Biodegradation of Poly(Ethylene Terephthalate) by Marine Bacteria, and Strategies for Its Enhancement

Biodegradation of poly(ethylene terephthalate) by marine bacteria, and strategies for its enhancement

Submitted in total fulfillment of the requirements for the degree of Doctor of Philosophy

by Hayden Webb

Environmental and Biotechnology Centre

Faculty of Life and Social Sciences

Swinburne University of Technology

May 2012

______

Abstract

Plastic accumulation, particularly in the world’s oceans is of increasing environmental concern. One of the major components of plastic waste is poly(ethylene terephthalate) (PET), a polymer frequently used in many applications, including textiles and food packaging. The current methods of disposal of PET waste, landfill, incineration and recycling, each have inherent drawbacks and limitations, and as such there is a need for efficient and cost-effective alternative.

Biodegradation is an attractive option for environmentally friendly and efficient disposal of plastic waste. To date, no protocol has yet been developed to feasibly dispose of PET by biodegradation con a commercial scale. The current works aims to investigate the potential of PET biodegradation as a plastic disposal procedure by providing fundamental knowledge of biodegradation processes, and to develop strategies for improving biodegradation efficiency.

PET samples were incubated in marine bacterial community enrichment cultures, and the dynamics of the polymer – bacterial interactions traced. Modifications to polymer surfaces were monitored using a variety of surface characterisation techniques, including atomic force microscopy (AFM), x-ray photoelectron spectroscopy (XPS) and infrared microspectroscopy using Synchrotron radiation. Taxonomic members of the bacterial enrichment cultures that developed in the presence of PET were recovered and identified via 16S rRNA gene sequencing.

Marine bacteria were shown to possess the ability to degrade PET surfaces. Larger topographical surface features were degraded preferentially, which resulted in overall smoothening of the polymer surface. Subtle changes in the chemistry of the PET surface were also detected, which were indications of polymer chain cleavage and oligomer excision.

Bacterial enrichment communities were dominated by members of α- and γ- proteobacteria, with contributions from β-proteobacteria. Among the isolates with the highest potential for PET biodegradation, many returned nearest taxonomic relatives of species of Thalassospira, Kordiimonas and Alteromonas. Commensal relationships

i appeared to be required for effective PET biodegradation, with several sulphur-cycling strains showing potential participation in PET degrading communities.

Several strategies for enhancement of the rate of PET biodegradation were also put forward. Increasing the size of bacterial enrichment communities by supplementation with essential elemental nutrients in nitrogen and phosphorus proved effective in increasing the rate of biodegradation, as did alteration of the composition of the communities by increasing the selective advantage of PET-degrading strains through incubation in dark conditions. An alternative approach through manipulation of the polymer was another effective option; pre-treatment of PET samples with surfactants increased their surface roughness, which enabled bacterial cells greater access to the polymer chains.

Limited cell attachment to PET surface was identified as one of the major barriers to biodegradation. To address this, several unrelated surface topographies were tested for the ability to promote cell adhesion, for eventual application on PET surfaces. Of those tested, a superhydrophobic topography based on the structure of the lotus leaf showed the most promise for enhancement of biodegradation through promotion of cell attachment.

The results of this study represent a significant contribution to the fundamental understanding of PET biodegradation mechanisms. The knowledge obtained here can be used for the further development of biodegradation as a technique for disposal of PET waste.

Acknowledgements

First and foremost, I would like to acknowledge and thank my primary supervisor, Professor Elena Ivanova for all her guidance and support over the course of this study. The inspiration, motivation and encouragement she provided was a key driving force behind the project, and I am particularly grateful that she (almost) always was able to take time out of her busy schedule to talk things over whenever I needed advice.

Similarly, I would also like to thank Professor Russell Crawford for co-supervising the project, and for providing feedback on the various manuscripts and papers that I prepared. His support and encouragement were vital for completion of the work.

To Vi Khanh Truong and Jafar Hasan, thank you for your co-operative efforts on countless individual experiments and analyses; we make a pretty good team. I value your friendship, and I’ll always remember the ‘deep philosophical discussions’ we had on a daily basis.

I’d like to express my thanks to Hooi Jun Ng, The Hong Phong Nguyen and Rebecca Flecknoe for their assistance in cultivating, counting and characterising bacterial strains. You guys really made my life a little easier.

Thank you to Dr. James Wang for his assistance in performing SEM and in early AFM experiments, without his expertise I would have been lost. Also to Dr. Natasa Mitik-Dineva for lighting the way with her earlier work on bacterial attachment.

A special thank you goes to Chris, Soula, Ngan, Savithri and all the laboratory technicians who have helped me throughout the course of this work.

To all my friends and colleagues at Swinburne for making my time here enjoyable and rewarding. You guys made my time here worthwhile, and although I can’t thank you all individually I would like to specifically thank Abdullah, Bita, Elisa, Fiona, Kaylass, Liz, Pete and Sarah.

And finally thank you to my family. It goes without saying, but I’ll say it here: your love and support have shaped me into the person I am today.

iii

Declaration

I, Hayden Webb, declare that this thesis is my original work and contains no material that has been accepted for the award of Doctor of Philosophy, or any other degree or diploma, except where due reference is made.

I declare that to the best of my knowledge this thesis contains no material previously published or written by any other person except where due reference is made. Wherever contributions of others were involved every effort has been made to acknowledge the contributions of the respective workers or authors.

Signature ______

List of publications

Online Book:

Webb, H. K., Truong, V. K., Jones, R. T., Crawford, R. J. & Ivanova, E. P. (2010). X- ray photoelectron spectroscopy as a tool in the study of nanostructured titanium and commercial PET surfaces in biotechnological applications. (Nova Science Publishers Inc., New York).

Book Chapters:

Peer-reviewed articles:

Ivanova, E. P., Truong, V. K., Webb, H. K., Baulin, V. A., Wang, J. Y., Mohammodi, N., Wang, F., Fluke, C. & Crawford, R. J. (2011). Differential attraction and repulsion of Staphylococcus aureus and Pseudomonas aeruginosa on molecularly smooth titanium films. Scientific Reports 1: 165-172.

Webb, H. K., Truong, V. K., Hasan, J., Fluke, C., Crawford, R. J. & Ivanova, E. P. (2011). Roughness parameters for standard description of surface nanoarchitecture. Scanning [In Press].

Webb, H. K., Hasan, J. Truong, V. K., Crawford, R. J. & Ivanova, E. P. (2011). Nature inspired structured surfaces for biomedical applications. Current Medicinal Chemistry 18: 3367-3375.

Webb, H. K., Truong, V. K., Hasan, J., Crawford, R. J. & Ivanova, E. P. (2011). Physico-mechanical characterisation of cells using atomic force microscopy - current research and methodologies. Journal of Microbiological Methods 86: 131-139.

Ivanova, E. P., Webb, H., Christen, R., Zhukova, N. V., Kurilenko, V. V., Kalinovskaya, N. I. & Crawford, R. J. (2010). Celeribacter neptunius gen. nov.

sp. nov., a new member of Alphaproteobacteria. International Journal of Systematic and Evolutionary Microbiology 60: 1620-1625.

Webb, H. K., Crawford, R. J., Sawabe, T. & Ivanova, E. P. (2009). Poly(ethylene terephthalate) polymer surfaces as a substrate for bacterial attachment and biofilm formation. Microbes and Environments 24: 39-42.

Peer-reviewed conference proceedings:

Webb, H. K., Crawford, R. J., and Ivanova, E. P. (2010). Bacterial modifications of the surface chemistry and nanotopography of poly(ethylene terephthalate). International Conference on Nanoscience and Nanotechnology. Sydney, NSW.

Conference presentations with published abstracts:

Webb, H. K., Crawford, R. J., and Ivanova, E. P. (2010). Biodegradation of poly(ethylene terephthalate) plastic by marine bacteria. Australian Society for Microbiology Annual Meeting. Sydney, NSW.

Webb, H. K., Crawford, R. J., and Ivanova, E. P. (2009). Characterisation of two dominating bacterial phylotypes associated with poly(ethylene terephthalate) polymer degradation. Australian Society for Microbiology Annual Meeting. Perth, WA.

Conference posters with published abstracts:

Webb, H. K., Crawford, R. J., Ivanova, E. P. (2011). Design of novel surface nanoarchitectures for controlling bacterial adhesion. Australian Society for Microbiology Annual Meeting. Hobart, Tas.

Ng, H. J., Webb, H. K., Crawford, R. J., Malherbe, F., Patterson, J., Ivanova, E. P. (2011). Alteromonas australica sp. nov., a poly(ethylene terephthalate) degrading bacterium, isolated from the Tasman Sea, Pacific Ocean, Victoria, Australia. Australian Society for Microbiology Annual Meeting. Hobart, Tas.

Ng, H. J., Webb, H. K., Malherbe, F., Patterson, J., Ivanova, E. P. (2011). Development of multilocus sequence analysis (MLSA) and MALDI-ToF mass spectrometry for Alteromonas species classification. Inaugural Meeting of Bergey’s International Society for Microbial Systematics. Beijing.

Webb, H. K., Crawford, R. J., and Ivanova, E. P. (2010). Cell surface wettability and thermodynamic potential for substrate adhesion by poly(ethylene terephthalate) degrading marine bacteria. Australian Colloids and Surface Science Student Conference. Roseworthy, SA.

Webb, H. K., Crawford, R. J., Sawabe, T. and Ivanova, E. P. (2009). Environmental marine bacteria attachment and growth on poly(ethylene terephthalate) polymer surfaces. Australian Colloid and Interface Symposium. Adelaide, SA.

vii

Table of contents

Abstract ...... i Acknowledgements ...... iii Declaration ...... iv List of publications ...... v List of Figures ...... xiii List of Tables ...... xx Chapter 1: Introduction ...... 1 1.1. Overview ...... 2 1.2. Aims of the study ...... 3 Chapter 2: Literature review ...... 5 2.1. Overview ...... 6 2.2. Plastics ...... 6 2.2.1. Environmental impacts of plastic pollution ...... 6 2.2.2. Levels of plastic in the marine environment ...... 7 2.2.3. Poly(ethylene terephthalate)...... 8 2.2.4. Chemistry, synthesis and manufacture ...... 9 2.2.5. Degradation ...... 11 2.2.6. Biodegradable polymers...... 13 2.3. Plastic disposal methods ...... 15 2.3.1. Landfill ...... 15 2.3.2. Incineration ...... 16 2.3.3. Recycling...... 16 2.3.4. Biodegradation ...... 19 2.4. Surface characterisation techniques ...... 20 2.4.1. Atomic force microscopy ...... 20 2.4.2. X-ray photoelectron spectroscopy...... 29 2.4.3. Infrared spectroscopy and Raman spectroscopy ...... 30 2.5. Bacteria ...... 31 2.5.1. Bacterial taxonomy ...... 31 2.5.2. Molecular classification techniques ...... 32 2.5.3. Bacterial attachment and biofilm formation ...... 34 Chapter 3: Materials and Methods ...... 36 viii

3.1. Overview ...... 37 3.2. PET sources ...... 37 3.3. Enrichment conditions ...... 37 3.4. Bacterial recovery ...... 41 3.5. 16S rRNA gene sequencing ...... 41 3.6. Phylogenetic analysis ...... 42 3.7. Biochemical characterisation ...... 43 3.8. Bacterial strains ...... 43 3.8.1. Alteromonas spp...... 43 3.8.1.1. A. addita R10SW13T ...... 44 3.8.1.2. A. genovensis LMG 24078T ...... 44 3.8.1.3. A. hispanica F-32T ...... 44 3.8.1.4. A. litorea TF-22T ...... 44 3.8.1.5. A. macleodii DSM 6062T ...... 45 3.8.1.6. A. marina SW-47T ...... 45 3.8.1.7. A. simiduii BCRC 17572T ...... 45 3.8.1.8. A. stellipolaris LMG 21861T ...... 45 3.8.1.9. A. tagae BCRC 17571T ...... 46 3.8.1.10. Alteromonas sp. H86 ...... 46 3.8.1.11. Alteromonas sp. H89 ...... 46 3.8.1.12. Alteromonas sp. H91 ...... 46 3.8.2. Other bacterial strains ...... 47 3.8.2.1. Celeribacter neptunius H14T ...... 47 3.8.2.2. Glaciecola polaris LMG 21857T ...... 47 3.8.2.3. Idiomarina baltica LMG 21691T ...... 47 3.8.2.4. Jannaschia rubra 4SM3T ...... 48 3.8.2.5. Marinobacter litoralis SW-45T ...... 48 3.8.2.6. Nautella italica LMG 24365T ...... 48 3.8.2.7. Nereida ignava 2SM4T ...... 49 3.8.2.8. Oceanicola marinus AZO-CT ...... 49 3.8.2.9. Pseudoalteromonas citrea NCMB 1889T ...... 49 3.8.2.10. Pseudoalteromonas ruthenica KMM 300T ...... 50 3.8.2.11. Pseudoalteromonas translucida KMM 520T ...... 50

3.8.2.12. Pseudoruegeria aquimaris SW-255T ...... 50 3.8.2.13. Roseisalinus antarcticus EL-88T ...... 50 3.8.2.14. Shewanella japonica KMM 3299T ...... 51 3.8.2.15. Shimia marina CL-TA03T ...... 51 3.8.2.16. Thalassobacter arenae GA2-M15T ...... 51 3.8.2.17. Thalassospira tepidiphila 1-1BT ...... 52 3.8.2.18. Thalassospira sp. H88 ...... 52 3.8.2.19. Thalassospira sp. H93 ...... 52 3.8.2.20. Thalassospira sp. H94 ...... 52 3.9. Biofilm characterisation using confocal laser scanning microscopy ...... 53 3.10. Scanning electron microscopy ...... 53 3.11. Surface characterisation ...... 54 3.11.1. Atomic force microscopy ...... 54 3.11.2. Surface wettability and surface free energy ...... 55 3.11.3. X-ray photoelectron spectroscopy ...... 55 3.11.4. Infrared microspectroscopy ...... 55 Chapter 4: Mechanisms of PET biodegradation ...... 57 4.1. Overview ...... 58 4.2. Long-term PET degradation studies – Enrichment 1 ...... 58 4.2.1. PET degradation after six months ...... 58 4.2.1.1. Topographical analysis ...... 58 4.2.1.2. Chemical analysis ...... 60 4.2.2. PET degradation after ten months ...... 66 4.2.2.1. Topographical analysis ...... 66 4.2.2.2. Chemical analysis ...... 68 4.3. The role of light in bacterial degradation of PET ...... 71 4.3.1. Topographical analysis...... 72 4.3.2. Chemical analysis...... 74 4.4. Degradation quantification and optimisation of conditions ...... 84 4.4.1. Measured loss of mass ...... 85 4.4.2. Topographical analysis...... 86 4.4.3. Chemical analysis...... 88 4.4.4. Determination of optimal degradation conditions...... 92

4.4.4.1. Topography ...... 92 4.4.4.2. Chemistry ...... 97 4.5. Summary ...... 103 Chapter 5: Taxonomic affiliation of bacteria associated with PET biodegradation ...... 104 5.1. Overview ...... 105 5.2. Colony morphology and abundance ...... 105 5.3. Bacterial taxonomy and phylogenetics ...... 109 5.3.1. Sequence similarity of 16S rRNA genes ...... 109 5.3.2. Phylogenetic analysis ...... 115 5.4. Celeribacter neptunius H14T ...... 124 5.5. Alteromonas australica H17 ...... 129 5.6. Summary ...... 133 Chapter 6: Design of surface topographies for enhancing cellular interactions with PET ...... 134 6.1. Overview ...... 135 6.2. Thermodynamics of cell attachment on PET ...... 135 6.3. Superhydrophobic surfaces ...... 139 6.3.1. Bacterial attachment on superhydrophobic titanium surfaces ...... 139 6.3.2. Bacterial attachment on superhydrophobic cicada wings ...... 143 6.4. Molecularly smooth surfaces ...... 146 6.5. Summary ...... 150 Chapter 7: General discussion ...... 152 7.1. Overview ...... 153 7.2. Mechanisms of bacterial biodegradation of PET ...... 154 7.2.1. Physical mechanisms of PET biodegradation ...... 154 7.2.2. Chemical mechanisms of PET biodegradation ...... 155 7.3. Bacterial strains responsible for PET biodegradation ...... 158 7.4. Barriers to PET biodegradation ...... 160 7.5. Strategies for optimisation of PET biodegradation ...... 160 Chapter 8: Conclusions and future directions ...... 164 8.1. Summary and conclusions ...... 165 8.2. Future directions ...... 166 8.3. Close ...... 167 xi

Bibliography ...... 168

xii

List of Figures

Figure 2.1. Chemical reactions of PET manufacturing ...... 10

Figure 2.2. Schematic representation of the effect of polymer crystallinity on enzymatic degradation...... 12

Figure 2.3. General molecular structure of PHAs. R side chains consist of alkyl groups up to 13 carbon atoms long, and the number of consecutive CH2 groups in the polymer backbone (x) ranges from 1 to 4...... 14

Figure 2.4. Technical configuration of an atomic force microscope. A laser is reflected off the back of a flexible cantilever onto a position sensitive detector. As the tip interacts with the sample the cantilever bends, causing deflection of the laser to different sections of the detector...... 21

Figure 2.5. Levels of topography. Topography can vary significantly depending on the scale of analysis...... 22

Figure 2.6. An example of the limitations of Ra for roughness characterization. The two sinusoidal profiles presented above have clearly different structures, however the value of Ra for each is identical...... 25

Figure 2.7. Calculation of bearing ratio, tp, core roughness, Rk, peak height, Rpk, and valley depth, Rvk...... 26

Figure 2.8. Life-cycle of a biofilm ...... 35

Figure 4.1. Two dimensional representations of control and degraded PET surface topography, with single line profiles of the indicated positions...... 59

Figure 4.2. XPS survey spectrum of a control PET surface...... 61

Figure 4.3. XPS survey spectrum of degraded PET surface 1...... 61

Figure 4.4. XPS survey spectrum of degraded PET surface 2...... 62

Figure 4.5. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of a control PET surface as determined by XPS, with fitted components...... 63

Figure 4.6. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of degraded PET sample 1 as determined by XPS, with fitted components...... 64

xiii

Figure 4.7. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of degraded PET sample 1 as determined by XPS, with fitted components...... 65

Figure 4.8. Two dimensional representations of AFM scans of 10 μm × 10 μm (A) and 5 μm × 5 μm (B) areas of PET degraded over a period of 10 months, and corresponding three-dimensional constructions (C, D respectively)...... 67

Figure 4.9. XPS survey spectrum of a PET surface degraded over a period of 10 months...... 68

Figure 4.10. High-resolution spectra of the C1s, O1s, N1s, S2p and Cl2p peaks of a PET surface degraded over a period of 10 months, with fitted components...... 70

Figure 4.11. Three-dimensional reconstructions of approximately 10 μm × 10 μm scanning areas of a PET film incubated in seawater, b PET bottle piece incubated in seawater and c PET film incubated in seawater in dark conditions. All incubations lasted eight months...... 73

Figure 4.12. XPS survey spectra of a PET film incubated in sterile seawater, b PET film incubated in sterile pure water, c PET film incubated in seawater, d PET bottle piece incubated in seawater and e PET film incubated in seawater in dark conditions. All incubations lasted for a period of eight months...... 77

Figure 4.13. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in sterile seawater for eight months. .. 79

Figure 4.14. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in sterile pure water for eight months...... 80

Figure 4.15. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in seawater for eight months...... 81

Figure 4.16. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET bottle pieces incubated in seawater for eight months...... 82

Figure 4.17. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in seawater in dark conditions for eight months...... 83

xiv

Figure 4.18. Average loss of mass of PET measured after three months incubation in various culture conditions. The loss of mass associated with PET samples incubated in bacterial enrichment community culture supplemented with nitrogen and phosphorus was statistically significant...... 86

Figure 4.19. Three-dimensional representations of AFM scan data on PET film samples degraded by the bacterial enrichment community a with, and b without supplementation with nitrogen and phosphorus. Films were degraded over a period of three months. ... 87

Figure 4.20. XPS survey spectra of PET films degraded by a bacterial enrichment community without nitrogen and phosphorus supplementation...... 89

Figure 4.21. XPS survey spectra of PET films degraded by a bacterial enrichment community with nitrogen and phosphorus supplementation...... 89

Figure 4.22. High resolution XPS spectra of PET films degraded by a bacterial enrichment community without nitrogen and phosphorus supplementation...... 91

Figure 4.23. High resolution XPS spectra of PET films degraded by a bacterial enrichment community with nitrogen and phosphorus supplementation...... 91

Figure 4.24. Surface topography of PET incubated under a various conditions over one month. Culture conditions were: a sterile seawater (control), b non-sterile seawater, c non-sterile seawater in dark conditions, d non-sterile seawater supplemented with nitrogen/phosphorus, e films pre-treated with SDS in sterile seawater, f SDS pre-treated films in non-sterile seawater and g pure culture of Pseudoalteromonas citrea...... 96

Figure 4.25. Spatial distribution of chemical groups on the surface of a control PET surface, generated by integration of selected peaks in the IR spectra. Maps are coloured according to variation in intensity, relative to the maximum...... 99

Figure 4.26. Spatial distribution of chemical groups on the surface of a PET surface degraded by bacteria in the presence of light, generated by integration of selected peaks in the IR spectra. Maps are coloured according to variation in intensity, relative to the maximum...... 100

Figure 4.27. Spatial distribution of chemical groups on the surface of a PET surface degraded by bacteria in dark conditions, generated by integration of selected peaks in the IR spectra. Maps are coloured according to variation in intensity, relative to the maximum...... 101 xv

Figure 4.28. Single point spectra of PET samples degraded under a variety of conditions...... 102

Figure 5.1. Abundance of individual bacterial phenotypes recovered from four separate enrichment cultures after eight months...... 107

Figure 5.2. Proportions of bacterial phenotypes recovered from each of the four enrichment cultures...... 108

Figure 5.3. Comparison of the proportions of bacterial phenotypes recovered from PET enrichment cultures in the presence and absence of light...... 109

Figure 5.4. Subtree showing the phylogenetic position of one isolated strain, extracted from a larger tree containing 125 members of the class α-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 116

Figure 5.5. Subtree showing the phylogenetic position of eight isolated strains, extracted from a larger tree containing 125 members of the class α-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 117

Figure 5.6. Subtree showing the phylogenetic position of fourteen isolated strains, extracted from a larger tree containing 125 members of the class α-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 118

Figure 5.7. Subtree showing the phylogenetic position of five isolated strains, extracted from a larger tree containing 34 members of the class β-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 119

Figure 5.8. Subtree showing the phylogenetic position of fourteen isolated strains, extracted from a larger tree containing 98 members of the class γ-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 120

xvi

Figure 5.9. Subtree showing the phylogenetic position of four isolated strains, extracted from a larger tree containing 98 members of the class γ-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 121

Figure 5.10. Subtree showing the phylogenetic position of four isolated strains, extracted from a larger tree containing 98 members of the class γ-proteobacteria. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 122

Figure 5.11. Subtree showing the phylogenetic position of one isolated strain, extracted from a larger tree containing 20 members of the class Flavobacteriia. Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively...... 123

Figure 5.13. Scanning electron micrographs of Celeribacter neptunius H14T at magnifications of 10000× and 100000× (inset)...... 125

Figure 5.14. Phylogenetic position of C. neptunius, based on 16S rRNA gene sequence analysis. Bootstrap percentages above 50% are shown at nodes. Reproduced with permission from Ivanova et al. 2010b...... 126

Figure 5.15. Phylogenetic position of A. australica, based on 16S rRNA gene sequence analysis. Bootstrap percentages above 50% are shown at nodes. The symbols + and X indicate branch agreement by maximum likelihood and maximum parsimony algorithms respectively...... 130

Figure 6.1. Static contact angle of a water droplet on a PET surface...... 136

Figure 6.2. Static water contact angles of droplets resting on bacterial lawns of six representative strains isolated from enrichment experiments...... 137

Figure 6.3. Three-dimensional reconstruction of the bacterial biofilm formed on PET surfaces after nine months of enrichment incubation. Viable cells are stained red xvii

(bottom left) while exopolysaccharides, which form the biofilm matrix are stained green (bottom right). The top image combines both channels and demonstrates the location of viable cells within the biofilm matrix...... 138

Figure 6.4. Staphylococcus aureus cell attachment on titanium surfaces with lotus- inspired hierarchical surface features...... 140

Figure 6.5. Schematic diagram of the hypothetical cause of cell accumulation and ‘bridge’ formation, based on retention of air bubbles within surface features...... 141

Figure 6.6. Scanning electron micrograph of lotus-like titanium surfaces (left) and Raman microspectroscopy map of the spatial distribution of air retained by the surface features (right). Raman map is coloured according to intensity of the water peak, with darker areas indicating larger amounts of air...... 142

Figure 6.7. Dynamics of cell adhesion and proportion of air trapped between surface features on lotus-structured titanium surfaces. Air retention is measured as a percentage of the amount of air present immediately upon immersion...... 143

Figure 6.8. Scanning electron micrograph of a cicada wing surface, titled at an angle of approximately 40°...... 144

Figure 6.9. Cicada wing chemical distribution maps generated based on integrated areas of Infrared absorption spectra. Colour maps indicated the percentage variation of the integrated areas relative to the highest value obtained...... 144

Figure 6.10. Wettability map of a cicada wing surface...... 145

Figure 6.11. Reconstructions of biofilms produced by Staphylococcus aureus (left) and Pseudomonas aeruginosa (right) on the surfaces of 3 nm (top), 12 nm (middle) and 150 nm (bottom) thick titanium films deposited on silicon wafers, based on CLSM analysis. Red areas represent viable cells, and green areas represent the biofilm matrix...... 148

Figure 6.12. Levels of Staphylococcus aureus and Pseudomonas aeruginosa cell retention on 3 nm, 12 nm and 150 nm titanium films sputtered on silicon substrata, expressed as a percentage of the total cells available for attachment in the bacterial suspensions...... 149

Figure 6.13. Schematic representation of the hypothetical mechanism of species- dependent bacterial cell adhesion on molecularly smooth titanium surface...... 150

xviii

Figure 7.1. Peak erosion model whereby larger surface features are preferentially worn away by bacterial enzymes, resulting in smaller surface features and overall smoother surfaces...... 155

Figure 7.2. Excision of ethylene glycol from PET by hydrolysis...... 156

Figure 7.3. Displacement of an ester bond in a PET molecule by formation of an amide bond to a nitrogen-containing molecule...... 157

xix

List of Tables

Table 2.1. Manufacturers and commercial names of PET ...... 9

Table 2.2. Intrinsic properties of PET polymers...... 9

Table 2.3. Minimum requirements of PET flakes to undergo mechanical recycling...... 18

Table 3.1. The range of experimental conditions used for enrichment and quantification/characterisation experiments...... 38

Table 3.2. Nucleotide sequences of primers used for 16S rDNA amplification ...... 42

Table 4.1. Comparative roughness analysis of control and degraded PET surfaces. .... 59

Table 4.2. Elemental composition of control and degraded PET samples...... 62

Table 4.3. Chemical component contributions to the C1s, O1s, N1s and S2p peaks of control and degraded PET surfaces...... 65

Table 4.4. Roughness analysis of different size fields of view of PET samples degraded over ten months ...... 67

Table 4.5. Elemental composition of a PET sample degraded over 10 months...... 68

Table 4.6. Chemical component contributions to the C1s, O1s and N1s peaks of a PET sample degraded over 10 months...... 71

Table 4.7. Comparative roughness analysis of PET samples degraded under various conditions over a period of eight months...... 74

Table 4.8. Elemental composition of PET samples incubated over a period of eight months under different conditions...... 77

Table 4.9. Chemical component contributions to the C1s, O1s, N1s, S2p, Si2p and Sn3d peaks of PET samples degraded over eight months under various incubation conditions...... 84

Table 4.10. Comparative surface roughness analysis of PET films degraded by the bacterial enrichment community with and without nitrogen and phosphorus supplementation...... 88

Table 4.11. Elemental composition of PET films degraded by a bacterial enrichment community with and without nitrogen and phosphorus supplementation...... 90

Table 4.12. Chemical component contributions to the C1s and O1s peaks of PET samples degraded by a bacterial enrichment culture with and without nitrogen and phosphorus supplementation...... 92

Table 4.13. Comparative roughness analysis of PET films degraded over a period of one month in a variety of culture conditions...... 97

Table 5.1. Morphological description of colonies recovered from enrichment cultures after eight months...... 106

Table 5.2. Taxonomic affiliations of isolates assigned to phenotype a...... 110

Table 5.3. Taxonomic affiliations of isolates assigned to phenotype b...... 111

Table 5.4. Taxonomic affiliations of isolates assigned to phenotype c...... 112

Table 5.5. Taxonomic affiliations of isolates assigned to phenotype d...... 113

Table 5.6. Taxonomic affiliations of strains assigned to phenotype e...... 114

Table 5.7. Taxonomic affiliation of strains assigned to low-abundance phenotypes. .. 115

Table 5.8. Susceptibility of Celeribacter neptunius H14T to various antibiotics...... 127

Table 5.9. Comparison of biochemical characteristics of Celeribacter neptunius to those of closely related strains...... 128

Table 5.10. Comparison of biochemical characteristics of Alteromonas australica to that of other species of Alteromonas...... 131

Table 5.11. Susceptibility of Alteromonas australica H17 to various antibiotics...... 132

Table 6.1. Quantitative roughness analysis of titanium thin films deposited on silicon wafer substrata...... 147

xxi

Chapter 1: Introduction

______

1.1. Overview

As a material, poly(ethylene terephthalate) (PET) possesses a number of properties that make it extremely popular for a wide range of applications. PET is strong and durable, chemically and thermally stable, has low gas permeability, high clarity and low electrical conductance, and is easily handled and processed (Kint and Muñoz-Guerra 1999, Awaja and Pavel 2005). These properties, among others, make PET an appealing material for use in a number of applications, such as synthetic fibres and textiles, packaging applications, sheets and films, electronics, automotive parts, houseware, lighting products, power tools, sports goods and photographic and x-ray materials (Awaja and Pavel 2005, Sinha et al. 2010). The primary functions of PET are in food packaging and synthetic fibres; PET accounts for more than half of the synthetic fibre production around the world, and global consumption of the polymer exceeds $17 billion (Sinha et al. .2010)

However some of the very properties that make PET such a popular material also lead to the polymer posing a significant environmental threat, particularly to the world’s oceans (Moore 2008, Ryan et al. 2009). The stability and durability of PET mean that it is exceedingly slow to degrade in the environment, and this coupled with the high popularity and manufacture of PET leads to its accumulation (Derraik 2002, Moore 2008). Also, as plastics often have a tendency to float, much of the PET in the environment is washed away via storm water etc. and ultimately makes its way to the ocean. Once there it can cause considerable harm to marine wildlife and damage to marine environments (Azzarello and Van Vleet 1987, Baird and Hooker 2000, Barreiros and Barcelos 2001, Rios et al. 2007, Gregory 2009, Denuncio et al. 2011, Ivar do Sul et al. 2011).

To date, only three plastic disposal methods are routinely used on a large scale: landfill, incineration and recycling. Each technique has disadvantages and drawbacks. Landfill and incineration both lead to the release of dangerous secondary pollutants into the environment, and landfill also has an additional drawback in the requirement of large portions of land space (Zhang et al. 2004, Simoneit et al. 2005, Urase et al. 2008, Valavanidis et al. 2008). Recycling addresses the environmental concerns of landfill and incineration however the process is relatively inefficient and less cost effective, and

2 subsequently there is less incentive for investment in recycling facilities (Paci and La Mantia 1999, Awaja and Pavel 2005).

Biodegradation is an attractive alternative plastic disposal method which has the potential to be both environmentally friendly and cost effective. However as of yet, no viable biodegradation protocol has been developed for PET. Several attempts have been made in the literature, however the only study in which substantial PET biodegradation was achieved used a protocol that would be unsuitable for large-scale commercial application due to prohibitive expense (Müller et al. 2005).

While considerable work has gone into understanding the abiotic environmental degradation mechanics of PET (Andrady 2011, Raquez et al. 2011), there is little literature available that specifically addresses biological processes in degradation of the polymer. Most studies simply assume that biodegradation of PET occurs by depolymerisation through breakage of the ester bonds, and although this is certainly a possibility there is a tendency not to consider alternatives. Many cases exist in nature where different organisms evolve different techniques to achieve the same goal.

1.2. Aims of the study

The primary objective of this study is to contribute toward the development of biodegradation as an effective and viable alternative PET disposal procedure, using marine bacteria as the degradative agents. As mentioned above, a large proportion of PET waste is found in the ocean, and as such the marine environment is a good prospective source for organisms which are already starting to develop PET degrading metabolic pathways. Three intermediate objectives were identified as stepping stones toward novel PET biodegradation techniques.

The first intermediate objective is to determine whether or not marine bacteria possess the capability for biodegradation of PET, and to assess their potential to do so. This involves characterisation and comparison of PET surfaces both before and after exposure to bacterial interactions, to identify what modification, if any, the bacteria can induce. Achievement of this objective should contribute significantly to the fundamental understanding of the mechanisms by which PET biodegradation occurs.

The second intermediate objective is to characterise the composition of any bacterial communities which prove to be capable of biodegradation of PET. Taxonomic identification of the bacterial species which are required for PET biodegradation will provide clues as to what culture conditions may help to promote degradation. Additionally, an important question to be answered under this objective concerns the number of individual taxonomic strains required for effect biodegradation of PET, i.e. can PET be degraded by a single bacterial species, or does it require the co-operation of an entire bacterial community.

The third intermediate objective is the development of one or more strategies to enhance the rate at which marine bacteria are able to degrade PET. Biodegradation processes need to be as efficient as possible for potential application on a commercial scale. This objective can only be addressed after fundamental knowledge of the mechanisms of biodegradation and the bacterial strains responsible, however depending on their results some possible approaches may be supplementation of growth media with trace nutrients or growth factors, or manipulation of PET surfaces to make them more conducive for biodegradation.

In the coming chapters, the current state of knowledge concerning PET biodegradation will be reviewed, after which an overview of the methods and techniques used in this study is presented. Following this, experimental results detailing the way in which PET biodegradation occurs, which bacterial strains are responsible and possible degradation enhancement strategies are presented, followed by general discussion of the topic and conclusions drawn. This work increases the fundamental understanding of PET biodegradation processes, and provides a significant contribution toward the development of new PET disposal protocols.

Chapter 2: Literature review

______

2.1. Overview

This chapter aims to present a detailed overview of the current state of knowledge in areas that related to biodegradation of poly(ethylene terephthalate) (PET). This includes an outline of the problems associated with plastic pollution in the marine environment, description of the properties and application of PET, analysis of the current state of research on the biodegradation of PET and discussion of some of the analysis techniques that can be used for tracking biodegradation.

2.2. Plastics

2.2.1. Environmental impacts of plastic pollution

Plastic pollution in the marine environment is causes several hazardous and ecologically damaging effects. Plastic debris poses a direct threat to wildlife, with many and varied species documented as being negatively impacted by plastic items. The main dangers associated with plastic objects for most species surround entanglement in and ingestion of said items (Derraik 2002). Juvenile animals in particular, often become entangled in plastic debris, which can result in serious injury as the animal grows (Pemberton et al. 1992, Sazima et al. 2002), not to mention restriction of movement, preventing animals from properly feeding, and, in the case of mammals, breathing (Gregory 2009). A wide variety of species have been reported to be impacted by plastic debris; marine birds (Azzarello and Van Vleet 1987, Blight and Burger 1997), sea turtles (Barreiros and Barcelos 2001), cetaceans (Baird and Hooker 2000), fur seals (Pemberton et al. 1992), sharks (Sazima et al. 2002), and filter feeders (Moore et al. 2001) are just a few of the affected species documented. Marine birds are particularly susceptible to ingestion of plastic objects that they mistake for food (Azzarello and Van Vleet 1987, Blight and Burger 1997). Plastic ingested by these animals persists in the digestive system, and can lead to decreased feeding stimuli, gastrointestinal blockage, decreased secretion of gastric enzymes, and decreased levels of steroid hormones, leading to reproduction problems (Azzarello and Van Vleet 1987).

Plastic particles in the ocean have been shown to contain quite high levels of organic pollutants. Toxic chemicals such as polychlorinated biphenyls (PCBs), nonylphenol (NP), organic pesticides such as DDT, polycyclic aromatic hydrocarbons

(PAHs), polybrominated diphenyl ethers (PBDEs), and bisphenol A (BPA) have been consistently found throughout oceanic plastic debris (Mato et al. 2001, Rios et al. 2007, Hirai et al. 2011). The presence of these compounds further increases the risks associated with ingestion of plastic debris by wildlife, and additionally many of these compounds can undergo significant biomagnification and may potentially pose a risk to humans (Hirai et al. 2011). PCBs are associated with developmental impairment (neurological impairment, growth abnormalities, and change in the onset of puberty), cancer and endocrine disruption (Schecter et al. 2010). PBDEs cause neurobehavioural changes and endocrine disruption (Trudel et al. 2011 ). Organic pesticides, especially DDT, have been linked to arthritis, breast cancer, diabetes, neurobehavioural changes and DNA hypomethylation (Schecter et al. 2010). PAHs are carcinogenic, cytotoxic and mutagenic (Chung et al. 2011), and NP and BPA have been reported to be endocrine disruptors (Zhou et al. 2011).

Plastic debris also provides increased opportunity for dispersal of many potentially invasive species to new environments (Wheeler 1916, Censky et al. 1998, Masó et al. 2003, Barnes and Milner 2005, Gregory 2009). Colonisation and subsequent dispersal of marine species is, unsurprisingly, common, with many reports of barnacles (Barnes and Milner 2005, Gregory 2009), bryozoans (Barnes and Milner 2005, Gregory 2009), polychaetes (Barnes and Milner 2005, Gregory 2009), dinoflagellates (Masó et al. 2003 ), algae (Gregory 2009), and molluscs (Gregory 2009) found adhered to plastic debris. Terrestrial animals are also capable of riding marine debris to new areas, ants have been reported to ride debris from the Brazilian mainland to San Sebastian Island, several kilometres away (Wheeler 1916). There are even examples of animals as big as iguanas riding flotsam to new islands in the Caribbean (Censky et al. 1998).

2.2.2. Levels of plastic in the marine environment

There have been many efforts at quantifying the level of plastic pollution in the marine environment, the vast majority of which focus on the deposition of debris on beaches (Frost and Cullen 1997, Walker et al. 1997, Martinez-Ribes et al. 2007, Morishige et al. 2007, Santos et al. 2008). This is probably due to the fact that plastics generally float, and therefore tend to accumulate on beaches (Santos et al. 2008). The bulk of debris washed onto beaches studied, in terms of numbers, is composed of plastic

7 items; often three quarters of items washed onshore are made of plastic (Morishige et al. 2007, Santos et al. 2008). Such high levels of plastic debris in the environment can be attributed to the high availability of plastic products and its ability to persist in the environment (Frost and Cullen 1997, Ivar do Sul and Costa 2007).

