N° d‘ordre : 4008 THESE

Présentée à L’UNIVERSITE BORDEAUX 1

Ecole Doctorale Sciences et Environnements Par Melle Ika PAUL-PONT

Pour obtenir le grade de DOCTEUR

Spécialité : Biogéochimie et écosystèmes

Sensibilité et adaptation de populations de bivalves marins soumis à des stress multiples : infestation parasitaire, contamination microbienne et pollution métallique

Soutenue le : 16 mars 2010

Après avis de :

M Michel FOURNIER, Directeur de recherche INRS Canada Rapporteur M Dario MORAGA, Maître de conférences, Université de Bretagne Occidentale Rapporteur

Devant la commission d‘examen formée de :

M Michel AUFFRET, Professeur, Université de Bretagne Occidentale Examinateur Mme Magalie BAUDRIMONT, Maître de Conférences, Université Bordeaux I Directrice de thèse M Jérôme CACHOT, Professeur, Université Bordeaux I Examinateur M Frédéric GARABETIAN, Professeur, Université Bordeaux I Président M Michel FOURNIER, Directeur de recherche INRS Canada Rapporteur M Xavier de MONTAUDOUIN, Maître de Conférences, Université Bordeaux I Directeur de thèse M Dario MORAGA, Maître de conférences, Université de Bretagne Occidentale Rapporteur Mme Michèle ROMEO, Chargée de recherche INSERM Examinatrice

Remerciements

D’abord LEESA et LOB, puis GEMA et ECOBIOC, j’ai eu la chance de bénéficier des moyens et des savoirs de chacune de ces équipes durant ma thèse. Je tiens à remercier vivement les directeurs d’équipe Jean-Charles Massabuau et Guy Bachelet ainsi que les directeurs de laboratoire Pierre Chardy et Antoine Grémare pour m’avoir accueillie et permis la réalisation de ce projet multi disciplinaire. Je remercie également le LEMAR (UMR 6539) et l’INSERM (U 876) grâce auxquels j’ai pu réaliser l’ensemble des analyses en cytométrie en flux.

Cette thèse a été réalisée grâce au support financier de la Région Aquitaine (Bourse doctorale région Aquitaine 2006-2009) et de l’Agence Nationale de la Recherche (Projet Multistress). Je tiens à adresser mes remerciements à ces deux structures qui ont permis la réalisation de ce projet dans d’excellentes conditions. J’exprime de plus ma reconnaissance { la SEPANSO pour nous avoir permis l’accès { la réserve naturelle du Banc d’Arguin.

J’exprime toute ma reconnaissance aux membres du jury qui ont accepté d’évaluer ce travail : - Michel Fournier, Directeur de recherche, INRS Canada ; - Dario Moraga, Maître de conférences, Université de Bretagne Occidentale ; - Michèle Roméo, Chargée de recherche INSERM, Université de Nice ; - Michel Auffret, Professeur, Université de Bretagne Occidentale ; - Jérôme Cachot, Professeur, Université Bordeaux 1 ; - Frédéric Garabétian, Professeur, Université Bordeaux 1

J’adresse un immense merci à mes directeurs de thèse qui m’ont emmené sur ce sujet multistress dès le master 2, Magalie Baudrimont et Xavier de Montaudouin. Merci { vous d’avoir accepté de m’encadrer et de me diriger pendant ces 4 années. Magalie merci d’avoir cru en moi dès le tout début, de m’avoir fait partager ton expérience et de m’avoir guidée avec une confiance qui a ensuite dépassé de loin le cadre scientifique de cette thèse. En grande partie grâce aux voyages en ta compagnie, j’ai découvert une autre Mag que Mag-ma-directrice, avec sa famille, sa douceur et son univers, et ça a été un véritable régal de partager tous ces moments avec toi. Les galères se transformant quasi-instantanément en situations comiques, tu as placé la barre haute pour les prochains colloques et ça va être dur de faire aussi bien! Affaire { suivre… Xavier-le-MDTTE, je ne crains même pas faire de jaloux en employant ce qualificatif si bien trouvée par Danguinette, tellement les gens qui travaillent avec toi ne peuvent que s’accorder sur tes nombreuses qualités professionnelles et humaines. Qualités aussi rares à trouver qu’elles sont immenses, j’ai eu énormément de chance d’être encadrée par quelqu’un comme toi. Tu mets toi aussi la barre haute et instinctivement on essaye d’être { la hauteur car on n’a pas envie de te décevoir ! Bref... Les mots me manquent pour t’exprimer toute l’amitié, la reconnaissance et le respect que je te porte. MDTTE je persiste et signe donc, et encore tu as de la chance Dang n’est pas l{ pour faire stéréo ! Pas sur le papier mais pleinement en réalité, tu as constitué Patrice le troisième pilier de cette thèse et je te remercie énormément pour ton investissement dans l’encadrement de ces travaux. Investissement là aussi à bras le corps, tant sur le terrain, que lors des longues soirées devant un cyto, ou en interminables corrections de thèse et d’articles. Merci de m’avoir aussi bien guidée sur les pas de la biologie moléculaire toutes ces années, en apprenant à tes côtés mille trucs et astuces, j’ai toujours progressé dans l’enthousiasme, la bonne humeur et avec un humour parfois… surprenant !! Allez je taquine mais tu sais comme ça va me manquer ! Je suis fière d’avoir été dirigée par vous trois dans le commencement de ma vie professionnelle. Et peut être parce que je n’avais pas de montre (!) le temps { vos côtés est passé très vite … Je vous attends donc fermement dans le Pacifique pour qu’enfin on la goûte ensemble cette piñacolada les fesses posées dans le lagon !

La réalisation de cette thèse a été possible grâce au travail et { la coopération de l’ensemble de l’équipe Multistress { la fois brestoise et arcachonnaise. Et pour faire face à un sujet aussi multistressant (avec une thésarde déjà un poil multistressée) il fallait bien tout ce monde là, je vous assure! Mes multistressés de Brest, Phi-Phi, Nelly, Fabienne, Christophe, Hansy et Christine, un immense merci (and so more) pour les savoirs et les compétences que vous m’avez enseigné sur l’immunologie des bivalves et la cytométrie en flux, pour les nombreuses discussions, conseils et autres astuces prodigués. Tant de moments partagés, en manip, à pas d’heure devant un FACS, dans le bureau du grand chef a triturer des données de ROS pour les moins « bizarres » (Fabienne !), { l’apéro, non, aux apéros en fait, { la recherche désespérée d’une pizzeria ouverte un lundi soir morose d’hiver { Arcachon ou lors de nos rencontres sur des aires d’autoroute suspectes, oui oui je le maintiens, je le dis même haut et fort, avec de la CAR-BO-GLACE ! Bon allez malgré ce petit désaccord persistant, et aussi le fait que vous ameniez du temps absolument dégueulasse avec vous à chaque fois, tous ces moments partagés n’ont été que du bonheur, merci pour tout !

De nombreux voyages à Brest donc, une dizaine en deux ans, 700 et quelques kilomètres entre Brest et Arcachon, 8 heures de routes. Même pas peur. Il faut dire, voila l’équipée !! Constituée des deux grandes copines de microbio, Florence et Natalie, de la maître de conf’ la plus avenante du laboratoire, en l’occurrence Alexia sans hésitation aucune, ainsi que des directeurs précédemment cités ! Avec cette équipe, allouée d’un véhicule motorisé de type Espace ou assimilé contenant cookies, boîte de haribo (avec beaucoup de réglisse), waders et tamis plein le coffre qui-cognent-la-tête, trop facile les 8h de route Arcachon-Brest ! Bon malgré le fait que ça déboite beaucoup plus de faire la route en BMW Série 1 (faut le dire) et que Nostalgie soit vraiment une torture pour mes petites oreilles, les arrêts crêpes beaucoup-de- beurre-beaucoup-de-sucre, la magie du Kyriad, et de son réceptionniste surtout, et les fins de manip Ouististi ont rendu ces voyages particulièrement agréables ! Je vous remercie « la fine équipe » pour votre bonne humeur, votre enthousiasme, votre investissement qui ont permis un travail riche scientifiquement et humainement.

J’adresse mes plus grands remerciements à mes stagiaires, Frances, Natalia, Zeljka et Hanane, que j’ai eu la chance d’encadrer, merci pour votre investissement, votre rigueur et votre efficacité au travail grâce auxquelles autant de travaux ont pu être effectués dans le temps imparti de la thèse. Mais aussi pour m’avoir fait partager vos pays et vos univers qui ont rendu ces encadrements de stage particulièrement riches, multiculturels et festifs !

Mes remerciements s’adressent également { tous ceux qui de près ou de loin ont contribué { l’aboutissement de ces travaux de thèse et à rendre ces 3 années aussi agréables. Aux techniciens et aux marins tout d’abord, Henri, Pascal, Christian, Francis et Laurent, merci pour votre aide précieuse sur le terrain et en labo, pour votre présence et votre bonne humeur constante sur terre comme en mer. Du côté GEMA, merci à Véro et Régine pour votre aide dans la formation au dosage des métaux et dans l’apprivoisemement des AMA, M6SOLAAR et autres spectrophotomètres indomptables… Merci à Bruno pour ton aide dans mes manips et ta patience entre mes appels panique le week end, mes circuits d’eau inréglables, mes oublis-Doris et j’en passe… Sache que ça a été un plaisir de t’embêter toutes ces années ! Pour ces deux derniers, sachez que les cafés du matin Flo-Régine-Bruno me manquent déjà ! Maman Flo, je n’ai pas de mot assez fort pour te remercier d’avoir été près de moi avec tant de chaleur et d’attention. La maman du labo tu l’es même si ça t’énerve parfois. En tout cas tu l’as été pour moi, avec une tendresse et un soutien qui m’ont énormément touchée, aidée et fait grandir durant ces trois ans de thèse. Tu es toi, entière et unique, et tu vas la garder ta place dans mon cœur malgré doutes et distance. Nath et Alex merci pour votre bonne humeur, votre pêche incroyable, quand on vous côtoie ça ne peut qu’être contagieux, et que c’est bon !!! Reste des soirées à faire avant que je parte pour de bon les cocottes ! Un merci et clin d’œil tout particulier { Jean-Paul grâce à qui je suis arrivée à la station marine en master 1 et qui par la suite, même de loin et malgré les conflits, a su garder sur moi un œil d’une extrême bienveillance, dur mais exigeant, complice, conscient et intelligent. Tu es quelqu’un à part et de grande qualité. Du côté ECOBIOC un grand merci à Line pour tout ce que tu as fait en microbio notamment dans la partie Vibrio pour que tout le matériel soit opérationnel et que les manips puissent être réalisées en temps et heure. Merci { Marie et Monique d’avoir animé nos midis avec parfois beaucoup d’histoires coquines devant nos chastes oreilles de thésards. Un grand merci aussi { MC et Cathy pour votre bonne humeur et m’avoir toujours sortie des plans galères inhérents aux commandes, sorties, missions, inscriptions, etc. avec lesquelles ma patience flanche si souvent…quel charmant caractère ! Un merci général { tous ceux que je n’ai pas nommé, je m’étais juré de faire court ces remerciements et, { l’image de ma thèse, je n’y arrive pas !

Remerciements courts ou pas courts, un grand merci quand même, et pas des moindres, à mes compagnons de labo qui ont plus que largement contribué à rendre ces années si chouettes et si rapides… 4 ans déj{ ?!!! Je salue ici mes grands copains de bureau : petit Seb mon grand maître zen devant l’éternel avec qui j’ai partagé cette thèse du début à la presque toute fin ; Mohamedou qui nous a beaucoup manqué cette dernière année, étant parti (pour des raisons d’algorithmes mathématiques encore obscures…) vers de froides latitudes ; puis le reste des Seb’s Girls Adeline et Audrey à qui je passe le flambeau en toute confiance ! Avec tout autant de plaisir, je remercie évidemment mes copains étudiants / CDD / thésards / mcf / badmintoniens / Master 2 / djeun’s d’en face : Corine & Béachou les inséparables, Val’, Hugues Blanchet and the famous Dr Lucia ^^, Flora (the win !!!), Sabrina, M et Mme Pacaly, Lilouliloulou, Fanfan, les GuillaumeS, Emilie, Christelle, et ma copine Fofie bien évidemment avec laquelle beaucoup de choses ont été partagées ces tous derniers mois, finalement peu de temps ! Un clin d’œil { mon poto Lolo, the sexiest PhD of the DGO (after FMY of course) pour son amitié, son sens de l’humour juste magique, son univers { l’image de sa musique, de ses dessins et de ses équations de géochimie : psychédélique !!! Enfin, mes GRAAAANDES copines Dang, Lavesque et Daffe, précieuses étoiles dans l’univers de ma thèse et depuis dans ma vie, merci d’avoir supporter mon caractère so charming, merci d’être l{ et d’être aussi chouettes, merci pour tout ce qui a été partagé pendant ces années, mais aussi et surtout pour tout ce qui est { suivre, ici ou ailleurs… Dang prépares le terrain car je te rejoins ! M et Mme copine, la géographie capricieuse va nous éloigner et vous allez terriblement me manquer bande de moules. Mais loin d’être une fin, ce n’est que le début, alors ne changez rien les cocos vous êtes parfaits !!!

Enfin je tiens { remercier du fond du cœur mes miens qui sont restés dans l’ombre et qui ont pourtant tellement partagé, souvent subi, cette thèse, avec ses investissements, ses concessions, ses joies, ses peines, ses paniques, ses coups de gueule, ses angoisses, ses victoires... Etant, tout le monde le sait, d’un naturel zen, serein et apaisé, une mer d’huile pourrait-on dire, je leur tire mon chapeau d’avoir supporté (dans tous les sens du terme) la ika- en-thèse pendant toutes ces années. Je remercie par là mon premier sponsor officiel, présent et aimant depuis plus de 27 ans, mes parents, grâce { qui j’ai pu me réaliser et entreprendre tant de choses déjà. Merci aussi à mes grands-parents qu’ils soient encore de ce monde ou pas, pour leur soutien et toutes les valeurs inculquées. Merci à mes frangins évidemment, Taram, Léo, Thomas, la prunelle de mes yeux ces trois là, vous faites ce que je suis chacun à votre manière. et la force que vous me donnez a été (est) une alliée inestimable ! Enfin un plus qu’énorme merci à mes copains, « mes Miens », qui ont constitué THE bouffée d’oxygène qui permet de tenir toute une thèse : mes tahitiens préférés mon petit soleil Marie & son marquisien chéri Poiti, Tat Patou & the G-family, Laurette, Julie, mais aussi Carlito et Pelle qui ont rendu, entre autres, les années fac tellement débiles, drôles et inoubliables; The Best Neighbours of the World Emilie et Jim of course ; mes Bros du bassin qui ont vraiment embelli cette fin de thèse parfois douloureuse (Chapitre III…), ainsi que les Sales Gosses Rennais qui ont constitué, par leur musique et leur univers, les grains de folie indispensables { l’équilibre ! Enfin, et au risque de me faire sévèrement réprimander, je ne peux pas finir sans remercier sobrement mais du plus fort que je peux celui qui a été près de moi tout au long de cette thèse, avec complicité, attention et un soutien sans faille, pendant la première, la deuxième mais aussi la troisième année… et ce même { l’autre bout du monde. Pour tout ça et tellement plus, tu vois j’avais raison, c’est bien toi le plus fort !!! Avec beaucoup d’amour, d’amitié, de respect et d’admiration, merci Jo !!!

Résumé

Dans les écosystèmes littoraux, l‘activité anthropique, mais aussi le contexte naturel induisent chez les organismes aquatiques des situations de « multistress ». Parmi les sources potentielles de perturbation, trois d‘entre elles ont été étudiées : contamination métallique, infection bactérienne et infestation parasitaire. Une approche intégrée de leurs interactions sur la réponse génétique, protéique, cellulaire et populationnelle chez la coque Cerastoderma edule et la palourde japonaise Ruditapes philippinarum a été entreprise sur le terrain et en laboratoire. In situ, un suivi de deux ans a été mené sur quatre populations de bivalves issues d‘environnements différents afin d‘estimer la santé des populations (paramètres de dynamique de population), de la mettre en rapport avec les niveaux de base des stress subis (métaux, parasites) et d‘évaluer les réponses adaptatives mises en place par les bivalves en termes de détoxication et de défenses immunitaires. Cette approche a permis de décrire un certain nombre de scenarii concernant les relations entre des populations de bivalves et leur environnement. En laboratoire, des expériences de co-infestation contrôlées (parasite trématode Himasthla elongata, bactérie Vibrio tapetis) en milieu indemne de métaux ou contaminé au cadmium ont été menées sur certaines de ces populations et montrent que la réponse des bivalves à ces agressions multiples dépend des espèces, des combinaisons subies, et ne se déduit pas intuitivement à partir des réponses « mono- stress », soulignant ainsi la notion d‘interactions. Malgré la complexité et la diversité de ces dernières, certains mécanismes semblent prédominer quelle que soit l‘espèce étudiée. Alors que l‘infestation par le parasite ne semble pas modulée par la présence du métal ou de la bactérie dans le milieu, la bioaccumulation métallique est fortement influencée par la présence d‘un ou plusieurs agents pathogènes. En plus de perturber l‘accumulation du polluant, la présence d‘organismes pathogènes interfère dans les mécanismes de détoxication cellulaire avec notamment une perturbation de la synthèse des métallothionéines (MT). Des travaux menés plus particulièrement sur la réponse des MT chez la coque exposée à un métal, à travers l‘expression de l‘isoforme Cemt1 et la synthèse de la protéine, confirment la complémentarité des observations (génétique et protéique) ainsi que la nécessité d‘observer les réponses des organismes à différentes échelles afin d‘avoir une vision globale des processus interactifs existants entre polluants et pathogènes. Des interactions fortes sont révélées au niveau génétique, protéique et immunitaire: « cadmium x trématode » chez la coque et « bactérie x trématode » chez la palourde. Enfin ces expériences, en parallèle du suivi in situ, mettent en évidence le rôle majeur de l‘histoire de vie dans la sensibilité des organismes aux interactions polluants- pathogènes.

Mots clés : Bivalves, métaux, pathogènes, interactions, métallothionéines, système immunitaire.

Abstract

In coastal ecosystems, man-mediated activity and natural context induce a ―multistress‖ situation in aquatic organisms. Amongst potential perturbation sources, three of them were studied: metal contamination, bacterial infection and parasite infestation. An integrated approach of their interactions on genetic, protein, cellular and population dynamics responses in the cockle (Cerastoderma edule) and the Manila clam (Ruditapes philippinarum) was undertaken through field and laboratory studies. A two-years monitoring was conducted in four bivalve populations from different environments to estimate the fitness of populations (parameters of population dynamics), in relation with the baseline levels of stressors (metals, parasites) and with the adaptive responses implemented by bivalves in terms of detoxification and immunity. This approach allowed describing different scenarii concerning the relationship between bivalve populations and their environment. In laboratory, co-infection experiments (trematode parasite Himasthla elongata, bacteria Vibrio tapetis) were conducted on these populations in controlled condition with or without metal contamination (cadmium). They showed that the response of bivalves to stress is species-dependent. Combinations were not intuitively deduced from the "mono - stress" responses underscoring the concept of interactions. Despite the complexity and diversity of these interactions, some mechanisms predominated regardless of the studied species. While parasite infestation was not modulated by the presence of metal or bacteria in the environment, metal bioaccumulation in turn was strongly influenced by the presence of one or several pathogens. Beyond disrupting the accumulation of pollutants, the presence of pathogens interfered with the cellular detoxification mechanisms including impairment of the metallothionein (MT) synthesis. A focus on the response of MT in cockles exposed to a metal through the expression of isoform Cemt1 and protein synthesis confirmed the complementary of these observations (gene and protein). They also highlighted the need to analyze responses at different scales to obtain an overview of existing interactive process between pollutants and pathogens. Strong interactions were found: ―Cd x parasite‖ in the cockle and" bacteria x parasite "in the clam, at the genetic, protein and immune levels. Finally, these experiments highlighted the important role of life history on the sensitivity of organisms to pollutant-pathogen interactions.

Keywords: Bivalves, metals, pathogens, interactions, metallothioneins, immune system.

Sommaire

Introduction générale 1

Chapitre I. Etat des connaissances 11

A. Les modèles biologiques 13 1. Les organismes hôtes 13 1.1. La coque commune Cerastoderma edule a. Généralités 13 b. Ecologie 13 c. Réglementation 15 1.2. La palourde Ruditapes philippinarum a. Généralités 15 b. Ecologie 15 c. Réglementation 17 2. Les organismes pathogènes 17 2.1. La bactérie Vibrio tapetis (Maladie de l‘anneau brun) a. Historique de la maladie 17 b. L‘agent pathogène 17 c. Impacts sur l‘hôte 19 2.2. Les parasites trématodes digènes a. Taxonomie 21 b. Cycle de vie 21 c. Impacts sur l‘hôte 23

B. Généralités sur les mécanismes de bioaccumulation des métaux chez les mollusques bivalves marins 26 1. Les métaux étudiés 26

2. Biodisponibilité et transferts des polluants métalliques chez les organismes aquatiques 2.1. Notion de biodisponibilité 28 2.2. Notion de spéciation 29 2.3. Transfert des métaux chez les organismes aquatiques 30 2.4. Les modèles de bioaccumulation aquatique 33 a. Le modèle de l‘ion libre 33 b. Le modèle du ligand biotique 33 c. Les autres modèles 34 2.5. Mécanismes de détoxication des métaux 35 a. Régulation des flux d‘entrée et de sortie des métaux 35 b. Séquestration des métaux sur des ligands cellulaires 37 c. Séquestration des métaux dans des structures compartimentées 37 2.6. Processus d‘élimination 39

C. Généralités sur les mécanismes de défense des mollusques bivalves marins 41 1. Les hémocytes, des cellules à activité multiple 41 1.1. Origine et types hémocytaires 41 1.2. Propriétés 43 a. Plasticité et mobilité 43 b. Chimiotactisme 43 c. Agrégation cellulaire et adhésion 44 1.3. Les réactions de défenses cellulaires 44 a. Apoptose ou mort cellulaire programmée 44 b. Inflammation et infiltration hémocytaire 45 c. Phagocytose 45 d. Encapsulation 48 1.4. Autres fonctions hémocytaires 48 2. Les facteurs humoraux responsables de l‘immunité 49 3. Les effets des métaux sur les paramètres hémocytaires 51

Chapitre II. Bases méthodologiques 53

A. Détermination et numération des organismes pathogènes 55 1. Analyse des communautés parasitaires de trématode 55 1.1. Objectif 55 1.2. Méthodologie 55 1.3. Variables mesurées 55 2. Numération de Vibrio sur milieu gélosé de culture sélective (T.C.B.S.) 55 2.1. Objectif 55 2.2. Méthodologie 56 2.3. Variables mesurées 57

B. Techniques d’analyse des métaux traces 59 1. Traitement des échantillons 59 1.1. Echantillons biologiques 59 1.2. Echantillons d‘eau 59 2. Principe général du dosage des métaux par spectrophotométrie d'absorption atomique 60 2.1. Dosage du zinc, du cuivre et du cadmium par spectrophotométrie d'absorption atomique en flamme (SAAF) 63 2.2. Dosage du cadmium par spectrophotométrie d'absorption atomique électrothermique (SAAE) 63 2.3. Dosage du mercure par spectrophotométrie d‘absorption atomique sans flamme 65 3. Validité des mesures 65 C. Techniques d’analyse des Métallothionéines 67 1. Rappels 67 2. Principe de la méthode de saturation par le mercure inorganique 69

D. Techniques d’analyse des paramètres hémocytaires 71 1. La cytométrie en flux 71 1.1. Historique 71 1.2. Principe 73 2. Protocoles de mesure des paramètres hémocytaires 75 2.1. Prélèvement d‘hémolymphe 75 2.2. Mesures 77 a. Taille / complexité 77 b. Viabilité 77 c. Concentration en hémocytes 77 d. Capacité d‘adhésion 79 e. Activité de phagocytose 80 f. Production d‘Espèces Réactives de l‘Oxygène (ERO) 80

E. Techniques de biologie moléculaire 81 1. Recherche et séquençage des régions codantes d'intérêt 81 1.1. Purifications des acides ribonucléiques (ARN) totaux 83 1.2. Rétro-transcription 84 1.3. PCR classique 84 1.4. Electrophorèse 85 1.5. Purification des fragments d'intérêt 86 1.6. Clonage et transformation bactérienne 87 2. PCR quantitative en temps réel 89 3. Cas particulier du gène de métallothionéine de coque – Cemt1 93 3.1. Amplification des extrémités non codantes (5‘ et 3‘ RACE) 93 3.2. Extraction de l‘ADN génomique 95

Chapitre III. Approche multistress sur le terrain 97

Introduction 99 Bivalve population health: multistress to identify hot spots 101 Conclusion 135

Chapitre IV. Approche multistress en laboratoire 137

Introduction 139

A. Influence du cycle de vie sur la réponse aux stress multiples 141 Seasonal modulated MT synthesis in the cockle (Cerastoderma edule) after parasite and cadmium contamination 141 Conclusion 159

B. Rôle de l’histoire de vie sur la réponse aux stress multiples 161 How life-history contributes to stress response in the Manila clam Ruditapes philippinarum 161 Conclusion 181

C. Influence de l’espèce sur la réponse aux stress multiples 183 1. Interactive effects of metal contamination and pathogenic organisms on the marine bivalve Cerastoderma edule 183

2. Interactive effects of metal contamination and pathogenic organisms on the marine bivalve Ruditapes philippinarum 209 Conclusion 231

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque Cerastoderma edule 239

Introduction 241

A. Caractérisation moléculaire d’une isoforme du gène de métallothionéine – Cemt1 Cloning, characterization and gene expression of a metallothionein isoform in the edible cockle Cerastoderma edule after cadmium or mercury exposure 243

B. Etude cinétique de son expression au regard de la réponse protéique suivant une exposition au Cd et au Hg 265 Short-term metallothionein inductions in the edible cockle Cerastoderma edule after cadmium or mercury exposure: discrepancy between mRNA and protein responses 265

Conclusion 285

Chapitre VI. Synthèse générale et perspectives de recherche 287

Synthèse générale et conclusions 289 Perspectives 299

Références bibliographiques 305

INTRODUCTION GENERALE

Photo aérienne du Banc d‘Arguin situé à l‘entrée du bassin d‘Arcachon.

Introduction générale

Les écosystèmes côtiers, en tant qu‘interfaces avec le continent, l‘atmosphère et l‘océan constituent des milieux riches mais à l‘équilibre fragile. Ils représentent 7 % (26.106 km2) de la surface totale des océans et prennent en compte principalement les estuaires, les lagunes, les récifs coralliens et les marais salés. En dépit de cette proportion relativement modeste, elles jouent un rôle considérable dans la productivité biologique marine totale et dans le maintien des cycles biogéochimiques, en raison de nombreux échanges de matières entre les zones continentales et l‘océan ouvert (sédiments, nutriments, éléments minéraux…). Ainsi, ces écosystèmes prennent en charge environ 15 % de la production primaire océanique et leur valeur économique a été estimée à 40 % de la valeur de service des écosystèmes mondiaux (Costanza et al., 1997 ; Glé, 2007). Ces milieux doivent alors relever un certain nombre de défis environnementaux relatifs à l'activité humaine. En effet, l‘UNESCO estime à environ 3,4 milliards le nombre de personnes vivant sur une côte ou à moins de 200 km d‘un littoral, soit la moitié de la population actuelle du globe (40 % de la population mondiale vit dans la zone des 100 km). En 2025, ce nombre devrait atteindre 6,3 milliards, soit 75 % de la population mondiale. Cette pression anthropique croissante entraîne un profond bouleversement qui porte atteinte à la structure, l‘intégrité et le fonctionnement de ces écosystèmes. C‘est pourquoi, depuis quelques décennies, se développe une prise en compte réelle de l‘impact des activités humaines sur la « résilience » des écosystèmes, c'est-à-dire leur capacité à supporter ces pressions anthropiques et à maintenir la richesse, la diversité et le fonctionnement des communautés biologiques à des niveaux comparables à d‘autres habitats naturels similaires non impactés. Les plus importantes perturbations anthropiques affectant les zones côtières sont la surexploitation des ressources naturelles (pêche et aquaculture intensives - Héral, 1993), la dégradation de l‘habitat et la fragmentation écologique (urbanisation, aménagement - Komatsu, 1997), ainsi que le rejet de nombreux contaminants (urbanisation, activités agricoles et industrielles - Tett et al., 2003). Ce dernier point requiert une attention particulière puisque l‘ensemble des contaminants rejetés dans l‘environnement, quel que soit le compartiment (air, eau, sol), finit par se retrouver dans les milieux aquatiques, en particulier estuariens ou côtiers, par le biais des eaux de ruissellements et de percolations. De ce fait, ces zones comptent parmi les plus exposées aux différents types de pollutions récurrentes, comme les matières organiques en excès qui induisent des phénomènes d‘eutrophisation, les polychlorobiphényles (PCB), les radionucléides, les pesticides, les hydrocarbures aromatiques polycycliques (HAPs), les éléments traces métalliques (ETM), mais également les organismes pathogènes (bactéries, virus, microorganismes…), qui peuvent avoir des conséquences écologiques majeures (Galloway &

1

Introduction générale

Depledge, 2001). Les organismes aquatiques présents dans ces milieux ont donc à faire face à de très nombreuses sources de perturbations qui agissent le plus souvent de manière simultanée engendrant ainsi des interactions plus ou moins néfastes. Parmi ces sources de perturbation, les interactions entre la contamination par les métaux ou ETM et la présence d‘organismes pathogènes chez des mollusques bivalves, ont été étudiées dans le cadre de ces travaux de thèse. La notion d‘ETM, anciennement appelés métaux lourd, demeure encore mal définie. Actuellement, ils sont considérés comme les métaux dont la concentration naturelle moyenne dans la croûte continentale supérieure est inférieure à 1000 mg.kg-1. Ces réserves naturelles ne constituent pas à proprement parler de danger en elles-même. Cependant, outre les processus naturels (érosion, volcanisme…) responsables du transfert de ces éléments dans l‘environnement, l‘activité humaine, même si elle ne crée pas d‘ETM, participe plus que largement à leur diffusion et leur concentration, via notamment les activités minières et industrielles. Parmi ces éléments traces, certains sont indispensables au fonctionnement des processus biologiques, les oligo- éléments. Toutefois, ces métaux essentiels peuvent s‘avérer toxiques pour diverses formes de vie, à des teneurs plus élevées (ex : cuivre et zinc). Il en est de même pour d‘autres métaux dont le caractère indispensable n‘est pas démontré (ex: cadmium, mercure et plomb). Les plus connus pour leur dangerosité sont le plomb (Pb), le mercure (Hg) et le cadmium (Cd), suivis par le chrome (Cr), le cuivre (Cu), le nickel (Ni), le zinc (Zn). Dans le cadre de ces travaux de thèse, nous nous sommes intéressés aux contaminations par le cuivre et le zinc en tant qu‘éléments métalliques essentiels, et aux expositions par le mercure et surtout par le cadmium, considérés comme particulièrement toxiques. Outre leur caractère fortement mutagène et cancérigène, les métaux peuvent entraîner à l‘échelle de l‘organisme d‘importantes perturbations du système nerveux, cardiovasculaire, immunitaire, gastro-intestinal, hépatique, rénal, respiratoire et reproducteur. Les atteintes structurales et fonctionnelles engendrées par les métaux sur la composante biologique des écosystèmes aquatiques se répercutent sur les différents niveaux d‘intégration, à l‘échelle des individus, des populations, des communautés et des écosystèmes. Dès lors, la prédiction de l‘impact environnemental de ces substances toxiques à des niveaux sub-létaux est un des challenges majeurs pour la recherche en toxicologie environnementale. La notion d‘ « écotoxicologie », née de cette dernière discipline, apparaît à la fin des années 1960 sous la plume du toxicologue René Truhaut. Alors que la toxicologie classique limite ses études aux organismes, l'écotoxicologie tente de mesurer l'impact des substances chimiques, physiques ou biologiques, non seulement sur les individus mais aussi sur les populations, les écosystèmes et

2

Introduction générale sur les équilibres dynamiques qui les caractérisent. Les effets des polluants à des niveaux d‘intégration élevés (population et écosystème) étant toujours précédés de changements précoces à l‘échelle individuelle, il est apparu indispensable de développer des biomarqueurs d‘alerte précoce d‘effets toxiques (Van der Oost et al., 2003). Ils vont en effet permettre de mesurer le degré des dommages engendrés par des polluants ou les conditions de l‘environnement sur les organismes, à l‘échelle moléculaire, biochimique, cellulaire ou physiologique (Adams, 2001). Parmi ces biomarqueurs, les métallothionéines (MT), protéines de faible poids moléculaire et ubiquistes dans le règne , ont été particulièrement étudiées du fait de leur réponse précoce aux contaminations métalliques. Leur biosynthèse en réponse aux métaux a été décrite pour la première fois en 1957 par Margoshe et Vallee, et elles constituent l'un des mécanismes les plus importants et les plus étudiés à l'heure actuelle concernant la détoxication des métaux à l‘échelle cellulaire et tissulaire. Ces protéines possèdent d‘importantes potentialités de fixation, basées sur les propriétés thioloprives des métaux (forte affinité pour les groupements thiols – SH). Ceci leur confère un rôle premier de régulation homéostasique des métaux essentiels (zinc et cuivre), ainsi qu‘un rôle secondaire de séquestration des métaux toxiques (cadmium, argent, mercure, …). Cette capacité de séquestration limite ainsi l'interaction des ions métalliques avec d'autres biomolécules cellulaires, et les atteintes toxiques associées (Roesijadi, 1992). En raison de l‘ensemble de ces caractéristiques, ces protéines ont longtemps été utilisées en tant que biomarqueurs d‘exposition métallique. Cependant, depuis quelques années de nombreuses études ont montré qu‘une variété d‘autres facteurs (stress oxydant, hormones, agents cytotoxiques, conditions de stress physique…) étaient également impliqués dans la régulation de la transcription des gènes codant pour les MT (Vallee, 1995 ; Nordberg, 1998). L‘induction de la biosynthèse des MT par de nombreux composés intervenant dans les processus inflammatoires est également un mécanisme reconnu (Anderson et al., 1999). Daniels et Andrews (2003) ont récemment mis en évidence la présence d‘éléments de régulation sensibles aux interférons dans la région 5‘ non codante du gène de MT de mammifère. Dans le cadre de ces travaux de thèse portant sur les effets interactifs des polluants métalliques et des infections par des organismes pathogènes, il est apparu pertinent d‘analyser le comportement de cette protéine chez des organismes soumis à des situations de stress multiples. Il est à noter qu‘une description et une analyse plus poussées de cette protéine chez un de nos modèles biologiques, le mollusque bivalve Cerastoderma edule, ont été développées à l‘occasion de cette étude. Les bivalves représentent des modèles particulièrement intéressants à différents titres, écologique, souvent dominants en termes de biomasse, ils ont un rôle majeur dans le

3

Introduction générale fonctionnement des écosystèmes ; et économique, certaines espèces étant commercialisées. De plus, leur abondance, leur large répartition géographique, associées à leur caractère robuste et résistant par rapport aux perturbations du milieu, constituent d‘importants atouts dans le cadre d‘études environnementales (Rittschof & McClellan-Green, 2005). Leur activité de filtration à des fins respiratoires et nutritionnelles leur permet d‘accumuler les polluants, et notamment les métaux, de manière très importante. Ils ne possèdent en outre qu‘une capacité limitée à métaboliser les contaminants organiques accumulés (Phillips, 1977). Enfin, les formes adultes étant généralement sessiles ou sédentaires permettent d‘avoir un bon reflet de l‘état qualitatif du milieu environnant. L‘ensemble de ces caractéristiques en font de bons indicateurs biologiques qualitatifs et quantitatifs de l‘état de santé et de la contamination du milieu (Cossa, 1985 ; Fournier et al., 2002). Ainsi, des programmes de surveillance de la qualité des écosystèmes côtiers, tels que le Réseau national d‘Observation de la Contamination CHimique du milieu marin (ROCCH, ancien RNO) en France ou le « Mussel Watch Program » aux Etats-Unis (Beliaeff et al., 2002) utilisent des mollusques bivalves (huîtres et moules) comme indicateurs biologiques. Ce dernier, débuté en 1986 sur plus de 250 estuaires, lacs et lagunes des Etats-Unis, analyse les variations spatio- temporelles des contaminants de l‘environnement (HAP, PCB, pesticides, ETM). Il présente, de plus, la particularité de coupler à cette analyse un suivi des contaminants biologiques à travers l‘observation de nombreux indices : prévalence et intensité des infections parasitaires, maladies et atteintes tissulaires. Les populations de bivalves sont en effet affectées par un large spectre d‘agents pathogènes présents dans le milieu pouvant causer de sévères impacts amenant jusqu‘à des mortalités massives de ces populations. Les bactéries pathogènes et les macroparasites, protozoaires et trématodes, font l‘objet d‘une attention particulière dans notre problématique puisqu‘ils constituent une des principales menaces sur les populations de mollusques souvent citée comme un facteur limitant la production de bivalves exploités ou non (de Montaudouin et al., 2009 ; Gam et al., 2009 ; Lauckner, 1983 ; OIE, 2006 ; Potasman et al., 2002). Parmi les bactéries pathogènes, les bactéries du genre Vibrio sont la cause de nombreuses affections chez les organismes aquatiques et peuvent agir aussi bien en tant qu‘agents infectieux primaires, comme dans le cas de la maladie de l‘anneau brun chez la palourde, qu‘en tant qu‘agents opportunistes dans le cadre d‘infections secondaires (Fisher et al., 1999). Le déclenchement de nombreuses vibrioses a par ailleurs été mis en relation avec des perturbations de l‘environnement et des infections antérieures mettant en jeu d‘autres pathogènes. D‘autre part, plusieurs études

4

Introduction générale soulignent l‘importance du parasitisme dans les écosystèmes lagunaires (Bartoli, 1974 ; de Montaudouin et al., 2000). Dans ces milieux, les trématodes digènes sont extrêmement bien représentés et sont souvent responsables de mortalités importantes (Jensen et Mouritsen, 1992 ; Blanchet et al., 2003 ; Desclaux, 2003). Ils entraînent de nombreux effets néfastes sur les individus hôtes, tels que la castration, la réduction du taux de croissance ou encore les modifications du comportement (Deltreil et His, 1970 ; Huxham et al., 1993 ; Jensen et al. 1999). Ces impacts délétères peuvent ainsi engendrer une diminution de la résistance des hôtes infestés face à des stress environnementaux nombreux et variés : conditions environnementales défavorables (hypoxie, température, dessication), pollutions multiples et infections bactériennes (Desclaux, 2003). Pour faire face à l‘agression de ces pathogènes, les mollusques bivalves possèdent un système immunitaire « inné » qui contribue au maintien de l‘intégrité de l‘organisme hôte en éliminant les composants étrangers (particules du non soi, bactéries, virus, parasites…). Ce système repose sur des cellules, appelées hémocytes, capables de proliférer et de se différencier sous le contrôle de facteurs humoraux, et pouvant montrer différentes activités, notamment la phagocytose ainsi que la production et la libération de molécules actives amenant à la destruction de corps étranger (Canesi et al., 2002 ; Chu, 2000) . Il est intimement relié aux systèmes nerveux et endocrinien, avec lesquels il communique par l‘intermédiaire de médiateurs solubles (Krzystyniak et al., 1995). Pour ces raisons, il est extrêmement vulnérable face aux polluants. Dans un contexte de pollution croissante, la prise en compte de l‘impact des contaminants sur le système immunitaire ainsi que sur le développement de maladie constitue une voie de recherche privilégiée (Oubella et Auffret, 1995 ; Lafferty, 1997 ; McKenzie et al., 1995). Selon Sniezko (1974), le déclenchement d‘une maladie chez un organisme résulte de l‘interaction entre l‘hôte, l‘organisme pathogène et l‘environnement (Figure 2). Dès lors, il apparaît indispensable d‘étudier et de replacer l‘apparition de maladies et d‘infections parasitaires dans un contexte environnemental, c'est-à-dire en interaction avec d‘autres sources de stress abiotiques, telles que les pollutions d‘origine anthropique. Les programmes de recherche s‘intéressant aux effets combinés des polluants et des agents pathogènes sur la santé des organismes aquatiques sont en plein essor (Sures, 2004, 2006 ; Morley, 2006, 2009). De nombreux travaux ont révélé une nette augmentation de la sensibilité aux infections virales, bactériennes et parasitaires en relation avec la présence de contaminants dans les milieux estuariens et côtiers (Coles et al., 1994, 1995 ; Pipe and Coles, 1995 ; Pipe et al., 1999).

5

Introduction générale

Figure 1. Schéma des interactions environnement-hôte-pathogène (modifié d‘après Snieszko, 1974).

La multiplicité des sources de stress représente une menace de plus en plus reconnue pour les écosystèmes, et face à la complexité des relations « hôte-pathogène-environnement », il est primordial de développer des approches méthodologiques permettant d‘intégrer les interactions qui en découlent, de manière à améliorer la compréhension et la gestion de nos écosystèmes littoraux.

Dans le cadre de ce travail de thèse, nous nous sommes particulièrement intéressés aux interactions existantes entre polluants métalliques et organismes pathogènes chez deux mollusques bivalves marins, la coque Cerastoderma edule et la palourde japonaise Ruditapes philippinarum. Ces deux espèces ont été choisies au regard de : - Leurs particularités écologiques propres : alors que la coque présente un statut d‘espèce indigène sur l‘ensemble des côtes européennes, la palourde japonaise constitue une espèce introduite relativement récemment (1972). - Leurs sensibilités différentes par rapport aux agents infectieux les plus communément retrouvés dans les écosystèmes littoraux, et étudiés dans le cadre de cette thèse: les bactéries pathogènes, et notamment les bactéries du genre Vibrio, ainsi que les macroparasites (protozoaires et trématodes). - Leurs capacités de bioaccumulation des polluants métalliques divergentes.

6

Introduction générale

Il existe de plus une carence d‘informations concernant les bivalves fouisseurs comme la coque et la palourde sur les niveaux de polluants et de pathogènes, la plupart des données disponibles in situ concernant principalement les huîtres et les moules.

De plus, si l‘on s‘intéresse aux études menées sur les associations pathogènes / polluants, aucun paradigme ne peut être avancé en raison de l‘immense diversité des systèmes « hôte- parasite » qui présentent chacun des traits de vie et des caractéristiques physiologiques propres (Blanar et al., 2009). Dès lors, cette démarche de travail sur deux modèles biologiques différents a été choisie afin d‘appréhender de manière plus large le panel d‘interactions sur plusieurs couples « hôte-pathogène » et de voir si une généralisation de certains processus était envisageable au sein d‘un groupe taxonomique. L‘approche scientifique de cette thèse repose sur le couplage d‘études descriptives (in situ) et expérimentales (en laboratoire). Un échantillonnage in situ sur trois sites présentants des environnements différents a été réalisé sur 2 ans à partir de populations naturelles et transplantées afin de déterminer l‘état de base des contaminants/pathogènes présents dans les bivalves, tout en intégrant l‘état de santé des populations ainsi que les réponses des organismes en termes de processus de détoxication et de défenses immunitaires. L‘objectif était donc de comparer les caractéristiques de populations aux histoires de vie différentes et dans des environnements à priori bien contrastés afin de comprendre comment elles composent avec les stress présents. En parallèle, des études expérimentales en laboratoire ont été réalisées afin de décortiquer les mécanismes qui peuvent expliquer le succès ou l‘échec du maintien des populations. Il apparaît en effet indispensable de déterminer les rôles directs et indirects de chaque facteur afin d‘appréhender les effets résultant de leur présence simultanée. Il est à noter que tout au long de ce manuscrit nous restreindrons la signification du terme « stress » à la contamination métallique et aux infestations parasitaires et microbiennes.

La complémentarité de ces approches (in situ / expérimental) permet d‘avoir une vision globale de la problématique de thèse à la fois au niveau écologique et mécanistique (Figure 3). Lors de ces différentes démarches, les réponses adaptatives des organismes ont été analysées à l‘échelle de l‘individu à différents niveaux d‘intégration biologique : moléculaire (expression génétique, synthèse protéique), cellulaire (activité immunitaire) et physiologique (indice de condition et croissance).

7

Introduction générale

Figure 2. Représentation schématique des différents niveaux d‘organisation pouvant être affectés par les contaminations biotiques et abiotiques, ainsi que leur intérêt au niveau écologique et mécanistique (modifié d‘après Snape et al., 2004).

Ce mémoire se découpe en six chapitres :

Après une description des modèles biologiques hôtes et parasites, une synthèse bibliographique sur les mécanismes de bioaccumulation des métaux chez les mollusques bivalves marins ainsi que sur leurs mécanismes de défenses immunitaires sera développée en Chapitre I.

Le Chapitre II s‘attachera à détailler l‘ensemble des techniques et méthodes analytiques qui ont été utilisées pour les études menées en laboratoire et sur le terrain tout au long de ce travail.

Le Chapitre III sera consacré à l‘approche menée in situ, à partir de l‘échantillonnage réalisé sur quatre populations de coques et de palourdes présentes sur la façade atlantique ouest (un site breton (Landéda) et deux sites aquitains (Bassin d‘Arcachon interne et externe). Ce chapitre sera presenté sous la forme d‘un article en préparation. La mise en relation de ces travaux avec les études expérimentales sera proposée en introduction du chapitre suivant.

Le Chapitre IV abordera les expériences de co-infestations contrôlées, en milieu contaminé ou non par les métaux, réalisées en laboratoire. Ce chapitre s‘efforcera d‘améliorer l‘état des connaissances sur la compréhension des interactions engendrées par la multiplicité des stress. L‘importance du cyle de vie ainsi que de l‘histoire de vie des espèces hôtes dans la réponse aux stress sera également mise en lumière dans cette quatrième partie. L‘ensemble de ces travaux sera

8

Introduction générale présenté sous la forme de quatre articles dont trois acceptés (Marine Pollution New Research, Marine Pollution Bulletin et Environmental Science and Pollution Research) et un soumis (Environmental Pollution).

Le Chapitre V étendra le champ d‘investigation à une étude plus approfondie de la réponse des métallothionéines chez la coque Cerastoderma edule. En effet, un effort tout particulier a été mis sur la caractérisation moléculaire d‘une isoforme du gène de MT chez cette espèce ainsi que sur son expression différentielle dans le cadre de contamination métallique. Ces travaux seront eux aussi présentés sous la forme de deux articles, le premier étant soumis (BioMetals) et le second accepté dans Aquatic Toxicology.

Enfin, le Chapitre VI aura pour objet de discuter l‘ensemble de ces résultats dans le cadre d‘une synthèse générale et de proposer des perspectives de recherche inscrites dans la continuité de ces travaux de thèse.

