THÈSE
Pour l'obtention du grade de
DOCTEUR DE L'UNIVERSITÉ DE POITIERS UFR des sciences fondamentales et appliquées Ecologie et biologie des interactions - EBI (Poitiers)
(Diplôme National - Arrêté du 7 août 2006)
École doctorale : Sciences pour l'environnement - Gay Lussac (La Rochelle) Secteur de recherche : Biologie des organismes - Interactions symbiotiques
Présentée par : Jessica Dittmer
Diversity of endosymbiotic bacterial communities in terrestrial isopods: the role of feminizing Wolbachia
and other major players Directeur(s) de Thèse : Didier Bouchon, Monique A. Johnson Soutenue le 16 décembre 2013 devant le jury
Jury :
Président Yann Héchard Professeur des Universités, Université de Poitiers Rapporteur Patrick Mavingui Directeur de recherche CNRS, Université de Lyon 1 Rapporteur Dieter Ebert Professeur, Université de Bâle, Suisse
Membre Didier Bouchon Professeur des Universités, Université de Poitiers Membre Monique A. Johnson Maître de conférences, Université de Poitiers Membre Fabrice Vavre Directeur de recherche CNRS, Université de Lyon 1
Pour citer cette thèse :
Jessica Dittmer. Diversity of endosymbiotic bacterial communities in terrestrial isopods: the role of feminizing Wolbachia and other major players [En ligne]. Thèse Biologie des organismes - Interactions symbiotiques. Poitiers : Université de Poitiers, 2013. Disponible sur Internet <http://theses.univ-poitiers.fr>
THÈSE
pour l’obtention du grade de
DOCTEUR DE L’UNIVERSITE DE POITIERS
FACULTE DES SCIENCES FONDAMENTALES ET APPLIQUEES
(Diplôme National – Arrêté du 7 août 2006)École Doctorale Sciences pour l’environnement Gay Lussac Spécialité: Biologie des organismes – Interactions symbiotiques
préparée au laboratoire Écologie et Biologie des Interactions, équipe Écologie Évolution Symbiose – Poitiers – France
présentée et soutenue publiquement le 16 décembre 2013 par
Jessica DITTMER
Host-associated Microbiota in Armadillidium
vulgare
: Feminizing Wolbachia and other major
players
Directeurs de thèse: Didier BOUCHON Monique JOHNSON
JURY
Pr. D. Ebert Professeur, Université de Bâle, Suisse Rapporteur
Dr. P. MAVINGUI Directeur de recherche CNRS, Université de Lyon Rapporteur
Pr. D. BOUCHON Professeur, Université de Poitiers Examinateur
"Life is like a box of chocolates –
you never know what you're going to get." Forrest Gump
"Les cloportes sont partout." Sagesse universelle
A mes amis.
Sans vous, ces pages
auraient été bien vides.
Remerciements
Acknowledgements
Danksagung
Je tiens avant tout à remercier mon directeur de thèse, Didier Bouchon. D'abord d'avoir accepté une thésarde qui a débarqué un jour au labo, avec un parcours qui ressemblait à rien, pour laquelle les bactéries étaient surtout quelque chose de très petit – et qui avait carrément refusé de travailler sur ce sujet de thèse quand il lui avait été proposé pour la première fois. Trois ans plus tard, les choses ont un peu changées et la thésarde en question ne regrette pas du tout d'être revenue sur sa décision. Merci également de m'avoir laissé la liberté de faire de ce projet de thèse mon projet de thèse, tout en m'apportant le soutien et les encouragements qu'il fallait. Et, le plus important, merci pour cette communication franche et le grand respect mutuel. Tout ça fait que je considère ce projet comme un bon travail d'équipe.
Merci à toute l'équipe EES pour leur accueil, d'abord en tant qu'étudiante vagabonde et ensuite en tant que doctorante – surtout à Mathieu Sicard, c'est aussi grâce à tes encouragements que j'ai postulé pour l'EMAE et osé faire une thèse. Merci également à l'équipe technique, en particulier à Maryline pour les belles photos de "l'intestin de Jessica" et à Jérôme pour mes belles séquences!
Mais avant tout, ce que je garderai toujours en mémoire de ces six ans à Poitiers, ce sont les gens géniaux que j'ai rencontrés ici et les amitiés qui en résultent. Ce n'est pas pour rien si cette thèse vous est dédiée, c'est grâce à vous que j'ai beaucoup plus confiance en moi qu'encore il y a quelques années et que je ne me suis pas perdue en cours de route.
Il y a d'abord la bande des anciens du 132: Vincent et Maureen, Lenka ("Deux idiots, une seule pensée"), Seb, Gaël, Isabelle et le petit Nicolas. Avec vous, je ferais volontiers le tour du monde – et en fait, on l'a déjà bien entamé avec un road trip épique au pays des barbares (i.e. l'Allemagne) et une mission spéciale dans la patrie du Slivovice – avec des conséquences
5
Ensuite, un grand merci aux trois mousquetaires: Sandrine, Myriam, mon p'tit gnome préféré (et que personne n'ose plus jamais t'appeler "petite crevette"!) et Lise – sisters in arms du premier au dernier jour de la thèse. C'était génial de passer ces années avec vous, qui aurait cru que quatre filles puissent travailler ensemble sans se massacrer!
Bien sûr, merci à vous, Monseigneur Vico Ier, Duc de Bretagne, et à Margot, nouvellement arrivée et déjà d'un grand soutien. J'espère qu'on t'as pas fait trop peur de la fin de thèse ce dernier temps… Et finalement, merci à nos voisins de la salle des Anonymous, y compris notre responsable en chef des services informatiques, Bouziane! A vous tous, merci pour toutes les conneries faites ensemble et les fous rires, ce qui fait que des expressions banales comme "vider la tête", "salade de fruits" ou "Honolulu" n'ont plus la même signification pour moi.
A titre plus personnelle, merci à toi, Sandrine – si cette thèse a été écrite, c'est en grande partie grâce à toi. Merci de m'avoir aidé à reprendre pied cet été, maintenant que j'ai écrit une thèse moi-même, je donne encore plus de valeur au temps que tu m'as consacré quasiment 24H/24 pendant ton rush final. J'espère pouvoir rendre cette grande preuve d'amitié un jour. De la même manière, merci à Joanne et Seb (t'étais un super grand frère!) pour vos soutiens lors de cette fin de thèse un peu plus turbulente que prévu.
Und das Beste zum Schluß: Danke an den besten Papa der Welt, Hanna, Coco, Tina und Vera – ihr seid das Beste von Deutschland und ich entschuldige mich hiermit für die dreijährige Funkstille und immer dieselbe Leier: Keine Zeit, zu viel zu tun… ab jetzt wird alles anders! Hoffe ich. Und die Hoffnung stirbt ja bekanntlich zuletzt…
Abstract
In recent years, there has been a shift of focus in symbiosis studies, away from the traditional 'one host-one symbiont' concept towards a more holistic, community-based approach. This concept takes into account that a host is not only associated with one bacterium, but harbours and interacts with a diverse bacterial community, the microbiota. Terrestrial isopods represent an excellent model system for the understanding of complex multipartite symbioses due to their well-characterised association with feminizing Wolbachia bacteria. To date, three different feminizing Wolbachia strains have been identified in Armadillidium vulgare, presumably representing different host-symbiont co-evolutionary histories. The aim of this PhD was to get a more complete picture of the terrestrial isopod microbiome and the role of the Wolbachia within the bacterial community. In order to achieve this, quantitative and metagenomic techniques were combined to characterize the microbiota of A. vulgare on multiple levels: (i) In field populations and laboratory lineages, (ii) in different host tissues, and (iii) in relation to Wolbachia infection status, i.e. presence/absence of Wolbachia as well as infection with different
Wolbachia strains. Surprisingly, no core microbiome was evidenced despite a relative homogeneity of
the bacterial consortia across host tissues. On the other hand, environmental bacteria had a strong impact on the symbiotic microbiota, which might account for the observed inter-individual variations.