In a study investigating the amount of plastic in the open ocean as opposed to on beaches, the levels of plastic particles was compared to zooplankton in surface waters in the North Pacific Ocean (Moore et al. 2001). Zooplankton outnumbered plastic particles approximately five to one, however, the total plastic collected outweighed zooplankton by approximately 6 times. Other studies have reported similar abundances of plastic pollution. During an expedition in the Atlantic Ocean, the total floating marine debris (including naturally occurring debris such as kelp) was characterised from as far south as the Southern Ocean, all the way north to the high Arctic (Barnes and Milner 2005). In most samplings plastic was the most abundant type of debris, reaching as high as 92% of items collected in one sampling, and plastic was present in every 10° latitudinal belt. A separate study monitoring plastic debris in the Atlantic over a period of 22 years found that 62% of all net tows contained plastic items large enough to be easily detected, and in some regions reaching an average of more than 20000 pieces km-2 (Law et al. 2010). It has been reported that in the North Pacific subtropical gyre plastic levels contain up to 335000 plastic items km-2, or 5.1 kg km-2 ( Ryan et al. 2009).

2.2.3. Poly(ethylene terephthalate)

Poly(ethylene terephthalate) (PET) is a semicrystalline, thermoplastic polyester (Kint and Muñoz-Guerra 1999, Levchik and Weil 2004, Awaja and Pavel 2005, Bergeret et al. 2009 ). It is manufactured by a number of separate companies, under different trade names (Table 2.1). PET is strong and durable, chemically and thermally stable, has low gas permeability, and is easily processed and handled (Kint and Muñoz- Guerra 1999, Awaja and Pavel 2005). This combination of desirable properties makes PET a suitable material for a wide range of properties, and a significant component of worldwide plastic consumption. More than 50% of synthetic fibres produced around the world consist of PET, and global consumption of PET has been reported to exceed $17 billion per year (Sinha et al. 2010). PET is primarily used as fibres, sheets and films, more specifically its uses include food and beverage packaging (esp. softdrink and

8 water bottles), electronics, automotive parts, houseware, lighting products, power tools, sports goods, photographic applications, x-ray sheets and textiles (Kint and Muñoz- Guerra 1999, Awaja and Pavel 2005, Sinha et al. 2010). Depending on the intended application and desired properties, PET can be manufactured to specification by controlling the polymerisation conditions. Some of the specific properties of PET are summarised in Table 2.2.

Table 2.1. Manufacturers and commercial names of PET

Manufacturer Commercial name DSM Engineering Plastics Arnitel® Du Pont De Nemours & Co., Inc. Mylar® Du Pont De Nemours & Co., Inc. Rynite® Eastman Chemical Company Eastapac® ENKA-Glazstoff Diolen® Farbwerke Hoescht AG Hostadur® Imperial Chemical Industries Ltd. Melinex®

Table 2.2. Intrinsic properties of PET polymers

Property Value* Average molecular weight 30000-80000 g mol-1 Density 1.41 g cm-3 Melting temperature 255-265 °C Glass transition temperature 69-115 °C Young’s modulus 1700 MPa Water absorption (24 hrs) 0.5 %

*Values taken from Awaja and Pavel, (2005). Values with ranges indicate properties which vary depending on crystallinity and degree of polymerisation.

2.2.4. Chemistry, synthesis and manufacture

Commercial synthesis of PET begins with one of two chemical reactions: ethylene glycol (EG) is reacted with either (1) terephthalic acid or (2) dimethyl terephthalate (DMT) (Kint and Muñoz-Guerra 1999, Awaja and Pavel 2005). Reaction 1 is performed 9 at 240-260 °C and 300-500 kPa, and reaction 2 is performed at 140-220 °C and 100 kPa, both reactions yield bis(hydroxyethyl) terephthalate (BHET) (Kint and Muñoz- Guerra 1999, Awaja and Pavel 2005). After the initial reaction, two or three polymerisation steps are then performed, depending on the required molecular weight. The first polymerisation step is transesterification between BHET molecules, displacing EG, at 250-280 °C and 2-3 kPa (Amari and Ozaki 2001, Awaja and Pavel 2005). The resulting oligomers are then polycondensed at 270-280 °C and 50-100 kPa (Kint and Muñoz-Guerra 1999, Awaja and Pavel 2005). At this stage the polymer is suitable for applications that do not require high molecular weight chains, however if higher molecular weight is required, the polymer is subjected to a third, solid state polymerisation, at 200-240 °C and 100 kPa (Kint and Muñoz-Guerra 1999, Awaja and Pavel 2005, Sinha et al. 2010). After synthesis of the raw polymer, it can then be processed into the required form, via extrusion, injection or blow moulding (Awaja and Pavel 2005).

Figure 2.1. Chemical reactions of PET manufacturing

2.2.5. Degradation

As a rule, widely used plastics do not naturally degrade to a large degree when released into the environment (Yamada-Onodera et al. 2001, Bonhomme et al. 2003, Zheng et al. 2005, Marqués-Calvo et al. 2006). This is perhaps unsurprising, as one of the primary reasons for the popularity and widespread application of many polymers is their exceptionally high stability and durability (Yamada-Onodera et al. 2001, Zheng et al. 2005). There are four mechanisms by which plastics degrade in the environment: photodegradation, thermooxidative degradation, hydrolytic degradation and biodegradation by microorganisms ( Andrady 2011). Generally speaking, natural degradation of plastic begins with photodegradation, which leads in to thermooxidative degradation. Ultraviolet light from the sun provides the activation energy required to initiate the incorporation of oxygen atoms into the polymer (Andrady 2011, Raquez et al. 2011). This causes the plastic to become brittle and to break into smaller and smaller pieces, until the polymer chains reach sufficiently low molecular weight to be metabolised by microorganisms (Zheng et al. 2005, Andrady 2011). These microbes either convert the carbon in the polymer chains to carbon dioxide or incorporate it into biomolecules (Yamada-Onodera et al. 2001, Andrady 2011). However this entire process is very slow, and it can take 50 or more years for plastic to fully degrade (Müller et al. 2001). This is not aided by the fact that the photodegradative effect is significantly decreased in seawater due to the lower temperature and oxygen availability, and that the rate of hydrolysis of most polymers is insignificant in the ocean (Andrady 2011).

Many studies have investigated the degradability of a wide range of polymers (Yamada-Onodera et al. 2001, Bonhomme et al. 2003, Zhang et al. 2004, Müller et al. 2005, Herzog et al. 2006, Artham and Doble 2009, Kondratowicz and Ukielski 2009). Zheng et al. (2005) observed that in most cases, polymers with pure carbon backbones are particularly resistant to most methods of degradation, but polymers that include heteroatoms in the backbone (e.g. polyesters, polyamines) show higher susceptibility to degradation. While this is often true, there is however a secondary qualifier in the latter case; aromatic polymers tend to be resistant to degradation, despite the presence of bonds that are normally readily hydrolysed (Müller et al. 2001, Zheng et al. 2005). PET is a classic example of such a polymer; the ester bonds that form part of the polymer chain could normally be quite easily broken by a number of mechanisms, however due

11 to its aromatic groups the polymer is essentially non-degradable under normal conditions.

While the PET polymer is particularly stable and durable, it has been shown that microbial communities are capable of utilising diethylene glycol terephthalate (DTP), a subunit of PET, as a sole carbon and energy source (Zhang et al. 2004). This suggests that the reason for the extreme stability arises from being in a polymeric state. There have been a few studies that have established a link between plastic degradability and the degree of crystallisation of the polymer (Kint and Muñoz-Guerra 1999, Marten et al. 2003, Müller et al. 2005, Herzog et al. 2006, Asakuma et al. 2009, Kondratowicz and Ukielski 2009). Increased crystallisation limits chain movement and decreases the availability of polymer chains for degradative agents such as microbial lipases or other ester lysing molecules (Fig. 2.1). Lipase catalysed degradation of poly(hydroxybutyrate- co-valerate) (PHBV) has been shown to occur preferentially in the amorphous regions of the polymer, exposing polymer crystals (Mueller 2006).

Crystalline Amorphous Crystalline

Polymer Bulk

Enzyme (Esterase/Lipase)

Figure 2.2. Schematic representation of the effect of polymer crystallinity on enzymatic degradation.

2.2.6. Biodegradable polymers

Over the last ten years, there has been a shift away from investigation of the degradability of traditional plastics, with more and more emphasis placed on the development of novel biodegradable polymers. Many biodegradable polymers currently exist, both natural and synthetic, however the two major barriers to their incorporation in current plastic-based applications are increased production costs and inferior material properties, e.g. decreased durability (Amass et al. 1998, Zheng et al. 2005). Production costs can be minimised through continued development of manufacturing protocols and increasing efficiency, but substantial research is still required to produce biodegradable polymers with comparable physical properties to conventional plastics. Regardless, some progress has been made in the field of biodegradable plastics, and a number of strategies have emerged for their development.

One technique for production of biodegradable plastics is to attempt to produce materials based on conventional plastics with increased degradability without compromising physical properties. For example, polymers with additional functionality on the polymer chains have been produced by both post-polymerisation treatments (Yamada-Onodera et al. 2001, Artham and Doble 2009) and copolymerisation with equivalent functionalised monomers (Marqués-Calvo et al. 2006). The rationale behind this was to create increased opportunities for microbial enzymes to attack the polymer chains. However, biodegradation of these polymers is still relatively limited, and for degradation to occur there is a requirement for significant energy input, especially in the post-polymerisation treatments. Better degradation rates have been achieved when producing block copolymers of conventional plastics with readily hydrolysable polymeric molecules. Common examples are starch, lactic acid, ethylene glycol, and caprolactone (Kint and Muñoz-Guerra 1999, Russo et al. 2009, Brandelero et al. 2011, Kim et al. 2011b, Raquez et al. 2011). Unfortunately these types of polymers are significantly less durable than many conventional plastics, and in many cases it is unclear whether these polymers are truly biodegraded, or whether they simply disintegrate into small pieces.

Development of plastics based on biological molecules has been a popular area of research. Polymers have been produced based partially or entirely on starch (Russo et al. 2009, Brandelero et al. 2011, Canché-Escamilla et al. 2011, Raquez et al. 2011),

13 lactic acid ( Nampoothiri et al. 2010, Ye et al. 2011), caprolactone ( Neppalli et al. 2011), proteins (Verbeek and van den Berg 2010), cellulose acetate (Puls et al. 2011) and other polysaccharides (Ghasemlou et al. 2011), and in many cases the mechanical durability of these polymers has been improved through addition of plasticisers (Verbeek and van den Berg 2010, Ghasemlou et al. 2011) or nanoparticles (Neppalli et al. 2011), or by controlling production conditions (Brandelero et al. 2011). However the most significant group of biopolymers is poly(hydroxyalkanoates) (PHAs). PHAs are polymeric materials naturally produced by many bacteria and some archaea, which can be processed into a number of forms suitable for packaging, coatings, and biomedical applications (Nitschke et al. 2011). The general structure of PHAs is given below in Figure 2.2. PHAs are produced commercially through bacterial fermentation, although quantities are somewhat limited due to increased production expenses compared to conventional plastics, and the lack of high-value applications (Chen 2009, Akaraonye et al. 2010, Kunasundari and Sudesh 2011). Manufacturers have been able to decrease expenses to some degree by utilising cheaper foodstocks for metabolism by bacteria; PHAs can be produced from waste materials such as whey, wheat and rice bran, molasses, vegetable oil, and even carbon dioxide (Akaraonye et al. 2010), however the main prohibitive expense remains the extraction procedure for recovery of the polymer (Kunasundari and Sudesh 2011). Extraction methods include solvent extraction, chemical digestion, enzymatic treatment, mechanical disruption, supercritical fluid disruption, flotation, gamma irradiation, and two-phase systems, however as of yet no sufficiently inexpensive extraction technique has been developed to allow PHAs to truly compete with conventional plastics in the market (Kunasundari and Sudesh 2011).

R O

CH C

(CH2)x O n

Despite the significant work that has been conducted into the development of biodegradable polymers, it seems unlikely that any polymer that undergoes significant environmental degradation will ever be able to compete with conventional synthetic plastics in terms of material properties and widespread application. The very reason conventional plastics are so popular across a wide range of applications is the fact that they are so physically and chemically stable, which is also the cause for their lack of degradability. Therefore any attempt to increase biodegradability will likely compromise the physical properties of the material to some degree. There is however potential for compromise between degradability and durability on an intended application basis.

2.3. Plastic disposal methods

Currently, there are three main methods for plastic waste handling: burying in landfill, incineration, and recycling (Zhang et al. 2004 ). Each has its own inherent limitations.

2.3.1. Landfill

The first drawback associated with disposal of plastic waste is the fact that landfill facilities occupy space that could be utilised for more productive means such as agriculture (Zhang et al. 2004). This is compounded by the slow degradability of most plastics, as this means the occupied land is unavailable for long periods of time. Plastic components of landfill waste have been shown to persist for more than 20 years (Tansel and Yildiz 2011). This is due to the limited availability of oxygen in landfills; the surrounding environment is essentially anaerobic (Massardier-Nageotte et al. 2006, Tollner et al. 2011). As discussed in Section 2.2.5, the limited degradation that is experienced by many plastics is largely due to thermooxidative degradation (Andrady 2011), and the anaerobic conditions in landfills only serves to further limit degradation rates.

Plastic debris in landfill also acts as a source for a number of secondary environmental pollutants (Zhang et al. 2004). Pollutants of note include volatile

15 organics such as benzene, toluene, xylenes, ethyl benzenes and trimethyl benzenes released both as gases and contained in leachate (Urase et al. 2008), and endocrine disrupting compounds, in particular BPA (Svenson et al. 2009, Tsuchida et al. 2011, Xu et al. 2011a). In addition to its endocrine disruption properties, BPA released from plastics in landfill has also been shown to lead to an increase in production of hydrogen sulphide by sulphate-reducing bacteria in soil populations (Tsuchida et al. 2011). High concentrations of hydrogen sulphide are potentially lethal (Tsuchida et al. 2011 ).

2.3.2. Incineration

Another technique routinely used for disposal of plastic waste is incineration (Zhang et al. 2004). Plastic incineration overcomes some of the limitations placed on landfill in that it does not require any significant space, and there is even the capability for energy recovery in the form of heat (Sinha et al. 2010), however there is a significant trade-off. Incineration of plastics leads to the formation of numerous harmful compounds, most of which is released to the atmosphere (Zhang et al. 2004). PAHs, PCB, heavy metals, toxic carbon- and oxygen-based free radicals, not to mention significant quantities of greenhouse gases, especially carbon dioxide are all produced and released when plastics are incinerated (Simoneit et al. 2005, Valavanidis et al. 2008, Astrup et al. 2009, Khoo and Tan 2010, Shen et al. 2010). The significant environmental drawbacks of plastic disposal via both landfill and incineration led to the development of plastic recycling process.

2.3.3. Recycling

There are two approaches currently widely used for the recycling of PET: chemical and mechanical processing (Awaja and Pavel 2005). Chemical processing of PET is performed by carrying out chemolysis with one of a number of compounds, resulting in depolymerisation of the plastic (Awaja and Pavel 2005). Depolymerisation can be carried out by hydrolysis (using water), methanolysis (methanol), glycolysis (EG), or aminolysis (methylamine, e.g.) ( Awaja and Pavel 2005, Sinha et al. 2010). Each results in different monomer units that can be recovered, but all of the monomers can be used as polymerisation material to produce new plastic. Mechanical processing, however, is

16 generally preferred to chemolysis, because even though it is a complicated process, it has the advantage of being much more cost effective (Awaja and Pavel 2005).

Mechanical recycling of PET consists of several individual steps (Awaja and Pavel 2005). The first of which is the removal of as much contaminating material as possible. The minimum requirements for PET flakes to undergo mechanolytic recycling are listed in Table 2.3. There are a number of contaminants that can significantly inhibit the recycling process and result in severe deterioration of the recovered plastic (Awaja and Pavel 2005). These contaminants will be discussed later in this section. The first step in contamination removal is sorting of plastic waste in order to separate the PET from other plastics. It is of particular importance that PVC be removed from the PET to be recycled, and this is typically done manually (Awaja and Pavel 2005). After sorting the PET is ground into flakes and washed, either using 2% NaOH and detergent at 80 ºC followed by rinsing in cold water, or using tetrachloroethylene (Awaja and Pavel 2005). After washing, the PET must be dried, usually under desiccation at ~170 ºC for six hours (Awaja and Pavel 2005). Washed and dried flakes are then ready to be melted down and extruded into new forms, however it is at this step that the real limitations of mechanolysis arise. Contaminants still reside within the PET flakes, and under high extrusion temperatures they lead to hydrolytic and thermal degradation, and subsequent decreases in both the molecular weight and intrinsic viscosity of the plastic (Awaja and Pavel 2005).

Table 2.3. Minimum requirements of PET flakes to undergo mechanical recycling.

Property Requirement Flake size 0.4 – 8 mm i.d. Melting temperature >240 °C Viscosity [η] >0.7 dl g-1 Water content <0.02 % Dye content <10 ppm PVC content <50 ppm Polyolefin content <10 ppm Metal content <3 ppm Yellowing index* <20

*Yellowing index is a measure of the tendency of polymers to turn yellow through photodegradation

There are a number of contaminants that can significantly limit the effectiveness of PET recycling processes. The first class of contaminants is acids, and acid-producing contaminants, for example PVC, which produces hydrochloric acid (Paci and La Mantia 1999, Awaja and Pavel 2005). Acids act as catalysts for chain cleavage reactions (Cardi et al. 1993, Paci and La Mantia 1998). Similarly, elevated water content can lead to chain breakage through hydrolysis. Most water contamination arises from the washing process, and can be removed by proper drying (Awaja and Pavel 2005). Dyes and colouring agents can lead to undesirable colours in recycled plastic (Awaja and Pavel 2005). Finally, contaminants such as acetaldehyde (a natural degradation product of PET) and other contaminants arising from misuse of PET by consumers (such as storage of fuel, pesticides, other dangerous materials) are potential health hazards in recycled PET products (Villain et al. 1995, Demertzis et al. 1997).

While recycling processes address the environmental shortcomings of both landfill and incineration, it is a relatively expensive and inefficient process (Yamada-Onodera et al. 2001, Zhang et al. 2004). The presence of additives and impurities can complicate the recycling procedure and decrease both the yield and quality of the recovered product (Zhang et al. 2004, Awaja and Pavel 2005).

2.3.4. Biodegradation

It is generally accepted that less than 20% of microorganisms are known to man (Muyzer et al. 1993 , Iranzo et al. 2001). This reason alone provides compelling evidence that there is a vast, largely untapped reservoir of metabolic pathways to be investigated for biotechnological applications. Bacteria, in particular, are diverse in metabolic activity, and highly adaptable (Iranzo et al. 2001, Synnes 2007), which makes them excellent sources of biotechnologically useful products.

The metabolic diversity of bacteria makes them a useful resource for remediation of pollution in the environment (Iranzo et al. 2001). Bacteria have been utilised in the clean-up of oil spills (Head and Swannell 1999, Piedad Díaz et al. 2002, Rosa and Triguis 2007, Hazen et al. 2010), polychlorinated biphenyls (PCBs) (Luigi et al. 2007), and heavy metals such as arsenic, mercury, cadmium and lead (Takeuchi et al. 2007, De et al. 2008). In fact, there are sufficient examples to suggest there are few if any substances that cannot be used at least in part by microbes for metabolic activities (Iranzo et al. 2001). Biodegradation is an attractive alternative to current practices for waste disposal, as it is generally a cheaper process, potentially much more efficient, and does not produce secondary pollutants such as those associated with incineration and landfill ( Head and Swannell 1999, Pieper and Reineke 2000, Iranzo et al. 2001, Luigi et al. 2007, Takeuchi et al. 2007). In some cases, it may even be possible to obtain useful end products with economic benefit from bacterial metabolism of pollutants, for example ethanol for use in biofuels ( Iranzo et al. 2001).

Bioremediation of hydrocarbons, crude oil for example, poses a number of practical difficulties. Bacteria prefer aqueous nutrients (Iranzo et al. 2001), and hydrocarbons are often immiscible with water. Also, hydrocarbons are largely deficient in certain essential elements, namely nitrogen, potassium and phosphorus (Iranzo et al. 2001, Rosa and Triguis 2007). It has been shown in the literature that manually adding these elements in the form of fertiliser or ammonium sulphate can significantly aid the degradation of hydrocarbons by bacteria (Rosa and Triguis 2007). Hydrocarbons also actively interfere with cell membranes, accumulating within and disrupting the phospholipid bilayer (Pieper and Reineke 2000), however some bacteria have even been isolated that resist organic solvents (Pieper and Reineke 2000).

Despite the difficulties and limitations that have been encountered in previous attempts, the potential environmental and economic benefits that would be associated with a viable protocol for plastic biodegradation make the pursuit of this kind of plastic disposal technique a worthwhile one. Couple this with the sheer adaptational potential of bacteria, and not only do you have a worthy cause, but an achievable one. But to maximise the chances of developing an efficient, green plastic waste disposal technique, a detailed understanding of the biodegradation processes need to be attained, and to this end there are two separate components that need to be analysed: i) the bacteria capable of plastic biodegradation and their metabolic pathways and ii) the changes they induce in the chemical and physical properties of the polymer in question.

2.4. Surface characterisation techniques

In order to trace plastic biodegradation, alterations made to the polymer being degraded need to be monitored. As interactions between the material and potential degrading bacteria occur almost exclusively on the outer surface, techniques that characterise material properties at the surface will provide the most informative data. Techniques for physical and chemical characterisation of surfaces are discussed below.

2.4.1. Atomic force microscopy

Atomic Force Microscopy (AFM) is a surface scanning analytical technique which allows for the measurement and mapping of the topographical features of a sample (Binnig et al. 1986, Dorobantu and Gray 2010). A sharp tip on a flexible cantilever is scanned across the sample surface while the interactions between the tip and the surface monitored by measuring the deflections of a laser that reflected off of the back of the cantilever (Fig. 2.3), enabling construction of a three-dimensional representation (Müller and Dufrêne 2008). AFM can be performed in a variety of media, including liquids such as buffers, at nanometre resolution, which make it excellent for use in cell biology applications (Müller and Dufrêne 2008, Dorobantu and Gray 2010, Liu and Wang 2010). There are two standard modes of AFM scanning: contact mode, where the tip is in constant contact with the surface and the piezoelectric scanner maintains constant force between the tip and the surface, and semi-contact mode (also known as

‘tapping’ mode), in which a driving voltage is applied to cause the tip to oscillate, and the amplitude and phase difference between the driving voltage and tip oscillation is monitored (Dorobantu and Gray 2010, Liu and Wang 2010). Tapping mode is generally preferred in microbiological contexts, as lateral shear forces are minimised; contact mode imaging can result in sweeping weakly adsorbed cells and molecules from the surface (Fritz et al. 1995, Engel et al. 1999).

Position-sensitive photodiode detector

Laser

Sample

The apparent topography of any surface, and the classification of its structure both depend on the scale upon which they are examined. For example, consider the surface of a melon. At first glance, the melon appears to be approximately spherical; this is the melon’s shape, or form, and would typically be measured in centimetres. Closer

21 examination of the melon rind reveals numerous bumps and/or cracks on the surface; this level of topography is referred to as surface waviness and is measured in millimetres. Detailed surface analysis of a single bump on the micron scale reveals yet another distinct topography, classified as surface roughness (Fig. 2.4). The thresholds between these three levels of surface structure are somewhat arbitrary and subjective depending on context (Thomas 1998), however when investigating the interactions between cells and material surfaces we are primarily concerned with micron and sub- micron levels of roughness.

Figure 2.5. Levels of topography. Topography can vary significantly depending on the scale of analysis.

In order for adequate characterization and comparison of roughness on any scale, surface topographical data needs to be summarized into one or more quantifiable parameters. A set of 14 roughness parameters was proposed by Stout et al. (Stout et al.

1993) for comprehensive roughness analysis, primarily intended for engineering applications, referred to throughout the literature as the Birmingham 14. These parameters, among others, will be addressed in this section.

A three dimensional surface can be considered to be composed of a number of discrete line profiles; indeed, this is in fact the method by which most analysis techniques reconstruct a surface profile. In turn a profile can be considered to be analogous to a sinusoidal wave, which can be adequately described in terms of just two properties, its wavelength and amplitude (Thomas 1998). Generally, roughness parameters are indicative of one or the other, however there are some that incorporate both. Additionally, while many parameters for describing the roughness of a profile have an equivalent counterpart for describing three dimensional surfaces, some do not, and some parameters exclusively describe surfaces. It should be noted that in most cases the conventional annotation for roughness parameters describing profiles is the letter R, followed by an identifying subscript, e.g. Ra, while surface parameters begin with the letter S (Stout et al. 1993). For the purposes of this review, most annotations will be restricted to profile parameters, and the reader should assume the given descriptions also apply to their surface parameter counterparts, unless otherwise stated.

Amplitude parameters are those that are calculated based on the height values of a given profile or surface, i.e. the ‘amplitude’ of the sinusoidal wave representing a topographical profile. The most commonly cited amplitude parameters describe either the range or distribution of heights. The maximum roughness (Rmax, also referred to as

Rt) is the vertical distance between the highest and lowest points on a topographical profile or surface. Sometimes instead of reporting Rmax, authors may divide it into two components: Rp, the distance from the mean line or plane to the maximum height value, and R m, the distance between the mean line and the lowest height value (Thomas 1998). These three parameters, however, provide minimal characterization of a profile or surface, and may not be truly representative of the data anyway as they are significantly affected by outliers such as surface damage or contamination. The ten point average roughness Rz was developed in response to this, which is calculated as the difference between the average of the five highest peaks and the average of the five lowest valleys (Brune et al. 1997, Thomas 1998). While this does improve the representativeness of the statistic, it does raise the question of how to define and distinguish between individual peaks and valleys, which will be discussed later.

The parameters discussed thus far are all extreme value parameters. However those parameters that describe the distribution of height values are generally more representative of the topographical amplitude. The average roughness, Ra, is the average deviation of the height values from the mean line/plane, and similarly the RMS roughness, Rq is root-mean-square deviation from the mean line/plane, i.e. the standard deviation. These two parameters are by far the most cited statistics for the description of roughness in microbiological papers (Švorčík et al. 2007, Kuo et al. 2008, Mitik-Dineva et al. 2008, Lee et al. 2010, Truong et al. 2010, Joergensen et al. 2011, Kim et al. 2011a, Li et al. 2011, Park and Jeong 2011, Samyn et al. 2011, Santra et al. 2011), and they are essentially interchangeable. For an ideal sinusoidal profile, the ratio of Ra:Rq is 3/2 equal to 2 π (Thomas 1998). In practice this is not always the case, and Rq is more susceptible to outliers (Thomas 1998). Statisticians generally prefer the use of Rq as standard deviation is considered more meaningful than average deviation, however in practical use there is essentially no difference.

Unfortunately, in most biological papers, characterization of roughness is limited to the presentation of Ra and/or Rq (An et al. 1995, Boulangé-Petermann et al. 1997, Li and Logan 2004). However it is common for two surfaces with clearly different structures to have similar values for these two parameters (Fig. 2.5). Ideally, to resolve this one or more spatial parameters would be cited, but there are amplitude parameters commonly available in roughness analysis software that give at least an indication of the spatial variation in height. The skewness, Rsk, and kurtosis, Rku, of the distribution of height values are two such parameters (Stout et al. 1993, Brune et al. 1997, Thomas 1998). Profiles with positive skewness tend toward having higher peaks with shallower, broader valleys, while negative skewness indicates smaller peaks with deep and narrow valleys (Brune et al. 1997). Kurtosis is a measure of the ‘sharpness’ of the height distribution, and although it can also be an indication of the sharpness of the peaks in a profile or surface, it has found little practical use (Thomas 1998). Even though both Rsk and Rku can be indicative of spatial variations in height, they are considered amplitude parameters as they are calculated based solely on the height distribution.

The final group of amplitude parameters that will be discussed here are those that are based on bearing area. Bearing area is the area a surface occupies in a single plane at a given height h above the mean plane, and is often expressed as a fraction of the projected area, i.e. the bearing area ratio tp (Abbott and Firestone 1933, Thomas 1998).

The bearing ratio can be used to find a further three roughness parameters, namely the core roughness, Rk, peak height, Rpk, and valley depth, Rvk (Schneider et al. 1988, Thomas 1998). First, the height is plotted as a function of the bearing area. The tangent with the minimum slope within the middle 40% of the bearing ratios is then extended in both directions until it intersects tp = 0 and tp = 1. The height difference between these two intercepts is the core roughness. To find Rpk, a right-angled triangle is constructed with an area equal to the area enclosed by the curve, the height axis and the upper limit of Rk. The peak height is equal to the length of the vertical edge of the triangle, when the right angle is placed on the intersection of the tangent and the height axis, and the lower right corner touches the curve. Valley depth is found by a similar method, at the lower end of the curve, substituting tp = 1 for the height axis. A graphical representation of the calculation of tp, Rk, Rpk and Rvk is presented in Figure 2.6. These parameters are able to distinguish between different surfaces with similar average roughness, however they have a significant drawback in that they are quite convoluted, and do not directly relate to the structures on the surface.

5 nm Ra = 3.16 nm

1000 nm 0

-5

5 nm Ra = 3.16 nm

1000 nm 0

-5

Figure 2.7. Calculation of bearing ratio, tp, core roughness, Rk, peak height, Rpk, and valley depth, Rvk.

While amplitude parameters are certainly mandatory when characterizing roughness and surface topography, they give adequate description of only the z- dimension. For comprehensive characterization there also needs to be some analysis of the spatial variations in roughness. A relatively simple, and easy to interpret example is summit density, Sds ( Stout et al. 1993). The summit density is simply defined as the number of summits per unit area on a surface, and a corresponding profile parameter

26 could be easily calculated as the number of peaks per unit length. However calculation of this parameter (and several others) can be significantly complicated by the definition of peaks/summits. Most commonly a peak is defined as any point that is higher than its nearest neighbours, i.e. the two points on either side in a profile, and the surrounding four or eight points on a surface (Greenwood 1984, Stout et al. 1993, Thomas 1998). The major drawback of this method is the tendency to overestimate the number of peaks, particularly on very smooth samples where noise can have a greater influence. An alternative method was described for use in calculating the mean spacing between peaks, where a peak was only included in the analysis provided it was more than 1% of

Rmax higher than the preceding valley (Brune et al. 1997).

The periodicity of a surface or profile can be assessed by the construction of one (or more) of several related statistical functions. These functions measure how the correlation of the heights of two points in the profile or surface varies with the distance between them (Brune et al. 1997, Thomas 1998). For example, the autocovariance function (ACVF) is given by:

where L is the length of a profile, τ is the separation interval between two points, and z is the height of a given point (Bennett and Mattsson 1989, Thomas 1998). The ACVF is usually normalized and expressed as the autocorrelation function (ACF) by dividing 2 R(τ) by the variance of the height distribution, i.e. Rq (Peklenik 1967, Thomas 1998, Zemek 2010). The ACF and ACVF both tend toward 0 as τ increases, with relatively random surfaces decaying more rapidly than those with well-defined structure. This enables them to be characterized by a single value, the correlation length, which is defined as the value of τ at which the ACF decays to a given fraction (1/10 and 1/e are sometimes used) of its original value (Thomas 1998). Points that are further apart than the correlation length are considered statistically independent (Thomas 1998).

Another function based on the ACVF is the power spectral density (PSD) function. The PSD is the Fourier transform of the ACVF (Bendat and Piersol 1966, Thomas 1998, Zemek 2010). Its units are height squared per unit frequency, and it is a function of frequency (Brune et al. 1997, Thomas 1998). A PSD of a single line profile generally appears as nothing more than a noisy line, however averaging several profile PSDs

27 often reveals a number of peaks in the function (Brune et al. 1997). These peaks indicate the frequencies (and wavelengths, as f = 1/λ) which have the highest power, or amplitude. The PSD, ACF and ACVF are all particularly useful for identifying patterns that may be obscured by other surface features, and each can be plotted in three dimensions.

Three of the Birmingham 14 parameters are calculated based on the correlation and spectral functions, i.e. the fastest decay autocorrelation length, Sal, the texture aspect ratio, Str, and the texture direction, Std ( Stout et al. 1993). The fastest decay autocorrelation length is the shortest length required for the three-dimensional ACF to decay to 0.2. The texture aspect ratio is also calculated from the 3-D ACF; Str = Sal/the longest correlation length. The texture aspect ratio describes the ratio of the shortest repeating pattern to the longest repeating pattern on the surface, and is always less than or equal to 1 (Tsukada and Sasajima 1983, Stout et al. 1993). The texture direction is defined as the angle between the major direction of the surface texture and the y-axis of the sampled area, and can be calculated from the 3-D PSD (Stout et al. 1993). However, as this parameter is entirely dependent on the orientation of the sampled field, it is of very limited use for characterization or comparison of surface nanoarchitecture.

A few roughness parameters exist which combine aspects of both amplitudinal and spatial characteristics. The average slope, Δa, and the RMS slope, Δq are two of these, and both are relatively simple to calculate (Spragg and Whitehouse 1974, Whitehouse

1994, Brune et al. 1997, Thomas 1998, Zemek 2010). They are equivalent to Ra and Rq respectively, but calculated based on the derivative of the height profile. The average and RMS wavelengths, λa and λq respectively, can easily be found from their corresponding slopes, according to λa = 2πRa/Δa and λq = 2πRq/Δq ( Spragg and Whitehouse 1974, Whitehouse 1994, Brune et al. 1997, Thomas 1998). Two further members of the Birmingham 14 are classified as hybrid parameters: the mean summit curvature, Ssc, and the developed area ratio, Sdr (Stout et al. 1993). The mean summit curvature is defined as the average of the principle curvatures of each of the summits on a surface, and as such requires the summits be previously defined. The developed area ratio is calculated as the ratio of the surface area to the projected surface area, and it fluctuates with both amplitude and spatial variations.

Requirements for the adequate characterization of roughness depend greatly on the specific application, however for determining a minimum standard for use in

28 microbiology it is useful to start with the Birmingham 14. The original 14 parameters chosen by Stout et al. were RMS roughness, Sq, ten point average roughness, Sz, skewness, Ssk, kurtosis, Sku, summit density, Sds, texture aspect ratio, Str, texture direction, Std, fastest decay autocorrelation length, Sal, RMS slope, SΔq, mean summit curvature, Ssc, developed area ratio, Sdr, and three additional parameters not discussed here (Stout et al. 1993). These three additional parameters have been developed primarily for engineering applications, and have no significant relevance to microbiology. Any minimum set of parameters should include descriptors of both the amplitudinal and spatial properties, e.g. Sq and Sds, in order to give an indication of the vertical and the horizontal dimensions of the surface. The developed area ratio should also be included in microbiological contexts, as the substratum contact area available to cells plays a key role in governing their adhesion. Based on this, it is recommended that when describing roughness in microbiological research, the roughness parameters Sq, Sds and Sdr be reported as the absolute minimum standard. Quotation of Sz, Ssk and Str is also encouraged, in addition to any relevant application specific parameters, to complement the minimum set.

2.4.2. X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS) is an analytical technique for determination of the surface chemistry of a material based on the ionisation energies of the atoms on or near the material surface (Hollander and Jolly 1970). The material in question is irradiated with monochromatic x-rays in a vacuum, which causes electrons in the material to escape from their orbitals. Electrons that are near the surface of the material are released into the vacuum where a detector simultaneously counts them and measures their kinetic energy. A spectrum is then constructed by plotting the number of electrons (often measured in counts per second) against their kinetic energy, or ‘binding’ energy. Different elements produce different characteristic peaks in the spectrum, and different chemical configurations of the atoms produce subtle differences in binding energies ( Hollander and Jolly 1970).

XPS is typically performed by first scanning a wide range of binding energies, to first determine the elements present on the samples surface (Leone et al. 2006, Amor et al. 2009). Once the specific elements have been identified, narrow range, high

29 resolution spectra are recorded to identify the contributing chemical species in a given elemental peak of interest. Generally the different contributing components produce significantly overlapping peaks that together form a characteristic peak of the element in question, and in order to be resolved they need to be mathematically determined using a peak-fitting algorithm ( Salvi et al. 1995, Hesse and Denecke 2011).

XPS is an excellent tool for the study of surface chemistry, as it has a very low depth of analysis; the data obtained are usually representative of no more than the first 10 nm of the sample surface ( Gentle and Barnes 2005, Ivanova et al. 2010a ). The major drawbacks associated with the technique lie within the limited lateral spatial resolution (Blomfield 2005), and the complication of spectra to arising from atmospheric contamination of samples (Johansson and Campbell 2004).

2.4.3. Infrared spectroscopy and Raman spectroscopy

An alternative chemical characterisation technique is infrared spectroscopy. Light from the infrared region of the electromagnetic spectrum is shone onto or through a sample, and the decrease in the energy of the beam due to interaction with the sample is measured as a function of the wavelength of the light (Coates 2000, Menzies et al. 2010). Absorption occurs when the frequency of the incident light matches one of the vibrational energies of a chemical bond in the sample. The result is a spectrum that contains absorption peaks that are characteristic of the vibrational energies of the chemical bonds in the sample. The major advantage of infrared spectroscopy is its nondestructive nature (Whelan et al. 2011). Additionally, by performing IR measurements in attenuated total reflection (ATR) mode, the analysis depth can be limited to just the first 1 to 2 µm, causing the generated signal to be more representative of the surface without the obscurement by the bulk material, as is the case in standard transmission mode. Regardless of whether ATR or transmission mode is used, two-dimensional spectral maps with micron-scale spatial resolution can be generated by collecting individual spectra in a grid or array, and integrating the area under a characteristic peak of interest (Croxford et al. 2011, St John et al. 2011). The spatial resolution can be further improved if a Synchrotron is used as the radiation source (Croxford et al. 2011, St John et al. 2011).

Similarly to infrared spectroscopy, Raman spectroscopy relies on the interaction of light with chemical bonds. Raman spectroscopy employs visible, near-infrared or near- ultraviolet light which is shone on the sample (Hartschuh et al. 2003, Sandt et al. 2009). The photons of the incoming light excite electrons in the sample causing them to move into higher orbitals. These electrons then drop back into lower orbitals, however they may not drop into the same orbital from which they originated, resulting in the emission of a photon of a different energy level to the excitation light (Dietzek et al. 2010). These photons are the measured signal in Raman spectroscopy. Like infrared spectroscopy, Raman spectroscopy is non-destructive and applicable to chemical mapping, and the two techniques are complementary to each other (Pappas et al. 2000, Ivleva et al. 2009).

2.5. Bacteria

While characterisation of the PET surfaces and determination of the biodegradation mechanisms is of great importance to the current project, another significant aspect lies in the identification and characterisation of the bacterial communities that have potential PET-degrading ability. There are a number of approaches that have been developed for taxonomical characterisation of bacteria.

