Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
FACULTY OF SCIENCE UNIVERSITY OF COPENHAGEN PhD thesis Dawei Ren Synergistic Interactions in Multispecies Biofilms Academic advisor: Søren Johannes Sørensen & Mette Burmølle Submitted: 07/02/2014 1 Synergistic Interactions in Multispecies Biofilms Ph.D. Thesis Dawei Ren Supervisors: Professor: Søren Johannes Sørensen Associate professor: Mette Burmølle Section for Microbiology, Department of Biology Faculty of Science University of Copenhagen Denmark February, 2014 This thesis has been submitted to the PhD School of The Faculty of Science, University of Copenhagen 2 Cover picture Illustration of interspecific interactions in multispecies biofilms (Burmølle et al., 2013). Top left: co-metabolism or niche generation Bottom left: coaggregation Top right: horizontal gene transfer (HGT; conjugation) Bottom right: quorum sensing (QS) Burmølle M, Ren D, Bjarnsholt T, Sørensen SJ (2013) Interactions in multispecies biofilms: do they actually matter? Trends in microbiology. doi: 10.1016/j.tim.2013.12.004 3 Preface This thesis embraces all the efforts that I have put into exploring synergistic interactions in multispecies biofilms during the last three years as a PhD student in Molecular Microbial Ecology (MME) group, Department of Biology, University of Copenhagen. This study was funded by Danish Council for Independent Research and China Scholarship Council. First and foremost, I would like to express my deepest appreciation to my supervisors Prof. Søren J. Sørensen and Associate Prof. Mette Burmølle. They always give me invaluable guidance, priceless advice and generous support. Thank you, Søren, for your optimism to create a highly motivated group, for your humor to make us feel relaxed and passionate in work and for your encouragement to boost my confidence in the face of difficulties. Thank you, Mette, for your patience when guiding my work, for your thoughtfulness when making an outline for my clear understanding and for your concern when helping me arrange journey and accommodation in Britain. I also wish to thank Karin Vestberg. She is the most professional, hardworking and helpful technician I have ever met. Her white hair, affable smile and enthusiasm have left a deep impression on me. Thank you, Karin, for your promptness in helping me to find the protocols and reagents for the experiment, for your carefulness in assisting me to use the new instruments and for your kindness in correcting the occasional improprieties in my lab work. I am especially grateful for the biofilm group members, including Jonas Madsen, Henriette Røder, Lea Hansen, Jakob Herschend, Wenzheng Liu and Jakob Russel. Thank you, Jonas, for your constructive comments on my manuscripts. Thank you, Lea, for your excellent work on transcriptomic analysis. Thanks to all of you for your active contribution in the biofilm group meeting that truly inspired me in work. I also want to give thanks sincerely to, Lasse Bergmark, for your great help in primers design and quantitative PCR work; Waleed Abu Al-Soud and Lars Behrendt, for your valuable suggestions in RNA extraction from biofilms; Sten Struwe and Annelise Kjøller, for your incredible kindness to Chinese students that have warmed my heart so much! Dr Jeremy S. Webb, Dr Robert P. Howlin and Caroline Duignan are appreciated for giving me the opportunity to have a wonderful experience in Southampton, Britain. I would like to especially mention Caroline Duignan for her generosity in donating time and sharing resource to guide me in the lab and take care of my life. Her genuine friendship is a precious gift to me! A big thank you to all the other MME group members: Lars Hansen, Bo Jensen, Anders Prieme, Tim Evison, Leise Riber, Stefan, Anette, Gisle, Martin Hansen, Zhuofei, Trine, Samuel, Witold, Tue, Jonas Stenbæk, Luo, Martin Mortensen, Peter, Tom, Claudia, Michael, Ines and the members who have left: Barbara, Luisa, Shanshan, Lili and Analia. You are so lovely, amazing and make my life in Denmark so enjoyable. Special thanks to all my Chinese friends, especially to Lili for her care throughout my first few months in Denmark, to Shanshan and Luo for being my listeners and sharing the joy and pain in my life here. Finally, I would like to extend my sincerest thanks to my parents. Without your unconditional support and love, these past three years have been impossible. As my best friends, you always timely enlighten me in spite of thousands of miles between us. This thesis is also dedicated to my 4 dear grandfather who passed away when I struggled with my work here. Your instruction and encouragement will never be forgotten. I appreciate the past three years spent in this beautiful country, full of laughter and tears and this will be precious treasure in my future life and engraved in my memory forever. Dawei Ren January 2014 in Copenhagen 5 Publications list 1. Ren D, Madsen JS, de la Cruz-Perera CI, Bergmark L, Sørensen SJ, Burmølle M (2013) High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCRBased Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. Microbial ecology. doi: 10.1007/s00248-013-0315-z Manuscript 1 2. Ren D, Madsen JS, Sørensen SJ, Burmølle M (2013) High prevalence of biofilm synergy among bacterial soil isolates in co-cultures indicates bacterial interspecific cooperation. Submitted to The ISME journal. Manuscript 2 3. Hansen LBS+, Ren D+, Sørensen SJ, Burmølle M (2014) Metatranscriptome analysis of multispecies biofilms indicates strain- and community- dependent changes in gene expression. In preparation. Manuscript 3 + Shared first authorship 4. Ren D, Ekelund F, Sørensen SJ, Burmølle M (2013) Effects of grazing by flagellate Neocercomonas jutlandica on mono- and multi-species biofilms. In preparation. Manuscript 4 5. de la Cruz-Perera CI, Ren D, Blanchet M, Dendooven L, Marsch R, Sørensen SJ, Burmølle M (2013) The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains. FEMS Microbiol Lett 338(1):95–100. Manuscript 5 6. Burmølle M, Ren D, Bjarnsholt T, Sørensen SJ (2013) Interactions in multispecies biofilms: do they actually matter? Trends in microbiology. doi: 10.1016/j.tim.2013.12.004 6 English summary The coexistence of hugely diverse microbes in most environments highlights the intricate interactions in microbial communities, which are central to their properties, such as productivity, stability and the resilience to disturbance. Biofilm, in environmental habitats, is such a spatially structured aggregation consisting of multiple species of bacteria whose function relies on a complex web of cooperative and/or competitive interactions between community members, indicating that research in “whole-entity” should not be based on the assembled results from “mono pieces”. As one of the best multispecies biofilm models, oral microbial community, also known as “dental plaque” is thoroughly investigated as a focal point to describe the interspecies interactions [1]. However, owing to the lack of a reliable high throughput and quantitative approach for exploring the interplay between multiple bacterial species, the study to elucidate the impact of interaction networks on the multispecies biofilms in natural ecosystems, especially in soil, is still at an early stage. The diverse patterns of interactions within the mixed communities as well as the predatorprey relationship between protozoa and biofilm are summarized in Sections 1, 2 and 3 of this thesis, where the state-of-the-art techniques developed to exploit such interactions, including precisely quantifying the numbers of individual species by quantitative PCR (qPCR) and monitoring gene expression changes during interactions by transcriptomic analysis are also presented. Due to the poor reproducibility of most biofilm quantification assays, the first part of my work is to develop a rapid, reproducible and sensitive approach for quantitative screening of biofilm formation by bacteria when cultivated as mono- and multispecies biofilms, followed by species specific qPCR based on SYBR Green I fluorescence to measure the relative proportion of individual species in mixed-species biofilms. The reported approach was described in Manuscript 1 which can be used as a standard procedure for evaluating interspecies interactions in defined microbial communities. By use of this valuable tool, a more than 3-fold increase in biofilm formation and dominance of Xanthomonas retroflexus and Paenibacillus amylolyticus over the other two species Stenotrophomonas rhizophila and Microbacterium oxydans were demonstrated, indicating the strong synergistic interactions in this four-species biofilm model community. Manuscript 2 presents the further application of this developed approach on evaluating the synergistic/antagonistic interactions in multispecies biofilms composed of seven soil isolates. 63% of the four-species biofilms were found to interact synergistically, indicating a prevalence of synergistic interaction in biofilm formation among these strains. Hereafter, the population dynamics in a multispecies biofilm composed of Stenotrophomonas rhizophila, Xanthomonas retroflexus, Microbacterium oxydans and Paenibacillus amylolyticus, was assessed using qPCRs with species specific primers. Despite of the high prevalence of X. retroflexus (> 97% of total biofilm cell number), the presence of the three other strains was indispensable for the strong synergism that occurs in this mixed-species biofilm. The dramatically increased cell numbers of each strain at 24 h proved all the individual strains gained benefits in the multispecies biofilms compared with in monospecies biofilms, that is, they would rather cooperate than compete with each other. 7 The significant synergistic interaction observed in the biofilm consisting of four soil bacteria make this consortium a powerful model to study development and interactions in multispecies biofilms. In Manuscript 3, the gene expression profile of Xanthomonas retroflexus in a single-species biofilm was compared to its expression profiles in dual-species biofilms with Stenotrophomonas rhizophila, Microbacterium oxydans or Paenibacillus amylolyticus as well as in a four-species biofilm. The strongest change in expression profile was observed in the dual-species biofilms of X. retroflexus and P. amylolyticus, while a distinct expression pattern (non-linear response) was detected in the four-species biofilm, indicating the significant effect of interspecies interactions on gene expression. This is consistent with the results presented in manuscript 2 where each species was demonstrated to be indispensable for the synergistic interactions in the biofilm formation. 70 genes were found differentially expressed when co-culturing X. retroflexus with other species, which include genes involved in membrane bound efflux system and MazE/MazF toxin-antitoxin system, suggesting the enhanced resistance of multispecies biofilms. Despite of the widespread existence of biofilms and protozoa in nature, the predator-prey relation between biofilms and protozoa is still poorly studied. Moreover, this relationship could be affected by interspecies interactions within multispecies biofilms. The study presented by Manuscript 4 was to test whether these interactions in the developed multispecies biofilm model are involved in the defense mechanism of bacterial biofilms against protozoan grazing. The presence of the flagellate Neocercomonas jutlandica was shown to increase or reduce the bacterial abundance in biofilms, depending on the co-cultured bacterial prey, which suggests the grazing ability is closely related with the predator-prey interactions, whereas, the synergistic interactions in the multispecies biofilm model did not confer more protection against predation compared with single-species X. retroflexus biofilm. The same ratio of cell numbers between three species regardless of protozoan grazing suggests they were spatially arranged in integrated communities in multispecies biofilm. However, these conclusions are based on the assumption that this flagellate predator prefers surface attached cells which needs to be confirmed by further studies. Horizontal gene transfer by conjugation occurs more efficiently in biofilms. The connection between plasmid host range and composition of the recipient community was investigated in Manuscript 5 by comparing plasmid permissiveness in single populations and in a microbial community composed of 15 soil strains. By use of flow cytometry (FCM) and 16S rRNA gene sequencing, the IncP1 plasmid, pKJK10, was found only to transfer from Pseudomonas putida to Stenotrophomonas rhizophila in a diparental mating. However, when hosted by Escherichia coli, transfer of this plasmid occurred only in the mixed community, with Ochrobactrum rhizosphaerae as the dominating plasmid recipient. This study demonstrates that the plasmid host range can be greatly affected by the surrounding bacterial community. This needs to be taken into account as many antibiotic resistance and virulence determinants are plasmid-encoded, which can spread further and raise antibiotic-resistant bacteria in soil. 8 Dansk resumé I de meget diverse mikrobielle samfund i naturen er der nogle yderst komplekse interaktioner, som er meget centrale med hensyn til både produktivitet, stabilitet og modstandsdygtighed overfor forandringer i miljøet. Biofilm i naturlige habitater er en rumlig samling af bakterier bestående af flere arter, hvis funktion er afhængig af et indviklet net af kooperative og / eller konkurrencemæssige relationer mellem dem, hvilket indikerer, at forskning i en biofilm ikke bør være baseret på resultater fra enkelt bakterier. Som en af de bedste biofilm modeller med flere arter er det orale mikrobielle samfund, også kendt som " plak ," der er grundigt undersøgt med fokus på at beskrive interspecies interaktioner. Men manglen på nyere forskning med kvantitativ tilgang til at udforske samspillet mellem flere bakteriearter for at belyse konsekvenserne af interaktionen på mange arts biofilm i naturlige økosystemer, især i jord, gør at forskningen stadig er på et tidligt stadium. Forskellige mønstre af interaktioner i de blandede samfund såvel som predator-prey forhold mellem protozoer og biofilm er sammenfattet i afsnit 1, 2 og 3 i denne afhandling, hvor state-of- the- art teknikker udviklet til at udnytte sådanne interaktioner, herunder præcist at kvantificere antallet af enkelte arter ved kvantitativ PCR (qPCR) og overvågning af genekspression ændringer i interaktioner med transkriptomic analysis, også er præsenteret. På grund af den ringe reproducerbarhed af de fleste biofilm kvantificerings analyser er den første del af mit arbejde gået ud på at udvikle en hurtig, reproducerbar og følsom metode til kvantitativ screening af biofilm dannelse af bakterier, når de dyrkes som mono-og flere arts biofilm, efterfulgt af artsspecifikke qPCR baseret på SYBR Green I fluorescens til at måle den relative andel af de enkelte arter i blandede arts biofilm. Resultaterne er beskrevet i Manuscript 1, og kan bruges som standard procedure for evaluering af interspecies interaktioner i definerede mikrobielle samfund. Ved brug af dette værdifulde værktøj, blev der påvist en mere end 3-fold stigning i biofilm dannelse og dominans af Xanthomonas retroflexus og Paenibacillus amylolyticus over de to andre arter Stenotrophomonas rhizophila og Microbacterium oxydans, med angivelse af de stærke synergistiske interaktioner i dette fire-arts biofilm model samfund. Manuskript 2 præsenterer den videre anvendelse af den ovenfor beskrevne metode til at evaluere de synergistiske / antagonistiske interaktioner i flere arts biofilm bestående af syv jord isolater. 63% af de fire arts biofilm interagerede synergistisk , hvilket indikerer en prævalens for synergistisk interaktion i biofilmdannelse blandt disse stammer . Populationsdynamik i en flere arts biofilm bestående af Stenotrophomonas rhizophila, Xanthomonas retroflexus , Microbacterium oxydans og Paenibacillus amylolyticus blev estimeret ved hjælp qPCR med artsspecifikke primere. På trods af den høje forekomst af X. retroflexus (> 97% af det samlede antal biofilm bakterier) , var tilstedeværelsen af de tre andre stammer essentiel for den kraftige synergi, der opstår i denne blandede arts biofilm . De dramatisk øgede celletal for hver stamme efter 24 timer viste, at alle de individuelle stammer opnåede fordele i flere arts biofilm sammenlignet med i monospecies biofilm, det vil sige at de hellere ville samarbejde end konkurrere med hinanden . Den signifikante synergistiske interaktion, der blev observeret i en biofilm bestående af fire jordbakterier gør dette konsortium til en vigtig model til at studere udviklingen og interaktioner i 9 flere arts biofilm. I Manuscript 3 blev genekspression profilen for Xanthomonas retroflexus i en enkelt arts biofilm sammenlignet med dens ekspressions profil i dual- arts biofilm med Stenotrophomonas rhizophila , Microbacterium oxydans eller Paenibacillus amylolyticus såvel som i en fire- arts biofilm . Den største ændring i ekspressions profilen blev observeret i dual- arts biofilm med X. retroflexus og P. amylolyticus , mens et klart ekspressionsmønster ( ikke-lineær respons) blev detekteret i fire arts biofilm , hvilket indikerer en signifikant effekt af interspecies interaktioner på genekspression . Dette er i overensstemmelse med de resultater, der præsenteres i manuskript 2, hvor hver art viste sig at være essentiel for de synergistiske interaktioner i biofilmdannelse. Der blev fundet 70 gener, som blev udtrykt når X. retroflexus blev dyrket med andre arter, der omfatter gener involveret i et membranbundet efflukssystem og i Maze / MazF toksin - antitoxin systemet, hvilket tyder på forbedret resistens i flere arts biofilm . På trods af den udbredte forekomst af biofilm og protozoer i naturen er predator-.prey relationen mellem biofilm og protozoer stadig dårligt undersøgt. Endvidere kan dette forhold blive påvirket af interspecies interaktioner indenfor flere arts biofilm. Undersøgelsen præsenteret i Manuscript 4 var at teste, om disse interaktioner i den udviklede flere arts biofilm model er involveret i en forsvarsmekanisme hos den bakterielle biofilm mod protozo græsning. Tilstedeværelsen af flagellaten Neocercomonas jutlandica viste sig at øge eller reducere den bakterielle forekomst i biofilm, afhængig af bakterien, hvilket tyder på, at græsningsevnen er nært beslægtet med predatorprey interaktioner; de synergiske interaktioner i flere arts biofilm giver ikke mere beskyttelse mod prædation end enkelt - arts X. retroflexus biofilm. Det samme forhold af antal celler mellem de tre arter uanset protozo græsning antyder, at de var rumligt placeret i et integreret samfund i flere arts biofilm. Men disse konklusioner er baseret på den antagelse, at den benyttede flagellat foretrækker overflade vedhæftede celler; dette forhold skal bekræftes af yderligere undersøgelser. Horisontal genoverførsel ved konjugering forekommer mere effektivt i biofilm. Forbindelsen mellem plasmid host range og sammensætningen af recipient samfundet blev undersøgt i Manuscript 5 ved at sammenligne plasmid tolerance i populationer af en enkelt art og i et mikrobielt samfund bestående af 15 jord-stammer. Ved brug af flowcytometri (FCM) og 16S rRNA-gen sekventering, blev IncP1 plasmidet pKJK10 kun vist at blive overført fra Pseudomonas putida til Stenotrophomonas rhizophila i en diparental mating. Når Escherichia coli var vært skete overførsel af dette plasmid kun i det blandede samfund, med Ochrobactrum rhizosphaerae som dominerende plasmid recipient. Denne undersøgelse viser, at plasmidets værtsspektrum kan blive kraftigt påvirket af den omgivende bakterielle samfund. Dette skal der tages hensyn til, da mange antibiotikaresistens og virulensdeterminanter er plasmidkodede, og kan sprede sig yderligere og øge antibiotikaresistente bakterier i jord. 10 11 Table of Contents 1 Interactions in multispecies biofilms .......................................................................................... 14 1.1 Biofilm................................................................................................................................. 14 1.2 Multispecies biofilms in soil ............................................................................................... 14 1.3 Interactions in multispecies biofilms ................................................................................... 16 1.3.1 Coaggregation, cross-species protection and co-metabolism ...................................... 17 1.3.2 Chemical signaling systems ......................................................................................... 18 1.3.3 Lateral gene transfer..................................................................................................... 20 1.3.4 Synergism or antagonism/Cooperation or competition ............................................... 20 2 How to study multispecies biofilms? .......................................................................................... 24 2.1 In vitro biofilm models ........................................................................................................ 24 2.2 Quantitative PCR................................................................................................................. 25 2.3 Transcriptomics ................................................................................................................... 27 3 Biofilms and protozoa................................................................................................................. 31 3.1 Protozoa ............................................................................................................................... 31 3.2 Biofilms- the response of cell consortia to protozoan grazing ............................................ 31 3.3 Protozoa and biofilms- reservoirs of pathogenic bacteria ................................................... 34 4 Where are we going with biofilms? - In the context of microbial ecology ................................ 35 5 References .................................................................................................................................. 38 6 Manuscripts................................................................................................................................. 50 Manuscript 1 Manuscript 2 Manuscript 3 Manuscript 4 Manuscript 5 12 13 1 Interactions in multispecies biofilms 1.1 Biofilm Already in the late 1600s, van Leeuwenhoek had observed biofilm in the plaque on his own teeth. The first appearance of “biofilm” theory can be traced to 1987 when Costerton et al. described biofilm as adherent population consisting of single cells and microcolonies of sister cells all embedded in a highly hydrated, predominantly anionic matrix of bacterial exopolymers and trapped extraneous macromolecules [2]. Over the course of the past 25 years, this concept has evolved to include not only the irreversible cell attachment but also physiological attributes including altered growth rate and gene transcription [3]. Biofilms, prevalent on the most inert or living surfaces [4], are the dominant communities on planet earth. It is estimated that 99% of bacteria in nature exist in biofilms, while biofilms account for more than 65% of hospital infections [5, 6]. And the same kind of bacteria are known to show profoundly different characteristics when they are in a biofilm compared with in planktonic cultures. The high resistance to harsh conditions, including pollutants [7] , desiccation [8], protozoan grazing [9], antimicrobial agents [10] in nature and host defenses [11] in chronic infections, is some/one of the most important features of biofilms. This resilience can be explained by several mechanisms which are summarized in Table 1. Moreover, the heterogeneity within biofilms offers the possibility of the joint action of these multiple resistance mechanisms in a single community. Despite the intensive research in batch cultures, extrapolations of these results in bulk to that in biofilm cells seems unwise due to the biofilm-specific physiological properties. Table 1 Summary of resistance mechanisms of biofilms. Biofilms Mechanisms Escherichia coli Pseudomonas aeruginosa Staphylococcus epidermidis Reduced permeability of matrix Pseudomonas aeruginosa Escherichia coli Escherichia coli Salmonella sp. Pseudomonas aeruginosa Pseudomonas aeruginosa Serratia marcescens Pseudomonas aeruginosa Slow growth Adverse factors β-lactam antibiotic desiccation metal host defenses References Ciprofloxacin [16] Persister cells Ciprofloxacin host defenses protozoa Ofloxacin, metal Quorum sensing protozoa Biofilm-specific phenotype [12] [13, 14] [15] [17] [18] [19] [20] [14] [21] [19] 1.2 Multispecies biofilms in soil Biofilms in nature habitats are complex communities where various types of microorganisms (e.g. bacteria, archaea, protozoa, fungi and algae) are held together and protected by self-excreted extracellular polymeric substance (EPS) [22]. For example, soil is such a potential environment where the population density and diversity of bacteria may be up to 109 cells/g soil and 106 14 species/g soil, respectively [23, 24]. Because of the spatial variability in nutrient concentration, microbial cells are not uniformly distributed through the soil [25]. Bacteria living close to the nutrient sources i.e., plant roots or decaying organic matter, are prone to attach various available surfaces (e.g. roots, litter or soil particles) and develop into multispecies biofilms. By organizing into a biofilm community, bacteria could gain highly resilience to adverse soil conditions. Water is by far the largest component of the biofilm matrix which can account up to 97%, whereas the remaining are 2-5% microbial cells, 3-6% EPS (polysaccharides, proteins, nucleic acids and lipids) and ions [26]. The highly hydrated matrix could therefore buffer the biofilm cells against desiccation stress which is an important challenge met by soil bacteria. Chang et al. provided the direct evidence that alginate production by Pseudomonas putida contributed to a hydrated microenvironment which protected residents from water-limiting stresses [27]. Moreover, biofilm has a great capacity for heavy metal biosorption and toxic compound degradation which has a significant impact on bioremediation [28, 29]. Additionally, the widespread exposure to antibiotics makes biofilm formation more favorable in soil. The results from Walker et al. suggested that upon root colonization, Pseudomonas aeruginosa gained resistance against root-secreted antibiotics by forming a biofilm [30]. Apart from these advantages, biofilms can also function as protective barriers against protozoan grazing which is a major mortality factor faced by bacteria in the soil environment [31]. Section 3 (Biofilms and protozoa) will be devoted to a coherent introduction of the relationship between biofilms and protozoa. These improved biofilm-associated fitnesses mentioned above suggest that the preferred mode of bacterial growth is in a biofilm. By being encased in the recalcitrant matrix, the bacteria grow in a relatively stable environment called microbial homeostasis [32], reflected not by the characteristics of resident individuals but by the balance imposed by the numerous microbial interactions, including examples of quorum sensing (QS) and horizontal gene transfer (HGT). By means of quorum sensing, the sessile cells in the biofilms can “talk” to each other. Due to the increased population density and constrained diffusion, the quorum sensing molecules are concentrated. Once reaching a threshold level, these quorum sensing molecules modulate the transcription of certain genes and trigger phenotypic changes, including swarming motility, biofilm formation and the production of virulence factors [33-35]. This issue will be elaborated further in the next part of Section 1.The dramatically increased horizontal transfer of plasmid-borne antibiotic resistance determinants was observed by Savage et al. in the Staphylococcus aureus biofilm [36]. Since many antibiotic resistance determinants are plasmid-encoded, this further spread of antibiotic resistance genes among bacteria allows us to conclude that soil represents a reservoir of antibiotic resistance genes [37] which probably increases the current arsenal of antibiotic resistance mechanisms in pathogens when gene transfer occurring from soil bacteria to pathogenic bacteria. This was confirmed by Forsberg et al. [38] with the finding that multidrug-resistant soil bacteria, containing resistance cassettes against five classes of antibiotics, have perfect nucleotide identity to genes from diverse human pathogens. Therefore, the enhanced efficiency of gene transfer in biofilms has a profound impact on the pathogenesis, persistence and hence the treatment of human disease. 15 Despite of the notorious resistance to various common antibiotics and host defenses, soil biofilms can also be exploited for their diverse application in agriculture. The biofilmed inocula can be used as biofertilisers (BFBF) to promote and stimulate plant growth as well as aid in disease control [39]. Furthermore, in the biofilm formed by bacteria and fungi, another natural inhabitant in soil, there is often generated synergistic interactions with possible consequences of a significant increase in nutrient acquisition and uptake of phosphorus, nitrogen and metal ion [40]. The fungal-bacterial biofilms (Penicillium frequentans and Bacillus mycoides) resulted in a 14-fold increase in the biodegradability of degradable polyethylene by P. frequentans [41]. The co-culture of Pseudomonas fluorescens and a mushroom fungus (Pleurotus ostreatus) increased the endophyte colonization of tomato by 1000% compared to inoculation with P. fluorescens alone [42]. A bradyrhizobial-fungal biofilm showed nitrogenase activity, whereas the bradyrhizobial strain alone did not, which improved the shoot and root growth, nodulation and nitrogen accumulation of soybean and directly contributed to soil nitrogen fertility in the long term [43]. Additionally, anaerobic degradation of complex organic matter into methane and carbon dioxide requires the progressive action of numerous species of microorganisms [44]. Biofilms can provide such an ideal environment for the interaction of these metabolically cooperative organisms, owing to their highlyorganized structure enhancing the nutrient availability as well as removal of potentially toxic metabolites. In summary, as the dominant growth form for bacteria in soil, mix-species biofilms play an essential role in maintaining the ecological balance, whereas from an evolutionary perspective, this role is further strengthened by the selective pressures which favor bacteria capable of forming biofilms in versatile soil environment. What is real is rational -- what is rational is real. --Hegel, 1821, Basic Outline of the Philosophy of Right 1.3 Interactions in multispecies biofilms Different species, exhibiting different growth and survival properties, encased in an extracellular polymeric network could lead to the spatial and functional heterogeneity within biofilms. Even in a single-species biofilm, the physical, chemical (e.g. gradients of nutrients, waste products and signaling compounds) and biological (distinct metabolic pathways and stress responses) heterogeneity can develop [45]. In environmental habitats, diverse bacteria and in many cases fungi, algae and protozoan, do not live independently in their local microenvironments. The interactions among these microorganisms and with the external environment critically influence the development, structure and function of the biofilm and conversely, the spatial heterogeneity and biodiversity clearly have a dramatic effect on the communication between different biofilm components, allowing for the development of a complex multispecies community (Figure 1). 16 Figure 1 Communication in a natural multispecies biofilm [46]. Biofilm communities are shaped by various interactions between microbial species, including (1) competition between bacteria populations and their neighbours such as fungi (A), bacteria (B) and microalgae (C), (2) quorum sensing derived from clonal growth in microcolonies (D) which may induce the protection against protozoa (E) and (3) interactions with second colonizers such as macroalgae spores (F) and invertebrate larvae (G). Furthermore, the bacteria-host interaction should also be taken into consideration in medical environment. 1.3.1 Coaggregation, cross-species protection and co-metabolism Various types of interactions within biofilms include coaggregation, cross-species protection, cometabolism, quorum sensing (QS) and genetic exchange. Coaggregation, defined as the specific cell to cell recognition among genetically distinct bacteria [47], is vital for both biofilm formation and the existence of certain bacterial species. This has been well described in numerous studies of the oral biofilms. An example is provided by Fusobacterium nucleatum which can coaggreate with species that can not bind to each other thus serves as bridge between early and late colonizers in a sequential process [48]. Another example is the ability of Escherichia coli O157:H7 to adhere and persist in a capillary flow cell which requires the colonizing partner Pseudomonas aeruginosa PAO1. Development of E. coli microcolonies occurred only along the outer 200 μm edge of the flow cell after P. aeruginosa migrating away into the center of the flow path, indicating that the conditioned surface by the later may facilitate attachment of the former [49]. In addition, coaggregation confers cross-species protection of anaerobic species from oxygen and of susceptible species from antimicrobials [50, 51]. Cross-species protection is also likely derived from extracellular polymer and this was demonstrated by the study from the mixed fungal-bacterial biofilm where the extracellular polymer produced by Staphylococcus epidermidis RP62A could inhibit fluconazole penetration and conversely, the presence of Candida albicans in this biofilm appeared to protect the slime-negative Staphylococcus against vancomycin [52]. Co-metabolism, where one species utilizes a metabolite produced by a neighboring species, presumably plays a major role in biodegradation of organic molecules and is advantageous to the entire microbial community. Boonchan et al. proved that inoculation of fungal-bacteria co-cultures 17 resulted in significantly improved co-metabolic degradation of polycyclic aromatic hydrocarbons (PAHs) in soil [53]. It has also been demonstrated that the acceleration of the remediation of chlorophenol- and phenol-contaminated groundwater by a sequencing batch biofilm reactor was probably due to the co-metabolism [54]. The efficient degradation when multiple species are present can be derived from the optimized substrate availability when growing attached to a surface and the close proximity of enzymes involved in degradation that may be retained in the biofilm matrix [55, 56]. Additionally, metabolic communications were also reported between bacteria within the oral cavity. One of them is the metabolic interaction through arginine between two oral bacteria Actinomyces naeslundii and Streptococcus gordonii, where S. gordonii genes involved in arginine biosynthesis and transport were induced when coaggregated with A. naeslundii, otherwise S. gordonii could not grow without sufficient arginine [57]. 1.3.2 Chemical signaling systems A cell-to-cell signaling mechanism known as quorum sensing (QS) has been shown in many studies to play crucial roles in biofilm development as mentioned in Section 1.2. And due to the social behaviors of bacteria, the term “sociomicrobiology” was introduced by Parsek MR et al. in 2005 to vividly describe the inextricable link between biofilms and quorum sensing [58]. It has been shown that QS can be induced by a few thousand bacteria, which size is analogous to the number of bacteria found in biofilm microcolonies [59]. Many bacterial behaviors are regulated by chemical autoinducer molecules that are produced and used by bacteria to sense one another. Bacteria can communicate both intraspecifically and interspecifically via autoinducers which alter gene expression and allow bacteria to respond coordinately to their environments, in a manner that is comparable to behavior and signaling in higher organisms. In gram-negative bacteria, acylated homoserine lactones (AHLs) are the most intensively investigated signal molecules and have been well described in Pseudomonas aeruginosa. There is also report that two different chemical languages: N-acyl homoserine lactones (AHLs) and cis-2-unsaturated fatty acids were utilized to control biofilm formation and virulence in Burkholderia cepacia complex (Bcc) [33]. Despite of the species specificity of AHL systems, the cross-species talk was reported in a biofilm composed of cystic fibrosis-associated P. aeruginosa and B. cepacia [60, 61]. In addition, Bacillus sp. and Variovorax paradoxus were reported to degrade AHLs and interfere with quorum sensing of other species [62, 63]. In gram-positive bacteria, such as Staphylococcus aureus, peptides operate generally by binding to receptors on the cell surface rather than diffusing back into the cell like AHLs [64, 65]. The signaling communication in multispecies biofilms are mainly mediated by autoinducer 2 (AI-2), which is synthesized by the enzyme LuxS and found in both gram-negative and –positive bacteria [1, 66]. AI-2 has been shown to promote the biofilm formation of two oral bacteria Actinomyces naeslundii T14V and Streptococcus oralis 34. Whereas, AI-2 of Fusobacterium nucleatum was reported to differentially regulate biofilm growth of two oral streptococci by producing a stimulatory effect on Streptococcus gordonii and an inhibitory effect on S. oralis [67]. Two redundant quorum sensing systems were verified in Vibrio harveyi, with AHL for intraspecies communication and AI-2 for interspecies cell–cell signaling [68]. In spite of the apparent universality of luxS (present in more than 40 bacterial species), the difficulties of obtaining purified AI-2 from species expressing AI-2 activity raise the doubts that whether AI-2 is a universal 18 signal or just may be a byproduct of the activated methyl cycle (AMC) [69]. Recently, Santiago‐ Rodriguez et al. reported luxS sequences in 25- to 40-million-year-old bacteria, such as Bacillus schakletonii and B. aryabhattai, two extant bacterial species that had not been previously reported as carrying luxS [21]. This in turn raises new questions on the specific role of luxS in ancient microorganisms and whether it is involved in the regulation of metabolism in amber bacteria. Another quorum sensing signal-diffusible signal factor (DSF), identified in Burkholderia cenocepacia [70] and Pseudomonas aeruginosa [71], was reported recently to be involved in interspecies communications by altering biofilm formation, architecture and resistance to antibiotic [72-74]. Although the underlying mechanism of DSF in mixed communities remains to be elucidated, this signal may play a crucial role in mediating cell-to-cell interactions in parallel with AI-2 and AHL signals, owing to its widespread existence in various species and niches. Despite the significant advances regarding to bacterial quorum sensing and group behaviors mentioned here, expounding the functional consequences of QS in multispecies biofilms is still a challenge. Although quorum sensing (QS) can be considered as diffusion sensing (DS) as QS induction or repression is based on the interaction between the diffused signal and the cognate receptor [75], QS enables bacteria to coordinate their behaviors for group benefit while DS depends on individual fitness benefits [76]. These conflicting concepts are unified by an alternate hypothesis efficiency sensing (ES), which suggests the role of autoinducers relies on both cell density and spatial distribution and thus is favored by both group and individual benefits [77]. Future research towards uncovering the genetic network induced by QS that deterministically controls biofilm adaption will undoubtedly provide new insights into biofilm manipulation. Another chemical signaling system 3', 5'-cyclic diguanylic acid (c-di-GMP), as a ubiquitous secondary messenger, is found in diverse bacteria. The levels of c-di-GMP are mediated in the cell by diguanylate cyclase (DGC) activity involved in c-di-GMP synthesis and phosphodiesterases (PDE) activity involved in c-di-GMP degradation. The well characterized example is in Vibrio cholera where 62 genes are predicated to encode proteins capable of producing or degrading c-diGMP and influence many phenotypes including motility, biofilm formation and virulence [78, 79]. Moreover, c-di-GMP was reported to reciprocally control biofilm formation and virulence with quorum sensing in V. cholera. These two signaling function antagonistically to regulate biofilms and synergistically to repress virulence factor expression, which are fundamental for V. cholerae survival ex vivo and in vivo, respectively [80]. The increased production of c-di-GMP could enhance biofilm formation and decrease swarming motility were also observed in Pseudomonas aeruginosa [81]. George O’Toole elaborated upon how c-di-GMP influenced biofilm formation via sensing environmental input, phosphate levels, in Pseudomonas fluorescens. This involved two proteins, LapA and LapD. Whereas LapA was required for stable surface attachment and biofilm formation, LapD served as a unique c-di-GMP effector protein that utilized an inside-out signaling to regulate LapA localization and thus surface commitment [82]. Despite the diverse components and molecular processes involved in biofilm formation throughout the bacterial kingdom, c-di-GMP signaling seems to represent a common principle, which suggests that the enzymes that controlling cellular c-di-GMP levels may be promising targets for anti-biofilm drugs. 