1993).

38 Chapter 1

 



Figure 1.14. Diagram showing nucleotide excision repair (NER) (Augusto-Pinto et al., 2003).

IV.2.1.3 Mismatch repair (MMR)

Mismatch-repair (MMR) recognizes DNA mismatches that are produced during

DNA replication, cytidine deamination or other forms of DNA oxidative damage.

MMR is carried out by MutS, MutL and MutH in prokaryotes. MutS and MutL bind to DNA mismatches in an ATP-dependent manner and direct MutH to cleave the unmethylated DNA strand and initiate the repair process (Lahue et al., 1989). The

MutS protein homodimers recognize and bind specifically to base-base mispairings 39 Chapter 1

and insertion/deletion loop-outs (IDL). Then, MutS, in association with the MutL protein homodimers, activates the MutH protein to make an excision-initiating nick in the unmethylated, newly synthesized strand. The nicked strand containing the mismatch or IDL is excised by exonucleases and resynthesized by DNA polymerase and DNA ligase (Fig. 1.15) (Marra and Schär, 1999). MMR in E. coli targets indicates that there are seven of eight possible base–base mismatches, C:C mismatches being refractory to this system.

Figure 1.15. Schematic representation of methyl directed DNA mismatch repair in E. coli (Joseph et al., 2006).

40 Chapter 1

IV.2.2 Photoenzymatic repair (PER)

An alternative to the multi-step pathway of excision repair is photoenzymatic repair

(PER), also known as photoreactivation. This is a direct repair mechanism that is a less error-prone way of reversing pyrimidine dimers into their monomeric form. The process is catalyzed by a single enzyme, photolyase (Fig. 1.16). Photolyase is present and functional in many species, from the bacteria to fungi and animals.

Photoreactivation is a light-dependent process using UVA (320–400 nm) and blue light (400–500 nm) to monomerize pyrimidine dimers (Friedberg, 2003).

or Photolyase (green) 2 cofactors (yellow and orange)

Figure 1.16. The steps of photoreactivation (Friedberg, 2003).

41 Chapter 1

These lesions are recognized by a specific photolyase, which absorbs light at wavelengths of > 300 nm and facilitates a series of photochemical reactions that monomerize the dimerized pyrimidines, restoring them to their native conformation.

IV.2.2.1 Types of photolyase

Photolyases specifically bind to CPDs (CPD photolyase) or 6–4PPs (6–4 photolyase) and reverse the damage using the energy of the photoreactivating light (300-500 nm).

CPD and 6-4 photolyases have similar structures, the same chromophores, and the same basic reaction mechanisms. The 6-4 photolyase was first discoverd in

Drosophila (Todo et al., 1993) and later in Xenopus and Arabidopsis (Kim et al.,

1994; Nakajima et al., 1998). However neither E. coli nor any of the other bacteria tested to date have a photolyase capable of repairing 6-4PPs.

CPD photolyases are monomeric flavoproteins of 450-550 amino acids that contain two noncovalently bound chromophore cofactors, one of which is always FAD.

There are two classes of CPD photolyases, I and II, based on differences in their amino acid sequences. Class-I photolyases are found in many microbial organisms, whereas most of the class-II photolyases have been identified and cloned from higher organisms, including animals. The class-I photoreactivating enzymes are further divided according to the light-harvesting cofactor into either folate-types or deazaflavin-types in a limited number of species, with 5,10- methenyltetrahydrofolylpolyglutamate (MTHF) or 8-hydroxy-5-deazaflavin (8-HDF) as the second chromophore in a limited number of species. Although only the FAD of photolyase is required for its catalytic activity, the second cofactor significantly accelerates the reaction under low-light conditions (Sancar, 2003).

42 Chapter 1

IV.2.2.2 Binding specificity of photolyase

The binding of photolyase to CPDs depends on the base composition as follows: TT

> TU > UU > CC, with TT being bound 10-times better than CC. However, a followup study suggests that UU may be the more efficient substrate (Sancar, 2003).

Although it seems harder for the photolyase to flip out a lesion in a G-C-rich sequence than in an A-T-rich sequence, experimental tests of this prediction have never clearly demonstrated this until now (Svoboda et al., 1993), suggesting that the neighboring sequences have no impact on specific photolyase binding.

IV.2.2.3 Regulation of photolyase

The phr gene of E. coli encoding the apoenzyme of photoylase is not induced by the

SOS response, and as a consequence, the number of photolyase proteins in the cell should not increase after UV damage (Payne and Sancar, 1990). However, the photolyase gene in S. cerevisiae is induced by DNA-damaging agents as well as other general stressors (Sebastian et al., 1990; Sebastian and Sancar, 1991; Jang et al., 1999). Similarly, blue light induces the transcription of the photolyase gene in goldfish Carassius auratus (Yasuhira and Yasui, 1992) and in the fungus

Trichoderma harzianum (Berrocal-Tito et al., 1999). However, it is not yet clear whether this induction aids cell survival after UV damage in a significant way

(Birrell et al., 2002).

IV.2.2.4 Dual function of photolyase

The light-dependent repair of pyrimidine dimers has long been the sole known function of photolyases. An important indicator of an alternative function of CPD

43 Chapter 1

photolyase has come from the observation that photolyase can bind dimers in the absence of photoreactivating light both in vivo (Yamamoto et al., 1983; Sancar and

Smith, 1989) and in vitro (Sancar et al., 1984) as well as the fact that photolyase is found in many bacteria, such as E. coli, which are rarely exposed to UV light. It is now well recognized that, in the dark, photolyase after binding to the lesion, increases the poor helical deformity induced by CPDs (Husain et al., 1988; Park et al., 2002) by flipping out the bipyrimidine photoproduct and this could stimulates the

NER pathway (for a comprehensive review see Sancar, 2003). This light independent function of photolyase as an enhancer of NER remains to be fully characterized.

IV.2.3 Repair of double strand breaks

Although the bulk of DNA damage affects only one strand of a duplex DNA segment, both DNA strands are occasionally damaged opposite each other, resulting in two-strand damage. While strand breaks may not alter the coding sequence, a fracture of the DNA may interfere with normal DNA function and threaten the viability of cells. Two kinds of repair mechanisms may be involved in counteracting these deep lesions. DNA double strand breaks (DSBs) can be repaired by homologous recombination (HR) or by non-homologous end joining (NHEJ)

(Shuman and Glickman, 2007).

As illustrated in Fig. 1.17, in HR, a second intact complementary copy of the broken chromosome segment (typically in bacteria a newly replicated sister chromatid) serves as a template for DNA synthesis across the break. One or both of the DSB ends is resected by an exonuclease to leave a 3a single stranded tail that can then be paired with its complementary DNA sequence. This crucial process of locating and

44 Chapter 1

pairing the homologous sequence is performed by the RecA protein in bacteria. DNA polymerase subsequently copies the sequence information from the sister chromatid across the break and the residual single strand nicks in the repaired duplex are eventually fixed by the replicative DNA ligase (Lig A in bacteria). RH results in a faithfully restored chromosome that has no mutations (Shuman and Glickman, 2007).

Figure 1.17. Homologous replication and non-homologous end-joining pathways of double-strand break (Shuman and Glickman, 2007).

In contrast, the second type of repair mechanism, NHEJ does not rely on a homologous DNA template, which means it can operate in situations when only one chromosomal copy is available (Fig 1.17). Unlike HR, which is most of the time

45 Chapter 1

error-free, NHEJ can be mutagenic, if the ends are remodelled by nucleases or polymerases before sealing by DNA ligase.

V. Affects of solar radiation on gene expression in microorganisms

V.1 Transcriptome changes

Microorganisms often regulate their gene expression at the level of transcription or translation in response to solar radiation. The “transcriptome” and “proteome” consists of all the mRNA and proteins expressed by an organism at a given time under a given set of conditions. The wide range of approaches used to characterize the changes in the transcriptome and proteome are referred as transcriptomic and proteomic, respectively.

V.1.1 General technical overview

DNA microarrays are often used to measure changes in mRNAsynthesis. This multiplex technology used in molecular biology consists of an arrayed series of thousands of microscopic spots that can be a short section of a gene or other DNA element that are used as probes to hybridize a cDNA sample (called target) under high-stringency conditions. In standard microarrays, the probes are attached to a solid surface (glass or a silicon chip) by a covalent bond to a chemical matrix. Probe- target hybridization is usually detected and quantified by detection of fluorophore- labeled targets to determine the relative abundance of nucleic acid sequences in the target (Fig. 1.18).

46 Chapter 1

Figure 1.18. DNA microarrays technology.

V.1.2 State-of-the-art

DNA microarray technology has become a powerful tool for studying gene expression and regulation at the genome level under either damaging radiation and/or repair conditions. This paragraph describes several studies using the transcriptomic approach on different prokaryotic microorganisms (cyanobacteria, gram negative and gram positive bacteria, archaea).

Two studies focused on cyanobacteria and their adaption to different light intensities or UVB radiation. One used the marine cyanobacterium Prochlorococcus marinus

MED4 that has the most compact genome of all free-living photoautotrophs (1716 protein-coding genes) and the peculiarity of being high-light adapted (Steglich et al.,

2006). A transcriptomic approach was used to explore the pattern of responses to

47 Chapter 1

different light qualities and intensities. The expressions of approximately 10% and

5% of all genes were significantly altered by at least 2-fold, when shifted from darkness to high white light and white light, respectively. Not surprisingly, high-light had the most dramatic effect on gene expression. The most differently expressed group of genes belonged to the high-light-inducible genes (hli). The most highly upregulated gene (61.2-fold upregulated upon darkness-to-high-light transition) unfortunately encoded a protein of unknown function, as did the most highly downregulated gene (11.3-fold reduced in high light). Although seven different conditions were tested, only blue light elicited a strong response. Bacterial cryptochromes seem to be good candidates for the blue-light sensors, since the majority of known light-sensing proteins are absent from its genome (Steglich et al.,

2006).

A transcript profiling methodology was used to elucidate the expression patterns of the cyanobacterium Synechocystis sp. strain PCC 6803, in order to investigate changes in gene expression induced by irradiation with UVB and high-intensity white light. Several families of transcripts were found to be altered by both high- intensity white light and UVB, with a subsequent down-regulation of the genes involved in the light-harvesting system, photosynthesis, photoprotection, and the heat shock response (Huang et al., 2002). These two profile comparisons also corroborated the regulation of many pathways, including the synchronized induction of D1 protein recycling and a coupling between decreased phycobilisome biosynthesis and increased phycobilisome degradation. However, the gene expression profiles produced by high-intensity white light and UVB differed mostly in the regulation of several transcriptional processes, and in the regulation of the

48 Chapter 1

ribosomal protein transcripts, which are only repressed by UVB radiation (Huang et al., 2002).

The response of enteric bacteria, particularly E. coli, to solar UV light has been well investigated for more than 60 years (Hollaender et al., 1943). The growth of E. coli is inhibited by continuous UVA radiation with a subsequent adaptation to stress

(Berney et al., 2006). The transcriptomic approach was used to assess short-time stress and UVA light adapted growth. More genes (312) were expressed in the cells irradiated for a short time (1 h) than in UVA-adapted cells (50 h) (193). The number of up- and down-regulated genes was almost the same for both times of irradiation.

The results indicate that UV-induced the upregulation of the synthesis of several amino acids, such as valine, leucine, isoleucine, phenylalanine, histidine and glutamate, suggesting a possible direct inhibition of enzymes involved in the uptake or/and inhibition of N-assimilation. The findings of Berney and coworkers (2006) corroborate earlier reports on the induction of the SOS response in UVA-irradiated cells (Quillardet et al., 2003). The induction of genes such as those encoding recA, recN, dinD, dinl and UmuD, strongly points to DNA damage as a result of exposure to UVA light. Also the involvement of oxidative stress was confirmed with the induction of alkylhydroperoxidase reductase, the enzyme that converts lipid hydroperoxides to their corresponding alcohols. The RpoS gene was unexpectedly repressed in the light-adapted cells. This gene encodes a global stress regulator; several acid stress resistance genes were also significantly downregulated. This latter repression might be the result of shutting down unneeded biosynthesis (Berney et al.,

2006).

49 Chapter 1

Shewanella oneidensis MR-1 is an environmental gamma proteobacterium that is extremely sensitive to all wavelengths of UVR, natural solar radiation and ionizing radiation (Qiu et al., 2004). Because of its great sensitivity to different kind of radiations, this bacterium has been used in several studies to elucidate the transcriptional features underlying this special feature (Qiu et al., 2005; Qiu et al,

2006). Indeed, this sensitivity could not be explained by its genome, which is very similar to that of E. coli, with the major DNA damage repair/tolerance systems, including SOS response, recombination repair, mutagenic repair, nucleotide excision repair, mismatch repair and a DNA photolyase (Heidelberg et al., 2002). Qiu and colleagues (2005) first investigated and compared the genomic responses to UVC,

UVB and UVA. The authors found that S. oneidensis MR-1 expressed twice as many genes after exposure to UVA light than after UVC treatment. This result points to the

UVA light-induced stress response being very complex. Although the SOS response was observed with all three treatments, the induction was more robust in response to short wavelengths. Similarly, more prophage-related genes were induced by UVC and to a lesser extent UVB. The synthesis of antioxidant enzymes and iron- sequestring proteins were increased in response to UVA, enabling the cell to scavenge reactive oxygen species, such as ferritin like protein Dps, TonB dependent receptor, bacterioferritin and ferrochelatase. As mentioned in [IV.1.2] paragraph, the intracellular concentration of iron can have a detrimental effect in near UV radiation.

Iron containing proteins may act as chromophores, becoming excited and thereby being damaged directly. Hence, the regulation of iron uptake and metabolism and iron sequestration are important protection mechanisms against UVA induced oxidative damage. Contrasting results were observed with the same bacterial strain exposed to natural sunlight (Qiu et al., 2005b). As previously observed under

50 Chapter 1

artificial UVA lamps, the authors observed an induction of DNA damage-repair genes, the SOS response and detoxification strategies (Qiu et al., 2005b). However, a greater number of genes involved in detoxification were induced by natural solar radiation than by UVA lamps. No prophage gene was found to be induced by natural solar radiation. This suggests that the biological effects induce by the whole spectrum of natural solar radiation are not the simple sum of UVA and UVB effects.

The broader genomic response to natural sunlight could be due to synergistic effects of various UV wavelengths or possible effects induced by visible light. Finally, S. oneidensis was exposed to ionizing radiation (IR) and its transcriptome analyzed

(Qiu et al., 2006). IR is often used to produce reactive oxygen species. The genomic response of S. oneidensis to IR was very similar to its response to UVA and UVC radiation, with induction of genes encoding antioxidant enzymes, systems involved in DNA repair, prophage synthesis and the absence of differential expression of tricarboxylic acid cycle activity (Qiu et al., 2006).

Deinococcus radiodurans R1 is a gram-positive aerobic bacterium best known for its extreme resistance to the lethal effects of ionizing radiation. The transcriptomic approach was used to elucidate its puzzling resistance (Liu et al., 2003). Surprisingly, the function of most regulated genes under ionizing radiation was found to be either similar to other genes commonly identified in other bacteria exposed to damaging radiation or functionally uncharacterized. Thus these results do not enable us to clearly understand the extraordinary resistance of D. radiodurans towards damaging radation. If we compare the different IR-induced responses of the very sensistive bacterium, S. oneidensis with those of the highly resistant D. radiodurans we should be able to identify the factors peculiar to radiation resistant and radiation-sensitive

51 Chapter 1

bacteria. A prominent feature of the transcriptome IR-response of D. radiodurans is the differential regulation of its TCA cycle, where the isocitrate-to-fumarate steps of the TCA cycle are repressed and the glyoxylate bypass genes strongly induced by irradiation (Liu et al., 2003). Isocitrate lyase converts isocitrate to succinate and glyoxylate, allowing the carbon that enters the TCA cycle to bypass the formation of

-ketoglutarate and succinyl coenzymeA. It has been proposed that such regulation might strongly suppress oxidative stress in D. radiodurans, perhaps as a mechanism for preventing additional loss of genome integrity (Liu et al., 2003). Global repression of metabolism during DNA damage removal has been reported in E. coli and in the halophilic archaeon Halobacterium NRC-1 (Courcelle et al., 2001; Baliga et al., 2004). These observations contrast with the response of S. oneidensis, which showed no significant change in its TCA cycle, including its glyoxylate bypass. On the other hand, unlike D. radiodurans, S. oneidensis showed early induction of catalase, maybe the result of greater oxidative stress generated in S. oneidensis during irradiation (Liu et al., 2003). D. radiodurans and S. oneidensis appear to use different strategies to fight against oxidative stress following IR. This comparison of the two bacteria seems to better explain the high sensitivity of S. oneidensis than the strong resistance of D. radiodurans.

The last study focuses on the transcriptomic changes during a light or dark treatment, following UVC radiation (Baliga et al., 2004). The extremophile halophilic archaeon

Halobacterium NRC-1 that is remarkably resistant to UV-radiation was studied for its ability to repair damage under light or dark conditions. Baliga and colleagues

(2004) used gene knockout of the active photolyase gene (phr2) to rule out any photoreactivation repair. Three other putative repair mechanisms were identified

52 Chapter 1

including d(CTAG) methylation-directed mismatch repair, oxidative damage repair enzymes, and proteases for eliminating damaged proteins. It was also observed that light not only provided energy for the photolyase, but also facilitated global changes in regulation during DNA repair. This is not surprising, given that Halobacterium

NRC-1 has evolved mechanisms to use light as a source of energy through the action of photoactive ion pumps, proteorhodopsin (Baliga et al., 2004). Moreover, there is

UV-induced down-regulation of many important metabolic functions, including the genes involved in the TCA cycle, down-regulation of cell cycle genes and repression of gas vesicle biogenesis during repair that might prevent cells from floating toward the source of damage (Baliga et al., 2004).

V.2 Proteome changes

V.2.1 General technical overview

Proteomic technologies are powerful tools for studying the physiological responses of bacteria to environmental stress conditions and are complementary to the generation of databases of mRNA by microarray hybridization. Protein expression mapping has an advantage over monitoring mRNA levels in that it is a direct measure of the protein product of a gene. A wide range of MS-based approaches for quantifying proteins are available today, they can be gel-based or gel-free approaches.

Two-dimensional gel electrophoresis (2DE) allows the routine separation of thousands of proteins, and is thus the dominant technique in proteomics, though it was developed 30 years ago by O’Farrell and Klose (1975). Proteins are first separated according to their isoelectric point (pI), by gel-based isoelectric focusing

53 Chapter 1

(IEF). They are then subjected to a second orthogonal separation according to their relative molecular mass, by sodium dodecylsulphatepolyacrylamide gel electrophoresis (SDS-PAGE). The proteins in the gel are then stained and appear as spots. Stained intensity of protein spot patterns is used to assess the amount of each protein syntesized.

The development of fluorescent dyes to visualize proteins from 2D gels has significantly decreased gel-to-gel variability by using a single gel for cross sample comparison. This has increased the confidence that they are attributable to real biological variations rather than experimental problems (Van den Bergh and

Arckens, 2004). This recent multiplexing technology, the fluorescent two dimensional difference gel electrophoresis (2D-DIGE), overcome the limitations of traditional 2D gels - mainly reproducibility, since it is based on the use of a single gel for the simultaneous separation of multiple protein samples, followed by the independent visualization of each individual sample. 2D-DIGE allows the simultaneous detection and quantification of paired samples on a single gel, by covalently tagging samples at lysines by distinct and spectrally resolvable fluorescent derivatives of the cyanine dyes Cy2, Cy3 and Cy5. Two of the dye reagents (Cy3 and

Cy5) are usually used for independent labeling of each sample, whereas the third

(Cy2) is used to label a pool of both samples, thus generating an internal standard. As all three dyes have distinct excitation and emission wavelengths, the three labeled samples can be mixed prior to IEF and analyzed simultaneously in a single 2DE experiment. Thus, the corresponding protein spot patterns can be visualized by illuminating the gel with the excitation wavelength of each of the dyes (Viswanathan et al., 2006). The resulting images can be analyzed using software specifically

54 Chapter 1

designed for 2D-DIGE to highlight proteins with different concentrations. Naturally, multiple gels remain necessary to obtain statistically significant results (Van den

Bergh and Arckens, 2004). However, two-dimensional electrophoresis still has several limitations, despite the development of the fluorescent dyes in the 2D-DIGE method that have strikingly improved the reproducibility (Rabilloud, 2002).

A primary limitation is the difficulty in detecting low protein concentrations.

Furthermore, certain classes of proteins are absent or underrepresented in a 2DE map, such as very acid/basic proteins, or very large/hydrophobic proteins associated with membranes. Because membrane proteins account for approximately 30% of total proteins (Wallin and Von Heijne, 1998), this is a serious problem for characterization of proteomes. In general, only proteins with a molecular weight of

10-100 kDa and a pI of 4-8 migrate well within 2D gels. Another problem is overlapping spots, due to a very abundant protein in a given sample. The best illustration of this drawback is the highly abundant ribulose bisphosphate carboxylase/oxygenase complex (RuBisCO) proteins found in all photosynthetic organisms, which overwhelms many of the less abundant proteins that would otherwise be clearly resolved in a 2D gel (Kim et al., 2001). Proteolytic degradation and posttranslational modifications of the sample can also result in the same protein appearing at several locations on a gel. If this protein is abundant, it can further obscure several low abundant proteins.

Stable isotope labeling is the most comprehensisve gel-free approach for measuring overall protein abundance and can be performed by in vitro (ICAT, iTRAQ) and in

55 Chapter 1

vivo (e.g., metabolic labeling) approaches (Gygi et al., 1999; Krijgsveld et al., 2003;

Zhong et al., 2004).

There are some problems with using ICAT for comparative expression studies in response to radiation conditions. Several amino acids are prone to oxidation upon solar radiation. Cysteine can be modified by damaging radiation to form oxidized cysteine sulfhydryl residues, preventing ICAT labeling and increasing unmodified cysteines residues. Gel-free studies on the impact of solar radiation will have to use a new kind of targeting labeling such as the amine-reactive isobaric tagging reagents available with the iTRAQ technology.

The iTRAQ (isobaric tag for relative and absolute quantitation) method is a gel-free technique for identifying and quantifying proteins in up to eight different conditions in one single experiment. The method involves labeling proteins in vitro with a set of isobaric reagents (isobar-coded affinity tag) that yield the N-terminus and side chain amines of peptides from tryptic digests (Fig. 1.19). Two reagents are currently mainly used: 4-plex and 8-plex, which can be used to label all peptides from different condition in a same experiment. These samples are then pooled and usually fractionated by liquid chromatography and analyzed by tandem mass spectrometry

(MS/MS). A database search is then performed using the fragmentation data to identify the labelled peptides and hence the corresponding proteins. The fragmentation of the attached tag generates a low molecular mass reporter ion that can be used to relatively quantify the peptides and the proteins from which they originated, using software such as ProQuant.

56 Chapter 1

A disadvantage of both ICAT and iTRAQ is that the stable isotope label is introduced into the sample only after several stages of sample preparation, such as cell lysis, protein extraction and proteolysis. Therefore, it is usually preferred to introduce the stable isotope label very early in the process, such as in the metabolic labeling approach. By using this method, protein can be metabolically labeled by growing cells in either 14N minimal or 15N-enriched media, as the sole nitrogen source that can be incorporated during protein synthesis (Goshe and Smith, 2003).

This technique provides a complete proteome coverage in which all peptides are labelled. So far, metabolic labeling has been applied mostly to unicellular organisms, such as yeast (Oda et al., 1999) and bacteria (Conrads et al., 2001), which can be easily grown on defined media in the laboratory.

57 Chapter 1

Protein exctracts Condition 1 Condition 2 Condition 3 Condition 4

C term

Denaturation N term

Trypsin digest + iTRAQ labeling 114 label 115 label 116 label 117 label

Combine

Quantification iTRAQ reporter ions

Strong cation exchange (SCX) + Reversed Phase (RP) fractionations

Identification MS/MS fragmentation

LC/MS-MS Analysis

Figure 1.19. The steps in the iTRAQ method.

58 Chapter 1

V.2.2 State-of-the-art

Unlike the transcriptomic approach, only a few quantitative proteomic studies have reported on the impact of damaging solar radiation on prokaryotic micoorganisms.

We will examine three of them that decribe the impact of ionizing radiation, UVB and various intensities of visible light on three different bacteria.

A quantitative proteomic approach with 2D gels was used to compare the proteomes of irradiated and unirradiated D. radiodurans bacteria to identify the mechanisms of its extreme radioresistance and DNA repair (Zhang et al., 2005). Only 26 protein spots showed significant changes between both conditions of irradiation and 21 proteins were correctly identified by mass spectrometry. Most of these proteins had cellular functions, but none of them was known to be relevant to radioresistance, except for the single-stranded DNA-binding protein (SSB) and PprA proteins, that was previously found to be critical for the radiation resistance of D. radiodurans as an enhancer of DNA ligation (Narumi et al., 2004).

The cyanobacterium Nostoc commune is a dessication-tolerant terrestrial cyanobacterium and its protein expression map was analysed during continous growing continuous culture under UVB radiation using 2D gels. The proteome was first fractionated into three sub-proteomes (intracellular, membrane, sectreted) before proteomic analysis. 493 protein spots showed significant changes (at least by a factor three compare to their corresponding unirradiated growing culture), rendering the

UVB stimulon of Nostoc commune the most complex one described to date (Ehling-

Shulz et al., 2002). It has two parts: an early shock response influencing 214 proteins and a late acclimiatation response involving 279 proteins, with no or few overlaps

59 Chapter 1

between the two responses. The shock response involves many membrane proteins, whereas the acclimation response mainly affects cytosolic proteins. UV irradiation induced superoxide dismutase and the water stress protein in the extracellular fraction. A total of 27 UVB-induced proteins were partially sequenced by mass spectrometry. They were mainly involved in lipid and carbohydrate metabolism and in regulatory pathways. As much as 50% of the sequenced proteins of the UVB acclimation response remained uncharacterized, with no known function.

This study supports the idea that short-term stress and acclimation responses to damaging radiation are two completely different and remarkably complex strategies

(Ehling-Shulz et al., 2002).

The last study is more linked to the thesis topic since it focused on the light adaptation of the well known marine cyanobacterium Procchlorococcus marinus

MED4 assessed by a quantitative proteomic approach using iTRAQ (Pandhal et al.,

2006). MED4 is adapted to an oceanic environment with low nutrient levels and high light intensities. The microorganism was cultured under three light intensities (low, medium and high). Approximately 11% of the proteome was identified. 15 proteins were deemed to be significantly influenced by changing light intensities, particularly the down-regulation of photosystem-related proteins, and the up-regulation of the stress related chaperones in high light compared to low light treatment. Six stress- related proteins were identified: GroEL, DnaK, HtpG, ClpC, and GroES. GroEL,

HtpG, and GroES were deemed to be significantly up-regulated in response to high light intensities. Chaperonin proteins prevent cell damage by preventing the accumulation of misfolded or unfolded proteins that have lost their function.

60 Chapter 1

V.3 UV-induced protein damage

Solar radiation can generate a wide range of protein damage due to oxidation, such as amino acid modifications (Tab. 1.3), carbonyl group formation, fragmentation, formation of protein-protein cross-links, formation of S–S bridges.

Table 1.3. Amino acids residues of proteins that are oxidized and products formed.

Amino acids Oxidation products

Arginine Glutamic semialdehyde

Cysteine Disulfides, cysteic acid

Glutamyl Oxalic acid, pyruvic acid

Histidine 2-Oxohistidine, aspragine, aspartic acid

Lysine 2-Aminoadipic semialdehyde

Methionine Methionine sulfoxide, methionine sulfone

Phenylalanine 2,3-Dihydroxyphenylalanine, 2-,3-and 4-hydroxyphenylalanine

Proline 2-Pyrrolidone, 4- and 5- hydroxyproline pryoglutamic acid, glutamic semialdehyde

Threonine 2-Amino-3-ketobutyric acid

2-, 4-, 5-, 6-, and 7-, hydroxytryptophan, nitrotryptophan, kynurenine, 3- Tryptophan hydroxylcynurenine, formylkynurenine

3,4-Dihidroxyphenylalanine, Tyr-Tyr cross-linkages, Tyrosine Tyr-O-Tyr, Cross-linked nitrotyrosine

Carbonyls are relatively difficult oxidative modifications to induce. Carbonylation is an irreversible oxidative process unlike methionine sulfoxide and cysteine disulfide bond formation (Dalle-Donne et al., 2003). Thus, a cell must rid itself of carbonylated proteins by degrading them. Biochemical analysis revealed that

61 Chapter 1

carbonyl groups introduced into the side chains of specific amino acids in the active center of a protein trigger its degradation (Levine, 1983, 2002). Carbonyl derivatives are formed by a direct metal-catalyzed oxidative (MCO) attack on the amino-acid side chains of proline, arginine, lysine, and threonine. Carbonyl derivatives of lysine, cysteine, and histidine can also be formed by secondary reactions with reactive carbonyl compounds on carbohydrates (glycoxidation products), lipids, and advanced glycation/lipoxidation end products. The quantitatively most important products of the carbonylation reaction are glutamic semialdehyde from arginine and proline, and aminoadipic semialdehyde from lysine (Requena et al., 2003).

It was found that the cell protein oxidation is not limited by available reactive oxygen species (ROS) alone, but also by the levels of aberrant proteins.

Mistranslated proteins (due to DNA/RNA mutations) or otherwise misfolded proteins

(due to DNA/RNA mutations and/or chaperone deficiency) may become more susceptible to carbonylation (Fig. 1.20) (Dukan et al., 2000).

Proteomics demonstrated that this carbonylation is closely associated with the production of aberrant protein isoforms (Ballesteros et al., 2001). The rapid carbonylation of mistranslated or otherwise aberrant proteins points to an important physiological role of carbonylation in protein quality control. Since carbonylated proteins are more susceptible to proteolytic degradation than their nonoxidized counterparts (Dukan et al., 2000; Bota and Davies, 2002; Grune et al., 2003, 2004), the rapid carbonylation of an erroneous protein may ensure that it is directed to the proteolyse process. Thus, carbonylation may act as a signal ensuring that damaged proteins enter the degradation pathway rather than the chaperone/repair pathway

62 Chapter 1

since carbonylation is an irreversible/unrepairable modification. However, highly carbonylated proteins can sometimes form high-molecular-weight aggregates that are proteolysis-resistant. Such aggregates appear to inhibit protease function (Fig. 1.20)

(Ballesteros et al., 2001).

DNA damage ROS

Carbonylation

Mistranslated or midfolded proteins

Aggregation Proteolyse Chaperones

Figure 1.20. Formation of carbonylated proteins and their fate in the cell.

The response of E. coli to oxidative stress differs according to the intracellular concentration of carbonylated proteins (Desnues et al., 2003). Indeed, the cells with low concentrations of carbonyl products remain reproductively competent, whereas cells with a high carbonyl load become genetically dead (unculturable).

ROS mediating protein oxidation could be as high as those of other ligand-stimulated

Fenton reactions (Cooke et al., 2003a). Some proteins appear to be more vulnerable

63 Chapter 1

than others, perhaps because of their tendency to bind certain metals. At present it is not clear what fraction of the vulnerable enzymes becomes modified during oxidative stress, or whether such modifications typically interfere with protein function. This information is necessary in order to estimate the impact that this damage might have on cell viability.

Sensitive methods have been developed for detecting and quantifying protein carbonyl groups, and most of these involve derivatization of the carbonyl group with

2,4-dinitrophenol hydrazine and subsequent immunodetection of the resulting hydrazone using monoclonal or polyclonal antibodies (Levine, 2002).

VI. Thesis outline

VI.1 Impact of GC content on DNA damage

The GC content of bacterial DNA varies greatly, from 25 to 75%, while that of eukaryotic cells varies very little from one species to another with a GC% of 40 to

45%. As shown in section IV.1.1, direct DNA damage can be induced at four positions in the genome, depending on the nature of the two adjacent pyrimidines involved. Thus the highly variable GC content of bacteria should influence the nature of the bipyrimidine photoproducts induced at positions TT, TC, CT and CC.

Furthermore, since TT CPD was the first to be identified and is the most frequent

DNA photoproduct in mammalian cells, it has been quickly assumed that a low TT

CPD content provides greater resistant to UV radiation. Therefore bacteria with different GC% should have different frequencies of TT CPDs, affecting their UV resistance.

64 Chapter 1

However, several studies have reported a poor correlation between UV resistance in bacterial species and their GC content (Gascon et al., 1995; Joux et al., 1999;

Agogué et al., 2005). These observations probably indicate that the resistance of microorganisms to UVB is a complex combination of the amount of DNA damage induced and the efficiency of their repair mechanisms. Furthermore, the hypothesis of a link between photosensitivity and thymine content in the genome also disregards the fact that cytosine gives rise to CPDs and 6-4PPs (Freeman et al., 1965; Liu and

Yang, 1978; Franklin et al., 1985). Many early studies failed to provide a complete distribution of UV-induced photoproducts because of analytical limitations. Chapter

2 describes the qualitative distribution of bipyrimidine photoproducts in six DNAs with GC contents from 28 to 72% in order to reinvestigate the GC content and DNA damage that has not been studied for the last three decades. We used the appropriate, highly sensitive technique of HPLC-MS/MS to detect and quantify all possible bipyrimidine CPDs and 6-4PPs.

Chapter 3 describes the responses of two marine bacteria with different GC% to

UVB damaging radiation or repair conditions to further understand the role of the

GC. The oligotroph ultramicrobacterium Sphingopyxis alaskensis (GC%=65.5) was compared to the copiotroph Photobacterium (formerly Vibrio) angustum S14 angustum (GC%=39.6).

VI.2 Factors modulating bacterial physiological states and responses to

UV

Temperature and growth phase may influence the response of marine bacteria to UV.

65 Chapter 1

VI.2.1 Temperature

The tempetature at the surface of oceans varies widely from about -1°C to 30°C.

Hence, bacteria are exposed to a range of temperatures in combination with solar radiation at the oceanic surface. The independent effects of temperature on the physiology of phytoplankton and bacterioplankton have been well established

(Raven and Geider, 1988; Davison, 1991). However, less is known about how temperature mediates the physiology, DNA damage, and DNA repair of oceanic plankton exposed to UVR. Chapter 4 examines the effects of temperature (12°C versus 24°C) and UVR on the survival and DNA damage in the oligotropic bacterium, Sphingopyxis alaskensis.

VI.2.2 Growth phase

Bacteria can modulate their UV response depending on their physiological state, and subsequently their growth phase. Several studies have shown that the sensitivity of E. coli to UVC and sunlight is greater during exponential phase than in the stationary phase (Gourmelon et al., 1997; Morton and Haynes, 1969; Reed, 1997).

A novel hypothesis has recently been advanced to explain this difference in sensitivity, the suicide response (Aldsworth et al., 1999). The suicide response predicts that rapidly growing and respiring cells will suffer growth arrest when subjected to relatively mild stresses, but that their metabolism will continue. A burst of free radical production results from this uncoupling of growth from metabolism and it is this free radical burst that is lethal to the cells, rather than the stress per se.

(Aldsworth et al., 1999). The second goal of the Chapter 4 will be to examine the

66 Chapter 1

effects of the early and late stationary phases in combination with UVR on the survival and DNA damage formation in the oligotropic bacterium, S. alaskensis.

A recent proteomic analysis of the stationary phase in the marine “Ca. Pelagibacter ubique” showed that the stationary phase significantly increases the synthesis of a suite of proteins (Sowell et al., 2008). The stationary phase induces the synthesis of

OsmC and thioredoxin reductase, which may mitigate oxidative damage in “Ca.

Pelagibacter ubique”, as well as that of molecular chaperones, enzymes involved in methionine and cysteine biosynthesis, proteins involved in rho-dependent transcription termination, and the signal transduction enzyme CheY-FisH. Therefore a significant change in the proteome content of an oligotrophic bacterium could modulate its response to UV radiation.