Wolbachia represented the predominant member of the bacterial community in infected individuals
and was identified as an important factor influencing bacterial community structure. Apart from
Wolbachia, we detected a second highly abundant bacterium: Candidatus Hepatoplasma
crinochetorum, a facultative symbiont previously reported from the midgut caeca, was for the first time observed in all tested host tissues. The potential interactions of Wolbachia and Ca. H. crinochetorum constitute an interesting example for symbiont-symbiont relationships between two highly abundant members of a diverse bacterial community. In order to get insights into the potential interactions between the host and this relatively unknown symbiont, the genome sequence of Ca. H. crinochetorum was obtained. A preliminary analysis of functional interactions was carried out, focusing on the metabolic pathways involved in cellulose digestion. Combining the symbiont genome with data from the host whole transcriptome allowed us to identify symbiont- as well as host-specific cellulotic enzymes. This work opens new perspectives for the study of isopod-bacteria symbioses within the holobiont concept.
7
Résumé
Les recherches sur les associations symbiotiques ont connu récemment un changement de paradigme, d’une conception d’un dialogue bilatéral d’un hôte avec son symbiote vers un concept plus holistique de l’association intégrant l’ensemble de la communauté symbiotique. Ce concept prend en compte le fait qu’un hôte est rarement associé à un seul symbiote, mais héberge et interagit avec une communauté bactérienne diverse, le microbiote. Dans ce cadre, la bonne connaissance que l’on a des interactions isopodes terrestres-Wolbachia féminisantes constitue un excellent modèle d’étude des interactions multipartites complexes. A ce jour, trois Wolbachia féminisantes, issues d’histoires évolutives différentes, ont été identifiées chez l’hôte Armadillidium vulgare. L’objectif de cette thèse est d’avoir une vue plus complète du microbiote d’A. vulgare et de l’impact des souches de Wolbachia sur cette communauté bactérienne. Dans ce but, des approches quantitative et métagénomique ont été combinées afin de caractériser le microbiote d’A.vulgare à plusieurs niveaux d’intégration: (i) dans des populations naturelles et des lignées contrôlées de laboratoire, (ii) au sein de différents tissus hôtes et (iii) en relation avec le statut d’infection par Wolbachia (présence versus absence et existence de différentes souches). De manière surprenante, aucun microbiote coeur n’a été mis en évidence malgré une relative homogénéité des communautés symbiotiques dans les tissus de l’hôte. De plus, les bactéries environnementales ont un impact important sur le microbiote symbiotique, ceci pourrait expliquer les variations inter-individus observées. Nous montrons que chez les animaux infectés,
Wolbachia représente la bactérie dominante et le facteur majeur de la structuration de la communauté
symbiotique. Nous avons également mis en évidence la présence d’un autre constituant majeur du microbiote dans tous les tissus de l’hôte: Candidatus Hepatoplasma crinochetorum, un symbiote facultatif réputé jusque-là pour être seulement associé aux glandes digestives de l’intestin. Les interactions potentielles de Wolbachia et de Ca. H. crinochetum constituent donc un exemple intéressant des relations entre deux composants majeurs du consortium symbiotique et des conséquences sur l’hôte. Afin d’éclairer les interactions potentielles entre hôte et ce symbiote relativement peu connu, la séquence du génome de Ca. H. crinochetorum a été obtenue. Une analyse préliminaire des interactions fonctionnelles des voies métaboliques impliquées dans la digestion de la cellulose a été conduite. Une comparaison avec les résultats disponibles au laboratoire du séquençage massif du transcriptome de l’hôte nous a permis d’identifier des enzymes cellulolytiques spécifiques de l’hôte et du symbiote. Dans le cadre conceptuel de l’holobiont, cette étude ouvre de nouvelles perspectives dans l’étude de la symbiose chez les isopodes.
Contents
Remerciements...4
Abstract...6
Résumé...7
Introduction...12
I. Symbiosis from an evolutionary perspective...14
1.1 Definition...14
1.2 The diversity of symbiotic interactions...16
1.2.1 Primary endosymbionts...17
1.2.2 Mutualistic secondary endosymbionts...21
1.2.3 Reproductive parasites ...23
1.2.4 Endosymbiotic communities ...23
II. Wolbachia, a major player in arthropods...26
2.1 Introduction...26
2.2 Wolbachia-induced host phenotypes...28
2.2.1 Cytoplasmic Incompatibility (CI) ...28
2.2.2 Parthenogenesis...30
2.2.3 Male-killing...30
2.2.4 Feminization...30
2.2.5 From parasitism to mutualism...31
2.3 Interactions between Wolbachia and other microorganisms...33
2.3.1 Protection against parasites...33
2.3.2 Wolbachia-Wolbachia interactions in multiply infected hosts...36
2.3.3 Wolbachia co-existing with other bacteria...36
III. Armadillidium vulgare and its bacterial symbionts...38
9
Chapter 1: Quantitative insights into bacterial community dynamics...50
I. Objective...52
II. Publication: Host tissues as microhabitats for Wolbachia and quantitative insights into the bacterial community in terrestrial isopods...55
2.1 Abstract...56
2.2 Introduction...57
2.3 Materials and Methods...61
2.3.1 Animal Sampling...61
2.3.2 DNA Extraction and PCR verification of Wolbachia infection status...62
2.3.3 Host Population Genetics...63
2.3.4 Electron Microscopy of digestive tissues...60
2.3.5 Quantification of Wolbachia and the total bacterial community by real-time qPCR...66
2.3.6 Screening for bacteria associated with the midgut caeca...67
2.3.7 Statistical Analysis ...68
2.4 Results...68
2.4.1 Genetic diversity of host populations...68
2.4.2 Wolbachia screening in different host tissues and faeces...70
2.4.3 Wolbachia strain-specific tissue distribution patterns...72
2.4.4 Quantification of the total bacterial community ...75
2.5 Discussion...81
III. Conclusions...87
Chapter 2: Diversity of the bacterial communities associated with
Armadillidium vulgare...90
I. Objective...92
II. Introduction...94
III. Materials and Methods...98
3.1 Animal & Soil Sampling………...98
3.2 DNA Extraction...100
3.3 PCR verification of Wolbachia infection status...101
3.4 Multilocus Sequence Typing (MLST) of a pathogenic Rickettsiella strain………...101
3.6 Temperature Gradient Gel Electrophoresis (TGGE)…………...104
3.7 454 amplicon pyrosequencing...106
3.8 Data Analysis of 454 Sequences...107
IV. Results... ...113
4.1 Wolbachia infection in field populations...113
4.2 Phylogenetic identification of the pathogenic Rickettsiella strain...114
4.3 Bacterial communities associated with A. vulgare ...116
4.3.1Bacterial community composition based on TGGE fingerprints ...116
4.3.2Bacterial community composition based on 454 amplicon pyrosequencing...121
4.3.2.1 Impact of environmental bacteria on the host-associated microbiota... 121
4.3.2.2 Predominant members of the host-associated microbiota...130
4.3.2.3 Impact of Wolbachia infection on bacterial community structure...135
V. Discussion...138
VI. Conclusions...145
Chapter 3: Genomic insights into the symbiosis of A. vulgare and
Candidatus Hepatoplasma crinochetorum...148
I. Objective...150
II. Publication : Phylogenomics of Candidatus Hepatoplasma crinochetorum...152
2.1 Abstract...153
2.2 Main text...154
2.2.1 Phylogenomic analyses...155
2.2.2 Evolution of tryptophan tRNAs...158
2.2.3 CRISPR/Cas system...161
2.3 Conclusion...163
2.4 Materials and Methods...164
2.4.1 Genome sequencing and assembly...164
2.4.2 Genome annotation...165
2.4.3 Orthology and phylogenomic analyses...165
11
III. Functional insights into cellulose digestion...169
3.1 Objective...169
3.2 Candidate genes coding for cellulolytic enzymes in the Hepatoplasma genome...169
3.3 Candidate genes coding for cellulolytic enzymes in the A. vulgare transcriptome ...175
3.4 Discussion ...177
Conclusions & Perspectives...180
References...190
Annexes...214
Technical Annexes...216
Annex 1: Summary of Wolbachia titers……...218
Annex 2: List of primers...219
Annex 3: List of primers used for 454 sequencing...220
Annex 4: Temperature Gradient Gel Electrophoresis (TGGE)...221
Annex 5: Amplion Library Preparation for 454 Sequencing Workflow...222
Annex 6: QIIME Analysis Workflow...225
Annex 7: Electron Microscopy of digestive tissues...226
Publication & Poster.../...228
Publication: Influence of changing plant food sources on the gut microbiota of saltmarsh detritivores……….……….………230
Poster: Bacterial communities influenced by Wolbachia? Community structure and major players in the terrestrial isopod microbiota……….………..…242
I.