2.5.1. Bacterial taxonomy

The traditional methods for classification of bacteria into species via observation of morphology and other phenotypic traits have some limitations. Firstly, there can be a great deal of ambiguity when classifying bacteria by characteristics such as morphology; there really is not much that can be deduced about a particular bacterium by observing its morphology (Olsen et al. 1986). Secondly, classification by phenotypic traits does not necessarily reflect genetic relatedness or evolutionary pathways (Joyce et al. 2002). One of the major reasons for this is bacteria undergo a significant amount of lateral transfer of genetic material, which can easily introduce new traits to a specific bacterium (Joyce et al. 2002). Another, practical difficulty in classifying organisms in this way is that it requires a pure culture of the microorganism in question to properly distinguish it from others (Amann et al. 1995). The vast majority of microorganisms are currently unable to be cultured properly in the laboratory (Olsen et al. 1986, Muyzer et

31 al. 1993, Ravenschlag et al. 2001, Harris et al. 2004, Carrigg et al. 2007, Horisawa et al. 2008, Malik et al. 2008). Despite these limitations, however, phenotypic classification is a useful approach for rapid and easy grouping of bacteria, provided that the validity of those groups is confirmed by other, more reliable techniques. It is for this reason that molecular methods of classification based on genetic analysis have been developed.

2.5.2. Molecular classification techniques

Molecular methods of characterisation focus on the comparison of various homologous molecules common to all species. Some methods have been developed which compare homologous proteins in order to determine genetic relatedness, however these are not ideal as there can be a large amount of variability even between homologous proteins, especially those from eukaryotes compared with those from prokaryotes (Olsen et al. 1986). Much more success has been experienced using nucleic acids, in particular sequence comparison of ribosomal RNA (rRNA) or genes encoding for rRNA (Olsen et al. 1986, Amann et al. 1995). Ribosomal RNA and DNA have at least five major advantages that make them particularly useful for taxonomic classification of bacteria:

i. rRNA is homologous among all organisms (Olsen et al. 1986) ii. rRNA is extremely well conserved from species to species (Olsen et al. 1986, Malik et al. 2008) iii. rRNA and rDNA are readily isolated (Olsen et al. 1986) iv. Enough data can be generated to draw statistically significant conclusions (Olsen et al. 1986) v. rRNA genes generally are not transferred laterally, which is very useful for tracing evolutionary pathways (Olsen et al. 1986).

The 5S and 16S ribosomal subunits are the most commonly employed (Olsen et al. 1986, Amann et al. 1995). The 5S subunit is particularly easy to sequence, however it is too small to provide useful information for phylogenetic classification (Olsen et al. 1986, Amann et al. 1995). On the other hand, the 16S subunit provides ample information for phylogenetic classification, and can be sequenced with relative ease (Olsen et al. 1986, Burgess et al. 1993, Amann et al. 1995). Also, by incorporating

32 polymerase chain reaction (PCR) protocols into molecular classification methods time consuming and labour-intensive techniques such as cloning into plasmids can be avoided ( Olsen et al. 1986, Amann et al. 1995, Carrigg et al. 2007). In addition to this, PCR techniques have increased sensitivity in comparison to plate count techniques (Devos et al. 2005), and are able to identify species that constitute as little as 1% of the microbial population (Muyzer et al. 1993). The highly conserved nature of the 16S gene allows for PCR amplification of the 16S rDNA of an entire microbial population using a single pair of universal primers (Muyzer et al. 1993 ). The resultant mixture of amplified DNA can be separated according to base composition via denaturing gradient gel electrophoresis (DGGE) (Muyzer et al. 1993, Carrigg et al. 2007). The resulting set of bands is characteristic of the microbial population, with each band representing a discrete species (Muyzer et al. 1993). There are however some limitations associated with DGGE; slight sequence variation within a species may result in multiple bands that correspond to a single species (Carrigg et al. 2007), and the amplified products of more than one bacterial species may have identical mobility (Carrigg et al. 2007, Malik et al. 2008).

Other useful applications of PCR include reverse-transcription PCR and real-time PCR. Reverse transcription PCR uses RNA as the initial template, and includes an initial transcription using reverse transcriptase, providing information on gene expression (Amann et al. 1995). Real-time PCR is a useful tool for quantification of microbial community constituents (Heid et al. 1996, Carrigg et al. 2007, Malik et al. 2008, De Gregoris et al. 2011). A fluorescent dye is incorporated into the amplification reaction that intercalates double stranded DNA. By measuring the increase in fluorescence over time and comparing it to a standard curve of known quantities of DNA, the original concentration of DNA before amplification can be determined (Heid et al. 1996). Real-time PCR is fast, sensitive, accurate, and can be automated, but does not provide information on gene expression (Heid et al. 1996).

Alternative characterisation methods have also been developed. Fluorescent in situ hybridisation (FISH) involves the addition of a fluorescently labelled oligonucleotide probe complimentary to rRNA to microorganisms as they occur naturally in the environment, i.e. without any cultivation required (Amann et al. 1995, Malik et al. 2008). Use of in situ hybridisation techniques allows for the detection of individual cells and observations of morphology and spatial distribution (Amann et al. 1995, Malik et

33 al. 2008). There are disadvantages associated with FISH; as it is often used to detect low numbers of cells, the signal generated is often of low intensity, and it is often difficult to find an individual cell on the microscope stage (Amann et al. 1995). Also, signal intensity does not entirely correlate with cell numbers, as rRNA copies vary between species, faster growing species tending to have far more, resulting in increased signal strength (Amann et al. 1995).

Microarrays are a powerful tool that can be used for comparing genomic composition, and subsequently genetic relatedness (Joyce et al. 2002). Microarrays screen for the presence of many genes at once via immobilised probes (Joyce et al. 2002). This genomic-scale analysis along with high throughput are the main advantages of microarrays, but as they rely on complimentary probes, it can be difficult to obtain useful information on bacteria that have not previously had their genomes sequenced (Joyce et al. 2002). Also, while microarrays are very useful when analysing pure cultures, environmental samples introduce problems in specificity, sensitivity and quantification (Joyce et al. 2002).

2.5.3. Bacterial attachment and biofilm formation

Bacterial interactions with PET are limited almost entirely to the material surface. As such, these interactions need to be characterised, and the methods of cell adhesion and colonisation investigated. One of the primary modes of bacterial colonisation of surfaces is the formation of biofilms. Biofilms are microbial communities enclosed in a matrix of extracellular polymeric substances (EPS), adhered to a biological or non- biological surface (Armstrong et al. 2001, Hall-Stoodley et al. 2004, Türetgen and Cotuk 2007). Biofilms are a common and ancient style of bacterial growth, with fossil evidence of biofilms dating back approximately 3.25 billion years (Hall-Stoodley et al. 2004). They provide a number of advantages for the bacteria they contain. Biofilm formation provides protection for microorganisms from hostile environments, be it extreme conditions such as high temperatures, or immune response from an infected host (Yan et al. 2002, Hall-Stoodley et al. 2004, Palmer et al. 2007, Türetgen and Cotuk 2007). Biofilms also allow for better access to nutrients, as potentially nourishing organic material tends to accumulate on surfaces (Amann et al. 1995, Hall-Stoodley et al. 2004, Palmer et al. 2007). The formation of channels also allows the bulk liquid to

34 infuse a biofilm and improve the exchange of nutrients and waste products (Hall- Stoodley et al. 2004). The close proximity in which cells are kept promotes genetic exchange (Palmer et al. 2007) as well as the development of cell-cell communication systems (Hall-Stoodley et al. 2004). Finally, detachment of a biofilm (or pieces thereof) aids in the colonisation of new areas and ecological niches (Hall-Stoodley et al. 2004, Landini et al. 2010, Nijland et al. 2010).

Figure 2.8. Life-cycle of a biofilm

The adhesion potential between a bacterial cell and a substratum surface is governed by two general factors: the physico-chemical properties of the bacterium, and the physico-chemical properties of the substratum (Blenkinsopp and Costerton 1991, An and Friedman 1998, Bos et al. 1999b, Burgess et al. 2003, Ubbink and Schär- Zammaretti 2007, Busscher et al. 2010). While substantial research has been conducted investigating the effects of substratum surface chemistry on bacterial adhesion, the recent trend has been toward analysing the role of surface topography (Bos et al. 1999b, Katsikogianni et al. 2008, Katsikogianni and Missirlis 2010, Ploux et al. 2010, Bazaka et al. 2011, Parreira et al. 2011). The main reason for this shift of focus arises from inconsistencies in colonization rates on chemically similar surfaces; two samples with virtually identical surface chemistry may exhibit vastly different cell adhesion profiles due to differences in surface architecture and roughness.

Chapter 3: Materials and Methods

______

3.1. Overview

The nature of this work requires the implementation of many different analytical techniques from multiple disciplines, ranging from molecular biology to surface chemistry. Experiments have focussed on the characterisation of the bacterial strains and PET polymers, as well as the interactions between these two entities.

3.2. PET sources

PET samples were obtained from two separate sources. The first source was standard PET water bottles (Mount Franklin Spring Water, 400 mL, Coca-Cola Amatil, Australia), commonly sold at many retail stores in Australia. PET bottles were cut into approximately 1 cm2 pieces, and sterilised by placing them in 70 % ethanol for 20 minutes before use. The second sample type was Mylar® PET films, kindly supplied by the Australian office of DuPont Teijin Films, Melbourne. Mylar® films were prepared and sterilised in the same manner as the PET bottles.

3.3. Enrichment conditions

Two separate experiments were performed, following the same general procedure, to enrich environmental populations of bacteria selecting for PET degrading strains. Environmental seawater samples were collected from Port Phillip Bay, Melbourne, at St. Kilda Beach, on December 12, 2007, and December 17, 2008. Samples were collected from the first metre of the water surface, into sterile glass bottles, and were used on the day of collection. Enrichment culture vessels (pre-sterilised conical flasks, 200 mL) were prepared by adding 100 mL of the collected seawater, and supplying the naturally present bacterial communities with approximately five grams of PET pieces (either bottle pieces or films), as their sole carbon and energy source. Several culture conditions were tested between the two enrichment experiments, shown in Table 3.1. Cultures were incubated statically at room temperature for up to a year. In addition to the two separate enrichment experiments, two quantification/characterisation experiments were performed in order to measure the biodegradation rate and determine the mechanisms of biodegradation. Individual PET pieces were weighed before and after incubation to detect any loss of mass. The latter two experiments also tested a

37 number of culture conditions and chemical supplements (Table 3.1) for their ability to enhance biodegradation.

Table 3.1. The range of experimental conditions used for enrichment and quantification/characterisation experiments.

Experiment Bacterial culture Type of PET Light/ Additives/ Replicates water sample Dark* supplements† Enrichment 1 Environmental Seawater Bottle Light None 2 community pieces Enrichment 1 Environmental Seawater None Light None 1 community (Control) Enrichment 2 Environmental Seawater Films Light None 3 community Enrichment 2 None (Control) Sterilised Films Light None 3 seawater Enrichment 2 None (Control) MilliQ Films Light None 3 water Enrichment 2 Environmental Seawater Bottle Light None 3 community pieces Enrichment 2 Environmental Seawater Films Dark None 3 community Enrichment 2 Environmental Seawater None Light None 3 community (Control) Quantification None (Control) Sterilised Single Light None 5 /characterisation seawater film, (1 × (Control) 3 cm)

Quantification None (Control) Sterilised Single Light NaNO3 5 /characterisation seawater film, (1 × KH2PO4 3 cm)

Quantification None (Control) Sterilised Single Light FeCl2 5 /characterisation seawater film, (1 × 3 cm) Quantification None (Control) Sterilised Single Light Tween 20 5 /characterisation seawater film, (1 × 3 cm)

Quantification None (Control) Sterilised Single Light NaNO3 5 /characterisation seawater film, (1 × KH2PO4 3 cm) FeCl2 Tween 20 Quantification Thalassospira sp. Sterilised Single Light None 5 /characterisation H94 seawater film, (1 × (Control) 3 cm)

Quantification Thalassospira sp. Sterilised Single Light NaNO3 5 /characterisation H94 seawater film, (1 × KH2PO4 3 cm)

Quantification Thalassospira sp. Sterilised Single Light FeCl2 5 /characterisation H94 seawater film, (1 × 3 cm) Quantification Thalassospira sp. Sterilised Single Light Tween 20 5 /characterisation H94 seawater film, (1 × 3 cm)

Quantification Thalassospira sp. Sterilised Single Light NaNO3 5 /characterisation H94 seawater film, (1 × KH2PO4 3 cm) FeCl2 Tween 20 Quantification Thalassospira Sterilised Single Light None 5 /characterisation tepidiphila 1-1BT seawater film, (1 × (Control) 3 cm)

Quantification Thalassospira Sterilised Single Light NaNO3 5 T /characterisation tepidiphila 1-1B seawater film, (1 × KH2PO4 3 cm)

Quantification Thalassospira Sterilised Single Light FeCl2 5 /characterisation tepidiphila 1-1BT seawater film, (1 × 3 cm) Quantification Thalassospira Sterilised Single Light Tween 20 5 /characterisation tepidiphila 1-1BT seawater film, (1 × 3 cm)

Quantification Thalassospira Sterilised Single Light NaNO3 5 T /characterisation tepidiphila 1-1B seawater film, (1 × KH2PO4 3 cm) FeCl2 Tween 20 Quantification Thalassospira Sterilised Single Light None 5 /characterisation /Alteromonas seawater film, (1 × (Control) consortium‡ 3 cm)

Quantification Thalassospira Sterilised Single Light NaNO3 5 /characterisation /Alteromonas seawater film, (1 × KH2PO4 consortium‡ 3 cm)

Quantification Thalassospira Sterilised Single Light FeCl2 5 /characterisation /Alteromonas seawater film, (1 × consortium‡ 3 cm) Quantification Thalassospira Sterilised Single Light Tween 20 5 /characterisation /Alteromonas seawater film, (1 × consortium‡ 3 cm)

Quantification Thalassospira Sterilised Single Light NaNO3 5 /characterisation /Alteromonas seawater film, (1 × KH2PO4 ‡ consortium 3 cm) FeCl2 Tween 20 Quantification Enrichment Sterilised Single Light None 5 /characterisation culture§ seawater film, (1 × (Control) 3 cm)

Quantification Enrichment Sterilised Single Light NaNO3 5 § /characterisation culture seawater film, (1 × KH2PO4 3 cm)

Quantification Enrichment Sterilised Single Light FeCl2 5 /characterisation culture§ seawater film, (1 × 3 cm)

Quantification Enrichment Sterilised Single Light Tween 20 5 /characterisation culture§ seawater film, (1 × 3 cm)

Quantification Enrichment Sterilised Single Light NaNO3 5 § /characterisation culture seawater film, (1 × KH2PO4 3 cm) FeCl2 Tween 20 Optimisation Environmental Seawater Single Light None 5 community film, (1 × 3 cm) Optimisation None (Control) Sterilised Single Light None 5 seawater film, (1 × 3 cm) Optimisation Environmental Seawater Single Dark None 5 community film, (1 × 3 cm) Optimisation None (Control) Sterilised Single Dark None 5 seawater film, (1 × 3 cm)

Optimisation Environmental Seawater Single Light NaNO3 5 community film, (1 × KH2PO4 3 cm)

Optimisation None (Control) Sterilised Single Light NaNO3 5 seawater film, (1 × KH2PO4 3 cm) Optimisation Environmental Seawater Single Light PET pre- 5 community film, (1 × treated with 3 cm) SDS||

Optimisation None (Control) Sterilised Single Light PET pre- 5 seawater film, (1 × treated with 3 cm) SDS||

Optimisation Pseudoalteromonas Sterilised Single Light None 5 citrea NCMB seawater film, (1 × 1889T 3 cm) Optimisation Environmental Seawater Single Light None 5 community film, (1 × 3 cm)

*Light/Dark refers to sample incubation in the presence/ absence of light, respectively.

†Concentration of additives were as follows: sodium nitrate – 0.4 µM, potassium dihydrogen phosphate – 0.04 µM, iron (II) chloride – 100 µM, Tween 20 – 0.04 % (v/v)

‡ Thalassospira/Alteromonas consortium consisted of a mixture of Thalassospira sp. H88, Thalassospira sp. H93, Thalassospira sp. H94, Alteromonas sp. H86,

Alteromonas sp. H89 and Alteromonas sp.H91, in a ratio of 10:10:10:1:1:1, at a final OD600 = 0.3.

§ Tubes were inoculated with 100 µL of enrichment culture taken from one of the light exposed PET + seawater replicates in enrichment experiment 2.

|| PET samples pre-treated with sodium dodecylsulfate (SDS) were cleaned by sonication in 70 % EtOH, rinsed, dried, and soaked in a solution of 2 g/L SDS for 24 hours.

3.4. Bacterial recovery

Subsamples of 30 µL volume were aseptically taken from incubation vessels after gentle swirling and spread on marine agar 2216 (Difco, BD, U.S.A.), or oligotrophic marine agar at 0.1× normal working concentration. A similar procedure was followed for recovery of strains attached to plastic pieces: a single piece of PET was removed from the incubation vessels and vortexed for 30 seconds with 3 mL of phosphate buffered saline. Thirty µL samples were taken from the resulting mixture. Three replicate plates were performed for each growth medium for each incubation vessel. Plates were incubated for up to 72 hours at 25 °C, before the resulting colonies were counted and grouped according to morphology. Multiple representatives of each phenotype were streaked on the same medium from which they were taken and incubated at 25 °C to assess purity, before stock cultures were prepared by inoculating 500 µL of marine broth 2216 (Difco, BD, U.S.A.) supplemented with 20 % glycerol, and storing at –80 °C until required.

3.5. 16S rRNA gene sequencing

Pure cultures of each strain to be sequenced were grown in 3 mL of marine broth. The cells were then centrifuged (12000 rpm for 3 minutes) and the genomic DNA was extracted using Wizard SV genomic DNA extraction kit (Promega, U.S.A.). Approximately 880 bp of the 16S rRNA genes for each strain was amplified via polymerase chain reaction (PCR) using the primer pair (sequences shown in Table 3.2) 27F ( Chun and Goodfellow 1995) and 907R (Tait et al. 2007), and a temperature program consisting of an initial denaturation step at 95 °C for 4 min, then 30

41 amplification cycles of 94 °C for 30 s, 52 °C for 40 s and 72 °C for 70 s, and a final extension step of 72 °C for 20 min. The resulting 16S rRNA gene fragments were sequenced at the Australian Genome Research Facility (AGRF) using BigDye® Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Melbourne) and capillary electrophoresis. Amplification and sequencing primers were purchased from Invitrogen (U.S.A.). Based on the degree of similarity of each the returned sequences to their nearest match, obtained via BLAST analysis, four strains were selected for full 16S rRNA gene sequencing based on their likelihood of being novel species. Full gene sequences (~1500 bp) were obtained by amplifying the first ~880 bp and last ~900 bp (~280 bp overlap), using the above amplification conditions for the former and the following conditions for the latter: primer pair 518 F (Tait et al. 2007) and 1482R (Tait et al. 2007) and temperature program denaturation – 95 ° C for 15 min, amplification – 30 cycles of 94 ° C for 1 min, 55 °C – 1 min and 72 °C – 2 min, and final extension – 72 °C for 10 min. Amplification products were again sequenced at AGRF, and reconstructed using BioEdit software (Hall 1999). The 16S rRNA nucleotide sequences were deposited to the GenBank DNA database.

Table 3.2. Nucleotide sequences of primers used for 16S rDNA amplification

Primer Sequence

27F 5′-AGAGTTTGATCCTGGCTCAG-3′

907R 5′-CCGTCAATTCMTTTGAGTTT-3′

518F 5′-CCAGCAGCCGCGGTAATACG-3′

1482R 5′-GGTTACCTTGTTACGACTT-3′

3.6. Phylogenetic analysis

For each sequence obtained, 30-50 of the most similar sequences from the EzTaxon database of type strains (Chun et al. 2007) were retrieved and aligned using ClustalW2 (Larkin et al. 2007). Alignments were then manually checked and sequences trimmed using BioEdit (Hall 1999). Neighbour-joining phylogenetic trees were constructed using

42 the software available as part of PHYLIP (Phylogeny Inference Package) (Felsenstein 1989). Large trees were first constructed including 100+ different taxonomic strains, before smaller trees were constructed by the same method using approximately 10 of the nearest phylogenetic neighbours identified by the larger trees. One thousand replicates were generated for bootstrap analysis using the SEQBOOT program, distance matrices calculated using DNADIST, neighbour joining analysis performed by NEIGHBOUR using the Kimura 2-parameter algorithm, and replicate trees condensed using CONSENSE, all of which are part of PHYLIP.

3.7. Biochemical characterisation

Phenotypic profiles were generated using Biolog GN2 microplates (Biolog, U.S.A.) and API ZYM test strips (bioMérieux, France). Biolog GN2 microplates test bacterial strains for oxidative utilisation of 95 different carbons sources. API ZYM strips test for activity of 18 separate enzymes, namely: alkaline phosphatase, esterase (C4), esterase lipase (C8), lipase (C14), leucine arylamidase, valine arylamidase, cysteine arylamidase, trypsin, α-chymotrypsin, acid phosphatase, napthol-AS-B1-phosphohydrolase, α- galactosidase, β-galactosidase, β-glucuronidase, α-glucosidase, N-acetyl-β- glucosaminidase, α-mannosidase and β-fucosidase. Both kits were utilised according to the manufacturer’s specifications.

3.8. Bacterial strains

Many different bacterial strains were employed throughout this work, including the type strains of all validly described species of Alteromonas. As so many of the strains to be described belong to the same genus, it is perhaps convenient to present them in two groups, i.e. Alteromonas spp., and other bacterial strains.

3.8.1. Alteromonas spp.

All nine validly described species of the genus Alteromonas were used in the course of this work. Alteromonas spp. belong to the class Gammaproteobacteria, order Alteromonadales, family Alteromonadaceae. They are Gram negative, rod-shaped,

43 chemoorganotrophic bacteria, and grow strictly in aerobic conditions. Cells are motile by means of single polar flagella. They do not produce spores or, in most cases, accumulate polyhydroxybutyrate, however many species produces buds and prosthecae. Most species were originally isolated from open sea or coastal waters (Baumann et al. 1972, Gauthier et al. 1995, Van Trappen et al. 2004a ). Specific descriptions are given in Sections 3.8.1.1-9 below.

3.8.1.1. A. addita R10SW13T

A. addita are rod-shaped cells 0.7 – 0.9 µm in diameter, originally isolated from a seawater sample. They can grow in 1 – 10 % (w/v) NaCl concentrations, between 4 – 37 °C and pH 6.0 – 10.0 (optimum 7.5 – 8.0). The DNA G + C content is 43.0 mol% (Ivanova et al. 2005a ).

3.8.1.2. A. genovensis LMG 24078T

A. genovensis was originally isolated from an electroactive, marine biofilm. Cells are 1.8 µm long and 0.9 µm wide, and produce buds and prosthecae. Growth conditions are 2 – 15 % (w/v) NaCl, 4 – 37 °C and pH 6.0 – 8.5 (optimum 7.0 – 8.0). The DNA G + C content is 44.5 mol% (Vandecandelaere et al. 2008).

3.8.1.3. A. hispanica F-32T

A. hispanica was originally isolated from a hypersaline lagoon in Spain. Cells are 1.0 – 2.0 µm in length and 0.75 µm wide. Cells also produce buds and prosthecae, and accumulate polyhydroxybutyrate. Growth conditions are 7.5 – 15 % (w/v) NaCl (optimum 7.5 – 10 %), 4 – 40 °C (optimum 32 °C) and pH 5.0 – 10.0 (optimum 7.0 – 8.0). The DNA G + C content is 46.3 mol% ( Martínez-Checa et al. 2005).

3.8.1.4. A. litorea TF-22T

A. litorea was originally isolated from intertidal sediment. Cells are 2.0 – 4.0 µm in length and 0.9 – 1.2 µm wide. Growth conditions are 2 – 14 % (w/v) NaCl (optimum 2

– 5 %), 10 – 43 °C (optimum 30 – 37 °C) and pH 5.5 – 8.0 (optimum 7.0 – 8.0). The DNA G + C content is 46.0 mol% (Yoon et al. 2004).

3.8.1.5. A. macleodii DSM 6062T

A. macleodii is the type species of the genera, and was originally isolated seawater samples. Cells are 2 – 3 µm in length and 0.7 – 1.0 µm wide, and produce buds and prosthecae. Growth conditions are 1 – 12 % (w/v) NaCl, and 12 – 45 °C. The DNA G + C content is 45.6 ± 0.8 mol% ( Baumann et al. 1972, Gauthier et al. 1995, Vandecandelaere et al. 2008).

3.8.1.6. A. marina SW-47T

A. marina was originally isolated from a seawater sample. Cells are 2.5 – 4.0 µm in length and 1.0 – 1.2 µm wide, and produce buds and prosthecae. Growth conditions are 2 – 15 % (w/v) NaCl (optimum 2 – 5 %), 4 – 44 °C (optimum 30 – 37 °C) and pH 5.0 – 8. (optimum 7.0 – 8.0). Their DNA G + C content is 45.0 mol% (Yoon et al. 2003a).

3.8.1.7. A. simiduii BCRC 17572T

A. simiduii was originally isolated from a Taiwanese estuary, and possesses the ability to reduce mercury (Hg2+ → Hg0). Cells are 1.2 – 2.5 µm in length and 0.4 – 0.8 µm wide. Cells also produce buds and prosthecae, and accumulate polyhydroxybutyrate. Growth conditions are 0.5 – 13 % (w/v) NaCl (optimum 2 – 4 %), 15 – 40 °C (optimum 30 °C) and pH 6.0 – 9.0. Their DNA G + C content is 45.3 mol% (Chiu et al. 2007).

3.8.1.8. A. stellipolaris LMG 21861T

A. stellipolaris was originally isolated from an Antarctic seawater sample. Cells are 2.0 – 7.0 µm in length and 0.4 µm wide, and produce buds and prosthecae. Growth conditions are 1 – 10 % (w/v) NaCl, 5 – 37 °C and pH 6.0 – 9.0 (optimum 7.0 – 8.5). The DNA G + C content is 44.0 ± 1.0 mol% (Van Trappen et al. 2004a).

3.8.1.9. A. tagae BCRC 17571T

A. tagae was originally isolated from a Taiwanese estuary, and possesses the ability to reduce mercury (Hg2+ → Hg0). Cells are 1.2 – 2.5 µm in length and 0.5 – 0.9 µm wide. Cells also produce buds and prosthecae, and accumulate polyhydroxybutyrate. Growth conditions are 0.5 – 13 % (w/v) NaCl (optimum 2 – 4 %), 15 – 40 °C (optimum 30 °C) and pH 6.0 – 9.0. Their DNA G + C content is 43.1 mol% (Chiu et al. 2007).

3.8.1.10. Alteromonas sp. H86

The bacterial strain designated H86 was isolated from the first enrichment experiment after six months of incubation. Colonies of H86 grown on marine agar after two days were white, opaque, convex and circular, 2 – 3 mm in diameter. Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Alteromonas.

3.8.1.11. Alteromonas sp. H89

The bacterial strain designated H89 was isolated from the first enrichment experiment after six months of incubation. Colonies of H89 grown on marine agar after two days were pale yellow, opaque, convex and circular, 2 – 3 mm in diameter. Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Alteromonas.

3.8.1.12. Alteromonas sp. H91

The bacterial strain designated H91 was isolated from the first enrichment experiment after six months of incubation. Colonies of H91 grown on marine agar after two days were yellow, opaque, convex and circular, 2 – 3 mm in diameter. Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Alteromonas.

3.8.2. Other bacterial strains

A number of additional bacterial strains were also employed during this work. Most of these additional strains belong to either the class Alphaproteobacteria or the class Gammaproteobacteria.

3.8.2.1. Celeribacter neptunius H14T

Celeribacter neptunius is an Alphaproteobacterium belonging to the family Rhodobacteraceae, first isolated during the course of this work. C. neptunius represents a novel genus and species not previously described. They are Gram-negative, rod- shaped cells, 0.8 – 1.8 µm in length and 0.4 – 0.9 µm wide. They are chemoorganotrophic, facultative anaerobes, and are motile by means of single polar or subpolar flagella. Requires 1 – 8 % NaCl, temperatures of 5 – 35 °C (optimum 25 °C) and pH ranging between 4.0 – 10.0 (optimum 7.5 – 8.0) for growth to occur. The DNA G + C content is 59.1 mol% ( Ivanova et al. 2010b ).

3.8.2.2. Glaciecola polaris LMG 21857T

Glaciecola polaris is a Gammaproteobacterium belonging to the family Alteromonadaceae. They are Gram-negative, rod-shaped cells, 2 – 3 µm in length and 0.4 µm wide. They are chemoheterotrophic, strict aerobes, and are motile by means of polar or sub-polar flagella. Cells produce prosthecae and buds, and can tolerate up to 10 % (w/v) NaCl. Growth occurs at temperatures between 5 – 30 °C. The DNA G + C content is 44.0 mol% (Van Trappen et al. 2004b).

3.8.2.3. Idiomarina baltica LMG 21691T

Idiomarina baltica is a Gammaproteobacterium belonging to the family Idiomarinaceae. They are Gram-negative, rod-shaped cells, 0.7 – 1.6 µm in length and 0.4 – 0.7 µm wide. They are chemoheterotrophic, strict aerobes, and are motile by means of a single polar flagellum. Salt is required for growth, with optimum concentrations between 3 – 6 % (w/v). Growth occurs at temperatures between 8 – 46

°C, with optimum temperatures at 30 – 40 °C. The DNA G + C content is 49.7 mol% (Brettar et al. 2003).

3.8.2.4. Jannaschia rubra 4SM3T

Jannaschia rubra is an Alphaproteobacterium belonging to the family Rhodobacteraceae. They are Gram-negative, rod-shaped cells, 1.0 – 2.0 µm in length and 0.5 µm wide. They are chemoorganotrophic, strict aerobes, and are motile by means of a multiple monopolar flagella. Cells require sodium and magnesium, and grow at marine salt concentrations between 0.34 – 9 % (w/v). Optimum incubation temperature for growth is 4 – 25 °C. The DNA G + C content is 64.6 % (Macián et al. 2005).

3.8.2.5. Marinobacter litoralis SW-45T

Marinobacter litoralis is a Gammaproteobacterium belonging to the family Alteromonadaceae. They are Gram-negative, rod-shaped cells, 1.5 – 3.0 µm in length and 0.5 – 0.8 µm wide. They are chemoorganotrophic, strict aerobes, and are motile by means of a single polar flagellum. Growth conditions are 0.5 – 18 % (w/v) NaCl (optimum 2 – 7 %), pH 7.0 – 8.5, and temperatures between 4 – 46 °C (optimum 30 – 37 °C). The DNA G + C content is 55 mol% (Gauthier et al. 1992, Yoon et al. 2003b).

3.8.2.6. Nautella italica LMG 24365T

Nautella italica is an Alphaproteobacterium belonging to the family Rhodobacteraceae, originally isolated from a marine electroactive biofilm. They are Gram-negative, rod-shaped cells, 1.5 – 2.1 µm in length and 0.7 – 0.11 µm wide. They are strict aerobes, and are motile by means of a single polar flagellum. The cells accumulate poly-β-hydroxybutyrate. Salt is required for growth, with optimum concentrations between 1 – 5 % (w/v). Growth occurs at pH 5.5 – 9.0, and temperatures between 4 – 45 °C, with optimum pH and temperatures of 6.5 – 8.0 and 20 – 28 °C. The DNA G + C content is 61.0 ± 0.8 mol% (Vandecandelaere et al. 2009).

3.8.2.7. Nereida ignava 2SM4T

Nereida ignava is an Alphaproteobacterium belonging to the family Rhodobacteraceae. They are Gram-negative, coccoid to elongated rod-shaped cells, 1 – 3 µm in length and 0.2 – 0.3 µm wide. They are chemoorganotrophic, strict aerobes, and are non-motile. Cells can grow in marine salt concentrations between 1.36 – 8 % (w/v), and at temperatures of 13 °C and 28 °C. The DNA G + C content is 56 mol% (Pujalte et al. 2005).

3.8.2.8. Oceanicola marinus AZO-CT

Oceanicola marinus is an Alphaproteobacterium belonging to the family Rhodobacteraceae. They are Gram-negative, rod-shaped cells, 0.9 – 1.0 µm in length and 0.5 µm wide. They are chemoheterotrophic, facultative anaerobes, and are non- motile. Growth occurs at NaCl concentrations between 2 – 8 %, pH 6 – 9 and temperatures between 4 – 42 °C, with optimum conditions of 3 – 5 % NaCl, pH 7 and 28 – 35 °C. The DNA G + C content is 70.9 % (Lin et al. 2007).

3.8.2.9. Pseudoalteromonas citrea NCMB 1889T

Pseudoalteromonas citrea is one of several species of Gammaproteobacteria that were originally classified as members of the genus Alteromonas, but later reclassified as Pseudoalteromonas (Gauthier 1977, Gauthier et al. 1995). They are Gram-negative, rod-shaped cells, 1 – 1.5 µm in length and 0.5 – 0.8 µm wide. They are chemoorganotrophic, strict aerobes, and are motile by means of a single polar flagellum. Cells produce non-carotenoid pigments that range from pale yellow to orange to brown, and do not produce endospores or capsules. Sodium ions are required for growth. Optimum growth occurs at concentrations between 0.5 – 0.6 M, pH 6 – 12 and at a temperature of 28 °C, but growth can proceed from 4 – 40 °C. The DNA G + C content ranges between 38.9 – 44.7%. Additionally, some strains are capable of producing autotoxic, polyanionic antibiotics (Gauthier 1977, Gauthier et al. 1995, Ivanova et al. 1998).

3.8.2.10. Pseudoalteromonas ruthenica KMM 300T

Pseudoalteromonas ruthenica is a Gammaproteobacterium belonging to the family Pseudoalteromonadaceae. They are Gram-negative, rod-shaped cells, 0.7 – 0.9 µm in diameter. They are chemoorganotrophic, strict aerobes, and are motile by means of a single polar flagellum. Requires 1 – 9 % (w/v) NaCl, pH 6.0 – 10.0 (optimum 7.5 – 8.0), and temperatures between 10 – 35 °C (optimum 25 – 30 °C) for growth. The DNA G + C content is 48.4 – 48.9 mol% ( Ivanova et al. 2002b).

3.8.2.11. Pseudoalteromonas translucida KMM 520T

Pseudoalteromonas translucida is a Gammaproteobacterium belonging to the family Pseudoalteromonadaceae. They are Gram-negative, rod-shaped cells, 0.7 – 0.9 µm in diameter. They are chemoorganotrophic, strict aerobes, and are motile by means of multiple flagella on both poles. Requires 1 – 6 % (w/v) NaCl, and grows at pH 6.0 – 10.0 (optimum 7.5 – 8.0), and temperatures between 4 – 30 °C (optimum 25 °C). The DNA G + C content is 46.3 % ( Ivanova et al. 2002a).

3.8.2.12. Pseudoruegeria aquimaris SW-255T

Pseudoruegeria aquimaris is an Alphaproteobacterium belonging to the family Rhodobacteraceae. They are Gram-negative, rod-shaped cells, 1 – 8 µm in length and 0.3 – 0.8 µm wide. They are strict aerobes, and are non-motile. Cells require NaCl for growth, and can tolerate up to 8 % (w/v). Growth occurs at temperatures of 15 °C and 49 °C, and optimum pH is 7 – 8. The DNA G + C content is 67.0 mol% (Yoon et al. 2007).

3.8.2.13. Roseisalinus antarcticus EL-88T

Roseisalinus antarcticus is an Alphaproteobacterium belonging to the family Rhodobacteraceae, isolated from a hypersaline lake in Antarctica. They are motile, Gram-negative, rod-shaped cells, 2.18 – 4.20 µm in length and 0.90 – 1.02 µm wide. They are heterotrophic and aerobic to microaerophilic. Cells produce bacteriochlorophyll a, and require artificial seawater for cultivation; NaCl alone was 50 insufficient. Optimum growth conditions were pH 7.0 – 7.8, temperatures of 16 – 26 °C and artificial seawater concentrations of 50 – 90 ‰. The DNA G + C content is 67 mol% (Labrenz et al. 2005).

3.8.2.14. Shewanella japonica KMM 3299T

Shewanella japonica is a Gammaproteobacterium belonging to the family Shewanellaceae. They are Gram-negative, rod-shaped cells, 1 – 2 µm in length and 0.6 – 0.8 µm wide. They are heterotrophic, facultative anaerobes, and are motile by means of a single polar flagellum. Does not require sodium ions for growth, but can tolerate up to 3 % (w/v) NaCl. Growth occurs at pH 6.0 – 9.0 (optimum 7.5), and temperatures between 10 – 37 °C (optimum 20 – 25 °C). The DNA G + C content is 43 – 44 mol% (Ivanova et al. 2001).

3.8.2.15. Shimia marina CL-TA03T

Shimia marina is an Alphaproteobacterium belonging to the family Rhodobacteraceae. Cells are Gram-negative rods, 0.8 – 3.6 µm in length and 0.3 – 0.6 µm wide. They are heterotrophic, strict aerobes, and are motile by means of multiple monopolar flagella. Sea salt is required for growth, with optimum concentrations between 3 – 7 % (w/v). Growth occurs at pH 6 – 10, and temperatures between 15 – 35 °C, with optimum temperatures at 30 – 35 °C. The DNA G + C content is 57.2 % (Choi and Cho 2006).

3.8.2.16. Thalassobacter arenae GA2-M15T

Thalassobacter arenae is an Alphaproteobacterium belonging to the family Rhodobacteraceae. They are Gram-negative, rod-shaped cells, 1.2 – 2.4 µm in length and 0.7 – 1.2 µm wide. They are chemoorganotrophic, strict aerobes, and are motile by means of polar flagella. The cells accumulate polyhydroxybutyrate. Salt is required for growth, within the range of 0.34 – 10 % (w/v). Growth occurs at pH 6 – 9, and temperatures between 5 – 35 °C, with optimum pH and temperature of 7 and 30 °C, respectively. The DNA G + C content is 56.0 % (Kim et al. 2009).

3.8.2.17. Thalassospira tepidiphila 1-1BT

Thalassospira tepidiphila is an Alphaproteobacterium belonging to the family Rhodospirillaceae, originally isolated from oil-contaminated seawater. They are Gram- negative, vibroid/spiral shaped cells, 0.9 – 2.0 µm in length and 0.17 – 0.33µm wide. The genus is composed of chemoorganotrophic aerobes (López-López et al. 2002). T. tepidiphila is motile by means of a single polar flagellum. Cells are capable of degrading polycyclic aromatic hydrocarbons. Growth occurs at NaCl concentrations between 2 – 14 %, pH 5 – 10 and temperatures between 8 – 45 °C, with optimum conditions of 3 % NaCl, pH 7 and 43 °C. The DNA G + C content is 55.1 ± 0.4 % (Kodama et al. 2008).