In order to validate this hypothesis, Chapter 5 documents the changes in protein synthesis in the oligotophic bacterium S. alaskensis, harvested in the midlog and stationary phase and subsequently exposed to solar radiation. Protein synthesis was assessed by a quantitative proteomic approach, using an iTRAQ method.

VI.3 Role of spectral composition and UV intensity

As decribed in section/paragraph III.1, marine bacteria are exposed to different spectral bands according to their position in the water column. Basically, bacteria at the surface are exposed to the full sun spectrum that is to say UVB+UVA+PAR; those lower in the water column are exposed to UVA+PAR; then PAR and ultimately Darkness.

67 Chapter 1

Chapter 5 also examines how the four spectral bands of solar radiation affect the regulation of protein synthesis by S. alaskensis. Different filters were used with a solar simulator to mimic experimentally the relevant ecologically exposures found in the field.

Several studies (Berney et al., 2006; Ehling- Schulz et al., 2002) have also shown that the duration of irradiation plays a pivotal role in the UV response. Chapter 3 presents the responses of two marine bacteria to various UVB doses. Chapter 5 deals with the proteomic part of the response of S. alaskensis to UV and takes into account the possible differences in its response to short-term radiation and UV- acclimatation. Because bacterial protein synthesis often depends on many different factors, it is important to analyze cells that are grown under several different conditions in order to acquire a comprehensive library of protein profiles.

68 CHAPTER 2

Effect of the GC content of DNA on the distribution

of UVB-induced bipyrimidine photoproducts Chapter 2

CHAPTER 2

Effect of the GC content of DNA on the distribution of UVB-induced bipyrimidine photoproducts

Abstract

Solar UV radiation is a major mutagen that damages DNA through the formation of dimeric photoproducts between adjacent thymine and cytosine bases. A major effect of the GC content of the genome is thus anticipated, in particular in prokaryotes where this parameter significantly varies among species. We quantified the formation of UV-induced photolesions within both isolated and cellular DNA of bacteria of different GC content. First, we could unambiguously show the favored formation of cytosine-containing photoproducts with increasing GC content (from 28 to 72%) in isolated DNA. Thymine–thymine cyclobutane dimer was a minor lesion at high GC content. This trend was confirmed by an accurate and quantitative analysis of the photochemical data based on the exact dinucleotide frequencies of the studied genomes. The observation of the effect of the genome composition on the distribution of photoproducts was then confirmed in living cells, using two marine bacteria exhibiting different GC content. Because cytosine-containing photoproducts are highly mutagenic, it may be predicted that species with genomes exhibiting a high GC content are more susceptible to UV-induced mutagenesis.

70 Chapter 2

I. Introduction

Solar UV radiation reaching the Earth’s surface, and especially the most energetic

UVB portion (290–320 nm), is widely recognized as a major environmental mutagenic agent. Although emphasis is often placed on the role of UVB in the etiology of skin cancer (Melnikova and Ananthaswamy, 2005), this radiation also significantly affects microorganisms. Indeed, UVB has been proposed to be a key evolutionary force (Tenaillon et al., 2004) and to mediate some of the deleterious effects of increased UV fluence resulting from ozone layer depletion (Buma et al.,

2003). These properties mostly result from the ability of UVB to modify the chemical structure of DNA through the formation of covalent bonds between the two adjacent bases of bipyrimidine dinucleotides following absorption of incident photons. For each bipyrimidine doublet, damage may be of two different types (Fig.

2.1), namely cis, syn cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6–4) pyrimidone photoproducts (6-4PPs) (Sage et al., 2005). At high UVB fluence, 6-

4PPs can be converted into Dewar valence isomers upon intramolecular cyclization of the pyrimidone ring. A total of twelve photoproducts are expected but they are not all produced with the same frequency. First, the yield of photoproducts depends on the nature of the two pyrimidine bases involved in the photoreaction. In addition, a number of structural (Demidov and Potaman, 1993; Potaman and Soyfer, 1995;

Kundu et al., 2004) and sequence effects (Drouin and Therrien, 1997) have been described. The GC and AT base pair composition of genomes has long been known as another important factor in the photochemistry of DNA (Setlow and Carrier,

1966). This specific feature is of limited impact for eukaryotic cells because their genomic GC content exhibits only small variations from one species to the other, in the range from 40 to 45%. In contrast, the GC content of DNA among prokaryotes is

71 Chapter 2

highly variable with values as low as 25% and as high as 75% GC base pairs

(Mooers and Holmes, 2000).

Figure 2.1. Chemical structure of thymine–cytosine dimeric photoproducts. Similar compounds are produced at TT, CT and CC dinucleotides. The bottom line shows the deamination of TC CPD that yields the related uracil–thymine CPD. The same reaction occurs for CT and CC CPDs.

Because TT CPD has been the first identified and was found to be the most frequent

DNA photoproduct in mammalian cells, it was hypothesized that DNA with lower thymine content would be less sensitive to UV radiation (Singer and Armes, 1970).

However, this hypothesis was quickly challenged (Bak et al., 1972). More recent studies have reported a poor correlation between UV resistance in bacterial species and their GC content (Gascon et al., 1995; Joux et al., 1999; Agogué et al., 2005).

These observations likely indicate that the resistance of microorganisms to UVB is a complex combination between the amount of DNA damage induced and the efficacy of the different repair mechanisms. The hypothesis of a link between photosensitivity and thymine content in the genome also disregards the fact that cytosine has also been shown to give rise to CPDs and 6-4PPs (Freeman et al., 1965; Liu and Yiang,

72 Chapter 2

1978; Franklin et al., 1985). Indeed, many early studies failed to provide the complete distribution of UV-induced photoproducts because of analytical limitations.

The most widely applied technique, based on the acidic release of CPDs from DNA by cleavage of the N-glycosidic bond, did not allow the individual quantification of

TC and CT CPDs (Varghese, 1972). In addition, the data obtained on the yield of 6-

4PPs are not reliable because these photoproducts were partially destroyed under the experimental conditions used (Douki et al., 1995). The effects of climate change and subsequent potential evolutionary influence of photo-induced DNA damage in prokaryotes (Häder et al., 2007) led us to reinvestigate the topic of GC content and

DNA damage that was left unstudied in the last three decades. For this purpose, we quantified DNA photoproducts using an assay involving mild enzymatic release of

DNA photoproducts followed by HPLC mass spectrometry analysis that makes possible quantification of all possible bipyrimidine CPDs and 6-4PPs (Douki and

Cadet, 2001a; Mouret et al., 2006). Another improvement of the present study is that, on the basis of the numerous and recent genome sequencing programs, accurate dinucleotide frequencies (Karlin and Burge, 1995) could be correlated with photoproducts formation, providing more accurate analysis than the average GC content.

73 Chapter 2

II. Materials and methods

II.1 Bacterial strains, media and culturing conditions

Sphingopyxis alaskensis RB2256 (DSMZ 13593T), and Photobacterium angustum

S14 strain (gift from Professor R. Cavicchioli, UNSW, Sydney, Australia) were maintained on VNSS agar using reported procedures (Eguchi et al., 1996)

Salmonella typhimurium LT2 (CIP 60.62T, Collection Institut Pasteur, France) was maintained on Nutrient agar. S. alaskensis and P. angustum were grown in artificial sea water (ASW) with 3 mM of D-glucose and S. typhimurium were grown in

Nutrient Broth at 24°C under magnetic stirring (~220 rev min−1).

II.2 UVB irradiation of isolated DNA

DNA extracted from microorganisms was made soluble in 1 ml of a 0.1 M NaCl aqueous solution in concentrations ranging between 0.5 and 2.3 mg ml−1 depending on the species. The solutions were centrifuged and the supernatant collected. Aliquot fractions of 300 µl were exposed for 5 or 30 min to UVB radiation in 1 × 1 cm quartz cells with the lamp placed above (2 × 15 W Vilber Lourmat 215 G, Bioblock,

Illkirch, France). The emission spectrum of the UVB lamp (normalized at 312 nm) was the following: 280 nm (5%), 290 nm (40%), 312 nm (100%), 330 nm (60%) 350 nm (20%), 370 nm (5%). The fluence rate was 0.0219 J cm−2 min−1 as measured with a VLX 3 W radiometer (Vilber Lourmat, Marne La Vallée, France) equipped with a

312 nm probe. After irradiation, DNA was precipitated by addition of ethanol. The resulting residue was made soluble in water prior to HPLC-MS/MS analysis.

Irradiation of high grade commercially available DNA was conducted as described above with additional steps. Solutions at 1 mg ml−1 were prepared in 0.1M NaCl. The samples were dialyzed overnight in 0.1 M NaCl using 10 000 Da molecular weight

74 Chapter 2

cut-off dialysis cassettes. Then, the three cassettes were dialyzed in the same 500 ml pyrex beaker containing a fresh solution of 0.1 M NaCl. This second dialysis step lasted 4 h. The concentration of the resulting DNA solutions was determined spectrophotometrically. Samples of purified DNA at 0.1 mg ml−1 in 0.1 M NaCl were then prepared and exposed in triplicate to UVB radiation (0.0219 J cm−2 min−1) for 2, 5 and 10 min. DNA was precipitated by addition of ethanol and dissolved in pure water before digestion and HPLC-MS/MS analysis. All experiments were done in triplicate and reported results represent the average ± standard deviation.

II.3 UVB irradiation of bacterial cells and DNA extraction

Bacteria cells were harvested at the beginning of stationary phase (i.e., after 12 h of growth for P. angustum and 60 h for S. alaskensis). Prior to UVB exposure, cultures were washed twice by centrifugation (8 000 x g, 8 min) and diluted with ASW to obtain an OD433 nm of 0.1. Washed cells (200 ml) were placed into cylindrical quartz flasks (flat top) covered with cellulose acetate (50% transmission at 280 nm) to remove residual UVC of the UVB lamps. Cells were maintained at 24°C by placing the flasks in a water bath connected to a cryothermostat, with continuous agitation

(magnetic stirring ~220 rev min−1). Cells were irradiated under two UVB lamps

(UVB 313, Q Panel) during 20 min with a mean fluence rate of 0.0042 J cm−2 min−1 measured with a spectroradiometer (RAMSES, TriOS). DNA from bacterial cells before and after UVB irradiation was extracted using Wizard® genomic DNA purification kit (Promega). DNA was stored at −20°C until use. All experiments were done in triplicate and reported results represent the average ± standard deviation.

75 Chapter 2

II.4 HPLC-MS/MS analysis of bipyrimidine photoproducts

DNA in solution in water was digested as previously described (Mouret et al., 2006)

Briefly, the protocol involved a 2 h incubation step at 37°C and pH 6 with nuclease

P1, phosphodiesterase II and DNase II. The pH was then raised to 8 and the samples were incubated for 2 h (37°C) in the presence of phosphodiesterase I and alkaline phosphatase. The resulting solution contained unmodified bases as nucleosides and bipyrimidine photoproducts in the form of dinucleoside monophosphates. All cytosine-containing CPDs as well as CT and CC 6-4PPs were analyzed as deaminated products. The dimeric photoproducts were quantified by HPLC coupled to tandem mass spectrometry in the multiple reaction monitoring mode. Under these conditions, the deprotonated pseudomolecular ion of each photoproduct was collected and fragmented in the spectrometer and a specific daughter ion was quantified as previously set-up (Douki and Cadet, 2001). Normal nucleosides were quantified by a HPLC UV detector placed before the mass spectrometer. Results were first expressed in number of lesions per either 104 or 106 bases. For comparison, data were also presented as relative yields. In such cases, the yield of specific photoproducts was expressed in percent with respect to the sum of all the individual yields.

II.5 Dinucleotide frequencies

Dinucleotide frequencies in various genomes were determined with the ‘wordcount’ program (EMBOSS) (Rice, 2000) and applied to sequences from the NCBI database

(http://www.ncbi.- nlm.nih.gov/genomes/lproks.cgi) except for P. angustum S14 which have been communicated by R. Cavicchioli. Data for calf thymus DNA are those reported for vertebrates by Karlin and Burge (1995). Since the genome of

76 Chapter 2

Micrococcus luteus has not been sequenced, dinucleotide frequencies were interpolated from the values for 8 other species with similar GC content. Such a calculation was validated by the linearity in the variation in frequencies in the GC content range (Appendix A).

77 Chapter 2

III. Results and discussion

III.1 Frequencies of bipyrimidine dinucleotides at different GC content differs from statistical distribution

The present investigation aimed at accurately determining the effect of the GC content on the formation of bipyrimidine photoproducts in isolated and cellular

DNA. Obviously, the proportion of the four bipyrimidine dinucleotides varies according to the GC content of the genome, for instance with few TT at high and much more at low GC content. Correlation between the yields of formation of the different photoproducts could thus be made with frequencies of dinucleotide calculated on the basis of a statistical distribution as presented in the early reports on

CPDs (Setlow and Carrier 1966). Interestingly, the efforts devoted to genome sequencing carried out in the last decade made available the exact content in dinucleotides for a wide array of species. These figures often significantly differ from a statistical distribution (Karlin and Burge, 1995; Karlin et al., 1994;

Nakashima et al., 1997). Therefore, we extracted from databases the full sequence of

99 bacteria and determined therein the frequency of bipyrimidine dinucleotides (Fig.

2.2 and Appendix B). Significant deviations were observed from a statistical distribution. The frequency of TT was higher than expected at low GC content, whereas that of CC was lower at high GC content. A large scatter of the data was observed for CT and TC at GC contents below 50% GC. However, TC was found to be overrepresented and CT underrepresented at higher GC content. These data clearly show that interpretation of the photochemical data in the light of a statistical distribution could be misleading.

78 Chapter 2

Figure 2.2. Frequencies of the bipyrimidine nucleotides in the genomes of 99 bacteria (see Appendix B). Top panel: frequency of TT and CC dinucleotides. Lower panel: frequency of TC and CT dinucleotides. On both figures, the solid line shows the variation in dinucleotide frequency calculated on the basis of a statistical distribution. Under these conditions, the frequency of TC and CT sites should be equal.

79 Chapter 2

III.2 GC content differently affects the yields of formation of the different

DNA photoproducts

In a first series of experiments, we collected accurate and quantitative data on the effect of the GC content of isolated DNA on the yield of UVB-induced DNA photoproducts. For this purpose, we used commercially available isolated DNA of highest available grade that we further purified by extensive dialysis. DNA from three different origins with different % GC was used: Clostridium perfringens (28%), calf thymus (42%) and Micrococcus luteus (72%). The concentration of the solutions was carefully adjusted to identical UV absorption and exposed to UVB radiation.

The content of all possible CPDs and 6-4PPs was then determined by HPLC-mass spectrometry following enzymatic digestion. The dose-course formation was linear for all photoproducts (data not shown), indicative of the absence of secondary photoreactions. In particular, no Dewar valence isomer of the (6–4) photoproducts that could be produced at high UVB dose (Douki and Cadet, 2001a) was detected.

The yields of photoproduct formation calculated from these data by linear regression are listed in Tab. 2.1.

80 Chapter 2

Table 2.1. Effect of the variation in GC content on the yield of formation of bipyrimidine photoproducts in isolated DNA exposed to UVB radiation. Results are expressed in number of photoproducts per 104 normal bases per J cm−2 irradiation. Error bars represent the standard error of the linear regression slope (n=12).

DNA C. perfringens Calf thymus M. luteus

GC% 28 42 72

TT CPD 49.8 ± 1.5 27.6 ± 0.5 2.8 ± 0.1

TT 6-4PP 2.7 ± 0.1 1.6 ± 0.1 0.2 ± 0.1

TC CPD 13.8 ± 0.4 15.0 ± 0.3 10.0 ± 0.2

TC 6-4PP 8.3 ± 0.2 11.3 ± 0.2 8.8 ± 0.2

CT CPD 7.4 ± 0.3 7.4 ± 0.1 3.7 ± 0.1

CC CPD 2.2 ± 0.1 4.7 ± 0.2 6.5 ± 0.1

Total 84.2 ± 8.9 67.6 ± 4.6 32.0 ± 2.1

In agreement with our previous observations (Douki and Cadet, 2001; Mouret et al.,

2006; Douki et al., 2003), CC and CT 6-4PPs were produced in much lower yield than the other photoproducts and were below the detection limit of the assay in the dose range applied. A first observation is the significant decrease in the overall yield of damage with increasing GC content (Fig. 2.3).

81 Chapter 2

Figure 2.3. Dose-course formation of the overall photoproducts within isolated DNA exposed to UVB radiation. The amount of all photoproducts was determined for each sample and added. Experiments were performed in triplicate. Results represent the average ± standard deviation.

This result mostly reflects the strong decrease in the yield of TT photoproducts by more than an order of magnitude (ratio: 17.8) when the GC content increased from

28 to 72%. This value correlates with that of the ratio between the frequencies of TT sites in the two genomes that is 10.3. Formation of CC CPD followed an opposite trend. A ratio of 3 was found between the yield of damage at 72 and 28% GC, identical to the ratio between the frequencies of the CC dinucleotide in the two genomes. The frequency of TC and CT photoproducts remained almost unaffected when comparing DNA from C. perfringens and M. luteus, in good agreement with the ratio between the dinucleotide frequencies in the two species (0.96 and 0.92 for

TC and CT, respectively).

82 Chapter 2

The bulk of these quantitative observations show that, on the average, the photoreactivity of TT, TC, CT and CC dinucleotides does not significantly vary with the genome composition in base pairs. Indeed, the amount of photoproducts arising from a given bipyrimidine dinucleotide is proportional to its frequency in the considered DNA. This observation indirectly provides some insight into basic features of DNA photochemistry. For instance, by analogy with the efficient charge transfer processes leading to predominant guanine oxidation as the result of its low ionization potential (Schuster and Landman, 2004), one could expect that energy transfer toward one of the four bipyrimidine dinucleotides with a lower excited state energy should be promoted and result in the major formation of the corresponding photoproducts independently of the base pair composition. Our results do not support this hypothesis for the excited states of dinucleotides. Anyhow, it should be emphasized that our observations do not mean that energy transfer does not occur at all. The numerous studies devoted to the quantification of photoproducts at the nucleotide level within specific sequences have revealed the existence of photoproducts hotspots that likely reflect sinks for the excitation energy (Drouin and

Therrien, 1997; Sage et al., 1992; Mitchell et al., 1992; You et al., 2003). Moreover, recent theoretical calculations showed that excited states in DNA are spread over several bases (Markovitsi et al., 2005; Emanuele et al., 2005a and 2005b). Our results show that, on the average, TT, TC, CT and CC are equally targeted by energy transfer processes.

An alternative or additional explanation to this relative independency of the average photoreactivity of TT, TC, CT and CC with the genome composition could be a limited energy redistribution resulting from the very short lifetime of the excited

83 Chapter 2

states involved, in agreement with the recent time-resolved observation of a reaction time in the picosecond range for the formation of TT CPD (Schreier et al., 2007). It may be added that such a short lifetime may reflect the involvement of singlet rather than triplet excited states that are longer lived. In agreement with this proposal, the triplet excited state of thymine has been shown to exhibit a lower energy than that of cytosine (Wood and Redmond, 1996; Bosca et al., 2006) as illustrated by the predominant formation of TT CPDs upon photosensitized triplet energy transfer

(Ben-Ishai et al., 1968; Lamola, 1970). Our results show that this trend is obviously not true in UVB irradiated DNA since TT CPDs is a minor lesion at high GC content. Interestingly, formation of CPDs through a singlet excited state would explain why formation of both CPDs and 6-4PPs, lesions known to arise from a singlet excited state, exhibited the same dependence toward light intensity in UV- laser irradiation studies (Douki et al., 2001).

III.3 The proportion of cytosine-containing photoproducts correlates with that of the cytosine containing dinucleotides in irradiated isolated bacterial

DNA

These first quantitative results clearly show that the GC content of a genome strongly impacts the distribution of UVB-induced photoproducts. In order to refine this first analysis, we examined the effects of the precise dinucleotide frequency on the photochemistry of DNA from different sequenced bacteria. For this purpose, we used, in addition to the three types of samples discussed above, DNA extracted from three microorganisms, namely Sphingopyxis alaskensis, Photobacterium angustum and Salmonella typhimurium. Unfortunately, the purity of the bacterial DNA samples was poorer than that of commercially available DNA as the result of the presence of

84 Chapter 2

insoluble contaminants and significant amounts of RNA (30–50%). It was thus not possible to have a precise estimation of the amount of UV radiation actually absorbed by the DNA. Although this drawback prevented the determination of accurate absolute yields of formation, reliable relative photoproducts distributions could still be obtained. We thus calculated, for each dinucleotide and in each genome, the ratio between the proportions of its photoproducts amongst bipyrimidine dimers and its frequency amongst bipyrimidine sites. This normalization procedure provides a way to compare the photoreactivity of dinucleotide independently of its abundance in a genome (Tab. 2.2).

Table 2.2. Ratio between the relative frequency of photoproducts and the frequency of bipyrimidine sites in genomes of different GC content.

% GC TT TC CT CC

C. perfringens 28 1.18 ± 0.02 1.40 ± 0.04 0.39 ± 0.04 0.41 ± 0.08

P. angustum 39.6 1.03 ± 0.07 1.83 ± 0.01 0.61 ± 0.08 0.27 ± 0.07

Calf thymus 42 1.10 ± 0.03 1.66 ± 0.03 0.52 ± 0.04 0.43 ± 0.05

S. typhimurium 52 1.14 ± 0.04 1.73 ± 0.06 0.78 ± 0.06 0.34 ± 0.05

S. alaskensis 65.5 1.29 ± 0.04 1.73 ± 0.06 0.77 ± 0.05 0.40 ± 0.04

M. luteus 72 1.38 ± 0.07 2.06 ± 0.09 0.56 ± 0.07 0.48 ± 0.07

Average ± standard deviation 1.19 ± 0.13 1.73 ± 0.21 0.61 ± 0.15 0.39 ± 0.07

Only limited variation in the value of the ratio for a given dinucleotide was observed, although the GC content of the studied genomes varied between 28 and 72%. The calculated ratios also reflect the intrinsic photoreactivity of the four bipyrimidine dinucleotides. They were found to be in the following decreasing order: TC > TT >

85 Chapter 2

CT > CC. This order was not only true for the average value but also for each of the studied genomes. The poor influence of the genome composition on the intrinsic reactivity of the four bipyrimidine dinucleotides results in a distribution of photoproducts depending almost exclusively of the frequency of the different dinucleotides in the DNA. Consequently, a close correlation is observed between the proportion of cytosine-containing photoproducts and that of cytosine-containing bipyrimidine dinucleotides (Fig. 2.4).

Figure 2.4. Correlation between the proportion of cytosine-containing bipyrimidine dinucleotides and the proportion of cytosine-containing photoproducts in isolated DNA exposed to UVB. DNAs were from the following origins in increasing order of proportion of C-containing dinucleotides: Clostridium perfringens, Photobacterium angustum, calf thymus, Salmonella typhimurium, Sphingopyxis alaskensis and Micrococcus luteus (n=6).

Another interesting observation on the relative yields of photoproducts in these 6 genomes deals with (6–4) photoproducts. For both TT and TC dinucleotides, the ratio between the yields of CPDs and 6-4PPs exhibited some variations from one

86 Chapter 2

DNA to the other. However, these fluctuations could not be correlated to the dinucleotide frequency (Tab. 2.3). The median value for the CPD to 6-4PP ratio were

16.0 (10.2–18.1) and 1.3 (1.1–1.7) for TT and TC, respectively. Interestingly, the proportion of TC 6-4PP has been reported to be one of the most sensitive features of

DNA photochemistry with respect to irradiation conditions in dry DNA (Douki and

Cadet, 2003), at high temperature (Douki, 2006a) and low ionic strength (Douki,

2006b). The lack of correlation between the GC content and the CPDs/6-4PPs ratio even at TC sites further illustrates the poor effect of the overall genome composition on the reactivity of individual dinucleotides. Therefore, a first major conclusion of the present work is that, after initial excitation of the DNA bases by UVB photons, the basic photochemical processes of each bipyrimidine dinucleotide are roughly identical in all genomes. The relative proportion of the final damage depends primarily on the frequency of the different dinucleotides. Obviously, our observations do not rule out a role for local sequence effects illustrated by the existence of hotspots for UV-induced dimers in specific genes revealed by sequencing techniques (Drouing and Therrien, 1997; Sage et al., 1992; Mitchell et al., 1992; You et al., 2000).

87 Chapter 2

Table 2.3. Ratio between the yields of CPD and 6-4PP at TT and TC sites in DNA exhibiting various GC content. Yields were calculated by linear regression from the dose course study.

% GC TT CPD/6-4PP TC CPD/6-4PP

C. perfringens 28 18.1 ± 0.8 1.7 ± 0.2

P. angustum 39.6 10.2 ± 0.6 1.2 ± 0.1

Calf thymus 42 17.7 ± 0.6 1.3 ± 0.1

S. typhimurium 52 14.0 ± 1.2 1.2 ± 0.1

S. alaskensis 65.5 16.0 ± 0.7 1.6 ± 0.1

M. luteus 72 15.8 ± 1.4 1.1 ± 0.1

III.4 Favored formation of cytosine-containing photoproducts at high GC content in vivo

The completely in vitro experiments reported above showed a strong impact of GC content on the distribution of photoproducts that directly reflects the frequency of each of the four bipyrimidine dinucleotides. However, DNA in pure aqueous solution does not reflect all the features of nucleic acids inside a cell. Therefore, we wanted to test whether our findings were still valid within living cells, beginning with two bacteria previously studied in our group for their response to UVB radiation. For this purpose we quantified the level of UVB-induced bipyrimidine photoproducts in two marine bacteria: P. angustum and S. alaskensis (Joux et al., 1999). These species exhibit a GC content of 39.6 and 65.5%, respectively. Bacteria were exposed to UVB in suspension in artificial seawater and DNA was extracted. Dimeric photoproducts were then quantified (Fig. 2.5).

88 Chapter 2

TT CPD TT 6-4PP TC CPD 80 TC 6-4PP CT CPD CC CPD

60 bases 6

40 lesion 10 per

20

0 S. alask ensis P. angustum

Figure 2.5. Level of bipyrimidine photoproducts within the DNA of S. alaskensis and P. angustum exposed to 20 min (dose=0.084 J cm−2) of UVB radiation (n=3). Cells were irradiated, DNA was extracted and the level of photoproducts was quantified by HPLC-MS/MS.

TC CPD was the main photoproduct within DNA extracted from irradiated S. alaskensis while TT CPD exhibited the highest yield in P. angustum. The frequency of CC CPD was two times higher in S. alaskensis than in P. angustum. It should be mentioned that the present determination of the level of DNA damage contrasts with that previously reported using an immunological approach (Joux et al., 1999).

However, the antibodies used in that study were raised against TT CPD and are likely to be specific for this lesion (Mitchell, 1996). Accordingly, HPLC-MS/MS measurements show a higher yield of TT CPD in P. angustum than S. alaskensis.

Interestingly, the proportion of damage at each of the four bipyrimidine dinucleotides is the same in vivo and ex vivo for each bacterium (Fig. 2.6). Differences between the distribution in isolated and cellular DNA was only observed for TC photoproducts where the ratio between the yields of CPD and 6-4PP was similar in both species of

89 Chapter 2

bacteria but higher than within isolated DNA. This likely reflects structural features as previously reported within naked DNA (Douki, 2006a).

S. alaskensis

P. angustum

Figure 2.6. Proportion of the photoproducts of the four bipyrimidine dinucleotides within DNA of two bacteria. DNA was exposed to UVB either in cells or following extraction. Bacteria species were either S. alaskensis (upper half) or P. angustum (lower half) (n=4).

It may be also emphasized that, although the GC content of DNA in S. alaskensis is higher than in P. angustum, the total level of lesions is slightly higher in the former bacteria than in the latter. This is in opposite trend with that observed in isolated

DNA and suggests the presence of UV absorbing compounds in P. angustum.

However, all of these results confirm that higher GC content is associated with increased formation of cytosine-containing photoproducts, like observed within isolated DNA. Interestingly, a larger yield of TC than TT photoproducts has been reported upon UVC- and UVB-irradiation of Deinococcus radiodurans cells that exhibit a high 67% GC content (Pogoda de la Vega et al., 2005), further confirming our findings. Obviously, a larger number of bacteria should be studied to definitively

90 Chapter 2

establish the link between % GC and DNA photoproduct distribution and confirm our positive preliminary observations.

III.5 Increased UV-induced mutation rate in bacteria with high GC content?

Our results show that GC content significantly affects the distribution of UVB- induced DNA damage both in solution and in vivo. The main trend is an increase in the contribution of cytosine-containing photoproducts for genomes containing high

GC content. A major consequence of this result is a potential variation in the photo- induced mutation rate among different bacterial species. Indeed, it is well established that, although a blocking lesion like other cyclobutane dimers, TT CPD is poorly mutagenic, as shown by using oligonucleotides bearing a single photoproduct that were replicated by either purified polymerases or upon transfection in cells (Taylor,

1994; Lawrence et al., 1993). Another support of the poor mutagenicity of TT CPDs is the almost complete lack of mutation at TT sites in UVB mutation spectra in mammalian cells (Brash et al., 1991; Ziegler et al., 1993; Sage et al., 1996). The mutagenic properties of cytosine-containing CPDs are completely different. Indeed, the induction of TC to TT and CC to TT mutations are considered as the hallmarks of

UVB radiation (Brash et al., 1991; Ziegler et al., 1993; Sage et al., 1996). A likely explanation for the high mutagenicity at cytosine-containing sites is the ability of cytosine moieties of CPDs to undergo deamination (Freeman et al., 1965; Douki and

Cadet, 1992; Lemaire and Ruzsicska, 1993). The result of this spontaneous reaction is a uracil-containing CPD that codes for adenine like a thymine (Jiang and Taylor,

1993; Peng and Shaw, 1996). Therefore, it may be expected that microorganisms with high GC content are more prone to UV-induced mutations than those exhibiting

91 Chapter 2

a higher frequency of TT dinucleotides in their genome. A link between the genome composition of prokaryotes and the amount of UV exposure experienced in their habitat has recently been ruled out (Palmeira et al., 2006). However, this last work was based on early values of the yield of DNA damage that did not include 6-4PPs and were from a single source of DNA (i.e., Escherichia coli). Our present observations could refine this approach and validate our prediction of a higher mutability among high GC containing genomes. Interestingly, prokaryotes with reduced genome, including abundant marine oligotrophic bacteria like “Ca.

Pelagibacter ubique” HTCC1062 (Giovannoni et al., 2005a) also have low GC (Xia et al., 2003). These bacteria may be consequently less affected by UV-induced mutation.

92 Chapter 2

IV. Conclusion

The present work unambiguously shows that the distribution of bipyrimidine photoproducts within UVB-irradiated DNA is greatly affected by the content in GC base pairs of the genome. Indeed, a very nice correlation was found between the proportion of C-containing photoproducts and that of C-containing dinucleotides both in vitro and in vivo. In terms of basic photochemistry of DNA, these results mean that the photoreactivity of TT, TC, CT and CC dinucleotides is largely independent of the genome base composition. From a biological perspective, our observations emphasize the need to take genome composition into account when comparing the distribution and even the yield of damage between different species.

These results raise some novel questions on the relationship between genome composition and UV-mutagenesis. It would be interesting to determine how variation in global GC content of a genome is translated in number of bipyrimidine dinucleotides in actual genes. These data could be correlated with photoproducts distribution at the sequence level in a large series of species exhibiting different GC contents. Such information could provide some insight on UVB mutagenesis and its role in the evolution of bacteria.

93 CHAPTER 3

Remarkable resistance to UVB of the marine bacterium Photobacterium angustum explained by an

unexpected role of photolyase Chapter 3

CHAPTER 3

Remarkable resistance to UVB of the marine bacterium Photobacterium angustum explained by an unexpected role of photolyase

Abstract

DNA damage and cell survival was assessed in the marine bacteria, Photobacterium angustum (GC% = 39.6) and Sphingopyxis alaskensis (GC% = 65.5) following UVB irradiation and recovery in the presence or absence of visible light. The extent of bipyrimidine photoproduct formation was analyzed by HPLC-MS/MS. S. alaskensis was chosen as a reference species since it was previously shown to be photoresistant.

Interestingly, P. angustum exhibited an even higher level of survival to UVB irradiation than S. alaskensis. This higher photoresistance was associated with a decrease in the rate of formation of cyclobutane pyrimidine dimers (CPDs) at high

UVB doses. Despite different distributions in UVB-induced lesions, the survival difference between the two marine bacteria could not be accounted for by qualitative differences in either photoreactivation or the rate of nucleotide excision repair of the photoproducts arising from the different bipyrimidine doublets (TT, CT, TC and

CC). Dark repair was found to be much more efficient for P. angustum than S. alaskensis but the corresponding rate of photoproduct removal was lower than that observed at high UVB doses. We propose that the increased resistance of P. angustum under high UVB doses results from a UVB-induction of CPD photolyase(s) that may directly repair DNA damage and/or act indirectly by enhancing the rate of nucleotide excision repair.

95 Chapter 3

I. Introduction

Heterotrophic marine bacteria are of great importance in global nutrient cycling especially at the sea surface where they process about half of primary production

(Azam and Malfatti, 2007). There is now strong evidence that UVR could affect bacterial activity in a range of aquatic ecosystems (Sommaruga et al., 1997), and that solar ultraviolet radiation (UVR, 280-400 nm) may penetrate deep into the clear oligotrophic marine waters influencing a large part of the euphotic layer (Tedetti and

Sempéré, 2006). UVR could be particularly deleterious for marine bacteria (Jeffrey et al., 1996), which have simple haploid genomes with little or no functional redundancy (Giovannoni et al., 2005a) and no efficient protective pigmentation

(Garcia-Pichel, 1994). UVB (280-320 nm) and UVA (320-400 nm) cause distinct but overlapping damage. UVB causes mainly direct effects on DNA by inducing dimerization of pyrimidine bases, leading to the formation of cyclobutane pyrimidine dimers (CPDs), pyrimidine (6-4) pyrimidone photoproducts (6-4PPs) and the related

Dewar valence isomers. These photoproducts may block DNA replication and transcription or induce mutations. UVA, and to a lesser extent photosynthetic active radiation (PAR, 400-700 nm), causes indirect damage by generating reactive oxygen species (ROS) that damage DNA, proteins and lipids. However, at the same time the

UVA as well as PAR are involved in the photoenzymatic repair (PER) of CPDs and

6-4PPs (Häder and Sinha, 2005). During PER, photolyases specifically bind either to

CPDs (CPD photolyase) or 6-4PPs (6-4 photolyase) and reverse the damage using the energy of the photoreactivating light (300-500 nm). The presence of 6-4 photolyase has, as yet, not been characterized in bacteria (Sancar and Sancar, 1987;

Sancar, 2008). CPD photolyases are monomeric enzymes with two non-covalently bound chromophore cofactors (Sancar and Sancar, 1987; Sancar, 2008). One of the

96 Chapter 3

cofactors is always FAD, and the second is either methenyltetrahydrofolate (MTHF) or 8-hydroxy-7,8-didemethyl-5-deazariboflavin (8-HDF) in a limited number of species (Eker et al., 1990; Kiener et al., 1989). Although only FAD of photolyase is required for its catalytic activity, the second cofactor significantly accelerates reaction rate under low-light conditions.