Symbiosis from an evolutionary perspective
1.1 Definition
According to the definition by Anton de Bary (1879), symbiosis represents the 'living together of unlike organisms'. In other words, it describes the intimate association of individuals of different species, independent of the fitness effects on the different partners. Thus, it includes mutualistic relationships (characterized by benefits for both partners) as well as commensalism (the symbiont benefits from the association without harming its host) and parasitism (the symbiont benefits from the association but negatively affects its host). In reality, however, host-symbiont interactions can be extremely dynamic and it may be more accurate to think of symbiotic interactions as anything along a continuum from parasitism to mutualism. This is illustrated by the fact that parasitic and mutualistic bacteria often use the same molecular mechanisms (secretion systems, mobile genetic elements or phages) to enter the host environment and to persist within it without being eliminated by the host's immune system (Dale & Moran 2006; Moran et al. 2005a; Oliver et al. 2009).
Apart from this distinction according to the fitness effect on the host, symbioses are also defined depending on the localization of the symbiont in relation to the host organism, distinguishing between symbionts living on the surface of the host (ectosymbiosis), within the host's body (endosymbiosis) or in the host's cytoplasm (endocytobiosis). Beyond this classification, symbioses can take diverse shapes, such that the relationship between
Introduction
15
host and symbiont can be eukaryotes (pollinating insects and plants), while animal-prokaryote endosymbioses are extremely widespread (Moran et al. 2008; Moya et al. 2008). A fascinating illustration of the diverse interactions that symbiosis can bring forth is the mealybug-bacteria-bacteria endosymbiosis: The mealybug Planococcus citri harbours endosymbiotic Betaproteobacteria (Candidatus Tremblaya princeps), which in turn harbour endosymbiotic Gammaproteobacteria (Candidatus Moranella endobia) (McCutcheon & von Dohlen 2011; von Dohlen et al. 2001). This constitutes the first described symbiosis between two bacterial taxa. Finally, the Alphaproteobacterium Candidatus Midichloria mitochondrii exhibits another particular endosymbiotic lifestyle, residing in the mitochondria of ovarian cells in the tick Ixodes ricinus (Beninati et al. 2004; Sassera et al. 2006; Sassera et al. 2011). This is of particular interest in the sense that mitochondria are themselves derived from ancient acquisitions of bacterial endosymbionts related to the extant Alphaproteobacteria (Sagan 1967).
The remainder of this introduction will focus on the diversity of stable, heritable, nonpathogenic interactions between arthropods and endosymbiotic bacteria. Symbiosis with bacteria is extremely widespread and well-researched in this group of animals, especially in insects (Fig. 1). Moreover, endosymbionts influence many aspects of host ecology and evolution, facilitating the exploitation of new ecological niches and interfering with host nutrition, reproduction, immunity and defense against natural enemies (Oliver et al. 2005; Pais et al. 2008; Ryu et al. 2008; Shigenobu et al. 2000; Tsuchida et al. 2004; Tsuchida et al. 2010; Werren et al. 2008; Wu et al. 2006; Zientz et al. 2004). These aspects will be illustrated in the following sections, with an emphasis on the recent shift of focus towards a more holistic, community-based approach to symbiosis (Feldhaar 2011; Ferrari & Vavre 2011; Zilber-Rosenberg & Rosenberg 2008).
Introduction
Fig. 1 Diversity of heritable endosymbionts of insects. Colours indicate different host-symbiont
relationships (from Moran et al. 2008).
1.2 The diversity of symbiotic interactions
Symbiotic associations with bacteria represent an important source of genetic variation to the host, thereby fuelling evolutionary innovations (Fraune & Bosch 2010; Gilbert et al. 2010; McFall-Ngai et al. 2013; Moran 2007; Rosenberg et al. 2007; Zilber-Rosenberg & Rosenberg 2008). This is already evidenced by the development of the eukaryotic cell itself, which likely originated from endosymbioses between several formerly free-living unicellular organisms (Sagan 1967). According to the endosymbiotic theory of cell evolution (Sagan 1967), the incorporation of photosynthetic and aerobic ancestors of existing Cyanobacteria and
Introduction
17
Interestingly, extant mutualistic endosymbionts of insects often have drastically reduced genome sizes (Akman et al. 2002; Bennett & Moran 2013; Gil et al. 2003; Hosokawa et al. 2006; McCutcheon & Moran 2012; Nakabachi et al. 2006; Perez-Brocal et al. 2006; Tamas et al. 2002). This genome reduction seems to go along with host-symbiont co-speciation over long time-scales, indicating the adaptation of endosymbiotic bacteria to a host-dependent lifestyle (Fig. 2). Thus, genome degeneration is caused by the accumulation of slightly deleterious mutations (Muller's ratchet) and less efficient purifying selection due to the small effective symbiont population size and frequent bottlenecks (transmission) inherent to intracellular confinement (McCutcheon & Moran 2012). As a result, these genomes retain only the most essential genes, most often representing functions that are required by the host, e.g. genes related to nutrient provisioning (Akman et al. 2002; McCutcheon & Moran 2012; Moran et al. 2005b). However, these minimal endosymbiont genomes remain distinct from organelles in that they are not present in all host cells and have not reached the same level of integration with the host cell, e.g. in terms of host protein reimportation mechanisms which are common for organelles (McCutcheon & Moran 2012).