3.8.2.18. Thalassospira sp. H88

The bacterial strain designated H88 was isolated from the first enrichment experiment after six months of incubation. Colonies of H88 grown on marine agar after two days were pale yellow, opaque, convex and circular, 1 – 2 mm in diameter. Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Thalassospira.

3.8.2.19. Thalassospira sp. H93

The bacterial strain designated H93 was isolated from the first enrichment experiment after six months of incubation. Colonies of H93 grown on marine agar after two days were white/colourless, transparent, convex and circular, <1 mm in diameter. Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Thalassospira.

3.8.2.20. Thalassospira sp. H94

The bacterial strain designated H94 was isolated from the first enrichment experiment after six months of incubation. Colonies of H94 grown on marine agar after two days were white/colourless, transparent, convex and circular, <1 mm in diameter.

Partial sequencing of the 16S rRNA gene and subsequent BLAST analysis identified this strain as being genetically most closely related to species of Thalassospira.

3.9. Biofilm characterisation using confocal laser scanning microscopy

In order to assess bacterial attachment and growth, confocal laser scanning microscopy (CLSM) was performed. Biofilms were imaged using a FluoViewTM FV1000 spectroscopic confocal system (Olympus, Japan), which included an inverted microscope system Olympus IX81 with 20×, 40×(oil) and 100× (oil) UIS objectives, multi Ar and HeNe lasers (458 nm, 488 nm, 515 nm, 543 nm and 633 nm wavelengths) and a transmitted differential interference contract attachment and CCD camera (Cool View FDI, U.K.). For visualisation, biofilms were stained with two fluorescent dyes: an AlexaFluor 488-conjugated concanavalin A (Invitrogen, U.S.A.), and Vybrant CFDA SE Cell Tracer kit (Invitrogen, U.S.A.), according to the manufacturer’s protocols. Concanavalin A is a lectin that binds to a number of polysaccharides, including common exopolysaccharides produced by many bacteria. Thus, conjugation of concanavalin A with a fluorescent dye is an effective technique for labelling of the biofilm matrix. The excitation and emission maxima of AlexaFluor 488 are 495 nm and 519 nm respectively. The Vybrant CFDA SE Cell Tracer kit contains carboxyfluorescein diacetate succinimidyl ester, which is colourless and does not fluoresce. However, when this compound is taken up via passive diffusion into viable, active cells it is metabolised by cell esterases, resulting in the formation of highly fluorescent carboxyfluorescein succinimidyl ester which binds to amine groups within the cell, causing it to be retained. The end result is fluorescent labelling of all viable cells. The excitation and emission maxima of carboxyfluorescein succinimidyl ester are 492 nm and 517 nm respectively.

3.10. Scanning electron microscopy

Celeribacter neptunius H14T cells were immobilised for observation of morphology by scanning electron microscopy (SEM). Cells were grown overnight at 25 °C, and 1 mL of the resulting culture was subsequently incubated on aluminium discs for 24 h at 25 °C. After incubation, discs were rinsed with distilled water and air dried, before

53 coating with a thin layer of gold using a Dynavac CS300 unit. Electron micrographs were obtained using a field emission scanning electron microscope (FESEM, Supra 40 VP, Carl-Zeiss, Germany) under 2000×, 5000×, 10000×, 15000×, 70000× and 100000× magnification.

3.11. Surface characterisation

Bacterial recovery and characterisation only accounts for half of the required experimental work of this project. In order to measure and assess the degradation effect bacteria have on PET surfaces, certain surface analytical tools needed to be employed. The physico-chemical characterisation techniques used to analyse the PET surfaces are described below.

3.11.1. Atomic force microscopy

Unless otherwise indicated, AFM analysis conditions were as follows. AFM scans were performed in air at ambient conditions using an Innova scanning probe microscope (Veeco, Bruker, U.S.A.) in tapping mode. Phosphorus-doped silicon cantilevers (MPP- 31120-10, Veeco, Bruker) with a spring constant of 0.9 N/m, tip radius of curvature of 8 nm and a resonance frequency of ~20 kHz were utilized for surface imaging. Scanning was performed perpendicular to the cantilever axis at a speed of 1 Hz. The resulting topographical data were processed with first order horizontal and vertical levelling before performing roughness analysis. Topographical analysis was performed by importing the resulting AFM data files into the software Nanoscope (v1.10, Veeco) and selecting the roughness tool. The entire scan field was encompassed in the roughness calculations. The roughness parameters included the average roughness (Sa), root-mean- square (rms) roughness (Sq), maximum height difference (Smax), skewness (Ssk), kurtosis

(Sku) and developed surface area (Sdr). To minimize the effect of noise on peak counts, surfaces were smoothed using a convolution algorithm, and the minimum threshold for definition of peaks was set at 10% of the average roughness above the mean height. Processed data files were exported to Avizo® (Visualization Sciences Group, France) for generation of three-dimensional surface profiles.

3.11.2. Surface wettability and surface free energy

The wettability and free energy of PET surfaces were determined using the sessile drop technique and the Lewis acid/base method (Van Oss et al. 1986, Crawford et al. 1987, Van Oss et al. 1988, Guy et al. 1996, Öner and McCarthy 2000). Contact angles of three diagnostic liquids (water, formamide and diiodomethane) were measured using an FTA1000 instrument (First Ten Angstoms, U.S.A.) on PET surfaces and bacterial lawns. Bacterial lawns were prepared by filtering 100 mL of bacterial culture grown overnight in marine broth, through cellulose acetate filters with a pore diameter of 0.2 µm. The resulting bacterial lawns were allowed to air-dry for approximately 30 minutes before performing contact angle experiments. Once a minimum of three repeated contact angle measurements had been taken for each diagnostic liquid for a given PET sample or bacterial lawn, the surface free energy was calculated based on the results, using the instrument software.

3.11.3. X-ray photoelectron spectroscopy

The surface chemistry of the PET plastic was analysed using an Axis Ultra x-ray photoelectron spectrometer (Kratos Analytical Ltd., UK) fitted with a monochromatic x- ray source (Al Kα, hv = 1486.6 eV) operating at 150W. The Au 4f7/2 photoelectron peak with binding energy of 83.98 eV was used to calibrate the spectrometer energy scale. Survey scans were first performed from 0 to 1400 eV, and the relative atomic concentrations of the elements detected were calculated from the areas under the peaks in the spectrum. High resolution scans were then performed across the peaks corresponding to C1s, O1s, N1s, Cl2p and S2p binding energies using scanning ranges of no more than 20 eV. Peaks in the high resolution spectra were fitted to Gaussian- Lorentzian components with subtraction of a linear background using Kratos Vision II software.

3.11.4. Infrared microspectroscopy

The spatial distribution of chemical functionalities present across selected areas of the PET samples was mapped by FTIR microspectroscopy at the Australian Synchrotron on the Infrared Microspectroscopy beamline. Samples were scanned in transmission mode and in

55 attenuated total reflection (ATR) mode over several scanned areas of approximately 50 µm × 50 µm, using a Bruker Hyperion 2000 FTIR microscope, (Bruker Optic GmbH, Germany). The microscope was equipped with a 36× (0.5 numerical aperture) reflecting objective and condenser, a narrow-band mercury cadmium telluride detector and a Bruker V80v FTIR spectrometer. Infrared map data sets were constructed by collecting square arrays of infrared absorbance spectra at spatial intervals of 3 µm. Maps of chemical functional groups were generated using the instrument software (OPUS v6.5, Bruker, Germany) by integrating the area under each absorption curve of a given array of infrared spectra, specifying the start and end wavenumber from which both the integration and a straight baseline were taken.

Chapter 4: Mechanisms of PET biodegradation

______

4.1. Overview

Interactions between PET and bacteria are localised to the interface between the two. Accordingly, it is expected that any degradative effect that bacteria are able to exert upon PET polymers occurs on the surface of the plastic. This chapter will present an analysis of the physical and chemical changes of PET surfaces induced by bacteria, and attempt to elucidate possible degradation mechanisms.

4.2. Long-term PET degradation studies – Enrichment 1

PET biodegradation is known to proceed exceedingly slowly in the environment. As such, allowances were made during the initial enrichment culture experiments to facilitate long periods of incubation. The data presented in Sections 4.2.1 and 4.2.2 were obtained after incubation periods of six and ten months, respectively.

4.2.1. PET degradation after six months

4.2.1.1. Topographical analysis

After six months of incubation in enrichment culture, individual PET pieces were aseptically removed from the enrichment vessels and subjected to topographical and chemical analysis. Topographical analysis was performed using a Solver P7LS scanning probe microscope (NT-MDT, Netherlands), fitted with carbon ‘whisker’ silicon cantilevers (NSC05, NT-MDT, Netherlands) with a spring constant of 11 N m-1, in semi-contact mode. Typical topographical images of control (undegraded PET samples) and degraded surfaces are shown in Figure 4.1, along with single line profiles extracted from the positions indicated on two-dimensional maps.

While both surfaces were in general quite smooth, there is quite an obvious difference observable between the two. The degraded surface lacks the particle-like features that are present on control surface, and appears to be somewhat smoother overall. This fact was confirmed by statistical roughness analysis (Table 4.1).

Figure 4.1. Two dimensional representations of control and degraded PET surface topography, with single line profiles of the indicated positions.

Table 4.1. Comparative roughness analysis of control and degraded PET surfaces.

Sample Scanned area Sa (nm) Sq (nm) Smax (nm) Ssk Sku Control 10 μm × 10 μm 3.29 4.9 64.4 2.46 13.2 Degraded 10 μm × 10 μm 1.81 2.68 79.8 4.92 87.4

The decreased average roughness, Sa, and RMS roughness, Sq, of the degraded sample relative to that of the control sample is a good indication that the degraded sample has in fact become smoother. The maximum roughness, Smax, appears contradictory as it is higher for the degraded sample, however as Smax describes the entire range of height values of a surface it can be substantially affected by the presence of outliers, as is the case here. The increased skewness, Ssk, and kurtosis, Sku, of the degraded sample indicate that the shape and spatial distribution of the peaks has also altered. Higher Ssk values indicate a tendency toward high, narrow peaks and shallow, broader valleys, while large kurtosis values indicate that the peaks are relatively sharp.

The decreased number and size of peaks, combined with increased sharpness of the remaining peaks may indicate that the protrusions on the surface of PET are degraded preferentially by bacteria. This is likely due to the larger exposed surface area of these sections, which makes them more readily available to bacterial cells.

4.2.1.2. Chemical analysis

Chemical analysis of the control and two separate degraded PET samples was performed using XPS, as described in Section 3.11.3. Survey scans were first performed on each sample (Figs 4.2 – 4.4) to determine the elemental composition of each, summarized in Table 4.2. High-resolution scans were then performed of the binding energy ranges containing the major elements detected on the two surfaces, and Gaussian-Lorentzian components were fitted to the resulting spectra to identify the contributing chemical species (Figs 4.5 – 4.7). Components contributing to the C1s, O1s,

N1s and S2p peaks are compared in Table 4.3.

The elemental composition of each of the PET samples was dominated by carbon and oxygen, which together comprised more than 97% of the surface in each case. This is unsurprising, as PET polymer chains are composed of only these two elements (and hydrogen, which cannot be detected by XPS). However the ratio of carbon to oxygen is not as expected; based on the stoichiometry of the PET molecule the ratio of C:O should be 5:2, but is in fact is above 3:1 for each of the samples. This is likely due to adsorption of various organic contaminants on the sample surfaces, which is a common, and difficult to resolve problem when performing XPS analysis (Henderson 1998).

Figure 4.2. XPS survey spectrum of a control PET surface.

Figure 4.3. XPS survey spectrum of degraded PET surface 1.

Figure 4.4. XPS survey spectrum of degraded PET surface 2.

Table 4.2. Elemental composition of control and degraded PET samples.

Relative atomic concentration (%) Element Control PET Degraded PET 1 Degraded PET 2 Carbon 76.1 74.1 78.5 Oxygen 21.3 23.2 19.3 Nitrogen 1.1 1.3 1.3 Sulphur – 0.1 0.2 Silicon 1.3 0.7 0.7 Magnesium 0.1 0.2 – Chlorine – 0.4 –

Generally speaking, the elemental composition of each of the degraded samples remained relatively unchanged in comparison to control sample. There were some small fluctuations in the proportions of many of the detected elements, however in most cases the two degraded samples deviated from the elemental composition of the control in different ways, indicating either a) that the two degraded samples were being modified in different ways, or b) that the elemental fluctuations were within the ranges of natural elemental variation for PET surfaces immersed in seawater for six months. The

62 degraded samples did however show consistent increases in both nitrogen and sulphur content, both of which are elements commonly cycled by bacteria (Gonzalez et al. 1999, Aminot and Kérouel 2004, Simjouw et al. 2004, Vila-Costa et al. 2006). This may or may not indicate bacterial modifications of the polymer, as the measured increases were quite small.

Figure 4.5. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of a control PET surface as determined by XPS, with fitted components.

Figure 4.6. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of degraded PET sample 1 as determined by XPS, with fitted components.

Figure 4.7. High-resolution spectra of the C1s, O1s, N1s and S2p peaks of degraded PET sample 1 as determined by XPS, with fitted components.

Table 4.3. Chemical component contributions to the C1s, O1s, N1s and S2p peaks of control and degraded PET surfaces.

Component contribution (%) Peak Component Control PET Degraded PET 1 Degraded PET 2

C1s Aromatic C 71.1 62.7 71.2 C–O 16.9 23.0 17.0 O=C–O 11.9 14.5 11.9

O1s O=C 56.8 52.8 54.1 O–C 43.2 47.2 45.9

N1s N–C 100 100 100 * S2p Component 1 – 66.8 66.7 Component 2* – 33.2 33.3

*Component 1 and 2 refer to unidentified peak components

Similarly to the elemental composition of the PET samples degraded over six months, the contributing molecular species of the C1s, O1s, N1s and S2p peaks are largely inconclusive, with one possible exception. The ratio between the carbonyl oxygen atom and the oxygen heteroatom incorporated into the polymer backbone has shifted slightly, with both degraded samples tending slightly more towards the latter. The carbonyl group may act as a nucleophile, leading to formation of new carbon – oxygen bonds with small organic molecules.

It should be noted that the components of the S2p peak could not be conclusively identified; the binding energies of each component correlated poorly with known sulphur-species binding energies. Additionally, while the data analysis software calculated two individual components to be contributing to the S2p peak, the peak is quite small and the spectrum is relatively noisy, which may lead to false identification.

4.2.2. PET degradation after ten months

4.2.2.1. Topographical analysis

AFM scans of a PET surface degraded over a period of 10 months were performed under the same conditions described in Section 4.2.1.1. Two different size scanning areas are presented in 2D and 3D form, i.e. a 10 μm × 10 μm field (Fig 4.8. A, C) and a 5 μm × 5 μm field (Fig 4.8. B, D). Performing roughness calculations for multiple different scanning areas of the same surface allows for a more comprehensive analysis of the surface, as surface topography is dependent on scale.

Compared to the degraded PET sample after six months, the 10 month-degraded samples became even smoother, with all calculated roughness parameters decreasing (Table 4.4). This suggests a progression of the ‘peak erosion’ scenario put forward in Section 4.2.1.1. The taller protrusions on the PET surface have been worn down further, decreasing the overall height of the peaks, as indicated by the decreases in Sa, Sq and

Smax. Skewness and kurtosis of the 10 month sample decreased, after the initial increase observed after six months. Erosion of the sides of the taller surface features would result in an initial increase in peak sharpness, however at some point the feature would become exceedingly thin and fragile and may be easily broken off, resulting in a small ‘stub’ being left behind on the surface. This would be observable by roughness analysis

66 as decreases in typical height parameters (i.e. Sa, S q) and decreased Ssk and Sku, which is consistent with the current data.

Table 4.4. Roughness analysis of different size fields of view of PET samples degraded over ten months

Scanned area Sa (nm) Sq (nm) Smax (nm) Ssk Sku 10 μm × 10 μm 1.20 1.86 35.0 2.65 20.8 5 μm × 5 μm 0.84 1.15 18.8 1.64 14.5

4.2.2.2. Chemical analysis

Chemical analysis of the 10 month degraded PET samples was performed using XPS, as described in Section 3.12.3. Survey scans detected the presence of some ions not detected on the six month degraded samples (Fig. 4.9). In addition to this, some interesting changes in elemental proportions were observed (Table 4.5).

Figure 4.9. XPS survey spectrum of a PET surface degraded over a period of 10 months.

Table 4.5. Elemental composition of a PET sample degraded over 10 months.

Element Relative atomic concentration (%) Carbon 72.9 Oxygen 21.5 Nitrogen 2.6 Sulphur 0.3 Sodium 0.9 Potassium 1.4 Chlorine 0.4

The most significant change in elemental composition observed is that of nitrogen. The proportion of nitrogen present on the surface of the 10 month sample is more than double the proportion of nitrogen on the six month sample. The increase in nitrogen that occurred during the additional four months far exceeds that of the first six, which suggests that it is not simply a result of adsorption or absorption processes that may occur naturally in seawater. Such processes would be fastest immediately upon immersion, and would progressively slow as the system neared equilibrium. Rather, it is more likely that the process is mediated by bacteria, or more specifically bacterial enzymes which are able to work with increasing efficiency due to positive feedback mechanisms, e.g. a polymer chain may protrude slightly further from the bulk material in the immediate area surrounding an introduced functionalisation, facilitating further functionalisation and subsequently increased chain availability.

The carbon, oxygen and nitrogen components that make up the 10 month degraded PET surface are significantly more complex than that of six months samples (Fig. 4.10, Table 4.6). First of all, new components of all three elements were detected on the 10 month samples which were not previously detected. New chemical species detected were alkyl carbon, carbonyl carbon (as opposed to the carbonyl carbon of the polyester 2- backbone), an unidentified oxide (most likely SO3 ), alcohol and/or ether and ammonium. In addition to the new chemical components, the proportions of the chemical groups that make up the polymer chain changed. The ratio of aromatic carbon : C–O : O=C–O (expected from stoichiometry to be 3:1:1) was approximately 5:2:1, and similarly the ratio of the two oxygen components (expected to be 1:1) was approximately 3:2.

Figure 4.10. High-resolution spectra of the C1s, O1s, N1s, S2p and Cl2p peaks of a PET surface degraded over a period of 10 months, with fitted components.

Table 4.6. Chemical component contributions to the C1s, O1s and N1s peaks of a PET sample degraded over 10 months.

Peak Component Component contribution (%)

C1s Aromatic C 53.1 C–C, C–H 13.3 C–O 21.1 C=O 3.0 O=C–O 9.5 * O1s Oxide 18.0

O=C 45.3 C–O–R† 7.9 O–C 28.7

N1s N–C 82.6 N+ 17.4

*This component was inconclusively identified, but is thought to be an oxide of some form, most likely sulphite.

† C–O–R refers to an alcohol or ether, and is to be distinguished from O–C, which is the oxygen atom between two carbons in an ester bond, such as is part of the PET chain.

4.3. The role of light in bacterial degradation of PET

Long-term enrichment culture experiments suggested that bacteria do have the capability for modifying and degrading PET surfaces. However for any practical application to be possible, the efficiency of the process needs to be significantly improved. While the culture conditions of the original enrichment experiment provide selective pressure for organisms with the capability to utilise PET as an energy source, organisms which employ other strategies for obtaining energy may also receive a selective advantage. Two examples are photoautotrophic bacteria and chemolithotrophic bacteria. Chemolithotrophs obtain energy by reduction of various inorganic substrates (Madigan et al. 2003), however in this instance are not of particular concern, as they are expected to deplete their energy sources relatively quickly, due to the confined nature of the experiment. Of greater concern are the photoautotrophs, which utilise light for

71 energy production (Madigan et al. 2003). In order to assess the role, if any, which photoautotrophic bacteria play in this particular enrichment culture, and to provide the true PET-degrading bacteria with an enhanced selective advantage, a second enrichment experiment was performed, which included an incubation condition that was maintained in the absence of light. Five different incubation conditions (including 2 controls) were prepared in triplicate: i) PET films in sterilised seawater, ii) PET films in sterile pure water (MilliQ), iii) PET films in seawater, iv) PET bottle pieces in seawater and v) PET films in seawater in the absence of light. The incubation period was eight months, with subsequent topographical and chemical characterisation.

4.3.1. Topographical analysis

AFM scans of degraded PET bottle pieces and PET films degraded in the presence and absence of light were performed after eight months of enrichment incubation. The degraded bottle sample showed a distinctly different topography, while the difference between the two films was more subtle (Fig. 4.11). The differences were all the samples were better defined and expounded by comparative roughness analysis (Table 4.7).

Table 4.7. Comparative roughness analysis of PET samples degraded under various conditions over a period of eight months.

Incubation Sa (nm) Sq (nm) Smax (nm) Ssk Sku Sdr (%) conditions PET film + 3.69 4.75 38.1 0.568 4.02 100.03 Sterile seawater PET film + 1.12 1.54 14.7 1.79 7.21 100.02 MilliQ water PET film + 3.48 4.50 38.0 0.878 4.73 100.03 seawater PET bottle + 1.04 1.51 17.6 2.25 12.7 100.02 seawater PET film + 2.95 3.83 28.2 0.782 4.10 100.02 seawater (dark)

For the most part, PET films samples displayed higher amplitude roughness parameters than the bottle sample. The exception to this was the MilliQ water control, which remained particularly smooth. This suggests some ‘roughening’ effect exerted on the PET films by seawater that does not occur on plastic bottles. Despite this, there was a partial compensation observed particularly in the dark sample, which saw a decrease in Sa, Sq and Smax, compared to that of the control seawater film. This suggests a scenario where upon immersion, PET film surfaces initially, and probably relatively quickly, increase in roughness, perhaps due to water absorption processes, before bacteria begin to erode the surface features. Erosion occurred to a greater degree in the dark conditions than in the light-exposed, which may indicate that inhibition of photoautotrophic growth does indeed provide PET-degrading organisms with additional selective advantage.

4.3.2. Chemical analysis

Chemical analysis was again performed using XPS. Survey scans of samples from each incubation condition are shown in Figure 4.12, and relative elemental concentration details are presented in Table 4.8.

Table 4.8. Elemental composition of PET samples incubated over a period of eight months under different conditions.

Relative atomic concentration (%) Element PET film + PET film + PET film + PET bottle PET film + sterile MilliQ water seawater pieces + seawater seawater seawater (dark) Carbon 73.1 75.1 73.2 72.5 74.7 Oxygen 25.1 23.9 25.4 25.2 22.3 Nitrogen 1.5 0.7 1.1 1.4 1.4 Sulphur – – 0.1 0.3 0.1 Sodium – – – – 1.2 Silicon 0.2 0.2 0.2 0.5 0.2 Tin 0.1 – 0.1 – –

Elemental analysis did not reveal many substantial changes in surface chemistry of the degraded samples, however there are two differences to note. The first is the presence of sulphur on the surfaces of PET samples from all three of the degradation conditions, while sulphur concentration on the control surfaces remained below the detectable level. This confirms that the appearance of sulphur on degraded samples is a result of bacterial modifications. The second point to note is the significant amount of sodium detected on the surface of the PET incubated in dark conditions, likely in the form of salt crystals.

The detection of silicon and tin on the surfaces of the PET samples was unexpected, and is unlikely that their presence is the result of any biological actions. However for certainty, both of these elements were included in chemical component analysis.

A number of revelations are made by chemical component analyses that are not apparent from elemental analysis (Figs 4.13 – 4.17, Table 4.9). First, the sulphur present on each of the degraded samples is in a single form, i.e. sulphate. Second, the silicon present on all samples was in oxide form, i.e. silica/sand. Strangely however, no oxide component was detected in the O1s peak of any sample except the dark condition sample. It may be that as silicon was detected in small amounts only just exceeding the detection limit, the corresponding contribution to the O1s peak may be overwhelmed by the more abundant oxygen components. Third, the tin detected on three of the samples was approximately equal proportions of Sn2+ and Sn4+ ionic species, and likely formed nitrides; the same three PET samples displayed nitride components of the N1s peak. Finally, the PET sample incubated under dark conditions showed an increase in the proportion of aromatic carbon relative to the other two carbon components that form the polymeric chains. This means that either new aromatic groups are being formed on the surface, or more likely, that oxygenated carbon species are being removed from the surface via an unspecified mechanism.

Figure 4.13. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in sterile seawater for eight months.

Figure 4.14. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in sterile pure water for eight months.

Figure 4.15. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in seawater for eight months.

Figure 4.16. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET bottle pieces incubated in seawater for eight months.

Figure 4.17. High-resolution spectra of the C1s, O1s, N1s, S2p, Si2p and Sn3d regions of the XPS survey spectrum of PET film incubated in seawater in dark conditions for eight months.

Table 4.9. Chemical component contributions to the C1s, O1s, N1s, S2p, Si2p and Sn3d peaks of PET samples degraded over eight months under various incubation conditions.

Component contribution (%) Peak Component PET film PET film PET film PET bottle PET film + + sterile + MilliQ + seawater pieces + seawater seawater water seawater (dark)

C1s CAr 60.7 62.7 60.9 64.9 70.1 C–O 22.3 21.6 22.2 20.3 17.3 O=C–O 16.8 15.7 16.6 13.5 10.4 C=O 0.2 – 0.3 1.3 2.2

O1s O=C 49.0 46.7 48.6 54.3 49.5 O–C 51.0 53.3 51.4 45.7 40.3 O2-/OH- * – – – – 10.2

N1s N–C 83.8 73.1 83.1 100 100 N3- 16.2 26.9 16.9 – –

S2p SO4 – – 100 100 100

Si2p Si–O 100 100 100 100 100 2+ Sn3d Sn 51.8 51.3 50.3 – – Sn4+ 48.2 48.7 49.7 – –

* The third oxygen component contributing to the O1s peak consists of oxide and/or hydroxide; distinction between the two could not be made conclusively.

4.4. Degradation quantification and optimisation of conditions

The previous sections in this chapter are all devoted to qualitative characterisation of the PET biodegradation mechanisms of bacteria. In this section, the first attempt at quantification of the degradation rate is presented. In addition to simply measuring the rate of PET biodegradation, various bacterial cultures and chemical additives were tested for their ability to enhance the rate of degradation.

4.4.1. Measured loss of mass

Four different bacterial cultures were used, plus a sterile control. Two cultures were single species cultures: Thalassospira sp. H94, isolated from early enrichment experiments, and Thalassospira tepidiphila 1-1B T, a known degrader of polycyclic aromatic hydrocarbons (Kodama et al. 2008). The other two cultures were mixed species; one was a consortium of three Thalassospira spp. and three Alteromonas spp., all of which were isolated during early enrichment experiments, and the other was a sample of the entire bacterial community developed in an enrichment experiment, taken directly from the incubation flask. For each culture, four different combinations of additives were tested, as well as a control without any additives. The four additive combinations are as follows:

Supplementation of the nitrogen and phosphorus reserves naturally present in seawater (designated as N/P samples). Previous studies have indicated that bacterial growth in hydrocarbon contaminated environments can be limited by low concentration of trace nutrients required for growth (Iranzo et al. 2001, Nikolopoulou and Kalogerakis 2008). Addition of Fe (II) to culture media. Detection of alkyl carbon presented in Section 4.2.2.2 suggested a possibility of aromatic ring-opening oxygenases, many of which require the presence of Fe2+ ions (Schlafli et al. 1994, Fukuhara et al. 2008). Addition of Tween 20 surfactant. Previous studies have suggested the idea that aromatic polyester degradation is limited by polymer chain mobility (Marten et al. 2003, Müller et al. 2005). By adding surfactants to the culture media it is hoped that chain mobility into the aqueous phase may be more favourable. Combination of all three conditions listed above.

Individual pieces of polymer film were weighed both before and after three months incubation in each of the above experimental conditions to trace any loss of mass (Fig. 4.18). Because of the nature of this kind of experiment, i.e. involving biological systems, relatively large error values were associated with the measured changes in mass of the PET films. Despite this, however, a significant loss of mass was recorded for films degraded by the enrichment community supplemented with additional nitrogen

85 and phosphorus. On average, PET films degraded under these conditions lost 1.43% of their original mass.

4.4.2. Topographical analysis

To determine any qualitative differences in the mode of PET biodegradation when bacteria are supplemented with extra nitrogen and phosphorus, samples from the above experiment were subjected to topographical and chemical analysis. AFM scans were performed on PET samples that had been degraded by the bacterial enrichment community both with and without nitrogen and phosphorus supplementation and subsequently compared (Fig. 4.19).

Comparison of the two surface topographies showed that architecture of both was quite similar. There is a small observable difference in the number and size of the major

87 peaks on the two surfaces; N/P supplementation appears to result in slightly smoother surfaces. Roughness analysis also appears to suggest this; samples degraded with nitrogen and phosphorus supplements had with slightly lower amplitude values (Table 4.10). The two samples appear to have been degraded via essentially the same mechanisms, although the degradation of the supplemented sample may be slightly more advanced. This supports the hypothesis that bacterial growth is limited by the availability of trace nutrients, and suggests that increasing the reserves of these nutrients helps to boost the metabolic activities of the bacterial community.

Table 4.10. Comparative surface roughness analysis of PET films degraded by the bacterial enrichment community with and without nitrogen and phosphorus supplementation.

Incubation Sa (nm) Sq (nm) Smax (nm) Ssk Sku Sdr (%) conditions No additional N/P 2.58 3.96 47.92 2.4 8.08 100.68 N/P supplemented 2.17 3.48 53.08 2.67 10.9 100.47 culture

4.4.3. Chemical analysis

In comparison to changes in topography, changes in the surface chemistry of the PET film degraded with nitrogen and phosphorus supplementation were more obvious (Figs 4.20, 4.21, Table 4.11). Perhaps the most interesting change observed was a shift in the ratio of carbon to oxygen, decreasing from 3.04:1 to 2.42:1. This corresponds to quite a substantial increase in the amount of oxygen on the surface, indicating that when supplemented with nitrogen and phosphorus, bacteria are able to oxidise the surface of PET at a faster rate.

Figure 4.20. XPS survey spectra of PET films degraded by a bacterial enrichment community without nitrogen and phosphorus supplementation.

Figure 4.21. XPS survey spectra of PET films degraded by a bacterial enrichment community with nitrogen and phosphorus supplementation.

Table 4.11. Elemental composition of PET films degraded by a bacterial enrichment community with and without nitrogen and phosphorus supplementation.

Relative atomic concentration (%) Element No additional N/P supplemented N/P culture Carbon 73.8 68.8 Oxygen 24.3 28.4 Nitrogen 1.2 1.6 Sulphur – 0.4 Sodium – 0.3 Calcium 0.2 0.4 Magnesium 0.1 0.1 Silicon 0.3 –

This increased degree of oxidation was also observed in the chemical component analysis (Figs 4.22, 4.23, Table 4.12). PET films degraded by nitrogen and phosphorus supplemented cultures showed a shift toward a lower proportion of aromatic carbon and an increase in the amount carbon/oxygen single bonds. There was also a small increase in the incidence of carbonyl groups, supporting the observations in Section 4.4.2. that the degradation mechanisms in the two culture conditions are similar, and that the PET sample degradation in supplemented cultures was more advanced.

Figure 4.22. High resolution XPS spectra of PET films degraded by a bacterial enrichment community without nitrogen and phosphorus supplementation.

Figure 4.23. High resolution XPS spectra of PET films degraded by a bacterial enrichment community with nitrogen and phosphorus supplementation.

Table 4.12. Chemical component contributions to the C1s and O1s peaks of PET samples degraded by a bacterial enrichment culture with and without nitrogen and phosphorus supplementation.

Component contribution (%) Peak Component No additional N/P supplemented N/P culture

C1s CAr 65.4 57.8 C–O 18.9 25.2 O=C–O 14.8 14.4 C=O 0.8 2.6

O1s O=C 52.7 52.6 O–C 47.3 47.4

4.4.4. Determination of optimal degradation conditions

Based on the data presented throughout this chapter, an experiment was performed to attempt to find the incubation conditions that result in optimal PET biodegradation. In most of the conditions PET film samples were incubated in environmental seawater samples containing natural bacterial communities. Among the experimental conditions tested were degradation in dark conditions and in nitrogen/phosphorus supplemented culture, for reproducibility. Two new conditions were also tested: surfactant pre- treatment and degradation by pure culture of Pseudoalteromonas citrea. Pre-treating films with sodium dodecyl sulphate (SDS) surfactant was a variation on addition of Tween 20 from the previous section, and was intended to increase chain mobility. P. citrea is a known producer of biosurfactants, which may increase chain mobility, and also secretes a number of extracellular enzymes which help may facilitate degradation (Kalinovskaya et al. 2004). Films were degraded for one month, before AFM and infrared microspectroscopic analysis.

4.4.4.1. Topography

Three-dimensional representations of AFM scans for each of the experimental conditions are presented below in Figure 4.24. PET film surfaces of samples degraded

92 in the presence and absence of light both exhibited similar topographical patterns to those observed in previous experiments. PET degraded in nitrogen\phosphorus supplemented culture however resulted in much rougher surface topography than observed previously (Table 4.13). This is probably due to the differences in bacterial community composition, which may contain members that degrade PET by different mechanisms.

An interesting effect was observed on surfactant pre-treated films. Surfactant pre- treatment resulted in a surface architecture that was not previously observed, which was relatively rough. Bacterial degradation of pre-treated films however, caused the surfaces to become very smooth. The increased surface area made available by surfactants may have allowed bacterial enzymes better access to the polymer chains, resulting in more effective peak erosion.

In addition to this, PET incubated in pure P. citrea culture exhibited a similar structure to the surfactant control, indicating that the biosurfactants produced by the strain had a similar effect on the surface to SDS. It is unclear whether P. citrea was capable of degrading PET, but at the very least co-culture of P. citrea with a proven degrader could lead to more effective biodegradation.

Table 4.13. Comparative roughness analysis of PET films degraded over a period of one month in a variety of culture conditions.

Experimental Sa (nm) Sq (nm) Smax (nm) Ssk Sku Sdr (%) conditions Sterile seawater 2.70 4.00 147.2 3.29 50.3 101.1 Non-sterile 2.65 3.92 65.1 2.20 10.6 100.9 seawater Non-sterile (dark) 2.41 3.54 109.4 1.99 9.94 100.9 Non-sterile, N/P 8.0 10.60 179.5 1.02 6.57 104.1 supplemented Sterile, SDS 6.48 8.08 94.1 0.325 3.07 101.0 pre-treatment Non-sterile, SDS 1.70 2.51 99.8 2.96 36.7 100.7 pre-treatment P. citrea culture 7.10 8.90 115.2 0.351 3.49 102.2

4.4.4.2. Chemistry

A different approach was taken for the chemical characterisation of PET films in this optimisation experiment. Infrared microspectroscopy was performed on PET samples degraded in the presence and absence of light, at the infrared microspectroscopy beamline of the Australian Synchrotron. IR microspectroscopy allows for mapping of the spatial distribution of chemical groups, not possible using standard XPS.

Individual spectra were collected in attenuated total-reflection (ATR) mode at 3 µm intervals across approximately 50 µm × 50 µm, before the areas under characteristic peaks of the spectra were integrated and mapped according to percent variation, relative to the maximum intensity (Figs 4.25 – 4.27). Five peaks were chosen for integration, with corresponding wavenumbers of approximately 1710 cm-1, 1415 cm-1, 1345 cm-1, 1100 cm-1 and 1020 cm-1.

For the most part, each of the PET films showed a relatively even distribution of each of the functional groups mapped. However, a clear variation was observed in the intensity of the C–O–C stretching peak (1113 – 1074 cm-1) of the sample degraded in the presence of light (Fig. 4.26). This peak forms a doublet with an adjacent peak at a

97 slightly lower wavenumber, therefore to ensure that the correct peak was analysed the ‘shoulder’ of the doublet at the higher wavenumber was integrated. The intensity of this peak indicates that this part of the molecule is in gauche conformation. PET molecules in crystalline sections tend to lay mostly in the one plane, in trans conformations (Zhu and Kelley 2005). Gauche conformations occur almost exclusively in amorphous regions, therefore integration of the C–O–C stretching peak is a good approximation of the level of crystallinity of the PET samples. The resulting map shows a boundary between two regions of differing crystallinity.

It should be noted that in the IR maps of the control PET surface, there are a number of atypically high intensity points just to the right of centre (Fig. 4.25). These are probably due to inconsistent degrees of contact between the surface and the ATR crystal. Some of the maps of the light-degraded and dark-degraded samples (Figs 4.26 and 4.27) have similar apparent outliers, however as these points aren’t observed across all maps for a given sample they are more likely to be truly representative of heterogeneity in the chemical composition of the surface. Inconsistent contact is usually an indication of microscale roughness on the sample surface.

Single point spectra were also acquired for each of the samples (Fig. 4.28). Generally speaking the spectra all have peaks in similar positions, and only their relative absorbances vary, indicating that each of the surfaces has quite similar chemistry. The spectrum of the PET film exposed to pure culture of P. citrea however was substantially more complex, which may be due to adsorption of one or more of the numerous extracellular substances secreted by the cells.

100

101

Figure 4.28. Single point spectra of PET samples degraded under a variety of conditions.

102

4.5. Summary

Marine bacteria did appear to have the ability to degrade PET surfaces and utilise them as carbon and energy sources. Degradation appeared to occur preferentially at the higher, more exposed points on the PET surface. These points have more exposed surface area, which allows for increased polymer chain mobility and relative ease of access for bacterial enzymes. This mode of peak erosion resulted in surfaces which were measurably smoother.

The chemical functional groups on the surfaces of degraded PET samples did not qualitatively change much with respect to the control surfaces. However one functionality that was commonly detected only on degraded surfaces was non-ester, carbonyl groups. The major changes between degraded and control surfaces was the quantities and proportions, rather than types, of chemical groups. This is an indication that degradation of PET occurs by excision of small sections of the polymer chains.

The rate of PET biodegradation was shown to be quite slow, with the optimum rate calculated to be approximately 5 nm per exposed face per day. However strategies to increase this rate proved promising; supplementation of bacterial culture with nitrogen and phosphorus, enrichment culture in dark conditions, pre-treatment with surfactants and culture of biosurfactant producing bacteria all presenting favourable results.

103

Chapter 5: Taxonomic affiliation of bacteria associated with PET biodegradation

______

104

5.1. Overview

Bacterial recovery and subsequent phenotypic and genetic characterisation was performed in order to identify bacterial strains with highest PET-degrading potential. Samples of enrichment cultures were taken after approximately eight months incubation and spread on agar plates for bacterial recovery. The resulting colonies were counted and grouped according to colony morphology, before representative colonies were picked and spread on agar plates to obtain pure cultures. Once pure cultures were obtained taxonomic characterisation was possible.

5.2. Colony morphology and abundance

Bacteria were recovered from four different enrichment cultures: bacteria grown in seawater with PET films, bacteria grown in seawater with PET bottle pieces, bacteria grown in seawater with PET films in dark conditions and a control culture without any PET. Recovery was performed as described in Section 3.4. In total, across all four conditions, 23 different colony morphologies were recovered and identified, and each was assigned a letter (Table 5.1). The number of colony forming units of each phenotype in each culture was quantified (Fig. 5.1).