PER has been reported to have varying levels of importance for the recovery of marine bacteria after sunlight exposure (Kaiser and Herndl, 1997; Pausz and Herndl,

2002). Moreover all the studies describing the impact of diel cycles on the induction of DNA damage in marine bacteria indicate that, even when PER occurs, it only plays a limited role (Buma et al., 2003). However, the efficiency of PER for different aquatic bacterial species has been reported to be highly variable (Joux et al., 1999;

Arrieta et al., 2000; Zenoff et al., 2006). Nucleotide excision repair (NER) is a light- independent repair pathway that involves the recognition, removal and resynthesis of damaged stretches of DNA. During NER, the multienzyme complex UvrABC endonuclease removes a wide range of DNA lesions including CPDs, 6–4PPs and single-strand DNA breaks. The NER pathway appears to maintain a constant basal level of activity under most growth conditions, while also being up-regulated by the

DNA damage-inducible SOS response (Friedberg et al., 1995)

In order to monitor the response of organisms to environmental parameters it is important to identify biomarkers diagnostic of cellular damage or physiological stress. Because TT CPDs were found to be the most frequent DNA photoproduct induced in mammalian cells (Mouret et al., 2006), it has generally been used as reference to quantify the DNA damage induced by UVB in bacteria using

97 Chapter 3

radioimmunoassay or enzyme-linked immunosorbent assay (ELISA) techniques in the laboratory (Joux et al., 1999) and in the field (Buma et al., 2003; Meador et al.,

2009). However, TT CPDs are not the sole UVB-induced DNA lesion since 12 different bipyrimidine photoproducts (CPD, 6-4PPs and Dewar valence isomer of the four bipyrimidine doublets) are potentially inducible (Douki et al., 2000). It was recently demonstrated that the GC content of bacterial DNA (ranging from 25% to

75%) (Mooers and Holmes, 2000) can greatly influence the distribution of pyrimidine dimers induced by UVB, with the formation of cytosine-containing photoproducts increasing with GC content (Matallana-Surget et al., 2008). This illustrates that monitoring TT CPDs may not be optimal for assessing UVB induced

DNA damage in bacteria.

In the present study we determined whether bipyrimidine dimer content could affect

DNA repair and survival by comparing the responses of two marine bacteria models with different trophic life-styles (Schut et al., 1997) and genomic GC content.

Photobacterium (formerly Vibrio) angustum S14 is a marine copiotrophic bacterium adapted to grow rapidly in relatively rich environments (e.g., estuaries, associated with particulate matter) and is able to form resting-stage cells when nutrients become depleted (Srinivasan and Kjelleberg, 1998). P. angustum possesses a genome size of

5.1 Mbp with 4558 genes and a GC content of 39.6% (Information (Genbank): http:// www.ncbi.nlm.nih.gov/. Access Number: AAOJ00000000). In contrast,

Sphingopyxis alaskensis RB2256 is a marine ultramicrobacterium (cell volume

<0.1)m3) that maintains a relatively slow growth rate across a large range of organic matter concentrations (Schut et al., 1993 and 1997; Eguchi et al., 1996) and possesses a smaller genome size (3.2 Mbp) than P. angustum with 3196 genes and a

GC content of 65.5% (Cavicchioli et al., 2003). S. alaskensis has many physiological 98 Chapter 3

and genetic properties that contrast with traditionally studied bacterial strains (e.g., E. coli, Vibrio/Photobacterium spp.) that reflect its adaptation to life in oligotrophic waters (Fegatella et al., 1998; Klappenbach et al., 2000). For instance, this bacterium possesses only one rRNA operon against 8 to 11 in Vibrio spp (Fegatella et al., 1998) andexhibits an inherently high level of resistance to different stresses (H2O2, temperature, ethanol) during logarithmic and stationary phase growth (Eguchi et al.,

1996). S. alaskensis was also used as a reference because it is known to be more photoresistant than a range of other marine bacteria (Joux et al., 1999). We studied the extent of bipyrimidine photoproduct formation during both UVB treatment and light/dark repair, by using high performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS), which provided quantitative data about individual photoproducts within test samples (Douki et al., 2000; Douki and Cadet, 2001). We also assessed the inherent photoresistance of P. angustum after exposure to high doses of UVB and found that it had a very high level of resistance that could be correlated with its ability to repair efficiently their DNA damage during UVB exposure.

99 Chapter 3

II. Materials and methods

II.1 Bacterial strains, media and culturing conditions

Photobacterium angustum S14 (UNSW, Sydney, Australia) and Sphingopyxis alaskensis RB2256 (DSMZ 13593T) were grown aerobically in Artificial Sea Water

(ASW) supplemented with 3 mM of D-glucose, vitamins and trace elements (Eguchi et al., 1996) in a rotary shaker (130 rpm) at 24°C. After two precultures, the growth was monitored by optical density (OD) measurements at 433 nm. Stationary grown cells (i.e., after 12 h for P. angustum and 60 h for S. alaskensis) from 100 ml cultures were harvested by centrifugation at 8 000 x g for 8 min at 24°C and then washed twice with sterile ASW (pH=6.5). Cells were resuspended in sterile ASW to obtain an OD433nm=0.1 and placed into 250-ml cylindrical quartz flasks with flat top.

II.2 UVB exposure and repair conditions

Quartz flasks were exposed to the light emitted by UVB lamps (UVB 313, Q-Panel) filtered through acetate cellulose film to remove residual UVC (Fig. 3.1). The mean fluence rate of UVB measured with a UV/visible spectroradiometer (RAMSES,

TriOS) was 0.9 W m-2. Cells were maintained under agitation during exposure with a continuous magnetic stirring (~220 rpm) and at 24°C by partial submersion of the quartz flasks in a water bath connected to a cryothermostat. Cells were sampled for survival and DNA damage analysis at different times of exposure up to 160 min

(UVB dose=8.64 kJ m-2). For repair experiments, culture flasks were first exposed to

UVB for 20 min (1.08 kJ m-2) and then placed under photoreactivating light (PRL) or in the dark. Repair experiments were conducted at room temperature (18-20°C) with a continuous magnetic stirring (~220 rpm) without addition of carbon source. Cells were maintained in dark or exposed under two CoolWhite lamps (F20/T12/CW,

100 Chapter 3

General Electric) with a fluence of 12.9 W m-2 (Fig. 3.1). After 3, 6 and 18 h, cells were sampled to determine survival and DNA damage.

Figure 3.1. The emission spectra for the lamps used to induce DNA damage (UVB) and to photoreactivate damage after UVB exposure (PER).

II.3 Survival and DNA damage analysis

Counts of viable bacteria were estimated from the number of colony forming units on

VNSS media (Eguchi et al., 1996) by plating in triplicate dilution series carried out in sterile ASW in triplicate. For the DNA damage analysis, genomic DNA was isolated in duplicate from 60 ml of sample using the Wizard® genomic DNA purification Kit from Promega according to manufacturer’s instructions. The amount of the different forms of cyclobutane pyrimidine dimers (CPDs; i.e., Thy< >Thy,

Thy< >Cyt, Cyt< >Thy and Cyt< >Cyt), 6-4 photoproducts (6-4PPs; i.e., TT 6-4, TC

6-4, CT 6-4 and CC 6-4) and Dewar valence isomers was determined by high performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS)

101 Chapter 3

(Douki and Cadet, 2001). Briefly, DNA was enzymatically digested by a mixture of endonucleases, exonucleases and phosphatase. Bipyrimidine photoproducts were recovered under the form of modified dinucleoside monophosphates. The mixture was separated on an HPLC column and the photoproducts quantified by online mass spectrometry analysis in the multiple reaction monitoring mode. Normal nucleosides were simultaneously quantified on a UV spectrophotometer. The data was expressed as the number of lesions per 106 bases.

102 Chapter 3

III. Results

III.1 Survival and DNA photoproducts following UVB radiation

Distinct patterns of response were observed for cell survival and total UVB-induced photolesions for the two marine bacteria. Whereas S. alaskensis showed an exponential decrease in viability, P. angustum was characterised by a distinctive plateauing of survival at ~10% for UVB doses ranging from 2 to 8.64 kJ m-2 (40 to

160 min) (Fig. 3.2). While S. alaskensis was more resistant than P. angustum to low

UVB doses up to 3 kJ m-2, P. angustum was more resistant to high UVB doses. The same trend was observed for the frequency of photolesions (Fig. 3.2). For both marine bacteria, a total of 6 different photoproducts were detected by HPLC-MS/MS during UVB exposure (Fig. 3.3). CT 6-4, CC 6-4 and Dewar valence isomers were not detected. The increase in the level of all photoproducts was linear with respect to the dose in S. alaskensis. In contrast, in P. angustum the formation of total DNA lesions reached a plateau of ~220 lesions per 106 bases. Although the formation of 6-

4PPs at TT and TC positions was linear with respect to the dose in P. angustum, a statistically significant decrease of the amount of TT and TC CPDs was observed for high UVB doses between 2.16 to 8.64 kJ m-2 (Student t-test, p < 0.05) (Fig. 3.3). As previously reported (Matallana-Surget et al., 2008), the two marine bacteria showed a distinct distribution of photoproducts with a predominance of TC CPD for S. alaskensis contrasting with an equal amount of TT and TC CPD dimers in P. angustum. These results reflect the difference in GC content of the two genomes.

103 Chapter 3

a 100 1000 oa ein per 10 lesions Total 800 10

600 1 400 6 Viability (%) bases 0,1 200

0,01 0 0246810 b 100 1000 oa ein per 10 lesions Total 800 10

600 1 400 6 Viability (%) bases 0,1 200

0,01 0 0246810

UVB dose (kJ m-2)

Figure 3.2. Changes in viability and quantity of total DNA lesions during UVB irradiation in P. angustum (a) and S. alaskensis (b). Survival (); total DNA lesions (). The data is expressed as the mean (± standard deviations) of three independent experiments.

104 Chapter 3

P. angustum S. alaskensis

Figure 3.3. Induction of the different CPD (upper figures) and 6-4 photoproducts (lower figures) in P. angustum and S. alaskensis. The data is expressed as the mean (± standard deviations) of three independent experiments.

III.2 Survival and DNA repair following photoreactivating light or darkness

Repair experiments were carried out after 1.08 kJ m-2 of exposure to UVB (20 min) because i) this dose is in the linear part of the formation of photoproducts for both bacteria and ii) the level of photoproducts was similar (140 ± 7 and 124 ± 26 total lesions per 106 bases in S. alaskensis and P. angustum, respectively). Our rationale

105 Chapter 3

was based in part on the observation in mammalian cells that the rate of repair depends on the initial amount of photoproducts (Greinert et al., 2000; Courdavault et al., 2004). Under these conditions, the survival was 75 ± 4% in S. alaskensis and 13

± 3% in P. angustum (Fig. 3.4).

These cells were used to determine the kinetics of light and dark repair in the absence of any carbon source. During incubations in darkness, P. angustum cells were able to repair up to 75% of the photoproducts after 3 h and achieved full repair after 6 h

(Fig. 3.4). However, 100% recovery of viability was not achieved even after 18 h

(maximumof 80%) (Fig. 3.4). Under PRL conditions, viability was fully restored and photoproducts completely repaired after only 3 h (Fig. 3.4). The percentage of viable counts in fact exceeded 100% (~150%), which is likely to reflect the capacity of cells to undergo final rounds of cell division (Abboudi et al., 2008).

For S. alaskensis, a greater disparity was observed between the kinetics of photoproducts repair during PRL and dark incubations. Whereas a maximum of

~85% of repair was reached after 6 h under PRL, this percentage only slowly increased during incubations in darkness to reach ~50% at 18 h (Fig. 3.4). The low rate of removal of DNA lesions in S. alaskensis (compared with P. angustum) during dark incubations, was accompanied by a small loss of viability from 80% to 50% within 18 h (Fig. 3.4).

106 Chapter 3

P. angustum S. alaskensis 200 200

160 160

120 120

80 80 Viability (%)

40 40

0 0 0 5 10 15 20 0 5 10 15 20 125 125

100 100

75 75

50 50 Photoproducts repair (%) 25 25

0 0 0 5 10 15 20 0 5 10 15 20 Time (hr) Time (hr)

Figure 3.4. Changes in viability and repair of DNA lesions during photoreactivation () and liquid holding () in P. angustum and S. alaskensis. The data are expressed as percentages of the initial counts before 20 min UVB irradiation and are means (± standard deviations) based on the data from two independent experiments.

107 Chapter 3

After 3 h of incubation in darkness, the removal of CPDs was 5 times more efficient in P. angustum than S. alaskensis (Fig. 3.5). Repair under PRL was also marginally higher in P. angustum. However, the extent to which the rates of repair can be quantified is complicated by the fact that repair was essentially complete in P. angustum after 3 h recovery.

It is noteworthy that the PRL repair was essentially as efficient towards CPDs as 6-

4PPs (Fig. 3.5). In contrast, during incubations in the dark, 6-4PPs were repaired more efficiently than CPDs for S. alaskensis; 80% of 6-4PPs were repaired after 3 h in comparison to ~15% for CPDs (Fig. 5).Moreover, for both bacteria, repair of individual lesionswas not affected by the nature of the dinucleotide involved (TT,

TC, CT or CC), contrasting with a recent study performed on human cells (Mouret et al., 2008).

PRL Darkness PRL Darkness

P. angustum S. alaskensis

Figure 3.5. Distribution of bypyrimidine photoproducts in P. angustum and S. alaskensis after 3 h of PRL or darkness. The data is expressed as the mean (± standard deviations) of two independent experiments.

108 Chapter 3

IV. Discussion

A range of responses to UVB resistance have been documented for marine bacteria

(Joux et al., 1999; Agogué et al., 2005). In previous work, S. alaskensis was shown to be resistant to UVB in comparison to other marine bacteria, including Vibrio natriegens (Joux et al., 1999). We observed here that S. alaskensis was more sensitive to high UVB doses than P. angustum (Fig. 3.2). This was surprising as

Vibrionaceae and Photobacteriaceae families are closely related (Thompson et al.,

2004) and it may be expected that P. angustum would be as sensitive as V. natriegens. To reconcile these two studies that used different culture medium and

UVB exposure conditions, we studied the survival of V. natriegens with our new protocol and confirmed its high sensitivity to UVB compared to both S. alaskensis and P. angustum (see Appendix B). Contrasting levels of resistance have been observed within the Pseudomonas and Acinetobacter genera (Zenoff et al., 2006), and our present observations provide good reason to explore the resistance of the more than 63 species in the Vibrionaceae and Photobacteriaceae families (Thompson et al., 2004).

Consistent with a previous study (Matallana-Surget et al., 2008), we showed that GC content of bacteria greatly influences the type of photoproducts induced by UVB, with an increased proportion of cytosine-containing photoproducts for genomes with high GC content. In the present study, the formation of DNA photoproducts was investigated over a larger range of UVB doses than in previous work. By doing so, we observed an unexpected decreased accumulation of TT and TC CPDs in P. angustum following high doses of UVB irradiation. While the CPD removal must be linked to photoprotection or repair mechanisms, it cannot be explained by direct

109 Chapter 3

photoreversion of CPDs, a well established process that occurs in the UVC and not the UVB range (Fisher and Johns, 1976; Rontó et al., 2002).

The correlation observed between the survival of P. angustum and its ability to counteract the increase in the formation of CPDs within DNA damage, could be explained by the production of UVB absorbing compounds in the cell reducing

UVB-induced DNA dimer formation. Consistent with this basic DNA photochemistry, we previously showed that more photoproducts were produced in

DNA purified from P. angustum compared with DNA from S. alaskensis (Matallana-

Surget et al., 2008). However, the opposite trend was observed for intact cells.

Photoprotection is a mechanism that minimizes the photon absorption by adjacent bipyrimidines and should generate an attenuated but constant rate of formation of both CPDs and 6-4PPs. As this was not observed for CPDs, our present results rule out a major role for photoprotection by UV-absorbing compounds. Repair is thus more likely to explain the lack of dose dependence in cell survival and the level of

CPDs in P. angustum.

If the linear increase in CPD formation that was observed before 2.16 kJ m-2 is extrapolated to higher doses, it is possible to calculate the expected photoproduct frequency at high dose. By subtraction of this value from the experimental data, after exposure to the highest UVB dose P. angustum is predicted to be able to efficiently repair as much as 132 and 90 lesions per 106 bases per hour for both TT and TC

CPDs, respectively (Tab. 3.1). Strikingly, the repair rate of CPDs determined after exposure at high UV doses (“UVB repair”) is higher than that of the dark repair mechanism (i.e. ~15 lesions per 106 bases per hour after incubation in the dark for

110 Chapter 3

both TT and TC CPDs) (Tab. 3.1). Consequently, some specific and efficient repair mechanisms seem to be present under UVB exposure and not during dark repair. The rates of “UVB repair” cannot be compared to those obtained during photoreactivation since, for the first point of the kinetics (3 h), P. angustum already repaired 100% of the lesions. It is noteworthy that P. angustum has previously been found to accumulate very low levels of CPDs when grown under simulated solar radiation, even though the UVB reduced the growth rate and induced the formation of filamentous cells (Abboudi et al., 2008).

As a result of these above findings we reasoned that photolyase may function to facilitate DNA repair during UVB exposure. PRL that is generated by the UVB lamps may be sufficient for photolyase I of P. angustum to break down the cyclobutyl ring of the pyrimidine dimers during UVB exposure, given that one dimer splits for every blue-light photon absorbed (Payne and Sancar, 1990). Consistent with a role for photolyase, by carefully examining the UVB spectrum, residual peaks at 360 nm and 440 nm were identified; specific wavelengths that could be absorbed by either MTHF or 8-HDF, respectively (Sancar and Sancar, 1987; Sancar, 2008). P. angustum possesses three genes encoding deoxyribodipyrimidine photolyase (EC

4.1.99.3) type I enzymes, compared to only one gene for S. alaskensis. The fact that high doses show the greatest response in P. angustum may therefore be explained by

P. angustum possessing photolyase gene(s) that are more efficiently induced at high

UVB doses than the single S. alaskensis photolyase gene. In other organisms, photolyase gene expression was shown to be induced by general stressors (Sebastian and Sancar, 1991; Sebastian et al., 1990; Jang et al., 1999), and over-expression in P. angustum may explain its UVB phenotype. It is also possible that the photolyase enzymes expressed from the three genes in P. angustum have different levels of

111 Chapter 3

efficiency for repairing DNA damage. It will be valuable to assess the regulation of the P. angustum gene expression, enzyme synthesis and enzyme activity (including post-translational modification) throughout the extended dose-response regime in order to assess these possibilities.

Table 3.1. Estimations of the rates of lesions repaired per 106 bases per hour in P. angustum during UVB radiation between 2 and 8.64 kJ m-2 (“UVB reapir”) and during the two first hours repair in the dark condition (“Dark repair”). The data expressed the mean (± standard deviations) of three independent experiments for the UVB repair rates and two independent experiments for the Dark repair rates.

TT TC CT CC CPD CPD CPD CPD UVB Repair 132.1 ± 1.5 90.0 ± 2.2 18.1 ± 2.8 15.1 ± 0.4 Dark Repair 15.5 ± 5.6 15.8 ± 4.6 1.9 ± 0.6 3.2 ± 1.9

In other biological systems it has been reported, both in vivo (Yamamoto et al., 1983;

Sancar and Smith, 1989) and in vitro (Sancar and Smith, 1989; Sancar et al., 1984), that in the absence of PRL, CPD photolyase can bind to CPD lesions and induce the removal of UV damage by stimulating NER. Photolyase could thus increase the poor helical deformity induced by CPDs (Husain et al., 1988; Park et al., 2002) by flipping out the bipyrimidine photoproduct and act as an enhancer of NER promoting the specific removal of CPD lesions (Sancar et al., 1984). Consistent with this, NER of P. angustum appeared to be very efficient for both 6-4PPs and CPDs, reaching after 3 h 97% and 71%, respectively, compared to 74% and 7%, respectively, for S. alaskensis. Collectively our results lead us to hypothesize that the high resistance of

112 Chapter 3

P. angustum during UVB exposure is due to effective photolyase activity that occurs under very low PRL and/or involved in the stimulation of NER.

S. alaskensis displayed efficient NER towards 6–4PPs and not CPD lesions. In E. coli, disruption of genes involved in the SOS response greatly reduced the efficiency of NER towards CPDs but had no effect on the repair of 6-4PPs (Crowley and

Hanawalt, 1998). While the efficiency of the removal of CPDs appeared to be dependent on the upregulation of key genes uvrA and uvrB by the DNA damage inducible SOS response, the basal levels of repair enzymes (UvrA, UvrB) were sufficient to repair 6-4PPs. It is therefore possible that S. alaskensis has an inefficient

SOS response. Consistent with this, two proteins in S. alaskensis that could be involved in an SOS response (RadA, RecN) were recently shown in a proteomics study to have equivalent cellular abundance in cells exposed to simulated solar irradiation and cells incubated in the dark (Matallana-Surget et al., 2008).

113 Chapter 3

V. Conclusion

In this study, we demonstrated that the marine bacterium P. angustum was very resistant to high UVB doses, and this phenotype could be explained by the efficient repair of all its CPD lesions. The presence of 3 genes coding for DNA photolyase type I enzymes in P. angustum compared to only 1 for S. alaskensis led us to suggest that the photoresistance strategy involved a capacity to utilize 3 distinct gene products, including the UVB-induced overexpression of the gene(s). We further suggested that photolyase activity not only led to the repair of DNA through a photochemical process, but may also enhance the efficiency of NER. This last process would be far more efficient in P. angustum than in S. alaskensis. The inefficient NER of CPD lesions in S. alaskensis could be linked to a reduced capacity to mount an SOS response. In rodent cells the absence of DNA damage-recognition protein (DDB2) appears to be responsible of CPDs repair-deficient system

(Hanawalt., 2001). In contrast, the reason of such deficiency has never been described in bacteria.

We also showed that, in both bacteria, neither dark repair nor photolyases are affected by the nature of the two pyrimidine bases involved in a given type of photoproduct. Moreover, for both bacteria UVB induction of DNA photoproducts correlated well with survival across the treatment regime that was employed. In contrast, the capacity to remove DNA lesions following exposure to either darkness or light did not necessarily correlate with recovery of viability. This latter effect may be due to mutations occurring in some essential genes, or damage to molecules other than DNA (e.g. lipids, proteins) as has been observed for other bacteria (Zenoff et al.,

2006)

114 CHAPTER 4

Temperature-dependent formation of bipyrimidine

photoproducts by UVB radiation in different conditions of starvation and light irradiation in the

marine bacterium Sphingopyxis alaskensis Chapter 4

CHAPTER 4

Temperature-dependent formation of bipyrimidine photoproducts by UVB radiation in different conditions of starvation and light irradiation in the marine bacterium Sphingopyxis alaskensis

Abstract

DNA damage in the form of cyclobutane pyrimidine dimers (CPDs) and (6-4) photoproducts (6-4PPs) induced by UVB radiation (280-320 nm) and simulated solar radiation (280-700 nm) in Sphingopyxis alaskensis, was investigated using HPLC-

MS/MS for cultures grown at 12°C and 24°C and for two different growth stages

(early and long stationary phases). Contrary to the commonly accepted view that

DNA damage induced by UVB radiation is temperature-independent because of its photochemical nature, we found in S. alaskensis a slower formation of total photoproducts at 12°C compared to 24°C. UVA and PAR seemed to reduce the induction level of the overall yield of formation of CPD lesions, by a process that certainly involves PER. Experiments that looked at survival rates following different conditions of UV exposure demonstrated an increased survival rates at 24°C, for the cells harvested at the beginning of stationary phase, under solar radiation. Our results from S. alaksensis also indicated that the time spent in stationary phase at 24°C, could remarkably modify the induction pattern of bipyrimidine photoproducts, with a favoured formation of TC 6-4PP for long-term starved cells compared to TC CPDs for cells entering the stationary phase, and we found that this result could be related to the level of viability upon UV irradiation.

115 Chapter 4

I. Introduction

Marine heterotrophic bacteria present at the surface of oceans are exposed to solar radiation with very different temperature conditions (i.e., from -1°C to 30°C) depending of the latitude and the seasons. Different studies have detailed the damaging effects of ultraviolet radiation (UVR) on marine bacteria (Jeffrey et al.,

2000), but little is known on the interactive effects of temperature and UVR on these microorganisms. Moreover, climate change is raising global surface temperatures and depletion of the stratospheric ozone layer is increasing levels of UVR reaching the

Earth’s surface (Madronich et al., 1998). Those environmental changes raise concerns about the response of aquatic microorganisms that could be differently altered by the interactions of higher surface temperature and increase of damaging UVR. Rae and

Vincent (1998) found that UVR exposure significantly inhibited the percentage of respiring bacteria at higher temperatures (20°C), up to 59%, whereas lower temperature treatments (10°C), did not show significant UVR effects. This suggests that with climate change and increases in temperature fields, the physiological effects of UVR damage may be greater.

UVB radiation (280-320 nm) promotes the formation of the two most frequent types of lesions, cyclobutane pyrimidine dimers (CPD), pyrimidine (6-4) pyrimidone photoproducts (6-4PP) between adjacent bipyrimidines bases on the same strand of

DNA. If the dimers are not repaired before replication they interfere with DNA synthesis and cause high rate of mutation. Cells have evolved several repair pathways to remove the DNA lesion. CPD repair can proceed along two pathways, involving either photoenzymatic repair (PER), a single-enzyme repair process driven by UVA (320-400 nm) and photosynthetically active radiation (PAR, 400-700 nm) energy, which is not energetically costly, or nucleotide excision repair (NER).

116 Chapter 4

It was previously thought that DNA damage induced by UVB radiation was a purely photochemical process and as a consequence would be relatively insensitive to temperature. However, several studies focusing on plant response, reported that the induction of CPD and 6-4PP was temperature dependent. Takeuchi et al. (1996) were the first to present that higher temperature would lead to an increased amount of both

CPDs and 6-4PPs in a greenhouse grown cucumber cotyledons. Similar results were obtained for tobacco cells (Li et al., 2002) and for Arabidopsis thaliana (Waterworth et al., 2002 ; Li et al., 2004).

In contrast, as most enzyme-catalyzed reactions are temperature dependent (Keeton et al., 1973), several studies showed that DNA repair rates could increase with increasing temperature in Antarctic zooplankton and for the crustacean Daphnia

(Malloy et al., 1997; Williamson et al., 2002; Macfayden et al., 2004). A study of a marine red alga provided evidence that the temperature optimum of PER was different for CPD and 6-4PP, suggesting that the enzymes involved in PER (i.e., photolyase) may be active at different temperature depending on the substrate specificity (Pakker et al., 2000).

Sphingopyxis alaskensis is a cold adapted marine ultramicrobacterium isolated from

Resurrection Bay in Alaska, the North Sea in Europe, and the Japanese North Pacific

(Eguchi et al., 1996; Schut et al., 1997; Eguchi et al., 2001). S. alaskensis has been well studied with respect to its UV response and DNA damage induction/repair (Joux et al., 1999; Matallana-Surget et al., 2008 and 2009a). This bacterium well-known for its ability to cope with nutrient limitation (Schut et al., 1993), presented other very interesting physiological properties as its high level of inherent stress resistance

117 Chapter 4

to a range of stress inducing agents (e.g., hydrogen peroxide, ethanol, heat and UVB radiation) (reviewed in Cavicchioli et al., 2003). S. alaskensis has been well studied with respect to its UV response and DNA damage induction/repair (Joux et al., 1999;

Matallana-Surget et al., 2008 and 2009a,b). S. alaskensis has been shown to suffer less DNA damage and have a higher level of resistance to UVB than a number of other marine bacteria (Joux et al., 1999), even if a more resistant marine bacteria has been described recently (Matallana-Surget et al., 2009a). Due to its high GC content,

UVB induces mainly cyclobutane pyrimidine dimers at TC site rather than TT in S. alaskensis (Matallana-Surget et al., 2008 and 2009a).

In this study, we reinvestigate the topic of the temperature dependence of the molecular processes by which UV damage DNA. For this purpose we exposed S. alaskensis cells grown at two different temperatures (i.e., 12°C and 24°C) and harvested in stationary phase, to observe the accumulation level of bipyrimidine photoproducts. Since it was previously shown in E. coli that the time spent in stationary phase could be crucial regarding deleterious mutational processes (Loewe et al., 2003), we chose to consider as well two different time of starvation: cells entering stationary phase (SP), and long term stationary phase (LSP). S. alaskensis was exposed to UVB radiation in presence and absence of longer-wavelength photoreactivating radiation (UVA, PAR) allowing photoenzymatic repair (PER), in order to estimate the relative importance and temperature dependence of possible

PER processes.

118 Chapter 4

II. Materials and methods

II.1 Bacterial strains, media and culturing conditions

Sphingopyxis alaskensis RB2256 (DSMZ 13593T) was grown aerobically in Artificial

Sea Water (ASW) supplemented with 3 mM of D-glucose, vitamins and trace elements (Eguchi et al., 1996) in a rotary shaker (130 rpm) at 24°C and 12°C. After two precultures, the growth was monitored by optical density (OD) measurements at

433 nm (Fig. 4.1). Maximum growth rates were 0.008 h-1 and 0.126 h-1 at 12°C and

24°C, respectively. Both stationary and late stationary grown cells (i.e., after 2.5 and

9.5 days for 24°C cultures; 22 and 29 days for 12°C, respectively) from 100 ml cultures were harvested by centrifugation at 8 000 x g for 8 min at 24°C and then washed twice with sterile ASW (pH=6.5). Cells were resuspended in sterile ASW to obtain an OD433nm = 0.1 and placed into 250-ml cylindrical quartz flasks with flat top.

1

SP LSPSP LSP 0,8 ) 0,6

0,4 OD (433 nm

0,2

0 0 3 6 9 12151821242730

Time (days)

Figure 4.1. Growth curves of S. alaskensis at 12°C () and 24°C ( ). The bars indicate the harvesting time for the stationary phase (SP) and late stationary phase (LSP).

119 Chapter 4

II.2 UVB exposure and repair conditions

Quartz flasks were either exposed to the light emitted by UVB lamps (UVB 313, Q-

Panel) or to simulated solar radiation produced by a solar-filtered (1.0 air mass) 1000-

W xenon arc lamp (Oriel Corp., Stratford, CT, USA). In both cases, quartz flasks were covered with acetate cellulose [50% transmission at 280 nm] to remove residual

UVC. The optical output irradiance under UVB lamps was 0.9 W m-2 and under solar simulator it was 0.85, 19.3 and 250W m-2 for UVB, UVA and PAR respectively, as measured with a UV/visible RAMSES spectroradiometer (TriOS, Germany). Cells were maintained under agitation during exposure with a continuous magnetic stirring

(~220 rpm) and a constant temperature of 12°C or 24°C, by partial submersion of the quartz flasks in a water bath connected to a cryothermostat.

Cells were sampled for survival analysis at different times of exposure up to 160 min

(UVB dose) and 320 min (solar simulator). Formation of bipyrimidine photoproducts was analyzed for short time exposure under UVB lamps (40 min) and solar simulator

(80 min).

II.3 Survival and DNA damage analysis

Counts of viable bacteria were estimated from the number of colony forming unit on VNSS media (Eguchi et al., 2001) by plating in triplicate dilution series carried out in sterile ASW and by incubating plates at 30°C during one week. For the

DNA damage analysis, genomic DNA was isolated in duplicate from 60 mL of sample, using the Wizard® genomic DNA purification Kit from Promega according to the protocol of the manufacturer. The amount of the different forms of cyclobutane pyrimidine dimers (CPDs; i.e., Thy< >Thy, Thy< >Cyt, Cyt< >Thy and Cyt< >Cyt), 6-4 photoproducts (6-4PPs; i.e., TT 6-4, TC 6-4, CT 6-4 and CC

120 Chapter 4

6-4) and Dewar valence isomers was determined by high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) (Douki et al., 2001).

Briefly, DNA was enzymatically digested by a mixture of endonucleases, exonucleases and phosphatase. Bipyrimidine photoproducts were recovered under the form of modified dinucleoside monophosphates. The mixture was separated on an HPLC column and the photoproducts quantified by online mass spectrometry analysis in the multiple reaction monitoring mode. Normal nucleosides were simultaneously quantified on a UV spectrophotometer. Results were expressed in number of lesions per 106 bases.

II.4 Hierarchical clustering analysis

Identification of cluster of conditions exhibiting similar patterns of DNA damage induction was performed using the complete linkage hierarchical clustering and

Euclidean distance metric, provided in PRIMER v6 software (Clarke et al., 1993).

This original comparative approach that enabled us to clarify the complex dataset in order to have a better understanding of the impact of each spectral band of solar radiation according to the growth phase and time of exposure.

121 Chapter 4

III. Results

III.1 Impact of temperature, starvation state and irradiation condition on viability

For short time exposure under both solar simulator (<80 min) and UVB lamps

(<40 min), similar loss in viablity (~60%) was observed for the different cultures of S. alaskensis used in this study (grey boxes in Fig. 4.2a and 4.2b). With respect to UVB radiation, we did not observe significant changes of tolerance with temperature or starvation condition for the whole duration of exposure, except for the last point of the kinetic (160 min), where the cells in SP appeared more resistant than LSP for cultures grown at 24°C, whereas the inverse was observed for cultures grown at 12°C (Fig. 4.2b). In contrast, upon a sunlight treatment extended time of radiation more contrasted responses were revealed with respect to the temperature and/or the starvation state considered. After 320 min of exposure, we observed that higher growth temperature (24°C) confered a better resistance to solar radiation compared to 12°C for both stationary phases (Fig.

4.2a). Furthermore, long-term starved cells (LSP) were more sensitive than the cells entering stationay phase for both temperatures. These differences were observed after 160 min of exposure and were reinforced after 320 min of exposure. By comparing the viability of cells from the different cultures after 160 min of irradiation, we observed a general higher resistance under the solar simulator (i.e., when UVA and PAR were also present) compared to the UVB lamps.

122 Chapter 4

a 24¡C SP 24¡C LSP 12¡C SP 12¡C LSP

b

Figure 4.2. Changes in viability under Solar Simulator (a) or UVB lamps (b) according different temperatures (12°C and 24°C) and growth phases (SP and LSP). The grey rectangles indicate the duration where the UV-induced photoproducts were measured under either solar simulator or UVB lamps. The data expressed the mean (± standard deviations) of three independent experiments.

123 Chapter 4

III.2 Quantitative changes in the yield of formation of the total UV- induced bipyrimidine photoproducts

The formation curves of the overall yield of bipyrimidine photoproducts were found to be linear with respect to the applied fluence for all quantified dimers under all irradiation treatments (data not shown), suggesting the absence of secondary photoreactions, such as photoreversion of CPD and photoisomerization of 6-4PPs

(Douki et al., 2000). Total DNA damage appeared to be induced differently according to the temperature and the nature of UV irradiation (Fig. 4.3a and c), but not by the growth phase considering a same temperature and type of irradiation (Fig.

4.3b). The temperature had a major effect for LSP cells exposed to UVB, with greater DNA damage induced at 24°C (t-student, *p<0.05) (Fig. 4.3a). It clearly appeared that the nature of irradiation influenced the most the induction of DNA lesions, with a significant higher yield of formation of photoproducts under UVB irradiation than observed under the solar radiation (t-student, *p<0.05 and **p<0.01)

(Fig. 4.3c). Comparison of SP and LSP did not revealed any quantitative difference in the overall yield of formation of the total bipyrimidine photoproducts. All together these data showed the involvement of temperature / UV irradiation conditions rather than growth conditions in the induction of the overall yield of formation of bipyrimidine photoproducts.

124 Chapter 4

12¡C 24¡C a 8 7

6

pb / min 5 6

4

3 Lesions per 10 2

1

0 SP SP LSP LSP

8 b SP LSP 7

6

pb / min 5 6

4

3 Lesions per 10 2

1

0 12¡C 12¡C 24¡C 24¡C

c 8 SOL UVB

7

6

5 pb / min 6 4

3

Lesions per 10 2

1

0 12¡C 12¡C 24¡C 24¡C

Figure 4.3. Distribution of the yield of total UV-induced photoproducts according (a) different temperatures (12°C and 24°C), (b) growth phases (SP and LSP) and (c) radiation sources (SOL and UVB). Student's t-test assuming equal variances: *p<0.05 and **p<0.01. The data expressed the mean (± standard deviations) of three independent experiments.

