Algae-Bacteria Interactions During Commercial Production of Nannochloropsis Sp. in Closed Photobioreactors

Algae-bacteria interactions during commercial production of Nannochloropsis sp. in closed photobioreactors

Javier Giraldo

Academic Year 2018-2019 Promoters: Prof. Dr. Wim Vyverman Dr. Luc Roef Dissertation submitted in fulfilment of the requirements for the degree of Doctor (Ph.D.) in Sciences: Marine Sciences (Ghent University)

Members of the examination board

Voting members:

Prof. Dr. Koen Sabbe, Ghent University, Belgium (Chair)

Prof. Dr. Anne Willems, Ghent University, Belgium (Secretary)

Prof. Dr. Ir. Sven Mangelinckx, Ghent University, Belgium

Prof Dr. Koenraad Muylaert, KU Leuven-Kulak, Belgium

Prof. Dr. Peter Kroth, Konstanz University, Germany

Dr. Peter Chaerle, Ghent University, Belgium

Non-voting members:

Dr. Luc Roef, Proviron Holding NV

Prof. Dr. Wim Vyverman, Ghent University

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642575. This work makes use of resources and facilities provided by UGent as part of the Belgian contribution to EMBRC-ERIC.

Contents Summary...... 9 Samenvatting ...... 13 Chapter I ...... 17 General introduction and thesis objectives ...... 17 1. General introduction to industrial algal cultivation ...... 18 1.1. Potential algae applications ...... 18 1.2. Cultivation of microalgae ...... 19 2. Algae production for aquaculture at Proviron ...... 22 2.1. Characteristics of Nannochloropsis sp...... 22 2.2. The ProviAPT system ...... 23 2.3. Exploiting algae potential and overcoming limitations ...... 24 2.3.1. Reactor design ...... 25 2.3.2. Operating conditions ...... 25 2.3.3. Dealing with biofouling at Proviron ...... 26 3. The algal microbiome: algae-bacteria interactions ...... 28 4. Objectives and outline of this thesis ...... 29 Chapter II ...... 33 Exploration of the algal microbiome during industrial cultivation of Nannochloropsis sp. in closed photobioreactors and its potential effect on lipids composition ...... 33 Abstract ...... 34 1. Introduction ...... 35 2. Materials and methods ...... 35 2.1. Algae cultivation and sample selection ...... 35 2.2. Sample selection and characterization ...... 36 2.3. Fatty acids profiling ...... 37 2.4. Microbial characterization ...... 38 2.5. Statistical analysis ...... 38 3. Results and discussion...... 39 3.1. Reactor characterization and sample selection ...... 39 3.2. Total fatty acids content and fatty acids profile ...... 42 3.3. Microbial diversity ...... 44 3.4. Correlations between microbial diversity and EPA content ...... 49 4. Acknowledgements ...... 51

5. Supplementary material ...... 52 Chapter III ...... 61 Influence of the algal microbiome on biofouling during industrial cultivation of Nannochloropsis sp. in closed photobioreactors ...... 61 Abstract ...... 62 Keywords ...... 62 Highlights ...... 62 1. Introduction ...... 63 2. Materials and methods ...... 65 2.1. Biofilm sampling protocol and selected samples ...... 65 2.2. Microbial isolation ...... 66 2.3. Bacterial identification by MALDI-TOF MS and 16S rRNA gene sequencing ...... 67 2.4. Isolation, treatment and identification of an invasive green algae species ...... 67 2.5. Algae cultivation in shake flasks ...... 68 2.6. Co-cultivation of Nannochloropsis sp. with the green alga species ...... 68 2.7. Biofilm formation assay ...... 68 2.8. Statistical analysis ...... 69 3. Results and discussion...... 70 3.1. Microbial isolation ...... 70 3.2. Effect of the microbiome on the productivity of Nannochloropsis sp...... 75 3.3. Poseidonocella sp. is the strongest biofilm inducer among 24 isolated bacterial strains 78 4. Conclusions ...... 79 5. Acknowledgements ...... 80 Chapter IV ...... 81 Towards identification of chemical signals involved in biofilm formation during non-axenic cultivation in closed photobioreactors ...... 81 Abstract ...... 82 1. Introduction and experimental approach ...... 83 2. Materials and Methods ...... 83 2.1. Cultivation of Nannochloropsis sp...... 83 2.2. Cultivation of Poseidonocella sp...... 84 2.3. Biofilm formation assay ...... 84 2.4. Washing step and SSN preparation ...... 85 2.5. Heat treatment on SSN ...... 85 2.6. Solid phase extraction protocol ...... 85

2.7. UHPLC measurements ...... 86 2.8. LC-HR-MS data analysis ...... 87 3. Results and discussion...... 88 3.1. Effect of Poseidonocella sp. on BF formation in Nannochloropsis sp. cultures ...... 88 3.2. Effect of glucose on biofilm formation in Nannochloropsis sp. cultures ...... 89 3.3. Effect of the sterile supernatant from Poseidonocella sp. on biofilm formation in Nannochloropsis sp. cultures ...... 90 3.4. Effect of solid phase extracts on biofilm formation in Nannochloropsis sp. cultures ...... 92 3.5. UHPLC measurements and LC-HR-MS data analysis ...... 95 4. General discussion ...... 99 5. General conclusions ...... 101 6. Acknowledgements ...... 102 Chapter V ...... 103 Towards microbiome engineering of industrial microalgae cultivation: effects of non-native bacteria on Nannochloropsis sp. growth, metabolome and biofouling ...... 103 Abstract ...... 104 Keywords ...... 104 Highlights ...... 104 1. Introduction ...... 105 2. Materials and methods ...... 106 2.1. Growth of Seminavis robusta in co-cultivation with isolated bacteria ...... 106 2.2. Nannochloropsis sp. cultivation and co-cultivation with isolated bacteria from S. robusta 107 2.3. Biofilm formation assay ...... 108 2.4. Sample preparation for endometabolomics of S. robusta ...... 109 2.5. Sample preparation for endometabolomics of Nannochloropsis sp...... 109 2.6. Endometabolite extraction, derivatization and GC-HRMS measurements ...... 110 2.7. GC-HRMS data analysis ...... 110 3. Results ...... 111 3.1. Growth of Nannochloropsis sp. in co-cultivation with bacteria isolated from S. robusta111 3.2. The endometabolome of microalgae is influenced by bacteria ...... 113 3.3. Bacteria isolated from S. robusta enhance biofilm induction of Poseidonocella sp. on Nannochloropsis sp...... 118 4. Discussion ...... 119 5. Acknowledgements ...... 123

Chapter VI ...... 129 General discussion ...... 129 1. Introduction ...... 130 2. Main outcome and relevance of this work ...... 130 2.1. Cultivation conditions modulate algal microbiome and product quality ...... 130 2.2. Deciphering the bacterial influence on algae cultivation by isolation and co-cultivation 131 2.3. Targeting early stage biofilms: Poseidonocella sp. induces biofilm formation through successfully extracted polar compounds...... 132 2.4. Towards microbiome optimization: bacteria modulates algal metabolism through species-specific interactions ...... 133 3. Recommendations and future perspectives ...... 135 References ...... 137 Acknowledgement ...... 173 Curriculum Vitae ...... 177

Summary

Microalgae cultivation has increased in attention in recent years due to its wide range of potential applications – among others: aquaculture. Microalgae may be produced in open ponds or closed photobioreactors (PBRs) typically in non-axenic conditions. Biofouling and biofilm formation are frequent problems during algae cultivation. In closed PBRs, biofilm formation diminishes the light penetration, thus reducing the productivity. Proviron Holding NV (Belgium) is a chemical company specialized in microalgae production. They designed the ProviAPT which is a proprietary closed photobioreactor system with promising features. This work aims to understand the role of the algal microbiome on biofouling and biofilm formation during industrial cultivation using the microalgae production plant from Proviron. This knowledge may be used to design and implement a prevention strategy against biofilm forming agents. To achieve this goal, it is necessary to characterize the microbial population in a wide range of samples including reactors with high performance and reactors with evidence of biofouling. This enables us to find correlations between the presence/absence of bacterial groups and reactor performance. However, this approach fails to confirm the role of individual species and its ability to induce or prevent biofilm formation. Therefore, bacterial isolation and identification is necessary. The co-cultivation of the isolated strains with the algae is a precise method to identify relevant organisms. Once the biofilm inducers are found, it is possible to elaborate strategies towards the prevention of this phenomenon on industrial scale. Following this approach, this thesis is divided into 3 research chapters and 1 appendix. Chapter II – The objective of this chapter is to identify microbial communities cohabiting the ProviAPT reactors at different conditions. A representative set of samples was selected and characterized based on reactor conditions and fatty acids profiles. The microbial communities of those samples were analyzed using next-generation DNA sequencing techniques. Correlations between reactor performance, fatty acid content and microbial community were studied. Chapter III – Microbial isolation and identification of microorganisms was performed after selecting a representative set of samples. Microbial isolates were co-cultivated with Nannochloropsis sp. in a biofilm formation assay in order to identify potential biofilm inducers. Isolation was performed by cultivation of samples in agar plates. Identification was performed by Maldi-TOF and Sanger DNA sequencing. Several species were identified as biofilm inducers, Poseidonocella sp. being the one with strongest effect. Chapter IV – In chapter 3, Poseidonocella sp. was identified as a strong biofilm inducer of Nannochloropsis sp. The biofilm forming system Nannochloropsis- Poseidonocella is represented as a simplification of biofilms formed in the ProviAPT reactors and thus can be used to study the molecular mechanisms behind the biofilm forming process are studied. As biofilm prevention mostly targets early stage biofilms,

10 signal molecules from the bacterial exometabolome with the ability of initiating biofilm formation were explored. Chapter V – Algae-bacteria interactions may induce different effects on algae. In this chapter, synthetic microbial communities are explored in order to find potential algae growth enhancer and biofilm preventing candidates. Well-studied bacterial strains, associated to the diatom Seminavis robusta, were co-cultivated with Nannochloropsis sp. The responses on growth rate and biofilm formation were studied and, in addition, metabolic changes were analyzed. Effects of bacterial strains on algae were found to be species specific. Chapter VI – In this chapter, main findings are discussed and different approaches towards the prevention of the effect of harmful microorganisms are presented. Mechanic approaches on reactor design and operational conditions have been already taken. However, biological approaches require longer experimental and implementation time. This chapter finalizes with suggestions on how to proceed after this thesis as well as overall conclusions.

Samenvatting

Het kweken van microalgen is de afgelopen jaren sterk toegenomen vanwege het brede scala aan mogelijke toepassingen, waaronder ook in de aquacultuur. Microalgen kunnen worden geproduceerd in open vijvers of gesloten fotobioreactoren (PBR's), typisch in niet-axenische omstandigheden. Biofilmvorming op de wanden van de reactoren is een frequent probleem tijdens de algenkweek. In een gesloten PBR vermindert de biofilmvorming de lichtpenetratie waardoor de productiviteit verminderd. Proviron Holding NV (België) is een chemisch bedrijf dat gespecialiseerd is in de productie van microalgen. Ze hebben met de ProviAPT een gesloten fotobioreactorsysteem met veelbelovende functies ontworpen. In dit werk hebben we gepoogd om de rol van het algen microbioom op biofilmvorming tijdens industriële kweek van microalgen beter te begrijpen, gebruikmakend van de productie site van Proviron. Deze kennis kan worden gebruikt voor het uitwerken en implementeren van een preventiestrategie tegen biofilmvorming. Om dit doel te bereiken, is het noodzakelijk de microbiële gemeenschap in diverse situaties te karakteriseren, zowel in productieve reactoren als in onproductieve reactoren met biofilms. Dit maakt het mogelijk om correlaties te vinden tussen de aanwezigheid / afwezigheid van bacteriële groepen en de productiviteit van de reactor. Deze benadering slaagt er echter niet in om de rol van individuele soorten in de biofilmvorming te bevestigen, waardoor het isoleren van bacteriën noodzakelijk blijft. De co-cultivatie van de geïsoleerde bacteriële stammen met de algen is een efficiënte methode om alsnog relevante organismen te identificeren. Zodra de biofilm-inducerende bacteriën zijn gevonden, is het mogelijk om gerichte strategieën uit te werken tegen deze bacteriën op een industriële schaal. Dit proefschrift is onderverdeeld volgens deze benadering, in 3 experimentele hoofdstukken en 1 appendix. Hoofdstuk II - Het doel van dit hoofdstuk is om microbiële gemeenschappen te karakteriseren die voorkomen in de ProviAPT-reactoren onder verschillende omstandigheden. Een representatieve reeks stalen werd geselecteerd op basis van reactoromstandigheden en vetzuurprofielen. De gemeenschappen van deze stalen werd gekarakteriseerd op basis van high-throughput DNA-sequeneringstechnieken. De correlaties tussen de reactorprestatie, het vetzurengehalte en de microbiële gemeenschap werden bestudeerd. Hoofdstuk III - Van een representatieve reeks stalen werden micro-organismen geïsoleerd. Deze isolaten werden vervolgens geco-cultiveerd met Nannochloropsis sp. om potentiële biofilm inducerende organismen te identificeren. De isolaties werd uitgevoerd door de reactorstalen uit te platen op agarplaten. De identificaties werd uitgevoerd met behulp van Maldi-TOF, gevolgd door Sanger sequencing. Verschillende isolaten werden geïdentificeerd als biofilminduceerders, waarvan Poseidonocella sp. het sterkste effect had.

Hoofdstuck IV - In hoofdstuk 3 werd Poseidonocella sp. geïdentificeerd als een sterke biofilm-induceerder bij Nannochloropsis sp. De biofilm vorming in de Nannochloropsis- Poseidonocella co-cultuur wordt gezien als een vereenvoudiging van biofilms gevormd in de ProviAPT-reactoren en kan dus worden gebruikt om de moleculaire mechanismen achter het biofilmvorming in dit systeem te bestuderen. Aangezien biofilm-preventie zich vooral richt op het verstoren van de vroege stadia in de biofilmvorming, werden signaalmoleculen van het bacteriële exometaboloom met het vermogen biofilmvorming te initiëren onderzocht. Hoofdstuk V - In dit hoofdstuk worden synthetische microbiële gemeenschappen gescreend op potentiële stimulatoren van algengroei en inhibitoren van biofilmvorming. Reeds goed bestudeerde bacteriestammen, afkomstig van de diatomee Seminavis robusta, werden gecultiveerd met Nannochloropsis sp. Naast het effect van deze stammen op de groei van de alg en de biofilmvorming, werden metabole veranderingen, die ze induceerden, geanalyseerd. De effecten van bacteriestammen op algen bleken soortspecifiek te zijn. Hoofdstuk VI - In dit hoofdstuk worden de belangrijkste bevindingen besproken en worden verschillende benaderingen voor de preventie van schadelijke effecten ten gevolge van micro-organismen toegelicht. Hoewel mechanistische benaderingen van reactorontwerp en operationele omstandigheden al zijn genomen, vereisen biologische benaderingen een langere experimentele en implementatietijd. Dit hoofdstuk eindigt met enkele suggesties over hoe verder gegaan dient te worden na dit proefschrift en formuleert de algemene conclusies.

Chapter I

General introduction and thesis objectives

1. General introduction to industrial algal cultivation

1.1. Potential algae applications Interest in algae cultivation has risen worldwide as they are a good source of different nutrients which make algae attractive for different market niches such as food, animal feed, aquaculture, fuels and cosmetics [1,2]. Algae are photosynthetic organisms which occur in nearly all ecosystems around the planet. Two main groups are commonly recognized: microalgae (unicellular) and macroalgae or seaweed (multicellular). In this thesis, the term microalgae refers to unicellular photosynthetic organisms including thus unicellular algae, diatoms and cyanobacteria. They make use of carbon dioxide, mineral nutrients and solar energy to produce the building blocks necessary for their cell‐ structure and physiological processes, but also some strains are capable of heterotrophic [3] and mixotrophic [4] metabolism. In this work, we focus on autotrophic cultivation of microalgae. Microalgae are marketed in a range of products from pure algal biomass in the form of tablets, capsules or liquids as well as part of other products like pastas, snack foods, and beverages [5] with multiple applications. The nutritional value of microalgae importantly varies with the species and also with different culture conditions [6]. Different commercial microalgae products are currently available in the market for human consumption [7] e.g. β-carotene extracts from Dunaliella salina, Chlorella tablets [8] or Spirulina (Arthrospira platensis) food supplements [9]. In aquaculture, the main application of microalgae is usage for functional feed as certain components present in algae are essential for the development of larvae [10]. Monospecies or mixtures of microalgae can offer a valuable source of nutrients for the larvae of fish, bivalves and crustaceans, three major groups in aquaculture [11]. Microalgae species such as Isochrysis galbana, Diacronema lutheri, Tetraselmis suecica, Pseudoisochrysis paradoxa, Chaetoceros calcitrans and Skeletonema costatum are successfully used for bivalve culture [12]. Most of the strains used for feeding bivalve larval stages are also used as direct feed for shrimp larval culture, especially species such as Skeletonema spp. and Chaetoceros spp. Diacronema lutheri, T. suecica or Nannochloropsis spp. are commonly fed to Artemia or rotifers, which are used as feed for later larval stages of crustacean and fish larvae [12]. Live microalgae are used to eliminate excess dissolved nutrients from aquaculture effluents [13]. This green water technique is an efficient and cost effective waste water treatment method. Moreover, pigments from microalgae are used for the flesh color enhancement of species such as salmon and trout [14]. Freshwater green alga Haematococcus pluvialis is a natural source of astaxanthin which is commonly used for color enhancement of fish flesh [15]. Microalgae are used as a source of valuable bioactive compounds such as polyunsaturated fatty acids (PUFAs), antioxidants, compounds with pharmaceutical and

18 nutraceutical value [16]. PUFAs are an ingredient of infant formulations and nutritional supplements. Microalgae have proved to be a promising source of PUFAs due to their high content[17]. Algal PUFAs have been reported with promising medical applications [18] as a viable sustainable alternative to replace PUFA production from fish oil [19]. Polysaccharides from microalgae are used as pharmaceuticals, food, cosmetics, fabrics, stabilizers and emulsifiers [20]. Polysaccharides like agar, alginates and carrageenan extracted from algae are economically important in the food industry as gelling or thickening agents in marmalade, ice creams and jellies [8]. Polysaccharide extracts from microalgae may be used as moisturizing agents and thickening agents in cosmetic industry [21]. Microalgae are rich in pigments [22] and therefore represent an interesting pigment source [23]. Among others, astaxanthin is used as an antioxidant and fucoxanthin is used in cosmetic industry to prevent oxidative stress caused by UV radiation [24]. Carotenoid pigments extracted from genera such as Dunaliella and Chlorella are used as natural color enhancers in cosmetic industry [21]. Microalgae extracts have become a common ingredient in face and skin care products (e.g., anti-aging cream, refreshing care products, emollient and anti-irritant in peelers) [25].

1.2. Cultivation of microalgae The concept of biorefinery has been increasing in importance during recent years [26,27]. Production of a large variety of products from low-cost raw materials has raised the enthusiasm for a bio-based economy. Microalgae cultivation has been proven to be a suitable option for the biorefinery concept as a large variety of products from microalgae within the simultaneous use of proteins, fractionation of lipids and residual nutrient feedstocks may be obtained [28]. As photosynthetic single-cells microorganisms they mostly require inexpensive substrates and sunlight for growing, while they could also reduce the freshwater consumption by growing halotolerant algae in seawater [29]. This could include the production of valuable compounds while waste water is treated [13]. Strain improvement can be achieved by genetic modifications [30]. It is also possible to explore the nature potential. Apart from the extensive catalogue of commercially available microalgae species, there is strong potential in finding relevant species by bio-prospecting [31,32], using selective environments [33] or adapting cell cultures [34]. To optimally exploit the microalgae potential, mild and efficient separation technologies are needed too [35]. Despite the great potential, these microorganisms present important limitations during industrial production. They exhibit relatively high growth rates but it is still very challenging to achieve high cell densities. This is mainly due to the light distribution in photobioreactors (PBR). While cell density increases, the self-shading effect increases too. Single cells absorb light and refract a fraction of the irradiated light. Therefore, the further from the light irradiated area, the chances of finding light limited cells increase.

Light limitation might produce changes in the algae life cycle, affecting morphology and biochemical stoichiometry [36,37]. This could be solved by increasing the light irradiation. However, this could lead to photoinhibition [38] which also results in a decrease of growth rate. Current industrial solutions for light distribution involve an array of different shapes for the PBRs [39]. Tubular reactors consist of an array of horizontal tubes that are ideally used in locations with mild light irradiation[40]. The tubes could also be vertical, which generally increases the efficiency since the areal productivity is increased [41]. Most promising industrial shapes are vertical flat panels due to a higher volume/area ratio and hence higher yield [39]. Furthermore, the production can be developed in closed and independent reactors which gives the possibility of producing different algae strains with a better control over contaminations. Harvesting microalgae is an energy demanding step [42]. Nowadays, the most common technique is centrifugation. However, high speed of rotation induces shear stress and this may have negative effects on sensitive cells leading into breaking cell walls in some cases. De-watering techniques are promising candidates to be used prior to centrifugation in order to reduce the energy demand. This step should be efficient and not compromise the final product. Typical de-watering steps are flocculation methods that may consist in pH modifications [43,44], bio-flocculation [45,46] or use of cationic polymers [47,48]. After centrifugation, the microalgae culture is harvested in an algae paste. For applications that simply require the algae biomass as it is generally the case in aquaculture, the product can be already commercialized at -20 °C. However, the algae paste can be freeze dried to preserve its nutritional value and increase the product stability. This technique prolongs the life of the product and enables the possibility to store it at room temperature. If it is desired to use a specific component of the algal cell – as could be pigments, lipids or proteins – further processing is needed. Due to the algal cell-wall, cell disruption techniques are needed. While generally algal cell walls are rigid, some species of the genus Dunaliella lack a rigid cell wall [49]. Depending on the cell wall rigidity, the cell disruption technique will be adapted. These techniques need to be mild in order to preserve algae biochemical integrity during disruption [50]. This procedure will also influence the extraction yields for lipids [51,52] and proteins [53]. The cell-wall is one of the main barriers in the downstream processing of microalgae and it is the main particularity compared to other bioprocesses. When the product is not naturally excreted, a cell-disruption method needs to be applied. Otherwise, the disrupted material can be used for extraction techniques. It can be the case that the algae only produce the precursor of the final product, as it is the case for biofuel production [54]. Therefore, conversion methods need to be applied and optimized. During algae cultivation, problems like biofouling and biofilm formation may occur. When it happens in in closed PBRs – typically under non-axenic conditions –it becomes a harmful phenomenon [55]. The negative consequences of biofouling include

20 accelerated degradation or corrosion and impaired function in culture systems, and significant costs associated with equipment maintenance, repair, or replacement [56]. Biofilm formation in PBRs reduces the light penetration to algae which inhibits the photosynthesis and as a result, reduces the cell growth and biomass yield [55,57]. Biofilms can be formed on both abiotic surfaces and living surfaces [58]. The extracellular polymeric substances (EPS) matrix is the primary structuring agent for microbial microenvironments and controls the physical properties of biofilms [59]. Polymeric substances of biofilm are typically constituted by exopolysaccharides (40- 95%), proteins (1-60%), nucleic acids (1-10%) and lipids (1-40%) [60]. Composition of EPS is very important in biofilm development since it determines the chemical and physical properties of EPS [61]. This diversity of EPS confers unique biofilm morphology. Furthermore, bacteria generate multiple EPS in the biofilms that protects the bacterial cells against various stresses such as desiccation, temperature and other competing microbes [62]. It is a dynamic, frequent and progressive process [63] where four phases may be distinguished (Figure 1). First, it starts with the conditioning of a surface, typically coated with EPS (figure 1.A). The ability of microorganisms to adhere to an abiotic surface is influenced by electrostatic charge, hydrophobicity, wettability, roughness, microtopography of the substrate and surface chemodynamic properties such as nutrient enrichment and charge accumulation [64]. Then, bacteria, cyanobacteria, protists, diatoms, and other unicellular organisms may attach and develop colonies [65] (figure 1.B). There are different algae and bacteria species which produce EPS abundantly. For instance, microalgae Chlorella vulgaris and Botryococcus braunii are among the abundant EPS producers [66]. In mature biofilms, symbiotic relationships may be established among different species (Figure 1.C) and nutrients may be retained (Figure 1.D). From mature biofilm, some bacterial cells convert to planktonic growth and disperse from the biofilm to find new substrate [64]. Thus, dispersal is the final stage of the biofilm lifecycle and at the same time, the start of a new biofilm lifecycle [62]. Dispersal can be active or passive depending on cell motility, degradation of EPS and nutrient limitation caused by nitrogen, phosphorous or iron scarcity while passive dispersal depends on physical factors such as shearing force under liquid flow conditions [62,64].

Figure 1 Development of a mixed community algal biofilm: (A) growth surfaces are first conditioned with bacteria cells that excrete the initial EPS matrix; (B) various species of algae cells present in the bulk medium then begin to colonize the EPS matrix; (C) the algae cells grow and reproduce, forming a symbiotic relationship with the bacteria present in the EPS matrix; and (D) a mature biofilm matrix is densely populated with algae cells, particularly cyanobacteria and chlorophytes, and retains nutrients in the EPS matrix (Figure and caption extracted from Schnurr et al. (2015) [67]). 2. Algae production for aquaculture at Proviron The Flemish company Proviron (Belgium) commercially produces microalgae for aquaculture purposes. Along this section, we revise the potential of algae in aquaculture and describe details about their production system.

2.1. Characteristics of Nannochloropsis sp. Nannochloropsis sp. is one of the microalgal genera produced by Proviron. This genus receives increasing attention due to its high photosynthetic efficiency, converting carbon dioxide into storage lipids mainly in the form of triacylglycerols and the ω-3 long-chain PUFA eicosapentaenoic acid (EPA) [68]. Apart from PUFAs, these algae cells produce high concentrations of valuable pigments such as chlorophyll a, zeaxanthin, canthaxanthin and astaxanthin [69]. Glucose is the most dominant sugar in the polysaccharide composition while in the protein composition, the amino acids aspartate, glutamate and proline are predominant [70]. Its nutritional value – particularly the high levels of EPA oscillating 2-3% dry biomass depending on cultivation conditions [71] – provides this genus increasing interest for human consumption as these nutrients may contribute on the prevention of several diseases [72]. Nannochloropsis is a genus of unicellular and non-mobile marine microalgae belonging to: Phylum Heterokontophyta, Class Eustigmatophyceae and family Eustigmataceae. It

22 was first named by Hibberd (1981) [73]. Currently, it comprises 5 known species (Nannochloropsis australis, Nannochloropsis granulata, Nannochloropsis limnetica, Nannochloropsis oceanica and Nannochloropsis oculata) as the species Nannochloropsis salina and Nannochloropsis gaditana recently have been transferred to the genus Microchloropsis [74]. The cell morphology consists in spheres with diameters varying around 3 μm [75]. Nannochloropsis is considered as a potential oleaginous model microalga because of the great photosynthetic efficiency, high lipid productivity, well established genetic toolbox and relatively mature technology for outdoor cultivation systems on a large scale [76–78].

2.2. The ProviAPT system Proviron Holding NV designed the ProviAPT system which is described as a promising closed photobioreactor (PBR) system for large scale microalgae production [79] and it has been used for successful microalgae production for years. The ProviAPT system includes an array of vertical flat-panel type reactors enclosed in a translucent plastic bag filled with water. This feature increases productivity compared to other systems. This system was used to cover a production site of ~1 Ha which has enabled successful microalgae production for now more than 12 years (Figure 2).

Figure 2 Outdoor ProviAPT reactors in microalgae production site of Proviron Holding NV during 2016 (Antwerp, Belgium)

Despite its high productivity compared to other systems [39], this type of PBR has room for improvement. One of the major drawbacks -inherent to closed photobioreactor system- is biofilm formation. During non-axenic cultivation, mixed biofilms often grow on reactor walls which diminish the productivity due to a reduction in light penetration. During 2017, transition from outdoor to indoor production in a vertical farming setup took place (figure 3) with a significant reduction in biofilm formation. It was presumed

23 that extremely high light irradiances during outdoor production in the summer contributed to the development of biofilms, most probably due to high light stress and consequential cell death. This is currently avoided as light can be controlled during indoor production. Indoor lighting conditions that contribute to reduce biofilm formation have been identified and applied for industrial cultivation. These are related to light spectrum and intensities. To achieve optimal lighting conditions, the algae photosynthetic efficiency, the influence on biofilm development and electricity costs need to be considered.

Figure 3 Indoor ProviAPT production plant with vertical algal cultivation under artificial light irradiation during 2018 (Hemiksem, Belgium).

Despite the transition from outdoor to indoor production, algal cultivation is still performed under non-axenic conditions and the role of the algal microbiome remains unexplored. The presence of undesired microorganisms has been importantly reduced. However, high bacterial loads and even presence of flagellates are occasionally detected. Understanding the variations of the algal microbiome and its influence on reactor performance will provide valuable information towards increasing productivity.

2.3. Exploiting algae potential and overcoming limitations The algal microbiome represents a key factor to consider for the enhancement of algae productivity in industry. Major breakthroughs applying these approaches are expected to be achieved in the following years. In the meantime, mechanical operations may be applied in order to maximize algae productivity. In this section, common reactor

24 configurations and operations are discussed as well as how these approaches have been successfully applied at Proviron.

2.3.1. Reactor design One of the major challenges in reactor design is to homogenize light distribution. Current industrial solutions involve an array of different shapes for the PBRs [39]. Additionally, most of the industrial production systems are designed for outdoor production as they aim to reduce the electricity input avoiding artificial lighting. The most common production system is the open pond. It is simple to build and can easily cover extensive areas. However, the Volume/Area ratio is too low since it is desired to prevent the shadow effect. Furthermore, it is an open system which limits the range of species that can be produced. To ensure successful algae growth in this case, they require specific features to outcompete other microorganisms e. g. halophilic algae growing at very high salinity. Besides contamination, during outdoor production the algae broth will be diluted during rainy periods or suffers from evaporation due to high temperatures. Open ponds may also be located in greenhouses, creating thus a hybrid situation between open and close systems. This allows the use of both natural and artificial light and partially protects cultures from environmental conditions and contaminants. Other possibilities are closed PBRs systems, which include tubular, bag-type or panel reactor designs. Tubular reactors consist in an array of horizontal tubes that are ideally used in locations with mild light irradiation [40]. The tubes could also be vertical, which generally increases the efficiency since the areal productivity is increased [41]. Most promising industrial shapes are vertical flat panels due to its higher yield and Volume/Area ratio [39]. Furthermore, the production can be developed in closed and independent reactors which give the possibility of producing different algae strains with a better control over contaminations.

2.3.2. Operating conditions Operating conditions may have an influence in the development of contaminations and biofilm formation. Industrial microalgae production typically operates in semi-continuous or batch mode as continuous mode is typically applied when only small quantities are needed. Semi-continuous and continuous operations allow the dilution of the reactors which could be beneficial as it increases light availability and contributes as a strategy to get rid of contaminants by decreasing their concentration below their survival threshold.

Algae may grow autotrophically, heterotrophically or mixotrophically depending on whether the carbon source is inorganic, organic or both simultaneously. Mixotrophic growth may increase the growth rate of algae as algal cells may use simultaneously inorganic and organic carbon [80,81]. However, the presence of organic carbon might stimulate the growth of harmful bacteria too during non-axenic cultivation. Furthermore, algal pigmentation might be affected with heteroprophic growth and although

25 pigmentation might be preserved in mixotrophic cultures, the energy conversion efficiency is reduced [4]. Media composition also affects bacterial population during non-axenic cultivation. Salinity needs to be considered as high level might reduce the chances of contaminations but usually affects algal growth and increase costs. Optimal salinity varies with every species affecting biomass yield and composition [82,83]. Furthermore, optimal media composition must find proper balance between every compound according to algae requirements as deficiency in any of them might have consequences in the metabolism. Nevertheless, this might be a desired feature if it is intended to obtain secondary metabolites only secreted under stress e. g. halophilic algae Dunaliella salina excretes β-carotene under stress conditions i.e. nitrogen starvation [84]. Diverse considerations may be applied to reduce the presence of contaminants and biofilm development during non-axenic cultivation. Mechanical treatments like adding granulate beads with different densities to the reactors are often use to have them mildly scraping the reactor walls and thus prevent biofilm development. Bactericidal compounds and biofilm destabilizers might be added for the same purpose but special attention needs to be paid on the algae since these compounds might affect their viability and affect the product quality. Biological strategies might also be applied as once the algal microbiome is understood it can be modified accordingly to increase biomass yield and robustness against contaminations.

2.3.3. Dealing with biofouling at Proviron Proviron Holding NV designed the ProviAPT system which is a promising closed photobioreactor (PBR) system for large scale microalgae production [79]. The ProviAPT system includes an array of vertical flat-panel type reactors which is a feature that increases productivity compared to other systems [39]. This system consists in single units that may cover any desired surface. Initial design was enclosed in a translucent plastic bag filled with water in order to provide stability to the panels and buffer temperature during outdoor algae cultivation. Recently, this production system has been adapted for indoor production, removing the water mattress and incorporating LED light. From the initial designs, special attention has been paid to aeration of the ProviAPT reactors. Aeration and liquid recirculation prevent microorganisms from attaching to the reactor walls due to high tangential flows across the reactor surface while keeping the reactor volume homogenized. Lack of homogenization creates local hot spots of older culture material which contributes to an increase in bacterial growth and hence development of biofilms during non-axenic cultivation. During this period, special measures have been designed and optimized, involving a proprietary system of nutrient application that guarantees a prolonged preservation of optimal air flows. Furthermore,

26 a new aeration driven method for culture recirculation was designed. The beneficial effects of these measures have become clear throughout the project. We have also identified suboptimal nutrient conditions in the recipes we were using. Especially existing K and Mg concentrations appeared to have the most negative effects on growth with increasing fouling. These have been adapted and this has shown immediate positive effect on growth. It is not believed that K and Mg had a direct effect on bacterial growth itself but rather acted on the health of the algae themselves. Consequently, algae cultures appeared less stressed, resulting in decreased cell death and biofouling also diminished. Moreover, a transition from outdoor to indoor production on 2017 has reduced biofilm formation. It was observed that extremely high light irradiances during outdoor production in the summer contributed to the development of biofilms, most probably due to high light stress and consequential cell death. This is currently avoided as light may be controlled during indoor production. Indoor lighting conditions that contribute to reduce biofilm formation have been identified and applied for industrial cultivation. These are related to light spectrum and intensities. Optimal light conditions need to consider the algae photosynthetic efficiency, the influence on biofilm development and electricity costs. During a coarse screening of light conditions involving 6 commercially available fixed spectrum solutions and one variable setup we identified conditions that proved beneficial against biofilm formation. These tests showed that the color composition greatly affects biofilm formation. “Red” light was identified as the most successful for the algae growth in our system. Consequently, this has been implemented by applying a combination of red and blue light at the production reactors as blue light is still essential for algae growth. Still, at high irradiances with 'optimized' light qualities, algae can still become stressed and more prone to biofilm formation. A preliminary screening of light intensities has already yielded good results; finding optimal values is a task still in progress. The ProviAPT reactors may be operated continuously in time. However, they need to be stopped, cleaned and re-inoculated when fouling becomes too high. Existing cleaning routines have been revised. Whereas initially reactors where operated until a point where biofilm became visually obvious, cleaning is now performed at earlier stages. We experienced that allowing a reactor to become too dirty results in longer and harsher cleaning conditions (detrimental to the reactor itself), worse cleaning results and continued negative effects in following campaigns. Consequently, milder cleaning conditions have been identified that can be applied more often, while still resulting in a decreased overall down time. Despite all the efforts towards biofilm prevention, they are still occasionally formed on the surface of reactors walls. With the transition to indoor production and the implementation of more stringent measures to control entry of foreign microorganisms,

27 less contamination is expected. Nevertheless, special attention is paid to the algal microbiome under these production conditions as seen in this thesis. 3. The algal microbiome: algae-bacteria interactions In the ocean, algae and bacteria have had relationships for millions of years [85,86]. Algae are primary producers in aquatic ecosystems and bacteria grow on organic matter produced by algae and recycle nutrients [87]. Marine bacteria interact with algae through both physical attachment and secretion of small molecules [88]. Algae and bacteria interactions for nutrient exchange and signal transduction are important to shape up the communities and geochemical cycles in the aquatic environment [87]. Mutual interactions of algae and bacteria are species specific as the microenvironments of different algae vary [86]. Algae and bacteria interact in many different ways, ranging from mutualism to parasitism [89]. Mutualism is an interaction in which two or more partners of different species exist in a relationship where all partners benefit from each other [86]. Different mutualistic interactions can be found among algae and bacteria including different types of nutrient exchange [86,90]. Generally, bacteria benefitted from dissolved organic carbon provided by autotrophic microalgae [91] and provide nutrients to the algae e.g. vitamin B12 [92]. Commensalism is a biological relationship in which only one partner benefits while other derives neither benefit nor harm [89]. As an example of commensalism, the microalgae Neochloris oleoabundans and Scenedesmus sp. BA032 were able to benefit from the bacteria Azotobacter vinelandii. More concisely, the algae strains were able to utilize the siderophore azotobactin from the bacterium as a nitrogen source [93]. Furthermore, certain bacteria might be commensal living on the algal sheath for carbon and shelter, while the algae do not benefit from the bacteria [94]. In parasitic interactions, bacteria negatively affect algae and vice versa. Bacteria can be parasitic on algae by lysing the algae cells using enzymes such as glucosidases, chitinases, cellulases and other enzymes and extracting intracellular algal compounds as nutrients [89]. Competition for essential nutrients such as nitrogen and phosphorous between microalgae and bacteria is another parasitic interaction that can result in slower growth rates of the algae [86]. Different microalgae show different parasitic mechanisms against bacteria. A variety of compounds belonging to several chemical classes including indoles, acetogenins, phenols, fatty acids and volatile halogenated hydrocarbons have been attributed as antimicrobial compounds of microalgae [95]. Fatty acids and esters secreted by diatoms act as antibacterial compounds and affect the bacterial community structure [85]. Molecular mechanisms behind algae-bacteria interactions are of a complex nature. Quorum sensing (QS) is a cell-cell communication used by certain bacteria that regulates gene expression in response to fluctuations in cell-population density of

28 bacteria [85,96,97]. QS systems play important roles in regulating many physiological functions of microorganisms including bioluminescence, antibiotic production [98]and biofilm formation [99,100]. QS communication systems rely on small, self-generated signal molecules that are known as autoinducers (AIs), to initiate a coordinated response across a population [101]. There are different categories of AIs such as AHLs, oligopeptide-based signals and furanone‐based system [102]. Among them, AHLs are the best-studied group of signaling molecules and are used by many Gram-negative bacteria. Oligopeptide-based signals are predominantly used by Gram-positive bacteria. Furanone‐based systems (autoinducer‐2, AI‐2) are used by both Gram-negative and Gram-positive bacteria [103]. Due to the differences in quorum sensing between Gram- negative and Gram-positive bacteria, certain signaling molecules that are produced by one group can be antagonistic on the other group [104]. Other QS signals have been described with less attention e.g. quinolones [105]. 4. Objectives and outline of this thesis Increasing attention is paid to industrial microalgae cultivation due to its wide range of potential applications. Algae production may occur in open ponds or closed PBRs typically in non-axenic conditions. Despite operational conditions and reactor conformations are well studied, little is known about the influence of the algal microbiome during industrial cultivation. Modifications on bacterial population may lead to biofouling and biofilm formation in closed PBRs. Reasons why this phenomenon occurs remains unknown. However, it is desired to avoid it. Moreover, optimal microbial flora may lead to higher productivities. In this thesis, we aim to understand the influence of the algal microbiome during commercial production of Nannochloropsis sp. The study of the microbial flora at different conditions may reveal potential growth inducers as well as harmful strains. Since biofilm formation is often an issue, special attention is paid to the study of the mechanisms behind biofilm formation during non- axenic cultivation of Nannochloropsis sp. at Proviron (Belgium). This knowledge may be used to design and implement strategies to improve reactors performance. The synthetic design of an optimal microbial flora may improve growth and prevent contaminants. Furthermore, the understanding of the initiation of biofilm formation and biofouling is essential to design a prevention strategy against biofilm forming agents. In order to achieve this ambitious objective, it was first established a sampling strategy. This means that the first intention was to identify which microorganisms cohabited with the algae at different conditions ranging from high productivity to contaminated reactors. In parallel, bacterial strains need to be isolated and identified. Despite the impossibility of isolating all microorganisms, this step was considered essential as bacterial isolates were needed to proceed with further experimentation. The co- cultivation of bacterial isolates with the algae determines the influence of the bacterium which could range from a growth stimulating strain to a biofilm inducer. Once the effect of bacterial strains is tested, it is possible to decipher the mechanisms behind it and

29 consequently to production interests. Following this approach, the experimental work of this thesis was divided as follows: Chapter 2 – The objective of this chapter is to identify microbial communities cohabiting the ProviAPT reactors at different conditions. A wide set of samples representing reactors with different productivities have been collected and deeply analyzed. Sampling occurred during indoor production of Nannochloropsis sp. on the years 2018 and 2019. A full characterization of the reactors where samples were taken is provided and information regarding total biomass, total fatty acids, fatty acids profile was estimated. Samples were divided based on performance of the reactors. The microbial population of these samples was analyzed based on next-generation DNA sequencing techniques. Reactor productivities and EPA content were used as an estimation of algae health and thus correlated to the study of their microbial flora. These correlations provide potential beneficial algae-bacteria interactions towards improving reactor performance. Chapter 3 – Algae-bacteria interactions may be studied by co-cultivation of algae with isolated bacterial strains. In this chapter, algae broth and reactor biofilms were sampled from reactors with high and low productivity. The low productivity reactors were discarded for production due to severe contamination. It was decided to use a severe contaminated reactor as a wide range of different biofilms could be found minimizing thus the number of required reactors. Microorganisms were isolated by cultivation in agar plates. Identification of microorganisms was performed by Maldi-TOF and Sanger DNA sequencing. The influence on biofilm formation of every identified microorganism was tested on a biofilm formation assay. Chapter 4 – In the previous chapter, Poseidonocella sp. was identified as a strong biofilm inducer of Nannochloropsis sp. In this chapter, the molecular mechanisms behind the biofilm forming process are studied. The biofilm forming system Nannochloropsis-Poseidonocella is suggested as a model system of biofilms formed in the ProviAPT reactors. To design a biofilm prevention strategy, early stage biofilms must be targeted. Therefore, the bacterial exometabolome was explored in order to screen for signal molecules that initiate the biofilm formation process. Chapter 5 – Algae-bacteria interactions may induce different effects on algae. In this chapter, synthetic microbial communities are explored in order to find potential algae growth enhancer and biofilm preventing candidates. Well-studied bacterial strains associated to the diatom Seminavis robusta were co-cultivated with Nannochloropsis sp. Responses on growth rate and biofilm formation were studied and metabolic changes were analyzed. Chapter 6 – In this chapter, a general discussion of the results is provided. Moreover, different approaches towards the prevention of the effect of harmful microorganisms are discussed. While mechanic approaches on reactor design and operational conditions

30 have been already taken, biological approaches require longer experimental and implementation time. Suggestions on how to proceed after this thesis are provided as well as overall conclusions.