Several phenotypes appeared to show preference for one or more growth conditions. For example, phenotypes r, s, t, u, v and w were all only recovered from the enrichment culture maintained in dark conditions. Each of these groups of bacteria may have been unable to gain a selective advantage in light-exposed cultures, but in cultures where no light was available for photoautotrophic bacteria they were better able to proliferate. Phenotype j showed similar behaviour; although it was recovered from all cultures it was more abundant in the dark culture. This phenotype may have experienced the same selective pressures as the other dark-preferring strains, but was more robust and better able to compete with photoautotrophs.

105

Table 5.1. Morphological description of colonies recovered from enrichment cultures after eight months.

Designation Colony morphological description a ~3 mm i.d., yellow, opaque, smooth, convex b ~3 mm i.d., white, opaque, smooth, convex c ~2 mm i.d., white, translucent d ~1 mm i.d., white, opaque, smooth, convex e <1 mm i.d., white/colourless f 2-3 mm i.d., pale yellow/white, transparent edges g ~1 mm i.d., orange/yellow, transparent h ~2 mm i.d., pale brown, opaque i Fluffy/filamentous, fungus j ~1 mm i.d., yellow, opaque k ~3 mm i.d., pale pink/orange, opaque l ~1 mm i.d., orange, transparent m 1-2 mm i.d., white, opaque n ~2 mm i.d., white, translucent centre with opaque edges o ~1 mm i.d., red, opaque p <1 mm i.d., yellow, transparent q ~2 mm i.d., irregular shape, white opaque centre with translucent edges r ~1 mm i.d., yellow, opaque with transparent edges s ~1 mm i.d., pale yellow, transparent t 1-2 mm i.d., yellow, transparent, possible agar digester u ~2 mm i.d., yellow, opaque v 1-2 mm i.d., orange, opaque w ~1 mm i.d., white, opaque

Phenotypes f and h were both recovered from all three plastic containing cultures, but not from the control. This is an indication that strains from these two groups may be the most versatile plastic degraders recovered from this experiment. Phenotype h in particular may be effective at degrading PET, as its abundance between the three cultures was similar to that of phenotype j, i.e. it showed maximum abundance in dark conditions where competition from photoautotrophs is lowest. Other phenotypes with potential for PET degradation are phenotypes m, n, o, p and q.

106

FilmPET degraders films BottlePET degraders bottles FilmPET degraders films (dark) (Dark) ControlControl (No (no Plastic) plastic) 100000 PhenotypePhenotypic abundance, abundance full strength medium

10000

1000 1 - cfu mL cfu 100

1 a b c d e f g h i j k l m n o p q r s t u v w Total Phenotype Figure 5.1. Abundance of individual bacterial phenotypes recovered from four separate enrichment cultures after eight months.

Even after eight months with extremely limited nutrient supplies, there were still a number of bacterial strains that could be isolated from the control culture. These strains are likely the most robust and resilient members of the natural bacterial community normally present in seawater, and are simply able to endure the longest in nutrient deficient conditions. Interestingly, most phenotype groups that were isolated from the control culture were also isolated from cultures containing PET, but in greater numbers. Their metabolic activities appear to be stimulated somewhat in these culture conditions. Bacterial strains from these groups are probably ‘metabolic scavengers’ that feed off the various organic compounds produced by other members of the enrichment cultures. The primary producers in these cultures are the strains capable of obtaining carbon from the PET samples.

A summary of the phenotypic proportions of the bacterial communities in each of the enrichment cultures is presented in Figure 5.2. Each of the cultures containing PET is dominated by phenotype e, with relatively substantial contributions from phenotype d. Phenotype e consists of colonies that are very small and are therefore difficult to distinguish from one another, which may mean that this group is actually comprised of a

107 relatively large number of taxonomically different bacterial species. This may account in part for the dominance of this group. This is not to say however, that the various members of this phenotype are not PET degraders; as PET degradation occurs at a slow rate, the strains capable of degrading PET likely grow slowly, and favour relatively oligotrophic conditions.

Figure 5.2. Proportions of bacterial phenotypes recovered from each of the four enrichment cultures.

The major component of the bacterial community in the control was phenotype a. This phenotype also contributed a sizable portion to the light-exposed culture with PET films. It is quite likely that members of this group possess photoautotrophic ability, as 108 they grew only in the presence of light, and the distinct lack of alternative energy sources in the control culture could provide an explanation for their proportional dominance. A comparison of the light-exposed and dark condition bacterial cultures is shown in Figure 5.3.

Figure 5.3. Comparison of the proportions of bacterial phenotypes recovered from PET enrichment cultures in the presence and absence of light.

5.3. Bacterial taxonomy and phylogenetics

Grouping bacterial isolates by colony morphology is a very broad and general classification system. For more accurate taxonomic assignment representative isolates form the major phenotypes and some of isolates of phenotypes in lower abundance were subjected to genetic analysis via partial 16S rRNA gene sequencing. The resulting sequences were compared against a database of 16S rRNA gene sequences of bacterial type strains to find their nearest taxonomic neighbour with the highest degree of 16S rRNA gene similarity (Tables 5.2 – 5.7).

5.3.1. Sequence similarity of 16S rRNA genes

Strains grouped under phenotype a consistently found to identify closest to species of Erythrobacter, from the class α-proteobacteria (Table 5.2). Several species of

109

Erythrobacter, and the closely related genera of Erythromicrobium and Porphyrobacter possess bacteriochlorophyll a, and are thus capable of photosynthetic growth (Yurkov et al. 1994, Denner et al. 2002, Hiraishi et al. 2002). This is consistent with the results of Section 5.2, in which isolates classified under phenotype a were found to favour conditions where photosynthetic ability would provide the most selective advantage.

Table 5.2. Taxonomic affiliations of isolates assigned to phenotype a.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class B1a1 a 631 Erythrobacter 96.8 α-proteobacteria pelagi B2a1 a 760 Erythrobacter 96.2 α-proteobacteria nanhaisediminis B2a2 a 137 Erythrobacter 94.1 α-proteobacteria seohaensis B3a1 a 723 Erythrobacter 97.5 α-proteobacteria citreus B3a2 a 524 Erythrobacter 60.3 α-proteobacteria seohaensis

Strains grouped under phenotype b primarily identified to members of the class α- proteobacteria, with two strains belonging to γ-proteobacteria (Table 5.3). Five of these strains returned nearest matches to Roseovarius nubinhibens, and an additional strain matched to Sulfitobacter litoralis. Both of these species are α-proteobacteria that participate in sulphur cycling (González et al. 2003, Park et al. 2007). Sulphur was identified as potentially important for bacterial biodegradation of PET in Sections 4.2.1.2 and 4.3.2.

All strains assigned to phenotype c returned nearest matches to members of the class γ-proteobacteria (Table 5.4). Most of these were Alteromonas macleodii, however one strain was matched to each of the genera Marinobacter and Glaciecola. Species of Alteromonas produce and secrete considerable extracellular polysaccharide (Raguenes et al. 1996, Hayase et al. 2003, Raguénès et al. 2003, Martínez-Checa et al. 2005). These polysaccharides can be quite viscous, and provide good protection to cells as a

110 biofilm matrix material. Some Alteromonas strains also produce some extracellular enzymes ( Ivanova et al. 2000, Ivanova et al. 2005a).

Table 5.3. Taxonomic affiliations of isolates assigned to phenotype b.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class B1b1 b 710 Thalassospira 97.6 α-proteobacteria lucentensis B1b2 b 516 Roseovarius 96.1 α-proteobacteria nubinhibens B2b1 b 438 Roseovarius 98.4 α-proteobacteria nubinhibens B2b2 b 833 Roseovarius 98.9 α-proteobacteria nubinhibens B3b1 b 834 Roseovarius 98.1 α-proteobacteria nubinhibens B3b2 b 837 Roseovarius 98.0 α-proteobacteria nubinhibens D2b1 b 148 Alteromonas 93.7 γ-proteobacteria addita D3b1 b 630 Sulfitobacter 97.0 α-proteobacteria litoralis E1b1 b 359 Cobetia 97.4 γ-proteobacteria crustatorum

111

Table 5.4. Taxonomic affiliations of isolates assigned to phenotype c.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class A3c1 c 337 Marinobacter 97.3 γ-proteobacteria sediminum B1c2 c 637 Alteromonas 97.5 γ-proteobacteria macleodii B2c1 c 787 Alteromonas 98.0 γ-proteobacteria macleodii B2c2 c 885 Alteromonas 98.8 γ-proteobacteria macleodii B3c1 c 791 Alteromonas 97.5 γ-proteobacteria macleodii B3c2 c 900 Alteromonas 98.0 γ-proteobacteria macleodii D3c1 c 480 Glaciecola 89.6 γ-proteobacteria lipolytica

Strains belonging to phenotype d mostly identified as γ-proteobacteria, with two strains of α-proteobacteria and a single strain of Actinobacteria (Table 5.5). The γ- proteobacteria strains came from two closely related genera in Alteromonas and Aestuariibacter. The two α-proteobacteria belong to the genus Thalassospira. To date only five species of Thalassospira have been validly described, and most of these are documented to degrade highly-hydrophobic aromatic hydrocarbons (López-López et al. 2002 , Liu et al. 2007, Kodama et al. 2008, Zhao et al. 2010). This is of some significance, as PET polymer chains contain aromatic constituents.

112

Table 5.5. Taxonomic affiliations of isolates assigned to phenotype d.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class A1d1 d 195 Alteromonas 86.3 γ-proteobacteria addita A1d2 d 137 Alteromonas 84.3 γ-proteobacteria genovensis A1d3 d 202 Alteromonas 89.4 γ-proteobacteria addita A1d4 d 203 Alteromonas 87.6 γ-proteobacteria addita A1d6 d 437 Aestuariibacter 92.0 γ-proteobacteria salexigens D2d1 d 202 Thalassospira 95.5 α-proteobacteria profundimaris D2d3 d 795 Thalassospira 97.0 α-proteobacteria tepidiphila D3d1 d 401 Micrococcus 98.7 Actinobacteria yunnanensis D3d3 d 437 Aestuariibacter 91.8 γ-proteobacteria halophilus D3d4 d 198 Alteromonas 88.3 γ-proteobacteria addita

Bacterial strains classified under phenotype e were primarily from the classes α- and β-proteobacteria (Table 5.6). Among the β-proteobacteria, nearest matches were commonly species of Limnobacter. Two validly described species of Limnobacter currently exist: L. thiooxidans and L. litoralis. Both of these species are chemolithoheterotrophs that oxidise thiosulfate for energy (Spring et al. 2001, Lu et al. 2011). Another two strains from phenotype e were matched to the genus Kordiimonas from the class α-proteobacteria. Like Limnobacter, there are only two validly described species of Kordiimonas (Kwon et al. 2005, Xu et al. 2011b). The type species of Kordiimonas, K. gwangyangensis, is capable of degrading high molecular weight, polycyclic aromatic compounds (Kwon et al. 2005).

113

Table 5.6. Taxonomic affiliations of strains assigned to phenotype e.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class A1e1 e 165 Limnobacter 91.1 β-proteobacteria litoralis A1e2 e 146 Burkholderia 92.5 β-proteobacteria andropogonis A1e3 e 142 Limnobacter 95.7 β-proteobacteria thiooxidans A1e7 e 357 Aestuariibacter 92.2 γ-proteobacteria halophilus A1e8 e 179 Limnobacter 87.8 β-proteobacteria litoralis A1e9 e 296 Limnobacter 98.3 β-proteobacteria thiooxidans A3e7 e 679 Erythrobacter 96.2 α-proteobacteria nanhaisediminis D2e1 e 643 Jannaschia 94.8 α-proteobacteria cystaugens D2e2 e 639 Oceanicola 93.2 α-proteobacteria nitratireducens D3e1 e 555 Methylonatrum 87.5 γ-proteobacteria kenyense D3e2 e 603 Kordiimonas 95.3 α-proteobacteria lacus D3e3 e 277 Kordiimonas 93.1 α-proteobacteria lacus

Table 5.7 summarises the nearest match found for each of the strains with less abundant phenotypes. Many of these strains returned similar matches to other strains from the major phenotypes. Three strains belonging to phenotype f are of note, as they each returned particularly low percentage similarity to Alteromonas macleodii (strains A3f2, A3f3, A3f4). The widely accepted 16S gene similarity threshold for description of new species is 97% (Stackebrandt et al. 2002, Tindall et al. 2010), which suggests that there is a high probability that these three strains belong to distinct, undescribed taxonomic species.

114

Table 5.7. Taxonomic affiliation of strains assigned to low-abundance phenotypes.

Strain Colony Sequence Nearest match Similarity Phylogenetic no. Phenotype length (bp) (%) class A3f2 f 592 Alteromonas 72.0 γ-proteobacteria macleodii A3f3 f 608 Alteromonas 76.2 γ-proteobacteria macleodii A3f4 f 722 Alteromonas 79.0 γ-proteobacteria macleodii D3f1 f 320 Micrococcus 97.0 Actinobacteria yunnanensis A1g1 g 558 Maribacter 97.2 Flavobacteriia forsetii D2g1 g 321 Micrococcus 99.3 Actinobacteria yunnanensis D2g2 g 638 Erythrobacter 97.5 α-proteobacteria citreus A1h1 h 198 Thalassospira 94.4 α-proteobacteria profundimaris D3h1 h 147 Lentibacter 96.8 α-proteobacteria algarum D3j1 j 557 Micrococcus 99.4 Actinobacteria yunnanensis D3j2 j 299 Micrococcus 97.6 Actinobacteria endophyticus D3j3 j 440 Micrococcus 99.5 Actinobacteria yunnanensis D2m1 m 481 Marinobacter 97.3 γ-proteobacteria sediminum D2o1 o 678 Arthrobacter 96.6 Actinobacteria agilis D3p1 p 263 Erythrobacter 98.5 α-proteobacteria citreus

5.3.2. Phylogenetic analysis

Based on the 16S rRNA gene similarity results shown in Section 5.3.1, phylogenetic trees for each class of bacteria isolated were constructed using

115 approximately 10 of the closest matches for each strain. Subtrees were then extracted showing the phylogenetic positions of each of the identified strains (Figs 5.4 – 5.12).

Figure 5.4 shows the phylogenetic position of strain D3b1 amongst its nearest neighbours in the class α-proteobacteria. This strain forms quite a stable cluster with other species of Sulfitobacter, with a bootstrap percentage of 61 and agreement between the three difference algorithms.

Paracoccus halophilus HN-182T {DQ423482} + X + X Paracoccus sphaerophysae Zy-3T {GU129567}

+ X Paracoccus sulfuroxidans LW36T {DQ512861}

+ X Paracoccus isoporae SW-3T {FJ593906}

+ X Paracoccus denitrificans DSM 413T {CP000489} + Paracoccus aestuarii B7T {EF660757} + Woodsholea maritima CM243T {AJ578476} + Paracoccus marinus KKL-A5T {AB185957}

Roseovarius nanhaiticus NH52JT {FJ403243}

Marivita byusanensis SMK-114T {FJ467624} X Sulfitobacter mediterraneus DSM 12244T {Y17387}

Sulfitobacter brevis Ekho Lake-162T {Y16425}

D3b1 + X Sulfitobacter litoralis Iso 3T {DQ097527} + Sulfitobacter marinus SW-265T {DQ683726}

Sulfitobacter pontiacus DSM 10014T {Y13155}

Tropicibacter naphthalenivorans C02T {AB302370} + X Primorskyibacter sedentarius KMM 9018T {AB550558}

Vadicella arenosi KMM 9024T {AB564595}

The phylogenetic positions of eight isolated strains of α-proteobacteria are presented in Figure 5.5. Five of these strains were classified to the one phenotype, and are shown to be phylogenetically closely related. These five strains form a relatively

116 stable cluster with members of Nautella, Roseovarius, Lentibacter and Ruegeria. Three more strains form an unstable cluster with Leisingera nanhaiensis, Jannaschia cystaugens and Thalassobius aestuarii.

Ruegeria scottomollicae LMG 24367T {AM905330} Thalassobius maritimus GSW-M6T {HM748766} Phaeobacter inhibens T5T {AY177712} + X Phaeobacter gallaeciensis BS107T {ABIF01000020} Phaeobacter daeponensis TF-218T {DQ981486} Phaeobacter caeruleus CCUG 55859T {AM943630} Leisingera methylohalidivorans MB2T {AY005463} Leisingera aquimarina CCUGT {AM900415} + X D2e2 D3h1 Leisingera nanhaiensis NH52FT {FJ232451} Jannaschia cystaugens CECT 5294T {AJ631302} + D2e1 Thalassobius aestuarii JC2049T {AY442178} + X Marivita cryptomonadis CL-SK44T {EU512919} + X Marivita hallyeonensis DPG-28T {JF260872} Thalassococcus halodurans UST050418-052T {DQ397336} Nautella italica CCUG 55857T {AM904562} Roseovarius halocynthiae MA1-10T {HQ852039} + X B1b2 B2b1 B2b2 B3b1 B3b2 Lentibacter algarum ZXM100T {FJ436732} Roseovarius nubinhibens ISMT {AALY01000002} Ruegeria faecimaris HD-28T {GU057915}

The phylogenetic positions of a further 14 strains belonging to α-proteobacteria are shown in Figure 5.6. Four of these strains form a very stable cluster with species of Thalassospira. A further two form a stable cluster with the genus Kordiimonas. The remaining eight strains form a somewhat stable cluster (bootstrap = 98%) that contains members of Citromicrobium, Erythrobacter, Altererythrobacter, Novosphingobium, Sphingomonas and Porphyromonas. These genera belong to the order Sphingomonadales, which is a complex group of closely related strains.

117

Citromicrobium bathyomarinum JF-1T {16267} Erythrobacter seohaensis SW-135T {AY562219} Erythrobacter pelagi UST081027-248T {HQ203045} Erythrobacter citreus RE35F/1T {AF118020} D3p1 D2g2 B3a2 B3a1 B2a2 B2a1 B1a1 Altererythrobacter marensis MSW-14T {FM177586} Altererythrobacter luteolus SW-109T {AY739662} Altererythrobacter epoxidivorans JCS350T {DQ304436} A3e7 + X Altererythrobacter indicus MSSRF26T {DQ399262} Novosphingobium tardaugens ARI-1T {AB070237} X Sphingomonas astaxanthinifaciens TDMA-17T {AB277583} + X Novosphingobium aromaticivorans DSM 12444T {CP000248} + X Novosphingobium subterraneum IFO 16086T {AB025014} Erythromicrobium ramosum DSM 8510T {AF465837} + X Porphyrobacter sanguineus IAM 12620T {AB021493} + Altererythrobacter aestuarii KYW147T {FJ997597} Erythrobacter aquimaris SW-110T {AY461441} Erythrobacter flavus SW-46T {AF500004} Erythrobacter litoralis DSM 8509T {AB013354} + X Erythrobacter longus DSM 6997T {AF465835} Erythrobacter nanhaisediminis T30T {FJ654473} Erythrobacter vulgaris 022 2-10T {AY706935} Sneathiella glossodoripedis IAM 15419T {AB289439} Nisaea denitrificans DR41 21T {DQ665838} Novispirillum itersonii subsp. itersonii LMG 4337T {EF612765} Novispirillum itersonii subsp. nipponicum LMG 7370T {EF612766} Marispirillum indicum B142T {EU642410} Maritalea mobilis E6T {EU255260} X Pseudovibrio ascidiaceicola F423T {AB175663} + X Pseudovibrio japonicus WSF2T {AB246748} + Pseudovibrio denitrificans DN34T {AY486423} + Labrenzia marina mano18T {AY628423} Polymorphum gilvum SL003B-26A1T {CP002568} Stappia indica B106T {EU726271} Maritalea porphyrae LCM-3T {AB583774} Hellea balneolensis 26III/A02/215T {AY576758} + X Roseospira goensis JA135T {AM283537} Roseospira navarrensis SE3104T {AJ298880} Pannonibacter phragmitetus C6/19T {AJ400704} X Kiloniella laminariae LD81T {AM913907} Terasakiella pusilla IFO 13613T {AB006768} A1h1 B1b1 D2d1 D2d3 Thalassospira lucentensis DSM 14000T {AM294944} + X Thalassospira profundimaris WP0211T {AY186195} Thalassospira tepidiphila 1-1BT {AB265822} Thalassospira xiamenensis M-5T {AY189753} Thalassospira xianhensis P-4T {EU017546} Kordiimonas gwangyangensis GW14-5T {AY682384} + X Kordiimonas lacus S3-22T {FJ847942} + X D3e2 X D3e3

Figure 5.6. Subtree showing the phylogenetic position of fourteen isolated strains, extracted from a larger tree containing 125 members of the class α-proteobacteria.

118

Numbers at nodes indicate the bootstrap percentage, and the symbols + and X denote agreement by maximum likelihood and maximum parsimony algorithms, respectively.

The five isolated strains of β-proteobacteria cluster together in between species of Limnobacter and Chromobacterium (Fig. 5.7). All but one of these strains returned a 16S rRNA gene similarity below 97%, which indicates a high probability of these strains belonging to distinct species. As a group they may form a distinct taxonomic lineage at the genus level, representing a phylogenetic link between Limnobacter and Chromobacterium.

Chromobacterium subtsugae PRAA4-1T {AY344056} + X Chromobacterium violaceum ATCC 12472T {AE016825} Chromobacterium pseudoviolaceum CCM 2076T {AJ871128} Chromobacterium piscinae CCM 3329T {AJ871127} + X Chromobacterium haemolyticum MDA0585T {DQ785104} Chromobacterium aquaticum CC-SEYA-1T {EU109734} A1e1 A1e2 A1e3 A1e8 + X A1e9 Limnobacter litoralis KP1-19T {AB366174} 67 + X Limnobacter thiooxidans CS-K2T {AJ289885} Burkholderia andropogonis ATCC 19311T {AB021422} Xenophilus azovorans KF46FT {AF285414} + X + Lautropia mirabilis ATCC 51599T {GL636064} X Rivibacter subsaxonicus BF49T {AM774413} + Aquabacterium parvum B6T {AF035052} + Ideonella dechloratans CCUG 30898T {X72724} + X Ideonella azotifigens 1a22T {EU542576} + X Piscinibacter aquaticus IMCC1728T {DQ664244}

The phylogenetic positions of fourteen strains of γ-proteobacteria are presented in Figure 5.8. The isolates shown in this tree grouped quite strongly according to phenotype. Five isolates from phenotype c grouped well with members of Alteromonas,

119 while three isolates each from phenotypes d and f formed small clusters near, but separate from the Alteromonas group. A further three strains formed a stable cluster with relatively long evolutionary length branches. Many of these isolates are good candidates for description of novel species, especially the group of phenotype f strains (16S rRNA gene similarity: 72 – 79 %).

B1c2 B3c2 B3c1 B2c2 + X B2c1 Alteromonas macleodii DSM 6062T {Y18228} + Alteromonas addita R10SW13T {AY682202} + X Alteromonas stellipolaris LMG 21861T {AJ295715} + Alteromonas genovensis LMG 24078T {AM885866} + X Alteromonas hispanica F-32T {AY926460} Alteromonas litorea TF-22T {AY428573} + Alteromonas marina SW-47T {AF529060} + + + X Alteromonas simiduii BCRC 17572T {DQ836766} Alteromonas tagae BCRC 17571T {DQ836765} + X A1d2 + X D2b1 A1d1 + X + A1d3 + X A1d4 D3d4 A3f2 A3f3 + X A3f4 Psychrosphaera saromensis SA4-48T {AB545807} Psychromonas ingrahamii 37T {CP000510} Psychromonas heitensis AK15-027T {AB365352} + Pseudoalteromonas mariniglutinosa KMM 3635T {AJ507251} Psychromonas arctica Pull 5.3T {AF374385} Pseudoalteromonas byunsanensis FR1199T {DQ011289} Pseudoalteromonas tetraodonis IAM 14160T {AF214730} + Pseudoalteromonas agarivorans KMM 255T {AJ417594} + X Pseudoalteromonas arctica A 37-1-2T {DQ787199} + X Pseudoalteromonas carrageenovora ATCC 12662T {X82136} Pseudoalteromonas nigrifaciens NCIMB 8614T {X82146}

120

Each of the isolates whose phylogenetic positions are shown in Figure 5.9 formed stable clusters with their nearest neighbours. Two strains formed a stable cluster with members of the genus Marinobacter. One strain formed a stable cluster with two species of Cobetia , and another strain made a stable cluster with Congregibacter litoralis.

+ X Congregibacter litoralis KT71T {AAOA01000004} + D3e1 + X Haliea rubra CM41 15aT {EU161717} Chromocurvus halotolerans EG19T {AM691086} + X Spongiibacter borealis CL-AS9T {HQ199599} + X Zhongshania antarctica ZS5-23T {FJ889619} + Spongiibacter marinus HAL40bT {AM117932} + Dasania marina KOPRI 20902T {AY771747} + X + X Oceanicoccus sagamiensis PZ-5T {AB545809} Saccharophagus degradans 2-40T {CP000282} Simiduia areninigrae M2-5T {GQ872422} Cobetia marina DSM 4741T {AJ306890} + X E1b1 Cobetia crustatorum JO1T {EU909460} X Halomonas denitrificans M29T {AM229317} + 69 Halomonas pacifica DSM 4742T {L42616} Halomonas mongoliensis Z-7009T {AY962236} + X + Halomonas aidingensis Ad-1T {GQ281062} X Halomonas maura S-31T {AJ271864} Halomonas taeanensis BH539T {AY671975} Modicisalibacter tunisiensis LIT2T {DQ641495} X Halomonas saccharevitans AJ275T {EF144149} + Halomonas shengliensis SL014B-85T {EF121853} Methylonatrum kenyense ATM 1T {DQ789390} Neptunomonas antarctica S322T {FJ713802} Marinobacter mobilis CN46T {EU293412} + X Marinobacter hydrocarbonoclasticus MBIC1303T {AB019148} Marinobacter lipolyticus SM19T {AY147906} + X Marinobacter lutaoensis JCM 11179T {AF288157} Marinobacter pelagius HS225T {DQ458821} X Marinobacter gudaonensis SL014B61AT {DQ414419} Marinobacter flavimaris SW-145T {AY517632} + X Marinobacter sediminum R65T {AJ609270} D2m1 A3c1 + Marinobacter adhaerens HP15T {AY241552} Marinobacter algicola DG893T {ABCP01000031} + Marinobacter lacisalsi FP2.5T {EU047505} + Marinobacter salsuginis SD-14BT {EF028328}

121

The four isolates whose phylogenetic positions are indicated in Figure 5.10 formed a stable cluster with the genus Aestuariibacter. These four strains are highly likely to be one or more novel species of Aestuariibacter, due to the stability of the cluster coupled with low 16S rRNA gene similarity; all four strains showed less than 93% similarity to their nearest matches.

Glaciecola nitratireducens FR1064T {CP003060} Glaciecola polaris LMG 21857T {AJ293820} + X Glaciecola psychrophila 170T {DQ007436} Glaciecola arctica BSs20135T {EU365479} Aestuariibacter litoralis KMM 3894T {AB473549} + X Glaciecola agarilytica NO2T {DQ784575} Glaciecola chathamensis S18K6T {AB247623} X Glaciecola mesophila KMM 241T {AJ488501} Glaciecola lipolytica E3T {EU183316} + X Aestuariibacter halophilus JC2043T {AY207503} Aestuariibacter salexigens JC2042T {AY207502} + X D3d3 + A1d6 + A1e7 D3c1

Most of the clusters formed by the Actinobacteria isolates lacked stability (Fig 5.11). One isolate grouped well with most of the species of Arthrobacter, although there was poor agreement between phylogenetic algorithms. Four strains did however cluster stably with Micrococcus luteus and Micrococcus yunnanensis.

122

Arthrobacter agilis DSM 20550T {X80748} Arthrobacter flavus JCM 11496T {AB299278} D2o1 Arthrobacter pigmenti LMG 22284T {AJ639827} Arthrobacter parietis LMG 22281T {AJ639830} Arthrobacter subterraneus CH7T {DQ097525} + X Arthrobacter tumbae LMG 19501T {AJ315069} Arthrobacter ilicis DSM 20138T {X83407} + X Arthrobacter nitroguajacolicus G2-1T {AJ512504} Arthrobacter alkaliphilus LC6T {AB248527} Arthrobacter phenanthrenivorans Sphe3T {CP002379} + Arthrobacter defluvii 4C1-aT {AM409361} + X Arthrobacter equi IMMIB L-1606T {FN673551} + Arthrobacter niigatensis LC4T {AB248526} Arthrobacter woluwensis DSM 10495T {X93353} Arthrobacter roseus CMS 90rT {AJ278870} + X Micrococcus niistensis NII-0909T {FJ897464} + Micrococcus terreus V3M1T {FJ423763} + D3f1 Arthrobacter oryzae KV-651T {AB279889} + X Arthrobacter pascens DSM 20545T {X80740} + X Arthrobacter nicotianae DSM 20123T {X80739} Arthrobacter soli SYB2T {EF660748} + Citricoccus parietis 02-Je-010T {FM992367} + X Citricoccus zhacaiensis FS24T {EU305672} + X Citricoccus muralis 4-0T {AJ344143} + X Citricoccus nitrophenolicus PNP1T {GU797177} + X D3d1 + D3j1 D2g1 X D3j3 Micrococcus luteus NCTC 2665T {CP001628} + Micrococcus yunnanensis YIM 65004T {FJ214355} Micrococcus lylae DSM 20315T {X80750} Micrococcus endophyticus YIM 56238T {EU005372} Micrococcus flavus LW4T {DQ491453} + X D3j2 + X Micrococcus antarcticus T2T {AJ005932}

A single strain amongst those isolated and identified belonged to the class Flavobacteriia, in the genus Maribacter. Phylogenetic analysis revealed that it clustered very well with other species of Maribacter (Fig. 5.11).

123

+ X Maribacter dokdonensis DSW-8T {AY960749} + X Maribacter ulvicola KMM 3951T {AY271626} A1g1 Maribacter sedimenticola KMM 3903T {AY271623} + + Maribacter forsetii KT02ds18-6T {AM712900} + X Maribacter orientalis KMM 3947T {AY271624} Maribacter stanieri KMM 6046T {EU246691} Maribacter aquivivus KMM 3949T {AY271625} + X Pibocella ponti KMM 6031T {AY576654} + Maribacter arcticus KOPRI 20941T {AY771762} Maribacter antarcticus CL-AP4T {EU512921} Maribacter polysiphoniae LMG 23671T {AM497875}

5.4. Celeribacter neptunius H14T

During the course of initial bacterial enrichment experiments, one bacterial isolate with the designated strain number H14 was found to belong to the Roseobacter lineage of α-proteobacteria. This strain was equally distantly related from three separate taxonomic genera, i.e. Nautella, Roseovarius and Pseudoruegeria. After phylogenetic, biochemical and physiological characterisation, strain H14 was classified as a novel species of a previously undefined genus, and accordingly a taxonomic paper was published describing the strain as Celeribacter neptunius H14T (Ivanova et al. 2010b).

Cell size and morphology were determined using scanning electron microscopy

(Fig. 5.13). Cells were grown in marine broth liquid culture to OD600 ≈ 0.3. One millilitre of this culture was then incubated for 24 hours on aluminium disks at a temperature of 25 °C. After incubation, disks were rinsed gently with sterile MilliQ water, air dried, and coated with a layer of gold using a Dynavac CS300 unit. Electron micrographs were obtained using a field emission scanning electron microscope (FESEM, Supra 40 VP, Carl-Zeiss, Germany) at 2000×, 5000×, 10000×, 15000×, 70000× and 100000× magnifications.

124

Figure 5.13. Scanning electron micrographs of Celeribacter neptunius H14T at magnifications of 10000× and 100000× (inset).

Taxonomic analysis was performed by 16S rRNA gene sequencing and construction of phylogenetic trees. The sequence was compared with three gene sequence databases (University of Nice BLAST server, http://bioinfo.unice.fr/blast, DDBJ gene sequence database, http://blast.ddbj.nig.ac.jp/top-e.html, and EzTaxon database, http://147.47.212.35:8080/) in order to retrieve the most similar sequences of validly described bacteria. The sequences were aligned using Clustal X 2.0 (Larkin et al. 2007) and checked using SeaView (Galtier et al. 1996). Trees were prepared using neighbour-joining, maximum likelihood and maximum parsimony algorithms. The nearest phylogenetic neighbour to C. neptunius is Nautella italica, which forms a cluster with bootstrap of only 77%. This level of relatedness is an indication, depending on biochemical profiles, that C. neptunius belongs to a distinct genus.

125

Figure 5.14. Phylogenetic position of C. neptunius, based on 16S rRNA gene sequence analysis. Bootstrap percentages above 50% are shown at nodes. Reproduced with permission from Ivanova et al. 2010b.

Comprehensive biochemical analysis was performed to complement genetic analysis. Biochemical tests performed included: motility, oxidation/fermentation of glucose, Gram stain, reduction of nitrate/nitrite, oxidase activity, catalase activity, gelatinase activity, arginine dihydrolase activity, lysine decarboxylase activity, ornithine decarboxylase activity, poly-β-hydroxybutyrate production, acetoin production, sodium requirements, indole production, H2S production, amylase activity, lipase activity and temperature range. C. neptunius was also tested for production of bacteriochlorophyll a and susceptibility to a number antibiotics, and was also subjected to phospholipid and fatty acid analysis. The results of biochemical characterisation are summarised in Table 5.8 and Table 5.9.

126

Table 5.8. Susceptibility of Celeribacter neptunius H14T to various antibiotics.

Antibiotic Concentration Susceptibility of C. neptunius H14T Chloramphenicol 30 µg per disc Susceptible Benzylpenicillin 10 U Susceptible Carbenicillin 100 µg per disc Susceptible Nalidixic acid 30 µg per disc Susceptible Lincomycin 15 µg per disc Resistant Rifampicin 5 µg per disc Susceptible Ofloxacin 5 µg per disc Susceptible Ampicillin 10 µg per disc Susceptible Cephalexin 30 µg per disc Susceptible Oleandomycin 15 µg per disc Resistant Cephazolin 30 µg per disc Susceptible Tetracycline 30 µg per disc Susceptible Kanamycin 30 µg per disc Susceptible Oxacillin 10 µg per disc Susceptible Streptomycin 30 µg per disc Susceptible Gentamicin 10 µg per disc Susceptible Erythromycin 15 µg per disc Susceptible Polymyxin 300 U Resistant Neomycin 30 µg per disc Susceptible

127

Table 5.9. Comparison of biochemical characteristics of Celeribacter neptunius to those of closely related strains.

Characteristic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

DNA G + C (mol%) 59 61 59-65 52-54 62-69 56 64-72 57 67 67 58-66 54-64 57 59 57-61 Bacteriochlorophyll a – ND – ND ND ND – – ND – ND – – + ND Motility + + v v v – – – + – v v + + v Pigment – Beige v Brown v, pink – Yellow – Yellow/ Red Pink/ v Beige Brown – grey beige Oxidase – + + + + + + – + + + + + + + Temp. range (°C) 5-35 4-45 4-32 27-31 4-37 13-28 4-40 5-37 15-49 3-33 4-40 5-37 15-35 13-37 13-37 NaCl range (%, w/v) 1-8 1-5 0.34-9 0-4 0-14 1.4-8 0-10 0-8 0-9 0.1-13 0-15 0.6-12 3-7 0.85-7 1-8 Nitrate reduction + – – ND v – w ND + + – + + – v Lipase production – + v ND v ND + ND – + v – + – v Amylase production – – – ND v – – ND – – – v w – – Gelatinase production v – – ND – ND – – – – v v + – v

D-glucose utilisation + – v + – w v – + + – + – – + Citrate utilisation – – + ND – w – – + v v + + – +

D-xylose utilisation – + + ND – – – – + + – v – – v

L-Arabinose utilisation – + – + – – v + – + – – – – v Major polar lipids PE, PG, ND PG, ND ND ND ND ND PG, DPG, ND PC, PG, ND PG, ND PC, DPG, DPG, PC DPG, DPG, LPE, AL PC, PE, PE, PL, PE, PL PC, AL AL GL

T T T Taxa: 1, H 14 ; 2, Nautella italica LMG 24365 (Vandecandelaere et al. 2009); 3, Jannaschia rubra 4SM3 (Macián et al. 2005); 4, Ketogulonicigenium (Urbance et al. 2001); 5, Loktanella (Ivanova et al. T T T 2005b); 6, Nereida ignava 2SM4 (Pujalte et al. 2005); 7, Oceanicola marinus AZO-C (Cho and Giovannoni 2004); 8, Octadecabacter (Gosink et al. 1997); 9, Pseudoruegeria aquimaris SW-255 (Yoon T et al. 2007); 10, Roseisalinus antarcticus EL-88 (Labrenz et al. 2005); 11, Roseovarius (Labrenz et al. 1999, Boettcher et al. 2005, Yoon et al. 2008); 12, Ruegeria (Uchino et al. 1998, Arahal et al. 2005, T T Martens et al. 2006); 13, Shimia marina CL-TA03 (Choi and Cho 2006); 14, Thalassobacter arenae GA2-M15 (Kim et al. 2009); 15, Thalassobius (Arahal et al. 2005). All species are catalase-positive. +, Positive; –, negative; w, weakly positive reaction; v, variable reaction depending on strain; ND, no data available; AL, unidentified aminolipid; LPE, lysophosphatidylethanolamine; DP, diphosphatidylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol. Adapted from Ivanova et al. (2010)b.

128

Based on the results of genetic and biochemical analyses, it was determined that Celeribacter neptunius H14T is representative of a novel taxonomic classification at the genus level, under the class α-proteobacteria, family Rhodobacteraceae. The cells are Gram-negative and rod-shaped, 0.8 – 1.8 µm in length and 0.4 – 0.9 µm wide. They are chemoorganotrophic, facultative anaerobes, and are motile by means of single polar or subpolar flagella. Requires 1 – 8 % NaCl, temperatures of 5 – 35 °C (optimum 25 °C) and pH ranging between 4.0 – 10.0 (optimum 7.5 – 8.0) for growth to occur. The DNA G + C content is 59.1 mol%.

5.5. Alteromonas australica H17

Another bacterial strain isolated from early enrichment experiments, designated H17, was subjected to comprehensive taxonomic analysis. Based on 16S rRNA gene sequence analysis it was determined that this strain belonged to the genus Alteromonas, of the class γ-proteobacteria. Based on comparisons of biochemical profiles of H17 with that of the type strains of each of the validly described species of the genus it is suggested that H17 represents a novel species of Alteromonas, for which the specific name australica is proposed.

Bacteria of the genus Alteromonas represent one of the oldest genera of Gram- negative aerobic heterotrophic marine bacteria (Baumann et al. 1972, Baumann and Baumann 1981, Baumann et al. 1984). Bacteria of this genus underwent detailed investigations and a number of reclassifications (Van Landschoot and De Ley 1983, Gauthier et al. 1995, Ivanova et al. 2004) leaving Alteromonas macleodii (Baumann et al. 1972) as a single species of the genus for about a decade. Recently, however, the number of validly described species increased rapidly, reaching nine species (Yoon et al. 2003a, Van Trappen et al. 2004a, Yoon et al. 2004, Ivanova et al. 2005a, Martínez- Checa et al. 2005, Chiu et al. 2007, Vandecandelaere et al. 2008).