9

10

CHAPITRE I. ETAT DES CONNAISSANCES

Cerastoderma edule

11

Stries de croissance

Umbo

Charnière

Dents latérales Dents cardinales

Ligne palléale

Cicatrice des muscles Cicatrice des muscles adducteurs Espace palléal adducteurs

Figure I.1. Détails de la coquille de Cerastoderma edule et Ruditapes philippinarum

Estomac Cœur Glande digestive Rein Muscle adducteur antérieur Muscle adducteur postérieur Anus

Siphon exhalant Bouche

Palpes Siphon inhalant

Intestin Branchies Pied Manteau Coquille

Figure I.2. Schéma de l‘anatomie générale d‘un mollusque bivalve

12

Chapitre I. Etat des connaissances

A. Les modèles biologiques

1. Les organismes hôtes

1.1. La coque Cerastoderma edule a. Généralités La coque commune, Cerastoderma edule (Linné, 1758) anciennement nommée Cardium edule est un mollusque bivalve de la famille des Cardiidae. C‘est une espèce exploitée sous plusieurs appellations : bucarde, sourdon, hémon, rigadeau, demoiselle ou maillot. Ce mollusque présente une coquille arrondie (Figure I.1), il est pourvu d'un manteau largement ouvert en avant et inférieurement pour le passage d'un pied long et coudé (Figure I.2) et de deux siphons courts, non extensibles et ciliés. C‘est l‘un des mollusques bivalves les plus abondants sur les estrans semi-abrités des lagunes et des estuaires (Bachelet et al. 1992). Elle se trouve essentiellement dans la zone intertidale, et plus rarement en position subtidale (Guevara et Niell, 1989) vivant enfouie dans des sédiments sablo-vaseux ou sableux, voire même dans des fins graviers (Hayward et Ryland, 1990). Sa distribution va de la mer de Barents (Norvège) au Maroc couvrant ainsi une grande partie de l‘Europe occidentale (Figure I.3). Sur une même zone géographique, la répartition des coques est agrégative, les densités peuvent aller jusqu‘à 1900 individus/m², comme en baie de Saint Brieuc (Ponsero et al. 2009) et mais sont de l‘ordre de 100 ind/m2 dans le bassin d‘Arcachon au niveau du banc d‘Arguin (Gam et al., 2010). b. Ecologie Alimentation C‘est un mollusque suspensivore actif qui filtre l‘eau grâce à l‘ouverture de deux siphons à la surface du sédiment. La coque filtre l‘eau à travers ses branchies et ses palpes puis retient le phytoplancton, zooplancton et les particules de matières organiques. Tous ces éléments sont agglomérés dans un mucus et dirigés vers la bouche grâce à de nombreux cils.

Reproduction et recrutement Les coques vivent en moyenne deux à quatre ans mais peuvent atteindre six ans (Cole, 1956). La maturité sexuelle est généralement atteinte dès la deuxième année (taille d‘environ 20 mm ; 15 à 18 mois). Cette espèce gonochorique, ovipare, a une fécondité élevée avec généralement une à deux pontes par an (Bachelet et al., 1992). 13

Chapitre I. Etat des connaissances

Figure I.3. Aire originale de répartition de la coque commune Cerastoderma edule (Sources: de Montaudouin et al., 2009 - http://www.marbef.org/).

La ponte est induit par un accroissement de la température plutôt qu‘à une valeur absolue de celle-ci et est printanière dans le bassin d‘Arcachon. L‘oeuf (50 à 60 μm) se développe pour former une larve planctonique nommée « trochophore » puis « véligère ». Cette sédentarisation se produit à une taille d‘environ 280 μm (Creek, 1960). Le recrutement est l‘installation des larves planctoniques dans le sédiment. Le sédiment doit présenter des caractéristiques granulométriques favorables à la fixation de l‘eau car la survie des jeunes dépend d‘une bonne rétention en eau. Pour cela elle va produire du byssus qui va permettre d‘augmenter sa surface portante et donc d‘être remise activement en suspension et transportée par les courants vers d‘autres lieux. Ce type de déplacement est nommé « migration bysso-pélagique » et constitue ainsi l‘occasion d‘un « recrutement secondaire » (« secondary settlement ») (de Montaudouin, 1997).

Croissance La coque peut atteindre une taille maximale de 40 mm. Un certain nombre de facteurs va affecter la croissance comme par exemple la saison, la localisation géographique (Ducrotoy et al., 1991), la température de l‘eau (Smaal, 1997), la disponibilité en nourriture, le temps d‘immersion ou la densité des adultes (Jensen, 1993; de Montaudouin & Bachelet, 1996).

14

Chapitre I. Etat des connaissances

c. Réglementation La pêche professionnelle implique de disposer d'un permis. Pour les amateurs, la loi impose en France aux particuliers de la pêcher à pied sans utilisation de râteaux ou d'autres gros outils autres que le « gratte-à-main », petit outil à main doté de trois doigts métalliques. En France la taille légale de capture est de 30 mm.

1.2. La palourde Ruditapes philippinarum a. Généralités La palourde japonaise, Ruditapes philippinarum (Adams & Reeves, 1850) est aussi un mollusque bivalve mais appartient à la famille des Veneridae. La palourde est une espèce exploitée dans de nombreux pays et se situe en deuxième position des espèces de bivalves marins les plus exploités au monde après l‘huître creuse Crassostrea gigas (3 138 514 tonnes, FAO 2007). Elle présente une coquille arrondie avec deux valves identiques mais inéquilatérales (Figure I.1), un pied large, comprimé, linguiforme et muni quelquefois d'un sillon byssifère. Les deux siphons sont extensibles et soudés sur deux tiers de leur longueur, contrairement à l‘espèce native, Ruditapes decussatus, qui présente deux siphons totalement séparés. La palourde japonaise est originaire de la province indopacifique : des mers des Philippines, de Chine et du Japon (Ponurovsky & Yakovlev, 1992). Depuis le début du 20ème siècle, cette espèce a été introduite dans différentes parties du monde accidentellement avec l‘importation du naissain de l‘huître japonaise Crassostrea gigas ou volontairement pour l‘aquaculture (Figure I.4) (Le Treut, 1986). Elle a été importée en France en 1972 à partir des populations nord américaines puis en Angleterre, Espagne, Irlande, Italie et Allemagne (Ponurovsky & Yakokvlev, 1992). En Europe, des populations sont observées des côtes britanniques et norvégiennes aux rias espagnoles, ainsi qu‘en Méditerranée et en Adriatique. L‘habitat de la palourde est constitué par des baies abritées (lagunes, estuaires) présentant des substrats très variables : vaseux, sablo-vaseux, et sableux (Le Treut, 1986). b. Ecologie Alimentation La palourde est suspensivore. Elle se nourrit principalement de phytoplancton et de matière organique particulaire (Dang, 2009).

15

Chapitre I. Etat des connaissances

Figure I.4. Aire de répartition de la palourde japonaise Ruditapes philippinarum (d‘après Flye- Sainte-Marie, 2007). En rouge, l‘aire originale de répartition. En vert, les zones où des populations introduites se sont développées.

Reproduction et recrutement Cette espèce gonochorique et ovipare a une fécondité élevée. Dans les populations naturelles, la palourde est sexuellement mature à partir de sa première année. Le cycle reproducteur est connu pour être variable et plusieurs pontes sont possibles dans une simple saison (Ponurovsky & Yakovlev 1992). Le processus de gamétogénèse est initié quand la température de l‘eau dépasse 8-10°C et s‘accélère en relation avec l‘accroissement de température. La ponte est initiée à partir de 12°C et est optimale lorsque la température de l‘eau est comprise entre 20 et 22°C. La phase pélagique larvaire dure deux à quatre semaines (Le Treut 1986). Une série de métamorphoses intervient et la larve va ensuite se fixer sur le substrat. Elle mesure alors 190 à 235 µm.

Croissance La longueur n‘excède pas 34,9 mm au Japon (Ohba, 1959) alors qu‘elle atteint 50 mm dans l‘état de Washington (Nosho & Chew, 1972). La température est un des facteurs écologiques le plus important dans la biologie de la palourde car elle détermine les périodes de croissance, les cycles sexuels et le niveau d‘alimentation (Le Treut, 1986).

16

Chapitre I. Etat des connaissances

c. Réglementation La pêche professionnelle implique de disposer d'un permis en règle. Parmi les sites français bénéficiant d‘une étude de stock, le bassin d‘Arcachon se situait jusqu‘en 2008 à la première place française en terme de stock total (4457 t) et de stock exploitable (1159 t). Pour les amateurs, la loi en France n‘interdit pas l‘utilisation d‘accessoires divers et variés (fourche à palourde, rateau, pelle, etc.) contrairement à la pêche à la coque. La taille légale de pêche était de 40 mm jusqu‘à début 2008, elle est maintenant passée à 35 mm (Dang, 2009).

2. Les organismes pathogènes

2.1. La bactérie Vibrio tapetis (maladie de l’anneau brun) a. Historique de la maladie A partir de l‘année 1987, des cas de mortalité massive ont été observés chez des palourdes (Ruditapes philippinarum) de divers sites vénéricoles des côtes françaises. Cette maladie a également été décrite en Norvège, en Irlande, en Grande Bretagne, en Italie, en Espagne et au Portugal (Figure I.5). Récemment des cas de maladies ont été rapportés en Corée du Sud, dans l‘aire de distribution original de la palourde japonaise (Park et al., 2006). Le développement de cette maladie semble étroitement contrôlé par les facteurs de l'environnement, et en particulier la température puisque la maladie se développe principalement en hiver et dans les zones les plus froides. b. L’agent pathogène L‘agent pathogène responsable de la maladie de l‘anneau brun (MAB) est une bactérie appartenant au genre Vibrio (Tableau I.1) et a été découvert pour la première fois au niveau du site de Landéda dans le Finistère Nord (Bretagne, France). Cette bactérie a été appelée Vibrio Predominant 1 (VP1) puis Vibrio tapetis (Paillard & Maes, 1990 ; Borrego et al., 1996). V. tapetis présente tous les caractères du genre Vibrio. Ce sont des coccobacilles (1 à 1,5 µm de longueur sur 0,5 µm de diamètre) mobiles grâce à un seul flagelle polaire, halophiles, donnant des colonies circulaires, translucides, non pigmentées. La culture est observée sur gélose trypticase soja contenant 2% NaCl (ou sur le milieu Difco marine agar) pour des températures d‘incubation comprises entre 4 et 22 °C avec un optimum thermique de 18 °C.

17

Chapitre I. Etat des connaissances

Figure I.5. Carte des différents sites ( ) où la Maladie de l‘Anneau Brun (MAB) a été signalée chez la palourde japonaise, Ruditapes philippinarum (d‘après Paillard, 2004; Lasalle et al., 2007 et Flye Sainte Marie, 2007).

Phyllum Proteobacteria (Stackebrandt et al., 1986) Classe Gammaproteobacteria Ordre Vibrionales Famille Vibrionaceae (Véron, 1965)

Tableau I.1. Position systématique

18

Chapitre I. Etat des connaissances

c. Impacts sur l’hôte

Cette maladie est principalement retrouvée chez la palourde japonaise Ruditapes philippinarum, aussi bien chez les palourdes d‘élevage que chez les palourdes sauvages et elle a également été décrite chez d‘autres mollusques bivalves (Ruditapes decussatus, Paphia aurea) et dans d‘autres taxons tels que des poissons (Hypoglossus hypoglossus, Symphodus melops). La maladie de l'anneau brun se caractérise par la formation d'un dépôt brun sur la face interne de la coquille. Lorsque ce dépôt est très développé, il forme un anneau sur la périphérie de la coquille d'où provient le nom de la maladie (Figure I.6). Ce dépôt est principalement constitué de conchyoline, matrice organique secrétée par le manteau lors des processus de calcification (Paillard, 2004). Chez les individus malades ce dépôt de conchyoline est mélanisé et sa composition biochimique diffère de la matrice organique d‘individus sains (Goulletquer et al., 1989). La colonisation par la bactérie débute au niveau de la lame périostracale. Sous certaines conditions, les bactéries peuvent provoquer des lésions voire une rupture de la lame périostracale (Paillard, 2004) ce qui permet la pénétration des bactéries dans les fluides extrapalléaux périphériques dans un premier temps, et éventuellement dans les fluides extrapalléaux centraux dans un deuxième temps. Lorsque ces compartiments sont atteints, la bactérie se trouve logée entre le manteau et la coquille et c‘est l‘interaction entre la bactérie et le manteau qui induit la formation du dépôt de conchyoline par l‘espèce hôte. A ce stade, la bactérie peut être considérée comme un microparasite. Si des lésions tissulaires se produisent chez les individus malades, la bactérie peut pénétrer dans l‘hémolymphe et les tissus de l‘hôte et entraîner sa mort par septicémie généralisée (Figure I.7a). Le principal symptôme associé à cette maladie est le dépôt progressif de conchyoline sur la coquille du bivalve (Figure I.7b). D‘autres types de symptômes peuvent aussi apparaître lors du développement de cette maladie tels que des déformations de la coquille et des dégénérations tissulaires, notamment du manteau et de la glande digestive (lésion, desquamation et hypertrophie) associées aux stades les plus avancés de la maladie (Paillard and Maes, 1994). Le développement de cette maladie est aussi associé à une diminution de la capacité d‘enfouissement, du poids, des réserves en glycogène, de la croissance, de la reproduction (Goulletquer, 1989), entraînant de manière plus générale un dérèglement de la balance énergétique de l‘hôte (Flye Sainte Marie, 2009). De nombreuses études se sont intéressées à la réponse du système immunitaire face à une infection par V. tapetis, ainsi qu‘aux relations existantes entre réponse immunitaire et établissement de la maladie chez l‘espèce hôte (Paillard et

19

Chapitre I. Etat des connaissances

al., 1994 ; Allam et al., 2000, 2002, 2006). Cette partie sera discutée de manière plus approfondie ultérieurement.

Figure I.6. Symptôme visible de la MAB : formation d‘un anneau brun correspondant au dépôt de conchyoline.

Temps

Coquille

4. Colonisation des tissus et Branchies mort de la palourde

Manteau

3. Colonisation du fluide extrapalléal central Fluide extrapalléal central 2. Colonisation du fluide Fluide extrapalléal extrapalléal périphérique périphérique

Lame périostracale 1. Colonisation de la lame périostracale Figure I.7a. Les stades de progression de la Maladie de l‘Anneau Brun : les différents stades de colonisation (14) (d‘après Paillard, 2004b).

Figure I.7b. Les stades de progression de la Maladie de l‘Anneau Brun : illustration de l‘échelle de classification des symptômes établie par Paillard et Maes (1994). CDS 2 : symptômes visibles à l‘œil nu. CDS 4 : le dépôt brun prend une place significative sur la face interne de la coquille.

20

Chapitre I. Etat des connaissances

2.2. Les parasites trématodes a. Taxonomie Parmi les cas d‘helminthose affectant les bivalves, les trématodes digènes représentent les parasites les plus importants en termes de prévalence et d‘abondance (Lauckner, 1983). Les trématodes digènes appartiennent au des plathelminthes. Ce phylum regroupe au moins 20 000 espèces (Caira & Littlewood, 2001). Justine (1998) divise ce phylum en deux groupes : les (présence d‘une axonème regroupant les organismes libres, en l‘occurrence les turbellariés) et les néodermata (renouvellement de l‘épiderme en fin de vie larvaire, il regroupe les plathelminthes parasites obligatoires ; les monogènes, les trématodes digènes et les cestodes) (Figure I.8). Les trématodes sont ecto- ou endoparasites. Le nom de cette classe fait référence à la cavité des ventouses caractérisant les trématodes (du grec trema-, « trou »). La classe des digènes est la plus diversifiée des groupes de plathelminthes, comprenant 17 superfamilles réparties sur 156 familles et comptant plus de 2500 genres.

Figure I.8. Place des digènes dans le groupe des plathelminthes. b. Cycle de vie Les cycles de vie des digènes marins sont encore à l‘heure actuelle moins bien connus que ceux du milieu terrestre ou d‘eau douce. Ils sont complexes et intègrent un à quatre hôtes (mais généralement trois) en alternant des stades libres et parasites. Leur développement fait intervenir

21

Chapitre I. Etat des connaissances

deux types de reproduction (sexuée et asexuée) et se fait en alternance dans des mollusques comme premiers hôtes intermédiaires et dans des vertébrés (poissons, oiseaux) comme hôtes définitifs. Une série de stades larvaires (miracidium, sporocystes, rédies, cercaires, métacercaires) assure la réplication du génome et sa transmission à travers les différents hôtes. Le cycle de vie général des parasites trématodes digènes est décrit figure I.9.

Figure I.9. Cycle classique des parasites trématodes digènes.

Le genre Himasthla est l‘un des genres les plus communs qui utilise la coque comme hôte intermédiaire (de Montaudouin et al., 2000). Lors des différentes expériences en laboratoire, le parasite Himasthla elongata a été particulièrement utilisé en raison de son ubiquité et de la relative facilité d‘obtenir des larves infestantes à partir du premier hôte intermédiaire, le bigorneau (Littorina littorea). Il s‘agit d‘un endoparasite hétéroxène et digène de la famille des Echinostomatidae (caractérisée par un collier d‘épines autour de la ventouse orale). Le cycle de ce parasite est décrit sur la figure I.13.

22

Chapitre I. Etat des connaissances

Figure I.10. Cycle du parasite Himasthla elongata impliquant un oiseau marin comme hôte définitif (généralement le goëland L. argenteus), ainsi que la littorine (Littorina littorea) et la coque (Cerastoderma edule) comme premier et second hôtes intermédiaires.

c. Impacts sur l’hôte La sévérité des atteintes dues aux infestations parasitaires dépend à la fois de l‘espèce de parasite et de la position de l‘hôte en tant que premier ou second hôte intermédiaire (Morley, 2006). Le stade parasite observé dans le premier hôte intermédiaire, le stade sporocyste ou le stade rédie, est généralement beaucoup plus délétère que le stade métacercaire, enkysté ou non, correspondant au second hôte intermédiaire (Lauckner, 1983). Toutefois, les métacercaires ne peuvent pas être considérées comme de simples kystes inertes, et leur impact est loin d‘être négligeable. En effet, de nombreux effets pathologiques, voire létaux, ont été observés lors d‘infestations importantes entraînant la présence d‘un grand nombre de métacercaires dans les tissus de l‘hôte (Lauckner, 1987 ; Desclaux, 2003 ; Gam et al., 2009).

23

Chapitre I. Etat des connaissances

Parmi les altérations dues aux trématodes digènes, la perturbation du système immunitaire est avérée. En effet, afin de faciliter la survie à long terme dans l‘hôte, ces organismes ont développé des stratégies de survie sophistiquées, y compris la capacité de moduler et de manipuler le système immunitaire de l‘hôte. Il a été démontré dans de nombreuses études une réduction des capacités de phagocytose et d‘adhésion des hémocytes de l‘espèce hôte (Humbert & Coustau, 2001 ; Iakovleva et al., 2006). Alors que ces effets ont principalement été rapportés chez des modèles trématodes-gastéropodes, peu d‘études concernent la réponse immunitaire des mollusques bivalves en tant que premier ou second hôte intermédiaire. Les trématodes peuvent interférer avec le système neuroendocrinien et perturber la croissance et la reproduction de l‘hôte (Morley, 2006). Les effets des trématodes digènes sur la reproduction sont toujours préjudiciables à l‘hôte (Curtis & Hurd, 1983 ; Sousa, 1983 ; Huxham et al., 1993 ; Mouritsen & Jensen, 1994 ; Tétreault et al., 2000 ; Krist, 2001). Un effet direct du parasite sur l‘hôte est la réduction ou la cessation de la production d‘œufs suite à une castration dite ‗mécanique‘. Selon Hurd (1990), la réduction de la fécondité jusqu‘à la castration totale, peut être directe ou indirecte en fonction de la proximité du parasite par rapport aux tissus gonadiques. L‘envahissement des tissus gonadiques par les sporocystes (chez le premier hôte intermédiaire) n‘est pas la première ou la seule cause de castration. Il peut subvenir d‘une part des mécanismes de compétition pour les nutriments entre les parasites et les gonades, et d‘autre part, la production d‘une substance endocrinienne antagoniste par les parasites qui va affecter directement le système hormonal de l‘hôte et indirectement le développement des gonades (Hurd, 1990). Une diminution du taux de croissance a été observée chez de nombreuses espèces en tant que premier ou second hôte intermédiaire (Sousa, 1983 ; Curtis et al., 2000 ; Wegeberg & Jensen 2003). Les auteurs expliquent cette réduction de croissance par une réallocation de l‘énergie dédiée à la réparation des dommages tissulaires. Chez les mollusques premiers hôtes intermédiaires, des phénomènes inverses d‘accélération de la croissance ont été observés. Un tel phénomène a été décrit chez des coques (Cerastoderma edule) infestées par des Bucephalidae (Lauckner, 1983). Ce phénomène est plus largement observé chez les gastéropodes (de Montaudouin et al., 2003 ; Curtis, 1995 ; Huxham et al., 1995 ; Mouristen et Jensen, 1994). Les auteurs expliquent le gigantisme des individus parasités par une réallocation de l‘énergie reproductive (suite à la castration parasitaire) en énergie allouée à la croissance. Des malformations de certains membres ont aussi été observées chez des vertébrés (salamander Ambystoma macrodactylum) infestés par des parasites trématodes (Ribeiroia 24

Chapitre I. Etat des connaissances

ondatrae) (Johnson et al., 2006). Les parasites peuvent enfin modifier le comportement de leur hôte, afin d‘augmenter les probabilités de prédation par l‘hôte aval et ainsi assurer la transmission parasitaire d‘un hôte à l‘autre. Combes (1991) qualifie cette stratégie de favorisation. Ce terme regroupe les processus qui, en réponse à des stimuli issus de l‘environnement ou de l‘hôte, vont amener les parasites à rencontrer leur hôte. Parmi les altérations les plus connues, certaines sont relativement bien documentées, comme l‘autonomie de certaines parties du corps et notamment des siphons (Maillard, 1976), une altération de la fermeture efficace des valves de la coquille (Bartoli, 1974 ; Bowers et al., 1996) ou encore une réduction des capacités d‘enfouissement (Jensen et al., 1999 ; Desclaux et al., 2004) et de nage (Helluy, 1982). In fine, les dommages causés par les parasites peuvent induire la mortalité de leur hôte soit directement (destruction des tissus), soit indirectement (prédation, compétition, affaiblissement par rapport à d‘autres stress environnementaux) (Wegeberg, 1998 ; Evans et al. 2001 ; de Montaudouin et al., 2003 ; Desclaux et al., 2004).

25

Chapitre I. Etat des connaissances

B. Bioaccumulation des métaux chez les mollusques marins

Le milieu marin est caractérisé à la fois par une remarquable stabilité de ses propriétés fondamentales et une grande variabilité de ses micro-constituants. L‘eau de mer contient en solution des combinaisons de tous les éléments chimiques mais seulement certains d‘entre eux, au nombre de douze, ont des concentrations égales ou supérieures au mg.L-1. Ces douze éléments majeurs interviennent pour 99,4 % en masse du total de la croûte terrestre (O, Si, Al, Fe, Ca, Na, K, Mg, Ti, H, P et Mn par ordre décroissant d‘abondance). Les éléments traces métalliques, au nombre de 68, ne représentent en masse que 0,6 % du total et sont à des concentrations inférieures à 10-6 M dans l‘eau de mer (Miquel, 2001). Toutefois, l‘apport de contaminants métalliques par l‘intermédiaire des effluents industriels et de l‘atmosphère, des fleuves et de leurs estuaires, peut localement modifier la composition de l‘eau de mer qui peut devenir toxique pour les plantes et les animaux.

1. Les métaux étudiés Dans ces travaux de thèse, quatre métaux ont été suivis dans le cadre du monitoring in situ (cuivre, zinc, cadmium) ou étudiés dans le cadre d‘expérimentations en laboratoire (cuivre, cadmium, mercure). Le choix de ces métaux s‘est fait selon plusieurs critères. Tout d‘abord, ces quatre éléments font parties des dix métaux considérés comme les plus préoccupants pour l‘environnement et les plus toxiques pour les écosystèmes aquatiques (Islam et Tanaka, 2004). Ces dix métaux sont, dans un ordre décroissant de toxicité : le mercure, le cadmium, l‘argent, le nickel, le sélénium, le plomb, le cuivre, le chrome, l‘arsenic et le zinc. Le cadmium et le mercure sont particulièrement toxiques du fait de leur grande affinité pour les ligands soufrés dont sont, entre autres, composés les protéines des membranes biologiques directement en contact avec le milieu environnant (épithélium tégumentaire, branchial et intestinal). En plus de leur caractère particulièrement toxique en relation avec leur forte affinité vis-à-vis de la composante biologique, ces métaux sont facilement transportés dans l‘environnement et peuvent revêtir de nombreuses formes chimiques plus ou moins nocives. Leurs propriétés physiques, avec notamment une bonne résistance thermique et une conductivité électrique élevée expliquent leur utilisation dans de très nombreuses activités humaines, favorisant ainsi leur dispersion à travers l‘environnement. Pour compléter cette étude, le cuivre et le zinc ont été aussi suivis lors du monitoring in situ (Chapitre III). Contrairement au cadmium et au mercure, ces deux derniers

26

Chapitre I. Etat des connaissances

métaux sont indispensables au métabolisme des êtres vivants, en participant au déroulement de nombreux processus biologiques. Ils peuvent aussi devenir particulièrement toxiques lorsque leur concentration dans l‘environnement dépasse un certain seuil (Roesijadi, 1992). Le cuivre est impliqué dans de nombreuses voies métaboliques, notamment pour la formation d‘hémoglobine, la maturation des polynucléaires neutrophiles, en tant que cofacteurs de nombreuses enzymes et métalloprotéines impliquées dans le métabolisme oxydatif, la respiration cellulaire, et la pigmentation. Les crustacés et certains mollusques possèdent de plus un pigment respiratoire à cœur de cuivre, l‘hémocyanine. A des concentrations élevées, l‘activité biocide du cuivre a donné lieu à de nombreuses applications telles que la fabrication de peintures antifouling pour les coques de bateau ou encore la synthèse de produit fongicide et de bactéricide dans le milieu agricole. Le zinc comme le cuivre, est un élément essentiel intervenant notamment au niveau de la croissance, du développement osseux et cérébral, de la reproduction et des fonctions immunitaires. Il est en effet impliqué dans de nombreuses voies de régulation cellulaires étant indispensable à la formation des protéines à doigt de zinc ou de certains facteurs de transcription régulant l‘activation des gènes. Ce métal ne constitue pas un des contaminants les plus dangereux, bien qu‘il entraîne à des concentrations élevées un certain nombre de dysfonctionnements notamment au niveau de la croissance et de la reproduction. Dans un contexte plus régional, le cadmium et le zinc sont responsables d‘une contamination importante du continuum Lot / Garonne / Estuaire de la Gironde mis en évidence dès 1970 par le Réseau National d‘Observation de la qualité des eaux côtières (RNO, IFREMER) avec des concentrations de Cd dans les huîtres sauvages 50 fois supérieures à celles des autres stations, et 10 fois supérieures à la norme de consommation humaine de l‘époque (10 µg.g-1, ps (poids sec)) (Boutier et al., 1989). La source de cette pollution provient de rejets miniers d‘une usine d‘extraction de minerai de zinc, située 400 Km en amont de l‘estuaire (Marie, 2005). Outre l‘impact écologique évident, cette pollution a aussi engendré un impact économique puisqu‘elle a entraîné le classement sanitaire des eaux estuariennes en zone D avec interdiction de reparcage, purification et ramassage des coquillages (huîtres, coques, palourdes…). Par ailleurs, il convient de mentionner que dans la zone estuarienne, le cuivre est également un des polluants métalliques majeurs. Les apports se font principalement à partir des bassins versants de la Dordogne et de la Garonne, fortement influencés par la culture de la vigne et les traitements phytosanitaires associés (Bourdais, 1998). Une étude menée sur des marais nord médocains à vocation aquacole alimentés par les eaux de l‘estuaire a mis en évidence des concentrations en cadmium, zinc et cuivre, dans la colonne d‘eau et les sédiments, supérieures aux valeurs moyennes retrouvées dans 27

Chapitre I. Etat des connaissances

d‘autres systèmes estuariens typiques (Baudrimont et al., 2005). Ces travaux ont de plus révélé la présence de mercure dans la colonne d‘eau et les sédiments, à des niveaux relativement faibles et comparables aux résultats de travaux plus anciens sur l‘estuaire de la Gironde (Cossa & Noël, 1987), mais pouvant potentiellement augmenter localement (Tseng et al., 2001). Malgré des concentrations relativement modérées en mercure, une très forte bioaccumulation de ce métal a été mise en évidence chez la coque C. edule, en comparaison aux autres espèces (C. gigas et R. philippinarum) et aux autres métaux (Cd, Zn, Cu) analysés dans cette étude (Baudrimont et al., 2005). Ainsi, dans la zone aval du continuum fluvioestuarien, plusieurs métaux, dont le cadmium, le zinc, le cuivre et le mercure, présentent une importante biodisponibilité à l‘égard des organismes aquatiques et en particulier des bivalves filtreurs.

2. Biodisponibilité et transfert des polluants métalliques chez les organismes aquatiques 2.1 Notion de biodisponibilité Au XVIème siècle l‘alchimiste Paracelse déclarait : « la dose fait le poison ». Cette affirmation n‘est pas entièrement vraie en écotoxicologie. En effet, un métal présent dans les milieux naturels constitue un risque pour les organismes uniquement dans la mesure où il est capable d‘accéder et de traverser les membranes biologiques. Ainsi, la détermination des concentrations métalliques dans les écosystèmes aquatiques doit obligatoirement être associée à l‘estimation de la biodisponibilité afin de pouvoir appréhender le risque réel. La biodisponibilité se traduit par l‘aptitude d'une fraction d'un élément à être absorbée par un organisme vivant. Cette fraction biodisponible varie considérablement suivant la nature de l‘organisme et de la barrière biologique rencontrée, les facteurs de contamination (nature et propriétés du métal), et les caractéristiques physicochimiques du milieu environnant. Chez les mollusques bivalves filtreurs, on distingue généralement deux voies d‘entrée des métaux : la voie directe, en relation avec l‘épithélium branchial et la voie trophique, en relation avec l‘épithélium intestinal. La fraction biodisponible d‘un élément vis-à-vis de ces différentes barrières biologiques est conditionnée par les interactions du métal avec l'ensemble des ligands dissous et particulaires présents dans le milieu. C‘est l'ensemble des formes physiques ou chimiques que peut revêtir un métal en interaction avec les ligands en présence que l‘on désigne sous le terme de spéciation (Templeton et al., 2000).

28

Chapitre I. Etat des connaissances

2.2. Notion de spéciation On définit deux grands types de spéciations : physique et chimique. La spéciation physique d‘un métal est sa répartition à travers les formes particulaires (> 220 nm), colloïdales (10-220 nm) et dissoutes "vraies" (< 10 nm). La spéciation chimique va quant à elle concerner la distribution d'un élément selon différentes catégories d'espèces chimiques dans un système (Campbell et Couillard, 2004). Dans ce dernier cas, pour un métal cationique, on peut considérer les formes suivantes : - L‘ion métallique libre (Cd2+, Hg2+, Cu2+, Zn2+) - Des complexes impliquant des anions inorganiques présents dans les milieux naturels - - - - 2- 2- 2- (HO , F , Cl , HCO3 , CO3 , SO4 , S2O3 …) - Des complexes impliquant des ligands organiques simples (acides aminés, hydroxamiques, polycarboxyliques…) ou complexes (acides humiques, fulviques, protéines…) L‘abondance relative de chaque espèce va être dépendante de plusieurs facteurs, et notamment de la composition ionique de l‘eau, qui va moduler les caractéristiques physico- chimiques telles que la salinité ou le pH, ainsi que la composition et la richesse en matière organique. La concentration en ligand ainsi que la constante d‘équilibre K entre le ligand et l‘élément métallique influencent aussi la proportion et la stabilité des espèces chimiques rencontrées (Templeton et al., 2000). Les milieux naturels présentent des compositions extrêmement variables qui moduleront la spéciation d‘un métal d‘un milieu à l‘autre. Cependant, pour un métal donné, les mêmes tendances générales se retrouveront dans les différents milieux en raison des propriétés intrinsèques de l‘élément métallique. Le tableau I.2 synthétise l‘état d‘oxydation ainsi que les espèces chimiques majoritaires sous lesquels sont retrouvés les métaux étudiés dans le milieu marin. Ainsi pour chaque métal et environnement aquatique considéré, c‘est la proportion relative des ces espèces chimiques coexistant dans le milieu, associées à leurs propriétés propres (charges électriques, encombrement stérique, solubilité…) qui détermine in fine la biodisponibilité de l‘élément métallique pour les membranes biologiques des organismes aquatiques.

29

Chapitre I. Etat des connaissances

Tableau I.2 Proportions relatives des espèces chimiques du cuivre, du zinc et du cadmium dans l'eau de mer (S = 30 psu ; T=25°C).

Cuivre Zinc Cadmium M2+ 3 39 1,9 MCl+ 1 15,8 29,1

MCl2/M(OH)Cl 0,6 10 37,2 - 2- MCl3 /MCl4 / 3,8 31

MSO4 0,4 15,8 / MOH+ 2,8 0,6 /

M(OH)2 1 0,1 0,8

MCO3 80 13 / 2- M(CO3)2 / 1,8 / M-ac. humique 11 / /

2.3. Transfert des métaux vers les organismes aquatique s La composition biochimique des membranes biologiques varie d‘une espèce à l‘autre et même d‘un tissu à l‘autre (Simkiss et Taylor, 1995). Elles présentent toutefois des caractéristiques communes comme leur caractère hydrophobe, l‘omniprésence de phospholipides et de protéines ainsi que l‘existence de transporteurs et/ou de canaux ioniques. Ces caractéristiques sont primordiales dans le processus de bioaccumulation métallique qui s‘effectue en trois étapes principales (Figure I.11) : Une phase advection/diffusion du métal (et ses complexes) dans la solution « baignant » l‘interface entre la membrane biologique et le milieu d‘exposition. Elle est fonction de la concentration du métal en solution (et ses complexes) et de la chimie des équilibres qui s‘établissent à l‘interface entre l‘organisme et son milieu. - Une phase de complexation avec le site de surface. Elle repose sur une forte électropositivité des métaux (M2+) qui vont pouvoir se fixer sur de nombreux constituants tels que les groupements sulfhydriles, les ponts disulfures des protéines, les têtes polaires des phospholipides, etc…

30

Chapitre I. Etat des connaissances

Figure I.11. Modèle conceptuel des interactions entre les métaux et les organismes aquatiques (Source Campbell et Couillard, 2004). Légende: MZ+: ion métallique libre; ML: complexe métallique en solution; M-X-Membrane: complexe métallique de surface; K1: constante d‘équilibre pour la formation de l‘espèce ML; Kf, Kf‘: constantes d‘équilibre de formation du complexe de surface; Kd, Kd‘: constantes d‘équilibre de dissociation du complexe de surface; Kint: constante de vitesse pour l‘assimilation du métal. Les charges sur les complexes ne sont pas indiquées.

Figure I.12. Mécanismes membranaires permettant l‘entrée des ions métalliques (modifié d‘après Simkiss et Taylor, 1989).

31

Chapitre I. Etat des connaissances

- Une étape de transport à travers la membrane biologique (Figure I.12). De nombreux mécanismes de transport des métaux à travers les membranes biologiques ont été mis en évidence : (i) La diffusion passive à travers la membrane des complexes métalliques lipophiles. (ii) Un transport facilité impliquant un transporteur transmembranaire protéique ou un canal transmembranaire. (iii) Un transport facilité d‘un complexe métallique (cations liés à des anions) via un transporteur d‘anion relativement peu sélectif. (iv) Un mécanisme d‘endocytose correspondant à l‘internalisation des complexes métalliques par un processus d‘invagination de la membrane, créant une vacuole. Pour les métaux présents en phase dissoute, c‘est généralement la voie (ii) qui domine, aussi bien pour les métaux essentiels pour lesquels il existe des systèmes de transports facilités (maintien de l‘homéostasie) que pour les métaux non essentiels qui réussissent à tromper les systèmes de transport cationique. Une exception notable à ce principe général concerne le mercure inorganique qui traverse aisément les membranes biologiques (Boudou et al., 1991) par simple diffusion passive (voie (i)) pour les formes neutres (les chlorocomplexes possédant les plus fortes aptitudes à franchir la barrière hydrophobe) ou grâce au transport des anions halogénés (voie (iii)) (Laporte et al., 2002). De plus, les transporteurs d‘acides aminés apparaissent comme les principaux mécanismes impliqués dans les flux trans-membranaires de la forme méthylée (Clarkson, 1994). Pour les métaux présents en phase particulaire, c‘est la voie (iv) qui prédomine. Les métaux particulaires vont en effet être réellement assimilés uniquement après ingestion de la particule et solubilisation du métal. Cette étape de solubilisation peut se produire directement dans le tractus digestif ou dans les vacuoles créées par invagination de la membrane biologique. Dans ces deux cas, les conditions physico-chimiques (pH, oxydo-réduction…) vont favoriser la solubilisation des métaux initialement liés à la particule. Une fois solubilisés, les métaux doivent franchir la membrane digestive ou vacuolaire selon les mêmes processus cités ci-dessus (Campbell et Couillard, 2004).

2.4. Les modèles de bioaccumulation métallique Au cours des trente dernières années, les chercheurs ont exploré les liens entre spéciation et biodisponibilité et ont essayé d‘établir des modèles prédictifs des effets biologiques utilisés à des fins aussi bien scientifiques que réglementaires.

32

Chapitre I. Etat des connaissances

a. Le modèle de l’ion libre Une des premières bases valables de prédiction de la biodisponibilité des métaux a été formulée par Morel (1983) sous le terme de Modèle de l‘Ion Libre (MIL ou FIAM en anglais pour Free Ion Activity Model). Ce modèle souligne l‘importance de l‘activité de l‘ion métallique libre (M2+) pour déterminer la prise en charge, la nutrition et la toxicité des métaux traces cationiques, et peut être illustré par les équations suivantes :

2+ - K1 Equation 1 M + X-C M-X-C

- 2+ [M-X-C] = K1 [ X-C][M ]

K2 - Equation 2 MLn + X-C M-X-C + n L

- 2+ [M-X-C] = K2 [ X-C][M ]/

Selon ce modèle, les sites -X-C sont constants et la réponse biologique varie en fonction du degré de saturation des sites -X-C, c'est-à-dire de la concentration [M-X-C]. Dans ce cas, la réponse biologique suivra la concentration de l‘ion métallique libre. Ceci n‘implique pas forcément que l‘ion libre soit « l‘espèce biodisponible » mais la connaissance de l‘activité et/ou de la concentration de l‘ion métallique permet de prédire la concentration de l‘espèce clé [M-X- C]. Ainsi, plusieurs études en laboratoire suggèrent que dans le milieu aquatique, l‘assimilation est proportionnelle à la concentration en ion libre (M2+) et non à la concentration totale du métal en solution. Ce modèle de l‘ion libre a été décrit en détail par Campbell (2004), le validant par de nombreuses expériences de terrain (Campbelle et Couillard, 2004). Il est cependant important de souligner que le modèle de l'ion libre ne s'applique pas au mercure : l'espèce chimique Hg2+ n'existe pas dans des conditions représentatives de celles rencontrées en milieu naturel, à la différence du cadmium par exemple pour lequel ce modèle a été confirmé de façon très probante en particulier chez les mollusques bivalves (Tessier et al., 1993). Sur les bases de ce MIL repose un modèle développé plus récemment : le Modèle du ligand biotique (ou Biotic Ligand Model, BLM). b. Le modèle du ligand biotique

33

Chapitre I. Etat des connaissances

Dans ce modèle, les bases sont identiques mais l‘espèce {-X-C} a été renommée {-Ligand biotique} et les accents sont moins portés sur l‘ion métallique libre lui-même que sur le ligand biotique où le métal et les autres ligands tels que la matière organique dissoute ou les autres cations (H+, Ca2+…) sont en compétition pour s‘y fixer. La principale différence par rapport au MIL est la prise en compte des phénomènes de compétition au niveau du ligand biotique ce qui permet d‘appréhender l‘effet protecteur des autres métaux cationiques présents dans le milieu ainsi que le rôle du pH (Di Toro et al., 2001). Enfin, la nature du ligand biotique (LB) auquel va se fixer l‘élément métallique va aussi déterminer les effets biologiques associés. Si le LB est un site physiologiquement actif, la fixation du métal va induire une réponse biologique directe alors que si le LB correspond plutôt à un site de transport, la fixation du métal entraînera son transport à travers la membrane, son entrée dans le cytosol et des réactions plus ou moins actives avec des ligands intracellulaires. Il est à noter que ce modèle, ainsi que le MIL, s‘applique lorsque la voie d‘absorption impliquant l‘intervention de transporteurs/canaux transmembranaires est prédominante. De plus ces modèles sont applicables pour les métaux cationiques (et leurs complexes) hydrophiles (Cd2+, Cu2+, Zn2+…) chez les organismes n‘assimilant pas la matière particulaire ou encore ceux pour lesquels la phase dissoute constitue le vecteur principal de prise en charge métallique (Campbell et Couillard, 2004). c. Les autres modèles Depuis la création du MIL et du MLB, largement utilisés dans les milieux scientifiques mais aussi réglementaires (Canada, Etats Unis, Europe), plusieurs études ont tenté de les valider sous différentes conditions avec de nombreux métaux cationiques et ont souligné les limites d‘application de ces modèles (Markich et al., 2000). Ainsi, des travaux de recherche ont tenté d‘élargir le champ d‘applicabilité du MIL et du MLB (conditions physico-chimiques, sédiments…) donnant ainsi naissance à de nouveaux modèles. Brown and Markich (2000) ont déterminé un nouveau modèle désigné sous le terme de « Biological Receptor Theory » permettant d‘étendre la validité du modèle LIM à un plus large panel de conditions physico- chimiques (concentration en MOD, pH et dureté variables). De la même façon, des travaux ont essayé de modéliser la réponse biologique des organismes vivants dans les sédiments soumis à une exposition métallique en intégrant dans leurs modèles les phénomènes de compétition entre le métal et les autres cations à l‘interface eau-sédiment (Cheng et Allen, 2001 ; Veltman et al., 2007). On peut aussi mentionner le modèle créé par Wang et Rainbow (2006) qui tend à prédire

34

Chapitre I. Etat des connaissances

la toxicité d‘un métal à travers sa répartition subcellulaire (« Subcellular Partioning Model »). Enfin, l‘apparition de modèles biodynamiques permettant d‘appréhender quantitativement tous les processus impliqués dans la bioaccumulation métallique et l‘élimination chez de nombreuses espèces (phytoplancton, invertébrés, poissons…) semble être une voie de recherche en plein essor (pour revue voir Wang et Rainbow, 2008).

2.5. Mécanismes de détoxication des métaux Une fois les barrières biologiques franchies, l‘ion métallique va se retrouver dans la cellule et va entraîner un large panel de dommages cellulaires à travers notamment sa liaison à des biomolécules essentielles, entraînant ainsi leur inactivation. Son changement d‘état d‘oxydation, génére la production d‘espèces réactives de l‘oxygène (ERO). Ces radicaux libres de l‘oxygène vont s‘attaquer à l‘intégrité des membranes et des éléments intracellulaires (réticulum endoplasmique, mitochondries, ADN…) (Campbell et Couillard, 2004). La détoxication comprend tous les processus, stratégies, ou mécanismes qui permettent de minimiser le potentiel d‘un métal à interagir avec des macromolécules essentielles (Campbell et Couillard, 2004). Ces différentes stratégies de détoxication sont mises en place au niveau cellulaire (Figure I.13) et concernent principalement :

a. Régulation des flux d’entrée et de sortie des métaux La régulation de la concentration intracellulaire de métal par limitation du flux d‘entrée et/ou augmentation du flux sortant est un mécanisme largement étudié. Parmi les mécanismes d‘efflux faisant intervenir des pompes transmembranaires, la famille des transporteurs à ATP binding cassette (ABC) est l‘une des familles de protéines les plus importantes en termes d‘ubiquité à travers le règne vivant, de taille et d‘implication dans la résistance aux maladies et aux xénobiotiques. Ces transporteurs sont caractérisés par une structure en quatre domaines : deux domaines hydrophobes, formant le site de reconnaissance du substrat, et deux domaines hydrophiles conservés, impliqués dans l‘hydrolyse de l‘ATP, source d‘énergie du transport pour l‘import ou l‘export d‘une variété considérable de substrats, des ions aux macromolécules (Mourez et al., 2000). La surexpression de certains de ces transporteurs ABC a fréquemment pour conséquence une résistance à des agents cytotoxiques n‘ayant pas de caractéristiques communes. Parmi eux, les antibiotiques, les antifongiques, les herbicides, et les médicaments anticancéreux. Ce phénomène est appelé « Multidrug resistance » (MDR). Il a été observé pour la première fois sur des lignées de cellules tumorales (pour revue voir Mourez et al., 2000). La possibilité qu‘une

35

Chapitre I. Etat des connaissances

protéine de type MDR puisse être impliquée dans l‘apparition d‘un phénotype MXR (multixenobiotic resistance) chez des organismes aquatiques vivant près de sites pollués a été pour la première fois proposée par Kurelec et Pivcevic (1991). Des travaux sur des modèles invertébrés, comme les moules, les éponges, et les vers, ont montré l‘implication de protéines de type MDR dans l‘efflux de polluants d‘origine anthropique (Minier et Moore, 1996) ou naturelle (Eufemia et al., 2000). L‘implication des transporteurs de type ABC dans la résistance aux xénobiotiques (polluants organiques, métaux…) a été mise en évidence chez de nombreuses espèces de mollusques bivalves (Eufemia & Epel, 2000; Kurelec et Pivcevic, 1991 ; Legeay et al., 2005). Enfin, depuis quelques temps de nombreuses études biodynamiques s‘intéressent à ces flux entrant et sortant afin d‘expliquer les différences inter-spécifiques de bioaccumulation chez les bivalves (Wang & Rainbow, 2008 ; Pan & Wang, 2009).

Figure I.13. Les différentes stratégies de séquestration mises en place au niveau cellulaire.

36

Chapitre I. Etat des connaissances

b. Séquestration des métaux sur des ligands cellulaires Trois classes de ligands intracellulaires (L) séquestrant les métaux ont été définies par Mason et Jenkins (1995) selon les effets cellulaires observés après fixation de l‘élément métallique:

- LE : effets bénéfiques une fois le métal lié. Cette classe concerne principalement les métalloenzymes nécessitant un métal essentiel dans leur cœur catalytique ;

- LT : effets négatifs une fois le métal lié. Cette classe concerne l‘ensemble des biomolécules (protéines, acides nucléiques, membranes…) dont l‘activité va être inhibée ou stoppée par la présence du métal (déformation de la molécule, blocage des corps catalytiques…) ;

- LI : aucun effet observé, ni négatif, ni positif. Cette classe concerne les ligands inertes représentant des formes de stockage du métal sous une forme non réactive dans le milieu intracellulaire. De tels ligands sont essentiellement caractérisés par la présence de groupements thiols leur permettant de complexer efficacement les métaux. Ils sont portés par les acides aminés libres (cystéine), les peptides (glutathion) et de nombreuses protéines dont les plus importantes et les plus étudiées à ce jour sont les métallothionéines. Ces protéines de faible poids moléculaire très riches en cystéine (30%), possèdent de fortes potentialités de fixation, basées sur les propriétés thioloprives des métaux (forte affinité pour les groupements thiols – SH). Ces protéines constituent ainsi des ligands intracellulaires majeurs dont nous développerons plus en détails la structure, les mécanismes d‘induction et les rôles biologiques dans le Chapitre 5. c. Séquestration des métaux dans des structures compartimentées La nature des structures compartimentées impliquées est à la fois tissu-spécifique et métal- spécifique. Parmi ces structures on distingue notamment :

- Les lysosomes La séquestration des métaux dans ces organites représente une voie majeure de stockage et de détoxication non spécifique vis-à-vis des métaux et dont le rôle est étudié chez les bivalves depuis fort longtemps. Deux mécanismes principaux permettent l‘accumulation des métaux dans les lysosomes : le métal dissous présent dans le cytosol peut être chélaté par une métallothionéine, et c‘est ce complexe MT-métal qui va être incorporé aux lysosomes. Selon la nature du métal, le complexe MT-métal va être plus ou moins stable dans le lysosome. Le complexe Cu-MT particulièrement stable et insoluble dans les lysosomes constitue une exception à l‘égard de la majorité des autres métaux liés aux MT qui seront facilement dissociés des 37

Chapitre I. Etat des connaissances

protéines par les conditions physico-chimiques acides régnant dans ces organites (pH=5), ce qui permet l‘hydrolyse ultérieure de la protéine et le relargage du métal par exocytose dans la système circulatoire (Viarengo et al., 1989). De plus, la présence dans les lysosomes de produits issus de la digestion de macromolécules cellulaires, telles que la lipofuscine, pigment lipoprotéique formant des granules insolubles, participent également à la séquestration intra- lysosomale. Les métaux se lient de manière passive à la lipofuscine par un mécanisme d‘adsorption. Ces mécanismes de séquestration lysosomale sont retrouvés dans les différents types cellulaires (épithelium branchial, digestif, tégumentaire mais aussi dans les hémocytes circulants). Cependant, selon les tissus, l‘accumulation dans ces organites va être plus ou moins transitoire. En effet, la distribution cellulaire et le maintien des métaux dans les lysosomes va varier en fonction des caractéristiques de rétention du lysosome en relation avec la formation et l‘accumulation de lipofuscine, ainsi qu‘en fonction du turn over des complexes MT-métal (Marigómez et al., 2002). Au final, les granules formés par la polymérisation des complexes métal-MT et les agrégats de lipofuscine du système vacuolaire lysosomal sont éliminés par excrétion des corps résiduels vers le système circulatoire (Viarengo et al., 1981). Il est important de mentionner que le nombre de lysosomes ainsi que la stabilité et l‘intégrité de leur membrane constituent des marqueurs d‘effets toxiques utilisés dans de nombreuses études (Sokolova et al., 2005)

- Les concrétions minérales insolubles Ces précipités minéraux de diamètre de 0,2 à 3 µm sont majoritairement composés de carbonates et de phosphates de calcium. Les granules de carbonates de Ca constituent surtout un stock remobilisable de calcium alors que les phosphates de Ca jouent un rôle important dans la détoxication des métaux, dans des proportions plus faibles pour le Cd que pour le Zn (Mason et Jenkins, 1995). Chez les mollusques bivalves marins, ils ne sont pas retrouvés dans les cellules branchiales contrairement aux bivalves d‘eau douce (Marigómez et al., 2002). Ils contribuent néanmoins à la séquestration de nombreux métaux (Ca, Cd, Fe, Cu, Zn…) dans les hémocytes, mais aussi dans les cellules du manteau et de la glande digestive de nombreux bivalves et présentent l‘avantage d‘être régulièrement éliminées par les fécès (pour revue voir Marigómez et al., 2002).