Chapter II

Exploration of the algal microbiome during industrial cultivation of Nannochloropsis sp. in closed photobioreactors and its potential effect on lipids composition

Javier B. Giraldo1, 2, Willem Stock2, Jorien Fret1, Anne Willems3, Sven Mangelinckx4, Luc Roef1, Wim Vyverman2 and Mark Michiels1

1 Proviron Holding NV, Georges Gilliotstraat 60, B- 2620 Hemiksem, Belgium

2 Laboratory of Protistology and Aquatic Ecology, Department of Biology, Ghent University, Krijgslaan 281-S8, B-9000 Ghent, Belgium

3 Laboratory of Microbiology, Department of Biochemistry and Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

4 SynBioC, Department of Green Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium

Abstract The microbial population of two sets of reactors commercially growing Nannochloropsis sp. was compared. Each set of reactors was characterized and classified under regular performance (MPC14) and irregular performance (MPC16). MPC14 reported a better performance with higher dried biomass content and higher total lipid and EPA content. In none of the samples, eukaryotic microorganisms were detected based on microscopic observations and DGGE. The Gammaproteobacterium Marinicella pacifica was highly abundant in samples from MPC14. Initial samples from MPC16 had low alphadiversity while this was higher in the later ones. The increase in biodiversity was observed after a heat peak. Alphaproteobacteria increased over time in MPC16. The increase in the relative abundance of two alphaprotebacterial strains coincided with a decrease in EPA concentration. These strains were assigned to the genus Thalassospira and Roseovarius. Species from both genera have been described with detrimental effect on algae cultures. The same strain of R. mucosus was found and isolated from production reactors two years before this sampling suggesting a persistent association with Nannochloropsis sp.

1. Introduction During the last years, the interest in industrial cultivation of the oleaginous eustigmatophyte Nannochloropsis has increased due to its high content in poly unsaturated fatty acids (PUFAs) and carotenoids [68,106]. Species from this genus are small unicellular cells, found in both marine environments and freshwater [73,107]. This microalga is characterized by spherical or slightly ovoid cells with a diameter of 2–5 μm [108,109] which remains constant during stress conditions [110]. Their high growth rates and valuable fatty acid composition [68] make them attractive species for industrial growth in photobioreactors (PBR) where critical operating variables can be controlled [111]. They can be used in applications ranging from biofuel production to food and feed products [112–114]. Nannochloropsis spp. can accumulate large amounts of lipids, varying with species and cultivation conditions [115,116]. The accumulation of PUFAs increases their commercial value [117], particularly due to high levels of eicosapentaenoic acid (EPA) [118]. PUFAs have shown great benefits for human health [119–121]. PUFAs are typically obtained from fish oil [122]. The increasing demand for PUFAs urges to find alternative sustainable production methods [123]. Algae have shown potential as an alternative PUFA source [124,125]. Algal monocultures have been the preferred production route [126], with the notion that industrial microalgae production is typically conducted under xenic conditions. Despite the initial lack of knowledge on the influence of the algal microbiome during growth, the interest on deciphering algae-bacteria interactions has risen over the past years [127,128]. Although microbial population of Nannochloropsis sp. varies with cell phase and environmental factors [129,130], little is known about its implications for productivity. In this chapter, samples coming from different algal reactors at different times are compared in terms of yield, PUFAs and accompanying microbial communities to extract conclusions towards the understanding of the dynamics of the microbiome during industrial cultivation of Nannochloropsis. Samples from high performance and lower performance reactors were compared. Furthermore, in the reactors with lower performance samples were taken over time in order to correlate modifications on the algal microbiome to reactor performance. 2. Materials and methods

2.1. Algae cultivation and sample selection Nannochloropsis sp. CCAP211/78 – obtained from the Culture Collection of Algae and Protozoa (CCAP, United Kingdom) – was cultivated indoor in closed PBRs, the so-called ProviAPTs, during industrial operation of the microalgae producer company Proviron NV (Belgium). The ProviAPT reactor is a closed system consisting of an array of 35

35 interconnected vertical flat panels. Each reactor has an effective surface of 7 m2 [131] with an upgraded operational volume of 200 L. Every reactor panel is irradiated with LED lights and light intensity can be controlled. During industrial production they are clustered in groups of up to 16 reactors, controlled by a central unit called Micro ProviAPT Controller (MPC). Reactors are operated in semi-continuous mode by harvesting approximately one third of its volume per day and feeding the same amount of fresh medium. Single reactors can be individually manipulated: they can be connected or disconnected, thus varying the total MPC active volume/effective surface. Every reactor is identified by a code consisting of the MPC number followed by their position in the header (H1 or H2) and reactor position in the vertical rack (R1 to R8). Algae cultures were grown in saline media enriched with inorganic nutrients [131] adapted from f/2 [132]. 2.2. Sample selection and characterization Samples were taken from two reactor units, respectively a reactor unit with an apparently low bacterial load which was considered productive (‘MPC14’) and a reactor unit that presented increasing evidence of biofouling and biofilm formation on the reactor walls (‘MPC16’). Three reactors from MPC14 were sampled on the 21st of January, 2019. A total of 15 samples were taken from 3 different reactors belonging to MPC16. Samples were taken every 2-3 days from June 25th to July 18th of 2018. At that time, microscopic observations revealed higher bacterial presence than usual when compared to daily observations and algae in MPC16 were suspected to grow suboptimal.

During the sampling periods information regarding CO2 consumption, active volume/effective reactor surface, temperature and light irradiation was collected. These samples were characterized microscopically and by the dry weight estimation, FAME analysis and microbial population analysis (see below).

Overall CO2 consumptions of every MPC were calculated by monitoring the difference of CO2 amounts entering and leaving the reactor unit. The CO2 consumption is the value used for estimations of biomass production. In this chapter, these estimations are not considered and only CO2 consumption values are reported. Light irradiation in the photosynthetically active radiation (PAR) region (400-700 nm) was measured using the Lighting Passport Pro (Asensetek) and Spectrum Genius Agricultural Lighting software at different spots on the reactor panel surface. Irradiation was averaged over the surface -2 -1 of the reactor panels and expressed as µmole ∙ m ∙ s . The estimation of the CO2 -1 consumption on light (gCO2 · molePhotons ) is used to determine the yield on light of the MPC. The temperature of the water used for the light cooling system was measured. Measurement points are closely located to the reactor surface. Temperature measurements at different spots along the reactor surface represented a good

36 estimation of the average temperature over the reactor surface. Registered maximum temperatures of the cooling water are reported. Each algal sample was examined microscopically as a first quality control and its dry weight (DW) was calculated and expressed as gram of biomass per unit of volume. Values for dry weights were obtained by measuring the weight of pelleted algae biomass of a known sample volume after 24h of drying at 100 °C. Dry pellets contained also salts from the medium. Therefore, salinity was measured and final estimations were corrected as indicated in Equation 1. Please note that in the formula the weight of each pellet includes the weight of the vial and that is why it needs to be subtracted.

Equation 1 Estimation of dry weight (DW) in g/L

(퐷푟푦 푝푒푙푙푒푡 (푔) − 퐷푟푦 푣푖푎푙(푔)) − (푊푒푡 푝푒푙푙푒푡(푔) − 퐷푟푦 푝푒푙푙푒푡(푔)) ∗ 푆푎푙푖푛푖푡푦 (‰) 퐷푊(푔⁄퐿) = 푉표푙푢푚푒 (퐿) 2.3. Fatty acids profiling Fatty acid methyl esters (FAME) were prepared from freeze-dried samples (5 mg) using a direct transesterification procedure with 2.5% (v:v) sulfuric acid in methanol as described by De Troch et al. (2012) [133]. The internal standard (C19:0, 15 µg) was added prior to the transesterification procedure. Fatty acid methyl esters were extracted twice with hexane. Composition analysis of fatty acids was carried out using a gas chromatograph (HP 7890B, Agilent Technologies, Diegem, Belgium) equipped with a flame ionization detector (FID) and connected to an Agilent 5977A Mass Selective Detector (MS; Agilent Technologies, Diegem, Belgium). The GC was further equipped with a PTV injector (CIS- 4, Gerstel, Mülheim an der Ruhr, Germany). A 60m×0.25mm×0.20μm film thickness HP88 fused-silica capillary column (Agilent Technologies, Diegem, Belgium) was used for the gas chromatographic analysis, at a constant Helium flow rate (2 mL/min). The injected sample is split equally between the MS and FID detectors at the end of the GC column using an Agilent capillary flow technology splitter. The oven temperature program was as follows: at the time of sample injection the column temperature was 50°C for 2 min, then gradually increased at 10°C/min to 150°C, followed by a second increase at 2°C/min to 230°C. The injection volume was 1 μL. The injector temperature was held at 30°C for 0.1 min and then ramped at 10°C/s to 250°C and held for 10 min. The transfer line for the column was maintained at 250°C. The quadrupole and ion source temperatures were 150 and 230°C, respectively. Mass spectra were recorded at 70 eV ionization voltage over the mass range of 50 - 550 m/z units. Data analysis was done with Agilent MassHunter Quantitative Analysis software (Agilent Technologies, Diegem, Belgium). The signal obtained with the FID detector was used to generate quantitative data of all the compounds. Peaks were identiﬁed based on their retention times, compared with external standards as a reference (Supelco 37

Component FAME Mix, Sigma-Aldrich, Overijse, Belgium) and by the mass spectra obtained with the Mass Selective Detector. Quantiﬁcation of fatty acid methyl esters was based on the area of the internal standard (19:0) and on the conversion of peak areas to the weight of the fatty acid by a theoretical response factor for each fatty acid [134,135]. 2.4. Microbial characterization Microbial composition was assessed using amplicon sequencing. The DNA extraction was performed following an adapted phenol chloroform extraction for marine samples [136]. The methodology described by Morrissey et al. (2019) was applied for the bacterial DNA isolation, amplification and sequencing [137]. As algae cultures had a high cell density and bacteria were expected to be in low amount, high interferences from chloroplasts were expected in bacterial DNA amplification. Therefore, primers were selected in order to reduce chloroplast amplification [137]. Amplicons were sequenced on an Illumina MiSeq platform giving 2 × 300 bp paired-end reads. The Illumina sequencing reads were processed as described by Tytgat et al. (2016) [138]. In order to verify the presence of other eukaryotic microorganisms, the eukaryotic community was characterized through the amplification of the 18S rDNA region followed by denaturing gradient gel electrophoresis (DGGE) [139], using the primers as in van Hannen et al. [140] for PCR amplification. The PCR products were loaded on an agarose electrophoresis gel and followed a DGGE protocol from van Hannen et al. (1998) [140]. Relevant visible bands were excised, amplified and sequenced on a Sanger platform. 2.5. Statistical analysis The OTU table, obtained from the 16S amplicon sequencing was processed in R (version 3.4.1). After removal of reads assigned to mitochondria or chloroplasts, samples with less than a thousand reads were discarded. The remaining samples were rarefied to 1591 reads. The diversity within samples was quantified by calculating the Shannon and Simpson diversity indices and as the number of observed OTUs. A PERMANOVA (adonis2, vegan 2.4; 999 permutations) was used to assess the compositional differences between ‘high’ and ‘low’ productivity samples. The discriminating OTUs between the two groups were identified with a SIMPER (vegan 2.4) analysis. Similarly, differences in fatty acid composition between ‘high’ and ‘low’ productivity samples were tested with a PERMANOVA (999 permutations) and the discriminating FAs were identified with a SIMPER analysis. A mantel test was performed on the Bray–Curtis based dissimilarity matrices of the rarefied OTUs and FA data to test for a multivariate correlation between the 16S rDNA amplicon data and FA profiles. Both the significance of the Pearson and Spearman based multivariate correlations were assessed. The relation between OTUs and EPA

38 concentrations in MPC16 samples was calculated with the Spearman correlation. Only the significant (uncorrected p-value≤0.05) correlations were retained. 3. Results and discussion

3.1. Reactor characterization and sample selection Figure 2 and Figure 3 show the performance of MPC14 and MPC16 respectively. In MPC14, the -2 daily light irradiation was slightly above 20 mole Photons · m . However, this decreased between days 15 and 25. This is explained due to the addition of new reactors to the MPC. When a new reactor is added, the light irradiation applied is lower and gradually increases as long as cell density increases too. The sampling of the MPC14 was performed on new added reactors that were able to maintain the growth in time. Information regarding the total MPC volume, number of connected reactors and daily harvest and feed of MPC14 and MPC16 can be found in Supplementary Figure 1 and Supplementary Figure 2 respectively. The maximum registered temperature decreased when light irradiation cropped but typically oscillated around 25 °C. The daily CO2 consumption and the CO2 consumption on light followed a similar trend. A peak was registered between days 9 and 18. This is explained due to a decrease in the number of connected reactors (Supplementary Figure 1). Total CO2 consumption values are accumulated.

This means that when reactors are disconnected, the accumulated total CO2 – which is influenced by previous day – is divided by a smaller surface reporting a larger value and thus failing to report an accurate CO2 consumption. During the light decrease between days 15 and

25, the daily CO2 consumption dropped. However, the CO2 consumption on light remained high and shortly stabilized afterwards indicating a good health of the algae during this time. Reported data between days 30 and 60 indicate a stable performance over a prolonged time.

In MPC16, the daily light irradiation was higher than in MPC14. MPC14 corresponded to a later period of the year when observations showed that similar yields could be achieved with lower energy input. Sampling of MPC16 was performed in the summer which in combination with a higher light irradiation caused higher temperature values. Daily CO2 consumption was higher than in MPC14. However, the daily CO2 consumption on light was considerably lower. Higher CO2 consumption does not correlate to an increase of algal biomass as reported dry weight values were higher in MPC14 (Table 1) than in MPC16 (Figure 4). Furthermore, a heat wave during that period increased the room temperature and during days 36-38 after inoculation, it was decided to decrease light irradiation to avoid extra heat coming from the lamps. This issue had consequences for temperature – when sampling MPC14 was slightly above to 20 °C while in MPC16 was closed to 30 °C – affecting thus reactor productivity and dry biomass values. Moreover, microscopic observations of daily samples reported that bacterial presence was higher in MPC16.

Figure 2 Daily PAR light irradiation, maximum reactor temperature, daily CO2 consumption and daily CO2 consumption based on light of MPC14. Number in figure indicates day of sampling.

Figure 3 Daily PAR light irradiation, maximum reactor temperature, daily CO2 consumption and daily CO2 consumption based on light of MPC16. Numbers in figure indicate days of sampling.

Table 1 Dry weight on samples from MPC 14 taken on day 23 after inoculation.

Reactor Dry weight

(g/L)

14H2R2 2.55

14H2R5 2.03

14H2R7 1.94

Figure 4 Dry weight on samples from MPC16. Black square indicates the sample selection for FA analysis and DNA sequencing.

3.2. Total fatty acids content and fatty acids profile The fatty acid profile was markedly different in both MPCs as shown by principal component analysis (Figure 5). Two fatty acids in particular contributed to this variation: 20:5n-3 (eicosapentaenoic acid (EPA)) and cis9-16:1 (palmitoleic acid). PCA did not reveal a clear temporal pattern on MPC16. In MPC14, the total FAME content was higher (Figure 6). This was mostly due to a higher amount of EPA (20:5n-3) in MPC14 (Figure 7). Palmitoleic (cis9-16:1) acid was also lower in MPC16 but to a lesser extent. While in MPC14 the most abundant fatty acid was EPA, on MPC16 this was palmitic (16:0) acid.

Figure 5 Principal component analyses on the fatty acids profile data. In grey color, the fatty acids which contributed to main differences among samples are represented. Color and shapes of the dots correspond to reactor origin and are followed with the sampling date.

140

120

100 Biomass

80 /g

60 Total FAME Total

40 mg

0 MPC14 MPC16

Figure 6 Total FAME content per gram of biomass. Values plotted are average of samples. Error bars represent standard deviation.

Biomass 30

/g MPC14 X 20

mg MPC16 10 0

Figure 7 Fatty acids profile of MPC14 and MPC16. Values plotted are average of samples. Error bars represent standard deviation.

Nannochloropsis sp. is considered an oleaginous algae species due to its high lipid content, which is estimated to oscillate between 10-60% of dry biomass depending on cultivation conditions, cell phase and the lipid extraction methods used for quantification [141–143]. Cell stress due to nitrogen starvation is known to be an efficient technique for increasing cell lipid content [75,144]. Both MPCs were operated with excess of nutrients to ensure fast algae growth. Therefore, no accumulation of lipids was expected and FAME values around 10% of dry biomass as observed in this study are normal. Total lipids values may be higher as other lipids fractions were not quantified. On MPC16, lower total FAME value could be related with the lower performance. The FAME profile of Nannochloropsis is known to vary with strain identity and cultivation conditions [68,142]. Our data confirm that the FAME profile of Nannochloropsis sp. CCAP211/78 found in this study is in line with published data [110]. In MPC16 the EPA concentration was only about half as high as in MPC14. Both reactor units were inoculated with the same original commercial strain. Differences in cultivation conditions may influence EPA levels [145], being typically higher with high levels of carbon dioxide and nitrate as well as lower salinities and temperatures [71,146,147]. In addition, higher light irradiations and temperatures are correlated with a decrease in PUFAs and a relative increase in 16:0 and 16:1 [148,149]. Therefore, we suggest that the lower EPA content of MPC16 might be explained by higher temperatures while high light irradiance may have contributed to a relative decrease of palmitoleic and an increase of EPA in plastid galactolipids [150]. 3.3. Microbial diversity The presence of relatively abundant eukaryotic organisms other than Nannochloropsis sp. was tested with DGGE (Supplementary Figure 3). Electrophoresis gels reveal no other eukaryotic organisms were present. This is in line with microscopic observations

44 on selected samples where no eukaryotic organisms besides Nannochloropsis sp. were detected. For the study of the prokaryotic DNA, the reads of single OTUs were expressed as fraction of total reads per sample (Supplementary Table 1 and 2). Average reads per sample per MPC were calculated and summary of number of reads and OTUs may be found in Table 2. In contrast to Supplementary Table 1 and 2 the results from Table 2 exclude samples 14H2R2, 16H1R1-29 and 16H2R2-20 as the number of reads of these samples were unexpectedly low. The fraction of bacterial Illumina reads was higher in MPC14, suggesting higher bacterial abundance in these samples. A total of 154 OTUs were found in MPC14 where one of them was mitochondria with 49.85% of the total reads. Only 10 out of the other 153 OTUs had a relative abundance higher than 1% of bacterial reads, while 58.13% of all bacterial reads belonged to one OTU (OTU3, Supplementary Table 1). MPC16 reported 212 OTUs where mitochondrial fraction was one of them with 82.03% of the reads (Supplementary Table 2). Only 3 OTUs were present with a relative abundance higher than 1% of bacterial reads. From all non- mitochondrial reads, 80.46% of the reads belonged to one OTU (OTU2). Identification of OTUs was conducted using EZBioCloud [151]. A search on the database matched OTU3 with the type strain of Marinicella pacifica [152] and OTU2 with a cryomorphaceae strain (Table 3). Relative abundances reported on Table 3 are estimated from the total bacterial fraction excluding thus the mitochondrial fraction.

Table 2 Number of reads and OTUs after processing DNA sequences.

MPC14 MPC16

Average reads per 48893 81434 sample

Non-mitochondrial 50.15% 17.97% fraction

Total OTUs 154 212

Table 3 Search on EZBioCloud [50] of the most relevant strains based on relative bacterial abundance per MPC. Percentages of MPC are estimated from the non-mitochondrial fraction.

OTU % in % in Match Strain Identity Max. Total Query E- MPC14 MPC16 designation Score Score cover value

OTU3 58.13% 9.12% Marinicella sw153 100.00% 682 682 100% 0 pacifica

OTU2 5.94% 80.46% Unclassified DG890 99.18% 658 658 100% 0

PCA revealed important differences between MPCs (Figure 8). Both samples from MPC14 clustered in the right side while the distribution of MPC16 samples is more heterogeneous. Mostly two OTUs contributed to this variation. A higher relative abundance of OTU3 differentiates samples from MPC14. Samples from MPC16 differentiate due to a lower relative abundance of OTU3 and higher presence of OTU2.

Figure 8 Principal component analyses on the rarefied DNA sequences data. In grey color, OTUs which contributed to main differences among samples are represented. Color and shapes of the dots correspond to reactor origin and are followed with the sampling date.

The alpha diversity in samples was studied through the Shannon and Simpson indices and the number of rarefied OTUs (Figure 9). While samples from MPC14 presented similar biodiversity, samples from MPC16 had differences in biodiversity over time. Earlier samples from MPC16 presented a lower biodiversity which explains the observed

46 differences on the PCA. Comparing three methods, MPC16 samples can be classified in three groups based on their biodiversity. Samples taken 22 and 25 days after inoculation had a low biodiversity while samples from days 34 and 39 presented higher biodiversity. Diversity in the samples from day 29 was variable which is in line with the earlier described hypothesis. Therefore, it is suggested that the increase in temperature triggered an increase in microbial diversity. It is known that an increase of temperature may lead to an increase on microbial abundance [153]. Higher temperature generally contributes to biofilm formation [154,155] since it stimulates fouling inducers species [155,156]. Moreover, an increase of temperature and light irradiation – as it occurs on MPC16 compared to MPC14– may lead to damage to photosystem II in Nannochloropsis sp [157]. Therefore, it is expected that an increase of temperature may lead to biofouling and a decrease on productivity as it is found here.

Figure 9 Biodiversity indices and rarefied OTUs of reactor samples. Codes correspond to reactor origin followed by the sampling date.

Gammaproteobacteria were dominant in MPC14 and Flavobacteria were the most abundant class on MPC16 (Figure 10). However, the relative abundance of Alphaproteobacteria was increasing over time on MPC16. Gammaproteobacteria were dominated by Marinicella pacifica (OTU3) and Flavobacteriia by the bacterium DG890 (OTU2). Alphaproteobacteria were represented by multiple OTUs with low relative

48 abundance. Alphaproteobacteria were more abundant 34 and 39 days after inoculation which is in line with previous observations about biodiversity. The genus Leptospira was only detected on samples from day 39 in very low relative abundance.

Figure 10 Class classification of bacterial DNA sequences of reactor samples. Samples code corresponds to reactor code followed by number of days after inoculation.

The observation that Alphaproteobacteria, Gammaproteobacteria and Flavobacteria were the most dominant classes is in line with previous observations on Nannochloropsis sp. cultures [129,130]. In these studies, it is suggested that Alphaproteobacteria and Flavobacteria are dominant in stationary phase while Gammaproteobacteria are dominant during the exponential phase. A lower performance of MPC16 could be related with a late stationary phase while MPC14 could have had its cells growing exponentially. In contrast to previous studies reporting a dominance of the order Alteromonadales in Gammaproteobacteria [130], we found the genus Marinicella (order Oceanospirillales) to be the dominant member of the Gammaproteobacteria.

3.4. Correlations between microbial diversity and EPA content Correlations between microbial diversity and EPA content were estimated by Spearman correlations (Table 4). By these correlation coefficients, it was possible to identify the OTUs that increased when EPA concentration decreased. Considering the low relative abundance of all OTUs and errors derived from the analytical method, it was decided to focus only on OTUs that have an average relative abundance higher than 0.05 % and

49 match with > 99.50 % identity with a result after a search on EZBioCloud [151]. This was an arbitrary decision to discard errors.

Table 4 Putative OTUs decreasing EPA concentration based on Spearman correlation coefficient.

OTU Correlation P value Average relative Identification (Identity match) coefficient abundance

OTU105 -0.55874 0.047158 0.00%

OTU130 -0.73193 0.004452 0.00%

OTU173 -0.58554 0.035505 0.00%

OTU18 -0.64539 0.017205 0.19% Thalassospira sp. (100%)

OTU218 -0.5691 0.042368 0.01% Uncultured bacterium (90.35%)

OTU26 -0.57736 0.038811 0.58% Roseovarius sp. (100%)

OTU323 -0.64279 0.017808 0.00%

OTU359 -0.65465 0.015179 0.15% Uncultured bacterium (93.26%)

OTU51 -0.5736 0.040404 0.06% Unclutured bacteroidetes (99.46%)

OTU66 -0.60344 0.028995 0.00%

OTU18 had an average relative abundance of 0.19%. The sample that contained the least EPA was taken from 16H1R2-39 (18.65 mg/g) and its relative abundance of this OTU was 1.19%. OTU26 had an average relative abundance of 0.58%. T. xiamenensis M-5T was first isolated and identified from the surface water of a waste-oil pool [158] and its genome has been sequenced [159]. A species from the Thalassospira genus reported algicidal effect against dinoflagellate Karenia mikimotoi [160]. Its killing mechanism is based on the secretion of benzoic acid which disrupts the algae cell walls. Roseovarius mucosus DSM17069 is an alphaproteobacteria member of the Roseobacter clade characterized for presenting traces of bacteriochlorophyll a [161]. It has been described as a non-forming biofilm strain associated with microalgae and/or solid surfaces [162]. The same strain of Roseovarius mucosus was found in reactors during production campaign of 2016 [163]. Its ability to induce biofilm formation on Nannochloropsis sp. was evaluated. Although average values were higher than controls suggesting thus ability of inducing biofilm formation, differences were not statistically significant. In Roseovarius, the ability for forming biofilms is strain dependent [162,164]. Moreover, some Roseovarius strains may possess antimicrobial activity [165]

50 e.g. Roseovarius strain TM1035 and Roseovarius sp. strain TM1042; ethyl acetate extracts that contained AHLs produced by these bacteria inhibited the growth of Vibrio anguillarum [165]. Considering the established criteria, the presence of OTU18 and OTU26 is correlated to a decrease in EPA (Figure 11). Therefore, the interaction of these bacterial strains with the algae could have a negative impact on the EPA levels. It is suggested that the increase of temperature in MPC16 led to an increase in certain alphaproteobacterial strains that diminish growth and EPA concentration of Nannochloropsis sp. In any case, both the algae performance and EPA concentration was suboptimal in MPC16. Further experimental work is needed to confirm the negative effect of these bacteria. Multiple experimental approaches can be followed for this purpose but more reliable ones need isolation of the bacterial strain and co-cultivation with the algae.

Figure 11 EPA concentration (left axis) and relative abundance of OTU26 and OTU18 (right axis) on selected samples. 4. Acknowledgements This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642575. This work makes use of resources and facilities provided by UGent as part of the Belgian contribution to EMBRC-ERIC.

5. Supplementary material

Supplementary Figure 1 Feed, harvest and total volume and number of connected reactors of MPC14

Supplementary Figure 2 Feed, harvest and total volume and number of connected reactors of MPC16

Supplementary Figure 3 Electrophoresis gels from DGGE showing two relevant bands. Position on gel from left to right: (left gel) Marker, 14-H2R2-0121, 14-H2R5-0121, 14-H2R7-0121, Empty, 16-H1R2-0629, 16-H1R2-0702, 16-H1R2-0706, 16-H1R2- 0711, 16-H1R2-0716, 16-H2R2-0629, 16-H2R2-0711, 16-H2R7-0706, Marker, (right gel) Marker, 14-H2R2-0121, 16-H1R2- 0711, 16-H2R2-0702, 16-H2R2-0706, Empty, 16-H2R2-0711, 16-H2R2-0716, 16-H2R7-0629, 16-H2R7-0702, 16-H2R7-0706, 16-H2R7-0711, 16-H2R7-0716, Marker

Supplementary Table 1 Number of reads per sample and proportion of OTUs in MPC14 samples. Samples with all relative abundances lower than 0.005% were rounded to 0.00% and are not shown in this table.

14H2R2 14H2R5 14H2R7 Average

Reads 51 35092 62694 32612.33 per sample

OTU1 78.43% 64.10% 41.88% 49.87%

OTU3 1.96% 16.43% 36.27% 29.13%

OTU2 5.88% 3.39% 5.24% 4.58%

OTU13 1.96% 10.10% 0.00% 3.62%

OTU24 0.00% 0.82% 2.31% 1.77%

OTU29 0.00% 0.15% 2.55% 1.69%

OTU32 0.00% 0.53% 1.94% 1.43%

OTU9 0.00% 0.51% 1.19% 0.95%

OTU37 0.00% 0.24% 1.26% 0.89%

OTU83 0.00% 0.12% 0.94% 0.65%

OTU58 0.00% 0.05% 0.96% 0.64%

OTU43 0.00% 0.27% 0.53% 0.43%

OTU26 0.00% 0.31% 0.50% 0.43%

OTU45 0.00% 0.42% 0.23% 0.30%

OTU18 0.00% 0.17% 0.33% 0.27%

OTU4 1.96% 0.23% 0.23% 0.23%

OTU46 0.00% 0.15% 0.27% 0.23%

OTU77 0.00% 0.18% 0.21% 0.20%

OTU87 0.00% 0.15% 0.22% 0.20%

OTU71 0.00% 0.06% 0.27% 0.19%

OTU28 0.00% 0.05% 0.22% 0.16%

OTU93 0.00% 0.01% 0.17% 0.11%

OTU303 0.00% 0.04% 0.14% 0.10%

OTU30 0.00% 0.07% 0.09% 0.08%

OTU100 0.00% 0.11% 0.06% 0.08%

OTU6 0.00% 0.07% 0.08% 0.08%

OTU15 0.00% 0.04% 0.08% 0.06%

OTU325 0.00% 0.06% 0.06% 0.06%

OTU95 0.00% 0.07% 0.06% 0.06%

OTU110 0.00% 0.01% 0.09% 0.06%

OTU140 0.00% 0.03% 0.07% 0.06%

OTU351 0.00% 0.08% 0.05% 0.06%

OTU359 0.00% 0.03% 0.07% 0.06%

OTU297 0.00% 0.01% 0.07% 0.05%

OTU283 0.00% 0.03% 0.06% 0.05%

OTU190 0.00% 0.04% 0.05% 0.05%

OTU14 1.96% 0.01% 0.07% 0.05%

OTU63 0.00% 0.04% 0.05% 0.05%

OTU198 0.00% 0.03% 0.05% 0.04%

OTU334 0.00% 0.04% 0.04% 0.04%

OTU47 0.00% 0.01% 0.05% 0.04%

OTU51 0.00% 0.05% 0.03% 0.04%

OTU8 0.00% 0.04% 0.03% 0.03%

OTU11 0.00% 0.03% 0.03% 0.03%

OTU50 0.00% 0.03% 0.03% 0.03%

OTU106 0.00% 0.05% 0.02% 0.03%

OTU307 0.00% 0.03% 0.03% 0.03%

OTU132 0.00% 0.00% 0.04% 0.03%

OTU182 0.00% 0.04% 0.02% 0.03%

OTU166 0.00% 0.00% 0.04% 0.03%

OTU238 0.00% 0.02% 0.03% 0.02%

OTU97 0.00% 0.02% 0.03% 0.02%

OTU111 0.00% 0.01% 0.03% 0.02%

OTU131 0.00% 0.01% 0.03% 0.02%

OTU134 0.00% 0.02% 0.02% 0.02%

OTU114 0.00% 0.01% 0.02% 0.02%

OTU139 0.00% 0.01% 0.02% 0.02%

OTU155 0.00% 0.03% 0.01% 0.02%

OTU315 0.00% 0.01% 0.02% 0.02%

OTU133 0.00% 0.02% 0.01% 0.02%

OTU187 0.00% 0.01% 0.01% 0.01%

OTU336 0.00% 0.01% 0.01% 0.01%

OTU20 0.00% 0.01% 0.01% 0.01%

OTU162 0.00% 0.01% 0.02% 0.01%

OTU135 0.00% 0.00% 0.02% 0.01%

OTU352 0.00% 0.01% 0.01% 0.01%

OTU108 0.00% 0.02% 0.01% 0.01%

OTU257 0.00% 0.03% 0.00% 0.01%

OTU364 0.00% 0.00% 0.02% 0.01%

OTU224 0.00% 0.00% 0.02% 0.01%

OTU239 0.00% 0.00% 0.01% 0.01%

OTU22 0.00% 0.01% 0.01% 0.01%

OTU34 0.00% 0.01% 0.01% 0.01%

OTU252 0.00% 0.00% 0.01% 0.01%

OTU267 0.00% 0.00% 0.01% 0.01%

OTU117 0.00% 0.01% 0.01% 0.01%

OTU144 0.00% 0.00% 0.01% 0.01%

OTU211 0.00% 0.00% 0.01% 0.01%

OTU275 0.00% 0.01% 0.01% 0.01%

OTU16 0.00% 0.01% 0.01% 0.01%

OTU184 0.00% 0.00% 0.01% 0.01%

OTU231 0.00% 0.00% 0.01% 0.01%

OTU183 0.00% 0.00% 0.01% 0.01%

OTU218 0.00% 0.01% 0.01% 0.01%

OTU242 0.00% 0.01% 0.01% 0.01%

OTU152 0.00% 0.00% 0.01% 0.01%

OTU165 0.00% 0.00% 0.01% 0.01%

OTU337 0.00% 0.00% 0.01% 0.01%

OTU41 0.00% 0.00% 0.01% 0.01%

OTU254 0.00% 0.00% 0.01% 0.01%

OTU36 0.00% 0.01% 0.01% 0.01%

OTU229 0.00% 0.01% 0.01% 0.01%

OTU7 0.00% 0.01% 0.00% 0.01%

OTU69 5.88% 0.00% 0.00% 0.01%

OTU181 0.00% 0.00% 0.01% 0.01%

OTU256 0.00% 0.00% 0.01% 0.01%

OTU249 0.00% 0.00% 0.01% 0.01%

OTU130 0.00% 0.01% 0.00% 0.01%

OTU292 0.00% 0.01% 0.00% 0.01%

OTU129 0.00% 0.01% 0.00% 0.01%

OTU356 0.00% 0.01% 0.00% 0.01%

OTU105 0.00% 0.00% 0.01% 0.00%

OTU138 0.00% 0.00% 0.01% 0.00%

OTU219 0.00% 0.00% 0.01% 0.00%

OTU84 0.00% 0.00% 0.01% 0.00%

OTU278 0.00% 0.01% 0.00% 0.00%

OTU136 0.00% 0.01% 0.00% 0.00%

OTU124 0.00% 0.01% 0.00% 0.00%

OTU128 0.00% 0.01% 0.00% 0.00%

OTU5 1.96% 0.00% 0.00% 0.00%

Supplementary Table 2 Number of reads per sample and proportion of OTUs in MPC16 samples. Samples with all relative abundances lower than 0.005% were rounded to 0.00% and are not shown in this table.