The 16S rRNA gene sequence was amplified and sequenced by the Australian Genome Research Facility (AGRF Laboratories) (Brisbane, Australia). The EzTaxon server 2.0 at http://147.47.212.35:8080 was used to identify the closest phylogenetic neighbours (Chun et al. 2007). Phylogenetic analyses were performed as described elsewhere (Ivanova et al. 2004, Ivanova et al. 2005a), using three different taxonomic algorithms: neighbour-joining (NJ), maximum likelihood (ML) and maximum

129 parsimony (MP). All three algorithms are available as part of the PHYLIP software package (Felsenstein 1993). Bootstrap analysis was performed using 1000 replications. For the NJ analysis, the distance matrix was calculated based on Kimura's two- parameter method using the DNADIST program. A subtree was extracted from the resulting phylogenetic tree, which demonstrates the taxonomic position of strain H17 among the Alteromonas (Fig. 5.15). The cluster formed by this group is very stable.

T 96 +X Alteromonas hispanica F-32 {AY926460} 61 +X Alteromonas genovensis LMG 24078T {AM885866} Alteromonas litorea TF-22T {AY428573} 97 +X Alteromonas australica H17 {FJ595485} Alteromonas macleodii DSM 6062T {Y18228} Alteromonas tagae BCRC 17571T {DQ836765} Alteromonas marina SW-47T {AF529060} Alteromonas simiduii BCRC 17572T {DQ836766} Alteromonas addita R10SW13T {AY682202}

100 +X T Alteromonas stellipolaris LMG 21861 {AJ295715} Figure 5.15. Phylogenetic position of A. australica, based on 16S rRNA gene sequence analysis. Bootstrap percentages above 50% are shown at nodes. The symbols + and X indicate branch agreement by maximum likelihood and maximum parsimony algorithms respectively.

Biochemical analysis of A. australica included tests of oxidation/fermentation of glucose, denitrification, oxidase and catalase activity, gelatin liquefaction, arginine dihydrolase, lysine decarboxylase, ornithine decarboxylase, sodium requirements, growth temperatures, indole and H2S production and the ability to hydrolyse starch, Tween-80 and casein. Susceptibility to various antibiotics was also tested. The results of biochemical analyses are summarised in Table 5.10 and Table 5.11.

130

Table 5.10. Comparison of biochemical characteristics of Alteromonas australica to that of other species of Alteromonas.

Characteristic H17 1 2 3 4 5 6 7 8 9 Production of – – – – – – – + – – pigments Growth at: 4 C + – – – + + + – + 10 C + + – + + + + + + + 40 °C + + + + + + – – + – Nitrate – v(–) + – – – – – – – reduction to nitrite Growth in NaCl at: 10% + – + + + + + + + + 15% – – – – + + – – – + Hydrolysis of: Gelatin – v(+) + + + + + + + w Agar – – – – – – + – – – Starch + v(+) – + + + + + + w Utilisation of: D-Mannitol – v (+) – – – + – + – – L-Lactate – v(+) + ND – – + – – + L-Xylose – + – ND + + – ND – + L-Arabinose – – – ND – – + – – +

Sucrose, – + + + + + + v + + maltose, acetate, glycerol G+C mol % 43 45-46 45 43 44-45 46 43 45 46 44.5

Taxa: H17, A. australica; 1, A. macleodii ( Baumann et al. 1972); 2, A. simiduii ( Chiu et al. 2007); 3, A. tagae (Chiu et al. 2007); 4, A. marina (Yoon et al. 2003a); 5. A. hispanica ( Martínez-Checa et al. 2005); 6, A. addita (Ivanova et al. 2005a ); 7, A. stellipolaris ( Van Trappen et al. 2004a ); 8, A. litorea (Yoon et al. 2004); 9, A. genovensis (Vandecandelaere et al. 2008). All species/strains are Gram-negative, motile, oxidase and catalase positive, negative for indole and H2S production, grow at 3-6% NaCl, produce lipase (Tween 80); +, positive; -, negative; v, variable reaction; w, weak reaction.

131

Table 5.11. Susceptibility of Alteromonas australica H17 to various antibiotics.

Antibiotic Concentration Susceptibility of C. neptunius H14T Levomycetin 30 µg per disc Susceptible Benzylpenicillin 10 U Resistant Vancomycin 30 µg per disc Resistant Carbenicillin 100 µg per disc Susceptible Nalidixic acid 30 µg per disc Susceptible Lincomycin 15 µg per disc Resistant Rifampicin 5 µg per disc Susceptible Ofloxacin 5 µg per disc Susceptible Ampicillin 10 µg per disc Susceptible Cephalexin 30 µg per disc Resistant Oleandomycin 15 µg per disc Susceptible Cephazolin 30 µg per disc Susceptible Tetracycline 30 µg per disc Resistant Kanamycin 30 µg per disc Susceptible Oxacillin 10 µg per disc Resistant Streptomycin 30 µg per disc Susceptible Gentamicin 10 µg per disc Susceptible Erythromycin 15 µg per disc Susceptible Polymyxin 300 U Susceptible Neomycin 30 µg per disc Susceptible

Alteromonas australica cells are Gram-negative and rod-shaped, about 0.7 – 0.9 µm in diameter. They are chemoorganotrophic aerobes, and are motile by means of single polar flagella. Requires 1 – 10 % NaCl, temperatures of 4 – 37 °C and pH ranging between 6.0 – 10.0 (optimum 7.5 – 8.0) for growth to occur. The DNA G + C content is 59.1 mol%. Based on the results of genetic and biochemical analyses, it was determined that Alteromonas australica H17 is likely a novel species of the genus Alteromonas, under the class γ-proteobacteria, however due to the relatively high 16S rRNA gene similarity of this strain with other members of species of Alteromonas, further analysis may be required, in the form of DNA-DNA hybridisation or multi-locus sequence analysis.

132

5.6. Summary

A total of 23 different phenotypes classified according to colony morphology were isolated from three different enrichment cultures supplied with PET as the sole carbon and energy source, and a control culture without any PET. Members of several of these phenotypes have high potential for PET biodegradation. Six phenotypes were isolated only from the PET enrichment culture in dark conditions, where the selective advantage for PET degrading organisms was greatest. A further two phenotypes were recovered from each of the cultures supplied with PET, but not from the control.

The composition of the bacterial communities that developed in the enrichment cultures that were supplied with PET was very different to that of the control condition. Each of the three PET enrichment communities were dominated by phenotype e, with a significant contribution from phenotype d, while the phenotype a comprised the largest proportion of the community developed in the control. This indicates that phenotypes d and e have high potential for containing PET degrading organisms.

Based on partial sequencing of the 16S rRNA gene, isolates recovered from the enrichment cultures belonged to the taxonomic classes α-, β- and γ-proteobacteria, Actinobacteria and Flavobacteriia. In the cases of several of the recovered strains, their nearest validly described taxonomic relatives are known to be important sulphur cyclers, which is an indication that sulphur may be an important element in the degradation of PET. Several more strains identified most closely to the genera Thalassospira and Kordiimonas, both of which are comprised of bacterial species that are known to degrade polycyclic aromatic hydrocarbons.

In addition to this, two strains that were isolated from enrichment cultures have been comprehensively characterised. As a result of this work one of these strains, designated H14, has been identified as belonging to a previously undescribed genus, and was subsequently named Celeribacter neptunius H14T. The other is thought to be a new species from the genus Alteromonas, which is expected to be confirmed upon further genetic analysis.

133

Chapter 6: Design of surface topographies for enhancing cellular interactions with PET

______

134

6.1. Overview

Bacterial interactions with PET, or any material for that matter occur almost exclusive on the outer surface of said material. As such, it is crucial that bacterial cells have as unrestricted access as possible to PET surfaces in order for them to have maximal biodegradative effect. In this chapter, the role of substratum surface architecture on attachment and adhesion of bacterial cells to a variety of materials will be covered, with particular emphasis on determining which surface structures allow for the maximum level of cell attachment.

6.2. Thermodynamics of cell attachment on PET

In the general case, the law of free energy of adhesion states that:

– Eq. 6.2.1. where ΔGadh is the free energy of adhesion, γS/B is the interfacial tension (IFT) between a substratum and a bacterial cell, γS/W is the IFT between the substratum and water, and

γB/W is the IFT between the bacterium and water (Bos et al. 1999a). Bacterial adhesion is favoured more as ΔGadh decreases, i.e. the system equilibrium has minimal free energy. According to equation 6.2.1, ΔGadh is lowest when γS/B is small and /or one or both of γS/W and γB/W are large. In the case of the latter two terms, IFT between a surface and water is directly related to hydrophobicity; highly hydrophobic surfaces have high IFT with water. It follows that cell attachment is strongly favoured when both the cell and the substratum are highly hydrophobic.

As a material, PET is moderately hydrophobic (Fig. 6.1). It exhibits a static water contact angle of approximately 81°. According to commonly accepted thresholds of hydrophobicity, PET is actually classified as a hydrophilic material; surfaces with water contact angles below 90° are considered hydrophilic, those between 90° and 150° are hydrophobic, and surfaces with contact angles above 150° are superhydrophobic. However hydrophobicity is a relative concept, and the above definitions are arbitrary.

135

Figure 6.1. Static contact angle of a water droplet on a PET surface.

Static water contact angles were also measured on lawns of bacterial cells immobilised on cellulose acetate filter papers, in order to estimate cell hydrophobicity (Fig. 6.2). Cells belonged to six representative strains isolated from early enrichment experiments. Three strains were identified as members of Alteromonas, while a further three were identified as Thalassospira spp. Water contact angles measured on bacterial lawns ranged from approximately 42° for Alteromonas sp. H89, to approximately 66° for Alteromonas sp. H91. None of the strains proved to be significantly hydrophobic, and were in fact more hydrophilic than the PET surface.

136

Figure 6.2. Static water contact angles of droplets resting on bacterial lawns of six representative strains isolated from enrichment experiments.

As the hydrophobicity of PET and each of the bacterial strains tested is relatively mediocre, it is expected that the level of cell attachment on PET will not be especially high. To test this, confocal laser scanning microscopy was carried out on bacterial biofilms formed on PET samples during enrichment experiments. PET samples were aseptically removed and stained with two fluorescent dyes to visualise viable cells and EPS attached to the surface. The resulting three-dimensional reconstructions show that even after nine months, the cells attached to the PET surface are quite sparse (Fig. 6.3).

137

The amount of EPS on the surface is also quite limited, and does not form a complete layer.

Figure 6.3. Three-dimensional reconstruction of the bacterial biofilm formed on PET surfaces after nine months of enrichment incubation. Viable cells are stained red (bottom left) while exopolysaccharides, which form the biofilm matrix, are stained green (bottom right). The top image combines both channels and demonstrates the location of viable cells within the biofilm matrix.

The restricted cell attachment observed on PET surfaces may be a contributor to the inhibited rate of PET biodegradation, as plastic/bacteria interactions are primarily limited to the surface. It therefore stands to reason that if the affinity of bacteria for PET surfaces can be enhanced, then PET biodegradability may also increase. By investigating the effect that a diverse range of surfaces structures have on regulating bacterial adhesion, it is thought that the specific surface properties that most favour bacterial attachment can be identified. Once these specific properties are known they can be introduced to PET through manipulation of the polymer surface, potentially maximising biodegradation.

138

6.3. Superhydrophobic surfaces

According to the equation of free energy of adhesion (Eq. 6.2.1), attachment of bacterial cells will be favoured on increasingly hydrophobic surfaces. Because of this, surfaces that exhibit extremely high hydrophobicity were chosen as the starting point for cell-substratum interaction studies.

6.3.1. Bacterial attachment on superhydrophobic titanium surfaces

Among superhydrophobic surfaces, the lotus leaf is perhaps the best known example. Much research has been conducted investigating the superhydrophobicity and surface structure of lotus leaves (Barthlott and Neinhuis 1997, Marmur 2004, Sun et al. 2005, Zhang et al. 2006b, Bhushan et al. 2009, Koch et al. 2009, Wang et al. 2009, Zhang et al. 2009). A certain amount of work has also gone into developing techniques for fabrication of lotus structures from a variety of materials (Sun et al. 2005, Bhushan et al. 2009, Koch et al. 2009, Fadeeva et al. 2011). One of these artificial lotus surfaces was selected for investigation of cell adhesion.

The surface in question was fabricated and supplied by a research group at Laser Zentrum Hannover e.V. in Germany using femtosecond laser ablation to create dual- scale hierarchical structures on titanium based on those found on the surfaces of lotus leaves ( Fadeeva et al. 2011). Staphylococcus aureus 65.8T was chosen as a model bacterium for adhesion studies, due to its well characterised nature. Cells were prepared by overnight culture in nutrient broth (Oxoid, U.K.) at 25 °C. The resulting logarithmic- phase cells were then harvested and resuspended in phosphate buffered saline to OD600 = 0.3. Approximately five millilitres of this cell suspension was then placed on lotus- replica titanium samples and incubated at 25 °C for 18 hours, after which time the samples were rinsed with MilliQ water and air dried. Once samples were dry they were sputter-coated with gold using a Dynavac CS300 gold-sputterer for subsequent scanning electron microscopy.

Lotus-structured titanium surfaces proved to be relatively favourable for attachment of S. aureus cells (Fig. 6.4). Attachment was localised primarily to the grooves and crevices between the microscale features on the surface, and encroaching up the sides of the micro-features. On some parts of the surface the bacterial cells appeared to form a

139

‘bridge’ of sorts; suspended across the microscale crevices and anchored to the micro- features on either side (Fig 6.4).

Figure 6.4. Staphylococcus aureus cell attachment on titanium surfaces with lotus- inspired hierarchical surface features.

It is not immediately clear by what mechanism this ‘bacterial suspension bridge’ might form. However, a possible explanation lies within the accepted theories of roughness induced surface wettability. The Cassie-Baxter model of surface wettability states that the apparent contact angle of a water droplet resting on a surface is dependent on the proportion of air trapped within the topographical features of the surface (Cassie and Baxter 1944). Thus as the proportion of trapped air approaches 100%, the apparent water contact angle approaches the contact angle between water and air, i.e. 180°. The air trapped on the surface may also be the key to explaining the cell attachment profile of S. aureus on lotus-like titanium. Nano-sized bubbles become trapped within the nanoscale surface features, and provided they are not spaced too far apart may provide a very low friction surface for bacterial cells to slide across (Fig. 6.5). This would cause the cells to slide down the sides of the microscale features and accumulate in the crevices between them. At the bottom of the crevices the trapped nanobubbles are in proximity to each other close enough to coalesce and form micro-sized air pockets. The

140 cells, lacking significant mass are unable to cross the air/water interface at the bottom of the crevice, and begin to form a layer on the interface, building inwards from the intersection of the three air/water/titanium interfaces.

Figure 6.5. Schematic diagram of the hypothetical cause of cell accumulation and ‘bridge’ formation, based on retention of air bubbles within surface features.

To test this hypothesis, Raman microspectroscopy was performed on the lotus-like titanium surfaces immersed in water. Individual Raman spectra were recorded in a 50 × 50 grid, over 100 µm × 100 µm. A map was then generated based on the intensity of the water peak in the individual spectra, indicating the spatial distribution of air trapped on the surface (Fig. 6.6). By comparing to scanning electron micrographs taken at a similar magnification, it can be seen that the areas of lowest intensity, i.e. the regions that are occupied by air correspond very well with the pattern of crevices between the microscale surface features (Fig. 6.6). While the spatial resolution and sensitivity of the technique is not sufficient to provide conclusive evidence of the presence of nanobubbles on the titanium surface, larger bubbles located at the bottom of the surface crevices were detected, supporting the idea of nano- and micro-bubble directed cell attachment.

141

Time-resolved Raman microspectroscopy was also performed to track the proportion of air retained on the lotus-like titanium surface over time. Titanium surfaces were immersed in water for discrete time intervals (0, 10, 30 and 60 minutes) to determine whether or not the amount of air trapped by surface features remained constant. It was found that the amount of air in fact does not remain constant, and that the trapped air is released over time until the surface nears equilibrium with the surrounding environment (Fig. 6.7). After one hour immersed in water, the amount of air trapped on the surface of the lotus-like titanium decreased to 6 % of the air on the surface immediately upon immersion. This has implications in cell attachment, because the greater the proportion of the surface occupied by air, the less surface area that is available for cell adhesion. Cell adhesion studies after, 0, 10, 30 and 60 minute incubations showed that this is indeed the case; the number of cells attached to the lotus titanium surfaces was very low initially, and increased substantially once the proportion of trapped air dropped below 40 % of the initial value (Fig. 6.7).

142

1.41.4 Adhered cells Retained air 120

2 -

1.21.2 100100 Airretention (%) 1.01 8080 0.80.8 6060 0.60.6 4040 adhered cells mm cells adhered

0.40.4 5

0.20.2 2020 × 10 0.00 00 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Incubation time (min)

6.3.2. Bacterial attachment on superhydrophobic cicada wings

Another type of superhydrophobic surface with a surface structure very different to that of the lotus-like titanium was tested for its favourability to bacterial cell attachment. The surfaces of the wings of the Clanger cicada (Psaltoda claripennis) possess an array of hexagonally-spaced, spherically-capped, conical nanopillars, which help to impart superhydrophobicity (Fig. 6.8). The nanopillars are approximately 200 nm tall, 60 nm in diameter at cap and are spaced approximately 170 nm apart, centre to centre. This surface structure contributes significantly to the hydrophobicity of the surface; the nanopillars are quite effective at trapping air between them, increasing the water contact angle according to the Cassie-Baxter model of wettability. The surface chemistry of the wing also plays a role in determining the hydrophobicity. The wings are composed primarily of three general components: protein, chitin and wax (Vincent and Wegst 2004), however it is mainly cuticular wax that is present on the outer surface of the wing. Figure 6.9 shows chemical distribution maps based on the Amide II C=O stretching peak and the CH2 stretching, which are representative of the protein and wax components of the wing. The relatively higher variation in the amount of wax on the

143 surface is due to the fact that is present as a thin layer across the surface, while protein comprises most of the bulk of the wing. Scans were performed in transmission mode, and thus there was a higher contribution to infrared absorption spectra from proteins.

Figure 6.8. Scanning electron micrograph of a cicada wing surface, titled at an angle of approximately 40°.

144

The net result of the surface structure and surface chemistry of the cicada wings, in terms of wettability, is a surface that is extremely hydrophobic. Spatial mapping of surface wettability of cicada wings showed that in some parts of the wing the water contact angle can be as high as 171°, while the average contact angle across the analysed area was 158.8° (Fig. 6.10). A small degree of spatial variation in surface hydrophobicity was observed on the cicada wing surface, which is likely due to the layer of cuticular wax being inconsistent and/or incomplete, however the structure of the surface ensures that the wing remains particularly hydrophobic.

Figure 6.10. Wettability map of a cicada wing surface.

Cell adhesion experiments on cicada wings using Pseudomonas aeruginosa ATCC 9027 showed that bacterial cells do have appreciable affinity for attaching to the surface. However the wing structures elicited an effect that would be extremely detrimental if they were to be applied for PET biodegradation: the cells were ruptured and killed by the nanopillar structures on the wing surface. For this reason, cicada wing nanopillar structures are not suitable for fabrication on PET, unless their bactericidal effect can be eliminated.

145

6.4. Molecularly smooth surfaces

Lotus-like titanium and cicada wings, like most superhydrophobic surfaces, have relatively high levels of roughness. Irrespective of hydrophobicity, rough surfaces are generally expected to show higher levels of cell attachment than smoother surfaces, due to the higher availability of anchor points, and through providing shelter from potential detaching forces, e.g. water turbulence (Rowan et al. 2002, Whitehead et al. 2005, Rozhok et al. 2006, Díaz et al. 2007). To contrast against the rough lotus-like titanium and cicada wings presented in Section 6.3, some exceptionally smooth, thin titanium films were analysed for their bacterial cell attachment properties. These films were smooth at a molecular level, and as such were expected to show very limited levels of bacterial adhesion.

Titanium films were fabricated using magnetron sputtering. Films were deposited using a Kurt J. Lesker CMS-18 magnetron sputtering system in DC mode, with the power set to 150 kW, argon gas pressure at 4 mTorr and base pressure below 5 × 10-5 mTorr. By modulating the exposure time, films of 3 nm, 12 nm and 150 nm thicknesses were deposited on pre-cleaned silicon wafers. After fabrication, films of each thickness were analysed using atomic force microscopy to assess their roughness and topology. Roughness analysis showed that each film was sub-nanometrically smooth, with the 3 nm film presenting the lowest average roughness at 0.19 nm (Table 6.1). Titanium naturally forms an oxide layer on its outer surface when exposed to atmospheric conditions, and considering that the bond length in a molecule of titanium dioxide is similar in magnitude to the average roughness of the 3 nm film, i.e. approximately 1.6 Å (Ivanova et al. 2011), it can be said that this film is molecularly smooth. The films were moderately hydrophobic, displaying water contact angles within the range of 95 – 105°.

146

Table 6.1. Quantitative roughness analysis of titanium thin films deposited on silicon wafer substrata.

* Sa (nm) Sq (nm) Smax (nm) Ssk Sku WCA (°) 3 nm Ti film 0.19 0.22 2.81 -0.09 3.47 96 12 nm Ti film 0.20 0.24 6.61 0.11 2.63 97 150 nm Ti film 0.66 0.83 7.59 0.63 3.79 104

*WCA – Water contact angle

Cell adhesion studies were carried out on the molecularly smooth titanium films using both Staphylococcus aureus CIP 65.8T and Pseudomonas aeruginosa ATCC 9027. P. aeruginosa behaved in a manner consistent with what is predicted by the literature, i.e. few P. aeruginosa cells were able to adhere to the molecularly smooth titanium surfaces (Figs 6.11, 6.12). However, the attachment profile of S. aureus was vastly different; many cells were able to attach to the smooth titanium surfaces in relatively high density (Figs 6.11, 6.12).

147

148

Given that the titanium surfaces were consistent for both of the strains tested, the differences between the attachment profiles of S. aureus and P. aeruginosa are probably due to cell-specific properties. A likely contributing factor is cell rigidity. S. aureus has higher rigidity than P. aeruginosa, which arises from a combination of several different cell properties, including morphology, thickness of peptidoglycan layer and turgor pressure (Whatmore and Reed 1990, Arnoldi et al. 2000, Koch 2003, Harold 2007). P. aeruginosa cells are relatively soft and flexible, and when these cells are in contact with molecularly smooth surfaces the flexibility of the membrane allows the cell to spread its load across a greater area, which causes the contact points between the cell and the substratum to be limited to just the tip of the sub-nanometric surface features. The cells are easily detached as they can slide across the surface with little resistance. S. aureus on the other hand is more rigid and less deformable, so the total load of the cell is concentrated on a much smaller contact area. The sub-nanoscale features penetrate the cell wall slightly and limit lateral movement, decreasing the likelihood of cell detachment (Fig. 6.13).

149

Figure 6.13. Schematic representation of the hypothetical mechanism of species- dependent bacterial cell adhesion on molecularly smooth titanium surface.

6.5. Summary

As bacterial degradation of PET occurs exclusively at the polymer surface, it is important to characterise how bacterial cells interact with and attach to the plastic. Visualisation of biofilms grown on PET surfaces shows that bacterial attachment to PET surfaces is quite limited, which may be a contributing factor to the inherent resistance of the polymer to biodegradation. If PET surfaces could be modified to be friendlier towards bacterial attachment it may result in an increase in biodegradability.

Bacterial attachment on several substrata with interesting surface properties was assessed as potential sources of inspiration for design of PET surfaces. Two of the substrata were superhydrophobic: titanium with surface features based of the structure of the lotus leaf, and the wings of a species of cicada. The lotus-like titanium did show a relatively high level of retention of bacterial cells; however the spatial distribution of the attached cells appeared to be directed by air bubbles trapped on the surface. As the

150 air was gradually replaced by water the overall cell attachment levels increased. Application of this type of surface structure not only has the potential to boost the level of cell attachment on the surface, but may also be useful for directing cells to attach to specific regions.

Cicada wings also retained quite a large number of cells, however in this case the surface also proved lethal to the cells. Every Pseudomonas aeruginosa cell to attach to the surface was ruptured and killed by the nanopillar structures on the wing surface. This would be detrimental to degradation processes if it were applied to PET.

Molecularly smooth surfaces were also shown to direct bacterial attachment. The number of cells that could attach to the surfaces of molecularly smooth titanium films was found to be dependent on the bacterial species. Spherical Staphylococcus aureus cells adhered to the titanium films in much greater number than the rod-shaped Pseudomonas aeruginosa. It is thought that the different attachment profiles of the two bacteria are based on their structural rigidity. If this is the case, fabrication of PET films with very low roughness could lead to selective attachment of specific bacteria.

151

Chapter 7: General discussion

______

152

7.1. Overview

Several studies have been conducted previously that investigate the potential of biodegradation of polymers, including PET, by microorganisms (Grima et al. 2000, Ishigaki et al. 2000, Müller et al. 2001, Yamada-Onodera et al. 2001, Bonhomme et al. 2003, Marten et al. 2003, Müller et al. 2005, Herzog et al. 2006, Marqués-Calvo et al. 2006, Mueller 2006, Artham and Doble 2009). In most cases, the level of polymer degradation that was achieved was too low to apply for polymer disposal on a commercial scale, and of the techniques that were able to achieve faster degradation rates, few if any could feasibly be implemented commercially due to poor cost- effectiveness. For example, Müller et al. (2005) measured up to 54% loss of mass of PET samples upon incubation in a 0.1 mg mL-1 solution of purified hydrolase from an actinomycete, Thermobifida fusca. The enzyme solution was completely refreshed every week. If such an approach were to be applied on a commercial scale, the expense of preparing and purifying such large amounts of the enzyme would prove inhibitory.

More recently, there has been a shift in the literature away from the development of techniques for biodegradation of resistant polymers, and towards the development of new, more readily degradable polymers (Chen 2009, Akaraonye et al. 2010, Chanprateep 2010, Nampoothiri et al. 2010, Verbeek and van den Berg 2010, Brandelero et al. 2011, Canché-Escamilla et al. 2011, Ghasemlou et al. 2011, Kim et al. 2011b, Kunasundari and Sudesh 2011, Liu et al. 2011, Neppalli et al. 2011, Puls et al. 2011, Raquez et al. 2011, Sui et al. 2011, Ye et al. 2011, Zhu et al. 2011). The major drawback to this approach is that the resistance of polymers like PET to degradation is inextricably linked to their stability and durability, which is one of the primary properties which makes them such popular materials for use in a wide range of applications. Replacement of these polymers with biodegradable materials in many cases may defeat the purpose of using polymers altogether.

For the most part, the limiting factors in PET biodegradation have not previously been determined, however the low polymer chain mobility due to the semi-crystalline nature of the polymer has been identified as a contributor to degradation resistance (Marten et al. 2003, Müller et al. 2005, Mueller 2006). Despite the fact that one of the barriers to PET degradation is known, there has been little attempt to circumvent it. In the previous chapters, it has been shown that marine bacteria do possess the ability to

153 degrade PET, and that some enhancement strategies do aid them in that cause. This chapter will summarise and discuss the results from the previous chapters, and identify the major barriers to PET biodegradation as well as the most effective strategies for developing an effective and efficient PET disposal technique.

7.2. Mechanisms of bacterial biodegradation of PET

Before one can realistically attempt to enhance the biodegradation of PET, something must be known about the way in which the polymer is degraded. By characterising the surface of PET samples that have been partially degraded and comparing them to control samples, the physical and chemical changes induced by the bacteria were determined, and are discussed below.

7.2.1. Physical mechanisms of PET biodegradation

It was consistently observed throughout the PET degradation experiments that the degraded surfaces were noticeably smoother than control surfaces. Rougher surface features, i.e. larger topographical peaks on the surfaces generally became smaller and fewer in number. Based on this observation a hypothesis is proposed for the physical mechanism of bacterial biodegradation of PET. The larger features on the PET surfaces have increased localised surface area relative to the rest of the sample. That makes these regions more available to bacteria, and also enables greater mobility of the polymer chains. Bacterial enzymes are thus better able to access the PET chains in order to induce cleavage of the molecules. Consequently the larger surface features are degraded preferentially, and are subsequently eroded away, resulting in a smoothening of the surface (Fig. 7.1).

154

Figure 7.1. Peak erosion model whereby larger surface features are preferentially worn away by bacterial enzymes, resulting in smaller surface features and overall smoother surfaces.

7.2.2. Chemical mechanisms of PET biodegradation

Generally speaking, the surface chemistry of the degraded PET samples was very similar to the control samples. This may simply be due to the fact that as degradation occurs slowly, the changes in surface chemistry are quantitatively small, and the corresponding differences in the detected signal are somewhat overwhelmed by the normal surface signal. Alternatively, degradation of PET surfaces may occur without much need for surface functionalisation, and instead small sections of the polymer chains are simply excised from the bulk material. Both explanations are equally probable, and likely both contribute.

Excision of small sections of the polymer chains is a particularly likely scenario. The results presented in Chapter 4 show that on some samples, the amount of alcohol groups on the surface increased. This is an indication that hydrolysis occurred at the ester bonds in the polymer molecule, which results in alcohol and carboxylate

155 functionalities on either side of the chain scission. Hydrolysis at two adjacent ester bonds would result in the extraction of a single molecule of ethylene glycol (Fig. 7.2), which could then be easily metabolised by bacterial cells. It is unlikely that the aromatic portion of the polymer chain could be excised in this manner, however, as the aromatic rings tend to stack quite tightly together, which is the primary cause of the limited chain mobility of PET (Marten et al. 2003, Müller et al. 2005).

O O

2H2O

O O

OH HO

OH HO Figure 7.2. Excision of ethylene glycol from PET by hydrolysis.

Aside from hydrolysis, the carbonyl group is a potential target for chain breakage by other mechanisms. X-ray photoelectron spectroscopy data suggest that the enzyme- catalysed formation of amide bonds may displace the ester bond, resulting in chain breakage. Several degraded samples showed an increase in the amount of nitrogen on their surfaces, and in most cases this nitrogen was in the form of C–N groups. In addition to this, an extra component was consistently detected in O1s which is not normally a part of the PET molecule: a non-ester carbonyl group. This suggests a possible degradation mechanism in which the ester carbon atom acts as an electrophile for the lone pair of electrons on a nitrogen atom of another molecule. The nitrogen- containing molecule may be a protein, or perhaps a single amino acid, e.g. lysine. The free electrons form a new bond to the carbon atom, inducing a tetravalent transition

156 state, before one of the oxygen atoms is displaced, breaking the polymer chain (Fig. 7.3).

O O

HN R2 O O n

O O Transition State O O n HN

O O

O N HO

R2 Figure 7.3. Displacement of an ester bond in a PET molecule by formation of an amide bond to a nitrogen-containing molecule.

One further characteristic several of the degraded surfaces had in common was the presence of small amounts of sulphur, usually in the form of sulphate. Sulphur is required as a trace nutrient by most organisms, and many marine bacteria are important cyclers of sulphur. It is unclear whether or not sulphur serves a specific function in PET biodegradation, or if it just a requirement of the cells that are responsible, but detection of sulphur only on degraded sample surfaces is an indication that, one way or another, sulphur plays an important role in bacterial degradation of PET.

157

7.3. Bacterial strains responsible for PET biodegradation

Taxonomic affiliations of several members of the bacterial communities that developed in the bacterial enrichment cultures were determined, in order to identify which marine bacteria might be capable of degrading PET and to provide some clues as to how biodegradation might be enhanced. Over 100 bacterial isolates were recovered from four different enrichment culture conditions: enrichment culture in seawater supplied with PET films, enrichment culture in seawater supplied with PET bottle pieces, enrichment culture in seawater supplied with PET films incubated in dark conditions, and control cultures not containing any plastic. Bacteria were initially grouped according to colony morphology and enumerated by plate counts. Twenty-three different colony morphologies were identified across all of the cultures.

Each of the three cultures that were incubated in the presence of PET was dominated by members of the phenotypes d and e. Of the strains belonging to phenotype d that were identified most belonged to the genus Alteromonas of the class γ- proteobacteria, with a couple of strains of the closely related Aestuariibacter. Several species of Alteromonas are known to be producers of a number of extracellular enzymes (Ivanova et al. 2000 , Ivanova et al. 2005a) that may be involved in PET biodegradation. Identified strains of phenotype e belonged to mostly α- and β-proteobacteria. Two strains of this phenotype identified most closely with Kordiimonas lacus, which is a bacterium known to be able to degrade polycyclic aromatic hydrocarbons (Xu et al. 2011b). The ability to degrade aromatic molecules is a useful ability in PET biodegradation, as the polymer chains possess a relatively high density of aromatic groups. Several other strains across multiple phenotypes identified to the genus Thalassospira of the α-proteobacteria, which are also known aromatic compound degraders (López-López et al. 2002, Liu et al. 2007, Kodama et al. 2008, Zhao et al. 2010).

Phenotypes f and h were isolated in relatively low numbers; however both of these phenotypes were isolated only from each of the three cultures that contained PET, and not from the control. These two phenotypes likely represent the most robust degraders of PET, as they were able to compete with photosynthetic bacteria. Strains of these phenotypes were most closely related to Alteromonas (phenotype f), Thalassospira and Lentibacter. The Alteromonas-related strains in particular were quite distantly related,

158 returning partial 16S gene sequence similarity below 80 %. Not only do these strains have high potential to be PET degrading organisms, they also have high probability for belonging to as yet undescribed species, and perhaps an undescribed genus.

The nearest taxonomic neighbours of several of the strains isolated are important sulphur cycling bacteria. Multiple strains, mostly from phenotype b, returned nearest matches to Roseovarius nubinhibens, which is capable of dimethylsulfoniopropionate and other related sulphur-containing molecules, while several other strains returned matches to Sulfitobacter and Limnobacter, which are capable of obtaining energy via oxidation of sulphide and thiosulphate respectively.

Incubation of PET samples in pure and mixed cultures of strains with high degradation potential such as Thalassospira and Alteromonas failed to produce any significant loss of plastic mass. This is strong indication that a commensal relationship is required for effective biodegradation of PET. The enrichment community taken straight from the culture flask was however capable of degrading PET. The culture likely contained a proportion of sulphur cycling bacteria (e.g. Limnobacter, which belonged to phenotype e, the dominating phenotype in each of the plastic containing cultures), which supplied the PET degrading strains with a source of utilisable sulphur, and in turn these strains produced additional nutrients for the sulphur-cyclers.

Even after eight months of incubation without a significant food source, a few bacteria were still able to be recovered from the control culture. These strains most likely represent the most durable and robust strains present in the initial seawater samples, and are simply able to exist for long periods with limited nutrients. Most of these strains actually grew better in the cultures that contained plastic, probably feeding of the metabolic by-products of other strains with PET-degrading or photosynthetic ability.

In this study, the contributions of non-cultivable bacteria to PET biodegradation were not assessed in detail. The reason for this is that for the eventual application of the knowledge obtained to plastic disposal procedures, the ability to culture adequate amounts of bacteria will be essential. If a non-cultivable bacterium were discovered to possess significant PET degradative ability, it would be extremely hard to grow enough of it for application on a commercial scale. For this reason it was decided to focus primarily on cultivable bacteria.

159

7.4. Barriers to PET biodegradation

From the results of this study it has been determined that marine bacteria do indeed possess the capability for biodegradation of PET. However, there are at least three barriers that limit the degree to which biodegradation occurs. These barriers will be addressed in this section.

One limiting factor in PET biodegradation that was identified previously in the literature is concerning the crystallinity of the polymer (Marten et al. 2003, Müller et al. 2005). PET has a tendency to orient primarily in the trans configuration near its surface, which favours increased crystallinity (Zhu and Kelley 2005). The infrared data presented in Section 4.4.4.2 showed that much of the surface of the PET samples was indeed in trans configuration. If the outer surface were to form a predominantly crystalline layer, then regardless of how much of the bulk material was in amorphous form the limitations imposed on chain mobility by the crystalline outer layer would prohibit bacteria causing significant biodegradation.

The results of topographical characterisation of PET surfaces presented in Chapter 4 indicated that biodegradation occurred preferentially on larger surfaces features and that subsequently degradation occurred fastest on rougher surfaces. However, upon degradation, the surfaces became smoother as the larger surface peaks were eroded away. Degradation of smoother surfaces occurred at a slower rate, thus biodegradation of PET may be a self-limiting process; i.e. degradation causes smoothening of the surface which in turn increases the resistance of the surface to further degradation.

The third barrier to PET biodegradation lies in its propensity to promote bacterial attachment on its surface. In Section 6.2 it was shown that as a material, PET is not a particularly favourable substratum for bacterial attachment. This likely arises from a combination of the mediocre hydrophobicity of the material and its roughness profile. As a result bacterial cells are not able to easily access the PET surface and must produce larger amounts of enzyme to achieve comparable levels of biodegradation.

7.5. Strategies for optimisation of PET biodegradation

Based on the identified barriers to PET biodegradation, a number of strategies can be developed in an attempt to enhance the biodegradability of the polymer. These

160 strategies can be grouped into two broad approaches: techniques aimed at enhancing the biodegradation potential of the bacterial cultures, and techniques for increasing the susceptibility of PET to biodegradation.

The first group of strategies mostly aims at increasing the size of the PET degrading community and increasing the proportion of PET degrading members of the bacterial community. Supplementation of the bacterial community with nitrogen and phosphorus proved effective at increasing PET biodegradation. The mechanisms of degradation appeared to remain unchanged when the bacterial cultures were supplied with these essential nutrients, indicating that the measured increase was simply a result of greater bacterial numbers.

Another strategy for improving PET biodegradation is to manipulate the culture conditions so that PET degrading bacteria have as much selective advantage as possible. The specific example from this study is culture in dark conditions; the absence of light limits the growth potential of photosynthetic organisms, reducing the competitive pressure on PET degraders. Over time the composition of the bacterial community shifts towards higher proportion of the bacterial strains which are capable of obtaining energy through degradation of PET. Culture in dark conditions has not been thoroughly explored in the literature, as one of the primary mechanisms of the limited amount of degradation experienced by PET in the environment is exposure to UV light. Researchers have been reluctant to eliminate one of the few stimuli which are known cause degradation of PET.

The other broad approach for enhancement of PET biodegradation surrounds manipulation of the polymer itself. When it comes to surface modification there are two properties which can be altered to attempt to increase PET degradation: the surface chemistry and topography. However changing the surface chemistry of the material is not always an acceptable approach, for example, many of the current applications of PET are in food packaging, and coating of the polymer with a given chemical may lead to its absorption into the contents. Changing the topography of PET is a more favourable alternative, as its effects on many of the material properties will be minimal.