38

Chapitre I. Etat des connaissances

- Les hémocytes et les cellules brunes Ces cellules sont présentes dans l‘hémolymphe des organismes invertébrés. Elles sont capables de phagocyter les corps étrangers et ainsi d‘accumuler les éléments métalliques présents dans l‘hémolymphe. Les cellules brunes sont des hémocytes modifiés qui possèdent les mêmes propriétés de détoxication. Ce mécanisme de nettoyage du milieu intérieur complète l‘action des organes excréteurs (Ballan-Dufrançais et al., 1985). Ils constituent ainsi un des systèmes les plus importants dans le transport des métaux entre les organes. Giamberini et al. (1996) ont en effet montré une migration des hémocytes vers les branchies et la glande digestive chez la moule d‘eau douce Dreissena polymorpha soumise à des expositions métalliques (Cd, Cu, Pb). Comme dans les autres types cellulaires, les hémocytes contiennent de puissants ligands intracellulaires (métallothionéines), des lysosomes et des granules insolubles permettant d‘accumuler efficacement les métaux (Marigómez et al., 2002). Enfin, il est important de noter le rôle essentiel de l‘accumulation de zinc et de cuivre dans l‘activité antimicrobienne des hémocytes, comme cela a été montré chez l‘huître C. virginica (Fisher, 2004).

2.6. Processus d’élimination Une fois les métaux extrudés de la cellule via entre autres la libération par exocytose des vacuoles lysosomales dans le système circulatoire, ils sont pris en charge par les hémocytes circulants (Figure I.14). L‘élimination s‘effectue par voie rénale via les processus d‘ultrafiltration qui ont lieu au niveau de la glande péricardique. La glande péricardique va en effet produire des « cellules brunes », encore appelées « podocytes ». Ces cellules vont s‘organiser en couche et vont sécréter une lame basale pour former une barrière d‘ultrafiltration séparant le milieu intérieur (hémolymphe) et l‘espace cœlomique appelé encore « espace urinaire ». La lame basale chargée négativement va alors constituer la barrière principale d‘ultrafiltration chez les mollusques bivalves, laissant passer des molécules comprises entre 40 et 400 KDa (Meyhöfer et Morse, 1996). L‘élimination peut aussi se faire par voie digestive avec la formation de pseudofécès qui correspond au relargage des corps résiduels du tissu digestif, par exocytose au niveau des cellules de la glande digestive. Enfin, il est à noter aussi que des phénomènes de mue chez les organismes aquatiques possédant une cuticule externe participent à l‘élimination de métaux, de même que la formation de la coquille, de la dentition ou encore du byssus chez les bivalves (Mason et Jenkins, 1995).

39

Chapitre I. Etat des connaissances

Figure I.14. Mécanismes d‘élimination des métaux chez les bivalves. Les métaux (M) peuvent être transportés vers la glande péricardique via les hémocytes (Hc) ou les protéines de l‘hémolymphe afin d‘être éliminés par ultrafiltration à travers la double couche « lame basale-podocytes » qui sépare l‘espace cœlomique (ou urinaire) de l‘hémolymphe. Les hémocytes réduisent aussi les teneurs métalliques par le processus de diapédèse. Enfin, le relargage des corps résiduels du tissu digestif, par exocytose au niveau des cellules de la glande digestive permet aussi l‘élimination des métaux à travers les pseudofécès.

40

Chapitre I. Etat des connaissances

C. Généralités sur les mécanismes de défenses immunitaires des mollusques bivalves

Les mollusques bivalves possèdent un système immunitaire « inné » qui contribue au maintien de l‘intégrité de l‘organisme hôte en éliminant les composants étrangers (particules du non soi, bactéries, virus, parasites…). Ce système repose sur des cellules, appelées hémocytes, capables de proliférer et de se différencier sous le contrôle de facteurs humoraux. En termes de défense immunitaire, nos travaux ont uniquement porté sur les activités de défenses cellulaires, au détriment des facteurs humoraux et de leur participation dans le maintien de l‘intégrité de l‘organisme. En effet, chez les mollusques, les mécanismes de défenses cellulaires détiennent un rôle majoritaire dans les processus immunitaires au regard des mécanismes humoraux considérés comme plus mineurs (Roch, 1999). Dès lors, ce chapitre abordera plus particulièrement la description des cellules hémocytaires, ainsi que leurs propriétés et leurs fonctions, avant de présenter de manière plus synthétique le rôle des facteurs humoraux dans l‘immunité des bivalves.

1. Les hémocytes, des cellules à activité multiple 1.1. Origine et types hémocytaires L‘existence d‘un ou de plusieurs sites d‘hématopoïèse chez les bivalves est supposée en l‘absence d‘apparente multiplication des hémocytes circulants dans l‘hémolymphe. Il est généralement admis que les hémocytes proviennent d‘une différenciation de cellules du tissu conjonctif. Toutefois, certains auteurs suggèrent l‘existence d‘un organe hématopoïétique particulier, tandis que d‘autres considèrent que le tractus digestif pourrait intervenir comme centre d‘hématopoïèse (Cheng, 1981 ; Lorteau et al., 1995). Il est par ailleurs communément admis que les divers types hémocytaires pourraient représenter des stades de développement différents d‘un même type cellulaire. A partir de cette hypothèse, différents modèles ont été proposés (Figure I.15) (Cheng, 1981 ; Mix, 1976). A l‘heure actuelle, aucun travail ne permet de confirmer ou d‘infirmer ces hypothèses, et de localiser de manière claire le lieu de l‘hématopoïèse chez les mollusques bivalves.

41

Chapitre I. Etat des connaissances

Figure I.15. Modèles généraux d‘ontogénèse chez les mollusques bivalves proposés par (a) Mix (1976) et (b) Cheng (1981) (Gagnaire, 2005).

Tableau I.3. Types hémocytaires chez les mollusques bivalves utilisés dans ces travaux de thèse (R. philippinarum et C. edule) (d'après Wotton et al., 2003; Cima et al., 2000).

Ruditapes philippinarum Cerastoderma edule Taille (µm) Proportion relative (%) Taille (µm) Proportion relative (%) Hémoblastes 3-5 18,97 (± 0,63) / / Granulocytes 3-16 48,05 (± 1,43) 4-7 75,12 (± 7,71) Hyalinocytes 4-12 32,18 (± 0,99) 3-4 15,75 (± 1,98) Cellules séreuses 7-15 0,8 (± 0,19) 10-12 9,13 (± 1,14)

42

Chapitre I. Etat des connaissances

Chez ces organismes, la classification des hémocytes est basée essentiellement sur des critères morphologiques. Ils correspondent à la taille, la forme, le rapport nucléoplasmique, les affinités tinctoriales (colorations) et la présence d‘organites intracellulaires observés en microscopie. A ce jour, deux principaux types cellulaires sont identifiés et acceptés par la plupart des auteurs (Hine, 1999) : - Les granulocytes, cellules présentant un petit noyau entouré d‘un cytoplasme abondant caractérisé par de nombreuses granulations (rapport nucléoplasmique (Noyau/Cytoplasme) faible). - Les hyalinocytes (appelés aussi agranulocytes ou macrophages) caractérisés par un rapport nucléoplasmique élevé ainsi que par l‘absence ou quasi absence de granules. Au sein de ces deux grands types, le nombre de type cellulaires est très variable. On distingue notamment les granulocytes basophiles, neutrophiles et acidophiles chez de nombreuses espèces de mollusques (Tiscar et Mosca, 2004). Chez certaines espèces, comme la coque et la palourde, un troisième type cellulaire est identifié et appelé « cellule séreuse » ou « cellule brune » en raison de la présence de globules cytoplasmiques pigmentés (Tableau I.3). Ce troisième type cellulaire, lorsqu‘il est présent, est retrouvé dans une proportion moindre que les granulocytes et les hyalinocytes. Ces cellules séreuses sont des hémocytes modifiés, non mobiles, qui semblent être synthétisés dans les glandes péricardiques. Ces cellules sont capables de s‘agréger et de réaliser la phagocytose, mais leur fonction la plus importante est l‘élimination des produits de dégradation et d‘excrétion (Cheng, 1981 ; Cima et al., 2000).

1.2. Propriétés a. Plasticité et mobilité Les hémocytes sont des cellules de forme amœboïde possédant la capacité d‘émettre des évaginations de leur membrane pour former des pseudopodes (Chu, 2000). Ces derniers sont impliqués dans les processus d‘internalisation de particules étrangères lors de la phagocytose. De plus, lorsque les hémocytes se trouvent en contact d‘un support, la formation de ces pseudopodes leur permet de se mouvoir (Fisher, 1986). Dans les stades de néoplasie disséminée avancée, maladie s‘apparentant à une leucémie chez les vertébrés, les hémocytes retrouvent une forme totalement sphérique et perdent leur plasticité et leur capacité à former des pseudopodes (Villalba et al., 2001).

43

Chapitre I. Etat des connaissances

b. Chimiotactisme Les hémocytes de mollusques peuvent être attirés par des substances produites par les organismes étrangers ou par le mollusque lui-même. L‘orientation du déplacement cellulaire est primordiale dans les phénomènes de défense, notamment pour la localisation et la capture de particules étrangères (Fisher, 1986). c. Agrégation cellulaire et adhésion Lorsque des conditions stressantes apparaissent, les hémocytes ont la capacité de s‘agréger les uns avec les autres. En l‘absence de toute lésion ou pathogène, on parle dans ce cas d‘agrégation dite « spontanée » comme cela est observé en condition d‘expérimentation in vitro. Dans le cas d‘atteinte de l‘organisme, ce mécanisme d‘agrégation va permettre à la fois l‘immobilisation de corps étrangers et la réparation des lésions. Ce phénomène d‘agrégation hémocytaire s‘apparente alors au mécanisme de coagulation des cellules sanguines des vertébrés à la différence notable que ce processus est totalement réversible pour les hémocytes (Feng, 1988). Le phénomène d‘adhésion est distinct de l‘agrégation et n‘intervient pas en même temps que ce dernier. Ce phénomène correspond à une interaction entre certaines molécules présentes à la surface des cellules (lectines), et les substrats, notamment les bactéries.

1.3. Les réactions de défenses cellulaires a. Apoptose ou mort cellulaire programmée L‘apoptose est un processus cellulaire actif par lequel des cellules déclenchent leur autodestruction en réponse à un signal. C'est une mort cellulaire physiologique, génétiquement programmée, nécessaire à la survie des organismes pluricellulaires. Elle est en équilibre constant avec la prolifération cellulaire et se différencie ainsi de la nécrose ou de la cytolyse. Les caractéristiques d‘une mort cellulaire programmée ou apoptose correspondent à des transformations comme le bourgeonnement des membranes au niveau cellulaire et l‘arrondissement des cellules, une condensation de la chromatine, une fragmentation de l‘ADN cellulaire, et une perturbation du métabolisme mitochondrial amenant in fine la destruction de la cellule (Kim & Sharma, 2004). L‘apoptose est une des principales réponses face à une infection virale (Renault et al., 2000), en empêchant ainsi la dissémination des pathogènes dans l‘organisme. Divers travaux ont permis de démontrer son existence chez Ruditapes philippinarum, Ostrea edulis et Crassostrea gigas (Renault et al., 2000; Lacoste et al., 2001). 44

Chapitre I. Etat des connaissances

b. Inflammation et infiltration hémocytaire La réponse inflammatoire constitue un des mécanismes immunitaires innés les plus précoces en cas de lésion tissulaire ou d‘infections par des agents pathogènes, aussi bien chez les vertébrés que les invertébrés (Janeway et al., 2001). Elle se traduit par une augmentation locale de la température et une vasodilatation des vaisseaux sanguins ou des sinus qui va permettre l‘infiltration des cellules compétentes sur le site de l‘infection (Cheng, 1983). Cette augmentation de la densité d‘ hémocytes, ou leucocytose, sur le lieu de l‘infection constitue un préambule à la mise en place des mécanismes de phagocytose afin d‘éviter la dissémination du pathogène, et a pour rôle d‘initier la cicatrisation par agrégation hémocytaire. La réparation des blessures implique en effet une infiltration hémocytaire sur le site de la blessure (leucocytose) suivi des phénomènes d‘agrégation et d‘adhésion des hémocytes, afin de former un « bouchon hémocytaire » permettant de limiter les pertes d‘hémolymphe et la pénétration de germes (Cheng, 1988). Les cellules lésées sont ensuite remplacées par des hémocytes qui se différencient en cellules fusiformes (Sparks & Morado, 1988) et du collagène est déposé entre les cellules du bouchon hémocytaire. Enfin, les débris de tissus nécrotiques sont éliminés par phagocytose. De la même manière, les hémocytes participent aussi aux mécanismes de réparation de la coquille en migrant à la surface du manteau dès les premiers stades de la réparation (Fisher, 1986). Ils transportent du calcium et des protéines sur le site de la blessure, indispensables à la réparation de la coquille (Cheng, 1996). Le facteur le plus important induisant une leucocytose chez les mollusques correspond à l‘infection par des organismes pathogènes. En effet, 72 h après l‘injection de Vibrio tapetis chez Ruditapes philippinarum et Ruditapes decussatus, le nombre d‘hémocytes circulants augmente de manière importante (Oubella et al., 1993). De la même façon, l‘augmentation de la densité des hémocytes circulants chez l‘huitre américaine, Crassostrea virginica, est corrélée positivement au niveau d‘infection par le protozoaire pathogène Perkinsus marinus (Anderson et al., 1995). c. La phagocytose Une fois que les cellules compétentes, circulantes ou tissulaires, ont été recrutées au niveau du tissu lésé ou infecté, la phagocytose peut se mettre en place pour éliminer des éléments endogènes (cellules mortes, lésées) ou exogènes (pathogènes, nourriture). Hormis quelques différences, le processus de phagocytose, élément essentiel de l‘immunité innée, est fondamentalement le même chez les invertébrés et les vertébrés.

45

Chapitre I. Etat des connaissances

Figure I.16. Les mécanismes de la phagocytose

46

Chapitre I. Etat des connaissances

Il a de plus été démontré que la capacité de phagocytose n‘est pas la même pour tous les types hémocytaires chez les bivalves selon les espèces. Chez R. philippinarum les deux types cellulaires sont capables de réaliser la phagocytose. Toutefois, les granulocytes semblent être plus actifs dans ce processus (Cima et al., 2000). Chez la coque C. edule, seuls les granulocytes semblent être capables de réaliser la phagocytose, alors que les hyalinocytes semblent être plus impliqués dans les mécanismes d‘agrégation (Russell-Pinto et al., 1994). La phagocytose est caractérisée par quatre étapes incluant la reconnaissance, l'adhésion, l'ingestion et la destruction des particules (Figure I.16) (Galloway et Depledge, 2001). La reconnaissance de particules étrangères peut être réalisée passivement par rencontres aléatoires, ou activement par chimiotactisme (Feng, 1988). L’adhésion est l'étape au cours de laquelle la membrane de la cellule phagocytaire adhère à la particule qu'elle va ingérer. Cette étape se fait grâce aux lectines qui se fixent à la surface du corps étranger permettant ainsi leur reconnaissance (opsonisation). Des liaisons se forment entre les glycoprotéines membranaires des cellules immunitaires et les oses du corps étranger. L’ingestion s‘effectue par modification du cytosquelette de la cellule permettant la formation de pseudopodes qui entourent la particule phagocytée et forme ainsi une nouvelle vacuole intracellulaire, le phagosome. Une fois à l‘intérieur de ce phagosome, la destruction de la particule phagocytée peut être de deux types : la digestion est consécutive à l'accolement et à la fusion des lysosomes avec la membrane du phagosome constituant ainsi un phagolysosome. Les divers enzymes lyzosomales vont alors se déverser et, selon leur spécificité, s'attaquer aux divers constituants de la particule ou du micro- organisme (Chu, 2000). La digestion peut aussi s‘effectuer par un mécanisme appelé « flambée oxydative » faisant intervenir des espèces réactives de l‘oxygène (ERO). La production de ces métabolites est initiée par l‘attachement de particules étrangères sur la membrane plasmique des hémocytes, ce qui va activer la NADPH oxydase membranaire qui va catalyser la transformation - de l‘oxygène moléculaire (O2) en anion superoxyde (O2 ). Cet anion superoxyde peut ensuite être dismuté en hydroperoxyde d‘hydrogène (H2O2) sous l‘action de superoxyde dismutase (SOD). Ce dernier métabolite peut donner naissance à d‘autres ERO comme les radicaux hydroxyles (OH-) 1 ou l‘oxygène singulet ( O2), et peut être transformé en acide hypochloreux (HOCl) sous l‘action de la myéloperoxydase (MPO) ou bien être réduit en H2O et O2 sous l‘action de la catalase (CAT) (Buggé et al., 2007). Cette flambée oxydative associée à la phagocytose a été mise en évidence chez de nombreuses espèces de bivalves comme Crassostrea virginica, Crassostrea gigas, Ostrea edulis, Mytilus edulis, Mytilus galloprovincialis, Mercenaria mercenaria, Pecten

47

Chapitre I. Etat des connaissances

maximus, Ruditapes philippinarum et Cerastoderma edule (Wotton et al., 2003 ; Cima et al., 2001 ; Gagnaire, 2005). d. L’encapsulation L‘encapsulation permet l‘élimination de particules de taille trop importante pour être phagocytées, comme un individu parasite (protozoaires, nématodes, trématodes…) (Marmaras et Lampropoulou, 2009). Les hémocytes ne pouvant pas phagocyter la particule se différencient en cellules de type fibroblastique et s‘organisent en couches concentriques afin de former une capsule fine autour du parasite. Les hémocytes granuleux ont été considérés comme les principaux responsables de la formation des capsules chez les mollusques. Le modèle le plus étudié est le mollusque gastéropode Biomphalaria glabrata (Mollusca: Pulmonata) chez qui les hémocytes sont impliqués dans la réponse immune vis-à-vis des larves de trématodes Schistosoma mansoni (Van der Knaap and Loker, 1990). Ce mécanisme est accompagné d‘un phénomène de mélanisation et la dégradation de la particule encapsulée s‘effectue par flambée oxydative ou asphyxie (Marmaras et Lampropoulou, 2009).

1.4. Autres fonctions hémocytaires En plus de maintenir l‘intégrité de l‘organisme à travers la mise en place de réactions de défenses contre les pathogènes, les hémocytes sont impliqués dans bien d‘autres processus vitaux. Ils participent par exemple aux mécanismes de digestion et de transport de nutriments grâce à leur capacité à migrer dans le tube digestif, à endocyter et dégrader les particules digestives via l‘action des enzymes lysosomales pour en relarguer les nutriments (Cheng, 1996). Les cellules séreuses participent aussi à l‘élimination et l‘excrétion des produits de dégradation par phagocytose dans les glandes péricardiques (Cheng, 1981). Enfin, des travaux réalisés principalement sur les huîtres Crassostrea gigas et Ostrea edulis ont mis en évidence le rôle des hémocytes dans les mécanismes de détoxication des polluants et notamment des métaux comme le fer, le plomb, le cuivre ou le zinc (Coombs et Geroge, 1977). L‘accumulation importante de zinc et de cuivre dans les hémocytes de C. virginica, même en l‘absence de toute contamination anthropique, jouerait un rôle important dans l‘activité antimicrobienne de ces cellules (Fisher, 2004).

48

Chapitre I. Etat des connaissances

2. Les facteurs humoraux responsables de l’immunité Les substances solubles présentes dans l‘hémolymphe jouent un rôle important dans l‘immunité innée, indirectement à travers la régulation certains mécanismes cellulaires, et directement à travers leurs activités antimicrobiennes diverses (Chu, 2000). Les mécanismes les plus importants mis en place particulièrement chez les mollusques bivalves concernent :

- La cascade phénoloxydase qui joue un rôle crucial dans l‘immunité innée en faisant intervenir une enzyme cuivre-dépendante détectée dans les hémocytes et l‘hémolymphe, la phénoloxydase (PO) (Muñoz et al., 2006 ; Bryan-Walker et al., 2007). Cette enzyme catalyse la réaction de conversion de phénols en quinones, molécules extrêmement réactives. Ces molécules peuvent à leur tour être polymérisées en mélanine en produisant différents intermédiaires à propriétés bactéricides (Coles & Pipe, 1994; Söderhäll et Cerenius, 1998). Ce processus de mélanisation est impliqué dans l‘encapsulation et l‘élimination de nombreux parasites (Bryan-Walker et al., 2007).

- L’oxyde nitrique synthétase qui consitue une voie concomitante ou alternative à la phagocytose dans l‘élimination des organismes pathogènes (Ottaviani et al., 1993). Cette enzyme présente dans les hémocytes catalyse la production d‘oxyde nitrique (NO) à partir de la L-arginine en utilisant le NADPH comme cofacteur. Les oxydes nitriques sont des molécules réactives qui interviennent dans la régulation des systèmes nerveux, cardiovasculaire et immunitaire, et sont utilisé en tant qu‘agents antibactériens chez les mollusques (Franchini et al., 1995).

- Les lectines qui sont des glycoprotéines dépourvues d‘activité enzymatique se liant spécifiquement et de façon réversible à un ose ou oligoside. Ces protéines participent grandement à l‘opsonisation des particules étrangères, du fait de leur spécificité pour certains sucres (mannose, fructose) présents à la surface des agents pathogènes (virus, bactéries, champignons et parasites) (Pipe, 1990), mais aussi aux processus d‘agrégation hémocytaire et d‘agglutination des particules étrangères, ce qui conduit à leur lyse via les mécanismes de phagocytose.

- Les peptides anti-microbiens (AMP) en tant que médiateurs humoraux importants de l‘immunité innée contre les infections virales, microbiennes et fongiques (Zasloff, 2002). Leur mécanisme d‘interaction repose sur la formation de pores transmembranaires qui perméabilisent les membranes bactériennes, préludant une mort certaine de la bactérie.

49

Chapitre I. Etat des connaissances

Chez les mollusques, les AMP, mis en évidence pour la première fois chez la moule Mytilus edulis, et ont depuis été identifiés chez d‘autres espèces telles M. galloprovincialis, C. gigas, C. virginica, R. philippinarum et R. decussatus (pour revue voir Cheng-Hua et al., 2009).

- Les enzymes lysosomales responsables de la dégradation des agents infectieux réalisée à l‘intérieur ou à l‘extérieur des cellules (Xue & Renault, 2000). Différentes hydrolases s'attaquant, selon leur spécificité, aux divers constituants de la particule ou du micro- organisme, ont été mises en évidence chez de nombreuses espèces de bivalves, et notamment chez la coque C. edule et la palourde R. philippinarum (Tableau I.4).

Tableau I.4: Activités enzymatiques dans les hémocytes de R. philippinarum et C. edule (d‘après Russell- Pinto et al., 1994; Cima et al., 2000; Wotton et al., 2003) (R.P. Ruditapes philippinarum ; C.E. Cerastoderma edule)

Granulocytes Hyalinocytes Cellules séreuses Enzymes hydrolytiques R.P. C. E. R.P. C.E. R.P. C.E. β-glucuronidase + - - - - - Phosphatase acide - + - - - - Phosphatase alkaline + nd + nd - nd 5'-nucléotidase - nd - nd - nd Esterase non spécifique* + + + nd - nd Esterase acide + - - - -

Arylsulphatase* + + - nd - nd Enzymes oxydantes Peroxydase + - + - - + Cytochrome C oxydase + nd + nd - nd * l‘activité de ces enzymes est présente dans les hémocytes de C. edule, sans précision sur la distinction du type cellulaire impliqué; Par défaut, le symbole + a été assigné aux granulocytes qui représentent la majorité des hémocytes chez cette espèce. nd: données non disponibles.

50

Chapitre I. Etat des connaissances

3. Les effets des métaux sur les paramètres hémocytaires Chez les mollusques bivalves, les mécanismes de défense sont modulés par les conditions environnementales (température, salinité, oxygène, etc.) mais aussi par la présence de polluants. Des études sur le terrain ont montré la relation entre présence de métaux lourds et paramètres hémocytaires chez les bivalves, mais les effets les plus importants sont principalement démontrés in vitro (Bouilly et al., 2006 ; Fisher et al., 2000; Oliver et al., 2003). Le cadmium et le mercure sont des métaux très toxiques et leurs effets sur les hémocytes de bivalves ont été largement décrits. En dépit de leur caractère essentiel, le zinc et surtout le cuivre peuvent être également responsables d‘altérations hémocytaires importantes. Le cadmium provoque in vivo l‘augmentation de la densité hémocytaire chez C. virginica (Cheng, 1988), de C. gigas (Auffret et Oubella, 1994) et de M. edulis (Coles et al., 1995) et diminue in vitro l‘agrégation hémocytaire spontanée chez C. gigas (Auffret et Oubella, 1994). Comme le cadmium, le cuivre est capable d‘induire une leucocytose in vivo à de faibles concentrations (Boutet et al., 2002 ; Oubella et Auffret, 1995; Pipe et al., 1999). De plus, l‘exposition in vitro à 10 µg.L-1 de cuivre diminue l‘agrégation hémocytaire spontanée chez C. gigas (Auffret et Oubella, 1994). La mortalité des hémocytes peut être augmentée par l‘exposition au Cd et au Hg in vitro et in vivo chez plusieurs espèces de bivalves (Fournier et al., 2001 ; Gagnaire, 2005). Ces métaux peuvent également inhiber in vivo le relarguage du lysozyme et d‘autres enzymes lysosomales à partir des granulocytes chez M. edulis (Coles et al., 1995). Ils peuvent aussi altérer in vivo la membrane lysosomale chez M. edulis (Coles et al., 1995), R. philippinarum (Matozzo et al., 2001) et chez M. edulis (Gagnaire, 2005). De manière générale, la phagocytose semble activée in vitro par de faibles doses de Cd, de Hg ou de Cu (10-9 et 10-8 M) alors qu‘elle est inhibée par de fortes doses (10-4 M) (Brousseau et al., 2000 ; Pipe et al., 1999 ; Parry and Pipe, 2004). Ceci a également été démontré pour la production d‘ERO après exposition au Cd. Les faibles doses de cet élément l‘activent, tandis que les fortes doses l‘inhibent chez C. gigas (Boutet et al., 2002). La production d‘anion superoxyde est également diminuée en présence de cet élément chez M. edulis (Pipe et al., 1999). Ces métaux peuvent de plus affecter les systèmes antioxydants : le Cd et le Cu provoquent une augmentation de la peroxydation lipidique chez C. gigas, M. edulis et C. virginica (Géret et al., 2002). Le Cd induit de plus une diminution de l‘activité de la SOD dans la glande digestive de M. edulis (Gagnaire, 2005).

51

Chapitre I. Etat des connaissances

52

CHAPITRE II. BASES METHODOLOGIQUES

Pied de coque présentant des métacercaires d‘Himasthla elongata.

53

54

Chapitre II. Bases méthodologiques A. Détermination et numération des organismes pathogènes

1. Analyse des communautés parasitaires de trématode 1.1. Objectif Estimer la structure des peuplements de parasite trématodes digènes de bivalves hôtes, la prévalence et l‘intensité de l‘infestation.

1.2. Méthodologie Après mesure de la longueur de coquille, le bivalve est ouvert. La poche contenant éventuellement les métacercaires de Meiogymnophallus minutus est recherchée à la loupe binoculaire au niveau de la charnière, près du coeur. Le cas échéant, cette poche est retirée, écrasée entre lames. Selon le nombre de métacercaires, celles-ci sont soit directement comptées, soit comptées à partir de photographies (l‘abondance peut excéder 1000 métacercaires). Les autres espèces de parasite sont identifiées et dénombrées par écrasement de la chair du bivalve entre deux lames de verre et observation à la loupe binoculaire (de Montaudouin et al., 2009).

1.3. Variables mesurées Richesse parasitaire = nombre d‘espèces par individu hôte ou par population hôte Abondance parasitaire = nombre de parasite par individu-hôte Intensité parasitaire = nombre de parasite par individu-hôte parasité Prévalence parasitaire = % d‘individus parasites

2. Numération de Vibrio sur milieu gélosé de culture sélective (T.C.B.S.)

2.1. Objectif Dénombrer les bactéries du genre Vibrio présents dans l‘hémolymphe et les fluides (extra et intrapalléaux) des coques et des palourdes. Il est à noter que cette partie des bases méthodologiques qui concerne la microbiologie n‘a pas été réalisée personnellement, mais par

55

Chapitre II. Bases méthodologiques une équipe avec laquelle j‘ai collaboré (F. Jude, N. Raymond, N. Le Goïc). La méthodologie est donc placée à titre indicatif sans avoir été réellement mise en pratique par moi-même.

2.2. Méthodologie Les Vibrio sont dénombrés sur milieu T.C.B.S. (Thiosulfate. Citrate. Bile. Saccharose. : extrait de levure(5.0 g), peptone de protéose n°3 (10.0 g), Citrate de sodium (10.0g), Thiosulfate de sodium (10.0 g), Fiel de boeuf (8.0 g), Saccharose (20.0 g), Chlorure de sodium (10.0 g), Citrate d‘ammonium ferrique (1.0 g), Bleu de Bromothymol (0.04 g), Bleu de Thymol (0.04 g), Gélose( 15.00 g), pH final 8.6 ± 0.2) conforme aux normes NF V 45-111 et NF ISO 8914. a. Préparation de la gélose T.C.B.S. Mettre 89 g de milieu T.C.B.S.1 Agar (DIFCO) dans 500 mL d‘Eau de Mer Naturelle Filtrée sur filtre 0.7 μm (GF/F, Whatman) et 0.2 μm (Nitrate de cellulose, Whatman) (EMNF 0.2) et 500 mL d‘eau distillée stérilisée à l‘autoclave (20 minutes à 120°C). Porter à ébullition pendant 1 à 2 minutes. Le milieu T.C.B.S. ne doit pas être autoclavé. Maintenir le milieu de culture en surfusion vers 45°C. Le milieu de culture est réparti en condition stérile dans des boîtes de Pétri. Les boîtes de gélose T.C.B.S. sont conservées à 18°C. Elles doivent êtres ensemencées dans les 48 heures qui suivent la préparation. b. Numération de Vibrio dans l'hémolymphe La numération de Vibrio est réalisée selon la technique de numération en surface. Cinquante μL d‘hémolymphe non diluée est déposé sur la gélose T.C.B.S. et étalé au râteau sur toute la surface de la boîte. Pour chaque inoculum 1 gélose T.C.B.S. est ensemencée. c. Numération de Vibrio dans les fluides Une gamme de dilution est réalisée avec de l‘EMNF 0.72 Deux cents millilitres d‘EMNF 0.7 (Eau de Mer Naturelle Filtrée sur filtre 0,7 μm (GF/F, Whatman)) sont répartis dans des flacons SCHOTT de 250 mL et stérilisés à l‘autoclave pendant 20 minutes à 120°C. L‘EMNF 0.7 est répartie dans des microtubes stériles à raison de 450 μL / microtube. Une gamme de dilution au 1/10 est réalisée. La numération de Vibrio est réalisée selon la technique de numération en surface. L‘inoculum de 100 μL pour les fluides dilués (10-1, 10-2, 10-3) est déposé sur la gélose T.C.B.S. et étalé au râteau sur toute la surface de la boîte. Pour chaque inoculum 2 géloses T.C.B.S. sont ensemencées.

56

Chapitre II. Bases méthodologiques

d. Lecture Après 72 heures d‘incubation à 18°C sont dénombrées les Unités Formant Colonies (U.F.C.) jaunes, les U.F.C. vertes, les U.F.C. noires ou au centre noir. A titre indicatif, les résultats possibles sur milieu gélosé de culture sélective T.C.B.S. sont: Colonies jaunes (Vibrions saccharose positif : V. cholerae, V. alginolyticus, V. fluvialis, V. furnisii). Colonies vertes (Vibrions saccharose négatif : V. parahaemolyticus:, V. vulnificus, V. mimicus). Petites colonies jaunes : bactéries n‘appartenant au genre Vibrio (E. coli, Klebsiella, Shigella). Petites colonies vertes : bactéries n‘appartenant pas au genre Vibrio (Enterococcus faecalis).

Colonies à centre noir : colonies H2S + (Vibrions ou non vibrions).

2.3. Variables mesurées Nombre d'U.F.C. du genre Vibrio par mL converti en nombre de bactéries du genre Vibrio par mL d'hémolymphe ou de fluides.

57

58

Chapitre II. Bases méthodologiques B. Techniques d‘analyse des métaux 1. Traitement des échantillons

1.1. Echantillons biologiques En premier lieu, il est indispensable de déterminer le poids des échantillons à doser de manière à pouvoir calculer la contamination de ces derniers non seulement en termes de teneur et de quantité de métal, mais aussi en termes de concentration. Dans le cadre de cette thèse, la plupart des résultats de concentration métallique dans les échantillons biologiques sont exprimés en fonction du poids sec de chair. En effet, travailler avec des échantillons secs permet de s'affranchir des problèmes de teneurs en eau variables d‘un individu à l‘autre et qui peuvent être modifiées, par exemple, par la congélation des échantillons. Cependant, certains travaux présentés dans ce manuscrit utilisent des valeurs de concentrations métalliques exprimées en fonction du poids frais de l‘échantillon. Cette prise en compte du poids frais à la place du poids sec a été effectuée le plus souvent pour des raisons pratiques et analytiques. Ces dernières seront développées directement dans chacune des parties concernées. La détermination du poids sec est réalisée par une pesée des tissus après un passage préalable des échantillons dans une étuve (température 45°C) pendant une durée minimum de 48 h. A l‘exception du dosage du mercure, qui peut être réalisé sur des échantillons solides, le dosage des concentrations en zinc, cuivre et cadmium par nos appareils de mesure requiert obligatoirement l‘obtention d‘une phase liquide. Pour ce faire, les échantillons biologiques secs (branchies, masse viscérale, reste du corps…) subissent une étape préalable de digestion par attaque acide (acide nitrique HNO3, 65%, Fluka) dans des tubes en polypropylène placés à 100°C pendant 3h. Les digestats sont ensuite dilués avec de l'eau ultra-pure (MilliQ, Millipore®) de manière à obtenir une concentration finale d‘acide nitrique comprise entre 2 et 20%. Une fois cette étape effectuée, les analyses peuvent être réalisées directement. Dans le cas contraire les échantillons sont stockés à l‘obscurité en chambre froide (+4°C) en attendant d‘être dosés.

1.2. Echantillons d’eau Les analyses effectuées sur des échantillons d‘eau dans le cadre de ces travaux de thèse (expériences en laboratoire ou prélèvements sur le terrain) concernent essentiellement des échantillons d‘eau de mer, naturelle ou synthétique. L‘eau de mer constitue une matrice riche en sels pouvant éventuellement poser des problèmes d‘interférence lors des dosages. Dans ce cadre, des mesures précises ont été employées selon l‘élément à doser. Dans tous les cas, les 59

Chapitre II. Bases méthodologiques

échantillons d‘eau de mer ont été préalablement acidifiés à 2 % d'acide nitrique afin de provoquer la dissociation des complexes du métal et de limiter l‘adsorption des cations métalliques sur les parois du tube de prélèvement et les matières en suspension par complétion avec les protons (H+). Pour le dosage du mercure, les échantillons sont introduits directement après cette étape d‘acidification, la présence de sels n‘interférant pas sur le dosage du mercure total. En ce qui concerne le dosage du cuivre, réalisé sur un spectrophotomètre d'absorption atomique en flamme, la gamme étalon employée lors de l‘analyse a été préparée et renouvelée tous les 2 jours avec la même matrice que celle de l‘échantillon de manière à éviter toute absorption non spécifique pouvant induire un biais lors de l‘analyse. Enfin, le dosage du Cd dans l‘eau de mer par spectrophotométrie d'absorption atomique électrothermique a nécessité la mise au point d'une nouvelle procédure de dosage, le chlorure de sodium (NaCl) perturbant fortement la mesure. Les échantillons sont tout d'abord dilués avec de l‘eau ultra-pure acidifiée à 2% pour abaisser leur salinité à 2 ‰ ce qui a pour effet de diminuer l'effet matrice du NaCl. Il est ensuite nécessaire d‘employer un modificateur de matrice supplémentaire, le nitrate d'ammonium (NH4NO3) qui, couplée à une procédure d'optimisation des températures du programme thermique du four, permet ainsi d'éliminer le NaCl lors de l'étape de minéralisation (l‘action de ce modificateur est présentée en détail ultérieurement).

2. Principe général du dosage des métaux par spectrophotométrie d'absorption atomique La quantification des métaux a reposé sur une seule et même méthode : la spectrométrie d'absorption atomique, communément nommée SAA, qui constitue un outil privilégié d'analyse en sciences environnementales. Couplée à un atomiseur, la spectrométrie SAA autorise le dosage d'éléments majeurs et traces dans divers types de substrats : matrices biologiques, eaux, sédiments, roches, déchets solides…

Dans son principe, la SAA consiste à vaporiser l'échantillon liquide et à le chauffer à l'aide d'un atomiseur (une flamme ou un four). Les éléments à doser sont alors portés à l'état d'atomes libres sous forme de vapeur. Les atomes à l'état fondamental vont alors absorber de façon spécifique le rayonnement émis par une source lumineuse (Figure II.1). La mesure de cette absorption qui correspond au rapport des intensités incidente et transmise, est mesurée à l'aide d'un prisme dispersif et d'une cellule photoélectrique : elle est directement proportionnelle à la concentration de l'élément (loi de Beer-Lambert) (Figure II.2). 60

Chapitre II. Bases méthodologiques

Notre laboratoire dispose d'un spectrophotomètre M6 Solaar AA spectrophotometer (Thermoptec) opérationnel en mode four graphite avec atomisation électrothermique, un spectrophotomètre SpectrAA 220FS (Varian®) avec atomisation par flamme, ainsi qu‘un spectrophotomètre d‘absorption atomique sans flamme (AMA 254, Leco France). La spectrophotométrie correspond à de la spectrométrie, à la différence qu‘elle se fait par comparaison de deux faisceaux (la substance à analyser et un blanc), alors que la spectrométrie est absolue.

Figure II.1. Principe de la spectrométrie d‘absorption atomique

Figure II.2. Principe général de la loi de Beer-Lambert

61

Chapitre II. Bases méthodologiques

Figure II.3. Spectrophotomètre d‘absorption atomique à flamme SpectrAA 220FS (Varian®) et principe du dosage des métaux (Cd, Cu, Zn) (Pierron, 2007).

Figure II.4. Spectrophotomètre d‘absorption atomique électrothermique M6 Solaar AA (Thermoptec) et principe du dosage du cadmium (Lucia, 2007).

62

Chapitre II. Bases méthodologiques

2.1. Dosage du zinc, du cuivre et du cadmium par spectrophotométrie d'absorption atomique en flamme (SAAF)

L'appareil utilisé est le SpectrAA 220FS (Varian®) (Figure II.3). L'échantillon liquide est aspiré par un passeur automatique et introduit dans l‘appareil de manière fragmenté en fines gouttelettes. L'atomiseur est constitué par un brûleur à fente produisant une flamme laminaire (température 2100°C), alimenté par un mélange air/acétylène. Le bruit de fond qui correspond à l'absorption non spécifique par des molécules non dissociées, des fumées et des particules est corrigé à l'aide d'une lampe deutérium. L'absorption du rayonnement émis par cette dernière correspond au fond et le rayonnement émis par les cathodes creuses correspond à la fois au fond et à l'élément à doser. La correction est alors obtenue par soustraction (Absorbance cathode creuse –

Absorbance deuterium).

En mode flamme, la limite de détection est de l'ordre du centième de mg.L-1 (ppm). En effet, sur le SpectrAA 220FS la limite de détection est de 50 µg.L-1 pour le Cu, 40 μg.L-1 pour le Zn, et 20 μg.L-1 pour le Cd. Au regard du caractère non essentiel du Cd, cette limite de détection trop élevée a posé des problèmes d‘analyse au niveau des échantillons faiblement concentrés. La sensibilité des dosages en mode flamme est limitée par des réactions secondaires (évaporation) et par le temps très court de passage dans la flamme. Pour pallier à ce problème et accroître la sensibilité du dosage, il est nécessaire de réduire ou d'éliminer ces deux facteurs par atomisation. Ceci est possible en spectrophotométrie d'absorption atomique électrothermique (SAAE) dans laquelle l‘atomisation est réalisée dans un four graphite d'un volume réduit sous atmosphère inerte. La limite de détection est alors de 0,1 μg.L-1 et permet notamment de doser le Cd présent à l‘état de traces dans les organismes.

2.2. Dosage du cadmium par spectrophotométrie d'absorption atomique électrothermique (SAAE) Le dosage du cadmium dans les échantillons peu contaminés est effectué par spectrophotométrie d'absorption atomique électrothermique avec le M6 Solaar (Thermo Elemental®) qui présente une limite de détection de 0,1 μg.L-1 pour le Cd (Figure II.4). Dans le cas présent, le volume de l‘échantillon à doser est considérablement réduit (20 µL) et l‘atomiseur est constitué par un four en graphite de forme cylindrique. L'intérieur du four, à l'exception de

63

Chapitre II. Bases méthodologiques l'étape de mesure, est en permanence balayé par un flux d'argon qui le protège de l'action de l'oxygène aux fortes températures. Lorsque l'échantillon est déposé dans le four, il va être soumis à une série de traitements thermiques permettant successivement la désolvatation (séchage), la pyrolyse (destruction d'un corps organique par la chaleur), la minéralisation (décomposition des molécules), l'atomisation (vapeur monoatomique) et le nettoyage du four. Contrairement à la SAAF, des modificateurs de matrice sont utilisés de manière à éviter toute absorption non spécifique due à des composés présents dans la matrice. Ces derniers peuvent en effet engendrer un biais très important lorsque des éléments sont dosés à l'échelle du μg.L-1. L‘action d‘un modificateur de matrice repose sur trois principes :

- Soit il réduit la volatilité de l‘analyte - Soit il augmente la volatilité de la matrice - Soit il réduit le bruit de fond de l‘absorption

Lors du dosage du Cd, nous avons utilisé le palladium (Pd) comme modificateur d'analyte. Ce dernier forme un alliage thermiquement plus stable avec le Cd, autorisant ainsi des températures d‘atomisation plus élevées (de 400°C sans palladium à 900°C en présence de ce dernier) et donc une destruction plus efficace de la matrice. Le nitrate de magnésium (Mg(NO3)2) est le deuxième modificateur de matrice utilisé lors du dosage du Cd. Ce modificateur convertit certains éléments de la matrice en de nouveaux éléments présentant des températures de décomposition/sublimation plus faibles permettant leur élimination plus précoce lors des étapes de pyrolyse ou de minéralisation. Le Mg(NO3)2 a aussi une petite action stabilisatrice de l‘analyte. Enfin, comme il est précisé plus haut, le nitrate d‘ammonium (NH4NO3) est utilisé lorsque le dosage de Cd est effectué dans une matrice salée contenant du NaCl, comme l‘eau de mer. Le NaCl a en effet une température de volatilisation très proche de celle du Cd et il est nécessaire de l‘éliminer ou le transformer en produit plus volatile afin d‘éviter toute absorption non spécifique. Dans ce cadre, le nitrate d‘ammonium a pour effet de dissocier et de diminuer la température de volatilisation des espèces séparées selon la réaction suivante :

NaCl + NH4NO3 NaNO3 + NH4Cl T° de volatilisation 1413°C 210°C 380°C 335°C

64

Chapitre II. Bases méthodologiques

Cette réaction, couplée à une procédure d'optimisation des températures du programme thermique du four (Pierron, 2007), permet ainsi d'éliminer le NaCl lors de l'étape de minéralisation.

2.3. Dosage du mercure par spectrophotométrie d’absorption atomique sans flamme

Dans l‘eau de mer et dans les matrices biologiques, le mercure total a été déterminé par spectrophotométrie d‘absorption atomique sans flamme (AMA 254, Leco France). Les échantillons (liquides ou solides) sont placés dans des nacelles en nickel et sont introduits dans l‘appareil à l‘aide d‘un passeur automatique (ASS 254). Une fois dans le tube catalytique, les échantillons se trouvent sous flux permanent d‘oxygène et vont être soumis à plusieurs traitements thermiques. Après une phase de séchage de 45 secondes à 200°C, la température est portée à 600°C pendant 180 secondes. A cette température, les différentes formes chimiques du mercure sont atomisées et volatilisées sous forme élémentaire gazeuse (Hg°). Le flux d‘oxygène pousse les gaz de combustion et le mercure à travers le catalyseur qui parfait la combustion et retient les composés halogénés, qui pourraient interférer avec le métal lors de la phase de lecture. Sous forme élémentaire gazeuse, le Hg° sortant du catalyseur est pré-concentré sur un amalgame à base d'or. Une fois le piégeage terminé, l‘amalgame est chauffé, et le mercure est relargué puis entraîné par le flux d‘oxygène vers une cellule de lecture, où l‘absorbance du Hg est détectée à 253,65 nm (Figure II.5). La limite de détection de l‘appareil est de 0,01 ng de mercure/échantillon, avec une variabilité analytique moyenne de 5 %.

3. Validité des mesures

Pour chaque appareil et chaque série de dosage, quatre échantillons certifiés (2 TORT : hépatopancréas de homard et 2 DOLT : foie de chien de mer fournis par le NRCCC-CNRC, National Research Council Canada – Conseil National de Recherches Canada) ainsi que deux blancs (acide pur, témoin d'une éventuelle contamination métallique) ont subi le même traitement que les échantillons biologiques afin de vérifier la validité de la méthode.

65

Chapitre II. Bases méthodologiques

Figure II.5. Principe du dosage du mercure avec le spectrophotomètre d‘absorption atomique sans flamme (AMA 254, Leco France).

66

Chapitre II. Bases méthodologiques C. Technique d‘analyse des métallothionéines

1. Rappels

Les métallothionéines sont des protéines extrêmement étudiées depuis plus de 60 ans, et les techniques de quantification se sont progressivement diversifiées en faisant le plus souvent intervenir leur principales caractéristiques : leur taille, leur poids, leur propriétés biochimiques (résistance à la chaleur et aux attaques acides) et colorimétriques, leur richesse en groupement thiols qui leur confère la capacité de fixer les métaux ainsi que leurs propriétés antigéniques (Tableau II.1).