OTU_ID 16H1R 16H1R 16H1R 16H1R 16H1R 16H2R 16H2R 16H2R 16H2R 16H2R 16H2R 16H2R 16H2R 16H2R 16H2R Avera 2-20 2-25 2-29 2-34 2-39 2-20 2-25 2-29 2-34 2-39 7-20 7-25 7-29 7-34 7-39 ge

Reads 96486 10813 39 80494 24021 48 13152 68377 56801 84663 13537 67841 39940 10259 62386 70582 per 5 8 7 5 .1 sample

OTU1 82.60 62.96 43.59 83.02 80.77 77.08 73.33 84.40 87.70 97.01 84.99 75.59 95.94 89.56 83.00 82.03 % % % % % % % % % % % % % % % %

OTU2 15.16 34.63 2.56% 11.44 2.53% 6.25% 24.58 13.50 1.88% 0.60% 13.63 23.33 1.79% 6.12% 10.80 14.46 % % % % % % % % %

OTU3 0.05% 0.15% 0.00% 1.38% 8.58% 2.08% 0.50% 0.73% 6.72% 0.53% 0.08% 0.13% 0.94% 1.49% 0.65% 1.07%

OTU21 0.77% 0.28% 7.69% 0.24% 0.17% 0.00% 0.16% 0.19% 0.03% 0.06% 0.24% 0.10% 0.04% 0.04% 0.04% 0.21%

OTU30 0.39% 0.41% 0.00% 0.11% 0.04% 0.00% 0.28% 0.12% 0.05% 0.02% 0.18% 0.11% 0.03% 0.07% 0.13% 0.18%

OTU13 0.05% 0.05% 2.56% 0.24% 0.25% 0.00% 0.07% 0.14% 0.95% 0.17% 0.12% 0.02% 0.35% 0.17% 0.18% 0.17%

OTU8 0.06% 0.08% 0.00% 0.73% 0.13% 0.00% 0.03% 0.04% 0.18% 0.04% 0.02% 0.03% 0.06% 0.14% 0.24% 0.13%

OTU36 0.03% 0.01% 0.00% 0.24% 0.94% 0.00% 0.01% 0.04% 0.11% 0.14% 0.00% 0.01% 0.03% 0.20% 0.65% 0.12%

OTU26 0.02% 0.10% 0.00% 0.38% 0.63% 0.00% 0.08% 0.01% 0.02% 0.08% 0.01% 0.01% 0.01% 0.12% 0.29% 0.10%

OTU6 0.01% 0.01% 0.00% 0.15% 0.46% 2.08% 0.01% 0.04% 0.22% 0.10% 0.01% 0.02% 0.04% 0.16% 0.52% 0.10%

OTU34 0.24% 0.14% 0.00% 0.06% 0.11% 0.00% 0.08% 0.05% 0.22% 0.06% 0.08% 0.05% 0.05% 0.07% 0.06% 0.10%

OTU15 0.03% 0.10% 0.00% 0.12% 0.48% 0.00% 0.07% 0.19% 0.19% 0.09% 0.01% 0.01% 0.01% 0.05% 0.12% 0.09%

OTU4 0.00% 0.00% 2.56% 0.17% 0.48% 6.25% 0.00% 0.00% 0.04% 0.08% 0.00% 0.00% 0.00% 0.10% 0.68% 0.08%

OTU9 0.03% 0.01% 0.00% 0.07% 0.28% 0.00% 0.04% 0.03% 0.30% 0.06% 0.05% 0.01% 0.07% 0.12% 0.11% 0.07%

OTU325 0.04% 0.03% 0.00% 0.10% 0.17% 0.00% 0.03% 0.04% 0.06% 0.11% 0.02% 0.02% 0.07% 0.10% 0.11% 0.06%

OTU61 0.01% 0.04% 0.00% 0.07% 0.22% 0.00% 0.02% 0.02% 0.07% 0.04% 0.07% 0.02% 0.08% 0.10% 0.05% 0.05%

OTU56 0.01% 0.01% 0.00% 0.04% 0.26% 0.00% 0.01% 0.00% 0.05% 0.11% 0.00% 0.01% 0.00% 0.05% 0.24% 0.04%

OTU18 0.01% 0.03% 0.00% 0.06% 0.23% 0.00% 0.02% 0.01% 0.02% 0.02% 0.01% 0.05% 0.00% 0.05% 0.11% 0.03%

OTU47 0.03% 0.04% 0.00% 0.08% 0.06% 0.00% 0.03% 0.01% 0.03% 0.03% 0.02% 0.01% 0.02% 0.05% 0.04% 0.03%

OTU10 0.01% 0.01% 0.00% 0.07% 0.12% 0.00% 0.01% 0.00% 0.06% 0.04% 0.01% 0.00% 0.02% 0.06% 0.15% 0.03%

OTU22 0.04% 0.08% 0.00% 0.04% 0.13% 0.00% 0.04% 0.01% 0.02% 0.00% 0.01% 0.03% 0.03% 0.03% 0.01% 0.03%

OTU198 0.04% 0.03% 0.00% 0.04% 0.02% 0.00% 0.03% 0.02% 0.05% 0.03% 0.02% 0.01% 0.02% 0.05% 0.03% 0.03%

OTU45 0.02% 0.02% 0.00% 0.03% 0.06% 0.00% 0.03% 0.02% 0.04% 0.03% 0.04% 0.01% 0.02% 0.02% 0.03% 0.03%

OTU24 0.01% 0.00% 0.00% 0.03% 0.21% 0.00% 0.02% 0.01% 0.12% 0.03% 0.00% 0.00% 0.01% 0.04% 0.04% 0.03%

OTU359 0.01% 0.07% 0.00% 0.02% 0.02% 0.00% 0.04% 0.01% 0.01% 0.00% 0.01% 0.05% 0.00% 0.01% 0.06% 0.03%

OTU46 0.03% 0.03% 0.00% 0.03% 0.04% 0.00% 0.02% 0.02% 0.03% 0.02% 0.02% 0.00% 0.01% 0.04% 0.04% 0.03%

OTU37 0.01% 0.07% 0.00% 0.04% 0.06% 0.00% 0.03% 0.01% 0.01% 0.00% 0.02% 0.02% 0.02% 0.01% 0.02% 0.03%

OTU50 0.00% 0.00% 0.00% 0.02% 0.11% 0.00% 0.01% 0.00% 0.03% 0.04% 0.00% 0.00% 0.01% 0.04% 0.18% 0.03%

OTU17 0.00% 0.00% 0.00% 0.02% 0.19% 0.00% 0.00% 0.00% 0.01% 0.03% 0.00% 0.00% 0.01% 0.03% 0.12% 0.02%

OTU334 0.01% 0.01% 0.00% 0.04% 0.02% 0.00% 0.02% 0.01% 0.01% 0.01% 0.01% 0.00% 0.01% 0.02% 0.09% 0.02%

OTU73 0.02% 0.04% 0.00% 0.01% 0.03% 0.00% 0.03% 0.00% 0.00% 0.02% 0.01% 0.03% 0.00% 0.00% 0.07% 0.02%

OTU48 0.00% 0.00% 0.00% 0.01% 0.17% 0.00% 0.00% 0.00% 0.14% 0.01% 0.00% 0.00% 0.02% 0.03% 0.01% 0.02%

OTU43 0.02% 0.03% 0.00% 0.00% 0.00% 0.00% 0.04% 0.01% 0.02% 0.00% 0.01% 0.01% 0.02% 0.01% 0.01% 0.02%

OTU85 0.02% 0.06% 0.00% 0.02% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.02% 0.02% 0.01% 0.01% 0.02% 0.02%

OTU232 0.01% 0.01% 0.00% 0.04% 0.02% 0.00% 0.01% 0.01% 0.02% 0.01% 0.01% 0.01% 0.00% 0.03% 0.08% 0.02%

OTU82 0.01% 0.06% 0.00% 0.03% 0.00% 0.00% 0.01% 0.01% 0.02% 0.00% 0.01% 0.02% 0.01% 0.01% 0.00% 0.02%

OTU78 0.01% 0.01% 0.00% 0.04% 0.17% 0.00% 0.00% 0.00% 0.03% 0.00% 0.02% 0.00% 0.02% 0.01% 0.01% 0.02%

OTU83 0.01% 0.03% 0.00% 0.01% 0.01% 0.00% 0.02% 0.02% 0.00% 0.01% 0.02% 0.01% 0.01% 0.00% 0.01% 0.01%

OTU115 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.01% 0.02% 0.03% 0.01% 0.01% 0.00% 0.02% 0.06% 0.00% 0.01%

OTU351 0.01% 0.01% 0.00% 0.01% 0.02% 0.00% 0.01% 0.01% 0.02% 0.01% 0.02% 0.01% 0.02% 0.01% 0.00% 0.01%

OTU114 0.01% 0.01% 0.00% 0.02% 0.01% 0.00% 0.01% 0.00% 0.00% 0.01% 0.01% 0.01% 0.00% 0.02% 0.06% 0.01%

OTU20 0.00% 0.00% 0.00% 0.02% 0.13% 0.00% 0.00% 0.01% 0.04% 0.00% 0.00% 0.00% 0.02% 0.02% 0.01% 0.01%

OTU81 0.01% 0.02% 0.00% 0.03% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.02% 0.01% 0.02% 0.02% 0.01%

OTU108 0.01% 0.01% 0.00% 0.05% 0.02% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.04% 0.03% 0.01%

OTU97 0.01% 0.01% 0.00% 0.02% 0.03% 0.00% 0.01% 0.01% 0.01% 0.00% 0.00% 0.01% 0.01% 0.01% 0.02% 0.01%

OTU41 0.01% 0.00% 0.00% 0.05% 0.01% 0.00% 0.00% 0.00% 0.02% 0.02% 0.00% 0.00% 0.00% 0.01% 0.03% 0.01%

OTU136 0.00% 0.00% 0.00% 0.01% 0.05% 0.00% 0.01% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.07% 0.01% 0.01%

OTU79 0.01% 0.02% 0.00% 0.02% 0.02% 0.00% 0.00% 0.01% 0.03% 0.01% 0.01% 0.01% 0.01% 0.01% 0.00% 0.01%

OTU57 0.00% 0.00% 0.00% 0.01% 0.08% 0.00% 0.00% 0.01% 0.03% 0.01% 0.00% 0.00% 0.02% 0.02% 0.03% 0.01%

OTU51 0.00% 0.01% 0.00% 0.01% 0.07% 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.01% 0.00% 0.01% 0.05% 0.01%

OTU144 0.01% 0.03% 0.00% 0.01% 0.00% 0.00% 0.02% 0.01% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.01%

OTU218 0.01% 0.03% 0.00% 0.01% 0.00% 0.00% 0.02% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.01% 0.01%

OTU80 0.00% 0.00% 0.00% 0.02% 0.03% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.01% 0.02% 0.07% 0.01%

OTU32 0.00% 0.01% 0.00% 0.01% 0.15% 0.00% 0.01% 0.01% 0.02% 0.00% 0.00% 0.00% 0.01% 0.01% 0.01% 0.01%

OTU63 0.00% 0.01% 0.00% 0.01% 0.02% 0.00% 0.01% 0.01% 0.00% 0.01% 0.01% 0.01% 0.01% 0.00% 0.02% 0.01%

OTU69 0.00% 0.00% 41.03 0.01% 0.02% 4.17% 0.00% 0.00% 0.01% 0.02% 0.01% 0.00% 0.02% 0.00% 0.00% 0.01% %

OTU88 0.00% 0.00% 0.00% 0.00% 0.06% 0.00% 0.00% 0.00% 0.00% 0.05% 0.00% 0.01% 0.00% 0.00% 0.02% 0.01%

OTU106 0.00% 0.01% 0.00% 0.02% 0.03% 0.00% 0.01% 0.00% 0.01% 0.00% 0.00% 0.01% 0.01% 0.01% 0.01% 0.01%

OTU100 0.00% 0.01% 0.00% 0.01% 0.05% 0.00% 0.01% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.02% 0.02% 0.01%

OTU124 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.01% 0.01% 0.03% 0.01%

OTU121 0.00% 0.01% 0.00% 0.01% 0.01% 0.00% 0.01% 0.01% 0.01% 0.01% 0.01% 0.00% 0.00% 0.01% 0.00% 0.01%

OTU28 0.00% 0.00% 0.00% 0.02% 0.03% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.03% 0.00% 0.01%

OTU55 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.02% 0.02% 0.01% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01%

OTU119 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.01% 0.02% 0.00% 0.01% 0.00% 0.02% 0.02% 0.00% 0.01%

OTU127 0.00% 0.00% 0.00% 0.02% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.01% 0.00% 0.01% 0.02% 0.01%

OTU11 0.00% 0.01% 0.00% 0.01% 0.01% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.01% 0.01%

OTU143 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.01% 0.00% 0.02% 0.00% 0.00% 0.00% 0.00% 0.03% 0.00% 0.01%

OTU70 0.00% 0.00% 0.00% 0.01% 0.05% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.04% 0.01%

OTU103 0.00% 0.00% 0.00% 0.00% 0.03% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.01% 0.01% 0.04% 0.01%

OTU5 0.00% 0.00% 0.00% 0.02% 0.04% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.01% 0.01% 0.01% 0.01%

OTU76 0.00% 0.00% 0.00% 0.01% 0.05% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.04% 0.01%

OTU126 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.01%

OTU130 0.01% 0.01% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU19 0.00% 0.00% 0.00% 0.01% 0.05% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU102 0.00% 0.00% 0.00% 0.01% 0.05% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.02% 0.00%

OTU336 0.00% 0.00% 0.00% 0.00% 0.07% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU133 0.01% 0.01% 0.00% 0.02% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU105 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00%

OTU292 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU44 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00%

OTU172 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00%

OTU129 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU116 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU111 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU132 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00%

OTU67 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU238 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU99 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU140 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00%

OTU148 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00%

OTU93 0.00% 0.00% 0.00% 0.01% 0.02% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU151 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU323 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU179 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00%

OTU173 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU189 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU59 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU96 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.03% 0.00%

OTU165 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00%

OTU277 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU350 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.01% 0.01% 0.00%

OTU118 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.03% 0.00%

OTU168 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU192 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU190 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU134 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU66 0.00% 0.00% 0.00% 0.00% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU7 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00%

OTU178 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00%

OTU109 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00%

OTU194 0.00% 0.00% 0.00% 0.01% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

OTU89 0.00% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

Chapter III

Influence of the algal microbiome on biofouling during industrial cultivation of Nannochloropsis sp. in closed photobioreactors

Javier B. Giraldo1, 2, Willem Stock2, Lachlan Dow3, Luc Roef1, Anne Willems4, Sven Mangelinckx5, Peter G. Kroth3, Wim Vyverman2 and Mark Michiels1

1 Proviron Holding NV, Georges Gilliotstraat 60, B- 2620 Hemiksem, Belgium

2 Laboratory of Protistology and Aquatic Ecology, Department of Biology, Ghent University, Krijgslaan 281-S8, B-9000 Ghent, Belgium

3 Department of Biology, University of Konstanz, D-78457 Konstanz, Germany

4 Laboratory of Microbiology, Department of Biochemistry and Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

5 SynBioC, Department of Green Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium

J.B. Giraldo, W. Stock, L. Dow, L. Roef, A. Willems, S. Mangelinckx, P.G. Kroth, W. Vyverman, M. Michiels, Influence of the algal microbiome on biofouling during industrial cultivation of Nannochloropsis sp. in closed photobioreactors, Algal Res. 42 (2019). doi:10.1016/j.algal.2019.101591.

Abstract

Industrial cultivation of microalgae is becoming increasingly important, yet the process is still hampered by many factors, including contamination and biofouling of the algal reactors. We characterized a subset of microorganisms occurring in the broth and different biofilm stages of industrial scale photobioreactors applied for the cultivation of Nannochloropsis sp. A total of 69 bacterial strains were isolated, belonging to at least 24 different species. In addition, a green microalga was isolated and identified as Chlamydomonas hedleyi. The effect of C. hedleyi and 24 of the bacterial isolates on the productivity of Nannochloropsis was evaluated through growth and biofilm assays. C. hedleyi was shown to reduce growth and induce biofilm formation in Nannochloropsis. These effects were however indirect as they could be attributed to the bacteria associated to C. hedleyi and not C. hedleyi itself. Although most bacterial strains reported no effect, several were able to induce biofilm formation.

Keywords

Industrial microalgae cultivation; Nannochloropsis; Biofouling; Biofilm formation; Photobioreactor

Highlights

 Extensive isolation of Nannochloropsis microbiome in PBR over gradient of biofilms  A pronounced biofilm inducing isolate – Poseidonocella sp. – was discovered  Bacteria and other contaminants play a major role in the biofilm formation in PBRs  A dedicated management of PBR microbiome might extend their lifetime

1. Introduction

The production of a large variety of products from low-cost raw materials has raised the enthusiasm for a biobased economy. Microalgae have been proven to be a suitable option for such a biorefinery concept: as photosynthetic single-celled microorganisms they mostly require inexpensive substrates and sunlight for growing. By growing them in seawater [29] or wastewater [13], the consumption of valuable clean freshwater can be reduced. It is expected that a large variety of products from microalgae could be obtained simultaneously, including proteins, fractionated lipids, pigments, and residual nutrient feedstocks [28]. Industrial cultivation of microalgae has become increasingly important during recent years [27].

Despite the great potential, important limitations during industrial production exist. For example, many microalgae have high growth rates, but it is still very challenging to achieve high algal densities. This is mainly due to the uneven light distribution in growing ponds or photobioreactors (PBR) [166]. Single cells absorb light and thus much less light penetration is possible at higher algal densities. Light limitation may produce changes in the algae life cycle, affecting morphology and biochemical composition [37]. This problem cannot be solved by increasing light irradiation, which in many cases, would lead to photoinhibition [38], leading to decreased efficiency of photosynthesis and thus decreased growth rates.

The ProviAPT system, designed by Proviron Holding NV, is a promising closed PBR system for large scale microalgae production [79]. The ProviAPT system includes an array of 35 vertical flat-panel type reactors attached to a common feed and aeration layer all enclosed in a translucent plastic bag filled with water (figure 1). The design increases productivity compared to other systems [39] and has thus far successfully been applied for large scale cultivation of different microalgae including Nannochloropsis and Isochrysis spp. The relatively small diameter of the flat-panels reduces problems with limited light penetration at higher densities, but biofilm formation, typical for closed reactor systems, can still strongly reduce light infiltration. Known causes of biofouling and biofilm formation include mechanical problems that are typically related with a lack of aeration, and bacteria [167].

Figure 1 View of ProviAPT in a production plant during 2016.

Algae-bacteria interactions are known to lead to biofilm formation [168]. The biofilm environment presents several benefits for the organisms e.g. they enhance the nutrient diffusion to the cells [169] and protect against antibiotics [170]. Bacteria typically initiate biofilm formation by the release of exopolymeric substances (EPS) in the so- called surface conditioning phase [171]. This EPS matrix is largely constituted of exopolysaccharides and proteins [172]. After the conditioning of the surface, cell attachment takes place [173], ultimately leading to the development of a mature biofilm [174]. These processes are mediated by a wide range of biochemical cascades [175]. It has been reasoned that the interference of theses cascades by chemical agents can be used to inhibit biofilm formation [176].

Industrial microalgae production is typically conducted in non-axenic conditions. Therefore, alterations in the algal microbiome may induce biofouling and biofilm formation initiated by bacterial adhesion to solid surfaces [177]. Due to lower light penetration and the subsequent lower yields, biofouling is highly undesirable in algal production reactors.

This study aims at identifying the effect of associated microorganisms from industrial bioreactors on the productivity of the system.

To this extent, microorganisms from an industrial bioreactor setup were isolated and co- cultured with a Nannochloropsis species. The experimental setup is summarized in

figure 2. The goal was to identify microorganisms which affect the algal growth rate and biofilm formation. While other studies have analysed and compared the algal microbiome in closed PBRs [178–181], here we used samples from a microalgae production plant to actually isolate microorganisms and test their individual contribution to biofouling and biofilm formation.

Figure 2 Summary of the experimental approach: Microorganisms were isolated and identified from various industrial samples and their contribution to biofouling and biofilm formation was estimated.

2. Materials and methods

2.1. Biofilm sampling protocol and selected samples

The sample set consisted of algae broth and biofilm samples from ProviAPT reactors. Each reactor has an updated operational volume of 200 L on an effective surface of 7 m2 [131]. Semi-continuous growth of Nannochloropsis sp. cultures was held in outdoor conditions. Algae cultures were grown in saline media adapted from f/2 [132] by enriching with inorganic nutrients [131].

The algae broth was collected from operative reactors during cultivation of Nannochloropsis sp. strain CCAP211/78, obtained from the Culture Collection of Algae and Protozoa (CCAP, United Kingdom). Biofilm sampling was a reactor destructive process and could therefore only be performed at the end of the production season. Observations on the production facilities showed, however, that biofilms do not necessarily form uniformly in a reactor. One single reactor contains biofilm patches from initial to more mature biofilms. The biofilm sampling reported here was performed during October 2015 from reactors which were discarded for algae production due to inhibitive biofouling.

A total of five samples from two reactors with different productivities were used for microbial isolation and identification. Productivity was estimated as high or low based on overall CO2 consumption, dry biomass weight and observations on reactor behaviour. Values for low productivity reactors oscillated around 5 g ∙ m-2∙ d-1 while high

productivities got up to 18 g ∙ m-2∙ d-1. From an indoor inoculum reactor with high productivity, one algae broth sample was taken. From an outdoor reactor already discarded for production due to low productivity, four samples were taken: one from algae broth and three from biofilms. For the latter, sampling sites in the reactor wall were selected corresponding to areas that were visually defined as early, developed and mature biofilm stages. First, three panels from the same reactor were cut. From each panel, a relevant area with a visually homogeneous biofilm surface was chosen and marked. Each delimited area had a surface of 35 cm2. The external surface of the panels was thoroughly cleaned with 70 % ethanol to prevent contamination of the sample. The biofilm surface was rinsed with sterile PBS-buffer [182] in order to remove the non- adhered cells. Afterwards, the biofilm was scraped off with a sterile spatula and re- dissolved in PBS. The content was homogenized by vigorous agitation and the homogenized solution was used for cell-culturing.

2.2. Microbial isolation

Artificial seawater with added inorganic nutrients and vitamins [131] was used to prepare agar medium for bacterial isolation. It was enriched with 5 g/L peptone (DUCHEFA), 1 g/L yeast extract (DUCHEFA) and 25 % v/v of filter-sterilized supernatant from the particular reactor samples were taken from. The agar (Phyto-agar, DUCHEFA) concentration was 10 g/L. In this way, the enrichment of organic matter was similar to the marine agar based on the medium of Zobell (1941) [183] and nutrient proportions were similar to reactor conditions.

For each sample, a dilution series was made and spread plated on medium that was supplemented with supernatant from the corresponding reactor as described above. All plates were incubated at room temperature (23 °C) until colonies became apparent.

After one week, different colony types were visible. Three representatives of each type were picked and streaked onto a fresh agar plate. This procedure was repeated until it was ensured that all isolates were monoclonal. After 3 additional weeks the initially inoculated plates were again screened for slower growing colonies. The same procedure was repeated with these newly detected colonies. Once the identification was performed, the bacterial strains were grown in the same agar recipe without peptone, yeast extract and sterile supernatant but enriched with 20 mM of glucose at room temperature.

2.3. Bacterial identification by MALDI-TOF MS and 16S rRNA gene sequencing

All bacterial isolates were subjected to MALDI-TOF mass spectrometry in order to identify duplicate strains and group isolates based on their protein profile [184]. Samples were prepared and bacterial fingerprints were produced as describe previously [185] using a 4800 Plus MALDI TOF/TOF™ Analyzer (Applied Biosystems) in linear positive ion mode. Raw spectra were imported into the Data Explorer 4.0 software (Applied Biosystems) and converted to text files which were imported in BioNumerics 7.5 (Applied Maths). Cluster analysis using Pearson’s product moment correlation coefficients and UPGMA clustering was performed in Bionumerics. Based on the MALDI clustering representative strains were selected for 16S rRNA gene sequencing, on the assumption that identical MALDI protein profiles corresponded to a single species [186].

From each selected isolate, DNA was extracted following an adapted phenol:chloroform extraction for marine samples [136]. The V1-V3 hypervariable region of the 16S rDNA from these extracts was amplified by PCR using primers pA (AGAGTTTGATCCTGGCTCAG, positions 8–27) [187] and BKL1 (GTATTACCGCGGCTGCTGGCA, positions 536–516) [188]. Fragments were purified and sequenced, yielding fragments that were approx. 450 bp long. Identification was conducted using the EZTaxon online server [189].

2.4. Isolation, treatment and identification of an invasive green algae species

A green motile algae species was isolated from the outdoor reactor, following the same procedure as the bacterial isolation described above. Based on the unique colony morphology with green color and observations under the microscope, this isolate was identified as an alga. After isolation, this green alga was grown in liquid saline medium enriched with inorganic nutrients [131]. After an initial transfer from plate to flask, the culture was divided in two sub-cultures. The first sub-culture was not treated with antibiotics and the second one was treated with antibiotics over 48h. The antibiotics (Ab) cocktail consisted of 170 µg/µL of penicillin G (Sigma-Aldrich), 85 µg/mL of streptomycin (Sigma-Aldrich), 17 µg/mL of chloramphenicol (Sigma-Aldrich), 10 µg/mL of tetracycline (Sigma-Aldrich) and 10 µg/mL of kanamycin (Sigma-Aldrich). After the Ab treatment, cells were centrifuged (10000 rpm, 5 minutes, Beckman Coulter) and the pellet was suspended in antibiotic-free medium. This was repeated twice in order to ensure removal of the antibiotics from the Ab-treated culture. This culture was checked for bacterial presence with two approaches: (1) SYBR green staining followed by

microscopy and (2) inoculation of algae samples in agar plates followed by incubation for one week at room temperature. In both approaches, no evidence of bacteria was detected. Despite the fact that the combination of these two methods cannot ensure absolute axenicity, the microbial population associated to the algae was considered to be dramatically reduced and therefore sufficient to prove the influence of the algal microbiome. To identify the alga, DNA was isolated using a phenol:chloroform based extraction [136]. The ‘internal transcribed spacer’ (ITS) region was amplified with ITS universal primers – 1800F [190], R-ITS4 (White et al., 1990), DITS3 and DITS2 [192] – reporting fragments of approximately 640 bp which were sent for sequencing. The obtained sequence was deposited on GenBank (accession number MH727704). A BLAST search in the NCBI-nr/nt database was performed with the resulting sequence. It produced a match with Chlamydomonas hedleyi (GenBank AJ297797.1) with a Max. score and total score of 1188 for both, a query cover of 100 %, an E-value of 0.0 and an identity value of 100 % (retrieved on May 14th, 2019).

2.5. Algae cultivation in shake flasks

The microalga Nannochloropsis sp., strain CCAP211/78 was obtained from the Culture Collection of Algae and Protozoa (CCAP, United Kingdom). Nannochloropsis sp. and the isolated C. hedleyi were cultivated in 100 mL Erlenmeyer flasks (Schott-Duran) on a shaker at 70 rpm with a culture volume of 50 mL of algae growing in saline medium at 20°C [131]. The light irradiation was 70 µmol ∙ m-2 ∙ s-1 and applied from a cool fluorescent light source in day-night cycles of 16:8.

2.6. Co-cultivation of Nannochloropsis sp. with the green alga species

The growth rates of Nannochloropsis sp. and the two subcultures (Xenic and Ab treated) of C. hedleyi were estimated by daily measurements of OD750 during 9 days. The initial

OD750 for Nannochloropsis sp. was 0.550±0.005 and for the two C. hedleyi cultures 0.140±0.005. The volume in every flask was 50 mL. For the co-cultivation flasks, 1 mL from the same C. hedleyi was inoculated in a Nannochloropsis sp. culture. All treatments were setup in triplicate. At the end of the 9 days, growth rates were calculated as the logarithmic ratio of cell densities divided by the time and expressed as d-1.

2.7. Biofilm formation assay

Biofilm formation assays were performed in flat bottomed 96-well plates (Sarstedt) that were inoculated with Nannochloropsis sp. and co-cultures of C. hedleyi or the bacterial isolates depending on the case [193]. Each well was inoculated with 150 µL of algae

suspension with an optical density of OD630 1.50±0.05 and OD750 1.40±0.05. Each bacterial culture was transferred from a colony growing on agar to liquid medium and used after three days. Bacterial isolates were cultivated in saline media [131] enriched with 20 mM glucose as an organic carbon source is necessary to maintain bacterial cultures. Since addition of glucose might alter the capacity of different microorganisms to induce biofilm formation, prior to co-cultivation bacterial cultures were washed twice with fresh medium via centrifugation (10000 rpm, 5 minutes, Beckman Coulter) in order to remove traces of glucose. A volume of 30 µL of the glucose-free bacterial suspension was used to inoculate each well which later was brought to a final volume of 250 µL with fresh sugar-free culture media. Well plates were manually prepared and the well position of each treatment was randomized. Plates were incubated for 10 days under the same conditions as those used for shake flasks, however without shaking.

Well plates were prepared for biofilm quantification using a Freedom EVO pipetting robot (Tecan) along with a Tecan Infinite F500 plate reader. Growth media and non- adherent cells were removed via both robotic and manual pipetting, and biofilm density measured using absorbance at wavelengths 630 nm and 750 nm, as proxies for chlorophyll c concentration [194] and biomass [195] respectively. In order to quantify extracellular polysaccharide secretion, well plates were then stained with 40 µL of 0.02 % (w/v) aqueous crystal violet (CV) solution (Sigma-Aldrich), and incubated for one minute. Excess CV solution was removed, well plates rinsed with water and excess liquid removed via manual pipetting. CV staining intensity was quantified via absorbance at 580 nm. The absorbance of each well was measured in twelve locations. The average value of these twelve measurements was used as a single replicate.

2.8. Statistical analysis

The growth rates and each wavelength used in biofilm absorbance measurements were separately compared by means of a one-way ANOVA followed by a Tukey HSD post-hoc test if significant (p-value < 0.05). The ODs from the biofilm formation assays using the different bacterial strains were not compared by means of an ANOVA since residuals were not normally distributed and group sizes were not always equal. A non-parametric Kruskal-Wallis test was performed for each wavelength followed by a Dunn's test (R- package RVAideMemoire 0.9-66) if significant (p-value < 0.05). Only p-values comparing each bacterial strain with the control were retained and corrected by means of false discovery rate multiple test correction. All statistical tests were run in R (3.4.1). Figures were also designed in this software with ggplot 2 (2.2.1).

3. Results and discussion

3.1. Microbial isolation

Sixty-nine bacterial isolates were obtained from the sampling of the reactor setups under study. Using MALDI-TOF MS profile clustering in combination with partial 16S rRNA gene sequencing, 24 different species-level taxa were identified among 64 strains. For 5 isolates, a good quality MALDI profile, nor good sequence, could be obtained. Most isolates belonged to the Gamma- and Alphaproteobacteria and the rest were Bacteroidetes (table 1).

Table 1 Overview of the 69 isolates that were identified by MALDI-TOF MS and/or 16S rRNA gene sequence analysis. Strains that were used in the co-cultivation experiments are shown in bold. Numbers in brackets at the identification column refer to different strains on the same genus. Asterisks in the accession number column refer to strains identified by MALDI-TOF MS.

Broth/ Biofilm Reactor Accession Strain Identification (BF) productivity number

R-67403 Idiomarina loihiensis Broth High MH974272

R-67404 Halomonas sp. (1) Broth High *

R-67405 Roseivirga ehrenbergii Broth High *

R-67406 Alcanivorax sp. Broth High *

R-67407 Alcanivorax sp. Broth High MH974273

R-67408 Halomonas sp. (1) Broth High MH974238

R-67409 Alcanivorax sp. Broth High MH974274

R-67410 Halomonas sp. (2) Broth High MH974239

R-67411 Halomonas venusta Broth High *

R-67413 Alcanivorax sp. Broth High MH974262

R-67415 Halomonas venusta Broth High *

R-67416 Halomonas sp. (1) Broth High MH974240

R-67417 Celeribacter Broth High MH974241 baekdonensis

R-67418 Devosia subequoris Broth High MH974242

R-67419 Alcanivorax sp. Broth High MH974243

R-67420 Paracoccus sp. Broth High MH974263

R-67421 Roseovarius pacificus Broth High *

R-67423 Stappia indica Broth High MH974244

R-67424 Halomonas venusta Broth High *

R-67425 Halomonas venusta Broth High MH974245

R-67426 Halomonas venusta Broth High MH974246

R-67427 Marinobacter Broth High MH974264 adhaerens

R-67428 Marinobacter adhaerens Broth High MH974247

R-67429 Alcanivorax sp. Broth High MH974265

R-67430 Roseivirga ehrenbergii Broth High MH974266

R-67431 Roseivirga ehrenbergii Broth High MH974267

R-67432 Halomonas venusta Broth High MH974248

R-67434 Muricauda sp. Broth High MH974249

R-67436 Alcanivorax sp. Broth Low MH974250

R-67437 Pseudoalteromonas sp. Broth Low MH974251

R-67438 Pseudoalteromonas sp. Broth Low MH974252

R-67439 Halomonas sp. (1) BF Low *

R-67440 Halomonas alkaliphila BF Low *

R-67441 Marinobacter adhaerens BF Low *

R-67442 Halomonas sp. (3) BF Low MH974253

R-67443 Halomonas sp. (3) BF Low *

R-67444 Marinobacter adhaerens BF Low *

R-67445 Halomonas alkaliphila BF Low *

R-67446 Idiomarina loihiensis BF Low MH974254

R-67447 Idiomarina loihiensis BF Low MH974255

R-67448 Halomonas alkaliphila BF Low *

R-67449 Idiomarina loihiensis BF Low MH974268

R-67450 Idiomarina loihiensis BF Low *

R-67451 Halomonas alkaliphila BF Low *

R-67452 Roseovarius pacificus BF Low *

R-67453 Idiomarina loihiensis BF Low *

R-67454 Halomonas alkaliphila BF Low MH974256

R-67471 Nitratireductor Broth Low * aquimarinus

R-67472 Labrenzia aggregata Broth Low MH974257

R-67473 Halomonas sp. (3) BF Low *

R-67474 Nitratireductor BF Low * aquimarinus

R-67475 Halomonas sp. (3) BF Low *

R-67476 Idiomarina sp. BF Low MH974258

R-67477 Roseovarius pacificus BF Low MH974269

R-67478 Alcanivorax sp. BF Low *

R-67479 Halomonas sp. (1) BF Low MH974259

R-67480 Donghicola sp. BF Low MH974270

R-67481 Roseovarius pacificus BF Low *

R-67482 Halomonas alkaliphila BF Low *

R-67483 Idiomarina loihiensis BF Low MH974260

R-67484 Roseovarius mucosus BF Low MH974261

R-67486 Celeribacter halophilus BF Low MH974271

R-67487 Nitratireductor BF Low MH974275 kimnyeongensis

R-67488 Poseidonocella sp. BF Low MH974276

Furthermore, a green motile alga strain was isolated from the low productivity reactor and identified as Chlamydomonas hedleyi. The presence of green algae at the production site is often correlated with an increase in biofouling as well as a dramatic drop in productivity. Observations in the production plant suggested that C. hedleyi is the principal species invading the reactors.

Due to the likely unculturability of certain strains, the isolates identified in this study may represent only part of the microbiome [196]. Our findings are however in line with previous efforts to identify bacteria present in Nannochloropsis cultures. A study based on microbial communities associated to Nannochloropsis salina in open ponds [129,130], concluded that the microbial community is highly dynamic, depending on growth phase and external conditions. The authors showed that Alteromonadales dominated in exponential phase while Alphaproteobacteria and Flavobacteria dominated in the stationary phase. This is in line with the results presented here, since the reactors were in late exponential to stationary phase when sampled.

The dominance of Alphaproteobacteria in all samples may be related to the secretion of signal molecules. Members of the Roseobacter clade have been described to produce acylated homoserine lactones (AHL) [197] which are quorum sensing compounds, which are well-known signals that can be involved in the regulation of biofilm formation processes [198]. Further studies are desirable to better understand signalling between bacteria and algae in closed PBRs.

3.2. Effect of the microbiome on the productivity of Nannochloropsis sp.

The detection of C. hedleyi in the PBRs was often correlated with a dramatic decrease of their productivity. Therefore, testing the influence of C. hedleyi on Nannochloropsis sp. growth was the first step to understand biofouling in the ProviAPT production reactors. It was decided to compare the growth of Nannochloropsis sp., an axenic and a non- axenic C. hedleyi culture, and both Nannochloropsis-C. hedleyi co-cultures (figure 3).

Figure 3 Estimated growth rates (d-1) of five different cultures: (N) Nannochloropsis sp.; (C-Xe) C. hedleyi (Xenic); (C-Ab) C. hedleyi (Ab treated); (N+C-Xe) Nannochloropsis sp. + C. hedleyi (Xenic); (N+C-Ab) Nannochloropsis sp. + C. hedleyi (Ab treated). Lower case letters indicate groups that are not significantly different (Tukey HSD post-hoc test, p-values > 0.05).

When the algal monocultures were compared, both C. hedleyi subcultures grew faster than Nannochloropsis sp. (p-value < 0.05). This was expected, since from previous observations it is known that C. hedleyi may overgrow the reactors shortly after its detection. However, the C. hedleyi subcultures behaved differently in co-cultivation with Nannochloropsis sp. The growth rate of the co-culture with Ab treated C. hedleyi is comparable to the Nannochloropsis sp. culture, while the growth rate of the Nannochloropsis sp. with xenic C. hedleyi is dramatically lower (p-value < 0.05) compared to all other treatments. These observations suggest that the microbiome associated to C. hedleyi might not affect its own growth but negatively affect the algal growth of Nannochloropsis sp. during co-cultivation.

In addition to monitoring growth effects, a biofilm formation assay was used to estimate the ability of C. hedleyi to induce biofilm formation of Nannochloropsis sp. (figure 4). Although growth rates of both C. hedleyi subcultures were similar (figure 3), the two

cultures differ in terms of biofilm formation (figure 4). Biofilm production was most pronounced in the xenic C. hedleyi cultures, while the lowest values were recorded for the Ab treated C. hedleyi. The presence of bacteria stimulated biofilm formation of C. hedleyi. It is concluded that the presence of bacteria contributed to the attachment of algae on the substrate, which is in line with the classical mechanisms described for algae biofilm formation, where it is suggested that bacteria play an important role in initiating and developing biofilms [67]. A similar pattern was observed when C. hedleyi was co-cultivated with Nannochloropsis sp. The xenic culture reported higher optical densities when compared to the Ab treated (p-value < 0.05). Only the xenic C. hedleyi and the co-cultivation of the xenic C. hedleyi and Nannochloropsis sp. reported higher values compared to axenic Nannochloropsis sp. at all three wavelengths (p-value < 0.05). These results together with those from the growth rate experiments highlight the effect of bacteria associated with C. hedleyi, suggesting they negatively affect the growth rate of Nannochloropsis sp. and initiate biofilm formation. Despite several Chlamydomonas sp. strains are used for industrial applications [199,200], the potential of C. hedleyi remains unexplored. This might be a promising candidate for harvesting microalgae by means of bioflocculation [201,202] although this suggestion requires further investigation.

Figure 4 Estimated biofilm densities of five different cultures: (N) Nannochloropsis sp.; (C-Xe) C. hedleyi (Xenic); (C-Ab) C. hedleyi (Ab treated); (N+C-Xe) Nannochloropsis sp. + C. hedleyi (Xenic); (N+C-Ab) Nannochloropsis sp. + C. hedleyi (Ab treated). Lower case letters indicate groups that are not significantly different (Tukey HSD post-hoc test, p-values > 0.05).

3.3. Poseidonocella sp. is the strongest biofilm inducer among 24 isolated bacterial strains

Twenty-four bacterial strains representative of the isolated diversity (table 1) were selected for the biofilm formation assay with the Nannochloropsis sp. (figure 5).

Figure 5 Biofilm formation assay of the bacterial-algae co-cultivations. The R-codes from bacterial names (x-axis) indicate co- cultivation of that strain together with Nannochloropsis sp. Nannochloropsis sp. was used as a control. Small letters next to bacterial names indicate p-values < 0.05 in comparison with the control (a for 580 nm, b for 630 nm and c for 750 nm). Strains are ordered by increasing biofilm formation capacity.

Most bacterial strains did not affect biofilm formation compared to the axenic Nannochloropsis sp. The strain R-67488 – closely related to Poseidonocella pacifica (table 1), an alphaproteobacterium first described from shallow sandy sediments of the Sea of Japan [203] – is the exception here. It very strongly induced biofilm formation in co-culture with Nannochloropsis sp. Although the different values obtained for the three wavelengths are not significant due to the high variation associated with this method at higher ODs, the high values obtained at OD 630 and 750 indicate the ability of this bacterium to induce pronounced biofilms within a short amount of time. Although to a lesser extent, R-67480 (Donghicola sp.) and R-67418 (Devosia subaequoris) also significantly induced the development of biofilms.

In this study, Poseidonocella sp. was only found in biofilms from low productivity reactors. The observed effect on Nannochloropsis sp. during laboratory co-cultivation is similar to that in low productivity reactors during industrial microalgae production. This bacterium was able to induce attachment of a considerable amount of algal biomass,

which can be concluded from the high 630/750 ratio, if this ratio is considered as an indication of chlorophyll content per algal biomass.

Co-cultivations resulting in lower biofilm formation than axenic Nannochloropsis sp. cultures could be interpreted as an indication that the bacterial strains involved may have prevented biofilm formation by Nannochloropsis sp. or could be indicative of an algicidal effect of the bacteria. This could be the case when values taken at 580 nm are much higher than the ones at 630 nm and 750 nm since this is an indication of a higher relative abundance of bacterial cells and EPS compared to algae due to CV staining. These hypotheses would need to be confirmed by experiments on larger scale including co-cultivation.

Moreover, a study on microbial communities associated to Nannochloropsis oceanica proved the importance of environmental parameters such as growth temperature [167]. These authors found that certain Halomonas sp., Muricauda sp., Rhodobacterales and an unknown gammaproteobacteria strain caused aggregation of N. oceanica. These conclusions emphasize the role of bacteria in inducing biofouling on Nannochloropsis sp. This is in line with this study, where certain Proteobacteria strains were found able to induce biofouling even at lower temperatures.