19 1.3.3 Lateral gene transfer Lateral gene transfer (LGT), also termed horizontal gene transfer (HGT), refers to the gene exchange among bacteria cells in a manner other than traditional reproduction, which can occur among conspecific strains [83] and strains in different species in biofilms [84]. There are two LGT mechanisms: transformation and conjugation. Transformation - the uptake of free DNA from the environment by a bacterial cell- requires exogenous DNA which can be easy to meet as a result of the large amount of extracellular DNA in biofilms. The other mechanism, conjugation, occurs when there is direct cell-to-cell contact or a bridge-like connection and transfers small pieces of DNA, usually plasmids which often carry virulence and antimicrobial resistance genes. Efficient transformation and conjugation in microbial consortium have been found in both natural and artificial environments. For example, lateral gene transfer of AHL synthase gene could facilitate cross talk between Burkholderia spp. and Pseudomonas spp. [85], whereas lateral gene transfer of ring-hydroxylating-dioxygenase (RHD) gene may improve aromatics’degradation by spreading the gene among different species [86]. Burmølle et al. reported a conjugative plasmid pOLA52, which confers resistance to olaquindox and other antimicrobial agents through a multidrug efflux pump, can also promote biofilm formation in Escherichia coli [87]. Also, there is example that transfer efficiency of plasmids in Pseudomonas putida biofilm depended on the type of antibiotics, suggesting biofilm bacteria may “sense” antibiotics to which they are resistant and enhance the spread of that resistance [88]. Overall, efficient gene transfer is both the cause and consequence of biofilm development. On the one hand, LGT is facilitated within biofilms as a result of the presence of extracellular DNA, close spatial juxtaposition of bacterial cells and stable habitat provided by EPS matrices [89]. On the other hand, both DNA transfer processes seem to have positive effects on biofilms, such as enhanced resistance against predators, toxins and antibiotic factors and the stabilization of biofilm structure mediated by conjugative pili which may act as cell adhesins in hydrodynamic biofilm systems [90]. 1.3.4 Synergism or antagonism / Cooperation or competition The interactions responsible for synergism in biofilms as exemplified above, often operate in concert and have been demonstrated to strengthen the protective effects of biofilms when multiple species are present compared with single species communities. This was verified by a substantial increase in the chlorine tolerance of a multispecies biofilm from drinking water, regarding to the planktonic cultures and monospecies biofilms [91]. Likewise, it was recently reported that a reproducible mixed-species biofilm comprising Pseudomonas aeruginosa, Pseudomonas protegens and Klebsiella pneumonia was more resistant to the antimicrobials sodium dodecyl sulfate and tobramycin than the single-species biofilms. Moreover, such community level resilience was found to come from the protection offered by the resistant species rather than selection for the resistant species, suggesting that the community-level interactions, such as the sharing of public goods, are unique to the structured biofilm community [92]. Nevertheless, some antagonistic interactions between bacteria also have been documented in the dental biofilm [93] as well as in marine [94] and soil environments [95]. Microbial antagonism can be caused by inhibition of microbial growth by diffusible antibiotics, toxins or biosurfactants [96], competition for colonization sites and nutrients [97] and degradation of quorum sensing molecules [98]. Diverse physical interactions between 20 bacteria and fungi have been associated with reduced fungal viability due to the antifungal molecules secreted by bacteria into the local environment. An example is Acinetobacter baumanniiCandida albicans in chronic infections. It was shown that A. baumannii could inhibit several important virulence determinants of C. albicans, including hyphae and biofilm formation via polymicrobial infection and conversely, the viability of A. baumannii is reduced when C. albicans cells adopt the quorum mode in a biofilm environment [99]. Thus, exploiting the mechanisms used by competing microorganisms could potentially contribute to combating detrimental biofilms in medical, industrial and natural environments. According to West et al. [100], interactions can be roughly classified into cooperation or competition, based on the effect of the microbial social behavior on each population in a binary system (actor and recipient). When recipient benefits from the presence of actor, the interaction is termed cooperation, which can be further subcategorized into mutualism (beneficial for both actor and recipient) and altruism (beneficial for recipient but costly to actor). On the contrary, when recipient is negatively affected, the interaction is identified as competition which is grouped into selfishness (beneficial for actor) or spite (negative effect on both recipient and actor). In addition, two other closely related terms are also frequently used to describe the social behavior in biofilms, i.e. synergism and antagonism. Synergism has been defined as the cooperative action of two or more organisms where the effect of their collective effort is greater than it would be by their individual effect [101]. In contrast, antagonism is the relationship in which one species of an organism is inhibited or adversely affected by another species in the same environment. When these definitions are applied to soil microorganisms, synergism, also known as protocooperation, is described as a facultative phenomenon that both populations can survive on their own but the association provides some mutual benefits [102]. In this thesis, we use “synergism” to refer to the enhanced overall productivity (biomass) and fitness (resistance against protozoa grazing) of the multispecies community as a whole compared with the individual species. Competition often indicates the active competition for nutrients and space. Thus, in a narrower sense, competition can also be defined as “the injurious effect of one organism on another because of the removal of some resource of the environment” [103]. While, antagonism may be used broadly to include the competition for limited substrate, the inhibition by antibiotics or metabolites produced by another organism, or exploitation which is either predation or direct parasitism [104]. When used practically, however, these definitions are not clearly defined and may vary somewhat. For example, Foster et al. applied the evolutionarily stringent definition of cooperation [97] when analyzing the productivity of two-species mixtures grown in aquatic microcosms, that is, only the interactions that can cause the increased productivity of both species in co-cultures, can be termed as cooperation. Moreover, in spite of the usefulness of this binary system mentioned above in defining interaction, the natural communities are far more complex than expected where more than one type of interactions probably simultaneously occurs and more than two species are involved. Hence, it may seem trivial to place the interaction within a biofilm in one category while exclude others [105]. A typical example described by Hansen et al. [106] showed that the increased biomass of the dualspecies biofilm in total and of one member (Pseudomonas putida) is at the expense of another member (Acinetobacter sp.) due to intensified competition for oxygen. Therefore, both the biomass 21 and/or function of the integrated multispecies community and each individual member need to be evaluated to define whether the cooperative or competitive interactions shape the community. This is what we have done in manuscript 1 and 2 [107]. The quantitative PCR developed in our study was applied to measure the absolute cell numbers of each species in a four-species biofilm and hence identified the dominance of cooperation interaction as each member could benefit, with respect to biomass, in this multispecies biofilm compared with when they grow alone. What we know from the research focusing on the link between genetic population structure and social behavior are still very limited. And most studies support the idea that while cooperation will occur within the same genotype, competition should prevail between different genotypes [108-110]. However, this prediction is drew with the implicit assumption that there is competition for resources among these interacting genotypes which may not always be the case in nature environment, exemplified by efficient biodegradation through co-metabolism in multispecies communities mentioned above. Zhang et al. demonstrated that the switch between synergy and competition is closely related to the flow conditions. Limited resource replenishment favors competition under low-flow conditions while high flow favoring synergy by providing greater resource and making biofilm easier to shear from the surface [111]. Moreover, cooperation between different species also may prone to emerge in communities with low niche overlap and high relatedness within each member. From the social evolution point of view, cooperation is challenging to explain as cooperative phenotypes are susceptible to exploitation by rapidly growing, non-cooperative cells. However, Nadell et al. reported that this condition can be reversed if cooperative cells are segregated in space and preferentially interact with each other, indicating the cooperation may evolve readily than naively expected [112]. The proposition that biofilm promotes altruism was evidenced by an individual-based model simulations which showed the often-observed structural organization into microcolonies and shedding of single cells in biofilms are necessary for the origin and maintenance of the altruistic strategy [113]. Moreover, a recent report by [114] pointed that intraspecies variation, despite the enhanced individual fitness of the generated morphotypic variants compared with their parental strains, was reduced in the mixed species biofilms, suggesting this sacrifice of self-produced diversity in the presence of other species probably represents altruism in multispecies biofilms [115]. The “Black Queen Hypothesis (BQH)” was put forward recently to better explain how and why cooperation in a multispecies community can evolve [115]. It presents a scenario whereby a division of labor in microbial communities is likely to occur by receivers deleting costly pathways that are provided by the surrounding bacteria which unavoidably produce public resources, thus bacteria might often develop interdependent cooperative interactions. This is opposite to the “Red Queen Hypothesis” which states the evolutionary conflict between coinhabitant organisms. Therefore, BQH provides a new framework for looking at interactions in the ecological context where these organisms are evolving, which needs to be carefully considered when applying social evolution theory to microbial communities. Viewed in this way, the conflict between group-level selection and individual-level selection, which favor cooperation and competition respectively, should be assessed on a case-by-case basis. Instead of discerning which is the ‘right’ model, it should be noted, however, that different types of selection are likely involved in shaping the evolutionary adaption of biofilm, leading to different types of molecular strategies 22 involved in regulating surface associated communities according to various environmental cues [116]. Since the majority of microorganisms present in the environment remain unculturable, the diversity of complex bacterial communities is inevitably underestimated. We can not exclude that other bacteria out there might depend to a great extent on cooperation with partners, and perhaps just this is one of the reasons why we failed so far to isolate them in the laboratory [117]. This also highlights the importance of the in situ studies of microbial interactions, which is a prerequisite for deeper understanding of social behavior in multispecies biofilms and hence provide better treatment for biofilm-associated infections and exploit biofilms for beneficial applications, such as waste water treatment, N2 fixation, pollutant degradation and so on. 23 2 How to study multispecies biofilms? The study of biofilms has proliferated in the past 15 years. And it is widely acknowledged that biofilm formation has a close correlation with the species and/or intra-/interspecies communication that is present, as well as the environmental conditions involved. 2.1 In vitro biofilm models In vitro biofilm experiments were brought into being in the mid-1980s [118, 119], followed by the development of several biofilm-forming devices, including Modified robbins device [120], Microtiter plates [121], Calgary device [122], Flow cell [123] and BioFilm ring test [124], which allow to explore the adhesion ability of different microorganisms and the effect of various environmental conditions on biofilm development with tightly controlled parameters. Using 96-well microtiter plates, the effects of coaggregation and quorum sensing molecules on biofilm formation were assessed respectively [125, 126]. Although in vitro studies have the advantages in evaluating the influence of pre-determined environmental, physiological and genetic variables on biofilm formation, large amounts of reproducible data under different conditions are still missing as the insufficient versatility and the difficulty to control all the complex variables especially when a heterotrophic consortium, such as a multispecies biofilm, is studied. Considering the lack of reproducibility, possibly due to the researcher’s dependent variable, the comparison of results obtained when using different protocols and biofilm growth systems, particularly between different research groups, is problematic. This is what we hopefully are contributing to in manuscript 1 [107]. A standard procedure was developed for evaluating interspecies interactions in defined microbial communities by comparing the reproducibilities of 96 Well Cell Culture Plate and Nunc-TSP lid system (also referred to Calgary methods)-based biofilm formation. The Calgary Biofilm Device (CBD) was first described and applied by Ceri et al. [127] for high-throughput determination of antibiotic susceptibilities of bacterial biofilms. The biofilm is formed on the pegs of a modified microtiter lid in this device instead of at the bottom of a well which can avoid non-specific bacterial sedimentation that may not accurately relate to biofilm, though the difference between this “upsidedown” biofilm and the biofilm formed at the bottom has not been addressed. Apart from the broad and robust applicability for many microorganisms, CBD, moreover, seems to be more amenable for combination with microscopy, e.g. epifluorescence microscopy or confocal laser scanning microscopy (CLSM), which makes it possible to analyze the structural heterogeneity of biofilms under diverse exposure conditions [128]. Thus, by introducing microscopy into biofilm studies, many parameters such as biofilm biomass, total and active number of cells as well as associated physiological activity, extracellular matrix and overall structure can be assessed. Especially for multispecies biofilms, the spatial organization of different species is very important which plays a vital role in determining biofilm function, including antimicrobial resistance [92] and driving metabolic reactions [129]. In order to gain a more comprehensive and in-depth understanding of complex interactions that drive biofilm development, the multispecies consortia should be investigated nondestructively, in real time and in situ which also enable studies of the species that are unculturable in the laboratory. However, it is still challenging until now to mimic the microenvironment as much as possible where microorganisms live in close proximity and interact with each other, especially for soil 24 biofilms, owing to the complex structure of soil environment and the strikingly high bacterial numbers. Therefore, the use of in vitro and in vivo biofilm model systems under controlled conditions is indispensable for the study of the complex communities. For instance, biofilms exposed to shear stress are studied in flow cells, and when combined with Fluorescence in situ hybridization (FISH) and CLSM, both quantitative data and interactions information can be provided [130]. FISH allows the detection not only of cultivable microorganisms but also of fastidious or uncultured species. However, the main target molecules, 16S rRNAs, could somewhat cause perplexity when measure viability and metabolic activity, as ribosomes can remain intact even in recently nonviable and/or non-metabolically active cells. This can be surmounted by targeting the short-lived intergenic space region (ISR) between the 16S and 23S rRNA segments, called Spacer-FISH [59]. The recent combination of Raman-FISH and secondary ion mass spectrometry can be applied for functional studies of biofilm microbes on a single-cell level by using stable isotopes as labels [131]. Nevertheless, the widely used FISH technique also has limitations, such as laboriousness and unsuitability for high-throughput screening. An alternative device, BioFlux, was presented recently which comprises disposable microplates with embedded microfluidic channels and a distributed pneumatic pump providing rate-controlled fluid flow [132]. BioFlux 1000, which integrates a high performance microscopy workstation, allows real-time, automated image capture. In combination with strain specific markers such as reporter genes, fluorescence-labeled antibodies and probes, flow cells and BioFlux are capable of exploring the structural organization and dynamics of biofilms. Despite the direct visualization of biofilm structures using microscopy techniques, this technique can only provide high quality but semi-quantitative results. Other molecular techniques, such as quantitative PCR and transcriptomic analysis using next-generation sequencing, are promising tools in quantitative assessment and have led to a clearer depiction of the patterns of transcriptional regulation of biofilm-specific genes and the signaling network employed by biofilms at various stages in their growth. 2.2 Quantitative PCR Quantitative PCR (qPCR) has been widely used to quantify microorganisms and measure functional gene markers in complex communities due to its accuracy, high sensitivity, specificity and speed. In contrast to the traditional end-point PCR, the amplification of the PCR products are recorded in “real-time” via a corresponding increase in fluorescent signals. By detecting the accumulation of amplicons during the early exponential phase of the PCR, gene/transcript numbers are quantified when these are proportional to the starting amount of nucleic acid, while the levels of expressed genes are measured when combined with a preceding reverse transcription reaction (RT-qPCR). There are two commonly used reporter systems, namely, SYBR Green assay and TaqMan probe assay. Since SYBR Green binds to all double-stranded DNA by intercalating between adjacent base pairs, including nonspecific PCR products, the target may be overestimated if the primers are not highly specific to their target sequence. Hence, a post-PCR melting (dissociation) curve analysis is needed to verify that the fluorescence signal only comes from target templates. However, the introduction of the reporter probe can significantly increase specificity of the detection. TaqMan 25 probes are dual labeled hydrolysis probes, incorporating a fluorescent reporter molecule at the 5' end and a quencher molecule at the 3' end [133]. During template extension, the reporter-quencher proximity is broken by the 5' to 3' exonuclease activity of the Taq polymerase, thereafter unquenched emission of fluorescence is detected after excitation with a laser. By using multiple TaqMan probes and primer sets, highly similar sequences can be differentiated in different qPCR assays [134]. Furthermore, TaqMan probes labeled with different fluorophores enable that different targets are amplified and quantified within a single reaction which is called multiplex qPCR [135]. In spite of the additional specificity afforded by Taqman probe, the relatively high cost of labeled probe limits its use to some extent. In addition, TaqMan amplicons need to be longer as additional conserved sites are required for designing probes, whereas identification of three conserved regions for primer set and probe within the short amplicon may not be always possible, especially for divergent gene sequences. But this dilemma can be alleviated to a certain extent by using TaqMan MGB probe as the DNA probes with conjugated minor groove binder (MGB) groups form extremely stable duplexes with single-stranded DNA targets, allowing to design shorter probes [136]. However, using MGB probes will further increase the running cost. Maeda et al. reported that both TaqMan and SYBR Green assays showed sufficient sensitivity and specificity for quantification of bacteria species in dental plaque [137]. Taking into account the lower running cost, ease in primer design and assay set-up, SYBR Green assay may be suitable for routine clinical examinations. As a powerful, convenient tool, quantitative PCR assay has been increasingly employed in the past few years to detect and quantify target bacteria in biofilms, or to perform gene expression analysis during biofilm development. By using SYBR Green based qPCR, the population dynamics of pathogenic salmonellas during 4-week period in both water and biofilm samples had been followed [138]. Zhang et al. identified three genes involved in Pseudomonas aeruginosa biofilm-specific resistance to antibiotics by performing mRNA-based qPCR to compare the expression of these genes in planktonic- and biofilm-grown cells [139]. The data from differential transcript levels coupled with quantification of DNA release and cell densities in mixed cultures of three streptococci provided new insights into ecological factors that influence the competition between pioneer colonizing oral species in oral biofilm. In addition, in combination with propidum monoazide (PMA), qPCR assay can overcome its main limitation of the inability to discriminate between live and dead cells [140, 141]. This is achieved through PMA penetrating the membranes that have lost their integrity and then binding to the dsDNA which prevents the use of dsDNA-PMA complex as a template for PCR reaction. The first research using qPCR-PMA technique together with TaqMan probe for analyzing the live and dead cells present in a multispecies oral biofilms was reported recently [142]. Moreover, by use of laser capture microdissection microscopy (LCMM), a small group of cells can be harvested at spatially resolved sites within biofilms, thus reflect heterogeneity inherent to biofilms. Combined LCMM with multiplex RT-qPCR, Franklin presented a stratified biofilm which showed the high amount of housing keeping, acpP, in the top 30 μm of the biofilm, with little or no mRNA at the base of the biofilms, suggesting that the transcription of individual genes varies dramatically in different regions of the biofilms [143]. 26 In order to make valid comparisons between different samples, there are a number of factors that should be taken into account before performing the qPCR assay. One of the important factors is the choice of method used for nucleic acid extraction which is a major determinant on the final quantification. As the extraction efficiencies vary significantly between different methods as well as different types of environmental samples [144], it is problematic to make direct comparison of absolute cell numbers between studies without ensuring that the same extraction procedure is used for each sample. The DNA extraction protocol optimized in our study, which allows DNA extraction from both the gram-negative and -positive cells in the multispecies biofilm with the same efficiency [107], is likely to have a wide application in DNA-based biofilm research. Additionally, PCR inhibitors are often found in environmental samples and interfere with the following qPCR performance. Hence, the equivalent amplification efficiencies between the environmental templates and external standard curves are necessary for absolute quantification. Other potential variables in qPCR assays include preparation and quantification of the standard curve, the subsequent qPCR efficiency, as well as different qPCR reagents and analysis software that are used [145, 146]. Therefore, only the ‘absolute’ numbers generated from the same single qPCR assay can be compared when using the same standard curve [147]. Despite of the unparalleled specificity and sensitivity provided by qPCR-based approaches to target the sequences from a mixed community sample, one noteworthy limitation is that qPCR-based approaches require prior knowledge of the specific target gene of interest. This inevitably results in the fact that any qPCR-based method can not be used to analyze the sequences of unknown species which are likely to account for the vast majority of the world's millions of species. The development of “omic” approaches in recent years, however, can circumvent this problem by providing a PCRindependent assessment of microbial diversity. Hence, combing qPCR technique with other approaches, such as metagenomic and metatranscriptomic analysis, can enable researchers to gain a more comprehensive and in-depth understanding of complex microbial communities. 2.3 Transcriptomics The development of next-generation sequencing techniques, allowing for the analysis of microbial population at a large-scale, has brought forth novel applications, such as metatranscriptomics and metaproteomics which revolutionize the study of complex microbial communities, such as biofilms. The transcriptome is the complete collection of transcribed elements of the genome present in a cell or tissue at a specific development stage or physiological condition. The difference in gene expression patterns between biofilm and planktonic bacteria modes of growth has been well established [148, 149]. Moreover, it is not surprising that the intricate interactions between species in multispecies biofilms can also bring about major changes in gene expression compared to single species biofilms. All existing technologies that have been developed to deduce and quantify the transcriptome can be attributed to either hybridization- or sequencing-based approaches. Hybridization-based approaches are typically based on the custom-made microarrays or commercial high-density oligo microarrays that are incubated with fluorescently labeled cDNA. Although microarray techniques are high throughput and relatively inexpensive, these methods have several limitations, such as high 27 background levels caused by cross-hybridization, difficulty in detecting and quantifying lowabundance species owing to the analog nature of the signal and challenges with comparing expression levels between laboratories and across platforms. Furthermore, microarray analyses rely on existing knowledge about genome sequence [150]. These limitations, however, have been surmounted by sequencing-based approaches since the remarkable sequencing technology have exploded onto the scene, offering dramatically lower per-base costs. Especially, the recently developed next-generation sequencing techniques have opened new doors in the field of transcriptomic analysis, prompting rapid emergence of RNA sequencing, termed RNA-Seq, which directly determines the cDNA sequence from an organism of interest. The major steps involved in bacterial transcriptome sequencing are depicted in Figure 2 [151]. In general, a population of RNA is converted to cDNA library and then sequenced in a high-throughput manner to obtain short reads. Thereafter, the resulting reads are assembled using either de novo or genome-guided approach to produce a genome-scale transcription map that consists of both the transcriptional structure and/or level of expression for each gene [152]. The output length of sequence reads and depth of coverage vary depend on the different DNA sequencing platform used. Three major commercial nextgeneration sequencing platforms: Roche's 454, Illumina's Solexa and Applied Biosystems' SOLiD are commonly used worldwide. Although RNA-Seq, as a novel field of research, is still under active development, it has clear advantages over existing technologies. First, it allows the detection of transcripts in non-model organisms with unknown genomic information. Given the tremendous diversity of uncultivated microorganisms, future transcriptomic studies have the potential to identify new non-coding RNA families. Second, it has a much larger dynamic range (spanning five orders of magnitude) compared with DNA microarrays, and its high accuracy for quantifying expression levels has been verified by quantitative PCR [153]. Third, it can reveal the precise location of transcription boundaries and identify single-nucleotide polymorphism (SNPs) in the transcribed regions which make RNA-Seq useful in complex transcriptomic studies. Overall, RNA-Seq is such a powerful and promising tool that it enables the entire transcriptome to be analyzed in a very high-throughput and quantitative manner and offers single-base resolution while concurrently, profiling gene expression levels at a genome scale. By the use of RNA-Seq technology, the difference of gene expression between mature P. aeruginosa biofilms and planktonic cells was firstly evaluated recently [154]. A set of genes that were specifically regulated in biofilms were identified, including genes involved in type three secretion, adaptation to microaerophilic growth and the production of extracellular matrix components, indicating that biofilms are not just surface attached cells in stationary phase. Moreover, the qualitative analysis of the RNA-Seq data revealed the enrichment of the 5'-ends of the original transcripts, enabling an accurate prediction of transcriptional start sites (TSS). In spite of many studies on expression profile of monospecies biofilms, metatranscriptomic analysis of multispecies biofilm opens up the possibilities for assessing gene expression profiles of whole microbial communities, facilitating the identification of genes of importance under different environmental conditions. Such a study was conducted recently to study patterns of community gene expression in a multispecies biofilm model composed of oral bacteria and periodontal pathogens. The metatranscriptomic data demonstrated the changes in gene expression profiles of the 28 organisms present in the healthy community after the addition of periodontal pathogens to this model and these changes can be accurately evaluated, focusing either on changes at the gene level or treating the transcriptome of the community as a whole [155]. Figure 2 Strategies used in RNA-Seq experiments for bacterial transcriptomic analysis. (a) Outline of the general steps involved in a typical RNA-Seq experiment. (b) Details of an RNA-Seq experiment used for whole-transcriptome profiling of Burkholderia cenocepacia [156]. (c) Procedure used to identify sRNAs associated with Hfq in Salmonella typhimurium [157]. (d) Differential RNA-Seq (dRNA-Seq) used to identify putative transcriptional start sites in Helicobacter pylori [158]. Despite the appealing advantages described above, metatranscriptomic studies of microbial assemblages in situ are still rare so far. This is due to several technological challenges associated with the processing of RNA samples, including the recovery of high-quality mRNA, high demand for computational power and biostatistical expertise and relatively high costs of more depth sequencing for more complex transcriptome. The newly emerging metraproteomics analysis, which aims at assessing the entire protein complement of environmental microbiota at a given time point [159], has a huge potential to link the diversity and activity of microbial communities with their impact on ecological functions. With the falling costs of sequencing, it can be expected that more and more transcriptomics and proteomics will contribute to a deeper understanding of functional dynamics of microbial communities and evolutionary processes. Thence, the query is proposed recently about if it is the time to start an integrated omics approach to biofilms - “biofomics”. This will then lead to the construction of a free on-line database where biofilm signatures are identified and interrogated, including the ability of a microorganism to attach to surfaces, interact with its neighbors and form biofilms, which therefore can provide comprehensive data sets about the overall behavior of the microorganism or system [160]. Nevertheless, before commencing such an action, a 29 versatile, reliable and high-throughput biofilm growing device and appropriate methods for biofilm analysis should be selected with the purpose of minimizing variations and the consequent need of taking a higher number of replicates. However, whether such a device and analytical methods are already fully developed, especially for multispecies biofilms, remains open to question. 30 3 Biofilms and protozoa 3.1 Protozoa Protozoa are unicellular, ubiquitous and colorless eukaryotes with size varying from a few to hundreds of micrometers. Most protozoa are heterotrophic and generally feed on bacteria and other smaller microorganisms, e.g. fungi and algae. A typical predator-prey interaction exhibits three stages of feeding process: contact, capture and ingestion (phagocytosis). Motility plays an important role for protozoa in capturing food and avoiding unfavorable environmental conditions. Based on the types of locomotion, protozoa can be divided into three major groups: amoebae, flagellates and ciliates. Amoebae produce pseudopodia for both locomotion and food-acquiring, while flagellates possess one or more flagella and ciliates use numerous small cilia. In agricultural soil, heterotrophic flagellates and naked amoebae predominate, with numbers ranging from l0,000 to 1,00,000 per gram of arable soil and by forming cysts in their life cycle, they can persist through adverse soil conditions, such as drought stress [31]. Protozoa are important for soil nutrient cycling by feeding on bacteria and releasing excess nitrogen into the soil environment which make them valuable in maintaining microbial equilibrium in the soil. Moreover, protozoa graze different bacteria to different degrees depending on the characteristics of bacteria, including cell size [161], cell surface properties [162], rate of motion [163], extent of biofilm formation [164] as well as the nutritional and biochemical status [165]. There is also cumulative evidence that secondary metabolites produced by bacteria may make them less susceptible to grazing [166-168]. Although amoebae grazing appears to be non-size-selective, flagellates and ciliates have preference for medium-size bacterial cells [169, 170]. By altering bacterial size distribution, grazing is likely to affect the bacterial community structure. In addition, another mechanism involved in the effect of protozoa has on the bacterial community structure is the lower edibility of gram-positive bacteria [171] due to the lower rate of digestion of cell wall [172]. However, this is not always the case as many gram-negative bacteria are inedible while many gram-positive bacteria are adequate food sources for protozoa [173, 174]. For instance, both Bacillus licheniformis and Pseudomonas aeruginosa were found to produce toxic substance to the amoebae [175] and some other features of bacterial cells like size, cell morphology and motility, are not related to gram status but have been shown to influence feeding selectivity. Additionally, the bacterial community structure may also be affected by non-selective grazing. For example, growth rates have been demonstrated to affect the survival of bacteria in environments with intense protozoan predation [176, 177], as a fast-growing species may survive at higher densities than the slow-growing organisms by compensating for cell loss faster. Moreover, by decreasing bacterial numbers and releasing nutrient immobilized by soil microorganisms, protozoa grazing may reduce interspecies competition for substrates [178]. There is also report that shifts in bacterial community composition resulted from the enhancement of grazing which disturbed the established balance between population-specific growth and mortality rates of bacteria by stimulating viral activity [179]. 3.2 Biofilms - the response of cell consortia to protozoan grazing Environmental bacteria, as an integral part of microbial food chain, are frequently confronted with consuming protozoa, which is considered to be a major cause of bacterial death in most freshwater, 31 marine and moist soil habitats [180]. Given their pronounced effects on prey fitness, such as reduced bacterial biomass and changes in both the composition and morphological structure, bacteria have developed diverse strategies against protozoa predation. One of the remarkably effective ways is the formation of cell clusters, e.g., biofilms, although it is not clear up to now that whether enhanced microcolony formation under grazing pressure is a direct defense strategy or an indirect stimulation as a result of nutrient recycling and/or chemical cues [181]. Generally, in natural environment, protozoan colonization of bacterial biofilms has three succession stages. First, early surface colonizers, heterotrophic flagellates, colonize the surface within minutes after exposure due to their high mobility and abundance in the environment. This is followed by ciliates and then the later amoebae. The efficient grazing protection provided by microcolony formation depends on the stage of biofilm development and the feeding mode of the grazer [181, 182], that is early biofilms are colonized mainly by generalists which feed on suspended and attached bacteria, while later biofilm stages are colonized by more permanently attached specialists feeding on surface-associated bacteria [183]. Clearly, biofilms provide a dramatically different feeding environment for protozoa in contrast to planktonic cells. For example, introducing ciliate grazers to biofilms formed by the yeast, Cryptococcus spp., could result in the 1.75 higher levels of biofilm metabolism compared with non-grazed controls. Furthermore, the preferential grazing on the noncellular biofilm matrix over cells embedding in the biofilm demonstrated that EPS could serve as source of nutrients and energy for protists which benefits the biofilm not only from physical protection against ingestion but also from the enhanced nutrient recycling [184]. In addition, the close proximity of bacteria within biofilms can result in enhanced opportunities for interactions such as horizontal gene transfer and quorum sensing (QS), which may induce multiple anti-predator mechanisms that are expressed at different stages of biofilm development. Evidences that QS is involved in regulating anti-predator biofilms have been reported in various microorganisms. Matz and colleagues proved the key roles of microcolonies formation (Figure 3) and inhibitor production induced by quorum sensing in the resistance of Pseudomonas aeruginosa biofilms to protozoan grazing [19]. And the further study showed that inhibitor production of mature P. aeruginosa biofilms is effective against a wider range of biofilm-feeding predators while microcolony-mediated protection is only beneficial in the early stages of biofilm formation [182]. A recent study suggested that protozoan resistance of pathogen P. aeruginosa did not result from activation of QS-regulated public goods, but from the larger and stronger biofilms and thus, concluded that protist predation can favor cooperation within bacterial species [185]. While in Serratia marcescens, QS-controlled, biofilm-specific differentiation of filaments and cell chains in biofilms provides an efficient mechanism against protozoan grazing [21], Vibrio cholera can resist a range of predators by QS-induced production of an unknown toxin [186], which has been proved to be more important in grazing resistance of late biofilms compared with the protection provided by EPS. In early river biofilms under semi-natural conditions, it was shown that the presence of heterotrophic flagellates significantly reduced the abundance of single bacterial cells, but stimulated the formation of bacterial microcolonies [187]. There is also report indicating that protozoa from healthy activated sludge could initially disturb the biofilm development in flow cells but later stimulate its growth [188]. This stimulation could be caused by release of nutrients or non-selective 32 protozoan grazing, which can result in total domination of faster-growing bacteria, whereas some strains are able to even outgrow predators [189-192]. Evidence that biofilm-specific chemical defenses against protozoan predators are a widespread phenomenon in a diverse set of marine bacteria has been presented in a study comparing efficacy of chemical defenses in biofilms and planktonic phases of growth [193]. Despite the intense grazing pressure faced by biofilms due to the lack of mobility to escape protozoan grazing, the fact that bacteria live as biofilms in many environments, could indicate that biofilms have a relatively high anti-predator fitness. And this increased fitness is likely to be favored as a grazing resistance mechanism that has evolved during the long history of close ecological associations between bacteria and protozoa. (a) (b) (c) Figure 3 The effects of flagellate Rhynchomonas nasuta grazing on Pseudomonas aeruginosa biofilms [9, 19]. (a) 3-day-old biofilms of P. aeruginosa PAO1 growing without flagellate. (b) 3-day-old biofilms of P. aeruginosa PAO1 growing with flagellate. (c) 7-day-old biofilms of P. aeruginosa PAO1. Biofilms were pregrown for 3 days before the addition of the flagellate. Scale bar = 50 µm. Nevertheless, some flagellates, such as Cafeteria roenbergensi, Bodo saliens and Caecitellus parvulus, as efficient bacterivores, are proved to be able to reduce the numbers of surface-deposited bacteria just as other protozoa restrict numbers of suspended bacteria [194]. Protozoan grazing have been reported to affect biofilm development by inducing fragmentation and sloughing [195, 196] as well as changing exopolymers distributions and chemical natures of biofilms [197], hence, result in the enhanced spatial and temporal heterogeneity within biofilm communities. Acanthamoeba castellanii and Colpoda maupasi were proved to significantly influence the development and population dynamics of mixed biofilm communities. A. castellanii, as a predominant biofilm grazer, integrated in biofilms, whereas, C. maupasi reduced biofilm thickness by up to 60% as a result of grazing and/or their movement causing sloughing, indicating biofilm growth may not provide total protection against protozoa grazing [198]. Böhme A et al. [21] demonstrated the predation by protozoa with different feeding modes and motility resulted in different morphological structures of multispecies biofilms, including smaller microcolonies with lower maximal and basal layer thickness, larger or mushroom-shaped microcolonies. Furthermore, protozoan grazing could improve mass transfer of nutrients into biofilms, thus accelerate microbial growth. Studies on feeding interactions of two contrasting ciliates with bacterial biofilms have shown that feeding preferences for spatially separated Pseudomonas costantinii biofilms over Serratia plymuthica 33 biofilms can be initiated by the detection of dissolved chemical cues or contact-based detection of bacterial attributes [199]. 