University of Groningen

The ecology and evolution of of mycosphere-inhabiting Paraburkholderia spp. Pratama, Akbar Adjie

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): Pratama, A. A. (2018). The ecology and evolution of bacteriophages of mycosphere-inhabiting Paraburkholderia spp. Rijksuniversiteit Groningen.

Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 29-09-2021 Chapter 1

General Introduction

Akbar Adjie Pratama and Jan Dirk van Elsas

Partly published in Book chapter: The in soil – potential roles, activities and impacts. Van Elsas J.D., Trevors J.T., Rosado A.S., and Nannipieri P., Taylor and Francis, 2018. Modern Soil III. In press

General introduction | Chapter 1

1

Essentially, all life depends upon the soil. There can be no life without soil and no soil without life; they have evolved together (Charles E. Kellogg)

The living soil, soil activity hot spots and phages The living soil is a very heterogeneous and dynamic system that influences microbial contact, attachment, niche differentiation and diversity. Its heterogeneity stems from the diversity of soil physicochemical characteristics and soil environmental conditions (e.g. pH, salinity) that have shaped both macro- and microhabitats across soil space and time (Fierer, 2017). The most dynamic activity in soil can be observed in microhabitats that are nutrient-rich and have been denominated hotspots. The factors that affect the life in soil hotspots include soil density, redox potential and nutrient contents, the latter spurring dynamically active and diverse microbiomes. Key known activity hotspots in the living soil are (i) the rhizosphere, i.e. the soil volume surrounding living roots; (ii) the detritusphere, i.e. organic litter that may consist of plant and/or animal residues, (iii) biopores, i.e. the soil surrounding animal-caused burrows (mainly earthworms), also known as the drillosphere. Other biopores such as the termitosphere (burrow caused by termites) and myrmecosphere (burrow caused by ants) should also be noted, yet the extent to which these affect microbial dynamics is understudied. (iv) aggregate surfaces, i.e. hotspots formed as the consequence of the movement of dissolved organic matter (DOM) through soil pores. Finally (v), the mycosphere, i.e. the region surrounding fungal hypae in soil (for details, see Zhang et al., 2014) stands out as a highly relevant yet hitherto understudied soil activity hot spot. All hotspots are ecologically significant for the functioning of the soil (e.g. decomposition, mineralization as well as build-up of soil organic matter). In generic terms, they increase microbial processes and interactions and consequently accelerate microbial exchange pools and evolution. With respect to the mycosphere, we may safely state that it is a highly dynamic microhabitat in terms of local conditions that are shaped by the microbially-mediated as well as physicochemical processes that shift over time. This is discussed further in the next section. The dynamic conditions are thought to shape the lifestyle of mycosphere inhabitants.

11 Chapter 1 | General introduction

In general, soil microbiomes include bacteria, fungi, archaea, protozoa, as well as their respective viruses. Compared to bacteria, the study of soil viruses is still in its infancy. Adding up to it, the role of soil viruses as shapers of the ecology and the evolution of soil microbiomes is still poorly understood, as compared to marine counterparts. In the study reported in this thesis, I focus on the importance of soil viruses, especially bacteriophages, for the ecology and evolution of mycosphere inhabitants. The phage like or related genes in there lay the basis of this thesis. The importance of horizontal gene transfer (HGT) in soil is undoubtedly great, as evidenced by major studies on soil-derived bacteria. I conclude this introduction by developing concepts on the roles of phages in mycosphere dwellers, which yield hypotheses underlying the work described in each of the chapters.

The mycosphere in soil and the key mycosphere inhabitant Paraburkholderia terrae In soil microbiomes, bacteria are often dominant, both in terms of numbers, diversity and activity. Their diversity across soil space and time is immense and has been well documented (Torsvik et al., 1990). Moreover, their role in global ecological processes such as degradation, mineralization and fixation processes is essential. Next to bacteria, fungi also (i) are abundant with respect to cell numbers (expressed as hyphal length), and (ii) play significant roles in soil functioning. Some of the fundamental questions with respect to the ecology of soil microbial communities are therefore: how are bacterial-fungal interactions (BFI) established? What are the key players and roles in these interactions? What are the molecular/ecological mechanisms behind it? And, what are the consequences of such interactions for the ecosystem functions in soil and the benefits between the partners. The mycosphere is known to provide carbonaceous compounds that are released by fungal cells, including oxalate, glycerol, formate, acetate, fumarate, mannitol, trehalose, erythritol, arabitol, citric acid and amino acids (Boersma et al., 2010; Frey, 1997; Haq et al 2018). These offer ecological opportunities to soil bacteria that are able to utilize, and thrive on, these released compounds. Here, I briefly describe the system used in this study, the mycosphere inhabitant Paraburkholderia terrae interacting with host fungi in the soil. It was found this organism has intricate strategies to interact with soil fungi (Haq et al., 2014; Nazir et al., 2012; Warmink et al., 2011; Yang et al., 2016). P. terrae has been shown to have “high-affinity” fungal-interactive behavior, with the ectomycorrhizal Laccaria proxima in soil under hazel trees (see Figure 1.1). Thus, assessments based on culture-dependent (isolation and colony-plate counting) and culture-independent

12 General introduction | Chapter 1

assays (molecular detection of a type-3 secretion system (T3SS) proxy, the hrcC gene) for three consecutive years on L. proxima mycospheres showed significant increases 1 of P. terrae (exemplified by strain BS110), as compared to the corresponding bulk soil. This suggested the occurrence of positive selection of these organisms presumably by the nutrients released by the fungi (Warmink and van Elsas, 2008). Further studies in soil microcosms revealed a strong profiency of another P. terrae strain, denoted BS001, for migration along the growing hyphae of the soil saprotroph Lyophyllum strain Karsten. This fungal-interactive behavior possibly involved a complex array of mechanisms, including motility, the T3SS and biofilm formation, which was observed surrounding the fungi hyphae (Warmink and Van Elsas, 2009). Strain BS001 was indeed a proficient single-strain migrator along the growing hyphae, next to P. terrae strains BS007, BS110, DSM 17804T and P. hospita DSM 17164T (Nazir et al., 2012). The mechanisms behind the interaction of P. terrae with soil fungi have also been addressed extensively (Haq et al., 2014, 2016, 2017, Yang et al., 2016, 2017, 2018). The involvement of the T3SS was investigated by comparing the knock-out strain P. terrae sctD with the wild-type strain, as regards its growth, nutrient utilization ability and migration profiency. No significant differences were observed in terms of BS001Δ growth, nutrient utilization (BIOLOG GEN III plates assay) and migratory ability in single inoculation experiments. However, mixed (1:1 ratio) inoculation experiments showed the wild-type to outcompete the mutant. Thus the T3SS possibly played a subtle role in the comigration along fungal hyphae, helping P. terrae in attachment to the host cell surface. Interestingly, this behaviour was shown with both L. sp. strain Karsten and Trichoderma asperellum 302 (Yang et al., 2016). The ability of P. terrae to adhere to fungal cell walls was then investigated (using ELISA), with a focus on the glycosphingolipids ceramide monohexosides (CMHs) that possibly play roles in fungal cell envelopes. The results showed adherence of the P. terrae wild-type to CMH of L. sp. strain Karsen but not to that of T. asperellum 302 and to a lesser extent of the sctD mutant. Subsequently, addition of P. terrae BS001 enhanced the total biomass with L. sp. strain Karsten but not with T. asperellum 302 (Haq et al., 2016). Δ fliP pliN) revealed the essential role of functional flagella for migration along with fungal hyphae, next to a minor Work with a flagellar mutant (Δ ) and a T4P one (Δ positive effect of the T4P. In conclusion, the migration of P. terrae BS001 along fungal hyphae was posited to be a strongly flagellar-driven process (Yang et al., 2017). The movement of P. terrae strain BS001 towards the fungal-released compounds glycerol and oxalic acid has also been addressed (Haq et al., 2016, 2018). Both compounds indeed promoted the movement of the cells. A transcriptome study of P. terrae strain BS001 in confrontation with L. sp. strain Karsten on soil extract agar plates supported the data with respect to the key role of chemotaxis. Confrontation