Biofouling and biofilm formation can lead to important losses during algae cultivation [55]. This can be prevented by improving technical aspects and biological control. The reactor wall composition influences the algal adhesion on the solid surface [204]. Nevertheless, this study emphasizes the relevance of the microbiome in non-axenic production and opens the exploration of ways to control it in order to prevent biofilm formation during cultivation in closed PBRs.

This study suggests that bacteria can both positively and negatively impact biofilm formation in a PBR. This has important implications for industrial algal cultivation, where working in axenic conditions is hardly feasible. The inoculation of bacteria which inhibit biofouling might suppress biofouling and extend the lifetime of the reactors. At the same time, culturing conditions can be tweaked to suppress biofilm inducers or their activity. Identification of the bacteria involved and their biochemical pathways will be crucial to develop such strategies.

4. Conclusions

Chlamydomonas hedleyi was identified as a typical contaminant at a microalgal production site and proved to have higher growth rate than Nannochloropsis sp., which allows it to rapidly overgrow the reactors. The contaminant had a microbiome itself,

which stimulated biofilm development, indicating a dual threat of C. hedleyi to the yield of the PBRs. The bacterial communities of the Nannochloropsis sp. reactors were dominated by Proteobacteria, of which some were able to induce biofilm formation, a Poseidonocella sp. in particular. Further studies on the isolates obtained in this study will help to identify the pathways involved in biofilm inhibition and formation.

5. Acknowledgements

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642575. W.S. was funded by the Fund for Scientific Research – Flanders (FWO-Flanders, Belgium). This work makes use of resources and facilities provided by UGent as part of the Belgian contribution to EMBRC-ERIC. The authors would like to thank the Konstanz University Screening Centre for their time and equipment.

Chapter IV

Towards identification of chemical signals involved in biofilm formation during non-axenic cultivation in closed photobioreactors

Javier B. Giraldo1, 2, Emilio Cirri3, Nuwanthi Samarasinghe5, Anne Willems4, Sven Mangelinckx5, Georg Pohnert3 and Luc Roef1, Wim Vyverman2

1 Proviron Holding NV, Georges Gilliotstraat 60, B- 2620 Hemiksem, Belgium

2 Laboratory of Protistology and Aquatic Ecology, Department of Biology, Ghent University, Krijgslaan 281-S8, B-9000 Ghent, Belgium

3 Institute of Inorganic and Analytical Chemistry, Friedrich Schiller University Jena, D- 07743 Jena, Germany

4 Laboratory of Microbiology, Department of Biochemistry and Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

5 Department of Green Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium

Abstract The mechanisms behind the ability of a Poseidonocella strain to induce biofilm formation on Nannochloropsis sp. were investigated in this chapter. A biofilm formation assay was used to perform different tests. The sterile supernatant from the bacterium was able to induce biofilm formation suggesting thus the presence of an exometabolite that triggers the mechanism. Glucose was able to induce biofilm formation on the algae. Although traces of glucose were present on the bacterial supernatant, the contribution of the sugars was discarded. A heat treatment of the supernatant followed by a biofilm formation assay suggested that bioactive compounds could be small molecules instead of extracellular enzymes discarding thus the possible contribution of glucose towards biofilm formation. A solid phase extraction method was established in order to extract bioactive compounds from the bacterial supernatant. Two different cartridges were used: hydrophilic-lipophilic balance for a wide range of compounds (HLB, Oasis) and a weak anion exchanger with capacity of binding more hydrophilic compounds (Easy CHROMOBOND). The Easy cartridge was more successful in extracting the bioactive compounds suggesting thus a polar character of the molecule. Preliminary results on UHPLC were obtained although the lack of sufficient biological replicates impedes the confirmation of these results. Therefore, it is suggested to repeat the extractions with more replicates followed of a bioassay guided fractionation to validate this hypothesis.

1. Introduction and experimental approach Non-axenic cultivation may occasionally lead to the development of undesired contaminants. This depends on unexpected environmental changes that may affect algae production by inducing biofouling and biofilm formation. As these biofilms reduce the light penetration to algae, it diminishes photosynthesis consequently reducing the cell growth and biomass yield. Bacterial strains have been isolated and identified from a wide range of reactor samples with different productivities (Chapter 3). The contributions of these bacterial isolates in biofilm formation have been individually tested. After successful microbial screening and evaluation, a strain of Poseidonocella was identified as a strong biofilm inducer and therefore selected as a model organism for studying algae-bacteria interactions. Poseidonocella sp. and Nannochloropsis sp. will be used as model organisms for the study of biofilms in the ProviAPT system. Understanding of biological and chemical processes behind the biofilm formation is crucial in order to develop preventive strategies. Since the ability of this bacterium to induce biofilm formation has been confirmed, chemical signals from this bacterium will be investigated in order to decipher the initial steps of the biofilm formation process. In this chapter, we confirmed the role of Poseidonocella sp. as a biofilm inducer. We successfully implemented in our research group the biofilm formation assay previously used in Konstanz (Germany). This allowed us to explore the chemical signaling between these two microorganisms by focusing on the detection of chemical signals secreted from Proseidonocella sp. The exometabolome of the bacterium was studied by the separation and sterilization of the supernatant where bacterial cells grew. This sterile supernatant contains traces of glucose as it is an essential requirement for bacterial growth. The sterile supernatant induces biofilm formation of Nannochloropsis sp. Despite glucose is able to induce biofilm formation, this chapter reveals that chemical signals present in the supernatant are responsible of this effect. By means of solid phase extraction (SPE), these signals have been identified as polar molecules. Preliminary identification suggests that these molecules may be dihydroquinoline isomers. Despite these molecules have been described as biofilm inducers in other microorganisms, further experimentation is needed to confirm this finding as it is explained later in this chapter.

2. Materials and Methods

2.1. Cultivation of Nannochloropsis sp. Nannochloropsis sp. strain CCAP211/78 was obtained from culture stock of Proviron which was initially obtained from Culture Collection of Algae and Protozoa (CCAP, United Kingdom). For sub-cultivation, 4 ml of algae cultures were inoculated to 50 mL of saline

medium [131] in 250 ml flasks (Greiner bio-one). Algae cultures were incubated at a temperature of 25 °C. The light irradiation was 30 µmol ∙ m-2 ∙ s-1 and applied from a cool fluorescent light source for 24 hours. Growth of algae cultures was assessed by taking daily absorbance measurements at 630 nm and 750 nm wavelengths using a spectrophotometer (UV-1601, SHIMADZU). An estimation of maximum chlorophyll concentration was measured at 630 nm [205] and total algae biomass was measured at 750 nm [206]. Continuous subcultivations were conducted every two weeks in order to ensure that algal cells were in the same cell stage when harvesting for performing the experiment. 2.2. Cultivation of Poseidonocella sp. Poseidonocella sp. MH974276 was isolated and identified from previous work of the research group [163] that involved screening of bacterial strains in algae cultures at Proviron and evaluation of their effect on biofilm formation. The same saline medium enriched with 20 mM of glucose was used for the bacteria culture. For sub-cultivation, 4 mL of bacteria culture were inoculated to 50 mL of medium in 250 mL flasks (Greiner bio-one). Bacteria were inoculated 5 days before the bioassay preparation. Bacteria cultures were also incubated at a temperature of 25 °C. 2.3. Biofilm formation assay Nannochloropsis sp. was cultivated under different treatments. The tested treatments are explained in the next sections (figure 1). In each bioassay, co-culture of Nannochloropsis sp. and Poseidonocella sp. was used as positive control and Nannochloropsis sp. culture, Poseidonocella sp. cell cultures, medium with glucose and glucose-free medium were used as negative controls. Biofilm formation assays were performed in flat bottomed 96-well plates (Greiner) that were inoculated with required combinations of algae, bacteria, extracts and mediums (Figure ). Every treatment was replicated six times in the same plate. Optical density of algae cultures that were used for biofilm bioassay was measured using a spectrophotometer at 630 nm and 750 nm. Each well of treatments was inoculated with

150 µL of algae suspension with an optical density of OD630 1.500 ± 0.05 and OD750 1.400 ± 0.05 except the negative controls. Poseidonocella sp. culture with an optical density of OD600 0.500 ± 0.05 was used in bioassay. After every treatment, the final volume was adjusted to 250 µL with glucose-free medium. For negative controls, 250 µL of Nannochloropsis sp. culture, Poseidonocella sp. cell cultures, medium with sugar and medium without sugar were used while three different volumes of Poseidonocella sp. cell cultures (10 µL, 30 µL and 50 µL) that co-cultured with Nannochloropsis sp. (150 µL) were used as positive control in bioassays. Well plates were manually prepared and the well position of each treatment was randomized. Initial absorbance of well plates was taken with the plate reader (CYTATIONǀ3, BioTek) using Gen5 software. Absorbance measurements in plate reader were taken as an average of absorbance at

nine points of each well at 630 nm and 750 nm. Plates were incubated for 10 days under the same conditions as used for algae and bacteria flasks. After 10 days, growth media and non-adherent cells were removed by manual pipetting. Biofilm density was measured using absorbance at wavelengths of 630 nm and 750 nm. 2.4. Washing step and SSN preparation Bacterial cultures were washed via centrifugation in order to remove traces of glucose. Volumes of 5 ml of Poseidonocella sp. were centrifuged at 5000 rpm for 10 minutes. Supernatants were separated and filtered through a filter (0.2 µm Supor® Membrane, PALL® Acrodisc® 32mm) to obtain sterile supernatant. Bacterial pellets were suspended in 5 ml of medium without glucose and centrifuged again. Supernatants were discarded and the bacterial cell pellet was suspended in 5 ml of glucose-free medium for use as washed bacterial cells. 2.5. Heat treatment on SSN Two different heat treatments were followed at 80 °C for 1 hour and 90 °C for 1 hour and 30 minutes. Sterile supernatant stored in glass bottles was heated. Glass bottles were capped during heating in order to minimize the evaporation which was not detected. After heat treatment, the supernatant was allowed to cool down to room temperature before use in bioassays.

2.6. Solid phase extraction protocol For fractionation, bacterial supernatant was separated using SPE. Two types of cartridges were used i.e. HLB, OASIS (60 mg, 3 cc, 30 µm particle size, copolymer) and CHROMOBOND Easy (200 mg, 6 cc, 80 µm particle size, modified polystyrene- divinylbenzene copolymer). Both sterile supernatants of bacteria culture and filtered fresh mediums without glucose were used for SPE. HLB cartridges are used for extraction of a wide range of molecules, ranged from extraordinary retention of hydrophilic molecules to relative retention of hydrophobic molecules (Oasis, Kinesis). Easy cartridges are specifically designed to extract hydrophilic molecules (SPE- Macherey-Nagel). First, cartridge columns were conditioned using methanol and water (2 ml for 60 mg cartridge and 3 ml for 200 g cartridge of both methanol and water). Then, prepared samples (50 ml for 60 mg and 125 ml for 200 mg) were passed through the column under vacuum. After the aspiration, columns were washed with methanol-water (5:95, v/v) (2 ml for 60 mg and 4 ml for 200 mg). Finally, analytes were eluted with methanol (2 ml for 60 mg cartridge and 4 ml for 200 mg cartridge). Flow through (Ft), washed phase (Wp) and eluate were collected separately and stored at -20 °C.

Figure 1 Diagrammatic representation of different treatments used in bioassay. 2.7. UHPLC measurements Samples were dried under a gentle stream of nitrogen and re-dissolved in 65 µL of methanol. Two quality control (QC) samples were prepared by pooling 5 µL of each sample.

After randomizing the order of the samples and including QC samples at the beginning and at the end of the sequences, 1 µL of each samples was analysed by UHPLC Dionex UltiMate® 3000 (Thermo Fisher Scientific, Dreieich, Germany), coupled to an ESI- Orbitrap MS Q-Exactive Plus (Thermo Fisher Scientific, Dreieich, Germany).

Liquid chromatography was performed on an Accucore® C18 column (2.1 × 100 mm, 2.6 μm particle size, Thermo Scientific, Dreieich, Germany). The composition of the mobile phase was set to 100% A (0.1% formic acid and 2% acetonitrile in water) for 0.2 min and ramped to 100% B (0.1% formic acid in acetonitrile) in a linear gradient within 12 min. The solvent composition was held at 100% B for 1 min, returned to 100% A in 0.1 min and held at 100% A for 0.9 min. The flow rate ramped from 0.4 mL min-1 to 0.7 mL min-1 from 0.5 min to 13.5 min.

Ionization was performed with a spray voltage of 3 kV and a capillary temperature of 360 °C. The sheath gas flow rate was kept at 60 au (arbitrary units); the auxiliary gas rate at 20 au and the sweep gas rate at 5 au. Nitrogen was used as desolvation gas.

For monitoring, the scanned mass range was between m/z 100 and 1500, at a resolution m/Δm 280,000 full width at half maximum (FWHM) (m/z 200) in positive mode and in negative mode separately, with automatic gain control (ACG) target 3 × 106, a maximum injection time (IT) of 200 ms.

For compound identification, full-scan MS/data dependent MS/MS (ddMS2) experiment was performed on QC samples. Each experiment was composed of one full MS and up to five ddMS2. The five ions with the most intense signal detected in the full MS scan (intensity threshold 1,6× 105) produced a specific MS/MS spectrum. For full MS, the settings were the ones described above, while for the data dependent MS/MS the settings were the following: positive and negative mode with a resolution of m/Δm 17,500 and an ACG target 1 × 105, a maximum IT of 50 ms, a stepped normalized collision energy (NCE, 15 ,30 , 45), an isolation window of 0.4 m/z.

Finally, a parallel reaction monitoring (PRM) experiment was performed to identify interesting compounds with the following settings: negative mode with m/Δm 70,000 resolving power and an ACG target 1× 106, a maximum IT of 100 ms, a stepped normalized collision energy (NCE, 35, 45, 65), an isolation window of 0.4 m/z.

All the data were acquired and processed with the software XcaliburTM version 3.0.63 (Thermo Fisher Scientific, Bremen, Germany).

2.8. LC-HR-MS data analysis XcaliburTM raw data file were imported into Thermo Compound Discoverer 2.1.0.398 (Thermo Fisher Scientific, Bremen, Germany) and analyzed following a standard pipeline for untargeted metabolomics for high resolution spectra. The important values for features extraction are the following: precursor ion deviation of 5 ppm, maximum retention time shift 0.5 min., signal to noise threshold (S/N) 3, minimum peak intensity for peak selection 1 × 106 au, retention time shift for grouping 0.5 min., and relative intensity tolerance for isotope search 30%. The exact masses of unknown compounds found in the samples were compared to online databases (PubChem, ChemSpider, mzCloud) and to an in-house library of 650 natural compounds (mass tolerance = 5 ppm) for identification.

After the analysis, a table with putative compound names and the molecular formula, exact masses, retention times and chromatographic area for each sample was exported for further processing with an in-house R-script (http://www.R-project.org/, R Core Team, 2017).

All features found in the medium blank samples were removed from the samples. Data were then filtered based on QCs coefficient of variation (CV): only features with CV < 20% were retained (Dunn et al., 2011).

Finally, data were normalized on diatom cell densities and Pareto scaled. The obtained .csv table was used to perform statistical analysis with MetaboAnalystR (Chong and Xia, 2018). PCA was performed to detect grouping and outliers in the samples. Significant features were selected by looking at PCAs loading plots and at the results of two-way ANOVA (adjusted p-value (FDR) cutoff = 0.05, FisherLSD post-hoc analysis), which were visualized on heatmaps (distance measure = euclidean, clustering algorithm = Ward).

After statistical analysis, significant features were selected in the Thermo Compound Discoverer molecule list and exported to SIRIUS v. 4.0 (Böcker et al. 2008) to confirm features identity.

3. Results and discussion

3.1. Effect of Poseidonocella sp. on BF formation in Nannochloropsis sp. cultures Three different doses of Poseidonocella sp. were co-cultured with Nannochloropsis sp. in bioassays. Figure 2 suggests a dose-dependent relationship between volumes of added Poseidonocella sp. and biofilm density. Increasing volumes of added Poseidonocella sp. result in increased biofilm density at both wavelengths of 630 nm and 750 nm. No substantial difference in biofilms was observed between 30 µL and 50 µL of Poseidonocella sp. However, a considerable increase in biofilm density was observed for all three doses of Poseidonocella sp. compared to negative controls except for the negative control with glucose medium. Unexpected high biofilm density was observed for the negative control of Nannochloropsis + medium with glucose. Accordingly, further experiments were carried out to confirm the effect of glucose on biofilm formation.

1 .2 6 3 0 n m

1 .0 7 5 0 n m

t i

s 0 .8

l 0 .6

i t

p 0 .4 O 0 .2

0 .0 a a o s i i 1 2 3 n o d d s s s n P e e o o o a m P P P N M + + + + G o o o o + n n n n o n n n n n a a a a n N a N N N N

Figure 2 Biofilm formation assay - bacterial-algae co-cultivations with different volumes of Poseidonocella. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp. G medium = Medium with glucose, Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation. 3.2. Effect of glucose on biofilm formation in Nannochloropsis sp. cultures Increasing glucose concentration from 0 to 8 mM reported a gradual increase of biofilm density (figure 3). However, the highest volume of glucose-containing medium gave a lower absorbance reading than all positive controls with Poseidonocella sp. On the other hand, the measured absorbance values of Nannochloropsis sp. in glucose-containing medium were considerably higher than negative controls indicating an effect of glucose on biofilm formation. Washed Poseidonocella sp. cells which contain no traces of glucose gave substantial levels of biofilm formation, further verifying the effect of Poseidonocella sp. The biofilm density from washed Poseidonocella cells was lower than from Poseidonocella cells in glucose medium. These results show that both glucose and Poseidonocella sp. have a positive effect on biofilm formation in Nannochloropsis sp. culture.

N e g a tiv e G lu c o s e P o s itiv e W a sh e d S te r ile c o n tr o ls tr e a tm e n ts c o n tr o ls c e ll s s u p e rn a ta n ts 1 .0

0 .8 6 3 0 n m

t i

s 7 5 0 n m

n 0 .6

a c

i 0 .4

p O 0 .2

0 .0 a a 1 2 3 4 1 2 3 ia i s i 1 2 3 1 2 3 d d o d ia ia ia ia s s s s s s s s s e e P e d d d d o o o o o o o o o m M e e e e P P P P P P P P P M + + + d d d G + M M M M o o o e e e S S S o n n n h h h S S S n G G G G n n n s s s + + + n + + + + a a a a a a o o o a o o o o n n n N n n n n N N N W W W n n n n n n n + + + a a a a a a a o o o n n n N N N N N N N n n n a a a N N N

Figure 3 Biofilm formation assay – Incubation of Nannochloropsis sp. in the presence of different volumes of glucose medium, Poseidonocella sp. culture and washed Poseidonocella sp. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp. Pos SS = Sterile supernatant, Washed Pos = Washed Poseidonocella sp., Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. G medium = Medium with glucose, G Media 1, 2, 3, and 4 represent 25 µL, 50 µL, 75 µL, and 100 µL of medium with glucose, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation. 3.3. Effect of the sterile supernatant from Poseidonocella sp. on biofilm formation in Nannochloropsis sp. cultures Biofilm formation of Nannochloropsis sp. was induced by incubation with the sterile supernatant of Poseidonocella sp. (figure 4). These results suggest a dose-dependent pattern when treated with different volumes of the sterile bacterial supernatant and reached the highest absorption after addition of 50 µL of the bacterial supernatant. Biofilm formation from sterile supernatant was lower than the biofilm formation from Poseidonocella cells, but considerably higher than negative controls.

To evaluate the thermal stability of the active substances in sterile supernatant and accordingly to obtain more information about the nature of involved molecule (large enzyme-like molecules or small signalling molecules), portions of sterile supernatant were subjected to two heat treatments i.e. 80°C for 1 hour as described by Windler et al. (2015) [207] and 90 °C for 1 hour and 30 minutes. Involved substances in the sterile supernatant were heat-resistant at 80 °C as biofilm density after heat treatment was the same as the effect of sterile supernatant (figure 5). Involved substances in the sterile supernatant were partially inactivated at 90 °C for 1 hour and 30 minutes. Under this

condition, biofilm density was reduced as compared to the biofilm density from sterile supernatant, but still higher than negative controls (figure 6).

N e g a tiv e P o s itiv e S te r ile c o n tr o ls c o n tr o ls s u p e rn a ta n ts 1.0 6 3 0 n m

0.8 7 5 0 n m

i s

n 0.6

a c

i 0.4

p O 0.2

0.0 a o s ia i a 1 2 3 1 2 3 n o d i s s s n P d e d s s s a e e o o o o o o m P P P P P P N M M + + + G o o o S S S G n n n S S S + n n n + + + o a a a o o o n n n n n N N N n n n a a a a N N N N

Figure 4 Biofilm formation assay – Incubation of Nannochloropsis sp. in the presence of different volumes of Poseidonocella sp. culture and sterile supernatant of Poseidonocella sp. Nanno = Nannochloropsis sp. Pos = Poseidonocella sp., G medium = Medium with the glucose, Pos SS = Sterile supernatant of Poseidonocella sp. Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation.

N e g a tiv e P o s itiv e S te r ile H e a te d s te r ile c o n tr o ls c o n tr o ls s u p e rn a ta n ts s u p e rn a ta n ts 0.5

6 3 0 n m

0.4 7 5 0 n m

i s

n 0.3

l a

c 0.2

p O 0.1

0.0

s a o ia i a a 1 2 3 1 2 3 1 2 3 n o d i i s s s n P d e d d s s s s s s a e e e o o o o o o o o o m P P P P P P P P P N M M M + + + G + o o o S S S S S S o G n n n S S S S S S n + + + + n n n n o o o H H H a o a a a n n n + + + n N N N n n n o o o N n a a a n n n a n n n N N N N a a a N N N

Figure 5 Biofilm formation assay - Incubation of Nannochloropsis sp. in the presence of different volumes of Poseidonocella sp. culture, sterile supernatant of Poseidonocella sp. and heated (80 °C for 1 hour) sterile supernatant of Poseidonocella sp. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp., G medium = Medium with glucose, Pos SS = Sterile supernatant of Poseidonocella sp., H SS = Heat treated sterile supernatant of Poseidonocella sp. Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation.

N e g a tiv e P o s itiv e S te r ile H e a te d s te r ile c o n tr o ls c o n tr o ls s u p e rn a ta n ts s u p e rn a ta n ts 1.0 6 3 0 n m

0.8 7 5 0 n m

i s

n 0.6

l a

c 0.4

p O 0.2

0.0 a o s a i 1 2 3 1 2 3 1 2 3 n o i d ia n P d e d s s s s s s s s s a e e o o o o o o o o o m P P P P P P P P P N M M + + + G o o o S S S S S S G S S S S S S n n n + + + + n n n o o o H H H o a a a n n n + + + n N N N n n n o o o n a a a n n n a n n n N N N N a a a N N N

Figure 6 Biofilm formation assay - Incubation of Nannochloropsis sp. in the presence of different volumes of bacteria culture, sterile supernatant of Poseidonocella and heated (90 °C for 1 hour and 30 minutes) sterile supernatant of Poseidonocella. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp., G medium = Medium with the glucose, Pos SS = Sterile supernatant of Poseidonocella sp., H SS = Heat treated sterile supernatant of Poseidonocella sp. Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation.

The effect of Poseidonocella sp. was confirmed with avoidance of the glucose effect using washed Poseidonocella sp. However, presence of glucose traces in the sterile supernatant cannot be discarded. In figure 4, the effect of the sterile supernatant is always more prominent than the lowest dosage of glucose (25 µL of culture medium on a total volume 250 µL). The lowest dosage of sterile supernatant was 10 µL and reported a similar effect on biofilm formation. Even unrealistically assuming no glucose consumption by Poseidonocella sp., a similar effect on biofilm formation has been reported, containing the sterile supernatant 2.5 times less glucose. We also found a stronger effect on biofilm formation when 50 µL of sterile supernatant were added to algal cells than when 50 µL of glucose containing medium were added. We cannot neglect the effect of glucose in biofilm formation but we can confirm the presence of other type of molecules secreted by Poseidonocella sp. that may lead to biofilm formation of Nannochloropsis sp.

3.4. Effect of solid phase extracts on biofilm formation in Nannochloropsis sp. cultures The eluate of sterile bacterial supernatant of both cartridges showed a substantial induction of bioﬁlm formation compared to the negative controls (figure 7 and 8). Nevertheless, the eluted fraction of Easy cartridges showed similar biofilm formation as the positive control with bacterial cells while eluate of HLB cartridges gave less biofilm than the positive control. However, Ft and Wp fractions of the sterile bacterial supernatant of HLB cartridges induced a stronger bioﬁlm formation as the positive

controls. Biofilm formations from Ft and Wp fractions of the Easy cartridge were comparatively low. A considerable bioﬁlm formation was not observed in most of the fractions of the SPE performed with algae growth medium except in eluate of HLB cartridge and Ft fraction and Wp fractions of Easy cartridge.

N e g a tiv e P o s itiv e S te r ile M e d ia B a c te ria M e d ia B a c te ria M e d ia B a c te ria c o n tr o ls c o n tr o ls s u p e rn a ta n ts e lu te e lu te F t F t W p W p 1.2

1.0 6 3 0 n m

y 7 5 0 n m t

i 0.8

l 0.6

i t

p 0.4 O

0.2

0.0 a a a 1 2 3 1 2 3 s ia i i i 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 o d d d d s s s s s s B B B P e e e e o o o o o o B B B B B B B B B B B B B B B P P P P P P L L L L L L L L L L L L L L L L L L M m M M H H H H H H H H H H + + + + S S S s H H H H H H H H G o G o o o S S S ia ia ia o s s a a a s s s a a a s s s n n n n o o i i i o o o i i i o o o + n n n + + + d d d P P P d d d d d d n o a a a o o o e e e e e e e P P P e e e P P P a n n n n t e e t t t p p p N n N N N n n n M M M u t t M M M F F F M M M a a a e e e l u u t t t a t t t E l l + + + p p p W W W N N N N u u u E E F F F o o o + + + l l l + + + + + + n n n W W W o o o o o o o n n n + + + n n n E E E n o o n n n a a a + + + n n n o o o n n n o o o a n n n n n N N N n n n a a a n n n a a a a a n n n N N N n n n N N N N N N a a a a a a N N N N N N

Figure 7 Biofilm formation assay - Incubation of Nannochloropsis sp. in the presence of different volumes of Poseidonocella sp. culture, sterile supernatant of Poseidonocella sp. and solid phase extracts of sterile supernatant of Poseidonocella sp. using HLB cartridges. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp., G medium = Medium with glucose, SS Pos = Sterile supernatant of Poseidonocella sp., Ft = Flow through, Wp = Washed phase. Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation.

N e g a tiv e P o s itiv e S te r ile M e d ia B a c te ria M e d ia B a c te ria M e d ia B a c te ria c o n tr o ls c o n tr o ls s u p e rn a ta n ts e lu te e lu te F t F t W p W p 1.2 630 nm 1.0

750 nm

y t

i 0.8

l 0.6

i t

p 0.4 O

0.2

0.0 a a a a 1 2 3 s i i i i 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 o d d d d s s s s s s y y y y y y y y y y y y y y y y y y P e e e e o o o o o o s s s s s s s s s s s s s s s s s s P P P a a a a a a a a a a a a a a a a a a M m M M P P P E E E + + + + S S S E E E E E E E E E E E E E E E G G o o o a s o n n n S S S i ia ia s s o a a a s s s a a a s s s n + + + + d o o i i i o o o i i i o o o n n n n o o o e d d P P P d d d d d d a o a a a e e e e e e P P P e e e P P P n N N N n n n M e e t t t t p p p N n n n n M M t t u M M M F F F M M M a a a a te e e u u l t t t t t l l E + + + p p p W W W N N N N lu u u E E F F F o o o + + + l l + + + + + + n n n W W W o o o E E E o o o o o o n n n + + + n n n + n n n n a a a o o o n n n o + + n n n n n n a a a n o o n n a a a a N N N n n n n n n a a n n n N N N a n n N N N N N N a a a a a N N N N N N

Figure 12 Biofilm formation assay - Incubation of Nannochloropsis sp. in the presence of different volumes of Poseidonocella sp. culture, sterile supernatant of Poseidonocella sp. and solid phase extracts of sterile supernatant of Poseidonocella sp. using Easy cartridges. Nanno = Nannochloropsis sp., Pos = Poseidonocella sp., G medium = Medium with glucose, SS Pos = Sterile supernatant of Poseidonocella sp., Ft = Flow through, Wp = Washed phase. Pos 1, Pos 2 and Pos 3 represent the 10 µL, 30 µL and 50 µL Poseidonocella sp. culture, respectively. Each bar represents the average of six replicates and error bars represent the standard deviation.

Since the presence of small molecules was confirmed, we opted for SPE to start their characterization. Characterization is important as identification of the biofilm-inducing substances is the first step to design a biofilm inhibiting strategy. We further separated the bacterial supernatant by using two types of cartridges i.e. HLB and Easy. SPE can be used for extraction of particular molecules by elution of concentrated compounds with a small volume of eluent [208]. SPE was performed with both bacterial supernatant and algae growth medium. Though eluted fractions of both cartridges showed an induction of biofilm formation, eluted fractions of Easy cartridges showed high biofilm formation and similar values as the positive control with bacterial cells (figure 7 and 8). The HLB cartridges have absorbent capacity for a wide range of molecules ranging from high retention of polar compounds to relative retention of non-polar compounds (Oasis, Kinesis). According to their manufacturers, the Easy cartridges are weak anion exchangers and more hydrophilic than conventional polystyrene-divinylbenzene. In present study, Easy cartridges were more effective in absorbing bioactive compounds from sterile bacterial supernatant. Ability of biofilm formation from both cartridges that use hydrophilic interactions suggested a polar character of the bioactive compounds. Ft and Wp fractions of the HLB cartridge induced stronger biofilm formation than the elute fractions of the Poseidonocella sp. cell culture (Figure 7), indicating that the biological activity had not been absorbed by the sorbent of HLB cartridge. Biofilm

formation values from Ft and Wp fractions of the Easy cartridge were comparatively low. Furthermore, no considerable amount of bioﬁlm formation was observed in most of the fractions of the SPE performed with algae growth medium except to the elute fraction of HLB cartridge and Ft and Wp fractions of Easy cartridge. Considering the observed low values, this could be due to the some residues attached to the sorbent that came out in each phase. Inadequate prewashing and conditioning steps could result in incomplete removal of residues such as fines and dust that can be created by physical abrasion of sorbent particles during shipping and storage [209]. 3.5. UHPLC measurements and LC-HR-MS data analysis UHPLC measurements and LC-HR-MS data analysis was performed on the Easy extracts, as these were more active. Unfortunately, we did not have three biological independent replicates as they are considered technical replicates for this purpose. For practical reasons, extracts were merged after elution and the decision of performing UHPLC was taken afterwards. Consequently, we count with one replicate of fresh medium against one replicate of sterile supernatant. Although no statistical conclusions can be achieved from this study, this is relevant information that can be used in future work towards the identification of the signals secreted by Poseidonocella sp.

As a general trend, the extracts corresponding to the sterile supernatant contained more compounds than the fresh medium in both positive and negative mode (figure 9). The flow-through did not contain anything remarkable, suggesting an efficient extraction method applied to the sterile supernatant.

Figure 9 PCAs of metabolomics sample. Positive modus on the left, negative mode on the right

In positive mode, no relevant molecule was identified. However, big peaks are present with a pattern closely related to polyethylenglycole (PEG) (figure 10 and 11). This is a

common polymer that could have been accidentally introduced during extraction procedure or due to long storage of methanol extracts in plastic culture flasks. However, the flasks that were used are made of polystyrene and PEG detection differs among samples being more prominent in the extracts from the sterile supernatant.

Molecular Weight RT Comp. 1 Comp. 2 Comp. 3 Comp. 4 Name Formula 700.38303 3.458 39 2.8083 1.7846 1.7418 1.7394 C26 H59 N10 O6 P3 475.2986 3.109 502 1.8052 1.0147 0.98836 0.98683 PEG n10 C20 H42 O11 502.29733 3.184 262 7.0345 4.8132 4.6883 4.6806 [Similar to: PEG n12; ΔMass: 44.0278 Da] C26 H48 O5 P2 634.37526 3.392 31 5.605 4.0006 3.8969 3.8903 [Similar to: PEG n15; ΔMass: 44.0285 Da] C32 H60 O8 P2 612.33159 3.346 317 3.371 1.8873 1.8384 1.8354 C22 H51 N10 O4 P3 480.25438 3.098 252 3.7349 2.2863 2.2304 2.227 Oxohongdenafil C25 H32 N6 O4 459.27436 3.101 57 3.6449 2.5734 2.5073 2.5031 524.28062 3.179 304 3.6364 2.1976 2.1424 2.1389 C14 H41 N10 O9 P 519.32494 3.187 55 1.7161 0.97024 0.94504 0.94385 PEG n11 C22 H46 O12 568.30563 3.274 284 3.5293 2.0075 1.9567 1.9537 C24 H51 N4 O5 P3 458.27104 3.101 106 7.4668 5.1591 5.0254 5.0171 [Similar to: PEG n11; ΔMass: 44.0279 Da] C24 H44 O4 P2 436.22784 3.007 336 3.8633 2.378 2.3167 2.3129 Flurandrenolide C24 H33 F O6 414.24481 3.011 128 7.1994 4.9471 4.8186 4.8105 [Similar to: PEG n10; ΔMass: 44.0279 Da] C18 H38 O10 546.32345 3.258 120 6.728 4.724 4.6017 4.5939 [Similar to: PEG n13; ΔMass: 44.0279 Da] C28 H52 O6 P2 262.08856 2.247 269 2.2484 3.9037 3.813 3.8071 [Similar to: 4-(2-Furyl)-6-(methylthio)-2-tetrahydro-1H-pyrrol-1-ylpyrimidine-5-carbonitrile;C12 H14 N4 O S ΔMass: 24.0003 Da] 470.35751 8.629 299 3.6369 6.7532 6.5907 6.5817 C23 H46 N6 O4 166.08555 0.752 11 3.1245 5.6274 5.4917 5.4843 JJO1TS4A8Y C7 H10 N4 O 346.30778 8.459 301 2.013 3.7306 3.7128 3.7243 [Similar to: Stearic acid; ΔMass: -62.0363 Da] C20 H42 O4 390.33372 8.039 81 1.7116 3.1683 3.1927 3.2112 C17 H43 N8 P 127.01204 0.79 103 3.1882 4.0068 3.9108 3.9155 239.10221 2.304 238 2.3531 2.9975 2.9248 2.9202 C9 H23 N P2 S 678.23076 0.72 399 1.7179 2.3976 2.9925 2.9966 [Similar to: Colchicine; ΔMass: -279.0626 Da] C23 H47 N6 O9 P S3 261.08402 0.854 185 1.992 2.3223 2.2757 2.3536 Flumequine C14 H12 F N O3 239.10204 0.816 412 2.8092 2.9551 2.9137 2.9137 7,8-Dihydrobiopterin C9 H13 N5 O3 974.36525 0.727 92 1.664 2.1469 2.1405 2.1538 [Similar to: Colchicoside; ΔMass: -427.1599 Da]C33 H73 N10 O9 P3 S4 Figure 10 Top features in positive mode based on Variable Importance in Projection (VIP) scores from PLS-DA. In read the putative PEG compounds.

Figure 11 Chromatogram and MS spectrum of the measured polymer. The space between the m/z peaks is 44.02629 m/z, which corresponds to the formula C2H4O. This is the formula of the monomer of PEG.

In negative mode, more biologically interesting compounds were found and we focused on a putative dihydroquinoline carboxylic acid. This was identified during MS/MS experiments using the program Sirius, which is indicating the best match using MS and MS/MS high resolution data a making fragmentation trees out of it. We found 2 peaks corresponding to the molecular weight of 189.04215 Da (RT=2.7 and 3.0 minutes), and all of them showed the same fragmentation (figure 12 and 13). This could be an indication of two dihydroquinoline isomers. The presence of two peaks with the same molecular weight could be an indication of tautomerism. However, these peaks are well separated making this hypothesis less probable. Other explanation could be the presence of two isomers. Since this is a library match, it is not possible to elucidate molecule structure. Therefore, these two isomers are putative molecules. To validate this hypothesis more replicates and a bioassay guided fractionation is needed to confirm if this compound induce biofilm formation on Nannochloropsis sp.

Molecula weight RT f.value p.value -LOG10(p) FDR Name Formula 970.39173 3.522 51 6433.7 7.25E-14 13.14 3.25E-12 C37H66N10O14S3 300.20774 7.866 206 4876 2.20E-13 12.658 6.81E-12 Tretinoin C20 H28 O2 272.16137 4.521 40 2928.2 1.69E-12 11.773 2.75E-11 C14 H24 O5 606.31946 3.153 213 3144 1.27E-12 11.897 2.44E-11 C24 H50 N2 O15 287.21614 6.666 63 4392.4 3.33E-13 12.477 8.48E-12 354.18797 6.668 46 3686.3 6.72E-13 12.173 1.35E-11 C11 H27 Cl N8 O3 142.05284 3.386 27 7513.9 3.90E-14 13.409 1.96E-12 C9 H6 N2 650.33634 3.448 126 20293 7.33E-16 15.135 9.85E-14 C24 H46 N10 O11 571.3317 3.535 164 12470 5.14E-15 14.289 5.18E-13 C25 H45 N7 O8 239.15913 5.947 116 49857 2.01E-17 16.696 8.11E-15 C7 H21 N5 O4 399.20067 5.516 172 36718 6.84E-17 16.165 1.38E-14 C18H29N3O7 610.16345 2.55 179 2940.7 1.66E-12 11.781 2.75E-11 724.41056 3.54 96 8147 2.82E-14 13.55 1.89E-12 C35 H64 O13 S 353.98962 7.899 180 4051.2 4.61E-13 12.337 9.77E-12 C16 H9 N2 O2 P3 528.3245 3.468 223 4381.3 3.37E-13 12.473 8.48E-12 C23H48N2O11 406.25331 6.121 273 5277.7 1.60E-13 12.796 5.86E-12 C16 H34 N6 O6 285.16775 3.159 347 7821.4 3.32E-14 13.479 1.91E-12 Leu-Gly-Pro C13 H23 N3 O4 474.24087 2.911 263 4651.8 2.65E-13 12.577 7.63E-12 [Similar to: Prop-2-ynyl N-[6-methoxy-5-(trifluoromethyl)pyridin-3-yl]carbamate;C20 H45 O6 P3 ΔMass: -200.1843 Da] 1014.41761 3.568 218 3036.9 1.46E-12 11.836 2.67E-11 C46H70N4O17S2 324.11503 5.367 123 4101.3 4.39E-13 12.358 9.77E-12 acetohexamide C15 H20 N2 O4 S 559.3212 3.957 26 9928.2 1.28E-14 13.893 1.03E-12 C26 H41 N9 O5 302.20848 4.227 130 5473.7 1.38E-13 12.859 5.57E-12 8-Hydroxyhexadecanedioic acid C16 H30 O5 770.46774 3.453 395 4979.7 2.02E-13 12.695 6.78E-12 C34H62N10O10 689.36989 3.823 341 4065.1 4.54E-13 12.343 9.77E-12 C26 H62 N9 O2 P3 S2 189.04215 2.893 338 2918.8 1.71E-12 11.768 2.75E-11 2-oxo-1_2-dihydroquinoline-4-carboxylate C10 H7 N O3 Figure 12 Top 25 hits (one-way Anova, α=0.05, Fisher LSD Post-Hoc analysis).

Figure 13 Chromatographic peak of putative dihydroquinoline carboxylic acid isomers. Molecular peak (M-H)- at 188.03485 m/z.