Tableau II.1. Principales techniques d‘étude pour la purification et la quantification des métallothionéines

Techniques Propriétés de la métallothionéine Références Purification Electrophorèse dénaturante ou non Taille des protéines Richards et Beattie, 1995

Méthodes chromatographiques Chromatotographie échangeuse d'ions Suzuki, 1987 Propriétés biochimiques: taille, affinité Chromatographie liquide haute performance pour les ions métalliques Kägiet al., 1984 (HPLC) chromatographie d'exclusion (IEC) Lacorn et al., 2001

Traitements à la chaleur Thermo-stabilité des MT Hensbergen et al., 2000

Quantification Méthode électrochimique Propriétés redox des complexes formés par les groupements thiols avec les Bordin et al., 2000 métaux dans la protéine

Méthodes colorimétriques Propriétés colorimétriques de différents Meloni et al., 2005 (Western Blot) produits en association avec des mercaptans

Méthodes immunologiques Immunoréactivité avec des anticorps Windge et Garvey, 1983 (épitopes des protéines)

Méthodes de spectroscopie Paris-Palacios et al., 2000 Présence des groupements thiols (spectrofluorimétrie, spectrométrie de masse) Dabrio et al., 2001

Méthodes de saturation + spectrophotométrie Ali et al., 2002 d'absorption atomique ou spectrophotométrie Capacité de saturation des MT et affinité Bodar et al., 1990 gamma (saturation au Hg, Cd ou Ag "froids" ou différentielle pour les métaux Campbellet al., 2005 radioactifs)

La technique de quantification employée au laboratoire correspond à la technique de saturation, utilisant l‘affinité différentielle des métaux pour ces protéines. A travers les différentes techniques de saturation existantes, trois métaux ont été principalement employés : le

67

Chapitre II. Bases méthodologiques

Figure II.6. Principe de la méthode de dosage des MT par saturation avec le Hg «froid».

[M] : Métal, GSH : Glutathion, P1 à 4 : Protéines cytosoliques

68

Chapitre II. Bases méthodologiques mercure, le cadmium et l‘argent, aussi bien sous forme d‘élément stable que sous forme d‘isotope radioactif du métal. Dans notre laboratoire, l‘utilisation préférentielle du mercure par rapport aux autres métaux a été choisie pour des raisons méthodologiques. En effet, le principe de cette technique repose sur le déplacement des métaux originellement fixés sur les résidus cystéine de la protéine par un métal de plus haute affinité ajouté en excès. La concentration des MT est alors estimée par quantification dudit métal, après séparation des autres protéines et du reste du métal en excès. L'affinité différentielle des métaux pour les sites thiols des MT correspond à la séquence suivante décrite par Vasák en 1991 :

Hg(II)>>Cu(I), Ag(I), Bi(III)>>Cd(III)>Pb(II)>Zn(I)>Co(II)>Fe(II)

Dès lors, l‘utilisation du Cd et de l‘Ag comme métal « saturant » ne semble pas applicable dans le cadre de contamination par le cuivre ou le mercure, du fait de leur affinité supérieure sur les sites thiols des MT. De plus, des travaux menés par Couillard (1997) ont révélé des précipitations incomplètes du complexe hémoglobine-Ag, chargé de récupérer l‘Ag en excès dans le cytosol. La méthode de saturation par le mercure s‘est donc révélée comme étant la plus pertinente et a été mise au point par Dutton et al. (1993) utilisant l‘isotope 203 du mercure (émetteur γ) chez un poisson, et adaptée par Couillard et al. (1993) chez un mollusque bivalve d‘eau douce. Cette technique a par la suite été mise en place dans notre laboratoire (Baudrimont, 1997a) et modifiée de manière à remplacer le mercure radioactif par du mercure froid, dont la quantification par SAA est réalisable au sein du laboratoire et s‘avère particulièrement sensible.

2. Principe de la méthode de saturation par le mercure inorganique

Le principe général de cette technique ainsi que les différentes étapes sont décrits Figure II.6.

L‘échantillon biologique (minimum de 50 mg) est homogénéisé dans 1 mL de tampon Tris- HCl 25 mM (Sigma, pH 7,2 à 20°C) à l‘aide d‘un broyeur mécanique (T 10 basic ULTRA- TURRAX, IKA®). Cette première étape s‘effectue sous atmosphère d‘azote (sac à gant en polyéthylène étanche, W12953-Fisher Scientific) afin d‘éviter l‘oxydation des groupements thiols des cystéines R–SH en ponts disulfures R-S-S-R, empêchant ainsi la séquestration des métaux par les MT, et dans la glace afin de limiter l‘action des protéases. Après homogénéisation, les échantillons sont centrifugés à 4°C pendant 1 heure à la vitesse de 20 000 g (Eppendorf). Cette étape permet de séparer la fraction cytosolique contenant les MT (surnageant) du reste (culot :

69

Chapitre II. Bases méthodologiques débris cellulaires, organites, micro granules, noyaux…). Un volume de 200 µL de surnageant est -1 prélevé et mis en contact avec une solution constituée de Hg inorganique (HgCl2 50 mg.L ) et d‘acide trichloracétique (TCA 10%) pendant 10 minutes. Le mercure en excès va se substituer aux autres métaux présents sur les MT, et saturer l‘ensemble des sites accessibles au sein du cytosol. La présence d‘acide trichloracétique va abaisser le pH de l‘échantillon ce qui entraîne la précipitation des protéines de haut poids moléculaire alors que les MT restent en solution. L‘excès de mercure présent dans le cytosol complexé ou non aux autres protéines est récupéré en ajoutant 400 µL d‘une solution d‘hémoglobine (Hb) de bœuf ou de porc. L‘affinité de cette protéine pour le métal reste inférieure à celle de la MT mais supérieure à celle de l‘ensemble des autres sites thiols présents dans le cytosol. Le complexe Hb-Hg précipite rapidement dans ces conditions acides et ce dernier est séparé par une deuxième centrifugation de 20 minutes à 20 000 g à température ambiante, empêchant dès lors la décomplexation éventuelle du Hg fixé aux MT. Le mercure présent dans le surnageant final correspond alors uniquement au mercure fixé sur les groupements thiols des MT. Ce surnageant est donc récupéré pour l‘étape finale du dosage du mercure par spectrophotométrie d‘absorption atomique sans flamme.

Afin de s‘assurer de la validité de la méthode, trois « blancs » (200 µL tampon Tris-HCl 25 mM, 200 µL de solution HgCl2-TCA et 400 µL de solution d‘hémoglobine) sont systématiquement préparés afin de contrôler l‘efficacité de complexation du mercure non lié aux MT par l‘hémoglobine et de sa précipitation. Le surnageant après centrifugation ne doit contenir qu‘une très faible quantité de mercure (< 2 %) et la moyenne des trois valeurs de blancs est soustraite aux quantités de Hg mesurées dans les échantillons.

Afin de contrôler l‘efficacité du dosage, trois échantillons contenant une MT purifiée de mammifère (MT de foie de lapin diluée dans le tampon d'homogénéisation Tris-HCl 30 mM, pH 7,2 à 20°C pour une concentration finale de 10 μg.mL-1) sont traités et 100 ± 20 % de récupération de cette MT doivent être retrouvés à l'issue de la saturation au Hg. Au cours de nos analyses, le pourcentage moyen de récupération était de 109,4 ± 5,8 %.

Les résultats des dosages du mercure permettent de calculer une « capacité de séquestration tissulaire ». La quantité exacte de sites de fixation des métaux par molécule de MT n‘étant pas connue chez les bivalves étudiés, la concentration ne peut être directement exprimée en nombre de moles de MT par gramme de tissu frais. Dès lors, la concentration en MT dans les échantillons biologiques est exprimée en nanomoles de sites de liaison de Hg par gramme de poids frais.

70

Chapitre II. Bases méthodologiques

D. Techniques d‘analyse des paramètres hémocytaires

Les activités cellulaires des hémocytes chez la coque et la palourde ont été étudiées de manière à caractériser :

- La variabilité inter-espèce de base existant au niveau des paramètres immunitaires observés. - Les réponses immunitaires de chaque espèce suite à des infections (bactéries et trématodes). - Le lien entre réponse immunitaire et sensibilité aux pathogènes/maladies. - Les potentiels effets immunotoxiques des métaux traces étudiés.

Les analyses ont été réalisées par cytométrie en flux en collaboration avec le Laboratoire des Sciences de l‘Environnement Marin (UMR 6539-LEMAR, Université de Bretagne Occcidentale (UBO)) et le laboratoire INSERM de l‘Université Bordeaux 2.

1. La cytométrie en flux

1.1. Historique

Cette technique est née du besoin d‘automatiser le comptage des constituants cellulaires du sang. En 1934, Moldovan conçoit le premier appareil réalisant des numérations cellulaires en faisant défiler les cellules dans un fin capillaire dans lequel elles sont détectées par un capteur photoélectrique. En 1949, W. Coulter met au point un appareil qui associe la numération des cellules et la mesure de la taille par variation de la résistance du courant liquidien. Au cours des 35 dernières années, de nombreux progrès techniques ont conduit à la mise au point de méthodes de plus en plus fines pour analyser des populations hétérogènes.

71

Chapitre II. Bases méthodologiques

72

Chapitre II. Bases méthodologiques

1.2. Principe

La cytométrie en flux (CMF) est définie comme l‘étude précise de particules isolées (molécules, cellules, bactéries...) entraînées par un flux liquide. C‘est une technique de caractérisation individuelle, quantitative et qualitative de particules en suspension dans un liquide. Un cytomètre de flux est constitué d‘une composante fluidique, d‘une composante optique et d‘une composante électronique qui sont gérées au moyen d‘une interface informatique (Cram, 2002) (Figure II.7). Le système fluidique permet d‘entraîner la suspension cellulaire à l‘intérieur de la chambre d‘analyse, dans laquelle les cellules traversent une à une le faisceau laser. La chambre d‘analyse comprend :

- une source lumineuse excitatrice : un laser à ions argon qui produit une lumière monochromatique (488 nm) unidirectionnelle et stable. - un ensemble de miroirs dichroïdes récepteurs, qui permettent d‘analyser les rayonnements émis par les cellules.

Après excitation des cellules par le laser, les signaux émis par la particule sont séparés par des filtres optiques et sont collectés par des photo-multiplicateurs (PMT) (Figure II.7). Ainsi, la lumière diffractée mesurée en face du rayon laser (angle < 10°) permet d'évaluer la taille des cellules (Forward Scatter Height, FSC). La lumière diffractée, mesurée à 90° (Size Scatter Height, SSC) donne une mesure de la granulosité et de la complexité de la cellule. Cet ensemble de capteurs spécifiques permet ensuite la mesure de la fluorescence naturelle (auto-fluorescence) ou induite (utilisation de fluorochromes). Les fluorochromes ont la capacité d‘absorber l‘énergie lumineuse à une longueur d‘onde donnée, puis de la réémettre à une longueur d‘onde différente après excitation. Le nombre de détecteurs permettant d‘analyser les différentes plages de longueurs d‘ondes peuvent varier selon le cytomètre. Généralement, un cytomètre possède quatre lentilles permettant de recueillir quatre plages de longueurs d‘ondes différentes : le vert de 500 à 550 nm (détecteur FL1), le jaune de 550 à 600 nm (détecteur FL2), l‘orange et le rouge de 560 à 670 nm (détecteur FL3), et le rouge uniquement de 640 à 700 nm (détecteur FL4) (Figure II.7). Les caractéristiques précises de chacun des cytomètres utilisés (BD FACSCalibur flow cytometer et Guava PCA-96 Flow cytometer) sont répertoriées Figure II.8.

73

Chapitre II. Bases méthodologiques

Figure II.9. Procédure de prélèvement d‘hémolymphe de coque et de palourde.

74

Chapitre II. Bases méthodologiques

L‘ensemble de ces signaux est ensuite amplifié, numérisé, traité, et stocké par un ordinateur. L‘ordinateur calcule les données statistiques associées aux distributions des paramètres mesurés et les représente sous la forme d‘histogrammes (un paramètre) ou de cytogrammes (deux paramètres) sur une ou plusieurs populations dont les propriétés cellulaires sont ainsi évaluées. Ce procédé d‘analyse individuelle (cellule par cellule) est multiparamétrique.

2. Protocoles de mesure des paramètres hémocytaires Les méthodes appliquées ici ont été décrites sur des modèles palourdes et huîtres dans des publications antérieures (Delaporte et al., 2003 ; Hégaret et al., 2003 ; Lambert et al., 2003) et ont été adaptées au modèle de la coque au cours de ce travail.

2.1. Prélèvement d’hémolymphe L‘hémolymphe est prélevée à travers la charnière dans le muscle adducteur postérieur du bivalve à l‘aide d‘une seringue en plastique de 1 mL surmontée d‘une aiguille stérile Terumo 25G (0.50 x 8 mm). Entre 0.5 et 1 mL d‘hémolymphe est prélevé par animal, et une observation au microscope est immédiatement effectuée pour éliminer : - Les échantillons «non conformes », présentant des débris extérieurs (bactéries, particules, phytoplancton, etc…), des débris de tissus, des gamètes, ou une trop grande dilution. - Les individus parasités par le macroparasite Bucephallus minimus qui se reproduit de manière asexuée dans la coque et dont les sporocystes sont facilement détectables dans les tissus et dans l‘hémolymphe du bivalve. Ce parasite se révélant particulièrement destructeur pour la coque, une attention particulière a été menée afin d‘éviter la prise en compte d‘individus parasités par B. minimus lors des diverses expériences en laboratoire. - Les coques présentant des stades néoplasiques avancés. Phénomène qui pourrait, de la même manière, induire des réponses tout à fait particulières des paramètres hémocytaires au regard des effets excessivement délétères de cette maladie. L‘hémolymphe est ensuite rapidement filtrée sur une maille de 80 µm, transférée dans un micro-tube stérile puis conservée dans la glace le temps de réaliser les analyses afin de limiter au maximum l‘agrégation des hémocytes (Figure II.9).

75

Chapitre II. Bases méthodologiques

76

Chapitre II. Bases méthodologiques

2.2. Mesures a. Taille / complexité Dans chacune des analyses présentées ci-dessous, il est possible de distinguer les cellules en fonction de leur taille et de leur complexité à l‘aide d‘un cytogramme FSC vs SSC. Ceci permet d‘une part, d‘éliminer les débris, et d‘autre part, de séparer les deux sous-populations hémocytaires présentes chez la coque et la palourde, les hyalinocytes (faible SSC, fort FSC) et les granulocytes (fort SSC, fort FSC) (Figure II.10-11-12). b. Viabilité Afin de mesurer la viabilité des hémocytes, un aliquot d‘hémolymphe (100 µL) est mis en présence de 200 µL d‘une solution anti-aggrégante (SAAH, Auffret et Oubella, 1994) et 100 µL d‘eau de mer filtrée stérile (EMFS). Les échantillons sont ensuite incubés à l‘obscurité pendant 2h à 18°C en présence de 4 µL de SYBR® green I (utilisé dilué 1/1000 dans du diméthylsulfoxide (DMSO) à partir de la solution commerciale (AMRESCO - Ref J976, Molecular probes, Oregon, USA)) et 4 µL d‘iodure de propidium à une concentration finale de 10 µg.mL-1 (IP, Sigma P4864). Ces deux fluorochromes sont des agents intercalants de l‘ADN. Toutefois, alors que le SYBR Green va s‘intercaler à la fois dans l‘ADN des cellules vivantes et mortes, l‘IP ne pénètre que les cellules ayant perdu leur intégrité membranaire, phénomène caractéristique de la nécrose indiquant donc la mort cellulaire. La fluorescence du SYBR Green est détectée par le détecteur FL1 alors que la fluorescence due à l‘IP est détectée par le détecteur FL3. Le cytogramme FL1 vs FL3 permet ainsi de visualiser le pourcentage de cellules viables dans chaque échantillon (Figure II.10). c. Concentration en hémocytes Lors de l‘analyse de la viabilité hémocytaire, un cytogramme SSC vs FL1 permet de distinguer les hémocytes marqués par le SYBR Green des autres particules présentes dans l‘hémolymphe (Figure II.11). Ce diagramme permet ainsi de calculer la concentration hémocytaire (CH) en prenant en compte le débit du fluide passant dans le capillaire. Ce débit est calculé automatiquement par l‘appareil (Guava) ou manuellement (FACScalibur) et le calcul de

77

Chapitre II. Bases méthodologiques

78

Chapitre II. Bases méthodologiques la concentration hémocytaire (exprimée en cellule / mL) correspond à la formule :

CH = NHT x Tm (sec) x dilution / débit (mL/sec)

NHT : Nombre d‘hémocytes détectés par l‘appareil dans le temps de mesure (Tm)

Tm : Temps de mesure

Ce calcul de la CH s‘effectue aussi sur des échantillons d‘hémolymphe fixée. Pour cela, un aliquot de 100 µL d‘hémolymphe dilué avec 100 µL d‘EMFS est fixé avec 200 µL de formol 6%. Les échantillons sont ensuite incubés avec 4 µL de SYBR Green à l‘obscurité et à température ambiante 30 minutes avant d‘être analysés par cytométrie en flux. Dès lors, le cytogramme SSC vs FL1 permet là aussi de distinguer les hémocytes marqués par le SYBR Green des autres particules présentes dans l‘hémolymphe et le comptage de la CH s‘effectue de la même manière. Cependant, la détermination de la CH sur des échantillons d‘hémolymphe fixée n‘est pas aussi rigoureuse que sur des échantillons frais puisqu‘on estime la perte de cellule à environ 30% de la quantité initiale. d. Capacité d’adhésion Les hémocytes ont naturellement la capacité d‘émettre des pseudopodes leur permettant d‘adhérer sur des surfaces telles que le plastique ou le verre (Bachère et al., 1991; Lambert & Nicolas, 1998). Afin de tester cette capacité d‘adhésion, un aliquot d‘hémolymphe (100µL) est mis en présence de 100 µL d‘EMFS dans un puit de microplaque en polypropylène. Après 3h d‘incubation à 18°C, le surnageant contenant les cellules qui n‘ont pas adhéré est prélevé et fixé avec 200 µL de formol 6%. Les échantillons sont ensuite incubés avec 4 µL de SYBR Green à l‘obscurité et à température ambiante 30 minutes avant d‘être analysés par cytométrie en flux. L‘analyse des résultats ainsi que le calcul de la concentration en hémocytes s‘effectue de la même manière que pour le calcul du CH fixé. Le pourcentage d‘adhésion correspond alors à la formule :

% adhésion = [(NHT – NHS) / NHT ]*100

NHT : Nombre d‘hémocytes présent dans l‘hémolymphe fixée

NHS : Nombre d‘hémocytes présent dans le surnageant récupéré

79

Chapitre II. Bases méthodologiques

e. Activité de phagocytose Un aliquot de 100 µL d‘hémolymphe préalablement dilué avec 200 µL d‘EMFS est mis en contact avec 30 µL d‘une solution de billes fluorescentes en latex (Ø 2 µm) diluée dans de l‘EMFS à 2% de la solution commerciale (Fluoresbrite microspheres YG 2.0 mm, polysciences, Eppelheim, Germany). Les échantillons sont alors incubés à l‘obscurité pendant 2h à 18°C. L‘analyse par cytométrie en flux permet de détecter la fluorescence des billes sur le détecteur FL1 ainsi que le pourcentage d‘hémocytes ayant phagocyté au moins 3 billes (Figure II.12). f. Production d’Espèces Réactives de l’Oxygène (ERO) La production d‘ERO par les hémocytes est estimée en utilisant le 2‘7‘-dichlorofluorescein diacetate (DCFH-DA) qui pénètre dans les cellules sous forme de DCFH et est oxydé par les ERO sous forme de DCF, molécule hautement fluorescente détectée par le capteur FL1 du cytomètre (Figure II.13). L‘intensité de fluorescence FL1 en unité arbitraire est proportionnelle à la production de radicaux libres oxydants. Pratiquement, un aliquot de 100 µL d‘hémolymphe est dilué dans 200 µL d‘EMFS auxquels sont ajoutés 4 µL d‘une solution de DCFH-DA (concentration finale 0.01 mM). Les tubes sont incubés à 18°C à l‘obscurité pendant 2h puis sont analysés au cytomètre. La production d‘ERO est exprimée par la fluorescence relative moyenne des cellules hémocytaires.

80

Chapitre II. Bases méthodologiques

E. Techniques de biologie moléculaire

Les analyses réalisées en biologie moléculaire ont pour but d'étudier l'effet de la contamination métallique et/ou des infestations parasitaires sur le niveau d'expression de certains gènes par PCR quantitative, autrement dit d'étudier la production différentielle d'ARN messager (ARNm) entre des individus témoins et des individus contaminés/parasités à un instant t. Une des conditions préalables indispensables à l'étude du niveau d'expression des gènes par PCR quantitative est la connaissance d'au moins une partie de la séquence codante du gène étudié. Un travail important de recherche et de séquençage des séquences nucléotidiques d'intérêt a donc dû être mené en début de thèse, les génomes de la coque et de la palourde demeurant à l'heure actuelle très peu connus.

1. Recherche et séquençage des régions codantes d'intérêt

La séquence de certains gènes d‘intérêt chez la coque et la palourde n‘ayant pas été précédemment identifiée, il a été nécessaire de rechercher les séquences codantes d'intérêt dans les banques de données ("GenBank" via pubmed-NCBI) chez des espèces de mollusques pour lesquelles cette séquence est connue. Les différentes séquences rapatriées sont alors alignées à l'aide du logiciel Custalw (Infobiogen), qui permet de mettre en évidence les régions codantes du gène les plus conservées entre les espèces, c'est-à-dire celles qui ont les plus fortes probabilités d'être retrouvées identiques chez la coque et la palourde. Des couples d'amorces sont alors déterminés dans ces régions « conservées » et synthétisés (Sigma-proligo®). Une fois cette démarche effectuée, le principe général de clonage et séquençage des régions codantes d‘intérêt fait appel à différentes techniques de biologie moléculaire (Figure II.14 - Etapes 1 à 6). Cette démarche est présentée ci-dessous, puis les différentes étapes seront développées dans un second temps, en respectant leur chronologie d'utilisation. La première étape consiste à extraire les ARN totaux (1) présents dans différents tissus de bivalve et les rétro-transcrire en ADNc (2). Les ADNc sont des molécules d'ADN simple brin qui sont les copies des ARNm. Ces molécules ne contiennent que la région codante du gène et sont donc totalement dépourvues d'introns. C‘est sur ces ADNc que l'amplification des régions codantes d'intérêt est réalisée par PCR classique (3) à l‘aide des couples d‘amorces

81

Chapitre II. Bases méthodologiques

Figure II.14. Démarche expérimentale pour la recherche et le séquençage d‘un gène d‘intérêt. 82

Chapitre II. Bases méthodologiques précédemment déterminés. Une électrophorèse en gel d'agarose (4) est alors réalisée avec les produits issus de la PCR afin de les visualiser et de les purifier. Un marqueur de poids moléculaire est utilisé lors de l‘électrophorèse afin d‘estimer la taille des fragments amplifiés, et de les comparer à la taille « attendue », c'est-à-dire la taille des fragments déjà séquencés chez d‘autres espèces et répertoriés dans la Genbank. En effet, puisque le couple d‘amorces est déterminé dans une région relativement conservée du gène, des tailles similaires entre les différentes espèces de mollusque sont attendues pour le gène étudié. Les fragments présentant la taille attendue sont alors découpés du gel sous table UV (Fisher Bioblock Scientific), purifiés (5) et utilisés pour l'étape suivante : le clonage et la transformation bactérienne (6) qui vont permettre d‘isoler le fragment d‘intérêt en l‘insérant dans un vecteur (plasmide) capable de se multiplier dans des bactéries. Le séquençage du fragment d‘ADN est réalisé à partir d'un clone bactérien contenant l'insert à la taille attendue. Dans notre cas, le séquençage a été sous-traité auprès de l'entreprise Millegen biotechnologies.

1.1. Purifications des acides ribonucléiques (ARN) totaux L‘extraction des ARN requiert certaines précautions de manipulation afin d‘éviter toute contamination de l‘échantillon par des ribonucléases (RNAse). Dès lors, toutes les manipulations sont effectuées avec des gants et l'eau utilisée pour la préparation des réactifs et pendant toutes les étapes d'extraction et d'analyse de l'ARN, est préalablement traitée contre les ribonucléases par le diéthyl pyrocarbonate (DEPC) 0,1% (v/v) pendant 12h à 37°C, puis autoclavée. De même, la paillasse doit être décontaminée (RNAse away, Molecular BioProducts) et l‘ensemble du matériel utilisé (tubes, potters…) lors de l‘extraction des ARN doit être certifié « RNAse free » par le fabricant ou décontaminé puis autoclavé.

Les ARN de bivalves ont été extraits à l‘aide de kit « Absolutely RNA RT-PCR miniprep kit » (Agilent) selon les instructions du fournisseur. En premier lieu, 20 à 40 mg de tissu sont broyés à l‘aide d‘un potter en polypropylène dans 600 µL de tampon de lyse contenant du thiocynate de guanidine auquel est ajouté 4,2 µL de β-mercaptoéthanol. Cette solution exerce aussi une activité inhibitrice des endonucléases présentes naturellement dans les tissus. Afin d‘éliminer les principaux débris cellulaires, le broyat est déposé sur une colonne de préfiltration et centrifugé 5 min à 14 500 rpm à température ambiante. Le filtrat, additionné à un volume de phénol : choloroforme : alcool isoamylique (25:24:1), est vortexé rapidement puis centrifugé

83

Chapitre II. Bases méthodologiques pendant 10 min à 14 500 rpm. Cette étape a pour but de déprotéiniser et délipidiser les filtrats. La phase aqueuse contenant les ARN est prélevée puis additionnée à un volume d‘éthanol 75%, l‘ensemble est vortexé et déposé sur une colonne d‘affinité ne retenant que les acides nucléiques (ADN et ARN). Les colonnes sont centrifugées 60 sec à vitesse maximale puis le filtrat est éliminé. Un premier lavage est effectué par ajout de 600 µL d‘une solution de lavage faiblement concentrée en sels, suivi d‘une centrifugation de 60 sec à vitesse maximale. Les ADN fixés à la colonne sont dégradés par incubation avec 55 µL de solution de DNAse I (Tampon d‘activité + 5 unités de DNAse I) pendant 15 min à 37°C. Les colonnes subissent alors deux lavages, un premier avec 600 µL d‘une solution fortement chargée en sels ayant pour but d‘éliminer la DNAse et les fragments d‘ADN, le second avec 300 µL de tampon faiblement chargés en sels. Entre chaque lavage une centrifugation de 60 sec à vitesse maximale est appliquée. La colonne est séchée par une centrifugation de 2 min puis transférée dans un tube neuf et finalement incubée 2 min à température ambiante avec 30 µL de tampon d‘élution. Les ARNs purifiés sont récupérés par centrifugation (60 sec, vitesse max.) et aliquotés à – 80°C avant d‘être utilisés lors des réactions de rétro-transcription si cette dernière n‘est pas effectuée directement.

1.2. Rétro-transcription La rétro-transcription a pour objectif de transformer les ARN en ADN complémentaires (ADNc), plus stables et indispensables pour effectuer les étapes de PCR ultérieures. La rétro- transcription est réalisée à l‘aide du kit « Stratascript first strand synthesis system » (Agilent) selon les recommandations du fournisseur. Brièvement, un mélange de tampon d‘activité 10X (2 µL), de dNTP (1 µL – 100 mM) et d‘amorces (1µL oligodT – 0,5 µg.µL-1 + 1 µL hexaprimers – 0,1 µg.µL-1) sont additionnés à 14 µL d‘ARN (≈ 5 µg) totaux puis l‘ensemble est incubé 5 min à 65°C afin de linéariser les ARN. Le mélange se refroidit progressivement jusqu‘à 42°C ce qui permet aux amorces de se fixer sur les ARN. Enfin 0,5 µL de RNAse block (40 U.µL-1) et 1 µL de Reverse-transcriptase (RT) sont ajoutés, l‘ensemble est homogénéisé puis incubé 1 heure à 42°C. Cette dernière étape permet la rétro-transcription des ARN en ADNc sous l‘action de la RT, et les ADNc ainsi obtenus seront conservés à -20°C jusqu‘aux étapes de PCR.

1.3. PCR classique La PCR classique permet d‘amplifier facilement et de manière importante un fragment d‘ADN d‘intérêt, même présent en très faible quantité dans l‘échantillon. Cette technique fait

84

Chapitre II. Bases méthodologiques appel à la Taq polymérase, une ADN polymérase provenant de la bactérie Thermus aquaticus vivant dans les sources d‘eau chaude. Cette ADN polymérase va permettre de synthétiser de façon fidèle un brin complémentaire à partir d‘une matrice ADN présente dans l‘échantillon, et ce de manière importante puisque cette enzyme est capable de polymériser plus de 1000 pb par minute. Les réactions de PCR ont été réalisées à l'aide de la Taq DNA polymerase (Promega) en respectant les recommandations du fournisseur. Brièvement, 2 μL d'ADNc sont mis en présence d‘un mélange contenant : - 0,5 μL de chaque amorce à 100 μM - 1 μL de dNTP à 10 mM

- 3 μL de MgCl2 à 25 mM (cofacteur de la Taq polymérase) - 5 μL de tampon d‘activité 10X - 0,2 μL de Taq polymérase à 5 u.μL-1 - 37,8 μL d'eau

Ce mélange est placé dans un thermocycleur (Mastercycler, Eppendorf) qui va réaliser les cycles de PCR décrits ci-dessous. Un cycle de PCR classique comporte 3 étapes successives : - La dénaturation de l‘ADN : séparation des deux brins complémentaires par une élévation de la température à 95°C pendant 30 secondes à 1 minute. - L‘hybridation : fixation des amorces au niveau des régions spécifiques, encadrant le segment à amplifier. La durée de cette étape est de 30 sec et la température est définie au cas par cas car elle dépend du Tm (température de fusion) du couple d‘amorces utilisé (comprise généralement entre 42 et 65°C). - L‘élongation : synthèse du brin complémentaire par la Taq polymérase pendant 1 minute à 72°C. Ce cycle est réitéré entre 30 et 50 fois, et en fin de réaction une étape de 10 min à 72°C va permettre à l‘enzyme de finaliser les élongations des brins d‘ADNc. Les amplifiats ainsi obtenus par PCR sont ensuite visualisés par électrophorèse.

1.4. Electrophorèse Afin de visualiser les amplifiats de PCR, des gels d'agarose à 1% (p/v) ont été réalisés par dissolution d‘une poudre d‘agarose dans du tampon TAE 1X (Tris base à 40 mM + acide acétique

85

Chapitre II. Bases méthodologiques

à 5,7 % + Na2EDTA.2H2O à 2 mM) par chauffage au micro-onde. Après quelques minutes, 2 µL de bromure d'éthidium (BET, 10 mg.mL-1) sont incorporés dans le gel avant sa solidification. Le BET est un agent interacalant de l‘ADN qui présente la propriété de fluorescer sous UV. Une fois le gel solidifié, il est plongé dans une cuve électrophorétique contenant un bain de TAE 1X. 20 µL d‘amplifiat de PCR sont prélevés, mélangés à 5 µL de tampon de charge et déposés dans un puit. Un champ électrique de 100 volts est alors appliqué pendant 20 à 40 minutes, puis le gel est retiré et placé sous une lampe à UV. Sous cette lumière, le BET va fluorescer et va permettre de révéler les différents fragments d‘ADN.

1.5. Purification des fragments d'intérêt La purification des fragments d'intérêt peut se faire à partir de l‘amplifiat de PCR à l‘aide du kit « QIAquick PCR amplification extraction kit » (Qiagen) ou à partir des morceaux de gel renfermant les ADN à l'aide du kit "QIAquick gel extraction kit" (Qiagen), selon les recommandations du fournisseur. La procédure de ces deux types de purification est similaire, seules les premières étapes diffèrent en raison de la nature de l‘échantillon contenant les fragments d‘intérêt. Un fragment d'agarose renfermant les ADN d'intérêt est découpé au scalpel sous UV puis pesé, et dissous par traitement thermique (50°C) dans 3 volumes de tampon QG (3 μL de QG pour 1 mg de gel). Un volume d'isopropanol est ajouté et l‘ensemble est vortexé puis déposé sur une colonne d'affinité retenant les acides nucléiques. Après centrifugation pendant 60 sec à 14 500 rpm, la colonne subit deux lavages successifs, le premier avec 500 μL de tampon QG puis le deuxième (*) avec 750 μL de tampon PE. Entre chaque lavage, le filtrat est jeté et une centrifugation de 60 sec à 14 500 rpm est réalisée. Finalement, 30 μL de tampon d'élution (tampon EB) sont déposés sur la colonne placée dans un nouveau tube de 1,5 mL. Après 2 min d‘incubation à température ambiante, les fragments sont récupérés par centrifugation (60 sec à 14 500 rpm). Lorsque la purification se fait directement sur l‘amplifiat PCR, cinq volumes de tampon PB sont ajoutés à ce mélange réactionnel et l‘ensemble est déposé sur la colonne d‘affinité retenant les acides nucléiques puis centrifugé 60 secondes à 14 500 rpm. Le reste de la purification se déroule ensuite de manière identique au protocole énoncé ci-dessus à partir de l‘étape (*): un lavage avec 750 µL de tampon PE, une étape de séchage de la colonne par centrifugation (60 sec

86

Chapitre II. Bases méthodologiques

à vitesse maximale) puis dépôt du tampon d‘élution EB (30 µL) et récupération des fragments purifiés.

1.6. Clonage et transformation bactérienne Le principe du clonage bactérien est d‘isoler un fragment d‘ADN en l‘insérant dans un vecteur capable de se multiplier dans les bactéries. Les vecteurs utilisés dans le clonage bactérien peuvent être des plasmides, comme c‘est le cas dans ces travaux, des bactériophages ou des cosmides. L‘insertion dans le vecteur est réalisée à l‘aide d‘enzymes de restriction qui permettent la linéarisation du vecteur (le site de restriction doit être unique dans le vecteur). a. Insertion des fragments dans le vecteur

Le plasmide pGEMT, "pGEM®-T easy vector systems", Promega®), a été utilisé en raison de sa capacité à confèrer aux bactéries transformées la résistance à un antibiotique : l'ampicilline

(gène Ampr) et la présence du gène de la ß-galactosidase dont l‘utilité est décrite plus bas. Le pGEMT initialement linéaire, présente à chaque extrémité une base azotée T libre, complémentaire à la base azotée A libre ajoutée à chaque extrémité 3' du fragment amplifié par la Taq polymérase pendant la réaction de PCR. Les liaisons phosphodiester entre le fragment à cloner, appelé insert, et le vecteur sont créées à l‘aide d‘une enzyme particulière, la ligase (réaction de ligation). On utilise généralement la T4 DNA ligase issue du bactériophage T4. -1 Brièvement, 1 μL de plasmide (50 ng.mL ) sont mélangés avec 3 μL de fragment purifié, 1 μL de ligase (3 U.μL-1) et 5 μL de tampon 2X. Le mélange réactionnel est incubé à température ambiante 1 h puis une nuit à 4°C. Le clonage ne présentant pas une efficacité de 100%, le mélange de ligation obtenu va donc à la fois contenir des plasmides ayant intégrer le fragment d‘intêret avec succès ainsi que des plasmides vides sans insert. b. Transformation des bactéries Le mélange de ligation est mis en présence de bactéries de façon à ce que le vecteur avec éventuellement l‘insert pénètrent dans la bactérie pour s‘y multiplier : c‘est la transformation bactérienne. Les bactéries Escherichia coli thermo-compétentes "JM 109 competent cells, high efficiency", Promega® sont utilisées en raison de leur capacité à incorporer des vecteurs par choc thermique et leur relativement bonne résistance. Les bactéries préalablement stockées à – 80°C

87

Chapitre II. Bases méthodologiques sont lentement décongelées sur glace puis homogénéisées avec précaution. Un aliquot de 2 μL du mélange de ligation est mélangé à 50 μL de bactéries et l‘ensemble est placé dans la glace pendant 20 min. Par la suite, le mélange est soumis à un choc thermique (42°C) pendant 45 s puis replacé immédiatement dans la glace pendant 2 min. A ce mélange est ajouté 950 μL de milieu de culture liquide LB (milieu LB, Luria Bertani, Sambrook et al., 1989) et l‘ensemble est incubé 1 h à 37°C, ce qui va permettre aux bactéries d‘exprimer le plasmide et ainsi acquérir la résistance à l'ampicilline (étape d' « expression phénotypique »). Finalement, 200 μL de cette suspension bactérienne sont étalés sur milieu LB solide contenant de l'ampicilline (LB-Amp - 100 µg.mL-1). La ligation puis ensuite la transformation bactérienne ne sont pas efficaces à 100%, c‘est à dire que des vecteurs peuvent ne pas posséder d‘insert après ligation et des bactéries peuvent ne pas avoir reçu de vecteurs après transformation. Il faut donc pouvoir distinguer ces différentes espèces. Cette distinction, appelée sélection, va être obtenue à l‘aide de gènes particuliers apportés par le vecteur. Le vecteur pGEMT possède un gène de résistance à un antibiotique, l‘ampicilline, qui va permettre de distinguer les bactéries qui ont reçu le vecteur (devenues résistantes à l‘ampicilline) des bactéries qui n‘ont pas reçu le vecteur et qui sont naturellement sensibles à cet antibiotique. Sur le milieu solide contenant de l‘antibiotique, seules les bactéries ayant incorporé et exprimé le pGEMT vont donc parvenir à croître et former une colonie. Par ailleurs, le site d‘insertion de l‘insert est situé à l‘intérieur du gène de la ß-galactosidase. L‘insertion va donc couper ce gène qui ne fonctionnera plus. Il est donc possible de distinguer les vecteurs qui ont inséré un fragment d‘ADN de ceux qui n‘ont rien inséré mais se sont re-circularisés à l‘aide de la ligase, simplement en testant le fonctionnement de la ß-galactosidase. Ceci est réalisé en additionnant le milieu de culture des bactéries d‘un substrat de l‘enzyme (Xgal - 20 µg.mL-1) qui, lorsque celle-ci fonctionne, est transformé en un produit qui colore les colonies bactériennes en bleu. Dès lors les colonies ayant incorporé un plasmide vide seront colorées en bleue, au contraire des colonies ayant incorporé un plasmide avec insert qui resteront blanchâtres. Après une nuit d'incubation à 37°C, un minimum de 10 colonies blanchâtres sont isolément repiquées dans 10 tubes remplis avec 3 mL de milieu LB-Amp liquide, puis incubées une nuit à 37°C sous agitation. La vérification de la présence de l‘insert dans le plasmide incorporé par les

88

Chapitre II. Bases méthodologiques bactéries sélectionnées se fait par une étape de PCR à l'aide d'amorces universelles présentes sur le plasmide (T7 et Sp6) puis révélation sur gel d‘électrophorèse en fonction de la taille attendue. Une fois les « bonnes » colonies repérées , c'est-à-dire celles contenant le plasmide avec l‘insert d‘intérêt, un aliquot de 1 mL de la suspension bactérienne correspondante est fixé dans 100 µL de glycérol puis envoyé à une entreprise privée chargée du séquençage du fragment d‘intérêt (MillGen).

2. PCR quantitative en temps réel

L‘utilisation de la PCR quantitative nous a permis d‘estimer l‘expression basale de plusieurs gènes d‘intérêt chez nos modèles biologiques ainsi que l‘impact de contamination métallique ou d‘infestation micro et macro parasitaire sur les niveaux d‘expression de ces derniers. De plus, la PCR quantitative permet, à l‘instar de la PCR classique, la détection et la quantification des produits amplifiés en temps réel en cours de réaction (Figure II.15). Pour cela, le SYBR Green I, fluorochrome ayant la propriété de s‘intercaler entre les bases d'une séquence nucléotidique double brin et de ne fluorescer qu‘uniquement dans cette condition, est utilisé lors des réactions PCR. Le SYBR Green émet une fluorescence dans le vert (530 nm) et l‘intensité de la lumière émise est directement proportionnelle au nombre de copies du fragment amplifié. Les PCR quantitatives ont été réalisées à l‘aide du LightCycler (Roche Molecular Biochemicals). Ce thermocycleur est équipé d'un spectrophotomètre qui mesure la fluorescence émise par chaque échantillon à l'issue de chaque étape d'élongation. Comme pour la PCR classique, les ARNm sont purifiés puis rétro-transcrits en ADNc. Les amorces, qui dans le cas présent sont rigoureusement complémentaires à la région étudiée, ont été définies à l'aide du logiciel LightCycler Probe design (version 1.0, Roche). Les réactions de PCR quantitative sont réalisées dans des capillaires en verre en présence de : - 1 μL de tampon 1b activé (tampon d'activité + SYBR Green + Taq polymérase + dNTPs),

- 3,2 μL de MgCl2 à 25 mM, - 2 μL d'amorces (sens et anti-sens) à 3 μM, - 1 μL d'ADNc - 12,8 μL d'eau ultra-pure.

89

Chapitre II. Bases méthodologiques

-

90

Chapitre II. Bases méthodologiques

Le programme thermique commence par une étape préalable d'activation de l'enzyme (contenue dans des microbilles de glycérol) de 10 min à 95°C, suivie de la répétition de 50 cycles d'amplification (95 °C, 5 s ; 60 °C, 5 s et 72 °C, 20 s). En fin de réaction, la spécificité de la PCR est analysée par réalisation d‘une courbe de fusion. En effet, chaque produit d‘amplification est caractérisé par une température de fusion (Tm) qui dépend de sa composition en bases G et C et qui lui est donc propre. Le lightcycler lance un cycle de température allant progressivement de 60 à 95 °C par pas de 0,5 °C par seconde et un suivi de continu de la fluorescence. Les Tm sont déterminés par analyse de la perte de la fluorescence lorsque les deux brins se séparent.

Au cours de nos différentes études, la quantification a été relative. En plus des gènes étudiés, un contrôle endogène était systématiquement amplifié. Dans notre cas, le gène de référence choisi est le gène codant pour la β-actine. Pour chaque échantillon analysé, le nombre de copies d'ADNc du gène d'intérêt a d'abord été normalisé par rapport au nombre de copies d'ADNc codant pour la β-actine avant d'être utilisé pour comparer l'expression des gènes d'intérêt chez les individus.

Les facteurs d‘induction et de répression des gènes ont été déterminés par la méthode 2-ΔCT décrite par Livak et Schmittgen (2001) où ΔCT est la différence entre le cycle de sortie du gène d‘intérêt et le cycle de sortie du gène de référence. Pour cela, pour chaque gène étudié, la valeur moyenne de l‘expression du gène déterminée chez les individus contaminés a été comparée avec celle observée chez les témoins pour ce même gène.

91

Chapitre II. Bases méthodologiques

5’ RT-RACE 3’ RT-RACE

5’ ARNm 3’ 5’ ARNm 3’ AAAAA AAAAA

+ -XXXX + TTTTT + TTTTT- SMARTer II A 5’CDS Primer A 3’ CDS Primer A (Oligo-dT)

+ Transcriptase inverse

5’ ADNc 3’ 5’ ARNm 3’ -XXXX AAAAA AAAAA TTTTT TTTTT- -XXXX 5’ 3’ 3’ 5’

Figure II.16a. Synthèse des 5‘-ADNc et 3‘-ADNc utilisés lors des réactions de 5‘RACE et 3‘RACE.

5’ RACE 3’ RACE

TTTTT- -XXXX TTTTT 5’ 3’ 3’ 5’ 5’ 3’ + UPM 10X 3’ 5’ + UPM 10X + Amorce (Universal Primer Mix) + Amorce spécifique spécifique sens anti-sens

+ Advantage 2 Polymerase Mix + Advantage 2 Polymerase Mix

3’ 5’ -XXXX TTTTT 5’ TTTTT- 3’ UPM--XXXX AAAAA--UPM

AAAAA--UPM UPM--XXXX AAAAA--UPM UPM--XXXX

Fragment 5’RACE double brin Fragment 3’RACE double brin

Figure II.16b. Principe général de la 5‘RACE et 3‘RACE

92

Chapitre II. Bases méthodologiques

3. Cas particulier du gène de métallothionéine de coque Cemt1

Lors de ces travaux de thèse, une attention particulière a été portée sur le gène de métallothionéine de la coque Cerastoderma edule. Ainsi, la détermination d‘une partie de la région codante de ce gène, indispensable pour les analyses de son expression par PCR quantitative, a été complétée par la détermination de la séquence nucléotidique des régions 3' et 5' non codantes, de manière à caractériser la totalité de la séquence codante du gène.

Par la suite, il est apparu intéressant de caractériser l‘organisation de ce gène au niveau génomique, avec notamment l‘organisation des séquences exoniques et des introniques. Pour cela, des amorces correspondantes aux régions 3‘ et 5‘ non codantes ont été désignés, et une PCR classique a été réalisée sur l‘ADN génomique de coque précédemment extrait.

L‘ensemble de ces techniques est décrite ci-dessous :

3.1. Amplification des extrémités non codantes

La technique utilisée pour amplifier les régions non codantes d‘un gène, appelée respectivement 3' ou 5' RACE (Rapid Amplification of cDNA Extremities), a été réalisée à l'aide du Kit "BD Smart Race cDNA amplification" (Clontech) en suivant les recommandations du fournisseur (Figure II.16).

La première étape assure la rétrotranscription des ARNm en ADNc mais aussi l'adjonction d'adaptateurs (séquences oligonucléotidiques universelles) en région 3' ou 5' des ADNc synthétisés :

- Dans le cadre d‘une 3‘ RACE (région 3' recherchée), l‘adaptateur (3'- race CDS primer A) est fixé à l'extrémité 5‘ de l'oligo-dT.

- Dans le cadre d‘une 5‘ RACE (région 5' recherchée), l‘adaptateur (5'-race CDS primer) contient des résidus G terminaux capables de s‘hybrider à des résidus C. La fixation de l'adaptateur est alors rendue possible par l'utilisation d'une enzyme particulière, la "BD PowerScript Reverse Transcriptase". Cette dernière présente en plus de son activité de rétro-transcription, une activité terminale transférase qui lui confère la propriété d'ajouter des résidus C (entre 3 et 5) à l'extrémité 3' de l'ADNc synthétisé. La transcriptase-inverse assure la formation de la liaison entre l'adaptateur et l'ADNc et l'élongation de l'ADN

93

Chapitre II. Bases méthodologiques

double brin contenant alors, dans sa région 3', la séquence complémentaire à celle de l'adaptateur.

Afin d‘amplifier ces régions avec succès, le mélange réactionnel des réactions de PCR contient alors des amorces dont la séquence est identique à celle des adaptateurs ainsi que des amorces complémentaires, soit sens (3‘ RACE) soit anti-sens (5‘ RACE) de la région recherchée.

Brièvement, 3 μL d'ARNm sont mélangés avec 1 μL de 3'-race CDS primer A (oligo-dT- adaptateur, 12 μM) et 1 μL d'eau ultra-pure ou avec 1 μL de 5'-race CDS (oligo-dT, 12 μM) et 1μL de BD Smart II A oligo (adaptateur-GGG, 12 μM) pour réaliser une 3' ou une 5' RACE, respectivement. Afin de linéariser les ARNm et surtout permettre l'hybridation des amorces, le mélange réactionnel est incubé 2 min à 70°C. Le mélange est ensuite placé 2 min dans la glace puis centrifugé, ces deux manipulations ayant pour objectif de concentrer les produits au fond du tube. Le contenu du fond du tube est alors prélevé à l'aide d'une pipette et homogénéisé avec 2 μL de tampon (5X First Strand Buffer), 1 μL dithiothreitol (DTT, 20 mM), 1 μL de dNTP (10 mM) et 1 μL de "BD PowerScript Reverse Transcriptase". Le mélange réactionnel ainsi obtenu est placé 1 h 30 à 42°C puis 7 min à 72°C. Les ADNc obtenus sont ensuite utilisés pour les réactions de PCR. Ces réactions se déroulent dans un volume réactionnel de 50 μL associant 2,5 μL d'ADNc, 5 μL d'Universal Primer Mix (amorces universelles, 1,2 μM), 1 μL d'amorce (10 μM) anti-sens (5' race) ou sens (3' race) spécifique de la région recherchée ainsi que 41,5 μL de master mix. Ce dernier est obtenu par le mélange de 34,5 μL d'eau ultra-pure, 5 μL de tampon (10X BD Advantage 2 PCR Buffer), 1 μL de dNTPs (10 mM) et 5 μL de Polymérase (50X BD Advantage 2 Polymerase Mix, 10 mM). Les échantillons sont ensuite placés au thermocycleur (94°C, 30 s; 68°C, 30 s; 72°C, 3 min; n = 30 cycles). Les fragments amplifiés ont ensuite été visualisés par électrophorèse, purifiés et clonés comme décrit précédemment.