13 Chapter 1 | General introduction

with fungal hyphae upregulated the expression of chemotaxis and flagellar motility genes, as well as of oxalate uptake and utilization gene clusters (i.e. oxalate/ formate antiporter oxalyl-CoA decarboxylase, formyl-CoA transferases, formyl- CoA transferase, tartronic semialdehyde reductase and glyoxylate carboligase).

G SOIL GRAM PER LOG-CFU A ? Mycosphere Abundance and diversity

Five-gene cluster 0.01

48 76 F BB 97

Genomic sequencing Biofilm O E and transcriptome C HO HO Oxalic acid HO O Glycerol D

HO HO

N.S. N.S.

Chemotaxis SWIMMING Mutants generated ***

WT ∆SctD ∆pilN ∆fliP

Mobility

Figure 1.1. Paraburkholderia terrae strains isolated from the mycosphere. (A) Abundance and diversity assessments using culture-dependent and culture-independent methods showed that P. terrae to be mycosphere inhabitant (Warmink and van Elsas, 2008), (B) P. terrae strain BS001 forming biofilms on Lyophyllum sp. Karsten hyphae (Haq et al., 2016; Nazir et al., 2012; Warmink and Van Elsas, 2009), (C) Mutants generated to investigate the mechanisms behind the interactions of P. terrae SctD pilN fliF (Yang et al., 2016, 2017, 2018), (D) Mobility/comigration along L. sp. hyphae is flagellar-driven (swimming mobility) and type 4 pili (T4P)BS001 andand fungaltype 3 hyphae; secretion Δ system, Δ (T3SS) and Δ offer an anchoring/ twitching devices (Yang et al., 2017, 2018), (E) P. terrae BS001 moved toward fungal released compounds, including glycerol and oxalic acid (Haq et al., 2016, 2018), (F) Genomic analyses and transcriptome studies enhanced our understanding on the fungal-interactive behaviour of P. terrae BS001, reveling the important of a five-gene cluster for energy generation at the fungus (Haq et al., 2014, 2017). (G) The “enigmatic role” of bacteriophages in shaping P. terrae genomes.

14 General introduction | Chapter 1

However, no enhancement of expression was observed for a glycerol uptake (GUP) gene. The latter finding indicated that (i) glycerol might have been at low concentration 1 and (ii) in the current setting, glycerol uptake might have occurred through another mechanism, such as passive diffusion through the membrane. Interestingly, a five-gene cluster was dynamically modulated and significantly upregulated upon physical contact with the fungal mycelium. This five-gene cluster includes genes for (i) alkyl hydroperoxidase AhpD family core domain , (ii) a cupin, (iii) a LysR transcriptional regulator, (iv) a putative nucleoside-diphosphate sugar epimerase (nmrR like) and (v) a conserved exported protein: alpha/beta hydrolase fold. To my surprise, this five-gene cluster, presumably involved in aerobic energy generation, was found to be conserved across several other fungal-interactive Paraburkholderia strains. In summary, to this point, it is hard to pinpoint to which extent and in what temporal manner P. terrae employs the plethora of capabilities in its interaction with soil fungi. However, the studies performed so far have certainly broadened the understanding of the interaction of P. terrae with its fungal counterparts in soil. Considering the soil microhabitat from which the P. terrae strains were obtained and in which the organism probably thrives, i.e. the mycosphere, one may posit that P. terrae needs to be phenotypically flexible. The reason is that the mycosphere can be short-lived, and so the organism may – following selection by active fungi – be delivered into the bulk soil, where it will be subjected to the vagaries of that system. It may thus be expected that such a flexible phenotype is supported by flexibility at the genetic level. Considering the latter, HGT was found to have contributed to the shaping of the genome of P. terrae strain BS001 (Haq et al., 2014). The comparative genome analyses performed indicated that only 4% of the P. terrae strain BS001 genome could be identified as the core genome, the remainder belonging to variable parts, including prophage and plasmid regions and genomic islands (Haq et al., 2014). However, the work by Haq et al, with one exception, stopped short of an in-depth analysis of the prophage/island regions. I therefore here placed a focus on these, using both phage culture-dependent (i.e. double-agar layer, spot-assay, induction) and –independent approaches (mining phage signal from bacterial genome and virome community metagenomics - viromics) as the method of choice.