4. General discussion The result of the present study validates once more the findings on previous chapter. All treatments with Poseidonocella sp. cells caused a higher biofilm density than the negative controls (figure 2). Previous study reported that alphaproteobacteria was among the most abundant classes that are involved in biofilm formation in the coastal waters [210]. The formation and proliferation of a biofilm are influenced by many factors including several complex regulatory systems that were shown to be involved in different stages of the biofilm life-cycle. Many Gram-negative bacteria possess several systems including the quorum sensing systems [211,212]. Signaling molecules participate in the initial stages of biofilm formation, attachment, proliferation and biofilm differentiation [213]. Self-excreted extracellular polymeric substances (EPS) influence biofilm formation, increasing optical density with the biofilm thickness [214,215]. Therefore, the biofilm- forming ability can be determined through optical density of developed biofilms after a specific period [215]. Increasing initial concentrations of Poseidonocella sp. in the co- cultures have increased the biofilm density at both wavelengths of 630 nm and 750 nm. These high biofilm densities at 630 nm and 750 nm indicate that Poseidonocella sp. can form thick biofilm together with high algae cell densities. Greater values in absorbance at 630 nm can be correlated to an increase of microalgae pigments in the sample, especially chlorophyll c [216]. As more biofilm was observed at 630 nm than at 750 nm, suggesting the ability of this bacterium to attach a great number of algae cells.

The relative importance of biological characteristics to biofilm formation was in the order of exopolysaccharide > flagella > AHLs > extracellular protein > swarming motility [217]. To date, no studies on Poseidonocella sp. have been conducted to study the biological characteristics that induce the biofilm formation. Further studies on characteristics of this bacterium such as production of signaling molecules, exopolysaccharide and extracellular proteins are necessary to understand the factors behind this higher biofilm formation capacity. Unexpectedly, Nannochloropsis sp. with medium containing glucose showed a considerable biofilm density compared to other negative controls (figure 2). The effect of glucose on biofilm formation was investigated by using an increasing gradient of glucose doses with Nannochloropsis sp. in the bioassay (figure 3). This experimental result showed that increasing volumes of glucose gave an increasing gradient of biofilm density. However, the effect of Poseidonocella sp. cell culture was further demonstrated by using washed Poseidonocella sp. cells suspended in glucose-free medium. Washed Poseidonocella sp. cells gave less biofilm density than Poseidonocella sp. cell culture containing glucose. This might be explained by growth promotion by glucose but also due to the stress on the cells after multiple centrifugations steps. However, the biofilm density with washed Poseidonocella sp. was considerably higher than the negative controls (figure 3). The effect of glucose on biofilm formation might be due to the presence of other bacteria in Nannochloropsis sp. cultures which were induced by the glucose. Although clean cultures of Nannochloropsis sp. were used, they might contain unharmful bacteria to algae, as they were maintained longer period after purchasing from CCAP. Therefore, glucose addition might develop this bacterial population and consequently induce biofilm formation. Bacterial biofilm formation is influenced by environmental signals. Nutrient availability is a common environmental signal that affects biofilm formation in different bacterial species [218]. A different study reported that flocculation of Chlorococcum nivale, Chlorococcum ellipsoideum and Scenedesmus sp. was enhanced at low pH resulting in disruption of the dispersing stability of cells and subsequent flocculation of cells [219]. Glucose utilization in microalgae culture of Chlorella resulted in a decrease of pH [220]. The hypothesis of biofilm formed due to pH drop because the presence of glucose is unlikely since medium contained a pH buffer. However, pH was not monitored during the experiment. We consider more plausible the hypothesis of the variation on microbial community although we still suggest to measure pH during the experiment in future occasions. These results conclusively demonstrated that bacterial cells and some bacterial extracellular metabolites play a major role in biofilm formation in co-culture of Nannochloropsis sp. and Poseidonocella sp. Still, bacterial cells might play a more

100

profound role in biofilm formation than bacterial extracellular metabolites alone. Windler et al. (2015) [221], have stated that filtered sterile supernatant of Bacteroidetes strain 32 can induce biofilm formation by incubation with an axenic culture of Achnanthidium minutissimum. Our results are in a good agreement with their findings, as A. minutissimum cultures showed a dose-dependent pattern when treated with different volumes of the sterile bacterial supernatant. The development, maturation and breakdown of many algal biofilms are regulated by signaling molecules of bacteria which help to control biofilm formation, development, maturation and dispersal [198]. To test if substances in sterile supernatant responsible for biofilm formation were heat- stable or heat-sensitive, portions were heat-treated at 80 °C for 1 hour as described by Windler et al. (2015) and at 90 °C for 1 hour and 30 minutes. Furthermore, the heat treatment was expected to have an effect on bacterial exometabolome and not on glucose, as decomposition of glucose starts at 220 °C and reacted completely at 280 °C [222]. Substances in the sterile supernatant were heat-resistant at 80 °C as biofilm formation was not abolished after heat treatment at this temperature while substances were partially inactivated at 90 °C for 1 hour and 30 (figure 5 and 6). A different study reported enzymatic activity of cell-free lysate of Bacillus firmus PT18 and Enterobacter asburiae PT39 could be confirmed by complete inactivation by heat treatment at 95 °C for 10 min [223]. Gutiérrez‐Barranquero et al. (2017) have reported in their study on disruption of AHL specific signaling and virulence in clinical pathogens by marine sponge bacteria that the effect of sterile bacterial supernatant on pathogens was not inactivated at 80 °C for 1 hour and 95 °C for 30 minutes suggesting thus that their target could be a small thermostable molecule rather than an extracellular enzyme [224]. As substances in sterile supernatant were heat resistant at 80 °C and only partially inactivated by strong heat treatment (90 °C for 1 hour and 30 minutes), active substance in the exometabolomes of Poseidonocella sp. could be small molecules instead of extracellular enzymes. Next steps should be taken towards the validation of the presence of dihydroquinoline derivatives in the sterile supernatant from Poseidonocella sp. and the confirmation of its role on biofilm formation. This could be achieved by bioassay guided fractionation but also by directly testing the identified compound on a biofilm formation assay if commercially available. Despite the limitations of this study, the finding of putative dihydroquinoline carboxylic acid isomers is an interesting finding as this belongs to a group of molecules related to different types of bacterial communication including quorum sensing and biofilm formation[105,225]. 5. General conclusions A study was developed to investigate the chemical signals involved in biofilm formation in algae cultures. Present results further confirmed that Poseidonocella sp. can induce biofilm formation in Nannochloropsis sp. culture. Glucose was found to be a biofilm

101

inducer in our Nannochloropsis sp. culture as demonstrated by the high absorbance values found in bioassays. However, effect of Poseidonocella sp. on biofilm formation was more prominent than the effect of glucose. Sterile supernatant of Poseidonocella sp. induced biofilm formation, proving the role of the exometabolomes of Poseidonocella sp. in biofilm formation. Bioactive compounds in sterile supernatant were heat stable at 80 °C for 30 minutes and partially inactivated at 95 °C for 1 hour and 30 minutes. The extracted bioactive compounds could be small molecules instead of extracellular enzymes. In SPE of sterile supernatant, elutes of both HLB and Easy cartridges that use hydrophilic interactions induced biofilm formation and therefore, polar character of bioactive compounds in sterile supernatant was suggested. Easy cartridges were more effective in absorbing bioactive compounds from sterile supernatant than HLB cartridges. Preliminary UHPLC results suggest that these bioactive compounds might be isomers of dihydroquinoline carboxylic acids. However, more replicates and bioassay guided fractionation would be needed to confirm this hypothesis.

6. Acknowledgements This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642575. This work makes use of resources and facilities provided by UGent as part of the Belgian contribution to EMBRC-ERIC. Special thanks to Nuwanthi Samarasinghe whose MSc thesis contributed to the elaboration of this chapter. The authors would like to thank the Institute of Analytical Chemistry of FSU Jena, in particular Emilio Cirri for his contribution to the experimental design and the experimental work and discussion related to the UHPLC part.

102

Chapter V

Towards microbiome engineering of industrial microalgae cultivation: effects of non-native bacteria on Nannochloropsis sp. growth, metabolome and biofouling

Emilio Cirri1#, Javier B. Giraldo2, 3#, Sandesh Neupane1,4, Anne Willems5, Sven Mangelinckx6, Luc Roef2, Wim Vyverman3 and Georg Pohnert1

1 Institute of Inorganic and Analytical Chemistry, Friedrich Schiller University Jena, D- 07743 Jena, Germany

2 Proviron Holding NV, Georges Gilliotstraat 60, B- 2620 Hemiksem, Belgium

3 Laboratory of Protistology and Aquatic Ecology, Department of Biology, Ghent University, Krijgslaan 281-S8, B-9000 Ghent, Belgium

4 Pharmaceutical Institute, Christian-Albrechts University of Kiel, Gutenbergstr. 76, D- 24118 Kiel, Germany

5 Laboratory of Microbiology, Department of Biochemistry and Microbiology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium

6 Department of Green Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium

# These authors contributed equally to the work.

Cirri, E., Giraldo, J. B., Neupane, S., Willems, A., Mangelinckx, Roef, L., Vyverman, W., Pohnert, G. Towards microbiome engineering of industrial microalgae cultivation: effects of non- native bacteria on Nannochloropsis sp. growth, metabolome and biofouling, in preparation.

103

Abstract Interactions between bacteria and microalgae shape aquatic communities and affect large-scale microalgae cultivation. Bacterial presence constitutes a parameter that is increasingly considered in algal farming in the context of growth control and product quality. In the present study, we transplanted bacteria associated with the biofilm- inhabiting pennate diatom Seminavis robusta to the commercially relevant microalga Nannochloropsis sp. in order to compare the effects of these bacteria on their hosts. Contrary to its effects on S. robusta, Maribacter sp. significantly sustained Nannochloropsis sp. growth, while Roseovarius sp. only affected growth of these algae at stationary phase. The specific effects of the bacteria were reflected in changes in the endometabolome of the microalgae. In the presence of either Maribacter sp. or Roseovarius sp., S. robusta contained more disaccharides and less terpenes. Nannochloropsis sp. extracts had higher concentrations of spermidine and aminobutyric acid when bacteria were present, while many α-amino acids and α-tocopherol were less abundant. Bacteria from S. robusta did not induce biofilm formation but enhanced the biofilm development in a tripartite microbial community containing in addition Poseidonocella sp., a bacterium isolated from bioreactor samples. This study reveals the potential of microbiome engineering in closed cultivation conditions of Nannochloropsis sp. with promise for the industrial application of Maribacter sp.

Keywords Metabolomics, microalgae, biofilm formation, inter-kingdom communication

Highlights  Bacteria isolated from the benthic marine diatom Seminavis robusta affect Nannochloropsis sp. growth and biochemical composition  Species specific effect of bacteria on algal growth and metabolome are observed  Biofilm formation can be modulated in complex engineered consortia  Maribacter sp. sustains Nannochloropsis growth and is a promising candidate for industrial applications

104

1. Introduction Interactions between microalgae and bacteria range from mutualism [226] to pathogenicity [227]. These interspecies cross-talks are fundamental for the co-evolution of microalgae and heterotrophic bacteria [86] and shape their population dynamics in the oceans [228,229]. Moreover, these interactions influence primary production and biogeochemical cycles [230,231], not only in the pelagic zone [232] but also in benthic environment [233], due to the formation of highly productive epipelic biofilms mainly constituted by diatoms and bacteria [234]. In recent years, the study of inter-kingdom communication between microalgae and bacteria has moved from a predominantly ecological interest to applied research for industrial purposes [235]. Multispecies consortia have been fruitfully employed in several applied fields, including bioremediation [236], aquaculture and microalgae cultivation in photobioreactors [127,237] for the production of high valuable compounds [89] and biofuels [238]. Mixed biofilms formed by these organisms can be successfully exploited for technological purposes, for example in wastewater treatment [239]. On the other hand, biofilm formation on technical surfaces, a process called biofouling, has a negative economic impact during industrial operations [56,240], since huge investments are needed to reduce and control it. Anti-fouling research has directed its efforts not only to develop products to counteract biofilm development [241,242] but also to use natural interspecific interaction mechanisms to inhibit fouling using ecologically robust approaches [243]. This research is particularly important in industrial microalgae cultivation in closed photobioreactors, which is typically conducted in non-axenic conditions. Alteration in the algal microbiome may induce biofilm growth on solid surfaces of the reactors which needs to be avoided since it may diminish productivity due to reduction of light penetration and increasing contamination [55]. In this study, we question if bacteria trigger similar effects in different microalgae and can be used universally for improving algae cultivation. We use Nannochloropsis sp., a commercially relevant microalga as a target organism [69,244]. Nannochloropsis sp. is an industrially-promising candidate due to its high protein and lipid content – particularly eicosapentaneoic acid (EPA) – [110] which makes it important to be used for aquaculture high quality feed [245], as well as a promising candidate for replacing fish oil as eicosapentaenoic acid (EPA) source [123]. We studied the effect of bacteria isolated from another Stramenopile, the marine benthic pennate diatom Seminavis robusta, on the growth, metabolome and biofilm formation of Nannochloropsis sp.. S. robusta is a biofilm forming diatom [246] and its life cycle and sexual pheromone chemistry are well studied [247–249]. It was recently reported that bacteria associated to S. robusta have different effects on its sexual development and on pheromone concentration, with Maribacter sp. inducing a reduction in diatom’s mating and Roseovarius sp. enhancing it [250]. These bacteria also influence the vegetative

105

propagation in a contrasting manner. Over a seven days co-cultivation Roseovarius sp. strongly sustained diatom growth while Maribacter sp. reduced it [250]. As these two bacterial strains influence the proliferation of S. robusta, we assessed their potential effects on the growth, endometabolome and biofouling in co-cultivation with Nannochloropsis sp. and compared this with their native host. We found that bacteria have a specific effect on the respective microalgae, triggering contrasting growth responses and metabolic changes in the algae dependent on the bacterial community composition. Tailored supplementation of bacteria can thus specifically be used to modulate algae in reactors with respect to growth, biofilm formation and value product generation. It can however not be assumed that effects are universally triggered in different microalgal species. Our results demonstrate the importance and need of studying and manipulating the microbiome associated to commercially relevant algae and offers the opportunity to use microbiome engineering as a tool to control microalgal cultivation and biofilm formation in closed systems.

2. Materials and methods

2.1. Growth of Seminavis robusta in co-cultivation with isolated bacteria S. robusta strain 85A (MT+) (BCCM: DCG 0105) with cell sizes below the sexual size threshold (SST= 22.4 μm) was obtained from the diatom culture collection of the Belgian Coordinated Collection of Micro-organisms (BCCM/DCG, http://bccm.belspo.be). Diatoms were inoculated in 250 mL f/2 medium in 250 cm² culture flasks (TC Flask T25, Sarstedt®, Nümbrecht, Germany) at an initial cell density of 104 cells/mL. Axenic cultures were prepared as explained in Cirri et al. (2018) [250]. Roseovarius sp. and Maribacter sp. were isolated from non-axenic cultures [250]. The bacteria were grown in 30 mL of marine broth medium for three days. Each bacterial strain culture was transferred into a 2 mL Eppendorf® tube, centrifuged for 3 min at 6,000 × g, washed three times with minimal medium (f/2 medium with 5 g/L glucose, 5 mL/L glycerol, 1.5 g/L NH4NO3) and transferred to 30 mL of minimal medium. After ten days of growth under shaking at room temperature, bacteria were inoculated with the diatoms (initial bacterial cell density = 104 cells/mL, initial diatom cell density = 103 cells/mL). Each treatment as well as an axenic control was repeated in triplicate. Diatom growth was followed for 7 days by counting the number of cells using a Leica DM IL LED inverted light microscope with a Leica DFC 280 camera system (Heerbrugg, Switzerland). For each culture flask, 5 pictures were taken and S. robusta cells were counted using ImageJ software [251]. Due to the attached growth of S. robusta, cell counts with flow cytometry and DNA staining was not possible, and therefore cells were counted under the microscope as described previously [247,250].

106

2.2. Nannochloropsis sp. cultivation and co-cultivation with isolated bacteria from S. robusta The Nannochloropsis sp., strain CCAP211/78 was obtained from the Culture Collection of Algae and Protozoa (CCAP, United Kingdom). Although this culture was not axenic, its bacterial population was never an impediment to grow in the production process using this alga strain. It was cultivated in 100 mL of saline medium [131] in 250 cm² culture flasks (Greiner bio-one) in a cultivation chamber at 20 °C. The irradiation was 70 µmol ∙ m-2 ∙ s-1 and applied from a cool fluorescent light source in day-night cycles of 16:8 h. The cultures were agitated twice a day to avoid precipitation of algal cells. Poseidonocella sp. MH974276 was isolated as described in Giraldo et al. (2019) [163]. It was grown in 50 mL saline medium enriched with 20 mM of glucose in 100 mL flasks (Greiner bio-one). Roseovarius sp. strain R-50241 and Maribacter sp. strain R-50239 isolated from S. robusta were grown in 30 mL of marine broth medium for three days. Afterwards, the cultures were transferred to 50 mL Falcon tubes, centrifuged for 3 min at 6,000 × g, washed three times with 10 mL minimal medium (saline medium with 5 g/L glucose, 5 mL/L glycerol, 1.5 g/L NH4NO3) and transferred to 100 mL of minimal medium in 250 cm² culture flasks (TC Flask T25, Sarstedt®, Nümbrecht, Germany). After ten days of growth at room temperature, bacteria reached an OD600 of ≈ 0.3-0.4 and were ready for inoculation.

Nannochloropsis sp. (initial OD630 = 0.62 ± 0.05) was inoculated with Poseidonocella,

Roseovarius sp. and Maribacter sp. to reach a final OD600 = 0.1 for each strain. Apart from the control with Nannochloropsis sp., five co-cultivations were established: Nannochloropsis sp. + Poseidonocella sp., Nannochloropsis sp. + Roseovarius sp., Nannochloropsis sp. + Maribacter sp., Nannochloropsis sp. + Poseidonocella sp. + Maribacter sp. (N+P+M) and Nannochloropsis sp. + Poseidonocella sp. + Roseovarius sp. (N+P+R). Each treatment was performed in triplicates.

The growth of Nannochloropsis sp. was estimated by daily measurements of OD630 – maximum chlorophyll concentration [252] – and OD750 – total algae biomass [206] – during 7 days using a Cary 50 UV-Visible spectrophotometer (Agilent Technologies). Although both wavelengths were measured in all Nannochloropsis sp. cultures, only

OD630 is presented as the other values followed the same trend. Estimated growth curves at OD750 may be found in Figure SF1. When cultures reached an OD >1, samples were diluted and the OD values were back-calculated in respect to the Lambert-Beer law. The growth monitored at both wavelengths was analyzed by means of a two-way ANOVA followed by Bonferroni’s multiple comparisons tests that were conducted using

107

GraphPad Prism version 7.00 for Windows (GraphPad Software, La Jolla California USA, www.graphpad.com).

2.3. Biofilm formation assay Biofilm formation assays were based on the protocol described by Leinweber et al. [193]. They were performed in flat bottomed 96-well plates (Sarstedt) that were inoculated with Nannochloropsis sp. and different bacterial cultures or their filter- sterilized spent medium (see Figure 1 for the treatment summary). Each well was inoculated with 150 µL of algae suspension with an OD630 = 1.50 ± 0.05. Poseidonocella sp. was inoculated at a final OD600 of 0.1, Roseovarius sp. and Maribacter sp. were inoculated at final OD600 of 0.1 and 0.01. Spent medium of each bacterial strain was obtained after centrifuging 5 mL of each culture at 5000 rpm for 5 min. Supernatants were separated and filter-sterilized (0.2 µm Supor® Membrane, PALL® Acrodisc® 32 mm). Sterile supernatant was added to well plates to reach for each strain an equivalent

OD600 as mentioned before. Volumes were adjusted to 250 µL with glucose free f/2 medium. In total, 20 treatments were tested in triplicates (see FIGURE 1 for experimental setup summary). Controls included empty wells (blank), wells filled with fresh medium (medium), Nannochloropsis culture with fresh medium (Nanno control) and with minimal medium (Nanno + MinMed). Initial absorbance of well plates was measured with a Cytation 3TM plate reader (BioTek® instruments, Winooski, VT) using Gen5TM software. For each well, average absorbance at 630 nm and 750 nm was calculated from nine measurement points. Afterwards, plates were incubated for 10 days at 20 °C, with a light irradiation of 30 µmol ∙ m-2 ∙ s-1 in day-night cycles of 16:8 h. After the incubation, the medium and non- adherent cells were removed by manual pipetting and biofilm absorbance was measured at 630 nm and 750 nm using the same parameter as reported above. Biofilm absorbance measurements were compared by means of a one-way ANOVA followed by a Bonferroni post-hoc test for multiple comparisons (p<0.05) using GraphPad Prism version 7.00 for Windows (GraphPad Software, La Jolla California USA, www.graphpad.com).

108

Figure 1 Experimental design for the monitoring of biofilm formation of Nannochloropsis sp. in presence of different bacteria.

2.4. Sample preparation for endometabolomics of S. robusta After seven days of co-cultivation, S. robusta cells were harvested 3 h after the onset of light to obtain extracts from the intra- and extracellular metabolites. Therefore, 250 mL of each culture were filtered through GF/C filters (ø 47 mm) at 650 mbar on NalgeneTM reusable bottle top filter units (Thermo Fisher Scientific, Bremen, Germany) connected to sterile 500 mL Duran® bottles (Schott, Jena, Germany). Two filtrations were done in parallel; each filtration took on average 30 min. Cultures that were not extracted immediately were kept under low light and low temperature regime (4 °C) to slow down metabolic processes. After filtration, the still wet filters were immediately transferred to 25 mL glass beakers. Cells were re-suspended in 2 mL extraction mix (methanol : ethanol : chloroform, 1:3:1, v: v: v) and transferred to 4 mL glass vials [253]. Samples were vortexed for 60 s and kept at -20 °C until further processing and analysis.

2.5. Sample preparation for endometabolomics of Nannochloropsis sp. After seven days of growths, (co-)cultures were extracted. 45 mL of each culture were transferred to 50 mL Falcon tubes and centrifuged in a Sigma® 4K15 centrifuge at 3500 rcf for 15 min. Afterwards, the supernatant was discarded and the cell pellet was immediately frozen in liquid nitrogen and stored at -80 °C until extraction. To extract the endometabolites, samples were thawed, vortexed and cells were re-suspended in 2 mL extraction mix (methanol : ethanol : chloroform, 1:3:1, v: v: v) and transferred to 4 mL glass vials. Samples were vortexed for 60 s and kept at -20 °C until further processing and analysis.

109

2.6. Endometabolite extraction, derivatization and GC-HRMS measurements Extracts were thawed and vortexed. 1 mL of S. robusta sample / 0.7 mL of Nannochloropsis sample were transferred into 1.5 mL centrifuge tubes (Eppendorf, Germany) and 20 nmol of aqueous ribitol (>99%, Sigma-Aldrich) were added as internal standard (IS). Suspensions were treated for 10 min in an ultrasonic bath (VWRTM Ultrasonic Cleaner), centrifuged (15 min, 30.000 rcf, 4 °C), and supernatants were transferred to 1.5 mL glass vials. Samples were evaporated under a stream of nitrogen and dried under vacuum overnight to eliminate the excess of water. Samples were then derivatized with methoxyamine and MSTFA [254]. Quality (QC) samples were prepared by pooling 5 µl from each sample and blank samples in one clean vial. Samples were run in random order with QC every 5 samples for S. robusta and every 7 samples for Nannochloropsis on a Thermo Scientific™ GC Ultra coupled to a QExactive™ Orbitrap (Thermo Fisher Scientific, Bremen, Germany) equipped with a DB-5ms column (30 m, 0.25 mm internal diameter, 0.25 µm film thickness, 10 m Duraguard pre-column, Phenomenex®). Helium 5.0 (Linde AG, Pullach, Germany) was used as carrier gas with a constant flow of 1 ml min-1. 1 µL of sample was injected in split mode (split ratio 5 for S. robusta, 100 for Nannochloropsis). The initial oven temperature of 80 °C was held for 2 min, ramped to 120 °C at 2 °C min-1, then to 250 °C at 5 °C min-1 and finally to 320 °C at 10 °C min-1. The final temperature was held for 2 min. The injector temperature was kept at 250 °C during the entire run. All transfer lines were heated at 280 °C, the ion source was at 300 °C. Electron-impact ionization was carried out at 70 eV. The scanned mass range was between 50 m/z and 600 m/z, at a resolution m/Δm 120,000 full width at half maximum (FWHM) (m/z 200), with automatic gain control (ACG) target 3 × 106, and a maximum injection time (IT) of 200 ms. Data were acquired and processed with the software XcaliburTM version 4.0.27.19 (Thermo Fisher Scientific, Bremen, Germany).

2.7. GC-HRMS data analysis XcaliburTM raw data files were converted to mzXML with the software ProteoWizard v. 3.0.11799 [255] and then uploaded on Workflow4Metabolomics v 3.0 (http://workflow4metabolomics.org/the-galaxy-environment/, Giacomoni et al., (2015) [256]), an online platform implemented in the Galaxy Environment [257] for pre- processing, analysis and statistics of GC-MS, LC-MS and NMR metabolomics. The GC-MS analysis is based on the R-package metaMS [258]. The important values for features extraction are the following: Full Width at Half Maximum (FWHM) 5, retention time (RT) difference 0.05 min, minimum features 5, similarity threshold 0.7, minimum class fraction 0.1, and minimum class size 1.

110

After the analysis, a table with exact masses, retention times and chromatographic area for each sample was exported for further processing with an in-house R-script available on request (http://www.R-project.org/, R Core Team, 2017). Features were normalized on the internal standard, medium features were deleted from the table and data were then filtered based on QCs coefficient of variation (CV): only features with CV < 35 % were retained [259]. Finally, data were normalized on cell densities. The obtained .csv table was used to perform statistical analysis with MetaboAnalystR [260]. The data were Pareto scaled to achieve a normal distribution of both features and samples. PCA was performed to detect grouping and outliers in the samples. Significant features were selected by inspection of PCAs loading plots and the results of two-way ANOVA analysis (adjusted p-value (FDR) cutoff = 0.05, FisherLSD post-hoc analysis) were visualized on heatmaps (distance measure = euclidean, clustering algorithm = Ward). After statistical analysis, significant features were manually checked in the chromatogram and identified using the software MS Search (version 2.0 d, NIST) by comparing the spectra with the following libraries: NIST 11 library version (mainlib, replib, nist_ri), Golm Metabolome Database libraries T_MSRI_ID (http://csbdb.de/csbdb/dload/dl_msri.html; 2004) and GMD_20111121_VAR5_ALK_MSP (http://gmd.mpimp-golm.mpg.de/download/; 2011), and an in-house library (175 compounds from several metabolite classes including algal extracts of Skeletonema marinoi, [254]. Mass spectra were considered identical with a reverse match factor (R.Match) >800, tagged with "?" if the RM was between 800 and 700, with "??" if the RM was between 700 and 600 and “???” if below 600. Metabolites with RM < 800 were indicated as putative.

3. Results

3.1. Growth of Nannochloropsis sp. in co-cultivation with bacteria isolated from S. robusta In the study of Cirri et al. (2018)[250], different bacteria from S. robusta exhibited a different effect on the growth of this diatom: Roseovarius sp. had a strong enhancing effect on algal proliferation, while Maribacter sp. only had a limited positive influence. To verify if this effect is more universal we tested the impact of these bacteria on the growth and metabolism of another microalga belonging to the Stramenopiles phylum, the commercially relevant Nannochloropsis sp. that was co-cultivated with the different bacteria for seven days (Figure 2 for p-values). Algal growth (initial OD630 = 0.62±0.05) was monitored daily by measuring OD630 (chlorophyll maximum) and OD750 (standard measurement for algal biomass). Both measurements show similar trends (Figure 2 and

111

Supplementary Figure 1), therefore only OD630 measurement will be further discussed. Despite all bacteria significantly supported Nannochloropsis sp. growth at the end of the seven days cultivation period (two way ANOVA interaction p-value < 0.0001 in all conditions), only Maribacter sp. was considered to enhance Nannochloropsis sp. growth as the effect is more prominent and occurred from day 3. The extent of this positive effect on algal proliferation was species-specific. Poseidonocella sp., an alphaproteobacterium isolated from Nannochloropsis sp. capable of inducing biofilm formation [163], significantly increased Nannochloropsis sp. growth only after 7 days of co-cultivation (p < 0.0001 on the seventh day of co-incubation in comparison to the control) (Figure 2 b and d, orange line). Bacteria isolated from S. robusta had different effects on Nannochloropsis sp. when compared to their influence on S. robusta [250] (Figure 2). Roseovarius sp., which supported S. robusta growth [250], increased Nannochloropsis growth significantly only at day 7 (p=0.001 at day 7) similar to Poseidonocella. Also the combination of the two bacteria did not significantly enhance algal growth compared to the single species treatments. In contrast, Maribacter sp. strongly and significantly enhanced Nannochloropsis sp. growth already after three days of co-cultivation (p=0.001) and this promoting effect was manifested until the end of the observations on day seven. The presence of Poseidonocella sp. partially reduced the Maribacter effect at the beginning of the growth curve, but the algal cell concentration at the end of the seven-day cultivation was the highest among the treatments (OD630 = 1.84 ± 0.12).

112

Figure 2 Nannochloropsis sp. growth curves (OD630). Co-cultivation with Maribacter sp. (a), Maribacter sp. and Poseidonocella sp. (b), Roseovarius sp. (c), Roseovarius sp. and Poseidonocella sp. (d). Two-way ANOVA, α = 0.05, Bonferroni´s correction for multiple comparisons. Asterisks refer to significance determined by multiple comparisons to the control cultures (red line); boxes give results from the two-way ANOVA.

3.2. The endometabolome of microalgae is influenced by bacteria After seven days of cultivation, extracts of both Nannochloropsis sp. and S. robusta from the different treatments were prepared and analyzed with an untargeted metabolomics approach to highlight changes in the metabolism when challenged with the different bacteria. Figure 3a shows a principal component analysis (PCA) plot with all the Nannochloropsis sp. / bacteria co-cultivation treatments. Principal component one and two explained more than 70 % of the total variation of the features. However, the separation between the different treatments was not significant; especially because co- cultivation with Roseovarius sp. showed a large deviation between the samples and dominated the variation. We therefore decided to divide the analysis in sub-groups in order to better visualize potential separation between treatments. The PCA plot in Figure 3b displays a partial overlap of Poseidonocella - Nannochloropsis co-cultivation with the control, therefore confirming that Poseidonocella had only a minor effect on the growth

113

and the metabolism of Nannochloropsis sp. Biological replicates of the co-cultivations with Roseovarius sp. exhibited a substantial variation and high overlap with the control (Figure 3f). The additional presence of Poseidonocella was reducing the variation among samples, leading to a better clustering and a better separation of the treatment from the control (Figure 3d). When Nannochloropsis was grown in presence of Maribacter (Figure 3c and 3e), the treated samples separated clearly from the control. They clustered together in the PCA and the presence or absence of additional Poseidonocella caused only minor additional effects. The effect of Maribacter sp. and Roseovarius sp. on the S. robusta endometabolome was more pronounced (Supplementary Figure 2), reflected in a better separation of the two treatments from the axenic control cultivation. Maribacter sp. had a stronger influence on the diatom metabolome and data were more separated from the control if compared to Roseovarius sp., although Roseovarius sp. had a higher growth-promoting effect on S. robusta compared to Maribacter sp. [250]. Taken together, the results from the metabolome investigation were congruent with the effects of bacteria on algal growth and confirmed the important growth supporting influence of Maribacter sp.

114

Figure 3 PCA plot analysis of the bacterial effect on the Nannochloropsis endometabolome. PCA analysis (a) between all treatments, (b) between Nannochloropsis (red) and Nannochloropsis + Poseidonocella co-cultivation (yellow), (c) Nannochloropsis + Poseidonocella + Maribacter co-cultivation (light green), (d) Nannochloropsis + Poseidonocella + Roseovarius co-cultivation (light blue), (e) Nannochloropsis + Maribacter co-cultivation (green) and (f) Nannochloropsis + Roseovarius co-cultivation (blue).

To identify the significant biomarkers that were determining group separation, we used a t-test for two way comparisons (α = 0.05) and a two-way ANOVA for multiple comparisons (adjusted p-value (FDR) cutoff = 0.05 and FisherLSD post-hoc analysis) and we visualized the most important up- and downregulated metabolites by using heatmaps. Most of the compounds influenced by the presence or absence of bacteria

115

were shared among all treatments (Figure 4 and Supplementary Figure 4). Many amino acids (especially valine, leucine, glycine and tyrosine, and the charged amino acid lysine) were less abundant in the presence of bacteria. Putrescine, a product of amino acid metabolism [261], was upregulated in absence of bacteria, while spermidine, a direct product of putrescine metabolism [262], was higher in presence of bacteria, especially Maribacter sp. (Figure 4b). Maribacter sp. also affected pyrrolinecarboxylic acid derivatives. Interestingly, the presence of Poseidonocella sp. alone induced a higher production of aminobutyric acid, which is a neurotransmitter in vertebrates. When Poseidonocella sp. was combined with other bacteria, proline and glutamine concentration increased in Nannochloropsis sp. cells (Figure 4c). Sterols exhibited a specific behavior dependent of the co-incubated species: Maribacter sp. resulted in an increase of the intracellular amount of some of sterols in Nannochloropsis sp. (Figure 4b and d), while Poseidonocella sp. caused a reduced concentration of cholesterol (Figure 4a). Interestingly, α-tocopherol was always downregulated in presence of bacteria, both in Nannochloropsis sp. (Figure 4) and S. robusta (Figure 5 and Supplementary Figure 5). Squalene, a crucial precursor in sterol biosynthesis [263], was always downregulated in S. robusta co-cultivated with bacteria.

116

Figure 4 Heatmaps of Nannochloropsis endometabolites abundance under different treatments. For two way comparisons, a t-test with α=0.05 was used. For multiple comparisons, a two-way ANOVA analysis with adjusted p-value (FDR) cutoff = 0.05 and FisherLSD post-hoc analysis was used. Color code legend is the same as Figure 3. Only annotated metabolites are displayed. Mass spectra were considered identical with a reverse match factor (R.Match) >800, tagged with "?" if the RM was between 800 and 700, with "??" if the RM was between 700 and 600 and “???” if below 600. Library searched metabolites can only be considered as structural suggestions, since no further co-injection experiments were performed. Upregulation is indicated in red and downregulation in blue with darker colors corresponding to higher relative concentrations.Legend on the horizontal axis corresponds to different replicates of Nannochloropsis (N), Nannochloropsis + Poseidonocella (NP), Nannochloropsis + Maribacter (NM), Nannochloropsis + Roseovariues (NR).

117

Figure 5 Heatmaps of endometabolites abundance of S. robusta axenic cultures vs co-cultivations with Maribacter (a) and Roseovarius (b). t-Test, α = 0.05. See legend of figure 4 for an explanation of the symbols. Only annotated metabolites are displayed.Legend on the horizontal axis corresponds to different replicates of axenic S. robusta cultures (Ax) and the co- cultivations with Maribacter (M) and Roseovarius (R).

Heatmaps (Figure 5) and additional PCA loadings plots (Supplementary Figure 3) showed that Roseovarius and Maribacter sp. effects on S. robusta metabolome were quite different from those on Nannochloropsis sp.: squalene and tocopherol were among the most upregulated metabolites in the axenic treatment, while bacterial presence strongly enhanced the concentration of several disaccharides, some amino acids (in particular alanine) and amino acid derivatives. Moreover, Roseovarius sp. increased the production of arabinose and a putative aminomalonic acid, while Maribacter sp. stimulated the synthesis of monosaccharides as putative xylulose and disaccharides in the form of putative laminaribiose (unknown disaccharide in figure 5).

3.3. Bacteria isolated from S. robusta enhance biofilm induction of Poseidonocella sp. on Nannochloropsis sp. We tested if bacteria isolated from S. robusta could induce or reduce biofilm formation in closed bioreactors by co-cultivating 20 different combinations of bacteria and bacteria spent medium with Nannochloropsis sp. in a biofilm formation assay (results are shown in Figure 6). One-way ANOVA with Bonferroni´s correction for multiple comparisons was used to test the significance of each treatment compared to the control sample (Nannochloropsis sp. alone).

118

Figure 6 Biofilm formation of Nannochloropsis sp. in presence of different bacteria. Each treatment was performed in triplicates. Data were statistically evaluated with a one-way ANOVA, α=0.05, Bonferroni´s correction for multiple comparisons. Asterisks refer to multiple comparisons to Nannochloropsis sp. control sample. OD600 values of 0.01 and 0.1 were selected for the inoculation of bacteria.

Our experiment confirmed the biofilm formation inducing effect of Poseidonocella sp. when co-cultivated with Nannochloropsis sp. (p = 0.025) (Figure 6). Spent medium from S. robusta associated bacteria did not elicit biofilm formation. Roseovarius sp. stimulated biofilm only at high OD600 (=0.1), while Maribacter sp. did not induce significant biofilm formation at any OD. However, when these two bacteria or their spent medium were combined with Poseidonocella sp., biofilm formation was maximal (OD630=1.2-1.3, p < 0.0001). Glucose and glycerol, two substrates that were present in our minimal medium used for bacterial growth, have a big impact on biofilm formation: Nannochloropsis sp. treated with f/2 minimal medium with glucose and glycerol formed a biofilm even without inoculation of bacteria (p < 0.0001). The same effect was observed when Poseidonocella sp. was added to the well plate together with glucose and glycerol enriched medium.

4. Discussion Following the work of Cirri et al. (2018)[250] on the effect of two bacteria (Roseovarius sp. and Maribacter sp.) on sexuality and growth of the biofilm-forming diatom S. robusta, we decided to study the effect of these bacteria isolated from a natural sample on a different, commercially relevant species belonging to the Stramenopile, the

119

oleaginous microalga Nannochloropsis sp. Recently, Giraldo et al. (2019) [163] have found that Poseidonocella sp., a Gram-negative alphaproteobacterium from the Rhodobacteraceae family and phylogenetically close to the Roseobacter clade [203], induces biofilm formation in closed bioreactors for Nannochloropsis sp. culturing. We therefore tested different bacterial combinations to explore changes in microalgal growth, endometabolome and biofilm formation in order to understand if an engineered microbiome with both native and alien bacteria could enhance biomass productivity of Nannochloropsis sp. while reducing biofouling side effects. In co-cultivation experiments, we observed different effects of bacteria on Nannochloropsis growth (Figure 2 and Supplementary Figure 1). The presence of Poseidonocella sp. and Roseovarius sp. did not affect Nannochloropsis sp. biomass increase until seven days of cultivation. On the other hand, Maribacter sp. significantly supported algal growth already 3 days after inoculation. This effect was reduced when Maribacter sp. was co-inoculated with Poseidonocella sp. In this treatment the growth was significantly increased only at the end of the seven-day co-cultivation, but the cell density reached higher values at this point when compared to all the other conditions. The effects on the non-associated alga are thus different to those on algae with which the bacteria co-exist in nature [250]. In the study of Cirri et al. (2018) [250], the diatom S. robusta grew better in presence of Roseovarius sp., while Maribacter sp. only had a small growth-enhancing effect. Many studies confirmed that bacteria support microalgal growth [226], for example by releasing auxins [264], or to reduce it by negatively influencing cell morphology [265]. Some bacteria may exhibit a two-stage behaviour, first promoting and then inhibiting algal growth, until becoming pathogenic [266]. Our work demonstrates that the growth-promoting effect is highly species-specific. To understand if bacteria also have an impact on the Nannochloropsis sp. endometabolome, we analyzed the different co-cultivation setups using a GC-HRMS- based untargeted endometabolomics approach. We then compared the endometabolome of Nannochloropsis sp. and S. robusta to find general and species- specific effects of bacterial isolates. Amino acids were the most affected metabolites in Nannochloropsis sp. (Figure 4), with many of them being downregulated in the presence of bacterial strains, especially glycine, valine, leucine, isoleucine and tyrosine. Glutamine and proline were upregulated in presence of Poseidonocella sp. in combination with one of the two bacteria from S. robusta. Moreover, Maribacter sp. displayed a particular effect on intermediates of the proline cycle: in its presence, pyrrole-2-carboxylic acid accumulated while pyrroline hydroxycarboxylic acid decreased. Several studies showed that bacteria have a significant impact on microalgal amino acid metabolism, inducing the production of specific compounds to sustain their own growth [264,267,268]. This interaction mechanism could be of particular interest for applied purposes, in regard to nutritional

120

value of algae for animal and human consumption [69]. The increase of proline and intermediates of its biosynthetic pathways may suggest a response of the alga to stress, since proline is involved in induced response to toxicity in plants and in some algae [269]. The hypothesis that amino acids are broken down under stress conditions is confirmed by the presence of several degradation byproducts. A putative spermidine, a polyamine that is a product of the putrescine pathway [262], a catabolic product of arginine [270] and involved in abiotic stress response in plants [271], was higher in Nannochloropsis sp. when Maribacter sp. was present (Figure 4). Since putrescine was higher in non- treated Nannochloropsis sp. cultures, these results show a species-specific stress response of the microalga towards bacteria that involves polyamines. Remarkably, the presence of Poseidonocella sp. increased the concentration of aminobutyric acid (Figure 4a), another product of amino acids transformation. Aminobutyric acid has been found, in the form of γ-Aminobutyric acid (GABA), in several microalgae [70]. A study by Pal et al. (2013) correlated the increase of GABA concentration in Nannochloropsis oculata to an osmotic stress response. This molecule is a neurotransmitter in mammals and humans [273] and an important defensive molecule in plants [271]. Bacteria also affect the concentration of amino acids in S. robusta: a putative ornithine was found upregulated in axenic conditions (Supplementary Figure 3), while alanine and other unknown amino acid derivatives were upregulated in presence of bacteria. Ornithine plays an important role in diatom proline biosynthesis [274], which is also involved in stress response mechanisms in many plants and some algae [275]. When investigating the endometabolome of Seminavis robusta (Figure 5), we observed that Maribacter sp. and Roseovarius sp. had a different influence on the native host compared to their effects on Nannochloropsis sp. In S. robusta axenic cultures, we detected a higher concentration of squalene and some unknown sugars and alditol (Supplementary Figure 3), while both Roseovarius sp. and Maribacter sp. treated cultures showed elevated levels of disaccharides and sugar acids. Disaccharide upregulation could be an indication for energy shortage and the increase in sugar acids concentration could be explained as a protective strategy against oxidative stress [276]. The unbiased metabolomic survey presented here demonstrates that different pathways are influenced by bacteria and most likely cascading effects on targeted value products have to be taken into account. This consideration is confirmed by the finding that the commercially important α-tocopherol is down regulated in similar amount in presence of bacteria (Figure 4 and 5). Nannochloropsis is known to produce Vitamin E in relatively high amounts depending on nutrient availability and growth stage [277], and studies on different microalgae [254,276] have shown that its concentration increases during stationary phase as a response to oxidative stress.