94

Chapitre II. Bases méthodologiques

3.2. Extraction de l’ADN génomique Il existe différents protocoles pour extraire l'ADN, qui suivent approximativement le même schéma de principe (Figure II.17): - Lyse des cellules - Elimination des protéines - Elimination des autres acides nucléiques (ARN...) - Concentration de l'ADN par précipitation à l'alcool

Les tissus sont incubés 1h à 55°C dans 800 µL d‘un tampon de lyse : 10 mM Tris pH 8, 100 mM EDTA pH 8, 0.5 % SDS, additionnés à 10 µL de protéinase K (10 mg.mL-1) et 2 µL de RNAse A (Qiagen, 100 mg.mL-1) de manière à lyser les cellules. Cette méthode permet de libérer le contenu cellulaire sans casser l‘ADN. La présence de détergent va disperser les bicouches lipidiques des membranes et dénaturer les protéines, et en particulier celles qui sont associées à l'ADN dans la chromatine. La solution obtenue est en général très visqueuse, car l'ADN ainsi libéré forme de très longs filaments qui s'opposent aux écoulements hydrodynamiques. Une solution de haute force ionique (500 µL de NaCl 5M) est ensuite rajoutée de manière à faire précipiter certaines protéines, et l‘ensemble est centrifugé pendant 10 min à 14 000 g. Le surnageant est récupéré afin d‘effectuer l‘étape suivante de déprotéinisation. Cette étape se fait au moyen de solvants organiques, en général du phénol additionné de plus ou moins de chloroforme. Un volume de phénol est ajouté au surnageant, l‘ensemble est vortexé puis centrifugé 10 minutes à 5000 g. L'ADN reste en solution dans la phase aqueuse qui est récupérée par pipetage. Cette étape est répétée deux fois. Sur le même principe, une troisième extraction au phénol : chloroforme : alcool isoamylique (25:24:1) est réalisée, le surnageant est récupéré pour l‘étape finale de précipitation de l‘ADN. Pour cela, deux volumes et demi d‘éthanol absolu sont additionnés, l‘ensemble est agité tout doucement, ce qui va entraîner la précipitation d‘ADN sous forme de pelote clairement visible à l‘œil nu. La pelote d‘ADN est récupérée avec précaution et subit deux lavages successifs dans 1 mL d‘éthanol absolu puis 1 mL d‘éthanol 70%. Chaque lavage est séparé par 5 minutes de centrifugation à 15 000 g. Finalement, la pelote est séchée (5 min à 50°C ou 15 min à température ambiante) puis resuspendue dans 200 à 300 µL d‘eau ultra- pure. La concentration en ADN est déterminée par spectrophotométrie avec une mesure de DO à 260 (1 unité de DO correspond à une concentration de 50 µg.mL-1). La solution est ensuite conservée à -20°C jusqu‘aux étapes de PCR.

95

Chapitre II. Bases méthodologiques

Tissu dans le tampon de digestion

Incubation sur 1h à 55 C

Lysat

Extraction au phénol chloroforme

Extraction au chloroforme

Précipitation éthanol

Pelote d’ADN

Lavages - séchage

Pelote d’ADN rincée et séchée

ADN remis en suspension (Tris EDTA) Solution d’ADN à 8 ng.µL-1 Quantification / dilution

Figure II.17. Protocole d‘extraction de l‘ADN à partir d‘un tissu biologique.

96

Chapitre III. Approche multistress in situ

Bivalve population health : multistress to identify hot spots

Article in preparation

Images satellite du site de Landéda et du bassin d‘Arcachon (Source Google Earth)

97

98

Chapitre III. Approche multistress in situ

Introduction

Il existe une carence d‘informations concernant les bivalves fouisseurs (type palourde ou coque) en ce qui concerne les niveaux de contamination métallique, de charge parasitaire, de pathologie. Le suivi mené pendant deux ans sur quatre populations présentes dans des environnements différents, à Arcachon (Aquitaine) et à Landéda (Bretagne), a été réalisé afin d‘obtenir un ensemble de données sur les niveaux des paramètres immunologiques, génétiques, écotoxicologiques, parasitaires, microbiologiques, physiologiques et écologiques (dynamique des populations). Le choix des sites a évidemment été conditionné par la présence des espèces, mais aussi par les caractéristiques permettant de tracer des histoires de vie variées. L‘objectif de ce travail a donc été de comparer les caractéristiques de quatre populations (palourdes et coques) afin de comprendre comment ces espèces composent avec les stress présents. Ce suivi a également été réalisé dans le cadre d‘un projet (Multistress) financé par l‘Agence Nationale pour la Recherche. Il s‘agissait d‘avoir une vue fine des niveaux de stress et de réponse des bivalves à l‘échelle saisonnière mais aussi à l‘échelle de l‘individu. En effet, la plupart des paramètres étaient mesurés sur chaque individu. De cette base de données à laquelle j‘ai pleinement contribué, j‘ai extrait pour la thèse les valeurs médianes et moyennes sur les deux années. Ainsi, ce chapitre présente une analyse globale, permettant d‘avoir une idée des grandes similitudes et différences entre les quatre situations étudiées en gommant la variabilité temporelle, mais aussi d‘appréhender l‘état de santé des populations et de le relier aux stress subis ainsi qu‘aux réponses développées par les individus. Ce travail est présenté sous forme d‘un article en préparation.

99

Chapitre III. Approche multistress in situ

100

Chapitre III. Approche multistress in situ

1. Introduction

The contamination of natural waters by heavy metals and microbial pathogens as a result of human activities affects aquatic life and poses a high environmental concern (Byrne and O‘Halloran, 2001). Thus, situations where ecosystems are confronted to solely one specific problem leading to the use of only one specific biomarker to answer is scarce or even inexistent. Generally, ecosystems are faced with multifactorial causes of disturbances: in this case a panel of complementary and ecologically relevant biomarkers would be necessary to assess and manage environmental quality.

The responses of estuarine and marine organisms to waste input are manifold, but can mainly be classified on five levels of biological organization, i.e. molecular, cellular, organismal, population and community basis (Verlecar et al., 2006). The present study covered the scale from the molecule to the population and aims to assess the correlation between environmental stress, bivalve response and overall population performances and to appraise environmental quality. Three categories of stressors were measured - metals (Cd, Cu, Zn), microparasites (Perkinsus olseni) or disease signs due or hypothesised to be due to microparasites (BRD: Brown Ring Diseases; BMD: Brown Muscle Disease) and macroparasites (digenean trematodes) - in two bivalve species, the edible cockle (Cerastoderma edule) and the Manila clam (Ruditapes philippinarum), and in three Atlantic French sites (Inner Arcachon bay (Andernos), Outer Arcachon bay (Arguin) and Landéda coast). The three metals were selected because our preliminary analyses reported significant concentrations in bivalves of one of our sites (Andernos, Inner Arcachon Bay). Perkinsosis, among molluscan diseases, has resulted in the most severe economic losses (Carnegie, 2005; Villalba et al., 2004). Perkinsus olseni occurs in Manila clams along the whole French coast (Lassalle et al., 2007). Manila clam culture in Western Europe has suffered a serious vibriosis problem since the late 1980s. The etiological agent was identified as a Vibrio sp., named P1 (Paillard and Maes, 1990) and later described as Vibrio tapetis (Borrego et al., 1996). Recently, the Brown Muscle Disease was described at Arcachon with local high prevalence and was thought to be due to viral particles (Dang and de Montaudouin, 2009). Digenean parasites are omnipresent in cockles (de Montaudouin et al., 2009) and may cause loss of production (Desclaux et al.,, 2004; Gam et al., 2009) while they remain scarce in Manila clams (Lassalle et al., 2007; Dang et al., 2009a). In common with many invertebrates, host defence in bivalves is largely non-specific, based on activities of circulating

101

Chapitre III. Approche multistress in situ hemocytes in the soft tissues and in extrapallial fluid (Allam et al., 2000; Canesi et al., 2002; Cheng, 1996; Chu, 2000). Metallothioneins (MT) responds to trace metals (Amiard et al., 2006; Baudrimont et al., 2003; Kägi, 1991) but can also be induced by other environmental tensions triggering inflammation and oxidative stress (Baudrimont et al., 1997b, 2006). Upstream, the expression of different genes involved in oxidative stress response like superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPX) is thought to contribute to detoxication as well as others contribute to functioning of the innate immunological system (Gestal et al., 2008).

Different scenarii are possible that can be summed up within two alternatives (Vasseur and Cossu-Leguille, 2003). 1) Sources of stress are considered as negligible, no particular adaptative response is attended and the population parameters (reproduction, growth) should be normal. If the population dynamics is weak, other regulation factors intervene. 2) Some stressors are identified at a level that is considered as high leading to adaptative physiological response with two alternatives: - short-term stress leading to rapid response of the organism and affecting bivalve physiology without impairing population dynamics parameters; - long-term stress affecting fitness of the organism. The capability (or not) for bivalves to overcome this situation will result in a more or less efficient population dynamics.

The strategy of the present work has been to monitor seasonally (four times a year, during two years) four populations (clams and cockles at Arcachon (Arguin, Andernos) and Landéda) that could illustrate these different scenarii. Seasons, and life-cycle of bivalves in particular, largely interact in the expression of biomarkers that are also influenced not only by stressors, but by the global metabolism of organisms. Depending on food availability, reproductive status, growth- dilution with season and some other factors, the levels of pollutants in tissues and biomarker responses may fluctuate extensively during the year. As a result, changes in biomarker levels may simply be a natural part of the annual physiological cycle and thus quite unrelated to changes in exposure to chemical pollution (Baudrimont et al.,, 2007; Leinio and Lehtonen, 2005). Consequently, prior to the correct use of biomarkers it is essential to know the ranges of natural variability in their levels (Galloway et al., 2004).

In this interdisciplinary approach, mean, maximum and minimum values of contaminant and parasite burdens and of stress biomarkers will be given and compared to what is known. In the following text, the terms ―stress‖ and ―adaptive response‖ will be restricted to the parameters that

102

Chapitre III. Approche multistress in situ were measured. All seasons will be pooled in order to obtain a chronic and averaged overview of stresses and bivalve responses within two years. Concomitantly, when available, population parameters have been estimated. When both ecological (population dynamic parameters) and ecotoxicological indicators (physiological biomarkers) are good, we could be confident in concluding that the ecosystem was healthy. Conversely, when both indicators traduce degradation in individuals and populations, we reach the conclusion as to the bad quality of the system (Vasseur and Cossu-Leguille, 2003). In case of disruption between both indicators, adaptation, compensation or genetic selection of organisms can be evoked (Gestal et al., 2008) as well as unfavourable ecological conditions (Vasseur and Cossu-Leguille, 2003).

2. Materiel and methods

2.1. Target bivalves

To evaluate specificity of mechanisms, two bivalve species were selected. The cockle Cerastoderma edule has previously be applied as a test organism in studies concerning metal accumulation (Baudrimont et al., 2003, 2006), biomarker expression (Desclaux-Marchand et al., 2007; Paul-Pont et al., 2007), parasite infection (de Montaudouin et al., 2005, 2009; Lassalle et al., 2007; Thieltges and Reise, 2006) and immunity (Wootton et al., 2003). The Manila clam Ruditapes philippinarum is an introduced (since 1972) and exploited species along the French coast. This is the second exploited bivalve in the world, after the Pacific Oyster, Crassostrea gigas. Its large geographical distribution makes it a good model in ecotoxicology of coastal systems (Fan and Wang, 2001; Ji et al., 2006; Veltman et al., 2008) and numerous studies about this biological model concerned parasite infection and associated immune responses (Da Silva et al., 2008; Hégaret et al., 2007, 2009).

In term of metal bioaccumulation, cockles have previously be applied as bioindicator of field metal contamination (Baudrimont et al., 2005; Bietz et al., 1999; Cheggour et al., 2001; Jung et al., 2006), as well as clams (Baudrimont et al., 2005; Martín-Díaz et al., 2008; Moschino et al., 2009; Sfriso et al., 2008). Both species have also been studied as a test organism in studies concerning inter-specific variation of metal accumulation (Saavedra et al., 2004, 2009; Wang and Rainbow, 2005).

103

Chapitre III. Approche multistress in situ

Figure III.1. Position of the sampling sites: ―Banc d‘Arguin‖ and ―Andernos‖ in Arcachon Bay, and Landéda.

104

Chapitre III. Approche multistress in situ

In terms of infectious pathology, Manila clams are severely impacted by several diseases, including brown ring disease (Allam et al., 2002; Paillard et al., 1997), perkinsosis (Elandalloussi et al., 2008; Elston et al., 2004; Hamaguchi et al., 1998; Park et al., 2008) and brown muscle disease (Dang and de Montaudouin, 2009; Dang et al., 2008). However, Manila clam is generally free of heavy trematode infection. The contrary was observed for cockles which are severely impacted by trematodes and remained exempt from BMD, BRD and almost from perkinsosis (Lasalle et al., 2007).

2.2. Sampling sites

Bivalves were seasonally collected in three different intertidal sites: two sites were located in Arcachon bay (―Banc d‘Arguin‖ and ―Andernos‖) and one site located in Britanny (―Landéda‖). Cockles were sampled at Banc d‘Arguin and Landéda, France (Figure III.1). Banc d‘Arguin (44°34‘N, 1°15‘W) is a sandy bank at the entrance of Arcachon bay where salinity remains high year-round (34-35 psu). Surface sediment temperature fluctuates between -1.4°C and 30.1°C (year average = 14.9°C) (own records). Sediment at the sampling station was medium sand (grain-size median = 360 µm) and immersion rate was 60%. Cockle abundance is ranging from 100 to 200 ind.m-2. Landéda (48°35‘N, 4°35‘W) site is a muddy sand bay (grain-size median: ca.300 µm). Salinity remains high year-round (34-35 psu) due to the oceanic position of this site. Mean surface water temperature fluctuates between 9°C in winter and 19°C in summer. Studied cockles come from a park wherein they have been seeded from Le Croisic (France) in 2006. Landéda was mostly well-known because of the important venerid culture of the Manila clams in this site before the BRD in 1987. Because of clam mass mortality in this site, we worked on recently introduced clams from Morbihan Gulf Island (―Ile aux Oiseaux‖). Finally, Manila clams were also sampled at Andernos (44°42‘N, 1°8‘W), an inner site of Arcachon Bay (Figure III.1). Surface sediment temperature fluctuates between -1.0°C and 36.6°C (year average = 15.3°C) (own records) and salinity ranges between 19 and 35 psu (year average = 30 psu) (Dang, 2009). Sediment at the sampling station is fine sand (grain-size median = 160 µm) covered by a Zostera noltii bed and immersion rate was 60%. Clam abundance is usually below 60 ind.m-2. The locations of Arguin and Andernos may both be called ―Arcachon‖ in the following text and opposed to Landéda.

2.3. Sampling and Methodology

105

Chapitre III. Approche multistress in situ

From January 2007 to December 2008, thirty clams and cockles were seasonally collected (eight seasons) in the three experimental sites (Arguin, Andernos and Landéda). First, the shell length

Table III.1: List of the parameters measured individually at each site. each date for the two target species (clam and cockle) and number of analyzed for each parameter (x: measured; -: unknow).

Series n 1 Series n 2 Individual parameters (20 animals) (10 animals) Condition index x x Disease status BRD* prevalence (shell) x x BMD* prevalence (muscle) x x Perkinsus load (gill) x - Trematodes load (whole flesh) x x

Metal dosage (gills and digestive gland) Cadmium (Cd) - x Copper (Cu) - x Zinc (Zn) - x

Partial oxygen pressure (PO2) (heart) x - Gene expression (gills and digestive gland) Cytochrome oxidase (cox) - x Superoxide dismutase (sod) - x metallothionein (mt) - x Metallothionein (gills and digestive gland) - x Hemocyte parameters Total hemocyte count (THC) - x Granulocyte concentration (GR) - x Granulocyte percentage (%GR) - x Hyalinocyte concentration (HY) - x Single cell concentration (S) - x Aggregated cell concentration (Agg) - x

106

Chapitre III. Approche multistress in situ structure of the different populations was seasonally determined. Six 0.25-m² quadrates were randomly collected and 1-mm mesh sieved. Individuals were sorted and shell length was measured with a calliper at the nearest mm. The mean annual population structure gives information about recruitment, abundance and maximal length.

Second, adults from the dominant length class were selected by hand, i.e. 31-35 mm range for cockles and 36-39 mm range for clams. Individuals were carefully brought back to the laboratory for immediate analysis. Two series of individuals were dissected: serie n°1 consists of 20 animals and serie n°2, 10 animals. Different parameters were analysed individually on animals according to their series (n°1 and n°2) (Table III.1).

Among analyzed parameters, some were not included in the present study: 1) we attempted to measure PO2 with optodes in heart blood (Series 1), just after opening the shell. However, basis levels in these bivalve species were too low to discern any variations between sites, seasons, species (<0.2 kPa). 2) We measured the expression of four genes that are involved in mitochondrial metabolism (cox I and 12S), response to oxidative stress (sod) and detoxification (mt). However, we considered that among physiological processes involved in response to stress, genes regulation set up precociously upstream from other biochemical pathways and transiently in response to acute stress. Consequently, that this is nonsense to average values at the seasonal level, in relation with chronic stresses. These two parameters were maintained in Table 1 to get an overview of the complete analysis methodology.

2.4. Condition index

Condition index was measured on 20 animals as the ratio of the flesh dry weight (in mg) on the shell dry weight (in g) (Walne and Mann, 1975).

2.5. Disease status

2.5.1. Brown ring disease

The severity of Brown Ring Disease is scored by a symptom, which is the extent of the conchiolin deposit and whether there is evidence of recovery in the form of calcium carbonate layered over the deposit (Paillard and Maes, 1994). The Conchiolin Deposit Stage (CDS) is scored from 0 to 7 with 7 being the most extensive deposits; the Shell Repair Stage (SRS) is scored from 0 to 3, with 3 being a completely repaired shell. The two parameters are ranked

107

Chapitre III. Approche multistress in situ separately, and then combined into an overall disease score (Disease Phase) that ranges from 0 to 3, with 3 being a completely recovered clam. BRD prevalence corresponds to the number of clams exhibiting CDS from 1 to 7 divided by the number of total clam examined.

2.5.2. Brown muscle disease

This disease concerns Manila clam only and was recently described (Dang et al., 2008). Necrosis of the posterior muscle, the most obvious clinical sign of BMD, was explored in all dissected animals.

2.5.3. Perkinsus olseni load

The two branchial lamellae of one side from each of the 20 clams or cockles sampled were excised, weighed, and incubated in Ray‘s fluid thioglycollate medium (RFTM, Ray, 1966) for seven days in the dark at room temperature. Gill tissues incubated in RFTM were then removed and prepared according to the protocol of Choi et al. (1989). Briefly, gill tissues were centrifuged (800×g, 10 min) to remove medium, digested with NaOH (2N, 1 h at 60°C, twice) and followed by two washes in PBS (0.1 M). Finally, the pellet was re-suspended in 1 mL of PBS. Ten microlitres of Lugol‘s solution were added, and the number of P. olseni hypnospores was assessed using a Nageotte chamber under a light microscope. The level of infection by P. olseni (= Perkinsus load) was presented as the number of hypnospores per wet gram of gill tissue.

2.5.4. Trematode diagnosis

The bivalve flesh was squeezed between two large glass slides and observed under stereomicroscope to identify and count trematode utilizing the shellfish as second intermediate host (metacercariae). Identification was made following de Montaudouin et al. (2009). When an individual was utilized as first intermediate host, its tissues were completely invaded by sporocysts. This generally occurs in less than 10% of the population (de Montaudouin et al., 2009; Desclaux et al., 2002). These individuals were discarded. Results were expressed as number of cysts per individual.

2.6. Metal analysis

108

Chapitre III. Approche multistress in situ

Metal quantification was carried out for gills and visceral mass of bivalves. Dry weights of tissues were determined before broken down in 1 to 1.5 mL (depending on weight of tissue) of nitric acid (Fluka; Buchs, Switzerland. 65% HNO3) at 100°C for 3h to dissolve metals in the liquid for quantification. After a six-fold dilution of digestates with ultrapure water (MilliQ®, Bedford, MA, USA), Cd, Cu and Zn concentrations were analyzed. Cadmium concentration was determined by electrothermic atomic absorption spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. Copper and zinc concentrations were determined by Atomic Absorption Spectrophotometry using an air-acetylene flame (Varian AA 220 FS, Victoria, Australia). The detection limits were 0.1 mg.L-1 and 0.04 mg.L-1 for Cu and Zn, respectively.

The analytical methods were simultaneously validated for each sample series by the analysis of standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Throughout our metal analyses, mean values were consistently in certified ranges of Tort-2 and Dolt-3 (data not shown).

2.7. Metallothioneins analysis

As soon as bivalves were dissected, gills and visceral mass were placed into polyethylene bags

(Whirl-Pak) under N2 atmosphere at -80°C to minimize metallothionein oxidation until the analysis. The concentration of total MT protein was determined in tissues of bivalves by the mercury-saturation assay, using cold inorganic mercury according to a protocol already described by Baudrimont et al. (2003). This technique is based on the quantification of Hg bound to the saturated MT. The denaturation of non-MT proteins was performed using trichloroacetic acid and the excess Hg not bound to the MT was removed by scavenging with lyophilized beef hemoglobin (Sigma) prepared in 30 mM Tris-HCl buffer (Sigma, pH 8.2 at 20°C). The final supernatant was then quantitatively recovered and used for Hg determination by flameless atomic absorption spectrometry (AMA 254, Altec, Prague, Czech Republic). The detection limit was estimated at 0.01 ng Hg. The exact quantity of Hg binding sites per MT molecule being unknown for these species, MT concentrations were expressed in nmol sites Hg.g−1 (fw).

At the same time, three references samples or ―blanks‖ were prepared to monitor the Hg complexation efficiency of the hemoglobin. The mean of the three blank values at each analytical run was deducted from the Hg burdens measured in each sample.

109

Chapitre III. Approche multistress in situ

A recovery percentage from purified rabbit liver MT (Alexis biochemicals ALX-202-071) was systematically determined. This ―internal standard‖ enabled us to determine the ratio between the binding sites measured after Hg saturation and the potential binding sites indicated by the supplier and previously verified by Cd and Zn determinations on purified MT solution samples. Throughout our metallothionein analyses, the mean recovery percentage was 91.2 ± 9.6 %. This value was consistently within the certified ranges (100 ± 20 %) of the method.

2.8. Immunological parameters

Hemolymph was withdrawn from the adductor muscle of each individual clam or cockle with an adapted needle attached to a 1-mL syringe, then passed through 80 µm mesh, stored temporarily in a microcentrifuge tube on crushed ice to retard cell clumping.

For flow cytometry (FCM) analysis, 200µL of a 6% formol solution in filtered sterile seawater (FSSW) and 100 µL of FSSW were added to 100µL of hemolymph from each individual in a clean 1.5 mL microtube. FCM analyses were performed on a FACScalibur (BD Biosciences, San Jose, CA) flow-cytometer equipped with a 488 nm argon laser. Procedures for characterization of hemocytes were adapted from those of Delaporte et al. (2003), Soudant et al. (2004) and Lambert et al. (2007). Briefly, fixed hemolymph samples were incubated with SYBR® green I (Molecular Probes, 10× final concentration, =1/1000 of the DMSO commercial solution), a nucleic acid- specific dye, in darkness at room temperature for 30 minutes before flow-cytometric analysis. A density plot visualization of SSC vs. FL1 allowed differentiation and gating of hemocytes stained by SYBR® green from other particles in the hemolymph. Two distinct sub-populations of hemocytes were identified on a FSC (size) -SSC (complexity) density plot according to their SSC and FSC ranges: Granulocytes (high SSC and high FSC) and hyalinocytes (low SSC, high FSC). The flow rate of the cytometer was measured as previously described (Marie et al., 1999) by weighting a tube containing distilled water before and after a timed analysis to determine the volume analyzed. Calculated flow rate was used to determine hemocyte concentrations. For each individual hemolymph sample, the following parameters were calculated: total hemocyte count (cell.mL-1), single cell concentration (cell.mL-1), aggregated cell concentration (cell.mL-1), granulocyte and hyalinocyte concentration (cell.mL-1) and finally the percentage of granulocyte among single cells (%).

2.9. Data analysis

110

Chapitre III. Approche multistress in situ

The objective was to obtain mean annual values, including natural seasonal range. Due to heterogeneity of variances, median and quartiles were calculated for each parameter and comparisons between sites/species were performed using Kruskal-Wallis non parametric test (Sokal & Rohlf, 1981). In order to obtain an overview of the in situ monitoring results, two Principal Component Analysis (PCA) were also conducted with median values, one to display stress factors (metacercariae of trematodes, Perkinsus, brown ring disease and Cu-Zn-Cd concentrations in gills and visceral mass) and the other to display response factors (metallothioneins in gills and visceral mass and immunological parameters).

3. Results and discussion

3.1. Stress factors

3.1.1. Trematode diagnosis

The median number of metacercariae was below 27 per cockle at Arguin and Landéda (Kruskal- Wallis (KW) test, p > 0.05) (Figure III.2A; Table III.2). In cockles, the higher variability recorded at Arguin was due to high abundance of Curtuteria arguinaea at the beginning of the survey. The dominant species was Meiogymnophallus minutus whereas at Landéda, the dominant species was another echinostomatid trematode, Himasthla elongata (Figure III.3). Levels of infestation in both populations remained low compared to hundreds of metacercariae that can be found in tissues of molluscs (de Montaudouin et al., 2009; Thieltges et al. 2006) and below the threshold beyond which mortality was observed (Desclaux et al., 2004; Gam et al., 2009). Indeed, the abundance of M. minutus can easily overpass 1000 metacercariae per cockle (de Montaudouin et al., 2009) and effects on populations may occur when it reaches 400 metacercariae per cockle (Gam et al., 2009). Manila clams were insignificantly infested by trematodes (KW test, p > 0.05) (Figure III.2A). These results were in accordance with knowledge about sensitivity of these bivalve species to trematode parasites. Cockles are known to host many different parasites species, especially digenean trematodes (Lauckner, 1983; de Montaudouin et al., 2009) whereas Manila clams constitute a resistant species to trematodes (Dang et al., 2009a). Densities of host tissues as well as parasite strategy to optimize the likelihood of trophic transmission were envisaged to explain this host specificity. Indeed, manila clams present some disadvantages for digenean in terms of successful parasite strategy, such as a deeper position in sediment and autumn recruitment,

111

Chapitre III. Approche multistress in situ

A 400 a 350

300

250

(cyst per bivalve) per(cyst 200

150

100 a

Metacercariae Metacercariae 50 b b 0 Cockle ARC Cockle LAN Clam ARC Clam LAN

B 160000 b 140000

wet gills) wet 120000

1 - 100000 b 80000

load (cells.gload 60000

40000

20000 Perkinsus a a 0 Cockle ARC Cockle LAN Clam ARC Clam LAN C 80 b 70

60

50

40

% BRD % 30 a,b 20

10 a a 0 Cockle ARC Cockle LAN Clam ARC Clam LAN Figure III.2. Infestation levels of bivalves by (A) metacercarie, (B) Perkinsus olseni and (C) pourcentage of bivalves affected by the Brown Ring disease (BRD) (median ± min/max). ARC: Arcachon; LAN: Landéda. Letters a, b indicate statistical differences between populations (Kruskal-Wallis test, p < 0.05).

112

Chapitre III. Approche multistress in situ when temperature was too low to cercarial emission (Wegeberg, 1998). In addition, several studies have been reported the lack infection of introduced bivalve species by natural parasites in comparison to native species (Bachelet et al., 2004; Calvo-Ugarteburu and McQuaid, 1998a,b). In the same way, even if some clams were slightly infested as second intermediate host (metacercariae stage, abundance max.: 9.8 metacercariae per clam), there was no report of clam infestation as first intermediate host (sporocyst stage) whatever the origin during all the study. Cockles only were infested as first intermediate host, by Bucephalus minimus, with a low prevalence of 1.6% and 2.7% at Banc d‘Arguin and Landéda, respectively. The higher host specificity of trematode at sporocyst stage rather than metacercarie explains the species difference in B. minimus infestation (Desclaux, 2003).

3.1.2. Perkinsus abundance

Perkinsus load is strongly linked to the target species. Clams are more infested (5.104 hypnospores per gram of gill tissue) than cockles wherein intensity of Perkinsus was negligible (1 cell.g-1 wet gill tissues) and could be considered as null (KW test, p < 0.05) (Figure III.2B). Infestation by Perkinsus sp. was species-dependent as it has been already observed by Lassalle et al. (2007). Gauthier and Vasta (2002) compared the effects of mollusc plasmas on the in vitro proliferation of Perkinsus marinus between highly susceptible bivalve (Crassostrea virginica) and other bivalve mollusc species that, although naturally exposed to the parasite or even carrying sublethal parasite loads, do not show obvious sign of disease (Mytilus edulis, Mercenaria mercenaria and Anadara ovalis ). Authors demonstrated higher inhibition of P. marinus in vitro proliferation by plasma from M. edulis, M. merceneria and A. ovalis than plasma from C. virgnica due to species difference of plasma humoral inhibitors. An experimental study conducted by Defer et al. (2009) demonstrated that C. edule had the broadest and the most active antibacterial activity in hemolymph compared to four other mollusc species including R. philippinarum. Consequently, differences in humoral and cellular components in hemolymph of cockles compared to clams might explain the difference of infection levels between both sympatric species.

Estimation of Perkinsus sp. burden in gill (cells.g-1 wet gill tissues) provides an over-estimation of the total body burden (cells.g-1 wet flesh) (Choi et al., 2002). Among clams, population from Andernos highlighted a 12 times more important Perkinsus load than those from Landéda (99000

113

Chapitre III. Approche multistress in situ

100% 90% 80% D. brusinae

70% P. brevicolle M. minutus 60%

host C. arguinaea 50% H. continua 40% H. elongata % metacercariae per % per metacercariae 30% H. interrupta 20% H. quissetensis 10% 0% Cockle-Arc Cockle-Lan Clam-Arc Clam-Lan

Figure III.3. Percentage of metacercariae species in bivalves.

114

Chapitre III. Approche multistress in situ vs. 8000 cells. g-1 wet gill tissues) (Figure III.2B; Table III.2). A previous study conducted by Lassalle et al. (2007) showed similar result with an increase of Perkinsus olseni density to the south, from 0 at Landéda to 90 000 hypnospores per gram in gills of clams at Andernos. In addition, Perkinsus olseni prevalence varied from 65 to 100% in clams from Andernos with an average prevalence of 84%, whereas prevalence of P. olseni in clams from Landéda was lower and varied from 30 to 80% with an average value of 57%. The most important environmental factors influencing Perkinsus spp. viability, metabolic activity and proliferation were temperature and salinity. Infection intensity of several Perkinsus sp. was positively correlated with temperature (La Peyre et al., 2008) and salinity (Villalba et al., 2004). In the present survey, salinities were similar between Landéda and Arcachon Bay. Consequently, this parameter may not affect Perkinsus sp. distribution between both sites. However, higher Perkinsus olseni abundance at Andernos than at Landéda might be linked with latitudinal gradient of temperature between both locations.

Level of infection was moderate in Manila clams from Landéda (mean 2x104 cells.g-1 wet gill tissues and median 7.9x103 cells.g-1 wet gill tissues) as it has been already reported at this site (Lassalle et al., 2007) and in Manila clams from another brittany location (Gulf of Morbihan) (Flye-Sainte-Marie et al., 2009). Higher values were reported for Manila clams from Andernos (mean 1x105 cells.g-1 wet gill tissues and median 9.9x104 cells.g-1 wet gill tissues) consistently with a previous study (Lassalle et al., 2007) (Table III.2). Higher infection levels have been reported in Korea (8.7x106 cells. g-1 flesh) and in Japan (2.3x106 cells.g-1, flesh) (Choi et al., 2002; Park and Choi, 2001). However, infection level of Andernos population remained important since several physiological perturbations were demonstrated from 105 cells. g-1 wet gill tissues such as inhibition of gametogenesis, decrease of host growth associated with a concomitant decrease of feeding rate and increase of energy consumption (Villalba et al., 2004). Nevertheless, such level was not yet considered as critical for theoretical sustainability of clam population since several studies demonstrated apparition of mortality events from 106 cells. g-1 wet gill tissues (Dang, 2009).

3.1.3. Brown ring disease

To date, four techniques are available for diagnosis of Brown Ring Disease and V. tapetis detection: 2 PCR assays with different set of primers: PCR assay of Rodríguez et al. (2006) and Paillard et al. (2006), a microbiological method and the shell valve analysis. Drummond et al.

115

Chapitre III. Approche multistress in situ

(2006) compared these four techniques and showed that in experimental conditions, none of the four techniques was sufficient on its own for effective BRD diagnosis. However, even if shell valve analysis was the least sensitive technique, detecting BRD in 7.7% of the sample, it was an essential diagnostic tool because it was the only technique that identified the disease, rather than the etiological agent. Only the Manila clams from Landéda were affected seriously by Brown Ring Disease. Apparition of BRD at Landéda goes back to 80‘s with record of massive mortalities of cultured Manila clams in this site (Paillard et al., 1989). The median BRD prevalence year-round was 33% in Landéda with a maximal prevalence of 75% (Figure III.2C; Table III.2). This result is in accordance with a previous study conducted on aquacultural clams beds of Northen Brittany (Brouennou, Finistère, France) in which prevalances ranged between 33 and 100% (Paillard, 2004) but higher compared to prevalences found in natural populations of the Gulf of Morbihan (4-30%, Lassalle et al. 2007; Paillard, 2004). The period of recovery appears at the end of the sampling, from the end of February to May 2008. In clam from Andernos, median BRD prevalence was 11% and disease prevalence never reached more than 25% during all the sampling. Clams infected by the disease at Andernos presented only microscopic signs, i.e. minute irregularities within the shell along the pallial line. Thus, BRD was present in Andernos with a low prevalence and a low stage of infection (stage 1) in accordance with previous study conducted by Dang et al. (2009b) which revealed low BRD prevalence and intensity in several sites of Arcachon Bay. Higher infection level in northwestern France compared to Arcachon Bay could be explained by temperature patterns, as BRD is classified as a cold-water disease (Paillard, 2004). In cockles from the two sites, no BRD signs were observed. However, even if Arguin does not present the same location than Andernos in Arcachon Bay, cockles from Landéda lived in sympatry with clams wherein BRD prevalence was relatively high. Cockles are known to be ―carrier species‖ for this bacterium since no BRD symptoms appeared both in natural environment and after experimental challenges (Maes and Paillard, 1992). Thus, no sign of BRD in cockles did not consequently mean that prevalence of Vibrio tapetis was null. Such hypothesis seemed to be obvious for Landéda but it was not necessarily intuitive for Arguin.

3.1.4. Brown muscle disease

116

Chapitre III. Approche multistress in situ

None of our samples presented a posterior muscle necrosis, the most obvious clinical sign of BMD whereas recent studies demonstrated the presence of BMD clinic signs in clam population of Andernos (Dang et al., 2008).

3.1.5. Metal analysis

Metals concentrations (Cd, Cu, Zn) found in bivalves living at Landéda and Arcachon Bay remained low (Table III.2) and similar to baseline levels already demonstrated in cockles and clams from numerous non-contaminated sites (Ji et al., 2006; Jung et al., 2006). Thus, the present study demonstrated that neither site can be considered highly contaminated. Nevertheless, Cd was enriched at Andernos as it was revealed by metals concentrations in clams from this site.

Metals concentrations in visceral mass ([Cd]VM, [Cu]VM, [Zn]VM) of bivalves displayed similar trends among metals analyzed. In this tissue, highest metal contents were systematically found in clams compared to cockles (Figure III.4, Table III.2) (KW test, p < 0.05). A significant site effect highlighted higher metal contents in clams of Arcachon compared to Landéda (KW test, p < 0.05). Thus, among the four populations examined in the present survey, Manila clams from Andernos presented the highest Cd, Cu and Zn concentrations in the visceral mass (Figure III.4, Table III.2).

On contrary, fluctuations of gills metal concentrations among species and sites were metal dependent. Although copper contents were similar between species (KW test, p > 0.05), cockles displayed highest Cd concentrations and lower Zn concentrations than clams (KW test, p < 0.05). Considering site differences, zinc concentrations in gills of bivalves were similar at Landéda and Arcachon whereas Cd and Cu contents were higher in gills of bivalves from Landéda compared to Arcachon (Figure III.4, Table III.2). It is obvious that metal concentration in both sites constituted the first clue to understand differences in metal bioaccumulation between populations. In Arcachon Bay, interspecific differences were demonstrated with higher metal concentrations in visceral mass of clams than cockles (Figure III.4, Table III.2). However, since Arguin and Andernos constituted two different sites within the bay, these highest metal contents in clams have to be mainly related to particular metal concentration in sediment from Andernos in contrast with cockles from Arguin. In Landéda, cockles and clams lived in the same area but displayed some differences in metal contents: cockles have higher Cd concentration in gills than clams whereas clams showed higher Cu concentration in visceral mass than cockles. Since both species

117

Chapitre III. Approche multistress in situ

2500 A a 1400 a 1200 2000

1000 Gills

- b 1500 a,b

Visceral mass 800 – 600 1000 b Cadmium b b b

400 Cadmium 500 200

0 0 Cockle ARC Cockle LAN Clam ARC Clam LAN Cockle ARC Cockle LAN Clam ARC Clam LAN

120000 B 100000 a 90000 100000 80000 b a,b 70000 80000 b

Gills 60000 - 60000 50000

Zinc 40000 Visceral mass

40000 – 30000 NS

Zinc 20000 20000 10000 0 0 Cockle ARC Cockle LAN Clam ARC Clam LAN Cockle ARC Cockle LAN Clam ARC Clam LAN

20000 C a,b 25000 a

15000 a a,b 20000 Gill - b 15000 a,b

10000 Visceral mass – 10000 Copper b,c

5000 Copper c 5000

0 0 Cockle ARC Cockle LAN Clam ARC Clam LAN Cockle ARC Cockle LAN Clam ARC Clam LAN

Figure III.4. Concentrations (in ng.g-1. dw) of cadmium (A), zinc (B) and copper (C) in gills and visceral mass of bivalves (median ± min/max). NS: no significant difference. ARC: Arcachon; LAN: Landéda. Letters a, b, c indicate statistical differences between populations (Kruskal- Wallis test, p < 0.05).

118

Chapitre III. Approche multistress in situ lived in the same geochemical environment, interspecific difference was caused by different abilities of bivalves to accumulate metals. Indeed, it is well known that metal concentrations vary among different species of marine bivalves whithin a same location, even between taxonomically closed species (Pan and Wang, 2009). First, subcellular repartition of metal has ecotoxicological implications and several studies demonstrated that association between metals and different subcellular pools were metal and species dependent (Marigómez et al., 2002). Some authors demonstrated implication of metal subcellular repartition in both inter and intraspecies differences on metal accumulation (Shi and Wang, 2004, 2005). However, such processes have been little considered among literature data probably due to the complexity of metal handling strategies in bivalves tissues (Wang and Rainbow, 2005). In addition, development of biokinetic models within the last decades enhanced the understanding of metal bioaccumulation in aquatic animals and consequently explained interspecies discrepancies. Such models take into account all processes leading to metal bioaccumulation such as metal concentration in dissolved and dietary phases, uptake rate from dissolved phase, assimilation efficiency from the dietary phase, ingestion rate and efflux rate. These biokinetic parameters were obviously species and metal dependent (Wang and Rainbow, 2008) and may explain interspecies variation in metal concentration of bivalves from Landéda. For example, Pan and Wang (2009) suggested that filtration activities explained the interspecies differences of the Cu dissolved uptake between five marine bivalve species. Authors also demonstrated that efflux rate and, in a lesser extent, assimilation efficiency strongly contributed to interspecies differences of total Cu body burden between the five bivalve species. Finally, differences between gills and visceral mass among these bivalve species could be related to different metal handling strategies leading to different prioritary pathways (trophic vs. direct contamination) as well as modification of transfer of trophic compounds from gills to visceral mass for detoxification processes (Bebianno et al., 1994).

Intraspecies differences in metal concentration were also shown in the present study. Cockles from Landéda displayed higher Cu concentration in gills than cockles from Arcachon Bay (KW test, p < 0.05). In addition, clams from Arcachon Bay presented higher metal concentration, particularly Cd concentration, in their visceral mass than clams from Landéda. Both metal concentration in water and sediment as well as environmental parameters (temperature, salinity, nature of sediment) which could modify metal chemical and physical speciation leading to differences in metal bioavailability (Campbell and Couillard, 2004) may explain the intraspecies

119

Chapitre III. Approche multistress in situ

1,0

%BRD [Zn]G

0,5

[Cu]VM 0,0 [Zn]VM

Fact. 2 : 18,27% 2 Fact. [Cd]G [Cd]VM Metacercariae -0,5 Perkinsus [Cu]G

-1,0

-1,0 -0,5 0,0 0,5 1,0 Fact. 1 : 39,26% 5

4 Clams Landéda 3 B-P-Wi2B-P-Su1 Cockles Arcachon B-P-Au1 2 B-P-Sp1 Cockles Landéda B-P-Sp2 B-P-Au2 A-C-Wi2 1 B-C-Wi2 A-P-Sp2 B-P-Su2 A-C-Sp2B-C-Su1 A-C-Au2 0 A-C-Su2 A-P-Au1 A-C-Au1 A-P-Su1 A-C-Su1 Fact. 2 : : 18,27% 2 Fact. A-P-Wi2 A-C-Sp1 B-C-Au1 B-C-Sp2 A-P-Au2 A-P-Su2 B-C-Su2 -1 B-C-Au2 A-P-Sp1 B-C-Sp1 B-C-Wi1 A-P-Wi1 -2 A-C-Wi1

-3 Clams Arcachon -4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5

Fact. 1 : 39,26% Figure III.5. Principal Component Analysis 1 (PCA-1). Loading of stress factors (a) and (b) bivalve populations (A-P: clams from Arcachon ; B-P: clams from Landéda ; A-C: cockles from Arcachon B-C: cockles from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008) on the two first principal components (F1-F2). 120

Chapitre III. Approche multistress in situ differences in metal bioaccumulation between bivalves from Landéda and Arcachon Bay. In addition, biokinetic modeling can also be used to assess processes accounting in intraspecific differences in metal bioaccumulation between different sites. Indeed, species specific biokinetic parameters may vary strongly following environmental conditions (temperature, salinity, oxygen, nature of sediment, food quantity/quality) (Wang and Rainbow, 2008). Finally, it is important to note that field contamination context may also modify metal accumulation abilities of bivalves and may play a large part in explaining intraspecies differences of metal bioaccumulation. Shi and Wang (2004) demonstrated that population of Manila clams collected from a contaminated site with higher Cd tissue concentration have higher Cd assimilation efficiency in relation to higher MT contents than clams population collected from less contaminated or clean site. In the same way, Mouneyrac et al. (2003) demonstrated that Cu-tolerant population of the polychaete worm Nereis diversicolor living in Cu contaminated site combined higher rate of Cu accumulation and detoxification than population from a less contaminated site. Thus, population living in contaminated environment may modify their metal handling physiology and biochemistry, leading to modification of trace metal bioaccumulation compared to population from clean site.

The present field survey highlighted differences in metal concentrations between clam populations as well as between cockles populations between both sites. However, we cannot discriminate the part of abiotic conditions of respective habitats and feeding ecology of bivalves from the part of specific metal handling strategies between populations in explaining these differences of metal bioaccumulation. Thus, laboratory contamination experiment of both populations under the same experimental settings may help us to assess whether specific metal detoxification strategy occurred between populations of bivalves to explain a part of these intraspecific differences.

3.1.6. Principal Component Analysis – Stress factors

Several PCA were performed to identify population repartition among stress. The Principal component analysis 1 (PCA-1) (Figure III.5) gathered both species in both sites. It extracted two components explaining more than one half of the total variance (57.5%). The first axis was mainly defined by concentration of metals in visceral mass of bivalves ([Cd]VM, [Cu]VM, [Zn]VM).

The second principal component was mostly defined by Zn concentration in gills ([Zn]G) and %

121

Chapitre III. Approche multistress in situ

3,0

1,0 [Cd]G 2,5 A-P-Sp1 2,0 B-P-Su1 1,5 A-P-Wi1 B-P-Sp1 0,5 Perkinsus %BRD 1,0 B-P-Au1 [Cu]G 0,5 B-P-Su2 [Cu]VM A-P-Au2 [Zn] 0,0 VM 0,0

A-P-Wi2 A-P-Su1 B-P-Wi2 . 2 : : 16,00%2 . . 2 : : 16,00%2 . [Cd] [Zn] VM G -0,5

. 2 : 16,00% : .2 A-P-Su2 Fact Fact B-P-Au2 B-P-Sp2

Fact -1,0 -0,5 A-P-Sp2 -1,5 A-P-Au1 -2,0

-1,0 -2,5

-1,0 -0,5 0,0 0,5 1,0 -3,0

Fact. 1 : 50,39% -3,5 -5 -4 -3 -2 -1 0 1 2 3 4 5 Fact. 1 : 50,39%

Figure III.6. Principal Component Analysis 2 (PCA-2). Loading of stress factors (a) and (b) clams populations (A-P: clams from Arcachon ; B-P: clams from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008) on the two first principal components (F1-F2).

4

1,0 [Zn]G 3

2 B-C-Su1 0,5 B-C-Wi2 A-C-Wi2 1 [Cu]G A-C-Sp2 [Cu]VM B-C-Au1 A-C-Au1 B-C-Su2 A-C-Su2 B-C-Sp2 [Zn]VM 0 0,0 A-C-Au2

B-C-Au2 A-C-Sp1 . 2 : : 20,35% 2 . . 2 : : 20,35% 2 . [Cd] B-C-Sp1

G B-C-Wi1 A-C-Su1

. 2 : 20,35% : .2 Fact Fact -1

[Cd]VM Fact -0,5 -2

Metacercariae -3 A-C-Wi1 -1,0 -4 -1,0 -0,5 0,0 0,5 1,0 Fact. 1 : 32,52% -5 -4 -3 -2 -1 0 1 2 3 4 5 6 Fact. 1 : 32,52%

Figure III.7. Principal Component Analysis 3 (PCA-3). Loading of stress factors (a) and (b) cockle populations (A-C: cockles from Arcachon ; B-C: cockles from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008) on the two first principal components (F1-F2).

122

Chapitre III. Approche multistress in situ

BRD whereas other stress (Perkinsus, metacercariae, [Cu]G, [Cd]G) have lower contribution on the two first factors. The present principal component analysis separated three groups according to stress factors (Figure III.5). Manila clams from Landéda were principally characterized by the development of brown ring disease but also by Zn contamination in gills. Manila clams from Andernos were mainly isolated by Perkinsosis infection and metal contamination in the visceral mass. Cockles from the two different geographical location, Landéda and Banc d‘Arguin formed a single group, characterized by trematode infection and Cd contamination in gills. Whereas global PCA-1 allowed us to distinguish both clam populations among stressors, we cannot assess intraspecies differences among cockle populations. Thus, supplementary PCA were performed in order to represent the repartition of each population within a species among stress factors

(Figures III.6 - III.7). PCA-2 confirmed repartition of clams among stressors with [Zn]G and %BRD on one hand (Landéda clams) and Cd, Cu and Zn concentrations in visceral mass

([Cd]VM, [Cu]VM, [Zn]VM) on the other hand (Andernos clams) (Figure III.6). PCA-3 extracted two factors accounting for 52.8% of inertia and was mostly defined by metal concentrations on axis 1, except [Zn]G which mainly contributed to axis 2 with the metacercariae load (Figure III.7). Among this PCA, cockles populations were mainly separated by higher Cd and Cu contamination in gills ([Cd]G, [Cu]G) of cockles from Landéda compared to those from Arguin. Metacercariae load was strongly linked to cockles from Arguin in Autumn (ACA) in accordance with high parasite load (Curtuteria arguinae) at this specific time point.