Evolution in the mycosphere: horizontal gene transfer Bacterial evolution is rarely a simple speciation event. Rather, it is a complex process, in which processes like HGT, duplications, phage host switching and phage/island gains and losses play roles (Conow et al., 2010; Hutchinson et al., 2017). In early work, van Elsas et al., (1988) showed that the rhizosphere of wheat constitutes a habitat

15 Chapter 1 | General introduction

that incites gene (plasmid) transfer among the local bacteria. Fairly recently, Zhang et al., (2014) provided arguments for the notion that the mycosphere also constitutes a soil hotspot that drives HGT. In general, HGT is facilitated by mobile genetic elements (MGEs), which include, in a broad sense, bacteriophages, plasmids, genomic islands, transposons and integrons (see Figure 1.2). HGT occurs through the three main mechanisms conjugation, transformation and transduction. Transformation is the uptake of naked DNA, whereas conjugation involves the transfer of genetic material from donor to recipient cells by cell-to-cell contact. Transduction (generalized and specialized) is phage- mediated gene transfer in which donor DNA can be incorporated into the genome of the recipient cell (or persist as a plasmid). There is an increased interest in the study of HGT in soil because of the concern about the spread of DNA from genetically- modified plants into microbiomes, next to the key role of HGT as an accelerator of bacterial evolution in soil. The major HGT mechanism transduction or effects of prophage presence on bacterial fitness have, unfortunately, not been extensively studied in soil. A study by Herron and Wellington, (1994) showed that in harst soil conditions Streptomyces lividans TK24 lysogenic for phage KC301 decreased in abundance as a consequence of phage excision/lysis. This indicated a possible strategy of prophage KC301, i.e. active excision, allowing subsequent transduction, in order to survive. However, the ecological significant of this putative phage-mediated HGT is still unclear. Currently, it is estimated that between 1.6 and 32.6 % of all genes that occur in bacterial genomes have been acquired by HGT (Boto, 2010). HGT clearly provides a venue for the of newly-introduced (as well as existing) genes in bacterial genomes. Such genomic plasticity is at the basis of diversification and adaptive evolution (see Box 1.1). An early study by Freeman (1951) already demonstrated that Corynebacterium diphtheriae infected with gives origin to enhanced virulence, providing an evolutionary asset for the newly emerged organism. Furthermore, Strauch et al., (2008) showed that phage 2851 of Shigella sonnei transfers (stx) virulence genes to ’naive’ S. sonnei cells, turning these into virulent forms. Thus, major changes in bacterial lifestyle are driven by phage- dependent HGT processes. Clearly, phages play important roles in the evolution of pathogenic bacteria, resulting in the intricate co-evolution of these partners (see review by Brüssow et al., 2004). Key factors that influence phage fate in soil are the spatial separation between these and recipient. Water availability aids the mobility (and survival) of bacterial cells, within a range for optimal survival of between (roughly) 25 and 70% of WHC – water-holding capacity (Yang and van Elsas, 2018). Excess water in the soil leads to

16 General introduction | Chapter 1

unfavourable anaerobic conditions, while water scarcity leads to stress, forcing cell to become inactive or silent. 1 With respect to genetic barriers, DNA:DNA homology is a major determinant of recombinational success and hence a lack of homology poses a ‘general’ barrier to recombination-dependent gene ‘fixation’ mechanisms. Other barriers encompass the aforementioned incompatibilities of incoming DNA with available molecular machinaries pertaining to fixation in the recipient bacterium (van Elsas et al., 1986). A review by Baltrus, (2013) defined the ‘cost’ of HGT as a sum of all the inefficiencies in the genotypic/phenotypic changes that determine the evolutionary potential. This includes chromosomal distruption, misfolded protein, distruption in transcription- translation and suboptimal protein function and system-level effects (e.g. loss of phenotype, altered metabolic flux).

Transduction

A Bacteriophage

B Transposon D Integron

C Plasmid E Mobile gene cassette

Prophage

Figure 1.2. Transfer of DNA between bacterial cells facilitated by some of the mobile elements. (A) bacteriophage, (B) transposon, (C) conjugative plasmid, (D) integron and (E) mobile gene cassette.

17 Chapter 1 | General introduction

Box 1.1. Proposal on how to assess phage-mediated evolution?

Precise and accurate methods to quickly identify phage-mediated evolution events, for example HGT, are needed to examine the processes that have shaped the genomes of bacteria in soil. The current low-cost and advanced next-generation sequencing technologies and the growth of whole genome sequence availability have enabled large-scale analyses of such events. A review by de Carvalho and Loreto, 2012 divided the methods that allow to identify evolutionary events in complete genomes into so-called ‘surrogate’ or ‘comparative’ methods. The former class does not require the construction of phylogenetic trees to identify evolutionary events. The base composition of bacterial genomes is relatively uniform and it takes time for newly introduced genes to adapt to the recipient host genome. The methods are based on the proposition that newly acquired sequences harbor nucleotide compositions that are different from those in the recipient genome. The method examines the best match of genes in a genome. Such surrogate methods provide quick scans of evolutionary events. However, high rates of false positive results have been obtained and so care is required in interpretation of the results. The second method uses phylogenetically-based analyses to detect the occurrence of evolutionary events. This method can produce more details, as the involved species, time of the evolutionary event and direction of transfer can be estimated. Using phylogenetically based analyses allows to locate the order, absence or loss of

methods based on inferring phylogenetic incongruence. A disagreement between agenes robust in closely-relatedgene tree and species. the known Evolutionary species phylogenyevents can isalso often be identifiedbest explained using by invoking the occurrence of HGT (Fitzpatrick et al., 2008). Phylogenetic incongruence is often quite successful and has been widely adopted in microbial relatedness studies (Hutchinson et al., 2017). The drawback of phylogenetic tree analysis to identify HGT events lies in the fact that it generates large data sets, is laborious and requires computational skills and resources. Additionally, the donor- recipient unipartite network has been used to characterize the extent of HGT mediated by phahe transduction (Popa et al., 2017). However, in the end, the choice of methods to identify HGT events should be clearly based on the availability of data sets. Surrogate methods may be only applicable for large-scale analyses, quick scans, and no data are suitable for the use of comparative methods. Furthermore, comparative methods using phylogenetic tree analysis should be employed when full-genome data sets are available for the analysis.

18 General introduction | Chapter 1

Evolution, diversity and ecological role of phages Viruses 1 Viruses are extremely abundant agents (Suttle, 2007) and they are also highly diverse. Viruses are able to infect three domains of life, i.e. Eukarya, Bacteria and Archaea (see Table 1.1). The origin and evolution of viruses are still rather enigmatic. Three major hypotheses have been proposed (see reviews by Forterre, 2006): (1) The -first hypothesis: viruses emerged as first ‘organisms’ in the primitive- soup/precellular world, (2) The evolutionary reduction hypothesis: viruses originated from a simple cell that lost its replication machinery and became dependent on another cell to survive and (3) The evolutionary escape hypothesis: viruses originated from genes of cellular organisms that became selfish elements. A close inspection of the existing theories of viral origin and evolution led Koonin et al., (2006) to propose “an ancient virus world hypothesis” (see Figure 1.3). This hypothesis was based on the notion that the primordial world may have allowed a rampant and extensive flow of genetic information through a variety of selfish elements. The continuation of the “virus world” theory from the primitive soup world may have enabled the evolution of the first primitive selfish genetic elements into fully-fledged viruses (RNA to DNA based) and other elements to the pre-cellular and eukaryogenesis stages (Figure 1.3A). Thus, this hypothesis offers a scenario in which the “ancient virus world” stages, to certain extent viruses, contribute to the evolution of cellular life forms (bacteria, archaea and eukarya) (Figure 1.3B). The ancient virus world may connect the origin of viral lineages and give a perspective of the origin of different types of cells (see details in Koonin et al., 2006; Koonin and Dolja, 2013). Koonin also emphasized the importance of compartmentalization (inorganic to the pre-cellular compartment/membrane) as a prerequisite for successful gene transfers, ensuring a sufficiently high concentration of substrates and genetic elements for effective replication and, in consequence, evolution. This scenario, in an era in which descent is uncertain, blurs the distinction between monophyly and polyphyly origins of viruses, as no genes are shared (there is no universal virus gene, such the small subunit RNA gene in all life forms) among viruses. This appears to rule out any conventional notion of common origin. However, homologies at the structural and mechanistic levels, for example, between viral RNA and DNA helicases and replication mechanisms (i.e. rolling circle replication - RCR), as well as commonalities in viral structures/hallmarkgenes (e.g. gene), have been observed (Krupovic and Koonin, 2017). At present, viruses are seen as the driving force of evolution. Examples may be found across the board, however need to be addressed in the mycosphere- see Figure 1.3C.