125 Chapter 4

III.3 Qualitative changes in the yield of formation of the different UV- induced bipyrimidine photoproducts

The use of HPLC-MS/MS for the quantification of UV-induced DNA damage allowed comparison of the yields of formation of all different bipyrimidine photoproducts. A total of 6 different lesions were characterized by HPLC-MS/MS

(TT, TC, CT, CC CPDs and TT, TC 6-4PPs). The most striking result was the similar yield of formation of 6-4PPs and CPD at TC sites, for 24LSP cultures upon a sunlight treatment (Fig. 4.4). The distribution of photoproducts in this condition was unambiguously different from the other conditions and this was clearly showed in the

HCA of the Fig. 4.5.

Qualitatively, except for the condition cited above (24LSP sunlight treatment), we always observed a predominant formation of TC CPD compare to other lesions.

However according to the HCA, two main clusters exhibited different pattern of

DNA damage induction and interestingly both groups were related to a same temperature treatment, 12°C versus 24°C (Fig. 4.5).

126 Chapter 4

aa TT CPD TC CPD CT CPD CC CPD TT 64 TC 64

2,5

2 pb/min 6 1,5

1 Lesions 10 per

0,5

0 12°C SP 12°C LSP 24°C SP 24°C LSP bb 4

3,5

3 pb/min 6 2,5

2

1,5 Lesions 10 per

1

0,5

0 12°C SP 12°C LSP 24°C SP 24°C LSP

Figure 4.4. Distribution of the yield of bipyrimidine photoproducts induced under Solar Simulator (a) or UVB lamps (b) according different temperatures (12°C and 24°C) and growth phases (SP and LSP). The data expressed the mean (± standard deviations) of three independent experiments.

127 Chapter 4

Figure 4.5. Dendrogram from hierarchical clustering analysis, based on the yield of individual bipyrimidine photoproducts induced.

128 Chapter 4

IV. Discussion

By exposing S. alaskensis to sunlight and UVB radiation, we were able to estimate the relative importance of photorepair processes (Fig. 4.3c). A remarkably slower overall yield of total DNA damage induction was observed upon sunlight compare to

UVB radiation. Those differences might be linked to the nature of irradiation.

Qualitatively, while induction rate of 6-4PPs was similar in both light treatments

(except for 24LSP), the yield of induction of CPDs at the four positions (TT, TC, CT,

CC) was significantly reduced in presence of photoreactivating light compare to the sole exposure to UVB radiation (Fig. 4.4). Taken together, it appears that PAR and

UVA would reduce the CPD photoproducts that are accumulated in S. alaskensis by a possible mechanism of photoreactivation. Furthermore, a recent study confirmed those findings, since S. alaskensis was found to present a quite efficient system of photoreactivation (Matallana-Surget et al, manuscript submitted).

Biological molecules are dynamic structures constantly in motion due to thermal changes in the environment. At low temperature, DNA becomes more negatively supercoiled resulting in the retardation of unwinding and of RNA polymerase access

(Mizushima et al., 1997). The cold induction of DNA-modulating proteins (e.g., gyrase A, HU-ß, H-NS) may assist in maintaining correct or functional DNA topology. Therefore, the interplay of biochemical reactions between DNA and DNA- binding proteins could fluctuate in a temperature dependent manner. Our results indicated that higher temperature led to net DNA damage to be greater with respect to both sunlight and UVB treatments, thus demonstrating that formation of bipyrimidine photoproduct is temperature dependent (Fig. 4.3a). Those results correlated with two previous studies showing that induction of CPDs and 6-4PPs is

129 Chapter 4

temperature dependent in cucumber cotyledon, Tobacco cells, Arabidopsis thaliana

(Takeuchi et al., 1996; Li et al., 2002, 2004; Waterworth et al., 2002). Contrasting to the commonly previous thought, DNA damage induction would be more complex than the simple results of a photochemical process and DNA dynamic could play an important functional role. A possible explanation for the temperature dependence of photoproducts induction could be related to DNA conformational state. Indeed it was previously reported that the distance between the two strands of DNA would exhibit large and fast fluctuation depending on temperature considered (Blagoev et al.,

2006). Furthermore, temperature would also qualitatively affect the induction of photoproducts by modulating the nature of the covalent bond (cyclobutyl ring or 6-4 bond) involved and the site of damage (TT, TC, CT, and CC) (Fig. 4.4 and Fig. 4.5).

In oceans, marine bacteria have to cope with nutritional constraints (Matallana-

Surget et al., 2007) and spend most of their time in some form of growth arrest, equivalent to stationary phase state (Morita, 1997). Our data indicated that the time bacteria spent in stationary phase was very important at 24°C, upon a sunlight treatment, influencing bacterial resistance and the nature of UV-induced bipyrimidine photoproducts. Although the overall yield of formation of total lesions was found to be identical at 24SP and 24LSP upon sunlight exposure, reliable different pattern of induction of photoproducts were observed, with a favoured formation of 6-4PPs at TC sites for the LSP cultures. By comparing the bacteria entering the stationary phase (SP) with long term starved cells, we showed that qualitative changes in the nature of UV-induced photoproducts could lead to different susceptibility towards sunlight radiation. As 6-4PPs are know to be much more cytotoxic than CPDs (Mitchell and Nairn, 1989), a higher abundance of 6-4PPs

130 Chapter 4

at TC sites compare to CPDs in LSP condition would explain a higher sensitivity of

S. alaskensis for extended time of radiation (320 min) during a sunlight exposure

(Fig. 4.2a).

131 CHAPTER 5

The response of the marine bacterium Sphingopyxis alaskensis to solar radiation assessed by quantitative

proteomics Chapter 5

CHAPTER 5

The response of the marine bacterium Sphingopyxis alaskensis to solar radiation assessed by quantitative proteomics

Abstract

The adaptive response of the marine bacterium Sphingopyxis alaskensis RB2256 to solar radiation (both visible and ultraviolet) was assessed by a quantitative proteomic approach using iTRAQ (isobaric tags for relative and absolute quantification). Both growth phase (mid-log and stationary phase) and duration (80 min or 8 h) of different light treatments (combinations of visible light, UVA and UVB) were assessed relative to cultures maintained in the dark. Rates of total protein synthesis and viability were also assessed. Integrating knowledge from the physiological experiments with quanitative proteomics of the 12 conditions tested provided unique insight into the adaptation biology of UV and visible light responses of S. alaskensis.

Quantitative data was generated for 119 proteins, representing 811 high confidence protein identifications (27% of the genome). Mid-log phase cultures produced twice as many proteomic changes as stationary phase cultures, while extending the duration of irradiation exposure of stationary phase cultures did not increase the total number of quantitative changes. Proteins with significant quantitative differences were identified that were characteristic of growth phase and light treatment, and cellular processes, pathways and interaction networks were determined. Key factors implicated in a solar radiation adaptive response included DNA-binding proteins implicated in reducing DNA damage, detoxification of toxic compounds such as glyoxal and reactive oxygen species, iron-sequestration to minimize oxidative stress, chaperones to control protein re/folding, alterations to nitrogen metabolism, and specific changes to transcriptional and translational processes. 133 Chapter 5

I. Introduction

Oceans comprise the world’s largest natural habitat and mainly consist of clear oligotrophic waters, characterized by low levels of nutrients. All organisms in oligotrophic waters can be exposed to the deleterious effects of ultraviolet radiation

(UVR, 280-400 nm) that penetrates deep into clear waters (Tedetti and Sempéré,

2006). Exacerbating the effect of UVR is the stratospheric depletion of ozone which is increasing UVB (280-320 nm) fluxes in high and mid latitude regions, and the impact that climate change is likely to have on the thermal stratification of ocean waters resulting in the increased exposure of organisms to solar radiation in surface waters (Häder et al., 2007; Zepp et al., 2007).

In open ocean waters, marine heterotrophic bacteria are the most abundant members of the microbial community and play a pivotal role in the cycling of carbon, nitrogen and other important nutrients (Fuhrman et al., 1989). Bacteria are among the marine organisms most severely affected by UVR. The severe impact on marine bacteria may be due to their high surface to volume ratio and inefficient protection by pigments. As a result, marine bacteria may be sensitive indicators of UVR changes at the surface of the Earth.

The response of marine bacteria to solar radiation has mainly been assessed by bulk analysis of activity or cellular damage. These studies have revealed a high vulnerability to UVR that is manifested as a reduction in metabolic activity, diminished ectoenzymatic activity, decreased viability, and increased DNA damage

(Jeffrey et al., 2000). Particularly highlighted from these studies is the damaging effects of UVR in contrast to the effects of visible light or photosynthetically active

134 Chapter 5

radiation (PAR, 400-700 nm) which may either stimulate or inhibit bacterial activity

(Morán et al., 2001 and references therein). In contrast to studies of phytoplankton, there is a lack of data characterizing the molecular mechanisms by which marine bacteria respond to solar radiation, including the metabolic pathways and fundamental biological processes that are specifically affected.

Sphingopyxis alaskensis was isolated by dilution to extinction as a numerically abundant bacterium in North Pacific Ocean waters (Schut et al 1993; Eguchi et al

2001). Its capacity to form colonies on plates has made it a useful model for physiological studies (reviewed in Cavicchioli et al 2003), including proteomic analyses (Fegatella et al., 1999; Fegatella and Cavicchioli 2000; Ostrowski et al

2004), and the genome sequence of strain RB2256 has recently been completed (F.

Lauro et al, manuscript submitted). One of the interesting physiological properties about S. alaskensis is its high level of inherent stress resistance to a range of stress inducing agents (e.g., UVB radiation, hydrogen peroxide, ethanol and heat)

(reviewed in Cavicchioli et al., 2003). S. alaskensis has been shown to suffer less

DNA damage and have a higher level of resistance to UVB than a number of other marine bacteria (Joux et al 1999). Due to its high GC content, UVB induces mainly cyclobutane pyrimidine dimers as TC rather than TT in S. alaskensis (Matallana-

Surget et al., 2008).

In the present work we extended upon previous analyses to probe the physiological response of S. alaskensis and develop quantitative proteomics using iTRAQ (isobaric tags for relative and absolute quantification) to assess the effects that solar radiation has on cellular processes, regulatory networks and pathways. We varied the

135 Chapter 5

wavelength and intensity of light that S. alaskensis was exposed to in an attempt to imitate the types of light conditions occuring throughout the depths of the water column that natural light normally penetrates. Using cut-off filters with a solar simulator S. alaskensis cultures were exposed to full sunlight (FS; which equates to

PAR+UVA+UVB), PAR+UVA, PAR and complete darkness (as a control). As previous studies showed that the growth phase of S. alaskensis impacted gene expression (proteomics using two-dimensional gel electrophoresis) and physiology

(Fegatella et al 1998; Fegatella and Cavicchioli, 2000; Ostrowski et al 2004), irradiation treatments were performed on both mid-logarithmic (ML) and stationary phase (SP) cultures. The development of iTRAQ proteomics for S. alaskensis enabled the simultaneous identification and quantification of proteins from the four different light treatments. To the best of our knowledge, this is the first proteomic analysis to assess the response of a marine bacterium to sunlight conditions that mimic those experienced in the natural environment. The findings provide a depth of new insight about the molecular mechanisms of bacterial adaptation to solar radiation.

136 Chapter 5

II. Materials and Methods

II.1 Culture conditions and solar radiation treatment

Sphingopyxis alaskensis RB2256 (DSMZ 13593T) was grown aerobically in

Artificial Sea Water (ASW) supplemented with 3 mM of D-glucose, vitamins and trace elements (Eguchi et al., 1996) in a rotary shaker (130 rpm) at 24°C (maximum growth rate 0.27 d-1). After two precultures, the growth was monitored by optical density (OD) measurements at 433 nm (the OD reading that provides the best sensitivity for monitoring these small-sized cells). Cells from 200 ml cultures were harvested at ML after 36 h of growth and SP after 60 h of growth by centrifugation at

8 000 x g for 8 min at 24°C. In order to maintain an equivalent cell concentration for both ML and SP cultures after centrifugation the volume was adjusted to OD433nm,

0.4 using sterile unsupplemented ASW medium. Cultures were placed into 250-mL cylindrical quartz flasks with flat tops and exposed to simulated sunlight produced by a solar-filtered (1.0 air mass) 1000-W xenon arc lamp (Oriel Corp., Stratford, CT,

USA) that produced an optical output irradiance of 250 W m-2 PAR (400-700 nm),

19.3 W m-2 UVA (320-400 nm), and 0.85 W m-2 UVB (280-320 nm) as measured by a UV/visible RAMSES spectroradiometer (TriOS, Germany). The sunlight intensity used in this study corresponds to ∼60% of the maximal solar intensity measured in summer in the Mediterranean region. Four light treatments were studied simultaneously: (i) FS radiation (PAR+UVA+UVB, culture flask covered with acetate cellulose [50% transmission at 280 nm] to remove residual UVC); (ii)

PAR+UVA (culture flask covered with MylarA film [50% transmission at 320 nm]);

(iii) PAR (culture flask covered with OP3 [50% transmission at 400 nm]); (iv) dark

(culture flask covered with black tape). The four culture flasks were maintained at

24°C by partial submersion in a water bath connected to a cryothermostat with

137 Chapter 5

continuous magnetic stirring (100 rpm). Culture flasks of both ML and SP cells were irradiated for 8 h under the solar simulator and viability and rates of protein synthesis determined (see below). For proteomics, ML cultures were harvested after 8 h of simulated sunlight exposure (ML8h) and SP cultures were harvested after 80 min

(SP80min) and 8 h (SP8h) of simulated sunlight exposure. ML and SP cultures maintained in darkness for either 80 min or 8 h were used as controls. Four separate experiments were performed for each combination of growth phase and irradiation treatment.

II.2 Viability and protein synthesis activity

Colony-forming units (CFU) were determined for each growth phase conditions (ML and SP) and their associated light treatments (FS, PAR+UVA, PAR) by plating appropriate dilutions in triplicate on VNSS agar (Eguchi et al., 1996). Counts were made after 5 d of incubation at 30°C and the percentage of CFU was determined as the ratio of CFU after irradiation compared to initial cell counts.

Protein synthesis activity was estimated from the incorporation rates of 3H-leucine into bacterial protein (Smith and Azam, 1992). Briefly, samples (1 ml in triplicate) were incubated in the dark at 24°C for 4 min (ML) or 15 min (SP) with 2 nM 3H-Leu

(specific activity 120 Ci mmole-1, Perkin Elmer) and 598 nM of unlabelled leucine.

Time course experiments were conducted to confirm the linearity of incorporation during incubations. The incorporated 3H-Leu was collected by micro-centrifugation after precipitation by trichloroacetic acid (TCA, 5% final concentration) and samples were rinsed with TCA and ethanol. The precipates were finally resuspended in 1 ml of liquid scintillation cocktail (FilterCount, Perkin Elmer) and radioactivity

138 Chapter 5

determined using a liquid scintillation counter (LS 5000CE Beckman). Rates of incorporation into live samples were corrected for adsorption of radioactivity using

TCA-killed controls.

II.3 Protein Extraction, iTRAQ labeling and peptide purification

Aliquots (50 ml) of ML8h, SP80min and SP8h irradiated cultures were harvested by centrifugation at 8 000 x g for 10 min at 4°C. Excess salt was removed from the cell pellet in ice cold 0.2 M sucrose and the cells were collected by centrifugation (8 000 x g, 10 min, 4°C). Pellets were resuspended in 100 µL of 0.2 M sucrose and freeze- dried for long term storage at -20°C based on methods previously developed for S. alaksensis (Fegatella et al., 1999). For protein extraction, pelleted cells were re- suspended in 1 ml of freshly made extraction buffer consisting of 10 mM Tris-HCl

(pH=8.0), 1 mM EDTA and 1 mM PMSF. Proteins were extracted by sonication for

4 min with tubes on ice using a digital Branson sonicator at an amplitude of 30% and

0.5 pulse rate. Subsequently samples were centrifuged at 10 000 x g at 4°C for 25 min. Spin dialysis was performed on the supernatant to remove the salts from the protein extract and the total protein concentration obtained was determined using a

Bradford Protein Assay kit (Bio-Rad, Hertfordshire, U.K.) according to the manufacturer’s instructions. A total of 100 µg of protein from all 4 light conditions was used for iTRAQ labeling. Each sample was reduced, alkylated and trypsin

(Promega, Madison, WI) digested overnight at 37°C and labeled with iTRAQ reagents (Applied Biosystems, Forster City, CA) according to the manufacturer’s protocol, with the exception that iodoacetamide was used as the block agent of cysteine during the alkylation step. The digested protein samples were then processed for iTRAQ labeling by adding iTRAQ reagents 114, 115, 116 and 117 to

139 Chapter 5

the dark control, FS, PAR+UVA and PAR samples, respectively. The four labeled samples were mixed by vortexing, incubated at room temperature for 1 h, the samples combined and dried in a vaccum centrifuge at room temperature. The purity of the sample mixture was improved using off-line strong cation exchange (SCX) fractionation, using a Opti-Lynx cartridge holder (Applied Biosystems). To achieve this the sample mixture was re-suspended in buffer containing 10 mM KH2PO4 in

25% ACN pH 3.0 (buffer A) and adjusted to a pH between 2.5 and 3.3 using glacial acetic acid, and subsequently loaded onto the column at a flow rate of 9.5 ml/h. The sample on the column was washed with a further 1 ml buffer A and the peptides were eluted with 500 µL of elution buffer containing 10 mM KH2PO4, 350 mM KCl in

25% ACN pH 3.0. The fractionated sample was then dried, re-suspended in 500 μL of 0.2% heptafluorobutyric acid (HFBA) (Pierce Biotechnology, Rockford, IL) and subsequently loaded onto a C18 RP macrotrap column (Michrom Bioresources,

Auburn, CA) and the column washed with 1.5 ml 0.2% HFBA to desalt the sample.

The sample was eluted by injecting 350 μL of 50% 0.1 % formic acid: 50% ACN, followed by 150 μL ACN, and the purified sample was vacuum dried.

II.4 Mass spectrometry and protein identification

Purified peptides from digested protein samples were identified using a QStar Pulsar i hybrid liquid chromatography tandem mass spectrometry (LC-MS/MS) system

(Applied Biosystems, MDS Sciex) coupled with an online capillary LC system

(Famos, Switchos and Ultimate from Dionex/LC Packings, Amsterdam, The

Netherlands). The instrument was calibrated (variation < 50 ppm) using a

Glufibrinopeptide (Glufib) standard (Sigma, St Louis, MO). Dried samples were re- suspended in 50 μL 0.05% HFBA/ 1% formic acid and loaded onto the QStar

140 Chapter 5

instrument to determine the sample dilution that gave intensity counts in the range of

1000 – 1500 cps for the most abundant peaks. Ten μL of appropriately diluted sample was loaded onto a SCX micro column (0.75 x ~15 mm, Poros S10, Applied

Biosystems) and eluted sequentially using 5, 10, 15, 20, 25, 30, 40, 50, 100, 250, 500 and 1000 mM ammonium acetate (20 µL). The unbound fraction and solution from each salt step were concentrated and desalted on a micro C18 precolumn (500 µm x

2 mm) (Michrom Bioresources) with buffer A (H2O:CH3CN (98:2, 0.1% formic acid)) at 20 µl min-1. After a 10 min wash the pre-column was switched (Switchos) into line with a fritless analytical column (75 µm x ~12 cm) containing C18 reverse phase packing material (Magic, 5 micron, 200Å) prepared as described by (Gatlin et al., 1998). Peptides were eluted using a linear gradient of buffer A to buffer B

-1 (H2O:CH3CN (36:64, 0.1% formic acid) at ~300 nL min over 90 min. High voltage

(2300 V) was applied through a low volume tee (Upchurch Scientific, Oak Harbour,

WA) at the column inlet, and the outlet positioned ~ 1 cm from the orifice of an API

QStar Pulsar i mass spectrometer. Positive ions were generated by electrospray and the instrument operated in information-dependent acquisition mode. A time of flight

MS survey scan was acquired (m/z 350-1700, 0.75 s) and the 3 largest multiply charged ions (counts > 20, charge state ≥2 and ≤4) sequentially selected by the quadrupole for MS-MS analysis. Nitrogen was used as collision gas and an optimum collision energy automatically chosen (based on charge state and mass). Tandem mass spectra were accumulated for 2s (m/z 65-2000). All MS/MS spectra were analyzed using Mascot (Matrix Science, London, UK; version 2.1), set up to search a

S. alaskensis tryptic digest database, using fragment precursor and product ion tolerances of ± 0.25 and 0.2 Da respectively, with the maximum number of missed trypsin cleavage sites set to 1. Oxidation of methionine, iodoacetamide derivative of

141 Chapter 5

cysteine (Cys-carboxamidomethylation) and Applied Biosystems iTRAQA multiplexed quantitation chemistry of lysine, tyrosine and the N-terminus were specified in Mascot as variable modifications. High scores indicated a likely match.

The identification of the overall set of proteins was validated by manual inspection of the MS/MS ion spectra, ensuring that a series of consecutive sequence-specific b- and y-type ions was observed. The proportion of false-positives derived from a shuffle database was estimated to be lower than 1%. The entire iTRAQ experiment was conducted 28 times using different conditions (ML8h, SP80min and SP8h) and 4 biological replicates. Scaffold (version Scaffold_2_01_00, Proteome Software Inc.,

Portland, OR) was also used to validate MS/MS based peptide and protein identifications. By combining all the MASCOT searches from different biological replicates within the same condition, this software enabled the identification of proteins unique to one experimental condition. Peptide and protein identifications were accepted with a score of 80% and 90% respectively, with at least 1 identified peptide per protein.

II.5 Protein quantification

The relative quantification of proteins was achieved during MS/MS by estimating the abundance of reporter ion peaks that corresponded to PAR (m/z 117), PAR+UVA

(m/z 116) and FS (m/z 115) compared to the dark control (m/z 114). MS/MS spectra were searched against the S. alaskensis database for protein quantification using

ProQuant software v1.1 (Applied Biosystems) and the search settings were similar to

Mascot. ProGroup Viewer software v1.0.5 (Applied Biosystems) was used to inspect the differential expression ratio, probabilistic p-value and error-factor (EF) for peptides. A protein was considered significantly differentially expressed when the p

142 Chapter 5

value < 0.05 and EF < 2. The average of iTRAQ-ratios for the proteins found in at least two of the four biological replicates were calculated in log space and the mean was converted back and represented in linear space. For this purpose, the protein list obtained from ProGroupViewer was exported to Microsoft Excel and protein ratios with their EF were first converted to log10 space. The average protein expression is weighted by the EF value and was calculated using the following equation:

N

5(wi . xi) i=1 Weighted mean of log ratio = N

5(wi) i=1

, where w is [1/Log (EF)] and 0 is the log10 (protein ratio). Similarly, the weighted standard deviation (SDw) was calculated using the following equations:

SD Weighted standard deviation, SD = w b0.5

, where SD is the unweighted standard deviation and

N

( 5 wi )² i=1 b = N

5w²i i=1

Data was recorded for proteins with increased abundance above 1.25 or below 0.8 relative to the dark control, and data distinguished as more (SDw < ± 0.1) or less

(SDw > ± 0.1) reproducible.

143 Chapter 5

II.6 Computational analysis

The predicted pI versus MW of proteins was generated using the JVirGel

(http://www.jvirgel.de) tool (Hiller et al., 2003). PSORTb v2.0 with the pre- computed list of S. alaskensis (NC_008048) was used to predict the cellular location of proteins (Gardy et al., 2005). Protein hydrophobicity was estimated by the [Grand average of hydropathicity (GRAVY) index (http://www.expasy.ch/)]. Proteins with positive GRAVY indices are considered hydrophobic (Kyte and Doolittle, 1982). All proteins identified by proteomics were classified into COGs

(http://genome.ornl.gov/microbial/sala/). Hierarchical cluster analysis is a popular method for microarray data analysis that enables groups of genes with similar expression patterns to be identified (Quackenbush, 2001). Hierarchical cluster analysis was developed for application to the iTRAQ for proteins with abundance profiles that were in common from at least two growth and irradiation conditions

(ML8h, SP80min and SP8h). Proteins that behaved similarly across a set of experiments, or withsimilar abundance profiles were grouped together.

Identifications of clusters were performed using the complete linkage hierarchical clustering and Euclidean distance metric, provided in PRIMER v6 software (Clarke,

1993). BioLayout software (Goldovsky et al., 2005) was used to visualize (in three dimensions) the functional relationships between proteins. Values of predicted protein interactions were determined by the STRING database

(http://string.embl.de/) (Jensen et al., 2009). Spheres were coloured according to the functional prediction of protein that was determined by manual annotation rather than COG assignment. Functional assignments for genes were manually evaluated against experimental data from the literature and the confidence of each gene’s predicted function was assigned an Evidence Rating (ER) value based on a system of

144 Chapter 5

manual annotation developed for Methanococcoides burtonii (Allen et al., 2009):

ER1, S. alaskensis protein had been experimentally characterized; ER2, the most closely related functionally characterized orthologue share 1 35% sequence identity along the entire length of the protein; ER3, the most closely related functionally characterized homologue shares < 35% sequence identity along the length of the protein, but all required motifs/ domains for function are present; ER4, an experimentally characterized full-length homologue is not available but conserved protein motifs or domains can be identified; ER5 (hypothetical protein), no functionally characterized homologue can be found, and no characterized protein domains above the Pfam and InterProScan cut-off thresholds can be identified

(Appendix D).

145 Chapter 5

III. Results

III.1 Effect of light treatment on viability and rates of protein synthesis

The viability and rates of protein synthesis of S. alaskensis were similar when ML or

SP cultures were exposed to simulated solar radiation. When exposed to FS, ML cells showed a gradual decrease in viability and protein synthesis compared with cultures maintained in the dark (Fig. 5.1a). PAR+UVA and PAR did reduce viability and rates of protein synthesis, although the impact was substantially less than FS treatment. For SP cultures exposed to FS, viability declined more rapidly than for

ML cultures (Fig. 5.1b). However, the viability and protein synthesis of SP cells was largely unaffected by PAR+UVA or PAR. These data indicate that some component of FS that was not present in PAR or UVA (i.e., UVB) was the main component of the solar radiation that harmed ML and SP cultures of S. alaskensis; with SP cultures being somewhat more sensitive. The data also indicate that the doses of PAR and

UVA used in these experiments was able to harm ML, but not SP cells.

146 Chapter 5

10 10 a

1 1 Relative CFU Relative Relative proteinrate

0,1 0,1 012345678 012345678 10 10 b

1 1 Relative CFU Relative Relative protein rate protein Relative

0,1 0,1 012345678 012345678

Time (hours) Time (hours)

Figure 5.1. Relative changes in viability and protein synthesis rates in S. alaskensis RB2256 exposed to four different light treatments: Full sun (P); PAR+UVA (); PAR (
) and dark (J). a: ML cultures. b: SP cultures. The data are expressed as ratio of the initial count (T0) before sunlight exposure and are means (± standard deviations) based on the data from four independent experiments.

147 Chapter 5

III.2 Properties of proteins identified by LC-MS/MS

A total of 811 proteins were identified by LC-MS/MS, corresponding to 27% of the theoretical proteome of S. alaskensis. Four biological replicates were carried out for the 3 different growth conditions (ML8h, SP80min and SP8h) and a total of 28 LC-

MS/MS runs were performed. The characteristics of the expressed proteome were very similar to the theoretical proteome, spanning the pI range 4-12 and MW up to

125 kDa (Fig. 5.2a).

ab

Figure 5.2. Characteristics of all proteins identified (811) in all combined iTRAQ runs. a: Protein map of MW (kDa) versus pI for all proteins in the theoretical proteome of S. alaskensis RB2256 (black, closed circles) and proteins observed in all iTRAQ runs (grey, closed circles). b: Plot of GRAVY-score versus pI for the total proteins observed in all iTRAQ runs. Proteins with positive GRAVY indices are considered hydrophobic (black, closed circles) and those with negative indices, hydrophilic (white, closed circles).

148 Chapter 5

Among the proteins with a well characterized cellular location (56.9%), the majority were predicted to be cytoplasmic (44.0%), with 12.9% being membrane, periplasmic or outer membrane proteins (Appendix E). A large proportion (33%) of the proteins in the pI range 4-7 was predicted to be hydrophobic (Fig. 5.2b).

Proteins in the expressed proteome were classified into Clusters of Orthologous

Groups of proteins (COGs) and separated into broad functional categories (Tab. 5.1).

All 21 functional classes were represented in the expressed proteome, and the distribution was similar to the theoretical proteome of S. alaskensis. The only noticeable difference was the under-representation of the Defense Mechanisms COG category in the expressed proteome.

III.3 Differential protein abundance

Using SCAFFOLD, proteins that were observed in only one growth condition

(ML8h, SP80min or SP8h) were identified. The majority of the expressed proteome

(733/811) was in common between ML and SP growth conditions. With the exception of 1 protein (Stress response protein Y), all 23 proteins unique to ML

(Appendix F) and 55 proteins unique to SP (Appendix G) had unchanged levels of abundance in response to irradiation (see “Transcription and translation” below). The unique proteins represented a wide range of functional categories. The changed abundance of Stress response protein Y during SP8h is indicative of it playing a role in both the response to irradiation and early stages of SP.

The distribution of COG categories among the significant differentially abundant proteins was different from the total expressed, or theoretical proteomes (Tab. 5.1).

149 Chapter 5

In particular, COG categories for Translation, Amino Acid Transport and

Metabolism, and Post-Translational Modification, Protein Turnover and Chaperones were over-represented (Tab. 5.1). Only a small proportion of the differentially abundant proteins were in the COG categories of Unknown Function or General

Function Prediction Only (Tab. 5.1).

Table 5.1. Distribution of S. alaskensis proteins into COG categories.

A total of 156 iTRAQ-labeled proteins had significant differential abundance in comparison to the dark control, and 119 were identified in at least two biological replicates, representing an overall coverage of 15% of the expressed proteome. For all conditions of growth phase and light exposure, the number of proteins with higher abundance compared to dark growth was similar to the number with lower abundance, with the exception of ML-PAR and SP80min-PAR+UVA where a somewhat higher proportion of proteins with higher abundance was observed (Fig.

5.3). The number of changes in differential abundance caused by the 3 irradiation

150 Chapter 5

exposure regimes (FS, PAR, PAR+UVA) was higher for ML cells (92 proteins), compared to either of the SP conditions (SP80min, 38; SP8h, 43 proteins) (Fig. 5.4a). up-regulated Number proteins of down-regulated Number of proteins of Number

Figure 5.3. Total number of proteins with increased or decreased abundance caused by growth phase and irradiation treatment. The three different irradiation treatments (FS, PAR+UVA, and PAR compared to the dark control) are shown for ML8h (dark bars), SP80min (white bars) and SP8h (grey bars).

The 3 irradiation conditions affected the number of changes in protein abundance differently for the 3 growth conditions. For example, for the SP80min cultures, the

FS treatment produced the largest number of protein abundance changes, for SP8h cultures the 3 different light treatments produced a similar number of changes (∼22 proteins), and for ML cultures, FS and PAR+UVA produced a similar number of abundance changes (69 and 60 proteins, respectively) that was about double the number for PAR (28 proteins) (Fig. 5.3 and Fig. 5.4b).

151 Chapter 5

a

b

Figure 5.4. Venn diagrams showing the relationship of proteins from specific treatments. a: Comparison of proteomics datasets for ML8h, SP80min and SP8h; b: Comparison of proteomics datasets for each irradiation treatment, FS, PAR+UVA or PAR for a single growth condition.

The average expression ratios for proteins with significant differential abundance were visually represented using a color code based on both the level of abundance and standard deviation (Fig. 5.5). The dynamic range of protein abundance changes for all irradiation treatments versus a dark control increased in the order SP80min

152 Chapter 5

(0.60-1.99) < SP8h (0.48-2.60) < ML (0.22-4.31) (Fig. 5.5). A total of 19 proteins from ML cultures exhibited a > 2-fold change in protein abundance compared to only

2 for SP8h and none for SP80min (Fig. 5.5).

Similar proportions of proteins were affected by the sunlight for the 3 different growth conditions, and were classified into three qualitative groups. The first group contained proteins that were over-expressed irrespective of the light treatment; these represented ~45% of the total number of proteins with differential abundance within the same growth condition. The second group contained proteins that were always under-expressed after light exposure and represented ~43%. The third group representing ~12% of proteins had a specific response to the spectrum of sunlight that was used (Fig. 5.5).