3.3 Protozoa and biofilms - reservoirs of pathogenic bacteria The response of bacteria living in a biofilm to the protozoa can change from defensive to exploitative, making protozoa an environmental reservoir for bacterial pathogens. For instance, the growth/survival of bacteria inside the amoebae could give rise to several facultative and obligate pathogens, e.g. Listeria, Mycobacterium and Legionella [200]. What's more, intracellular bacterial growth in protozoa or as a biofilm in soil could lead to distinct phenotypes. Legionella pneumophila and Mycobacterium avium grown in amoebae were reported to be more invasive for macrophages and/or epithelial cells than those grown in vitro [201, 202], moreover, L. pneumophila, replication in amoebae could cause large morphological and biochemical changes and induce an antibiotic and biocide-resistant phenotype [203]. The fact that Legionellosis, in some cases, results from inhalation of aerosols of biofilm- and amoebae-containing L. pneumophila, suggests that protozoa probably have a crucial role in the maintenance and dissemination of human pathogenic bacteria in the environment. Similar to the resistance mechanisms of Pseudomonas aeruginosa biofilms against protozoan grazing, the resistance of biofilm infections to phagocyte-mediated host defense also results from the presence of matrix-enclosed aggregates and the quorum sensing-induced secretion of virulence factors [204, 205]. From a population dynamic-ecological perspective, the control of the replication of the pathogens at the stage of nonspecific, constitutive host defenses is analogous to that of a predator-prey system, with the pathogens being the prey and the phagocytic cells the predators [206]. Future studies directed at unraveling the role of protozoan grazing in the structural and functional stability of biofilms could contribute to a better understanding of the environmental persistence and continuous evolution of bacterial pathogens. 34 4 Where are we going with biofilms? - In the context of microbial ecology Since ecological interaction is one of the main drivers for shaping microbial community, it is obvious that understanding the mechanisms underlying microbial ecology of biofilm is the core of research. As a scientific discipline, microbial ecology examines three aspects of a community [207]: Structure – Which microorganisms are present? Function – Which metabolic capabilities are available/expressed that can support adaptation of this community to environmental conditions? Interaction – How microorganisms interact with one another and with the surrounding environment? These questions can be answered by the more recent and exciting application of the molecularbiology tools, which have marked a major breakthrough in transition from reductionism to holism, that is to elucidating the microbial community as an integral, coevolving system instead of identifying independent individuals. Transcriptomics and proteomics have come on the scene after genomics to allow a comprehensive study of the community function at the transcription and translation levels. Furthermore, metabolomics, described as the comprehensive analysis of the complete set of metabolites (metabolome) produced within each bacterial species and the entire microflora [208], can potentially provide a more accurate snapshot of the actual physiological state of the microorganisms [209, 210] and thus complement transcriptomics and proteomics to assess genetic function as metabolic pathway fluxs [211]. Since one factor affecting the dynamics of microbial community in an environmental niche is the metabolic activity from each member, metabolomics play a crucial role in exploring the biofilm dynamics and interactions within the microbial ecosystems. A good illustration of linking metabolomics with proteomics to assess functional differentiation and interactions was provided by studying a dual-species biofilm composed of Leptospirillum groups II and III. The findings demonstrated that strong metabolomic segregation exhibited organism-specific correlation patterns which reflected the functional differentiation of these two species, indicating that the evolutionary divergence had lessened competition between co-existing microorganisms and allowed them to occupy distinct niches [212]. Moreover, from reconstruction of near-complete or partial recovery of genomes, metabolic network could be deduced in a natural acidophilic biofilm which provided insights into community interactions and functions [213]. Besides, metabolic cooperation that is achieved by synergistic relationship between microorganisms or mutual exchange of metabolites within a community, can be inferred by analyzing metabolomic data [214]. The emergence of various high-throughput technologies now allows more “meta-omics”, e.g. metagenomics, metatranscriptomics and metametabolomics that aim to characterize complete microbial ecosystems. Thus, eco-systems biology was proposed by combining these microbiolomics to interrogate the diversity, function and ecology of the microbial community [214]. Biofilms, as the dominate lifestyle of microorganisms in nature, are undoubtedly ideal for conducting these studies which will provide a more complete picture of biofilm community behavior in the face of environmental stresses. However, the complementary utilization of different “omics” methods also put forward the challenge in meta-data analysis as well as technical challenges, including nucleic acids and protein extraction from environmental samples, mRNA 35 instability and low abundance of certain gene transcripts in total RNA. Especially, compared with genomic, transcriptomic and proteomic analysis, the simultaneous determination of metabolome at a given physiological state is extremely difficult due to the more variable products generated from metabolic reactions [215]. Besides, this may be further complicated when confronted with multispecies biofilms in the environment as a result of the structural and physiological complexity and unevenness of the microbial community. Although the “meta-omics” technologies are expected to revolutionize our understanding of the microbial community, due to the heterogeneity of microniches within a biofilm, particular attention should also be paid to transcriptomic analysis at single cell resolution in order to identify cell-to-cell spatiotemporal variations. Furthermore, prior to sequencing, hypothesis-driven research should be taken into consideration as it may allow more efficient identification and verification of predictions generated by the large-scale data [216]. In parallel, the high-resolution microscopic approaches also continue to shed light on the physical and structural properties of biofilms which enable the in situ investigation of cell attachment, organization and succession within microbial communities. Nevertheless, the three-dimensional, computer-based visualization and quantification of microbial communities in natural ecosystems are still challenges because of the insufficient discrimination of different species and the limited oligonucleotide probes available for various species. When reporter genes are applied, genetic manipulation is also involved for the microorganism being studied which appears to be practically infeasible for unknown species in nature, resulting in the restricted application of these methods to only single species or simple mixed-species population. This, thus, calls for combining multiple technologies or even combining interdisciplinary research efforts in biofilm science. Moreover, the need to standardize experimental approaches, which would allow for more reliable and comparable results, is also increasingly appreciated. More attention is being given to multispecies biofilms than ever before, as evidenced by a surge of publications in this field since 2003. While some molecular mechanisms underlying intra- and interspecies interactions between individual strains have been clarified, mostly are only applicable to dual-species biofilms, which is notably in contrast to the composition of biofilms in environmental habitats. The biofilms in natural environment are much more complex entities than what we have examined in the laboratory owing to the structural and physiological heterogeneity as well as the extensive and striking interrelations between their components. And the coordinated behaviors of biofilm have led to the idea that biofilm acts more like an orchestrated multicellular organism than a collection of organisms. This view has been supported by the observation that the altruistic selfsacrifice of the majority of the population when exposed to adverse conditions, such as programmed cell death [217], could ultimately enhance the survival of the whole community [113], which violates the principles of Darwinian evolution. This can be explained by the statement that Darwin’s evolutionary theory focused intensely on one level of existence (organic species) while failing to conceive the existence of multispecies communities and their potential role in the origin of species [21]. However, objections are also raised to this fascinating analogy, mainly because upon their decay, the biofilm cells can turn back to planktonic form in response to environmental signals in contrast to the irreversible differentiation of the cells in a multicellular organism. Additionally, whether the biofilm-specific genes that hierarchically regulate transition through specific stages in 36 biofilm formation have been identified successfully, is still be questioned [116]. Hence, the term “biofilm phenotype” is commonly used rather than “biofilm genotype”. From another point of view, [218]“city of microbes” was put forward to describe the association of microorganisms in biofilms, where inhabitants select location, limit settlements of too many bacteria, store energy in exopolysaccharide, transfer genetic material horizontally and communicate with their neighbors. When conditions in the “city” are less comfortable, inhabitants migration occurs [218, 219]. In the ecological context, the perception of biofilms as microbial landscapes has been proposed by Battin TJ et al. [220] to emphasize the spatially explicit dimension and unify numerous facets of biofilms, such as biodiversity, dynamics and ecosystem function. In summary, multispecies biofilm provides a valuable model for testing ecological and evolutionary theories relating to interspecies interactions. By integrating ecological and evolutionary frameworks with the promising molecular technologies mentioned in this thesis, a genetic and biochemical understanding of the interactions between biofilm landscape components and the operating ecological process is supposed to be achieved, and thus helps us to manage biofilms to provide services to society. 37 5 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Kolenbrander PE, Palmer RJ, Periasamy S, Jakubovics NS (2010) Oral multispecies biofilm development and the key role of cell–cell distance. Nature Reviews Microbiology 8: 471-480. Costerton JW, Cheng K, Geesey GG, Ladd TI, Nickel JC, Dasgupta M, Marrie TJ (1987) Bacterial biofilms in nature and disease. Annual Reviews in Microbiology 41: 435-464. Donlan RM, Costerton JW (2002) Biofilms: survival mechanisms of clinically relevant microorganisms. Clinical microbiology reviews 15: 167-193. Costerton J, Stewart PS, Greenberg E (1999) Bacterial biofilms: a common cause of persistent infections. Science 284: 1318-1322. Potera C (1996) Biofilms invade microbiology. Science 273: 1795-1797. Davey ME, O'toole GA (2000) Microbial biofilms: from ecology to molecular genetics. Microbiology and molecular biology reviews 64: 847-867. Almås ÅR, Mulder J, Bakken LR (2005) Trace metal exposure of soil bacteria depends on their position in the soil matrix. Environmental science & technology 39: 5927-5932. McKew BA, Taylor JD, McGenity TJ, Underwood GJ (2010) Resistance and resilience of benthic biofilm communities from a temperate saltmarsh to desiccation and rewetting. The ISME journal 5: 30-41. Matz C, Kjelleberg S (2005) Off the hook–how bacteria survive protozoan grazing. Trends in microbiology 13: 302-307. Stewart PS, William Costerton J (2001) Antibiotic resistance of bacteria in biofilms. The Lancet 358: 135-138. Jesaitis AJ, Franklin MJ, Berglund D, Sasaki M, Lord CI, Bleazard JB, Duffy JE, Beyenal H, Lewandowski Z (2003) Compromised host defense on Pseudomonas aeruginosa biofilms: characterization of neutrophil and biofilm interactions. The Journal of Immunology 171: 4329-4339. Jaffe A, Chabbert YA, Semonin O (1982) Role of porin proteins OmpF and OmpC in the permeation of beta-lactams. Antimicrobial agents and chemotherapy 22: 942-948. Roberson EB, Firestone MK (1992) Relationship between desiccation and exopolysaccharide production in a soil Pseudomonas sp. Applied and Environmental Microbiology 58: 1284-1291. Harrison JJ, Turner RJ, Ceri H (2005) Persister cells, the biofilm matrix and tolerance to metal cations in biofilm and planktonic Pseudomonas aeruginosa. Environmental Microbiology 7: 981994. Vuong C, Voyich JM, Fischer ER, Braughton KR, Whitney AR, DeLeo FR, Otto M (2004) Polysaccharide intercellular adhesin (PIA) protects Staphylococcus epidermidis against major components of the human innate immune system. Cellular microbiology 6: 269-275. Evans D, Allison D, Brown M, Gilbert P (1991) Susceptibility of Pseudomonas aeruginosa and Escherichia coli biofilms towards ciprofloxacin: effect of specific growth rate. Journal of Antimicrobial Chemotherapy 27: 177-184. Hengge-Aronis R (1996) Regulation of gene expression during entry into stationary phase. Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed ASM Press, Washington, DC: 1497-1512. Bagge N, Hentzer M, Andersen JB, Ciofu O, Givskov M, Høiby N (2004) Dynamics and spatial distribution of β-lactamase expression in Pseudomonas aeruginosa biofilms. Antimicrobial agents and chemotherapy 48: 1168-1174. Matz C, Bergfeld T, Rice SA, Kjelleberg S (2004) Microcolonies, quorum sensing and cytotoxicity determine the survival of Pseudomonas aeruginosa biofilms exposed to protozoan grazing. Environmental microbiology 6: 218-226. Lewis K (2001) Riddle of biofilm resistance. Antimicrobial agents and chemotherapy 45: 999-1007. Queck SY, Weitere M, Moreno AM, Rice SA, Kjelleberg S (2006) The role of quorum sensing mediated developmental traits in the resistance of Serratia marcescens biofilms against protozoan grazing. Environmental microbiology 8: 1017-1025. Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nature Reviews Microbiology 2: 95-108. 38 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. Torsvik V, Goksøyr J, Daae FL (1990) High diversity in DNA of soil bacteria. Applied and environmental microbiology 56: 782-787. Gans J, Wolinsky M, Dunbar J (2005) Computational improvements reveal great bacterial diversity and high metal toxicity in soil. Science 309: 1387-1390. Nunan N, Wu K, Young IM, Crawford JW, Ritz K (2003) Spatial distribution of bacterial communities and their relationships with the micro-architecture of soil. FEMS Microbiology Ecology 44: 203-215. Sutherland IW (2001) The biofilm matrix–an immobilized but dynamic microbial environment. Trends in microbiology 9: 222-227. Chang W-S, van de Mortel M, Nielsen L, de Guzman GN, Li X, Halverson LJ (2007) Alginate production by Pseudomonas putida creates a hydrated microenvironment and contributes to biofilm architecture and stress tolerance under water-limiting conditions. Journal of bacteriology 189: 82908299. Lee S-E, Kim S-H, Lee J-U, Chon H-T (2013) Technical development of geomicrobiological fixation of heavy metals in soils and sediments by biosorption and bioreduction-case histories in Korea. EGU General Assembly Conference Abstracts, vol. 15, pp. 2702. Sandoval-Carrasco CA, Ahuatzi-Chacón D, Galíndez-Mayer J, Ruiz-Ordaz N, Juárez-Ramírez C, Martínez-Jerónimo F (2013) Biodegradation of a mixture of the herbicides ametryn, and 2, 4dichlorophenoxyacetic acid (2, 4-D) in a compartmentalized biofilm reactor. Bioresource technology 145: 33-36. Walker TS, Bais HP, Déziel E, Schweizer HP, Rahme LG, Fall R, Vivanco JM (2004) Pseudomonas aeruginosa-plant root interactions. Pathogenicity, biofilm formation, and root exudation. Plant physiology 134: 320-331. Ekelund F, Rønn R (1994) Notes on protozoa in agricultural soil with emphasis on heterotrophic flagellates and naked amoebae and their ecology. FEMS Microbiology Reviews 15: 321-353. Marsh PD (2006) Dental plaque as a biofilm and a microbial community–implications for health and disease. BMC Oral Health 6: S14. Suppiger A, Schmid N, Aguilar C, Pessi G, Eberl L (2013) Two quorum sensing systems control biofilm formation and virulence in members of the Burkholderia cepacia complex. Virulence 4: 1918. Huber B, Riedel K, Hentzer M, Heydorn A, Gotschlich A, Givskov M, Molin S, Eberl L (2001) The cep quorum-sensing system of Burkholderia cepacia H111 controls biofilm formation and swarming motility. Microbiology 147: 2517-2528. Schmid N, Pessi G, Deng Y, Aguilar C, Carlier AL, Grunau A, Omasits U, Zhang L-H, Ahrens CH, Eberl L (2012) The AHL-and BDSF-dependent Quorum sensing systems control specific and overlapping sets of genes in Burkholderia cenocepacia H111. PloS one 7: e49966. Savage VJ, Chopra I, O'Neill AJ (2013) Staphylococcus aureus biofilms promote horizontal transfer of antibiotic resistance. Antimicrobial agents and chemotherapy 57: 1968-1970. Martínez JL (2008) Antibiotics and antibiotic resistance genes in natural environments. Science 321: 365-367. Forsberg KJ, Reyes A, Wang B, Selleck EM, Sommer MO, Dantas G (2012) The shared antibiotic resistome of soil bacteria and human pathogens. Science 337: 1107-1111. Seneviratne G, Kecskés ML, Kennedy IR (2008) Biofilmed biofertilisers: novel inoculants for efficient nutrient use in plants. In: Kennedy, I.R., Choudhury, A.T.M.A., Kecskés, M.L., Rose, M.T. (Eds.), Efficient Nutrient Use in Rice Production in Vietnam Achieved Using Inoculant Biofertilisers. Proceedings of a Project (SMCN/2002/073) Workshop Held in Hanoi, Vietnam, 12e13 October 2007. ACIAR Proceedings No. 130, Canberra, pp. 126e130. Bianciotto V, Bonfante P (2002) Arbuscular mycorrhizal fungi: a specialised niche for rhizospheric and endocellular bacteria. Antonie van Leeuwenhoek 81: 365-371. Seneviratne G, Zavahir J, Bandara W, Weerasekara M (2008) Fungal-bacterial biofilms: their development for novel biotechnological applications. World Journal of Microbiology and Biotechnology 24: 739-743. 39 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. Jayasinghearachchi H, Seneviratae G (2010) A Mushroom-fungus helps improve endophytic colonization of tomato by Pseudomonas fluorescens through biofilm formation. Research Journal of Microbiology 5: 689-695. Jayasinghearachchi H, Seneviratne G (2004) A bradyrhizobial-Penicillium spp. biofilm with nitrogenase activity improves N2 fixing symbiosis of soybean. Biology and fertility of soils 40: 432434. LoVLEY DR, Klug MJ (1986) Model for the distribution of sulfate reduction and methanogenesis in freshwater sediments. Geochimica et Cosmochimica Acta 50: 11-18. Stewart PS, Franklin MJ (2008) Physiological heterogeneity in biofilms. Nature Reviews Microbiology 6: 199-210. Matz C (2011) Competition, Communication, Cooperation: Molecular Crosstalk in Multi-species BiofilmsBiofilm Highlights. In: Flemming, H.-C., Wingender, J., Szewzyk, U. (Eds.), Biofilm Highlight. Springer Series on Biofilms, vol. 5. Springer-Verlag Berlin, Heidelberg, pp. 29-40. Rickard AH, Gilbert P, High NJ, Kolenbrander PE, Handley PS (2003) Bacterial coaggregation: an integral process in the development of multi-species biofilms. Trends in microbiology 11: 94-100. Kolenbrander PE, Andersen RN, Blehert DS, Egland PG, Foster JS, Palmer RJ (2002) Communication among oral bacteria. Microbiology and molecular biology reviews 66: 486-505. Klayman BJ, Volden PA, Stewart PS, Camper AK (2009) Escherichia coli O157: H7 requires colonizing partner to adhere and persist in a capillary flow cell. Environmental science & technology 43: 2105-2111. Bradshaw DJ, Marsh PD, Watson GK, Allison C (1998) Role of Fusobacterium nucleatum and coaggregation in anaerobe survival in planktonic and biofilm oral microbial communities during aeration. Infection and immunity 66: 4729-4732. Gilbert P, Maira-Litran T, McBain AJ, Rickard AH, Whyte FW (2002) The physiology and collective recalcitrance of microbial biofilm communities. Advances in microbial physiology 46: 203-256. Adam B, Baillie GS, DOUGLAS LJ (2002) Mixed species biofilms of Candida albicans and Staphylococcus epidermidis. Journal of medical microbiology 51: 344-349. Boonchan S, Britz ML, Stanley GA (2000) Degradation and mineralization of high-molecularweight polycyclic aromatic hydrocarbons by defined fungal-bacterial cocultures. Applied and Environmental Microbiology 66: 1007-1019. Farabegoli G, Chiavola A, Rolle E (2008) Remediation of chlorophenol-and phenol-contaminated groundwater by a sequencing batch biofilm reactor. Water Science and Technology 58: 295–301. Flemming H-C, Wingender J (2010) The biofilm matrix. Nature Reviews Microbiology 8: 623-633. Richard G, Burns, Richard P. Dick (2002) Microbes and Enzymes in Biofilms. Enzymes in the Environment: Activity, Ecology, and Applications 86: 307. Jakubovics NS, Gill SR, Iobst SE, Vickerman M, Kolenbrander PE (2008) Regulation of gene expression in a mixed-genus community: stabilized arginine biosynthesis in Streptococcus gordonii by coaggregation with Actinomyces naeslundii. Journal of bacteriology 190: 3646-3657. Parsek MR, Greenberg E (2005) Sociomicrobiology: the connections between quorum sensing and biofilms. Trends in microbiology 13: 27-33. Häussler S, Parsek MR (2010) Biofilms 2009: new perspectives at the heart of surface-associated microbial communities. Journal of bacteriology 192: 2941-2949. Riedel K, Hentzer M, Geisenberger O, Huber B, Steidle A, Wu H, Høiby N, Givskov M, Molin S, Eberl L (2001) N-acylhomoserine-lactone-mediated communication between Pseudomonas aeruginosa and Burkholderia cepacia in mixed biofilms. Microbiology 147: 3249-3262. Eberl L, Tümmler B (2004) Pseudomonas aeruginosa and Burkholderia cepacia in cystic fibrosis: genome evolution, interactions and adaptation. International journal of medical microbiology 294: 123-131. Leadbetter JR, Greenberg E (2000) Metabolism of acyl-homoserine lactone quorum-sensing signals by Variovorax paradoxus. Journal of Bacteriology 182: 6921-6926. 40 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. Liu X, Jia J, Popat R, Ortori CA, Li J, Diggle SP, Gao K, Cámara M (2011) Characterisation of two quorum sensing systems in the endophytic Serratia plymuthica strain G3: differential control of motility and biofilm formation according to life-style. BMC microbiology 11: 26. De Kievit T (2009) Quorum sensing in Pseudomonas aeruginosa biofilms. Environmental microbiology 11: 279-288. Suntharalingam P, Cvitkovitch DG (2005) Quorum sensing in streptococcal biofilm formation. Trends in microbiology 13: 3-6. Saito Y, Fujii R, Nakagawa KI, Kuramitsu H, Okuda K, Ishihara K (2008) Stimulation of Fusobacterium nucleatum biofilm formation by Porphyromonas gingivalis. Oral microbiology and immunology 23: 1-6. Jang Y-J, Sim J, Jun H-K, Choi B-K (2013) Differential effect of autoinducer 2 of Fusobacterium nucleatum on oral streptococci. Archives of oral biology 58: 1594-1602. Schauder S, Bassler BL (2001) The languages of bacteria. Genes & Development 15: 1468-1480. Reading NC, Sperandio V (2006) Quorum sensing: the many languages of bacteria. FEMS microbiology letters 254: 1-11. Deng Y, Wu Je, Eberl L, Zhang L-H (2010) Structural and functional characterization of diffusible signal factor family quorum-sensing signals produced by members of the Burkholderia cepacia complex. Applied and environmental microbiology 76: 4675-4683. Davies DG, Marques CN (2009) A fatty acid messenger is responsible for inducing dispersion in microbial biofilms. Journal of bacteriology 191: 1393-1403. Ryan RP, Dow JM (2011) Communication with a growing family: diffusible signal factor (DSF) signaling in bacteria. Trends in microbiology 19: 145-152. Boon C, Deng Y, Wang L-H, He Y, Xu J-L, Fan Y, Pan SQ, Zhang L-H (2007) A novel DSF-like signal from Burkholderia cenocepacia interferes with Candida albicans morphological transition. The ISME journal 2: 27-36. Ryan RP, Fouhy Y, Garcia BF, Watt SA, Niehaus K, Yang L, Tolker‐Nielsen T, Dow JM (2008) Interspecies signalling via the Stenotrophomonas maltophilia diffusible signal factor influences biofilm formation and polymyxin tolerance in Pseudomonas aeruginosa. Molecular microbiology 68: 75-86. Keller L, Surette MG (2006) Communication in bacteria: an ecological and evolutionary perspective. Nature Reviews Microbiology 4: 249-258. West SA, Winzer K, Gardner A, Diggle SP (2012) Quorum sensing and the confusion about diffusion. Trends in microbiology 20: 586-594. Hense BA, Kuttler C, Müller J, Rothballer M, Hartmann A, Kreft J-U (2007) Does efficiency sensing unify diffusion and quorum sensing? Nature Reviews Microbiology 5: 230-239. Tischler AD, Camilli A (2004) Cyclic diguanylate (c‐di‐GMP) regulates Vibrio cholerae biofilm formation. Molecular microbiology 53: 857-869. Beyhan S, Odell LS, Yildiz FH (2008) Identification and characterization of cyclic diguanylate signaling systems controlling rugosity in Vibrio cholerae. Journal of bacteriology 190: 7392-7405. Waters CM, Lu W, Rabinowitz JD, Bassler BL (2008) Quorum sensing controls biofilm formation in Vibrio cholerae through modulation of cyclic di-GMP levels and repression of vpsT. Journal of bacteriology 190: 2527-2536. Caiazza NC, Merritt JH, Brothers KM, O'Toole GA (2007) Inverse regulation of biofilm formation and swarming motility by Pseudomonas aeruginosa PA14. Journal of bacteriology 189: 3603-3612. Newell PD, Monds RD, O'Toole GA (2009) LapD is a bis-(3′, 5′)-cyclic dimeric GMP-binding protein that regulates surface attachment by Pseudomonas fluorescens Pf0–1. Proceedings of the National Academy of Sciences 106: 3461-3466. Tribble GD, Rigney TW, Dao D-HV, Wong CT, Kerr JE, Taylor BE, Pacha S, Kaplan HB (2012) Natural competence is a major mechanism for horizontal DNA transfer in the oral pathogen Porphyromonas gingivalis. MBio 3. Wang B, Chi B, Kuramitsu H (2002) Genetic exchange between Treponema denticola and Streptococcus gordonii in biofilms. Oral microbiology and immunology 17: 108-112. 41 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. Case RJ, Labbate M, Kjelleberg S (2008) AHL-driven quorum-sensing circuits: their frequency and function among the Proteobacteria. The ISME journal 2: 345-349. Huang Y, Zeng Y, Yu Z, Zhang J, Feng H, Lin X (2013) In silico and experimental methods revealed highly diverse bacteria with quorum sensing and aromatics biodegradation systems-a potential broad application on bioremediation. Bioresource technology 148: 311-316. Burmølle M, Bahl MI, Jensen LB, Sørensen SJ, Hansen LH (2008) Type 3 fimbriae, encoded by the conjugative plasmid pOLA52, enhance biofilm formation and transfer frequencies in Enterobacteriaceae strains. Microbiology 154: 187-195. Ma H, Bryers JD (2013) Non-invasive determination of conjugative transfer of plasmids bearing antibiotic-resistance genes in biofilm-bound bacteria: effects of substrate loading and antibiotic selection. Applied microbiology and biotechnology 97: 317-328. Ehrlich GD, Ahmed A, Earl J, Hiller NL, Costerton JW, Stoodley P, Post JC, DeMeo P, Hu FZ (2010) The distributed genome hypothesis as a rubric for understanding evolution in situ during chronic bacterial biofilm infectious processes. FEMS Immunology & Medical Microbiology 59: 269-279. Ghigo J-M (2001) Natural conjugative plasmids induce bacterial biofilm development. Nature 412: 442-445. Schwering M, Song J, Louie M, Turner RJ, Ceri H (2013) Multi-species biofilms defined from drinking water microorganisms provide increased protection against chlorine disinfection. Biofouling 29: 917-928. Lee KWK, Periasamy S, Mukherjee M, Xie C, Kjelleberg S, Rice SA (2013) Biofilm development and enhanced stress resistance of a model, mixed-species community biofilm. The ISME Journal 114. Kreth J, Merritt J, Shi W, Qi F (2005) Competition and coexistence between Streptococcus mutans and Streptococcus sanguinis in the dental biofilm. Journal of bacteriology 187: 7193-7203. Rao D, Webb JS, Kjelleberg S (2005) Competitive interactions in mixed-species biofilms containing the marine bacterium Pseudoalteromonas tunicata. Applied and environmental microbiology 71: 1729-1736. Costa R, Gomes N, Krögerrecklenfort E, Opelt K, Berg G, Smalla K (2007) Pseudomonas community structure and antagonistic potential in the rhizosphere: insights gained by combining phylogenetic and functional gene‐based analyses. Environmental microbiology 9: 2260-2273. Tait K, Sutherland I (2002) Antagonistic interactions amongst bacteriocin‐producing enteric bacteria in dual species biofilms. Journal of applied microbiology 93: 345-352. Foster KR, Bell T (2012) Competition, not cooperation, dominates interactions among culturable microbial species. Current Biology 22: 1845-1850. Molina L, Constantinescu F, Michel L, Reimmann C, Duffy B, Défago G (2003) Degradation of pathogen quorum‐sensing molecules by soil bacteria: a preventive and curative biological control mechanism. FEMS Microbiology Ecology 45: 71-81. Peleg AY, Tampakakis E, Fuchs BB, Eliopoulos GM, Moellering RC, Mylonakis E (2008) Prokaryote–eukaryote interactions identified by using Caenorhabditis elegans. Proceedings of the National Academy of Sciences 105: 14585-14590. West SA, Diggle SP, Buckling A, Gardner A, Griffin AS (2007) The social lives of microbes. Annu Rev Ecol Evol Syst 38: 53-77. Jass J, Surman S, Walker JT (2003) Medical biofilms. In: J. Jass, S. Surman, J. Walker (Eds.), pp. 51–72. Kuenen J, Robertson LA, van Gemerden H (1985) Microbial interactions among aerobic and anaerobic sulfur-oxidizing bacteria. Adv Microb Ecol 8: 1-59. Clark F (1969) Ecological associations among soil micro-organisms. In Soil Biology 125-161. Park D (1960) Antagonism—the background to soil fungi. The ecology of soil fungi: 148-159. Burmølle M, Ren D, Bjarnsholt T, Sørensen SJ (2013) Interactions in multispecies biofilms: do they actually matter? Trends in microbiology. doi: 10.1016/j.tim.2013.12.004 Hansen SK, Rainey PB, Haagensen JA, Molin S (2007) Evolution of species interactions in a biofilm community. Nature 445: 533-536. 42 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. Ren D, Madsen JS, de la Cruz-Perera CI, Bergmark L, Sørensen SJ, Burmølle M (2013) HighThroughput Screening of Multispecies Biofilm Formation and Quantitative PCR-Based Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. Microbial ecology. doi: 10.1007/s00248-013-0315-z Mitri S, Foster K (2013) A Genotypic View of Social Interactions in Microbial Communities. Annual review of genetics 47: 247-273. Nadell CD, Xavier JB, Foster KR (2009) The sociobiology of biofilms. FEMS microbiology reviews 33: 206-224. West SA, Griffin AS, Gardner A, Diggle SP (2006) Social evolution theory for microorganisms. Nature Reviews Microbiology 4: 597-607. Zhang W, Sileika T, Packman AI (2013) Effects of fluid flow conditions on interactions between species in biofilms. FEMS microbiology ecology 2: 344-354. Nadell CD, Foster KR, Xavier JB (2010) Emergence of spatial structure in cell groups and the evolution of cooperation. PLoS computational biology 6: e1000716. Kreft J-U (2004) Biofilms promote altruism. Microbiology 150: 2751-2760. Kelvin LKW (2013) Role of cell-cell signaling, differentiation and fitness in multi-species biofilms. Dissertation, Nanyang Technologial University. Morris JJ, Lenski RE, Zinser ER (2012) The Black Queen Hypothesis: evolution of dependencies through adaptive gene loss. MBio 3 doi: 10.1128/mBio.00036–12 Monds RD, O’Toole GA (2009) The developmental model of microbial biofilms: ten years of a paradigm up for review. Trends in microbiology 17: 73-87. Schink B (2002) Synergistic interactions in the microbial world. Antonie Van Leeuwenhoek 81: 257-261. Bakke R, Trulear M, Robinson J, Characklis W (1984) Activity of Pseudomonas aeruginosa in biofilms: steady state. Biotechnology and bioengineering 26: 1418-1424. Nickel J, Wright J, Ruseska I, Marrie T, Whitfield C, Costerton J (1985) Antibiotic resistance of Pseudomonas aeruginosa colonizing a urinary catheter in vitro. European journal of clinical microbiology 4: 213-218. Kharazmi A, Giwercman B, Høiby N (1999) Robbins device in biofilm research. Methods in enzymology 310: 207-215. Stepanović S, Vuković D, Dakić I, Savić B, Švabić-Vlahović M (2000) A modified microtiter-plate test for quantification of staphylococcal biofilm formation. Journal of Microbiological Methods 40: 175-179. Harrison J, Turner R, Ceri H (2005) High-throughput metal susceptibility testing of microbial biofilms. BMC microbiology 5: 53. Bragança S, Azevedo N, Simões LC, Keevil C, Vieira M (2006) Use of fluorescent in situ hybridization for the visualization of Helicobacter pylori in real drinking water biofilms. Water Science and Technology 55:387–393. Chavant P, Gaillard-Martinie B, Talon R, Hébraud M, Bernardi T (2007) A new device for rapid evaluation of biofilm formation potential by bacteria. Journal of microbiological methods 68: 605612. Silagyi K, Kim S-H, Martin Lo Y, Wei C-i (2009) Production of biofilm and quorum sensing by Escherichia coli O157: H7 and its transfer from contact surfaces to meat, poultry, ready-to-eat deli, and produce products. Food microbiology 26: 514-519. Min K, Rickard A (2009) Coaggregation by the freshwater bacterium Sphingomonas natatoria alters dual-species biofilm formation. Applied and environmental microbiology 75: 3987-3997. Ceri H, Olson M, Stremick C, Read R, Morck D, Buret A (1999) The Calgary Biofilm Device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms. Journal of clinical microbiology 37: 1771-1776. Harrison JJ, Ceri H, Yerly J, Stremick CA, Hu Y, Martinuzzi R, Turner RJ (2006) The use of microscopy and three-dimensional visualization to evaluate the structure of microbial biofilms cultivated in the Calgary Biofilm Device. Biological procedures online 8: 194-215. 43 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. Schubotz F, Meyer‐Dombard D, Bradley A, Fredricks H, Hinrichs KU, Shock E, Summons R (2013) Spatial and temporal variability of biomarkers and microbial diversity reveal metabolic and community flexibility in Streamer Biofilm Communities in the Lower Geyser Basin, Yellowstone National Park. Geobiology 11: 549-569. Zhu Y, Dashper SG, Chen Y-Y, Crawford S, Slakeski N, Reynolds EC (2013) Porphyromonas gingivalis and Treponema denticola Synergistic Polymicrobial Biofilm Development. PloS one 8: e71727. Wagner M (2009) Single-cell ecophysiology of microbes as revealed by Raman microspectroscopy or secondary ion mass spectrometry imaging. Annual review of microbiology 63: 411-429. Benoit MR, Conant CG, Ionescu-Zanetti C, Schwartz M, Matin A (2010) New device for highthroughput viability screening of flow biofilms. Applied and environmental microbiology 76: 41364142. Livak KJ, Flood S, Marmaro J, Giusti W, Deetz K (1995) Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. Genome Research 4: 357-362. Smith CJ, Nedwell DB, Dong LF, Osborn AM (2007) Diversity and abundance of nitrate reductase genes (narG and napA), nitrite reductase genes (nirS and nrfA), and their transcripts in estuarine sediments. Applied and environmental microbiology 73: 3612-3622. Baldwin BR, Nakatsu CH, Nies L (2003) Detection and enumeration of aromatic oxygenase genes by multiplex and real-time PCR. Applied and Environmental Microbiology 69: 3350-3358. Kutyavin IV, Afonina IA, Mills A, Gorn VV, Lukhtanov EA, Belousov ES, Singer MJ, Walburger DK, Lokhov SG, Gall AA (2000) 3′-minor groove binder-DNA probes increase sequence specificity at PCR extension temperatures. Nucleic Acids Research 28: 655-661. Maeda H, Fujimoto C, Haruki Y, Maeda T, Kokeguchi S, Petelin M, Arai H, Tanimoto I, Nishimura F, Takashiba S (2003) Quantitative real‐time PCR using TaqMan and SYBR Green for Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, tetQ gene and total bacteria. FEMS Immunology & Medical Microbiology 39: 81-86. Sha Q, Vattem DA, Forstner MR, Hahn D (2013) Quantifying Salmonella population dynamics in water and biofilms. Microbial ecology 65: 60-67. Zhang L, Fritsch M, Hammond L, Landreville R, Slatculescu C, Colavita A, Mah T-F (2013) Identification of Genes Involved in Pseudomonas aeruginosa Biofilm-Specific Resistance to Antibiotics. PloS one 8: e61625. Pan Y, Breidt F (2007) Enumeration of viable Listeria monocytogenes cells by real-time PCR with propidium monoazide and ethidium monoazide in the presence of dead cells. Applied and environmental microbiology 73: 8028-8031. Chen NT, Chang CW (2010) Rapid quantification of viable legionellae in water and biofilm using ethidium monoazide coupled with real‐time quantitative PCR. Journal of applied microbiology 109: 623-634. Àlvarez G, González M, Isabal S, Blanc V, León R (2013) Method to quantify live and dead cells in multi-species oral biofilm by real-time PCR with propidium monoazide. AMB Express 3: 1-8. Lenz AP, Williamson KS, Pitts B, Stewart PS, Franklin MJ (2008) Localized gene expression in Pseudomonas aeruginosa biofilms. Applied and environmental microbiology 74: 4463-4471. Martin-Laurent F, Philippot L, Hallet S, Chaussod R, Germon J, Soulas G, Catroux G (2001) DNA extraction from soils: old bias for new microbial diversity analysis methods. Applied and environmental microbiology 67: 2354-2359. Love JL, Scholes P, Gilpin B, Savill M, Lin S, Samuel L (2006) Evaluation of uncertainty in quantitative real-time PCR. Journal of microbiological methods 67: 349-356. Smith CJ, Osborn AM (2009) Advantages and limitations of quantitative PCR (Q‐PCR)‐based approaches in microbial ecology. FEMS microbiology ecology 67: 6-20. Smith CJ, Nedwell DB, Dong LF, Osborn AM (2006) Evaluation of quantitative polymerase chain reaction‐based approaches for determining gene copy and gene transcript numbers in environmental samples. Environmental Microbiology 8: 804-815. 44 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. Lo A, Seers C, Boyce J, Dashper S, Slakeski N, Lissel JP, Reynolds E (2009) Comparative transcriptomic analysis of Porphyromonas gingivalis biofilm and planktonic cells. BMC microbiology 9: 18. Shemesh M, Tam A, Steinberg D (2007) Differential gene expression profiling of Streptococcus mutans cultured under biofilm and planktonic conditions. Microbiology 153: 1307-1317. Shendure J (2008) The beginning of the end for microarrays? Nature Methods 5: 585-587. Febrer M, McLay K, Caccamo M, Twomey KB, Ryan RP (2011) Advances in bacterial transcriptome and transposon insertion-site profiling using second-generation sequencing. Trends in biotechnology 29: 586-594. Wang Z, Gerstein M, Snyder M (2009) RNA-Seq: a revolutionary tool for transcriptomics. Nature Reviews Genetics 10: 57-63. Nagalakshmi U, Wang Z, Waern K, Shou C, Raha D, Gerstein M, Snyder M (2008) The transcriptional landscape of the yeast genome defined by RNA sequencing. Science 320: 1344-1349. Dötsch A, Eckweiler D, Schniederjans M, Zimmermann A, Jensen V, Scharfe M, Geffers R, Häussler S (2012) The Pseudomonas aeruginosa transcriptome in planktonic cultures and static biofilms using RNA sequencing. PloS one 7: e31092. Frias-Lopez J, Duran-Pinedo A (2012) Effect of periodontal pathogens on the metatranscriptome of a healthy multispecies biofilm model. Journal of bacteriology 194: 2082-2095. Yoder-Himes D, Chain P, Zhu Y, Wurtzel O, Rubin E, Tiedje JM, Sorek R (2009) Mapping the Burkholderia cenocepacia niche response via high-throughput sequencing. Proceedings of the National Academy of Sciences 106: 3976-3981. Sittka A, Lucchini S, Papenfort K, Sharma CM, Rolle K, Binnewies TT, Hinton JC, Vogel J (2008) Deep sequencing analysis of small noncoding RNA and mRNA targets of the global posttranscriptional regulator, Hfq. PLoS genetics 4: e1000163. Sharma CM, Hoffmann S, Darfeuille F, Reignier J, Findeiß S, Sittka A, Chabas S, Reiche K, Hackermüller J, Reinhardt R (2010) The primary transcriptome of the major human pathogen Helicobacter pylori. Nature 464: 250-255. Wilmes P, Bond PL (2004) The application of two‐dimensional polyacrylamide gel electrophoresis and downstream analyses to a mixed community of prokaryotic microorganisms. Environmental Microbiology 6: 911-920. Azevedo NF, Lopes SP, Keevil CW, Pereira MO, Vieira MJ (2009) Time to “go large” on biofilm research: advantages of an omics approach. Biotechnology letters 31: 477-485. Posch T, Jezbera J, Vrba J, Šimek K, Pernthaler J, Andreatta S, Sonntag B (2001) Size selective feeding in Cyclidium glaucoma (Ciliophora, Scuticociliatida) and its effects on bacterial community structure: a study from a continuous cultivation system. Microbial ecology 42: 217-227. Monger BC, Landry MR, Brown SL (1999) Feeding selection of heterotrophic marine nanoflagellates based on the surface hydrophobicity of their picoplankton prey. Limnology and oceanography 44: 1917-1927. Matz C, Jürgens K (2005) High motility reduces grazing mortality of planktonic bacteria. Applied and environmental microbiology 71: 921-929. Hahn M, Moore E, Höfle M (2000) Role of microcolony formation in the protistan grazing defense of the aquatic bacterium Pseudomonas sp. MWH1. Microbial ecology 39: 175-185. Shannon SP, Chrzanowski TH, Grover JP (2007) Prey food quality affects flagellate ingestion rates. Microbial ecology 53: 66-73. Pedersen AL, Nybroe O, Winding A, Ekelund F, Bjørnlund L (2009) Bacterial feeders, the nematode Caenorhabditis elegans and the flagellate Cercomonas longicauda, have different effects on outcome of competition among the Pseudomonas biocontrol strains CHA0 and DSS73. Microbial ecology 57: 501-509. Jousset A, Lara E, Wall LG, Valverde C (2006) Secondary metabolites help biocontrol strain Pseudomonas fluorescens CHA0 to escape protozoan grazing. Applied and environmental microbiology 72: 7083-7090. 