19 Chapter 1 | General introduction Colson et al., 2017; Wang and Wang, 1991 Guo et al., 2014 Calisher and Higgs, 2018; Li et al., 2015 Ghabrial et al., 2015 References Salmond and 2015 Fineran, 2015 Rybicki, Prangishvili Prangishvili et al., 2017 . ), Sipoviridae Mycomononegaviridae viruses. vaginalis Trichomonas infects viruses (TVV), lamblia Giardia (giant amoebae virus) infects Caenorhabditis elegans Caenorhabditis virus infect Orsay , Arboviruses, Togaviruses, Togaviruses, Arboviruses, Flavivirus, Reoviruses, Mesoniviruses, Bunyaviruses, and Birnviruses Negiviruses Rhabdoviruses, , , , Chrysoviridae, Partitiviridae, Totiviridae, , , , , , Hypoviridae, , , Narnaviridae , and Examples Many examples: Many the Most phages belong to and ( , (TMV), Tomato mosaic Tobacco (TSWV), Tomato wilt tospovirus spotted Potato (TYLCV), leaf curl yellow mosaic virus Y (, PVY), Cauliflower virus X (CaMV) and Potato PVX). (, Inoviridae, and Leviviridae Inoviridae, types: viral Archaeal-specific , , and viruses: Cosmopolitan-Archaeal Podoviridae, Sipoviridae, Myoviridae, and Sphaerolipoviridae ii ) ) archaea-specific, without structural or without structural i ) archaea-specific, The knowledge of protozoal viruses is still in its infancy. Viruses viruses is still in its infancy. of protozoal The knowledge example, viruses. For RNA mostly are Protozoa that infect viruses. icosahedral dsRNA, Cryspoviruses, Knowledge still in its infancy. Nematode: natural vector for some for vector natural Nematode: still in its infancy. Knowledge virus that naturally virus is the first known plant viruses. Orsay It has or other nematodes. elegans Caenorhabditis infects genome. and ssRNA(+) viron icosahedral Arthropods are vectors of viruses that infect plants and mammals. of viruses that infect vectors are Arthropods (-)). viruses (ssRNA(+)/ssRNA RNA Mostly Lack extracellular phase, transmitted via hyphal anastomosis anastomosis via hyphal phase, transmitted Lack extracellular change host phenotype, may symptomless, Often or vertically. genome structures: Diverse used in biocontrol. hypovirulent, and circular linear ssRNA(-) linear ssRNA(+), linear dsRNA, ssDNA. Main Features Most studied viral type termed bacteriophage. Diverse genomes: Diverse bacteriophage. type termed Most studied viral Diverse linear and circular. and +ssRNA, dsRNA ssDNA, dsDNA, capsid, with or without icosahedral morphotypes, most have also pleomorphic forms and spherical to Filamentous envelope. versus lytic into divided phages are Based on their lifestyle, found. ones. lysogenic and non-extreme extreme as hosts occur across Widespread, Morphotypes divided or ssDNA. either dsDNA Have environments. ( categories: two into domains ( and eukaryote in prokaryote genetic counterparts (RT) ssRNA ssRNA(-), ssRNA(+), dsRNA, genomes: ssDNA, Diverse viruses. (RT) and dsDNA twinned (geminate) round, non-enveloped, Morphotypes diverse: rigid helical double-capsid, filamentous, icosahedral, incomplete spherical and allantoid, enveloped, Others have bacilliform. rods, form. flexuous highly cosmopolitan, with structure and genetic make-up similar to similar to and genetic make-up cosmopolitan, with structure viruses in other domains. -

The main features of soil viruses from three domains life. Protozoal Protozoal viruses Nematode Nematode viruses Arthropod Arthropod viruses Mycoviruses Viruses Bacterio Archaea virus Plant viruses phages

Bacteria Archaea Eukarya Domain Table 1.1.

20 General introduction | Chapter 1

A C Bacteria Archaea Eukarya (i) 1 (ii) 1

5 prophage genomes: 1

2

3 2 5

4 LUCA HGTs 5 P.hospita 1 3 4 species-cluster 4 2 3 Primoldial “soup” (RNA world) Host cell genome

The emergence of capsid genes, The emergence of B more complex molecules and RNA to DNA LUCA RNA world RNA virus DNA virus Modern virus world (i) (ii) (iii) (iv) (v) (vi)

Time Pre-Darwinian Darwinian Figure 1.3. . The ancient viral world hypothesis, as proposed by Koonin et al., (2006); (A) Depiction of pre-evolutionary (blue) and evolutionary (pink) time. Only after a “threshold” (LUCA- last universal common ancestor) is positioned, Darwinian evolution takes place. (B) Ancient viral world scenario, in which viruses evolve and contribute to the evolution of cellular life forms. (i) Initially, selfish genetic elements (e.g. group I introns) are formed, as a consequence of rampant gene and RNA flow in the primordial soup; flow related to compartmentalization, e.g. inorganic compounds, (ii) continuation of this event, including emergence of capsid genes and more complex molecules; small genomes, that later evolve into RNA viruses, (iii) RNA viruses using primitive enzymes, transforming RNA into DNA, (iv) Mixed RNA-DNA systems promote retroid elements (i.e. group II introns), later DNA viruses, (v) Viruses (RNA and DNA) involved in “birth” of early life forms (LUCA) (vi) Emergence of three domains of life: bacteria, archaea and eukarya. Viruses as the “effectors” of ecology and evolution. (C) Present-day impact of viruses and phages on evolution, (i) the integration of prophages, rearranging host cell genomes, adding moron genes that spur ecological fitness and (ii) phages as mobile genetic elements (MGEs) in the mycospere.