153 Chapter 5

PAR PAR UVA Full Sun 117:114 116:114 115:114 a Ratio Ratio Ratio Sala_2154_Putative Glutathione-dependent formaldehyde-activating GFA (ER3) 4.31 2.17 1.83 Sala_1216_NADPH-glutathione reductase EC 1.8.1.7 (ER2) 1.81 1.43 1.87 Sala_0685_Putative Glyoxalase (ER3) 1.27 1.07 2.05 Sala_2058_Chaperone protein DnaK (Heat shock protein 70) (ER2) 1.15 1.67 1.33 Sala_1359_Acetylornithine aminotransferase EC 2.6.1.11 (ER2) 1.14 1.06 1.85 Sala_0149_L-glutamine synthetase EC 6.3.1.2 (ER2) 1.35 1.14 1.03 Sala_0997_Putative Beta-Ig-H3/fasciclin [Precursor] (ER4) 1.76 2.54 Sala_2230_Malate dehydrogenase EC 1.1.1.37 (ER2) 1.35 1.72 Sala_0504_Hypothetical protein (ER4) 1.82 3.65 Sala_0047_Putative branched chain amino acid aminotransferase EC 2.6.1.42 (ER3) 1.49 1.84 Sala_2727_SSU ribosomal protein S16P (ER2) 1.18 1.64 Sala_2816_50S ribosomal protein L23 (ER2) 1.65 1.93 Sala_2326_Nitrogen regulatory protein P-II (ER2) 2.36 3.23 Sala_2799_50S ribosomal protein L15 (ER2) 2.02 2.08 Sala_2807_50S ribosomal protein L24 (ER2) 1.45 1.58 Sala_0453_10 kDa chaperonin_groES protein 1 (ER2) 1.68 2.34 Sala_2872_Putative pfkB family carbohydrate kinase_Putative adenosine kinase EC 2.7.1.20 (ER3) 1.87 1.40 Sala_0178_Electron transfer flavoprotein (Etf), beta-subunit (ER2) 1.38 1.13 Sala_1495_30S ribosomal protein S1P (ER2) 1.12 1.25 Sala_0156_Thioredoxin (ER2) 1.53 Sala_1441_50S ribosomal protein L1 (ER2) 1.46 Sala_1694_Putative uncharacterized protein [Precursor] (ER4) 1.47 Thyhf$!%fQ’ ˆ‰h‡rqru’q ‚trh†r@ p‚€ƒ‚r‡ir‡h†ˆiˆv‡@S! 2.06 Sala_0588_Bacterioferritin (ER2) 2.45 Sala_0843_Trigger factor (ER3) 1.75 Sala_0616_D-3-phosphoglycerate dehydrogenase EC 1.1.1.95 (ER2) 1.91 Sala_0829_Citrate synthase EC 2.3.3.1 (ER2) 1.42 Sala_0148_Nitrogen regulatory protein P-II Signal transducing (ER2) 3.38 Thyhf&'f9I6ƒ‚y’€r h†rDDDir‡h†ˆiˆv‡@8!&&&@S! 1.49 Sala_1048_Putative SapC protein (ER4) 1.62 Sala_2806_50S ribosomal protein L5 (ER2) 2.00 Sala_0617_Putative Phosphoserine aminotransferase EC 2.6.1.52 (ER3) 2.82 Sala_2162_Acetyl-CoA C-acyltransferase / 3-ketoacyl-CoA thiolase EC 2.3.1.16 (ER2) 2.15 Sala_2091_Putative TonB-dependent siderophore receptor (ER3) 2.01 Sala_2387_Putative metallopeptidase M16 like family protein (ER4) 1.62 Sala_0619_Putaive Lysine Decarboxylase (ER3) 1.80 Sala_0845_[2Fe-2S] ferredoxin (ER2) 2.98 Sala_2247_ATP-independent serine endopeptidase_Putative DegS stress sensor of the periplasm (ER2) 2.16 Sala_2741_4Fe-4S ferredoxin, iron-sulfur binding (ER2) 2.98 Sala_2288_ATP synthase F1, beta subunit EC 3.6.3.14 (ER2) 1.30 Sala_1444_Transcription antitermination protein NusG (ER2) 1.18 Sala_1602_Putative DNA-directed RNA polymerase subunit omega EC 2.7.7.6 (ER4) 1.01 Sala_0190_Glucose-6-phosphate 1-dehydrogenase EC 1.1.1.49 (ER2) 1.09 Sala_2797_Adenylate kinase EC 2.7.4.3 (ER2) 1.11 Sala_0565_DNA-directed RNA polymerase subunit alpha EC 2.7.7.6 (ER2) 1.07 Sala_2938_Flagellin-like protein (ER2) 0.54 0.48 0.46 Sala_3108_TonB-dependent receptor with predicted cobalamin (vitamin B12) specificity (ER2) 0.40 0.39 0.40 Sala_1988_Outer membrane Peptidoglycan-associated lipoprotein (PAL) (OmpA family) (ER2) 0.47 0.41 0.55 Sala_2719_50S ribosomal protein L19 (ER2) 0.99 0.71 0.64 Sala_0564_50S ribosomal protein L17 (ER2) 0.57 0.67 Sala_2815_50S ribosomal protein L2 (ER2) 0.63 0.58 Sala_0244_30S ribosomal protein S9 (ER2) 0.58 0.49 Sala_2812_30S ribosomal protein S3 (ER2) 0.63 0.42 Sala_2817_50S ribosomal protein L4 (ER2) 0.51 0.53 Sala_1901_30S ribosomal protein S18 (ER2) 0.57 0.67 Sala_0610_NusA antitermination factor (ER2) 0.41 0.22 Sala_0543_Polyribonucleotide nucleotidyltransferase (PNPase) (EC 2.7.7.8) (ER2) 0.70 0.47 Sala_2818_50S ribosomal protein L3 (ER2) 0.76 0.62 Sala_1148_Peptidyl-prolyl cis-trans isomerase EC 5.2.1.8 (ER2) 0.92 0.52 Sala_1417_Integration host factor subunit alpha (ER2) 0.61 0.95 Sala_0172_Aspartate aminotransferase (EC 2.6.1.-) (ER2) 0.69 Sala_2811_50S ribosomal protein L16 (ER2) 0.47 Sala_1068_Adenylosuccinate lyase EC 4.3.2.2 (ER2) 0.44 Sala_1523_Putative uncharacterized protein (ER4) 0.59 Sala_2312_Membrane dipeptidase EC 3.4.13.19 (ER2) 0.51 Sala_2810_50S ribosomal protein L29 (ER2) 0.44 Sala_1413_Putative Xaa-Pro dipeptidase (ER4) 0.69 Sala_2041_50S ribosomal protein L21 (ER2) 0.78 Sala_2801_30S ribosomal protein S8 (ER2) 0.67 Sala_0791_Serine hydroxymethyltransferase EC 2.1.2.1 (ER2) 0.51 Sala_1319_Glyceraldehyde-3-phosphate dehydrogenase EC 1.2.1.12 (ER2) 0.58 Sala_2096_Aconitase EC 4.2.1.3 (ER2) 0.41 Sala_3029_PutativeTonB-dependent receptor precursor (ER3) 0.36 Sala_0807_Putative MotA/TolQ/ExbB proton channel family protein (ER4) 0.63 Sala_1902_50S ribosomal protein L9 (ER2) 0.64 Sala_2291_Putative uncharacterized protein 0.82 Sala_0681_Gamma-glutamyltransferase EC 2.3.2.2 (ER2) 0.81 Sala_1701_50S ribosomal protein L10 (ER2) 0.98 Sala_1959_Elongation factor Ts (ER2) 0.98 Sala_0976_Putative Base-induced periplasmic YceI-like family protein (ER3) 0.84 Sala_0456_Dieptidyl-carboxydipeptidase A EC 3.4.15.1(ER2) 0.83 Sala_0252_Nucleotide Exchange Factor GrpE protein (ER2) 1.22 0.97 1.54 Sala_0799_Histone-like DNA-binding (ER2) 0.95 2.09 2.71 Sala_2813_LSU ribosomal protein L22P (ER2) 1.28 0.59 1.74 Sala_1472_Ketol-acid reductoisomerase EC 1.1.1.86 (ER2) 1.53 1.82 0.41 Sala_1124_Hypothetical protein (ER4) 1.32 0.93 1.03 Sala_0523_Enolase / (2-phosphoglycerate dehydratase) (EC 4.2.1.11) (ER2) 1.32 0.97 0.61 Sala_1094_Hypothetical protein (ER4) 1.31 0.88 0.50 Sala_1964_Isocitrate dehydrogenase (NADP) EC 1.1.1.42 (ER2) 2.08 0.96 0.58 Sala_0786_30S ribosomal protein S4 0.80 1.17 1.15 Sala_0452_60 kDa chaperonin 1_(groEL protein 1) [groL1] (ER2) 1.17 1.14 0.95 Sala_2344_Cold-shock DNA-binding protein family (ER2) 1.26 0.95

154 Chapter 5

PAR PAR UVA Full Sun b 117:114 116:114 115:114 Ratio Ratio Ratio Sala_0997_Putative Beta-Ig-H3/fasciclin [Precursor] (ER4) 1.85 1.36 Sala_1951_Putative uncharacterized protein containing a signal peptide (ER4) 1.57 Sala_0540_Putative uncharacterized protein [Precursor] (ER4) 1.56 Sala_0799_Histone-like DNA-binding (ER2) 1.25 Sala_2807_50S ribosomal protein L24 (ER2) 1.39 Sala_2387_Putative metallopeptidase M16 like family protein (ER4) 1.31 Sala_0564_50S ribosomal protein L17 (ER2) 1.74 Sala_2291_Putative uncharacterized protein (ER4) 1.55 Sala_2321_Putative uncharacterized protein (ER4) 1.30 Sala_2938_Flagellin-like protein (ER2) 1.39 Sala_2041_50S ribosomal protein L21 (ER2) 1.19 Sala_2725_Putative outer membrane protein OmpA (ER3) 1.19 Sala_1602_Putative DNA-directed RNA polymerase subunit omega EC 2.7.7.6 (ER4) 1.23 Sala_1523_Putative uncharacterized protein (ER4) 1.02 Sala_0976_Putative Base-induced periplasmic YceI-like family protein (ER3) 1.08 Sala_1396_Putative uncharacterized protein (ER4) 1.07 Sala_2344_Cold-shock DNA-binding protein family (ER2) 0.96 0.74 Sala_0280_Acyl carrier protein (ACP) (ER2) 0.73 Sala_2058_Chaperone protein DnaK (Heat shock protein 70) (ER2) 0.79 Sala_2801_30S ribosomal protein S8 (ER2) 0.76 Sala_2820_Translation elongation factor Tu (EF-Tu) (ER2) 0.65 Sala_2980_Putative Ankyrin (ER3) 0.74 Sala_0522_Putative uncharacterized protein (ER4) 0.59 Sala_2872_Putative pfkB family carbohydrate kinase_Putative adenosine kinase EC 2.7.1.20 (ER3) 0.74 Sala_1982_Putative Gamma-glutamyl transferase EC 2.3.2.2 (ER3) 0.75 Sala_1988_Outer membrane Peptidoglycan-associated lipoprotein (PAL) (OmpA family) (ER2) 0.80 Sala_2096_Aconitase EC 4.2.1.3 (ER2) 0.74 Sala_1472_Ketol-acid reductoisomerase EC 1.1.1.86 (ER2) 0.78 Sala_0047_Putative branched chain amino acid aminotransferase EC 2.6.1.42 (ER3) 0.83 Sala_2818_50S ribosomal protein L3 (ER2) 0.82 Sala_0986_Transcription elongation factor GreA (ER2) 0.91 Sala_1417_Integration host factor subunit alpha (ER2) 0.98 Sala_1434_Putative CsbD-like protein (ER4) 1.99 1.81 0.61 Sala_1069_Hypothetical protein (ER4) 0.60 1.10 1.50 Sala_1694_Putative uncharacterized protein [Precursor] (ER4) 0.76 0.98 1.20 Sala_1289_Putative uncharacterized protein (ER4) 0.63 1.50 Sala_2719_50S ribosomal protein L19 (ER2) 1.08 0.81 Sala_1959_Elongation factor Ts (ER2) 1.31 0.77

155 Chapter 5

PAR PAR UVA Full Sun 117:114 116:114 115:114 c Ratio Ratio Ratio Sala_1396_Putative uncharacterized protein (ER4) 1.01 1.49 1.42 Sala_3068_3-hydroxyacyl-CoAdehydrogenase/short chain enoyl-CoA hydratase EC 1.1.1.35/EC 4.2.1.17 (ER2) 1.49 1.65 1.40 Sala_3108_TonB-dependent receptor with predicted cobalamin (vitamin B12) specificity (ER2) 1.19 1.37 1.24 Sala_0453_10 kDa chaperonin_groES protein 1 (ER2) 1.03 1.03 1.31 Sala_2719_50S ribosomal protein L19 (ER2) 1.03 1.09 1.34 Sala_0291_Iron Superoxide dismutase EC 1.15.1.1 (ER2) 2.16 1.41 Sala_1700_50S ribosomal protein L7/L12 (ER2) 1.05 1.25 Sala_0843_Trigger factor (ER3) 1.15 1.40 Sala_2742_ATP-dependent Clp protease proteolytic subunit (ER2) 1.43 Sala_0786_30S ribosomal protein S4 (ER2) 1.47 Sala_1959_Elongation factor Ts (ER2) 1.65 Sala_1444_Transcription antitermination protein NusG (ER2) 1.90 Sala_2820_Translation elongation factor Tu (EF-Tu) (ER2) 1.34 Sala_2817_50S ribosomal protein L4 (ER2) 1.37 Sala_1182_Molybdenum cofactor biosynthesis protein B (ER2) 2.60 Sala_0801_Two component signal OmpR/PhoB transcriptional regulator (ER2) 1.33 Sala_2837_Putative Ribosome (S30EA) associated stress response protein Y_Sigma 54 modulation protein (ER3) 1.61 Sala_1902_50S ribosomal protein L9 (ER2) 1.23 Sala_1048_Putative SapC protein (ER4) 1.18 Sala_2291_Putative uncharacterized protein (ER4) 0.74 0.70 0.73 Sala_0799_Histone-like DNA-binding (ER2) 0.70 0.75 0.85 Sala_0997_Putative Beta-Ig-H3/fasciclin [Precursor] (ER4) 0.55 0.48 Sala_0456_Dieptidyl-carboxydipeptidase A EC 3.4.15.1(ER2) 0.80 0.74 Sala_1069_Hypothetical protein (ER4) 0.78 0.78 Sala_1398_L-threonine synthase EC 4.2.3.1(ER2) 0.74 Sala_0494_Putative uncharacterized protein [Precursor] (ER4) 0.80 Sala_0681_Gamma-glutamyltransferase EC 2.3.2.2 (ER2) 0.70 Sala_0685_Putative Glyoxalase (ER3) 0.67 Sala_0781_Acetyl-CoA C-acetyltransferase (Thiolase II) (EC 2.3.1.9) (ER2) 0.77 Sala_1951_Putative uncharacterized protein (ER4) 0.78 Sala_2091_Putative TonB-dependent siderophore receptor (ER3) 0.74 Sala_2725_Putative outer membrane protein OmpA (ER3) 0.67 Sala_0216_Putative Peptidase M20 (ER3) 0.50 Sala_1472_Ketol-acid reductoisomerase EC 1.1.1.86 (ER2) 0.71 Sala_0185_Hypothetical protein (ER4) 0.67 Sala_0976_Putative Base-induced periplasmic YceI-like family protein (ER3) 0.66 Sala_1694_Putative uncharacterized protein [Precursor] (ER4) 0.85 Sala_0191_Hypothetical protein (ER4) 0.94 Sala_2066_Ferric uptake regulator, Fur family (ER2) 0.89 Sala_0178_Electron transfer flavoprotein (Etf), beta-subunit (ER2) 0.82 Sala_2247_ATP-independent serine endopeptidase_Putative DegS stress sensor of the bacterial periplasm (ER2) 1.11 0.93 0.56 Sala_0452_60 kDa chaperonin 1_(groEL protein 1) [groL1] (ER2) 0.86 1.26 Sala_0526_Pyruvate dehydrogenase E1 component beta subunit (ER2) 0.75 1.27

Figure 5.5. Quantitative proteomics grid of proteins with differential abundance. a: ML8h; b: SP80min; c: SP8h. Each cell of the grid was colored based on increased abundance above 1.25 or below 0.8 relative to the dark control, and the level of reproducibility of the data; more reproducible (SDw < ± 0.1) or less reproducible (SDw > ± 0.1): Increased abundance ratio < 1.25 (SDw > ± 0.1 ratio values) Increased abundance ratio < 1.25 (SDw < ± 0.1 ratio values) Increased abundance ratio > 1.25 (SDw > ± 0.1 ratio values) Increased abundance ratio > 1.25 (SDw < ± 0.1 ratio values) Decreased abundance ratio > 0.8 (SDw > ± 0.1 ratio values) Decreased abundance ratio > 0.8 (SDw < ± 0.1 ratio values) Decreased abundance ratio < 0.8 (SDw > ± 0.1 ratio values) Decreased abundance ratio < 0.8 (SDw < ± 0.1 ratio values)

156 Chapter 5

From the 119 proteins with differential abundance, 46 were common between at least two growth conditions and only 8 were detected in all 3 conditions (Fig. 5.4a).

Among the 46 common proteins, seven distinct response patterns of protein abundance were revealed by HCA (Fig. 5.6). Clusters 1, 2, 3 and 7 were mainly represented by proteins with highest abundance during ML growth, in contrast to cluster 6 which mainly contained proteins with increased abundance during SP.

Cluster 5 contained proteins with increased abundance resulting from starvation followed by a short period of FS (SP80min-FS). Cluster 4 was the largest cluster with most of the 16 proteins involved in transcription and translation; the cluster did not represent a particular growth or light condition. This HCA of the 46 common proteins revealed that the physiological state of the cell (ML versus SP), rather than the nature of the light treatment or time of irradiation, was the main factor determining the proteomic response of the cells. This is the first report of the use of

HCA for the analysis of quantitative proteomics data, and proved to be very useful for identifying relationships between proteins in our complex dataset.

Even though the number of proteins linked to different durations of sunlight exposure for starved cells (SP80min and SP8h) was similar (38 and 43 proteins, respectively; Fig. 5.4a), most of the proteins were specific to duration of exposure.

Twenty five proteins were associated with an early response (80 min) and 30 to a late response (8 h), with only 13 of the total 68 in common (Fig. 5.4a). The proteins associated with higher protein abundance during an early sunlight response were ribosomal proteins or had uncertain functional annotations (i.e., 6/9 were ER4; see

Materials and Methods) (Fig. 5.5). For SP8h, 13/16 proteins had good evidence for function (ER2) and 11 had abundance levels greater than 1.25-fold. Seven of the 15

157 Chapter 5

were associated with an oxidative stress response (discussed below) and the 6 others with metabolic pathways (Sala_3068, Sala_0526, Sala_1182) and transcription/translation processes (Sala_1959, Sala_1444, Sala_2820) (Fig. 5.5c).

Some of the 13 proteins in common to SP (Fig. 5.4a) were involved in translation

(Sala_1959, ER2; Sala_2820,ER2; Sala_2719, ER2), DNA repair pathways

(Sala_2344, ER2; Sala_0799, ER2), or were periplasmic proteins (Sala_0976, ER3)

(Fig. 5.6). Their abundance was not regulated the same way by duration of irradiation. Seven of the 13 proteins had higher abundance following 80 min irradiation, while 9/13 had lower abundance following 8 h exposure. Seven of the proteins were regulated in the same way by irradiation treatment, whereas 4 had contrasting responses; 3 of these 4 in response to FS treatment. These latter 3 proteins had similar levels of differential abundance following 80 min or 8h exposure and were proteins annotated with uncertain functions (ER4).

Fifty one proteins were specific to ML and 27 were specific to SP (combined 80 min and 8 h exposure) (Fig. 5.4a). Forty one were in common between ML and at least one of the SP conditions and 27 of the 41 proteins were found in common for same irradiation treatment. Eleven of the 27 proteins exhibited a common increase or decrease in abundance, whereas 14 exhibited opposed changes. The 11 proteins included chaperone proteins (Trigger factor: Sala_0843, ER3; GroES: Sala_0453,

ER2), histone-like DNA-binding proteins (Sala_0799, ER2) and Beta-Ig-H3-

Fasciclin protein (Sala_0997, ER4) with higher abundance following ML and SP growth, with FS treatment (Fig. 5.6). With the exception of 2 proteins (Beta-Ig-H3-

Fascisclin and histone-like protein; Fig. 5.6, Cluster 7), the 8 proteins in common to

158 Chapter 5

all growth conditions (Fig. 5.4a) had common increases or decreases in protein abundance. Both Beta-Ig-H3-Fascisclin and histone-like protein had increased abundance during SP80min and decreased abundance for SP8h for PAR and FS irradiation.

To assess functional associations between the 119 proteins with significant differential abundance, interaction networks were constructed and visualized using

BioLayout (Appendix H). Eighty two proteins were found to be functionally associated, either through direct physical interaction or indirect association (e.g., sharing a substrate in the same metabolic pathway, having common gene regulation or participating in a common protein assembly). Fifteen different functional categories were identified, with only 4 proteins being annotated with unknown functions (ER4). This functional association and visualization method was useful for drawing inferences about the molecular mechanisms of adaptation occurring within the cell (Fig. 5.7; discussed below).

159 Chapter 5 0.41 0.82 0,53 0.64 0.59 1.18 1.62 0.95 0.95 1.47 2.01 1.40 1.18 0.41 0.81 1.02 1.62 0.570.51 0.67 0.71 0.76 0.62 1.68 2.34 1.45 1.58 0.84 0.64 0.61 0.78 0.67 1.26 1.17 1.15 0.98 1.76 2.54 1.49 1.84 1.75 2.06 2.16 1.38 1.13 2.09 2.71 0.540.47 0.480.40 0.41 0.46 0.39 0.55 0.40 0.83 1.170.80 1.14 0.95 1.27 1.071.15 2.05 1.87 1.67 1.33 2.19 1.53 1.82 0.95 0.73 1.23 1.26 1.47 1.42 0.48 0.67 1.43 0.56 0.70 0.85 0.70 0.66 0.86 1.49 1.65 0.55 1.40 1.27 0.71 70 0.75 . 0.67 0.78 1.03 1.09 1.34 0.99 1.191.03 1.37 1.03 1.24 1.31 0.74 1.37 1.34 1.90 1.18 0.74 0.80 0.75 0.82 0.85 1.15 0.75 1.11 0.93 0 0.74 0.80 0.60 0.78 0.78 1.23 0.98 1.07 1.01 1.31 1.85 0.76 0.78 1.08 0.96 5 FS P+UVA P P PUVA FS P PUVA FS 1.39 1.19 1.57 1.74 0.81 0.82 1.02 1.2 1.501.55 1.10 1.39 1.19 0.76 0.74 0.65 0.83 0.79 0.74 1.08 0.77 1.36 1.31 1.20 0.98 SP80m SP80m SP80m SP8h SP8h SP8h ML8h ML8h ML8h 1 2 3 4 5 6 7 Sala_2387_Putative metallopeptidase M16 like family protein (ER4) protein family like M16 metallopeptidase Sala_2387_Putative (ER4) [Precursor] protein uncharacterized Sala_1694_Putative (ER4) protein SapC Sala_1048_Putative L24 (ER2) protein ribosomal Sala_2807_50S (ER3) receptor siderophore TonB-dependent Sala_2091_Putative (ER3) Glyoxalase Sala_0685_Putative (ER3) EC 2.6.1.42 aminotransferase acid amino chain branched Sala_0047_Putative Thyhf!$'f8uhƒr ‚rƒ ‚‡rv9hFCrh‡†u‚pxƒ ‚‡rv&@S! !@S" Thyhf!'&!fQˆ‡h‡v‰rƒsx7sh€vy’ph i‚u’q h‡rxvh†r0ƒˆ‡h‡v‰r6qr‚†vrxvh†r@8!& Thyhf!&#!f6UQqrƒrqr‡8yƒƒ ‚‡rh†rƒ ‚‡r‚y’‡vp†ˆiˆv‡@S! (ER3) factor Trigger Sala_0843 p‚€ƒ‚r‡ir‡h†ˆiˆv‡@S! Thyhf$!%fQ’ ˆ‰h‡rqru’q ‚trh†r@ (ER2) periplasm bacterial the of sensor stress DegS Putative endopeptidase. serine Sala_2247_ATP-independent (ER2) EC 1.1.1.86 reductoisomerase Sala_1472_Ketol-acid (ER2) EC 2.3.2.2 Sala_0681_Gamma-glutamyltransferase (ER4) 2.7.7.6 EC omega subunit polymerase RNA DNA-directed Sala_1602_Putative (ER3) protein family YceI-like periplasmic Base-induced Sala_0976_Putative EC 3.4.15.1(ER2) A. Sala_0456_Dieptidyl-carboxydipeptidase L9 (ER2) protein ribosomal Sala_1902_50S (ER2) alpha subunit factor host Sala_1417_Integration L21 (ER2) protein ribosomal Sala_2041_50S (ER2) S8 protein ribosomal Sala_2801_30S (ER2) family protein DNA-binding Sala_2344_Cold-shock &'f@yrp‡ ‚‡ h†sr syh‰‚ƒ ‚‡rv@S! Thyhf 1) (ER2) protein 1_(groEL chaperonin kDa Sala_0452_60 (ER2) S4 protein ribosomal Sala_0786_30S (ER4) protein uncharacterized Sala_1396_Putative Ts (ER2) factor Sala_1959_Elongation (ER2) Tu (EF-Tu) factor elongation Sala_2820_Translation (ER2) nusG protein antitermination Sala_1444_Transcription (ER4) protein Sala_1069_Hypothetical fQˆ‡h‡v‰rˆpuh hp‡r v“rqƒ ‚‡rv@S# Thyhf!!( Thyhf!&!$fQˆ‡h‡v‰r‚ˆ‡r €r€i hrƒ ‚‡rvP€ƒ6@S" ($ fQˆ‡h‡v‰rˆpuh hp‡r v“rqƒ ‚‡rv@S# Thyhf &@S! Thyhf$%#f$T vi‚†‚€hyƒ ‚‡rvG L4 (ER2) protein ribosomal Sala_2817_50S L19 (ER2) protein ribosomal Sala_2719_50S L3 (ER2) protein ribosomal Sala_2818_50S (ER4) protein uncharacterized Sala_1523_Putative (ER2) 4.2.1.3 EC Sala_2096_Aconitase (ER2) protein Sala_2938_Flagellin-like (''fPˆ‡r €r€i hrQrƒ‡vq‚ty’phh††‚pvh‡rqyvƒ‚ƒ ‚‡rvQ6GP€ƒ6sh€vy’@S! Thyhf specificity(ER2) B12) (vitamin cobalamin predicted with receptor Sala_3108_TonB-dependent @S! Thyhf#$"f x9hpuhƒr ‚vft ‚@Tƒ ‚‡rv (ER2) protein HU DNA-binding Sala_0799_Histone-like (ER4) [Precursor] Beta-Ig-H3/fasciclin Sala_0997_Putative

Figure 5.6. HCA of quantitative proteomics data. The colored grid is linked by a dendrogram representing clusters formed from the different experimental treatments (top), and protein clusters (side). The colour code is the same as Fig. 5.5.

160 Chapter 5

IV. Discussion

IV.1 Overviewing the relationship between sunlight exposure and cell stress

For ML cells, the rate of protein synthesis decreased as the spectrum of irradiation increased (e.g., 27% for PAR, 54% for PAR+UVA, 89% for FS treatment), and viability mirrored this trend (Fig. 5.1a). In contrast, the number of proteomic changes increased (Fig. 5.3). This indicates that while the overall rate of protein synthesis decreased and cells experienced increased levels of stress, cells responded by changing the protein abundance levels of an increasing number of individual proteins. It is noteworthy that the abundance of proteins can change not only as a result of gene expression, but also by augmenting protein stability and turnover. This type of response is reminiscent of stress responses in a broad range of bacteria (e.g., cold shock or heat shock response in E. coli).

The extent of proteomic changes for starved cells (SP80min, SP8h) with different light exposures (Fig. 5.3) did not correlate with the inferred stress levels of the cell

(Fig. 5.1b). In particular, FS (but not PAR+UVA or PAR) greatly impacted viability and protein synthesis (Fig. 5.1b), whereas the number of proteomic changes for SP8h were similar for all irradiation treatments. In comparison to ML, SP cells are likely to be constrained in their capacity to mount a gene regulatory response; this is consistent with the overall lower number of proteomic changes observed for SP versus ML cells (Fig. 5.3).

The major proteomic differences between ML and SP cells related to the proportion of proteins involved in an oxidative stress response (Fig. 5.5). In response to

161 Chapter 5

irradiation, ML cells synthesized a wider range of proteins involved in oxidative stress compared to SP cells, including antioxidant enzymes such as thioredoxin, proteins involved in the detoxification of toxic UV-induced glyoxal and proteins involved in reducing the generation of hydroxyl radicals (see “Oxidative Stress” below).

Comparing the regulatory pattern of SP proteins arising from short (80 min) versus long (8 h) exposure to irradiation revealed different sets of responsive proteins.

Proteins that decreased during short exposure included those involved in the TCA cycle (Sala_2096, ER2), fatty acid biosynthesis (Sala_0280, ER2), osmolarity control

(Sala_1988, ER2), and both cold and heat shock proteins (Sala_2344, ER2 and

Sala_2058, ER2), which may be indicative of the minimization of non-essential biosynthetic and non-specific stress response pathways (Fig. 5.5b). In comparison,

15 proteins had higher abundance after 8 h solar radiation, that are likely to play a role in adaptation to more pronounced oxidative stress, including superoxide dismutase (Sala_0291, ER2), chaperone proteins (GroES: Sala_0453, ER2; trigger factor: Sala_0843, ER3; GroEL: Sala_0452, ER2), two associated stress response proteins (Protein Y: Sala_2837, ER3 and OmpR: Sala_0801, ER2) and an iron homeostasis control protein (Ton-dependent receptor: Sala_3108, ER2 and

Sala_2091, ER3). This is consistent with other studies identifying different responses to short versus long term UV stress (Ehling-Schulz et al., 2002).

The ability of S. alaskensis to perform photoreactivation of DNA damage has been well characterized (Joux et al., 1999; Matallana-Surget et al., 2009). However, the deoxyribodipyrimidine photolyase (Sala_0998) was not detected during proteomic

162 Chapter 5

analysis. Two proteins involved in light-independent DNA repair (DNA repair

RadA: Sala_1806; DNA repair RecN: Sala_0546) were identified in the expressed proteome, but did not have significant differential abundance for any of the conditions tested. This may indicate that both proteins play an equally important role in DNA repair under both dark and light conditions (Booth et al., 2001).

IV.2 Describing the roles of specific proteins in the adaptive responses to sunlight.

IV.2.1 DNA-binding proteins

Proteins representing 2 classes of DNA-binding proteins were responsive to irradiation; a histone-like protein (HU) (Sala_0799, ER2) and a cold-shock domain

(Csp) protein (Sala_2344, ER2) (Fig. 5.6, Cluster 7 and 4 respectively). The histone- like protein had increased protein abundance from SP80min (1.26-fold) and ML8h

(2.71 and 2.09-fold) cells in response to FS and PAR+UVA, and decreased abundance (0.70 and 0.75-fold) from SP8h cells following PAR and PAR+UVA irradiation (Fig. 5.6). The cold-shock domain protein had 1.26-fold increase during

ML8h treatment with FS.

Histone-like proteins play diverse roles in E. coli, including the initiation of DNA repair by displacing the LexA repressor binding to the SOS operon (Preobrajenskaya et al., 1994), DNA recombination (Dri et al., 1992), oriC-dependent DNA replication

(Bramhill and Kornberg, 1988) and transcriptional regulation (Aki and Adhya,

1997). Cells lacking HU are extremely sensitive to ionizing and UV radiation

(Boubrik and Rouvière-Yaniv, 1995; Li and Waters, 1998). The increased abundance in S. alaskensis during ML long exposure (8 h) and SP short exposure (80 min), but

163 Chapter 5

not SP long exposure, may indicate that this particular histone-like protein plays a functional role (e.g., recruit proteins involved in DNA repair) primarily under conditions when the cell is capable of continuing to mount an adaptive response; i.e.,

8 h of starvation and irradiation may exceed the cell’s capacity to sustain this response.

Bacterial proteins with cold-shock domains, including Csps, can function as nucleic acid chaperones that bind RNA and DNA and fulfill roles in replication, transcription and translation (Horn et al., 2007). In E. coli Csps have been reported to modulate resistance to ionizing radiation, including UV light (Verbenko et al., 2003). The major consequence of long term UV irradiation is DNA damage (Evan and Cooke,

2004), and the effects on S. alaskensis have been well documented (Matallana-Surget et al., manuscript submitted). Our proteomic data are consistent with this, indicating that both histone and Csp proteins play roles in protecting the cell against nucleic acid damage (Fig. 5.7).

IV.2.2 Oxidative stress

It is well established that solar radiation can lead to oxidative stress (Sies, 1997).

Many proteins involved in counteracting oxidative stress were detected in ML8h cells, although some (e.g., superoxide dismutase, Sala_0291, ER2) had elevated levels during SP growth (Fig. 5.5a and 5.6c). Thioredoxin (Sala_0156, ER2) was

1.53-fold higher in ML8h following FS treatment. Thioredoxin may promote the reduction of intracellular protein disulfide in S. alaskensis, thereby reducing oxidative damage. Superoxide dismutase (SOD) was 2.16 and 1.41-fold higher in

SP8h cells following PAR and PAR+UVA treatments, respectively (Figure 5.6c). S.

164 Chapter 5

alaskensis encodes an Iron-SOD (ER2) that is likely to facilitate defence against free

radicals (e.g., hydroxyl radical that can damage DNA and proteins) and catalyze the

dismutation of toxic superoxide into oxygen and hydrogen peroxide (Fig. 5.7).

Lipid Degradation of Oxidative stress peroxidation glycated proteins Chaperones ClpB Cell- to-cell Glucose O O [GrpE-DnaK] adhesion Glyoxal Beta-Ig-H3 Folding Fasciclin R-SH Chaperones GSR GSH Disaggregating Thioredoxin Complex ATP 3+ Glyoxalase I Fe Siderophor R-SS-R GrpE GSSG ATP ADP S hydroxymethyl GrpE glutathione ADP Active form [Fe] Glyoxalase II Homeostasis Native GSR Ferric Uptake Glycolate Regulator (Fur) GSSG ADP OmpA TonB dependent GSH Siderophore SOD GroES ATP ? Proteins reduction storage GroEL 2+ O . GSH Fe 2 O2 H2O2 H O peroxidase 2 Fe2+ /Fe3+ [Fe-S] ROS-induced ROS-induced DnaK Ferredoxin . DNA damage HO protein damage Bacterioferritin Trigger Factor DNA binding proteins RNApol Ribosomal [SOS response] LexA/RecA proteins

Histone like * Csp Antitermination * Translation DNA binding factor elongation protein Protein Y protein NusA factor Ts/Tu NusG Transcription Translation

Figure 5.7. Cartoon depicting important proteins and cellular processes that are important in S. alaskensis for its adaptive response to sunlight.

165 Chapter 5

Similar to the histone-like protein, SP cells had higher abundance of Beta-Ig-H3 fasciclin protein (Sala_0997, ER4) following a short exposure (1.36 and 1.85-fold) to

FS and PAR treatments and decreased abundance following a long exposure (0.55 and 0.48-fold) to FS and PAR+UVA irradiation (Fig. 5.6, Cluster 7). Beta-Ig-H3 fasciclin is a membrane anchored protein that possesses a FAS1 extracellular domain that can play a role in cellular adhesion in animals and plants, and is found in members of all 3 domains of life (Kim et al., 2000). Furthermore, according to the

STRING database Beta-Ig-H3 fasciclin was predicted to interact with a thioredoxin- like protein precursor (Sala_0074) and a deoxiribopyrimidine photolyase

(Sala_0998) (data not shown). It is possible that thioredoxin and photolyase in association with Beta-Ig-H3 fasciclin constitute a UV-stress, signal-response system

(Fig. 5.7).

A complete set of enzymes that carry out the detoxification of glyoxal, including glutathione (Sala_2154, ER3), glutathione reductase (Sala_1216 ER2) and glyoxalase (Sala_0685, ER3) (Fig. 5.7), had increased abundance in ML8h cells under all irradiation treatments (Fig. 5.5a). Protein levels increased ~2-fold for each enzyme, with the exception of the glutathione-dependent formaldehyde-activating enzyme which increased markedly to 4.31-fold during PAR treatment (Fig. 5.5a).

Glyoxal can form covalent bonds with free amino groups of proteins (Shangari et al.,

2003), and in the presence of iron may form formaldehyde that can impair the function of membranes, proteins and DNA through alkylation, mutation and cross- linking reactions (Heck et al., 1990; Ma and Harris, 1988). In S. alaskensis, glyoxal may be produced through pathways including glucose oxidation, degradation of glycated proteins or lipid peroxidation, and the glyoxal removed by the glyoxalase

166 Chapter 5

system. This may involve glyoxalase combining with the reduced form of glutathione to form S-2-hydroxyethylglutathione, which can then be further hydrolysed by glyoxalase into glycolate (Shangari et al., 2003).

IV.2.3 Iron homeostasis

For ML8h growth conditions and all irradiation treatments, the protein abundance was reduced ~2-fold for a TonB-dependent receptor (Sala_3108, ER2) and an OmpA homologue (outer membrane porin) (Sala_1988, ER3) (Fig. 5.5c). In addition, for

ML8h cells, two ferredoxins (Sala_2741, ER2 and Sala_0845, ER2) and a bacterioferritin (Sala_0588, ER2) had elevated levels (2.98 and 2.45-fold, respectively) during PAR+UVA treatment. While iron is essential for growth, it is generally limiting in many environments. Fe3+ is the major form in the marine environment (Rue and Bruland, 1997), but is not directly assimilatable. It can be transported into the cell through the action of siderophores that capture Fe3+, and the combined effects of TonB receptors and porins, such as OmpA (Higgs et al., 2002;

Wandersman and Delepelaire, 2004). Fe3+ is immediately reduced into the highly cytotoxic Fe2+ form, which can react via the Fenton reaction to produce the hydroxyl radical causing oxidative stress (see “Oxidative Stress” above). To protect against reactive Fe2+, the iron can be sequestered into ferritin and in heme containing bacterioferritin (Wandersman and Delepelaire, 2004). Ferredoxin may directly interact with bacterioferritin (and other iron complexes) to regulate iron storage or release (Andrews, 1998). The proteomic response of S. alaskensis is indicative of a comprehensive response to reduce uptake and increase sequestration of iron as a mechanism of reducing iron-linked oxidative damage to the cell (Fig. 5.7).