1.2.1 Primary endosymbionts
Many insects entertain long-lasting associations with obligate mutualistic endosymbionts (primary symbionts, Fig. 1) (Buchner 1965; Koga et al. 2013; Moran et al. 2008; Moya et al. 2008). These associations are the result of ancient symbiont acquisitions (up to 270 mya) followed by co-speciation of the two partners, as indicated by congruent host-symbiont phylogenies (Aksoy et al. 1997; Munson et al. 1991; Tamas et al. 2002) and reduced symbiont genomes (see above) (Akman et al. 2002; Shigenobu et al. 2000). Moreover, host and symbiont are interdependent, i.e. the symbiont has lost the capacity to live outside of its host and elimination of the symbiont negatively affects host fitness (Fukatsu & Hosokawa
Introduction
2002; Pais et al. 2008). The close relationship between the two partners is exemplified by the fact that the symbiont population is usually confined to specialized cells (the bacteriocytes) forming an additional host organ, the bacteriome (Buchner 1965). Symbionts are vertically transmitted from mother to offspring, in most cases via oocytes. However, other transmission routes exists, e.g. via the milk glands in viviparous tsetse flies (Aksoy & Rio 2005; Attardo et al. 2008) or via symbiont capsules deposited with the egg clutch in stinkbugs (Fukatsu & Hosokawa 2002; Hosokawa et al. 2006). In contrast to direct vertical transmission via the oocyte route, the symbiont capsules represent a means for stable environmental vertical transmission. These associations have generally evolved on a nutritional basis in that the symbiont provides essential nutrients lacking from the host's diet, thereby allowing the host to exploit new ecological niches by feeding on nutrient-deficient diets such as plant sap or vertebrate blood (Zientz et al. 2004).
Fig. 2 Genome dynamics for different host adaptation stages. Arrows pointing directly to the genome
indicate gene acquisition by horizontal gene transfer, arrows that loop back indicate changes within the genome, arrows that point away from the genome indicate gene loss or transfer to the host genome. Arrow weight represents the relative importance of each type of event (from Toft & Andersson 2010)
Introduction
19
One of the most famous examples is the aphid primary symbiont Buchnera (Gammaproteobacteria). Aphids feed on phloem sap, which is rich in sugars but lacks vitamins and several essential amino acids. In most aphid species, Buchnera synthesizes several essential amino acids from metabolites provided by the host, highlighting the metabolic interdependency between host and symbiont (Douglas 1998; Moran et al. 2005b; Shigenobu et al. 2000; Tamas et al. 2002). An exception is the aphid A. cedri. In this case, the primary symbiont (representing the smallest genome size known for Buchnera) has lost the metabolic capacity to synthesize tryptophan due to genome erosion. Hence, it has been suggested that Serratia symbiotica (Gammaproteobacteria), usually a facultative symbiont from the aphid's perspective, may compensate for the lost function required by the host. In the long run, this may result in the replacement of Buchnera by Serratia, underlining the dynamic nature of host-symbiont interactions (Gomez-Valero et al. 2004; Perez-Brocal et al. 2006). Another well-studied insect-primary symbiont association is the tsetse fly-Wigglesworthia (Gammaproteobacteria) symbiosis. In contrast to other primary symbionts, Wigglesworthia provides the host with vitamins and co-factors lacking from the vertebrate blood diet but has lost most amino acid biosynthesis pathways (Akman et al. 2002; Aksoy & Rio 2005; Pais et al. 2008; Zientz et al. 2004). Moreover, this symbiont is even more intricately linked to host fitness, since its absence results in decreased female fecundity, longevity and blood meal digestion, but a higher susceptibility to trypanosome infection (Pais et al. 2008).
These examples are by no means exhaustive. In fact, similar types of nutritional symbioses are known from carpenter ants (Blochmannia), cockroaches (Blattabacterium), psyllids (Carsonella), whiteflies (Portiera), among others, suggesting that these intimate associations are extremely widespread in insects (de Souza et al. 2009; López-Sánchez et al. 2009; Nakabachi et al. 2006; Sabree et al. 2009; Thao & Baumann 2004) (Fig. 1). However,
Introduction
the situation is slightly different concerning the most ancient nutritional symbiosis known to date, i.e. the association of numerous insects of the suborder Auchenorrhyncha (comprising plant- and leafhoppers, spittlebugs and cicadas) with the bacterium Candidatus Sulcia muelleri (Bacteroidetes, hereafter referred to as "Sulcia") (McCutcheon & Moran 2007; Moran et al. 2005d; Wu et al. 2006). Based on the congruent phylogenies of hosts and symbionts, it has been estimated that the initial acquisition of Sulcia by the common ancestor of all extant Auchenorrhyncha occurred > 270 mya (Moran et al. 2005d). Similar to other nutritional endosymbionts in the insect world, Sulcia provides its host with 7 or 8 out of 10 essential amino acids lacking from the hosts' diets (Bennett & Moran 2013; McCutcheon et al. 2009; McCutcheon & Moran 2007, 2010; Wu et al. 2006). The unusual aspect of these symbioses is that Sulcia is always accompanied by a second co-primary symbiont which is harboured in a second type of bacteriocytes (Bennett & Moran 2013; Koga et al. 2013; McCutcheon & Moran 2010; Wu et al. 2006). It has been proposed that the initial co-symbiont was a member of the Betaproteobacteria, which has been lost or replaced in some hosts in the course of speciation, resulting in the current distribution of different co-symbionts in a wide range of host species (Bennett & Moran 2013; Koga et al. 2013): Depending on the host, co-symbionts have been identified as Alphaproteobacteria (Candidatus Hodgkinia cicadicola), Betaproteobacteria (Candidatus Nasuia deltocephalinicola, Candidatus Vidania fulgoroideae, Candidatus Zinderia insecticola), Gammaproteobacteria (Candidatus Baumannia cicadellinicola, Sodalis-like). In all cases, however, the two co-symbionts have evolved a perfect metabolic complementarity, in that the second symbiont synthesizes the set of essential amino acids not provided by Sulcia, thereby establishing a stable tripartite symbiosis (Bennett & Moran 2013; McCutcheon et al. 2009;
Introduction
21
therefore represent striking examples of repeated convergent evolution of several bacterial partners to meet the host's metabolic needs.
1.2.2 Mutualistic secondary endosymbionts
In contrast to primary symbionts, secondary symbionts have been acquired more recently, as evidenced by the lack of congruence between host and symbiont phylogenies. Hence, the symbionts are facultative for the host, although their genomes already present clear signs of adaptation to the intracellular environment (e.g. in terms of genome reduction) compared to free-living bacteria (Aksoy & Rio 2005; Moran et al. 2005a). However, although secondary symbionts can also occur in bacteriocytes as it is the case in aphids and whiteflies, they may also exhibit wider tissue tropisms (Cheng & Aksoy 1999; Sakurai et al. 2005; Tsuchida et al. 2002). Due to the less intimate association of secondary symbionts with their hosts, these bacteria must evolve mechanisms that allow them to persist in host populations. One possible strategy consists in providing a fitness benefit to the host, analogous to (but not redundant with) obligate primary symbionts. Indeed, obligate primary symbionts often coexist with an array of secondary mutualistic symbionts, establishing highly specialised, small-scale endosymbiotic communities with synergistic effects on host fitness (Chiel et al. 2007; Gueguen et al. 2010; Russell et al. 2013; Tsuchida et al. 2002). These effects include increased thermal tolerance (Serratia symbiotica in aphids), host plant speciation (Regiella insecticola in aphids) and predator avoidance due to a change in body colour (Rickettsiella in aphids) (Ferrari et al. 2004; Montllor et al. 2002; Tsuchida et al. 2004; Tsuchida et al. 2010). However, the most frequently encountered symbiont-induced host fitness benefit consists in the defence against natural enemies, be it parasitic wasps, Plasmodium parasites, fungal pathogens or viruses. Hence, Hamiltonella defensa and, to a lesser degree, Serratia
Introduction
symbiotica as well as Regiella insecticola increase resistance to several parasitoids in aphids (Oliver et al. 2005; Oliver et al. 2003; Vorburger et al. 2010; Vorburger et al. 2009). In addition, Regiella insecticola, Rickettsia, Rickettsiella and Spiroplasma all reduce infection with the fungal pathogen Pandora neocephidis (Ferrari et al. 2004; Lukasik et al. 2013a; Lukasik et al. 2013b; Scarborough et al. 2005). Apart from the extensively studied aphids, the mollicute Spiroplasma confers resistance to sterilization caused by a parasitic nematode to its host Drosophila neotestacea (Jaenike et al. 2010b) and a midgut-associated Enterobacter inhibits infection with the human malaria parasite Plasmodium falciparum in Anopheles gambiae mosquitoes (Cirimotich et al. 2011). Finally, Wolbachia, well-known reproductive parasites, confer resistance to various RNA viruses (including dengue and chikungunya) as well as to several Plasmodium sp. in natively infected Drosophila as well as in transfected mosquito vectors (Bian et al. 2013; Blagrove et al. 2012; Hedges et al. 2008; Hughes et al. 2011; Moreira et al. 2009; Osborne et al. 2009; Teixeira et al. 2008).