Bacteriophages Independent research in 1915 by Frederick Twort, and in 1917 by Félix d’Hérelle on micrococci and Shigella, respectively, discovered viruses of bacteria, denoted as bacteriophages (bacterium-eaters) (D’Herelle, 1917; Twort, 1915). From comparative genomics, we have learned that phages have significantly contributed to bacterial genome over evolutionary time (Touchon et al., 2009). Thus, often-new bacterial traits have emerged, allowing genetic flexibility to the host, which eventually helps

21 Chapter 1 | General introduction

to adapt quickly to conditions in local environments. A case in point is given by E. coli in which a plethora of genomic islands – often including prophages - have been discovered that each confer different ecological behaviour to the host (Touchon et al., 2009). Since their discovery, phages have significantly revolutionalized our fundamental understanding of biology, i.e. (i) establishing the central dogma of molecular biology, (ii) shown to be essential in ecological functioning and evolutionary of microbes and (iii) reservoir of accessory genes for modern biology techniques (e.g. CRISPR-Cas9) (see review by Salmond and Fineran, 2015). Although studies of soil viruses, including phages, have emerged only as from the early 1970s (see Table 1.2), they are still understudied. It is expected that soil viruses play significant roles in the ecological functioning and evolution of their hosts in soil.

Current understanding of, and new insights in, phage biology The two known phage lifestyles, lytic action and lysogeny, govern the ecology of host cells. They also significantly contribute to the fate of cell populations in soil. For example, lysis (caused by lytic phages) essentially controls bacterial population densities – the phenomenon fits the “kill-the-winner” theory that stipulates that lytic “predation” suppresses the most abundant cell populations in a microbial community, thus modifying host diversity (see Box 1.2 on the phage-bacteria dynamics in highly dense bacterial populations). Lysis of local cells by phages in a biofilm releases extracellular DNA (eDNA) and polysaccharide that strengthens the structural integrity of the biofilm (see Obeng et al., 2016; Secor et al., 2015) for details on the role of phages in bacterial biofilm). In contrast, integrated prophages can offer beneficial genes that may advance the fitness of host cells. The abundance of soil viruses has been determined in varying soil systems. A study by Srinivasiah et al., (2015) examined the mechanisms affecting the production and abundance of soil viruses as related to the extant bacterial community. Using microcosms, the authors focused on bacteriophage community dynamics. To avoid the growth of fungi, cycloheximide (100 µg/gdw) was applied. Three different carbon sources were added, i.e. chitin, yeast extract (YE) and methanol. Over time, soil samples were collected to assess both bacterial and viral community abundance and diversity, using fingerprinting assays. The results showed that soil viral titers in yeast 7 8 extract increased from ~1 × 10 to 3.4 × 10 viruses/gdw on day 7 and then declined to 1.6 × 108 viruses/ gdw on day 30. In parallel, the soil bacterial populations went from about 1 × 106 to 8 × 107 bacterial cells/gdw on day 3, after which these gradually increased to 11 × 107 bacterial cells/gdw on day 30. Similar results were found with chitin, albeit with some delay for both bacterial and viral community increases.

22 General introduction | Chapter 1

However, only subtle changes occurred in both communities with the methanol- treated and control microcosms. The fingerprint analyses (16S rRNA-based DGGE and 1 RAPD-PCR) across treatments for both communities showed temporal changes in the community structures. The bacterial community compositions shifted in accordance with the amendments, with regards to 16S rRNA DGGE; the chitin and YE treatments showed greatest changes in 30 days, revealing ~34% similarity with the initial point. The lowest change occurred in the control and methanol treatments, with 56% and 49% similarity between initial and final points (30 days), respectively. These results showed that the ability of particular bacterial populations to utilize certain carbon sources spurs the growth of specific phage hosts, and, as a consequence, specific phage populations.

Box 1.2. Phage-bacterial host dynamics in dense bacterial populations

Lytic viruses regulate bacterial populations in accordance with the classical predator-prey model (called kill-the-winner; KtW). In this model, lytic viruses lyse the most abundant (presumably active) bacterial type in a community, promoting population control and bacterial diversity. However, in situations in which the bacterial host population is highly dense and therefore less actively growing, the rate of lysis by phages is decreased and - consequently - the virus-to-microbe ratio (VMR) lowered. This model is known as the ‘piggy-back-the-winner’ (PtW) model (Knowles et al., 2016). Indeed, in highly-dense bacterial populations, VLPs have been less observed. On the other hand, the abundances of integrases and excisionases were shown to be increased. The PtW strategy has emerged in lysogenic viruses to mitigate the highly-dense host condition. The energy cost to build up resistance in the host through superinfection to viral reproduction is lower than that in mutation ((Knowles et al., 2016). The strategy is universally observed in environments like soil and marine settings, as well as the human gut. We still ignore the extent to which the mechanism operates in the mycosphere.

An evolutionary arms race often exists between a bacterial host and its phage. In consequence, a considerable diversity of (both bacterial and phage) defensive and offensive strategies has evolved. Thus, the restriction modification (R-M) system, the abortive bacterial infection (ABI) and clustered regularly interspaced short palindromic repeats and Cas associated (CRISPR-Cas) have emerged. CRISPR array and spacers have been used to study the dynamic interaction and evolutionary history between host/phage partners (Puschnik et al., 2017; Touchon et al., 2016). The heritable immunity conferred by CRISPR-Cas systems offers a major fitness advantage for bacterial cells against specific foreign DNA or, in some cases, RNA.

23 Chapter 1 | General introduction References Hegazi and Jensen, Hegazi 1973 Van Elsas and Van 1982; Vary Penido, 1980 and Halsey, Reanney and Reanney Marsh, 1973 Williamson et al. 2005 Ashelford et al. Ashelford 2003 Swanson et al. Swanson 2009 Azotobacter phages are agents that phages are Remarks This study assessed This study phages in soils lytic The papers highlight the The papers highlight of B. megaterium role potential phages in soil This study highlights that highlights This study Bacillus evolution. accelerate This study showed that land use showed This study influencing is a significant factor phages abundance and diversity in soils Developed methods for methods for Developed numbers of counting the total in soil bacteriophages Reported that majority of Reported are virus particles observed indigenous soil viruses