121

All these observations suggest that the bacterial composition can deeply influence not only the biomass growth, but also the quality of it, and engineered bacterial communities or their bioactive compounds, when known, can be employed to modulate the nutritional value and chemical composition, e.g. amino acids, sterols or sugar contents, of cultivated microalgae. Bacteria had also a substantial influence on biofilm formation of Nannochloropsis sp (Figure 6). Although biofilm formation mechanism and dynamics in marine environments are still a subject of intense study, the general accepted theory is that bacteria are initiating biofilm formation by releasing exopolymeric substances (EPS) in the so-called surface conditioning phase [171]. EPS are usually composed of exopolysaccharides (40- 95%), proteins (1-60%), nucleic acids (1-10%) and lipids (1-40%) [172,278]. These substances create a protecting and resource-rich environment for both bacteria and microalgae [279] and shape multifaceted communities [280] that have an important and broad ecological role [281]. Interestingly, Maribacter sp. (or its spent medium) did not significantly increase biofilm formation when co-cultivated with Nannochloropsis sp., while Roseovarius sp. induced biofilm development, but only at high cell concentration (Figure 6). However, when either bacteria or their spent media were added to the algal cultures in the presence of Poseidonocella sp., biofilm formation was significantly higher as compared to Nannochloropsis sp. control cultures and the effect was even larger when compared to the effect caused by Poseidonocella sp. alone. The same promoting effect was seen when only saline medium, enriched in glucose and glycerol, was added to a co- cultivation of Nannochloropsis sp. and Poseidonocella sp. Since our Nannochloropsis sp. cultures were not axenic, it could be that bacteria could take advantage of the fresh organic carbon resources to proliferate and induce an increased biofilm production. However, the effect of residual glucose and glycerol on biofilm formation in co- cultivation of bacteria and Nannochloropsis can be excluded, since the spent medium of these bacteria did not trigger any significant biofilm formation (Figure 6), although the amount of inoculated medium, and therefore the potential residual glucose and glycerol, were the same in all treatments. Carbon sources (especially glucose) are important to sustain biofilm formation in various bacterial species, such as Gram-positive bacteria from the Bacillales order [282]. Dissolved sugars appear therefore to be a strong trigger for biofilm formation and have to be monitored in closed bioreactor cultures of Nannochloropsis sp. in order to prevent unwanted growth of thick biofilm.

In conclusion, we demonstrate that microbiome engineering is a powerful tool to direct algal growth, biofilm formation and the algal metabolome. While the first two parameters are directly connected to the yield of a bioreactor, the metabolomics tailoring offers the opportunity to engineer consortia optimized for high value product

122

production. Our experiments transplanting key species of the microbiome from one algae to another reveal that no universal effects can be observed but that microbiome engineering has to take into account prevailing species specific effects.

5. Acknowledgements This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 642575. We acknowledge funding by the state of Thuringia 2015 FGI0021 co-supported by the EU EFRE program. This work makes use of resources and facilities provided by UGent as part of the Belgian contribution to EMBRC-ERIC. We also thank Lachlan Dow for the productive scientific discussion.

123

Supplementary material

Supplementary Figure 1: Nannochloropsis sp. growth curves (OD750). Co-cultivation with Maribacter sp. (a), Maribacter sp. and Poseidonocella sp. (b), Roseovarius sp. (c), Roseovarius sp. and Poseidonocella sp. (d). Two-way ANOVA, α = 0.05, Bonferroni´s correction for multiple comparisons. Asterisks refer to significance determined by multiple comparisons to the control cultures (red line); boxes give results from the two-way ANOVA.

124

Supplementary Figure 2: PCA plot of bacterial effect on S. robusta endometabolome. Roseovarius and Maribacter have a clear influence on S. robusta endometabolome, with Maribacter having a more pronounced effect and therefore grouping further away from the control.

Supplementary Figure 3: PCA loadings plots for S. robusta axenic cultures vs co-cultivations with Maribacter (a) and Roseovarius (b).

125

Supplementary Figure 4: Complete heatmaps of Nannochloropsis endometabolites abundance under different treatments. For two way comparisons, a t-test with α=0.05 was used. For multiple comparisons, a two-way ANOVA analysis with adjusted p- value (FDR) cutoff = 0.05 and FisherLSD post-hoc analysis was used. Color code legend is the same as Figure 3. Mass spectra were considered identical with a reverse match factor (R.Match) >800, tagged with "?" if the RM was between 800 and 700, with "??" if the RM was between 700 and 600 and “???” if below 600. Library searched metabolites can only be considered as structural suggestions, since no further co-injection experiments were performed.

126

Supplementary Figure 5: Complete heatmaps of endometabolites abundance of S. robusta axenic cultures vs co-cultivations with Maribacter (a) and Roseovarius (b). t-Test, α = 0.05. See legend Figure SF4 for an explanation of the symbols.

127

128

Chapter VI

General discussion

129

1. Introduction This work aimed to contribute to the understanding of the algal microbiome in closed photobioreactor systems during non-axenic industrial cultivation and how this is related to biofouling and biofilm formation. The selected industrial environment was the production site of the microalgae producer Flemish company Proviron (Belgium). In chapter 2, production samples were characterized to evaluate reactor conditions and the microbial population was studied. Although this provides insights on the algal microbiome at different conditions, this work remained insufficient to determine the agents responsible of inducing biofilm formation. Therefore, microbial isolation and co- cultivation in a biofilm formation assay was performed in chapter 3. A bacterial strain from the Poseidonocella genus was identified as a strong biofilm inducer. The biofilm forming system Nannochloropsis sp. – Poseidonocella sp. was selected as a model for the study of the molecular mechanisms behind biofilm formation on closed photobioreactors. As seen in chapter 4, this process is initiated by polar signal molecules which might be related to quorum sensing compounds. The final objective is to use this knowledge to enhance productivity and this can be done by the optimization of the algal microbiome during industrial cultivation. In chapter 5, bacterial strains were co- cultivated with Nannochloropsis sp. reporting species specific metabolic responses. Preliminary results suggest a Maribacter sp. strain as a potential growth inducer of Nannochloropsis sp. In this chapter, the main findings of this thesis are discussed and placed in context based on their relevance. Moreover, conclusions and possible recommendations are suggested. 2. Main outcome and relevance of this work

2.1. Cultivation conditions modulate algal microbiome and product quality The influence of operation conditions on the algal microbiome during industrial cultivation has been proven throughout this thesis. During this period, the production site where samples were taken was moved from outdoor to indoor. This transition had important consequences on the algal microbiome. During outdoor cultivation, reactor failures were more frequent and the presence of an undesired green algae species was frequently detected. This alga did not affect the product quality but its presence in the reactor was often correlated with reactor failure. In this thesis, we proved that its microbial flora had a stronger effect on inducing biofilm formation than the green alga. After the transition to indoor production, this green algae contaminant was never detected at the production facilities confirming the hypothesis that location influences the algal microbiome during industrial cultivation. Moving to indoor production implied controlled artificial light conditions and constant room temperature. This prevented uncontrolled environmental events that affect the microbial communities. It is known

130

that environmental factors like peaks of high temperature may have an influence on the bacterial population and the algae [89,283]. As it was seen during this thesis, these environmental modifications may influence not only the microbiome but the product quality as the total fatty acids and particularly EPA can be affected (chapter 2). This could be due to environmental conditions affecting algae metabolism as well as presence/absence of certain bacterial strains. In chapter 5, it was observed downregulation on certain fatty acid when Nannochloropsis sp. was co- cultivated with different bacterial strains. The methodology used failed to determine precisely the nature of this fatty acid. However, this fact supports the evidence that stress may induce a rewiring of fatty acids metabolism on this alga. In chapter 2, Marinicella pacifica was highly abundant in high productivity reactors while its presence was diminished in reactors with lower performance. By means of next generation sequencing, microbial communities can be described and correlated to reactor conditions and productivities. However, the confirmation of certain bacterial strains can only be achieved after bacterial isolation and co-cultivation. During this thesis we suggest the beneficial effect of Marinicella pacifica but this is not confirmed as it was not possible to isolate this microorganism. It is suggested to isolate this bacterium and proceed with co-cultivations experiments with the algae to validate these findings. Alternatively, a commercially available strain may be acquired in case bacterial isolation fails. Moreover, the effect of Poseidonocella sp. in inducing biofilm formation was confirmed since it was possible to isolate and co-cultivate it with the algae (chapter 3). 2.2. Deciphering the bacterial influence on algae cultivation by isolation and co-cultivation In order to identify the effect of bacteria on algae, it is necessary to isolate the strains first. Although by means of next generation sequencing a most representative view of the microbial diversity can be obtained; this will not provide information regarding the effect of each bacterial strain. On the contrary, bacterial isolation fails to provide representative information about the relative abundance of every strain. However, this is not essential information when validating the effect of single isolates as relative abundance is not related with the effect the strains might induce. Consequently, bacterial isolation followed by co-culturing was the only way to experimentally test and manipulate specific algae-bacteria interactions. Limitations about bacterial isolation due to imposed selection conditions were known in advance. Nevertheless, it was decided to proceed as such since pure cultures were a requirement to address the research questions regarding the interaction effects. Furthermore, that the use of NGS techniques to determine relative abundances of microbial groups in the reactors presents two particular challenges which make their use difficult to address our research objective. First, NGS analysis suffers from amplification and sequencing biases that can seriously skew the relative abundances detected. Second, relative abundances are not related to the ability of inducing biofouling as a bacterial strain may induce biofouling independent

131

of the bacterial density. Microbial population differs from early stage biofilms to more mature ones [284]. Furthermore, species-specific synergies between bacterial strains play a major role when forming developed biofilm [285], and this factor is more important than bacterial concentrations. Therefore, it was decided to test all possible single isolates in order to find strong biofilm initiators instead of pre-selecting bacterial strains based on relative abundance which could have led to failing in detection of potential candidates. Despite these known limitations, it has been possible to identify novel biofilm inducers as the isolated Poseidonocella sp. strain. This was achieved by algae-bacteria co-cultivation in a biofilm formation assay. By the same method, biofilms preventers may be found as well if a bacterial strain is tested in co-cultivation with a biofilm forming system. By the follow-up of different co-cultures, further relations might be explored. As described also during this thesis, Maribacter sp. was found to potentially stimulate algae growth. Although that study needs further experimental work to achieve representative conclusions; it sets a starting point for the exploration of beneficial algae- bacteria interactions for industrial algae production. 2.3. Targeting early stage biofilms: Poseidonocella sp. induces biofilm formation through successfully extracted polar compounds Algae-bacteria interactions may lead to biofilm formation in closed photobioreactors. This is a highly regulated process. It has several steps and each stage is affected by several environmental factors [174]. First, bacterial cells attach to the surface using cell appendixes such as flagella, pili, and fimbriae [177]. This form of reversible attachment is followed by irreversible attachment. Surface proteins and EPS support the adhesion of cells [174]. Intracellular second messengers such as bis-(3′,5′)-cyclic dimeric guanosine monophosphate (c-di-GMP) and cAMP are required for the transition from reversible to irreversible attachment [286]. Attached bacterial cells develop microcolonies that may grow by cell proliferation and produce EPS [278]. Then, biofilms form a structured architecture with the assistance of EPS in the maturation stage [174]. Composition of EPS is very important in biofilm development since composition determines the chemical and physical properties of EPS [61]. This diversity of EPS confers unique biofilm morphology. Furthermore, bacteria generate multiple EPS in the biofilms that protect the bacterial cells against various stresses such as desiccation, temperature and competing microbes [174]. From mature biofilm, some bacterial cells convert to planktonic growth and disperse from the biofilm to find new substrate [64]. Thus, dispersal is the final stage of the biofilm lifecycle and at the same time, the start of a new biofilm lifecycle [174]. Currently at Proviron, reactors where well developed biofilms are formed may be stopped, cleaned and re-inoculated. However, this is a costly time-consuming procedure which is preferably avoided. Consequently, the main objective when dealing with biofilms should be biofilm prevention. In order to design a prevention strategy, it is necessary to identify the initiation signals that trigger the formation of early stage

132

biofilms. These signals may be part of the exometabolomome of algae and/or bacteria and thus, this must be the target to screen for the biofilm initiator signals. Growth aspects of Nannochloropsis sp. strains are well studied [287] and they do not naturally form biofilms. During this thesis, it was found that Poseidonocella sp. is able to induce biofilm formation (chapter 3). The Nannochloropsis-Poseidonocella co-cultivation was suggested as a model for the study of the molecular mechanisms behind biofilm formation. Cell responses to initiate biofilm formation may be triggered after first contact with a signal. This hypothesis allows two possibilities: either a signal secreted by bacteria in contact with Nannochloropsis leads to biofilm formation or a signal secreted from Nannochloropsis in contact with the bacteria initiates the biofilm formation. In this thesis, the first option was proved. The second option was discarded based on the non- biofilm forming outcome of the co-cultivation of the bacterium with sterile supernatant from Nannochloropsis sp. (data not shown in this thesis). Exudates from Poseidonocella sp. may induce biofilm formation on Nannochloropsis sp. without presence of bacterial cells (chapter 4). Therefore, the biofilm initiator agents are dissolved in the bacterial supernatant which was the starting point for a comparative metabolomics approach to decipher which signals may be involved in the initiation of biofilms. A solid phase extraction protocol was established with different cartridges with different adsorption properties. All extracts need to be tested on a bioassay to confirm that only the bioactive fraction was extracted. During this thesis we confirmed that small polar signals secreted from Poseidonocella sp. may induce biofilm formation on Nannochloropsis sp. Although the preliminary exploration seems to indicate that quorum sensing compounds might be involved, further experimental work needs to be performed to confirm these observations. The confirmation of this finding would be in line with current studies that involve quorum sensing in biofilm regulation. Within marine biofilms, bacteria are present at high densities which enhance cell-cell interactions and affect their gene expression and metabolism [288]. Biofilm formation is a QS-regulated process in all phases, including initial microbial surface attachment, initiation of biofilm formation, and biofilm development [96]. In addition to that, QS may also be involved in mediating the interactions of surface colonizers [289]. 2.4. Towards microbiome optimization: bacteria modulates algal metabolism through species-specific interactions Synthetic microbial communities may be the key to enhance productivity and robustness against contaminants during industrial microalgae cultivation. This is already a trend in other industrial biotechnological applications [290]. Despite extensive research work on engineering microbial consortia, successful industrial applications remain limited. Such industrial applications include wastewater treatment, biogas production, and the production of traditional foods e.g. dairy products [291] or vitamin C [292].

133

Therefore, the understanding of the algal microbiome is the first step towards its control and possible optimization. Algae-bacteria interactions may occur in different forms from mutualism to parasitism [86]. Despite the general perception in industry that relates bacteria presence as a contaminant, it is known that it can result also in beneficial effects [89]. These biologic interactions in which two or more partners of different species benefit each other are called mutualism [86,293]. Mutualism may involve nutrient exchange from micronutrients like vitamin B12 [92] to macronutrients –nitrogen fixing bacteria may contribute to algal growth [294,295]. Algae-bacteria interactions are complex and dynamics and may switch from mutualisms to parasitism depending on the physiological circumstances of the alga [296]. Differences in microbial population might influence the algal composition too, as seen in chapter 2. As bacteria may contribute to better growth and production of certain metabolites, they may also contribute to a defensive barrier towards contaminations. The principle of exclusion, also known as Gause’s law, states that two species cannot co-exist within the same ecological niche, within the same habitat at the same time [297]. Based on this principle, it is possible that an optimal algal microbiome prevents proliferation of microorganisms with negative consequences in reactor productivity. As an example, the cultures of Tetraselmis spp. secrete antimicrobial agents against pathogens [298]. These type of modifications on the algal microbiome may result in benefits during applications of algae in aquaculture as higher algae quality implies higher survival rate of larvae [299]. Although first steps have been taken, the industrial potential of microbial communities remained unexplored. Consequently, best methods to achieve optimal microbiome for algae production are undefined. In this thesis, marine bacterial isolates were co- cultivated with algae. In chapter 3, this was used to identify their contribution on biofouling and biofilm formation. In chapter 5, bacterial strains with certain effect on an algae species induced a different response on different species, suggesting thus species specificity in algae bacteria interactions. Selection of bacterial isolates based on their effect and co-cultivation may lead to synthetic microbial communities with beneficial effects on algae growth [94]. Other possibilities towards optimization of algal microbiome may consist in selective environments [33]. The idea behind this method is to convert a desired feature into a competitive advantage, converting competition and evolution into an advantage instead of a threat. Mooij et al. (2015) discuss a method based on the natural selection principle that succeeded to increase lipid production although this approach may be used for other characteristics. In this way, the whole ecosystem can be selected by a desired feature including algae strain with its associated bacterial flora. Nevertheless, the obtainment of this selective environment is a challenging task and the stability of the achieved cultures needs to be tested.

134

3. Recommendations and future perspectives In chapter 2, the fatty acids composition and the microbial population of reactors at different conditions were studied. Sampling was performed in consecutive dates to validate the stability of the results in time. Due to modifications from the initial experimental design, this was only possible in the samples from MPC16. However, it is still recommended to include samples at different dates also in high productivity reactors. Despite these were 3 representative samples which are sufficient to achieve statistical conclusions, unexpected issues occurred during DNA sequencing and therefore we recommend increasing the number of samples in future occasions. In this study, correlations between microbial community and fatty acids composition were found. Following studies may include quantification of other relevant compounds like total protein content and pigment analysis as the changes in microbial population might influence these parameters too. Lastly, it is recommended to complement the microbial population analysis with bacterial isolation from the same samples. In this way, it is possible to discern which fraction of microbial population can be cultivated and used for co-cultivation experiments. Despite microbial population analysis suggests an effect from certain strains; this cannot be proven without isolation of those strains an co- cultivation with the algal cells.

In chapter 3, a strain of Poseidonocella sp. was found to be a strong biofilm inducer of Nannochloropsis sp. This co-culture was suggested as a simplified model system to study the mechanisms behind biofilm formation. Besides this possibility, this system could also be used for the study of potential biofilm inhibitors. Bacterial strains with potential benefits may be co-cultivated with Nannochloropsis sp. and Poseidonocella sp. on a biofilm formation assay to validate a biofilm reduction. During the scope of this project, none of the 33 bacterial strains (out of 39) that were isolated from our own reactor setups, and that showed no biofouling activity, proved to have a positive effect on Nannochloropsis sp. growth or on biofilm formation. Apart from bacterial strains, anti-fouling compounds may be tested as well. Similarly, exudates from other microorganisms where antimicrobial effects are suspected may be added to the biofilm forming system. Based on observations at Proviron, industrial production of Chaetoceros sp. and Tetraselmis sp. is less susceptible to produce biofilms than Nannochloropsis sp. or Isochrysis galbana. Chaetoceros didymus is known to secrete proteases that may inhibit the effect of algicidal bacteria [300]. Extracellullar compounds from Tetraselmis chuii have reported antibacterial effect although precise information on the bioactive compounds remains unclear [301]. Contaminants are related with the production location. Therefore, it is expected to have similar contaminations when producing at the same location under the same conditions. When this is the case, different techniques can be applied to detect and prevent this contaminant to develop. Early detection of bacterial contaminants may be determined by 16S rRNA and qPCR assay [302].

135

Furthermore, novel techniques like pH shocks [302] or pulsed electromagnetic field (PEF) [303,304] have reported beneficial effects towards contamination control. In chapter 4, further recommendations are described throughout the chapter. The main objective is the repetition of the experiments with sufficient replicates to validate the results.

In chapter 5, co-cultivation experiments were performed suggesting Maribacter sp. as a potential growth enhancer of Nannochloropsis sp. However, experimental setup presented certain limitations. It is recommended to prolong the experimental time till results are stable. The variation of the algae-bacteria proportion may be followed over time. This can be achieved by precise cell counts for both algae and bacteria and supported by qPCR. This will provide valuable insights on the algae growth and confirm if this bacterium promotes cell division. Furthermore, it would be interesting to test the species-specificity hypothesis by selecting different bacterial strains with known effects on certain algae species and co-cultivate them on other algae.

136

References

137

[1] G. Rajauria, L. Cornish, F. Ometto, F.E. Msuya, R. Villa, Identification and selection of algae for food, feed, and fuel applications, Seaweed Sustain. (2015) 315–345. doi:10.1016/B978-0-12-418697-2.00012-X. [2] M.P. Caporgno, A. Mathys, Trends in Microalgae Incorporation Into Innovative Food Products With Potential Health Benefits., Front. Nutr. 5 (2018) 58. doi:10.3389/fnut.2018.00058. [3] R.M. Gladue, J.E. Maxey, Microalgal feeds for aquaculture, J. Appl. Phycol. 6 (1994) 131–141. doi:10.1007/BF02186067. [4] J. Zhan, J. Rong, Q. Wang, Mixotrophic cultivation, a preferable microalgae cultivation mode for biomass/bioenergy production, and bioremediation, advances and prospect, Int. J. Hydrogen Energy. 42 (2017) 8505–8517. doi:10.1016/J.IJHYDENE.2016.12.021. [5] P. Sharma, N. Sharma, Industrial and Biotechnological Applications of Algae: A Review, J. Adv. Plant Biol. 1 (2017) 1–25. doi:10.14302/issn.2638- 4469.japb-17-1534. [6] A.C. Guedes, F.X. Malcata, Nutritional Value and Uses of Microalgae in Aquaculture, in: Z.A. Muchlisin (Ed.), Aquaculture, IntechOpen, Rijeka, 2012. doi:10.5772/30576. [7] P. Spolaore, C. Joannis-Cassan, E. Duran, A. Isambert, Commercial applications of microalgae, J. Biosci. Bioeng. 101 (2006) 87–96. doi:10.1263/jbb.101.87. [8] D. Kovač Blagojević, J. Simeunović, O. Babić, A. Mišan, I. Milovanović, Algae in food and feed, Food Feed Res. 58226 (2013) 641. [9] R.A. Soni, K. Sudhakar, R.S. Rana, Spirulina – From growth to nutritional product: A review, Trends Food Sci. Technol. 69 (2017) 157–171. doi:10.1016/J.TIFS.2017.09.010. [10] R.J. Shields, I. Lupatsch, Algae for Aquaculture and Animal Feeds, 2012. https://pdfs.semanticscholar.org/2742/197d01cad4e28767fdb1b36288be60 7b066c.pdf (accessed June 12, 2019). [11] S. Hemaiswarya, R. Raja, R. Ravi Kumar, V. Ganesan, C. Anbazhagan, Microalgae: a sustainable feed source for aquaculture, World J. Microbiol. Biotechnol. 27 (2011) 1737–1746. doi:10.1007/s11274-010-0632-z.

138

[12] M.R. Brown, L.E. Cruz-Suárez, D. Ricque-Marie, M. Tapia-Salazar, M.G. Gaxiola-Cortés, Nutritional value of microalgae for aquculture, 2002. https://www.uanl.mx/utilerias/nutricion_acuicola/VI/archivos/A19.pdf (accessed April 26, 2019). [13] N. Abdel-Raouf, A.A. Al-Homaidan, I.B.M. Ibraheem, Microalgae and wastewater treatment., Saudi J. Biol. Sci. 19 (2012) 257–75. doi:10.1016/j.sjbs.2012.04.005. [14] C. Kalidas, L. Edward, Role of micro algae pigments in aquaculture, (2005). http://eprints.cmfri.org.in/6843/ (accessed April 26, 2019). [15] I. Sirakov, K. Velichkova, S. Stoyanova, Y. Staykov, C. Katya Velichkova, International Journal of Fisheries and Aquatic Studies 2015; 2(4): 81-84 The importance of microalgae for aquaculture industry. Review, (2015). www.fao.org/es/ESA/riga/english/index_en.htm (accessed April 26, 2019). [16] M. Bhattacharjee, PHARMACEUTICALLY VALUABLE BIOACTIVE COMPOUNDS OF ALGAE, Asian J. Pharm. Clin. Res. 9 (2016) 43. doi:10.22159/ajpcr.2016.v9i6.14507. [17] I. Khozin-Goldberg, S. Leu, S. Boussiba, Microalgae as a Source for VLC- PUFA Production, in: Subcell. Biochem., 2016: pp. 471–510. doi:10.1007/978- 3-319-25979-6_19. [18] R.-E. Go, K.-A. Hwang, G.-T. Park, H.-M. Lee, G.-A. Lee, C.-W. Kim, S.-Y. Jeon, J.-W. Seo, W.-K. Hong, K.-C. Choi, Effects of microalgal polyunsaturated fatty acid oil on body weight and lipid accumulation in the liver of C57BL/6 mice fed a high fat diet., J. Biomed. Res. 30 (2016) 234–42. doi:10.7555/JBR.30.2016K0004. [19] E. Ryckebosch, C. Bruneel, R. Termote-Verhalle, K. Goiris, K. Muylaert, I. Foubert, Nutritional evaluation of microalgae oils rich in omega-3 long chain polyunsaturated fatty acids as an alternative for fish oil, Food Chem. 160 (2014) 393–400. doi:10.1016/j.foodchem.2014.03.087. [20] M.G. de Morais, B. da S. Vaz, E.G. de Morais, J.A.V. Costa, Biologically Active Metabolites Synthesized by Microalgae, Biomed Res. Int. 2015 (2015) 1–15. doi:10.1155/2015/835761. [21] M. Mourelle, C. Gómez, J. Legido, M.L. Mourelle, C.P. Gómez, J.L. Legido, The Potential Use of Marine Microalgae and Cyanobacteria in Cosmetics and

139

Thalassotherapy, Cosmetics. 4 (2017) 46. doi:10.3390/cosmetics4040046. [22] B. Serive, E. Nicolau, J.-B. Bérard, R. Kaas, V. Pasquet, L. Picot, J.-P. Cadoret, Community analysis of pigment patterns from 37 microalgae strains reveals new carotenoids and porphyrins characteristic of distinct strains and taxonomic groups., PLoS One. 12 (2017) e0171872. doi:10.1371/journal.pone.0171872. [23] A.C. Guedes, H.M. Amaro, F.X. Malcata, Microalgae as sources of carotenoids., Mar. Drugs. 9 (2011) 625–44. doi:10.3390/md9040625. [24] R. Sathasivam, J.-S. Ki, A Review of the Biological Activities of Microalgal Carotenoids and Their Potential Use in Healthcare and Cosmetic Industries, Mar. Drugs. 16 (2018) 26. doi:10.3390/md16010026. [25] D. Gangl, J.A.Z. Zedler, P.D. Rajakumar, E.M.R. Martinez, A. Riseley, A. Włodarczyk, S. Purton, Y. Sakuragi, C.J. Howe, P.E. Jensen, C. Robinson, Biotechnological exploitation of microalgae, J. Exp. Bot. 66 (2015) 6975– 6990. doi:10.1093/jxb/erv426. [26] H.-W. Yen, I.-C. Hu, C.-Y. Chen, S.-H. Ho, D.-J. Lee, J.-S. Chang, Microalgae- based biorefinery – From biofuels to natural products, Bioresour. Technol. 135 (2013) 166–174. doi:https://doi.org/10.1016/j.biortech.2012.10.099. [27] I. Hariskos, C. Posten, Biorefinery of microalgae - opportunities and constraints for different production scenarios, Biotechnol. J. 9 (2014) 739– 752. doi:10.1002/biot.201300142. [28] R.H. Wijffels, M.J. Barbosa, M.H.M. Eppink, Microalgae for the production of bulk chemicals and biofuels, Biofuels, Bioprod. Biorefining. 4 (2010) 287–295. doi:10.1002/bbb.215. [29] P.T. Pienkos, A. Darzins, The promise and challenges of microalgal-derived biofuels, Biofuels, Bioprod. Biorefining. 3 (2009) 431–440. doi:10.1002/bbb.159. [30] W.J. Henley, R.W. Litaker, L. Novoveská, C.S. Duke, H.D. Quemada, R.T. Sayre, Initial risk assessment of genetically modified (GM) microalgae for commodity-scale biofuel cultivation, Algal Res. 2 (2013) 66–77. doi:10.1016/J.ALGAL.2012.11.001. [31] G.S. Araujo, L.J.B.L. Matos, L.R.B. Gonçalves, F.A.N. Fernandes, W.R.L. Farias,

140

Bioprospecting for oil producing microalgal strains: Evaluation of oil and biomass production for ten microalgal strains, Bioresour. Technol. 102 (2011) 5248–5250. doi:10.1016/J.BIORTECH.2011.01.089. [32] L. Sanchez Rizza, M.E. Sanz Smachetti, M. Do Nascimento, G.L. Salerno, L. Curatti, Bioprospecting for native microalgae as an alternative source of sugars for the production of bioethanol, Algal Res. 22 (2017) 140–147. doi:10.1016/J.ALGAL.2016.12.021. [33] P.R. Mooij, G.R. Stouten, M.C. van Loosdrecht, R. Kleerebezem, Ecology- based selective environments as solution to contamination in microalgal cultivation, Curr. Opin. Biotechnol. 33 (2015) 46–51. doi:10.1016/J.COPBIO.2014.11.001. [34] T. Ishika, N.R. Moheimani, P.A. Bahri, D.W. Laird, S. Blair, D. Parlevliet, Halo- adapted microalgae for fucoxanthin production: Effect of incremental increase in salinity, Algal Res. 28 (2017) 66–73. doi:10.1016/J.ALGAL.2017.10.002. [35] M. Vanthoor-Koopmans, R.H. Wijffels, M.J. Barbosa, M.H.M. Eppink, Biorefinery of microalgae for food and fuel, Bioresour. Technol. 135 (2013) 142–149. doi:10.1016/J.BIORTECH.2012.10.135. [36] P.J. Harrison, P.A. Thompson, G.S. Calderwood, Effects of nutrient and light limitation on the biochemical composition of phytoplankton, J. Appl. Phycol. 2 (1990) 45–56. doi:10.1007/BF02179768. [37] K. Spilling, P. Ylöstalo, S. Simis, J. Seppälä, Interaction Effects of Light, Temperature and Nutrient Limitations (N, P and Si) on Growth, Stoichiometry and Photosynthetic Parameters of the Cold-Water Diatom Chaetoceros wighamii, PLoS One. 10 (2015) e0126308. doi:10.1371/journal.pone.0126308. [38] A. Nikolaou, A. Bernardi, F. Bezzo, T. Morosinotto, B. Chachuat, A Dynamic Model of Photoproduction, Photoregulation and Photoinhibition in Microalgae using Chlorophyll Fluorescence, IFAC Proc. Vol. 47 (2014) 4370– 4375. doi:10.3182/20140824-6-ZA-1003.00904. [39] J.H. de Vree, R. Bosma, M. Janssen, M.J. Barbosa, R.H. Wijffels, Comparison of four outdoor pilot-scale photobioreactors, Biotechnol. Biofuels. 8 (2015) 215. doi:10.1186/s13068-015-0400-2.

141

[40] I. Dogaris, M. Welch, A. Meiser, L. Walmsley, G. Philippidis, A novel horizontal photobioreactor for high-density cultivation of microalgae, Bioresour. Technol. 198 (2015) 316–324. doi:10.1016/j.biortech.2015.09.030. [41] M. Cuaresma, M. Janssen, C. Vílchez, R.H. Wijffels, Horizontal or vertical photobioreactors? How to improve microalgae photosynthetic efficiency, Bioresour. Technol. 102 (2011) 5129–5137. doi:10.1016/J.BIORTECH.2011.01.078. [42] A.J. Dassey, C.S. Theegala, Harvesting economics and strategies using centrifugation for cost effective separation of microalgae cells for biodiesel applications, Bioresour. Technol. 128 (2013) 241–245. doi:10.1016/j.biortech.2012.10.061. [43] D. Vandamme, I. Foubert, I. Fraeye, B. Meesschaert, K. Muylaert, Flocculation of Chlorella vulgaris induced by high pH: role of magnesium and calcium and practical implications., Bioresour. Technol. 105 (2012) 114–9. doi:10.1016/j.biortech.2011.11.105. [44] Z. Wu, Y. Zhu, W. Huang, C. Zhang, T. Li, Y. Zhang, A. Li, Evaluation of flocculation induced by pH increase for harvesting microalgae and reuse of flocculated medium, Bioresour. Technol. 110 (2012) 496–502. doi:10.1016/J.BIORTECH.2012.01.101. [45] S. Salim, R. Bosma, M.H. Vermuë, R.H. Wijffels, Harvesting of microalgae by bio-flocculation., J. Appl. Phycol. 23 (2011) 849–855. doi:10.1007/s10811- 010-9591-x. [46] T. Ndikubwimana, X. Zeng, T. Murwanashyaka, E. Manirafasha, N. He, W. Shao, Y. Lu, Harvesting of freshwater microalgae with microbial bioflocculant: a pilot-scale study, Biotechnol. Biofuels. 9 (2016) 47. doi:10.1186/s13068-016-0458-5. [47] D. Bilanovic, G. Shelef, A. Sukenik, Flocculation of microalgae with cationic polymers — Effects of medium salinity, Biomass. 17 (1988) 65–76. doi:10.1016/0144-4565(88)90071-6. [48] G.P. ’t Lam, M.H. Vermuë, G. Olivieri, L.A.M. van den Broek, M.J. Barbosa, M.H.M. Eppink, R.H. Wijffels, D.M.M. Kleinegris, Cationic polymers for successful flocculation of marine microalgae., Bioresour. Technol. 169 (2014) 804–7. doi:10.1016/j.biortech.2014.07.070.

142

[49] A. Oren, A hundred years of Dunaliella research: 1905-2005., Saline Systems. 1 (2005) 2. doi:10.1186/1746-1448-1-2. [50] E. Günerken, E. D’Hondt, M.H.M. Eppink, L. Garcia-Gonzalez, K. Elst, R.H. Wijffels, Cell disruption for microalgae biorefineries, Biotechnol. Adv. 33 (2015) 243–260. doi:10.1016/J.BIOTECHADV.2015.01.008. [51] A.R. Byreddy, A. Gupta, C.J. Barrow, M. Puri, Comparison of Cell Disruption Methods for Improving Lipid Extraction from Thraustochytrid Strains., Mar. Drugs. 13 (2015) 5111–27. doi:10.3390/md13085111. [52] C.S. Theegala, Algal Cell Disruption and Lipid Extraction: A Review on Current Technologies and Limitations, in: Algal Biorefineries, Springer International Publishing, Cham, 2015: pp. 419–441. doi:10.1007/978-3-319- 20200-6_13. [53] C. Safi, A.V. Ursu, C. Laroche, B. Zebib, O. Merah, P.-Y. Pontalier, C. Vaca- Garcia, Aqueous extraction of proteins from microalgae: Effect of different cell disruption methods, Algal Res. 3 (2014) 61–65. doi:10.1016/J.ALGAL.2013.12.004. [54] Y. Tang, J.N. Rosenberg, P. Bohutskyi, G. Yu, M.J. Betenbaugh, F. Wang, Microalgae as a Feedstock for Biofuel Precursors and Value-Added Products: Green Fuels and Golden Opportunities, BioResources. 11 (2015) 2850–2885. doi:10.15376/biores.11.1.Tang. [55] O. Zeriouh, J.V. Reinoso-Moreno, L. López-Rosales, M. del C. Cerón-García, A. Sánchez-Mirón, F. García-Camacho, E. Molina-Grima, Biofouling in photobioreactors for marine microalgae, Crit. Rev. Biotechnol. 37 (2017) 1006–1023. doi:10.1080/07388551.2017.1299681. [56] I. Fitridge, T. Dempster, J. Guenther, R. de Nys, The impact and control of biofouling in marine aquaculture: a review, Biofouling. 28 (2012) 649–669. doi:10.1080/08927014.2012.700478. [57] A. Nawar, A.H. Khoja, N. Akbar, A.A. Ansari, M. Qayyum, E. Ali, Physical abrasion method using submerged spike balls to remove algal biofilm from photobioreactors, BMC Res. Notes. 10 (2017) 666. doi:10.1186/s13104-017- 2995-9. [58] M. Wahl, Ecological lever and interface ecology: epibiosis modulates the interactions between host and environment, Biofouling. 24 (2008) 427–438.

143

doi:10.1080/08927010802339772. [59] A. Decho, Decho AW.. Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanogr Mar Biol Ann Rev 28: 73-153, 1990. [60] M.A. Kumar, K.T.K. Anandapandian, K. Parthiban, Production and characterization of exopolysaccharides (EPS) from biofilm forming marine bacterium, Brazilian Arch. Biol. Technol. 54 (2011) 259–265. doi:10.1590/S1516-89132011000200006. [61] D.H. Limoli, C.J. Jones, D.J. Wozniak, Bacterial Extracellular Polysaccharides in Biofilm Formation and Function, Microbiol. Spectr. 3 (2015). doi:10.1128/microbiolspec.MB-0011-2014. [62] M. Toyofuku, T. Inaba, T. Kiyokawa, N. Obana, Y. Yawata, N. Nomura, Environmental factors that shape biofilm formation, Biosci. Biotechnol. Biochem. 80 (2016) 7–12. doi:10.1080/09168451.2015.1058701. [63] W.G. Characklis, Biofouling: Effects and Control, in: Biofouling Biocorrosion Ind. Water Syst., Springer Berlin Heidelberg, Berlin, Heidelberg, 1991: pp. 7– 27. doi:10.1007/978-3-642-76543-8_2. [64] H. Dang, C.R. Lovell, Microbial Surface Colonization and Biofilm Development in Marine Environments, Microbiol. Mol. Biol. Rev. 80 (2016) 91–138. doi:10.1128/MMBR.00037-15. [65] L. Harris, S. Tozzi, P. Wiley, C. Young, T.-M.J. Richardson, K. Clark, J.D. Trent, Potential impact of biofouling on the photobioreactors of the Offshore Membrane Enclosures for Growing Algae (OMEGA) system, Bioresour. Technol. 144 (2013) 420–428. doi:10.1016/J.BIORTECH.2013.06.125. [66] Y. Shen, W. Zhu, C. Chen, Y. Nie, X. Lin, Biofilm formation in attached microalgal reactors, Bioprocess Biosyst. Eng. 39 (2016) 1281–1288. doi:10.1007/s00449-016-1606-9. [67] P.J. Schnurr, D.G. Allen, Factors affecting algae biofilm growth and lipid production: A review, Renew. Sustain. Energy Rev. 52 (2015) 418–429. doi:10.1016/j.rser.2015.07.090. [68] X.-N. Ma, T.-P. Chen, B. Yang, J. Liu, F. Chen, Lipid Production from Nannochloropsis, Mar. Drugs. 14 (2016). doi:10.3390/MD14040061.