The roughness of PET surfaces has been identified in this study as a contributor to the resistance of the polymer to biodegradation. When the surface is particularly smooth the rate of biodegradation is decreased. The obvious solution to this problem would be

161 to fabricate a rough topography on the surface of PET during manufacture. However a limitation arises once the bacteria have degraded the material sufficiently to erode the surface features, revealing a smooth surface. Treatment of PET samples with surfactants was shown in Chapter 4 to increase the surface roughness of the polymer, although the selection of the correct surfactant is imperative. The surfactant needs to be adequately non-toxic to the degrading bacteria so as not to be antibacterial, but at the same time not so biocompatible that the bacteria are able to utilise the surfactant for energy over the polymer. In addition to this, a one-time pre-treatment of the polymer will result in the same problem that would arise from fabrication of a rough surface, and multiple treatments would increase costs and decrease efficiency. A possible solution to this is to incorporate a bacterium into the degradation community that is capable of producing biosurfactants. This would result in a dynamic process in which the biosurfactant continuously induces a roughening effect, while the PET degrading organisms erode the peaks.

Another surface modification strategy for enhancing biodegradation is to fabricate surface topographies which promote bacterial attachment. Increasing the amount of cells that can attach to the surface will increase the availability of the surface to bacterial cell. The question remains however as to what surface features enable the highest level of cell attachment. Surface structures based on the topography of the lotus leaf were shown in Chapter 6 to be quite favourable for bacterial attachment, with Staphylococcus aureus cells found to attach in relatively high numbers. The lotus structures were also able to direct bacterial attachment to specific areas by retaining air bubbles between the surface features. Molecularly smooth titanium films were also shown to promote attachment of S. aureus. However the limitation of these surface topographies is that the adhesion levels of other cells, in particular more flexible rod- shaped cells was extremely limited. Therefore they may not be applicable for fabrication on PET to enhance biodegradation. Many of the potential candidates for PET biodegradation identified in Chapter 5 have rod-shaped morphologies, and may be repelled by molecularly smooth surfaces.

As was mentioned previously, any effects that a fabricated PET surface may have on the rate of biodegradation will only last until those surface features have been degraded. However at the least they may stimulate the metabolic pathways of the bacteria which are involved in biodegradation of the polymer. At any rate, the

162 fabrication techniques that have been developed to date are not yet capable of adequately producing and controlling nanoscale surface topographies. Nanoimprinting and mask-mould lithography techniques currently show the most potential for producing highly controlled nano-topographies (Zhang et al. 2006a, Xie et al. 2008, Hong et al. 2009, Koch et al. 2009, Kostovski et al. 2009), but both require more development for application in the fabrication of PET surfaces with enhanced biodegradability.

163

Chapter 8: Conclusions and future directions

______

164

8.1. Summary and conclusions

The ability and potential of marine bacteria to biodegrade PET was assessed via bacterial enrichment cultures selecting for bacteria that could utilise PET as an energy source. The surface modifications to the polymer induced by bacterial interactions were traced with a variety of analytical techniques, and at the same time the composition of the bacterial enrichment community was monitored. It was found that marine bacteria do indeed possess the capability for PET biodegradation.

Bacteria were shown to preferentially degrade larger features on the PET surfaces, and to favour rougher surfaces. The increased exposed surface area of these sections allowed bacterial cells better access to the polymer chains. Based on these observations a model for the physical mechanisms of biodegradation was proposed, in which the larger, more exposed features were degraded faster than other regions of the surfaces, leading to erosion of the surface peaks and overall smoothening of the surfaces. No physical mechanism of degradation relating the rate of biodegradation of PET to the roughness of the polymer surface has been proposed before.

Subtle changes in the chemistry of the PET surfaces led to the proposal of two potential chemical mechanisms of biodegradation. The first assumed that the minor chemical changes detected on the surface were somewhat incidental, and suggested that degradation occurred via hydrolysis of ester bonds and excision of small sections of the molecule. The second proposed mechanism accounted for some of the surface chemistry changes, and suggested the possibility of displacement of an alcohol from the ester bonds, via formation of an amide bond to a secondary substrate or protein.

The bacterial enrichment communities that developed in the presence of PET were found to be dominated by members of the classes α-proteobacteria and γ- proteobacteria. Consistently recovered isolates included strains of Alteromonas, Thalassospira, Roseovarius and Limnobacter. The mostly likely candidates to be PET degrading organisms were strains that returned nearest relatives to species of Thalassospira and Kordiimonas; members of both genera are known to be capable of utilisation of polycyclic aromatic hydrocarbons. Biodegradation however appeared to require commensalism between multiple bacterial species. Other extracellular enzyme producers such as Alteromonas, and sulphur-cycling organisms such as Roseovarius and Limnobacter may play contributing roles.

165

A number of biodegradation enhancement strategies were proposed, based on either manipulation of the bacterial culture or the polymer itself. Bacterial manipulation strategies included techniques for increasing the size of the bacterial community and adjusting its composition towards higher proportions of polymer-degrading members. This was achieved primarily by manipulating culture conditions; supplementation of essential elemental nutrients nitrogen and phosphorus were shown to slightly increase the rate of biodegradation by enabling increase bacterial growth, and increasing the selective advantage of PET degrading organisms by inhibiting photosynthetic also proved effective.

Enhancement strategies based on manipulation of the polymer properties focussed on surface topography and roughness. Bacteria were shown to degrade rougher surfaces faster than smoother surfaces; however degradation resulted in smoothening of the surface, limiting the degree of degradation that could take place. Treatment of the polymer with sodium dodecyl sulphate surfactant increased the surface roughness of the material temporarily, before degradation caused the surface to become smoother again. Biosurfactants produced by the bacterial species Pseudoalteromonas citrea had a very similar surface roughening effect, and thus incorporation of this strain into the bacterial degradation community may help to enhance biodegradation through continuous exposure to surfactants.

Low cell attachment was identified as one of the barriers that inhibit biodegradation of PET. Fabrication of the surface topography of the polymer has the potential to boost the level of attachment cells can achieve, while not affecting the chemical properties of the surface. Lotus-inspired surface features and molecularly smooth surfaces both showed potential to enhance the level of cell attachment on PET; however fabrication techniques are not yet sufficiently developed to adequately reproduce nanoscale surface features.

8.2. Future directions

While the current work has provided a significant contribution to the fundamental understanding of the biodegradation mechanisms of PET, further work is still required to apply this knowledge on a commercial scale. Firstly, the strategies for biodegradation enhancement presented here represent a starting point for further development; however

166 in their current state they are not efficient enough to be commercially viable. Further work needs to be conducted, particularly in the investigation of additional surface topographies and their ability to promote cell adhesion, and in the development of fabrication techniques that can accurately and reliably produce nanoscale topographies to specification.

The other major area in which further study is required is in the cellular metabolism and regulation of the plastic degrading bacteria. The primary focus of this study was to characterise the physical and chemical changes made to PET surfaces in the process of biodegradation, and as such metabolism of bacteria was not explored. Proteomic and genetic analysis of PET degrading bacteria could provide further inspiration for biodegradation enhancement strategies.

8.3. Close

Development of biodegradation as a truly viable alternative plastic waste disposal protocol is extremely beneficial for the environment. Biodegradation is an environmentally friendly process, and has the potential to be much more efficient and cost-effective than recycling, increasing the incentive to move away from traditional, environmentally hazardous techniques.

The data presented here support the premise that biodegradation has the potential to become feasible for application of a large scale. Marine bacteria are indeed capable of degrading PET surfaces, and this work provides significant and novel contributions to the fundamental understanding of biodegradation mechanisms. For the first time a relationship between the rate of biodegradation of PET and its surface roughness has been identified, and a number of bacterial strains have been isolated and identified as having high potential as plastic degrading organisms. Also, several enhancement strategies for the biodegradation of PET were presented.

167

Bibliography

______

168

Abbott, E. J. and Firestone, F. A. (1933). Specifying surface quality. Mechanical Engineering 55: 569-572.

Akaraonye, E., Keshavarz, T. and Roy, I. (2010). Production of polyhydroxyalkanoates: the future green materials of choice. Journal of Chemical Technology & Biotechnology 85: 732-743.

Amann, R. I., Ludwig, W. and Schleifer, K.-H. (1995). Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews 59: 143-169.

Amari, T. and Ozaki, Y. (2001). Real-time monitoring of the initial oligomerization of bis(hydroxyethyl) terephthalate by attenuated total reflection/infrared spectroscopy and chemometrics. Macromolecules 34: 7459-7462.

Amass, W., Amass, A. and Tighe, B. (1998). A review of biodegradable polymers: uses, current developments in the synthesis and characterization of biodegradable polyesters, blends of biodegradable polymers and recent advances in biodegradation studies. Polymer International 47: 89-144.

Aminot, A. and Kérouel, R. (2004). Dissolved organic carbon, nitrogen and phosphorus in the N-E Atlantic and the N-W Mediterranean with particular reference to non- refractory fractions and degradation. Deep Sea Research Part I: Oceanographic Research Papers 51: 1975-1999.

Amor, S. B., Jacquet, M., Fioux, P. and Nardin, M. (2009). XPS characterisation of plasma treated and zinc oxide coated PET. Applied Surface Science 255: 5052- 5061.

An, Y. H. and Friedman, R. J. (1998). Concise review of mechanisms of bacterial adhesion to biomaterial surfaces. Journal of Biomedical Materials Research 43: 338-348.

An, Y. H., Friedman, R. J., Draughn, R. A., Smith, E. A., Nicholson, J. H. and John, J. F. (1995). Rapid quantification of staphylococci adhered to titanium surfaces using image analyzed epifluorescence microscopy. Journal of Microbiological Methods 24 : 29-40.

169

Andrady, A. L. (2011). Microplastics in the marine environment. Marine Pollution Bulletin 62 : 1596-1605.

Arahal, D. R., Macián, M. C., Garay, E. and Pujalte, M. J. (2005). Thalassobius mediterraneus gen. nov., sp. nov., and reclassification of Ruegeria gelatinovorans as Thalassobius gelatinovorans comb. nov. International Journal of Systematic and Evolutionary Microbiology 55: 2371-2376.

Armstrong, E., Yan, L., Boyd, K. G., Wright, P. C. and Burgess, J. G. (2001). The symbiotic role of marine microbes on living surfaces. Hydrobiologia 461: 37-40.

Arnoldi, M., Fritz, M., Bäuerlein, E., Radmacher, M., Sackmann, E. and Boulbitch, A. (2000). Bacterial turgor pressure can be measured by atomic force micrscopy. Physical Review E 62: 1034-1044.

Artham, T. and Doble, M. (2009). Biodegradation of Physicochemically Treated Polycarbonate by Fungi. Biomacromolecules 11: 20-28.

Asakuma, Y., Nakagawa, K., Maeda, K. and Fukui, K. (2009). Theoretical study of the transesterification reaction of polethylene terephthalate under basic conditions. Polymer Degradation and Stability 94: 240-245.

Astrup, T., Møller, J. and Fruergaard, T. (2009). Incineration and co-combustion of waste: accounting of greenhouse gases and global warming contributions. Waste Management & Research 27: 789-799.

Awaja, F. and Pavel, D. (2005). Recycling of PET. European Polymer Journal 41: 1453-1477.

Azzarello, M. Y. and Van Vleet, E. S. (1987). Marine birds and plastic pollution. Marine Ecology- Progress Series 37: 295-303.

Baird, R. W. and Hooker, S. K. (2000). Ingestion of plastic and unusual prey by a juvenile harbour porpoise Marine Pollution Bulletin 40: 719-720.

Barnes, D. K. A. and Milner, P. (2005). Drifting plastic and its consequences for sessile organism dispersal in the Atlantic Ocean. Marine Biology 146: 815-825.

170

Barreiros, J. P. and Barcelos, J. (2001). Plastic ingestion by a leatherback turtle Dermochelys coriacea from the Azores (NE Atlantic). Marine Pollution Bulletin 42: 1196-1197.

Barthlott, W. and Neinhuis, C. (1997). Purity of the sacred lotus, or escape from contamination in biological surfaces. Planta 202: 1-8.

Baumann, P. and Baumann, L. (1981). The marine Gram-negative eubacteria; genera Photobacterium, Beneckea, Alteromonas, Pseudomonas, and Alcaligenes. In: Starr, M. P., Stolp, H., Trüper, H. G., Balows, A. and Schlegel, H. G. The Prokaryotes. A Handbook on Habitats, Isolation, and Identification of Bacteria. pp: 1302-1330. (Springer-Verlag, Berlin).

Baumann, P., Baumann, L., Mandel, M. and Allen, R. D. (1972). Taxonomy of aerobic marine eubacteria. Journal of Bacteriology 110: 402-429.

Baumann, P., Gauthier, M. J. and Baumann, L. (1984). Genus Alteromonas. Baumann, Baumann, Mandel and Allen 1972. In: Krieg, N. R. and Holt, J. G. Bergey's Manual of Systematic Bacteriology. pp: 243-354. (Williams & Wilkins Co., Baltimore).

Bazaka, K., Jacob, M. V., Crawford, R. J. and Ivanova, E. P. (2011). Plasma-assisted surface modification of organic biopolymers to prevent bacterial attachment. Acta Biomaterialia 7: 2015-2028.

Bendat, J. S. and Piersol, A. G. (1966). Measurement and Analysis of Random Data. ed. (Wiley, New York).

Bennett, J. M. and Mattsson, L. (1989). Intorduction to Surface Roughness and Scattering. ed. (Optical Society of America, Washington D.C.).

Bergeret, A., Ferry, L. and Ienny, P. (2009). Influence of the fibre/matrix interface on ageing mechanisms of glass fibre reinforced thermoplastic composites (PA-6,6, PET, PBT) in a hygrothermal environment. Polymer Degradation and Stability 94: 1315-1324.

171

Bhushan, B., Jung, Y. C., Niemietz, A. and Koch, K. (2009). Lotus-like biomimetic hierarchical structures developed by the self-assembly of tubular plant waxes. Langmuir 25: 1659-1666.

Binnig, G., Quate, C. F. and Gerber, C. (1986). Atomic force microscope. Physical Review Letters 56: 930-933.

Blenkinsopp, S. A. and Costerton, J. W. (1991). Understanding bacterial biofilms. Trends in Biotechnology 9: 138-143.

Blight, L. K. and Burger, A. E. (1997). Occurrence of plastic particles in sea-birds from the eastern north pacific. Marine Pollution Bulletin 34: 323-325.

Blomfield, C. J. (2005). Spatially resolved X-ray photoelectron spectroscopy. Journal of Electron Spectroscopy and Related Phenomena 143: 241-249.

Boettcher, K. J., Geaghan, K. K., Maloy, A. P. and Barber, B. J. (2005). Roseovarius crassostreae sp. nov., a member of the Roseobacter clade and the apparent cause of juvenile oyster disease (JOD) in cultured Eastern oysters. International Journal of Systematic and Evolutionary Microbiology 55: 1531-1537.

Bonhomme, S., Cuer, A., Delort, A.-M., Lemaire, J., Sancelme, M. and Scott, G. (2003). Environmental degradation of polyethylene. Polymer Degradation and Stability 81: 441-452.

Bos, R., van der Mei, H. C. and Busscher, H. J. (1999a). Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study. FEMS Microbiology Reviews 23 : 179-230.

Bos, R., Van Der Mei, H. C. and Busscher, H. J. (1999b). Physico-chemistry of initial microbial adhesive interactions - Its mechanisms and methods for study. FEMS Microbiology Reviews 23 : 179-229.

Boulangé-Petermann, L., Rault, J. and Bellon-Fontaine, M.-N. (1997). Adhesion of Streptococcus thermophilus to stainless steel with different surface topography and roughness. Biofouling 11: 201-216.

172

Brandelero, R. P. H., Grossmann, M. V. E. and Yamashita, F. (2011). Effect of the method of production of the blends on mechanical and structural properties of biodegradable starch films produced by blown extrusion. Carbohydrate Polymers 86 : 1344-1350.

Brettar, I., Christen, R. and Höfle, M. G. (2003). Idiomarina baltica sp. nov., a marine bacterium with a high optimum growth temperature isolated from surface water of the central Baltic Sea. International Journal of Systematic and Evolutionary Microbiology 53: 407-413.

Brune, D., Hellborg, R., Whitlow, H. J. and Hunderi, O. (1997). Surface Characterization. ed. (Wiley-VCH, Weinheim).

Burgess, J. G., Boyd, K. G., Armstrong, E., Jiang, Z., Yan, L., Berggren, M., May, U., Pisacane, T., Granmo, Å. and Adams, D. R. (2003). The development of a marine natural product-based antifouling paint. Biofouling 19: 197-205.

Burgess, J. G., Kawaguchi, R., Sakaguchi, T., Thornhill, R. H. and Matsunaga, T. (1993). Evolutionary relationships among Magnetospirillum strains inferred from phylogenetic analysis of 16S rDNA sequences. Journal of Bacteriology 175: 6689-6694.

Busscher, H. J., Norde, W., Sharma, P. K. and van der Mei, H. C. (2010). Interfacial re- arrangement in initial microbial adhesion to surfaces. Current Opinion in Colloid and Interface Science 15: 510-517.

Canché-Escamilla, G., Canché-Canché, M., Duarte-Aranda, S., Cáceres-Farfán, M. and Borges-Argáez, R. (2011). Mechanical properties and biodegradation of thermoplastic starches obtained from grafted starches with acrylics. Carbohydrate Polymers 86: 1501-1508.

Cardi, N., Po, R., Giannotta, G., Occhiello, E., Garbassi, F. and Messina, G. (1993). Chain extension of recycled poly(ethylene terephthalate) with 2,2'-bis(2- oxazoline). Journal of Applied Polymer Science 50: 1501-1509.

173

Carrigg, C., Rice, O., Kavanagh, S., Collins, G. and O'Flaherty, V. (2007). DNA extraction method affects microbial community profiles from soils and sediments. Applied Microbiology and Biotechnology 77: 955-964.

Cassie, A. B. D. and Baxter, S. (1944). Wettability of porous surfaces. Transactions of the Faraday Society 40: 546-551.

Censky, E. J., Hodge, K. and Dudley, J. (1998). Over-water dispersal of lizards due to hurricanes. Nature 395: 556.

Chanprateep, S. (2010). Current trends in biodegradable polyhydroxyalkanoates. Journal of Bioscience and Bioengineering 110: 621-632.

Chen, G.-Q. (2009). A microbial polyhydroxyalkanoates (PHA) based bio- and materials industry. Chemical Society Reviews 38: 2434-2446.

Chiu, H.-H., Shieh, W. Y., Lin, S. Y., Tseng, C.-M., Chiang, P.-W. and Wagner-Döbler, I. (2007). Alteromonas tagae sp. nov. and Alteromonas simiduii sp. nov., mercury-resistant bacteria isolated from a Taiwanese estuary. International Journal of Systematic Bacteriology 57: 1209-1216.

Cho, J.-C. and Giovannoni, S. J. (2004). Oceanicola granulosus gen. nov., sp. nov. and Oceanicola batsensis sp. nov., poly-β-hydroxybutyrate-producing marine bacteria in the order 'Rhodobacteriales'. International Journal of Systematic and Evolutionary Microbiology 54: 1129-1136.

Choi, D. H. and Cho, B. C. (2006). Shimia marina gen. nov., sp. nov., a novel bacterium of the Roseobacter clade isolated from biofilm in a coastal fish farm. International Journal of Systematic and Evolutionary Microbiology 56: 1869- 1873.

Chun, J. and Goodfellow, M. (1995). A phylogenetic analysis of the genus Nocardia with 16S rRNA gene sequences. International Journal of Systematic Bacteriology 45: 240-245.

Chun, J., Lee, J.-H., Jung, Y., Kim, M., Kim, S., Kim, B. K. and Lim, Y. W. (2007). EzTaxon: a web-based tool for the identification of prokaryotes based on 16S

174

ribosomal RNA gene sequences. International Journal of Systematic and Evolutionary Microbiology 57: 2259-2261.

Chung, S. Y., Yettella, R. R., Kim, J. S., Kwon, K., Kim, M. C. and Min, D. B. (2011). Effects of grilling and roasting on the levels of polycyclic aromatic hydrocarbons in beef and pork. Food Chemistry 129: 1420-1426.

Coates, J. (2000). Interpretation of infrared spectra, a practical approach. In: Meyers, R. A. Encyclopedia of Analytical Chemistry. pp: 10815-10837. (John Wiley & Sons Ltd., Chichester).

Crawford, R., Koopal, L. K. and Ralston, J. (1987). Contact angles on particles and plates. Colloids and Surfaces 27: 57-64.

Croxford, A. M., Nandakumar, K. S., Holmdahl, R., Tobin, M. J., McNaughton, D. and Rowley, M. J. (2011). Chemical changes demonstrated in cartilage by synchrotron infrared microspectroscopy in an antibody-induced murine model of rheumatoid arthritis. Journal of Biomedical Optics 16: Art no. 066004.

De Gregoris, T. B., Aldred, N., Clare, A. S. and Burgess, G. (2011). Improvement of phylum- and class-specific primers for real-time PCR quantification of bacterial taxa. Journal of Microbiological Methods 86: 351-356.

De, J., Ramaiah, N. and Vardanyan, L. (2008). Detoxification of toxic heavy metals by marine bacteria highly resistant to mercury. Marine Biotechnology: Accessed online 17/14/08, http://www.springerlink.com/content/551t007558163385/fulltext.pdf.

Demertzis, P. G., Johansson, F., Lievens, C. and Franz, R. (1997). Studies on the development of a quick inertness test procedure for multi-use PET containers - sorption behaviour of bottle wall strips. Packaging Technology and Science 10: 45-58.

Denner, E. B. M., Vybiral, D., Koblížek, M., Kämpfer, P., Busse, H.-J. and Velimirov, B. (2002). Erythrobacter citreus sp. nov., a yellow-pgimented bacterium that lacks bacteriochlorophyll a, isolated from the western Mediterranean Sea.

175

International Journal of Systematic and Evolutionary Microbiology 52: 1655- 1661.

Denuncio, P., Bastida, R., Dassis, M., Giardino, G., Gerpe, M. and Rodríguez, D. (2011). Plastic ingestion in Franciscana dolphins, Ponotporia blainvillei (Gervais and d'Orbigny, 1844), from Argentina. Marine Pollution Bulletin 62: 1836-1841.

Derraik, J. G. B. (2002). The pollution of the marine environment by plastic debris: a review. Marine Pollution Bulletin 44: 842-852.

Devos, L., Clymans, K., Boon, N. and Verstraete, W. (2005). Evaluation of nested PCR assays for the detection of Legionella pneumophila in a wide range of aquatic samples. Journal of Applied Microbiology 99: 916-925.

Díaz, C., Schilardi, P. L., Salvarezza, R. C. and Fernández Lorenzo de Mele, M. (2007). Nano/microscale order affects the early stages of biofilm formation on metal surfaces. Langmuir 23: 11206-11210.

Dietzek, B., Cialla, D., Schmitt, M. and Popp, J. (2010). Introduction to the fundamentals of Raman spectroscopy. In: Dieing, T., Hollricher, O. and Toporski, J. Confocal Raman Microscopy. pp: 21-42. (Springer, Heidelberg).

Dorobantu, L. S. and Gray, M. R. (2010). Application of atomic force microscopy in bacterial research. Scanning 32: 74-96.

Engel, A., Lyubchenko, Y. and Müller, D. (1999). Atomic force microscopy: a powerful tool to observe biomolecules at work. Trends in Cell Biology 9: 77-80.

Fadeeva, E., Truong, V. K., Stiesch, M., Chichkov, B. N., Crawford, R. J., Wang, J. and Ivanova, E. P. (2011). Bacterial retention on superhydrophobic titanium surfaces fabricated by femtosecond laser ablation. Langmuir 27: 3012-3019.

Felsenstein, J. (1989). PHYLIP (Phylogeny Inference Package) version 3.2. Cladistics 5: 164-166.

Felsenstein, J. (1993). PHYLIP (Phylogenetic Inference Package). Version 3.68. Seattle, Department of Genome Sciences, University of Washington.

176

Fritz, M., Radmacher, M., Cleveland, J. P., Allersma, M. W., Stewart, R. J., Gieselmann, R., Janmey, P., Schmidt, C. F. and Hansma, P. K. (1995). Imaging globular and filamentous proteins in physiological buffer solutions with tapping mode atomic force miccroscopy. Langmuir 11.

Frost, A. and Cullen, M. (1997). Marine debris on northern New South Wales beaches (Australia): sources and the role of beach usage. Marine Pollution Bulletin 34: 348-352.

Fukuhara, Y., Kasai, D., Katayama, Y., Fukuda, M. and Masai, E. (2008). Enzymatic properties of terephthalate 1,2-dioxygenase of Comamonas sp. strain E6. Bioscience, Biotechnology, and Biochemistry 72: 2335-2341.

Galtier, N., Gouy, M. and Gautier, C. (1996). Sea View and Phylo_win, two graphic tools for sequence alignment and molecular phylogeny. Computer Applications in the Biosciences 12: 543-548.

Gauthier, G., Gauthier, M. and Christen, R. (1995). Phylogenetic analysis of the genera Alteromonas, Shewanella, and Moritella using genes coding for small-subunit rRNA sequences and division of the genus Alteromonas into two genera, Alteromonas (Emended) and Pseudoalteromonas gen. nov. and twelve new species combinations. International Journal of Systematic Bacteriology 45: 755- 761.

Gauthier, M. J. (1977). Alteromonas citrea, a new Gram-negative, yellow-pigmented species from seawater. International Journal of Systematic Bacteriology 27: 349-354.

Gauthier, M. J., Lafay, B., Christen, R., Fernandez, L., Acquaviva, M., Bonin, P. and Bertrand, J.-C. (1992). Marinobacter hydrocarbonoclasticus gen. nov., sp. nov., a new, extremely halotolerant, hydrocarbon-degrading marine bacterium. International Journal of Systematic Bacteriology 42: 568-576.

Gentle, I. and Barnes, G. (2005). Interfacial Science. ed. (Oxford University Press, New York).

177

Ghasemlou, M., Khodaiyan, F., Oromiehie, A. and Yarmand, M. S. (2011). Development and characterisation of a new biodegradable edible film made from kefiran, an exopolysaccharide obtained from kefir grains. Food Chemistry 127: 1496-1502.

González, J. M., Covert, J. S., Whitman, W. B., Henriksen, J. R., Mayer, F., Scharf, B., Schmitt, R., Buchan, A., Fuhrmann, J. A., Kiene, R. P. and Moran, M. A. (2003). Silicibacter pomeroyi sp. nov. and Roseovarius nubinhibens sp. nov., dimethylsulfoniopropionate-demethylating bacteria from marine environments. International Journal of Systematic and Evolutionary Microbiology 53: 1261- 1269.

Gonzalez, J. M., Kiene, R. P. and Moran, M. A. (1999). Transformation of Sulfur Compounds by an Abundant Lineage of Marine Bacteria in the alpha -Subclass of the Class Proteobacteria. Applied and Environmental Microbiology 65: 3810- 3819.

Gosink, J. J., Herwig, R. P. and Staley, J. T. (1997). Octadecabacter arcticus gen. nov., sp. nov., and O. antarcticus, sp. nov., nonpigmented, psychrophilic gas vacuolate bacteria from polar sea ice and water. Systematic and Applied Microbiology 20: 356-365.

Greenwood, J. A. (1984). A unified theory of surface roughness. Proceedings of the Royal Society A 393: 133-157.

Gregory, M. R. (2009). Environmental implications of plastic debris in marine settings - entanglement, ingestion, smothering, hangers-on, hitch-hiking and alien invasions. Philosophical Transactions of the Royal Society B 364: 2013-2025.

Grima, S., Bellon-Maurel, V., Feuilloley, P. and Silvestre, F. (2000). Aerobic biodegradation of polymers in solid-state conditions: a review of environmental and physicochemical parameter settings in laboratory simulations. Journal of Polymers and the Environment 8: 183-195.

Guy, D. W., Crawford, R. J. and Mainwaring, D. E. (1996). The wetting behaviour of several organic liquids in water on coal surfaces. Fuel 75: 238-242.

178

Hall-Stoodley, L., Costerton, J. W. and Stoodley, P. (2004). Bacterial biofilms: from the natural environment to infectious diseases. Nature Reviews Microbiology 2: 95- 108.

Hall, T. A. (1999). BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95-98.

Harold, F. M. (2007). Bacterial morphogenesis: learning how cells make cells. Current Opinion in Microbiology 10: 591-595.

Harris, J. K., Kelley, S. T. and Pace, N. R. (2004). New perspective on uncultured bacterial phylogenetic division OP11. Applied and Environmental Microbiology 70: 845-849.

Hartschuh, A., Sánchez, E. J., Xie, X. S. and Novotny, L. (2003). High-resolution near- field Raman microscopy of single-walled carbon nanotubes. Physical Review Letters 90: 095503.

Hayase, N., Sogabe, T., Itou, R., Yamamori, N. and Sunamoto, J. (2003). Polymer film produced by a marine bacterium. Journal of Bioscience and Bioengineering 95: 72-76.

Hazen, T. C., Dubinsky, E. A., DeSantis, T. Z., Andersen, G. L., Piceno, Y. M., Singh, N., Jansson, J. K., Probst, A., Borglin, S. E., Fortney, J. L., Stringfellow, W. T., Bill, M., Conrad, M. E., Tom, L. M., Chavarria, K. L., Alusi, T. R., Lamendella, R., Joyner, D. C., Spier, C., Baelum, J., Auer, M., Zemla, M. L., Chakraborty, R., Sonnenthal, E. L., D'haeseleer, P., Holman, H.-Y. N., Osman, S., Lu, Z., Van Nostrand, J. D., Deng, Y., Zhou, J. and Mason, O. U. (2010). Deep-sea oil plume enriches indigenous oil-degrading bacteria. Science 330: 204-208.

Head, I. M. and Swannell, R. P. J. (1999). Bioremediation of petroleum hydrocarbon contaminants in marine habitats. Current Opinion in Biotechnology 10: 234-239.

Heid, C. A., Stevens, J., Livak, K. J. and Williams, P. M. (1996). Real time quantitative PCR. Genome Research 6: 986-994.

179

Henderson, M. A. (1998). Evidence for bicarbonate formation on vacuum annealed

TiO2(110) resulting from a precursor-mediated interaction between CO2 and

H2O. Surface Science 400: 203-219.

Herzog, K., Müller, R.-J. and Deckwer, W.-D. (2006). Mechanism and kinetics of the enzymatic hydrolysis of polyester nanoparticles by lipases. Polymer Degradation and Stability 91: 2486-2498.

Hesse, R. and Denecke, R. (2011). Improved Tougaard background calculation by introduction of fittable parameters for the ineleastic electron scattering cross- section in the peak fit of photoelectron spectra with UNIFIT 2011. Surface and Interface Analysis 43: 1514-1526.

Hirai, H., Takada, H., Ogata, Y., Yamashita, R., Mizukawa, K., Saha, M., Kwan, C., Moore, C., Gray, H., Laursen, D., Zettler, E. R., Farrington, J. W., Reddy, C. M., Peacock, E. E. and Ward, M. W. (2011). Organic micropollutants in marine plastic debrisfrom the open ocean and remote and urban beaches. Marine Pollution Bulletin 62: 1683-1692.

Hiraishi, A., Yonemitsu, Y., Matsushita, M., Shin, Y. K., Kuraishi, H. and Kawahara, K. (2002). Characterization of Porphyrobacter sanguineus sp. nov., an aerobic bacteriochlorophyll-containing bacterium capable of degrading biphenyl and dibenzofuran. Archives of Microbiology 178: 45-52.

Hollander, J. M. and Jolly, W. L. (1970). X-ray photoelectron spectroscopy. Accounts of Chemical Research 3: 193-200.

Hong, S.-H., Hwang, J. and Lee, H. (2009). Replication of cicada wing's nanopatterns by hot embossing and UV nanoimprinting. Nanotechnology 20: Art. no. 385303.

Horisawa, S., Sakuma, Y., Nakamura, Y. and Doi, S. (2008). Profiling of a microbial community under confined conditions in a fed-batch garbage decomposer by denaturing gradient gel electrophoresis. Bioresource Technology 99 : 3084-3093.

Iranzo, M., Sainz-Pardo, I., Boluda, R., Sánchez, J. and Mormeneo, S. (2001). The use of microorgansims in environmental remediation. Annals of Microbiology 51: 135-143.

180

Ishigaki, T., Sugano, W., Ike, M., Kawagoshi, Y., Fukunaga, I. and Fujita, M. (2000). Abundance of polymers degrading microorganisms in a sea-based solid waste disposal site. Journal of Basic Microbiology 40: 177-186.

Ivanova, E. P., Bowman, J. P., Lysenko, A. M., Zhukova, N. V., Gorshkova, N. M., Sergeev, A. F. and Mikhailov, V. V. (2005a). Alteromonas addita sp. nov. International Journal of Systematic and Evolutionary Microbiology 55: 1065- 1068.

Ivanova, E. P., Flavier, S. and Christen, R. (2004). Phylogenetic relationships among marine Alteromonas-like proteobacteria: emendation of the family Alteromonadaceae and proposal of Pseudoalteromonadaceae fam. nov., Collwelliaceae fam. nov., Shewanellaceae fam. nov., Moritellaceae fam. nov., Ferrimonadaceae fam. nov., Idiomarinaceae fam. nov. and Psychromonadaceae fam. nov. International Journal of Systematic and Evolutionary Microbiology 54: 1773-1788.

Ivanova, E. P., Kiprianova, E. A., Mikhailov, V. V., Levanova, G. F., Garagulya, A. D., Gorshkova, N. M., Vysotskii, M. V., Nicolau, D. V., Yumoto, N., Taguchi, T. and Yoshikawa, S. (1998). Phenotypic diversity of Pseudoalteromonas citrea from different marine habitats and emendation of the description. International Journal of Systematic Bacteriology 48 : 247-256.

Ivanova, E. P., Sawabe, T., Gorshkova, N. M., Svetashev, V. I., Mikhailov, V. V., Nicolau, D. V. and Christen, R. (2001). Shewanella japonica. sp. nov. International Journal of Systematic and Evolutionary Microbiology 51: 1027- 1033.

Ivanova, E. P., Sawabe, T., Lysenko, A. M., Gorshkova, N. M., Hayashi, K., Zhukova, N. V., Nicolau, D. V., Christen, R. and Mikhailov, V. V. (2002a). Pseudoalteromonas translucida sp. nov. and Pseudoalteromonas paragorgicola sp. nov., and emended description of the genus. International Journal of Systematic and Evolutionary Microbiology 52: 1759-1766.

Ivanova, E. P., Sawabe, T., Lysenko, A. M., Gorshkova, N. M., Svetashev, V. I., Nicolau, D. V., Yumoto, N., Taguchi, T., Yoshikawa, S., Christen, R. and

181

Mikhailov, V. V. (2002b). Pseudoalteromonas ruthenica sp. nov., isolated from marine invertebrates. International Journal of Systematic and Evolutionary Microbiology 52: 235-240.

Ivanova, E. P., Truong, V. K., Wang, J. Y., Berndt, C. C., Jones, R. T., Yusuf, I. I., Peake, I., Schmidt, H. W., Fluke, C., Barnes, D. and Crawford, R. J. (2010a). Impact of nanoscale roughness of titanium thin film surfaces on bacterial retention. Langmuir 26: 1973-1982.

Ivanova, E. P., Truong, V. K., Webb, H. K., Baulin, V. A., Wang, J. Y., Mohammodi, N., Wang, F., Fluke, C. and Crawford, R. J. (2011). Differential attraction and repulsion of Staphylococcus aureus and Pseudomonas aeruginosa on molecularly smooth titanium films. Scientific Reports 1: 165-172.

Ivanova, E. P., Webb, H., Christen, R., Zhukova, N. V., Kurilenko, V. V., Kalinovskaya, N. I. and Crawford, R. J. (2010b). Celeribacter neptunius gen. nov. sp. nov., a new member of Alphaproteobacteria. International Journal of Systematic and Evolutionary Microbiology 60: 1620-1625.

Ivanova, E. P., Zhukova, N. V., Lysenko, A. M., Gorshkova, N. M., Sergeev, A. F., Mikhailov, V. V. and Bowman, J. P. (2005b). Loktanella agnita sp. nov. and Loktanella rosea sp. nov., from the north-west Pacific Ocean. International Journal of Systematic and Evolutionary Microbiology 55: 2203-2207.

Ivanova, E. P., Zhukova, N. V., Svetashev, V. I., Gorshkova, N. M., Kurilenko, V. V., Frolova, G. M. and Mikhailov, V. V. (2000). Evaluation of phospholipid and fatty acid compositions as chemotaxonomic markers of Alteromonas-like proteobacteria. Current Microbiology 41: 341-345.

Ivar do Sul, J. A. and Costa, M. F. (2007). Marine debris review for Latin America and the wider Caribbean region: from the 1970s until now, and where do we go from here? Marine Pollution Bulletin 54: 1087-1104.

Ivar do Sul, J. A., Santos, I. R., Friedrich, A. C., Matthiensen, A. and Fillmann, G. (2011). Plastic pollution at a sea turtle conservation area in NE Brazil:

182

contrasting developed and undeveloped beaches. Estuaries and Coasts 34: 814- 823.

Ivleva, N. P., Wagner, M., Horn, H., Niessner, R. and Haisch, C. (2009). Towards a nondestructive chemical characterization of biofilm matrix by Raman microscopy. Analytical and Bioanalytical Chemistry 393: 197-206.

Joergensen, L., Klösgen, B., Simonsen, A. C., Borch, J. and Hagesaether, E. (2011). New insights into the mucoadhesion of pectins by AFM roughness parameters in combination with SPR. International Journal of Pharmaceutics 411 : 162-168.

Johansson, L.-S. and Campbell, J. M. (2004). Reproducible XPS on biopolymers: cellulose studies. Surface and Interface Analysis 36: 1018-1022.

Joyce, E. A., Chan, K., Salama, N. R. and Falkow, S. (2002). Redefining bacterial populations: a post-genomic reformation. Nature Reviews Genetics 3: 462-473.

Kalinovskaya, N. I., Ivanova, E. P., Alexeeva, Y. V., Gorshkova, N. M., Kuznetsova, T. A., Dmitrenok, A. S. and Nicolau, D. V. (2004). Low-molecular-weight, biologically active compounds from marine Pseudoalteromonas species. Current Microbiology 48: 441-446.

Katsikogianni, M., Amanatides, E., Mataras, D. and Missirlis, Y. F. (2008). Staphylococcus epidermidis adhesion to He, He/O2 plasma treated PET films and aged materials: Contributions of surface free energy and shear rate. Colloids and Surfaces B: Biointerfaces 65: 257-268.

Katsikogianni, M. G. and Missirlis, Y. F. (2010). Bacterial adhesion onto materials with specific surface chemistries under flow conditions. Journal of Materials Science: Materials in Medicine 21: 963-968.

Khoo, H. H. and Tan, R. B. H. (2010). Environmental impacts of conventional plastic and bio-based carrier bags: Part 2:End-of-life options. International Journal of Life Cycle Assessment 15 : 338-345.