Methods used on Azotobacter DAL C32 chroococcum on Bacillus.megaterium DAL F4 and B. megaterium I2872), QM BI55I (ATCC respectively Bacillus cereus, B. on Bacillus cereus, DAL B. licheniformis, coagulans, B. pumilus, B. megaterium, B. sphaericus , subtilis, B.stearothermophilus TEM and EFM count (No significant difference both methods) from Direct counting TEM, on Direct CP6 and P.S. quinivorans aeruginosa PU21 Direct counting (TEM) Direct . a Viral Viral diversity Tailed phages Tailed Tailed phage MJ-1 Tailed and phages ) MP50 (Myoviridae MP9 to and Siphoviridae) Tailed phages Tailed Tailed phages Tailed Tailed phages Tailed Tailed phages Tailed spherical, (Caudovirales), filamentous bacilliform, viruses and rod-shaped

9 -1

-1 -1

8 - 3.4 - 1.5 - dry l0 dry 8 7 7

-1 -1 -1 pfu ml VLPs g VLPs pfu ml 10 8 6 - 10 9 Viral Viral abundance 1.2 x 10 10 suspension VLPs g VLPs suspension 1.5 x 10 > l0 suspension 1.1 - 1.2 x 10 soil 4.0 x 10 dry soil pfu ml x 10 x 10 2.15 x 10 VLPs g VLPs soil . Overview of studies on soil viruses Senecio jacobaea Soil devoid of vegetation of vegetation Soil devoid the sun, to and exposed New Zealand Soil type / habitat Illinois, USA and tropical and tropical Illinois, USA soil, Brazil Beta ( Beta Rhizosphere Amythest, var. vulgaris L, pratensis Poa L. Epibolium tetragonum Griseb. L., and Cardamine ) and respective flexuos bulk soils, UK Nile Valley, Egypt Nile Valley, Triticum ( Triticum Rhizosphere ), bulk soil, UK aestivum Agricultural soil, USA Agricultural Table 1.2

24 General introduction | Chapter 1

1 Narr et al. 2017 Pratama and van and van Pratama Elsas 2017 References Zablocki et al. 2014 Scola et al. 2017 et al. Paez-Espino 2016 This study showed that showed This study factor a key was sampling site shaping the abundance and for of soil community structure viruses Soil (inducible) prophage from from Soil (inducible) prophage BS437 terrae P. Remarks This study highlighted the highlighted This study viral to that contribute factors in abundance and diversity deserts ecosystem the role evaluated This study on gradient of aridity and xeric and phages the both microbial and community structure function in Namib desert soil, viral showed This study diverse distribution across habitat strong ecosystems; majority the vast specificity for some to of viruses, next cosmopolitan viruses. Microscopic enumeration, enumeration, Microscopic virus community fingerprinting approach Mitomycin C induction Mitomycin Methods used Virus community metagenomics (virome) study Virus community metagenomics (virome) study signal from Mining viral metagenomes bacterial . (Continued) a Not mentioned Tailed phages (ϕ437) Tailed Viral Viral diversity , phages Iridoviridae, Tailed , ssDNA viruses groups 18,470 viral Dominant-Caudovirales; Dominant-Caudovirales; gene: signature , Bicaudaviridae, , , and Retroviridae Rudiviridae

8 VLP 8 9 dry -1 -1 - 10 8 dry soil -1 Up to 10 Up to 2.3 - 6.4 x 10 Viral Viral abundance g 22 Mb virome 22 Mb virome reads 10,009 viral contigs identified (IMG/VR) 10 soil copies of major capsid gene ml VLPs g VLPs . Overview of studies on soil viruses Mycosphere Hyper-arid desert soil, Antarctic Soil type / habitat Namib desert soil, Africa Soil metagenomics land, soil, pasture Forest cropland Notes: DAL: double agar layer; PFU: Plaque-forming unit; TEM: transmission electron microscope; EFM: Epi-fluorescence Microscope; Mitomycin C: prophage Mitomycin C: Microscope; EFM: Epi-fluorescence electron microscope; Plaque-forming unit; TEM: transmission double agar layer; PFU: DAL: Notes: Table 1.2 a induction agent; IMG/VR: Integrated- microbial genome/Virus (https://img.jgi.doe.gov/cgi-bin/vr/main.cgi); VLP: Virus-like particles; ssDNA: single- Virus-like particles; ssDNA: (https://img.jgi.doe.gov/cgi-bin/vr/main.cgi); VLP: induction agent; IMG/VR: Integrated- microbial genome/Virus stranded DNA virus.

25 Chapter 1 | General introduction

CRISPR spacers harbor a legacy of past infections by phages and/or other mobile elements. However, according to the current database, CRISPR-Cas systems are only present in about 45% of the bacterial genomes sequenced so far. The reasons for this meager prevalence are currently unknown. It is possible that, given that defense systems other than CRISPR-Cas have evolved, some bacteria have depended on these. Thus, systems like BREX (Goldfarb et al., 2015) and type IV pilus (T4P) glycosylation have recently been discovered (Harvey et al., 2018). A very recent study even described the existence of nine novel antiphage systems, next to one antiplasmid system (for details see Doron et al., 2018). Hence, a full account of all key systems that allow bacteria to deal with phages lies still ahead of us. It is also important to give attention to non-model bacteria and archaea, as it might lead to the discovery of other novel antiphage mechanisms. In counter-defense, for example, anti-CRISPR proteins have evolved in phages. To date, 21 unique families of such anti-CRISPR proteins have been discovered (Pawluk et al., 2018). These are known as phage “moron” genes (Maxwell, 2017). Key “anti- CRISPRs” were recently discovered in Pseudomonas spp. phages (Maxwell, 2017; Pawluk et al., 2018). The genes for these “anti-CRISPR” proteins, such as AcrF1-5 and AcrE1-4, block CRISPR-Cas types I-F and I-E, respectively. The other “anti-CRISPRs” are encoded by anti-CRISPR associated gene 1 (aca1), giving the distinct proteins denoted AcrIIC1, AcrIIC2 and AcrIIC3 and AcrF6 to AcrF10 (Pawluk et al., 2018). The life cycle of E. coli phage lambda has long served as a simplified model for the lysis/lysogeny decision-making process, known as the “genetic switch”. Upon infection, phage lambda can make its host cell enter one of two distinct pathways, i.e. lysis or lysogeny. However, the trigger that ultimately determines the decision Bacillus subtilis rely on the host communication system to decide whether to go through has been unknown. A recent study discovered that the Spβ group phages of lytic or lysogeny phases upon infecting the bacterial host (Erez et al., 2017). This communication system, named arbitrium, is reminiscent of bacterial quorum sensing, in terms of function (as a cue of bacterial population activity is used) and protein sequence. It is regulated by a six-residue peptide known as arbitrium. High population densities of the bacterial host, leading to low per-cell arbitrium levels, induce the system to the lytic cycle. However, after several infection cycles, when the bacterial cell number is reduced, raised concentrations of arbitrium emerge. This signal leads phages to progressively enter the lysogenic state (for details see Erez et al., 2017). In summary, phage-host interactions are very intricate, and abundance-diversity, arms races and defense systems play key roles with bearing on ecological and evolutionary processes. Especially in soil, we are only at the beginning of exploring this phage-host dynamics and hence soil virome studies need greater attention.