144

[69] M.M. Rebolloso-Fuentes, A. Navarro-Pérez, F. García-Camacho, J.J. Ramos- Miras, J.L. Guil-Guerrero, Biomass nutrient profiles of the microalga Nannochloropsis., J. Agric. Food Chem. 49 (2001) 2966–72. http://www.ncbi.nlm.nih.gov/pubmed/11409994 (accessed November 26, 2018). [70] M.R. Brown, The amino-acid and sugar composition of 16 species of microalgae used in mariculture, J. Exp. Mar. Bio. Ecol. 145 (1991) 79–99. doi:10.1016/0022-0981(91)90007-J. [71] H. Hu, K. Gao, Response of Growth and Fatty Acid Compositions of Nannochloropsis sp. to Environmental Factors Under Elevated CO2 Concentration, Biotechnol. Lett. 28 (2006) 987–992. doi:10.1007/s10529-006- 9026-6. [72] D.A. Martins, L. Custódio, L. Barreira, H. Pereira, R. Ben-Hamadou, J. Varela, K.M. Abu-Salah, Alternative sources of n-3 long-chain polyunsaturated fatty acids in marine microalgae., Mar. Drugs. 11 (2013) 2259–81. doi:10.3390/md11072259. [73] D.J. HIBBERD, Notes on the taxonomy and nomenclature of the algal classes Eustigmatophyceae and Tribophyceae (synonym Xanthophyceae), Bot. J. Linn. Soc. 82 (1981) 93–119. doi:10.1111/j.1095-8339.1981.tb00954.x. [74] M.W. Fawley, I. Jameson, K.P. Fawley, The phylogeny of the genus Nannochloropsis (Monodopsidaceae, Eustigmatophyceae), with descriptions of N. australis sp. nov . and Microchloropsis gen. nov ., Phycologia. 54 (2015) 545–552. doi:10.2216/15-60.1. [75] D. Simionato, M.A. Block, N. La Rocca, J. Jouhet, E. Maréchal, G. Finazzi, T. Morosinotto, The Response of Nannochloropsis gaditana to Nitrogen Starvation Includes De Novo Biosynthesis of Triacylglycerols, a Decrease of Chloroplast Galactolipids, and Reorganization of the Photosynthetic Apparatus, Eukaryot. Cell. 12 (2013) 665. doi:10.1128/EC.00363-12. [76] L. Rodolfi, G. Chini Zittelli, N. Bassi, G. Padovani, N. Biondi, G. Bonini, M.R. Tredici, Microalgae for oil: Strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-cost photobioreactor, Biotechnol. Bioeng. 102 (2009) 100–112. doi:10.1002/bit.22033. [77] R. Radakovits, R.E. Jinkerson, S.I. Fuerstenberg, H. Tae, R.E. Settlage, J.L.

145

Boore, M.C. Posewitz, Draft genome sequence and genetic transformation of the oleaginous alga Nannochloropsis gaditana, Nat. Commun. 3 (2012) 686. doi:10.1038/ncomms1688. [78] A. Vieler, G. Wu, C.-H. Tsai, B. Bullard, A.J. Cornish, C. Harvey, I.-B. Reca, C. Thornburg, R. Achawanantakun, C.J. Buehl, M.S. Campbell, D. Cavalier, K.L. Childs, T.J. Clark, R. Deshpande, E. Erickson, A. Armenia Ferguson, W. Handee, Q. Kong, X. Li, B. Liu, S. Lundback, C. Peng, R.L. Roston, Sanjaya, J.P. Simpson, A. TerBush, J. Warakanont, S. Zäuner, E.M. Farre, E.L. Hegg, N. Jiang, M.-H. Kuo, Y. Lu, K.K. Niyogi, J. Ohlrogge, K.W. Osteryoung, Y. Shachar-Hill, B.B. Sears, Y. Sun, H. Takahashi, M. Yandell, S.-H. Shiu, C. Benning, Genome, Functional Gene Annotation, and Nuclear Transformation of the Heterokont Oleaginous Alga Nannochloropsis oceanica CCMP1779, PLoS Genet. 8 (2012) e1003064. doi:10.1371/journal.pgen.1003064. [79] L. Roef, M. Jacqmain, M. Michiels, 13 Case study: Microalgae production in the selfsupported ProviAPT vertical flat-panel photobioreactor system, in: Microalgal Biotechnol. Potential Prod., DE GRUYTER, Berlin, Boston, n.d. doi:10.1515/9783110225020.243. [80] K. Paranjape, G.B. Leite, P.C. Hallenbeck, Strain variation in microalgal lipid production during mixotrophic growth with glycerol, Bioresour. Technol. 204 (2016) 80–88. doi:10.1016/j.biortech.2015.12.071. [81] Y. Liang, N. Sarkany, Y. Cui, Biomass and lipid productivities of Chlorella vulgaris under autotrophic, heterotrophic and mixotrophic growth conditions, Biotechnol. Lett. 31 (2009) 1043–1049. doi:10.1007/s10529-009- 9975-7. [82] S.H. CHO, S.-C. JI, S.B. HUR, J. BAE, I.-S. PARK, Y.-C. SONG, Optimum temperature and salinity conditions for growth of green algae Chlorella ellipsoidea and Nannochloris oculata, Fish. Sci. 73 (2007) 1050–1056. doi:10.1111/j.1444-2906.2007.01435.x. [83] A.R. Rao, C. Dayananda, R. Sarada, T.R. Shamala, G.A. Ravishankar, Effect of salinity on growth of green alga Botryococcus braunii and its constituents, Bioresour. Technol. 98 (2007) 560–564. doi:10.1016/J.BIORTECH.2006.02.007. [84] K. Phadwal, P.. Singh, Effect of nutrient depletion on β-carotene and glycerol accumulation in two strains of Dunaliella sp., Bioresour. Technol. 90 (2003) 55–58. doi:10.1016/S0960-8524(03)00090-7.

146

[85] S.A. Amin, M.S. Parker, E. V Armbrust, Interactions between Diatoms and Bacteria, Microbiol Mol Biol Rev. 76 (2012). doi:10.1128/MMBR.00007-12. [86] R. Ramanan, B.-H. Kim, D.-H. Cho, H.-M. Oh, H.-S. Kim, Algae–bacteria interactions: Evolution, ecology and emerging applications, Biotechnol. Adv. 34 (2016) 14–29. doi:10.1016/J.BIOTECHADV.2015.12.003. [87] A. Kouzuma, K. Watanabe, Exploring the potential of algae/bacteria interactions, Curr. Opin. Biotechnol. 33 (2015) 125–129. doi:10.1016/j.copbio.2015.02.007. [88] J.J. Cole, Interactions Between Bacteria and Algae in Aquatic Ecosystems, Annu. Rev. Ecol. Syst. 13 (1982) 291–314. doi:10.1146/annurev.es.13.110182.001451. [89] J.L. Fuentes, I. Garbayo, M. Cuaresma, Z. Montero, M. González-Del-Valle, C. Vílchez, Impact of Microalgae-Bacteria Interactions on the Production of Algal Biomass and Associated Compounds., Mar. Drugs. 14 (2016). doi:10.3390/md14050100. [90] L.M. Lubián, O. Montero, I. Moreno-Garrido, I.E. Huertas, C. Sobrino, M. González-del Valle, G. Parés, Nannochloropsis (Eustigmatophyceae) as source of commercially valuable pigments, J. Appl. Phycol. 12 (2000) 249– 255. doi:10.1023/A:1008170915932. [91] H. Grossart, M. Simon, Interactions of planktonic algae and bacteria: effects on algal growth and organic matter dynamics, Aquat. Microb. Ecol. 47 (2007) 163–176. doi:10.3354/ame047163. [92] M.T. Croft, A.D. Lawrence, E. Raux-Deery, M.J. Warren, A.G. Smith, Algae acquire vitamin B12 through a symbiotic relationship with bacteria, Nature. 438 (2005) 90–93. doi:10.1038/nature04056. [93] J.A. Villa, E.E. Ray, B.M. Barney, Azotobacter vinelandii siderophore can provide nitrogen to support the culture of the green algae Neochloris oleoabundans and Scenedesmus sp. BA032, FEMS Microbiol. Lett. 351 (2014) 70–77. doi:10.1111/1574-6968.12347. [94] D.-H. Cho, R. Ramanan, J. Heo, J. Lee, B.-H. Kim, H.-M. Oh, H.-S. Kim, Enhancing microalgal biomass productivity by engineering a microalgal– bacterial community, Bioresour. Technol. 175 (2015) 578–585. doi:10.1016/j.biortech.2014.10.159.

147

[95] H.M. Amaro, A.C. Guedes, F.X. Malcata, Antimicrobial activities of microalgae: an invited review, n.d. http://www.formatex.info/microbiology3/book/1272-1284.pdf (accessed April 26, 2019). [96] L.R. Hmelo, Quorum Sensing in Marine Microbial Environments, Ann. Rev. Mar. Sci. 9 (2017) 257–281. doi:10.1146/annurev-marine-010816-060656. [97] M.B. Miller, B.L. Bassler, Quorum Sensing in Bacteria, Annu. Rev. Microbiol. 55 (2001) 165–199. doi:10.1146/annurev.micro.55.1.165. [98] W. Chi, L. Zheng, C. He, B. Han, M. Zheng, W. Gao, C. Sun, G. Zhou, X. Gao, Quorum sensing of microalgae associated marine Ponticoccus sp. PD-2 and its algicidal function regulation, AMB Express. 7 (2017) 59. doi:10.1186/s13568-017-0357-6. [99] Y.-H. Li, X. Tian, Quorum sensing and bacterial social interactions in biofilms., Sensors (Basel). 12 (2012) 2519–38. doi:10.3390/s120302519. [100] J. Zhou, Y. Lyu, M. Richlen, D.M. Anderson, Z. Cai, Quorum sensing is a language of chemical signals and plays an ecological role in algal-bacterial interactions., CRC. Crit. Rev. Plant Sci. 35 (2016) 81–105. doi:10.1080/07352689.2016.1172461. [101] A. Deep, U. Chaudhary, V. Gupta, Quorum sensing and bacterial pathogenicity: From molecules to disease, J. Lab. Physicians. 3 (2011) 4. doi:10.4103/0974-2727.78553. [102] Y.-H. Li, X. Tian, Quorum Sensing and Bacterial Social Interactions in Biofilms, Sensors. 12 (2012) 2519–2538. doi:10.3390/s120302519. [103] K. Saurav, V. Costantino, V. Venturi, L. Steindler, Quorum Sensing Inhibitors from the Sea Discovered Using Bacterial N-acyl-homoserine Lactone-Based Biosensors, Mar. Drugs. 15 (2017) 53. doi:10.3390/md15030053. [104] K. Ivanova, M.M. Fernandes, T. Tzanov, Current advances on bacterial pathogenesis inhibition and treatment strategies, 2013. https://pdfs.semanticscholar.org/003e/c4ac049becc5506738576148d5a7cc6 24dca.pdf (accessed April 26, 2019). [105] M.P. Storz, C.K. Maurer, C. Zimmer, N. Wagner, C. Brengel, J.C. de Jong, S. Lucas, M. Müsken, S. Häussler, A. Steinbach, R.W. Hartmann, Validation of

148

PqsD as an Anti-biofilm Target in Pseudomonas aeruginosa by Development of Small-Molecule Inhibitors, J. Am. Chem. Soc. 134 (2012) 16143–16146. doi:10.1021/ja3072397. [106] W.A. Faé Neto, C.R. Borges Mendes, P.C. Abreu, Carotenoid production by the marine microalgae Nannochloropsis oculata in different low-cost culture media, Aquac. Res. 49 (2018) 2527–2535. doi:10.1111/are.13715. [107] B. Karlson, D. Potter, M. Kuylenstierna, R.A. Andersen, Ultrastructure, pigment composition, and 18S rRNA gene sequence for Nannochloropsis granulata sp. nov. (Monodopsidaceae, Eustigmatophyceae), a marine ultraplankter isolated from the Skagerrak, northeast Atlantic Ocean, Phycologia. 35 (1996) 253–260. doi:10.2216/i0031-8884-35-3-253.1. [108] J.-C. Gwo, J.-Y. Chiu, C.-C. Chou, H.-Y. Cheng, Cryopreservation of a marine microalga, Nannochloropsis oculata (Eustigmatophyceae), Cryobiology. 50 (2005) 338–343. doi:10.1016/J.CRYOBIOL.2005.02.001. [109] H. Hu, K. Gao, Optimization of growth and fatty acid composition of a unicellular marine picoplankton, Nannochloropsis sp., with enriched carbon sources, Biotechnol. Lett. 25 (2003) 421–425. doi:10.1023/A:1022489108980. [110] C.J. Hulatt, R.H. Wijffels, S. Bolla, V. Kiron, Production of Fatty Acids and Protein by Nannochloropsis in Flat-Plate Photobioreactors., PLoS One. 12 (2017) e0170440. doi:10.1371/journal.pone.0170440. [111] E. Sforza, A. Bertucco, T. Morosinotto, G.M. Giacometti, Photobioreactors for microalgal growth and oil production with Nannochloropsis salina: From lab-scale experiments to large-scale design, Chem. Eng. Res. Des. 90 (2012) 1151–1158. doi:10.1016/J.CHERD.2011.12.002. [112] S. Archambault, C.M. Starbuck Downes, W. Van Voorhies, C.A. Erickson, P. Lammers, Nannochloropsis sp. algae for use as biofuel: Analyzing a translog production function using data from multiple sites in the southwestern United States, Algal Res. 6 (2014) 124–131. doi:10.1016/J.ALGAL.2014.10.005. [113] S.P. Alves, S.H. Mendonça, J.L. Silva, R.J.B. Bessa, Nannochloropsis oceanica, a novel natural source of rumen-protected eicosapentaenoic acid (EPA) for ruminants, Sci. Rep. 8 (2018) 10269. doi:10.1038/s41598-018-28576-7. [114] M. Sørensen, Y. Gong, F. Bjarnason, G.K. Vasanth, D. Dahle, M. Huntley, V. Kiron, Nannochloropsis oceania-derived defatted meal as an alternative to

149

fishmeal in Atlantic salmon feeds., PLoS One. 12 (2017) e0179907. doi:10.1371/journal.pone.0179907. [115] Y. Sun, Y. Huang, Q. Liao, A. Xia, Q. Fu, X. Zhu, J. Fu, Boosting Nannochloropsis oculata growth and lipid accumulation in a lab-scale open raceway pond characterized by improved light distributions employing built-in planar waveguide modules, Bioresour. Technol. 249 (2018) 880–889. doi:10.1016/J.BIORTECH.2017.11.013. [116] J. Janssen, R. Wijffels, M. Barbosa, Lipid Production in Nannochloropsis gaditana during Nitrogen Starvation, Biology (Basel). 8 (2019) 5. doi:10.3390/biology8010005. [117] L. Krienitz, M. Wirth, The high content of polyunsaturated fatty acids in Nannochloropsis limnetica (Eustigmatophyceae) and its implication for food web interactions, freshwater aquaculture and biotechnology, Limnologica. 36 (2006) 204–210. doi:10.1016/J.LIMNO.2006.05.002. [118] C.-Y. Chen, D. Nagarajan, W.Y. Cheah, Eicosapentaenoic acid production from Nannochloropsis oceanica CY2 using deep sea water in outdoor plastic-bag type photobioreactors, Bioresour. Technol. 253 (2018) 1–7. doi:10.1016/j.biortech.2017.12.102. [119] J. Wang, T. Luo, S. Li, J. Zhao, The powerful applications of polyunsaturated fatty acids in improving the therapeutic efficacy of anticancer drugs, Expert Opin. Drug Deliv. 9 (2012) 1–7. doi:10.1517/17425247.2011.618183. [120] F. Sahena, I.S.M. Zaidul, S. Jinap, N. Saari, H.A. Jahurul, K.A. Abbas, N.A. Norulaini, PUFAs in Fish: Extraction, Fractionation, Importance in Health, Compr. Rev. Food Sci. Food Saf. 8 (2009) 59–74. doi:10.1111/j.1541- 4337.2009.00069.x. [121] A.F.G. Cicero, A. Reggi, A. Parini, C. Borghi, Application of polyunsaturated fatty acids in internal medicine: beyond the established cardiovascular effects., Arch. Med. Sci. 8 (2012) 784–93. doi:10.5114/aoms.2012.31613. [122] S. Maqsood, S. Benjakul, A. Kamal-Eldin, Extraction, processing, and stabilization of health-promoting fish oils., Recent Pat. Food. Nutr. Agric. 4 (2012) 141–7. http://www.ncbi.nlm.nih.gov/pubmed/22741928 (accessed April 17, 2019). [123] G. Lenihan-Geels, K.S. Bishop, L.R. Ferguson, Alternative sources of omega-3

150

fats: can we find a sustainable substitute for fish?, Nutrients. 5 (2013) 1301– 15. doi:10.3390/nu5041301. [124] E.M. Grima, J.A.S. Pérez, F.G. Camacho, A.R. Medina, A.G. Giménez, D. López Alonso, The production of polyunsaturated fatty acids by microalgae: from strain selection to product purification, Process Biochem. 30 (1995) 711–719. doi:10.1016/0032-9592(94)00047-6. [125] M.L. Hamilton, J. Warwick, A. Terry, M.J. Allen, J.A. Napier, O. Sayanova, Towards the Industrial Production of Omega-3 Long Chain Polyunsaturated Fatty Acids from a Genetically Modified Diatom Phaeodactylum tricornutum, PLoS One. 10 (2015) e0144054. doi:10.1371/JOURNAL.PONE.0144054. [126] G. Padmaperuma, R.V. Kapoore, D.J. Gilmour, S. Vaidyanathan, Microbial consortia: a critical look at microalgae co-cultures for enhanced biomanufacturing, Crit. Rev. Biotechnol. 38 (2018) 690–703. doi:10.1080/07388551.2017.1390728. [127] J. Lian, R.H. Wijffels, H. Smidt, D. Sipkema, The effect of the algal microbiome on industrial production of microalgae, Microb. Biotechnol. 11 (2018) 806–818. doi:10.1111/1751-7915.13296. [128] G. NAKASE, M. EGUCHI, Analysis of bacterial communities in Nannochloropsis sp. cultures used for larval fish production, Fish. Sci. 73 (2007) 543–549. doi:10.1111/j.1444-2906.2007.01366.x. [129] H. Geng, M.B. Tran-Gyamfi, T.W. Lane, K.L. Sale, E.T. Yu, Changes in the Structure of the Microbial Community Associated with Nannochloropsis salina following Treatments with Antibiotics and Bioactive Compounds., Front. Microbiol. 7 (2016) 1155. doi:10.3389/fmicb.2016.01155. [130] H. Geng, K.L. Sale, M.B. Tran-Gyamfi, T.W. Lane, E.T. Yu, Longitudinal Analysis of Microbiota in Microalga Nannochloropsis salina Cultures, Microb. Ecol. 72 (2016) 14–24. doi:10.1007/s00248-016-0746-4. [131] J. Fret, L. Roef, R. Blust, L. Diels, S. Tavernier, W. Vyverman, M. Michiels, Reuse of rejuvenated media during laboratory and pilot scale cultivation of Nannochloropsis sp., Algal Res. 27 (2017) 265–273. doi:10.1016/J.ALGAL.2017.09.018. [132] R.R.L. Guillard, Culture of Phytoplankton for Feeding Marine Invertebrates, in: Cult. Mar. Invertebr. Anim., Springer US, Boston, MA, 1975: pp. 29–60.

151

doi:10.1007/978-1-4615-8714-9_3. [133] M. De Troch, P. Boeckx, C. Cnudde, D. Van Gansbeke, A. Vanreusel, M. Vincx, M. Caramujo, Bioconversion of fatty acids at the basis of marine food webs: insights from a compound-specific stable isotope analysis, Mar. Ecol. Prog. Ser. 465 (2012) 53–67. doi:10.3354/meps09920. [134] R.G. Ackman, J.C. Sipos, Application of specific response factors in the gas chromatographic analysis of methyl esters of fatty acids with flame ionization detectors, J. Am. Oil Chem. Soc. 41 (1964) 377–378. doi:10.1007/BF02654818. [135] R.L. Wolff, C.C. Bayard, R.J. Fabien, Evaluation of sequential methods for the determination of butterfat fatty acid composition with emphasis on trans - 18:1 acids. Application to the study of seasonal variations in french butters, J. Am. Oil Chem. Soc. 72 (1995) 1471–1483. doi:10.1007/BF02577840. [136] G. Muyzer, A. Teske, C.O. Wirsen, H.W. Jannasch, Phylogenetic relationships ofThiomicrospira species and their identification in deep-sea hydrothermal vent samples by denaturing gradient gel electrophoresis of 16S rDNA fragments, Arch. Microbiol. 164 (1995) 165–172. doi:10.1007/BF02529967. [137] K.L. Morrissey, L. Çavaş, A. Willems, O. De Clerck, Disentangling the Influence of Environment, Host Specificity and Thallus Differentiation on Bacterial Communities in Siphonous Green Seaweeds, Front. Microbiol. 10 (2019) 717. doi:10.3389/fmicb.2019.00717. [138] B. Tytgat, E. Verleyen, M. Sweetlove, S. D’hondt, P. Clercx, E. Van Ranst, K. Peeters, S. Roberts, Z. Namsaraev, A. Wilmotte, W. Vyverman, A. Willems, Bacterial community composition in relation to bedrock type and macrobiota in soils from the Sør Rondane Mountains, East Antarctica, FEMS Microbiol. Ecol. 92 (2016) fiw126. doi:10.1093/femsec/fiw126. [139] S.J. Green, M.B. Leigh, J.D. Neufeld, Denaturing Gradient Gel Electrophoresis (DGGE) for Microbial Community Analysis, in: Handb. Hydrocarb. Lipid Microbiol., Springer Berlin Heidelberg, Berlin, Heidelberg, 2010: pp. 4137– 4158. doi:10.1007/978-3-540-77587-4_323. [140] E.J. van Hannen, M.P. van Agterveld, H.J. Gons, H.J. Laanbroek, REVEALING GENETIC DIVERSITY OF EUKARYOTIC MICROORGANISMS IN AQUATIC ENVIRONMENTS BY DENATURING GRADIENT GEL ELECTROPHORESIS, J.

152

Phycol. 34 (1998) 206–213. doi:10.1046/j.1529-8817.1998.340206.x. [141] S. Ohse, R. Bianchini Derner, R. Ávila Ozório, R. Gordo Corrêa, E. Badiale Furlong, P.C. Roberto Cunha, Lipid content and fatty acid profiles in ten species of microalgae, Idesia (Arica). 33 (2015) 93–101. doi:10.4067/S0718- 34292015000100010. [142] Y. Ma, Z. Wang, C. Yu, Y. Yin, G. Zhou, Evaluation of the potential of 9 Nannochloropsis strains for biodiesel production, Bioresour. Technol. 167 (2014) 503–509. doi:10.1016/J.BIORTECH.2014.06.047. [143] F.J.L. Baumgardt, A.Z. Filho, M. V. Brandalize, D.C. da Costa, N.R. Antoniosi Filho, P.C.O.V. Abreu, M.L. Corazza, L.P. Ramos, Lipid content and fatty acid profile of Nannochloropsis oculata before and after extraction with conventional solvents and/or compressed fluids, J. Supercrit. Fluids. 108 (2016) 89–95. doi:10.1016/J.SUPFLU.2015.11.003. [144] J.H. Janssen, J. Kastenhofer, J.A. de Hoop, P.P. Lamers, R.H. Wijffels, M.J. Barbosa, Effect of nitrogen addition on lipid productivity of nitrogen starved Nannochloropsis gaditana, Algal Res. 33 (2018) 125–132. doi:10.1016/J.ALGAL.2018.05.009. [145] H. Safafar, M.Z. Hass, P. Møller, S.L. Holdt, C. Jacobsen, High-EPA Biomass from Nannochloropsis salina Cultivated in a Flat-Panel Photo-Bioreactor on a Process Water-Enriched Growth Medium, Mar. Drugs. 14 (2016). doi:10.3390/MD14080144. [146] H. Hoshida, T. Ohira, A. Minematsu, R. Akada, Y. Nishizawa, Accumulation of eicosapentaenoic acid in Nannochloropsis sp. in response to elevated CO2 concentrations, J. Appl. Phycol. 17 (2005) 29–34. doi:10.1007/s10811-005- 5512-9. [147] M. Hoffmann, K. Marxen, R. Schulz, K.H. Vanselow, TFA and EPA productivities of Nannochloropsis salina influenced by temperature and nitrate stimuli in turbidostatic controlled experiments., Mar. Drugs. 8 (2010) 2526–45. doi:10.3390/md8092526. [148] A. Sukenik, Y. Carmeli, T. Berner, REGULATION OF FATTY ACID COMPOSITION BY IRRADIANCE LEVEL IN THE EUSTIGMATOPHYTE NANNOCHLOROPSIS SP.1, J. Phycol. 25 (1989) 686–692. doi:10.1111/j.0022- 3646.1989.00686.x.

153

[149] C.W. Kim, M.-G. Sung, K. Nam, M. Moon, J.-H. Kwon, J.-W. Yang, Effect of monochromatic illumination on lipid accumulation of Nannochloropsis gaditana under continuous cultivation, Bioresour. Technol. 159 (2014) 30–35. doi:10.1016/j.biortech.2014.02.024. [150] A. Alboresi, G. Perin, N. Vitulo, G. Diretto, M.A. Block, J. Jouhet, A. Meneghesso, G. Valle, G. Giuliano, E. Maréchal, T. Morosinotto, Light Remodels Lipid Biosynthesis in Nannochloropsis gaditana by Modulating Carbon Partitioning Between Organelles, Plant Physiol. 171 (2016) pp.00599.2016. doi:10.1104/pp.16.00599. [151] S.-H. Yoon, S.-M. Ha, S. Kwon, J. Lim, Y. Kim, H. Seo, J. Chun, Introducing EzBioCloud: a taxonomically united database of 16S rRNA gene sequences and whole-genome assemblies, Int. J. Syst. Evol. Microbiol. 67 (2017) 1613– 1617. doi:10.1099/ijsem.0.001755. [152] Y. Wang, Y. Liu, Z. Zhang, Y. Zheng, X.-H. Zhang, Marinicella pacifica sp. nov., isolated from seawater, Int. J. Syst. Evol. Microbiol. 66 (2016) 2313–2318. doi:10.1099/ijsem.0.001027. [153] A.R. Currie, K. Tait, H. Parry, B. de Francisco-Mora, N. Hicks, A.M. Osborn, S. Widdicombe, H. Stahl, Marine Microbial Gene Abundance and Community Composition in Response to Ocean Acidification and Elevated Temperature in Two Contrasting Coastal Marine Sediments., Front. Microbiol. 8 (2017) 1599. doi:10.3389/fmicb.2017.01599. [154] N.M. Farhat, J.S. Vrouwenvelder, M.C.M. Van Loosdrecht, S.S. Bucs, M. Staal, Effect of water temperature on biofouling development in reverse osmosis membrane systems, Water Res. 103 (2016) 149–159. doi:10.1016/j.watres.2016.07.015. [155] J.P. Lord, Impact of seawater temperature on growth and recruitment of invasive fouling species at the global scale, Mar. Ecol. 38 (2017) e12404. doi:10.1111/maec.12404. [156] V. Diaz Villanueva, J. Font, T. Schwartz, A.M. Romani, Biofilm formation at warming temperature: acceleration of microbial colonization and microbial interactive effects., Biofouling. 27 (2011) 59–71. doi:10.1080/08927014.2010.538841. [157] B. Tamburic, S. Guruprasad, D.T. Radford, M. Szabó, R.M. Lilley, A.W.D.

154

Larkum, J.B. Franklin, D.M. Kramer, S.I. Blackburn, J.A. Raven, M. Schliep, P.J. Ralph, The effect of Diel temperature and light cycles on the growth of nannochloropsis oculata in a photobioreactor matrix., PLoS One. 9 (2014) e86047. doi:10.1371/journal.pone.0086047. [158] C. Liu, Y. Wu, L. Li, Y. Ma, Z. Shao, Thalassospira xiamenensis sp. nov. and Thalassospira profundimaris sp. nov., Int. J. Syst. Evol. Microbiol. 57 (2007) 316–320. doi:10.1099/ijs.0.64544-0. [159] Q. Lai, Z. Shao, Genome Sequence of Thalassospira xiamenensis Type Strain M-5, J. Bacteriol. 194 (2012) 6957–6957. doi:10.1128/JB.01904-12. [160] X. Lu, B. Zhou, L. Xu, L. Liu, G. Wang, X. Liu, X. Tang, A marine algicidal Thalassospira and its active substance against the harmful algal bloom species Karenia mikimotoi, Appl. Microbiol. Biotechnol. 100 (2016) 5131– 5139. doi:10.1007/s00253-016-7352-8. [161] H. Biebl, M. Allgaier, H. Lünsdorf, R. Pukall, B.J. Tindall, I. Wagner-Döbler, Roseovarius mucosus sp. nov., a member of the Roseobacter clade with trace amounts of bacteriochlorophyll a, Int. J. Syst. Evol. Microbiol. 55 (2005) 2377–2383. doi:10.1099/ijs.0.63832-0. [162] V. Michael, O. Frank, P. Bartling, C. Scheuner, M. Göker, H. Brinkmann, J. Petersen, Biofilm plasmids with a rhamnose operon are widely distributed determinants of the “swim-or-stick” lifestyle in roseobacters., ISME J. 10 (2016) 2498–513. doi:10.1038/ismej.2016.30. [163] J.B. Giraldo, W. Stock, L. Dow, L. Roef, A. Willems, S. Mangelinckx, P.G. Kroth, W. Vyverman, M. Michiels, Influence of the algal microbiome on biofouling during industrial cultivation of Nannochloropsis sp. in closed photobioreactors, Algal Res. 42 (2019). doi:10.1016/j.algal.2019.101591. [164] I. Doghri, S. Rodrigues, A. Bazire, A. Dufour, D. Akbar, V. Sopena, S. Sablé, I. Lanneluc, Marine bacteria from the French Atlantic coast displaying high forming-biofilm abilities and different biofilm 3D architectures., BMC Microbiol. 15 (2015) 231. doi:10.1186/s12866-015-0568-4. [165] J.B. Bruhn, L. Gram, R. Belas, Production of antibacterial compounds and biofilm formation by Roseobacter species are influenced by culture conditions., Appl. Environ. Microbiol. 73 (2007) 442–50. doi:10.1128/AEM.02238-06.

155

[166] K. Kumar, A. Sirasale, D. Das, Use of image analysis tool for the development of light distribution pattern inside the photobioreactor for the algal cultivation, Bioresour. Technol. 143 (2013) 88–95. doi:10.1016/J.BIORTECH.2013.05.117. [167] H. Wang, H.D. Laughinghouse, M.A. Anderson, F. Chen, E. Willliams, A.R. Place, O. Zmora, Y. Zohar, T. Zheng, R.T. Hill, Novel bacterial isolate from Permian groundwater, capable of aggregating potential biofuel-producing microalga Nannochloropsis oceanica IMET1., Appl. Environ. Microbiol. 78 (2012) 1445–53. doi:10.1128/AEM.06474-11. [168] B.O. Ortega-Morales, M.J. Chan-Bacab, S. del C. De la Rosa-García, J.C. Camacho-Chab, Valuable processes and products from marine intertidal microbial communities, Curr. Opin. Biotechnol. 21 (2010) 346–352. doi:10.1016/J.COPBIO.2010.02.007. [169] D. Taherzadeh, C. Picioreanu, H. Horn, Mass Transfer Enhancement in Moving Biofilm Structures, Biophys. J. 102 (2012) 1483–1492. doi:10.1016/J.BPJ.2012.02.033. [170] P.S. Stewart, Mechanisms of antibiotic resistance in bacterial biofilms, Int. J. Med. Microbiol. 292 (2002) 107–113. doi:10.1078/1438-4221-00196. [171] G.S. Lorite, C.M. Rodrigues, A.A. de Souza, C. Kranz, B. Mizaikoff, M.A. Cotta, The role of conditioning film formation and surface chemical changes on Xylella fastidiosa adhesion and biofilm evolution, J. Colloid Interface Sci. 359 (2011) 289–295. doi:10.1016/J.JCIS.2011.03.066. [172] H.C. Flemming, J. Wingender, Relevance of microbial extracellular polymeric substances (EPSs)--Part I: Structural and ecological aspects., Water Sci. Technol. 43 (2001) 1–8. http://www.ncbi.nlm.nih.gov/pubmed/11381954 (accessed March 30, 2018). [173] J.M. Koczan, B.R. Lenneman, M.J. McGrath, G.W. Sundin, Cell surface attachment structures contribute to biofilm formation and xylem colonization by Erwinia amylovora., Appl. Environ. Microbiol. 77 (2011) 7031–9. doi:10.1128/AEM.05138-11. [174] M. Toyofuku, T. Inaba, T. Kiyokawa, N. Obana, Y. Yawata, N. Nomura, Environmental factors that shape biofilm formation, Biosci. Biotechnol. Biochem. 80 (2015) 1–6. doi:10.1080/09168451.2015.1058701.

156

[175] G.A. O’Toole, G.C. Wong, Sensational biofilms: surface sensing in bacteria, Curr. Opin. Microbiol. 30 (2016) 139–146. doi:10.1016/j.mib.2016.02.004. [176] N. Rabin, Y. Zheng, C. Opoku-Temeng, Y. Du, E. Bonsu, H.O. Sintim, Agents that inhibit bacterial biofilm formation, Future Med. Chem. 7 (2015) 647– 671. doi:10.4155/fmc.15.7. [177] T.R. Garrett, M. Bhakoo, Z. Zhang, Bacterial adhesion and biofilms on surfaces, Prog. Nat. Sci. 18 (2008) 1049–1056. doi:10.1016/J.PNSC.2008.04.001. [178] A.-M. Lakaniemi, V.M. Intihar, O.H. Tuovinen, J.A. Puhakka, Growth of Chlorella vulgaris and associated bacteria in photobioreactors, Microb. Biotechnol. 5 (2012) 69–78. doi:10.1111/j.1751-7915.2011.00298.x. [179] K. Watanabe, N. Takihana, H. Aoyagi, S. Hanada, Y. Watanabe, N. Ohmura, H. Saiki, H. Tanaka, Symbiotic association in Chlorella culture, FEMS Microbiol. Ecol. 51 (2005) 187–196. doi:10.1016/j.femsec.2004.08.004. [180] L.T. Carney, S.S. Reinsch, P.D. Lane, O.D. Solberg, L.S. Jansen, K.P. Williams, J.D. Trent, T.W. Lane, Microbiome analysis of a microalgal mass culture growing in municipal wastewater in a prototype OMEGA photobioreactor, Algal Res. 4 (2014) 52–61. doi:10.1016/J.ALGAL.2013.11.006. [181] S.P. Fulbright, A. Robbins-Pianka, D. Berg-Lyons, R. Knight, K.F. Reardon, S.T. Chisholm, Bacterial community changes in an industrial algae production system, Algal Res. 31 (2018) 147–156. doi:10.1016/J.ALGAL.2017.09.010. [182] R. DULBECCO, M. VOGT, Plaque formation and isolation of pure lines with poliomyelitis viruses., J. Exp. Med. 99 (1954) 167–82. doi:10.1084/JEM.99.2.167. [183] I. Joint, M. Mühling, J. Querellou, Culturing marine bacteria - an essential prerequisite for biodiscovery., Microb. Biotechnol. 3 (2010) 564–75. doi:10.1111/j.1751-7915.2010.00188.x. [184] N. Singhal, M. Kumar, P.K. Kanaujia, J.S. Virdi, MALDI-TOF mass spectrometry: an emerging technology for microbial identification and diagnosis., Front. Microbiol. 6 (2015) 791. doi:10.3389/fmicb.2015.00791. [185] A.D. Wieme, F. Spitaels, M. Aerts, K. De Bruyne, A. Van Landschoot, P. Vandamme, Effects of growth medium on matrix-assisted laser desorption-

157

ionization time of flight mass spectra: a case study of acetic acid bacteria., Appl. Environ. Microbiol. 80 (2014) 1528–38. doi:10.1128/AEM.03708-13. [186] J. Ghyselinck, K. Van Hoorde, B. Hoste, K. Heylen, P. De Vos, Evaluation of MALDI-TOF MS as a tool for high-throughput dereplication, J. Microbiol. Methods. 86 (2011) 327–336. doi:10.1016/j.mimet.2011.06.004. [187] U. Edwards, T. Rogall, H. Blöcker, M. Emde, E.C. Böttger, Isolation and direct complete nucleotide determination of entire genes. Characterization of a gene coding for 16S ribosomal RNA., Nucleic Acids Res. 17 (1989) 7843–53. doi:10.1093/NAR/17.19.7843. [188] I. Cleenwerck, N. Camu, K. Engelbeen, T. De Winter, K. Vandemeulebroecke, P. De Vos, L. De Vuyst, Acetobacter ghanensis sp. nov., a novel acetic acid bacterium isolated from traditional heap fermentations of Ghanaian cocoa beans, Int. J. Syst. Evol. Microbiol. 57 (2007) 1647–1652. doi:10.1099/ijs.0.64840-0. [189] B.K. Kim, Y.-W. Lim, M. Kim, S. Kim, J. Chun, J.-H. Lee, Y. Jung, EzTaxon: a web-based tool for the identification of prokaryotes based on 16S ribosomal RNA gene sequences, Int. J. Syst. Evol. Microbiol. 57 (2007) 2259–2261. doi:10.1099/ijs.0.64915-0. [190] T. Friedl, Evolution of the polyphyletic genus Pleurastrum (Chlorophyta): inferences from nuclear-encoded ribosomal DNA sequences and motile cell ultrastructure, Phycologia. 35 (1996) 456–469. doi:10.2216/i0031-8884-35-5- 456.1. [191] T. J White, T. Bruns, S. Lee, J. Taylor, M. A Innis, D. H Gelfand, J. Sninsky, Amplification and Direct Sequencing of Fungal Ribosomal RNA Genes for Phylogenetics, Pcr Protoc. a Guid. to Methods Appl. 31 (1990) 315–322. [192] F.W. Zechman, E.A. Zimmer, E.C. Theriot, USE OF RIBOSOMAL DNA INTERNAL TRANSCRIBED SPACERS FOR PHYLOGENETIC STUDIES IN DIATOMS1, J. Phycol. 30 (1994) 507–512. doi:10.1111/j.0022- 3646.1994.00507.x. [193] K. Leinweber, S. Müller, P. G. Kroth, A semi-automated, KNIME-based workflow for biofilm assays, BMC Microbiol. 16 (2016) 61. doi:10.1186/s12866-016-0676-9. [194] S.W. Jeffrey, G.F. Humphrey, New spectrophotometric equations for

158

determining chlorophylls a, b, c1 and c2 in higher plants, algae and natural phytoplankton, Biochem Physiol Pfl. 167 (1975). [195] L. de Winter, A.J. Klok, M. Cuaresma Franco, M.J. Barbosa, R.H. Wijffels, The synchronized cell cycle of Neochloris oleoabundans and its influence on biomass composition under constant light conditions, Algal Res. 2 (2013) 313–320. doi:10.1016/J.ALGAL.2013.09.001. [196] E.J. Stewart, Growing unculturable bacteria., J. Bacteriol. 194 (2012) 4151–60. doi:10.1128/JB.00345-12. [197] T. Martens, L. Gram, H.-P. Grossart, D. Kessler, R. Müller, M. Simon, S.C. Wenzel, T. Brinkhoff, Bacteria of the Roseobacter Clade Show Potential for Secondary Metabolite Production, Microb. Ecol. 54 (2007) 31–42. doi:10.1007/s00248-006-9165-2. [198] J. Zhou, Y. Lyu, M. Richlen, D.M. Anderson, Z. Cai, Quorum sensing is a language of chemical signals and plays an ecological role in algal-bacterial interactions., CRC. Crit. Rev. Plant Sci. 35 (2016) 81–105. doi:10.1080/07352689.2016.1172461. [199] R. Kothari, R. Prasad, V. Kumar, D.P. Singh, Production of biodiesel from microalgae Chlamydomonas polypyrenoideum grown on dairy industry wastewater, Bioresour. Technol. 144 (2013) 499–503. doi:10.1016/J.BIORTECH.2013.06.116. [200] M. Chen, L. Zhang, S. Li, S. Chang, W. Wang, Z. Zhang, J. Wang, G. Zhao, L. Qi, W. Xu, Characterization of cell growth and photobiological H2 production of Chlamydomonas reinhardtii in ASSF industry wastewater, Int. J. Hydrogen Energy. 39 (2014) 13462–13467. doi:10.1016/J.IJHYDENE.2014.03.132. [201] S.B. Ummalyma, E. Gnansounou, R.K. Sukumaran, R. Sindhu, A. Pandey, D. Sahoo, Bioflocculation: An alternative strategy for harvesting of microalgae – An overview, Bioresour. Technol. 242 (2017) 227–235. doi:10.1016/J.BIORTECH.2017.02.097. [202] S. Salim, R. Bosma, M.H. Vermuë, R.H. Wijffels, Harvesting of microalgae by bio-flocculation., J. Appl. Phycol. 23 (2011) 849–855. doi:10.1007/s10811- 010-9591-x. [203] L.A. Romanenko, N. Tanaka, V.I. Svetashev, N.I. Kalinovskaya, Poseidonocella

159

pacifica gen. nov., sp. nov. and Poseidonocella sedimentorum sp. nov., novel alphaproteobacteria from the shallow sandy sediments of the Sea of Japan, Arch. Microbiol. 194 (2012) 113–121. doi:10.1007/s00203-011-0736-3. [204] O. Zeriouh, J.V. Reinoso-Moreno, L. López-Rosales, B. Sierra-Martín, M.C. Cerón-García, A. Sánchez-Mirón, A. Fernández-Barbero, F. García-Camacho, E. Molina-Grima, A methodological study of adhesion dynamics in a batch culture of the marine microalga Nannochloropsis gaditana, Algal Res. 23 (2017) 240–254. doi:10.1016/J.ALGAL.2017.02.008. [205] R. Kandilian, E. Lee, L. Pilon, Radiation and optical properties of Nannochloropsis oculata grown under different irradiances and spectra, Bioresour. Technol. 137 (2013) 63–73. doi:10.1016/J.BIORTECH.2013.03.058. [206] M.J. Griffiths, C. Garcin, R.P. van Hille, S.T.L. Harrison, Interference by pigment in the estimation of microalgal biomass concentration by optical density, J. Microbiol. Methods. 85 (2011) 119–123. doi:10.1016/j.mimet.2011.02.005. [207] M. Windler, K. Leinweber, C. Rio Bartulos, B. Philipp, P. Kroth, Biofilm and capsule formation of the diatom Achnanthidium minutissimum are affected by a bacterium, J Phycol. 51 (2015). doi:10.1111/jpy.12280. [208] N. Ozbek, A. Baysal, S. Akman, M. Dogan, Solid-Phase Extraction, in: Anal. Sep. Sci., Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2015: pp. 1571–1594. doi:10.1002/9783527678129.assep057. [209] M.J. Telepchak, T.F. August, G. Chaney, Solid Phase Extraction Troubleshooting, in: Forensic Clin. Appl. Solid Phase Extr., Humana Press, Totowa, NJ, 2004: pp. 147–168. doi:10.1007/978-1-59259-292-0_8. [210] S. Rampadarath, K. Bandhoa, D. Puchooa, R. Jeewon, S. Bal, Early bacterial biofilm colonizers in the coastal waters of Mauritius, Electron. J. Biotechnol. 29 (2017) 13–21. doi:10.1016/J.EJBT.2017.06.006. [211] T. Jakobsen, T. Tolker-Nielsen, M. Givskov, Bacterial Biofilm Control by Perturbation of Bacterial Signaling Processes, Int. J. Mol. Sci. 18 (2017) 1970. doi:10.3390/ijms18091970. [212] C. Zhai, P. Zhang, F. Shen, C. Zhou, C. Liu, Does Microcystis aeruginosa have quorum sensing?, FEMS Microbiol. Lett. 336 (2012) 38–44. doi:10.1111/j.1574-6968.2012.02650.x.