167 Chapter 5

IV.2.4 Nitrogen metabolism

Two nitrogen metabolism regulatory proteins (PII) (Sala_0146, ER2 and Sala_2326,

ER2) exhibited large increases in abundance in ML8h cells following FS (3.38 and

3.23) and PAR+UVA (2.36 for one of them) treatments. PII proteins are signal transduction proteins that regulate nitrogen metabolism coupled to central carbon metabolism. The nitrogen assimilatory capacity of bacteria is typically, tightly controlled in order to coordinate effective nitrogen and carbon assimilation (Ninfa and Atkinson, 2000). The marked changes in S. alaskensis caused by the more severe irradiation treatments may reflect a broad impact on cellular function, and an adaptive response to rebalance central metabolic processes (Fig. 5.7).

IV.2.5 Protein folding and processing

Six proteins involved in protein folding and processing, DnaK (Sala_2058, ER2;

1.67 and 1.33-fold), GroEL (Sala_0452, ER2;1.26-fold), GroES (Sala_0453, ER2;

1.34-fold), GrpE (Sala_0252, ER2; 1.54-fold), trigger factor (Sala_0843, ER3; 1.75- fold) and Clp (proteolytic subunit) (Sala_2742, ER2; 1.43-fold), exhibited increased abundance as a result of SP8h and/or ML8h treatment with PAR+UVA and/or FS

(Fig. 5.5b and c). The chaperones play critical roles in the folding of newly synthesized proteins and the refolding of mis-folded proteins, and Clp proteolytically degrades mis-folded proteins. The enhanced cellular level of these proteins in response to irradiation indicates that an important mechanisim for coping with UV- induced stress in S. alaskensis is to maintain proper protein function (Fig. 5.7).

Previous studies have associated a role for GroE and DnaK in coping with UV- induced damage in both E. coli and mammalian skin (Krueger and Walker, 1984;

Trautinger et al., 1996).

168 Chapter 5

IV.2.6 Transcription and translation

A homologue of the bacterial transcriptional termination/antitermination protein,

NusG (Sala_1444, ER2) (Pasman and von Hippel, 2000), had higher cellular levels

(1.9-fold) following SP8h treatment with PAR. The stress response protein Y (pY)

(Sala_2387, ER3) that plays a role in translational regulation in E. coli (Ye et al.,

2002), appears to be important for adaptation to UV in S. alaskensis (1.61-fold increase following SP8h treatment with FS), in addition to playing a specific role in stationary phase adaptation (see “Differential protein abundance” above). pY has been shown to be able to bind to the small ribosomal subunit and stabilize ribosomes when bacteria are stressed (Ye et al., 2002).

Both translation elongation factors, EF-Ts (Sala_1959, ER2; 1.31 and 1.65-fold) and

EF-Tu (Sala_2820, ER2; 1.34-fold) had higher abundance following SP80min and/or

SP8h treatment with PAR and/or PAR+UVA (Fig 5.6b and c). It is noteworthy that in SP, neither PAR nor PAR+UVA reduced rates of protein synthesis (Fig. 5.1b), suggesting that the increased cellular levels of the elongation factors may relate to alternative cellular functions. In S. alaskensis this may relate to a role in the refolding and renaturation of proteins, as has been suggested for E. coli (Dantas Caldas et al.,

1998).

Following ML8h treatment with PAR+UVA or FS, 14 of the 20 ribosomal proteins detected, and EF-Tu, exhibited decreased cellular levels (Fig. 5.5a). Previous studies have shown that ribosome synthesis does not directly correlate with protein synthesis, indicating that ribosomes themselves are not limited to a role in translation in S. alaskensis (Fegatella and Cavicchioli, 2000).

169 Chapter 5

These data suggest that important components associated with transcriptional (NusG) and translational (pY) processes contribute to an adaptive response, presumably by helping to maintain normal gene expression. In contrast, other core translational components (EF-Ts and EF-Tu) appear to take on additional auxiliary roles. Our proteomics assessment highlights the complex role that translational proteins play in the adaptative response of S. alaskensis to irradiation.

170 Chapter 5

V. Conclusion

Photosynthetically active radiation and UV-A have a major impact on the expressed proteome of S. alaskensis without greatly affecting viability and protein synthesis, underscoring the importance of the molecular responses for maintaining growth and survival. In addition to oxidative stress responses, specific aspects of bacterial metabolism (e.g. nitrogen and iron homeostasis) are important physiological traits that may conceivably be linked to ecologically relevant factors (e.g. diurnal cycle, nutrient limitation). The study has prompted specific lines of future study at the molecular and ecological levels, and provides a platform for comparative analyses with other important marine microorganisms.

171 CHAPTER 6

General discussion Chapter 6

CHAPTER 6 General discussion

I. Summary of the findings

Stratospheric ozone reduction results in increasing biologically harmful ultraviolet B radiation (UVB) and catches our attention since this phenomenon will continue at least to be present until the middle of this century. UV response is a complex system that integrates damaging affects and protection/repair processes. Until now, the majority of research focusing on the general understanding of UV response in marine heterotrophic bacteria was accomplished in the field, where multiple biotic and abiotic factors can interfere. If the studies based on single organism in the laboratory can be considered as reductive, they allow to describe in details the physiological responses to UV in controlled conditions. In this thesis we reported that the impact of different temperature (12°C versus 24°C), bacterial physiological state (ML versus

SP), UV/light treatment (Solar simulator: FS, PAR+UVA, PAR or UVB lamps) and variable time of exposure (80 min versus 8 h) remarkably modulate the impact of solar radiation in term of viability, DNA damage, protein synthesis activity and protein regulation. Physiological, molecular and proteomics approaches were used to dissect the molecular response of the marine oligotrophic bacterium S. alaskensis to sunlight. The overall set of experiments was performed under controlled ecologically relevant condition, so that the results provide ecologically meaningful data that can be related to ecosystem level effects.

In this thesis, we first revealed an original pattern of DNA damage induction in S. alaskensis with a main formation of CPD at TC than TT position both in vivo and in vitro, which led us to reinvestigate the topic of the impact of GC content on the formation of DNA damage (Chapter 2). Therefore, we concluded a favoured

173 Chapter 6

formation of mutagenic C-containing dinucleotide photoproducts in bacteria with increasing GC content (Chapter 2). This research presented here also suggested that higher temperature could increase the formation of bipyrimidine photoproducts in S. alaskensis (Chapter 4). The level of starvation experienced by the cells, would be a determining factor that would modulate the nature of UVB-induced bipyrimidine photoproducts, with a favoured formation of 6-4PPs at TC position than CDP for cells acclimated to long term starvation compared to cells entering stationary phase

(Chapter 4). A model of marine copiotrophic bacterium, Photobacterium angustum

S14, was exposed under the same conditions of UVB than S. alaskensis and was found to possess responses to damaging radiation that were distinctly different to those of S. alaskensis (Chapter 3). The work presented in this thesis investigated also the impact of UV response on the gene regulation at the translational level. We thus reported that while exponential cells (ML) actively responded to solar radiation by up-regulating a large set of proteins known to confer UV tolerance, starved cells

(SP) appeared to experience more passively the inferred stress, as shown by the overall lower number of proteomic changes for SP versus ML (Chapter 5). We also showed that an extended time of radiation did not necessarily lead to a higher set of differentially regulated proteins during stationary phase, although noteworthy qualitative changes were revealed (Chapter 5). Different strategies of responses appeared after a short time of exposure compared to extended time of irradiation in term of viability, DNA damage induction and protein regulation (Chapters 3 and 5).

Regarding the impact of the different spectral bands, it was interestingly found that the sole exposure of S. alaskensis to visible light (PAR) led to significant proteomic changes. As a reminder, the stress-related glutathione exhibited the largest up- regulation in PAR treatment, demonstrating that, constant detoxification of either

174 Chapter 6

ROS or glyoxal is required even under visible light. Diverse scavenging mechanisms against toxic oxidative compounds were successfully identified in S. alaskensis under solar radiation (Chapter 5).

II. Comparative genome analysis: a tool to predict UV response

The multitude of completed genome sequences that are now accessible has opened up a new field of research in comparative genomics and is revolutionizing our understanding of marine microbial ecology. Multi species genome comparison is a useful tool for exploring the relationship between genotype, phenotype and environment of diverse marine microbial species.

The Fig. 6.1 presents the number of complete chromosomes, organelles and plasmids as well as draft genome assemblies obtained from the Genbank database, organized in six major organism groups: Archaea, Bacteria, Eukaryota, Viruses, Viroids, and

Plasmids.

Viruses (2074)

Eukaryota (1694)

Bacteria (987)

Archaea (71)

Viroids (39)

Plasmids (38)

Figure 6.1. Distribution of the total number (4903) of fully sequenced genomes accessible to date in the Genbank database.

The tremendous power of comparative genomics is well illustrated with example of the marine bacterium Silicibacter pomeroyi, member of the Roseobacter clade.

175 Chapter 6

Indeed, its genome analysis revealed the presence of enzymes encoding for the oxidation of reduced inorganic coumpounds (e.g., carbon monoxide and sulphide) and laboratory experiments confirmed that carbon monoxide and reduced sulphur are important substrates for S. pomeroyi metabolism (Moran et al., 2004). Thus, comparison of DNA sequences from different species could be an extremely efficient way to identify new/original functional genes.

The Integrated Microbial Genomes (IMG) system (img.jgi.doe.gov) provides graphical viewers that enable comparison of the relative abundance of genes harbouring specific functions across multi-species genomes (Markowitz et al., 2008).

Therefore, this software would allow us to visualize the number of protein/functional families and extract biological meaning of S. alaskensis by comparing it with other genomes of interest, such as for instance some model of oligotrophic (“Candidatus

Pelagibacter ubique”) and copiotrophic (P. angustum) bacterium or microorganisms known to be highly resistant to damaging radiation (Deinococcus radiodurans) (Fig.

6.2). For example, it appears that Photobacterium angustum possesses a higher number of genes encoding for DNA-binding proteins as well as stress related proteins than the four other bacteria considered and S. alaskensis presents a large functional cluster related to iron transport (Fig. 6.2). This kind of figure illustrates how comparative genomics can help to formulate biochemical and physiological hypotheses that can be tested in the laboratory and/or in the field.

Finally, the detailed sequence-based genomic datasets can also provide informatics tools for building molecular probes in order to better track and quantify specific microorganisms and microbial processes.

176 Chapter 6 R1 RB2256 S14 Genes abundance Pelagibacter ubique HTCC1062 Pelagibacter ubique HTCC1002 9  0 5 10 13 >15 13 10 0 5 Deinococcus radiodurans Sphingopyxis alaskensis Vibrio angustum Candidatus Candidatus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
'       /1 4   %    % *  " 5!6   ( 9!  &# ' '  )% '!" . 0!'% ' 1 2% '   " ' 6 7 ?  1 0  '   '   ' =!'  '  #  .' '    '   #1   ' -   '    &&       '   '   '%! ' +,   ' '&/ '    %  ;48 ' 1 4   %    % " 5!6   ' 0!'%   (  !  $        '  ")    ' 7 '   '  -  '   / .%     '  !  '.%   ;" < '    (( '     '    ! ' "'%    &1 $   %  

Figure 6.2. Extract of an “Abundance Profile Map” comparing the genomes of “Ca. Pelagibacter ubique” (1-2), Deinococcus radiodurans (3) Sphingopyxis alaskensis RB2256 (4) and Photobacterium angustum S14 (5). Each column on the map corresponds to a genome, and each row corresponds to a gene. CAUTION: This figure shows only those genes that vary most in the five genomes.

177 Chapter 6

III. Distinct strategies of UV response

S. alaskensis and P. angustum showed different behaviours towards UV radiation with distinct response depending on the time of exposure considered. While S. alaskensis did present a higher UVB-resistance (both in term of viability and DNA damage) compare to Photobacterium angustum for a short time of irradiation, this trend was completely reversed for an extended exposure to UVB. Thus, both model of oligotrophic and copiotrophic bacterium would possess distinct adaptative strategies to solar radiation.

III.1 S. alaskensis

The present thesis first aimed at accurately determining the complete distribution of

UV induced photoproducts in S. alaskensis and thus the previous conclusion of it

UVB resistant mediated by low level of TT CPD photoproducts was reassessed. S. alaskensis was found to accumulate TC CPDs under UVB, due to its high GC content and thus for this reason would be more prone to mutagenesis than other bacteria with low GC content. Regarding to its ability to remove DNA lesions, S. alaskensis seemed to use efficiently photoreactivating light to repair the CPD lesions either simultaneously or subsequently to UVB radiation. Interestingly, although S. alaskensis presented quite efficient PER systems, no significant evidence of over- expression of proteins directly involved in DNA repair (i.e., photolyase) were observed in the proteomics approach. This result suggests that key enzyme involved in DNA repair could rather be constitutively expressed, irrespectively to the conditions of irradiation. Contrastingly, elevated cellular levels of chaperones and proteins associated to protection against oxidative stress were characterized.

Therefore, S. alaskensis would preferentially generate a general stress protective

178 Chapter 6

mechanism, as it was previously described following nutrient limitation (Ostrowski et al., 2004). The overall results of DNA damage and proteomics imply that S. alaskensis has selected a strategy of protection/repair, rather than replacement of cellular components, susceptible to be damaged under solar radiation. Additionally, the replacement of damaged cellular components by de novo synthesis would be rather limited by resource availability encountered in oligotrophic environment, thus suggesting the importance of oligotrophic bacteria to evolve efficient strategies of protection against oxidative stress, commonly faced by microorganisms in aquatic environments (Cooper et al., 1983; Gourmelon et al., 1994; Schut et al., 1997;

Ostrowski et al., 2001). However, those established strategies would not be sufficient to survive long term irradiation even in presence of photoreactivating light, since the accumulation of CPD and 6-4PP lesions overtime would fatally lead S. alaskensis to cell death.

III.2 P. angustum

Unlike oligotrophic bacteria and in comparison to S. alaskensis, P. angustum presented a strategy of UV tolerance completely different.

It is well established that copiotrophic bacteria are geared for carbon starvation in oligotrophic waters, and are likely to survive in such nutrient limited environment, by producing stress resistance, miniaturisation and resting stage cells (Nyström et al.,

1992). It was previously reported that carbon starved cells of P. angustum S14 became gradually less sensitive to potentially harmful stresses, with a maximal protection after about 40 h of starvation (Nyström et al., 1992). In addition, a recent study showed that growing cells of P. angustum exposed to solar radiation

179 Chapter 6

accumulate very low amount of CPD lesions (Abboudi et al., 2008). In this thesis, we demonstrated that P. angustum presented a strong resistance to long-term UVB radiation for cells entering stationary phase, as it was evident that an inflection point occured at about 2 kJ m-² in the dose response curves. Taken together, P. angustum would be a very resistant marine bacterium to UVB radiation in both growing and non-growing phases and would have evolved some puzzling strategies that we will discuss below.As previously described in the thesis, the photoresistance strategy of

P. angustum could be explained by the dual role of photolyase, allowing both DNA repair through the well-known photochemical process as well as enhancing the light independent repair pathway (NER). Since the photoreactivating light existing under

UVB lamps would be too low to promote efficient CPD removals, we thus suggested the “dark function” of photolyase.

UVB

Residual hv (360 nm)

«Dark function» «Light function» of Photolyase of Photolyase

hv (360 nm)

DNA DNA damage damage Photolyase Photolyase

+ ABC nuclease

NER PER

Resistance of P. angustum

Figure 6.3. The dual role of photolyase in P. angustum S14.

180 Chapter 6

Considering the fact that NER pathway is ATP dependent, and that UVB exposure was conducted with a carbon free medium, we could suggest that P. angustum would possess some forms of nutrient storage and/or would have evolved alternative pathways to produce ATP. Interestingly, according its genome, P. angustum harbours genes encoding for archaeal homolog bacteriorhodopsin, known to function as a light-driven proton pump. Genes of proteorhodopsin variants (homolog bacteriorhodopsin) have been identified in samples from the Mediterranean and Red

Seas and the Sargasso Sea and the Sea of Japan (Béjà et al., 2001). As previously reported, genetically modified E. coli with incorporated proteorhodopsin was able to pump protons in the presence of light (Béjà et al., 2000). It was further demonstrated that the proton gradient generated by proteorhodopsin could be used to generate ATP

(Martinez et al., 2007).

In order to explore further the mechanism of photoresistance of P. angustum, a quantitative proteomic approach could be used to determine the expression level of photolyase and bacteriorhodopsin genes under UVB stress.

IV. Impact of nutrient limitation on protein expression of S. alaskensis

The examination of protein identification derived from MS/MS profiles showed that the majority of the detected proteome was similar between mid-logarithmic and stationary phase (ML and SP). This observation confirms the conclusion of a recent study performed on “Ca. Pelagibacter ubique”, showing that most of the metabolic processes still remained active in early stationary phase (Sowell et al., 2008).

Proteomics approach enabled us to characterize carbon starvation-induced proteins.

181 Chapter 6

We successfully identified unique proteins of stationary phase that could be well related to their physiological state of growth, such as prophage-related genes, antibiotic biosynthesis monoxygenase, signal transduction response regulator and transposases. Indeed, for example transpositional activity can be induced by different environmental stresses, such like carbon starvation (Lamrani et al., 1999). It is also well-known that antibiotic biosynthesis in bacteria is generally elicited as a physiological response to a variety of environmental stimuli including nutritional imbalance and/or presence of stress-inducing agents (Bystrykh et al., 1996; Doull et al., 1990). Among the unique proteins of stationary phase, only one associated stress response protein (Stress response protein Y) also came out within the iTRAQ- datasets, thus suggesting a dual role of this protein in both UV response and stationary phase entrance. Carbon limitation experienced by S. alaskensis at the beginning of stationary phase lead to the expression of a specific set of proteins and as demonstrated in the thesis could modulate the UV response. We therefore also addressed the notion that a carbon starvation-specific signal transduction pathway, may exist in S. alaskensis.

V. Future prospects

As reported in this thesis, proteomics is a very well suitable approach to investigate the detection of potential biomarkers candidates associated to a UV response. In order to complete this work, it could be valuable to perform some other experiments in the proteomics field. First of all, abundance changes of some proteins of interest, observed with the iTRAQ methodology, could be confirmed by using another independent assay such as Western Blot. However among the potential biomarkers that we wish to control, only a small fraction of proteins would probably have

182 Chapter 6

commercially available antibodies. In addition to assess the question of protein activity it would be necessary to perform enzymatic assays for some proteins of interest.

Protein identification from individual conditions of irradiation would enable with a search engine such as Scaffold software to detect unique proteins, only expressed upon a specific spectral band, similarly as it was obtained for the two growth conditions (ML and SP).

The recent introduction of 8-plex iTRAQ reagents enabling to perform relative quantitation of up to eight samples in a single LC-MS/MS run, would allow us to multiply sample replicates to either improve statistically data obtained or exploring other conditions (such as different UV irradiation sources, different growth conditions, temperatures). For a better representativeness of UV response, hydrophobic protein extractions would enhance the proteome coverage of S. alaskensis. Secretome of S. alaskensis (i.e, extracellular, outer membrane proteins and proteins exported to the periplasm) could give us some new insights on cell-to- cell communication under UV stress.

Finally, since UV mediated oxidative stress cause irreversible oxidative modification to proteins it would be very interesting to focus on the qualitative and quantitative induction of protein carbonylation. Detection of carbonyl-modified proteins was achieved most commonly using hydrazide-based chemistry. Hydrazides react specifically with proteins carbonyls via Schiff base formation, which can be further stabilized by reduction with sodium cyanoborohydride (Yan et al., 1998). Recently, a

183 Chapter 6

new strategy for the identification of carbonylated proteins from complex mixture was established (Meany et al., 2007). This new approach combines biotin hydrazide labelling of protein carbonyl groups, avidin affinity chromatography and multiplexed iTRAQ reagent stable isotope labeling. This strategy would allow us to distinguish and quantify carbonylated proteins across a set of different conditions and furthermore would present the strong advantage of characterizing the site of carbonylation. Focusing on those proteins modifications induced by UV radiation would be a very original approach since this methodology was only described very recently and would give us new insight on the overall datasets that we obtained with iTRAQ. This work is at the present time underway.

VI. Concluding remarks

This research demonstrated the effects of solar radiation by recognizing potential synergistic relationships between environmental factors, such as temperature, nutrient limitation that may modulate the harmful affects of damaging radiation. The knowledge obtained through this research led us to better estimate the impact of global change on a numerically dominant bacterium S. alaskensis. This study also showed that the response to UV may be governed by a variety of protective strategies that are species dependent. Clearly more in-depths physiological and molecular analysis of environmentally important isolates, will allow to link the roles of physiologically relevant protein markers with the UV stress.

184 REFERENCES Abboudi M., S. Matallana-Surget, J. F. Rontani, R. Sempéré and F. Joux, Physiological alteration of the marine bacterium Vibrio angustum S14 exposed to simulated sunlight during growth, Curr. Microbiol., 2008, 57, 412–417.

Agogué H., F. Joux, I. Obernosterer and P. Lebaron, Resistance of marine bacterioneuston to solar radiation, Appl. Environ. Microbiol., 2005, 71, 5282– 5289.

Aki T. and S. Adhya, Repressor induced site-specific binding of HU for transcriptional regulation, EMBO J., 1997, 16, 3666–3674.

Aldsworth T. G., R. L. Sharman and C. E. R. Dodd, Bacterial suicide through stress, Cell. Mol. Life Sci., 1999, 56, 378–383.

Allen M., F. M. Lauro, T. J. Williams, D. Burg, K. S. Siddiqui, D. De Francisci, et al. The genome sequence of the psychrophilic archaeon, Methanococcoides burtonii: the role of genome evolution in cold-adaptation. ISME J (in press) 2009.

Amann R. and W. Ludwig, Ribosomal RNA-targeted nucleic acid probes for studies in microbial ecology, FEMS Microbiol. Rev., 2000, 24, 555–565.

Amann R. I., W. Ludwig and K. H. Schleifer, Phylogenetic identification and in situ detection of individual microbial cells without cultivation, Microbiol. Rev. 1995, 59, 143–169.

Ames B. N. and L. Swirsky Gold, Endogenous mutagens and the causes of aging and cancer, Mut. Res., 1991, 250, 3–16.

Andrews J. H. and R. F. Harris, r- and K-selection and microbial ecology, Adv. Microb. Ecol., 1986, 9, 99–147.

Andrews S. C., Iron storage in bacteria, Adv. Microb. Physiol., 1998, 40, 281–351.

Anesio A. M., W. Granéli, G. R. Aiken, D. J. Kieber and K. Mopper, Effect of humic substance photodegradation on bacterial growth and respiration in lake water, Appl. Environ. Microbiol., 2005, 71, 6267–6275.

Applegate L. A., C. Scaletta, R. Panizzon and E. Frenk, Evidence that ferritin is UV inducible in human skin: Part of a putative defense mechanism, J. Invest. Dermatol., 1998, 111, 159–163.

Arrieta J. M., M. G. Weinbauer and G. J. Herndl, Interspecific differences in sensitivity to UV radiation and subsequent recovery in selected isolates of marine bacteria, Appl. Environ. Microbiol., 2000, 66, 1468–1473.

Augusto-Pinto L., C. G. Regis da Silva, D. de Oliveira Lopes, A. Machado-Silva and C. R. Machado, Escherichia coli as a model system to study DNA repair genes of eukaryotic organisms. Genet. Mol. Res., 2003, 2, 77–91.

186 Azam F., T. Fencel, J. G. Field, J. S. Gray, T. A. Meyer-Reil and F. Thingstad, The ecological role of water column microbes in the sea, Mar. Ecol. Prog. Ser., 1983, 10, 257–263.

Azam F. and F. Malfatti, Microbial structuring of marine ecosystems, Nat. Rev. Micro., 2007, 5, 782–791.

Bak L., J. F. Atkins, C. E. Singer and B. N. Ames, Evolution of DNA base compositions in microorganisms, Science, 1972, 175, 1391–1393.

Baliga N. S., S. J. Bjork, R. Bonneau, M. Pan, C. Iloanusi, M. C. H. Kottemann, L. Hood and J. DiRuggiero, Systems level insights into the stress response to UV radiation in the halophilic archaeon Halobacterium NRC-1, Genome Res., 2004, 14, 1025–1035.

Ballesteros M., A. Fredriksson, J. Henriksson and T. Nyström, Bacterial senescence: protein oxidation in non-proliferating cells is dictated by the accuracy of the ribosomes, EMBO J., 2001, 20, 5280–5289.

Bauer J. E., P. M. Williams and E. R. M. Druffel, 14C activity of dissolved organic carbon fractions in the north-central Pacific and Sargasso Sea, Nature, 1992, 357, 667–670.

Beck B. D., P. G. Arscott and A. Jacobson, Novel properties of bacterial elongation factor Tu, Proc. Natl. Acad. Sci. USA, 1978, 75, 1250–1254.

Béjà O., L. Aravind, E. V. Koonin, M. T. Suzuki, A. Hadd, L. P. Nguyen, S. B. Jovanovich, C. M. Gates, R. A. Feldman, J. L. Spudich, E. N. Spudich and E. F. DeLong, Bacterial rhodopsin: evidence for a new type of phototrophy in the sea, Science, 2000, 289, 1902–1904.

Béjà O., E. N. Spudich, J. L. Spudich, M. Leclerc and E. F. DeLong, Proteorhodopsin phototrophy in the ocean, Nature, 2001, 411, 786–789.

Belzile C., S. Demers, G. A. Ferreyra, I. Schloss, C. Nozais, K. Lacoste, B. Mostajir, S. Roy, M. Gosselin, E. Pelletier, S. M. F. Gianesella and M. Vernet, UV effects on marine planktonic food-webs: a synthesis of results from mesocosm studies, Photochem. Photobiol., 2006, 82, 850–856.

Ben-Ishai R., E. Ben-Hur and Y. Hornfeld, Photosensitized dimerization of thymine and cytosine in DNA, Isr. J. Chem., 1968, 6, 769– 775.

Berney M., H.-U. Weilenmann, J. Ihssen, C. Bassin and T. Egli, Specific growth rate determines the sensitivity of Escherichia coli to thermal, UVA, and solar disinfection, Appl. Env. Microbiol., 2006, 72, 2586–2593.

Berrocal-Tito G., L. Sametz-Baron, K. Eichenberg, B. A. Horwitz, and A. Herrera- Estrella, Rapid Blue Light Regulation of a Trichoderma harzianum Photolyase Gene, J. Biol. Chem., 1999, 274, 14288–14294.

187 Bertoni R. and C. Callieri, Effects of UVB radiation on freshwater autotrophic and heterotrophic picoplankton in a subalpine lake, J. Plankton Res., 1999, 21, 1373– 1388.

Birrell G. W., J. A. Brown, H. I. Wu, G. Giaever, A. M. Chu, R.W. Davis and J. M. Brown, Transcriptional response of Saccharomyces cerevisiae to DNA-damaging agents does not identify the genes that protect against these agents, Proc. Natl. Acad. Sci. USA, 2002, 99, 8778–8783.

Bjørnsen P. I., Automatic determination of bacterioplankton biomass by image analysis, Appl. Environ. Microbiol., 1986, 51, 1199–1204.

Blagoev K. B., B. S. Alexandrov, E. H. Goodwin and A. R. Bishop, Ultra-violet light induced changes in DNA dynamics may enhance TT-dimer recognition, DNA repair, 2006, 5, 863–867.

Boelen P., A. F. Post, M. J. W. Veldhuis and A. G. J. Buma, Diel patterns of UVBR- induced DNA damage in picoplankton size fractions from the Gulf of Aqaba, Red Sea, Microb. Ecol., 2002, 44, 164–174.

Bopp L., P. Monfray, O. Aumont, J. L. Dufresne, H. Le Treut, G. Madec, L. Terray and J. C. Orr, Potential impact of climate change on marine export productions, Global Biogeochem. Cycles, 2001, 15, 81–99.

Børsheim K. Y., G. Bratback and M. Heldal, Enumeration and biomass estimation of planktonic bacteria and viruses by transmission electron microscopy, Appl. Environ. Microbiol., 1990, 56, 352–356.

Bosca F., V. Lhiaubet-Vallet, M. C. Cuquerella, J. V. Castell and M. A. Miranda, The triplet energy of thymine in DNA, J. Am. Chem. Soc., 2006, 128, 6318–6319.

Bota D. A. and K. J. Davies, Lon protease preferentially degrades oxidized mitochondrial aconitase by an ATP-stimulated mechanism, Nature Cell. Biol., 2002, 4, 674–680.

Boubrik F. and J. Rouvière-Yaniv, Increased sensitivity to gamma-irradiation in bacteria lacking protein HU, Proc. Natl Acad. Sci. USA, 1995, 92, 3958–3962.

Bramhill D. and A. Kornberg, A model for initiation at origins of DNA replication, Cell, 1988, 54, 915–918.

Brash D. E., J. A. Rudolph, J. A. Simon, A. Lin, G. J. McKenna, H. P. Baden, A. J. Halperin and J. Ponten, A role for sunlight in skin cancer: UV-induced p53 mutations in squamous cell carcinoma, Proc. Natl. Acad. Sci. USA, 1991, 88, 10124–10128.

Buck J., Effects of medium composition on the recovery of bacteria from seawater, J. Exp. Mar. Biol., 1974, 15, 25–34.

Buechter D. D., Free radicals and oxygen toxicity, Pharm. Res., 1988, 5, 253–260.

188 Buma A. G. J., P. Boelen and W. H. Jeffrey, UVR-induced DNA damage in aquatic organisms, in UV effects in aquatic organisms and ecosystems, ed. W. Helbling and H. Zagarese, Comprehensive series in Photosciences, The Royal Society of Chemistry, Cambridge, UK, 2003, pp. 291–327.

Button D. K., Isolation of typical marine bacteria by dilution culture: growth, maintenance, and characteristics of isolates under laboratory conditions, Appl. Environ. Microbiol., 1993, 59, 2150–2160.

Button D. K. and B. Robertson, Determination of DNA content of aquatic bacteria by flow cytometry, Appl. Environ. Microbiol., 2001, 67, 1636–1645.

Bystrykh L., M. Fernandez-Moreno, J. Herrema, F. Malpartida, D. Hopwood and L. Dijkhuizen, Production of actinorhodin-related "blue pigments" by Streptomyces coelicolor A3(2), J Bacteriol., 1996, 178, 2238–2244.

Cadet J., M. Berger, T. Douki and J. L. Ravanat, Oxidative damage to DNA: formation, measurement and biological significance, Rev. Physiol. Biochem. Pharmacol., 1997, 131, 1–87.

Cadet J., E. Sage and T. Douki Ultraviolet radiation-mediated damage to cellular DNA. Mut. Res., 2005, 571, 3–17.

Cass A. E. G., Superoxide dismutases, Top. Mol. Struct. Biol., 1985, 6, 121–156.

Cavicchioli R., M. Ostrowski, F. Fegatella, A. Goodchild and N. Guixa-Boixereu, Life under nutrient limitation in oligotrophic marine environments: and eco- physiological perspective of Sphingopyxis alaskensis (formerly Sphingomonas alaskensis), Microb. Ecol., 2003, 45, 203–217.

Clarke K. R. Non-parametric multivariate analyses of changes in community structure. Aus. J. Ecol. 1993, 18, 117–143.

Cole J. J., S. Findlay and M. L. Pace, Bacterial production in fresh and saltwater ecosystems: a cross-system overview, Mar. Ecol. Prog. Ser., 1998, 43, 1–10.

Connon S. A. and S. J. Giovannoni, High-throughput methods for culturing microorganisms in very low nutrient media yield diverse new marine isolates, Appl. Environ. Microbiol., 2002, 68, 3878–3885.

Conrads T. P., K. Alving, T. D. Veenstra, M. E. Belov, G. A. Anderson, D. J. Anderson, M. S. Lipton, L. Pasa-Tolic, H. R. Udseth, W. B. Chrisler, B. D. Thrall and R. D. Smith, Quantitative analysis of bacterial and mammalian proteomes using a combination of cysteine affinity tags and 15N-metabolic labeling, Anal. Chem., 2001, 73, 2132–2139.

Cooke M. S., M. D. Evans, M. Dizdaroglu and J. Lunec, Oxidative DNA damage: mechanisms, mutation, and disease, FASEB Journal, 2003a, 17, 1195–1214.

189 Cooke M. S., I. D. Podmore, N. Mistry, M. D. Evans, K. E. Herbert, H. R. Griffiths and J. Lunec, Immunochemical detection of UV-induced DNA damage and repair, J. Immunol. Methods, 2003b, 280, 125–133.

Cooper S., Checkpoints and restriction points in bacteria and eukaryotic cells, BioEssays, 2006, 28, 1035–1039.

Cooper W. J., and R. G. Zika, Photochemical formation of hydrogen peroxide in surface and ground waters exposed to sunlight, Science, 1983, 220, 711–712.

Courcelle J., A. Khodursky, B. Peter, P. O. Brownand and P. C. Hanawalt, Comparative gene expression profiles following UV exposure in wild-type and SOS-deficient Escherichia coli, Genetics, 2001,158, 41–64.

Courdavault S., C. Baudouin, S. Sauvaigo, S. Mouret, S. Candéias, M. Charveron, A. Favier, J. Cadet and T. Douki, Unrepaired cyclobutane pyrimidine dimers do not prevent proliferation of UVB-irradiated cultured human fibroblasts, Photochem. Photobiol., 2004, 79, 145–151.

Cox P. M., R. A Betts, C. D. Jones, S. A. Spall and I. J. Totterdell, Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model, Nature, 2000, 408, 184–187.

Crowley D. J. and P. C. Hanawalt, Induction of the SOS response increases the efficiency of global nucleotide excision repair of cyclobutane pyrimidine dimers, but not 6–4 photoproducts, in UV-irradiated Escherichia coli, J. Bacteriol., 1998, 180, 3345–3352.

Dalle-Donne I., D. Giustarini, R. Colombo, R. Rossi and A. Milzani, Protein carbonylation in human diseases, Trends Mol. Med., 2003, 9, 169–176.

Dantas Caldas T., A. El Yaagoubi and G. Richarme, Chaperone properties of bacterial elongation Factor EF-Tu, J. Biol. Chem., 1998, 273, 11478–11482.

Davison I., Environmental effects on algal photosynthesis: temperature- a minireview, J. Phycol., 1991, 27, 2–8.

Demidov V. V. and V. N. Potaman, HPLC photofingerprinting of conformational pecularities and transitions in oligonucleotide duplexes, Nucleic Acids Res., 1993, 21, 2691–2696.

Desnues B., G. Gregori, S. Dukan, H. Aguilaniu and T. Nyström, Differential oxidative damage and expression of stress regulons in culturable and nonculturable cells of Escherichia coli, EMBO Rep., 2003, 4, 400–440.

Dizdaroglu M., Mechanisms of free radical damage to DNA, in DNA & Free Radicals: Techniques, Mechanisms & Applications, ed. O. I. H. Aruoma, 1998, pp. 3–26.

Doetsch P. W., Translesion synthesis by RNA polymerases: occurrence and biological implications for transcriptional mutagenesis, Mut. Res., 2002, 510, 131–140.

190 Douki T. and J. Cadet, Far-UV photochemistry and photosensitization of 2'- deoxycytidylyl-(3'-5')-thymidine: isolation and characterization of the main photoproducts. J. Photochem. Photobiol. B:Biol., 1992, 15, 199–213.

Douki T., L. Voituriez and J. Cadet, Measurement of pyrimidine (6–4) pyrimidone photoproducts in DNA by a mild acidic hydrolysis-HPLC fluorescence assay, Chem. Res. Toxicol., 1995, 8, 244–253.

Douki T., T. Zalizniak and J. Cadet, Far-UV-induced dimeric photoproducts in shorts oligonucleotides: sequence effects, Photochem. Photobiol., 1997, 66, 171–179.