26 General introduction | Chapter 1

The effect of bacteriophages on the host Here I will briefly give some examples of the eco-evolutionary effect of bacteriophages 1 on host bacteria. The current growing interest in phage studies is spurred by our increased understanding of the ’fitness advantages’ that can be provided by lysogenic conversion to the host cell. This includes alteration of pathogenicity, of stress tolerance, of biofilm formation capacity and of metabolism of the host. For example, Vibrio cholerae chromosome in a site-specific manner (Faruque et al., 2002). Acquisition filamentous phage CTXϕ was found to be able to become integrated in the V. cholerae was key to the emergence of V. cholerae as a pathogen. V. cholerae of the CTXϕ element by non-toxigenic and CTXϕ have likely coevolved for mutual infected animals and promotes the transmission of V. cholerae benefit. The production of cholera toxin by CTXϕ lysogens results in diarrhea in able to utilize host functions for reproduction without compromising V. cholerae. . In return, CTXϕ is Integrated phages also offer protection to secondary phage infection, thus benefiting the host (Bondy-Denomy et al., 2016). The costs of the Rex system in E. coli carrying E. coli HK97 were assessed in the absence and presence of the phage. As expected, E. coli HK97 populations were completely eradicated within short time, phage λ against even with a very low phage load. In contrast, E. coli

λ populations with the Rex system Rex system was clearly to strengthen the host defense mechanism, thus protecting showed decreasing growth as the phage titer was higher. The benefit of the phage λ the host cell populations. Phages thus mediate survival, diversification and selection of host populations, by a range of mechanisms (Refardt et al., 2013). In summary, the intricate relationship between phages and bacterial hosts is subjected to co-evolution –the reciprocal adaptation between interacting species in specific ecological niches (Obeng et al., 2016). It may shape microbial genotypic and phenotypic diversity. The examination of E.coli K-12 from which cryptic prophages were delected (CP4- prophages) has shown that host fitness in adverse environments can be promoted. 6, DLP12, e14, rac, Qin, CP4-44, CPS-53, CPZ-55, CP4-57 and Δ9 – deleted all cryptic For instance, osmotic pressure, oxidative and acid stresses were survived better, and biofilm formation was affected. Furthermore, prophages enhance host resistance to

(i.e. rac and Qin) (Wang et al., 2010). Also, phages have been found to affect host sub-lethal concentrations of quinolone (i.e. CP4-6 and rac) and β-lactam antibiotics metabolism, energy and nutrient acquisition (auxiliary metabolic genes – AMGs), as well as increase the early stages of biofilm formation. In summary, the presence of both lytic (i.e. population and diversity control, provide readily available nutrient and enhance hosts’ biofilm formation/structure) and lysogenic (i.e. AMGs, pathogenocity) phages affects the ecological functioning of individual host cells to the community level.

27 Chapter 1 | General introduction

This thesis General hypotheses of this thesis Given the current state of knowledge with respect to the lifestyle and genomics of the mycosphere inhabitant Paraburkholderia terrae, I hypothesized that mobile genetic elements, notably phages, have shaped the genomes of P. terrae strains in soil and the mycosphere, and thus modulated the lifestyle of this organism. I also hypothesized that such lifestyle modulation has, to a considerable extent, contributed to the capacity of P. terrae to interact with soil fungi.

Specific hypotheses In the light of the general hypotheses, I pose the following specific hypotheses for further exploration in this thesis: (1) Mycosphere inhabitants such as P. terrae share genetic potential that enable them to, for example, interact with fungal mycelium that emerges in a soil setting, (2) Given that fungal mycelium is dynamic in time and space, P. terrae mycosphere inhabitants need to be genetically ‘flexible’; they thus foster the presence of prophages as key contributors to the flexibility. Such prophages occasionally give rise to phage progeny in biofilms, which – to a certain extent – improves the ‘health’ of the biofilm and so benefits the mycosphere P. terrae community, (3) Phage communities have shaped the evolutionary trajectory of Paraburkholderia spp. with different life styles in different ways. Phage relatedness reflects host relatedness, (4) The dynamics of gene flow across Paraburkholderia spp in the mycosphere can be followed through CRISPR-Cas system analyses in the respective genomes and (5) The mycosphere Paraburkholderia community contains unexplore viral communities that, in conjunction with the host communities, differ from those in corresponding bulk soil.

Aim of the thesis and research questions The aim of this thesis was to assess the putative role of bacteriophages in the ecology and evolution of the mycosphere inhabitant Paraburkholderia terrae. Given that no data on this topic existed prior to this work, I ignored the extent to which lytic phages would be prevalent. Hence, I used both culture-dependent and culture-independent approaches to enable the detection of both lytic and temperate phages in the P. terrae mycosphere communities. The culture-dependent approach consisted of phage isolation on the basis of the classical double agar layer (DAL) method and a spot-assay method. In addition, I investigated the prevalence of prophage sequences

28 General introduction | Chapter 1

across a variety of Paraburkholderia genomes using sequencing and molecular tools. Subsequently, I used a combined culturing/metagenomics approach to analyze 1 viromes produced from the mycosphere as well as soil by various bioinformatics tools. This thesis aims to provide a foundation for the integration of Paraburkholderia phages into models that depict host ecology as well as genetic processes in the mycosphere. I focused on the following research questions: (1) How similar are the genomes of mycosphere inhabitants belonging to Paraburkholderia terrae and related species? I used genomic information obtained from P. terrae strains BS001, BS110, BS007 and BS4367; next to three type strains (P. terrae DSM 17804T, P. hospita DSM 17164T and P. caribensis DSM 13236T). Do these bacteria share the genetic potential that drives the fungal- interactive behaviour in the mycosphere? (2) What is the state-of-the-art of knowledge on the diversity and role of soil phages? Do phages play a role in the ecology and evolution of Paraburkholderia spp.? (3) Are prophages present in diverse mycosphere/rhizosphere/soil-inhabiting Paraburkholderia spp.? (4) How is the interaction between mycosphere dwellers and their phages in the mycosphere? (5) How diverse is the soil virome of a Paraburkholderia community selected from the mycosphere and to what extent did they contribute to host ecology and evolution?