160

[213] D.G. Davies, M.R. Parsek, J.P. Pearson, B.H. Iglewski, J.W. Costerton, E.P. Greenberg, The involvement of cell-to-cell signals in the development of a bacterial biofilm., Science. 280 (1998) 295–8. http://www.ncbi.nlm.nih.gov/pubmed/9535661 (accessed February 8, 2019). [214] P. Di Donato, A. Poli, V. Taurisano, G. Abbamondi, B. Nicolaus, G. Tommonaro, P. Di Donato, A. Poli, V. Taurisano, G.R. Abbamondi, B. Nicolaus, G. Tommonaro, Recent Advances in the Study of Marine Microbial Biofilm: From the Involvement of Quorum Sensing in Its Production up to Biotechnological Application of the Polysaccharide Fractions, J. Mar. Sci. Eng. 4 (2016) 34. doi:10.3390/jmse4020034. [215] M. Yun, Y.-K. Oh, R. Praveenkumar, Y.-S. Seo, S. Cho, Contaminated bacterial effects and qPCR application to monitor a specific bacterium in Chlorella sp. KR-1 culture, Biotechnol. Bioprocess Eng. 22 (2017) 150–160. doi:10.1007/s12257-016-0584-8. [216] R. Goss, C. Wilhelm, G. Garab, Organization of the pigment molecules in the chlorophyll a/b/c containing alga Mantoniella squamata (Prasinophyceae) studied by means of absorption, circular and linear dichroism spectroscopy, Biochim. Biophys. Acta - Bioenerg. 1457 (2000) 190–199. doi:10.1016/S0005- 2728(00)00101-8. [217] M.-Y. LI, J. ZHANG, P. LU, J.-L. XU, S.-P. LI, Evaluation of Biological Characteristics of Bacteria Contributing to Biofilm Formation, Pedosphere. 19 (2009) 554–561. doi:10.1016/S1002-0160(09)60149-1. [218] N.R. Stanley, B.A. Lazazzera, Environmental signals and regulatory pathways that influence biofilm formation, Mol. Microbiol. 52 (2004) 917–924. doi:10.1111/j.1365-2958.2004.04036.x. [219] J. Liu, Y. Tao, J. Wu, Y. Zhu, B. Gao, Y. Tang, A. Li, C. Zhang, Y. Zhang, Effective flocculation of target microalgae with self-flocculating microalgae induced by pH decrease, Bioresour. Technol. 167 (2014) 367–375. doi:10.1016/J.BIORTECH.2014.06.036. [220] A. Huang, L. Sun, S. Wu, C. Liu, P. Zhao, X. Xie, G. Wang, Utilization of glucose and acetate by Chlorella and the effect of multiple factors on cell composition, J. Appl. Phycol. 29 (2017) 23–33. doi:10.1007/s10811-016- 0920-6.

161

[221] M. Windler, K. Leinweber, C.R. Bartulos, B. Philipp, P.G. Kroth, Biofilm and capsule formation of the diatom Achnanthidium minutissimum are affected by a bacterium, J. Phycol. 51 (2015) 343–355. doi:10.1111/jpy.12280. [222] Z. Fang, R.L. Smith, J.A. Kozinski, T. Minowa, K. Arai, Reaction of d-glucose in water at high temperatures (410 °C) and pressures (180 MPa) for the production of dyes and nano-particles, J. Supercrit. Fluids. 56 (2011) 41–47. doi:10.1016/J.SUPFLU.2010.11.009. [223] P.S. Rajesh, V. Ravishankar Rai, Quorum quenching activity in cell-free lysate of endophytic bacteria isolated from Pterocarpus santalinus Linn., and its effect on quorum sensing regulated biofilm in Pseudomonas aeruginosa PAO1, Microbiol. Res. 169 (2014) 561–569. doi:10.1016/j.micres.2013.10.005. [224] J.A. Gutiérrez-Barranquero, F.J. Reen, M.L. Parages, R. McCarthy, A.D.W. Dobson, F. O’Gara, Disruption of N -acyl-homoserine lactone-specific signalling and virulence in clinical pathogens by marine sponge bacteria, Microb. Biotechnol. (2017). doi:10.1111/1751-7915.12867. [225] S. Heeb, M.P. Fletcher, S.R. Chhabra, S.P. Diggle, P. Williams, M. Cámara, Quinolones: from antibiotics to autoinducers, FEMS Microbiol. Rev. 35 (2011) 247–274. doi:10.1111/j.1574-6976.2010.00247.x. [226] J.R. Seymour, S.A. Amin, J.-B. Raina, R. Stocker, Zooming in on the phycosphere: the ecological interface for phytoplankton–bacteria relationships, Nat. Microbiol. 2 (2017) 17065. doi:10.1038/nmicrobiol.2017.65. [227] N. Meyer, A. Bigalke, A. Kaulfuß, G. Pohnert, Strategies and ecological roles of algicidal bacteria, FEMS Microbiol. Rev. 41 (2017) 880–899. doi:10.1093/femsre/fux029. [228] F. Azam, F. Malfatti, Microbial structuring of marine ecosystems, Nat. Rev. Microbiol. 5 (2007) 782–791. doi:10.1038/nrmicro1747. [229] H. Teeling, B.M. Fuchs, D. Becher, C. Klockow, A. Gardebrecht, C.M. Bennke, M. Kassabgy, S. Huang, A.J. Mann, J. Waldmann, M. Weber, A. Klindworth, A. Otto, J. Lange, J. Bernhardt, C. Reinsch, M. Hecker, J. Peplies, F.D. Bockelmann, U. Callies, G. Gerdts, A. Wichels, K.H. Wiltshire, F.O. Glockner, T. Schweder, R. Amann, Substrate-Controlled Succession of Marine Bacterioplankton Populations Induced by a Phytoplankton Bloom, Science

162

(80-. ). 336 (2012) 608–611. doi:10.1126/science.1218344. [230] H.-P. Grossart, G. Czub, M. Simon, Algae-bacteria interactions and their effects on aggregation and organic matter flux in the sea, Environ. Microbiol. 8 (2006) 1074–1084. doi:10.1111/j.1462-2920.2006.00999.x. [231] A.Z. Worden, M.J. Follows, S.J. Giovannoni, S. Wilken, A.E. Zimmerman, P.J. Keeling, Rethinking the marine carbon cycle: Factoring in the multifarious lifestyles of microbes, Science (80-. ). 347 (2015) 1257594–1257594. doi:10.1126/science.1257594. [232] P.G. Falkowski, The role of phytoplankton photosynthesis in global biogeochemical cycles, Photosynth. Res. 39 (1994) 235–258. doi:10.1007/BF00014586. [233] K. Haynes, T.A. Hofmann, C.J. Smith, A.S. Ball, G.J.C. Underwood, A.M. Osborn, Diatom-Derived Carbohydrates as Factors Affecting Bacterial Community Composition in Estuarine Sediments, Appl. Environ. Microbiol. 73 (2007) 6112–6124. doi:10.1128/AEM.00551-07. [234] D.J. Smith, G.J.C. Underwood, Exopolymer production by intertidal epipelic diatoms, Limnol. Oceanogr. 43 (1998) 1578–1591. doi:10.4319/lo.1998.43.7.1578. [235] A. Kouzuma, K. Watanabe, Exploring the potential of algae/bacteria interactions, Curr. Opin. Biotechnol. 33 (2015) 125–129. doi:10.1016/j.copbio.2015.02.007. [236] S. Abinandan, S.R. Subashchandrabose, K. Venkateswarlu, M. Megharaj, Microalgae–bacteria biofilms: a sustainable synergistic approach in remediation of acid mine drainage, Appl. Microbiol. Biotechnol. 102 (2018) 1131–1144. doi:10.1007/s00253-017-8693-7. [237] F.M.I. Natrah, P. Bossier, P. Sorgeloos, F.M. Yusoff, T. Defoirdt, Significance of microalgal-bacterial interactions for aquaculture, Rev. Aquac. 6 (2014) 48– 61. doi:10.1111/raq.12024. [238] H. Wang, R.T. Hill, T. Zheng, X. Hu, B. Wang, Effects of bacterial communities on biofuel-producing microalgae: stimulation, inhibition and harvesting, Crit. Rev. Biotechnol. 36 (2016) 341–352. doi:10.3109/07388551.2014.961402. [239] X. Ji, M. Jiang, J. Zhang, X. Jiang, Z. Zheng, The interactions of algae-bacteria

163

symbiotic system and its effects on nutrients removal from synthetic wastewater, Bioresour. Technol. 247 (2018) 44–50. doi:10.1016/j.biortech.2017.09.074. [240] M.P. Schultz, J.A. Bendick, E.R. Holm, W.M. Hertel, Economic impact of biofouling on a naval surface ship, Biofouling. 27 (2010). doi:10.1080/08927014.2010.542809. [241] E. Almeida, T.C. Diamantino, O. de Sousa, Marine paints: The particular case of antifouling paints, Prog. Org. Coatings. 59 (2007) 2–20. doi:10.1016/J.PORGCOAT.2007.01.017. [242] M.S.-L. Yee, P.-S. Khiew, W.S. Chiu, Y.F. Tan, Y.-Y. Kok, C.-O. Leong, Green synthesis of graphene-silver nanocomposites and its application as a potent marine antifouling agent, Colloids Surfaces B Biointerfaces. 148 (2016) 392– 401. doi:10.1016/J.COLSURFB.2016.09.011. [243] H. Dahms, S. Dobretsov, Antifouling Compounds from Marine Macroalgae, Mar. Drugs. 15 (2017) 265. doi:10.3390/md15090265. [244] J. Liu, Y. Song, W. Qiu, Oleaginous microalgae Nannochloropsis as a new model for biofuel production: Review & analysis, Renew. Sustain. Energy Rev. 72 (2017) 154–162. doi:10.1016/J.RSER.2016.12.120. [245] Z. Yaakob, E. Ali, A. Zainal, M. Mohamad, M. Takriff, An overview: biomolecules from microalgae for animal feed and aquaculture, J. Biol. Res. 21 (2014) 6. doi:10.1186/2241-5793-21-6. [246] V.A. Chepurnov, D.G. Mann, P. von Dassow, P. Vanormelingen, J. Gillard, D. Inzé, K. Sabbe, W. Vyverman, In search of new tractable diatoms for experimental biology, BioEssays. 30 (2008) 692–702. doi:10.1002/bies.20773. [247] J. Gillard, J. Frenkel, V. Devos, K. Sabbe, C. Paul, M. Rempt, D. Inzé, G. Pohnert, M. Vuylsteke, W. Vyverman, Metabolomics Enables the Structure Elucidation of a Diatom Sex Pheromone, Angew. Chemie Int. Ed. 52 (2013) 854–857. doi:10.1002/anie.201208175. [248] J. Frenkel, C. Wess, W. Vyverman, G. Pohnert, Chiral separation of a diketopiperazine pheromone from marine diatoms using supercritical fluid chromatography, J. Chromatogr. B. 951–952 (2014) 58–61. doi:10.1016/j.jchromb.2013.12.040.

164

[249] S. Moeys, J. Frenkel, C. Lembke, J.T.F. Gillard, V. Devos, K. Van den Berge, B. Bouillon, M.J.J. Huysman, S. De Decker, J. Scharf, A. Bones, T. Brembu, P. Winge, K. Sabbe, M. Vuylsteke, L. Clement, L. De Veylder, G. Pohnert, W. Vyverman, A sex-inducing pheromone triggers cell cycle arrest and mate attraction in the diatom Seminavis robusta, Sci. Rep. 6 (2016) 19252. doi:10.1038/srep19252. [250] E. Cirri, W. Vyverman, G. Pohnert, Biofilm interactions—bacteria modulate sexual reproduction success of the diatom Seminavis robusta, FEMS Microbiol. Ecol. 94 (2018). doi:10.1093/femsec/fiy161. [251] C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis, Nat. Methods 2012 97. (2012). [252] R. Kandilian, E. Lee, L. Pilon, Radiation and optical properties of Nannochloropsis oculata grown under different irradiances and spectra, Bioresour. Technol. 137 (2013) 63–73. doi:10.1016/J.BIORTECH.2013.03.058. [253] C. Kuhlisch, G. Califano, T. Wichard, G. Pohnert, Chapter 18. Metabolomics of intra- and extracellular metabolites from micro- and macroalgae using GC- MS and LC-MS, (2018). doi:10.5281/ZENODO.2578249. [254] C. Vidoudez, G. Pohnert, Comparative metabolomics of the diatom Skeletonema marinoi in different growth phases, Metabolomics. 8 (2012) 654–669. doi:10.1007/s11306-011-0356-6. [255] M.C. Chambers, B. Maclean, R. Burke, D. Amodei, D.L. Ruderman, S. Neumann, L. Gatto, B. Fischer, B. Pratt, J. Egertson, K. Hoff, D. Kessner, N. Tasman, N. Shulman, B. Frewen, T.A. Baker, M.-Y. Brusniak, C. Paulse, D. Creasy, L. Flashner, K. Kani, C. Moulding, S.L. Seymour, L.M. Nuwaysir, B. Lefebvre, F. Kuhlmann, J. Roark, P. Rainer, S. Detlev, T. Hemenway, A. Huhmer, J. Langridge, B. Connolly, T. Chadick, K. Holly, J. Eckels, E.W. Deutsch, R.L. Moritz, J.E. Katz, D.B. Agus, M. MacCoss, D.L. Tabb, P. Mallick, A cross-platform toolkit for mass spectrometry and proteomics., Nat. Biotechnol. 30 (2012) 918–20. doi:10.1038/nbt.2377. [256] F. Giacomoni, G. Le Corguille, M. Monsoor, M. Landi, P. Pericard, M. Petera, C. Duperier, M. Tremblay-Franco, J.-F. Martin, D. Jacob, S. Goulitquer, E.A. Thevenot, C. Caron, Workflow4Metabolomics: a collaborative research infrastructure for computational metabolomics, Bioinformatics. 31 (2015) 1493–1495. doi:10.1093/bioinformatics/btu813.

165

[257] E. Afgan, D. Baker, B. Batut, M. van den Beek, D. Bouvier, M. Čech, J. Chilton, D. Clements, N. Coraor, B.A. Grüning, A. Guerler, J. Hillman-Jackson, S. Hiltemann, V. Jalili, H. Rasche, N. Soranzo, J. Goecks, J. Taylor, A. Nekrutenko, D. Blankenberg, The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update, Nucleic Acids Res. 46 (2018) W537–W544. doi:10.1093/nar/gky379. [258] R. Wehrens, G. Weingart, F. Mattivi, metaMS: An open-source pipeline for GC–MS-based untargeted metabolomics, J. Chromatogr. B. 966 (2014) 109– 116. doi:10.1016/j.jchromb.2014.02.051. [259] W.B. Dunn, I.D. Wilson, A.W. Nicholls, D. Broadhurst, The importance of experimental design and QC samples in large-scale and MS-driven untargeted metabolomic studies of humans, Bioanalysis. 4 (2012) 2249– 2264. doi:10.4155/bio.12.204. [260] J. Chong, J. Xia, MetaboAnalystR: an R package for flexible and reproducible analysis of metabolomics data, Bioinformatics. (2018). doi:10.1093/bioinformatics/bty528. [261] D.A. Bender, Amino Acid Metabolism, John Wiley & Sons, Ltd, Chichester, UK, 2012. doi:10.1002/9781118357514. [262] C.W. Tabor, H. Tabor, 1,4-Diaminobutane (Putrescine), Spermidine, and Spermine, Annu. Rev. Biochem. 45 (1976) 285–306. doi:10.1146/annurev.bi.45.070176.001441. [263] S.W. Rampen, B.A. Abbas, S. Schouten, J.S. Sinninghe Damste, A comprehensive study of sterols in marine diatoms (Bacillariophyta): Implications for their use as tracers for diatom productivity, Limnol. Oceanogr. 55 (2010) 91–105. doi:10.4319/lo.2010.55.1.0091. [264] S.A. Amin, L.R. Hmelo, H.M. van Tol, B.P. Durham, L.T. Carlson, K.R. Heal, R.L. Morales, C.T. Berthiaume, M.S. Parker, B. Djunaedi, A.E. Ingalls, M.R. Parsek, M.A. Moran, E. V. Armbrust, Interaction and signalling between a cosmopolitan phytoplankton and associated bacteria, Nature. 522 (2015) 98–101. doi:10.1038/nature14488. [265] H.M. van Tol, S.A. Amin, E.V. Armbrust, Ubiquitous marine bacterium inhibits diatom cell division, ISME J. 11 (2017) 31–42. doi:10.1038/ismej.2016.112. [266] M.R. Seyedsayamdost, R.J. Case, R. Kolter, J. Clardy, The Jekyll-and-Hyde

166

chemistry of Phaeobacter gallaeciensis, Nat. Chem. 3 (2011) 331–335. doi:10.1038/nchem.1002. [267] T. Chen, Q. Zhao, L. Wang, Y. Xu, W. Wei, Comparative Metabolomic Analysis of the Green Microalga Chlorella sorokiniana Cultivated in the Single Culture and a Consortium with Bacteria for Wastewater Remediation, Appl. Biochem. Biotechnol. 183 (2017) 1062–1075. doi:10.1007/s12010-017-2484-6. [268] C. Paul, M.A. Mausz, G. Pohnert, A co-culturing/metabolomics approach to investigate chemically mediated interactions of planktonic organisms reveals influence of bacteria on diatom metabolism, Metabolomics. 9 (2013) 349– 359. doi:10.1007/s11306-012-0453-1. [269] L. Szabados, A. Savouré, Proline: a multifunctional amino acid, Trends Plant Sci. 15 (2010) 89–97. doi:10.1016/j.tplants.2009.11.009. [270] T.M. Hildebrandt, A. Nunes Nesi, W.L. Araújo, H.-P. Braun, Amino Acid Catabolism in Plants, Mol. Plant. 8 (2015) 1563–1579. doi:10.1016/j.molp.2015.09.005. [271] K. Gupta, A. Dey, B. Gupta, Plant polyamines in abiotic stress responses, Acta Physiol. Plant. 35 (2013) 2015–2036. doi:10.1007/s11738-013-1239-4. [272] D. Pal, I. Khozin-Goldberg, S. Didi-Cohen, A. Solovchenko, A. Batushansky, Y. Kaye, N. Sikron, T. Samani, A. Fait, S. Boussiba, Growth, lipid production and metabolic adjustments in the euryhaline eustigmatophyte Nannochloropsis oceanica CCALA 804 in response to osmotic downshift, Appl. Microbiol. Biotechnol. 97 (2013) 8291–8306. doi:10.1007/s00253-013-5092-6. [273] M. Watanabe, K. Maemura, K. Kanbara, T. Tamayama, H. Hayasaki, GABA and GABA receptors in the central nervous system and other organs., Int. Rev. Cytol. 213 (2002) 1–47. http://www.ncbi.nlm.nih.gov/pubmed/11837891 (accessed November 27, 2018). [274] M.A. Bromke, Amino Acid biosynthesis pathways in diatoms., Metabolites. 3 (2013) 294–311. doi:10.3390/metabo3020294. [275] M.I. Dar, M.I. Naikoo, F. Rehman, F. Naushin, F.A. Khan, Proline Accumulation in Plants: Roles in Stress Tolerance and Plant Development, in: Osmolytes Plants Acclim. to Chang. Environ. Emerg. Omi. Technol., Springer India, New Delhi, 2016: pp. 155–166. doi:10.1007/978-81-322-2616-1_9.

167

[276] M.A. Mausz, G. Pohnert, Phenotypic diversity of diploid and haploid Emiliania huxleyi cells and of cells in different growth phases revealed by comparative metabolomics, J. Plant Physiol. 172 (2015) 137–148. doi:10.1016/j.jplph.2014.05.014. [277] Y. Durmaz, Vitamin E (α-tocopherol) production by the marine microalgae Nannochloropsis oculata (Eustigmatophyceae) in nitrogen limitation, Aquaculture. 272 (2007) 717–722. doi:10.1016/J.AQUACULTURE.2007.07.213. [278] M.E. Davey, G.A. O’toole, Microbial biofilms: from ecology to molecular genetics., Microbiol. Mol. Biol. Rev. 64 (2000) 847–67. http://www.ncbi.nlm.nih.gov/pubmed/11104821 (accessed March 30, 2018). [279] R. Xiao, Y. Zheng, Overview of microalgal extracellular polymeric substances (EPS) and their applications, Biotechnol. Adv. 34 (2016) 1225–1244. doi:10.1016/j.biotechadv.2016.08.004. [280] J. Bohórquez, T.J. McGenity, S. Papaspyrou, E. García-Robledo, A. Corzo, G.J.C. Underwood, Different Types of Diatom-Derived Extracellular Polymeric Substances Drive Changes in Heterotrophic Bacterial Communities from Intertidal Sediments, Front. Microbiol. 8 (2017) 245. doi:10.3389/fmicb.2017.00245. [281] C. Van Colen, G.J.C. Underwood, J. Serôdio, D.M. Paterson, Ecology of intertidal microbial biofilms: Mechanisms, patterns and future research needs, J. Sea Res. 92 (2014) 2–5. doi:10.1016/j.seares.2014.07.003. [282] E. Mhatre, R.G. Monterrosa, Á.T. Kovács, From environmental signals to regulators: Modulation of biofilm development in Gram-positive bacteria, J. Basic Microbiol. 54 (2014) 616–632. doi:10.1002/jobm.201400175. [283] N.S. Webster, R. Soo, R. Cobb, A.P. Negri, Elevated seawater temperature causes a microbial shift on crustose coralline algae with implications for the recruitment of coral larvae., ISME J. 5 (2011) 759–70. doi:10.1038/ismej.2010.152. [284] J. Luo, P. Lv, J. Zhang, A.G. Fane, D. McDougald, S.A. Rice, Succession of biofilm communities responsible for biofouling of membrane bio-reactors (MBRs), PLoS One. 12 (2017) e0179855. doi:10.1371/journal.pone.0179855. [285] M. Burmølle, J.S. Webb, D. Rao, L.H. Hansen, S.J. Sørensen, S. Kjelleberg, Enhanced biofilm formation and increased resistance to antimicrobial

168

agents and bacterial invasion are caused by synergistic interactions in multispecies biofilms., Appl. Environ. Microbiol. 72 (2006) 3916–23. doi:10.1128/AEM.03022-05. [286] C.D. Boyd, G.A. O’Toole, Second Messenger Regulation of Biofilm Formation: Breakthroughs in Understanding c-di-GMP Effector Systems, Annu. Rev. Cell Dev. Biol. 28 (2012) 439–462. doi:10.1146/annurev-cellbio-101011-155705. [287] J.M.S. Rocha, J.E.C. Garcia, M.H.F. Henriques, Growth aspects of the marine microalga Nannochloropsis gaditana, Biomol. Eng. 20 (2003) 237–242. doi:10.1016/S1389-0344(03)00061-3. [288] N. Mangwani, S. Kumari, S. Das, Bacterial biofilms and quorum sensing: fidelity in bioremediation technology, Biotechnol. Genet. Eng. Rev. 32 (2016) 43–73. doi:10.1080/02648725.2016.1196554. [289] J.D. Shrout, R. Nerenberg, Monitoring Bacterial Twitter: Does Quorum Sensing Determine the Behavior of Water and Wastewater Treatment Biofilms?, Environ. Sci. Technol. 46 (2012) 1995–2005. doi:10.1021/es203933h. [290] K. Jawed, S.S. Yazdani, M.A. Koffas, Advances in the development and application of microbial consortia for metabolic engineering, Metab. Eng. Commun. (2019) e00095. doi:10.1016/J.MEC.2019.E00095. [291] J. Bader, E. Mast-Gerlach, M.K. Popović, R. Bajpai, U. Stahl, Relevance of microbial coculture fermentations in biotechnology, J. Appl. Microbiol. 109 (2010) 371–387. doi:10.1111/j.1365-2672.2009.04659.x. [292] S. Guleria, J. Zhou, M.A.G. Koffas, Nutraceuticals (Vitamin C, Carotenoids, Resveratrol), in: Ind. Biotechnol., Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2016: pp. 309–336. doi:10.1002/9783527807833.ch10. [293] M.B. Cooper, A.G. Smith, Exploring mutualistic interactions between microalgae and bacteria in the omics age, Curr. Opin. Plant Biol. 26 (2015) 147–153. doi:10.1016/j.pbi.2015.07.003. [294] B.-H. Kim, R. Ramanan, D.-H. Cho, H.-M. Oh, H.-S. Kim, Role of Rhizobium, a plant growth promoting bacterium, in enhancing algal biomass through mutualistic interaction, Biomass and Bioenergy. 69 (2014) 95–105. doi:10.1016/J.BIOMBIOE.2014.07.015.

169

[295] J.-P. Hernandez, L.E. de-Bashan, D.J. Rodriguez, Y. Rodriguez, Growth promotion of the freshwater microalga Chlorella vulgaris by the nitrogen- fixing, plant growth-promoting bacterium Bacillus pumilus from arid zone soils, Eur. J. Soil Biol. 45 (2009) 88–93. doi:10.1016/J.EJSOBI.2008.08.004. [296] M.R. Seyedsayamdost, G. Carr, R. Kolter, J. Clardy, Roseobacticides: Small Molecule Modulators of an Algal-Bacterial Symbiosis, J. Am. Chem. Soc. 133 (2011) 18343–18349. doi:10.1021/ja207172s. [297] E. Kazamia, D.C. Aldridge, A.G. Smith, Synthetic ecology – A way forward for sustainable algal biofuel production?, J. Biotechnol. 162 (2012) 163–169. doi:10.1016/J.JBIOTEC.2012.03.022. [298] F. Tulli, G. Chini Zittelli, G. Giorgi, B.M. Poli, E. Tibaldi, M.R. Tredici, Effect of the Inclusion of Dried Tetraselmis suecica on Growth, Feed Utilization, and Fillet Composition of European Sea Bass Juveniles Fed Organic Diets, J. Aquat. Food Prod. Technol. 21 (2012) 188–197. doi:10.1080/10498850.2012.664803. [299] K.I. Reitan, J.R. Rainuzzo, G. Øie, Y. Olsen, A review of the nutritional effects of algae in marine fish larvae, Aquaculture. 155 (1997) 207–221. doi:10.1016/S0044-8486(97)00118-X. [300] C. Paul, G. Pohnert, Induction of Protease Release of the Resistant Diatom Chaetoceros didymus in Response to Lytic Enzymes from an Algicidal Bacterium, PLoS One. 8 (2013) e57577. doi:10.1371/journal.pone.0057577. [301] H.S. El-Sayed, H.A.H. Ibrahim, E.A. Beltagy, H.M. Khairy, Effects of short term feeding of some marine microalgae on the microbial profile associated with Dicentrarchus labrax post larvae, Egypt. J. Aquat. Res. 40 (2014) 251–260. doi:10.1016/J.EJAR.2014.08.001. [302] E. Ganuza, C.E. Sellers, B.W. Bennett, E.M. Lyons, L.T. Carney, A Novel Treatment Protects Chlorella at Commercial Scale from the Predatory Bacterium Vampirovibrio chlorellavorus, Front. Microbiol. 7 (2016) 848. doi:10.3389/FMICB.2016.00848. [303] M.A. Kempkes, Pulsed Electric Fields for Algal Extraction and Predator Control, in: Handb. Electroporation, Springer International Publishing, Cham, 2016: pp. 1–16. doi:10.1007/978-3-319-26779-1_215-1. [304] M.T. Pereira, D.S. Rego, L.M.S. Redondo, Advantages of Pulsed Electric Field

170

Use for Treatment of Algae, in: Handb. Electroporation, Springer International Publishing, Cham, 2017: pp. 2355–2368. doi:10.1007/978-3- 319-32886-7_171.

171

172

Acknowledgement

173

174

The elaboration of this thesis has been a challenging experience only possible thanks to the many people involved. I counted with the valuable input from so many people in different ways. It is impossible to summarize my gratitude in these lines but I will try to do my best.

First, I should start with the people who trusted in me for this position. These are my promotors Luc Roef and Wim Vyverman. Thank you for giving me this opportunity and for your support during these 4 years. Sven Mangelinckx and Anne Willems have been also part of this adventure. Your feedback and input during uncountable meetings have been always appreciated. Thank you for that.

This project was only possible thanks to the ALFF network. I want to thank Claire Gachon for the enormous work in coordinating this group of passionate scientists. Thank you all ALFFies for all the good moments during these years. Meeting you in so many places became an enriching exercise in so many ways. Thank you Andrea, Noreen, Emilio, Miriam, Hetty, Tatiana Kathryn, Frederike, Frederik, Johannes, Sri, Shu-Min, Gianmaria and Lachlan. I wish you all the best and hope we can continue seeing each other during unofficial ALFF meetings.

The ALFF project gave me the opportunity to perform two long research stays in Konstanz and Jena (Germany). I would like to thank Peter Kroth and Georg Pohnert for their valuable input on several parts of my project and for giving me the opportunity of working with them.

I want to thank all co-authors, colleagues and students I had during this thesis. This was possible thanks to your scientific and extra-academic support. Willem, I feel lucky that I had the chance to work with you. Thank you for everything. Lachlan, I will always remember my stay in Konstanz as one of the most fruitful periods. Thank you for making me feel like at home during those months. Emilio, you inspired me in so many ways. Thank you so much.

I want to thank all my colleagues from Proviron for all the lessons learned: Mark, Michiel, Koen, Jorien, Luc and Tom. To all my colleagues from the Sterre: Frederike, Alex, Luz, Willem, Kathryn, Sam, Andrea, Renaat, Heidi, Maxime, Kenny, Quinten, Darja, Sien, Katerina, Eveline, Lotte, Eli, Sofie, Anja, Ilse and all the people I ever interacted with but are not mentioned in this list. I also want to thank all students I had, hoping they learned from me at least as much as I learned from them: Nuwanthi, Maarnix, Anke, Arthur, Griet and Leonore.

I want to thank my family for their enormous support. Thanks to my parents for caring about me in every situation and to my brother for his confidence on me. Gracias también a mi abuelo Pepe y a mi abuela Flor quienes a su manera siempre están al día de mis andadas apoyándome en todo. In general, I want to thank every relative, friend or random person who ever cared about my project. You all were inspiring even if you were not aware.

Last but not least, I want to use this chance to raise awareness of the alarmingly increasing mental issues among the scientific community. This is an every time more common phenomenon often derived from multiple factors like excess of pressure or lack of guidance. I invite the scientific community to think about this topic and talk openly about it to help to prevent more cases. This issue cannot be accepted as normal and it is our duty to improve the situation.

175

176

Curriculum Vitae

177

178

GIRALDO BRITO, Javier – MSc. Biotechnology

Date of birth: 04th of February 1991 +32 492 46 24 98 +34 665 26 33 28 [email protected] www.linkedin.com/in/javier-giraldo-brito

Work experience

From September 2015 PhD Researcher – The ALgal Microbiome: Friends or Foes (ALFF): MSC Initial Training Network funded by Horizon 2020, Marie Sklodowska-Curie grant agreement number 624575 Proviron Holding NV, Hemiksem, Belgium/Ghent University, Belgium Focus on microbial interactions during industrial microalgae cultivation Research stays during PhD at: Plant Ecophisiology Group, Konstanz University, Germany (Nov 2016 – Feb 2017) Institute for Inorganic and Analytic Chemistry, Jena, Germany (Mar – May 2017) May – September 2015 MSc Intern Netherlands Institute of Ecology (NIOO-KNAW), Wageningen (NL) Optimization of N:P uptake by microalgae fed with digested black water Sept 2014 – April 2015 MSc Thesis Bioprocess Engineering Department/Biorefinery, Wageningen UR, (NL) Novel flocculant techniques for harvesting marine microalgae Oct 2012 – April 2013 BSc Intern Chemical Engineering Department/GICOM (Composting group), Autonomous University of Barcelona, Spain Use of lipase produced by Solid State Fermentation for increasing biogas production

Education

2013 – 2015 Wageningen University and Research (WUR), Wageningen, The Netherlands MSc. In Biotechnology – Process technology MSc Thesis: Elucidating the mechanism behind flocculation of marine microalgae 2009 – 2013 Autonomous University of Barcelona (UAB), Barcelona, Spain BSc. In Biotechnology – Specialization in Biotechnology of bioprocesses BSc. Thesis: Process design for TPA production using SuperPro Designer

Scientific contribution

G.P. ‘t Lam, J.B. Giraldo, M.H. Vermuë, G. Olivieri, M.H.M. Eppink, R.H. Wijffels, Understanding the salinity effect on cationic polymers in inducing flocculation of the microalga Neochloris oleoabundans, Journal of Biotechnology, Volume 225, 2016, Pages 10-17, ISSN 0168-1656, https://doi.org/10.1016/j.jbiotec.2016.03.009. J.B. Giraldo, W. Stock, L. Dow, L. Roef, A. Willems, S. Mangelinckx, P.G. Kroth, W. Vyverman, M. Michiels, Influence of the algal microbiome on biofouling during industrial cultivation of Nannochloropsis sp. in closed photobioreactors, Algal Res. 42 (2019) 101591. doi:10.1016/J.ALGAL.2019.101591.

Giraldo, J. B., Stock, W., Fret, J., Willems, A., Mangelinckx, S.,Roef, L., Vyverman, W., Michiels, M.. Exploration of the algal microbiome during industrial cultivation of Nannochloropsis sp. in closed photobioreactors and its potential effect on lipids composition (in preparation). Cirri, E., Giraldo, J. B., Neupane, S., Willems, A., Mangelinckx, Roef, L., Vyverman, W., Pohnert, G. Towards microbiome engineering of industrial microalgae cultivation: effects of non-native bacteria on Nannochloropsis sp. growth, metabolome and biofouling (in preparation).

J. Giraldo, L. Roef, S. Mangelinckx, A. Willems, W. Vyverman and M. Michiels, Understanding the mechanisms behind biofilm formation in a closed photobioreactor system. Poster presented at Young Algaeneers Symposium, April 2016, Qawra, Malta J. Giraldo, L. Roef, S. Mangelinckx, A. Willems, W. Vyverman and M. Michiels, Key agents for biofilm formation in non-axenic closed photobioreactor systems. Student talk at Summer School Molecular and physiological regulation of medical and environmental microbial biofilms, September 2016, Leuven, Belgium J. Giraldo, L. Roef, S. Mangelinckx, A. Willems, W. Vyverman and M. Michiels, A look into potential applications of microbial consortia in the microalgae industry. Guest speaker at SynComBio Workshop, December 2016, Dusseldorf, Germany J. Giraldo, L. Roef, S. Mangelinckx, A. Willems, W. Vyverman and M. Michiels, Bio-fouling in closed photobioreactor systems during microalgae cultivation as consequence of its associated microbiome. Poster presented at 11th IPC, August 2017, Szczecin, Poland J. Giraldo, L. Roef, S. Mangelinckx, A. Willems, W. Vyverman and M. Michiels, Towards biological control during industrial microalgae cultivation in closed photobioreactors. Oral presentation at Young Algaeneers Symposium, May 2018, Oban, United Kingdom

179