3.2. Response parameters

3.2.1. Metallothioneins

Natural range of MT concentration in both populations of cockles observed during the present survey (Figure III.8; Table III.4) was in accordance with results of a previous one-year field study conducted on this species in Arcachon Bay (Baudrimont et al., 2006). However, no field study using the same MT protocol analysis was performed on Manila clams making impossible potential comparison with other data. Indeed, several methods of MT quantification were demonstrated in literature but any provide an absolute value and the results are often expressed in different units (Amiard et al., 2006). Thus, intercalibration of the different existing analytical techniques for MT quantification was virtually impossible and data can be compared only when they have been obtained with the same analytical technique. Thus, in the present study only relative comparison was done on MT responses of clam populations. Further laboratory

123

Chapitre III. Approche multistress in situ

A 14 a 12

10

Gills - 8 b 6 a,b a,b

Metallothionein 4

2

0 Cockle ARC Cockle LAN Clam ARC Clam LAN B 70 a 60

50 a

Visceral Visceral mass 40 – a,b 30 b

20

Metallothionein 10

0 Cockle ARC Cockle LAN Clam ARC Clam LAN

Figure III.8. Metallothionein concentrations (in Eq nmol sites Hg.g-1. ww) in gills (A) and visceral mass (B) of bivalves (median ± min/max). ARC: Arcachon; LAN: Landéda. Letters a, b indicate statistical differences between populations (Kruskal-Wallis test, p < 0.05).

124

Chapitre III. Approche multistress in situ experiment may help us to assess whether MT concentrations of both Manila clam populations were under natural ranges or if they were due to identified stressors.

Within each site, there was no difference in the averaged concentration of MT between species (Figure III.8). However, for both species, a strong site effect occurred on MT concentrations: highest values were found at Arcachon bay both in gills and visceral mass (Figure III.8; Table III.4) (KW test, p < 0.05). In addition, higher variability in MT concentration was observed in cockles and clams from Arcachon bay, compared to Landéda. Higher variability of this metalloprotein in both bivalves from Arcachon Bay compared to Landéda might be related to gametogenesis and reproductive processes. Indeed, among factors likely able to induce MT biosynthesis except metals, hormones are frequently cited, especially in relation to reproductive phenomena (Amiard et al., 2006; Baudrimont et al., 1997b, 2006) and numerous studies have demonstrated that MT seasonal variations were partly linked to the reproductive cycle of bivalves (Baudrimont et al., 1997b, 2006; Cheggour et al., 2001; Serafim and Bebianno, 2001). On contrary, MT concentration in bivalves from Landéda displayed little fluctuations suggesting less intense gametogenesis. High fluctuation of condition index (CI) for cockles and clams from Landéda during the two years monitoring was in contradiction with the latter hypothesis. Indeed, even mean CI of cockles were similar at both location as were condition index of Manila clams, bivalves at Landéda displayed higher fluctuations suggesting more intense gametogenesis. However, since no mature individuals were found throughout the field survey (data not shown) and no recruitment occurred for both species, variation of CI might be only due to variation of food quality/quantity in Landéda leading to accumulation or loss of organic matter in bivalves. Finally, it might be interesting to wonder whether higher MT concentration in tissues of clams from Arcachon compared to clams from Landéda were due to differences in tissue metal concentrations and consequently metal handling strategies. Since too many environmental factors occurred in field, laboratory experiments may be necessary in order to go further in such interpretation.

3.2.2. Immunological parameters

Among the different immunological parameters that were analysed, total hemocyte count (THC) was correlated with granulocytes, hyalinocytes, single cells and aggregated cells concentrations (R>0.82, p<0.05). Therefore, THC and granulocyte percentage (% GR) only were kept for the further analysis.

125

Chapitre III. Approche multistress in situ

) 1 A - 2000000 1800000 a,b 1600000 1400000 1200000 a 1000000 800000 b 600000 a,b 400000 200000

Total Hemocyte Count (THC. in cells.mL (THC. Hemocyte Count Total 0 Cockle ARC Cockle LAN Clam ARC Clam LAN

B 90 80

70

60

50

40

% Granulocytes % 30

20 NS 10

0 Cockle ARC Cockle LAN Clam ARC Clam LAN

Figure III.9. Total hemocyte counts (A) and percentage of granulocytes (B) in bivalves (median ± min/max). NS: no significant difference. ARC: Arcachon; LAN: Landéda. Letters a, b indicate statistical differences between populations (Kruskal-Wallis test, p < 0.05).

126

Chapitre III. Approche multistress in situ

A site effect can be shown looking at THC: hemolymph from bivalves reared at Landeda showed 1.8 times more hemocytes than Arcachon (6.8x105 vs. 3.8x105 cell.mL-1) (Figure III.9A; Table III.4). Conversely, no significant difference can be showed among species within a same site. Looking at %GR, no difference is noticed between species and sites although clam hemolymph shows a slightly higher percentage of granulocytes as compared to cockles (median value 46.1% vs. 30.8%; Figure III.9B; Table III.4).

Total hemocyte counts found for clams in the present study were in accordance with the usual fluctuations of hemocyte counts in field clam populations according to season and sites (approximately from 1.1x105 to 1.8x106 cells.mL-1, (Flye-Sainte-Marie et al., 2009; Soudant et al., 2004). Even if some hemocyte parameters and activities of Cerastoderma edule and Cerastoderma glaucum have been characterized by Wotton et al. (2003) and Matozzo et al. (2007), actually no data on natural fluctuations of cockle hemocyte parameters in field was available in the literature. In cockle populations from the present field study, THC fluctuates between 1x105 and 1.7x106 cells.mL-1 in accordance with mean THC values observed in Cerastoderma spp. (Matozzo et al., 2007; Wotton et al., 2003). However, high percentage of granulocyte (85%) was reported for C. edule in contradiction with %GR values obtained from the present study which fluctuated between 21 and 57%, with a median value of 36%, among site and seasons (Figure III.9B; Table III.4).

Over the past decades relation between contaminant and bivalves hemocytes parameters have been widely studied under in vitro and in vivo laboratory experiments (Coles et al., 1994, 1995; Pipe and Coles, 1995). Only few field studies were performed to assess such relationship and revealed positive correlation between metals and THC, phagocytosis, hemocyte locomotion ability and hemolymph protein content in several bivalve species (Auffret et al., 2006; Cajaraville et al., 2000; Fisher et al., 2000; Oliver et al., 2003). Increase of THC was also demonstrated under in vivo contamination experiment using low concentration of Cd and Cu (Auffret and Oubella, 1994; Coles et al., 1995, Pipe et al., 1999). However, despite higher metal contents in clams from Andernos, no significant increase of THC was revealed in the present study. In addition, highly Perkinsus sp. infection (> 1.105 cells.g wet gill tissues) is known to increase THC and granulocyte concentration both in field (Flye-Sainte-Marie et al., 2009) and in laboratory (Villalba et al., 2004). Once more, neither THC nor %GR increased in clams population from Andernos severly impacted by Perkinsus olseni compared to clams from

127

Chapitre III. Approche multistress in situ

Landéda. Nevertheless, several experimental studies demonstrated an increase of THC in bivalves following Vibrio tapetis infection (Allam et al., 2000, 2006; Paillard, 2004) and the high BRD infection may explain higher THC in clams from Landéda. This hypothesis is in contradiction with the field study of Flye-Sainte-Marie et al. (2009) which showed no significant influence of BRD on any hemocyte parameters. However, prevalence (9,7%) and intensity of BRD (90% of clams < CDS 4) was lower than in Manila clams from Landéda in the present study. Calculation of correlation between THC and quantitative data on level of infection performed at individual scale would be necessary to carefully answer this matter. In addition to contaminants and pathogens, environmental (mainly temperature and salinity) and internal factors (mainly size/age and gametogenesis) are known to strongly influence immunological parameters (Auffret and Oubella, 1994; Flye-Sainte-Marie et al., 2009; Soudant et al., 2004). A field study conducted by Flye-Sainte-Marie et al. (2009) on Manila clams from Bailleron island in Gulf of Morbihan (France) demonstrated that temperature was the major environmental factor positively influencing hemocyte parameters whereas salinity (within the life range of the species) and food quality/quantity in natural conditions did not induce any modification of hemocytes parameters. Authors also demonstrated that a decrease of granulocyte concentration may occur during spawning period. In this context it was necessary to be careful to overstep the descriptive scale and to go further in the interpretation of the immune fluctuations between sites and species.

3.2.3. Principal Component Analysis - Responses

PCA-4 concerned both species in both sites. Despite a large variability between the different seasons, the first Axis of PCA-4 (48.6 % inertia) tends to separate both sites with Arcachon on the left and Landéda rather on the right part (Figure III.10).

128

Chapitre III. Approche multistress in situ

2,5

1,0 2,0 1,5 A-C-Wi2 A-C-Su2 A-C-Sp2 A-C-Wi1 B-C-Wi2 1,0 A-C-Au1 A-P-Au1 B-C-Sp1B-C-Su2 0,5 A-P-Sp1 A-P-Su2 B-C-Au2 0,5 A-C-Sp1 A-P-Au2 A-C-Su1A-C-Au2 0,0 A-P-Su1 B-P-Au1 A-P-Wi1 [MT] B-C-Au1 VM A-P-Sp2 0,0 -0,5 B-P-Au2 [MT]G B-P-Su1 B-C-Su1 . 2 : : 21,33% 2 . B-P-Wi2 -1,0 B-P-Sp2

. 2 : 21,33% : .2 B-P-Su2 Fact A-P-Wi2 B-P-Sp1

Fact -1,5 B-C-Wi1 -0,5 %Gr -2,0

THC -2,5 B-C-Sp2

-1,0 -3,0

-1,0 -0,5 0,0 0,5 1,0 -3,5

Fact. 1 : 48,63% -4,0 -5 -4 -3 -2 -1 0 1 2 3 4 Fact. 1 : 48,63%

Figure III.103.10. . Principal Component Analysis 4 (PCA-4). Loading of (a) MT concentrations and hemocyte counts and (b) bivalve populations (A-P: clams from Arcachon ; B-P: clams from Landéda ; A-C: cockles from Arcachon ; B-C: cockles from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008)on the two first principal components (F1-F2).

In regard with variables and individuals main contributions on each axis (data not shown), some trends might be observed. At Landéda, clams and cockles were relatively closed (in terms of response) due to higher THC and lower MT concentrations in gills ([MT]G) and visceral mass

([MT]VM) (which corresponds to the most contributive variables on axis 1). Clams from Andernos were characterized by high MT concentrations in gills and, at a lesser extent, a higher percentage of granulocytes (%GR) than in other populations. Arguin cockles were isolated because they exhibited the lowest THC and the highest MT concentrations in visceral mass. As well as for stress PCA, supplementary PCA were performed among each species in order to represent intraspecies differences in terms of responses. PCA-5 extracted to factors accounting for 77.5% of inertia and demonstrated that most points closely related to MT concentrations corresponded to winter and spring of both years in relation with the main gametogenesis period of clams (Figure III.11). Nevertheless, even if summer and autumn points were less related to MT concentration it is important to note that even in these periods, MT concentration in clams from Arcachon were higher than those in clams from Landéda. Clams from Landéda were highly related to higher THC and lower MT concentrations in gills, which constituted the most contributive variables on axis 1. Finally, PCA-6 extracted two components explaining more than one half of the total variance (60.5%) and confirmed that cockles from Arguin exhibited the

129

Chapitre III. Approche multistress in situ

3,0

1,0 2,5 A-P-Sp1 2,0

[MT] 0,5 VM 1,5 B-P-Au1 1,0 A-P-Su2

[MT]G B-P-Sp2 THC A-P-Wi1 A-P-Au2 0,5 0,0

A-P-Sp2 B-P-Su2

. 2 : : : 22,51% 22,51% 22 . .

. 2 : 22,51% : 2 . 0,0 Fact Fact B-P-Sp1

Fact -0,5 A-P-Au1 -0,5 A-P-Su1 B-P-Au2 -1,0 B-P-Wi2 B-P-Su1 %Gr A-P-Wi2 -1,5 -1,0 -2,0 -1,0 -0,5 0,0 0,5 1,0

Fact. 1 : 54,97% -2,5 -4 -3 -2 -1 0 1 2 3 4 Fact. 1 : 54,97%

Figure III.11. Principal Component Analysis 5 (PCA-5). Loading of (a) MT concentrations and hemocyte counts and (b) clams populations (A-P: clams from Arcachon ; B-P: clams from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008) on the two first principal components (F1-F2).

4,0

1,0 3,5 B-C-Wi1 3,0 THC [MT]G 2,5 0,5 2,0

B-C-Sp2 %Gr 1,5

0,0 1,0

[MT]VM . 2 : : 25,18% 2 .

0,5 A-C-Sp1 A-C-Su1

. 2 : 25,18% 25,18% : : ..2 2

Fact B-C-Su2

A-C-Au1 A-C-Au2 B-C-Au2 Fact Fact 0,0 B-C-Sp1 B-C-Au1 -0,5 A-C-Sp2 -0,5 A-C-Wi1 A-C-Su2 B-C-Su2 -1,0 A-C-Wi2 B-C-Wi2

-1,0 -1,5

-1,0 -0,5 0,0 0,5 1,0 -2,0

Fact. 1 : 35,32% -2,5 -4,0 -3,5 -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 Fact. 1 : 35,32%

Figure III.12. Principal Component Analysis 6 (PCA-6). Loading of (a) MT concentrations and hemocyte counts and (b) cockle populations (A-C: cockles from Arcachon ; B-C: cockles from Landéda ) among the eight seasons (Wi: winter; Sp: spring; Su: summer; Au: autumn; associated with 1: 2007 or 2: 2008) on the two first principal components (F1-F2).

130

Chapitre III. Approche multistress in situ

lowest THC and the highest MT concentrations in visceral mass, mainly during the first year (Figure III.12). Two points (B-C-Wi1 and B-C-Sp2) of cockles population from Landéda were closely related to the THC variable, indicating particularly high values in these seasons whereas other points of cockles from Arguin were almost related to lower %GR and lower MT concentration in visceral mass.

3.3. Population dynamic parameters

Mean condition index of cockles were similar at both locations as were condition index of Manila clams (Figure III.13). However, as we have previously mentioned bivalves at Landéda displayed higher fluctuations suggesting more intense gametogenesis. However, since no mature individuals were found at Landéda throughout the field survey (data not shown) and no recruitment occurred for both species, variation of CI might be only due to variation of food quality/quantity in Landéda leading to accumulation or loss of organic matter in bivalves. Condition index (CI) of adult cockles from Arguin fluctuated with reproduction status but was rather low (median = 58 ‰, max = 73 ‰) compared to Landéda (median = 67 ‰) (Figure III.13) and other sites where it easily reached 100-130 ‰ (Guillou et al., 1990). However, this CI can display lower values (median ≈ 50 ‰) in other systems like in the contaminated Oued Souss estuary (Bergayou et al., 2009). In clams from Andernos and Landéda, condition index median was 48 and 57 ‰ with a maximum of 65 and 78 ‰, respectively, while it reached 131-142 ‰ in other sites (Laruelle et al., 1994).

Population structure (year average) differed between the four studied populations (Figure III.13). Cockles from Arguin displayed a recruitment peak, a mean abundance of 130 ind.m-2 and an asymptotic 40-mm shell length (Figure III.13A). Recruitment peak during both investigated year reached about 45 juveniles.m-2 for the 2-mm shell length class when it reached several thousand (Ducrotoy et al., 1991). As well, mean abundance (130 ind.m-2) was moderate compared to other populations where it attained more than 1000 ind.m-2 (Bachelet et al., 1992; Gam, 2008; Jensen, 1992). Conversely, asymptotic length was in the high range (40 mm) of what is generally expected (Bourget and Brock, 1990; Gam et al., 2009). At Landéda, the introduced population was dense in terms of adult population (146 ind.m-2) and asymptotic shell length displayed high values but recruitment was null or did not succeed, both years (Figure III.13B). For clams from Andernos, all measured population parameters were low as already described in this area (Dang et al., 2009b): condition index median was 48 ‰ with a maximum of 65 ‰

131

Chapitre III. Approche multistress in situ

Coque - Arcachon 40 2 40 Coque - Landéda 2 Mean abundanceAbondance moy.= = 130 130 ind.m -2ind.m A Mean abundanceAbondance moy.= = 146 146 ind.m ind.m-2 B 35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0

1 4 7

1 4 7

10 13 16 19 22 25 28 31 34 37 40

10 13 16 19 22 25 28 31 34 37 40

Palourde - Arcachon Palourde - Landéda 20 2 20 2 Mean abundanceAbondance moy.= =46 48ind.m -2ind.m C Mean abundanceAbondance moy.= = 3638 ind.m ind.m-2 D 18 18

16 16 14 14

12 12 10 10

8 8

6 6 4 4

2 2

0 0

1 4 7

1 4 7

10 13 16 19 22 25 28 31 34 37 40 43 46 49

10 13 16 19 22 25 28 31 34 37 40 43 46 49

100 a 90 a,b 80 a,b CI – Cockle ARC (A) CI – Cockles LAN (B) 70 b Median = 58 Median = 67 60 Min = 49 Min = 57 50 Max = 73 Max = 94 40

30 CI – Clam ARC (C) CI – Clam LAN (D) ConditionIndex 20 Median = 48 Median = 57 Min = 43 Min = 39 10 Max = 65 Max = 79 0 Cockle ARC Cockle LAN Clam ARC Clam LAN

Figure III.13. Size structure and condition index (CI) of bivalve populations: cockles from Arguin (A), cockles from Landéda (B), clams from Andernos (C) and clams from Landéda (D). ARC: Arcachon; LAN: Landéda. Letters a, b indicate statistical differences between populations (Kruskal-Wallis test, p < 0.05).

132

Chapitre III. Approche multistress in situ

when it reached 131-142 ‰ in other sites (Laruelle et al., 1994) (Figure III.13). Recruitment consisted in few individuals per square meter, against 2000-2500 juveniles.m-2 in Musaka shoal, Japan (Ohba, 1959).Average density was 48 ind.m-2 when population can reach 300-900 ind.m-2 (Breber, 2002; Ponurovsky and Selin, 1988; Toba, 2004). Finally, Manila Clam at Arcachon gathered rather small adults (most of them < 40-mm shell length) (Figure III.13C). Asymptotic length hardly reached 44 mm, with a potential of 48-85 mm (Bourne, 1982). However, asymptotic length can be worse as in Musaka shoal with 35 mm (Ohba, 1959). All size classes were represented by few individuals.

Finally, population dynamic parameters of Manila clams that were introduced at Landéda are very bad: condition index and shell length were rather like at Andernos. Abundance was artificial (transplanted population) and recruitment was null or did not succeed (Figure III.13D).

4. Conclusion The aim of the present analysis was to assess the average status of four different populations (including two bivalve species, the cockle Cerastoderma edule and the Manila clam Ruditapes philippinarum) in terms of assimilated stressors (parasites, metals), levels of response at molecular (metallothionein) and cellular (haemocytes) scales, and population health (sustainability). By pooling seasonal results from a two-year survey, a baseline of the different measured parameters was obtained for different situations. Even if the present study suffers from numerous limitations, i.e. many other stressors can be measured (PAH, pesticides, other metals and pathogens) and the number of analysed biomarker was also reduced compared to batteries that can be employed (Auffret et al., 2006; Cravo et al., 2009; Roméo et al., 2003), this work gave a comprehensive view of the different populations. By computing literature data, it is completely workable to quantify the level of identified stressors (low, medium, high) and to determine in which extent these levels may lead to severe physiological perturbation and/or mortality events. In addition, population dynamic parameters (CI, growth, density, recruitment) gave us a good idea of the population health. Thus, both informations were relevant to identify ―hot spots‖ where stressors but also other environmental features pose problems to the survival of populations. For buried bivalves, Landéda appeared as a ―hotspot‖ where stressors and other environmental features pose problems to the survival of populations. At Andernos, environmental features have certainly a role (Dang et al., 2009b) but some identified stressors may contribute to deficiency in condition index, growth, recruitment.

133

Chapitre III. Approche multistress in situ

Thus, this site did not provide the best conditions for Manila clams but local population seem to develop resistance patterns. Banc d‘Arguin was the most appropriate spot allowing rather prosperous bivalve population dynamics. However, this situation encouraged the development of many benthic suspension feeders leading in fine to intra and interspecific competition which may explain the moderate state of population. Nevertheless, the determination of adaptative responses in relation to identified stressors remained a black box in the present study. Interpretation of molecular (MT) and cellular (hemocytes) response levels of each population was complex and required critical precautions, especially when literature data were not available to realize comparison. For example, concerning immune responses, several questions remained such as whether variations of hemocytes activities/densities observed in field were a cause or a consequence of an identified disease. Stresses sometimes produce an early transient decrease of THC, but it is not clear whether they lyse or leave the haemolymph in an orchestrated response (Hooper et al., 2007). These parameters are also very sensitive to physical factors like temperature (Hégaret et al., 2003; Yu et al., 2009) or salinity (Reid et al., 2003). It is consequently perilous to make direct links between stressors and bivalves molecular and cellular responses in such average analysis. Examination of fluctuations by season-per-season processing at individual scale within a site may be necessary in order to go further in such interpretation.

Despite many limitations, these averaged results provided potent baselines parameters for multistress experiments in order to better understand the interactions between pollutants, micro- and macroparasites, in bioaccumulation and infestation processes, and associated cellular and sub-cellular responses such as gene expression, immunity and proteomic processes. Multistress experiments working on metals, trematodes, bacteria and their interactions under controlled conditions of laboratory may also give us some clues to determine: - The importance of biological cycle on bivalve responses to multiple stress and whether such variations may partially mask the effects of moderate levels of contaminants and pathogens. - Whether local environmental conditions may induce adaptative response in natural populations. Finally, one of the main perspectives of this Multistress analysis corresponds to a more detailed approach at seasonal scale but also at individual scale with and between sites using the same data base in order to understand biological mechanisms variability in relation to season and species life-cycle.

134

Chapitre III. Approche multistress in situ Conclusion chapitre III

L‘objectif de suivi était de tenter d‘évaluer et de mettre en rapport la santé des populations (paramètres de dynamique) avec les stress subis (métaux, parasites) et la réponse apportée. Les mollusques bivalves disposent en effet de mécanismes leur permettant de lutter contre des agressions, dont la spécificité fait parfois débat (métallothionéines) (Amiard et al., 2006), tandis que d‘autres paraissent plus généralistes (système immunitaire) (Allam et Paillard, 1998). Cette étude a permis de décrire un certain nombre de situation plus ou moins favorables pour les bivalves en termes de dynamique de populations et de stress présents. Quatre situations sont envisagées :

Situation 1 - Coques d’Arcachon (Banc d’Arguin). Les coques sont parasitées par les trématodes, qui constituent apparemment le stress dominant du site. Cependant les niveaux d‘infestation, au vu de nos suivis hôtes-parasites et de la bibliographie, sont au-dessous des seuils critiques. Les systèmes de défense étudiés (métallothionéines, paramètres immunitaires) ne semblent pas sollicités et in fine la santé de la population est satisfaisante (recrutement, croissance, indice de condition).

Situation 2 - Coques de Landéda. Le même stress apparent (trématodes) que dans le cas précédent affecte les coques, mais également à un niveau relativement inoffensif. Pourtant, les niveaux moyens des paramètres immunitaires et des métallothionéines (MT) sont différents. Par ailleurs, la santé de cette population de coque est médiocre (recrutement quasi nul). Ce scénario suggère que l‘échec du maintien de la population est plus lié à un environnement naturel peu favorable qu‘à un stress type polluant ou agent pathogène.

Situation 3 - Palourdes d’Arcachon (Andernos). Alors que les parasites trématodes les plus opportunistes sont présents dans l‘environnement, les palourdes semblent résistantes puisqu‘elles en sont quasi-dépourvues. En revanche, ces palourdes sont largement infectées par Perkinsus olseni et ont un niveau de contamination au cadmium significatif. Elles présentent des concentrations en MT plus élevées et une résistance, impliquant entre autres cette réponse protéique, doit être mise en place puisque la population de palourdes se maintient.

Situation 4 - Palourdes de Landéda. Ces palourdes sont sévèrement atteintes par la Maladie de l‘Anneau Brun et ne présentent manifestement pas de résistance comme cela est observé chez des palourdes d‘autres origines. Le paramètre immunitaire THC (Total Hemocyte Count : nombre

135

Chapitre III. Approche multistress in situ total d‘hémocytes) paraît particulièrement sollicité. Il est cependant possible que ce paramètre soit plutôt sensible à l‘environnement global puisqu‘il est aussi remarquable chez les coques du même site. Le défaut de recrutement des palourdes de Landéda est donc peut-être, comme chez les coques, lié à des d‘autres paramètres physico-chimiques ou hydrodynamiques.

En complémentarité de cette approche, des expériences en laboratoire apparaissent nécessaires afin d‘appréhender l‘implication et la réponse des paramètres observés (détoxication, réponse immunitaire) dans des conditions contrôlées.

136

Chapitre IV. Approche multistress en laboratoire

Unités expérimentales

137

138

Chapitre IV. Approche multistress en laboratoire

Introduction

L‘échantillonnage in situ a été réalisé sur 2 ans à partir de populations naturelles et transplantées afin de suivre nos espèces modèles dans des situations caractérisées par des stress contrastés et multiples. Une telle analyse in situ a permis d‘avoir une vision représentative des stress et de l‘état de santé des populations associées. Toutefois, cette réponse adaptative a été étudiée dans un contexte naturel au niveau duquel la réponse à l‘agression des xénobiotiques et des organismes pathogènes se mêle à la variabilité des mécanismes liés au cycle de vie des espèces (reproduction, saisonnalité) et aux facteurs environnementaux (variables climatiques, conditions physicochimiques…). Ainsi, malgré une bonne représentativité, la complexité de l‘ensemble de ces variables rend l‘interprétation des paramètres moléculaires et cellulaires observés périlleuse. De la même manière, il est difficile de décrypter in situ les possibles interactions entre ces sources de stress ainsi que la réponse des organismes à ces interactions spécifiques. Les expériences en laboratoire se situent à une échelle plus réductionniste à travers le contrôle des conditions d‘expérimentation qui permettent une meilleure reproductibilité et simplicité d‘interprétation. De telles expériences en conditions contrôlées permettent de décrypter finement les processus d‘adaptation et de défense pouvant être activés chez une espèce suite à l‘exposition à un polluant et/ou un pathogène, et de quantifier les rôles directs et indirects de chaque facteur. Au sein des disciplines de l‘écotoxicologie et de la parasitologie, la réponse des organismes a le plus souvent été étudiée dans des contextes expérimentaux à un facteur (polluant ou pathogène). L‘objectif des travaux présentés dans ce chapitre est d‘étudier les mécanismes de défense des bivalves dans un contexte associant une, deux ou trois sources de stress, ceci à travers des approches expérimentales multifactorielles permettant une approche intégrée de leurs interactions. Le choix de multiplier les facteurs de stress, avec notamment une source de contamination bactérienne, a pour but de s‘approcher autant que possible des conditions de l‘environnement tout en identifiant la responsabilité de chaque facteur: les polluants sont en effet rarement présents de manière isolée dans les milieux littoraux caractérisés par une multitude d‘organismes pathogènes de nature diverse et variée.

139

Chapitre IV. Approche multistress en laboratoire

Plusieurs expériences double stress (polluant x parasite) ont été tout d‘abord réalisées afin d‘appréhender à la fois : - L‘influence du cycle de vie d‘une espèce (Chapitre A) : à différentes périodes de l‘année, dans des conditions physiologiques et de température identiques à celles du terrain, la réponse des organismes va-t-elle être dominée par le cycle de vie ou la source de perturbation appliquée ? - L‘influence de l‘histoire de vie d‘une population (Chapitre B): une même espèce soumise à des sources de stress multiples va-t-elle être affectée et répondre de manière identique quel que soit son passé ?

Enfin, des expériences à trois facteurs (métal x parasite x bactérie) ont été menées sur les deux espèces de bivalves afin d‘appréhender le type d‘interaction pouvant découler de ces sources de stress : additive ? synergique ? antagoniste ? (Chapitres C-D) De telles interactions pourraient-elles être intuitives ou à l‘inverse contre-intuitives ? L‘étude comparative de ces deux espèces repose sur la nécessité d‘appréhender les différences interspécifiques chez les bivalves afin de déterminer si une généralisation d‘un ou plusieurs processus est envisageable au sein d‘un groupe taxonomique.

L‘ensemble de ces travaux sont présentés dans ce chapitre sous la forme d‘articles récemment soumis ou déjà publiés dans les journaux Environmental Science and Pollution Research (sous presse), Environmental Pollution (soumis) et Marine Pollution Bulletin (sous presse), ainsi que sous la forme d‘un chapitre de livre publié dans Marine Pollution New Research (Ed. T.N. Hoher).

140

A. Influence du cycle de vie sur la réponse aux stress multiples

Seasonal modulated MT synthesis in the cockle (Cerastoderma edule) after parasite and cadmium contamination.

I. Paul-Pont; M. Baudrimont; P. Gonzalez and X. de Montaudouin Published in: Marine Pollution: New Research (Ed. T.N. Hoher). Nova Publisher, 323-338 (2008)

141

142

Chapitre IV. Approche multistress en laboratoire

Abstract

Among mechanisms involved in resistance to metallic stress, production of metallothioneins (MT) is often studied. The induction of these low molecular weight cytosolic proteins was investigated in the cockle Cerastoderma edule after parasite infection and cadmium contamination in relation to life cycle and target organ. Cockles were experimentally infested by trematode larvae (Himasthla elongata) and exposed to 15 µg Cd.L-1 for 7 and 14 days. The experiment was performed every two months at field temperature. Infection success and sexual maturity were assessed prior to MT and Cd measurements in two tissue samples: the gills and the visceral mass. Results suggested marked differences at an organ level both in Cd bioaccumulation and MT response. The effect of parasite on MT levels varied hugely between months leading to the absence of any significant effect of this factor over the whole year. Therefore, cadmium contamination and host reproductive cycle appeared to be the main factors which regulated the MT response.

Keywords: Cerastoderma edule, metallothionein, cadmium, parasite, temperature, reproduction.

143

144

Chapitre IV. Approche multistress en laboratoire

INTRODUCTION

Coastal zone ecosystems have strong ecological and socio-economical value. However, in addition to numerous anthropogenic disturbances, these regions are often contaminated by pollutants and especially metals, because of their proximity to heavily industrialized metropolitan centers. Bivalve molluscs have been frequently shown to accumulate high concentrations of metal owing to their large water filtration capabilities, their capacity to concentrate pollutants and their sedentary life (Cheggour et al., 2001). These properties lead to their use in many monitoring networks. Molluscs have several subcellular systems for accumulation and regulation of the excess of essential and non-essential metals. Among toxic metals detoxification mechanisms, metallothionein (MT) are common among marine organisms and have been reported for approximatively fifty aquatic invertebrate species, notably molluscs and crustaceans (Tanguy et al., 2003; Amiard et al., 2006). MT are low molecular weight cysteine rich proteins, which have a high affinity for metals such as Hg, Cd, Cu, Zn and Ag. Therefore, MT are inducible by a large number of metals, especially cadmium (Cd), and consequently they have been proposed as a specific biomarker for metal exposure. In many species, induction of MT synthesis by metals has been demonstrated, but literature data presents many contradictions and inconsistencies in MT induction. It seems that numerous confounding factors unrelated to metal contamination could interfere with the mechanisms regulating MT biosynthesis such as exogenous factors (temperature, physical stress) and endogenous factors (size, hormones, second messengers, sexual maturity and ) (Amiard et al., 2006). Few studies have reported parasitism as a stress factor in the literature while they are widely spread in mollusc populations. Furthermore, parasites are known to cause a large amount of damages leading sometimes to host population mortalities (Desclaux et al., 2004). Parasites using bivalves as their first intermediate host impose severe pathogenicity due to their reproductive stages in host‘s tissues. In contrast, parasites using bivalves as second intermediate host seem to be more benign because it has been demonstrated that bivalves tolerate high concentrations of metacercariae in their tissues. Nevertheless, these parasite life stages are considered harmless under standard conditions but could act as detrimental stress agents under extreme environmental conditions (Desclaux et al., 2002). Therefore they can interfere with physiological responses, in particular on MT response under metal contamination. MT production alteration by parasites (using cockles Cerastoderma edule both as first and second intermediate host) in the whole organism of cockles has been previously observed

145

Chapitre IV. Approche multistress en laboratoire

(Baudrimont et al., 2006; Baudrimont and de Montaudouin, 2007). Results showed first a higher MT content in parasitized cockles during the rest period whereas their MT levels are reduced during gametogenesis compared with healthy cockles. Secondly, in trematode infected cockles, a significant alteration of the protective effect of MT under metal contamination in comparison with unparasitized cockles was observed. However, these authors conducted their experiment at fixed temperature, sometimes with values inconsistent with the physiological status of the bivalves in the environment at a given period. Therefore, the objective of this study was to assess the interaction between parasitism and reproductive cycle on Cd bioaccumulation and MT synthesis in the cockle C. edule. Experiments were performed every other month at field temperature and responses were observed in two different organs of the cockle: the gills and the visceral mass, in regard with the strong variations of MT response between both tissues.

MATERIALS AND METHODS

Sampling site Adult cockles (mean shell length = 29.8 mm) were collected at the Banc d‘Arguin, a moderately sheltered inter-tidal sand flat. It is the most oceanic part of the Arcachon Bay, an Atlantic coastal lagoon in south-west France. It is moderately affected by freshwater input and the salinity ranges between 32 and 35psu, all year round (de Montaudouin et al., 2000). The surface water temperatures range from 9°C in winter to 21.1°C in the summer. The sediment is largely dominated by medium sand (median grain-size ≈ 350μm). Banc d‘Arguin is also a national nature reserve with several marine bird species and many fishes which could act as potential hosts for trematode parasites (de Montaudouin et al., 2000). This site was chosen particularly due to the low Himasthla elongata prevalence, the parasite species used in this experiment, and to the absence of heavy metal contamination in cockles.

Biological models The common cockle, Cerastoderma edule is one of the most common bivalve molluscs of the Western European coasts, spreading from the Barents Sea to Morocco (Tebble, 1966). It is an infaunal filter-feeder of intertidal flats. Rygg (1970) described it as a euryhaline marine species with a considerable tolerance for extreme salinities, in particular when juvenile. Temperature and food availability influence spawning of C. edule. Consequently, spawning period vary following

146

Chapitre IV. Approche multistress en laboratoire locations, although it usually occurs between March and October (Guillou et al., 1990, 1992). In Arcachon Bay, cockles spawn twice a year, with a main peak in spring and a lesser one in autumn (Guillou et al., 1990).

The parasite Himasthla elongata is a digenean trematode. Himasthla genus is the most prevalent in Arcachon Bay (de Montaudouin et al., 2000) but H. elongata remains very rare at our collecting site. Himasthla elongata uses the periwinkle Littorina littorea as first intermediate host before parasitizing its second intermediate host: the cockle C. edule. These parasitized cockles are then eaten by seagull which acts as the final host of H. elongata. The main concentrations of metacercariae in the natural environment are most commonly found in the foot (de Montaudouin et al., 1998). From the various parasites which use the cockle as second intermediate host, H. elongata was chosen for three reasons: (1) it can be easily manipulated in order to experimentally infect cockles (Wegeberg et al., 1999); (2) H. elongata is scarce in Arcachon Bay allowing a distinction between controls and experimentally parasitized cockles; (3) H. elongata induces moderate deleterious effects on cockles which allow metal contamination of parasitized individuals without too much mortality.

Cockle treatment Cockles were sampled in January, March, May , September and November 2006, transferred into experimental units (four cockles / EU) in synthetic sea water (instant ocean salts: 30 psu) controlled at in situ temperature, and left for 3 days to acclimatize before starting the experiment. Cadmium contamination was achieved by adding every day small quantities of a 100 mg Cd/L solution into the EUs, in order to obtain a mean and constant Cd concentration of 15 μg/L. In order to obtain the parasites for infection, infected periwinkles (Littorina littorea) were placed in small quantities of water at room temperature (~20 °C). This induced cercariae emission into the water from which aliquotes of ten were sampled with a pipette and released into the EUs until each contained between 80 and 100 cercariae. Four different exposure conditions were applied to EUs: uncontaminated-healthy (control), Cd contaminated-healthy, uncontaminated-parasitized and Cd contaminated-parasitized. For each exposure condition (duration = 7 and 14 days), samples were carried out in triplicate leading to 24 EUs. During experiments, several physico- chemical parameters were daily measured as mentioned in Table 1. The photoperiod was 12h per day and the cockles were fed twice a week with diatoms (Thalassiosira weissflogii).

147

Chapitre IV. Approche multistress en laboratoire

Table 1. Mean cadmium concentrations and physico-chemical parameters in the water column of the experimental units. Abiotic parameters (mean ± SD) Cd (µg/l) T°C S (‰) pH O2 (mg/l) January 15.2 ± 0.2 10.7 ± 0.2 29.8 ± 0.3 8.3 ± 0.1 9.7 ± 0.3

March 15.5 ± 0.4 9.3 ± 0.2 29.0 ± 0.4 8.4 ± 0.1 10.6 ± 0.1

May 15.3 ± 0.1 14.3 ± 0.1 29.2 ± 0.3 8.4 ± 0.1 10.5 ± 0.5

September 15.3 ± 0,2 21.3 ± 0.5 30.4 ± 0.5 8.3 ± 0.1 10.2 ± 0.4

November 14.8 ± 0.3 15.6 ± 0.4 30.4 ± 0.3 8.3 ± 0.1 10.5 ± 0.3

Cockles were sampled at random and removed from the EUs after the acclimatization period, seven and fourteen days under their corresponding stress conditions. As soon as the cockles were removed from the EUs, the foot was dissected out of the individuals and was squeezed between two glass slides. Using a stereomicroscope the number of H. elongata cysts were counted and recorded. The whole body was then removed from the shell and dissected further to separate the gills and the visceral mass. As small section of both organs was removed to be used in genetic analyses (not detailed in this study), the rest of the organs were then placed into separate polyethylene bags (Whirl-Pak) under N2 at -80°C to minimize metallothionein oxidation until they were ready to be analyzed. The shell of each individual was dried and weighed.

For the experiment conducted in July, the infestation success by Himasthla elongata was so low compared with all the other experiments that we decided to take of this month of analysis.

Reproductive state Condition index (CI) was calculated for cockles from the experiment. This CI was calculated in order to verify the homogeneity of reproductive state between cockles from different treatments. It also allowed us determining any accumulation or loss of organic matter in relation to reproduction. The CI was calculating using the following equation:

CI = (Wet soft-body weight / Wet shell weight) x 1000

The use of the wet weight of the tissues instead of the dry weight is explained by the need to analyse MT on the same samples. Consequently the values cannot be directly compared with theoretical values of CI in others studies. Nevertheless, we verified that the ratio between wet weight and dry weight was constant and equal to 5.6 ± 0.8 (n=120, data not shown).

148

Chapitre IV. Approche multistress en laboratoire

Cadmium and metallothionein determination The organs were thawed for analysis and dried on absorbent paper. Each organ was separated into two: one part (minimum of 50 mg) was used to determine MT concentration and the rest of the organ (all remaining tissue – weight recorded) allowed the determination of the Cd concentration. For Cd determination, the tissues were dissected, dried with absorbent paper, weighed and digested in high purity nitric acid at 95°C for 3h. Samples were diluted using de- ionised water (MilliQ®, Bedford, MA, USA) to obtain a final acid concentration between 2 and 20% and analysed by electrothermic atomic absorption spectrophotometer with Zeeman correction (EAAS Thermoptec M6Solaar). The detection limit was 0.1 µg.L-1 Cd. To ensure the accuracy and reliability of this determination technique, certified biological reference samples were systematically analyzed (TORT-2: Lobster Hepatopancreas and DOLT-2: Dogfish liver, Reference Material for Trace Metals, NRC-CRNC, Ottawa, Canada).

The levels of total MT proteins in gills and visceral mass were determined by mercury-saturation assay as previously described, using cold inorganic mercury (Dutton et al. 1993 ; Baudrimont et al. 2003 ; Gonzalez et al., 2006). This technique relies on mercury‘s affinity to metallothioneins and its ability to displace other heavy metals which can also fix to metallothioneins (Baudrimont et al., 1997a, Amiard et al., 2006). The denaturation of non-MT proteins was performed with trichloroacetic acid and excess Hg not bound to the MTs was removed by scavenging with lyophilised pig hemoglobin (Sigma) prepared in 30 mM Tris-HCl buffer (pH 8.2 at 20°C). The final supernatant was then quantitatively recovered and used for Hg determination by flameless atomic absorption spectrometry (AMA 254, Altec, Prague, Czech Republic). The detection limit was estimated at 1 ng Hg. Owing to the fact that the exact quantity of Hg binding sites per MT molecule is unknown for this species, MT concentrations cannot be directly expressed in nmol MT.g-1 (wet wt), but in nmol Hg.g-1 (wet wt).

Data analysis Due to strong heterogeneity of variances, non parametric statistics were used. A Burt table was made in order to perform with the same matrix, non-parametric analysis of variance and multiple correspondence analysis. The first step to build a Burt table consisted in transforming quantitative variables in qualitative variables. For each parameters, the quartiles were calculated in order to obtain four classes, from 1 (the lowest quartile) to 4 (the highest quartile). The six variables were : cadmium concentration in gills (1=[0-124]; 2=[124-664]; 3= [664-1730] and 4=[1730-3984] ng

149

Chapitre IV. Approche multistress en laboratoire g-1 fw), cadmium concentration in the visceral mass (1=[0-43]; 2=[43-201]; 3= [201-567] and 4=[567-1737] ng g-1 fw), MT concentration in gills (1=[0-2]; 2=[2-4]; 3= [4-6] and 4=[6-37] nmol sites Hg g-1 fw), MT concentration in the visceral mass (1=[0-12]; 2=[12-23]; 3= [23-39] and 4=[39-93] nmol sites Hg g-1 fw), fresh condition index (1=[0-190]; 2=[190-221]; 3= [221- 262] and 4=[262-542]), parasite intensity (1=[0]; 2=[0-7]; 3= [7-50] and 4=[50-110] metacercariae per cockle). A fixed four-ways ANOVA was performed to compare MT concentrations with the above transformed data. Factors were 1) exposure duration (two levels: t7 and t14), 2) months (five levels: January, March, May, September, November), 3) Cd contamination (two levels: contaminated and non-contaminated), 4) parasite infection (two levels: parasitized and non-parasitized). A fixed three-ways ANOVA was performed to compare Cd concentrations (as previously described but without the cadmium contamination factor). A correlation test was also completed between the six measured variables. Eventually, a multiple correspondence analysis was performed in order to assess the main reactions of cockles in terms of Cd accumulation, parasite infection and MT synthesis. This procedure smoothes results but appears as the best way to highlight the major trends in such a complex experimental design. Statistical analysis was performed using STATISTICA 6.0 for Windows ‗Statsoft, Inc. 2002).

RESULTS

Infection success The cercariae larvae of Himasthla elongata enter in the cockle using the inhalant siphon. They possess penetrating glands which they use to enter C. edule tissue where they transform into metacercariae, the latent stage of the parasite.Wegeberg et al. (1999) noted that in natural infections metacercariae of H. elongata were exclusively found in the foot whereas in experimental infection a small part of cysts were also found in the siphon and the mantle. Therefore, the presence of H. elongata metacercariae in the siphon and the mantle seems to be a laboratory artefact. The presence of cysts in the mantle and the siphon was also verified in these experiments (data not shown). However, for these experiments, parasites were only counted in the foot leading to an underestimation of the total parasite intensity. There is no metacercariae of H. elongata in the foot of healthy cockles both for contaminated and no contaminated conditions (figure 1). Infection success for experimentally infected cockles is significantly homogeneous between treatments (contaminated and not contaminated units) and

150

Chapitre IV. Approche multistress en laboratoire seasons. However, a high inter-individual variability was observed which could be due to differences in bivalve behaviour related to burying in sediment, ventilation rate and valve opening.

90 January 80 March 70 60 May 50 September 40 November 30 20 Parasiteabundance 10 0 Control Parasite Cd Cd + Parasite

Experimental conditions Figure 1. Parasite abundance (mean number of metacercariae per cockle ± standard error, n = 12) for the different months in function of experimental conditions.

Reproductive state For each month, no difference in condition index (CI) was found between treatments and between time of experiment (data not shown). The evolution of the reproductive status showed a small decrease in the condition index in May (figure 2). Moreover January and November showed the maximal values. A high CI corresponds to a late gametogenesis state with an important accumulation of gonadic elements (Guillou et al., 1990). Discharge of accumulated material owing to partial or total spawning explains the decrease of CI in May.

151

Chapitre IV. Approche multistress en laboratoire

350 300 250 200 150 100

Condition(CI) Index 50 0 January March May September November Months

Figure 2. Monthly evolution of condition index (± standard error) corresponding to average values of cockles from T7 and T14, and for all treatments (n = 24).

Cadmium bioaccumulation Globally, cadmium concentrations in gills remained similar between t7 and t14 and was not affected by seasonality and parasite infection (p>0.05, Table 2). The mean concentrations were 1928 ng g-1 fw in contaminated EU and 159 ng g-1 fw in non-contaminated EU. In the visceral mass, the mean Cd concentration was three times lower than in gills. It varied between months, with the lowest values in November-January with a mean concentration of 369 ng g-1 fw against 865 ng g-1 fw in March-May-September (p<0.001). Morever, there was a clear accumulation pattern from t7 to t14 in most treatments (x 1.3) (p<0.001).

Table 2. Results of the three-ways ANOVA comparing the effect of fixed factors on Cd concentrations in gills and visceral mass. Factors were 1) experiment duration (two levels: t7 and t14), 2) months (five levels: January, March, May, September, November), 3) parasite infection (two levels: parasitized and non- parasitized).

Gills Visceral mass Treatments df F p df F p Experiment duration (a) 1 0.94 0.338 1 14.40 < 0.001 Months (b) 4 1.09 0.375 4 8.75 < 0.001 Parasitism (c) 1 0.94 0.338 1 0.40 0.531 a x b 4 3.74 0.011 4 1.65 0.181 a x c 1 0.24 0.630 1 1.60 0.213 b x c 4 0.79 0.536 4 0.15 0.962 a x b x c 4 0.38 0.820 4 0.85 0.502 Error 40 40

152

Chapitre IV. Approche multistress en laboratoire

Metallothionein Synthesis Metallothioneins displayed a more complex pattern. In gills, MT concentrations varied following the different investigated factors (Table 3). The strongest effect was related to contamination when Cd-contaminated cockles reached 7 nmol sites Hg g-1 fw against 3 nmol sites Hg g-1 fw in non-contaminated cockles (p<0.001). This accumulation took more time during spring (March- May). Indeed, it was only observed at t14 (interaction ‗Experiment duration x months‘ p<0.05). A temporal variability separated the highest values in September-November-January (7 nmol sites Hg g-1 fw) against the lowest values (5 nmol sites Hg g-1 fw) in March-May (p<0.001). Mean MT concentration increased between T7 (3.7 nmol sites Hg g-1 fw) and T14 (7 nmol sites Hg g-1 fw). Parasitism generated a significant interaction with experiment duration (p<0.01). The non-parasitized cockles synthesized more MT between t7 and t14 (x 2.0) than the parasitized cockles (x 1.8) (p<0.01).