The most diverse assemblages of secondary symbionts have been observed in aphids (8 genera in addition to the primary symbiont Buchnera (Moran et al. 2005c; Russell et al. 2013; Tsuchida et al. 2002)) and whiteflies (6 genera in addition to Portiera (Gueguen et al. 2010)). Since these symbionts are facultative for the host, different populations of a given host species can harbour various assemblages of secondary symbionts, depending on the prevailing environmental conditions (e.g. in terms of parasitic pressure or temperature) (Montllor et al. 2002; Oliver et al. 2008). In turn, the prevalence of secondary symbionts usually diminishes rapidly in populations where their presence does not represent a selective advantage (Oliver et al. 2008).
Introduction
23
1.2.3 Reproductive parasites
Other facultative bacteria use a different strategy to establish stable symbiotic associations without conferring an obvious benefit to their host: They interfere with the host's reproduction in order to enhance their own vertical transmission to the next generation (Duron et al. 2008; Werren et al. 2008). By far the best-studied and the most frequently encountered reproductive parasites are bacteria of the genus Wolbachia (Alphaproteobacteria) (Hilgenboecker et al. 2008; Werren et al. 2008). They are mainly maternally transmitted and have developed various strategies to manipulate host reproduction: Cytoplasmic incompatibility (CI) is a reproductive incompatibility preventing normal mitosis when a Wolbachia-infected male mates with a Wolbachia-uninfected female or when both carry different Wolbachia-strains (Serbus et al. 2008). Other strategies result in female-biased sex-ratio distortions caused by parthenogenesis, male-killing or the feminization of genetic males (Bouchon et al. 2008; Hurst et al. 2003; Narita et al. 2007a; Stouthamer et al. 1993). Other reproductive manipulators are Rickettsia (Alphaproteobacteria), Arsenophonus (Gammaproteobacteria), Cardinium and Flavobacterium (Bacteroidetes) and Spiroplasma (Mollicutes) (Duron et al. 2008). Currently, Wolbachia represents the only known bacterial genus capable of inducing all four reproductive phenotypes, although closely followed by Cardinium (all phenotypes except male-killing) (Hunter et al. 2003; Weeks et al. 2003; Zchori-Fein et al. 2001). Together, these bacteria infect a wide range of arthropod hosts, including insects, mites, spiders and isopod crustaceans.
1.2.4 Endosymbiotic communities
The diversity of symbiont-mediated effects on hosts and the existence of endosymbiotic consortia representing fitness advantages in certain conditions (e.g. aphid
Introduction
secondary symbionts) have been increasingly recognised in recent years. Along with this recognition came a change of paradigm in symbiosis research and the perception of an animal as a "holobiont", i.e. an individual harbouring and interacting with a diverse bacterial community, the microbiota (Feldhaar 2011; Gilbert et al. 2010; Zilber-Rosenberg & Rosenberg 2008). Along the same lines, the "hologenome" stands for the wealth of genetic information present in the holobiont, i.e. in the host genome as well as in the genomes of all symbionts taken together (Zilber-Rosenberg & Rosenberg 2008). These concepts have given rise to the (still somewhat controversial) "Hologenome Theory of Evolution" (Rosenberg et al. 2007; Zilber-Rosenberg & Rosenberg 2008): According to this theoretical framework, a host and its symbiotic community form a functional entity and natural selection would favour the most advantageous holobiont depending on the prevailing environmental conditions. Considering the enormous amount of genetic variation within the microbiota, the holobiont can adapt more rapidly to changing environmental conditions than the host genome on its own, e.g. via changes in community composition, horizontal gene transfers or mutations (Gilbert et al. 2010; Moran 2007; Zilber-Rosenberg & Rosenberg 2008). From this perspective, the association with a stable, heritable symbiotic bacterial community not only complements the host's metabolic capabilities but also represents a source of evolutionary novelty. This highlights the need to consider a host with its associated microbiota in order to fully understand the multipartite host- symbiont-microbiota interactions that shape such complex symbiotic relationships.
II.
Wolbachia, a major player in arthropods
2.1 Introduction
Bacteria of the genus Wolbachia are obligate intracellular, Gram-negative Alphaproteobacteria. Their closest relatives within the order Rickettsiales are the equally intracellular Anaplasma, Ehrlichia and Rickettsia (Fig. 3), genera that establish parasitic,
mutualistic and commensal relationships with both vertebrate and invertebrate hosts.
Historically, Wolbachia was first discovered in the ovaries of Culex pipiens mosquitoes almost a century ago by Hertig and Wolbach (Hertig & Wolbach 1924), hence the name of the type species Wolbachia pipientis (Hertig 1936). Wolbachia has attracted tremendous interest in symbiosis research during the last decades, due to the multitude of interactions along the parasitism-mutualism continuum that these endosymbionts establish in many different host species. Wolbachia is extremely widespread, infecting filarial nematodes and a diverse range of arthropods, with up to 65% of insect species estimated to be infected (Bandi et al. 1998; Hilgenboecker et al. 2008; Werren et al. 2008). Apart from insects, Wolbachia also infects mites (Breeuwer & Jacobs 1996), spiders (Rowley et al. 2004) and isopod crustaceans (Bouchon et al. 1998; Cordaux et al. 2012). Different Wolbachia strains are divided into clades, referred to as "supergroups" (Lo et al. 2007; Werren et al. 1995). However, the question whether different supergroups should be considered different "species" still remains unanswered (Ellegaard et al. 2013). To date, 13 supergroups (A-N) have been proposed, with supergroups A and B restricted to arthropod-Wolbachia (Werren et al. 1995) and supergroups
Introduction
27
respectively (Bordenstein & Rosengaus 2005; Czarnetzki & Tebbe 2004; Rowley et al. 2004). Supergroup F is of interest since it contains strains from both arthropods and filarial nematodes (Fig. 3) (Casiraghi et al. 2005). Several other strains could not be assigned to any of the existing supergroups.
Due to the predominantly maternal transmission of Wolbachia, one might imagine that Wolbachia would specifically target the reproductive tissues. While Wolbachia indeed seems to be restricted to the germ line tissues in some host species (e.g. many Glossina species) (Cheng et al. 2000; Doudoumis et al. 2012), Wolbachia can be much more widely distributed in somatic tissues, depending on the Wolbachia-host association. In fact, Wolbachia has been observed in virtually all major tissues including the brain, muscles, fat body, Malpighian tubules, salivary glands, haemolymph and midgut in several organisms (e.g. Drosophila, the mosquitoes Aedes albopictus and Culex pipiens, the butterfly Eurema hecabe, the bean beetle Callosobruchus chinensis and leafcutter ants) (Andersen et al. 2012; Dobson et al. 1999; Goto et al. 2006; Ijichi et al. 2002; Min & Benzer 1997; Narita et al. 2007b; Osborne et al. 2012; Zouache et al. 2009b).