45 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. Matz C, Deines P, Boenigk J, Arndt H, Eberl L, Kjelleberg S, Jürgens K (2004) Impact of violaceinproducing bacteria on survival and feeding of bacterivorous nanoflagellates. Applied and environmental microbiology 70: 1593-1599. Hahn MW, Höfle MG (2001) Grazing of protozoa and its effect on populations of aquatic bacteria. FEMS Microbiology Ecology 35: 113-121. Jürgens K, Güde H (1994) The potential importance of grazing-resistant bacteria in planktonic systems. Marine ecology progress series Oldendorf 112: 169-188. Rønn R, McCaig AE, Griffiths BS, Prosser JI (2002) Impact of protozoan grazing on bacterial community structure in soil microcosms. Applied and Environmental Microbiology 68: 6094-6105. Iriberri J, Azua I, Labirua‐Iturburu A, Artolozaga I, Barcina I (1994) Differential elimination of enteric bacteria by protists in a freshwater system. Journal of Applied Microbiology 77: 476-483. Weekers PH, Bodelier PL, Wijen JP, Vogels GD (1993) Effects of grazing by the free-living soil amoebae Acanthamoeba castellanii, Acanthamoeba polyphaga, and Hartmannella vermiformis on various bacteria. Applied and environmental microbiology 59: 2317-2319. Singh B (1941) Selectivity in bacterial food by soil amoebae in pure mixed culture and in sterilized soil. Annals of Applied Biology 28: 52-64. Groscop JA, Brent MM (1964) The effects of selected strains of pigmented microorganisms on small free-living amoebae. Canadian journal of microbiology 10: 579-584. Gurijala KR, Alexander M (1990) Effect of growth rate and hydrophobicity on bacteria surviving protozoan grazing. Applied and environmental microbiology 56: 1631-1635. Sinclair JL, Alexander M (1989) Effect of protozoan predation on relative abundance of fast-and slow-growing bacteria. Canadian Journal of microbiology 35: 578-582. Petropoulos P, Gilbride KA (2005) Nitrification in activated sludge batch reactors is linked to protozoan grazing of the bacterial population. Canadian journal of microbiology 51: 791-799. Šimek K, Pernthaler J, Weinbauer MG, Horn k K, Dolan JR, Nedoma J, Ma n M, Amann R (2001) Changes in bacterial community composition and dynamics and viral mortality rates associated with enhanced flagellate grazing in a mesoeutrophic reservoir. Applied and environmental microbiology 67: 2723-2733. Fenchel T (1987) Ecology of protozoa: the biology of free-living phagotrophic protists. . In: Brock TD (ed) Brock/Springer series in contemporary bioscience. Science and Technology, Madison, WI. Erken M, Farrenschon N, Speckmann S, Arndt H, Weitere M (2012) Quantification of individual flagellate-bacteria interactions within semi-natural biofilms. Protist 163: 632-642. Weitere M, Bergfeld T, Rice SA, Matz C, Kjelleberg S (2005) Grazing resistance of Pseudomonas aeruginosa biofilms depends on type of protective mechanism, developmental stage and protozoan feeding mode. Environmental microbiology 7: 1593-1601. Arndt H, Schmidt-Denter K, Auer B, Weitere M (2003) Protozoans and biofilms. In Fossil and Recent Biofilms. Krumbein, W.E., Paterson, D.M., and Zavarzin, G.A., (eds). Dordrecht: Kluwer Academic Publishers, in press. Joubert L-M, Wolfaardt G, Botha A (2006) Microbial exopolymers link predator and prey in a model yeast biofilm system. Microbial ecology 52: 187-197. Friman V-P, Diggle SP, Buckling A (2013) Protist predation can favour cooperation within bacterial species. Biology letters 9: 20130548. Sun S, Kjelleberg S, McDougald D (2013) Relative contributions of Vibrio polysaccharide and quorum sensing to the resistance of Vibrio cholerae to predation by heterotrophic protists. PloS one 8: e56338. Wey JK, Scherwass A, Norf H, Arndt H, Weitere M (2008) Effects of protozoan grazing within river biofilms under semi-natural conditions. Aquatic Microbial Ecology 52: 283. Rychert K, Neu TR (2010) Protozoan impact on bacterial biofilm formation. Biological Letters 47: 3-10. Cole JJ (1999) Aquatic microbiology for ecosystem scientists: new and recycled paradigms in ecological microbiology. Ecosystems 2: 215-225. 46 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. Pernthaler J, Posch T, Simek K, Vrba J, Amann R, Psenner R (1997) Contrasting bacterial strategies to coexist with a flagellate predator in an experimental microbial assemblage. Applied and Environmental Microbiology 63: 596-601. Sieburth J, Davis P (1982) The role of heterotrophic nanoplankton in the grazing and nurturing of planktonic bacteria in the Sargasso and Caribbean Seas. Ann Inst Oceanogr 58: 285-296. Sherr BF, Sherr EB, Andrew TL, Fallon RD, Newell SY (1986) Trophic interactions between heterotrophic protozoa and bacterioplankton in estuarine water analyzed with selective metabolic inhibitors. Marine Ecology-Progress Series 32: 169-179. Matz C, Webb JS, Schupp PJ, Phang SY, Penesyan A, Egan S, Steinberg P, Kjelleberg S (2008) Marine biofilm bacteria evade eukaryotic predation by targeted chemical defense. PloS one 3: e2744. Zubkov MV, Sleigh MA (2000) Comparison of growth efficiencies of protozoa growing on bacteria deposited on surfaces and in suspension. Journal of Eukaryotic Microbiology 47: 62-69. Jackson S, Jones E (1991) Interactions within biofilms: the disruption of biofilm structure by protozoa. Kieler Meeresforsch 8: 264-268. Pedersen K (1982) Factors regulating microbial biofilm development in a system with slowly flowing seawater. Applied and environmental microbiology 44: 1196-1204. Wolfaardt G, Lawrence J, Headley J, Robarts R, Caldwell D (1994) Microbial exopolymers provide a mechanism for bioaccumulation of contaminants. Microbial ecology 27: 279-291. Huws S, McBain A, Gilbert P (2005) Protozoan grazing and its impact upon population dynamics in biofilm communities. Journal of applied microbiology 98: 238-244. Dopheide A, Lear G, Stott R, Lewis G (2011) Preferential feeding by the ciliates Chilodonella and Tetrahymena spp. and effects of these protozoa on bacterial biofilm structure and composition. Applied and environmental microbiology 77: 4564-4572. Brown MR, Barker J (1999) Unexplored reservoirs of pathogenic bacteria: protozoa and biofilms. Trends in microbiology 7: 46-50. Cirillo JD, Falkow S, Tompkins LS (1994) Growth of Legionella pneumophila in Acanthamoeba castellanii enhances invasion. Infection and immunity 62: 3254-3261. Cirillo JD, Falkow S, Tompkins LS, Bermudez LE (1997) Interaction of Mycobacterium avium with environmental amoebae enhances virulence. Infection and Immunity 65: 3759-3767. Barker J, Scaife H, Brown M (1995) Intraphagocytic growth induces an antibiotic-resistant phenotype of Legionella pneumophila. Antimicrobial agents and chemotherapy 39: 2684-2688. Singh PK, Schaefer AL, Parsek MR, Moninger TO, Welsh MJ, Greenberg E (2000) Quorum-sensing signals indicate that cystic fibrosis lungs are infected with bacterial biofilms. Nature 407: 762-764. Høiby N, Krogh Johansen H, Moser C, Song Z, Ciofu O, Kharazmi A (2001) Pseudomonas aeruginosa and the in vitro and in vivo biofilm mode of growth. Microbes and Infection 3: 23-35. Levin BR, Antia R (2001) Why we don't get sick: the within-host population dynamics of bacterial infections. Science 292: 1112-1115. Rittmann BE (2007) Where are we with biofilms now? Where are we going? Water science and technology 55: 1-8. Washio J, Mayanagi G, Takahashi N (2010) Challenge to Metabolomics of Oral Biofilm:—From “What Are They?” to “What Are They Doing?”—. Journal of Oral Biosciences 52: 225-232. Nicholson JK, Lindon JC (2008) Systems biology: metabonomics. Nature 455: 1054-1056. Garcia DE, Baidoo EE, Benke PI, Pingitore F, Tang YJ, Villa S, Keasling JD (2008) Separation and mass spectrometry in microbial metabolomics. Current opinion in microbiology 11: 233-239. Hollywood K, Brison DR, Goodacre R (2006) Metabolomics: current technologies and future trends. Proteomics 6: 4716-4723. Wilmes P, Bowen BP, Thomas BC, Mueller RS, Denef VJ, VerBerkmoes NC, Hettich RL, Northen TR, Banfield JF (2010) Metabolome-proteome differentiation coupled to microbial divergence. MBio 1. Tyson GW, Chapman J, Hugenholtz P, Allen EE, Ram RJ, Richardson PM, Solovyev VV, Rubin EM, Rokhsar DS, Banfield JF (2004) Community structure and metabolism through reconstruction of microbial genomes from the environment. Nature 428: 37-43. 47 214. 215. 216. 217. 218. 219. 220. Raes J, Bork P (2008) Molecular eco-systems biology: towards an understanding of community function. Nature Reviews Microbiology 6: 693-699. Ohashi Y, Hirayama A, Ishikawa T, Nakamura S, Shimizu K, Ueno Y, Tomita M, Soga T (2008) Depiction of metabolome changes in histidine-starved Escherichia coli by CE-TOFMS. Molecular BioSystems 4: 135-147. Jansson JK, Prosser JI (2013) Microbiology: The life beneath our feet. Nature 494: 40-41. Rice KC, Bayles KW (2003) Death's toolbox: examining the molecular components of bacterial programmed cell death. Molecular microbiology 50: 729-738. Watnick P, Kolter R (2000) Biofilm, city of microbes. Journal of Bacteriology 182: 2675-2679. Nikolaev YA, Plakunov V (2007) Biofilm—“City of microbes” or an analogue of multicellular organisms? Microbiology 76: 125-138. Battin TJ, Sloan WT, Kjelleberg S, Daims H, Head IM, Curtis TP, Eberl L (2007) Microbial landscapes: new paths to biofilm research. Nature Reviews Microbiology 5: 76-81. 48 49 6 Manuscripts Manuscript 1 High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCR-Based Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. 50 51 Microb Ecol DOI 10.1007/s00248-013-0315-z METHODS High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCR-Based Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions Dawei Ren & Jonas Stenløkke Madsen & Claudia I de la Cruz-Perera & Lasse Bergmark & Søren J. Sørensen & Mette Burmølle Received: 26 April 2013 / Accepted: 14 October 2013 # Springer Science+Business Media New York 2013 Abstract Multispecies biofilms are predominant in almost all natural environments, where myriads of resident microorganisms interact with each other in both synergistic and antagonistic manners. The interspecies interactions among different bacteria are, despite the ubiquity of these communities, still poorly understood. Here, we report a rapid, reproducible and sensitive approach for quantitative screening of biofilm formation by bacteria when cultivated as mono- and multispecies biofilms, based on the Nunc-TSP lid system and crystal violet staining. The relative proportion of the individual species in a four-species biofilm was assessed using quantitative PCR based on SYBR Green I fluorescence with specific primers. The results indicated strong synergistic interactions in a fourspecies biofilm model community with a more than 3-fold increase in biofilm formation and demonstrated the strong dominance of two strains, Xanthomonas retroflexus and Paenibacillus amylolyticus. The developed approach can be used as a standard procedure for evaluating interspecies interactions in defined microbial communities. This will be of significant value in the quantitative study of the microbial composition of multispecies biofilms both in natural Electronic supplementary material The online version of this article (doi:10.1007/s00248-013-0315-z) contains supplementary material, which is available to authorized users. D. Ren : J. S. Madsen : C. I. de la Cruz-Perera : L. Bergmark : S. J. Sørensen : M. Burmølle Section of Microbiology, Department of Biology, University of Copenhagen, Copenhagen, Denmark M. Burmølle (*) Section of Microbiology, University of Copenhagen, Universitetsparken 15, bygn 1, 2100 Copenhagen Ø, Denmark e-mail: [email protected] environments and infectious diseases to increase our understanding of the mechanisms that underlie cooperation, competition and fitness of individual species in mixed-species biofilms. Introduction Biofilms are defined as polymeric matrix-enclosed bacterial communities associated with surfaces or interfaces [1]. They are considered the dominant lifestyle of bacteria both in environmental ecosystems and human hosts, and typically comprise a large number of different bacteria living together [2]. Soil is an example of an environment that contains a large number of very diverse bacteria and numerous available surfaces for multispecies biofilm formation [3]. Coresidence of diverse bacteria in multispecies biofilms is likely to catalyse complex interactions, resulting in increased or decreased biofilm biomass [4–7], which may in turn affect the overall function of the biofilm community. As example, Pseudomonas aeruginosa PAO1 and Burkholderia sp. NK8 showed enhanced biofilm formation in a dual-species biofilm, directly benefitting bioremediation potential, as chlorobenzoates were more efficiently degraded [8] and similar biofilm synergies were observed in drinking water systems [7]. Despite an increase in biofilm-related studies over the past decades, most of these have focused on monospecies biofilms or specific ecological niches such as mixed-species oral biofilms. Our current knowledge regarding the prevalence, physiology and complexity of multispecies biofilm is still incomplete, partly due to the lack of reproducible screening methods. Two measurements are of particular importance when exploring interactions in biofilms; namely, the biomass (or D. Ren et al. productivity) of the total biofilm and of the individual strains. Measurements of the overall biomass can, when compared with the amount that each of the residing species are able to produce as monospecies biofilms, be used to distinguish the effect of interactions on the extent of biofilm that is formed, as to whether this is synergistic, neutral or antagonistic. Being able to differentiate the success of each different bacterial member of the biofilm can furthermore resolve interactions as being mutual, commensal or parasitic, which is fundamental for exploring and understanding the selective forces operating in and shaping multispecies communities. Several methods and protocols have been developed for studying biofilms. Flow cells combined with confocal scanning laser microscopy [9] is the most favoured tool and has the advantage of enabling one to obtain quantitative information on both the overall biomass and of the individual strains if combined with correct labelling techniques (e.g. fluorescence in situ hybridization, FISH) but is unfortunately not suited for high-throughput comparative screening studies. Furthermore, identification of specific strains in multispecies biofilms by this approach is highly labour-intensive and requires expert handling in order to avoid pitfalls such as uneven staining due to the limited probe penetration into biofilms, artefacts caused by hybridization and dehydration procedures in FISH [10, 11]. Microtiter plates are suitable for performing biofilm quantification, usually based on crystal violet retention, owing to their high-throughput screening capability and the simplicity of protocols [12]. However, the reproducibility of this assay is problematic, especially when multispecies consortia are analysed. In addition, molecular analysis of biofilm cells requires efficient and complete disruption of the cells followed by extraction of the target molecules, which still needs to be improved in multispecies biofilms. In this study, we report an easily applicable and reproducible approach for consistently quantifying multispecies biofilm formation and to evaluate interactions, in regard to the overall biofilm formation and relative proportions of individual species. We present an optimized DNA extraction protocol and SYBR Green-based quantitative PCR (qPCR) assay for the selective, rapid and sensitive detection of four species in multispecies biofilm. The reproducibility and broad applicability of this specific detection procedure makes this method useful for most types of defined biofilms, not limited to the soil isolates used in this study. Materials and Methods Soil Isolates and Culture Conditions The bacterial strains used in this study (Table 1) were obtained from agricultural soil as described previously [13]. From the total strain pool isolated by de la Cruz-Perera et al. [13], we selected seven strains based on growth compatibility (see below). Two of the selected strains (6 and 7) were not described by de la Cruz-Perera et al., but these were isolated and identified by procedures identical to those referred to above. Optimization of Growth Media To determine the optimal growth conditions and evaluate the biofilm-forming capabilities of the seven selected soil isolates, each strain was grown individually in Minimal Medium (basal medium 500 mL: NaH2PO4 ·H2O 0.5 g, K2HPO4 ·3H2O 2.125 g, NH 4Cl 1.0 g, pH 7.2; trace metals 500 mL: nitrilotriacetic acid 0.0615 g, MgSO4 ·7H2O 0.1 g, FeSO4 · 7H 2 O 0.006 g, ZnSO 4 · 7H 2 O 0.0015 g, MnSO 4 · H 2 O 0.0015 g, pH 7.0) supplemented with 0.2 % D (+)-Glucose, nutrient-low R2B medium (yeast extract 0.5 g, proteose peptone 0.5 g, casamino acids 0.5 g, glucose 0.5 g, soluble starch 0.5 g, sodium pyruvate 0.3 g, K2HPO4 0.3 g, MgSO4 ·7H2O 0.05 g, in 1 L distilled water) and nutrient-rich TSB medium (tryptic soy broth, Merck KgaA, Germany). When solid medium was needed, 1.5 % (wt/vol) agar was added. The strains were subcultured from frozen glycerol stocks onto tryptic soy agar (TSA) plates for 48 h at 24 °C, and then colonies were transferred onto Minimal Medium agar plates containing 0.2 % glucose, R2B agar plates or TSA plates. Colonies from solid type media were inoculated into 5 mL liquid media of the same type and incubated with shaking (250 rpm) at 24 °C overnight. Cultivation of Biofilms Both 96-well cell culture plates (cat. no. 655 180, Greiner BioOne, Germany) and Nunc-TSP lid system (cat. no. 445497, Thermo Scientific, Denmark), which comprises a 96-well plate lid with pegs that extend into each well, were used to cultivate biofilms. The seven selected strains were screened Table 1 Identification of the seven soil isolates by 16S rRNA analysis Strain no.a GenBank accession no. Closest relativeb 1 2 3 4 5 6 7 JQ890536 JQ890538 JQ890537 JQ890542 JQ890539 JQ890541 JQ890540 Pseudomonas lutea Stenotrophomonas rhizophila Xanthomonas retroflexus Ochrobactrum rhizosphaerae Microbacterium oxydans Arthrobacter nitroguajacolicus Paenibacillus amylolyticus a The numbers 1 to 7 were designated to the seven strains for simplification b The sequences had 98 to 100 % base identity to the closest relative in GenBank High-Throughput Screening and Species Abundance Analysis in Multispecies Biofilms for biofilm formation as single species and in four-species combinations as described below. The colonies (grown on agar plates for 24 h) were inoculated into 5 mL of TSB medium and incubated overnight at 24 °C with shaking (250 rpm). One hundred to 600 μL of these stationary phase bacterial cultures were transferred to fresh TSB medium and grown until an optical density at 600 nm (OD600) of ~1.0 was reached. The cell suspensions were then adjusted to an OD600 of 0.15 by dilution in TSB medium. A total of 160 μL of monospecies or four mixed species (40 μL of each species) exponentially growing cultures were added to each well. To some wells, the same volume of fresh TSB medium was added to obtain a background value, which was subtracted from values obtained from the wells containing cells. The plates were sealed with Parafilm and incubated with shaking (200 rpm) at 24 °C for 24 h. The biofilm assays were performed three times on different days (biological replicates) with four technical replicates every time. Quantification of Biofilm Formation by Use of CV and TTC A modified version of the crystal violet (CV) method for detection of biofilms using 96-well cell culture plate [12] was applied as previously described. Additionally, the assay was further modified for quantifying biofilms formed on pegs of the Nunc-TSP lid system, previously referred to as the Calgary method [14]. After 24 h incubation, in order to wash off planktonic cells, the peg lid was transferred successively to three microtiter plates containing 200 μL phosphate buffered saline per well, followed by staining of the biofilms formed on the pegs with 180 μL of an aqueous 1 % (w/v) CV solution. After 20 min of staining, the lid was rinsed again three times and then placed into a new microtiter plate with 200 μL of 96 % ethanol in each well. After allowing 30 min for the stain to dissolve into the ethanol, the absorbance of CV at 590 nm was determined by using an EL 340 BioKinetics reader (BioTek Instruments, Winooski, Vt.). The CV-ethanol suspension was diluted with 96 % ethanol when the OD590 was above 1.1. Parallel with the CV-based biofilm assays described above, an alternative method, based on 2, 3, 5-triphenyltetreazolium chloride (TTC) reduction, was also implemented and evaluated. After 24-h incubation, the pegs with attached cells were rinsed three times as described above. Thereafter, the peg lid was mounted on a new microtiter plate with 200 μL fresh medium containing 0.01 % TTC. The plate was then sealed with Parafilm, wrapped in foil and incubated with shaking (200 rpm) for another 24 h. In order to resuspend the formed formazan, the peg lid was transferred to a new microtiter plate containing 200 μL of 96 % ethanol per well. Finally, the absorbance was measured at 490 nm. The statistical analyses were conducted using ANOVA test (SPSS version 17.0 for Windows). A p value of ≤0.05 was regarded as a statistically significant difference. DNA Extraction from Four-Species Biofilms The 24-h biofilms formed on peg lids were rinsed three times with PBS to remove non-adherent bacteria as described above. Then pegs were broken from the lid from the back (i.e. without direct contact to the biofilm-covered part of the peg) using sterile forceps and transferred into Lysing Matrix E tubes (provided by the FastDNA™ SPIN Kit for soil). Each peg was placed in one tube. Aliquots of 882 µl of sodium phosphate buffer was added to each tube and biofilms were disrupted from pegs by bead beating using the Savant FastPrep FP120 for 40s at setting 6.0. The pegs were removed and stained with crystal violet to verify that after bead beating, all the cells were detached from the pegs. Ninety-eight microlitres of lysozyme solution (20 mg/ml), dissolved in sodium phosphate buffer was added, and the samples were incubated for 1 h at 37 °C. Next, 122 μL of MT buffer was added to each sample, followed by bead beating twice for 40s at setting 6.0 with cooling on ice during the short time interval. The complete cell lysis was visually confirmed with an optical microscope. Genomic DNA was then extracted using FastDNA™ SPIN Kit for soil (Qbiogene, Illkirch, France) according to the manufacturer's instructions and quantified with a Qubit fluorometer (Invitrogen, Carlsbad, CA, USA). DNA Sequence Analysis and Design of the Species-Specific Primers Multiple alignment of sequences from the four strains 2, 3, 5 and 7 was done using the DNAMAN software (version 7, Lynnon corporation). The species-specific primer pairs for SYBR Green qPCR were designed manually based on the variable regions of 16S rRNA genes (Table 2) according to the guidelines set by Primer Express (version 3.0) from Applied Biosystems with an approximate maximum amplicon size of 300 bp. The obtained best fitting primers were checked by DNAMAN to avoid hairpins or primer–dimer formations. The specificity of each pair of primers was checked against the other three, non-target strains using conventional PCRs as follows. Reaction mixtures of 25 μL contained 16 μL H2O, 1 μL genomic DNA, 5 μL 5× Phusion HF Buffer, 0.5 μL 10 mM dNTPs, 1 μL 10 μM of each primer, and 0.5 μL Phusion DNA polymerase (Phusion high-fidelity PCR kit; Finnzymes, Espoo, Finland). Amplifications were performed with the following cycling protocol: 5 min at 95 °C, followed by 30 cycles of 30s at 94 °C, 30s at 61 °C/62 °C, 20s at 72 °C and a final elongation step of 5 min at 72 °C in DNA Engine Dyad Peltier Thermal Cycler (MJ Research Inc.). The D. Ren et al. Table 2 Species-specific primers based on 16S rRNA gene sequences Strain no. Identity Primer position Sequence (5′–3′) Length (bp) Product size (bp) 2 S. rhizophila X. retroflexus 5 M. oxydans 7 P. amylolyticus GCCTTGCGCGGATAGATG CGGGTATTAGCCGACTGCTT GCCTTGCGCGATTGAATG CCGTCATCCCAACCAGGTATT TCAACTCTTTGGACACTCGTAAACA CATGCGTGAAGCCCAAGAC GATACCCTTGGTGCCGAAGTT CGGTCAGAGGGATGTCAAGAC 18 20 18 21 25 19 21 21 240 3 FP 159 RP 399 FP 135 RP 387 FP 935 RP 1148 FP 800 RP 945 252 213 145 The numbers in the primer position show the positions of the target sites of each species FP Forward primer, RP Reverse primer amplified products were separated by 1.0 % agarose gel electrophoresis, stained with ethidium bromide and photographed under UV illumination. 20s/30s, followed by a standard melting/dissociation curve segment. Each sample was run in triplicates and a negative control was included in every run. Plasmid Standards Used for Absolute Quantification Results A plasmid standard containing the target region was prepared for each primer pair as follows. The 16S rRNA gene fragments were amplified by conventional PCR using the corresponding primers as mentioned above. The products were cloned using the TOPO TA cloning kit (Invitrogen, Carlsbad, CA, USA). Plasmids were isolated with QIAprep Spin Miniprep Kit (Qiagen Gmbh, Hilden, Germany). The qualities were evaluated by agarose gel electrophoresis and concentrations were measured by Qubit Fluorometer (Invitrogen, Carlsbad, CA, USA). 16S rRNA gene copy numbers were calculated assuming that the average molecular mass of a double-stranded DNA molecule is 660 g/mol. Four standard curves were generated using triplicate 5-fold dilutions of plasmid DNAs, and the corresponding slope was used to calculate PCR amplification efficiency (E) according to the equation of E =10(−1 slope). 16S rRNA gene copy numbers in unknown samples were then determined by interpolation from the standard curve using their respective threshold cycle (Ct) values. The Ct value represents the number of PCR amplification cycles needed to produce fluorescence intensity above a pre-defined threshold. Quantitative PCR Based on SYBR Green I Fluorescence For each sample, four separate qPCR reactions were performed with the Mx3000PTM Real-time PCR System (Stratagene, USA) using each pair of species-specific primers. Each of the reaction components per 20 μL were 7 μL H2O, 10 μL 2× Brilliant III Ultra-Fast SYBR® Green QPCR Master Mix (Agilent Technologies), 1 μL 10 μM of each primer and 1 μL genomic DNA (either standard or sample). The PCR programmes were carried out as follows: 95 °C for 10 min, 40 cycles of 95 °C for 30s, 59–62 °C for 30s and 72 °C for Isolates and Media Optimization The seven selected strains, representing common soil bacteria, were all able to grow on Minimal Medium supplemented with 0.2 % D (+)-Glucose (data not shown). These strains were selected for further media optimization and multispecies biofilm formation. The strain identities and the accession numbers of the 16S rRNA gene sequences, used in this study for designing specific primers (strain 2, 3, 5 and 7), are shown in Table 1. All seven strains showed slower growth on Minimal Medium+Glucose (0.2 %) and R2A plates than on TSA plates and took much more time to reach exponential growth phase in liquid media at 24 °C (data not shown). Since all seven strains grew well on TSA/TSB, this medium was chosen for further cultivation of mono- and four-species biofilms. Assay for Biofilm Formation Monospecies biofilm formation was assayed both with and without the peg lid system using both the CV retention and the TTC reduction assays (Fig. 1a–c). To evaluate the reproducibility within technical replicates and between individual experimental series (biological replicates), all biofilm assays were performed in four replicates and conducted three individual times on different days under the same conditions. Compared with the 96-well cell culture plate, the NuncTSP lid system gave less variation between replicates and between individual days, which were directly reflected in the biofilm formation of strains 1, 2 and 4. These three strains, which were identified as poor biofilm formers in the NuncTSP lid system (Fig. 1b), showed significantly different High-Throughput Screening and Species Abundance Analysis in Multispecies Biofilms 5 day1 day2 day3 Biofilm formation (OD 590) 4 3 2 1 0 1 2 3 4 5 6 7 5 6 7 5 6 7 Isolates (b) 7.0 6.0 Biofilm formation (OD 590) abilities for biofilm formation from day to day (P <0.05) when using the 96-well cell culture plate (Fig. 1a). Another difference between the assays conduced in the two types of microtiter plates was the amount of biofilm formed by strain 6, which displayed biofilm formation in the 96-well cell culture plate, but was incapable of attaching to the pegs. The NuncTSP lid system could not only provide more reproducible results, but it also removed the possibility that aggregation may be linked to sedimentation of the microorganisms in test wells [15], which might have been the case for strain 6 (subpanels a vs. b of Fig. 1). Based on these characteristics, the Nunc-TSP lid system was chosen as a better device for high-throughput screening for synergistic interactions in multispecies biofilm formation. In addition to the CV method, which quantifies total attached material, we also used the TTC method, which evaluates cell activity. Respiring cells reduce the colourless TTC solution to the red insoluble formazan, which can be dissolved in 96 % ethanol and measured spectrophotometrically. The data shown in Fig. 1c represents the TTC absorbance at 490 nm. Low day-to-day variation was observed, but the low output signals resulted in less distinct difference between biofilm formers and non-formers. In addition, some absorbance measurements were close to the resolution limit of the Bioreader. Therefore, the TTC reduction assay was not applied for the following mixed-species biofilm formation. The results obtained by the Nunc-TSP lid system showed that only strain 3 was able to establish biofilm on its own; CV values obtained for this strain were more than 10-fold higher than for any of the other strains (Fig. 1b). Based on these results, strain 3 was identified as being a good biofilm former, whereas the remaining six strains were characterized as poor biofilm formers. (a) day1 day2 day3 5.0 4.0 3.0 0.4 0.2 0.0 1 2 3 4 Isolates (c) 0.25 day1 day2 day3 0.20 Biofilm formation (OD 590) Fig. 1 Biofilm formation in 96-well cell culture plate (a) and Nunc-TSP lid system (b, c) by the seven isolates. After 24 h of incubation, the biofilm formation was quantified by staining with crystal violet followed by absorbance measurements at 590 nm (a, b) or the presence of reduced TTC was quantified by measuring the absorbance at 490 (c). Three parallel bars represent means ± standard error for four replicates on three different days representing three biological replicates (day 1, day 2 and day 3). 1, Pseudomonas lutea; 2, Stenotrophomonas rhizophila; 3, X. retroflexus; 4, Ochrobactrum rhizosphaerae; 5, Microbacterium oxydans; 6, Arthrobacter nitroguajacolicus; 7, P. amylolyticus 0.15 0.10 0.05 0.00 Biofilm Formation by Single-Species and Four-Species Consortia The seven selected strains were screened for biofilm formation, by using the Nunc-TSP lid system, as single species and in all possible combinations of four, in order to identify a fourspecies model consortium, suitable for studies of interspecies interactions. As shown in Fig. 2, the combination of strains 2, 3, 5 and 7 gave no significant variation between replicates and 1 2 3 4 Isolates between individual days, indicating the high reproducibility of the assay. Obviously, with strain 3 as exception, the strains showed weak ability to form biofilm. However, when the four isolates coexisted in the biofilm, the biofilm biomass increased by >300 % compared to the single-species biofilms, D. Ren et al. the standard curves ranged from 3×106 to 3×107 copies/μL and were used in 5-fold dilution series. The DNA extracted from biofilms was diluted appropriately to ensure that all the unknown samples were within the range of the standard curves. 16S rRNA gene copy number of each species was calculated and is shown in Table 3. As the exact copy numbers of the 16S rRNA gene per cell in the four species are currently unknown, the cell numbers were estimated based on other species in the same genus [16–19]. 24 Biofilm formation (OD 590) 21 day 1 day 2 day 3 18 15 12 9 6 3 0 2 3 5 7 2357 Discussion Isolates Fig. 2 Biofilms formed by single-species and four-species communities (2, Stenotrophomonas rhizophila; 3, X. retroflexus; 5, M. oxydans; and 7, P. amylolyticus) in Nunc-TSP lid system. After 24 h of incubation, the biofilm formation was quantified by staining with crystal violet. Three parallel bars represent the average absorbance at OD 590 nm for four replicates on three different days representing three biological replicates (day 1, day 2 and day 3). The bars represent means ± standard error for four replicates indicating a strong synergy in this multispecies biofilm. This four-species consortium was therefore selected as a multispecies biofilm model for further studies of the species dynamics. Quantification of Four Strains by Quantitative PCR in Multispecies Biofilm DNA was successfully extracted from the biofilms by excision of entire pegs followed by cell disruption. The biofilms were completely removed from the pegs (confirmed by CV staining, data not shown), and complete cell lysis was confirmed by microscopy. The amount of DNA extracted from multispecies biofilms composed of strains 2, 3, 5 and 7 was 1.9– 2.1 μg/peg. Based on the variable regions of 16S rRNA genes, four pairs of specific primers were designed for SYBR Green qPCR (Table 2). For simplification and to enable quantification based on SYBR Green, the primers were designed for application in four separate reactions for each sample. The specificity of the primers was confirmed by conventional PCRs, verifying that primers were strictly species specific. The high linearity of the Ct values plotted in the standard curves was verified by the correlation coefficient (RSq) values of >0.99. The amplification efficiencies (E) ranged from 80 to 90 % (Supplementary Fig. 1). No detectable peaks that were associated with primer–dimer or other non-specific PCR products were observed in the melting curves (data not shown), and only the single bands of the expected size amplicon in each qPCR assay were detectable by agarose gel electrophoresis (data not shown). The standard plasmid DNA used for Multispecies biofilms are prevalent in almost all environments. A pressing need, therefore, exists to better the understanding of the social interactions and selective forces that drive bacterial communities in multispecies biofilm. During the past few decades, simultaneous staining of multiple species by FISH has been widely used in combination with confocal laser scanning microscopy (CLSM) for species differentiation in oral biofilms [20–22]. This approach is, however, not easily applicable to many different isolates from various environments and it is only partly quantitative. In this study, we have developed an approach to consistently quantify biofilm formation, which enables high-throughput screening of the prevalence of synergistic interactions and assessment of the proportions of individual bacterial species in biofilms. The general procedure of this approach is illustrated in Fig. 3 which can be used as a standard procedure for evaluating Table 3 The copy numbers of 16S rRNA gene and the estimated cell numbers derived from four separate qPCRs Strain Identity no. 16S rRNA Estimated copy Estimated gene copies number of the number of (per μL) 16S rRNA gene cells (per μL) (per cell) 2 3 5 7 7.91E+006 1.98E+009 8.89E+006 9.95E+008 S. rhizophila X. retroflexus M. oxydans P. amylolyticus 4a 2b 2c 12d 1.98E+006 9.90E+008 4.45E+006 8.29E+007 a The copy number was estimated to be 4, as Stenotrophomonas maltophilia is known to have four copies [17] b The copy number of the 16S rRNA gene in X. retroflexus is estimated to be 2, as Xanthomonas axonopodis , Xanthomonas campestris , Xanthomonas oryzae, Xanthomonas citri, Xanthomonas albilineans are known to have two copies, respectively [18] c The copy number of the 16S rRNA gene in M. oxydans is estimated to be 2, as Microbacterium testaceum is known to have two copies [19] d The copy number of the 16S rRNA gene in P. amylolyticus is estimated to be 12, as other Paenibacillus species are known to have an average 12 copies [16] High-Throughput Screening and Species Abundance Analysis in Multispecies Biofilms Strain isolation Media optimization DNA extraction Mono- and multi-species biofilm cultivation 16S rRNA gene sequencing Target selection DNA extraction Species-specific primer design Quantitative PCR Fig. 3 General procedure for high-throughput screening and evaluation of interactions in multispecies biofilms. The right-hand part illustrates the parallel strain identification and primer design for use in qPCR interspecies interaction in multispecies biofilms. The righthand part of this figure illustrates the parallel strain identification and primer design. Clearly, this protocol is applicable for multispecies biofilms composed of a defined number of known species. Other approaches, including metagenomic analysis, are more suitable when interactions in more complex communities are explored. However, the standard method presented in this study could be much useful in some natural settings where species diversity is more restricted e.g. chronic infections or when limited key species are the research focus. Multispecies Biofilm Assay Mixed species may facilitate synergistic or antagonistic interactions between consortia members in biofilms. The coculturing of four strains in this study provided an easy way to detect changes in biofilm formation from mono- to multispecies biofilms and allowed high-throughput screening of many isolates with high reproducibility. Biofilms have been proposed to exist in a fine balance between competition and cooperation [23], which can be tipped by various types of influences such as surrounding environmental factors and quorum sensing-dependent gene regulation. Inconsistent results in biofilm formation assays have previous been reported [24–27] due to different substrates, media, inoculum size and culture conditions, and the inconsistency is most likely to be worsened by the highly heterogeneous nature of multispecies biofilms. In this study, the three independent repetitive runs and four technical replicates per run were performed to assess the reproducibility of the quantitative analysis. Poor reproducibility of the microtiter plate assay (reflected by higher standard errors (Fig. 1a)) can be caused by pipetting uncertainties, pieces of biofilm detaching, etc. In the Nunc-TSP lid system, the number of pipetting steps is dramatically reduced, which decreases the handling-induced variability. This was best reflected by the lower standard errors, indicative of relatively uniform biofilms throughout the four replicate wells (Fig.1b, c and Fig. 2). TTC reduction, as a simple colorimetric method, has been widely used to evaluate the cell activity in plant tissue [28], fungal spores, yeast and bacteria cultures [29]. The low TTC values observed in this study were probably due to the relatively low metabolic activities of cells and the enhanced levels of non-biological material within mature biofilm compared to the CV assay where the extracellular matrix and all bacterial cells are stained. The combined use of CV and the respiratory indicator CTC (5-cyano-2,3-ditolyl tetrazolium chloride) in high-throughput biofilm assays was previously reported by Pitts et al., who also found cells that had lost metabolic activity could still contribute to the total amount of biomass [30]. Therefore, in this study, TTC was less suitable for quantitative biofilm measurements of microbial biomass. DNA Extraction and qPCR Analysis of Species Distribution Comparative studies of gene and protein expression of biofilm-associated cells have proven that these are significantly different from planktonically grown cells [31, 32]. The protocol optimized in the present study successfully detached and lysed both the Gram-negative and Gram-positive cells and is likely to become useful in many applications of DNA-based biofilm research. With the appropriate modification, the protocol is additionally suitable for RNA and protein extractions. qPCR has been successfully applied for quantifying bacterial abundance in plaque biofilm [33], faucet biofilm [34] and biofilms in wounds [35] due to its speed, sensitivity and specificity. The targets of qPCR can be species-specific genes or 16S ribosome RNA genes [36, 37]. 16S rRNA gene proved to be an excellent target for both quantitative broad-range PCR [38] and group-/species-specific PCR [39, 40] in complex communities. The qPCR assay offers several advantages over FISH: low risk of contamination by amplified products, the simplicity and rapidity of data analysis and low detection limit. Using the real-time TaqMan assay, Price et al. [41] were able to examine specific bacterial populations in biofilms grown from human saliva and their susceptibility to chlorhexidine. Ren et al. reported a consistent function-related bacterial distribution in anode biofilms using both FISH and qPCR analyses [42]. In the present study, we describe a specific qPCR assay to examine the population dynamics in multispecies biofilms. D. Ren et al. The significant synergistic interaction observed in a biofilm consisting of four soil bacteria make this consortium a powerful model to study development and interactions in multispecies biofilms. For this work, 16S rRNA gene was targeted for SYBR Green assay profiling of the four strains run in separate reactions. The hypervariable regions interspersed with the conserved regions make 16S rRNA gene an attractive target for both universal and specific primers. In addition, the public databases of 16S rRNA gene sequences are easily accessible, including GenBank, Greengenes and Ribosomal Database Project, which are valuable for bacterial identification and investigation in microbial ecology and evolution. TaqMan and SYBR Green qPCR are two frequently used assays. Despite of the significantly high specificity of the detection with TaqMan probe, SYBR Green qPCR is widely used due to the low cost and the ease in designing primers and optimizing assays. Maeda et al. [43] have reported that there are no significant differences between the TaqMan and SYBR Green chemistry in their specificity and sensitivity. However, the effect of SYBR Green qPCR is also limited in terms of the number of species it is practical to analyse. In such cases, the use of TaqMan would be preferred in a complex community to enable more species to be analysed, as it opens up for multiplex qPCR, which only requires one specific probe per target (instead of two specific primers) and reduces the amount of samples being run. Concluding Remarks Overall, we present a sensitive and reliable high-throughput method to investigate the interspecific interactions between bacterial isolates with the static co-culture assay followed by a qPCR assay that uses species-specific primers to measure the 16S rRNA gene numbers of each species in multispecies biofilms. By quantification and comparisons of the biomass of each species when grown alone and in the multispecies biofilm, understanding of the interspecific interactions (cooperative, mutualistic, competitive) that operate in the multispecies biofilm, is obtainable. To our knowledge, this presented approach applied in screenings for overall synergism and antagonism within multispecies biofilms composed of soil isolates is firstly reported in this study. As the ubiquity of biofilm formation is receiving increasingly more attention among scientists, this developed method will be a valuable tool for studying the social interactions and selective forces that drive complex bacterial communities in natural environments as well as infectious diseases. Acknowledgments This research was supported by funding to Mette Burmølle by The Danish Council for Independent Research, Technology and Production, ref no: 09-090701, to Søren Sørensen by The Danish Council for Independent Research, Natural Sciences and the Danish Innovation Consortium, SiB, ref no: 11804520 to Jonas Stenløkke Madsen. References 1. Costerton JW, Lewandowski Z, Caldwell DE, Korber DR, LappinScott HM (1995) Microbial biofilms. Annual Reviews in Microbiology 49(1):711–745 2. Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2(2):95–108 3. Burmølle M, Kjøller A, Sørensen SJ (2012) An invisible workforce: biofilms in the soil. In: Lear G, Lewis G (eds) Microbial biofilms: current research and applications. Caister Academic press, Norfolk, UK 4. Burmølle M, Webb JS, Rao D, Hansen LH, Sørensen SJ, Kjelleberg S (2006) Enhanced biofilm formation and increased resistance to antimicrobial agents and bacterial invasion are caused by synergistic interactions in multispecies biofilms. Appl Environ Microb 72(6): 3916–3923 5. Pathak AK, Sharma S, Shrivastva P (2012) Multi-species biofilm of Candida albicans and non-Candida albicans Candida species on acrylic substrate. Journal of Applied Oral Science 20(1):70–75 6. Wen ZT, Yates D, Ahn SJ, Burne RA (2010) Biofilm formation and virulence expression by Streptococcus mutans are altered when grown in dual-species model. BMC microbiology 10(1):111 7. Schwering M, Song J, Louie M, Turner RJ, Ceri H (2013) Multispecies biofilms defined from drinking water microorganisms provide increased protection against chlorine disinfection. Biofouling (ahead-of-print):1-12 8. Yoshida S, Ogawa N, Fujii T, Tsushima S (2009) Enhanced biofilm formation and 3-chlorobenzoate degrading activity by the bacterial consortium of Burkholderia sp. NK8 and Pseudomonas aeruginosa PAO1. J Appl Microbiol 106(3):790–800 9. Tolker-Nielsen T, Brinch UC, Ragas PC, Andersen JB, Jacobsen CS, Molin S (2000) Development and dynamics of pseudomonas sp. Biofilms. J Bacteriol 182(22):6482–6489 10. Evans RP, Nelson CL, Bowen WR, Kleve MG, Hickmon SG (1998) Visualization of bacterial glycocalyx with a scanning electron microscope. Clin Orthop Relat Res 347:243 11. Morck D, Raybould T, Acres S, Babiuk L, Nelligan J, Costerton J (1987) Electron microscopic description of glycocalyx and fimbriae on the surface of Pasteurella haemolytica-A1. Can J Vet Res 51(1): 83 12. O'Toole GA, Kolter R (1998) Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic analysis. Mol Microbiol 28(3): 449–461 13. de la Cruz-Perera CI, Ren D, Blanchet M, Dendooven L, Marsch R, Sørensen SJ, Burmølle M (2013) The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains. FEMS Microbiol Lett 338(1):95–100 14. Ceri H, Olson M, Stremick C, Read R, Morck D, Buret A (1999) The Calgary biofilm device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms. J Clin Microbiol 37(6):1771–1776 15. Harrison JJ, Stremick CA, Turner RJ, Allan ND, Olson ME, Ceri H (2010) Microtiter susceptibility testing of microbes growing on peg lids: a miniaturized biofilm model for high-throughput screening. Nature protocols 5(7):1236–1254 16. Shin SH, Kim S, Kim JY, Song HY, Cho SJ, Lee K-I, Lim HK, Park N-J, Hwang IT, Yang K-S (2012) Genome sequence of Paenibacillus terrae HPL-003, a xylanase-producing bacterium isolated from soil found in forest residue. J Bacteriol 194(5):1266–1266 17. Lira F, Hernández A, Belda E, Sánchez MB, Moya A, Silva FJ, Martínez JL (2012) Whole-genome sequence of Stenotrophomonas maltophilia D457, a clinical isolate and a model strain. J Bacteriol 194(13):3563–3564 High-Throughput Screening and Species Abundance Analysis in Multispecies Biofilms 18. da Silva AR, Ferro JA, Reinach F, Farah C, Furlan L, Quaggio R, Monteiro-Vitorello C, Van Sluys M, Almeida N, Alves L (2002) Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417(6887):459–463 19. Morohoshi T, Wang W-Z, Someya N, Ikeda T (2011) Genome sequence of Microbacterium testaceum StLB037, an N acylhomoserine lactone-degrading bacterium isolated from potato leaves. J Bacteriol 193(8):2072–2073 20. Al-Ahmad A, Wunder A, Auschill TM, Follo M, Braun G, Hellwig E, Arweiler NB (2007) The in vivo dynamics of Streptococcus spp., Actinomyces naeslundii, Fusobacterium nucleatum and Veillonella spp. in dental plaque biofilm as analysed by five-colour multiplex fluorescence in situ hybridization. J Med Microbiol 56(5):681–687 21. Dige I, Nilsson H, Kilian M, Nyvad B (2007) In situ identification of streptococci and other bacteria in initial dental biofilm by confocal laser scanning microscopy and fluorescence in situ hybridization. Eur J Oral Sci 115(6):459–467 22. Thurnheer T, Gmür R, Guggenheim B (2004) Multiplex FISH analysis of a six-species bacterial biofilm. J Microbiol Meth 56(1):37–47 23. Nadell CD, Xavier JB, Foster KR (2009) The sociobiology of biofilms. FEMS Microbiol Rev 33(1):206–224 24. Bechet M, Blondeau R (2003) Factors associated with the adherence and biofilm formation by Aeromonas caviae on glass surfaces. J Appl Microbiol 94(6):1072–1078 25. Dag I, Kiraz N, Yasemin O (2010) Evaluation of different detection methods of biofilm formation in clinical Candida isolates. Afr J Microbiol Res 4(24):2763–2768 26. Nett JE, Cain MT, Crawford K, Andes DR (2011) Optimizing a Candida biofilm microtiter plate model for measurement of antifungal susceptibility by tetrazolium salt assay. J Clin Microbiol 49(4): 1426–1433 27. Zara S, Gross MK, Zara G, Budroni M, Bakalinsky AT (2010) Ethanol-independent biofilm formation by a flor wine yeast strain of Saccharomyces cerevisiae. Appl Environ Microb 76(12):4089– 4091 28. Gupta S, Kaur S, Sehgal S, Sharma A, Chhuneja P, Bains NS (2010) Genotypic variation for cellular thermotolerance in Aegilops tauschii Coss., the D genome progenitor of wheat. Euphytica 175(3):373–381 29. Savaroglu F, Filik I, Oztopcu V, Kabadere S, Ilhan S, Uyar R (2011) Determination of antimicrobial and antiproliferative activities of the aquatic moss Fontinalis antipyretica Hedw. Turk J Biol 35:361–369 30. Pitts B, Hamilton MA, Zelver N, Stewart PS (2003) A microtiterplate screening method for biofilm disinfection and removal. J Microbiol Meth 54(2):269–276 31. Moreno-Paz M, Gómez MJ, Arcas A, Parro V (2010) Environmental transcriptome analysis reveals physiological differences between biofilm and planktonic modes of life of the iron oxidizing bacteria Leptospirillum spp. in their natural microbial community. BMC genomics 11(1):404 32. Resch A, Leicht S, Saric M, Pásztor L, Jakob A, Götz F, Nordheim A (2006) Comparative proteome analysis of Staphylococcus aureus biofilm and planktonic cells and correlation with transcriptome profiling. Proteomics 6(6):1867–1877 33. Kirakodu S, Govindaswami M, Novak M, Ebersole J, Novak K (2008) Optimizing qPCR for the quantification of periodontal pathogens in a complex plaque biofilm. The open dentistry journal 2:49 34. Beumer A, King D, Donohue M, Mistry J, Covert T, Pfaller S (2010) Detection of Mycobacterium avium subsp. paratuberculosis in drinking water and biofilms by quantitative PCR. Appl Environ Microb 76(21):7367–7370 35. Dalton T, Dowd SE, Wolcott RD, Sun Y, Watters C, Griswold JA, Rumbaugh KP (2011) An in vivo polymicrobial biofilm wound infection model to study interspecies interactions. PloS one 6(11): e27317 36. Suzuki N, Nakano Y, Yoshida A, Yamashita Y, Kiyoura Y (2004) Real-time TaqMan PCR for quantifying oral bacteria during biofilm formation. J Clin Microbiol 42(8):3827–3830 37. Tsushima I, Kindaichi T, Okabe S (2007) Quantification of anaerobic ammonium-oxidizing bacteria in enrichment cultures by real-time PCR. Water Res 41(4):785–794 38. Nadkarni MA, Martin FE, Jacques NA, Hunter N (2002) Determination of bacterial load by real-time PCR using a broadrange (universal) probe and primers set. Microbiology 148(1):257– 266 39. Matsuki T, Watanabe K, Tanaka R, Fukuda M, Oyaizu H (1999) Distribution of bifidobacterial species in human intestinal microflora examined with 16S rRNA-gene-targeted species-specific primers. Appl Environ Microb 65(10):4506–4512 40. Matsuki T, Watanabe K, Fujimoto J, Takada T, Tanaka R (2004) Use of 16S rRNA gene-targeted group-specific primers for real-time PCR analysis of predominant bacteria in human feces. Appl Environ Microbiol 70(12):7220–7228 41. Price R, Viscount H, Stanley M, Leung KP (2007) Targeted profiling of oral bacteria in human saliva and in vitro biofilms with quantitative real-time PCR. Biofouling 23(3):203–213 42. Ren Z, Steinberg L, Regan J (2008) Electricity production and microbial biofilm characterization in cellulose-fed microbial fuel cells. Water Sci Technol 58(3):617 43. Maeda H, Fujimoto C, Haruki Y, Maeda T, Kokeguchi S, Petelin M, Arai H, Tanimoto I, Nishimura F, Takashiba S (2006) Quantitative real-time PCR using TaqMan and SYBR Green for Actinobacillus actinomycetemcomitans , Porphyromonas gingivalis , Prevotella intermedia, tetQ gene and total bacteria. FEMS Immunology & Medical Microbiology 39(1):81–86 Manuscript 2 High prevalence of biofilm synergy among bacterial soil isolates in co-cultures indicates bacterial interspecific cooperation. High prevalence of biofilm synergy among bacterial soil isolates in co-cultures indicates bacterial interspecific cooperation Dawei Ren, Jonas Stenløkke Madsen, Søren J. Sørensen*, Mette Burmølle* Section of Microbiology, Department of Biology, University of Copenhagen *Corresponding authors Postal address: Universitetsparken 15, Bygn. 1, 2100 København Ø, Denmark e-mail Mette Burmølle: [email protected] e-mail Søren J. Sørensen: [email protected] Running title: High prevalence of synergy in multispecies biofilms Key words: Cooperation / Multi-species biofilm / Social evolution / Soil bacteria / Synergy Abstract Biofilms that form on roots, litter and soil particles typically contain multiple bacterial species. Currently, little is known about the interactions transpiring in these multi-species biofilms as only few studies have been based on environmental isolates. In this study, we assessed the prevalence of synergism and antagonism in biofilm formation among seven soil-isolates, which were co-cultured in combinations of four species. In 63% of the four-species biofilms, an increase of biofilm biomass was observed when compared to those formed by single-species. In contrast, only 6% of the multispecies biofilms displayed a decreasing trend, thus demonstrating a high prevalence of synergism in multispecies biofilm formation. One four-species consortium, composed of Stenotrophomonas rhizophila, Xanthomonas retroflexus, Microbacterium oxydans and Paenibacillus amylolyticus, showed particularly strong biofilm synergy and was selected for further studies. X. retroflexus was the only strain, out of the four, capable of forming abundant biofilm on its own, under the in vitro conditions provided. In accordance, strain specific quantitative PCRs revealed that X. retroflexus was highly dominant in the four-species consortium (> 97% of total biofilm cell number). Although the three other strains were present in low abundances, they were all indispensable for the strong synergistic effect to occur in the four-species biofilms. Moreover, absolute cell numbers of each strain were significantly enhanced when comparing multi- to singlespecies biofilms. This indicated that all the individual strains benefited from joining the multispecies community. Our results show a high prevalence of synergy in biofilm formation in multispecies consortia isolated from a natural bacterial habitat and suggest that interspecific cooperation occurred. Introduction Biofilms in environmental systems are comprised of a large number of different bacteria living together (Hall-Stoodley et al., 2004). Soil, for example, are systems that typically are high in bacterial numbers, diversity and readily available surfaces, which suggest a good setting for multispecies biofilm formation (Burmølle et al., 2012). The members of multispecies biofilms may influence each other antagonistically; e.g. through resource competition or production of inhibitory compounds (Rao et al., 2005), or synergistically; via mechanisms such as syntrophy, biofilm induction or improved resistance in biofilms (Burmølle et al., 2013). We have previously observed strong synergy among four epiphytic isolates from a marine environment. When comparing single species biofilms with four-species biofilms, enhanced biofilm biomass was observed in the latter, in addition to higher resistance to antibacterial agents and better protection from bacterial invasion (Burmølle et al., 2006). Diaz et al., likewise, revealed a synergistic partnership between Candida albicans and streptococci where C. albicans promoted the ability of streptococci to form biofilms on abiotic surfaces or on the surface of an oral mucosa analogue (Diaz et al., 2012). Additionally, several studies have shown that some species, which were unable to form a biofilm alone, could promote the biomass of mixed-species biofilms (Filoche et al., 2004, Klayman et al., 2009, Sharma et al., 2005, Yamada et al., 2005). In a recent study, Lee et al., (2013) addressed the protective effect of the resistant species in a threespecies biofilm. These resistant species protected the more sensitive ones from inhibitory compounds and the overall species ratio remained constant, implying that resistance mechanisms may serve as public goods in multispecies biofilms (Lee et al., 2013). Multi-species biofilms play an essential role in maintaining the ecological balance in soil (Burmølle et al., 2012) and when compared with single-species biofilms and planktonic counterparts, multispecies biofilm formation seem to promote certain advantages, including increased resistance to antibacterial compounds, enhanced protection from desiccation and protozoan predation as well as high rate of horizontal gene transfer (Davey and O'toole 2000, Jefferson 2004, Sørensen et al., 2005). Therefore a pressing need exists for more research directed towards understanding of the social interactions and selective forces that drive bacterial biofilm communities. Specifically, the net effects and gains/losses regarding biofilm biomass and protection level of each strain in a mixed community must be evaluated in order to characterize the underlying interactions as being either cooperative or competitive. In this study we assessed the prevalence of synergism in four-species biofilm formation of different soil isolates. Using quantitative PCR (qPCR), we determined the presence and progression dynamics of the individual strains at different stages of biofilm development in a selected four-species biofilm community. We demonstrate a high prevalence of synergistic effects in multispecies biofilm, indicative of cooperative forces shaping these communities. Materials and methods Bacterial strains used in this study Seven agricultural, bacterial isolates, previously identified and characterized, were used in this study: 1) Pseudomonas lutea, 2) Stenotrophomonas rhizophila, 3) Xanthomonas retroflexus, 4) Ochrobactrum rhizosphaerae, 5) Microbacterium oxydans, 6) Arthrobacter nitroguajacolicus and 7) Paenibacillus amylolyticus (de la Cruz Perera et al., 2013, Ren et al., 2013). Biofilm quantification by use of crystal violet assay Biofilm formation was assayed and quantified as previously described (Ren et al., 2013). Briefly, exponential phase cultures of the seven selected strains were adjusted to an OD 600 of 0.15 in TSB (Tryptic Soy Broth, Merck KgaA, Germany) medium and then inoculated into Nunc-TSP plate (Cat. No. 445497, Thermo Scientific, Denmark). The inoculum volumes were 160 μL for monospecies biofilms and 40 μL for each species in four-species biofilms. After 24-hour incubation at 24°C with shaking (200 rpm), biofilm formation was quantified by a modified crystal violet assay based on the Calgary device. Results The four-species consortium composed of strains 2, 3, 5 and 7 was examined for synergy in singlespecies and all possible combinations of two-, three- and four-species biofilms. The inoculum volumes of each strain were equivalent and added up to a total of 160 μL. Biofilm assays were performed as described above. The prevalence of synergistic and antagonistic effects in biofilm formation among the seven soil isolates was examined by co-culturing all possible combinations of four strains. A total of 35 different four-species consortia were screened for biofilm formation by using the Nunc-TSP lid system and quantified by the modified CV staining method, previously demonstrated to be suitable and reproducible (Ren et al., 2013). The biofilm experiments were repeated three times on three independent days with four replicates every time. The statistical analyses were conducted using ANOVA test (SPSS version 17.0 for Windows). P values < 0.05 were regarded as statistically significant. Strain specific qPCR on biofilm and planktonic fractions Cell numbers of mono- and co-cultures of strain 2, 3, 5 and 7 were assessed by qPCR as follows. The four-species biofilms attached on the pegs and planktonic cells in wells were collected at six time points (4h, 8h, 12h, 16h, 20h and 24h after coinoculation). In addition, both single-species biofilms and associated planktonic fractions were sampled at 24 h. Three replicates were prepared at each time point. Cells were lysed by lysozyme digestion and bead beating, followed by DNA extraction using FastDNA™ SPIN Kit for soil (Qbiogene, Illkirch, France) and species specific qPCRs as previously reported (Ren et al., 2013). The PCR programs were adjusted as follows: 95°C for 2 min, 40 cycles of 95°C for 15s, 64°C for 20s and 72°C for 20s, followed by a standard melting/dissociation curve segment. Each sample was run in duplicate wells and a no template control was included in every run. As the exact copy numbers of the 16S rRNA gene per cell in the four species are currently unknown, the cell numbers were calculated based on other species in the same genus as previously described (Ren et al., 2013). Biofilm formation by single-species and fourspecies consortia Whether synergistic or antagonistic effects dominated in the individual four-species biofilm were assessed by relating the measured absorbance of the multi-species biofilm (Abs590 MS) to that of the best single-species biofilm former present in the relevant combination (Abs590 BS) as follows: (Abs590 MS - St dev) > (Abs590 BS + St dev) = Synergism (Abs590 MS + St dev) < (Abs590 BS - St dev) = Antagonism This is based on the assumptions that in the absence of interactions 1) the cell density of single- and multi-species biofilms are equal so neither more nor less biofilm is expected to be formed by multiple species compared to single species with similar nutrient availability unless interactions causing synergistic or antagonistic effects occur and 2) the best biofilm former dominates the biofilm. Figure 1 shows data from a representative dataset; data from the three biological replicates are presented as Supplementary material (Table S1). Figure 1 Four-species biofilm formation of seven soil isolates. The observed data points (dark grey dots) were collected by quantifying biofilm formation of all four-species combinations using the crystal violet assay. Error bars represent standard deviations of four replicates. Light grey bars indicate the amount of monospecies biofilm produced by the best biofilm former present in the combination. Data points (incl. standard deviations) above grey areas indicate synergistic effects in four species biofilms and data points (incl. standard deviations) below grey areas indicate antagonism (see text for further details). Strains: 1- P. lutea, 2- S. rhizophila, 3- X. retroflexus, 4- O. rhizosphaerae, 5M. oxydans, 6- A. nitroguajacolicus, 7- P. amylolyticus. A total of 22 four-species combinations, which accounted for 63% of all combinations (35), showed synergy in biofilm formation, while only 6% (2 of 35) of the four-species combinations showed antagonistic effects owing to the diminished biomass compared with single-species biofilms formation. Furthermore, 10 of the 22 combinations showing synergy were composed only of poor biofilm-forming strains, especially combinations 12-5-7 and 1-2-6-7, in which biofilm biomass had increased by more than 5-fold (P < 0.05). This may indicate that interspecific interactions had led to cooperative biofilm formation by these strains that were unable to form biofilm when grown individually. In total, 31% (11 of 35) of all combinations were categorized as having no significant change in biomass as (Abs590 MS - St dev) < (Abs590 BS + St dev) and (Abs590 MS + St dev) > (Abs590 BS - St dev). Figure 2 Biofilms formed by four isolates 2Stenotrophomonas rhizophila, 3Xanthomonas retroflexus, 5- Microbacterium oxydans and 7Paenibacillus amylolyticus when equal aliquots of the diluted cultures were incubated in co-cultures of two, three and four isolates. Assays for the detection of synergistic effects were performed three times (Experiment 1, 2 and 3) with 4 replicates each time. Error bars represent ± SEM of four replicates. Synergism in biofilm formation among S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus An example of strong synergy was found in strain combination 2-3-5-7 (S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus, respectively, Figure 1 and Supplementary Table S1). To determine whether all these four strains contributed to the enhanced biomass, each strain was grown as single-species biofilm and in all possible combinations as two, three and four-species biofilm and analyzed for synergy in biofilm formation as described above (Figure 2). Two- and three-species biofilms did, with the exception of combinations 3-5 (P = 0.026) and 3-5- 7 (P = 0.001), not differ significantly (P > 0.05) compared with the amount of biofilm produced by the single species. Strain 7, P. amylolyticus, which grew much slower in TSB medium than the other three strains (data not shown), did not produce any biofilm alone, but as strain 2, it stimulated biofilm formation of the other three species. When the four isolates were co-cultured, the biomass increased by more than 4-fold compared to that of single-species biofilms and 2-5 fold compared to that of threespecies biofilms. This indicates that each of these four strains was indispensable to induce strong synergy, and interestingly, this applied both to the good biofilm former, strain 3 X. retroflexus as well as the three other strains that did not produce biofilm in monocultures. Figure 3 The estimated cell numbers per peg/well of each strain in four-species biofilms and associated planktonic fractions of S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus at six time points (4 h, 8 h, 12 h, 16 h, 20 h and 24 h after co-incubation) based on qPCR. MP – Multispecies planktonic cells, MB – Multispecies biofilms. Each point represents the mean of three replicates, with vertical lines representing ± standard deviations. The cell numbers of each species in the multispecies biofilm (peg) and in the associated planktonic fraction (well) are shown in Figure 3. In general, a marked increase in cell numbers was observed from 4h to 12h for all of the four strains in the biofilm fraction. After 12 hours, only the cell numbers of strain 3 increased continuously, whereas those of the other three species in the multi-species biofilm remained constant or decreased. Cell numbers of all four species in the planktonic fraction increased during the first 12-16 hours, hereafter planktonic growth leveled off, indicating nutrient depletion and transition to a steady state condition. Cell numbers in mono- and co-cultures Specific qPCRs were performed in order to evaluate the total cell numbers of each strain in both biofilms and planktonic communities at different stages of multi-species biofilm development. Additionally, the cell numbers in single-species biofilms and planktonic cells at 24 h were measured. The DNA samples from different time points were diluted appropriately in order to yield results within the dynamic range of the standard curves, which were generated using duplicate 10-fold dilutions of plasmid DNAs (Ren et al., 2013). R-Square (RSq) values, based on threshold cycles of the standard curves, ranged from 0.988 to 1 and application efficiencies (E) ranged from 84.7 - 98.0 % (Supplementary material, Table S2). 10 Multispecies biofilms Single-species biofilms Cell numbers (log10 ) 8 6 4 2 0 S. rhizophila X. retroflexus M. oxydans P. amylolyticus Strains Figure 4 The estimated cell numbers for each strain (S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus) in four-species and single-species biofilms at 24h. Bars represent means ± standard deviation for three replicates. The cell numbers of each strain in the multispecies biofilm and in the single-species biofilms at 24h are shown in Figure 4. For all of the four species, cell numbers were significantly higher (P < 0.05) in four-species biofilms compared to single-species biofilms, indicative of individual fitness gains when joining the multispecies biofilm, when interpreting enhanced cell numbers in biofilms as fitness gain. The ratio of total cell numbers at 24h of the four strains S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus was 4:900:9:15 in multispecies biofilms (Table 1). Thus, there is a strong dominance of X. retroflexus in the fourspecies biofilm, as this strain constitutes > 97% of the total cell number. It is worth noticing that, in contrast to biofilms (Figure 4), planktonic cell numbers of the four species, except for P. amylolyticus, are lower in the four-species coculture compared to single-species planktonic biomass (Table 1). The summarized cell numbers from the biofilm and the planktonic fraction from a specific well are not representative of the cell number of that well because there are other surfaces available for bacterial attachment in the wells besides the pegs. Discussion High prevalence of synergism in biofilm formation Single-species biofilms are rare in natural environments, especially in agricultural soil where micro-communities exposed to plenty of organic matter have the potential to develop into multispecies biofilms with high bacterial density and diversity (Burmølle et al., 2010, Narisawa et al., 2008, Rodríguez and Bishop 2007). Such conditions in bacterial habitats are likely to facilitate the development of intricate relationships between different species. While many previous studies have focused on interspecies interactions within oral microbial communities (Kuramitsu et al., 2007, Palmer Jr et al., 2001, Saito et al., 2008, Sharma et al., 2005, Wang and Kuramitsu 2005), research on multispecies biofilms composed of soil bacteria is still in its infancy. In this study, seven soil bacteria, isolated from one micro-habitat, were screened in order to assess the prevalence of interactions leading to synergism in biofilm formation of fourspecies co-cultures. Table 1 Absolute cell numbers per peg/well measured by quantitative PCRs at 24h in multispecies planktonic cells, single-species planktonic cells, multispecies biofilms and single-species biofilms. Each value is the mean of three replicates. S. rhizophila X. retroflexus Multispecies planktonic cells 2.33E+07 3.38E+08 Single-species planktonic cells 5.79E+07 5.46E+08 Multispecies biofilms 1.35E+07 2.97E+09 Single-species biofilms 9.14E+03 4.92E+07 M. oxydans 6.42E+06 1.58E+08 2.93E+07 4.73E+04 P. amylolyticus 2.81E+07 1.33E+07 4.78E+07 2.10E+03 Strains In total, 63% of four-species biofilms showed synergistic effects in biofilm formation while only 6% displayed antagonistic activity. This is in agreement with the assumption that the driving force in bacterial community development is the selforganization and cooperation rather than competition of individual microorganisms (Daniels et al., 2004, Davies et al., 1998, Parsek and Greenberg 2005). The recently presented “Black Queen Hypothesis”, explains how multispecies cooperation can evolve (Morris et al., 2012). This hypothesis considers cooperation in complex bacterial communities as being a consequence of species adapting to the presence of each other. In order to enhance own fitness, species delete vital functions or pathways that are provided by the surrounding bacteria. This leads to a dependency and is therefore an irreversible commitment to living in close association with other species, which may often require development of more complex systems to ensure that the coexistence is maintained. When applying the Black Queen Hypothesis on the results of the present study, the strains may prefer living in the multispecies biofilm in order to keep vital partners in close contact, or the ability of biofilm formation could be the function that is lost by some of the species. In contrast to our results, Foster et al., recently showed that the great majority of interactions in pairwise species combinations of bacterial isolates from tree-hole rainwater pools were net negative and very few strong higher-order positive effects arose from combinations composed of more than two species (Foster and Bell 2012). The two studies differ with respect to bacterial habitats targeted for isolation (soil vs. tree-holes), productivity parameter assessed (biofilm formation vs. CO2 production) and different definitions of synergism/cooperation, which may explain the observed differences. The tree-hole species tend to use similar resources which may be the key factor that lead to competition among microbes (Lawrence et al., 2012), while in nutrient rich agricultural soil, bacteria are likely to be tightly associated and by their collective metabolic activities, individual bacterial consortia stabilize their environment and fertilize the soil. The key to whether bacterial species compete or cooperate may lie in their possibility of long-term co-adaptation and degree of niche overlap. We are currently investigating the significance of these parameters for the net interactions of bacterial communities. The synergistic interactions were also observed when four poor biofilm formers were co-cultured (Figure 1), which verified the previous findings wherein the biofilm forming ability of individual bacteria is not necessarily an indicator of their potential in multispecies biofilms (Bharathi et al., 2011, Burmølle et al., 2006). This shifting of biofilm pattern from weak to moderate or strong can be the result of metabolic interactions (Møller et al., 1998), enhanced coaggregation (Rickard et al., 2003), organized spatial distribution (Skillman et al., 1998) and/or facilitated initial surface attachment (Klayman et al., 2009, Simões et al., 2008), e.g. bridging bacteria may facilitate the association of other species that do not coaggregate directly with each other. Thus, species that do not form biofilms as single strains may benefit from the advantages associated to biofilm formation, including enhanced protection from outside stress and expanded niche availability, by engaging in multispecies communities. When exploring strain dynamics of the selected four-species community, the strongest synergy was induced only when all the four strains S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus were co-cultivated (Figure 2). This demonstrates that each strain played an important role in the enhanced biofilm formation, regardless of the capability of single-species biofilm formation. Divergence in resource use may be one of the reasons that lead to increased productivity of the entire community. By adapting to consume different resources, and to metabolize waste products produced by other species, the consortium collectively decompose substrates in the medium more effectively. In the present study, the synergistic enhancement of biomass of the four co-cultured species was only observed when grown as a biofilm; Planktonic coculturing did not lead to changes in overall biomass (data not shown). This indicates that a structured environment is highly important for the synergy to occur. Thus co-metabolism (syntrophy) may Enhanced biofilm formation in co-cultures explain the observed synergy, but only when coupled to a structured environment that enables tight cell-cell associations optimizing product/substrate availability. Similar observations have previously been reported (Hansen et al., 2007, Stewart et al., 1997). qPCR results showed a strong dominance (> 97%) of X. retroflexus in the four-species biofilm, which is consistent with the results from the crystal violet assay where X. retroflexus was the only good biofilm former out of the four strains. Despite its monospecies biofilm forming abilities, cell numbers of strain X. retroflexus were enhanced approximately 60-fold when co-cultured in biofilms with the other strains. The remaining three strains constituted < 3% of the total cell number of the four-species biofilms, however they all showed enhanced cell numbers in the multispecies biofilm by over three orders of magnitude when compared to cell numbers in monospecies biofilms. Thus, S. rhizophila, M. oxydans and P. amylolyticus, although present in low abundances and inability of monospecies biofilm formation, stimulate of X. retroflexus and obtain higher cell numbers when present in the multispecies biofilm. This significant change in capacity of biofilm formation may be explained by the fact that, species evolving in communities may have higher growth rates when assayed in the presence of other species (Lawrence et al., 2012), especially in structured communities (Hansen et al., 2007, Stewart et al., 1997). P. amylolyticus, of which cell numbers was most strongly enhanced, has previously been reported to produce antibacterial agents with broad spectrum activity against both Gram-negative and Grampositive bacteria (DeCrescenzo Henriksen et al., 2007). However, in this study, this strain appeared to stimulate rather than inhibit the growth of other three strains. Further studies to identify the changes in gene expression patterns in this four-species biofilm will lead to a deeper understanding of the underlying molecular mechanisms of synergism in bacterial communities. Synergism in cooperation biofilm formation indicate Several observations in this study indicate that bacteria increase fitness from joining multispecies biofilms. If this fitness advantage applies to all of the species present, the underlying interaction is categorized as being cooperative (West et al., 2007). Four such observations are discussed here. (i) We observed a high prevalence of biofilm synergy in four-species biofilms, strongly dominating over antagonistic effects. This implies selection for living in multispecies communities, indirectly indicating that bacteria may enhance fitness when joining multispecies biofilms. The fitness advantage may be caused by growth promotion that enhances bacterial biomass and thereby the direct fitness, but also the advantages gained from the biofilmassociated bacterial protection may be of major significance in natural environments with high levels of stress. (ii) Each of the four species in the selected community was indispensable for the synergistic effect observed on biofilm biovolume. These vital interdependencies may evolve over many years under continuous selective pressures, whereby only fitness enhancing relationships are favored. (iii) The cell numbers of all four species were higher in multi-species biofilms when compared to biofilms composed of single species. Thus, when focusing on biofilm-associated growth, there is a direct gain in fitness for all strains from joining the four-species biofilm. (iv) The observed synergistic effect only applied to the four-species biofilm; cell numbers in the planktonic fraction decreased for three out of the four strains. Based on this, there seem to be selection for or forces driving bacteria to form multi-species biofilms. Alternatively, metabolic cooperation occurs in the biofilm, whereby the growth rates of biofilmresident strains are enhanced. In conclusion, the observations presented in this study points in the direction of multispecies biofilms as being a favorable bacterial habitat; they gain protection and opportunities of engaging in mutually beneficial cooperative interactions. Acknowledgements This study was partly funded by grants to SJS and MB from the Danish Innovation Consortium SiB, ref no: 11804520, and The Danish Council for Independent Research; ref no: DFF-1335- 00071 and ref no: DFF1323-00235 (SIMICOM). DeCrescenzo Henriksen E, Phillips D, Peterson J (2007). Polymyxin E production by P. amylolyticus. Letters in applied microbiology 45: 491-496. References Diaz PI, Xie Z, Sobue T, Thompson A, Biyikoglu B, Ricker A et al., (2012). Synergistic interaction between Candida albicans and commensal oral streptococci in a novel in vitro mucosal model. Infection and immunity 80: 620-632. Bharathi P, Bhowmick P, Shekar M, Karunasagar I (2011). Biofilm formation by pure and mixed culture of Lactobacillus isolates on polystyrene surface in varying nutrient conditions. Biotechnol. Bioinf. Bioeng. 1: 93-98. Filoche S, Anderson S, Sissons C (2004). Biofilm growth of Lactobacillus species is promoted by Actinomyces species and Streptococcus mutans. Oral microbiology and immunology 19: 322-326. Burmølle M, Webb JS, Rao D, Hansen LH, Sørensen SJ, Kjelleberg S (2006). Enhanced biofilm formation and increased resistance to antimicrobial agents and bacterial invasion are caused by synergistic interactions in multispecies biofilms. Applied and environmental microbiology 72: 3916-3923. Foster KR, Bell T (2012). Competition, not cooperation, dominates interactions among culturable microbial species. Current Biology 22: 1845-1850. Burmølle M, Thomsen TR, Fazli M, Dige I, Christensen L, Homøe P et al., (2010). Biofilms in chronic infections – a matter of opportunity–monospecies biofilms in multispecies infections. FEMS Immunology & Medical Microbiology 59: 324-336. Burmølle M, Kjøller A, Sørensen SJ (2012). An Invisible Workforce: Biofilms in the Soil. Microbial Biofilms: Current Research and Applications: 61. Burmølle M, Ren D, Bjarnsholt T, Sørensen SJ (2013). Interactions in multispecies biofilms: do they actually matter? Trends in Microbiology, In Press: DOI: 10.1016/j.tim.2013.12.004 Daniels R, Vanderleyden J, Michiels J (2004). Quorum sensing and swarming migration in bacteria. FEMS microbiology reviews 28: 261-289. Davey ME, O'toole GA (2000). Microbial biofilms: from ecology to molecular genetics. Microbiology and molecular biology reviews 64: 847-867. Davies DG, Parsek MR, Pearson JP, Iglewski BH, Costerton J, Greenberg E (1998). The involvement of cell-to-cell signals in the development of a bacterial biofilm. Science 280: 295-298. de la Cruz Perera CI, Ren D, Blanchet M, Dendooven L, Marsch R, Sørensen SJ et al., (2013). The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains. FEMS Microbiology Letters 338: 95-100. Hall-Stoodley L, Costerton JW, Stoodley P (2004). Bacterial biofilms: from the natural environment to infectious diseases. Nature Reviews Microbiology 2: 95108. Hansen SK, Rainey PB, Haagensen JA, Molin S (2007). Evolution of species interactions in a biofilm community. Nature 445: 533-536. Jefferson KK (2004). What drives bacteria to produce a biofilm? FEMS microbiology letters 236: 163-173. Klayman BJ, Volden PA, Stewart PS, Camper AK (2009). Escherichia coli O157: H7 requires colonizing partner to adhere and persist in a capillary flow cell. Environmental science & technology 43: 2105-2111. Kuramitsu HK, He X, Lux R, Anderson MH, Shi W (2007). Interspecies interactions within oral microbial communities. Microbiology and molecular biology reviews 71: 653-670. Lawrence D, Fiegna F, Behrends V, Bundy JG, Phillimore AB, Bell T et al., (2012). Species interactions alter evolutionary responses to a novel environment. PLoS biology 10: e1001330. Lee KWK, Periasamy S, Mukherjee M, Xie C, Kjelleberg S, Rice SA (2013). Biofilm development and enhanced stress resistance of a model, mixed-species community biofilm. The ISME journal 1-14. Morris JJ, Lenski RE, Zinser ER (2012). The Black Queen Hypothesis: evolution of dependencies through adaptive gene loss. MBio 3: DOI: 10.1128/mBio.00036– 12 Møller S, Sternberg C, Andersen JB, Christensen BB, Ramos JL, Givskov M et al., (1998). In situ gene expression in mixed-culture biofilms: evidence of metabolic interactions between community members. Applied and environmental microbiology 64: 721-732. Narisawa N, Haruta S, Arai H, Ishii M, Igarashi Y (2008). Coexistence of antibiotic producing and antibiotic-sensitive bacteria in biofilms is mediated by resistant bacteria. Applied and environmental microbiology 74: 3887-3894. Palmer Jr RJ, Kazmerzak K, Hansen MC, Kolenbrander PE (2001). Mutualism versus independence: strategies of mixed-species oral biofilms in vitro using saliva as the sole nutrient source. Infection and immunity 69: 57945804. Parsek MR, Greenberg E (2005). Sociomicrobiology: the connections between quorum sensing and biofilms. Trends in microbiology 13: 27-33. Rao D, Webb JS, Kjelleberg S (2005). Competitive interactions in mixed-species biofilms containing the marine bacterium Pseudoalteromonas tunicata. Applied and environmental microbiology 71: 1729-1736. Ren D, Madsen J, Cruz-Perera C, Bergmark L, Sørensen S, Burmølle M (2013). High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCRBased Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. Microbial ecology: DOI: 10.1007/s00248-013-0315-z Rickard AH, Gilbert P, High NJ, Kolenbrander PE, Handley PS (2003). Bacterial coaggregation: an integral process in the development of multi-species biofilms. Trends in microbiology 11: 94-100. Rodríguez SJ, Bishop PL (2007). Three-dimensional quantification of soil biofilms using image analysis. Environmental Engineering Science 24: 96-103. Saito Y, Fujii R, Nakagawa KI, Kuramitsu H, Okuda K, Ishihara K (2008). Stimulation of Fusobacterium nucleatum biofilm formation by Porphyromonas gingivalis. Oral microbiology and immunology 23: 1-6. Sharma 420 A, Inagaki S, Sigurdson W, Kuramitsu H (2005). Synergy between Tannerella forsythia and Fusobacterium nucleatum in biofilm formation. Oral microbiology and immunology 20: 39-42. Simões LC, Simoes M, Vieira MJ (2008). Intergeneric coaggregation among drinking water bacteria: evidence of a role for Acinetobacter calcoaceticus as a bridging bacterium. Applied and environmental microbiology 74: 1259-1263. Skillman LC, Sutherland IW, Jones MV, Goulsbra A (1998). Green fluorescent protein as a novel speciesspecific marker in enteric dual-species biofilms. Microbiology 144: 2095-2101. Stewart PS, Camper AK, Handran S, Huang C-T, Warnecke M (1997). Spatial distribution and coexistence of Klebsiella pneumoniae and Pseudomonas aeruginosa in biofilms. Microbial ecology 33: 2-10. Sørensen SJ, Bailey M, Hansen LH, Kroer N, Wuertz S (2005). Studying plasmid horizontal transfer in situ: a critical review. Nature Reviews Microbiology 3: 700710. Wang BY, Kuramitsu HK (2005). Interactions between oral bacteria: inhibition of Streptococcus mutans bacteriocin production by Streptococcus gordonii. Applied and environmental microbiology 71: 354-362. West SA, Diggle SP, Buckling A, Gardner A, Griffin AS (2007). The social lives of microbes. Annual Review of Ecology, Evolution, and Systematics 38: 53-77. Yamada M, Ikegami A, Kuramitsu HK (2005). Synergistic biofilm formation by Treponema denticola and Porphyromonas gingivalis. FEMS microbiology letters 250: 271-277. Supplementary Table 1 Biofilms formed by single species and combinations (four strains). 