Table 3. Results of the four-ways ANOVA comparing the effect of fixed factors on MT concentrations in gills and visceral mass. Factors were 1) experiment duration (two levels: t7 and t14), 2) months (five levels: January, March, May, September, November), 3) Cd contamination (two levels: contaminated and non- contaminated), 4) parasite infection (two levels: parasitized and non-parasitized). A fixed three-ways ANOVA was performed to compare Cd concentrations (as previously described but without the cadmium contamination factor).

Gills Visceral mass Treatments df F p df F p Experiment duration (a) 1 30,02 < 0,001 1 16,33 < 0,001 Months (b) 4 12,14 < 0,001 4 5,69 < 0,001 Cd contamination (c) 1 100,45 < 0,001 1 524,48 < 0,001 Parasitism (d) 1 0,88 0,352 1 0,33 0,565 a x b 4 2,61 0,042 4 1,43 0,233 a x c 1 9,45 0,002 1 0,33 0,565 b xc 4 0,98 0,422 4 2,72 0,035 a x d 1 7,88 0,006 1 0,93 0,339 b x d 4 2,30 0,066 4 0,61 0,656 c x d 1 0,16 0,690 1 0,04 0,848 a x b x c 4 2,66 0,039 4 0,61 0,656 a x b x d 4 1,27 0,290 4 1,02 0,403 a x c x d 1 0,88 0,352 1 0,33 0,565 b x c x d 4 1,86 0,126 4 0,69 0,604 a x b x c x d 4 2,39 0,057 4 0,43 0,789 Error 80 80

153

Chapitre IV. Approche multistress en laboratoire

Mean MT concentration in the visceral mass was much higher than in gills (x 5) and was mainly influenced by Cd contamination (32 nmol sites Hg g-1 fw in contaminated cockles vs 24 nmol sites Hg g-1 fw in non-contaminated cockles) (p<0.001). May displayed the lowest MT concentrations (19 nmol sites Hg g-1 fw) compared to November-January when the highest values were reached (38.0 nmol sites Hg g-1 fw) (p<0.001). Between t7 and t14, mean MT concentration increased from 26 to 30 nmol sites Hg g-1 fw (p<0.001). Parasitism did not interfere (p>0.05). When pooling data, correlations were significant between Cd concentration in gills (Cd concentration in the visceral mass) and MT concentration in gills (Figure 3). There was another correlation between MT concentrations in gills and the visceral mass. Finally, the MT concentration in the visceral mass was inversely correlated to the temperature.

1,0 CI:4 Mo:November [Cd]VM:3 T:2 Mo:January [MT]VM:4 Mo:May 0,5 * [MT]G:1 CI:3 ¤ T:1 [MT]G:4 [Cd]G:1 Time:7 [Cd]G:3 [Cd]VM:2 PI:3 *PI:4 0,0 NC [MT]VM:2 H ¤ P [MT]G:3 C Mo:March PI:1-2 ¤ [Cd]VM:1 [MT]VM:3 Time:14 [MT]G:2 CI:2 [Cd]G:4 [Cd]G:2 ¤ -0,5 [MT]VM:1

CI:1 [Cd]VM:4

Dimension 9,8%2: inertia ofDimension -1,0

-1,5 Mo:September T:3 * -1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 Dimension 1: 13,4 % of inertia Figure 3. Multiple correspondence analysis discriminating investigated factors on a factorial plan. Symbols represent each variables: ( ) months; ( ) time with 7 or 14 correspond to the time of experiment; ( ) Cd contamination ‗C‘ for contaminated and ‗NC‘ for no contaminated; ( ) Parasite infection with ‗H‘ for healthy and ‗P‘ for parasitized; ( ) temperature with T1 for low temperature (9- 11°C), T2 for intermediate temperature (14-15°C) and T3 for high temperature (21°C); ( ) cadmium concentration in gills with ‗1‘=[0-124], ‗2‘=[124-664], ‗3‘= [664-1730] and ‗4‘=[1730-3984] ng g-1 fw); ( ) cadmium concentration in the visceral mass with ‗1‘=[0-43], ‗2‘=[43-201], ‗3‘= [201-567] and ‗4‘=[567-1737] ng g-1 fw; ( ) MT concentration in gills with ‗1‘=[0-2], ‗2‘=[2-4], ‗3‘= [4-6] and ‗4‘=[6- 37] nmol sites Hg g-1 fw; (¤) MT concentration in the visceral mass with ‗1‘=[0-12], ‗2‘=[12-23], ‗3‘= [23-39] and ‗4‘=[39-93] nmol sites Hg g-1 fw; ( ) fresh condition index with ‗1‘=[0-190], ‗2‘=[190-221], ‗3‘= [221-262] and ‗4‘=[262-542]; ( ) parasite intensity with ‗1‘=[0], ‗2‘=[0-7], ‗3‘= [7-50] and ‗4‘=[50- 110] metacercariae per cockle.

154

Chapitre IV. Approche multistress en laboratoire

The multiple correspondence analyses gave an overview of the major constraints and MT‘s answer in this four-factor design experiment. Axis 1 clearly discriminated the Cd-contaminating treatment as the most efficient in term of MT stimulation. The positive part of the axis isolated cockles which undergone Cd contamination and displayed the highest values in term of Cd accumulation and, in a lesser extent, in term of MT concentrations. The negative part of the axis isolated the opposite situation. Axis 2 rather discriminated the temporal variability. The positive part isolated the coldest temperature with the highest CI, stimulating (when cockles are contaminated) MT synthesis in the visceral mass (but not in gills which seem independent of these factors). In absence of Cd, MT synthesis remains low and independent from CI and temperature. Parasitism treatment had a very small impact on MT concentration (close to axes‘ origin) but was a little closer to high MT values (right part of axis 1) than non-parasitized treatment (left part of axis 1).

DISCUSSION

Gills epithelium is in direct contact with the water and constitutes the main target organ for metal uptake and accumulation. Consequently, this tissue is often considered as the most suitable for contamination analyses after direct exposure. Cadmium concentrations in gills were three-fold higher than in visceral mass. Nevertheless, the gills have been noticed as a storage organ for a short time, whereas absorption through the digestive gland leads to an accumulation of toxic metals for a longer time (Amiard et al., 1989). Similarly, after Cd exposure of clams (Ruditapes decussatus) for 14 days, Bebianno et al. (1994) observed a transport of Cd from gills to digestive gland for storage. These hypotheses could explain that no effect of time or season were found in the gills whereas the visceral mass showed both variations of Cd accumulation between months and a clear accumulation pattern between T7 and T14. Temperature is estimated to be a major environmental factor influencing Cd accumulation in molluscs (Fischer, 1986). Modification of environmental temperature could influence the metabolism of aquatic organisms (Mubiana and Blust, 2007; Serafim et al., 2002). Owing to this fact, the lowest value of Cd accumulation should be found in March where the water temperature was very low. However, in visceral mass, the lowest values were found in January and November both for T7 and T14. These lowest values of Cd bioaccumulation appeared when the CI was the highest. These decreases of Cd concentrations in the visceral mass coincided with the increase of

155

Chapitre IV. Approche multistress en laboratoire weight due to the increase of CI. The visceral mass containing gonads and digestive gland —the organ where the main part of the absorption occurs— undergoes continuous weight changes due to the existence of cycles of accumulation or loss of food or reserve substances (Ibarrola et al., 1998). Thus, the increase of visceral mass weight leads to a diluting effect of Cd concentration. Results showed no significant effect of parasitism on Cd bioaccumulation both for gills and visceral mass after 7 and 14 days of exposure. Baudrimont and de Montaudouin (2006) showed similar results for Cd-contaminated cockles experimentally parasitized by H. elongata. The difference of Cd accumulation between parasitized and unparasitized cockles was not apparent after 7 days. Nevertheless, this difference became significant after 14 days. This inconsistency could be due both to the sampling period of cockles and the tissue observed. Indeed, the authors worked on the whole soft body of cockles whereas the present study showed results in different organs of the cockle. Due to its relative weight and its relative Cd and MT burden, the visceral mass could be representative of the whole body of cockle (data not shown). However, a direct comparison was not appropriate. Moreover, cockles were sampled in the beginning of February 2002 and were placed at 15°C in the laboratory. This procedure was really different from here where cockles stayed at field temperature during all the experiment. In the same study, Baudrimont and de Montaudouin worked with another parasite, Labratrema minimus, which used cockle as first intermediate host. After 7 days of Cd-exposure at 15 µg/L, they showed a significant accumulation of Cd in parasitized cockles in comparison with healthy ones. In this case, fundamental differences existed between the two types of parasite and parasite infection. As first intermediate host, the cockle suffered much more than as second intermediate host (Desclaux et al., 2002). L. minimus reproduced asexually into sporocysts which invaded the whole body of cockles. This process leads to important weakening of bivalve and a massive cell destruction which considerably facilitated the passive diffusion of cadmium. Since the infection was more severe, the difference of Cd bioaccumulation was logically more obvious.

The strongest effect on metallothionein synthesis was related to cadmium contamination both for visceral mass and gills. Although the biological functions of MTs have not been fully elucidated, they are involved in essential metals (Cu, Zn) homeostasis. They also play an important role in detoxification of toxic elements such as cadmium (Cd). Indeed, it was well documented in the literature that metal contamination, and notably cadmium contamination, leads to a strong induction of metallothionein in aquatic organisms (Bordin et al., 1997; Amiard et al., 2006).

156

Chapitre IV. Approche multistress en laboratoire

The levels measured in the visceral mass of C. edule controls were greater than that seen for the gills which corresponds to the findings of many other bivalve studies (Amiard et al., 2006). The visceral mass represents not only the largest proportion of the bivalve body mass, but is also the main site for MT biosynthesis (Baudrimont et al., 1997a). Therefore, this organ presents a high relative weight and high Cd and MT contents (data not shown), and can be representative of the whole organism. In visceral mass, MT concentrations were significantly higher in January and November whereas the concentrations of Cd in visceral mass were the lowest. Furthermore, the lowest values of metallothionein concentrations appeared in May both in gills and visceral mass. These results were correlated to the variations of CI corresponding to the pre and post-spawning period of cockles. Numerous studies demonstrated that MT seasonal variations were most likely related to the reproductive cycle of bivalves. MT levels show a peak during maximum gametogenesis followed by a strong decrease during the post-spawning period (Cheggour et al., 2001; Baudrimont et al., 2006; Baudrimont et al., 1997b; Serafim and Bebianno, 2001). This decrease in MT concentration could be due both to the deficit of reproductive hormones, to a global weakening of organisms after spawning and to the release of MT-rich gametes, as previously observed in Crassostrea gigas and in Corbicula fluminea (Baudrimont et al., 2007). These facts could partially explain the negative correlation found between temperature and MT concentration in this organ. Evolution of the reproductive state of the cockles used in these experiments did not follow the evolution of temperature as well as other field studies. This result was clearly showed in the ACM in which Axis 2 of the positive part isolated the coldest temperature with the highest CI. When temperature increased from March to May, we did not found an increase of CI corresponding to a gametogenesis evolution. In contradiction, CI decreased on May indicating a partial or total spawn. In a same manner, despite an increase of temperature for the month of September, low CI of cockles showed that they stayed in a relative post-spawning status. An apparent negative correlation between MT concentration and temperature was also found in a study of seasonal variation of MT in Corbicula fluminea after the post spawning period. MT concentrations seemed to be more related to reproductive phenomena than to temperature variation (Baudrimont et al., 1997). Globally, parasitism treatment had a very small impact on MT synthesis both in gills and visceral mass after 7 and 14 days of exposure. Baudrimont and de Montaudouin (2006) showed also no significant differences on MT response after 7 days of Cd-exposure between parasitized and unparasitized cockles (with the two different parasites). With H. elongata, marked differences on

157

Chapitre IV. Approche multistress en laboratoire

MT responses clearly appeared after 14 days of Cd-exposure. In the present study, it is important to note that MT concentrations significantly increased between 7 and 14 days of experiment for all treatments. These increases both in control group and in contaminated group - without any increase of Cd concentration - revealed the clear apparition of stressful conditions in units between these two exposure times. These increases were more marked in gills than in visceral mass so it could be suggested that detrimental conditions appeared at the water-sediment interface probably because of releases of waste products or faeces. Gills were the first organs concerned in relation to their direct contact with environmental surrounding. Thus, this experimental artifact, leading to the increase of MT concentrations whatever the experimental conditions between T7 and T14, completely masked an eventual difference of MT response between parasitized and unparasitized cockles after 14 days of exposure. In gills, the increase of MT concentrations was higher for unparasitized cockles than parasitized ones both in contaminated and no contaminated units. This fact might be sign possible weakening of bivalves such as what it was found in previous study of Baudrimont and de Montaudouin (2006). Therefore, in these stressful circumstances it was quite risked to interpret MT response at T14.

CONCLUSION

The present study suggested no significant effect of parasitism on Cd accumulation whatever seasons in the cockle Cerastoderma edule. In terms of metallothionein response, Cd contamination was the most discrimated factor, with marked differences at the organ level. No significant effect of parasite infection on MT concentrations was found troughout the whole year. This fact was in contradiction with others studies which analysed impact of parasitism on MT response. This inconsistency could be due to the strong effect of Cd contamination which might mask a potential impact of parasite infection. It could be also due to the important variability of metallothionein response between months. In addition, results evidenced a strong seasonal modulation of MT concentrations mainly due to the host reproductive processes. Consequently, cadmium contamination and reproductive cycle seemed to be the major factors in metallothionein regulation during this one-year study.

158

Chapitre IV. Approche multistress en laboratoire

Conclusion chapitre IV-A Influence du cycle de vie sur la réponse aux stress multiples

Cette expérience avait pour but de déterminer l‘influence du cycle de vie sur la réponse des organismes soumis à des stress multiples, ceci à travers une expérience double stress réitérée à différentes périodes de l‘année, dans des conditions physiologiques et de température identiques à celles du terrain. Cela constituait une originalité forte car dans de nombreux cas expérimentaux la température est fixée et ne correspond pas forcément à une réalité vis-à-vis de l‘état physiologique du bivalve, notamment par rapport à son état reproductif. Une forte variabilité saisonnière des concentrations en Cd et en MT dans les tissus de la coque est mise en évidence alors que le succès d‘infestation parasitaire demeure similaire quelque soit la saison considérée. Alors que l‘infestation parasitaire ne semble avoir aucun effet tant sur la bioaccumulation de Cd que sur la synthèse des MT, l‘exposition au Cd constitue un facteur important influençant la réponse de cette protéine. De plus, l‘influence de la reproduction chez la coque constitue un des principaux facteurs expliquant la forte variabilité saisonnière de cette protéine. En effet, l‘évolution de l‘indice de condition en lien avec la mise en place des processus de gamétogénèse et de ponte chez la coque apparaît fortement corrélée aux variations saisonnières de MT. Ces résultats expliquent la variabilité intra-annuelle observée lors du suivi in situ sur la valeur de ces paramètres (Chapitre III). Ainsi, le cycle de reproduction et, dans une moindre mesure, l‘exposition au cadmium semblent constituer les principaux facteurs intervenant dans la régulation de la réponse des métallothionéines au cours de l‘année. L‘effet du parasitisme, plus ténu, apparaît dans des cas d‘interactions avec le Cd, mais aussi avec les saisons. Ce type d‘expérience souligne la complexité des mécanismes régulant la synthèse de cette protéine, a forciori dans des situations de multistress. L‘influence de l‘histoire de vie, liée au contexte environnemental, sur la réponse au double stress « Cd x parasite » sera l‘objet du chapitre suivant.

159

160

B. Rôle de l’histoire de vie sur la réponse aux stress multiples

How life-history contributes to stress response in the Manila clam Ruditapes philippinarum.

I. Paul-Pont, X. de Montaudouin, P. Gonzalez, P. Soudant and M. Baudrimont. Accepted in: Environmental Science and Pollution Research (in press).

161

162

Chapitre IV. Approche multistress en laboratoire

Abstract

Within the last decade numerous studies have investigated the role of environmental history on tolerance to stress of many organisms. This study aims to assess if Manila clams Ruditapes philippinarum may react differently to cadmium exposure and trematode parasite infection (Himasthla elongata) depending on their origin and environmental history in Arcachon Bay (France). Clams were exposed to Cd (15 µg.L-1) and parasites (25 cercariae per clam), alone or in combination, at 15°C under controlled laboratory conditions for 7 days. Metal accumulation and success of parasite infestation were examined, also physiological parameters such as metallothionein response and hemocyte counts and activities (phagocytosis, oxidative burst, viability, adhesion). Sensitivity of Manila clams to both stressors differed from one site to another suggesting local adaptation of populations. Clams from the more parasitized site presented better resistance to trematodes than the others in terms of first line defense, i.e. avoidance of infection. On the other hand, clams that adapted to chronic Cd contamination showed better detoxification mechanisms, both in a faster transfer of metal from gills to visceral mass and in a higher metallothionein baseline, than clams which had never experienced Cd contamination. Finally, hemocyte concentration and viability differed between clam origin site, highlighting the fact that populations living in different environments may adapt their physiological and biochemical responses to environmental stressors. It is therefore important to be cautious when extrapolating results from field studies of one species and one site, if the life history of the organisms is not taken into account.

Key words: Ruditapes philippinarum / cadmium / trematode parasite / metallothionein / immunity / life-history

163

Chapitre IV. Approche multistress en laboratoire

164

Chapitre IV. Approche multistress en laboratoire

Introduction

Individuals within populations often display a range of responsiveness to local environmental conditions. Recent field and laboratory studies have documented the fact that environmental history can considerably affect physiological and biochemical processes, leading to a change in sensitivity to stressful conditions (Blackmore and Wang 2002; Boisson et al. 1998; Shi et al. 2003). Among sources of perturbations affecting marine populations in coastal ecosystems, two of the most important are the release of pollutants, such as trace metals, and the presence of pathogenic organisms. Marine bivalves, like the Manila clam Ruditapes philippinarum, have been successfully employed throughout the world as biological monitors of metal pollution in coastal waters. Most bioaccumulation studies have used animals collected from a single population in the field but few studies have examined the differences in metal accumulation among different field populations and the underlying physiological and biochemical mechanisms (Blackmore and Wang 2003; Wang and Rainbow 2005). Klerks and Weis (1987) indicated that the responses of animals to metal pollution is the result of either physiological acclimatation, representing a degree of tolerance obtained during a relatively short-time exposure, or adaptation, representing the genetically based resistance obtained by whole life-time exposure. Specifically, recent studies have demonstrated that populations exposed to metals show a better tolerance to metal contaminations than populations which have never experienced selected contaminants (Bervoets et al. 2004; Lapanje et al. 2008; Ross et al. 2002). Similar local adaptation was also highlighted in populations experiencing pathogenic organisms, such as toxic algae (Hégaret and Wikfors 2005) or trematode parasite (Bryan-Walker et al. 2007) compared to populations without prior exposures.

Moreover, relationships between pollution and pathogens in aquatic organisms are highly complex and have received increasing attention during the last two decades in ecological parasitology and ecotoxicology, since under natural conditions organisms rarely experience these factors separately (MacKenzie et al. 1995; Sures 2008; Valtonen et al. 1997). In fact, pollutants impair organisms‘ endocrine and immune systems and could thus increase host sensitivity to pathogens and diseases (Pipe and Coles 1995).

In this context, we focused on two populations of Manila clam sampled from different sites in Arcachon Bay (France), Andernos and Arguin, where environmental features, including stressors,

165

Chapitre IV. Approche multistress en laboratoire are different. In Andernos, sediments were characterized by metal contamination, particularly Cd, and wild populations of Manila clam sustain perkinsosis, Brown Ring Disease (BRD) (Dang et al. 2009b; Lassalle et al. 2007) and Brown Muscle Disease (BMD) (Dang et al. 2008). In contrast, Arguin was characterized by a relatively low level of stress in terms of disease (with a lack or a low prevalence and intensity of perkinsosis, BRD and BMD (Dang 2009)) and metal contamination (lack of any trace metal contamination in sediment). In Arguin, the main potential source of perturbation was the high abundance of trematode parasite since thirteen species were encountered in the cockle Cerastoderma edule (de Montaudouin et al. 2009). Trematodes exhibit a heteroxenous life cycle, i.e. they need several host species. Parasite sexual reproduction takes place in a vertebrate host (final host) and leads to the production of eggs that are released into water and transform into miracidium larvae which infests the first intermediate host (always a mollusc). They grow and produce a new type of larvae, the cercariae, which are shed in the water and will infect a second intermediate host. Cercariae, transformed into metacercariae, complete their cycle when their host is eaten by the final host. In the present work, we focused on the metacercarial stage in the second intermediate host which can be deleterious when metacercariae are numerous (Desclaux et al. 2004; Gam et al. 2009). In the sympatric bivalve Cerastoderma edule, a mean prevalence of 90% was found at Arguin (Gam et al. 2009). Previous studies showed that the Manila clam, an introduced species, was resistant to trematodes (Dang et al. 2009a) but there are some reports of such infestation in their native area (Lee et al. 2001; Ngo and Choi 2004). In Arcachon Bay, Manila clam infestation by trematodes is also very low, but slightly higher at Arguin (Dang et al. 2009a). Nevertheless, in a context of poor environmental conditions or poor individual health parasite infestation could increase considerably even in non- specifically sensitive species.

In studies focusing on potential local adaptation of different populations several endpoints have been analyzed, such as feeding activity, growth, survival, EC50, tolerance index, etc. Few studies have focused on physiological parameters at cell and molecular level. In a context of metal contamination and pathogen infection, two kinds of physiological parameters appeared to be relevant in relation to sensitivity of populations to stresses. First, metallothionein (MT) responses could be observed both at gene and protein levels in relation to its implication in the detoxication of toxic metals and its ability to respond to inflammatory processes (Amiard et al. 2006; Gagné et al. 2007). Previous field and experimental studies highlighted a significant modulation of these

166

Chapitre IV. Approche multistress en laboratoire proteins by digenean parasites, both in non-contaminated and Cd-contaminated bivalves (Baudrimont et al. 2006; Baudrimont and de Montaudouin 2007). In addition, the main line of defense in molluscan shellfish to toxic or pathogenic agents is the immune system (Cheng 1996) mediated by chemical agents and specialized cells called hemocytes. Therefore, immune response was analyzed by determining hemocyte concentration and activity (phagocytosis, oxidative burst, viability, adherence capability) in order to assess the modulation of immune function in relation to the bivalve environmental history. The aim of this study was to assess the impact of cadmium contamination, trematode parasite infestation and potential interactive effects on two populations of Manila clam from two different origins (Arguin and Andernos) in Arcachon Bay. Accordingly, a laboratory experiment was performed with these sources of stress applied singly or in combination. Success of infestation, Cd accumulation and several physiological and biochemical parameters (MT concentrations, hemocytes counts and activities) were analyzed after seven days of experiment.

Materials and methods Sampling site and bivalve contamination Arcachon Bay is a 180-km2 lagoon on the Atlantic coast of southwest France (Figure 1). The inner part comprises 110 km2 of mud flats colonized by a vast seagrass bed. The molluscan biomass on these flats is dominated by the Manila clam Ruditapes philippinarum (Blanchet et al. 2004). Two sites were sampled: Andernos, which is located in the inner part of the bay, and Arguin which is much closer to the ocean (Figure 1). Environmental variables characterizing the two sites are summarized in Table 1. Andernos

Table 1 Environmental variables of the sampling sites (Dang, 2009)

Mean Characteristics Arguin Andernos Median size sediment (µm) 360 163 Sediment silt (%) 3,5 14,5 Sediment organic matter (%) 1 3,3

Sediment temperature ( C) 16,1 15,8 Arguin Tidal level (m) 2 2,1 Salinity (psu) 34,2 30 Cadmium concentration in 1,9 41,7 sediment (ng.g-1, dry weight)

Figure 1. Location of sampling sites (Arguin and Andernos) in Arcachon Bay. Adapted from Blanchet et al. (2005).

167

Chapitre IV. Approche multistress en laboratoire

Sixty clams (38.3 ± 0.3 mm shell length, mean ± SD) from each site were collected during early spring 2007. Prior to the experiment, homogeneity of condition index (CI) between both clam populations was verified. CI was measured on 10 clams per population as the ratio of the flesh dry weight (in mg) on the shell dry weight (in g) (Walne & Mann, 1975). Clams were maintained in the laboratory for three days prior to the experiment, and they were fed with a concentrated diatom culture (Thalassiosira weissflogii) (≈ 105 algae.mL-1 per experimental unit) every two days during both acclimatization and exposure periods in experimental units (EU). Diatom culture was carried out at the laboratory under controlled conditions, allowing us to obtain an axenic culture without trace metal contamination. The experiment was conducted in the laboratory using glass aquaria of 12x12x24 cm filled with 1 L of synthetic sea water (Instant Ocean, salinity = 30‰) aerated by a diffuser system. Plastic coverings were placed over the inside walls to avoid cadmium contamination and to minimize cadmium adsorption on the walls. Four clams were placed in each unit. Temperature was fixed at 15°C (± 0.7°C, mean ± SD), pH was regularly monitored and remained stable (8.1 ± 0.3, mean ± SD) and the photoperiod was fixed at 12h using a timer and artificial light sources. Four experimental conditions were performed: control; parasitized condition (H); Cd-contaminated condition (C) and a parasitized and Cd-contaminated condition (CH). Three EU replicates were carried out for each experimental condition, giving a total of 24 EU running simultaneously for a period of seven days.

Infestation and contamination protocol

Among the trematode species that occur in Arcachon Bay, Himasthla elongata was collected because 1) many bivalves are its potential second intermediate host (Werding 1969), 2) it is a widespread species (de Montaudouin et al. 2009), 3) experiment process is controlled (Wegeberg et al. 1999) and 4) our host populations are free of this trematode (Dang et al. 2009a). The Himasthla elongata life cycle involves marine birds as final host, periwinkles (Littorina littorea) as first intermediate host and bivalves (mostly cockles and mussels) as second intermediate host. Himasthla elongata cercariae were collected from infected periwinkles (Littorina littorea). Infected snails were kept at 15°C and fed with macroalgae (Ulva spp.), then placed in a dish with seawater at room temperature; this condition induces the release of cercariae into the water which were transferred in batches of ten into the EU until each EU contained 100 cercariae. To maintain a constant infection pressure, cercariae were added at T0 and after 1, 3 and 4 days.

168

Chapitre IV. Approche multistress en laboratoire

-1 A single nominal Cd concentration at 133 nM (corresponding to 15 μg Cd.L added as CdCl2) was selected. After the first contamination (T0) and in order to maintain constant contamination throughout the exposure period, aqueous Cd solution was added daily, adjusted according to the decrease in Cd concentration in each EU measured as described hereafter.

Sampling procedure

After 7 days (T7), one clam was removed from each EU (nEU = 3). Hemolymph was withdrawn from the adductor muscle of each individual with a 25-gauge needle attached to a 1-mL sterile syringe (TERUMO). A recorded quantity of hemolymph (500 µL) was used for immunological analyses. Foot and mantle were dissected and squeezed between two sterilized glass slides. H. elongata metacercariae were rapidly counted under a stereomicroscope. The remaining tissues were dissected to separate gills and visceral mass. A small section of both organs (>50 mg) was removed for metallothionein analyses and placed into polyethylene bags (Whirl-Pak) under N2 atmosphere at -80°C to minimize metallothionein oxidation. The rest of both organs was used for cadmium analysis.

Metal and metallothionein quantification

Cadmium quantification was carried out for the water in the EU and in the different organs of each clam. Sampling water was acidified at 2% with nitric acid (Fluka; Buchs, Switzerland. 65%

HNO3) before Cd analysis. Tissues were digested in 1 to 1.5 mL (depending on tissue weight) of nitric acid at 100°C for 3 h to dissolve metals in the liquid for quantification. The liquid underwent a six-fold dilution with ultrapure water (MilliQ®, Bedford, MA, USA). Cd concentrations were determined by electrothermic atomic absorption spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. For Cd determination in EU, water samples were diluted and acidified at 2% HNO3. The analytical methods were simultaneously validated for each sample series by analyzing standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Values were consistently within the certified ranges (data not shown).

MT protein concentration in the gills and the visceral mass was determined by mercury-saturation assay, using cold inorganic mercury (Dutton et al. 1993; Baudrimont et al. 2003). Metallothionein analysis was conducted on three replicates per exposure condition from a

169

Chapitre IV. Approche multistress en laboratoire minimum of 50 mg of fresh tissue, the saturation assay being repeated twice per sample. This technique is based on the quantification of Hg bound to the saturated MT. The denaturation of non-MT proteins was performed using trichloroacetic acid, and the excess Hg not bound to the MT was removed by scavenging with lyophilized beef hemoglobin (Sigma) prepared in 30 mM Tris-HCl buffer (pH 8.2 at 20°C). The final supernatant was then quantitatively recovered and used for Hg determination by flameless atomic absorption spectrometry (AMA 254, Altec, Prague, Czech Republic). At the same time, three reference samples or ―blanks‖ were prepared to monitor the Hg complexation efficiency of the hemoglobin. The mean of the three blank values at each analytical run was deducted from the Hg burdens measured in each sample. The detection limit was estimated at 0.01 ng Hg. As the exact number of Hg binding sites per MT molecule was unknown for this species, MT concentrations were expressed in nmol sites Hg.g−1 (wet weight). Immunological analyses As soon as the hemolymph was withdrawn it was passed through an 80-µm mesh and stored temporarily in microtubes on ice to avoid cell clumping. Immunological parameters measured were (i) total hemocyte count (THC / number of hemocytes.mL-1); (ii) hemocyte viability (percentage of viable hemocytes); (iii) hemocyte reactive oxygen species (ROS) production that measures the potential to kill non-self particles; (iv) phagocytosis activity (ingestion of fluorescence beads by hemocytes), and (v) adhesion of hemocytes. Hemocyte analyses were adapted from Delaporte et al. (2003), Lambert et al. (2003), Soudant et al. (2004) and were performed using a FACScalibur flow-cytometer (BD Biosciences, San Jose, CA).

Statistical analyses All variables were analyzed using a three way ANOVA (independent variables: origin of clams, metal contamination and parasite infection) in order to determine possible interactive effects between clam origin, Cd contamination and parasite on clams‘ responses. Previously, normality was assumed and homogeneity of variance was verified with Cochran‘s test (data were log- transformed when homogeneity of variance was not achieved). Percentages of phagocytic and dead hemocytes were also arcsin-transformed according to homogeneity requirements. Whenever ANOVA was significant, differences between treatments were determined by comparison of means using the Tukey test. Statistical analysis was performed using STATISTICA 7.1 software for Windows.

170

Chapitre IV. Approche multistress en laboratoire

Results Success of infestation All control clams were totally free of the parasite H. elongata. The success of infestation was null for clams from Arguin. On the contrary, clams from Andernos were moderately infected after seven days of experiment with 3 ± 1.1 and 1.3 ± 0.9 metaceracriae in the soft tissue of clams submitted to parasite (condition H) and to both cadmium and parasite (CH), respectively. Cadmium contamination was not shown to have any effect on experimental infestation success by H. elongata (p > 0.05) (Table 2).

Cadmium accumulation

Non-contaminated bivalves displayed low concentrations of cadmium in the gills after 7 days of experiment, with similar values whatever the origin site (Figure 2a). In contaminated clams, Cd concentration was significantly lower in the gills of clams from Andernos than in those from Arguin (x 1.5). Finally, no effect of experimental parasite infection was observed on Cd accumulation in either population. Indeed, for a given population, Cd levels in clams contaminated with Cd alone (C) and in combination with trematode (CH) were similar (p > 0.05).

Non-contaminated clams from Andernos presented higher Cd concentrations in visceral mass than clams from Arguin (Figure 2b). The means comparison Tukey test revealed that clams from Andernos presented higher Cd concentrations in visceral mass than clams from Arguin, in non- contaminated conditions (p = 0.0013). In addition, in contaminated conditions, clams from Andernos accumulated significantly more cadmium in the visceral mass than clams from Arguin (x 6) (Figure 2b) (p < 0.001). Parasites did not interfere in Cd accumulation in the visceral mass of either population (p > 0.05). a b 900 1800 Arguin Arguin 800 1600 Andernos Andernos 700 1400

600 1200

1, 1, dw

1, dw - - 500 1000 400 800

300 600

[Cd] in ng.g in [Cd] [Cd] in ng.g in [Cd] 200 400 100 200 0 0 Control H C CH Control H C CH Experimental conditions Experimental conditions Figure 2. Cadmium concentrations (mean ± SE) in gills (a) and visceral mass (b) of clams after 7 days of experiment. Experimental conditions: Control, parasitized with Himasthla elongata (H), Cd contaminated (C), parasitized and Cd contaminated (CH). 171

Chapitre IV. Approche multistress en laboratoire

Table 2 Results of three-way ANOVA comparing the effect of Origin (O), cadmium contamination (C) and parasite infestation (P) (fixed factors) on different variables measured in Ruditapes philippinarum after 7 days of experiment.

Source of Source of Variable variation dF F p Variable variation dF F p Trematode Origin 1 4.558 0.049 Total Origin 1 0.080 0.777 infestation Cadmium 1 0.372 0.550 hemocyte Cadmium 1 4.640 0.048 Parasite 1 4.558 0.049 count Parasite 1 1.360 0.262 O*C 1 0.372 0.550 O*C 1 1.940 0.184 O*P 1 4.558 0.049 O*P 1 2.050 0.173 C*P 1 0.372 0.550 C*P 1 9.830 0.007 O*C*P 1 0.372 0.550 O*C*P 1 2.490 0.136

Cadmium Origin 1 10.439 0.005 Viability Origin 1 10.797 0.005 Gills Cadmium 1 151.517 < 0.001 Cadmium 1 2.117 0.165 Parasite 1 0.382 0.545 Parasite 1 0.001 0.980 O*C 1 19.414 < 0.001 O*C 1 0.116 0.738 O*P 1 0.249 0.625 O*P 1 1.170 0.295 C*P 1 0.996 0.333 C*P 1 0.516 0.483 O*C*P 1 0.100 0.755 O*C*P 1 1.267 0.277

Cadmium Origin 1 23.247 < 0.001 Adhesion Origin 1 0.910 0.144 Visceral mass Cadmium 1 39.209 < 0.001 Cadmium 1 0.084 0.058 Parasite 1 0.154 0.700 Parasite 1 3.028 0.368 O*C 1 0.073 < 0.001 O*C 1 0.688 0.121 O*P 1 2.408 0.142 O*P 1 0.733 0.126 C*P 1 0.681 0.422 C*P 1 0.292 0.080 O*C*P 1 0.747 0.401 O*C*P 1 0.159 0.066

MT Origin 1 33.900 < 0.001 Phagocytosis Origin 1 0.044 0.836 Gills Cadmium 1 1.977 0.179 Cadmium 1 0.507 0.487 Parasite 1 2.147 0.162 Parasite 1 1.617 0.222 O*C 1 0.160 0.694 O*C 1 1.896 0.188 O*P 1 3.426 0.083 O*P 1 0.031 0.863 C*P 1 0.033 0.858 C*P 1 0.071 0.793 O*C*P 1 0.047 0.830 O*C*P 1 0.143 0.071

MT Origin 1 17.338 0.001 Oxidative Origin 1 1.203 0.290 Visceral mass Cadmium 1 1.371 0.260 Burst Cadmium 1 1.229 0.285 Parasite 1 7.412 0.016 Parasite 1 1.025 0.327 O*C 1 0.439 0.518 O*C 1 0.373 0.551 O*P 1 0.812 0.382 O*P 1 0.461 0.507 C*P 1 1.347 0.264 C*P 1 1.278 0.276 O*C*P 1 0.538 0.475 O*C*P 1 0.036 0.853

172

Chapitre IV. Approche multistress en laboratoire

In contrast to clams from Arguin, the visceral mass of Andernos clams was three times more contaminated than the gills, in both contaminated and non-contaminated treatments, as revealed by a three factor ANOVA test (p = 0.001) despite the fact that cadmium was applied via the direct rather than the trophic route.

Metallothionein concentrations

Clam origin had a significant effect on MT concentrations both in gills (p < 0.0001) and visceral mass (p = 0.0008) (Table 2). In the control group, clams from Andernos presented higher MT concentrations in the gills (x 6) and visceral mass (x 2) than clams from Arguin. In addition, highest MT levels were shown in clams from Andernos, whatever the treatment, both in gills and visceral mass (Figure 3). No significant cadmium effect was shown in MT response of either population (p > 0.05, Table 2). Finally, a significant parasite effect was shown in MT concentration in visceral mass (Table 2). In clams from Andernos, MT levels decreased in parasitized conditions compared to non-parasitized ones (Figure 3), as the mean Tukey comparison test revealed (control vs. H p = 0.0206; C vs. CH p = 0.0017).

a b 16 Arguin 80 Arguin Andernos 14 70 Andernos

1, 1, ww 12 1, 1, ww

- 60 -

10 50

8 40

6 30

4 20 [MT] in nmol sites sites in Hg.g nmol [MT]

2 sites in nmol Hg.g [MT] 10

0 0 Control H C CH Control H C CH Experimental conditions Experimental conditions

Figure 3. Metallothionein concentrations (mean ± SE) in gills (a) and visceral mass (b) of clams after 7 days of experiment. Experimental conditions: Control, parasitized with Himasthla elongata (H), cadmium contaminated (C), parasitized and Cd contaminated (CH).

Immune responses

In control conditions, both populations displayed similar values for THC and hemocyte viability (Figures 4, 5).

173

Chapitre IV. Approche multistress en laboratoire

No effect of treatment was shown on THC for clams from Andernos, whereas Cd contamination affected THC of clams from Arguin, decreasing in parasite treatment and increasing in non- parasite treatment (interaction Cd-parasite p < 0.05, Figure 4, Table 2).

Hemocyte viability in the control group was similar in the two populations (Figure 5), whereas the presence of parasites, Cd or both led to a decrease in hemocyte viability for the Andernos clams (from 91.7 ± 1.9 in control to 87.0 ± 2.3, 85.3 ± 3.8 and 87.0 ± 2.9 in stressed individuals from H, C and CH conditions, respectively). In clams from Arguin there was no significant difference in hemocyte viability after 7 days of cadmium and/or parasite exposure.

Cadmium and parasite in combination or separately had no significant effect on hemocyte adhesion capability (94.9 ± 1.1%, mean value ± SE), oxidative burst (128 ± 14 A.U., mean value ± SE) or phagocytosis activity (35.8 ± 2.1%, mean value ± SE) after 7 of experiment (data not shown).

4000000 100

Arguin Arguin 1)

- 3500000 Andernos Andernos 95 3000000 90 2500000

2000000 85 1500000 80

1000000 Hemocyte viability (%) viability Hemocyte 75

500000 Total hemocyte hemocyte (cell.mL count Total 0 70 Control H C CH Control H C CH Experimental conditions Experimental conditions Fig.4 Total hemocyte count (cell/mL) according to experimental Fig.5 Percentage of hemocyte viability according to condition (mean ± SE). Experimental conditions: Control, experimental condition (mean ± SE). Experimental parasitized with Himasthla elongata (H), cadmium conditions: Control, parasitized with Himasthla elongata (H), contaminated (C), parasitized and Cd contaminated (CH). Cd contaminated (C), parasitized and Cd contaminated (CH).

Discussion

This study showed that, depending on their origin and environmental history, individuals from a single species may react differently to certain sources of perturbation. Although it is not certain that clams from Arguin and Andernos are strictly separate and have not experienced genetic exchanges, i.e. form a different population, we hypothesized that, from larval recruitment to death, individuals experience contrasted environmental conditions. The bodies of water at both

174

Chapitre IV. Approche multistress en laboratoire sites are sufficiently remote to exhibit different physical parameters, like temperature and salinity (Dang 2009), different benthic communities (Blanchet et al. 2005) or planktonic assemblages (Glé 2007). It was also demonstrated that clams from Arguin and Andernos fed on different trophic sources (Dang et al. 2009c) and that they were infested, like their surrounding cockles (C. edule), by different parasite propagules (Dang et al. 2009a). Consequently, we presume that our adult individuals from Arguin and Andernos spent their life-time (38 mm corresponds to an age of 3.4 years, Dang 2009) under different stressors. Among analyzed factors in collected individuals, clams from Andernos exhibited higher concentrations of Cd (visceral mass) and MT than individuals from Arguin.

Obviously, this life experience induced contrasted reactions when clams from both origins were challenged with similar stressors. Clams from Arguin resisted better to trematode experimental infestation compared to those from Andernos. One possible explanation is that throughout their life-time clams from Arguin experienced an environment with a higher abundance of cercariae compared to what is shown at Andernos (Dang et al. 2009a; Desclaux 2003). Similar local adaptation was also highlighted in populations with the trematode parasite (Bryan-Walker et al. 2007) compared to populations without prior exposures. Intensity and abundance of experimental infestation in clams from Andernos remained low, in agreement with a previous study arguing that clam flesh was too hard for cercariae perforation (Dang et al. 2009a). Due to the low number of cysts (1-7) in Manila clam, it was difficult to assess the real impact of such infestation on this species. Moreover, within the same clam population, the similarity of the mean number of cysts per clam between experimental treatments suggested that cadmium did not modulate the infestation ability of Himasthla elongata. In our experiment, cadmium concentration was lower than that used in other studies which showed significant impact of metals, and particularly cadmium, on trematode survival, orientation behavior, encystement, transmission (Morley et al. 2003a,b, 2004, 2005). It also underlined that the sensitivity of the clam to this parasite infection - in terms of first line of defense, i.e. avoiding the parasite - was not modified by experimental cadmium contamination.

In addition, the most obvious difference between both populations was that at Andernos, clams were exposed to higher Cd concentration than at Arguin. Visceral mass of control clams from Andernos contained significantly more Cd than control clams from Arguin, despite no significant difference being shown in gills. In addition, these Cd concentrations in clams from Andernos

175

Chapitre IV. Approche multistress en laboratoire were higher in visceral mass than in gills. Suspension-feeding bivalves such as clams can accumulate metals by assimilating sediment-bound metals that are ingested (Griscom and Fisher 2004). Thus, sediment Cd contamination at Andernos (Table 1) was certainly responsible of Cd dietary exposure leading to higher Cd concentration in visceral mass than in gills.

Metallothionein level in gills and visceral mass of Andernos control clams was 5.6 and 2.5 times higher than those of Arguin clams, respectively. Such difference between controls may be explained by both external and internal factors. Chronic Cd contamination of Andernos may lead to higher MT content in natural clams from this site. Indeed, it is well known that Cd is one of the most potent inducers of MT synthesis and numerous studies have highlighted the role of MT in Cd detoxification processes in many marine invertebrates, such as crustaceans (Fraysse et al. 2006) or molluscs (Chan et al. 2002; Serafim et al. 2002). Among other external factors, the presence of diseases might in a lesser extent enhance MT synthesis throught inflammatory processes. Finally, hormones are frequently cited, especially in relation to reproductive cycle (Amiard et al. 2006; Baudrimont et al. 1997, 2006) with a peak of MT levels during maximum gametogenesis followed by a global decrease during the post-spawning period. Homogeneity of condition index between Arguin and Andernos population was performed prior to the present experiment (46.1 ± 9.6 vs. 51.7 ± 8.6 ‰ for Andernos and Arguin clams respectively, mean ± SD) and allowed us to compare both populations in regard with experimental Cd exposure.

The Arguin clams had no experience with cadmium and accumulated more in the gills than the Andernos clams (x 1.5) when suddenly challenged with this metal. They did not synthesize MT to defend themselves, whereas the Andernos clam MT content was already higher. Nevertheless, the literature data presented many contradictions and inconsistencies in MT induction and at sites where metals are bioavailable in high concentrations, with some species showing no increased MT concentrations, at least in some organs (for review see Amiard et al. 2006). More specifically, studies working on MT induction by metals in Ruditapes genus explained the lack of MT response following Cd exposure by (i) the relatively low accumulation levels reached in Arguin clams after 7 days of experiment; (ii) the presence of other cytosolic ligands such as high molecular mass protein, enzymes, amino acids and peptides like glutathione, which can interfere in the metal sequestration process (Viarengo 1989; Langston et al. 1998); (iii) the presence of antioxidizing enzymes, such as superoxide dismutase, catalase or glutathione-S-transferase, which act as first mechanism of defense against the excess of Cd in Ruditapes decussatus (Géret

176

Chapitre IV. Approche multistress en laboratoire et al. 2002, 2003); (iv) an insoluble distribution of Cd (more than 60%) leading to the unavailability of Cd for MT (Smaoui-Damak et al. 2009). Consequently, Ng and Wang (2004) hypothesized that this insoluble fraction (metal-rich granules, cellular debris and organelles) plays a role in Cd detoxification in R. philippinarum.

In contrast, the visceral mass of experimentally contaminated clams from Andernos was 4 times more contaminated than in clams from Arguin at the end of the experiment. Moreover, Cd concentrations in Andernos clams were three times higher in visceral mass than in gills, despite the fact that Cd contamination was applied directly via water (direct route) and not via ingested food (trophic route). Thus, this high Cd concentration in visceral mass of contaminated Andernos clams may be due to a faster Cd transfer from gills to the visceral mass in this population. Smaoui-Damak et al. (2004) noted that gills are short-term storage organs whilst the visceral mass is a more long-term storage organ, which suggests that these differences in Cd concentration are due to a Cd transfer from gills to visceral mass in relation to detoxification processes. Similarly, after Cd exposure of clams (Ruditapes decussatus) for 14 days, Bebianno et al. (1994) observed a transport of Cd from gills to digestive gland for storage. This transfer of toxic compounds corresponds to a detoxification process which seems to be more efficient in Andernos clams, suggesting a local adaptation of this population to chronic Cd contamination. Indeed, Shi and Wang (2004) demonstrated that a R. philippinarum population with a higher Cd tissue concentration (like the Andernos population) assimilated Cd and Zn more efficiently in association with a high metallothionein-like protein content. In an extended study conducted on mussels, these authors demonstrated a close positive relationship between Cd body burden, MT level and Cd assimilation efficiency, and concluded that nature and tissue location of metal binding ligands may explain differences of metal accumulation (Shi and Wang 2005). One possible explanation is that MT sequestered the incoming Cd, thus reducing the intracellular free Cd ion concentration and promoting a gradient for metal uptake. Our results were in agreement with these previous assumptions. Clams from Andernos quickly transferred Cd from gills to visceral mass leading to a higher and Cd accumulation at whole body scale, probably in relation with a higher MT content than clams from Arguin.

While the trematode parasite H. elongata did not modify Cd accumulation, a significant parasite effect on MT in the visceral mass was observed in the Andernos population. Induction of MT synthesis after trematode infestation has been already reported in cockles both as first and second

177

Chapitre IV. Approche multistress en laboratoire intermediate host (Baudrimont et al. 2006; Baudrimont and de Montaudouin 2007). In addition, Baudrimont et al. (2006) demonstrated lower MT concentration in parasitized individuals than in non infected ones during the gametogenesis period. This phenomenon was demonstrated in cockles infested as first intermediate host by Labratrema minimus which reproduces asexually in the host tissues (gonad, digestive gland). Thus, decrease of MT concentration had been related to gonad damage by parasites. In the present study, a significant decrease of MT concentration after parasite infestation has been also demonstrated in clams from Andernos both in non-Cd and Cd contaminated conditions. However, parasites occurred as relatively inactive metacercariae and castration mechanism was not envisaged to explain MT decrease. Nevertheless, despite the few number of cysts in clams tissues, it is possible that few metacercariae, or even just an attempt at penetration, can have a detrimental effect on some physiological parameters in a species that is not used to trematode attacks. Further experiments are required to go further in the explanation of such mechanism. Components of the hemolymph constitute the internal defense mechanism of bivalve molluscs. The most frequently measured parameters are concentration and nature of circulating hemocytes, cell viability, phagocytosis activity, and more recently, reactive oxygen species production by hemocytes (Donagly et al. 2009). It is well known that hemolymph parameters vary naturally according to environmental parameters such as temperature, food supplies, dissolved oxygen or salinity, but they also vary with season, location, and the presence of an anthropogenic stressor or disease agent (Allam et al. 2002; Chu 2000; Fisher et al. 1999; Reid et al. 2003; Soudant et al. 2004). In our experiment, there was no significant difference in any of the tested parameters between Andernos and Arguin clams in the control group. Regarding the major differences between Arguin and Andernos in terms of environmental stresses (notably parasites and metal contamination) and in terms of global clam health (i.e. population dynamic parameters such as condition index and asymptotic growth (Dang 2009)), it was really surprising to observe a similar immune background in control clams in both populations. Such results raised the question of the relevance of these immune parameters to chronic stress in environmental studies. The immunological effects of pollutants can differ considerably according to the nature and concentration of contaminants (Cajaraville et al. 1996). The effects of cadmium on immune parameters are widely described in the literature and differ between species and condition of exposure. In the present study, the decrease in hemocyte viability for clams from Andernos following Cd and/or parasite exposures compared to those from Arguin highlighted the

178

Chapitre IV. Approche multistress en laboratoire increasing sensitivity of Andernos clams under stressed conditions. Higher sensitivity of population from Andernos may also be due to its environmental history since this population was affected by Perkinsosis and in a lesser extent by BMD and BRD.