Fig. 3 Phylogeny of Wolbachia. (a) Phylogenetic relationship of Wolbachia within the Rickettsiales. (b) Unrooted phylogenetic tree of the main Wolbachia supergroups (from Werren et al. 2008).
Introduction
2.2 Wolbachia-induced host phenotypes
The detection of Wolbachia in numerous host species has been accompanied by an appreciation of the diversity of interactions that exist between Wolbachia and their various host species. While Wolbachia are obligate mutualists in filarial nematodes (Bandi et al. 1998; Darby et al. 2012; Foster et al. 2005), they act as reproductive parasites in many arthropods, interfering with their host's reproduction in various ways in order to enhance their own vertical transmission (Werren et al. 2008). The reproductive phenotypes induced by Wolbachia are either cytoplasmic incompatibility (CI) or female-biased sex-ratio distortions caused by parthenogenesis, male-killing or the feminization of genetic males (Hurst et al. 1999; O'Neill et al. 1992; Rigaud et al. 1991; Rousset et al. 1992a; Stouthamer et al. 1993). These and several Wolbachia-host interactions less commonly observed in arthropods will be briefly described in the following sections.
2.2.1 Cytoplasmic Incompatibility (CI)
Cytoplasmic incompatibility is a reproductive incompatibility between sperm and egg, resulting in embryonic death in diploid species (Serbus et al. 2008). CI represents the most widespread Wolbachia-induced reproductive phenotype and might thus be the most ancestral effect from which all others evolved at later stages (Rousset et al. 1992b). It can be unidirectional, resulting in unviable crosses between infected males and uninfected females, or bidirectional, when males and females do not harbour the same Wolbachia strain. Therefore, infected females have a fitness advantage compared to uninfected females since
Introduction
29
females since they are at lower risk to suffer from CI (Mouton et al. 2004; Mouton et al. 2003).
CI is caused by a Wolbachia-induced sperm modification during spermatogenesis, resulting in improper condensation of the paternal chromosome set during mitosis, unless the modification is "rescued" by the presence of the same Wolbachia strain in the fertilized egg (Serbus et al. 2008; Werren 1997). Recently, Zheng et al. (Zheng et al. 2011) proposed the following explanation for this modification/rescue mechanism: Decreased expression levels of the regulator gene HIRA in the presence of Wolbachia would lead to a reduced histone deposition in the male pronucleus, resulting in the abnormal condensation of paternal chromosomes in CI embryos, unless compensated by increased histone deposition in infected females. Bidirectional incompatibility would then be due to different Wolbachia strain-specific patterns of gene expression modification.
Fig. 4 Wolbachia-induced reproductive phenotypes in arthropods. Wolbachia-infected individuals are
Introduction
2.2.2 Parthenogenesis
Thelytokous parthenogenesis, less common than CI, has been observed in haplodiploid hymenopterans (e.g. wasps) and in mites (Stouthamer et al. 1993; Weeks & Breeuwer 2001). In these species, sex determination is based on ploidy: Fertilized diploid eggs develop into females, while unfertilized haploid eggs develop into males. In infected females, Wolbachia induces a diploidisation of unfertilized eggs in the early embryonic development, resulting in female progenies (Fig. 4). This interference with host reproduction can have far-reaching consequences for host evolution: The fixation of Wolbachia-induced parthenogenesis may result in the loss of sexual reproduction even in the absence of Wolbachia, e.g. due to counter-selection of costly reproduction-related traits such as pheromone production, thereby establishing a dependence on Wolbachia for host reproduction (Kremer et al. 2009a).
2.2.3 Male-killing
The selective killing of male progeny during embryogenesis is a very common type of reproductive parasitism caused by all known reproductive parasites except Cardinium (Bandi et al. 2001; Duron et al. 2008). This strategy presents an advantage for females in species where siblings compete for resources or feed on unhatched eggs after birth. Wolbachia-induced male-killing has been found in coleopterans (e.g. ladybird beetles), lepidopterans and dipterans (Fig. 4) (Hurst et al. 2003; Hurst & Jiggins 2000; Hurst et al. 1999).
Introduction
31
Hiroki et al. 2004; Narita et al. 2007a; Narita et al. 2007b) and the leafhopper Zyginidia pullula (Negri et al. 2006). The three host species differ in terms of sex determination systems: While females are the heterogametic sex in isopods and E. hecabe (ZW/ZZ), Z. pullula has an XX/X0 sex determination system. Accordingly, Wolbachia has evolved different feminization mechanisms: In terrestrial isopods, Wolbachia interferes with the development and functioning of the androgenic gland and, hence, androgenic hormone production, thereby preventing the development of male characters (Bouchon et al. 2008; Cordaux et al. 2011). The exact mechanism and the target of Wolbachia, however, are not yet fully understood. This is also the case concerning the feminization in the butterfly E. hecabe. Nevertheless, Wolbachia needs to act throughout the entire larval development in order to achieve complete feminization (Narita et al. 2007a). Consequently, Wolbachia does not seem to interfere in the process of embryonic sex determination but rather to target some downstream mechanism responsible for the expression of the female phenotype. In contrast, the mechanism underlying the feminization of male Z. pullula leafhoppers has been deciphered: Wolbachia interferes with the host genetic imprinting, altering the gender-specific pattern of genome methylation (Negri et al. 2009). Hence, feminized males have the same genome imprinting as genetic females, while intersexes maintain the male methylation pattern, probably due to lower Wolbachia titers.
2.2.5 From parasitism to mutualism
In some particular cases, the Wolbachia-host interaction does not correspond to any of these most widespread phenotypes. On the one hand, this applies to wMelPop, a highly proliferating, pathogenic Wolbachia strain causing tissue degeneration and a drastically reduced life span in its native host Drosophila melanogaster (Min & Benzer 1997). In contrast, other cases show a more pronounced tendency towards mutualistic relationships.
Introduction
The parasitic wasp Asobara tabida, for instance, harbours stable triple Wolbachia infections. While two of these strains (wAtab1 and wAtab2) induce cytoplasmic incompatibility, the third strain (wAtab3) is required for the production of mature oocytes (Dedeine et al. 2005; Dedeine et al. 2001; Dedeine et al. 2004). The reason for this transition from a facultative to an obligatory symbiosis is that successful oogenesis in A. tabida has become dependent on the presence of the symbiont: Wolbachia down-regulates apoptotic processes in the ovaries via iron sequestration, thereby decreasing the levels of cytotoxic reactive oxygen species (ROS) (Kremer et al. 2009b; Pannebakker et al. 2007). The host, on the other hand, has evolved in response to Wolbachia infection, by enhancing apoptosis to compensate for the bacterial interference. Hence, the presence of Wolbachia is necessary to prevent apoptosis of nurse cells during oocyte maturation (Pannebakker et al. 2007). This underlines that parasites and mutualists often use the same tools in terms of molecular interactions with the host (in this case the host's apoptotic processes), in order to evade the host immune system and thus to persist in the intracellular environment.