Scope of this thesis This thesis is divided into three parts. In the first part, I focus on the system-of-interest in this study, i.e. mycosphere-inhabiting Paraburkholderia strains. In the second part, I examine the conceptual and theoretical perspective on soil viruses. Finally, in the last part, I study the newly discovered inducible prophage from P. terrae strain Paraburkholderia. Also,I provide insight into the viral community in the mycosphere. Hereafter, I briefly BS437, ϕ437, and its eco-evolutionary significance in the genus introduce the aims and findings of each thesis chapter.

Section I: Genomics-based dissection of mycosphere-inhabiting Paraburkholderia strains, yielding a basis for bioinformatics- assisted searches for prophages In chapter 2, in an attempt to give ecological significance to the current classification of soil Burkholderia spp., we reclassified the former B. terrae strains BS007, BS110 and BS437 to Paraburkholderia terrae. This reclassification was based on

29 Chapter 1 | General introduction

our molecular genome analyses and on the fact that these bacteria are excellent mycosphere colonizers. We, thus, report the draft genome sequences of these three fungal-interactive Paraburkholderia terrae strains. We compared the three genomes and showed that all strains possess the genetic repertoire necessary for adaptation to diverse soil niches, including those influenced by soil fungi, as well as plants. Furthermore, all genomes were found to harbor highly diverse metabolic (primary and secondary) systems, mobility, biofilm formation, secretion systems (T1SS, T2SS, T3SS, T4SS and T6SS), glycerol uptake systems and energy generating gene clusters. Finally, our analyses also highlight the presence of prophage-like sequences, complete or incomplete, in the three P. terrae genomes. In chapter 3, I used whole-genome sequencing to explore the evolutionary relationships and ecological versatilities of P. terrae DSM 17804T, P. hospita DSM 17164T and P. caribensis DSM 13236T, next to selected own P. terrae isolates. The type strains have been used as taxonomical and ecophysiological references for a subgroup within the genus Paraburkholderia. Our study showed that the previously designated P.terrae strains BS001, BS007, BS110 and BS437 were closely related to P.hospita DSM 17164T. They share the recommended threshold for species circumscription (using 95-96% threshold for ANIm and >0.99 for TETRA values) of the taxon. Moreover, the P. terrae type strain DSM 17804T also adhered to the criterion, but P. caribensis DSM 13236T did not. Through genomic analyses of all type strains we showed that they all share highly similar features with respect to fungal- (and plant-) interactive traits. This includes: secretion systems, biofilm formation, glycerol and oxalate uptake, IAA production, ACC deaminase and nodulation genes. Overall, the results confirm the tight relatedness of the type strains P. terrae DSM 17804T and P. hospita DSM 17164T, together with all own P. terrae isolates. This, next to the more remote species P. caribensis DSM 13236T.

Section II: Conceptualisation and theoretical perspective on (soil) viruses Chapter 4 examines the current huge gap in virome studies between soil and marine ecosystems. We examined the state-of-the-art of the environmental virome studies, with a focus on soil in comparison to marine systems. After a description of the selective forces exerted on organisms in marine versus soil settings, we examine the current understanding of phage abundance, diversity, evolution and putative role in these ecosystems. We also identify the current challenges in this area and suggest potential strategies to boost our knowledge with respect to the relevant questions on the soil virome.

30 General introduction | Chapter 1

In chapter 5, together with co-authors, I provide a perspective on the ecological and evolutionary consequences for bacteria interacting with phages, when phage 1 and host interests are aligned. Temperate phages can introduce a plethora of genes that provide functions of different kinds to their hosts. Such lysogenic conversions can take the form of stable phenotypic changes (e.g. auxiliary metabolic genes (AMGs), virulence factors) or increased host plasticity (e.g. stress tolerance, quorum sensing), thereby improving bacterial survival under changing conditions. We argue that phage–host antagonism and mutualism get blurred at the population scale. The negative consequences of infection that an individual may experience may actually lead to population-level benefits or vice versa. Furthermore, we examine the question whether bacterial hosts can lower immune barriers to phage infection, thereby facilitating infection by beneficial phages. Taking recent evidence together, we suggest that in many cases temperate phages are to be considered as being mutualistic as well as parasitic, at the same time.

Section III: Paraburkholderia viruses and their evolutionary trajectory In chapter 6, I present the first study on a prophage induced from the mycosphere- isolated P. terrae for the contention that all of the analyzed Paraburkholderia sequences contain BS437, denoted ɸ437. Moreover, our analyses showed evidence substantial amounts of prophage regions. Most of these regions were remnants of prophages that have been subjected to host selective pressure (deletion bias). As a consequence, prophages often lose their abilities to produce progeny. The inducible raised by successful induction with mitomycin C. The experimental results and phage ɸ437 was apparently ‘spontaneously’ released and its particle numbers were its host. Importantly, I found one moron gene (orf64) that potentially endows the host genomic analysis of ɸ437 showed evidence for a lysogenic lifestyle of this phage with with a superinfection defense mechanism against other phage infections, enhancing host fitness and biofilm formation. However, experimental works still need to be done to support this finding. In chapter 7, using the prophage found in chapter 6 and other prophages identified across the genus Paraburkholderia, we examined to what extent the relationship between Paraburkholderia and their prophages can be taken to have bearing on evolutionary processes. We found the genome structures of the full prophages in the selected Paraburkholderia species to be highly diverse, yet classifiable into five defined groups. Moreover, phage remnants were frequently found. We also found some genes (e.g. DNA adenine methylase) that might be advantageous for the host cell. Cophylogenetic analyses showed complex coevolutionary scenarios between the

31 Chapter 1 | General introduction

Paraburkholderia hosts and their prophages that include cospeciation, duplication, host-switching and phage losses. These coevolutionary scenarios also yielded insight into the interaction between these partners. Analysis of CRISPR-Cas systems observed in the Paraburkholderia genomes showed the record of past phage infections. In summary, our results revealed the importance and extent of the role of phages with respect to the coevolution with their Paraburkholderia host. In chapter 8, I performed a viral metagenomics (virome) study with focus on the mycosphere. Two fungi collected at a forest floor were selected for the study, next to a forest floor bulk soil. Using the respective microbiomes as inocula, we selectively grow Paraburkholderia/Burkholderia in semi-selective medium (PCAT), followed by growth of the resulting community and identification of the respective virome upon induction. This study thus addressed the dynamic diversity of the Paraburkholderia/ Burkholderia community from two forest habitats, the mycosphere versus the bulk soil, next to the diversity of its virome upon induction. In chapter 9, I present a synthesis and discussion of all results obtained throughout this thesis. I provide insight in the importance of integration of viruses into considerations of the soil microbial ecosystem. Phages play a significant role in the ecology and evolution of soil microbiomes, especially with respect to mycosphere inhabitants. Future research avenues in, and practical applications of, soil viromes for host cell and soil microbial community-level questions are also discussed.

32