Douki T., M. Court, S. Sauvaigo, F. Odin and J. Cadet, Formation of the main UV induced thymine dimeric lesions within isolated and cellular DNA as measured by high performance liquid chromatography-tandem mass spectrometry. J. Biol. Chem., 2000, 275, 11678–85.

Douki T. and J. Cadet, Individual determination of the yield of the main-UV induced dimeric pyrimidine photoproducts in DNA suggests a high mutagenicity of CC photolesions, Biochemistry, 2001, 40, 2495– 2501.

Douki T., D. Angelov and J. Cadet, UV Laser photolysis of DNA: effect of duplex stability on charge-transfer efficiency, J. Am. Chem. Soc., 2001, 123, 11360– 11366.

Douki T. and J. Cadet, Formation of the spore photoproduct and other dimeric lesions between adjacent pyrimidines in UVC-irradiated dry DNA, Photochem. Photobiol. Sci., 2003, 2, 433–436.

Douki T., A. Reynaud-Angelin, J. Cadet and E. Sage, Bipyrimidine photoproducts rather than oxidative lesions are the main type of DNA damage involved in the genotoxic effect of solar UVA radiation, Biochemistry, 2003, 42, 9221–9226.

Douki T., Effect of denaturation on the photochemistry of pyrimidine bases in isolated DNA, J. Photochem. Photobiol., B, 2006a, 82, 45–52.

Douki T., Low ionic strength reduces cytosine photoreactivity in UVC irradiated isolated DNA, Photochem. Photobiol. Sci., 2006b, 5, 1045–1051.

Doull J. L. and L. C. Vining, Nutritional control of actinorhodin production by Streptomyces coelicolor A3(2): suppressive effects of nitrogen and phosphate, Appl. Microbiol. Biotechnol., 1990, 32, 449–454.

Dri A. M., P. Moreau and J. Rouviere-Yaniv, Involvement of the histone-like proteins OsmZ and HU in homologous recombination, Gene, 1992, 120, 11–16.

Drouin R., S. Gao and G. P. Holmquist, Technologies for Detection of DNA Damage and Mutations, ed. G. P. Pfeifer, Plenum Press, 1996, pp. 37–43.

191 Drouin R. and J. P. Therrien, UVB-induced cyclobutane pyrimidine dimer frequency correlates with skin cancer mutational hotspots in p53, Photochem. Photobiol., 1997, 66, 719–726.

Duarte C. M. and S. Agusti, The CO2 balance of unproductive aquatic ecosystems, Science, 1998, 281, 234–236.

Ducklow H. W., Bacterioplankton production and biomass in the oceans, in Microbial Ecology of the Oceans, ed. D. Kirchman, New York, Wiley, 2000, 4, pp. 85–120.

Dukan S., A. Farewell, M. Ballesteros, F. Taddei, M. Radman and T. Nyström, Protein oxidation in response to increased transcriptional or translational errors, Proc. Natl. Acad. Sci. USA, 2000, 97, 5746–5749.

Eguchi M., T. Nishikawa, K. MacDonald, R. Cavicchioli, J. C. Gottschal and S. Kjelleberg, Responses to stress and nutrient availability by the marine ultramicrobacterium Sphingomonas sp. strain RB2256, Appl. Environ. Microbiol., 1996, 62, 1287–1294.

Eguchi M., M. Ostrowski., F. Fegatella., J. Bowman., D. Nichols., T. Nishino and R. Cavicchioli, Sphingomonas alaskensis strain AFO1, an abundant oligotrophic ultramicrobacterium from the North Pacific, Appl. Environ. Microbiol., 2001, 67, 4945–4954.

Ehling-Schulz M., S. Schulz, R. Wait, A. Görg and S. Scherer, The UV-B stimulon of the terrestrial cyanobacterium Nostoc commune comprises early shock proteins and late acclimation proteins, Mol. Microbiol., 2002, 46, 827–843.

Eilers H., J. Pernthaler, F. O. Glockner and R. Amann, Culturability and in situ abundance of pelagic bacteria from the North Sea, Appl. Environ. Microbiol., 2000, 66, 3044–3051.

Eker A. P. M., P. Kooiman, J. K. C. Hessels and A. Yasui, DNA photoreactivating enzyme from the cyanobacterium Anacystis nidulans, J. Biol.Chem., 1990, 14, 8009–8015.

Ellis R. J. and S. M. van der Vies, Molecular chaperones, Annu. Rev. Biochem., 1991, 60, 321–347.

Emanuele E., K. Zakrzewska, D. Markovitsi, R. Lavery and P. Milli´e, Exciton states of dynamic DNA double-helices: alternating dCdG sequences, J. Phys. Chem. B, 2005a, 109, 16109–16118.

Emanuele E., D. Markovitsi, P.Millie and K. Zakrzewska, UV spectra and excitation delocalization in DNA: influence of the spectral width, ChemPhysChem, 2005b, 6, 1387–1392.

Evan M. D. and M. S. Cooke, Factors contributing to the outcome of oxidative damage to nucleic acids, Bioessays, 2004, 26, 533–542.

192 Falkowski P. G. and C.de Vargas, Shotgun sequencing in the sea: a blast from the past?, Science, 2004, 304, 58–60.

Farman J. C., B. G. Gardiner and J. D. Shanklin, Large losses of total ozone in Antarctica reveal seasonal ClO x /NO x interaction, Nature, 1985, 315, 207–210.

Fegatella F., J. Lim, S. Kjelleberg and R. Cavicchioli, Implications of rRNA operon copy number and ribosome content in the marine oligotrophic ultramicrobacterium Sphingomonas sp. strain RB2256, Appl. Environ. Microbiol., 1998, 64, 4433–4438.

Fegatella F., M. Ostrowski and R Cavicchioli, An assessment of protein profiles from the marine oligotrophic ultramicrobacterium, Sphingomonas sp. strain RB2256, Electrophoresis, 1999, 20, 2094–2098.

Fegatella F. and R. Cavicchioli, Physiological responses to starvation in the marine oligotrophic ultramicrobacterium Sphingomonas sp. strain RB2256, Appl. Env. Microbiol., 2000, 66, 2037–2044.

Ferguson R. L., E. N. Buckley and A. V. Palumbo, Response of marine bacterioplankton to differential filtration and confinement, Appl. Environ. Microbiol., 1984, 47, 49–55.

Ferreyra G. A., B. Mostajir, I. R. Schloss, K. Chatila, M. E. Ferrario, P. Sargian, S. Roy, J. Prod’homme and S. Demers, Ultraviolet-B radiation effects on the structure and function of lower trophic levels of the marine planktonic food web, Photochem. Photobiol., 2006, 82, 887–897.

Fisher G. J. and H. E. Johns, Pyrimidine photodimers, in Photochemistry and Photobiology of Nucleic Acids, ed. S. Wang, Academic Press, New York, 1976, 1, pp. 225–294.

Franklin W. A., P. W. Doetsch and W. A. Haseltine, Structural determination of the ultraviolet light-induced thymine-cytosine pyrimidinepyrimidone (6–4) photoproduct, Nucleic Acids Res., 1985, 13, 5317– 5325.

Freeman K. B., P. V. Hariharan and H. E. Johns, The ultraviolet photochemistry of cytidylyl-(3'-5')-cytidine, J. Mol. Biol., 1965, 13, 833–848.

Freeman S. E., A. D. Blackett, D. C. Monteleone, R. B. Setlow, B. M. Sutherland and J. C. Sutherland, Quantitation of radiation-, chemical-, or enzyme-induced single strand breaks in nonradioactive DNA by alkaline gel electrophoresis: Application to pyrimidine dimers. Anal. Biochem. 1986, 158, 119–129.

Friedberg E. C., G. C. Walker and W. Siede, in DNA repair and Mutagenesis, ASM Press Washington, 1995.

Friedberg E. C., DNA damage and repair, Nature, 2003, 421, 436–440.

193 Fuhrman J. A., T. D. Sleeter, C. A. Carlson and L. M. Proctor, Dominance of bacterial biomass in the Sargasso Sea and its ecological implications, Mar. Ecol. Prog. Ser., 1989, 57, 207–217.

Gajewsji E., G. Rao, Z. Nackerdien, M. Dizdaroglu, Modification of DNA bases in mammalian chromatin by radiation-generated free radicals, Biochemistry, 1990, 29, 7876–7882.

Gan C. S., P. K Chong., T. K. Pham and P. C. Wright, Technical, Experimental, and biological variations in isobaric tags for relative and absolute quantitation (iTRAQ), J. Proteome Res., 2007, 6, 821–827.

Garcia-Pichel F., A model for internal self-shading in planktonic organisms and its implications for the usefulness of ultraviolet sunscreen, Limnol. Oceanogr., 1994, 39, 1704–1717.

Gardy J. L., M. R. Laird, F. Chen, S. Rey, C. J. Walsh, M. Ester and F. S. L. Brinkman, PSORTb v.2.0: Expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis, Bioinformatics, 2005, 21, 617–623.

Gascon J., A. Oubina, A. Perez-Lezaum and J. Urmeneta, Sensitivity of selected bacterial species to UV radiation, Curr. Microbiol., 1995, 30, 177–182.

Gasol J. M., U. L. Zweifel, F. Peters, J. A. Fuhrman and A. Hagström, Significance of size and nucleic acid content heterogeneity as measured by flow cytometry in natural planktonic bacteria, Appl. Environ. Microbiol., 1999, 65, 4475–4483.

Giovannoni S. J., E. F. DeLong, G. J. Olsen and N. R. Pace, Phylogenetic group- specific oligodeoxynucleotide probes for identification of single microbial cells, J. Bacteriol., 1988, 170, 720–726.

Giovannoni S. J., T. B. Britschgi, C. L. Moyer and K. G. Field, Genetic diversity in Sargasso Sea bacterioplankton, Nature, 1990, 345 , 60–63.

Giovannoni S. and M. Rappé, Evolution, diversity and molecular ecology of marine prokaryotes, in Microbial ecology of the oceans, ed. D. L. Kirchman, Wiley-Liss, New York, NY, 2000, pp. 47–84.

Giovannoni S. J., H. J. Tripp, S. Givan, M. Podar, K. L. Vergin, D. Baptista, L. Bibbs, J. Eads, T. H. Richardson, M. Noordewier, M. S. Rappé, J. M. Short, J. C. Carrington and E. J. Mathur, Genome streamlining in a cosmopolitan oceanic bacterium, Science, 2005a, 19, 1242–1245.

Giovannoni S. J., L. Bibbs, J. C. Cho, M. D. Stapels, R. Desiderio, K. L. Vergin, M. S. Rappé, S. Laney, J. W. Lawrence, H. J. Tripp, E. J. Mathur and D. F. Barofsky, Proteorhodopsin in the ubiquitous marine bacterium SAR11, Nature, 2005b, 483, 82–85.

194 Giovannoni S. J. and U. Stingl, The importance of culturing bacterioplankton in the ‘omics’ age, Nature Rev., 2007, 5, 820–826.

Goldovsky L, I. Cases, A. J. Enright and C. A. Ouzounis, BioLayout(Java): versatile network visualisation of structural and functional relationships, Appl. Bioinformatics., 2005, 4, 71–74.

Goltekar R. C., K. P. Krishnan, M. J. B. D. De Souza, A. L. Paropkari and P. A. Loka Bharathi, Effect of carbon source concentration and culture duration on retrievability of bacteria from certain estuarine, coastal and offshore areas around peninsular India, Curr. Sci., 2006, 90, 103–106.

Goshe M. B. and R. D. Smith, Stable isotope-coded proteomic mass spectrometry, Curr. Opin. Biotechnol., 2003, 14, 101–109.

Gourmelon M., J. Cillard and M. Pommepuy, Visible light damage to Escherichia coli in seawater: oxidative stress hypothesis, J. Appl. Bacteriol., 1994, 77, 105–112.

Gourmelon M., D. Touati, M. Pommepuy and M. Cormier, Survival of Escherichia coli exposed to visible light in seawater: analysis of rpoS dependent effects, Can. J. Microbiol., 1997, 43, 1036–1043.

Greinert R., O. Boguhn, D. Harder, E. W. Breitbart, D. L. Mitchell and B. Volkmer, The dose dependence of cyclobutane dimer induction and repair in UVB- irradiated human keratinocytes, Photochem. Photobiol., 2000, 72, 701–708.

Grune T., K. Merker, G. Sandig and K. J. Davies, Selective degradation of oxidatively modified protein substrates by the proteasome, Biochem Biophys. Res. Commun. 2003, 305, 709–718.

Grune T., T. Jung, K. Merker and K. J. A. Davies, Decreased proteolysis caused by protein aggregates, inclusion bodies, plaques, lipofuscin, ceroid, and ‘aggresomes' during oxidative stress, aging, and disease, Int. J. Biochem. Cell. Biol., 2004, 36, 2519–2530.

Gygi S. P., B. Rist, S. A. Gerber, F. Turecek, M. H. Gelb and R. Aebersold, Quantitative analysis of complex protein mixtures using isotope-coded affinity tags, Nature Biotechnol., 1999, 17, 994–999.

Häder D. P. and R. P. Sinha, Solar ultraviolet radiation-induced DNA damage in aquatic organisms: potential environmental impact, Mut. Res., 2005, 571, 221–223.

Häder D. P., H. D. Kumar, R. C. Smith and R. C. Worrest, Effects of solar UV radiation on aquatic ecosystems and interactions with climate change, Photochem. Photobiol. Sci., 2007, 6, 267–285.

Hagström A., J. Pinhassi and U. L. Zweifel, Biogeographical diversity among marine bacterioplankton, Aquat. Microb. Ecol., 2000, 21, 231–244.

195 Hagström A., T. Pommier, F. Rohwer, K. Simu, W. Stolte, D. Svensson and U. L. Zweifel, Use of 16S ribosomal DNA for delineation of marine bacterioplankton species, Appl. Environ. Microbiol., 2002, 68, 3628–3633.

Hanawalt P. C., Controlling the efficiency of excision repair, Mutat. Res., 2001, 485, 3– 13.

Heck H. D., M. Casanova, and T. B. Starr, Formaldehyde toxicity–new understanding, Crit. Rev. Toxicol., 1990, 20, 397–426.

Heidelberg, J. F., I. T. Paulsen, K. E. Nelson, R. J. Gaidos, W. C. Nelson, T. D. Read, J. A. Eisen, R. Seshadri, N. Ward, B. Methe, R. A. Clayton, T. Meyer, A. Tsapin, J. Scott, M. Beanan, L. Brinkac, S. Daugherty, R. T. DeBoy, R. J. Dodson, A. S. Durkin, D. H. Haft, J. F. Kolonay, R. Madupu, J. D. Peterson, L. A. Umayam, O. White, A. M. Wolf, J. Vamathevan, J. Weidman, M. Impraim, K. Lee, K. Berry, C. Lee, J. Mueller, H. Khouri, J. Gill, T. R. Utterback, L. A. McDonald, T. V. Feldblyum, H. O. Smith, J. C. Venter, K. H. Nealson and C. M. Fraser, Genome sequence of the dissimilatory metal ion-reducing bacterium Shewanella oneidensis, Nature Biotechnol., 2002, 20, 1118–1123.

Helbling E. W., E. R. Marguet, V. E. Villafañe, and O. Holm-Hansen, Bacterioplankton viability in Antarctic waters as affected by solar ultraviolet radiation, Mar. Ecol. Prog. Ser., 1995, 126, 293–298.

Helmstetter C. E. and O. Pierucci, Cell division during inhibition of deoxyribonucleic acid synthesis in Escherichia coli, J. Bacteriol., 1968, 95, 1627–1633.

Herndl G. J., G. Müller-Niklas and J. Frick, Major role of ultraviolet-B in controlling bacterioplankton growth in the surface layer of the ocean, Nature, 1993, 361, 717–719.

Herndl G. J., Role of ultraviolet radiation on bacterioplankton activity, in The Effects of Ozone Depletion on Aquatic Ecosystems, Ch. 8, ed. D.-P Häder, Environmental Intelligence Unit, Academic Press and R.G. Landes Company, Austin, 1997, pp. 143–154.

Higgs P. L., T. F. Letain, K. K. Merriam, N. S. Burke, H. Park, C. Kang and K. Postle, TonB interacts with nonreceptor proteins in the outer membrane of Escherichia coli, J. Bacteriol., 2002, 184, 1640–1648.

Hiller K., M. Schobert, C. Hundertmark, D. Jahn, R. Munch, JVirGel: calculation of virtual two-dimensional protein gels, Nucleic Acids Res., 2003, 31, 3862–3865.

Hollaender A., Effect of long ultraviolet and short visible radiation on Escherichia coli, J. Bacteriol., 1943, 46, 531–541.

Holmgren A., Thioredoxin, Annu. Rev. Biochem., 1985, 54, 237–271.

Horn G., R. Hofweber, W. Kremer and H. R. Kalbitzer, Structure and function of bacterial cold shock proteins, Cell. Mol. Life. Sci., 2007, 64, 1457–1470.

196 Horwich A. L., G. W. Farr and W. A. Fenton, GroEL-GroES-mediated protein folding, Chem Rev., 2006, 106, 1917–1930.

Huang L., M. P. McCluskey, H. Ni and R. A. LaRossa, Global gene expression profiles of the cyanobacterium Synechocystis sp. strain PCC 6803 in response to irradiation with UV-B and white light, J. Bacteriol., 2002, 184, 6845–6858.

Husain I., J. D. Griffith and A. Sancar, Thymine Dimers Bend DNA, Proc. Natl. Acad. Sci. USA, 1988, 85, 2258–2262.

Jang Y. K., L Wang and G. B. Sancar, RPH1 and GIS1 Are Damage-Responsive Repressors of PHR1, Mol. Cell. Biol., 1999, 19, 7630–7638.

Jannasch H. W. and G. E. Jones, Bacterial populations in seawater as determined by different methods of enumeration, Limnol. Oceanogr., 1959, 4, 128–139.

Jeffrey W. H., P. Aas, M. M. Lyons, R. B. Coffin, R. J. Pledger and D. L. Mitchell, Ambiant solar radiation-induced photodamage in marine bacterioplankton, Photochem. Photobiol. Sci., 1996, 64, 419–427.

Jeffrey W. H., J. P. Kase and S. W. Wilhelm, Ultraviolet radiation effects on bacteria and viruses in marine ecosystems, in The effects of UV radiation on marine ecosystems, ed. S. de Mora, S. Demers and M. Vernet, Cambridge University Press. 2000, pp. 206–236.

Jeffrey W. H., J. A. Meador, J. D. Pakulski, T. A. Douki, A. J. Baldwin and D. L. Mitchell, Solar induced DNA photoproducts across a latitudinal gradient in the eastern Pacific Ocean, Annual Meeting, American Society for Limnology and Oceanography, Honolulu, HI. Feb, 2004, 15–20.

Jensen L. J., M. Kuhn , M. Stark , S. Chaffron , C. Creevey , J. Muller , T. Doerks , P. Julien , A. Roth , M. Simonovic, P. Bork and C. von Mering, STRING 8–a global view on proteins and their functional interactions in 630 organisms, Nucleic Acids Res., 2009, 37, 412–416.

Jiang N. and J.-S. Taylor, In vivo evidence that UV-induced C->T mutations at dipyrimidine sites could result from the replicative bypass of cis-syn cyclobutane dimers or their deamination products, Biochemistry, 1993, 32, 472–481.

Joseph N., V. Duppatla and D. N. Rao, Prokaryotic DNA Mismatch Repair, Prog. Nucleic Acid Res., 2006, 81, 1–49.

Joux F., W. H. Jeffrey, P. Lebaron and D. L.Mitchell, Marine bacterial isolates display diverse responses to UV-B radiation, Appl. Environ. Microbiol., 1999, 65, 3820– 3827.

Joux F. and P. Lebaron, Use of fluorescent probes to assess physiological functions of bacteria at the single cell level, Microb. Infect., 2000, 2, 1523–1535.

197 Kaiser E. and G. J. Herndl, Rapid recovery of marine bacterioplankton activity after inhibition by UV radiation, Appl. Environ. Microbiol., 1997, 63, 4026–4031.

Kamashev D., A. Balandina and J. Rouvière-Yaniv, The binding motif recognized by HU on both nicked and cruciform DNA. EMBO J., 1999, 18, 5434–5444.

Karentz D., J. E. Cleaver and D. L. Mitchell, Cell survival characteristics and molecular responses of Antarctic phytoplankton to ultraviolet-B radiation, J. Phycol., 1991, 27, 326–341.

Karlin S., I. Ladunga and B. E. Blaisdell, Heterogeneity of genomes: measures and values, Proc. Natl. Acad. Sci. USA, 1994, 91, 12837– 12841.

Karlin S. and C. Burge, Dinucleotide relative abundance extremes: a genomic signature, Trends Genet., 1995, 11, 283–290.

Keeton W. T., J. L. Gould and C. G. Gould, Biological Science, 5th edn., WW Norton and Company, New York, 1973, pp. 80–85.

Keil R. G. and D. L. Kirchmann, Dissolved combined amino acids in marine waters as determined by a vapor-phase hydrolysis method, Mar.Chem., 1991, 33, 243–259.

Kiener A., I. Husain, A. Sancar and C. Walsh, Purification and properties of Methanobacterium thermoautotrophicum DNA photolyase, J. Biol. Chem. 1989, 264, 13880–13887.

Kim J. E., S. J. Kim, B. H. Lee, R.W. Park, K. S. Kim and I. S. Kim, Identification of motifs for cell adhesion within the repeated domains of transforming growth factor-beta-induced gene, betaig-h3, J. Biol. Chem., 2000, 275, 30907–30915.

Kim J. K., D. Patel and B. S. Choi, Contrasting structural impacts induced by cis-syn cyclobutane dimer and (6-4) adduct in DNA duplex decamers: implication in mutagenesis and repair activity, Photochem. Photobiol., 1995, 62, 44–50.

Kim S. T., K. Malhotra, C. A. Smith, J.-S. Taylor and A. Sancar, Characterization of (6- 4) photoproduct DNA photolyase, J. Biol. Chem., 1994, 269, 8535–8540.

Kim S. T., K. S. Cho, Y. S. Jang and K. Y. Kang, Two-dimensional electrophoretic analysis of rice proteins by polyethylene glycol fractionation for protein arrays, Electrophoresis, 2001, 22, 2103−2109.

Klappenbach J. A., J. M. Dunbar and T. M. Schmidt, rRNA operon copy number reflects ecological strategies of bacteria, Appl. Environ. Microbiol., 2000, 66, 1328–1333.

Kogure Y., E. Yabuuchi, T. Naka, N. Fujiwara and N. Taga, A tentative direct microscopic method for counting living marine bacteria, Can. J. Microbiol., 1979, 25, 415–420.

198 Krijgsveld J., R. F. Ketting, T. Mahmoudi, J. Johansen, M. Artal-Sanz, C. P. Verrijzer, R. H. A. Plasterk and A. J. R. Heck, Metabolic labeling of C. elegans and D. melanogaster for quantitative proteomics, Nature Biotechnol., 2003, 21, 927–931.

Krueger J. H. and G. C. Walker, groE and dnaK genes of Escherichia coli are induced by UV irradiation and nalidixic acid in an htpR dependent fashion, Prec. Natl. Acad. Sci. USA, 1984, 81, 1499–1503.

Kundu L. M., U. Linne, M. Marahiel and T. Carell, RNA is more UV resistant than DNA: the formation of UV-induced DNA lesions is strongly sequence and conformation dependent, Chem.–Eur. J., 2004, 10, 5697–5705. Kyte J. and R. F. Doolittle, A simple method for displaying the hydropathic character of a protein, J. Mol. Biol., 1982, 157, 105–132.

Lahue R. S., K. G. Au, and P. Modrich, DNA mismatch correction in a defined system. Science, 1989, 245, 160–164.

Lamola A. A., Triplet photosensitization and the photobiology of thymine dimers in DNA, Pure Appl. Chem., 1970, 24, 599–610.

Lamrani S., C. Ranquet, M. J. Gama, H. Nakai, J. A. Shapiro, A. Toussaint and G. Maenhaut-Michel, Starvation-induced Mucts62-mediated coding sequence fusion: a role for ClpXP, Lon, RpoS and Crp, Mol. Microbiol., 1999, 32, 327–334.

Landry S. J., R. Jordan, R. McMacken and L. M. Gierasch, Different conformations for the same polypeptide bound to chaperones DnaK and GroEL, Nature, 1992, 355, 455–457.

Larsen L. A., M. Christiansen, J. Vuust and P. S. Andersen, Single-strand conformation polymorphism analysis using capillary electrophoresis, Curr. Protoc. Hum. Genet., 2003, Chapter 7, Units 7.12.

Lawrence C. W., P. E. M. Gibbs, A. Borden, M. J. Horsfall and B. J. Kilbey, Mutagenesis induced by single UV photoproducts E. Coli and yeast, Mutat. Res., 1993, 299, 157–163.

Lee S. and J. A. Fuhrman, Relationships between biovolume and biomass of naturally derived marine bacterioplankton, Appl. Environ. Microbiol., 1987, 53, 1298– 1303.

Lemaire D. G. E. and B. P. Ruzsicska, Kinetic analysis of the deamination reactions of cyclobutane dimers of thymidylyl-3',5'-2'-deoxycytidine and 2'-deoxycytidine- 3,5'-thymidine, Biochemistry, 1993, 32, 2525–2533.

Levine R. L., Oxidative modification of glutamine synthetase. I. Inactivation is due to loss of one histidine residue, J. Biol. Chem., 1983, 258, 11823–11827.

Levine R. L., Carbonyl modified proteins in cellular regulation, aging, and disease, Free Radic. Biol. Med., 2002, 32, 790–796.

199 Li S. and R.Waters, Escherichia coli strains lacking HU are UV sensitive due to a role for HU in homologous recombination, J. Bacteriol., 1998, 180, 3750–3756.

Li S. S., M. Paulsson and L. O. Björn, Temperature-dependent formation and photo repair of DNA damage induced by UV-B radiation in suspension-cultured tobacco cell, J. Photochem. Photobiol. B, 2002, 66, 67–72.

Li S. S., Y. Wang and L. O. Björn, Effects of temperature on UVB-induced DNA damage and photorepair in Arabidopsis thaliana, J. Env. Sci., 2004, 16, 173–176.

Liberek K., J. Marszalek, D. Ang, C. Georgopoulos and M. Zylicz. Escherichia coli DnaJ and GrpE heat shock proteins jointly stimulate ATPase activity of DnaK, Proc. Natl. Acad. Sci. USA, 1991, 88, 2874–2878.

Liu F. T. and N. C. Yang, Photochemistry of cytosine derivatives. 1 Photochemistry of thymidylyl-(3'-5')-deoxycytidine, Biochemistry, 1978, 17, 4865–4876.

Liu Y., J. Zhou, M. V. Omelchenko, A. S. Beliaev, A. Venkateswaran, J. Stair, L. Wu, D. K. Thompson, D. Xu, I. B. Rogozin, E. K. Gaidamakova, M. Zhai, K. S. Makarova, E. V. Koonin and M. J. Daly, Transcriptome dynamics of Deinococcus radiodurans recovering from ionizing radiation, Proc. Natl Acad. Sci. USA, 2003, 100, 4191–4196.

Lo H. L., S. Nakajima, L. Ma, B. Walter, A. Yasui, D. W. Ethell and L. B. Owen, Differential biologic effects of CPD and 6-4PP UV-induced DNA damage on the induction of apoptosis and cell-cycle arrest, BMC Cancer, 2005, 5, 135–162.

Loewe L., V. Textor and S. Scherer, High deleterious genomic mutation rate in stationary phase of Escherichia coli, Science, 2003, 302, 1558–1560.

Long R. A. and F. Azam, Antagonistic interactions among marine pelagic bacteria, Appl. Environ. Microbiol., 2001, 67, 4975–4983.

Ma T. H. and M. M. Harris, Review of the genotoxicity of formaldehyde, Mutat. Res., 1988, 196, 37–59.

Macfayden E. J, C. E. Williamson, G. Grad, M. Lowery, W. H. Jeffrey and D. L. Mitchell, Molecular response to climate change: temperature dependence of UV- induced DNA damage and repair in the freshwater crustacean Daphnia pulicaria, Glob. Change Biol. 2004, 10, 408–416.

Malloy K. D., M. A. Holman and D. Mitchell, Solar UVB-induced DNA damage and photoenzymatic DNA repair in Antarctic zooplankton, Proc. Nat. Acad. Sci, 1997, 94, 1258–1263.

Markovitsi D., D. Onidas, T. Gustavsson, F. Talbot and E. Lazzarotto, Collective behavior of Franck–Condon excited states and energy transfer in DNA double helices, J. Am. Chem. Soc., 2005, 127, 17130–17131.

200 Markowitz V. M. and N. C. Kyrpides, Comparative genome analysis in the integrated microbial genomes (IMG) system, in Comparative genomics, ed. N. H. Bergman, Humana Press, 2008, 395, pp. 35-56.

Marra G. and P. Schär, Recognition of DNA alterations by the mismatch repair system, J.Biochem., 1999, 338, 1–13.

Martinez A., A. S. Bradley, J. R. Waldbauer, R. E. Summons and E. F. DeLong, Proteorhodopsin photosystem gene expression enables photophosphorylation in a heterologous host, Proc. Natl. Acad. Sci. USA, 2007, 104, 5590–5595.

Matallana-Surget S., F. Joux, P. Lebaron and R. Cavicchioli, Isolement et caractérisation de bactéries marines oligotrophes, J. Soc. Biol., 2007, 201, 41–50.

Matallana-Surget S., J. A. Meador, F. Joux and T. Douki, Effect of the GC content of DNA on the distribution of UVB-induced bipyrimidine photoproducts, Photochem. Photobiol. Sci., 2008, 7, 794–801.

Matallana-Surget S., T. Douki, R. Cavicchioli and F. Joux, Remarkable resistance to UVB of the marine bacterium Photobacterium angsutum explained by an unexpected role of photolyase, Photochem Photobiol Sci., 2009a, 8, 1313-1320.

Matallana-Surget S., F. Joux, M. J. Raftery and R. Cavicchioli, The response of the marine bacterium Sphingopyxis alaskensis to solar radiation assessed by quantitative proteomics, Environ. Microbiol. 2009b, 11, 2660–2675.

Matsunaga T., Y. Hatakeyama, M. Ohta, T. Mori and O. Nikaido, Establishment and characterization of a monoclonal-antibody recognising the Dewar isomers of (6-4) photoproducts, Photochem. Photobiol., 1993, 57, 934–940.

Meador J. A., W. H. Jeffrey, J. P. Kase, J. D. Pakulski, S. Chiarello and D. L. Mitchell, Seasonal fluctuation of DNA photo damage in marine plankton assemblages at Palmer Station, Antarctica, Photochem. Photobiol., 2002, 75, 266–271.

Meador J. A., A. J. Baldwin, P. Catala, W. H. Jeffrey, F. Joux, J. A. Moss, J. D. Pakulski, R. Stevens and D. L. Mitchell, Sunlight-induced DNA damage in marine microorganisms collected along a latitudinal gradient from 70°N to 68°S, Photochem. Photobiol., 2009, 85, 412–420.

Meany D. L., H. Xie, L. V. Thompson, E. A. Arriaga and T. J. Griffin, Identification of carbonylated proteins from enriched rat skeletal muscle mitochondria using affinity chromatography-stable isotope labeling and tandem mass spectrometry, Proteomics, 2007, 7, 1150-1163.

Melnikova V. O. and H. N. Ananthaswamy, Cellular and molecular events leading to the development of skin cancer, Mutat. Res., 2005, 571, 91–106.

Middelboe M., A. Hagström, N. Blackburn, B. Sinn, U. Fischer, N. H. Borch, J. Pinhassi, K. Simu and M. G. Lorenz, Effects of bacteriophages on the population

201 dynamics of four strains of pelagic marine bacteria, Microb. Ecol., 2001, 42, 395– 406.

Miller D. L. and H. Weissbach, in Molecular Mechanisms of Protein Biosynthesis, ed. S. Pestka and H. Weissbach, Academic Press, New York, 1977, pp. 323–373.

Mitchell D. L. and B. S. Rosenstein, The use of specific radioimmunoassays to determine action spectra for the photolysis of (6-4) photoproducts, Photochem. Photobiol., 1987, 45, 781– 786.

Mitchell D. L. and R. S. Nairn, The biology of the (6–4) PD photoproduct, Photochem. Photobiol., 1989, 49, 805–819.

Mitchell D. L., J. Jen and J. E. Cleaver, Sequence specificity of cyclobutane pyrimidine dimers in DNA treated with solar (ultraviolet B) radiation, Nucleic Acids Res., 1992, 20, 225–229.

Mitchell D. L., Ultraviolet radiation damage to DNA, in Molecular Biology and Biotechnology, ed. R. A. Meyers, VCH Publishers, 1995, pp. 939–943.

Mitchell D. L., Radioimmunoassay of DNAdamaged by ultraviolet light, in Technologies for Detection of DNA Damage and Mutations, ed. G. Pfeifer, Plenum Publishing Corp., New York, 1996, pp. 73–85.

Mizushima T., K. Kataoka, Y. Ogata, R. Inoue and K. Sekimizu, Increase in negative supercoiling of plasmid DNA in Escherichia coli exposed to cold shock, Mol. Microbiol., 1997, 23, 2381–2386.

Molina M. J. and F. S. Rowland, Stratospheric sink for chlorofluoromethanes: chlorine atomc-atalysed destruction of ozone, Nature, 1974, 249, 810–812.

Mooers O. and E. C. Holmes, The evolution of base composition and phylogenetic interference, Trends Ecol. Evol., 2000, 15, 365–369.

Moran M. A., A. Buchan, J. M. González, J. F. Heidelberg, W. B. Whitman, R. P. Kiene, J. R. Henriksen, G. M. King, R. Belas, C. Fuqua, L. Brinkac, M. Lewis, S. Johri, B. Weaver, G. Pai, J. A. Eisen, E. Rahe, W. M. Sheldon, W. Ye, T. R. Miller, J. Carlton, D. A. Rasko, I. T. Paulsen, Q. Ren, S. C. Daugherty, R. T. Deboy, R. J. Dodson, A. S. Durkin, R. Madupu, W. C. Nelson, S. A. Sullivan, M. J. Rosovitz, D. H. Haft, J. Selengut and N. Ward, Genome sequence of Silicibacter pomeroyi reveals adaptations to the marine environment, Nature, 2004, 432, 910–913.

Morán X. A. G., R. Massana and J. Gasol, Light conditions affect the measurement of oceanic bacterial production via leucine uptake. Appl. Environ. Microbiol., 2001, 67, 3795–3801.

Morita R. Y., Physiology of starvation-survival, in Bacteria in oligotrophic environments: starvation-survival lifestyle, ed. R. Y. Morita, Chapman and Hall, New York, 1997, pp. 274–350.

202 Morris R. M., M. S. Rappé, S. A. Connon, K. L. Vergin, W. A. Siebold, C. A. Carlson and S. J. Giovannoni, SAR11 clade dominates ocean surface bacterioplankton communities, Nature, 2002, 420, 806–810.

Morton R. A. and R. H. Haynes, Changes in the ultraviolet sensitivity of Escherichia coli during growth in batch cultures, J. Bacteriol., 1969, 97, 1379–1385.

Moter A. and U. B. Göbel, Fluorescence in situ hybridization (FISH) for direct visualization of microorganisms, J. Microbiol. Methods, 2000, 41, 85–112.

Mouret S., C. Baudouin, M. Charveron, A. Favier, J. Cadet and T. Douki, Cyclobutane pyrimidine dimers are predominant DNA lesions in whole human skin exposed to UVA radiation, Proc. Natl. Acad. Sci. USA, 2006, 103, 13765–13770.