In other cases, Wolbachia is in some ways involved in host nutrition. Similar to the A. tabida-wAtab3 symbiosis, the Wolbachia strain wMel also plays a role in iron homeostasis in Drosophila melanogaster, representing a fitness benefit during periods of nutritional stress (Brownlie et al. 2009). Another example implies Wolbachia infecting Acromyrmex leafcutter ants: Based on the extracellular localization of live Wolbachia in the ant gut and faecal droplets, a novel (albeit as yet undefined) mutualistic nutritional relationship in this ant-fungus symbiosis has been proposed (Andersen et al. 2012). In contrast, there is no doubt concerning the nutritional role of Wolbachia in the bedbug Cimex lectularius: Wolbachia has evolved a bacteriocyte-associated nutritional symbiosis, providing its hematophagous host
Introduction
33
2.3 Interactions between Wolbachia and other microorganisms
The previous sections have illustrated the important impact of Wolbachia on their diverse arthropod hosts. However, much less research has been devoted to the potential interactions between Wolbachia and other microorganisms they might encounter within the same host. The following sections will summarize the current knowledge regarding Wolbachia-induced protection against parasites, interactions between different Wolbachia strains in multiply infected hosts and interactions between Wolbachia and other bacteria co-existing in the same host.
2.3.1 Protection against parasites
A growing number of studies describe an effect of some Wolbachia strains that induce no or only weak reproductive manipulations: Similar to some facultative insect endosymbionts (e.g. Hamiltonella defensa), these Wolbachia strains have evolved in order to benefit their host, via protection against parasites such as viruses or Plasmodium spp. (Hedges et al. 2008; Kambris et al. 2010; Teixeira et al. 2008). However, no Wolbachia-induced protection against bacterial pathogens has been observed (Wong et al. 2011). In naturally infected Drosophila melanogaster and Drosophila simulans, Wolbachia induces resistance to numerous RNA viruses (Drosophila C virus, Flockhouse virus, Cricket paralysis virus, Nora virus) but not to the DNA virus IIV-6 (Insect Iridescent Virus 6) (Hedges et al. 2008; Teixeira et al. 2008). Depending on the Wolbachia strain, there are two mechanisms of antiviral protection: When infected with the Wolbachia strains wMel or wRi, this protection is due to Wolbachia interfering with virus proliferation, while viral titers in flies infected with wAu are similar to those in Wolbachia-uninfected hosts, rather indicating an increased host tolerance to virus infection (Osborne et al. 2009; Teixeira et al. 2008).
Introduction
Fig. 5 Localisation of Wolbachia and dengue virus (DENV-2) in Ae. aegypti mosquitoes. Double
immunofluorescence staining of mosquito sections showing the localisation of dengue virus (red) and
Wolbachia (green). (A) Dengue virus in fat tissue of Wolbachia-uninfected mosquitoes 14 days
post-DENV-2 thoracic injection. (B) Wolbachia in fat tissue of Wolbachia-infected mosquitoes 14 days post-DENV-2 thoracic injection. No dengue virus was detected. (C) Cellular exclusion of DENV-2 by
Wolbachia. The presence of both Wolbachia and DENV-2 was observed at very low frequency in a
small number of Wolbachia-infected outcrossed mosquitoes, 14 days post-DENV-2 infection. Dengue is only apparent in cells lacking Wolbachia. Scale bars: A, B = 20 µm, C = 50 µm (from Moreira et al. 2009).
Wolbachia-mediated protection against parasites also occurs in mosquitoes, some of which representing important vectors of major human pathogens such as the dengue and chikungunya viruses and Plasmodium parasites. However, in naturally Wolbachia-infected mosquitoes, the resistance phenomenon does not reduce within-host viral replication but
Introduction
35
strains from Drosophila (wMelPop) and Aedes albopictus (wAlbB) into naturally Wolbachia-uninfected Aedes aegypti or Anopheles mosquitoes results in drastically reduced infection levels with dengue, chikungunya and Plasmodium (Bian et al. 2013; Bian et al. 2010; Hughes et al. 2011; Kambris et al. 2010; Moreira et al. 2009). In contrast to naturally Wolbachia-infected mosquitoes, the resistance in transfected mosquitoes is accompanied by an up-regulation of the innate immune system (Kambris et al. 2010; Kambris et al. 2009; Moreira et al. 2009), e.g. an activation of certain immune genes of the Toll pathway via elevated oxidative stress (Bian et al. 2010; Pan et al. 2012; Rances et al. 2012). Similarly, transfection of the wMel strain into Ae. albopictus induced resistance to dengue, together with only a modest and transient up-regulation of the immune system (Blagrove et al. 2012).
These protective phenomena indicate that Wolbachia can have an impact on other microorganisms in the same host environment and several lines of evidence suggest that this would be due to a direct interaction, e.g. via some effector molecules or competition for resources and space, rather than to a Wolbachia-induced activation of the host's immune response. First of all, no immune stimulation has been observed in naturally infected Drosophila or mosquito hosts (Blagrove et al. 2012; Osborne et al. 2009; Rances et al. 2012). In addition, antiviral protection seems to depend on a certain Wolbachia titer, since the Wolbachia strains wHa and wNo, occurring at lower titers in Drosophila simulans than the three protective strains, do not interfere with virus infection (Osborne et al. 2009). Moreover, resistance provided by the protective strain wAu was lost when its titers were reduced to levels comparable to those of the two nonprotective strains (Osborne et al. 2012). Finally, wAlbB titers are much lower in its native host than after transfection into Ae. aegypti, suggesting that native Wolbachia titers in Ae. albopictus are too low to induce antiviral resistance (Lu et al. 2012).
Introduction
2.3.2 Wolbachia-Wolbachia interactions in multiply infected hosts
Relatively benign parasites like Wolbachia can be expected to regulate their titers based on a trade-off between titers high enough to ensure efficient vertical transmission, at the same time keeping them as low as possible to reduce infection costs for the host. The situation is more complex when several Wolbachia strains co-exist: On the one hand, the different strains might compete for the limited available resources and space but, on the other hand, multiple Wolbachia infections might also represent a fitness advantage when several CI strains co-exist in the same population. Several examples of Wolbachia-host associations involving multiple CI-inducing Wolbachia strains in the same host organism indicate that the different Wolbachia strains may or may not compete with each other. The first scenario has been observed in parasitoid wasps and moths, with Wolbachia titers being highly strain-specific and unaffected by the presence of other strains (Ikeda et al. 2003; Mouton et al. 2004; Mouton et al. 2003). However, co-existing Wolbachia strains have been found to compete for resources and space in the beetle Callosobruchus chinensis and Acromyrmex leafcutter ants (Andersen et al. 2012; Ijichi et al. 2002; Kondo et al. 2005). Competition between several Wolbachia strains can be expected to be most strongly pronounced in early developmental stages, e.g. when competing for niche space in the egg (Ijichi et al. 2002), while different strains may show different tissue tropisms in adults, resulting in some degree of within-host niche partitioning (Andersen et al. 2012; Ijichi et al. 2002).2.3.3 Wolbachia co-existing with other bacteria
Introduction
37
flies, whiteflies or mosquitoes (Cheng et al. 2000; Doudoumis et al. 2012; Gottlieb et al. 2008; Gueguen et al. 2010; Zouache et al. 2011; Zouache et al. 2009a; Zouache et al. 2009b). Being part of the microbiota in these and other hosts also implies that Wolbachia frequently co-exists with other well-known arthropod symbionts such as Buchnera, Cardinium,
Rickettsia, Sodalis, SOPE, Spiroplasma or Wigglesworthia (Doudoumis et al. 2013;
Gomez-Valero et al. 2004; Goto et al. 2006; Heddi et al. 1999; Jaenike et al. 2010a; Martinez-Rodriguez et al. 2013; Ros & Breeuwer 2009; White et al. 2009).