40μL of each of the four diluted cultures was inoculated in each well. The plate was incubated for 24 h followed by crystal violet staining and absorbance measurements at 590 nm. The ratio of (Abs 590 multispecies biofilm – St dev) / (Abs 590 best strain + St dev) or (Abs 590 multispecies biofilm + St dev) / (Abs 590 best strain - St dev) was calculated and the average value from three experiments is presented as the “Fold change”. Strains 1c 2 3 4 5 6 7 1234 1245 1256 1267 1235 1236 1237 1246 1247 1257 1345 1356 1367 1346 1347 1357 1456 1467 1457 1567 2345 2356 2367 2346 2347 2357 Experiment-1 0.143 ± 0.038 b 0.126 ± 0.011 3.769 ± 0.589 0.381 ± 0.009 0.157 ± 0.009 0.055 ± 0.014 0.004 ± 0.003 3.763 ± 0.289 0.245 ± 0.021 0.593 ± 0.084 0.668 ± 0.176 4.642 ± 0.689 13.734 ± 0.433 6.290 ± 1.043 0.633 ± 0.241 0.882 ± 0.049 1.146 ± 0.207 5.246 ± 0.806 8.454 ± 1.014 9.950 ± 1.818 10.528 ± 0.676 13.542 ± 0.905 13.078 ± 1.068 0.168 ± 0.049 1.053 ± 0.298 1.257 ± 0.155 0.876 ± 0.143 3.452 ± 0.655 7.292 ± 1.099 9.224 ± 0.588 7.436 ± 1.394 14.468 ± 1.888 21.067 ± 1.335 Biofilm formation (Abs 590) a Experiment-2 Experiment-3 0.180 ± 0.106 0.256 ± 0.036 0.077 ± 0.004 0.077 ± 0.016 6.606 ± 1.171 6.245 ± 0.502 0.190 ± 0.037 0.302 ± 0.011 0.094 ± 0.048 0.166 ± 0.018 0.049 ± 0.028 0.149 ± 0.073 0.017 ± 0.004 0.033 ± 0.010 4.649 ± 0.435 3.777 ± 0.144 0.281 ± 0.037 0.247 ± 0.030 0.992 ± 0.490 0.390 ± 0.061 1.895 ± 0.320 2.564 ± 0.589 10.889 ± 0.905 6.050 ± 1.544 12.697 ± 2.195 13.790 ± 1.572 6.937 ± 1.727 6.142 ± 0.742 1.972 ± 0.621 1.978 ± 0.093 0.987 ± 0.253 1.840 ± 0.511 1.947 ± 0.464 1.782 ± 0.187 10.877 ± 2.125 5.198 ± 0.961 11.587 ± 0.575 9.646 ± 0.526 11.033 ± 0.339 10.226 ± 1.822 12.201 ± 1.132 12.650 ± 1.069 12.526 ± 1.317 9.806 ± 0.983 13.601 ± 0.622 11.502 ± 0.738 0.103 ± 0.013 0.268 ± 0.005 0.707 ± 0.175 0.839 ± 0.394 0.676 ± 0.173 1.169 ± 0.334 0.498 ± 0.220 1.375 ± 0.694 6.246 ± 0.559 3.674 ± 0.091 7.646 ± 1.411 6.153 ± 1.172 9.102 ± 1.934 9.813 ± 1.833 6.762 ± 0.894 7.757 ± 1.833 15.250 ± 0.894 13.813 ± 0.588 22.885 ± 2.816 19.045 ± 2.158 Fold change 0.910 Nd 1.767 5.325 N 1.910 N 3.001 3.113 5.292 N 1.460 1.442 1.721 1.730 1.893 0.741 1.862 2.624 2.229 N N 1.260 N 2.127 3.002 2456 2467 2457 2567 3456 3467 3457 3567 4567 a 0.330 ± 0.040 0.200 ± 0.070 0.310 ± 0.060 0.794 ± 0.480 7.020 ± 0.180 8.340 ± 1.117 8.908 ± 2.148 11.436 ± 1.773 0.129 ± 0.034 1.023 ± 0.084 0.526 ± 0.136 0.281 ± 0.070 1.028 ± 0.467 6.542 ± 0.370 6.710 ± 1.350 9.374 ± 1.759 11.006 ± 1.709 0.085 ± 0.063 0.514 ± 0.083 0.413 ± 0.119 0.471 ± 0.115 1.502 ± 0.444 7.325 ± 1.049 6.533 ± 0.973 11.777 ± 2.374 10.053 ± 2.061 0.399 ± 0.154 1.785 N N 3.929 1.021 N 1.259 1.427 N Assays for the detection of biofilm formation were performed three times (Experiment 1, 2 and 3). Values represent the means ± standard deviation of four replicates in each experiment. c 1- Pseudomonas lutea, 2- Stenotrophomonas rhizophila, 3- Xanthomonas retroflexus, 4- Ochrobactrum rhizosphaerae, 5- Microbacterium oxydans, 6- Arthrobacter nitroguajacolicus, 7- Paenibacillus amylolyticus. d Combinations which showed (Abs 590 multispecies biofilm – St dev) < (Abs 590 best strain + St dev) and (Abs 590 multispecies biofilm + St dev) > (Abs 590 best strain - St dev) were identified as no significant change in biomass. b Supplementary Table 2 Standard curves used to measure the cell numbers of strains S. rhizophila, X. retroflexus, M. oxydans and P. amylolyticus in mono- and multi-species biofilms and planktonic cell fractions. a Species RSq Application efficiency(E) Estimated copy number of the 16S rRNA gene (per cell)a S. rhizophila 1.000 87.1% 4 X. retroflexus 0.999 94.3% 2 M. oxydans 0.988 98.0% 2 P. amylolyticus 0.999 84.7% 12 The copy numbers of the 16S rRNA gene were estimated according to other species in the same genus (Ren et al. 2013). Manuscript 3 Metatranscriptome analysis of multispecies biofilms indicates strain- and communitydependent changes in gene expression Metatranscriptome analysis of multispecies biofilms indicates strain- and community- dependent changes in gene expression Lea Benedicte Skov Hansen+, Dawei Ren+, Søren J. Sørensen*, Mette Burmølle* + Shared first authorship Section of Microbiology, Department of Biology, University of Copenhagen *Corresponding authors Postal address: Universitetsparken 15, Bygn. 1, 2100 København Ø, Denmark e-mail Mette Burmølle: [email protected] e-mail Søren J. Sørensen: [email protected] Key words: Metatranscriptome / Multispecies biofilm / Soil bacteria / Synergistic interactions Abstract It has gradually become more apparent that bacteria often exist in naturally formed multispecies biofilms. Within these biofilms interspecies interaction seems to play an important role in shaping the function and activities of these dynamic communities. However, little is known about the effect of interspecies interaction on a gene expression level in these multispecies biofilms. This study represents a comparative gene expression analysis of the Xanthomonas retroflexus transcriptomes when grown in a single-species biofilm and in dual- and four-species biofilms with Stenotrophomonas rhizophila, Microbacterium oxydans and Paenibacillus amylolyticus. The result revealed a species specific response to coexistence in the dual-species biofilm, where the overall strongest change in expression profile was observed in the dual-species biofilms of X. retroflexus and P. amylolyticus. Furthermore, a distinct expression pattern was detected in the four-species biofilm including changes in expression of genes, which were not observed differentially expressed in any of the dual-species biofilms. This non-linear response in the four-species biofilm indicates a complex interaction pattern at the gene regulation level when increasing the number of species in the co-cultures. This is in correspondence with the previously described emergent behaviour of enhanced biofilm formation by this specific four-species consortia isolated from soil [1]. 70 genes were found differentially expressed when co-culturing X. retroflexus with other species, which include genes involved in amino acid metabolism, membrane bound efflux system and MazE/MazF toxin-antitoxin system, suggesting the enhanced resistance of multispecies biofilms. Introduction It is now acknowledged that biofilms, as the dominant microbial lifestyle in nature, are composed of multiple species, where extensive interactions between different species are bound to play a crucial role in shaping these dynamic communities [2]. In general, interspecies interactions, either synergistic or antagonistic, involve physical contact, metabolic communication, quorum sensing and genetic exchange [3]. The synergistic effect of living as a mixed community, which often exhibits enhanced biomass compared with either single species grown alone, is implicated in the development of several beneficial phenotypes, including promoted cell attachment due to coaggregation [4], co-metabolism where one species takes advantage of the metabolite produced by a neighbouring species [5], and cross-species protection against antimicrobials or host immune responses [6, 7]. The synergistic interactions in multispecies biofilms are considered to be an effective strategy in degrading organic matter [8, 9] in nature and evading host clearance in chronic infections [10], indicating the profound impact interactions have both on major ecological processes and on clinical settings. While the reductionist approach has advanced biofilm research by analysing single-species biofilms or individual species in a complex microbial community, interactions between different organisms within multispecies biofilms as a key area for providing insight into evolutionary ecology and developing new therapeutics, is still in its infancy. The rise of metatranscriptomic analysis to study the global gene expression profiling of the whole microbial communities, has been greatly accelerated by RNA-Seq in the past few years, and thus opens new possibilities for clarifying the functional interaction networks within multispecies microbial communities. This new technology has been recently applied to an oral biofilm model [11], where cell-cell interactions are believed to play integral roles in the development of biofilm architecture and pathogenicity. The dramatic changes in gene expression profiles of this multispecies biofilm model in the presence of periodontal pathogens were accurately assessed. Moreover, small noncoding RNAs with important gene regulatory roles were identified, showing promise of metatranscriptomic analysis in complex microbial communities [11]. The synergism in a four-species biofilm composed of Stenotrophomonas rhizophila, Xanthomonas retroflexus, Microbacterium oxydans and Paenibacillus amylolyticus, was presented in our previous studies with increased biofilm formation over 3-fold relative to singlespecies biofilms [1]. As the only good biofilm former, the prevalence (> 97% of total biofilm cell number) of strain X. retroflexus was demonstrated by species specific quantitative PCR [12]. Moreover, despite of the low abundance of three other strains, they were all indispensable for the strong synergy that occurred in the four-species biofilm which make this consortium a powerful model to study intricate interactions in multispecies biofilms. In the present study, the gene expression profile of X. retroflexus in a single-species biofilm was compared to its expression profiles in dualspecies biofilms with S. rhizophila, M.oxydans or P. amylolyticus as well as in a four-species biofilm. This represents, to our knowledge, the first evaluation of gene expression changes due to the interspecies interactions in a multispecies biofilm model composed of soil isolates. We utilized an RNA-Seq-based metatranscriptomic approach, and found a significant effect of the interspecies interactions on the gene expression. Materials and methods Bacterial strains and growth conditions The four species used in this study were Stenotrophomonas rhizophila (JQ890538), Xanthomonas retroflexus (JQ890537), Microbacterium oxydans (JQ890539), and Paenibacillus amylolyticus (JQ890540) [1]. Strains were streaked from frozen glycerol stocks onto TSA (Tryptic Soy Agar) plates and incubated for 48h at 24°C. Hereafter, the freshly isolated colonies were transferred into 5 mL of TSB (Tryptic Soy Broth) and incubated with shaking (250 rpm) at 24°C overnight. Biofilm cultivation Biofilms for gene expression analysis by RNAsequencing were grown in six-well polystyrene plates (Greiner Bio-One, Germany). Overnight cultures of each strain were subcultured to exponential phase and adjusted to an OD600nm of 0.15 in fresh TSB, as described by Ren et al. 2013 [1]. A total of 4 mL of single-species, dual-species or four-species diluted cultures (admixtures of each species at equal cell density) were inoculated in each well and allowed to grow for 24h at 24°C. Single- and dual-species biofilms were prepared in triplicate wells, while four-species biofilm were prepared in five replicate wells. DNA and RNA extraction Isolation of genomic DNA from overnight cultures of four bacteria was performed as previously described using FastDNA™ SPIN Kit for soil (Qbiogene, Illkirch, France) [1]. For RNA extraction, planktonic cells from biofilm-containing wells were gently removed, wells were rinsed three times using phosphate buffer solution (PBS) and then the biofilms were scraped with sterile pipette tips into 1 mL of PBS. The biofilm suspensions were immediately centrifuged and resuspended in 100 μL of PBS followed by addition of 500 μL of RNAlater® (Ambion). The RNAlaterpreserved samples were kept at 4 °C overnight, after which RNAlater was removed by adding 600 μL cold PBS and centrifuging at 8000 rpm for 5 min at 4 °C. The pellet was resuspended in 200 μL of lysozyme solution (20 μg/mL) and incubated at room temperature for 10 min, with vortexing for 10s every 2 min. After 700 μL of buffer RLT was added and vortexed for 10s, the obtained solution was transferred into beadbeating tube (provided by the FastDNA™ SPIN Kit for soil) and bead beaten using the Savant FastPrep FP120 for 30s at setting 5.0, followed by centrifugation at 13000 rpm for 8 min at 4 °C. 0.85 mL of supernatant was transferred to new microcentrifuge tube. 0.47 mL of 100% ethanol was added and mixed by pipetting. 0.7mL of lysate was transferred on to RNeasy mini spin column. Total RNA was then further purified from each biofilm sample using the RNeasy mini kit (Qiagen) according to manufacturer’s instructions. Genomic DNA removal was conducted according to the instructions in DNAfree™ Kit (Ambion), except the incubation time was extended to 1 hour. The complete removal of DNA was confirmed by RT-qPCR using 20 µL SYBR Green reactions on Mx 3000 (Stratagene, Cedar Creek, Texas). The qPCR targeted the 16S rRNA gene by the eubacterial primers EUB338 and EUB518 [13]. All qPCR reactions were run in technical duplicates and contained 10 µL of Brilliant III SYBR Green QPCR Master Mix (Stratagene, Cedar Creek, Texas), 1 µL forward primer (final concentration 385 nM), 1 µL reverse primer (final concentration 385 nM), 1 µL template DNA (100-fold diluted to avoid PCR inhibitors in accordance to Bustin SA et al. [14], and 7 µL ddH2O. The program was modified from Bergmark L et al. [15], combining the annealing and extension step: 95 °C for 3 min followed by 40 cycles of 95 °C for 10 s, 60 °C for 20 s, and a final dissociation curve. The standard curve for bacteria was made from a pure culture of Pseudomonas putida kt2440 [16, 17]. mRNA was subsequently isolated using the MicroExpress Bacterial mRNA Purification Kit (Ambion) following the manufacturer’s instructions. Sequencing The genomic DNA was prepared for sequencing using the NEBNext Quick DNA Sample Prep. Master Mix 2 (New England BioLabs Inc., Ipswitch, MA, USA). DNA was amplified by PCR and sequenced as 250 bp paired end libraries on an Illumina MiSeq (Illumina, San Diego, CA, USA). The RNA transcripts were prepared using ScriptSeq™ v2 RNA-Seq Library Preparation Kit (Epicentre Biotechnologies, Madison, WI, USA) and sequenced as 50 bp single reads on an Illumina HiSeq 2000 (Illumina, San Diego, CA, USA). Data analysis The quality filtering of the genomic and metatranscriptomic libraries was carried out in similar ways. Only reads approved by the CASAVA v1.8.2 (Illumina, San Diego, CA, USA) were included and adaptor remnants were removed. End nucleotides with a quality Phred score below 20 were filtered. After the end trim process, sequences shorter than 50 nucleotides (nt) or 45 nt were removed for the genomic and metatranscriptome libraries, respectively. Reads were discarded if they displayed a mean Phred score below 15 or a mean Phred score below 10 in a sliding window of 10 nt for the genomic libraries and a mean Phred score below 15 across a sliding window of 5 nt for the metatranscriptomes. The quality filtering was carried out using Biopieces. Velvet v.1.2.07 was used to assemble the genomic data [18] and Prodigal v2.50 was used for gene calling [19]. A gene catalogue was created including the genes from all four genomes. All metatranscriptome libraries were mapped to the four genomes using Bowtie2 v.2.0.5 [20] and number nt mapping to genes in the gene catalogue was calculated, which yielded an expression matrix. The 1000 most expressed genes from X. retroflexus were extracted from the matrix and the Bray-Curtis dissimilarity metric was calculated between samples and visualized in a NMDS plot. EdgeR was utilized to infer any statistical significant differential expression of the top 1000 genes [21]. The average number of nt mapping to a gene within biofilm types was calculated and the logarithm to base 2 fold changes (log FC) was calculated between single-species biofilm and the four different multispecies biofilms. Genes showing log FCs between -3 and 3 were excluded and infinite values were treated according to the expression levels of the gene in the singlespecies biofilm. The remaining genes were annotated using protein Blast and the nonredundant protein database maintained by National Center for Biotechnology Information (NCBI - 2014-01-09), furthermore, conserved Pfam domains were located using the Pfam web based batch search tool and signal peptides were identified using the signalP 4.1 server [2224]. Results Sequencing statistics In order to conduct a comparative gene expression analysis of X. retroflexus (Xr) grown in single-species, dual-species and four-species biofilms with S. rhizophila (Sr), M. oxydans (Mo) and P. amylolyticus (Pa), the genomes of these four species were sequenced to create a mapping template for the RNA-Seq transcripts. The raw output from the 250 bp paired end Illumina MiSeq DNA sequencing were quality filtered and between 53.47% and 68.83% nt remained, which yielded between 210 Mb and 1,475 Mb for assembly. The largest estimated genome size of 6.3 Mb and the highest number of coding sequence (CDS) belonged to Pa (Table S1). The smallest genome size of 2.5 Mb and lowest number of CDS belonged to Sr, however the assembly statistics for Sr indicated a poor assembly of the genome with a low N50 and high number of contigs (Table S1). The transcriptome and metatranscriptome libraries were sequenced, quality filtered and mapped against the genome templates. After quality filtering the RNA transcripts, 75.82% nt remained (Table S2). The filtered reads were mapped against all four genomes and 92.39% nt could be remapped, however only 2.46% mapped against CDSs (Table S2). The remaining 97.54% nt were mainly from rRNAs and few mapped to intergenic regions. In general, a higher percentage of the RNA transcripts mapped to CDSs from biofilms containing Pa (2.32% to 4.86%) (Table S2). Expression activity The proportion of transcripts from the four genomes was estimated to assess the relative species specific expression activity in each of the metatranscriptomes from the dual- and four- species biofilms (Figure 1). A noise level of 2.20% was detected, which represents the average percentage of transcripts that mapped unspecifically to conserved regions in the genomes. The abundances of Sr and Mo were close to the detection limit, making their true expression activities difficult to assess. Pa transcripts dominated the transcriptomes with more than 50% and showed the highest percentage in the dual-species biofilms (Xr+Pa). A decrease in abundance could be observed for Pa in the four-species biofilm while Xr showed a relative increase compared to their abundances in the dual-species biofilms, however this was not statistically significant (Student’s t-test, p-values > 0.05). Comparative expression analysis of the single-, dual-, and four-species biofilms In order to investigate gene expression changes in Xr induced by interspecies interactions, Xr expression profiles of dual- and four-species biofilms were compared to that of the Xr single-species biofilm. For this purpose, all transcripts from Xr were extracted from the metatranscriptomes and expression profiles of Xr in each sample were obtained. To infer the correlation between the sample replicates, a pearson correlation matrix was created including all 17 samples and the replicates showed good correlations with values between 0.85 and 0.98 (Table S3). The high rRNA content in the samples resulted in low sequence coverage of the CDSs and higher deviation for low abundant transcripts. Hence, only the 1000 highest expressed genes in Xr were used for further analysis, in order to minimize errors caused by the low genome coverage and low abundant transcripts (Table S1). deviation and did not cluster together. The second cluster consisted of the three replicates of Xr+Pa in close proximity of each other and the third tight cluster consisted of the five replicates of the four-species biofilms, Xr+Sr+Mo+Pa. Figure 1 The species specific transcript abundance in dual- and four-species biofilms. The broken line represents the average percentage of transcript that map unspecifically to conserved regions in the genome and represents the detection limit (2.20%). A) The relative abundance of transcripts that mapped against the Xr genome, in the Xr+Sr, Xr+Mo, Xr+Pa and Xr+Sr+Mo+Pa biofilms. B) The relative abundance of transcripts that mapped against the Sr genome in the Xr+Sr and Xr+Sr+Mo+Pa biofilms. C) The relative abundance of transcripts that mapped against the Mo genome in the Xr+Mo and Xr+Sr+Mo+Pa biofilms. D) The relative abundance of transcripts that mapped against the Pa genome in the Xr+Pa and Xr+Sr+Mo+Pa biofilms. After the samples were normalized to the Xr transcript library sizes and transformed using the natural logarithm, a similarity analysis of the biofilms was conducted based on the expression of the top 1000 genes. The BrayCurtis dissimilarity metric was applied and the results are reflected in an NMDS plot (Figure 2). Three clusters appeared in the plot. The first cluster consisted of the single-species biofilms of Xr and the dual-species biofilms of Xr+Sr and Xr+Mo. Within this cluster the Xr and Xr+Mo replicates created separate sub-clusters, however the replicates of Xr+Sr showed larger The 1000 highest expressed genes in Xr were tested for significantly different expression across the five sample groups using statistics based on empirical Bayes methods [25]. This analysis resulted in 587 differentially expressed genes (p-values < 0.05 with a Benjamini and Hochberg correction). To detect the active genes in the transition from growing in a single-species biofilm to dual- and four-species biofilms, the ratio of the average number of nt mapping to a specific gene in the single-species biofilms vs. each of the four different multispecies biofilms was calculated. A total of 70 genes had an logarithm to base 2 fold change (log FC) above 3 or below -3, which correspond to a 8 times up- or down-regulation. The log FCs of the 70 genes are displayed in a heatmap (Figure 3) and the result correlated well with the observations from the NMDS plot. Many expression ratios for Xr+Sr and Xr+Mo were close to zero, which indicated a closer relation to the single-species biofilms. However, two genes (Xr21 and Xr22) showed an upregulation in both dual-species biofilms (Xr+Sr and Xr+Mo) and four genes (Xr23, Xr24, Xr25 and Xr26) were down-regulated in the Xr+Sr biofilms. The Xr+Pa and Xr+Sr+Mo+Pa biofilms clustered together and most genes seemed to be up- and down-regulated in both biofilm types. However, a substantial number of genes were regulated in the four-species biofilms only. However, 3 genes (Xr3, Xr4 and Xr27) displayed unique regulation in the Xr+Pa biofilms. Figure 2 NMDS plot of the Bray-Curtis dissimilarity measure between the Xanthomonas retroflexus (Xr) expression profiles of the 1000 most abundant transcripts, when grown in dual- and four-species biofilms with Stenotrophomonas rhizophila (Sr), Microbacterium oxydans (Mo) and Paenibacillus amylolyticus (Pa). Functional analysis of the differentially expressed genes The 70 genes with highly responding expression profiles were annotated using Blast against the non-redundant database at NCBI and a search for conserved domains was conducted and signal peptides were detected [22-24]. These annotation results can be observed in Table S4. Some genes annotated to functions outside the cytoplasm but did not contain a signal peptide, however some genes were truncated and the signal peptide could be missing. The differentially expressed genes were divided into functional groups, where the most prominent were membrane proteins, genes involved in regulation, and amino acid metabolisms. Table 1 lists genes encoding membrane proteins and many of these can be assigned into subgroups of potential efflux systems (Xr14, Xr18, Xr38 and Xr66), receptors (Xr19, Xr27 and Xr67), transport proteins (Xr1, Xr5, Xr10, Xr31 and Xr41), and chemotaxis proteins (Xr7 and Xr34). Genes involved in efflux systems and receptors displayed a consistent up-regulation in the Xr+Pa and the four-species biofilms compared to the singlespecies biofilms. However, Xr27 only showed an up-regulation in the Xr+Pa biofilms (Figure 3). The transport and chemotaxis subgroups showed more variable expressions with both up- and down-regulations in the Xr+Pa and four-species biofilms, except for Xr31 and Xr34, which were up-regulated in the four-species biofilms only. Few of the membrane protein-encoding genes annotated to more specific functions. A domain for CsgG (PF03783.9) was identified in Xr44, which was up-regulated in both the Xr+Pa and four-species biofilms. The CgsG domain is found in an outer membrane lipoprotein and it is thought to be limiting factor in the assembly of curli, which are adhesive fibres [26]. The Xr2 gene encodes a domain (PF08813.6), which is connected to bacterial integrins and this gene was down-regulated in the Xr+Pa and four-species biofilms. Integrins are also connected to cell adhesion [27]. Furthermore, a domain found in an outer membrane lysozyme type-c inhibitor (PF09864.4) was located in the Xr30 gene. This gene was up-regulated in the four-species biofilms, indicating resistance toward lysozyme [28]. Table 1 Differential expressed genes of Xanthomonas retroflexus (Xr) encoding for membrane proteins ID name e-val Pfam e-val sigPa Xr+Sr Xr+Mo Xr+Pa Xr+Sr+ Mo+Pa Xr1 major facilitator superfamily transmembrane nitrite extrusion 0 PF07690.11 4.7E-13 N -1.33 -0.90 -3.35 -4.11 Xr 2 hypothetical protein SMD_0239 2.00E-151 PF08813.6 5.4E-42 Y -1.05 -1.43 -3.30 -2.81 Xr5 transport-associated protein 1.00E-66 PF04972.12 5.1E-17 Y -0.19 -0.02 -3.57 -3.14 Xr7 methyl-accepting chemotaxis protein 3.00E-99 PF13426.1 2.5E-16 N -0.45 -0.63 -8.60 -5.31 Xr10 membrane protein 7.00E-112 PF08212.7 3.7E-47 N 1.10 0.50 -3.14 -5.82 Xr14 membrane protein 0 PF02321.13 3.9E-35 Y 0.76 1.80 2.63 3.41 Xr18 RND family efflux transporter MFP subunit 0 PF00529.15 1.6E-59 Y 0.84 2.29 4.76 4.92 Xr 19 TonB-dependent receptor 0 PF00593.19 1.6E-24 Y 1.88 2.10 4.21 4.83 Xr 27 TonB-dependent siderophore receptor 0 PF00593.19 4.4E-23 Y 1.10 0.34 3.07 0.30 Xr30 hypothetical protein 1.00E-131 PF09864.4 3.9E-13 Y 0.24 -0.07 0.18 3.28 Xr31 TolA protein 0 PF13103.1 1.6E-10 N 0.04 -2.01 1.08 3.56 Xr32 membrane protein 9.00E-41 - - N 0.00 -2.22 1.72 3.36 Xr34 methyl-accepting chemotaxis protein 0 PF00015.16, PF00672.20 2.6E-55, 7.2E-11 N -1.16 -1.62 1.71 3.24 Xr38 efflux transporter, RND family, MFP subunit 0 PF12700.2 1.7E-30 N -1.50 0.19 3.33 4.99 Xr41 Vitamin B12 transporter btuB 0 PF00593.19 2.2E-28 N -0.45 2.08 2.67 4.24 Xr44 hypothetical protein, partial 0 PF03783.9 1.9E-13 N 0.38 1.51 3.05 3.90 Xr66 multidrug transporter 1.00E-68 PF00893.14 9.3E-21 N -0.66 -0.03 2.96 3.52 Xr67 TonB-denpendent receptor 0 PF00593.19 1.8E-27 Y -1.19 -0.91 2.12 3.60 a Identified signal peptides in the gene: yes (Y) or no (N). Table 2 Differential expressed genes of Xanthomonas retroflexus (Xr) encoding functions in regulation a ID name e-val Pfam e-val sigPa Xr+Sr Xr+Mo Xr+Pa Xr+Sr+ Mo+Pa Xr4 hypothetical protein 2.00E-41 PF05532.7 6.1E-20 N -0.95 -0.27 -3.10 -1.26 Xr12 AbrB family transcriptional regulator 2.00E-39 PF04014.13 5E-09 N 2.62 0.10 6.83 7.47 Xr17 hypothetical protein SMD_2382 6.00E-64 PF02604.14 2.2E-09 N 1.73 1.45 4.27 3.58 Xr20 leucyl aminopeptidase 0 PF00883.16 2.1E-115 N 2.40 1.40 3.64 4.54 Xr22 HTH-type transcriptional regulator betI 4.00E-98 PF00440.18 1.9E-11 N 3.75 3.34 0.27 -0.85 Xr39 hypothetical protein Smlt2244 1.00E-84 PF01850.16 1.3E-08 N -1.83 -0.74 4.28 5.00 Xr58 ArsR family transcriptional regulator 0 PF08241.7, PF01022.15 1.1E-22, 2.5E-14 N -1.89 -0.28 1.81 5.24 Xr63 NusA antitermination factor 0 PF08529.6, PF13184.1, PF14520.1, PF00575.18 2.4E-41, 5.5E-27, 9.5E-10, 2.4E-09 N -0.27 0.49 1.75 3.42 Xr65 DEAD/DEAH box helicase 0 PF00270.24 2.4E-46 N -0.86 0.43 3.43 3.07 Xr70 ArsR family transcriptional regulator 2.00E-65 PF12840.2 2.7E-19 N -0.50 0.10 1.65 3.95 Identified signal peptides in the gene: yes (Y) or no (N). Table 3 Differential expressed genes of Xanthomonas retroflexus (Xr) encoding functions in amino acid metabolisms a ID name e-val Pfam e-val sigPa Xr+Sr Xr+Mo Xr+Pa Xr+Sr+ Mo+Pa Xr23 methylcrotonoyl-CoA carboxylase 0 PF02786.12, PF00289.17, PF00364.17 6.5E-73, 1.8E-38, 3.4E-15 N -3.85 -1.03 0.43 2.02 Xr28 aromatic amino acid aminotransferase 0 PF00155.16 2.8E-80 N 1.32 -3.09 1.91 2.67 Xr37 aminotransferase 0 PF00733.16 2E-58 N 0.00 0.58 4.05 4.19 Xr49 branched-chain amino acid aminotransferase 0 - - N 0.15 0.43 1.19 3.52 Xr51 glycine system protein H 2.00E-83 PF01597.14 2.2E-45 N 0.34 0.87 1.93 3.20 Xr56 biotin synthase 0 PF06968.8, PF04055.16 1.5E-28, 2E-15 N -3.33 0.68 2.93 3.05 Identified signal peptides in the gene: yes (Y) or no (N). Figure 3 Heatmap displaying the logarithmic (to base 2) fold change (log FC) of the expression of Xanthomonas retroflexus (Xr) genes between the single-species biofilm and each of the four different multispecies biofilms. All genes were among the 1000 most abundant Xr transcripts. They displayed a logarithmic fold change above 3 (equivalent to 8 times difference in expression levels) compared to the single-species biofilm and were tested significantly different (p-values < 0.05 with Benjamini Hochberg correction). A euclidean clustering of both the samples and the genes are displayed as similarity trees on the x- and y-axes, respectively. Table 2 summarizes all differentially expressed genes that encode regulatory functions. Five genes are involved in transcription regulation (Xr22, Xr58, Xr63, Xr65 and Xr70) and these genes were mainly up-regulated in the fourspecies biofilms compared to the single-species biofilm. An exception was Xr65, which was also up-regulated in the Xr+Po biofilms and Xr22, which was one of the two genes upregulated in the Xr+Sr and Xr+Mo biofilms (Figure 3). Toxin-antitoxin systems were also present among the differentially expressed genes and these systems have been shown to play a role in regulation [29]. An antitoxin domain (PF02604.14) was identified in Xr17. Furthermore, Xr12 encodes the AbrB/MazE antitoxin and Xr39 contains a PIN domain (PF01850.16) that is often present in toxins [29]. Xr12 and Xr39 are consecutive genes in the genome of Xr, and this indicates the presence of a complete toxin-antitoxin system in the differentially expressed genes [29]. All three genes seemed to be up-regulated in both the Xr+Pa and four-species biofilms and particular Xr12 showed a high shift in regulation level (Figure 3). Genes participating in other regulation mechanisms were also identified (Table 2). Xr4 was down-regulated only in the Xr+Pa biofilms; this gene encodes a hypothetical protein with a CsbD domain that has been connected to stress response [30]. Leucyl aminopeptidase, involved in the Glutathione metabolism, is encoded by Xr20 and up-regulation both in the Xr+Pa and the four-species biofilms was observed [31]. Glutathione is known to participate in responses to environmental factors [32]. Table 3 lists genes involved in amino acid metabolism. Branched chain amino acid aminotransferase and methylcrotonoyl-CoA carboxylase (Xr49 and Xr23) are both known for their participation in Leucine, Valine and Isoleucine metabolisms and were up-regulated in the four-species biofilms. Furthermore, X23 displayed a down-regulation in the Xr+Sr biofilm. Biotin synthase, encoded by Xr56, participates in the production of biotin; a cofactor for methylcrotonoyl-CoA and limiting to amino acid decomposition [33]. This gene was up-regulated in both the Xr+Pa and fourspecies biofilms. Additionally, Xr28 encodes an aromatic amino acid aminotransferase, which catalyses the last step in Tyrosine and Phenylalanine synthesis. Xr28 was slightly down-regulated in the Xr+Mo biofilms and upregulated in both the Xr+Pa and four-species biofilms. The aminotransferase (Xr37) contains an AsnB domain, representing an alternative pathway for Asparagine synthesis from Aspartate in the absence of AsnA and this gene was up-regulated in both the Xr+Pa and fourspecies biofilms. Xr51 encodes a glycine cleavage system protein H, which participates in the catabolism of Glycine and was upregulated in the four-species biofilms. Besides genes coding for membrane proteins or participating in regulation and amino acid metabolisms, four differentially expressed genes were annotated to peptidoglycan systems. Three of these genes (Xr24, Xr60 and Xr64) were up-regulated in the four-species biofilm, indicating an increased peptidoglycan production, furthermore, Xr24 was downregulated in the Xr+Sr biofilms. The fourth gene (Xr8), which was down-regulated in the Xr+Pa and four-species biofilms, encodes a domain (PF03734.9) often found in an L,Dtranspeptidase that has been shown to provide an alternative pathway for peptidoglycan synthesis [34]. Furthermore, the gene (Xr15) encoding a cell division protein (FtsK), which plays an important role in cell division, seemed to be up-regulated in both the Xr+Pa and fourspecies biofilm. mechanisms in multispecies biofilms [11, 35, 36]. In general, many genes involved in cell defense systems have been identified as being differentially expressed, including the membrane bound efflux systems involved in multidrug exclusion and the lysozyme type-c inhibition (Table 1). In connection to this, a beta-lactamase domain was identified in Xr52, which was up-regulated in the four-species biofilm. Additionally, an up-regulation of Xr29 in the four-species biofilm was observed, which encodes a DNA processing protein (DprA) that is involved in DNA recombination. Pa displayed high gene expression activity when coexisting with Xr in biofilms. Over 50% of the mRNA transcripts in the metatranscriptomes of the Xr+Pa and fourspecies biofilms annotated back to the genome of Pa, indicating high activity levels. Our previous study showed that Pa did not display good biofilm production abilities [1] and the biofilm production decreased in the dualspecies biofilm Xr+Pa compared to the singlespecies biofilm of Xr according to Ren et al., in review [12]. However, the cell abundance of Pa showed a significantly increase in four-species biofilm which is consistent with the high activity level of Pa in the present study. But, we should note that the relative abundance of mRNA can vary between 1-5% of the total cellular RNA and the percentage of transcripts mapping to CDSs increased in samples containing Pa, which indicated a higher mRNA/rRNA ratio in Pa compared to Xr, Sr and Mo [37]. An increased mRNA/rRNA ratio would yield more mRNAs per cell and this could introduce a bias in the assessment of the relative activity. Furthermore, Pa displayed the largest estimated genome size and this could also affect the proportion of transcripts from Pa. Discussion This study demonstrates the effect of interspecies interactions in multispecies biofilms on the gene expression in X. retroflexus (Xr). Comparisons of the expression profiles of X. retroflexus grown in single- and dual-species biofilms with S. rhizophila (Sr), M. oxydans (Mo) and P. amylolyticus (Pa), revealed shifts in gene expression, however, these changes seemed to be species dependent. Through further investigations of the fourspecies biofilms, indication of a non-linear response to multiple species was identified, suggesting complex and dependent interaction patterns. This is in good correspondence with the previously described emergent behaviour of enhanced biofilm formation by this four-species consortium [12]. The genes identified in the differential expression analysis revealed changes in expression of regulatory elements, membrane proteins and genes involved in amino acid metabolism. Only few studies exist that investigated the effect of coexistence in multispecies biofilms on a gene expression level and the results presented here contributes to further understanding of the interaction The proportion of transcripts annotating back to Sr and Mo was close to the detection limit and indicated a low activity level. The low activity level in the four-species biofilms correlated with the low cell copy numbers found by the qPCR analysis in our previous study [12]. Coexistence of different species in biofilms seemed to alter the gene expression, however, the response was species specific [11, 35]. The presence of Pa clearly induced a change in the gene expression profiles of Xr, compared to the profiles observed in the single-species biofilm and was responsible for the induced regulation of the majority of the 70 highly differentially expressed genes. It did show high activity levels in the biofilms and this could be causing the response observed in the Xr expression. The induced gene regulation in the dual-species biofilm of Xr+Sr and Xr+Mo was less evident. However, the Xr+Mo replicates did create a cluster, which separated from the single-species biofilm and some genes were differently regulated in their presence. This was consistent with the biofilm formation of Xr+Mo which significantly increased compared with the single-species biofilm of Xr, while the Xr+Sr biofilms did not show any significantly increase in biomass [12]. Transcripts from Sr and Mo constituted a smaller proportion of the metatranscriptomes and showed low abundance in the previous study of the four-species biofilm composition [12]. The relative low abundance and activity could explain the low effect of the presence of Sr and Mo on the Xr gene expression. Interestingly, when inferring the four-species biofilms, a unique expression pattern for Xr emerged, with regulations uniquely observed for this combination. The five biological replicates of the four-species biofilms created a tight cluster that separated from the other biofilm samples (Figure 2) and several genes displayed expression shifts, which were not observed in any of the dual-species biofilms (Figure 3). Hence, the co-occurrence of Sr, Mo and Pa induced a different expression pattern in Xr, compared to the gene expression in the dual-species biofilms, which is consistent with our previous results that each strain is indispensable for the synergistic interactions in this four-species biofilm [12]. This observation was not evident in a gene expression study of Saccharomyces cerevisiae exposed to multiple physiological transitions [38]. They showed that when exposed to different simultaneous stresses, the cellular response approximated the sum of the responses for each individual stress and indicated a linear response to multiple physiological changes [38]. The current result differs from these observations and suggests a non-linear response to multiple species, indicating that gene expression is regulated by a complex interspecies interaction network. The functional annotation of the 70 significantly differentially expressed genes revealed expression changes of regulatory elements in the presence of Pa [11]. Genes coding for transcription factors and toxinantitoxin system displayed an increase in expression in Xr. Especially the potential MazE/MazF toxin-antitoxin system was upregulated and has been reported to be a general regulatory element and involved in biofilm formation [39]. Also, the MazE/MazF system has been connected with programmed cell death and lysis, which was widely reported in Xanthomonas [40, 41]. Furthermore, genes encoding transcriptional regulators show unique expression levels in the four-species biofilm only. These changes in expression of regulatory elements possibly control the phenotypic adaptation of the cells to the community composition in the biofilms [42]. The regulation of genes encoding membrane proteins with roles in transport, efflux systems, receptors and chemotaxis indicated that many of the phenotypic changes induced by coexistence in biofilm were in relation to the surrounding microenvironment [11]. Some of the identified transporters could potentially play a role in secreting extracellular polymeric substances (EPS). Furthermore, the upregulation of the potential cell division protein, the potential participant in curli production and chemotaxis systems in the presence of Pa could also indicate higher growth rate, adhesion and specific growth direction. An increase in EPS excretion and directional growth could be a competitive response to coexistence and an advance for Xr [43]. A further indication of enhanced growth was the observed upregulation of peptidoglycan production in the four-species biofilm only. All these observations are correlated with previous quantitative PCR (qPCR) measurements which demonstrated the increase in cell number of Xr when co-cultured with other three species [12]. Up-regulation of potential defense system like efflux pump was evident in the presence of Pa and a potential lysozyme type-c inhibitor together with a potential beta-lactamase were up-regulated in the four-species biofilm. The up-regulation of these genes indicates an upgrade in defense mechanisms against toxins in this multispecies biofilm [44], which could positively affect the total resistance of the biofilm. One of the main emergent behaviour of bacterial biofilm is enhanced antibiotic resistance. This is usually thought to be due to lack of diffusion and low growth rate [45, 46]. However our results may indicate that this is also due to specific induction of resistance trait. The expression changes in amino acid metabolisms represented a potential interspecies cooperation in the multispecies biofilms [36]. The aromatic amino acid aminotransferase displayed an up-regulation in the presence of Pa. Through further inspection, the function of this gene is not immediately present in the genome of Pa, hence, an enhanced production and export of tyrosine and phenylalanine could constitute an example of interspecies cooperation. Furthermore, the catabolism of valine, leucine, isoleucine and glycine seemed to be up-regulated in the fourspecies biofilm and this could indicate an increased exchange of amino acids or polypeptides. The interspecies interaction could be further investigated by investigating the expression profiles of Sr, Mo and Pa. The current comparative gene expression study of X. retroflexus (Xr) grown in single-, dualand four-species biofilms with S. rhizophila (Sr), M. oxydans (Mo) and P. amylolyticus (Pa) displayed phenotypic adaptation according to the species composition in the biofilms. By inferring the differentially expressed genes, the adaptations seemed to respond to both positive and negative relations. It can be hypothesized that interspecies interactions in biofilms are balances between cooperation and competition. To further investigate the effects of interspecies interactions, the phenotypic changes in knockout mutants of candidate genes identified in the comparative gene expression analysis might further elucidate their functional effect on interspecies interaction in multispecies biofilms. References 1. Ren D, Madsen JS, de la Cruz-Perera CI, Bergmark L, Sørensen SJ, Burmølle M (2013) High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCR-Based Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. Microbial ecology: 1-9. 