Kim, B.-Y., Weon, H.-Y., Son, J.-A., Lee, C.-M., Hong, S.-B., Jeon, Y.-A., Koo, B.-S. and Kwon, S.-W. (2009). Thalassobacter arenae sp. nov., isolated from sea sand

183

in Korea. International Journal of Systematic and Evolutionary Microbiology 59: 487-490.

Kim, K. S., Shin, M. K., Ahn, J.-J., Haw, C.-R. and Park, H.-K. (2011a). Investigation of hair shaft in seborrheic dermatitis using atomic force microscopy. Skin Research and Technology 17: 288-294.

Kim, S. H., Tan, J. P. K., Fukushima, K., Nederberg, F., Yang, Y. Y., Waymouth, R. M. and Hedrick, J. L. (2011b). Thermoresponsive nanostructured polycarbonate block copolymers as biodegradable therapeutic delivery carriers. Biomaterials 32: 5505-5514.

Kint, D. and Muñoz-Guerra, S. (1999). A review on the potential biodegradability of poly(ethylene terephthalate). Polymer International 48: 346-352.

Koch, A. L. (2003). Bacterial wall as a target for attack: past, present, and future research. Clinical Microbiology Reviews 16: 673-687.

Koch, K., Bhushan, B., Jung, Y. C. and Barthlott, W. (2009). Fabrication of artificial lotus leaves and significance of hierarchical structure for superhydrophobicity and low adhesion. Soft Matter 5: 1386-1393.

Kodama, Y., Stiknowati, L. I., Ueki, A., Ueki, K. and Watanabe, K. (2008). Thalassospira tepidiphila sp. nov., a polycyclic aromatic hydrocarbon-degrading bacterium isolated from seawater. International Journal of Systematic and Evolutionary Microbiology 58: 711-715.

Kondratowicz, F. L. and Ukielski, R. (2009). Synthesis and hydrolytic degradation of poly(ethylene succinate) and poly(ethylene terephthalate) copolymers. Polymer Degradation and Stability 94: 375-382.

Kostovski, G., White, D. J., Mitchell, A., Austin, M. W. and Stoddart, P. R. (2009). Nanoimprinted optical fibres: biotemplated nanostructures for SERS sensing. Biosensors and Bioelectronics 24: 1531-1535.

Kunasundari, B. and Sudesh, K. (2011). Isolation and recovery of microbial polyhydroxyalkanoates. Express Polymer Letters 5: 620-634.

184

Kuo, C. C., Liu, C. C., Lin, C. C., Liou, Y. Y., Lan, Y. F. and He, J. L. (2008). Effects of oxygen flow rate on microstructure and properties of indium molybdenum oxide films by ion-beam-assisted deposition. Vacuum 82: 441-447.

Kwon, K. K., Lee, H.-S., Yang, S. H. and Kim, S.-J. (2005). Kordiimonas gwangyangensis gen. nov., sp. nov., a marine bacterium isolated from marine sediments that forms a distinct phyletic lineage (Kordiimonadales ord. nov.) in the 'Alphaproteobacteria'. International Journal of Systematic and Evolutionary Microbiology 55: 2033-2037.

Labrenz, M., Collins, M. D., Lawson, P. A., Tindall, B. J., Schumann, P. and Hirsch, P. (1999). Roseovarius tolerans gen. nov., sp. nov., a budding bacterium with variable bacteriochlorophyll a production from hypersaline Ekho Lake. International Journal of Systematic Bacteriology 49: 137-147.

Labrenz, M., Lawson, P. A., Tindall, B. J., Collins, M. D. and Hirsch, P. (2005). Roseisalinus antarcticus gen. nov., sp. nov., a novel aerobic bacteriochlorophyll a-producing α-proteobacterium isolated from hypersaline Ekho Lake, Antarctica. International Journal of Systematic and Evolutionary Microbiology 55: 41-47.

Landini, P., Antoniani, D., Burgess, J. G. and Nijland, R. (2010). Molecular mechanisms of compounds affecting bacterial biofilm formation and dispersal. Applied Microbiology and Biotechnology 86: 813-823.

Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. A., McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R., Thompson, J. D., Gibson, T. J. and Higgins, D. G. (2007). ClustalW and ClustalX version 2. Bioinformatics 23: 2947-2948.

Law, K. L., Morét-Ferguson, S., Maximenko, N. A., Proskurowski, G., Peacock, E. E., Hafner, J. and Reddy, C. M. (2010). Plastic accumulation in the North Atlantic subtropical gyre. Science 329: 1185-1188.

185

Lee, G.-J., Park, K.-H., Park, Y.-G. and Park, H.-K. (2010). A quantitative AFM analysis of nano-scale surface roughness in various orthodontic brackets. Micron 41: 775-782.

Leone, L., Loring, J., Sjöberg, S., Persson, P. and Shchukarev, A. (2006). Surface characterization of the Gram-positive bacteria Bacillus subtilis-an XPS study. Surface and Interface Analysis 38: 202-205.

Levchik, S. V. and Weil, E. D. (2004). A review on thermal decomposition and combustion of thermoplastic polyesters. Polymers for Advanced Technologies 15: 691-700.

Li, B. and Logan, B. E. (2004). Bacterial adhesion to glass and metal-oxide surfaces. Colloids and Surfaces B: Biointerfaces 36: 81-90.

Li, Z., Qian, S. and Wang, W. (2011). Influence of superalloy substrate roughness on adhesion and oxidation behavior of magnetron-sputtered NiCoCrAlY coatings. Applied Surface Science 257: 10414-10420.

Lin, K.-Y., Sheu, S.-Y., Chang, P.-S., Cho, J.-C. and Chen, W.-M. (2007). Oceanicola marinus sp. nov., a marine alphaproteobacterium isolated from seawater collected off Taiwan. International Journal of Systematic and Evolutionary Microbiology 57: 1625-1629.

Liu, C., Wu, Y., Li, L., Ma, Y. and Shao, Z. (2007). Thalassospira xiamenensis sp. nov. and Thalassospira profundimaris sp. nov. International Journal of Systematic and Evolutionary Microbiology 57: 316-320.

Liu, J., Jiang, Z., Zhang, S., Liu, C., Gross, R. A., Kyriakides, T. R. and Saltzman, W. M. (2011). Biodegradation, biocompatibility, and drug delivery in poly(ω- pentadecalactone-co-p-dioxanone) copolyesters. Biomaterials 32: 6646-6654.

Liu, S. and Wang, Y. (2010). Application of AFM in microbiology: a review. Scanning 32: 61-73.

López-López, A., Pujalte, M. J., Benlloch, S., Mata-Roig, M., Rosselló-Mora, R., Garay, E. and Rodriguez-Valera, F. (2002). Thalassospira lucentensis gen. nov.,

186

sp. nov., a new marine member of the α-Proteobacteria. International Journal of Systematic and Evolutionary Microbiology 52: 1277-1283.

Lu, H., Sato, Y., Fujimura, R., Nishizawa, T., Kamijo, T. and Ohta, H. (2011). Limnobacter litoralis sp. nov., a thiosulfate-oxidising heterotrophic bacterium isolated from a volcanic deposit, and emended description of the genus Limnobacter. International Journal of Systematic and Evolutionary Microbiology 61: 404-407.

Luigi, M., Gaetano, D. M., Vivia, B. and Angelina, L. G. (2007). Biodegradative potential and characterization of psychrotolerant polychlorinated biphenyl- degrading marine bacteria isolated from a coastal station in the Terra Nova Bay (Ross Sea, Antarctica). Marine Pollution Bulletin 54: 1754-1761.

Macián, M. C., Arahal, D. R., Garay, E., Ludwig, W., Schleifer, K. H. and Pujalte, M. J. (2005). Jannaschia rubra sp. nov., a red-pigmented bacterium isolated from sea water. International Journal of Systematic and Evolutionary Microbiology 55: 649-653.

Madigan, M. T., Martinko, J. M. and Parker, J. (2003). Brock Biology of Microorganisms. 10 ed. (Pearson Education LTD., London).

Malik, S., Beer, M., Megharaj, M. and Naidu, R. (2008). The use of molecular techniques to characterize the microbial communities in contaminated soil and water. Environment International 34: 265-276.

Marmur, A. (2004). The lotus effect: superhydrophobicity and metastability. Langmuir 20: 3517-3519.

Marqués-Calvo, M. S., Cerdà-Cuéllar, M., Kint, D. P. R., Bou, J. J. and Muñoz-Guerra, S. (2006). Enzymatic and microbial biodegradability of poly(ethylene terephthalate) copolymers containing nitrated units. Polymer Degradation and Stability 91 : 663-671.

Marten, E., Müller, R.-J. and Deckwer, W.-D. (2003). Studies on the enzymatic hydrolysis of polyesters I. Low molecular mass model esters and aliphatic polyesters. Polymer Degradation and Stability 80: 485-501.

187

Martens, T., Heidorn, T., Pukall, R., Simon, M., Tindall, B. J. and Brinkhoff, T. (2006). Reclassification of Roseobacter gallaeciensis Ruiz-Ponte et al. 1998 asn Phaeobacter gallaeciensis gen. nov., comb. nov., description of Phaeobacter inhibens sp. nov., reclassification of Ruegeria algicola (Lafay et al. 1995) Uchino et al. 1999 as Marinovum algicola gen. nov., comb. nov., and emended descriptions of the genera Roseobacter, Ruegeria and Leisingera. International Journal of Systematic and Evolutionary Microbiology 56: 1293-1304.

Martínez-Checa, F., Béjar, V., Llamas, I., del Moral, A. and Quesada, E. (2005). Alteromonas hispanica sp. nov., a polyunsaturated-fatty-acid-producing, halophilic bacterium isolated from Fuente de Piedra, southern Spain. International Journal of Systematic Bacteriology 55: 2385-2390.

Martinez-Ribes, L., Basterretxea, G., Palmer, M. and Tintoré, J. (2007). Origin and abundance of beach debris in the Balearic Islands. Scientia Marina 71: 305-314.

Masó, M., Garcés, E., Pagès, F. and Camp, J. (2003). Drifting plastic debris as a potential vector for harmful algal bloom (HAB) species. Scientia Marina 67: 107-111.

Massardier-Nageotte, V., Pestre, C., Cruard-Pardet, T. and Bayard, R. (2006). Aerobic and anerobic biodegradability of polymer films and phsyico-chemical characterization. Polymer Degradation and Stability 91: 620-627.

Mato, Y., Isobe, T., Takada, H., Kanehiro, H., Ohtake, C. and Kaminuma, T. (2001). Plastic resin pellets as a transport medium for toxic chemicals in the marine environment. Environmental Science & Technology 35: 318-324.

Menzies, D. J., Cowie, B., Fong, C., Forsythe, J. S., Gengenbach, T. R., McLean, K. M., Puskar, L., Textor, M., Thomsen, L., Tobin, M. and Muir, B. W. (2010). One- step method for generating PEG-like plasma polymer gradients: chemical charaterization and analysis of protein interactions. Langmuir 26: 13987-13994.

Mitik-Dineva, N., Wang, J., Mocanasu, R. C., Stoddart, P. R., Crawford, R. J. and Ivanova, E. P. (2008). Impact of nano-topography on bacterial attachment. Biotechnology Journal 3: 536-544.

188

Moore, C. J. (2008). Synthetic polymers in the marine environment: a rapidly increasing long-term threat. Environmental Research 108: 131-139.

Moore, C. J., Moore, S. L., Leecaster, M. K. and Weisberg, S. B. (2001). A comparison of plastic and plankton in the north Pacific central gyre. Marine Pollution Bulletin 42: 1297-1300.

Morishige, C., Donohue, M. J., Flint, E., Swenson, C. and Woolaway, C. (2007). Factors affecting marine debris deposition at French Frigate Shoals, northwestern Hawaiian islands marine national monument, 1990-2006. Marine Pollution Bulletin 54: 1162-1169.

Mueller, R.-J. (2006). Biological degradation of synthetic polyesters- enzymes as potential catalysts for polyester recycling. Process Biochemistry 41: 2124-2128.

Müller, D. J. and Dufrêne, Y. F. (2008). Atomic force microscopy as a multifunctional molecular toolbox in nanobiotechnology. Nature Nanotechnology 3: 261-269.

Müller, R.-J., Kleeberg, I. and Deckwer, W.-D. (2001). Biodegradation of polyesters containing aromatic constituents. Journal of Biotechnology 86: 87-95.

Müller, R.-J., Schrader, H., Profe, J., Dresler, K. and Deckwer, W.-D. (2005). Enzymatic degradation of poly(ethylene terephthalate): rapid hydrolyse using a hydrolase from T. fusca. Macromolecular Rapid Communications 26: 1400- 1405.

Muyzer, G., de-Waal, E. C. and Uitterlinden, A. G. (1993). Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Applied and Environmental Microbiology 59: 695-700.

Nampoothiri, K. M., Nair, N. R. and John, R. P. (2010). An overview of the recent developments in polylactide (PLA) research. Bioresource Technology 101: 8493-8501.

189

Neppalli, R., Causin, V., Marega, C., Saini, R., Mba, M. and Marigo, A. (2011). Structure, morphology and biodegradability of poly(ε-caprolactone)-based nanocomposites. Polymer Engineering & Science 51: 1489-1496.

Nijland, R., Hall, M. J. and Burgess, J. G. (2010). Dispersal of biofilms by secreted, matrix degrading, bacterial DNase. PLoS One 5: Art no. e15668.

Nikolopoulou, M. and Kalogerakis, N. (2008). Enhanced bioremediation of crude oil utilizing lipophilic fertilizers combined with biosurfactants and molasses. Marine Pollution Bulletin 56: 1855-1861.

Nitschke, M., Costa, S. G. V. A. O. and Contiero, J. (2011). Rhamnolipids and PHAs: recent reports on Pseudomonas-derived molecules of increasing industrial interest. Process Biochemistry 46: 621-630.

Olsen, G. J., Lane, D. J., Giovannoni, S. J., Pace, N. R. and Stahl, D. A. (1986). Microbial ecology and evolution: a ribosomal RNA approach. Annual Reviews of Microbiology 40: 337-365.

Öner, D. and McCarthy, T. J. (2000). Ultrahydrophobic surfaces. Effects of topography length scales on wettability. Langmuir 16: 7777-7782.

Paci, M. and La Mantia, F. P. (1998). Competition between degradation and chain extension during processing of reclaimed poly(ethylene terephthalate). Polymer Degradation and Stability 61: 417-420.

Paci, M. and La Mantia, F. P. (1999). Influence of small amounts of polyvinylchloride on the recycling of polyethyleneterephthalate. Polymer Degradation and Stability 63: 11-14.

Palmer, J., Flint, S. and Brooks, J. (2007). Bacterial cell attachment, the beginning of a biofilm. Journal of Industrial Microbiology & Biotechnology 34: 577-588.

Pappas, D., Smith, B. W. and Winefordner, J. D. (2000). Raman imaging for two- dimensional chemical analysis. Applied Spectroscopy Reviews 35: 1-23.

190

Park, B.-D. and Jeong, H.-W. (2011). Effects of acid hydrolysis on microstructure of cured urea-formaldehyde resins using atomic force microscopy. Journal of Applied Polymer Science 122: 3255-3262.

Park, J. R., Bae, J.-W., Nam, Y.-D., Chang, H.-W., Kwon, H.-Y., Quan, Z.-X. and Park, Y.-H. (2007). Sulfitobacter litoralis sp. nov., a marine bacterium isolated from the East Sea, Korea. International Journal of Systematic and Evolutionary Microbiology 57: 692-695.

Parreira, P., Magalhaes, A., Gonaçalves, I. C., Gomes, J., Vidal, R., Reis, C. A., Leckband, D. E. and Martins, M. C. L. (2011). Effect of surface chemistry on bacterial adhesion, viability, and morphology. Journal of Biomedical Materials Research - Part A 99 A: 344-353.

Peklenik, J. (1967). Investigation of surface typology. CIRP Annals 15: 381-385.

Pemberton, D., Brothers, N. P. and Kirkwood, R. (1992). Entanglement of Australian fur seals in man-made debris in Tasmanian waters. Wildlife Research 19: 151- 159.

Piedad Díaz, M., Boyd, K. G., Grigson, S. J. W. and Burgess, J. G. (2002). Biodegradation of crude oil across a wide range of salinities by an extremely halotolerant bacterial consortium MPD-M, immobilized onto polypropylene fibers. Biotechnology and Bioengineering 79: 145-153.

Pieper, D. H. and Reineke, W. (2000). Engineering bacteria for bioremediation. Current Opinion in Biotechnology 11: 262-270.

Ploux, L., Ponche, A. and Anselme, K. (2010). Bacteria/material interfaces: Role of the material and cell wall properties. Journal of Adhesion Science and Technology 24: 2165-2201.

Pujalte, M. J., Macián, M. C., Arahal, D. R., Ludwig, W., Schleifer, K. H. and Garay, E. (2005). Nereida ignava gen. nov., sp. nov., a novel aerobic marine α- proteobacterium that is closely related to uncultured Prionitis (alga) gall symbionts. International Journal of Systematic and Evolutionary Microbiology 55: 631-636.

191

Puls, J., Wilson, S. A. and Hölter, D. (2011). Degradation of cellulose acetate-based materials: a review. Journal of Polymers and the Environment 19: 152-165.

Raguénès, G., Cambon-Bonavita, M. A., Lohier, J. F., Boisset, C. and Guezennec, J. (2003). A novel, highly viscous polysaccharide excreted by an Alteromonas isolated from a deep-sea hydrothermal vent shrimp. Current Microbiology 46: 448-452.

Raguenes, G., Pignet, P., Gauthier, G., Peres, A., Christen, R., Rougeaux, H., Barbier, G. and Guezennec, J. (1996). Description of a new polymer-secreting bacterium from a deep-sea hydrothermal vent, Alteromonas macleodii su bsp. fijiensis, and preliminary characterization of the polymer. Applied and Environmental Microbiology 62: 67-73.

Raquez, J.-M., Bourgeois, A., Jacobs, H., Degée, P., Alexandre, M. and Dubois, P. (2011). Oxidative degradations of oxodegradable LDPE enhanced with thermoplastic pea starch: thermo-mechanical properties, morphology, and UV- ageing studies. Journal of Applied Polymer Science 122: 489-496.

Ravenschlag, K., Sahm, K. and Amann, R. (2001). Quantitative molecular analysis of the microbial community in marine arctic sediments (Svalbard). Applied and Environmental Microbiology 67: 387-395.

Rios, L. M., Moore, C. and Jones, P. R. (2007). Persistent organic pollutants carried by synthetic polymers in the ocean environment. Marine Pollution Bulletin 54: 1230-1237.

Rosa, A. P. and Triguis, J. A. (2007). Bioremediation process on Brazil shoreline. Laboratory experiments. Environmental Science and Pollution Research 14: 470-476.

Rowan, B., Wheeler, M. A. and Crooks, R. M. (2002). Patterning bacteria within hyperbranched polymer film templates. Langmuir 18: 9914-9917.

Rozhok, S., Fan, Z., Nyamjav, D., Liu, C., Mirkin, C. A. and Holz, R. C. (2006). Attachment of motile bacterial cells to prealigned holed microarrays. Langmuir 22: 11251-11254.

192

Russo, M. A. L., O'Sullivan, C., Rounsefell, B., Halley, P. J., Truss, R. and Clarke, W. P. (2009). The anaerobic degradability of thermoplastic starch: Polyvinyl alcohol blends: Potential biodegradable food packaging materials. Bioresource Technology 100: 1705-1710.

Ryan, P. G., Moore, C. J., van Franeker, J. A. and Moloney, C. L. (2009). Monitroing the abundance of plastic debris in the marine environment. Philosophical Transactions of the Royal Society B 364: 1999-2012.

Salvi, A. M., Castle, J. E., Watts, J. F. and Desimoni, E. (1995). Peak fitting of the chromium 2p XPS spectrum. Applied Surface Science 90: 333-341.

Samyn, P., Van Erps, J., Thienpont, H. and Schoukens, G. (2011). Paper coatings with multi-scale roughness evaluated at different sampling sizes. Applied Surface Science 257 : 5613-5625.

Sandt, C., Smith-Palmer, T., Comeau, J. and Pink, D. (2009). Quantification of water and biomass in small colony variant PAO1 biofilms by confocal Raman microspectroscopy. Applied Microbiology and Biotechnology 83: 1171-1182.

Santos, I. R., Friedrich, A. C. and Ivar do Sul, J. A. (2008). Marine debris contamination along undeveloped tropical beaches from northeast Brazil. Environmental Monitoring and Assessment 148: 455-462.

Santra, T. S., Bhattacharyya, T. K., Patel, P., Tseng, F. G. and Barik, T. K. (2011). Structural and tribological properties of diamond-like nanocomposite thin films. Surface & Coatings Technology 206: 228-233.

Sazima, I., Gadig, O. B. F., Namora, R. C. and Motta, F. S. (2002). Plastic debris collars on juvenile carcharhinid sharks (Rhizoprionodon lalandii) in southwest Atlantic. Marine Pollution Bulletin 44: 1147-1149.

Schecter, A., Colacino, J., Haffner, D., Patel, K., Opel, M., Päpke, O. and Birnbaum, L. (2010). Perfluorinated compounds, polychlorinated biphenyls, and organochlorine pesticide contamination in composite food samples from Dallas, Texas, USA. Environmental Health Perspectives 118: 796-802.

193

Schlafli, H. R., Weiss, M. A., Leisinger, T. and Cook, A. M. (1994). Terephthalate 1,2- dioxygenase system from Comamonas testosteroni T-2: purification and some properties of the oxygenase component. Journal of Bacteriology -176: 6644 6652.

Schneider, U., Steckroth, A., Rau, N. and Hübner, G. (1988). An approach to the evaluation of surface profiles by separating them into functionally different parts. Surface Topography 1: 343-355.

Shen, C., Tang, X., Yao, J., Shi, D., Fang, J., Khan, M. I., Cheema, S. A. and Chen, Y. (2010). Levels and patterns of polycyclic aromatic hydrocabons and polychlorinated biphenyls in municipal waste incinerator bottom ash in Zhejiang province, China. Journal of Hazardous Materials 179: 197-202.

Simjouw, J.-P., Mulholland, M. and Minor, E. (2004). Changes in dissolved organic matter characteristics in Chincoteague Bay during a bloom of the pelagophyte Aureococcus anophagefferens. Estuaries and Coasts 27: 986-998.

Simoneit, B. R. T., Medeiros, P. M. and Didyk, B. M. (2005). Combustion products of plastics as indicators for refuse burning in the atmosphere. Environmental Science & Technology 39: 6961-6970.

Sinha, V., Patel, M. R. and Patel, J. V. (2010). PET waste management by chemical recycling: a review. Journal of Polymers and the Environment 18: 8-25.

Spragg, R. C. and Whitehouse, D. J. (1974). An average wavelength parameter for surface metrology. Revue-M: Mecanique 20: 293-300.

Spring, S., Kämpfer, P. and Schleifer, K. H. (2001). Limnobacter thiooxidans gen. nov., sp. nov., a novel thiosulfate-oxidizing bacterium isolated from freshwater lake sediment. International Journal of Systematic and Evolutionary Microbiology 51: 1463-1470.

St John, A. M., Best, S. P., Wang, Y., Tobin, M. J., Puskar, L., Siegele, R., Cattrall, R. W. and Kolev, S. D. (2011). Micrometer-scale 2D mapping of the composition and homogeneity of polymer inclusion membranes. Australian Journal of Chemistry 64: 930-938.

194

Stackebrandt, E., Frederiksen, W., Garrity, G. M., Grimont, P. A. D., Kämpfer, P., Maiden, M. C. J., Nesme, X., Rosselló-Mora, R., Swings, J., Trüper, H. G., Vauterin, L., Ward, A. C. and Whitman, W. B. (2002). Report of the ad hoc committee for the re-evaluation of the species definition in bacteriology. International Journal of Systematic and Evolutionary Microbiology 52: 1043- 1047.

Stout, K. J., Sullivan, P. J., Dong, W. P., Mainsah, E., Luo, N., Mathia, T. and Zahouani, H. (1993). Development of methods for the characterisation of roughness in three dimensions. ed. (Publication no. EUR 15178 EN of the Commission of the European Communities Dissemination of Scientific and Technical Knowledge Unit, Luxembourg).

Sui, X., Fu, Z. and Shi, Y. (2011). Synthesis of degradable copolymer networks containing hemiacetal components and well-defined backbones. Journal of Applied Polymer Science 121: 1860-1865.

Sun, M., Luo, C., Xu, L., Ji, H., Ouyang, Q., Yu, D. and Chen, Y. (2005). Artificial lotus leaf by nanocasting. Langmuir 21: 8978-8981.

Svenson, A., Sjöholm, S., Allard, A.-S. and Kaj, L. (2009). Antiestrogenicity and estrogenicity in leachates from solid waste deposits. Environmental Toxicology 26: 233-239.

Švorčík, V., Siegel, J., Slepička, P., Kotál, V., Švorčíková, J. and Špirková, M. (2007). Au nanolayers deposited on polyethyleneterephthalate and polytetrafluoroethylene degraded by plasma discharge. Surface and Interface Analysis 39: 79-85.

Synnes, M. (2007). Bioprospecting of organisms from the deep sea: scientific and environmental aspects. Clean Technologies and Environmental Policy 9: 53-59.

Tait, E., Carman, M. and Sievert, S. M. (2007). Phylogenetic diversity of bacteria associated with ascidians in Eel Pond (Woods Hole, Massachusetts, USA). Journal of Experimental Marine Biology and Ecology 342: 138-146.

195

Takeuchi, M., Kawahata, H., Gupta, L. P., Kita, N., Morishita, Y., Ono, Y. and Komai, T. (2007). Arsenic resistance and removal by marine and non-marine bacteria. Journal of Biotechnology 127: 434-442.

Tansel, B. and Yildiz, B. S. (2011). Goal-based waste management strategy to reduce persistence of contaminants in leachate at municipal solid waste landfills. Environment, Development and Sustainability 13: 821-831.

Thomas, T. R. (1998). Rough Surfaces. 2nd ed. (Imperial College Press, London).

Tindall, B. J., Rosselló-Móra, R., Busse, H.-J., Ludwig, W. and Kämpfer, P. (2010). Notes of the characterization of prokaryote strains for taxonomic purposes. International Journal of Systematic and Evolutionary Microbiology 60: 249- 266.

Tollner, E. W., Annis, P. A. and Das, K. C. (2011). Evaluation of strength properties of polypropylene-based polymers in simulated landfill and oven conditions. Journal of Environmental Engineering 137: 291-296.

Trudel, D., Scheringer, M., von Goetz, N. and Hungerbühler, K. (2011). Total consumer exposure to polybrominated diphenyl ethers in North America and Europe. Environmental Science & Technology 45: 2391-2397.

Truong, V. K., Lapovok, R., Estrin, Y. S., Rundell, S., Wang, J. Y., Fluke, C. J., Crawford, R. J. and Ivanova, E. P. (2010). The influence of nano-scale surface roughness on bacterial adhesion to ultrafine-grained titanium. Biomaterials 31: 3674-3683.

Tsuchida, D., Kajihara, Y., Shimidzu, N., Hamamura, K. and Nagase, M. (2011). Hydrogen sulfide production by sulfate-reducing bacteria utilizing additives eluted from plastic resins. Waste Management & Research 29: 594-601.

Tsukada, T. and Sasajima, K. (1983). An assessment of anisotropic properties in three- dimensional asperities. Bulletin of the Japan Society of Precision Engineering 17: 261-262.

196

Türetgen, I. and Cotuk, A. (2007). Monitoring of biofilm-associated Legionella pneumophila on different substrata in model cooling tower system. Environmental Monitoring and Assessment 125: 271-279.

Ubbink, J. and Schär-Zammaretti, P. (2007). Colloidal properties and specific interactions of bacterial surfaces. Current Opinion in Colloid and Interface Science 12 : 263-270.

Uchino, Y., Hirata, A., Yokota, A. and Sugiyama, J. (1998). Reclassification of marine Agrobacterium species: Proposals of Stappia stellulata gen. nov., comb. nov., Stappia aggregata sp. nov., nom. rev., Ruegeria atlantica gen. nov., comb. nov., Ruegeria gelatinovora comb. nov., Ruegeria algicola comb. nov., and Ahrensia kieliense gen. nov., sp. nov., nom. rev. Journal of General and Applied Microbiology 44: 201-210.

Urase, T., Okumura, H., Panyosaranya, S. and Inamura, A. (2008). Emission of volatile organic compounds from solid waste disposal sites and importance of heat management. Waste Management & Research 26: 534-538.

Urbance, J. W., Bratina, B. J., Stoddard, S. F. and Schmidt, T. M. (2001). Taxonomic characterization of Ketogulonigenium vulgare gen. nov., sp. nov., and Ketogulonigenium robustum sp. nov., which oxidise L-sorbose to 2-keto-L- gulonic acid. International Journal of Systematic and Evolutionary Microbiology 51: 1059-1070.

Valavanidis, A., Iliopoulos, N., Gotsis, G. and Fiotakis, K. (2008). Persistent free- radicals, heavy metals and PAHs generated in particulate soot emissions and residue ash from controlled combustion of common types of plastics. Journal of Hazardous Materials 156: 277-284.

Van Landschoot, A. and De Ley, J. (1983). Intra- and intergeneric similarities of the rRNA cistrons of Alteromonas, Marinomonas (gen. nov.) and some other Gram- negative bacteria. Journal of General Microbiology 129: 3057-3074.

197

Van Oss, C. J., Good, R. J. and Chaudhury, M. K. (1986). The role of van der Waals forces and hydrogen bonds in "hydrophobic interactions" between biopolymers and low energy surfaces. Journal of Colloid and Interface Science 111: 378-390.

Van Oss, C. J., Good, R. J. and Chaudhury, M. K. (1988). Additive and nonadditive surface tension components and the interpretation of contact angles. Langmuir 4: 884-891.

Van Trappen, S., Tan, T.-L., Yang, J., Mergaert, J. and Swings, J. (2004a). Alteromonas stellipolaris sp. nov.: novel budding and prosthecate bacteria from Antarctic seas. International Journal of Systematic and Evolutionary Microbiology 54: 1157-1163.

Van Trappen, S., Tan, T.-L., Yang, J., Mergaert, J. and Swings, J. (2004b). Glaciecola polaris sp. nov., a novel budding and prosthecate bacterium from the Arctic Ocean, and emended description of the genus Glaciecola. International Journal of Systematic and Evolutionary Microbiology 54: 1765-1771.

Vandecandelaere, I., Nercessian, O., Segaert, E., Achouak, W., Mollica, A., Faimali, M., De Vos, P. and Vandamme, P. (2008). Alteromonas genovensis sp. nov. isolated from a marine electroactive biofilm and emended description of Alteromonas macleodii Baumann et al. 1972 (Approved lists 1980). International Journal of Systematic and Evolutionary Microbiology 58: 2589- 2596.

Vandecandelaere, I., Nercessian, O., Segaert, E., Achouak, W., Mollica, A., Faimali, M. and Vandamme, P. (2009). Nautella italica gen. nov., sp. nov., isolated from a marine electroactive biofilm. International Journal of Systematic and Evolutionary Microbiology 59: 811-817.

Verbeek, C. and van den Berg, L. E. (2010). Extrusion processing and properties of protein-based thermoplastics. Macromolecular Materials and Engineering 295: 10-21.

198

Vila-Costa, M., Simo, R., Harada, H., Gasol, J. M., Slezak, D. and Kiene, R. P. (2006). Dimethylsulfoniopropionate Uptake by Marine Phytoplankton. Science 314: 652-654.

Villain, F., Coudane, J. and Vert, M. (1995). Thermal degradation of polyethylene terephthalate: study of polymer stabilization. Polymer Degradation and Stability 49: 393-397.

Vincent, J. F. V. and Wegst, U. G. K. (2004). Design and mechanical properties of insect cuticle. Arthropod Structure & Development 33: 187-199.

Walker, T. R., Reid, K., Arnould, J. P. Y. and Croxall, J. P. (1997). Marine debris surveys at Bird Island, South Georgia 1990-1995. Marine Pollution Bulletin 34: 61-65.

Wang, J., Chen, H., Sui, T., Li, A. and Chen, D. (2009). Investigation on hydrophobicity of lotus leaf: experiment and theory. Plant Science 176: 687- 695.

Whatmore, A. M. and Reed, R. H. (1990). Determination of turgor pressure in Bacillus subtilis: A possible role for K+ in turgor regulation. Journal of General Microbiology 136: 2521-2526.

Wheeler, W. M. (1916). Ants carried in a floating log from the brazilian mainland to San Sebastian Island. Psyche 23: 180-183.

Whelan, D. R., Bambery, K. R., Heraud, P., Tobin, M. J., Diem, M., McNaughton, D. and Wood, B. R. (2011). Monitoring the reversible B to A-like transition of DNA in eukaryotic cells using Fourier transform infrared spectroscopy. Nucleic Acids Research 39: 5439-5448.

Whitehead, K. A., Colligon, J. and Verran, J. (2005). Retention of microbial cells in substratum surface features of micrometer and sub-micrometer dimensions. Colloids and Surfaces B: Biointerfaces 41: 129-138.

Whitehouse, D. J. (1994). Handbook of Surface Metrology. ed. (Institute of Physics Publishing, Bristol).

199

Xie, G., Zhang, G., Lin, F., Zhang, J., Liu, Z. and Mu, S. (2008). The fabrication of subwavelength anti-reflective nanostructures using a biotemplate. Nanotechnology 19: Art. no. 095605.

Xu, S.-Y., Zhang, H., He, P.-J. and Shao, L.-M. (2011a). Leaching behaviour of bisphenol A from municipal solid waste under landfill environment. Environmental Technology 32: 1269-1277.

Xu, X.-W., Huo, Y.-Y., Bai, X.-D., Wang, C.-S., Oren, A., Li, S.-Y. and Wu, M. (2011b). Kordiimonas lacus sp. nov., isolated from a ballast water tank and emended description of the genus Kordiimonas. International Journal of Systematic and Evolutionary Microbiology 61: 422-426.

Yamada-Onodera, K., Mukumoto, H., Katsuyaya, Y., Saiganji, A. and Tani, Y. (2001). Degradation of polyethylene by a fungus, Penicillium simplicissimum YK. Polymer Degradation and Stability 72 : 323-327.

Yan, L., Boyd, K. G. and Burgess, J. G. (2002). Surface attachment induced production of antimicrobial compounds by marine epiphytic bacteria using modified roller bottle cultivation. Marine Biotechnology 4: 356-366.

Ye, R.-R., Wang, Z.-Y., Wang, Q.-F., Yang, K. and Luo, S.-H. (2011). Synthesis of biodegradable material poly(lactic acid-co-aspartic acid) via direct melt polycondensation and its characterization. Journal of Applied Polymer Science 121: 3662-3668.

Yoon, J.-H., Kang, S.-J. and Oh, T.-K. (2008). Roseovarius aestuarii sp. nov., isolated from a tidal flat of the Yellow Sea in Korea. International Journal of Systematic and Evolutionary Microbiology 58: 1198-1202.

Yoon, J.-H., Kim, I.-G., Kang, K. H. and Park, Y.-H. (2003a). Alteromonas marina sp. nov., isolated from seawater of the East Sea in Korea. International Journal of Systematic and Evolutionary Microbiology 53: 1625-1630.

Yoon, J.-H., Lee, S.-Y., Kang, S.-J., Lee, C.-H. and Oh, T.-K. (2007). Pseudoruegeria aquimaris gen. nov., sp. nov., isolated from seawater of the East Sea in Korea.

200

International Journal of Systematic and Evolutionary Microbiology 57: 542- 547.

Yoon, J.-H., Shin, D.-Y., Kim, I.-G., Kang, K. H. and Park, Y.-H. (2003b). Marinobacter litoralis sp. nov., a moderately halophilic bacterium isolated from sea water from the East Sea in Korea. International Journal of Systematic and Evolutionary Microbiology 53: 563-568.

Yoon, J.-H., Yeo, S.-H., Oh, T.-K. and Park, Y.-H. (2004). Alteromonas litorea sp. nov., a slight halophile isolated from an intertidal sediment of the Yellow Sea in Korea. International Journal of Systematic Bacteriology 54: 1197-1201.

Yurkov, V., Stackebrandt, E., Holmes, A., Fuerst, J. A., Hugenholtz, P., Golecki, J., Gad'on, N., Gorlenko, V. M., Kompantseva, E. I. and Drews, G. (1994). Phylogenetic positions of novel aerobic, bacteriochlorophyll a-containing bacteria and description of Roseococcus thiosulfatophilus gen. nov., sp. nov., Erythromicrobium ramosum gen. nov., sp. nov., and Erythrobacter litoralis sp. nov. International Journal of Systematic Bacteriology 44: 427-434.

Zemek, J. (2010). Electron spectroscopy of corrugated solid surfaces. Analytical Sciences 26 : 177-186.

Zhang, G., Zhang, J., Xie, G., Liu, Z. and Shao, H. (2006a). Cicada wings: a stamp from nature for nanoimprint lithography. Small 2: 1440-1443.

Zhang, J., Sheng, X. and Jiang, L. (2009). The dewetting properties of lotus leaves. Langmuir 25: 1371-1376.

Zhang, J., Wang, X., Gong, J. and Gu, Z. (2004). A study on the biodegradability of polyethylene terephthalate fiber and diethylene glycol terephthalate. Journal of Applied Polymer Science 93: 1089-1096.

Zhang, L., Zhou, Z., Cheng, B., DeSimone, J. M. and Samulski, E. T. (2006b). Superhydrophobic behavior of a perfluoropolyether lotus-leaf-like topography. Langmuir 22: 8576-8580.

201

Zhao, B., Wang, H., Li, R. and Mao, X. (2010). Thalassospira xianhensis sp. nov., a polycyclic aromatic hydrocarbon-degrading marine bacterium. International Journal of Systematic and Evolutionary Microbiology 60: 1125-1129.

Zheng, Y., Yanful, E. K. and Bassi, A. S. (2005). A review of plastic waste biodegradation. Critical Reviews in Biotechnology 25: 243-250.

Zhou, Q., Gao, Y. and Xie, G. (2011). Determination of bisphenol A, 4-n-nonylphenol, and 4 -tert-octylphenol by temperature-controlled ionic liquid dispersive liquid- phase microextraction combined with high performance liquid chromatography- fluorescence detector. Talanta 85: 1598-1602.

Zhu, W., Zhou, W., Li, C., Xiao, Y., Zhang, D., Guan, G. and Wang, D. (2011). Synthesis, characterization and degradation of novel biodegradable poly(butylene-co-hexamethylene carbonate) copolycarbonates. Journal of Macromolecular Science, Part A: Pure and Applied Chemistry 48: 583-594.

Zhu, Z. and Kelley, M. J. (2005). IR spectroscopic investigation of the effect of deep UV irradiation on PET films. Polymer 46: 8883-8891.

202