Mouret S., M. Charveron, A. Favier, J. Cadet and T. Douki, Differential repair of UVB- induced cyclobutane pyrimidine dimers in cultured human skin cells and whole human skin, DNA Repair, 2008, 7, 704–712.

Müller-Niklas G., A. Heissenberger, S. Puskaric and G. J. Herndl, Ultraviolet-B radiation and bacterial metabolism in coastal waters, Aquat. Microb. Ecol., 1995, 9, 111–116.

Muyzer G., E. C. de Waal and A. G. Uitterlinden, Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA, Appl. Environ. Microbiol., 1993, 59, 695–700.

Muyzer G., Structure, function and dynamics of microbial communities: the microbial communities: the molecular biological approach, in Advances in molecular ecology, ed. G. R. Carvalho, Nato Science Series: Series A: Life Sciences. IOS Press, Amsterdam, 1998, pp. 87–117.

Nagata T. and Y. Watanabe, Carbon- and nitrogen-to-volume ratios of bacterioplankton grown under different nutritional conditions, Appl. Environ. Microbiol., 1990, 56, 1303–1309.

Nakajima S., M. Sugiyama, S. Iwai, K. Hitomi, E. Otoshi, S. T. Kim, C. Z. Jiang, T. Todo, A. B. Britt and K. Yamamoto, Cloning and characterization of a gene (UVR3) required for photorepair of 6-4 photoproducts in Arabidopsis thaliana, Nucleic Acids Res., 1998, 26, 638–644.

Nakashima H., K. Nishikawa and T. Ooi, Differences in dinucleotide frequencies of human, yeast and Escherichia coli genes, DNA Res., 1997, 4, 185–192.

Narumi I., K. Satoh, S. Cui, T. Funayama, S. Kitayama and H. Watanabe, PprA: a novel protein from Deinococcus radiodurans that stimulates DNA ligation, Mol. Microbiol., 2004, 54, 278–285.

Ninfa A. J. and M. R. Atkinson, PII signal transduction proteins, Trends Microbiol., 2000, 8, 172–179.

203 Nyström T., R. M. Olsson and S. Kjelleberg, Survival, stress resistance, and alterations in protein expression in the marine Vibrio sp. strain S14 during starvation for different individual nutrients, Appl. Env. Microbiol., 1992, 58, 55–65.

Nyström T., Translational fidelity, protein oxidation, and senescence: lessons from bacteria, Ag. Res. Rev., 2002, 1, 693–703.

Obernosterer I., R. Sempéré and G. J. Herndl, Ultraviolet radiation induces a reversal of the bioavailability of DOM to marine bacterioplankton, Aquat. Microb. Ecol., 2001, 24, 61–68.

O'Brien P. A. and J. A. Houghton, UV-induced DNA degradation in the cyanobacterium Synechocystis PCC 6308, Photochem. Photobiol., 1982, 36, 417–422.

Oda Y., K. Huang, F. R. Cross, D. Cowburn and B. T. Chait, Accurate quantitation of protein expression and site-specific phosphorylation, Proc. Natl. Acad. Sci. USA, 1999, 96, 6591–6596.

Ogawa H., Y. Amagai, I. Koike, K. Kaiser and R. Benner, Production of refractory dissolved organic matter by bacteria, Science, 2001, 292, 917–919.

Ostrowski M., R. Cavicchioli, M. Blaauw and J. C. Gottschal, Specific growth rate plays a critical role in hydrogen peroxide resistance of the marine oligotrophic ultramicrobacterium Sphingomonas alaskensis strain RB2256, Appl. Env. Microbiol., 2001, 67, 1292–1299.

Ostrowski M., F. Fegatella, V. Wasinger, M. Guilhaus, G. Corthals and R. Cavicchioli, Cross-species identification of proteins from proteome profiles of the marine oligotrophic ultramicrobacterium, Sphingopyxis alaskensis, Proteomics, 2004, 4, 1779–1788.

Pakker H. S., R. S. T. Martins and P. Boelen, Effects of temperature on the photoreactivation of ultraviolet-B-induced DNA damage in Palmaria palmata (Rhodophyta), J. Phycol., 2000, 36, 334–341.

Palmeira L., L. Gu´eguen and J. R. Lobry, Mol. Biol. Evol., 2006, 23, 2214–2219.

Pandhal J., P. C. Wright and C. A. Biggs, A Quantitative proteomic analysis of light adaptation in a globally significant marine cyanobacterium Prochlorococcus marinus MED4, J. Proteome Res., 2007, 6, 996–1005.

Pandya G. A., I. Y. Yang, A. P. Grollman and M. Moriya, Escherichia coli responses to a single DNA adduct, J. Bacteriol., 2000, 182, 6598–6604.

Park H., K. Zhang, Y. Ren, S. Nadji, N. Sinha, J. S. Taylor and C. Kang, Crystal structure of a DNA decamer containing a cis-syn thymine-dimer, Proc. Natl. Acad. Sci. USA, 2002, 99, 15965–15970.

204 Pasman Z. and P. H. von Hippel, Regulation of rho-dependent transcription termination by NusG is specific to the Escherichia coli elongation complex, Biochemistry, 2000, 39, 5573–5585.

Paul J. H. and D. Carlson, Genetic material in the marine environment: implication for bacterial DNA, Limnol. Oceanogr., 1984, 29, 1091–1097.

Pausz C. and G. H. Herndl, Role of nitrogen versus phosphorus availability on the effect of UV radiation on bacterioplankton and their recovery from previous UV stress, Aquat. Microb. Ecol, 2002, 29, 89–95.

Payne G. and A. Sancar, Absolute action spectrum of E-FADH2 and E-FADH2-MTHF forms of Escherichia coli DNA photolyase, Biochemistry, 1990, 29, 7715–7727.

Peng W. and B. R. Shaw, Accelerated deamination of cytosine residues in UV-induced cyclobutane dimers leads to CC-TT transitions, Biochemistry, 1996, 35, 10172– 10181.

Pfeifer G. P., Formation and processing of UV photoproducts: effects of DNA sequence and chromatin environment, Photochem. Photobiol., 1997, 65, 270–283.

Plante A. J. and M. T. Arts, Effects of chronic, low levels of UV radiation on carbon allocation in Cryptomonas erosa and competition between C. erosa and bacteria in continuous cultures, J. Plankton Res., 2000, 22, 1277–1298.

Pogoda de la Vega U., P. Rettberg, T. Douki, J. Cadet and G. Horneck, Sensitivity to polychromatic UV-radiation of strains of Deinococcus radiodurans differing in their DNA repair capacity, Int. J. Radiat. Biol., 2005, 81, 601–611.

Poindexter J. S., Oligotrophy: fast and famine existence, Adv. Microb. Ecol., 1981, 5, 63–89.

Porter K. G. and Y. S. Feig., The use of DAPI for identifying and counting aquatic microflora, Limnol. Oceanogr., 1980, 25, 943–948.

Potaman V. N. and V. N. Soyfer, Oligonucleotide model with nonidentical complentary strands for chromatographic studies of structuredependent photosusceptibility, J. Chromatogr. A, 1995, 690, 65–70.

Preobrajenskaya O., A. Boullard, F. Boubrik, M. Schnarr and J. Rouviere-Yaniv, The protein HU can displace the LexA repressor from its DNA binding sites, Mol. Microbiol., 1994, 13, 459–467.

Protic-Sablic S. M. and K. H. Kreamer, One pyrimidine dimer inactivates expression of a transfected gene in xeroderme pigmentosum cells, Proc. Natl. Acad. Sci. USA., 1985, 82, 6622–6626.

Qiu, X., G. W. Sundin, B. Chai and J. M. Tiedje, Survival of Shewanella oneidensis MR-1 after UV radiation exposure, Appl. Environ. Microbiol., 2004. 70, 6435– 6443.

205 Qiu, X., G. W. Sundin, L. Wu, J. Zhou and J. M. Tiedje, Comparative analysis of differentially expressed genes in Shewanella oneidensis MR-1 following exposure to UVC, UVB, and UVA radiation, J. Bacteriol., 2005, 187, 3556–3564.

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

Quackenbush J., Computational analysis of microarray data, Nat. Rev. Genet., 2001, 2, 419–427.

Quillardet P., M.-A. Rouffaud and P. Bouige, DNA array analysis of gene expression in response to UV irradiation in Escherichia coli, Res. Microbiol., 154, 2003, 559– 572.

Rabilloud T., Two-dimensional gel electrophoresis in proteomics: old, old fashioned, but it still climbs up the mountains, Proteomics, 2002, 2, 3–10.

Rae R. and W. F. Vincent, Effects of temperature and ultraviolet radiation on microbial foodweb structure: potential responses to global change, Fres. Biol., 1998, 40, 747–758.

Rai P., T. D. Cole, D. E. Wemmer and S. Linn, Localization of Fe(2+) at an RTGR sequence within a DNA duplex explains preferential cleavage by Fe(2+) and H2O2. J. Mol. Biol. 2001, 312, 1089–101.

Rappé M. S., S. A. Connon, K. L. Vergin and S. J. Giovannoni, Cultivation of the ubiquitous SAR11 marine bacterioplankton clade, Nature, 2002, 418, 630–633.

Ravanat J. L., T. Douki and J. Cadet, Direct and indirect effects of UV radiation on DNA and its components, J. Photochem. Photobiol. B: Biol., 2001, 63, 88–102.

Raven J. and R. Geider, Temperature and algal growth, New Phytol., 1988, 110, 441– 461.

Reed R. H., Solar inactivation of faecal bacteria in water: the critical role of oxygen, Lett. Appl. Microbiol., 1997, 24, 276–280.

Requena J. R., R. L. Levine and E. R. Stadtman, Recent advances in the analysis of oxidized proteins, Amino Acids, 2003, 25, 221–226.

Rice P., I. Longden and A. Bleasby, EMBOSS: the European Molecular Biology Open Software Suite, Trends Genet., 2000, 16, 276–277.

Riesenfeld C. S., P. D. Schloss and J. Handelsman, Metagenomics: Genomic analysis of microbial communities, Ann. Rev. Gen., 2004, 38, 525–552.

Rontó G., S. Gáspár, A. Fekete, T. Kerékgyártó, A. Bérces and P. Gróf, Stability of nucleic acid under the effect of UV radiation, Adv. Sp. Res., 2002, 30, 1533–1538.

206 Rosen C. F., R. Poon and D. J. Drucker, UVB radiation-activated genes induced by transcriptional and posttranscriptional mechanisms in rat keratinocytes, Am. J. Physiol., 1995, 268, 846–855.

Rue E. L. and K. W. Bruland, The role of organic complexation on ambient iron chemistry in the equatorial Pacific Ocean and the response of a mesoscale iron addition experiment, Limnol. Oceanogr., 1997, 42, 901–910.

Sage E., E. Cramband and B.W. Glickman, The distribution of UV damage in the lacI gene of Escherichia coli: correlation with mutation spectrum, Mutat. Res., 1992, 269, 285–299.

Sage E., B. Lamolet, E. Brulay, E. Moustacchi, A. Chateauneuf and E. A. Drobetsky, Mutagenic specificity of solar UV light in nucleotide excision repair-deficient rodent cells, Proc. Natl. Acad. Sci. USA, 1996, 93, 176–180.

Sage E., D. Perdiz, P. Grof, A. Reynaud-Angelin, T. Douki, J. Cadet, P. Rochette, N. Bastien and R. Drouin, DNA damage induced by UVA radiation: role in solar mutagenesis, in From DNA Photolesions to Mutations, Skin Cancer and Cell Death, ed. E. Sage, R. Drouin and M. Rouabhia, Cambridge, UK, 2005.

Sancar A., F. W. Smith and G. B. Sancar, Purification of Escherichia coli DNA photolyase, J. Biol. Chem., 1984, 259, 6028–6032.

Sancar G. B. and A. Sancar, Structure and function of DNA photolyases, Trends Biochem. Sci., 1987, 12, 259–261.

Sancar G. B. and F. W. Smith, Interactions between yeast photolyase and nucleotide excision repair proteins in Saccharomyces cerevisiae and Escherichia coli, Mol. Cell. Biol., 1989, 9, 4767–4776.

Sancar A., Structure and function of DNA photolyase and cryptochrome blue-light photoreceptors, Chem. Rev., 2003, 103, 2203–2237.

Sancar A., Structure and function of photolyase and in vivo enzymology: 50th Anniversary, J. Biol. Chem., 2008, 47, 32153–32157.

Schreier W. J., T. E. Schrader, F. O. Koller, P. Gilch, C. E. Crespo-Hernandez, V. N. Swaminathan, T. Carell, W. Zinth and B. Kohler, Thymine dimerization in DNA is an ultrafast photoreaction, Science, 2007, 315, 625–629.

Schuster G. B. and U. Landman, The mechanism of long-distance radical cation transport in duplex DNA: ion-gated hopping of polaron-like distortions, in Longe- range charge transfer in DNA I, ed. G. B. Schuster, Springer Berlin, Heidelberg, 2004, pp. 139–161.

Schut F., E. J. de Vries, J. C. Gottschal, B. R. Robertson, W. Harder, R. A. Prins and D. K. Button, Isolation of typical marine bacteria by dilution culture : growth, maintenance, and characteristics of isolates under laboratory conditions, Appl. Environ. Microbiol., 1993, 59, 2150–2160.

207 Schut F., M. Jansen, T. M. Pedro Gomes, J. C. Gottschal, W. Harder and R. A. Prins, Substrate uptake and utilization by a marine ultramicrobacterium, Microbiol., 1995, 141, 351–361.

Schut F., R. A. Prins and J. C. Gottschal, Oligotrophy and pelagic marine bacteria: facts and fiction, Aquat. Microb. Ecol., 1997, 12, 177–202.

Seawell P. C., C. A. Smith and A. K. Ganesan, den V gene of bacteriophage T4 determines a DNA glycosylase specific for pyrimidine dimers in DNA. J. Virol., 1980, 35, 790–796.

Sebastian J., B. Kraus and G. B. Sancar, Expression of the yeast PHR1 gene is induced by DNA-damaging agents, Mol. Cell. Biol., 1990, 10, 4630–4637.

Sebastian J. and G. B. Sancar, A damage-responsive DNA binding protein regulates transcription of the yeast DNA repair gene PHR1, Proc. Natl. Acad. Sci. USA, 1991, 88, 11251–11255.

Seedberg E., L. Eide and M Bjoras, The base excision pathway, Trends Biochem. Sci., 1995, 20, 391–397.

Selby C. P. and A. Sancar, Molecular mechanism of transcription-repair coupling, Science, 1993, 260, 53–58.

Setlow R. B. and W. L. Carrier, Pyrimidine dimers in ultravioletirradiated DNAs, J. Mol. Biol., 1966, 17, 237–254.

Shangari N., W. R. Bruce, R. Poon and P. J. O’Brien, Toxicity of glyoxals – role of oxidative stress, metabolic detoxification and thiamine deficiency, Biochem. Soc. Transac., 2003, 31, 1390–1393.

Shuman S. and M. S. Glickman, Bacterial DNA repair by non-homologous end joining, Nat. Rev. Microbiol., 2007, 5, 852–861.

Sieracki M. E. and J. M. Sieburth, Sunlight-induced growth delay of planktonic marine bacteria in filtered seawater, Mar. Ecol. Prog. Ser., 1986, 33, 19–27.

Sies H., Oxidative stress: oxidants and antioxidants, Exp. Physiol., 1997, 82, 291–295.

Simpson G. G., in Principles of animal taxonomy, Columbia University Press, New- York. 1961.

Simu K. and A. Hagström, Oligotrophic bacterioplankton with a novel single-cell life strategy, Appl. Environ. Microbiol., 2004, 70, 2445–2451.

Simu K., K. Holmfeldt, U. L. Zweifel and A. Hagström, Culturability and coexistence of colony-forming and single-cell marine bacterioplankton, Appl. Environ. Microbiol., 2005, 71, 4793–4800.

208 Singer C. E. and B. N. Ames, Sunlight ultraviolet and bacterial DNA base ratios, Science, 1970, 170, 822–826.

Singh N. P., M. T. McCoy, R. R. Tice and E. L. Schneider, A simple technique for quantitation of low levels of DNA damage in individual cells, Exp. Cell Res., 1988, 175, 184–191.

Sinha R. P. and D. P. Häder, UV-induced DNA damage and repair: a review, Photochem. Photobiol. Sci., 2002, 1, 225–236.

Smith D. C. and F. Azam, A simple, economical method for measuring bacterial protein synthesis rates in seawater using 3H-leucine, Mar. Microb. Food Webs, 1992, 6, 107–114.

Sommaruga R., I. Obernosterer, G. J. Herndl and R. Psenner, Inhibitory Effect of solar radiation on thymidine and leucine incorporation by freshwater and marine bacterioplankton, Appl. Environ. Microbiol., 1997, 63, 4178–4184.

Sommaruga R., B. Sattler, A. Oberleiter, A. Wille, S. Sommaruga-Wögrath, R. Psenner, M. Felip, L. Camarero, S. Pina, R. Gironés and J. Catalán, An in situ enclosure experiment to test the solar UVB impact on plankton in a high altitude mountain lake: Effects on the microbial food web, J. Plankton Res., 1999, 21, 859–876.

Sonntag V., in The Chemical Basis of Radiation Biology, ed. Taylor and Francis, New York, 1987.

Sowell S. M., A. D. Norbeck, M. S. Lipton, C. D. Nicora, S. J. Callister, R. D. Smith, D. F. Barofsky and S. J. Giovannoni, Proteomic analysis of stationary phase in the marine bacterium “Candidatus Pelagibacter ubique”, App. Env. Microbiol., 2008, 74, 4091–4100.

Srinivasan S. and S. Kjelleberg, Cycles of famine and feast: the starvation and outgrowth strategies of a marine Vibrio. J. Biosci., 1998, 23, 501–511.

Staley J. T. and A. Konopka, Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats, Ann. Rev. Microbiol., 1985, 39, 321–346.

Steglich C., M. Futschik, T. Rector, R. Steen and S. W. Chisholm, Genome-wide analysis of light sensing in Prochlorococcus, J. Bacteriol., 2006, 188, 7796–7806.

Svoboda D. L., C. A. Smith, J. S. Taylor and A. Sancar, Effect of sequence, adduct type, and opposing lesions on the binding and repair of ultraviolet photodamage by DNA photolyase and (A)BC excinuclease, Biol. Chem., 1993, 268, 10694– 10700.

Takahashi M., H. Okamiya, F. Furukawa, K. Toyoda, H. Sato, K. Imaida and Y. Hayashi, Effects of glyoxal and methyiglyoxal administration on gastric carcinogenesis in Wistar rats after initiation with N-methyl-N'-nitro-N- nitrosoguanidine, Carcinogenesis, 1989, 10, 1925–1927.

209 Takeuchi Y., M. Murakami, N. Nakajima, N. Kondo and O. Nikaido, Induction and repair of damage to DNA in cucumber cotyledons irradiated with UV-B, Plant Cell Physiol., 1996, 37, 181–187.

Tang M., X. Shen, E. G. Frank, M. O'Donnell, R. Woodgate and M. F. Goodman, UmuD'(2)C is an error-prone DNA polymerase, Escherichia coli pol V, Proc. Natl. Acad. Sci. USA, 1999, 96, 8919–8924.

Taylor J., H. Lu and J. J. Kotyk, Quantitative conversion of the (6-4) photoproduct of TpdC to its Dewar valence isomer upon exposure to simulated sunlight, Photochem. Photobiol., 1990, 51, 161–167.

Taylor J. S., Unraveling the molecular pathway from sunlight to skin cancer, Acc. Chem. Res., 1994, 27, 76–82.

Tedetti M. and R. Sempéré, Penetration of Ultraviolet radiation in the marine environment. A review, Photochem. Photobiol., 2006, 82, 389–397.

Tenaillon O., E. Denamur and I. Matic, Evolutionary significance of stress-induced mutagenesis in bacteria, Trends Microbiol., 2004, 12, 264–270.

Thekaekara M. P., Solar irradiance: total and spectral and its possible variations, Appl. Opt., 1976, 15, 915–920.

Thompson F. L., T. Iida and J. Swings, Biodiversity of Vibrios, Microbiol. Mol. Biol. Rev., 2004, 68, 403–431.

Todo T., H. Takemori, H. Ryo, M. Ihara, T. Matsunaga, O. Nikaido, K. Sato and T. Nomura, A new photoreactivating enzyme that specifically repairs ultraviolet light-induced (6-4) photoproducts, Nature, 1993, 361, 371–374.

Tornaletti S. and P. C. Hanawalt, Effect of DNA lesions on transcription elongation, Biochimie, 1999, 81, 139–146.

Touati D., Iron and oxidative stress in bacteria, Arch. Biochem. Biophys., 2000, 373, 1– 6.

Trautinger F., I. Kindås-Mügge, R. M. Knobler and H. Hönigsmann, Stress proteins in the cellular response to ultraviolet radiation, J. Photochem. Photobiol. B., 1996, 35, 141–148.

Van den Bergh G. and L. Arckens, Fluorescent two-dimensional difference gel electrophoresis unveils the potential of gel-based proteomics, Curr. Op. Biotechnol., 2004, 15, 38–43.

Van Guemerden H. and G. J. Kuenen, Strategies for growth and evolution of microorganisms in oligotrophic habitats, in Heterotrophic activity in the sea, ed. J. E. Ieb Hobbie and P. J. Williams, Nato Conference series, series IV. Plenum. Press, New York, 1984, pp. 25–44.

210 Varghese A. J., Photochemistry of nucleic acids and their constituents, in Photophysiology, Academic Press, New-York, 1972, pp. 207–274.

Veldhuis M. J. W., K. Timmermans, P. Croot and B. Vand der Wagt, Picophytoplankton: a comparative study of their biochemical composition and phtosynthetic propersties, J. Sea Res., 2005, 53, 7–24.

Verbenko N., L. V. Kuznetsova, A. V. Smolnikova and V. L. Kalinin, Effect of insertional mutation in the cspA gene encoding the major cold-shock protein on radiation resistance of Escherichia coli, Russ. J. Genet., 2003, 39, 748–757.

Vile G. F. and R. M. Tyrrell, Oxidative stress resulting from ultraviolet A irradiation of human skin fibroblasts leads to a heme oxygenase-dependent increase in ferritin, J. Biol. Chem., 1993, 268, 14678–14681.

Vincent W. F. and S. Roy, Solar UV-B radiation and aquatic primary production: damage, protection and recovery, Environ. Rev., 1993, 1, 1–12.

Viswanathan A. and P. W. Doetsch, Effects of nonbulky DNA base damages on Escherichia coli RNA polymerase-mediated elongation and promoter clearance, J. Biol. Chem., 1998, 273, 21276–21281.

Viswanathan S., M. Ünlü and J. S. Minden, Two-dimensional difference gel electrophoresis, Nature Prot., 2006, 1, 1351–1358.

Wallin E. and G. von Heijne, Genome-wide analysis of integral membrane proteins from eubacterial, archaean, and eukaryotic organisms, Protein Sci, 1998, 7, 1029– 1038.

Wandersman C. and P. Delepelaire, Bacterial iron sources: from siderophores to hemophores, Annu. Rev. Micriobiol., 2004, 58, 611–647.

Wang C. I. and J. S. Taylor, Site-specific effect of thymine dimmer formation on dAn.dTn tract bending and its biological implications, Proc. Natl. Acad. Sci. USA, 1991, 88, 9072–9076.

Ward D. M., A natural species concept for prokaryotes, Curr. Op. Microbiol., 1998, 1, 271–277.

Waterworth W. M., Q. Jiang, C. E. West, M. Nikaido and C. M. Bray, Characterization of Arabidopsis photolyase enzymes and analysis of their role in protection from ultraviolet-B radiation, J. Exp. Bot., 2002, 53, 1005–1015.

Wayne L. G., D. J. Brenner, R. R. Colwell, P. A. D. Grimont, O. Kandler, M. I. Krichevsky, L. H. Moore, G. E. Murray, E. Stackebrandt, M. P. Starr and G. H. Trüper, Report of the ad hoc committee on reconciliation of approaches to bacterial systematics, Int. J. Syst. Bacteriol., 1987, 37, 463–464.

Weibezahn J., B. Bukau and A. Mogk, Unscrambling an egg: protein disaggregation by AAA+ proteins, Microb. Cell. Fact., 2004, 3, 1–12.

211 Weibezahn J., C. Schlieker, P. Tessarz, A. Mogk and B. Bukau, Novel insights into the mechanism of chaperone-assisted protein disaggregation, Biol. Chem., 2005, 386, 739–744.

Whitman W. B., D. C. Coleman and W. J. Wiebe, Prokaryotes: the unseen majority, Proc. Natl. Acad. Sci. USA, 1998, 95, 6578–6583.

Wickham S. and M. Carstens, Effects of ultraviolet-B radiation on two arctic microbial food webs, Aquat. Microb. Ecol., 1998, 16, 163–171.

Wilhelm S. W., W. H. Jeffrey, C. A. Suttle and D. L. Mitchell, Estimation of biologically damaging UV levels in marine surface waters with DNA and viral dosimeters, Photochem. Photobiol. 2002, 76, 268–273.

Wilhelm S. W., W. H. Jeffrey, A. L. Dean, J. Meador, J. D. Pakulski and D. L. Mitchell, UV radiation induced DNA damage in marine viruses along a latitudinal gradient in the southeastern Pacific Ocean, Aquat. Microb. Ecol., 2003, 31, 1–8.

Williams P. M. and E. R. M. Druffel, Radiocarbon in dissolved organic matter in the central North Pacific Ocean, Nature, 1987, 330, 246–248.

Williamson C. E., G. Grad and H. J. De Lange, Temperature-dependent ultraviolet responses in zooplankton: implications of climate change, Limnol. Oceanogr. 2002, 47, 1844–1848.

Wood P. D. and R. W. Redmond, Triplet state interactions between nucleic acid bases in solution at room temperature: intramolecular energy and electron transfer, J. Am. Chem. Soc., 1996, 118, 4256–4263.

Woodgate R., A plethora of lesion-replicating DNA polymerases, Genes Dev., 1999, 13, 2191–2195.

Xenopoulos M. A. and D. F. Bird, Effect of acute exposure to hydrogen peroxide on the production of phytoplankton and bacterioplankton in a mesohumic lake, Photochem. Photobiol., 1997, 66, 471–478.

Xia X., Z. Xie and W.-H. Li, Effects of GC content and mutational pressure on the lengths of exons and coding sequences, J. Mol. Evol., 2003, 56, 362–370.

Yamamoto K., Y. Fujiwara and H. Shinagawa, Evidence that the phr_gene enhances the ultraviolet resistance of Escherichia coli recA strains in the dark, Mol. Gen. Genet., 1983, 192, 282–284.

Yan L.-J. and R. S. Sohal, Gel electrophoretic quantitation of protein carbonyls derivatized with tritiated sodium borohydride, Anal. Biochem., 1998, 265, 176– 182.

Yasuhira S. and A. Yasui, Visible light-inducible photolyase gene from the goldfish Carassius auratus, J. Biol. Chem., 1992, 267, 25644–25647.

212 Ye K., A. Serganov, W. Hu, M. Garber and D. J. Patel, Ribosome-associated factor Y adopts a fold resembling a double-stranded RNA binding domain scaffold, Eur. J. Biochem., 2002, 269, 5182–5191.

You Y.-H., P. E. Szabo and G. P. Pfeifer, Cyclobutane pyrimidine dimers form preferentially at the major p53 mutational hotspot in UVBinduced mouse skin tumors, Carcinogenesis, 2000, 21, 2113–2117.

Zenoff V. F., F. Siñeriz and M. E. Farías, Diverse Responses to UV-B Radiation and repair mechanisms of bacteria isolated from high-altitude aquatic environments, Appl. Environ. Microbiol., 2006,72, 7857–7863.

Zepp R. G., D. J. E. III, N. D. Paul and B. Sulzberger, Interactive effects of solar UV radiation and climate change on biogeochemical cycling, Photochem. Photobiol. Sci., 2007, 6, 286–300.

Zhang C., W. Jianfeng, Z. Zhiguo, Y. Nanjiao, S. Duohong and H. Yuejin, Proteomic analysis of Deinococcus radiodurans recovering from gamma-radiation, Proteomics, 2005, 5, 138–143.

Zhong H, S. L. Marcus and L. Li, Two-dimensional mass spectra generated from the analysis of 15N-labeled and unlabeled peptides for efficient protein identification and de novo peptide sequencing, J. Prot. Res., 2004, 3, 1155–1163.

Ziegler A., D. J. Leffel, S. Kunala, H. W. Sharma, P. E. Shapiro, A. E. Bale and D. E. Brash, Mutation hotspots due to sunlight in the p53 gene of nonmelanoma skin cancers, Proc. Natl. Acad. Sci. USA, 1993, 90, 4216–4220.

213 APPENDICES Appendix A

Determination of the dinucleotide frequency for M. luteus (72% GC) in the high GC content region. Values were interpolated from those of eight species exhibiting a similar GC content: Nocardia farcinica IFM 10152 (70% GC), Frankia sp. CcI3 (70.1), Saccharopolyspora erythraea, NRRL 2338 (71.1), Nocardioides sp. JS614 (71.4), Streptomyces coelicolor A3(2) (72), Clavibacter michiganensis subsp. michiganensis NCPPB 382 (72), Frankia alni ACN14a (72.5), Anaeromyxobacter dehalogenans 2CP-C (74.9).

215 Appendix B

Frequency of the four bipyrimidine dinucleotides in a series of 99 bacteria with increasing GC content (% GC). These values were used to draw Figure 2.2.

Photobacterium angustum S14

216 217 Appendix C

Effects of UVB radiation on the survival of V. natriegens ( ), P. angustum () and S. alaskensis (<). The data expressed the mean (± standard deviations) of three independent experiments for P. angustum and S. alaskensis and two independent experiments for V. natriegens.

218 Appendix D

Evidence rating system for manual annotation of proteins.

Blast similarity Blast Experimental evidence Evidence score rating High similarity with protein of > 90% Experimental evidence in the ER1 studied organism studied organism High similarity with a protein of > 35% Experimental evidence in ER2 known function another organism Low similarity with a protein of < 35% Experimental evidence in ER3 known function another organism Low similarity with a protein of < 35% No experimental evidence in any ER4 known function organism No similarity / / ER5

219 Appendix E

Cellular location predicted using PSORTb v.2.0 software for the total proteins observed in all iTRAQ runs (http://www.psort.org/psortb/).

Total number of Percentage of proteins Cellular location proteins observed observed

Cytoplasmic 357 44.0

Cytoplasmic/Membrane 62 7.7

Periplasmic 17 2.1

Outer Membrane 23 2.9 Extracellular 2 0.2 Unknown (may have multiple location sites) 30 3.7 Unknown 320 39.4 Total 811 100

220 Appendix F

Unique proteins of mid-logarithmic phase identified by Scaffold.

Accesion number Name of proteins Sala_2785 UspA Sala_3076 Short-chain_dehydrogenase_OR_reductase_SDR Sala_2262 Type_I_secretion_outer_membrane_protein_TolC Sala_0763 Cytidine_deaminase_homotetrameric Sala_2050 Uncharacterised_P-loop_ATPase_protein_UPF0042 Sala_1677 OmpW Sala_1650 Type_IV_pilus_assembly_PilZ Sala_2743 ATP-dependent_Clp_protease_ATP-binding_subunit_ClpX Sala_1881 UDP-N-acetylglucosamine Sala_0129 Putative_integral_membrane_protein Sala_1808 Enoyl-CoA_hydratase_OR_isomerase Sala_2755 Peptidase_S41 Sala_2775 Transglutaminase-like Sala_0303 Transcriptional_regulator_LacI_family Sala_0346 Metallophosphoesterase Sala_0790 Hypothetical_protein Sala_1494 Cytidylate_kinase Sala_1837 Formyltetrahydrofolate_deformylase Sala_2112 Protein_of_unknown_function_DUF1285 Sala_2114 Hypothetical_protein Sala_2680 RNA_polymerase,_sigma-24_subunit_ECF_subfamily Sala_2831 GCN5-related_N-acetyltransferase Sala_2878 Conserved_hypothetical_protein

221 Appendix G

Unique proteins of stationary phase identified by Scaffold.

Accesion number Name of proteins Sala_2167 General_stress_protein Sala_2837 Sigma_54_modulation_protein_OR_ribosomal_protein_S30EA Sala_0892 Protein_of_unknown_function_DUF35 Sala_0910 AMP-dependent_synthetase_and_ligase Sala_0911 Conserved_hypothetical_protein Sala_0361 Beta-lactamase-like Sala_2707 Putative_signal-transduction_protein_with_CBS_domains Sala_2632 Histone-like_DNA-binding_protein Sala_0328 TolB-like Sala_2164 Antibiotic_biosynthesis_monooxygenase Sala_1596 Transcriptional_regulator_XRE_family Sala_1890 Protein_of_unknown_function_UPF0040 Sala_1979 Putative_GAF_sensor_protein Sala_2993 4-hydroxybenzoyl-CoA_thioesterase Sala_2184 Alkyl_hydroperoxide_reductase_OR_Thiol_specific_antioxidant Sala_2188 Hemerythrin_HHE_cation_binding_region Sala_1566 NAD-dependent_epimerase_OR_dehydratase Sala_0544 Peptidyl-prolyl_cis-trans_isomerase_cyclophilin_type Sala_1616 GumN Sala_0684 Hydroxyacylglutathione_hydrolase Sala_0672 Acetyltransferase_GNAT_family Sala_0612 Translation_initiation_factor_IF-2 Sala_2194 Polyprenyl_synthetase Sala_1358 Ornithine_carbamoyltransferase Sala_0119 Transcriptional_regulator_MarR_family Sala_0834 Molybdenum_cofactor_biosynthesis_protein_C Sala_1883 UDP-N-acetylmuramoylalanine-D-glutamate_ligase Sala_1480 Cold-shock_DNA-binding_domain_protein Sala_1657 3,4-dihydroxy-2-butanone_4-phosphate_synthase Sala_1597 Conserved_hypothetical_protein Sala_0527 Hypothetical_protein Sala_0757 Conserved_hypothetical_protein Sala_1838 Glyoxalase_OR_bleomycin_resistance_protein_OR_dioxygenase Sala_0039 Conserved_hypothetical_protein Sala_1365 Dienelactone_hydrolase Sala_3182 Phosphopantothenoylcysteine_decarboxylase Sala_0679 Protein_of_unknown_function_DUF885 Sala_2673 Response_regulator_receiver_protein Sala_1726 Response_regulator_receiver_protein Sala_2052 HPrNtr Sala_0372 Twin-arginine_translocation_pathway_signal Sala_2491 Hypothetical_protein Sala_0601 Bacteriophage_N4_adsorption_protein_B Sala_0518 Conserved_hypothetical_protein Sala_0944 Transcriptional_regulator_XRE_family Sala_0011 3-demethylubiquinone-9_3-methyltransferase Sala_1378 Conserved_hypothetical_protein Sala_1065 Phosphoribosylanthranilate_isomerase Sala_1399 SurF1_family_protein Sala_0594 Conserved_hypothetical_protein Sala_1634 Enoyl-CoA_hydratase_OR_isomerase Sala_0740 Tryptophan_halogenase Sala_1771 Diguanylate_phosphodiesterase Sala_2535 Phage_integrase Sala_0734 Putative_transposase

222 Appendix H

Representation of predicted protein interactions using BioLayout. Nodes represent differentially expressed proteins and links represent interactions.

Ribosomal proteins Proteins turnover, chaperones Amino acid transport and metabolism Nucleotide transport and metabolism Lipid metabolism Oxidative stress defense DNA binding protein Transcription regulators Translation regulators Krebs cycle Glycolysis metabolic pathway Periplasmic protein Nitrogen regulatory pathway Iron transport and metabolsim Hypothetical proteins

223