2. Hall-Stoodley L, Costerton JW, Stoodley P (2004) Bacterial biofilms: from the natural environment to infectious diseases. Nature Reviews Microbiology 2: 95-108. 3. Burmølle M, Ren D, Bjarnsholt T, Sørensen SJ (2014) Interactions in multispecies biofilms: do they actually matter? Trends in microbiology. doi: 10.1016/j.tim.2013.12.004 4. Vornhagen J, Stevens M, McCormick DW, Dowd SE, Eisenberg JN, Boles BR, Rickard AH (2013) Coaggregation occurs amongst bacteria within and between biofilms in domestic showerheads. Biofouling 29: 53-68. 5. Farabegoli G, Chiavola A, Rolle E (2008) Remediation of chlorophenol-and phenolcontaminated groundwater by a sequencing batch biofilm reactor. 6. Schwering M, Song J, Louie M, Turner RJ, Ceri H (2013) Multi-species biofilms defined from drinking water microorganisms provide increased protection against chlorine disinfection. Biofouling 29: 917-928. 7. Høiby N, Bjarnsholt T, Givskov M, Molin S, Ciofu O (2010) Antibiotic resistance of bacterial biofilms. International journal of antimicrobial agents 35: 322-332. 8. Seneviratne G, Zavahir J, Bandara W, Weerasekara M (2008) Fungal-bacterial biofilms: their development for novel biotechnological applications. World Journal of Microbiology and Biotechnology 24: 739-743. 9. LoVLEY DR, Klug MJ (1986) Model for the distribution of sulfate reduction and methanogenesis in freshwater sediments. Geochimica et Cosmochimica Acta 50: 11-18. 10. Harrison F (2007) Microbial ecology of the cystic fibrosis lung. Microbiology 153: 917-923. 11. Frias-Lopez J, Duran-Pinedo A (2012) Effect of periodontal pathogens on the metatranscriptome of a healthy multispecies biofilm model. Journal of bacteriology 194: 2082-2095. 12. Ren D, Madsen JS, Sørensen SJ, Burmølle M (2013) High prevalence of biofilm synergy among bacterial soil isolates in co-cultures indicates bacterial interspecific cooperation. The ISME journal. Under review. 13. Fierer N, Jackson JA, Vilgalys R, Jackson RB (2005) Assessment of soil microbial community structure by use of taxon-specific quantitative PCR assays. Applied and Environmental Microbiology 71: 4117-4120. 14. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL (2009) The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clinical chemistry 55: 611-622. 15. Bergmark L, Poulsen PHB, Al‐Soud WA, Norman A, Hansen LH, Sørensen SJ (2012) Assessment of the specificity of Burkholderia and Pseudomonas qPCR assays for detection of these genera in soil using 454 pyrosequencing. FEMS microbiology letters 333: 77-84. 16. Park J-W, Crowley DE (2005) Normalization of soil DNA extraction for accurate quantification of target genes by real-time PCR and DGGE. BioTechniques 38: 579-586. 17. VanGuilder HD, Vrana KE, Freeman WM (2008) Twenty-five years of quantitative PCR for gene expression analysis. Biotechniques 44: 619. 18. Zerbino DR, Birney E (2008) Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome research 18: 821-829. 19. Hyatt D, Chen G-L, Locascio PF, Land ML, Larimer FW, Hauser LJ (2010) Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC bioinformatics 11: 119-119. 20. Langmead B, Salzberg SL (2012) Fast gappedread alignment with Bowtie 2. Nature methods 9: 357-359. 21. Robinson MD, McCarthy DJ, Smyth GK (2010) edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics (Oxford, England) 26: 139140. 22. Altschup SF, Gish W, Pennsylvania T, Park U (1990) Basic Local Alignment Search Tool. J mol biol 215: 403-410. 23. Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, Boursnell C, Pang N, Forslund K, Ceric G, Clements J, Heger A, Holm L, Sonnhammer ELL, Eddy SR, Bateman A, Finn RD (2012) The Pfam protein families database. Nucleic acids research 40: D290-301. 24. Petersen TN, Brunak S, von Heijne G, Nielsen H (2011) SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat Meth 8: 785-786. 25. McCarthy DJ, Chen Y, Smyth GK (2012) Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic acids research 40: 4288-4297. 26. Loferer H, Hammar M, Normark S (1997) Availability of the fibre subunit CsgA and the nucleator protein CsgB during assembly of fibronectin-binding curli is limited by the intracellular concentration of the novel lipoprotein CsgG. Molecular microbiology 26: 11-23. 27. Kelly G, Prasannan S, Daniell S, Fleming K, Frankel G, Dougan G, Connerton I, Matthews S (1999) Structure of the cell-adhesion fragment of intimin from enteropathogenic Escherichia coli. Nature structural biology 6: 313-318. 28. Callewaert L, Aertsen A, Deckers D, Vanoirbeek KGa, Vanderkelen L, Van Herreweghe JM, Masschalck B, Nakimbugwe D, Robben J, Michiels CW (2008) A new family of lysozyme inhibitors contributing to lysozyme tolerance in gram-negative bacteria. PLoS pathogens 4: e1000019-e1000019. 29. Anantharaman V, Aravind L (2003) New connections in the prokaryotic toxin-antitoxin network: relationship with the eukaryotic nonsense-mediated RNA decay system. Genome biology 4: R81-R81. 30. Prágai Z, Harwood CR (2002) Regulatory interactions between the Pho and sigma σBdependent general stress regulons of Bacillus subtilis. Microbiology (Reading, England) 148: 1593-1602. 31. Cappiello M, Lazzarotti A, Buono F, Scaloni A, D'Ambrosio C, Amodeo P, Méndez BL, Pelosi P, Del Corso A, Mura U (2004) New role for leucyl aminopeptidase in glutathione turnover. The Biochemical journal 378: 35-44. 32. Riccillo PM, Muglia CI, Bruijn FJD, Roe AJ, Booth IR, Aguilar OM, Bruijn FJDE, Booth IANR (2000) Glutathione Is Involved in Environmental Stress Responses in Rhizobium tropici, Including Acid Tolerance Glutathione Is Involved in Environmental Stress Responses in Rhizobium tropici, Including Acid Tolerance. doi: 10.1128/JB.182.6.1748-1753.2000. 33. Wood HG, Barden RE (1977) Biotin enzymes. Ann. Rev. Bioehem. 46: 385-413. 34. Biarrotte-Sorin S, Hugonnet J-E, Delfosse V, Mainardi J-L, Gutmann L, Arthur M, Mayer C (2006) Crystal structure of a novel beta-lactaminsensitive peptidoglycan transpeptidase. Journal of molecular biology 359: 533-538. 35. Wen ZT, Yates D, Ahn S-J, Burne Ra (2010) Biofilm formation and virulence expression by Streptococcus mutans are altered when grown in dual-species model. BMC microbiology 10: 111-111. 36. Jakubovics NS, Gill SR, Iobst SE, Vickerman MM, Kolenbrander PE (2008) Regulation of gene expression in a mixed-genus community: stabilized arginine biosynthesis in Streptococcus gordonii by coaggregation with Actinomyces naeslundii. Journal of bacteriology 190: 36463657. 37. Niedhardt FC, Umbarger HE (1996) Chemical composition of Escherichia coli. In: Curtiss, R, III, Ingraham, JL, Lin, ECC, Low, KB, Magasanik, B, Regnikoff, WS (eds.). American Society of Microbiology, Washington D. C., pp. 13-16. 38. Gasch aP, Spellman PT, Kao CM, Carmel-Harel O, Eisen MB, Storz G, Botstein D, Brown PO (2000) Genomic expression programs in the response of yeast cells to environmental changes. Molecular biology of the cell 11: 4241-4257. 39. Wang X, Wood TK (2011) Toxin-antitoxin systems influence biofilm and persister cell formation and the general stress response. Applied and environmental microbiology 77: 5577-5583. 40. Engelberg-Kulka H, Hazan R, Amitai S (2005) mazEF: a chromosomal toxin-antitoxin module that triggers programmed cell death in bacteria. Journal of cell science 118: 4327-4332. 41. Rice KC, Bayles KW (2008) Molecular control of bacterial death and lysis. Microbiology and molecular biology reviews: MMBR 72: 85-109, table of contents 42. López-Maury L, Marguerat S, Bähler J (2008) Tuning gene expression to changing environments: from rapid responses to evolutionary adaptation. Nature reviews Genetics 9: 583-593. 43. Xavier JB, Foster KR (2007) Cooperation and conflict in microbial biofilms. Proceedings of the National Academy of Sciences of the United States of America 104: 876-881. 44. Yang L, Liu Y, Wu H, Hóiby N, Molin S, Song Z-j (2011) Current understanding of multispecies biofilms. International journal of oral science 3: 74-81. 45. Harrison JJ, Turner RJ, Ceri H (2005) Persister cells, the biofilm matrix and tolerance to metal cations in biofilm and planktonic Pseudomonas aeruginosa. Environmental Microbiology 7: 981-994. 46. Evans D, Allison D, Brown M, Gilbert P (1991) Susceptibility of Pseudomonas aeruginosa and Escherichia coli biofilms towards ciprofloxacin: effect of specific growth rate. Journal of Antimicrobial Chemotherapy 27: 177-184. Supplementary tables and figures Table S1 Genome assembly statistics a Strain Trim (Mb) N50a Max. contig length (CL) Min. CL Mean CL contigs Genome size (Mb) CDSb Xanthomonas retroflexus (Xr) 441.58 107,385 410,416 984 49,439 94 4.6 4,205 Stenotrophomonas rhizophila (Sr) 210.48 5351 28,238 1,577 4,957 501 2.5 2,607 Microbacterium Oxydans (Mo) 531.54 1,156,050 1,214,131 540 307,568 13 4 3,799 Paenibacillus amylolyticus (Pa) 1,475.01 36,803 168,316 508 18,823 335 6.3 5,584 N50: 50% of the entire assembly is contained in contigs or scaffolds. Coding sequences. b Table S2 Metatranscriptome statistics Sample Raw (Mb)* Trim (Mb)** % Mapped (Mb)*** % CDS (Mb)**** % Xr a# 440.14 345.11 78.41% 337.03 97.66% 3.88 1.15% Xr b 551.73 432.69 78.42% 418.96 96.83% 4.20 1.00% Xr c 641.36 512.55 79.92% 498.34 97.23% 5.34 1.07% Xr+Sr a 720.76 524.87 72.82% 505.63 96.33% 8.02 1.59% Xr+Sr b 881.80 654.41 74.21% 631.82 96.55% 15.59 2.47% Xr+Sr c 805.52 573.65 71.21% 549.14 95.73% 5.03 0.92% Xr+Mo a 778.30 519.54 66.75% 492.34 94.76% 7.07 1.44% Xr+Mo b 1,129.83 807.85 71.50% 773.56 95.76% 6.01 0.78% Xr+Mo c 593.42 413.70 69.71% 395.37 95.57% 3.76 0.95% Xr+Pa a 1,058.75 797.95 75.37% 712.08 89.24% 34.40 4.83% Xr+Pa b 998.45 690.20 69.13% 611.14 88.54% 23.82 3.90% Xr+Pa c 951.04 736.32 77.42% 671.19 91.16% 28.66 4.27% Xr+Sr+Mo+Pa a 1,753.43 1,392.47 79.41% 1236.95 88.83% 39.89 3.23% Xr+Sr+Mo+Pa b 1,202.92 969.95 80.63% 847.00 87.32% 24.61 2.91% Xr+Sr+Mo+Pa c 980.82 785.00 80.03% 724.87 92.34% 21.70 2.99% Xr+Sr+Mo+Pa d 1,324.05 1,044.41 78.88% 935.45 89.57% 23.82 2.55% Xr+Sr+Mo+Pa e 895.13 708.44 79.14% 661.58 93.38% 15.33 2.32% * ** The number of mega bases from the raw output of the sequencer. Mega bases remaining after the quality filtering process. Mega bases mapped to the genomes. ****Mega bases maped to protein coding sequences. *** # 3 replicates for Xr, Xr+Sr, Xr+Mo and Xr+Pa biofilms; 5 replicates for Xr+Sr+Mo+Pa biofilm. Table S3: Pearson correlation between the samples Xr a Xr b Xr c Xr+Sr Xr+Sr Xr+Sr Xr+Mo Xr+Mo Xr+Mo Xr+Pa Xr+Pa Xr+Pa Xr+Sr Xr+Sr Xr+Sr Xr+Sr Xr+Sr a b c a b c a b c +Mo+ +Mo+ +Mo+ +Mo+ +Mo+ Pa a Pa b Pa c Pa d Pa e Xr a# 1.00 0.95 0.95 0.88 0.91 0.87 0.96 0.92 0.97 0.89 0.89 0.89 0.25 0.30 0.24 0.29 0.27 Xr b 0.95 1.00 0.97 0.87 0.85 0.92 0.93 0.93 0.92 0.87 0.87 0.85 0.33 0.38 0.32 0.37 0.35 Xr c 0.95 0.97 1.00 0.92 0.89 0.95 0.93 0.92 0.93 0.84 0.86 0.84 0.32 0.36 0.31 0.36 0.32 Xr+Sr a 0.88 0.87 0.92 1.00 0.95 0.92 0.85 0.91 0.88 0.72 0.74 0.73 0.23 0.28 0.23 0.27 0.25 Xr+Sr b 0.91 0.85 0.89 0.95 1.00 0.85 0.89 0.87 0.93 0.77 0.80 0.77 0.15 0.21 0.15 0.19 0.19 Xr+Sr c 0.87 0.92 0.95 0.92 0.85 1.00 0.87 0.91 0.88 0.80 0.80 0.79 0.42 0.44 0.41 0.46 0.39 Xr+Mo a 0.96 0.93 0.93 0.85 0.89 0.87 1.00 0.91 0.98 0.89 0.90 0.85 0.30 0.34 0.28 0.33 0.31 Xr+Mo b 0.92 0.93 0.92 0.91 0.87 0.91 0.91 1.00 0.90 0.80 0.80 0.75 0.32 0.36 0.31 0.36 0.33 Xr+Mo c 0.97 0.92 0.93 0.88 0.93 0.88 0.98 0.90 1.00 0.90 0.91 0.88 0.26 0.31 0.25 0.30 0.28 Xr+Pa a 0.89 0.87 0.84 0.72 0.77 0.80 0.89 0.80 0.90 1.00 0.98 0.96 0.44 0.46 0.42 0.46 0.42 Xr+Pa b 0.89 0.87 0.86 0.74 0.80 0.80 0.90 0.80 0.91 0.98 1.00 0.95 0.40 0.43 0.39 0.42 0.39 Xr+Pa c 0.89 0.85 0.84 0.73 0.77 0.79 0.85 0.75 0.88 0.96 0.95 1.00 0.43 0.44 0.42 0.45 0.40 Xr+Sr+Mo+Pa a 0.25 0.33 0.32 0.23 0.15 0.42 0.30 0.32 0.26 0.44 0.40 0.43 1.00 0.92 0.98 0.98 0.87 Xr+Sr+Mo+Pa b 0.30 0.38 0.36 0.28 0.21 0.44 0.34 0.36 0.31 0.46 0.43 0.44 0.92 1.00 0.93 0.96 0.98 Xr+Sr+Mo+Pa c 0.24 0.32 0.31 0.23 0.15 0.41 0.28 0.31 0.25 0.42 0.39 0.42 0.98 0.93 1.00 0.98 0.91 Xr+Sr+Mo+Pa d 0.29 0.37 0.36 0.27 0.19 0.46 0.33 0.36 0.30 0.46 0.42 0.45 0.98 0.96 0.98 1.00 0.93 Xr+Sr+Mo+Pa e 0.27 0.35 0.32 0.25 0.19 0.39 0.31 0.33 0.28 0.42 0.39 0.40 0.87 0.98 0.91 0.93 1.00 # 3 replicates for Xr, Xr+Sr, Xr+Mo and Xr+Pa biofilms; 5 replicates for Xr+Sr+Mo+Pa biofilm. Table S4 70 significant differentially expressed genes with a logFC above 3 ID Acc. name e-val Pfams e-val sig-P Xr+Sr Xr+Mo Xr+Pa Xr+Sr +Mo+ Pa Xr1 YP_001972537 major facilitator superfamily transmembrane nitrite extrusion 0.00E+00 PF07690.11 4.7E-13 N -1.33 -0.90 -3.35 -4.11 Xr2 YP_006183023 hypothetical protein SMD_0239 2.00E-151 PF08813.6 5.4E-42 Y -1.05 -1.43 -3.30 -2.81 Xr3 YP_004794138 hypothetical protein 4.00E-99 PF09537.5 4.6E-29 N -0.96 0.04 -3.13 -1.78 Xr4 WP_019661831 hypothetical protein 2.00E-41 PF05532.7 6.1E-20 N -0.95 -0.27 -3.10 -1.26 Xr5 WP_005419358 transport-associated protein 1.00E-66 PF04972.12 5.1E-17 Y -0.19 -0.02 -3.57 -3.14 Xr6 WP_019184261 hypothetical protein 5.00E-110 PF05974.7 1.1E-49 N 0.01 -0.12 -3.28 -2.49 Xr7 YP_001972524 methyl-accepting chemotaxis protein 3.00E-99 PF13426.1 2.5E-16 N -0.45 -0.63 -8.60 -5.31 Xr8 CCP12806 hypothetical protein SMSKK35_4264 0 PF03734.9, PF01471.13 3.2E-17, 6.6E-10 Y -0.13 -0.07 -3.91 -8.60 Xr9 WP_019659358 pyridoxamine 5'-phosphate oxidase 5.00E-94 PF01243.15 0.00017 N 0.16 0.33 -4.24 -5.19 Xr10 WP_021201924 membrane protein 7.00E-112 PF08212.7 3.7E-47 N 1.10 0.50 -3.14 -5.82 Xr11 WP_010484413 glyoxalase 1.00E-73 PF12681.2 7.6E-8 N 0.26 -0.24 -3.06 -5.25 Xr12 YP_002028224 AbrB family transcriptional regulator 2.00E-39 PF04014.13 5.0E-9 N 2.62 0.10 6.83 7.47 Xr13 WP_021203640 azurin 2.00E-100 PF00127.15 1.1E-19 Y -0.13 3.07 4.46 6.72 Xr14 WP_008266809 membrane protein 0.00E+00 PF02321.13 3.9E-35 Y 0.76 1.80 2.63 3.41 Xr15 YP_002028320 cell division protein FtsK 0.00E+00 PF01580.13 1.2E-65 N 1.55 1.91 2.83 3.80 Xr16 YP_001973969 30S ribosomal protein S9 1.00E-86 PF00380.14 6.3E-47 N 1.08 0.76 3.53 3.55 Xr17 YP_006185095 hypothetical protein SMD_2382 6.00E-64 PF02604.14 2.2E-09 N 1.73 1.45 4.27 3.58 Xr18 YP_002029850 RND family efflux transporter MFP subunit 0.00E+00 PF00529.15 1.6E-59 Y 0.84 2.29 4.76 4.92 Xr19 WP_019662033 TonB-dependent receptor 0.00E+00 PF00593.19 1.6E-24 Y 1.88 2.10 4.21 4.83 Xr20 YP_002026924 leucyl aminopeptidase 0 PF00883.16 2,1E-115 N 2,40 1,40 3,64 4,54 Xr21 YP_002029333 phospholipase D/transphosphatidylase 0,00E+00 PF13091.1 1,3E-15 N 3,44 3,62 0,44 -0,75 Xr22 WP_006397171 HTH-type transcriptional regulator betI 4,00E-98 PF00440.18 1,9E-11 N 3,75 3,34 0,27 -0,85 Xr23 YP_004790752 methylcrotonoyl-CoA carboxylase 0 PF02786.12 6.5E-73, 1.8E-38, 3.4E-15 N -3,85 -1,03 0,43 2,02 Xr24 YP_002027511 integral membrane protein MviN 0,00E+00 PF03023.9 3,2E-122 N -3,96 -0,31 -0,42 2,00 Xr25 YP_006186711 porphobilinogen synthase 0,00E+00 PF00490.16 1,6E-146 N -4,03 -2,32 0,62 2,51 Xr26 YP_004791464 PepSY-associated TM helix domain-containing protein 0,00E+00 PF13703.1 6,5E-19 N -3,17 -1,56 1,36 2,95 Xr27 YP_004791875 TonB-dependent siderophore receptor 0,00E+00 PF00593.19 4,4E-23 Y 1,10 0,34 3,07 0,30 Xr28 YP_002026409 aromatic amino acid aminotransferase 0,00E+00 PF00155.16 2,8E-80 N 1,32 -3,09 1,91 2,67 Xr29 WP_019660782 DNA processing protein DprA 0,00E+00 PF02481.10 2,3E-72 N -0,34 -0,82 0,35 3,23 Xr30 WP_006363899 hypothetical protein 1,00E-131 PF09864.4 3,9E-13 Y 0,24 -0,07 0,18 3,28 Xr31 WP_005414275 TolA protein 0,00E+00 PF13103.1 1,6E-10 N 0,04 -2,01 1,08 3,56 Xr32 WP_017354885 membrane protein 9,00E-41 - - N inf -2,22 1,72 3,36 Xr33 WP_005415399 4-diphosphocytidyl-2-Cmethyl-D-erythritol kinase 4,00E-175 PF00288.21 9,5E-10 N -0,68 -1,57 2,15 3,96 Xr34 WP_005418317 methyl-accepting chemotaxis protein 0 PF00015.16, PF00672.20 2.6E-55, 7.2E-11 N -1,16 -1,62 1,71 3,24 Xr35 NP_639162 30S ribosomal protein S21 7,00E-41 PF01165.15 2,7E-24 N -0,14 -0,38 3,15 6,35 Xr36 WP_004149880 methionyl aminopeptidase 4,00E-175 PF00557.19 4,3E-47 N inf -0,45 3,93 5,07 Xr37 WP_021203383 aminotransferase 0 PF00733.16 2E-58 N inf 0,58 4,05 4,19 Xr38 WP_005414536 efflux transporter, RND family, MFP subunit 0,00E+00 PF12700.2 1,7E-30 N -1,50 0,19 3,33 4,99 Xr39 YP_001972041 hypothetical protein Smlt2244 1,00E-84 PF01850.16 0,00000001 3 N -1,83 -0,74 4,28 5,00 Xr40 WP_006402337 cytochrome bd-I oxidase subunit I 0,00E+00 PF01654.12 1,5E-159 N -1,75 -0,85 3,37 4,45 Xr41 CCP09567 Vitamin B12 transporter btuB 0 PF00593.19 2,2E-28 N -0,45 2,08 2,67 4,24 Xr42 YP_002029700 inorganic pyrophosphatase 3,00E-125 PF00719.14 5,4E-59 N 0,62 1,80 3,45 3,84 Xr43 YP_002029231 hypothetical protein Smal_2848 0 - - N 0,65 1,52 2,68 4,55 Xr44 WP_005415097 hypothetical protein, partial 0,00E+00 PF03783.9 1,9E-13 N 0,38 1,51 3,05 3,90 Xr45 WP_019661298 cytidylate kinase 2,00E-152 PF02224.13 1,9E-52 N 0,21 1,32 2,77 3,94 Xr46 YP_001973281 angiotensin-converting peptidyl dipeptidase protein 0 PF01401.13 1,5E-110 Y 0,37 1,54 1,87 3,02 Xr47 YP_002030341 hypothetical protein Smal_3959 0,00E+00 PF05960.6 1,7E-159 Y 0,56 1,49 1,64 3,12 Xr48 CCP09956 NAD synthetase 0,00E+00 PF02540.12, PF00795.17 1.7E-69, 4E-16 Y 0,75 1,58 1,72 3,24 Xr49 WP_004144913 branched-chain amino acid aminotransferase 0,00E+00 - - N 0,15 0,43 1,19 3,52 Xr50 YP_006184128 macrophage infectivity potentiator 1,00E-162 PF01346.13 1,4E-29 Y 0,12 0,79 2,18 3,25 Xr51 YP_002029454 glycine cleavage system protein H 2,00E-83 PF01597.14 2,2E-45 N 0,34 0,87 1,93 3,20 Xr52 WP_010340581 hypothetical protein 4,00E-159 PF00753.22 0,000013 N 0,66 0,91 1,81 3,25 Xr53 WP_017356008 30S ribosomal protein S12 6,00E-83 PF00164.20 1E-53 N 0,86 0,55 2,10 4,81 Xr54 WP_004142671 phosphoribosylformylglycinami dine synthase 0,00E+00 PF13507.1, PF02769.17 5.7E-100, 5.1E-28 N 1,14 0,25 2,01 3,12 Xr55 YP_002029200 hypothetical protein Smal_2817 8,00E-132 PF02576.12 1,8E-41 N 0,34 0,07 2,78 3,67 Xr56 YP_006186679 biotin synthase 0,00E+00 PF06968.8, PF04055.16 1.5E-28, 2E-15 N -3,33 0,68 2,93 3,05 Xr57 YP_003017703 sulfatase 0,00E+00 PF00884.18 1,2E-47 N -2,51 -0,96 2,62 3,02 Xr58 YP_002028993 ArsR family transcriptional regulator 0,00E+00 PF08241.7, PF01022.15 1.1E-22, 2.5E-14 N -1,89 -0,28 1,81 5,24 Xr59 CCP13965 3-hydroxydecanoyl-(acyl carrier protein) dehydratase 3,00E-120 PF07977.8 1,8E-47 N -2,32 -0,17 2,27 5,04 Xr60 YP_001970648 D-alanine--D-alanine ligase 0,00E+00 PF07478.8 3,5E-53 N -1,63 0,66 1,93 3,79 Xr61 WP_010481806 alkyl hydroperoxide reductase subunit F 0,00E+00 PF07992.9, PF01063.14, PF13192.1 6.7E-32, 5.3E-29, 6.9E-08 N -0,42 1,02 1,48 3,27 Xr62 YP_001970812 30S ribosomal protein S4 3,00E-150 PF00163.14 4,2E-25 N -0,88 0,57 1,79 3,28 Xr63 YP_004793386 NusA antitermination factor 0,00E+00 PF08529.6, PF13184.1, PF14520.1, PF00575.18 2.4E-41, 5.5E-27, 9.5E-10, 2.4E-09 N -0,27 0,49 1,75 3,42 Xr64 YP_002029828 serine-type D-Ala-D-Ala carboxypeptidase 0,00E+00 PF00768.15, PF07943.8 2.6E-75, 9.2E-24 N -0,33 0,40 2,04 3,18 Xr65 YP_002026868 DEAD/DEAH box helicase 0 PF00270.24 2,4E-46 N -0,86 0,43 3,43 3,07 Xr66 WP_006450325 multidrug transporter 1,00E-68 PF00893.14 9,3E-21 N -0,66 -0,03 2,96 3,52 Xr67 WP_005408347 TonB-denpendent receptor 0,00E+00 PF00593.19 1,8E-27 Y -1,19 -0,91 2,12 3,60 Xr68 YP_002029000 FKBP-type peptidylprolyl isomerase 0,00E+00 PF01346.13 8,3E-26 Y -1,12 -0,44 2,43 3,53 Xr69 WP_005408268 trigger factor 0,00E+00 PF05697.8, PF00254.23 2.6E-38, 1.2E-04 N -0,76 -0,47 1,63 3,47 Xr70 YP_002027450 ArsR family transcriptional regulator 2,00E-65 PF12840.2 2,7E-19 N -0,50 0,10 1,65 3,95 Manuscript 4 Effects of grazing by flagellate Neocercomonas jutlandica on mono- and multispecies biofilms Effects of grazing by flagellate Neocercomonas jutlandica on mono- and multispecies biofilms Dawei Rena, Flemming Ekelundb, Søren J. Sørensena,*, Mette Burmøllea,* a Section of Microbiology, Department of Biology, University of Copenhagen b Section of Terrestrial Ecology, Department of Biology, University of Copenhagen *Corresponding authors Postal address: Universitetsparken 15, Bygn. 1, 2100 København Ø, Denmark e-mail Mette Burmølle: [email protected] e-mail Søren J. Sørensen: [email protected] Key words: Biofilm / protozoan grazing / synergism / predator-prey interaction Abstract Protozoan grazing is considered to be a major regulating factor for bacterial populations in agricultural soil. However, many soil inhabiting bacteria form biofilms and little is known about the impact of predation as well as predator-bacterial prey interactions on biofilm dynamics. Here, we used species specific quantitative PCR, to test effects of the heterotrophic flagellate Neocercomonas jutlandica on biofilm formation. We worked with both a monospecies biofilm formed by Xanthomonas retroflexus and a multispecies biofilm formed by soil isolates Stenotrophomonas rhizophila, Xanthomonas retroflexus, Microbacterium oxydans and Paenibacillus amylolyticus. The presence of N. jutlandica co-cultured with X. retroflexus increased the abundances of X. retroflexus in both mono- and multi-species biofilms. While, N. jutlandica co-cultured with the mixture of four bacteria could reduce the abundance of each species significantly in the four-species biofilm. These conflicting observations indicate that the performance of protozoan grazing greatly relies on predator-prey interactions, which probably induces protozoa with different grazing abilities. Additionally, the synergistic interactions in this multispecies biofilm did not afford more protection against predation compared with X. retroflexus biofilm. The constant ratio of cell numbers among X. retroflexus, M. oxydans and P. amylolyticus was observed regardless of protozoan grazing, suggesting these three species were spatially arranged in integrated communities rather than individual microcolonies. However, these conclusions are based on the assumption that this flagellate predator prefers surface attached cells. The experimental design needs to be optimized and further studies should be conducted to verify these preliminary results. Furthermore, by incorporating other types of protozoa, a more comprehensive assessment of protozoa-biofilm interactions can be achieved. Introduction Biofilms, i.e. microbial aggregates associated with surfaces, are ubiquitous in nature [1]. Typically, they contain multiple species of bacteria, as well as archaea, fungi, algae and protozoa. Protozoa occur ubiquitously and are an important regulating factor for bacteria growth and survival in nature [2]. The dominant groups of protozoa in soil are normally naked amoebae and heterotrophic flagellates (HF) [3]. Heterotrophic flagellates which utilize long flagella for propulsion are the pioneer colonizer of the surfaces and bacterial biofilms due to their high mobility and abundance [4]. In agricultural soil, heterotrophic flagellates are able to graze bacteria in small pores because of their small size and flexible cells and thus, play an important role in the control of bacterial communities as well as mineralization and nutrient cycling by releasing the nitrogen taken by bacteria [2]. The evidences that protozoan grazing could shape biofilm morphology and spatial arrangement of bacterial cells, such as stimulating microcolonies formation, reducing maximal and basal layer thickness and altering mass transfer of nutrients, have been widely reported [5, 6]. A generally held opinion is that by growing in biofilms, bacteria can be offered a protective niche against protozoan grazing, owing to the physical protection by secreted polymeric matrix [7] and enhanced coordination by horizontal gene transfer and quorum sensing [6, 8]. However, the counter argument is also presented by the reports that some protozoa showed marked preferences for attached and aggregated bacteria, such as the naked amoebae and many of the common heterotrophic flagellates, as demonstrated for Rhyncomonas nasuta and species of the genus Bodo [9, 10]. Moreover, some pathogens, such as Legionella, could multiply and become more virulent and resistant after ingested by protozoa in biofilms [11, 12], indicating the crucial role of protozoa and biofilms in the evolution of virulence factors in pathogenic bacteria. Despite the feeding interactions between protozoa and planktonic bacteria are well understood [13-15], only a handful of studies have attempted to assess the grazing impact on biofilms, especially multispecies biofilms, where interspecies interactions play an important role in determining the structure, function and dynamics of biofilms and probably are involved in the defense mechanism of bacterial biofilms against protozoan grazing [16, 17]. Here, we combine species specific quantitative PCR and a multispecies biofilm model, which consists of four soil isolates and shows synergism in biofilm formation. Then, the effects of grazing by a flagellate on the population dynamics of this multispecies biofilm and a single-species biofilm were quantified and compared, which could provide valuable information about the impact of interspecies interactions on the biofilm resilience to grazing pressure in soil. Materials and methods Organisms and growth conditions The bacterial strains Stenotrophomonas rhizophila (JQ890538), Xanthomonas retroflexus (JQ890537), Microbacterium oxydans (JQ890539), and Paenibacillus amylolyticus (JQ890540) were isolated from agricultural soil as described previously [18]. Bacteria were routinely grown from frozen glycerol stocks on Tryptic Soy Agar plates for 48h at 24°C. Thereafter, colonies were transferred into full strength TSB (Tryptic Soy Broth), incubated with shaking (250 rpm) at 24°C overnight for biofilm growth assay, or into one-tenth strength TSB, incubated with shaking (250 rpm) at 24°C for 48 h and then used as bacterial prey for protozoan predators. Neocercomonas jutlandica was originally isolated from soil sampled at Research Center Foulum (Jutland, Denmark), where it was the most abundant flagellate species occurring in microtiter plates used for enumeration of soil protozoa [19]. Since then it has been kept in the culture collection at Section for Terrestrial Ecology, and has been used in several experiments. Its growth kinetics was determined in [20], and its susceptibility to the fungicide propiconazole was assessed in [21] where it was named Cercomonas crassicauda. Ekelund F et al. [22] sequenced it and analysed it phylogenetically, and hence gave it the name Neocercomonas jutlandica. Finally, Pedersen AL et al. [23, 24] showed that N. jutlandica was among the more tolerant protozoa when exposed to potentially toxic Pseudomonas spp. We notice that N. jutlandica is synonymous with Cercomonas jutlandica [25]. The derived protozoan cultures growing on specific strains were prepared as follows. The pure bacterial 48 h-cultures and the equalvolume mixed culture of these four strains were diluted 20-fold in weak phosphate buffer (Neff ’s modified amoeba saline) [26], then the stepwise dilution technique [27, 28] was performed to provide protozoan cultures on each strain separately and on the mixed bacteria. In short, 100 μL protozoan cultures were repeatedly transferred to 10 mL bacterial cultures in cell culture flasks (Nunc A/S, Roskilde, Denmark, # 156367, 25 cm3) produced as described above and incubated in darkness, at 9 °C, until late exponential phase. More reproducible results could be obtained from this approach than using a fixed cell number as food source for protozoa when making a standard comparison between cultures according to Pedersen AL et al. [24]. Flagellate growth on bacterial cultures We determined growth rates and peak abundance of N. jutlandica on the four bacteria separately and as a mixture (equal-volume mixed) in 96-well microtiter plates (Cat. No. 655 180, Greiner Bio-One, Germany). For each of the four bacteria and a mixture, 125 μL of the 20-fold bacterial dilution and 25 μL protozoan cultures (prepared as mentioned above) were mixed and inoculated into each well. Each particular combination of bacteria and protozoa was set up in four replicates. The wells containing only 150 μL bacteria were blank controls. The batch cultures were then left at 9 °C in darkness. We used an inverted microscope (Olympus CK30, 200 × magnification, phase contrast) to quantify the protozoa. For each of the five bacterial treatments, we counted the protozoa in the four replicate wells every day from day 0 to day 8, and on days 10, 12, 14, 16, 19 and 21, where the cultures had approached the stationary phase. At each counting, protozoa spread five microscopic fields were counted in each well. The protozoan numbers in each well were calculated and plotted against times (day) in SigmaPlot 12.5. The nonlinear regression analyses were performed to fit these points to a Sigmoidal, 3 Parameter equation f = a/(1+exp((x-x0)/b)), where a, b and x0 are constants that are determined by the nonlinear regression. The intrinsic growth rate (r) defined as slope of the sigmoid curve in the point of inflexion (r = a/4b), and peak abundance (p) defined as the limit of f when t →∞ (p = a) were calculated as previous reported [29]. Quantitative PCR detection of the effects of flagellate on both mono- and multi-species biofilms The biofilm formation assay was conducted in 96-well microtiter plates (Cat. No. 655 180, Greiner Bio-One, Germany) as described previously [30]. As displayed in Figure 1, two (a) 24 h X. retroflexus culture X. retroflexus biofilms experiments were conducted for assessing the effects of the flagellate N. jutlandica on biofilm development. Since X. retroflexus was categorized as the only good biofilm former among these four soil isolates in a previous study [30], flagellate grazing on the singlespecies biofilm formed by this strain was performed to compare the abilities of flagellate N. jutlandica to grazing monospecies and multispecies biofilms. after day-1, day 1, 2 and 3 DNA extraction from biofilms X. retroflexus + derived protozoa qPCR to determine cell numbers of X. retroflexus (b) X. retroflexus culture 24 h Four-speices biofilms after day 1, 2 and 3 DNA extraction from biofilms qPCR to determine cell numbers of X. retroflexus after afterday-3 day 3 DNA extraction from biofilms qPCR to determine cell numbers of four strains X. retroflexus + derived protozoa four-speices bacterial culture bacterial culture + derived protozoa Figure 1 Experimental setup for assessing the grazing effects of flagellate N. jutlandica on both singlespecies (a) and multispecies biofilms (b). In short, 24-h biofilms in the wells were introduced with bacteriaprotozoan cultures or only bacterial cultures, followed by removal of planktonic cells and DNA extraction from biofilms after 1, 2 and 3 days/3 days incubation. Thereafter, SYBR Green qPCRs were performed to measure the cell numbers of only strain X. retroflexus or all the four strains. In short, overnight bacterial cultures were subcultured, grown to exponential phase in TSB and then adjusted to give an OD600 of 0.15. Aliquots of 150 μL of diluted X. retroflexus culture were then added to each well to form single-species biofilms. For multispecies biofilm study, equal volumes of the four isolate cultures were thoroughly mixed and 150 μL was added to each well. After 24 h incubation, the biofilms were rinsed three times with weak phosphate buffer to remove planktonic cells and hereafter, the bacteriaflagellate cultures prepared as in flagellate growth experiment, that is 125 μL of 20-fold diluted bacterial cultures and 25 μL protozoan cultures in the flasks, was added to each well. To the wells with pre-established single-species biofilms, only single-species bacterial culture and derived protozoan culture were added. To the wells containing multispecies biofilms, single-species or four-species mixed bacterial cultures and respectively derived protozoan cultures were added. In parallel, 150 μL of single-species or mixed-species bacterial cultures without protozoa was also inoculated into some wells as controls. For all the experiments investigating the effects of protozoa feeding on biofilms, the 5-day-old protozoan cultures in the flasks at 9 °C were used to ensure that N. jutlandica was in the exponential growth phase when inoculated in the wells. Three replicate wells were prepared for each treatment. The plates were then sealed with Parafilm and incubated with shaking (100 rpm) at 24°C for 3 days. The planktonic cells were removed after 1, 2 and 3 days/3 days, followed by biofilms were scratched with 200 μL plastic pipette tips in 150 μL PBS. The wells were then stained with crystal violet to verify that all the biofilm cells were detached. Multispecies biofilm samples before exposed to protozoan grazing were also collected using the same procedure. Bacterial DNA was extracted using FastDNA™ SPIN Kit for soil (Qbiogene, Illkirch, France) according to the method described in [30]. Three replicates of DNA samples extracted from biofilms incubated with/without protozoa were quantified by SYBR Green qPCR using standard curves generated by serial 10-fold dilutions of plasmid DNAs [30]. The cell numbers of only strain X. retroflexus or all the four strains were measured using species specific primers and thermal profile setup reported in [30]. All samples were run in duplicate and a no template control was included in every run. Statistical analysis One way ANOVA test (SPSS version 17.0 for Windows) was conducted to evaluate the effects of the flagellate predation on population dynamics in both single-species and multispecies biofilms. P values < 0.05 were reported as statistically significant. Results N. jutlandica growth curves The growth of N. jutlandica followed a similar pattern on all the bacterial cultures, with an exponential phase followed by a stationary phase (Figure 2), which demonstrates that all four strains are suitable as food for this protozoan. Non-linear regression was used to describe the growth pattern instead of simple linear regression that is only applicable to exponential growth phase. The data points S. rhizophila 4.5 M. oxydans 4.5 Protozoan numbers (log10 ) 4.0 3.5 3.0 2.5 2.0 0 5 10 15 20 25 15 20 25 15 20 25 Time (day) P. amylolyticus 4.5 4.0 Protozoan numbers (log10 ) fitted very well to the sigmoid model, with Rsqr values in the range of 0.9464 to 0.9786. Except for P. amylolyticus, N. jutlandica growing on the other bacterial cultures all could reach stationary phase after 10 days. The flagellate growth rates and peak abundances were calculated and shown in Table 1. When feeding on the four pure cultures, N. jutlandica in coculture with S. rhizophila yielded the highest growth rates and cell numbers, indicating the feeding preference of N. jutlandica for S. rhizophila over the other three species. However, when preying on the mixed bacterial culture, N. jutlandica showed more rapid growth even though the peak abundance was compromised. 