In this context, the co-existence of the two reproductive parasites Wolbachia and
Cardinium is of interest. While relatively little is known about their interactions, the outcome
seems to vary depending on the host-symbiont-symbiont association. For instance, Cardinium causes strong CI in the spider mite Bryobia sarothamni (Ros & Breeuwer 2009). Wolbachia, although not causing CI itself, interferes with Cardinium-induced CI in such a way that
Cardinium can neither modify nor rescue in doubly-infected hosts (Ros & Breeuwer 2009). In
contrast, Wolbachia causes CI in doubly-infected Encarsia inaron wasps, while Cardiunium does not (White et al. 2009). Finally, Goto et al. (Goto et al. 2006) observed asymmetrical interactions between the Wolbachia strain wMel and male-killing Spiroplasma (Mollicutes) co-existing in Drosophila melanogaster. More precisely, both symbionts exhibited distinct tissue tropisms and the more abundant Spiroplasma negatively affected Wolbachia titers compared to singly-infected individuals. In this host, Spiroplasma occurred both intra- and extracellularly, which might explain its negative impact on the obligate intracellular
III.
Armadillidium vulgare and its bacterial symbionts
3.1 Introduction
Among the Crustacea, oniscidean isopods (Malacostraca, Peracarida) are the most successful colonizers of terrestrial habitats, comprising around 3700 species (Schmalfuss 2003). Starting from the marine littoral zone, they have colonized all types of terrestrial environments on Earth, including deserts. This makes them excellent model organisms regarding evolutionary adaptations to a terrestrial lifestyle in terms of reproduction, respiration, excretion systems and protection against desiccation (reviewed in Hornung 2011). The main morphological and physiological changes compared to marine species include (i) a reduction in body size, (ii) a water-resistant cuticle, (iii) pleopodal lungs, (iv) a water conducting system and (v) a closed marsupium (reviewed in Hornung 2011). In addition, they had to cope with the nutritional conditions in terrestrial environments. Extant terrestrial isopods are saprophagous, i.e. they feed on dead and decaying organic matter, and play a key role as plant litter decomposers in terrestrial ecosystems. Considering that plant material mainly consists of lignocellulose and that the vast majority of animals do not possess the enzymatic machinery necessary for its degradation, the role of bacteria in terrestrial isopod nutrition has attracted considerable attention. Therefore, the following chapter will focus on potential nutritional symbionts before addressing the particular association of A. vulgare and feminizing Wolbachia bacteria.
Introduction
39
remnants of the midgut) and a tube-like hindgut. The latter can be roughly subdivided into an expandable anterior region containing a typhlosole (internal folds forming two dorsal channels) and a long tubular posterior region (Hassall & Jennings 1975; Palackal et al. 1984). Foregut and hindgut are of ectodermal origin and renewed with each moult, while the midgut caeca are endodermal.
Research investigating the role of bacteria in terrestrial isopod nutrition has roughly followed two main axes: (i) The identification of an autochthonous gut microbiota specifically adapted to this environment and (ii) identification of bacteria in the caeca that might be involved in cellulose degradation. Most of this work is based on the terrestrial isopod Porcellio scaber as model system. Early studies trying to identify an autochthonous gut microbiota described a bacterial community consisting of taxa commonly observed in animal gastrointestinal tracts (e.g. Enterococcus, Bacteroides) or typical soil bacteria such as Bacillus or Pseudomonas (Kostanjsek et al. 2001; Kostanjsek et al. 2004a; Kostanjsek et al. 2002). However, these studies were based on very few 16S rRNA gene sequences and cannot be considered representative of the entire gut microbiota.
Fig. 6 Digestive tract of Philoscia muscorum
(dorsal view). an: anus; cm: caeca; oe: oesophagus; pap: papillate region of the hindgut; prv: proventriculus; rm: rectum; sph: sphincter; typh: typhlosole (from Hassall & Jennings 1975).
Introduction
Nevertheless, these studies detected several bacterial taxa specifically associated with either the hindgut or the midgut caeca (Fig. 7). For instance, the recently identified Mollicute lineage Candidatus Bacilloplasma has been found to densely colonise the hindgut surface in Porcellio scaber via attachment to the cuticular spines protruding into the gut lumen (Fig. 7a) (Kostanjsek et al. 2001; Kostanjsek et al. 2007). This bacterium shares less than 83% 16S rRNA gene sequence similarity with known Mollicutes but its role in the terrestrial isopod hindgut remains unknown (Kostanjsek et al. 2007). In contrast, Candidatus Rhabdochlamydia porcellionis presumably represents an intracellular parasite (Kostanjsek et al. 2004b; Sixt et al. 2013). Sharing less than 92% 16S rRNA gene sequence similarity with their closest relatives of the order Chlamydiales, these bacteria were exclusively detected in the midgut caeca where they were found to multiply in vacuoles following a complex life-cycle (Fig. 7b) (Kostanjsek et al. 2004b). Finally, vacuoles containing bacteria are released into the lumen, thereby damaging the cell (Kostanjsek et al. 2004b; Sixt et al. 2013).
Apart from Ca. Rhabdochlamydia porcellionis, several studies have specifically searched for bacterial symbionts associated with the midgut caeca. The reason for this is that this tissue constitutes the site of highest cellulolytic activity and glucose absorption (Zimmer & Topp 1998). Moreover, it was hypothesized that the necessary enzymes were produced by endosymbiotic bacteria in the caeca rather than derived from bacteria ingested with the food (Zimmer et al. 2002a; Zimmer et al. 2001; Zimmer & Topp 1998). Subsequent screens of numerous terrestrial isopod species along an environmental gradient from marine to terrestrial and freshwater habitats revealed that bacteria were absent from the caeca in marine species but present in semi-terrestrial, terrestrial and freshwater species (Wang et al. 2007b; Zimmer & Bartholmé 2003; Zimmer et al. 2001; Zimmer et al. 2002b). This has led to the hypothesis
Introduction
41
Fig. 7 Symbiotic bacteria associated with digestive tissues in Porcellio scaber. (A) Candidatus
Bacilloplasma attached to the cuticular spines on the hindgut surface. Arrows indicate cuticular spines, arrowheads the bacteria-derived spherical attachment structures. (B) Vacuoles containing several developmental stages of Candidatus Rhabdochlamydia porcellionis. White arrowheads indicate spherical reticulate bodies, black arrowheads mature elementary bodies and black arrows intermediate bodies. (C) Candidatus Hepatoplasma crinochetorum in the caeca lumen. Candidatus Hepatoplasma crinochetorum (D) and Candidatus Hepatincola porcellionum (E) can be closely associated with the microvilli on the surface of the caeca cells. Scale bars: A, B = 2 µm; C, E = 1 µm; D = 0.2 µm (from Kostanjsek et al. 2007; Kostanjsek et al. 2004b; Wang et al. 2004a; Wang et al. 2004b).
preadaptation allowing the colonization of terrestrial habitats (Wang et al. 2007b; Zimmer et al. 2002a; Zimmer et al. 2001). However, different bacterial taxa were observed depending on host ecology. While the freshwater isopod Asellus aquaticus, which also feeds on terrestrial food sources such as leaf litter, harboured a relatively diverse bacterial community containing Alpha-, Beta- and Gammaproteobacteria (Wang et al. 2007b), terrestrial and intertidal species carried either the Mollicute Candidatus Hepatoplasma crinochetorum (hereafter referred to as "Hepatoplasma") (Wang et al. 2004a) or the Rickettsiales Candidatus Hepatincola