3.5 3.0 2.5 Protozoan numbers (log10 ) 4.0 2.0 3.5 0 10 Time (day) 3.0 Mixed strains 2.5 4.5 2.0 5 10 15 20 4.0 25 Protozoan numbers (log10 ) 0 Time (day) X. retroflexus 4.5 3.5 3.0 2.5 4.0 Protozoan numbers (log10 ) 5 2.0 3.5 0 5 10 Time (day) 3.0 2.5 2.0 0 5 10 Time (day) 15 20 25 Figure 2 N. jutlandica growth curves for the four replicate batch cultures. The points indicate the actual numbers of flagellate plotted against time of day. The solid line represents the simple sigmoid curve when data were fitted to the equation f = a/(1+exp(-(x-x0)/b)). Table 1 Growth rates (r) and peak abundances (p) of N. jutlandica consuming different planktonic bacteria. Growth rate (day)a Peak abundance (cell number)b a b S. rhizophila X. retroflexus M. oxydans P. amylolyticus Mixed strains 0.96 1.17 0.97 1.70 0.89 1.74E+04 1.15E+04 1.16E+04 8.69E+03 1.41E+04 Growth rate (r) is presented as the days it takes the protozoan to double in size (cell number). Peak abundance (p) is presented as the cell numbers of protozoan when t →∞, that is p = a. Impact of N. jutlandica grazing on biofilms under static condition R-Square (RSq) values, based on threshold cycles of the standard curves, ranged from 0.988 to 0.998 and application efficiencies (E) ranged from 83.5-100.0%. The log 10 of bacterial cell numbers were presented in Figure 3 and used in significance analysis. As can be seen in Figure 3 (a and b), the results obtained by qPCR showed that the presence of flagellate N. jutlandica, which was co-cultured with strain X. retroflexus, could increase the cell numbers of this strain in both single- and multi-species biofilms. The significant enhancement was observed after 1 (P = 0.003) and 2 (P = 0.008) days of incubation with N. jutlandica in single-species biofilms, while this difference gradually narrowed until day 3 (P = 0.787). In multispecies biofilms, this increase in cell numbers of strain X. retroflexus was not significantly different during the three days between grazing and non-grazing biofilms (P > 0.6). However, this bacterial growth-promoting effect of N. jutlandica had changed (Fig. 3c) when this flagellate was co-cultured with fourspecies bacterial culture and then introduced into multispecies biofilms, as evidenced by the reductions in cell numbers of all the four strains after 3 days’ predation. Moreover, these reductions were significant for S. rhizophila (P = 0.017), M. oxydans (P = 0.03) and P. amylolyticus (P = 0.036), while a marginally significant difference was revealed for X. retroflexus (P = 0.06). These contrary observations suggest that the grazing impact of the flagellate N. jutlandica also depends on their bacterial prey that co-cultured with before inoculated into pre-established biofilms. It is worth noting that the ratios of cell numbers of the three strains X. retroflexus, M. oxydans and P. amylolyticus remained almost constant (180: 24: 1) regardless of grazing or non-grazing treatments, indicating the possible existence of these three species as mixed microcolonies within the multispecies biofilm. A greater reduction was seen in the abundance of S. rhizophila cells under predation pressure by N. jutlandica, which is consistent with the prey preference of this flagellate over strain S. rhizophila derived from planktonic cultures. As shown in Figure 3c, even without protozoan grazing, the majority of bacterial cells still dispersed from this multispecies biofilm after 3 days (P < 0.007) probably due to the lownutrient availability. We notice unfortunately, contaminant, Pseudomonas rRNA gene. that the added protozoa apparently contained a which we determined to be sp. from its sequence of 16S Such contaminations are very difficult to avoid when working with protozoa. It never made up more than 20% of the bacterial numbers; thus we do not consider it to (a) 8 have had significant impact on the outcome of the experiments. (b) 10 With protozoan 6 4 2 Cell numbers of X. retroflexus (log10) Cell number of X. retroflexus (log10) Without protozoan Without protozoan With protozoan 8 6 4 2 0 0 day 1 day 2 day 1 day 3 Time day 2 day 3 Time (c) 10 0 day w ithout protozoan 3 day w ithout protozoan 3 day w ith protozoan Cell numbers (log 10) 8 6 4 2 0 S. rhizophila X. retroflexus M. oxydans P. amylolyticus Strains Figure 3 Effects of grazing by N. jutlandica upon the population dynamics of both single-species (a) and multispecies biofilms (b and c). (a) X. retroflexus culture with/without protozoan feeding on X. retroflexus was introduced into wells with pre-established X. retroflexus biofilm. Absolute cell numbers from biofilms were measured using qPCR after 1, 2 and 3 days of treatment. (b) X. retroflexus culture with/without protozoan feeding on X. retroflexus was introduced into wells with pre-established four-species biofilm and X. retroflexus cell numbers from biofilms were measured using qPCR after 1, 2 and 3 days of treatment. (c) Four-species culture with/without protozoan feeding on mixed strains was introduced into wells with preestablished four-species biofilm. Absolute cell numbers of each species were measured using qPCR only after 3 days of treatment. Bars represent means ± standard deviation for three replicates. Discussion Bacterial biofilms and protozoa are prevalent in natural environments. Protozoa shape the microbial communities, influence nutrient availability and act as reservisors and vectors of pathogenic bacteria [7, 31, 32]. Although, the underlying mechanisms have been partly elucidated from limited researches focusing on single-species models [6, 33, 34], there is still a deficiency of studies regarding the shaping impact of protozoa on mixed-species biofilms, which represent the predominate lifestyle in most ecosystems. In a previous study, a strong synergy was observed in a biofilm composed of four soil isolates, which could contribute to three-fold more biofilm biomass than the single-species biofilms. Moreover, by employing quantitative PCR, interspecific cooperation and the dominance of X. retroflexus in this microbial community had been further demonstrated [30], indicating the usefulness of qPCR assay in monitoring the population dynamics. Here we combined this powerful tool with the developed multispecies biofilm model to evaluate the role of synergistic interactions played in the impact of protozoan grazing on the mixed-species biofilms. Bacterial features, including cell size, surface properties, speed of movement and ability of biofilm formation [14, 35-37] could affect their vulnerability towards grazers. Additionally, some gram-positive bacteria showed lower edibility than gram-negative bacteria due to their thick and rigid cell walls [38].The larger size (0.7 to 0.9 by 3.0 to 5.0 μm) compared with other three species, motility by flagella and gram status of strain P. amylolyticus [39] may presumably explain the lower growth rate and abundance of N. jutlandica when feeding on this strain grown in planktonic culture. Although high biofilm-forming capacities were demonstrated in X. retroflexus and in co-culture of the four species previously [30], these were not observed under the oligotrophic condition in the present study (data not shown), which could be largely responsible for their vulnerability towards N. jutlandica. The highest growth rate of flagellate was found in mixed culture of these four species which may result from the enhanced prey density due to the synergistic growth effect among these bacteria. This synergistic interaction was observed in biofilm formation with higher nutrient concentration previously [30], thus, it requires further validation as bacterial cultures were diluted 20-fold in non-nutritious buffer (Neff ’s modified amoeba saline) in the present study. In order to maintain the vitality of N. jutlandica, diluted bacterial cultures were inoculated together with protozoan into the pre-established biofilms. Even so, the added bacteria were unlikely to result in the additional biofilm formation, as confirmed by their inability to produce any single- or multi-species biofilm in the low-nutrient medium with or without the presence of N. jutlandica (data not shown). Hence, the changes in biofilm biomass after addition of protozoa compared with addition of diluted bacteria cultures only should result solely from the influence of protozoan grazing on biofilms. A recent study has shown that while early-stage biofilms are mainly regulated by random attachment, mature biofilms seem to be largely regulated by interactions within the bacterial communities and with the abiotic environment, whereas grazing by heterotrophic flagellates may only has a detectable but limited effect [17]. However, in this study, we demonstrate that biofilm dynamics could be greatly affected by both the abiotic condition (e.g. nutrient starvation) and flagellate grazing. Previous studies from Wey JK et al. [17] have shown that within semi-natural river biofilms, while some bacterial species were especially susceptible to grazing by HF, others could benefit from the resulting decreased competition. This was not observed in the present study, as the cell numbers of all species in the presence of grazer decreased significantly (Figure 3c). Interestingly, although the cell number of strain S. rhizophila showed reduction to a greater extent, an approximate 3-fold reduction in cell numbers was shown in all other three species, indicating that these three species probably exist as integrated communities rather than form individual microcolonies. The evidences that grazing pressure is positively correlated with the formation of cell clusters have been supported by the results obtained both from monospecies laboratory biofilms [6, 40] and from natural/semi-natural multispecies biofilms [32, 41]. This could result from an active defense mechanism [42] or a passive process that the movement of HF and their flagella drive the bacterial cells to the substratum [17]. On the contrary, the conclusion that total protection against predation could not be fully met by growing as a biofilm was also widely reported, as exemplified by enhanced sloughing [43] and/or markedly altered population dynamics [44] in the presence of grazers. In this study, we aimed to test the hypothesis that multispecies biofilms should be more grazing-resistant compared with monospecies biofilms owing to the synergisitc interactions between community members. However, conflicting results were found. The increased bacterial abundance in the presence of protozoa co-cultured with X. retroflexus was reflected in both mono- and multi-species biofilms. However, when used the same flagellate but co-cultured with the mixed bacteria, the significant reductions of cell numbers were observed for all the four bacteria. This can be explained by the higher growth rate of N. jutlandica when feeding on mixed bacteria compared with feeding on a pure culture of X. retroflexus. These different prey preferences suggest that this flagellate probably has evolved better adaption to the mixed bacteria which results in higher density and vitality and thus higher ‘offensiveness’. When co-cultured only with strain X. retroflexus, this flagellate might become less active grazer or the synergistic interactions between biofilm members could enhance the resistance of this multispecies biofilms against protozoan grazing. However, the former explanation seems more reasonable as in single-species biofilms absent of interspecies interactions, the decrease in cell numbers of X. retroflexus was not observed either. It is very important to note that we only have been looking at the surface attached bacterial cells. The conclusions obtained in this study are based on the assumption that the chosen flagellate have a preferencial grazing on surface attached cells. However if this assumption is wrong then we may have a situation where majority of cells in planktonic phase will be grazed and only surface attached cells will be protected. In this situation we expect that the monospecies cultures of the three bacteria unable to produce biofilm by themselves will be heavily reduced by grazing and that they actually benefit greatly by being together in the four-species consortium where biofilms formation will provide them protections against protozoan grazing. In order to test this we need to conduct further experiments where we also count bacterial cells in planktonic phase both in mono cultures +/- protozoan grazing and in four-species co-cultures. Thus, from the present results, we can not support or reject this hypothesis, and further experiments are needed before it can be concluded which of the above listed scenarios is supported. Besides, it is not within the scope of this study to predict whether the flagellate-prey interactions obtained here, namely, the effects of flagellate N. jutlandica on biofilms, could be extrapolated to other protozoa. In future studies, incorporation of other protozoa would certainly be interesting; moreover, experimental approach and axenizing procedure need optimization to make this setup a valuable tool for research on protozoa-biofilm interactions. References 1. Donlan RM, Costerton JW (2002) Biofilms: survival mechanisms of clinically relevant microorganisms. Clinical microbiology reviews 15: 167-193. 2. Ekelund F, Rønn R (1994) Notes on protozoa in agricultural soil with emphasis on heterotrophic flagellates and naked amoebae and their ecology. FEMS Microbiology Reviews 15: 321-353. 3. Ekelund F, Rønn R, Griffiths BS (2001) Quantitative estimation of flagellate community structure and diversity in soil samples. Protist 152: 301-314. 4. Arndt H, Schmidt-Denter K, Auer B, Weitere M (2003) Protozoans and biofilms. Fossil and recent biofilms 1: 161-179. 5. Böhme A, Risse‐Buhl U, Küsel K (2009) Protists with different feeding modes change biofilm morphology. FEMS microbiology ecology 69: 158-169. 6. Matz C, Bergfeld T, Rice SA, Kjelleberg S (2004) Microcolonies, quorum sensing and cytotoxicity determine the survival of Pseudomonas aeruginosa biofilms exposed to protozoan grazing. Environmental microbiology 6: 218-226. 7. Joubert L-M, Wolfaardt G, Botha A (2006) Microbial exopolymers link predator and prey in a model yeast biofilm system. Microbial ecology 52: 187-197. 8. Queck SY, Weitere M, Moreno AM, Rice SA, Kjelleberg S (2006) The role of quorum sensing mediated developmental traits in the resistance of Serratia marcescens biofilms against protozoan grazing. Environmental microbiology 8: 1017-1025. 9. Caron DA (1987) Grazing of attached bacteria by heterotrophic microflagellates. Microbial Ecology 13: 203-218. 10. Sibbald MJ, Albright LJ (1988) Aggregated and free bacteria as food sources for heterotrophic microflagellates. Applied and environmental microbiology 54: 613-616. 11. Taylor M, Ross K, Bentham R (2009) Legionella, protozoa, and biofilms: interactions within complex microbial systems. Microbial ecology 58: 538-547. 12. Rowe M, Grant I (2006) Mycobacterium avium ssp. paratuberculosis and its potential survival tactics. Letters in applied microbiology 42: 305311. 13. Jürgens K, Matz C (2002) Predation as a shaping force for the phenotypic and genotypic composition of planktonic bacteria. Antonie van Leeuwenhoek 81: 413-434. 14. Matz C, Jürgens K (2005) High motility reduces grazing mortality of planktonic bacteria. Applied and environmental microbiology 71: 921-929. 15. Roberts EC, Legrand C, Steinke M, Wootton EC (2011) Mechanisms underlying chemical interactions between predatory planktonic protists and their prey. Journal of plankton research 33: 833-841. 16. Matz C (2011) Competition, Communication, Cooperation: Molecular Crosstalk in Multispecies Biofilms. Biofilm Highlights. Springer, pp. 29-40. 17. Wey JK, Jürgens K, Weitere M (2012) Seasonal and successional influences on bacterial community composition exceed that of protozoan grazing in river biofilms. Applied and environmental microbiology 78: 2013-2024. 18. de la Cruz‐Perera CI, Ren D, Blanchet M, Dendooven L, Marsch R, Sørensen SJ, Burmølle M (2013) The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains. FEMS microbiology letters 338: 95-100. 19. Rønn R, Ekelund F, Christensen S (1995) Optimizing soil extract and broth media for MPN enumeration of naked amoebae and heterotrophic flagellates in soil. Pedobiologia 39: 10-19. 20. Ekelund F (1996) Growth kinetics of five common heterotrophic soil flagellates. European journal of soil biology 32: 15-24. 21. Ekelund F, Westergaard K, Søe D (2000) The toxicity of the fungicide propiconazole to soil flagellates. Biology and fertility of soils 31: 7077. 22. Ekelund F, Daugbjerg N, Fredslund L (2004) Phylogeny of Heteromita, Cercomonas and Thaumatomonas based on SSU rDNA sequences, including the description of Neocercomonas jutlandica sp. nov., gen. nov. European journal of protistology 40: 119-135. 23. Pedersen AL, Ekelund F, Johansen A, Winding A (2010) Interaction of bacteria-feeding soil flagellates and Pseudomonas spp. Biology and Fertility of Soils 46: 151-158. 24. Pedersen AL, Winding A, Altenburger A, Ekelund F (2011) Protozoan growth rates on secondary‐metabolite‐producing Pseudomonas spp. correlate with high‐level protozoan taxonomy. FEMS Microbiology Letters 316: 16-22. 25. Karpov SA, Bass D, Mylnikov AP, CavalierSmith T (2006) Molecular Phylogeny of Cercomonadidae and Kinetid Patterns of Cercomonas and Eocercomonas gen. nov. (Cercomonadida, Cercozoa). Protist 157: 125158. 26. Page FC (1988) A New Key to Freshwater and Soil Gymnamoebae: With Instructions for Culture pp. 91-97. Freshwater Biological Association Cumbria 27. Pelegri S, Christaki U, Dolan J, Rassoulzadegan F (1999) Particulate and dissolved organic carbon production by the heterotrophic nanoflagellate Pteridomonas danica Patterson and Fenchel. Microbial ecology 37: 276-284. 28. Mohapatra B, Fukami K (2004) Comparison of the numerical grazing response of two marine heterotrophic nanoflagellates fed with different bacteria. Journal of Sea Research 52: 99-107. 29. Koch TA, Ekelund F (2005) Strains of the Heterotrophic Flagellate Bodo designis from Different Environments Vary Considerably with Respect to Salinity Preference and SSU rRNA Gene Composition. Protist 156: 97-112. 30. Ren D, Madsen JS, de la Cruz-Perera CI, Bergmark L, Sørensen SJ, Burmølle M (2013) High-Throughput Screening of Multispecies Biofilm Formation and Quantitative PCR-Based Assessment of Individual Species Proportions, Useful for Exploring Interspecific Bacterial Interactions. Microbial ecology. doi: 10.1007/s00248-013-0315-z 31. Brown MR, Barker J (1999) Unexplored reservoirs of pathogenic bacteria: protozoa and biofilms. Trends in microbiology 7: 46-50. 32. Wey JK, Scherwass A, Norf H, Arndt H, Weitere M (2008) Effects of protozoan grazing within river biofilms under semi-natural conditions. Aquatic Microbial Ecology 52: 283. 33. Dopheide A, Lear G, Stott R, Lewis G (2011) Preferential feeding by the ciliates Chilodonella and Tetrahymena spp. and effects of these protozoa on bacterial biofilm structure and composition. Applied and environmental microbiology 77: 4564-4572. 40. Matz C, McDougald D, Moreno AM, Yung PY, Yildiz FH, Kjelleberg S (2005) Biofilm formation and phenotypic variation enhance predation-driven persistence of Vibrio cholerae. Proceedings of the National Academy of Sciences of the United States of America 102: 16819-16824. 41. Rychert K, Neu TR (2010) Protozoan impact on bacterial biofilm formation. Biological Letters 47: 3-10. 34. Friman V-P, Diggle SP, Buckling A (2013) Protist predation can favour cooperation within bacterial species. Biology letters 9: 20130548. 42. Matz C, Kjelleberg S (2005) Off the hook–how bacteria survive protozoan grazing. Trends in microbiology 13: 302-307. 35. Posch T, Jezbera J, Vrba J, Šimek K, Pernthaler J, Andreatta S, Sonntag B (2001) Size selective feeding in Cyclidium glaucoma (Ciliophora, Scuticociliatida) and its effects on bacterial community structure: a study from a continuous cultivation system. Microbial ecology 42: 217227. 43. Jackson S, Jones E (1991) Interactions within biofilms: the disruption of biofilm structure by protozoa. Kieler Meeresforsch 8: 264-268. 36. Monger BC, Landry MR, Brown SL (1999) Feeding selection of heterotrophic marine nanoflagellates based on the surface hydrophobicity of their picoplankton prey. Limnology and oceanography 44: 1917-1927. 37. Hahn M, Moore E, Höfle M (2000) Role of microcolony formation in the protistan grazing defense of the aquatic bacterium Pseudomonas sp. MWH1. Microbial ecology 39: 175-185. 38. Rønn R, McCaig AE, Griffiths BS, Prosser JI (2002) Impact of protozoan grazing on bacterial community structure in soil microcosms. Applied and Environmental Microbiology 68: 6094-6105. 39. Shida O, Takagi H, Kadowaki K, Nakamura LK, Komagata K (1997) Emended Description of Paenibacillus amylolyticus and Description of Paenibacillus illinoisensis sp. nov. and Paenibacillus chibensis sp. nov. International journal of systematic bacteriology 47: 299-306. 44. Huws S, McBain A, Gilbert P (2005) Protozoan grazing and its impact upon population dynamics in biofilm communities. Journal of applied microbiology 98: 238-244. Manuscript 5 The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains RESEARCH LETTER The ability of soil bacteria to receive the conjugative IncP1 plasmid, pKJK10, is different in a mixed community compared to single strains Claudia I. de la Cruz-Perera1,2, Dawei Ren1, Marine Blanchet1,3, Luc Dendooven2, Rodolfo Marsch2, Søren J. Sørensen1 & Mette Burmølle1 1 Department of Biology, University of Copenhagen, Copenhagen, Denmark; 2Laboratory of Soil Ecology, Department of Biotechnology and Bioengineering, Cinvestav, Mexico City, Mexico; and 3Laboratoire d’Oc eanographie Microbienne, Observatoire Oc eanologique de Banyuls, Banyuls-sur-mer, France Correspondence: Mette Burmølle, Section of Microbiology, University of Copenhagen, Universitetsparken 15, bygn 1, 2100 Copenhagen Ø, Denmark. Tel.: +45 40220069; e-mail: [email protected] Received 18 October 2012; accepted 25 October 2012. Final version published online 22 November 2012. DOI: 10.1111/1574-6968.12036 MICROBIOLOGY LETTERS Editor: Paolina Garbeva Keywords horizontal gene transfer; bacterial community; conjugation; host range; soil bacteria. Abstract Horizontal gene transfer by conjugation is common among bacterial populations in soil. It is well known that the host range of plasmids depends on several factors, including the identity of the plasmid host cell. In the present study, however, we demonstrate that the composition of the recipient community is also determining for the dissemination of a conjugative plasmid. We isolated 15 different bacterial strains from soil and assessed the conjugation frequencies of the IncP1 plasmid, pKJK10, by flow cytometry, from two different donors, Escherichia coli and Pseudomonas putida, to either 15 different bacterial strains or to the mixed community composed of all the 15 strains. We detected transfer of pKJK10 from P. putida to Stenotrophomonas rhizophila in a diparental mating, but no transfer was observed to the mixed community. In contrast, for E. coli, transfer was observed only to the mixed community, where Ochrobactrum rhizosphaerae was identified as the dominating plasmid recipient. Our results indicate that the presence of a bacterial community impacts the plasmid permissiveness by affecting the ability of strains to receive the conjugative plasmid. Introduction Horizontal gene transfer (HTG) is a driving force in bacterial evolution as it allows bacteria to rapidly acquire complex new traits. Plasmids are one of the key vectors of HTG, enabling genetic exchange between bacterial cells across species and domain barriers (Poole, 2009; Boto, 2010), and they very often encode genes that confer adaptive traits to their host, such as antibiotic resistance, biodegradation pathways and virulence (de la Cruz & Davies, 2000). Transfer of these traits by conjugation requires the donor and the recipient cells to be in direct contact. Different abiotic and biotic factors affect the range of conjugal exchange of genetic material between environmental bacteria, such as nutrient availability, spatial architecture of the bacterial community, plasmid donor and recipient relatedness and plasmid host type (van Elsas & Bailey, 2002; De Gelder et al., 2005; Sørensen et al., 2005; FEMS Microbiol Lett 338 (2013) 95–100 Seoane et al., 2011). The fraction of the cells in a community capable of receiving and maintaining conjugative plasmids is highly dependent on several of these factors and has been described as the plasmid permissiveness (Musovic, 2010). It has been shown that conjugative plasmids express factors that favor the establishment of planktonic bacteria in biofilm communities, thereby increasing the chances for horizontal gene transmission (Ghigo, 2001; Reisner et al., 2006; Madsen et al., 2012). Complex interspecies communities facilitate synergistic interactions between populations, affecting the function, stability and flexibility of the community (James et al., 1995; Burmølle et al., 2006). In the present work, HTG by conjugation between single populations and microbial communities isolated from soil were investigated. The plasmid transfer frequencies and the identities of the recipients of the plasmid, when ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 96 hosted by different donors, were compared. The bacterial population was analyzed based on fluorescence properties and sorted by flow cytometry (FCM) to detect and quantify the plasmid transfer to the individual isolates and the mixed community (Muller & Nebe-von-Caron, 2010). Sequencing of the 16S rRNA gene from sorted transconjugant cells was used to evaluate the host range of the plasmid when a mixed microbial community was used as recipient. Materials and methods Soil and leaf sampling Soil samples were collected from an agricultural field in T astrup, Denmark, in the late summer of 2009. Soil was sampled from the 5- to 10-cm layer. The soil water content upon sampling was 14.2%, and the water holding capacity (WHC) was 60%. The soil was classified as sandy loam with pH 7.2. Leaves of baby maize seedlings were used for bacterial isolation. The seedlings were grown for 2 weeks in T astrup soil before harvesting. Bacterial strains, plasmids, and growth media Escherichia coli CSH26::lacIq and Pseudomonas putida KT2440::lacIq1, carrying pKJK10, a conjugative, green fluorescent protein (GFP) tagged IncP1 plasmid, originally isolated from soil (Sengeløv et al., 2001; Bahl et al., 2007) were used as donor strains. These strains were cultured in Luria Bertani (LB) broth supplemented with kanamycin monosulfate (50 mg mL 1); 1.5% (w/v) agar was added when solid medium was needed. The recipient strains (see below) were cultured in Tryptic Soy Broth medium (TSB; 17 g peptone from casein, 3 g peptone from soymeal, 2.5 g D(+)-Glucose, 5 g NaCl, 2.5 g K2HPO4 in 1 L distilled water, pH 7.3). Isolation and identification of recipient strains from leaves incubated in soil A 15 mg sample of a baby maize leaf was placed in 5 g T astrup soil adjusted to 40% WHC and incubated in triplicate at room temperature for 17 days. After 7, 12, and 17 days, the leaves were picked up from the soil, washed with PBS (8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, and KH2PO4, adjusted to 1 L distilled water and pH 7.4), placed in a microfuge tube, added 1 mL PBS and vortexed for 1 min. DNA was extracted from the cell suspension as described below. Dilutions to 10 6 were made and 100 lL were plated in triplicate onto Tryptic Soy Agar (TSA; Difco) 10% supplemented with cycloheximide (50 mg mL 1) and incubated at 25 °C for 2–5 days. Sixteen colonies from ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved C.I. de la Cruz-Perera et al. each triplicate looking phenotypically different were isolated and purified for DNA extraction. A denaturation gradient gel electrophoresis (DGGE) analysis was performed with 16S rRNA gene PCR fragments. From the total of 48 strains from day 7, 15 morphologically different strains were selected for the use as recipients. The strains were grown overnight (ON) in 5 mL TSB, the DNA was extracted using ‘Genomic Mini for Universal Genomic DNA Isolation Kit’ (A&A Biotechnology) and the 16S rRNA gene sequences were amplified with primers 27F and 1492R (Lane, 1991) for identification. The PCR mixture contained 0.5 lL DNA, 1XPhusion GC buffer, 0.2 mM dNTP mixture, 1 U Phusion Hot Start DNA Polymerase (FinnzymesOy, Espoo, Finland) and 0.5 lM of each primer (TAG Copenhagen A/S, Denmark). The final volume was adjusted with DNA-free water to 50 lL. Amplification was as follow: initial denaturation at 98 °C for 30 s, followed by 35 cycles at 98 °C for 10 s, at 55 °C for 30 s and at 72 °C for 45 s. A final primer extension reaction was performed at 72 °C for 6 min. The resulting sequence (1480 bp) was compared with reference sequences by BLAST search (Altschul et al., 1997) and aligned with them using CLUSTALX 1.7 program (Thompson et al., 1997). Maximum-likelihood analyses were performed using PhyML (Guindon & Gascuel, 2003). MODELTEST 3.06 (Posada, 2008) was used to select appropriate models of sequence evolution by the Akaike Information Criterion. The confidence at each node was assessed by 500 bootstrap replicates. Similarities among sequences were calculated using the MatGAT v.2.01 software (Campanella et al., 2003). Taxonomic assignment was carried out based on the Rosell o-Mora and Aman criteria (Rossell o-Mora & Amann, 2001). DGGE The cells from the leaves-PBS solution and from the 48- to 15-strain pools were lysed by bead beating followed by DNA extraction as specified above. The DNA was used for a 16S rRNA gene PCR as described above and 1 lL of the product was used as a template for a new PCR using internal primers with a GC clamp 341F and 518R (Muyzer et al., 1993) and a polymerization step at 72 °C for 20 s. This PCR product was loaded onto the DGGE gel, containing a denaturation gradient of 30–70% acrylamide, and an electrophoresis was run in a Dcode system (Biorad) at 60 °C and 70 V for 17 h. The gel was stained with SYBRGold (Invitrogene) in the dark for 45 min. Filter mating assays Prior to filter matings, the donor strains were grown in 5 mL LB broth at 250 r.p.m. at 30 °C (P. putida) and FEMS Microbiol Lett 338 (2013) 95–100 97 Gene transfer in mixed and single species communities 37 °C (E. coli) for 18 h. These ON cell cultures were then diluted 1 : 10 in fresh LB medium and grown under similar conditions for three more hours to reach exponential growth phase (OD600 0.6). The cells were then recollected, washed twice, and resuspended in sterile PBS. The recipient strains were cultured similarly in TSB at 25 °C. The lack of background fluorescence of the donor and recipient strains was verified in the flow cytometer (see specifications below) prior to their use in the filter mating assay. For the single-strain mating experiments, 10 lL of donor and recipient, respectively, were spotted onto 0.2 lm nitrocellulose filters in triplicate, mixed, placed on TSA and R2A (yeast extract 0.5 g, proteose peptone 0.5 g, casamino acids 0.5 g, glucose 0.5 g, soluble starch 0.5 g, sodium pyruvate 0.3 g, K2HPO4 0.3 g, MgSO47H2O 0.05 g, agar 15 g in 1 L distilled water) plates and incubated at 25 °C for 20 h. The cells were then harvested from the filter followed by resuspension in 1 mL PBS, and FCM analysis as specified below. For the microbial community, we spotted 5 lL of each isolate (OD600 0.3–0.7) and 75 lL of donor strain (either P. putida or E. coli, prepared as described above) onto the filter, incubated and analyzed by FCM at the same conditions as for the single-strain matings. Controls with only donors or recipients were included. Quantification and sorting by FCM Flow cytometric enumeration of cells was carried out with a FACScalibur flow cytometer (Becton Dickinson, San Jose, CA) equipped with a 15 mW argon laser (488 nm). The following settings and voltages were used during analysis: forward scatter = E01, side scatter (SSC) = 350, and the fluorescent detectors FL1 (530/ 30 nm), FL2 (585/42 nm), FL3 (650/30 nm) were set at 510 V. A threshold was set on the SSC, and no compensation was used. All parameters were on logarithmic mode. Samples were run at the ‘low’ flow rate setting for 1 min. All the samples were diluted in PBS before flow enumeration to assure optimal bacterial counts to 2000 events s 1. In part of the sample (100 lL), gfpexpression was induced by incubation in LB with 1 mM of isopropyl-b-D-1-thiogalactopyranoside (IPTG, SIGMA) for 3 h at 30 °C (P. putida) and 37 °C (E. coli) to determine the number of donor cells (Musovic et al., 2006). To isolate and identify recipients from the E. coli-community mating, one subsample of each replicate of the cell extract was diluted to 1000 events s 1 to flow-sorted (MoFlo; DAKO) at a flow rate of 400–1000 events s 1, with an optimal setting of the sheath pressure of ca. 60 psi and drop drive frequency to ca. 95 kHz, using a FEMS Microbiol Lett 338 (2013) 95–100 70-lm CytoNozzle tip on an enrichment sort option of single-mode per single drop envelope. Dilutions up to 10 3 were made from approximately 70 000 cells of each replicate, and 100 lL of each dilution were plated on TSA plates supplemented with kanamycin, streptomycin (100 mg mL 1) and tetracycline (20 mg mL 1) and incubated at 25 °C for 2–5 days. Four green colonies of each replicate were selected for DNA extraction and identified by sequencing after the amplification of the 16S rRNA gene as described above. Data analysis was carried out with the CELLQUEST software package. Two polygonal gates were defined in bivariate FL1 vs. FL2 to count for green cells and in bivariate SSC vs. FL2 density plot as a double check. Statistical analysis All microcosmic experiments were carried out in triplicate. Standard deviations were calculated with Excel (Microsoft®). A Student’s t-test was applied and probabilities less than 0.05 were considered significant. Results and discussion Isolation of the bacterial community Bacterial strains established on the leaves embedded in soil were isolated to obtain a highly diverse bacterial community with the capability of attachment. We used maize leaves as the plants grow fast with no specific requirements. The diversity of the established community over time was followed by DGGE analysis. DGGE is a simple and fast method to screen and compare the diversity of a bacterial community, well suited for this study. The number of bands in a DGGE lane reflects the degree of bacterial diversity and lanes from the same gels can be compared to explore changes in diversity (Muyzer et al., 1993). Based on the number of bands associated to the sampling days 7, 12, and 17 (Fig. 1, lane 1, 4, and 7, respectively), a highly diverse community was observed from day 7 and onwards. This was confirmed by comparing colony morphology of the 48 isolated strains from the different sampling points (data not shown). Due to this, and the fact that the leaves were at this time point highly decomposed (data not shown), the day 7-samples were chosen for strain isolation. To select a manageable, yet still diverse, subcommunity from all of the isolated strains of the day 7 sampling, the colony morphology of all the 48 isolates (from the three replicates) was visually compared. Fifteen isolates appearing morphologically different were chosen, and their DGGE profile was compared with that of the 48 isolates (Fig. 1, lane 2, and 3). Based on the number of bands, a ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 98 C.I. de la Cruz-Perera et al. Day Lane 7 1 2 12 3 4 5 17 6 7 8 9 Table 1. Identification of the isolated strains by 16S rRNA gene analysis Strain name Similarity (%)* Flavobacterium psychrolimnae Pseudomonas lutea Pseudomonas brassicacearum Pseudomonas fluorescens Ochrobactrum rhizosphaerae Chryseobacterium soldanellicol Chryseobacterium letacus Sphingobacteriaceae Xanthomonas retroflexus Micrococcaceae Chryseobacterium ginsengisoli Stenotrophomonas rhizophila Microbacterium oxydans Ensifer adherens Janthinobacterium lividum 97.8 98.1 99.6 99.7 100 98.2 98.5 93.7 99.6 94.2 99.2 99.6 100 98.9 99.7 *Similarity from sequenced 16S rRNA genes calculated with the MATGAT v.2.01 software (see text for more details). Fig. 1. Denaturing gradient gel electrophoresis (DGGE) analysis of the diversity of the bacterial communities isolated from maize leaves after 7, 12, and 17 days of incubation in soil. The DGGE gel shows the PCR amplified products of the 16S rRNA genes of the total bacterial consortia present on the leaves on the sampling days 7, 12, and 17 (lane 1, 4, and 7, respectively). The 16S rRNA gene profiles of a total of 48 cultured strains from each day are presented in lane 2 (day 7), lane 5 (day 12), and lane 8 (day 17). Of these 48, 15 morphologically different strains were selected to constitute representative communities (lanes 3, 6, and 9) of sampling days 7, 12, and 17, respectively. The DGGE analysis indicated that most bacterial diversity was preserved when reducing the community from 48 to 15 strains (by comparing lane 2–3, 5–6, and 8–9). The 15-strain community from sampling day 7 was used for gene transfer analysis. low number of strains were lost when the subcommunity was selected, but most of the initial diversity was represented in the selected community (lane 3). From the 16S rRNA gene sequencing analysis (Table 1), the isolates were identified as typical soil bacteria, mostly gram negatives, with the Pseudomonas and Chryseobacterium genera as the most abundant. Based on this, the strains of the selected subcommunity were considered as well-suited potential recipients of the pKJK10 plasmid. Transfer of pKJK10 to the individual soil isolates and the mixed community The donor strains used in this study encode the lacIq1 repressor gene from the chromosome, repressing GFP expression from pKJK10 when present in these donor ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved strains, as the lac promoter regulates GFP expression in this plasmid. Due to the lack of the lacIq1 repressor in the soil isolates, GFP will be expressed if the plasmid is transferred into these cells. This system thus allows enumeration of transconjugants and donors by direct sample analysis and after IPTG induction, respectively (Sørensen et al., 2003). Detection by FCM has several advantages in such approaches because enumeration of transconjugant cells is based solely on fluorescence markers. There is therefore no need for only including strains with specific antibiotic resistance profiles in the recipient community, and the strains do not need to be capable of expressing the resistance traits encoded by the plasmid to be characterized as transconjugants. The transfer frequency of the conjugative plasmid from the two different donors to the soil isolates was calculated as a transconjugant/donor ratio. No green cells were observed in the negative controls with only the recipient strains present (data not shown), indicating that none of the soil isolates produced auto-fluorescence and that green cells represented plasmid transfer events. When E. coli was used as donor, no transfer of pKJK10 was detected to any of the individual 15 soil isolates, but P. putida was observed to transfer pKJK10 to Stenotrophomonas rhizophila. The plasmid transfer frequency from P. putida to S. rhizophila was higher when the filters were placed on TSA medium (1.07 3.05 9 10 1) compared with R2A medium (0.33 2.32 9 10 2, Table 2), supporting the fact that the metabolic state of the cells may in some cases influence conjugation frequencies (van Elsas & Bailey, 2002). These results reflect the fact that the host range of plasmids depends on the identity of the donor strain (De Gelder et al., 2005). FEMS Microbiol Lett 338 (2013) 95–100 99 Gene transfer in mixed and single species communities Table 2. Transfer frequencies of pKJK10 by different donors to individual recipients and mixed communities* Donor E. coli CSH26::lac Iq P. putida KT2440::lacIq Recipient Frequency S. rhizophila Mixed community S. rhizophila –† 1.56 3.43 9 10 1.07 3.05 9 10 0.33 2.32 9 10 –† Mixed community 1 1 2‡ *All values are transconjugants/donor ratios standard deviation of triplicate experiments. † No transconjugants were obtained on the filter (below detection limit). ‡ Number of transconjugants obtained when placing the filters on R2A plates. In contrast to the results observed when transferring pKJK10 to individual isolates, no plasmid transfer events were observed from P. putida to the mixed community consisting of the same 15 strains applied individually above. Transconjugants were, however, obtained when applying E. coli as donor of pKJK10. The green fluorescent transconjugant cells were sorted by FACS and cultured on TSA agar plates. By sequence analysis of the 16S rRNA gene from four colonies from each replicate, the selected transconjugants were shown all to be identical and identified as Ochrobactrum rhizosphaerae. This does not exclude the possibility that other isolates may also have received the plasmid, but it does show that O. rhizosphaerae in fact did so and that it was the most dominant strain among the plasmid recipients. Interestingly, O. rhizosphaerae was not able to receive the plasmid in the individual mating experiment, indicating that the plasmid permissibility does not only depend on the abilities of the plasmid, host and recipient strains, but also on the surrounding microbial community, which may reduce or enhance plasmid transfer. Both of these scenarios were observed in this study; transfer of pKJK10 from P. putida to S. rhizophila was observed in diparental mating experiments, but not in a mixed community, possibly caused by reduced survival/competition ability of the strains or by the fact that the donor and this specific recipient populations had less opportunity for interaction in the mixed community. In contrast, the presence of a mixed community induced pKJK10 transfer from E. coli to O. rhizosphaerae, which may be due to altered physical cell–cell interaction or the presence of one or several intermediate plasmid host(s). These ‘plasmid step-stones’ may facilitate plasmid transfer from E. coli to O. rhizosphaerae, but are unable to establish and stabilize the plasmid in their own population. Because it was not possible to isolate the strains individually after growth in the community, the fraction of O. rhizosphaerae herein could not be determined; It is possible that O. rhizosphaerae is the FEMS Microbiol Lett 338 (2013) 95–100 dominating strain in the consortium or the most metabolically active, explaining its enhanced abilities as plasmid recipient. Regardless of this strain being dominant or representing a minor population of the community, it is still intriguing that no plasmid transfer was observed in the dual-strains mating from E. coli to O. rhizosphaerae. The results of this study indicate that the surrounding bacterial community strongly impacts the plasmid host range, which needs to be considered when analyzing potential plasmid dissemination in natural environments in association to risk assessment. Plasmid mediated traits, including antibiotic resistance and virulence, may spread to natural bacterial populations in situ, in spite of an apparent narrow host range detected in simple, dual-strain-mating experiments. Acknowledgements This research was supported by funding to Søren Sørensen by The Danish Council for Independent Research (Natural Sciences), The Danish Council for Independent Research (Technology and Production) (ref no: 09-090701, Mette Burmølle) and the Department of Biotechnology and Bioengineering (Cinvestav, Mexico). Claudia I. de La Cruz-Perera received grant-aided support from ‘ConsejoNacional de Ciencia y Tecnologia’ (CONACyT, Mexico) scholarship 166878. References Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W & Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402. Bahl MI, Hansen LH, Goesmann A & Sørensen SJ (2007) The multiple antibiotic resistance IncP-1 plasmid pKJK5 isolated from a soil environment is phylogenetically divergent from members of the previously established alpha, beta and delta sub-groups. Plasmid 58: 31–43. Boto L (2010) Horizontal gene transfer in evolution: facts and challenges. Proc Biol Sci 277: 819–827. Burmølle M, Webb JS, Rao D, Hansen LH, Sørensen SJ & Kjelleberg S (2006) Enhanced biofilm formation and increased resistance towards antimicrobial agents and bacterial invasion are caused by synergistic interactions in multispecies biofilms. Appl Environ Microbiol 72: 3916–3923. Campanella JJ, Bitincka L & Smalley J (2003) MatGAT: an application that generates similarity/identity matrices using protein or DNA sequences. BMC Bioinformatics 4: 29. De Gelder L, Vandecasteele FP, Brown CJ, Forney LJ & Top EM (2005) Plasmid donor affects host range of promiscuous IncP-1beta plasmid pB10 in an activated-sludge microbial community. Appl Environ Microbiol 71: 5309–5317. ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 100 de la Cruz F & Davies J (2000) Horizontal gene transfer and the origin of species: lessons from bacteria. Trends Microbiol 8: 128–133. Ghigo JM (2001) Natural conjugative plasmids induce bacterial biofilm development. Nature 412: 442–445. Guindon S & Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 52: 696–704. James GA, Beaudette L & Costerton JW (1995) Interspecies bacterial interactions in biofilms. J Ind Microbiol 15: 257–262. Lane DJ (1991) 16S/23S rRNA sequencing. Nucleic Acid Techniques in Bacterial Systematics (Stackebrandt E & Goodfellow M, eds.), pp. 115–174. John Wiley & Sons, Inc., New York. Madsen JS, Burmølle M, Hansen LH & Sørensen SJ (2012) The interconnection between biofilm formation and horizontal gene transfer. FEMS Immunol Med Microbiol 65: 183–195. Muller S & Nebe-von-Caron G (2010) Functional single-cell analyses: flow cytometry and cell sorting of microbial populations and communities. FEMS Microbiol Rev 34: 554–587. Musovic S, Oregaard G, Kroer N & Sørensen SJ (2006) Cultivation-independent examination of horizontal transfer and host range of an IncP-1 plasmid among Gram-positive and Gram-negative bacteria indigenous to the barley rhizosphere. Appl Environ Microbiol 72: 6687–6692. Musovic S, Dechesne A, Sørensen J & Smets BF (2010) Novel assay to assess permissiveness of a soil microbial community toward receipt of mobile genetic elements. Appl Environ Microbiol 76: 4813–4818. Muyzer G, de Waal EC & Uitterlinden AG (1993) Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction- ª 2012 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved C.I. de la Cruz-Perera et al. amplified genes coding for 16S rRNA. Appl Environ Microbiol 59: 695–700. Poole AM (2009) Horizontal gene transfer and the earliest stages of the evolution of life. Res Microbiol 160: 473–480. Posada D (2008) jModelTest: phylogenetic model averaging. Mol Biol Evol 25: 1253–1256. Reisner A, Holler BM, Molin S & Zechner EL (2006) Synergistic effects in mixed Escherichia coli biofilms: conjugative plasmid transfer drives biofilm expansion. J Bacteriol 188: 3582–3588. Rossell o-Mora R & Amann R (2001) The species concept for prokaryotes. FEMS Microbiol Rev 25: 39–67. Sengeløv G, Kristensen KJ, Sørensen AH, Kroer N & Sørensen SJ (2001) Effect of genomic location on horizontal transfer of a recombinant gene cassette between Pseudomonas strains in the rhizosphere and spermosphere of barley seedlings. Curr Microbiol 42: 160–167. Seoane J, Yankelevich T, Dechesne A, Merkey B, Sternberg C & Smets BF (2011) An individual-based approach to explain plasmid invasion in bacterial populations. FEMS Microbiol Ecol 75: 17–27. Sørensen SJ, Sørensen AH, Hansen LH, Oregaard G & Veal D (2003) Direct detection and quantification of horizontal gene transfer by using flow cytometry and gfp as a reporter gene. Curr Microbiol 47: 129–133. Sørensen SJ, Bailey M, Hansen LH, Kroer N & Wuertz S (2005) Studying plasmid horizontal transfer in situ: a critical review. Nat Rev Microbiol 3: 700–710. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F & Higgins DG (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25: 4876–4882. van Elsas JD & Bailey MJ (2002) The ecology of transfer of mobile genetic elements. FEMS Microbiol Ecol 42: 187–197. FEMS Microbiol Lett 338 (2013) 95–100