Positive correlations between hemocyte density and cadmium exposure have been shown in vivo in Mytilus edulis (Coles and Pipe 1994), Crassostrea gigas or Crassostrea virginica (Auffret and Oubella 1994; Cheng 1988). In our experiment, an interaction between Cd and parasites occurred with the THC response since Cd induced an increase in THC in clams from Arguin only in single Cd condition and not in double stress condition (CH). This increase in THC following Cd exposure is probably in accordance with the fact that these clams had never experienced this metal before. Antagonistic effect of macroparasite and metal on bivalve defense mechanisms is a process that has already been observed (Baudrimont and de Montaudouin 2007; Kim and Powell 2007). However, since no successful infestation by H. elongata was observed in clams from Arguin, the difference in circulating hemocyte concentrations in clams from Arguin between cadmium (C) and cadmium-parasite (CH) conditions remained elusive. Finally, neither parasite nor cadmium contamination had any significant effect on other immune responses in clams from Arguin. This increase in THC was the only significant variation in all the immune parameters of clams from Arguin.

Conclusion

Within the last decade, numerous studies have investigated the potential role of environmental history on tolerance to stress in several organisms such as plant (Eränen 2006), bryozoan (Piola and Johnston 2006), insect (Shirley and Sibly 1999), fish (Nacci et al. 2002; Ownby et al. 2002), crustacean (Ross et al. 2002; Lapanje et al. 2008) and mollusc (Bervoets et al. 2004). In the present study, clams from a Cd-enriched site (Andernos) showed a stronger Cd accumulation in relation to a higher MT concentration and a faster transfer of Cd from gills to visceral mass than in clams from Arguin (non-contaminated site). Thus, detoxification processes seem to be more efficient in clam population used to undergo chronic Cd exposure. Such result confirms that organisms living in contaminated environment may modify their metal handling physiology and biochemistry, leading to modification of trace metal bioaccumulation compared to population from clean site. In addition, the clam population from Arguin, which was used to experiencing trematode parasites, showed better resistance in terms of first line of defense (i.e. avoidance of

179

Chapitre IV. Approche multistress en laboratoire parasite penetration) than clams which had never encountered these parasites (Andernos). Overall, these results confirmed an increase in tolerance in populations used to undergoing particular stress (metal or pathogen).

180

Chapitre IV. Approche multistress en laboratoire

Conclusion chapitre IV-B Influence de l’histoire de vie dans la réponse aux stress multiples

L‘expérience réalisée sur des palourdes provenant de deux sites distincts au sein du bassin d‘Arcachon semble confirmer l‘influence de l‘histoire de vie d‘une population sur la sensibilité et la réponse des organismes à des sources de stress particulières, ici une exposition au Cd et aux parasites trématodes. Klerks et Weis (1987) suggéraient que la réponse des organismes à une contamination métallique était à la fois le résultat d‘une acclimatation physiologique rapide obtenue au cours d‘une exposition à relativement court terme, ainsi que d‘une possible adaptation génétique de la population face à une exposition chronique et continue. Bien qu‘il ne soit absolument pas avéré que les populations de palourdes d‘Arguin et d‘Andernos forment des populations génétiquement différentes, nous émettons l‘hypothèse que, depuis le recrutement jusqu‘à la mort de l‘animal, les organismes subissent des conditions environnementales contrastées qui ont conduit à une adaptation locale des populations par rapport aux stress ambiants. Les palourdes d‘Arguin vivant dans un environnement riche en parasite trématodes digènes semblent montrer une meilleure résistance à l‘infestation parasitaire en termes de première ligne de défense (évitement du parasite) puisqu‘elles demeurent moins infectées que les palourdes d‘Andernos suite à l‘exposition expérimentale à H. elongata. A l‘inverse, les palourdes d‘Andernos semblent posséder des mécanismes de détoxication plus efficaces à travers notamment un transfert rapide vers la masse viscérale du Cd accumulé dans les branchies, en lien avec de fortes teneurs en métallothionéines dans ce tissu. Une adaptation physiologique au stress ambiant semble être ainsi développée par les populations de palourdes du bassin d‘Arcachon, soulignant l‘importance de l‘histoire de vie dans la compréhension de la réponse des organismes face à différentes sources de perturbation. En effet, au cours des dix dernières années on observe une réelle expansion de la prise en compte de « l‘histoire environnementale » sur la tolérance des organismes face aux stress biotiques et abiotiques dans la littérature scientifique, alors que ces mécanismes semblaient globalement peu étudiés auparavant. L‘objectif suivant est, à l‘échelle d‘une même population, d‘étudier l‘impact de trois facteurs de stress sur les mécanismes de réponse de la coque dans un premier temps

181

Chapitre IV. Approche multistress en laboratoire

(Chapitre C), puis de la palourde (Chapitre D) pour aboutir à une discussion comparant la réponse des deux espèces.

182

C. Influence de l’espèce sur la réponse aux stress multiples

1. Interactive effects of metal contamination and pathogenic organisms on the marine bivalve Cerastoderma edule. Paul-Pont I., Gonzalez P., Baudrimont M., Jude F., Raymond N., Bourrasseau L., Le Goïc N., Haynes F., Legeay A., Paillard C., de Montaudouin X. Accepted in: Marine Pollution Bulletin (in press).

183

184

Chapitre IV. Approche multistress en laboratoire

Abstract

The present study evaluated the interactive effects of cadmium contamination and pathogenic organisms (trematodes Himasthla elongata and bacteria Vibrio tapetis) singularly and in combination during 7 days on the bivalve Cerastoderma edule. Some defense-related activities were analyzed such as genetic expression, metallothionein and immune responses. Trematode metacercarial infection, similar whatever the treatment, induced the strongest responses of immune parameters. Particularly, the interaction between cadmium and parasite exposures induced unusual responses on gene expression and immune responses. No effect of bacterial challenge appeared on bivalve responses, nevertheless a strong mortality of V.tapetis infected cockles occured between 7 and 14 days. Cadmium bioaccumulation was significantly modulated by both pathogenic organisms. Furthermore, an antagonistic effect of trematodes and bacteria was shown on metal bioaccumulation of co-infected cockles. These results highlighted the importance of considering the multiplicity of perturbation sources in coastal ecosystems to assess the health status of organisms.

Key words: Cerastoderma edule / metal / parasite / bacteria / metallothionein / immune response

185

Chapitre IV. Approche multistress en laboratoire

186

Chapitre IV. Approche multistress en laboratoire

1. Introduction In coastal ecosystems, aquatic organisms undergo a variety of biotic and abiotic stressors that rarely occur independently (MacKenzie et al., 1995; Sures, 2008). Among those, interactions between pollutant and pathogen in target organisms have received increasing attention over recent years. Pollutants impair the endocrine and immune systems of organisms and thus can increase host sensitivity to pathogens and diseases (Pipe and Coles, 1995). Pollutants also directly affect pathogenic organisms, changing their occurrence, distribution and virulence in a positive or negative way depending on an uncountable number of interactive variables (Sures, 2008). Previous studies demonstrated that pathogens interfere in the bioaccumulation of toxic compounds, and thus modulate the corresponding toxic effects for the host (Cross et al., 2003; Evans et al., 2001). Interactions between these stressors are dependent on both nature of pollutant and host-pathogen system and may vary from case to case, highlighting the difficulty of drawing a overarching scheme. Therefore, interactive effects on physiology and general metabolism of organisms might be synergistic, antagonistic or additive, highlighting the complexity of such interactions (Desclaux-Marchand et al., 2007; Klar and Sures, 2004). The present study focused on the impact of multiple stresses on the edible cockle Cerastoderma edule. This bivalve is abundant in tidal flats of bays and estuaries of the North African and European coast of Atlantic Ocean. It is a major prey for crustaceans, fishes and wading birds and a commercially important species (Ponsero et al., 2009). Its wide repartition, its capacity to accumulate pollutant and to host different parasites species, especially digeneans (Lauckner, 1983; de Montaudouin et al., 2009) and bacteria (Colwell and Liston, 1962), make it a relevant model for this investigation. Conversely to most studied bivalve models that are introduced species (e.g. Crassostrea gigas, Ruditapes philippinarum, Corbicula fluminea in Europe), the cockle is a native species of the north eastern Atlantic coast. Among anthropogenic pollutants, heavy metals are dominant contaminants in marine and estuarine environments and represent an increasing area of concern within environmental fields. Marine organisms may be exposed to abnormal concentrations of metals leading to numerous adverse effects on physiology and defense related activities of organisms (DNA damage, hormonal response, immunodepletion, etc…) (Pipe and Coles, 1995; Valko et al., 2005). Among heavy metals, cadmium is of particular interest because it has anthropogenic sources that are likely to impact coastal and estuarine ecosystems (Roesijadi, 1996). Cadmium is principally used in the production of stabilizers and pigments in plastic and electroplating industries and is also

187

Chapitre IV. Approche multistress en laboratoire released as a by-product from other anthropogenic activities including mining, metal industries and agriculture. Cadmium levels in the sea and particularly in estuarine sediments are increasing and the total input of cadmium to the world oceans is approximately estimated at 8000 t.yr-1 (Coles et al., 1995). Moreover, cadmium could be bioaccumulated to high concentrations in some ecologically and commercially important molluscs such as Cerastoderma edule (Anajjar et al., 2008; Jung et al., 2006). Indeed, bivalve molluscs are particularly susceptible to this form of pollution in relation to their biology and ecology (ubiquitous, sedentary, filter-feeders…).

As filter feeders, bivalves are also exposed to a constant challenge by various pathogenic and/or opportunistic bacteria naturally present in the microflora of coastal environments. Some bacteria, belonging to the Vibrio genus are often associated with diseases and mortalities in larval and adult stages of bivalves (Jeffries, 1982; McGladdery, 1999). Particularly, Vibrio tapetis which is a Gram-negative, non-sporulating motile bacterium that causes the Brown Ring Disease (BRD) in Manila clams (Ruditapes philippinarum). It was firstly reported in North West Brittany on the French Atlantic coastline in the spring and summer of 1987 before spreading to several other countries including England, Ireland, Italy, Spain, Norway and Korea. BRD affects the shells of infected individuals and is characterized by ring shaped brown conchiolin deposits on the inner surfaces of shell valves (Allam et al., 2002, 2006; Paillard et al., 1994, 2004, 2008). BRD can induce mass mortalities of Manila clams whereas some bivalve species, such as C. edule, appear to be less sensitive to infection or even completely unaffected by V. tapetis (―Healthy carrier‖) (Paillard, 2004; Allam et al., 2006). In Cerastoderma glaucum, Ruditapes decussatus and Ruditapes aureus, shell growth perturbations, similar to BRD, have been also associated with metacercarial infection of Gymnophallus fossarum (Bartoli, 1974). But to date, no synergy between Vibrios and trematodes has been mentioned in bivalves causing disease or mortalities. On the public health point of view, relationships between bivalves and macroparasites have received much less attention than relationships between bivalves and bacteria. Nevertheless, these macroparasites are often cited as factors susceptible of affecting population dynamics in marine bivalves (Jensen and Mouritsen, 1992). The cockle C. edule is known to host many different parasites species, especially digenean trematodes which are the most common metazoan parasite in marine invertebrates (Lauckner, 1983). Trematodes have a complex lifecycle with usually a vertebrate as definitive host for adult parasites, a mollusc (gastropod or bivalve) as first intermediate host and a mollusc or other vertebrate/invertebrate as a second intermediate host.

188

Chapitre IV. Approche multistress en laboratoire

The study focused on the parasite Himasthla elongata which uses cockle as second intermediate host. The cercariae larvae of Himasthla elongata possess penetrating glands which they use to perforate C. edule tissue (foot and mantle) where they transform into metacercariae, the latent stage of the parasite (Haas, 2003). Parasites using bivalves as second intermediate hosts, instead of first intermediate hosts, are considered probably as more benign (latent stage). However, they may still induce severe damages in host physiology, in particular when occurring in high numbers and in combination with poor health of organisms or in synergy with critical environmental parameters such as temperature, desiccation and pollution (Desclaux et al., 2004; Jensen et al., 1996; de Montaudouin et al., 2003). Under the combined exposure to pollutant and pathogen, the general metabolism of an organism, as measured by immunological and physiological processes, plays an important role in ensuring survival and reproductive potential at both individual and population levels. Previous studies highlighted the simultaneous effects of pollutants and biological stressors (mainly macroparasites and toxic algae) on these defense-related activities such as detoxification processes, endocrine system and immune responses (Baudrimont and De Montaudouin, 2007; Gestal et al., 2008; Hégaret et al., 2007; Sures and Radszuweit, 2007). However, studies which integrated the impact of pathogenic bacteria in interaction with pollutants are scarce, probably due to their inconspicuous character (Parry and Pipe, 2004; Wang et al., 2008). Our objective was to assess the interactive effects of cadmium contamination, bacterial challenge (V. tapetis) and trematode parasite infestation (H. elongata) on the defense related activities of the cockle Cerastoderma edule. Laboratory experiment was performed with these sources of stress applied in single, double and triple combinations. Several physiological parameters were analyzed after 7 days of experiment. Metallothionein responses both at gene and protein levels were determined in relation with its implications in detoxification of toxic metals and its ability to respond to inflammatory processes (Gagné et al., 2007). Indeed, previous experimental studies highlighted a significant modulation of these proteins by digenean parasites both in non contaminated and Cd-contaminated cockles (Baudrimont and De Montaudouin, 2007). In addition, we focused on the expression of several genes implied in mitochondrial metabolism (coxI, 12S) and in response to oxidative stress (sod). Toxicological studies at cellular level have shown that metals such as Cd can affect mitochondrial metabolism and inhibit the mitochondrial electron transfer chain and induces reactive oxygen species (ROS) production (Wang et al., 2004). These disturbances are not negligible for the survival of invertebrates especially when

189

Chapitre IV. Approche multistress en laboratoire they were exposed to other sources of stress. Finally, numerous studies have been performed to estimate the impact of V. tapetis and some protozoans such as Perkinsus spp. on immune parameters (Allam et al., 2006; Paillard et al., 1994, 2004; Villalba et al., 2004) but only a few of them integrated the interactive effect of toxic compounds on the sensitivity of organisms to these pathogens (Chu, 1999; Da Silva et al., 2008; Hégaret et al., 2007; Volety, 2008). Therefore, analysis of hemocytes concentration and activities (phagocytosis, oxidative burst, viability, adherence capability) were performed in order to assess the modulation of immune function in relation to multiple environmental stressors.

2. Material and methods

1. Bivalve treatments

Cockles (31.8 ± 0.3 mm shell length, mean ± SD) were collected at Banc d‘Arguin, a moderately sheltered inter-tidal sand flat, in Arcachon Bay (France, 44°40‘N 1°10‘W). This site was selected because of the lack of heavy metal contamination and parasite Himasthla elongata parasite in cockles and the null prevalence of BRD in the surrounding Manila clams suggesting the absence of V. tapetis (Dang, 2009). These cockles were consequently considered as ―control‖. Individuals were maintained in the laboratory for three days prior to experiments. They were fed every other day with a concentrated diatom culture, Thalassiosira weissflogii, (≈ 105 algae.mL-1 per experimental unit) during acclimatization period and exposure.

The experiment procedure followed a three-factor design with three treatments (C: cadmium; H: Himasthla elongata; V: Vibrio tapetis) and the four possible interactions (CH; CV; HV; CHV). This experiment was conducted in controlled conditions of laboratory using glass aquaria of 12x12x24 cm filled with 1L of synthetic sea water (Instant Ocean, salinity = 30,2 ± 0.1psu), aerated by a diffuser system. Plastic coverings were placed over the inside walls to avoid cadmium contamination and to minimize cadmium adsorption on walls. Four cockles per unit were introduced. The temperature was fixed at 15°C (± 0.7°C, mean ± SD), pH was regularly monitored and remained stable (8.1 ± 0.3, mean ± SD) and the photoperiod was fixed at 12h using a timer and artificial light sources. No sand was added in order to minimize bacterial contamination as the bacteria may adsorb to the sediment surface and biases bacterial load calculation at the end of the experiment. Three experimental unit (EU) replicates were carried out for each experimental condition. Therefore a total of 24 EU were running simultaneously for a

190

Chapitre IV. Approche multistress en laboratoire period of seven days. Then, one cockle per EU was sampled for the analysis whereas the remaining individuals were kept for another 7 days in order to calculate a mortality rate on a larger delay (14 days).

2. Contamination protocol

Vibrio tapetis (CTC4600) was isolated from diseased R. philippinarum (Borrego et al. 1996). Bacterial contamination was achieved by adding a culture of Vibrio tapetis to the water column to 9 -1 reach a concentration of 1.10 bacteria.mL at T0. This recreation of natural conditions was selected rather than injection of the culture into the palleal cavity or directly into the muscle to minimize stress. In order to keep a relatively constant load of V. tapetis, concentration of 1.109 bacteria.mL-1 was also adjusted after 3, 7 and 10 days.

Himasthla elongata cercariae were collected from infected periwinkles (Littorina littorea). Infected snails were kept at 15°C and fed with macro algae (Ulva spp.). For the experiment, periwinkles were placed at 20°C under artificial light source. This condition induces a cercariae release into the water from which batches of ten were transferred into the EU until each EU contained 100 cercariae. To maintain a constant pressure of infection, cercariae were added at T0 and after 1, 3 and 4 days.

-1 A single nominal concentration of Cd at 133nM (corresponding to 15 μg Cd.L added as CdCl2) was selected. After the first contamination (T0) and in order to maintain constant contamination throughout the exposure period, a daily addition of aqueous Cd solution was performed, adjusted according to the decrease in Cd concentration in each EU.

3. Sampling procedure

At T7, one cockle per EU was removed. Hemolymph was withdrawn from the adductor muscle of each individual with a 25-gauge needle attached to a 1-mL sterile syringe (TERUMO). Hemolymph was observed under a stereomicroscope and the absence of disseminated neoplasia was systematically verified in relation to its strong negative impact on host physiology (Carballal et al., 2001). An aliquot of hemolymph was used for immunological analyses. The remaining hemolymph, palleal and extra-pallial fluids were used to determine the bacterial load (Vibrio spp. and total bacteria). Foot and mantle were dissected and squeezed between two sterilized glass slides under a steromicroscope. When a cockle was already infected by the deleterious trematode

191

Chapitre IV. Approche multistress en laboratoire

Bucephalus minimus, it was eliminated. Then, H. elongata metacercariae were rapidly counted. At the end of this procedure both tissues were recovered to be used in microbiological analyses. The remaining tissues were dissected to separate gills and visceral mass. A small section of both organs was put in 500µL of RNA later (Qiagen) and kept at -80°C in order to perform genetic analyses. The rest of both organs were used for metallothionein and cadmium dosages. They were placed into polyethylene bags (Whirl-Pak) under N2 atmosphere at -80°C to minimize metallothionein oxidation.

4. Metals and metallothionein quantification

Cadmium quantification was carried out in the water and the different organs of each cockle. Tissues were digested in 1 to 1.5 mL (depending on weight of tissue) of nitric acid (Fluka; Buchs,

Switzerland. 65% HNO3) at 100°C under high pressure for 3h to dissolve metals in the liquid for quantification. After a six-fold dilution of digestates with ultrapure water (MilliQ®, Bedford, MA, USA), Cd concentrations were determined by electrothermic atomic absorption spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. For Cd determination in EU, water samples were diluted and acidified at 2% HNO3. The analytical methods were simultaneously validated for each sample series by the analysis of standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Values were consistently within the certified ranges (data not shown).

To quantify metallothionein we used an inorganic mercury saturation method followed by total mercury determination using flameless atomic absorption spectrometry (LECO AMA 254, ALTEC, Prague, Czech Republic) was used. This technique was first described by (Baudrimont et al., 1997). Inorganic mercury saturation relies on mercury‘s affinity to metallothioneins and its ability to displace other heavy metals which can also fix to metallothioneins. Therefore the amount of mercury detected in each sample is proportional to MT concentration. As the number of metal molecule attachment sites per MT molecule is currently unknown in studied species MT concentrations will be expressed in equivalent nmol of Hg-binding sites per gram (fresh weight). The detection limit was estimated at 1 ng Hg.

5. Immunological analyses

192

Chapitre IV. Approche multistress en laboratoire

Hemolymph was withdrawn from the adductor muscle, using a 1-mL plastic syringe (25-gauge needle) and was filtered through an 80-μm mesh and then transferred into an individual micro- centrifuge tube held on ice to minimize cell clumping prior to analysis. Immunological analyses were performed using a FACScalibur flow-cytometer (BD Biosciences, San Jose, CA). Hemocytes viability, phagocytic activity, reactive oxygen species production were measured according to protocols described by Delaporte et al. (2003). The total hemocyte count and the percentage of hemocyte adhesion were measured according to protocols described by Lambert et al. (2007). a. Viability Hemocytes viability was determined by addition 100 μL of anti-aggregant solution for bivalve hemocytes (AASH; Auffret and Oubella, 1994) and 50 μL of filtered sterile seawater (FSSW) to50 μL of hemolymph from each individual. Samples were incubated 2 h at 18 °C in dark conditions with 2 µL of SYBR green I and 2µL of propidium iodide (PI; 10 μg mL− 1). Analysis by flow cytometry allowing us to estimate precisely the percentage of dead cells in each sample (Delaporte et al., 2003). b. Total hemocyte count and hemocyte adhesion Total hemocyte count (THC) was measured by the fixation of 50 μL of each hemolymph with100 μl of a 6% formalin solution and 50 µL of FSSW. Samples were incubated with 2 µL of SYBR green I in darkness at room temperature for 30 minutes before flow-cytometric measurement. THC was calculated according to protocol described by Lambert et al. (2007). To calculate the percentage of hemocyte adhesion, a subsample of each hemolymph was distributed (100 μL) in 24-well microplates and 2 fold diluted with 100 μL of FSSW. After 3 h of incubation at 18 °C, the supernatant containing cells not adhering was transferred into a flow cytometer tube and fixed by addition of 200 μL of a 6% formalin solution. After 30 min of incubation with SYBR Green I, cell concentration was then evaluated as described above. c. Phagocytosis An aliquot of 100 μL of hemolymph, primarily diluted with 200 μL of FSSW, was brought in contact with 30 μL of the solution of 2 μm diameter latex fluorescent beads. Tubes were incubated for 2 h at 18 °C in dark condition. Analysis by flow cytometry allowed detecting hemocytes containing fluorescent beads on the FL1 detector. Phagocytic activity was calculated as the percentage of hemocytes that have ingested three fluorescent beads or more (Delaporte et al., 2003).

193

Chapitre IV. Approche multistress en laboratoire

d. Reactive oxygen species production

A 100 μL aliquot of hemolymph was diluted with 200 μL of FSSW and 4 μL of the 2′7′- dichlorofluorescein diacetate (DCFH-DA) solution (final concentration of 0.01 mM) was added to each tube. Tubes were then incubated for 2h at 18 °C. After the incubation period, DCF fluorescence, quantitatively related to the ROS production of hemocytes, was measured at 500– 530 nm by flow-cytometry. Results are expressed as the mean geometric fluorescence (in arbitrary units, AU).

6. Genetic analyses Total RNAs were extracted from 40 mg of fresh tissue using the Absolutely RNA Miniprep kit (Agilent), according to the manufacturer‘s instructions. For each exposure condition and each organ, samples were analyzed in triplicate. First strand cDNA was synthesized from 5 μg of total RNA using the Stratascript First Strand Synthesis System (Agilent) according to the manufacturer‘s instructions. The amplification of cDNA was monitored using the DNA binding dye SybrGreen I. Real time PCR reactions were performed in a LightCycler (Roche) following the manufacturer‘s instructions, one cycle at 95 °C for 10 min and 50 amplification cycles at 95 °C for 5 s, 60 °C for 5 s, and 72 °C for 20 s. Primer pairs used are listed in Table 1. Relative quantification of each gene expression level was normalized according to the β actin gene expression. Relative gene expression was generated using the 2-ΔCT method as described by Livak and Schmittgen (2001) and ΔCT represents the difference between the cycle threshold (cycle at which the fluorescence from a sample crosses the threshold) of specific gene and the cycle threshold of the β actin. Therefore, induction factor (IF) of each gene in comparison with control corresponds to the following equation: IF = 2-ΔCT (Treatment) / 2-ΔCT (control)

Table 1 Nucleotide sequences of specific primer pairs used in this study. Gene Function Sequence 5'-3' CTGTTGTATGTGGTCTCATGGATb β-actin Cytoskeletal gene (house keeping gene) GCTACGTTGCCCTTGACTa sub unit 1 of the cytochrome C oxidase CTCGCTAATACGACTCCAGTCAb coxI (complex 4 of the mitochondrial respiratory channel) GGTCGGCTTGGACGTAGAa AATACGGAAGTGTTGGGCGa 12S Small subunit 12S of the ribosomal RNA AGAAGAATGGCGAAGCTCTTTb GGCGTGCTCCCAGACATCb sod Mitochondrial superoxide dismutase (Mn) CAGGGATCAGGATGGGGa CGATGTGGATCAGGCTGTGGGa mt1 Metallothionein gene (isoform 1) CGACAGATACTCCCACCGACb Abbreviations: a: upstream primer / b: forward primer

194

Chapitre IV. Approche multistress en laboratoire

7. Statistical analyses Statistical analysis was performed using STATISTICA 7.1 software for Windows. All variables were analyzed statistically using a three way ANOVA in order to determine possible interactive effects between Cd contamination, parasite and bacterial infection on cockles‘ responses (Sokal and Rohlf, 1981). Previously, normality was assumed and homogeneity of variance was verified with Cochran‘s test (data were log transformed when homogeneity of variance was not achieved). Percentages of phagocytic and dead hemocytes were also arcsin transformed to follow homogeneity requirements. Whenever ANOVA was significant, differences between treatments were separated by Tukey means comparison test. Finally, a principal component analysis (PCA) was performed in order to identify possible correlation between the measured variables.

3. Results 1. Cockles mortality

During the first week there was no mortality in any experimental conditions. Between 7 and 14 days a strong mortality occurred in cockles infected by Vibrio tapetis leading to 100% mortality in all units challenged with V. tapetis (V, HV, CV and CHV). Any mortality has been reported for other experimental conditions (control, H, C, CH) between 7 and 14 days.

2. Success of infestation

The average number of metacercariae in cockle mantle and siphon was significantly higher for experimentally infected cockles with an average of 8.3 ± 0.92 (mean ± SE) metacercariae per individual vs. 0 metacercariae per individual for non experimentally infected cockles (Figure 1, Table 2) after 7 days of experiment. In experimentally infested cockles, there was no difference (p > 0.05) in relation with cadmium or V. tapetis contamination (Table 2).

14

12 inthe

10 cyst

cockles 8 of

of 6

number 4 mantle

2 Mean 0 Control H HV CH CHV Experimental conditions Figure 1 Mean intensity of Himasthla elongata cysts in the mantle of Cerastoderma edule (mean ± SE) after 7 days of experiment 195

Chapitre IV. Approche multistress en laboratoire

3. Bacterial challenge No quantitative results were available for the bacterial concentration in the body tissue of individuals as count of bacterial colonies in TCBS medium can not allow us to discriminate Vibrio tapetis from all Vibrio spp. Consequently, analyses of results were considered in a ―binary‖ sense i.e. either EUs were infected by V. tapetis or not.

4. Cadmium bioaccumulation Cd concentration in non contaminated cockles remained low and similar between tissues (Figure 2a, b). The pattern of Cd accumulation was different between single and double stress conditions. In gills, Cd bioaccumulation was significantly lower in cockles exposed to double stresses (1033 ± 30 ng.g-1, dw (mean ± SE) and 1269 ± 52 ng.g-1, dw (mean ± SE) for CH and CV conditions, respectively) than in single Cd condition (1982 ± 166 ng.g-1, dw) (Tukey means comparison test between C vs. CH and C vs. CV: p < 0.001 and 0.002, respectively). Moreover, this decrease of Cd accumulation was not found in the tissues of cockles exposed to triple stress condition (2138 ± 106 ng.g-1, dw) wherein the concentrations were similar to single stress ones. The same accumulation pattern was found in the visceral mass with a significant triple interaction between Cd, bacteria and parasite suggesting that accumulation of Cd in this tissue differed between cockles only infected by one pathogen and cockles infected by both of them (Table 2). a 2500 b 1600 1400

2000 w d

, , 1200

1

dw

-

, , 1

- 1500 1000 ng.g

ng.g 800

1000 600 [Cd] in [Cd]

[Cd] in [Cd] 400 500 200 0 0 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions Figure 2 Cadmium concentrations (mean ± SE) in gills (a) and visceral mass (b) of cockles after 7 days of experiment

5. Metallothionein response

In gills of non contaminated cockles the presence of pathogens did not induced a significant increase in MT concentration after 7 days of experiment (Figure 3a). Cd contamination led to a significant increase in MT concentration in gills of cockles from C and CHV conditions (Table 2,

196

Chapitre IV. Approche multistress en laboratoire

Table 2 Results of the three-way ANOVA comparing the effect of Cd contamination, parasite and bacterial infection (fixed factors) on different variables measured in Cerastoderma edule after 7 days of experiment.

Variable Source of variation dF F p Variable Source of variation dF F p Trematode Cd 1 0.165 0.689 THC Cd 1 1.252 0.281 Bacteria 1 < 0.001 1.000 Bacteria 1 0.937 0.348 Parasite 1 47.724 < 0.001 Parasite 1 9.492 0.008 Cd*Bacteria 1 0.293 0.595 Cd*Bacteria 1 5.804 0.029 Cd*Parasite 1 0.018 0.894 Cd*Parasite 1 0.052 0.823 Bacteria*Parasite 1 0.293 0.595 Bacteria*Parasite 1 0.879 0.363 Cd*Bacteria*Parasite 1 0.661 0.428 Cd*Bacteria*Parasite 1 3.264 0.091

Cd-Gills Cd 1 543.535 < 0.001 Hemocyte Cd 1 5.416 0.033 Bacteria 1 0.632 0.439 viability Bacteria 1 8.091 0.012 Parasite 1 0.332 0.573 Parasite 1 0.362 0.557 Cd*Bacteria 1 1.270 0.277 Cd*Bacteria 1 0.531 0.476 Cd*Parasite 1 4.979 0.041 Cd*Parasite 1 12.115 0.003 Bacteria*Parasite 1 34.630 < 0.001 Bacteria*Parasite 1 1.899 0.187 Cd*Bacteria*Parasite 1 42.469 < 0.001 Cd*Bacteria*Parasite 1 0.053 0.821

Cd-Visceral Cd 1 448.120 < 0.001 Phagocytosis Cd 1 0.047 0.830 mass Bacteria 1 0.705 0.414 Bacteria 1 0.015 0.903 Parasite 1 0.632 0.439 Parasite 1 0.997 0.332 Cd*Bacteria 1 0.460 0.508 Cd*Bacteria 1 0.111 0.742 Cd*Parasite 1 0.015 0.905 Cd*Parasite 1 0.002 0.962 Bacteria*Parasite 1 1.373 0.259 Bacteria*Parasite 1 0.187 0.671 Cd*Bacteria*Parasite 1 13.492 0.002 Cd*Bacteria*Parasite 1 0.561 0.464

Oxidative MT-gills Cd 1 13.805 0.002 burst Cd 1 74.388 < 0.001 Bacteria 1 0.239 0.633 Bacteria 1 7.668 0.014 Parasite 1 0.277 0.607 Parasite 1 23.515 < 0.001 Cd*Bacteria 1 2.081 0.172 Cd*Bacteria 1 8.991 0.008 Cd*Parasite 1 0.259 0.619 Cd*Parasite 1 19.433 < 0.001 Bacteria*Parasite 1 0.290 0.599 Bacteria*Parasite 1 12.691 0.003 Cd*Bacteria*Parasite 1 5.341 0.038 Cd*Bacteria*Parasite 1 54.362 < 0.001

1.518 0.237 MT-Visceral Cd 1 0.416 0.527 Hemocyte Cd 1 mass Bacteria 1 0.374 0.549 adhesion Bacteria 1 6.620 0.021 Parasite 1 2.737 0.117 Parasite 1 0.038 0.848 Cd*Bacteria 1 0.013 0.911 Cd*Bacteria 1 1.078 0.315 Cd*Parasite 1 0.014 0.909 Cd*Parasite 1 8.437 0.011 Bacteria*Parasite 1 0.692 0.418 Bacteria*Parasite 1 0.442 0.516 Cd*Bacteria*Parasite 1 0.022 0.883 Cd*Bacteria*Parasite 1 7.115 0.018

197

Chapitre IV. Approche multistress en laboratoire

Tukey test p = 0.013 and 0.031, respectively). Indeed, highest MT concentrations were shown in single Cd (24.7 ± 11.4 nmol sites Hg.g-1, ww) and triple stress conditions (10.6 ± 3.4 nmol sites Hg.g-1, ww) compared to double stress CH (7.7 ± 0.4 nmol sites Hg.g-1, ww) and CV (5.9 ± 0.5 nmol sites Hg.g-1, ww) conditions (mean ± SE) (Figure 3a). Moreover, the significant interaction between Cd, bacteria and parasite suggested that MT responses in gills of Vibrio infected cockles were modified by the presence of Himasthla elongata in both treatments with and without Cd (Table 2).

MT concentration was higher in visceral mass than in gills in control group (1.71 ± 0.61 nmol sites Hg.g-1, ww in gills and 18.4 ± 4.2 nmol sites Hg.g-1, ww in visceral mass) (Figure 3b). Metal contamination and infection or co-infection by pathogens did not induce a significant increase in MT concentration in visceral mass of stressed cockles in comparison to control group as any statistically significant effect was demonstrated (Table 2).

a a 40 b 100

90

ww ww

35 ,

, ,

1 1

- 80 -

30 70 Hg.g Hg.g 25 60 50

20 sites sites 15 40

nmol 30 nmol 10 20

5 10 [MT] in[MT] [MT] in[MT] 0 0 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions

Figure 3 Metallothionein (MT) concentrations (mean ± SE) in gills (a) and visceral mass (b) of cockles after 7 days of experiment

6. Gene expression

Expression of MT gene was analyzed by real time quantitative PCR in order to compare with protein induction. Concentration of MT protein in function of induction factor of MT gene for each experimental condition in gills and visceral mass was represented Figure 4a,b. Discrepancies between MT gene and protein levels were shown for most treatments and determination coefficient between gene expression and protein induction was very low (r2=0.13 and r2=0.27, in gills and visceral mass, respectively). The expression of mitochondrial genes was

198

Chapitre IV. Approche multistress en laboratoire

analyzed by real time quantitative PCR, induction factors of each gene in comparison to control were represented Table 3. In cockles‘ gills, significant inductions of 12S (x 3-8), coxI (x 2-3) and sod (x 2-4) were shown in all conditions compared to control except for Cd-Himasthla treatment. In this condition expression of all genes was similar to control group suggesting an antagonistic effect of Cd x parasite on mitochondrial gene expression in gills.

In visceral mass, H. elongata led to strong induction of coxI (x 21) and sod (x 23) suggesting an increase of mitochondrial activity leading to the apparition of an oxidative stress. The antagonistic effect of Cd x parasite on mitochondrial gene expression was also demonstrated in visceral mass since induction factors of these genes were significantly lower than in parasite treatment (H). The down regulation of 12S in this treatment suggested a negative impact on mitochondria integrity. Mitochondrial genes were also down regulated by V. tapetis in all Vibrio treatments except in Cd-bacteria (CV) condition. Thus, gene expression was highly tissue dependent in relation with their specific role in bivalve metabolism. a b C C

CH V CHV H HV CH CHV CV V CV H HV

Figure 4 Correlation between MT protein and MT gene induction factors in gills (a) and visceral mass (b) of cockles for each experimental condition after 7 days of experiment

Table 3. Comparative basal expression for the selected genes observed in gills and visceral mass from control cockles after 7 days of experiment. Gills H V HV C CH CV CHV coxI 3 2 3 2 / 2 3 12S 8 6 8 3 / 5 6 Sod (Mn) / 2 4 2 / 2 3 Visceral mass coxI 21 1/2 4/9 5 4 11 2/3 12S / 1/3 1/9 3 1/8 4 1/5 Sod (Mn) 23 2 / 6 6 7 / /: no significant difference in comparison with control group; <1: significant inhibition; >1: significant induction (p<0.05)

199

Chapitre IV. Approche multistress en laboratoire

7. Immune responses

No significant effect of cadmium, parasite and bacterial challenge, combined or not, was shown on phagocytosis activity after 7 days of experiment (26.4 ± 1.1%, mean value ± SE). An important production of ROS by hemocytes was observed in cadmium-parasite (CH) condition (167.3 ± 22.9 A.U., mean ± SE) in comparison with control (9.4 ± 2.2 A.U., mean ± SE) (Figure 5a). The Tukey means comparison test revealed that the ROS value in CH condition was significantly higher than control and others treatments (p < 0.001). Moreover, the three treatments (C: cadmium; H: Himasthla elongata; V: Vibrio tapetis) induced significant effect on oxidative burst and significant interactions occurred between all these treatments (Table 2). Concentration of circulating hemocytes (THC) was not directly affected by the single exposure to V. tapetis whereas a significant individual effect of parasite was shown on THC (Table 2) since hemocyte concentrations were 1.8, 2.7 and 3.8 times higher in H, HV and CH conditions respectively than in control condition (Figure 5b, Table 2). Such increase of THC was not found in CHV treatment. Moreover, a clear interaction (Cd x bacteria) was demonstrated since THC response following V. tapetis infection differed between contaminated and non contaminated cockles (Table 2). In addition, the triple interaction (Cd x Bacteria x parasite) revealed that these differences on THC response were also dependent of the presence of parasites H. elongata. This additional interaction highlighted once more the impact of H. elongata on THC response in cockles.

200 a 3,5E+06 b

3,0E+06

) 1 (A.U.) 150 - 2,5E+06 2,0E+06 100 1,5E+06 1,0E+06 50 (cell.ml THC

5,0E+05 ROS productionROS 0 0,0E+00 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions 100 c 100 d

90 98

80 96

70 94 % viability %

60 Adhesion % 92

50 90 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions Figure 5 Response of immune parameters in function of experimental conditions (mean ± SE) after 7 days of experiment: (a) ROS production (Arbitrary Unit (A.U.)); (b) Total hemocyte count (THC) (cell/mL); (c) Hemocytes viability (%); (d) Hemocytes adhesion (%). 200

Chapitre IV. Approche multistress en laboratoire

A significant effect of Cd contamination and bacteria was demonstrated on hemocytes viability, following by a significant interaction between Cd and parasite (Figure 5c, Table 2). In sum, the Tukey means comparison test demonstrated that percentage of viability in triple stress condition was significantly lower (74.5 ± 4.8 %, mean ± SE) than control one (89.5 ± 3.7 %, mean ± SE) (p = 0.025). Finally, percentage of hemocyte adhesion was significantly affected by the presence of Vibrio tapetis (Table 2). Hemocyte adhesion was lower in V treatment (94.0 ± 0.5 %, mean ± SE) in comparison with control group (98.2 ± 0.1 %, mean ± SE) and significant interactions between combined stressors of cadmium and parasite occurred (Figure 5d, Table 2). The high variability found in triple stress condition was probably related to the occurrence of an important mortality of hemocytes.

8. Principal component analysis

When plotting the different treatments in a Principal Component Analysis, the two first Factors contributed to 67% of the total variance (Figure 6). The most contributive treatment of Factor 1 is the control (32%) and Factor 1 can consequently be assimilated to a reactivity gradient to stress with Cd x H at the other extremity (30%). In this gradient, Vibrio treatment appears close to control. Three treatments are not ‗reactive‘ (C, H, CV) and two have a modest reactivity (HV and triple stress). Immune parameters, notably ROS production and THC, were well correlated to the parasite load in this gradient. Metallothionein concentration, but not MT gene expression, was correlated with cadmium concentration. Finally, the expressions of mitochondrial genes (coxI, 12S, sod) were well correlated to each other but seemed to be independent from others parameters.

201

Chapitre IV. Approche multistress en laboratoire

Figure 6 Principal Component Analysis (PCA). Results are shown to relate experimental treatments (control, H, V, HV, C, CH, CV, CHV), stressors body burden (cadmium concentration and parasite load) and physiological parameters (MT, genes expression and hemocyte count and activities) of cockles. Treatments are italicized and symbolized by stars ( ), stressors were bordered in dotted line and represented by dotted arrows and physiological parameters were bordered and represented by arrows in full lines. Axis 1 corresponds to a reactivity gradient to stress with ―control‖ and ―Cd x H‖ treatment (the most reactive) at both extremities and axis 2 was discriminated by the type of physiological response (immune, protein, gene).

202

Chapitre IV. Approche multistress en laboratoire

4. Discussion

The present study demonstrated that parameters, such as heavy metal contamination and the presence of opportunistic pathogens, can affect defense related activities in the cockle C. edule. Results lend support to the view that interaction between pollutants and pathogens is much more complex than the direct single effects of pollution and parasitism alone on aquatic organisms, and cannot be deduced by simple addition.

1. Trematode infection

Effects of pollutants on parasites may be positive or negative in relation with the balance between the effects of toxicants on host immune system and on pathogen virulence (Sures, 2008). In this study, the infestation success by H. elongata in the cockle under cadmium exposure and bacterial challenge was investigated. Although this digenean parasite mainly encysts in the foot of bivalves, results only showed the number of cysts per host in the mantle and the siphon of cockles and consequently the total number of cysts was underestimated. Nevertheless, it allows us to be sure that these cysts are solely a result of experimental infestation (Wegeberg et al., 1999). The lack of difference in the average number of cysts in cockle‘s tissues between experimental treatments suggested that cadmium or bacterial infection did not modulate the infestation ability of Himasthla elongata cercariae. Toxicity of heavy metal, and in particular cadmium, has been already demonstrated on digenean cercariae survival (Cross et al., 2003), emergence and transmission (Abd Allah et al., 1996), activity and orientation behavior (Morley et al., 2003a, b) and encystement/excystement activity (Morley et al., 2001) but the most significant effects were shown at much higher Cd concentration (from 100 to 50 000 µg/L) than the one used in this present study. It also apparent that the sensitivity of the cockle to this parasite (in term of first line of defense, i.e. avoiding parasite infection) was not modified by cadmium contamination or Vibrio tapetis infestation. These results were in accordance with previous studies dealing with the same host parasite model (Baudrimont and De Montaudouin, 2007; Desclaux-Marchand et al., 2007).

2. Cadmium accumulation

Although neither cadmium contamination nor bacterial infection influenced parasite infectivity, in contrast this experiment demonstrated that the presence of pathogens had led to a significant decrease in cadmium bioaccumulation both in gills and visceral mass under double stress

203

Chapitre IV. Approche multistress en laboratoire conditions. Decreased pollutants accumulation in parasitized individuals has been demonstrated in several host-parasite models by direct pathways (Evans et al., 2001; Sures, 2008). In this experiment, a modification of bivalve behavior, leading to a change of filtration activity or valve opening, was suspected in order to avoid infections rather than direct absorption of cadmium by pathogens. In addition, previous study showed that loss of hemolymph due to parasite encystement in host tissues led to significant changes in filtration activity (Wegeberg, 1998). Since infected hosts exhibit lower metal concentrations, and thus lower associated toxic effects, parasitism could be beneficial in cases of environmental pollution if the negative effects of parasitism do not overcome the positive impact of reduced contaminant levels. Surprisingly, the presence of two pathogens led to antagonistic effect rather than a hypothesized additional effect on Cd accumulation in cockles‘ tissues. The addition of the third sources of stress might exceed bivalve defense mechanisms leading to a global weakness of individuals. The early mortality in this condition from 9 days of exposure supported this hypothesis. Therefore, bivalves became unable to avoid pathogen infection or metal contamination. This result highlighted the complexity of interactions between pathogens and contaminants which cannot be easily predicted.

3. Metallothionein protein and gene

Pathogens may not only affect the accumulation of pollutants by bivalves but also some of hosts functional biology, such as detoxification processes or repair mechanisms (Baudrimont and de Montaudouin, 2007; Sures and Radszuweit, 2007). Metallothioneins were implicated in heavy metal detoxification and were induced by a variety of chemical or physical stresses (Stillman, 1995). In this context, it was interesting to focus on their response to multiple stresses. In non cadmium contaminated conditions, the presence of pathogens did not induce a significant MT concentration increase in comparison with the control at T7 although it is well documented that inflammatory state related to wounds, infections or diseases can induce MT (Gagné et al., 2007). In Cd contaminated cockles, maximal MT concentrations in single Cd treatment and triple stress condition compared to double stress conditions have to be related to results of Cd accumulation. Indeed, cadmium is known to be one of the most powerful inducers of metallothionein (Stillman, 1995). Therefore, the pattern of Cd accumulation could partly explain differences in MT induction between experimental conditions. The expression of the MT gene (mt1) was also investigated in order to compare with protein induction in gills. There was no correlation between

204

Chapitre IV. Approche multistress en laboratoire induction of Cemt1 gene and MT protein response (R2= 0.12) indicating a potential time-lag between gene and protein regulation. This discrepancy suggested that MT synthesis in cockles was regulated both at transcriptional and translational levels. A time delay was already reported in MT from fish Danio rerio (Gonzalez et al., 2006) and mussels Mytilus edulis (Lemoine and Laulier, 2003).

4. Mitochondrial genes

In gills of cockles, there was an induction of both genes involved in mitochondrial metabolism in all conditions compared to control indicating an increase in energetic supply probably in relation to the cost of resistance to pathogen or metal. Indeed, while an energetic failure was classicaly observed during the first phase of sepsis (whole-body inflammatory state due to a blood stream infection), which encompasses a spectrum of illness that ranges from minor signs and symptoms through to organ dysfunction and shock, a clear link between an increasing mitochondrial gene expression and biogenesis, and recovery from spesis has been described in a second step (for review see Carré and Singer, 2008). Moreover, the induction of the mitochondrial superoxide dismutase was the sign of the occurrence of oxidative stress in cells. Both genes were significantly induced in all conditions compared to control, except for Cd-Parasite condition. Indeed, interaction between cadmium and parasite led to an antagonistic effect on mitochondrial gene expression. Once more the combined effects of parasite and metal led to an unusual response, underlying the particularity of this interaction (Sures, 2008). Even though this negative interaction was also found in visceral mass, important inductions of mitochondrial genes were not systematically found in this tissue in contrast with gills. Mitochondrial genes were widely down regulated in visceral mass in contradiction with significant inductions observed in gills in Vibrio treatments. In agreement with results from Carré and Singer (2008), we hypothetized that bacterial infection could affect gills in an earlier time than visceral mass. These observations highlight the specificity of each tissue in gene responses. Indeed, gene response in gills directly reflected modification of environment (such as the occurrence of metal contamination) while visceral mass integrated more intense metabolic activities such as nutrition, reproduction, or inflammatory processes.

5. Immune system

205

Chapitre IV. Approche multistress en laboratoire

The immune system is not isolated and is able to communicate with the neuroendocrine system to maintain homeostasis. This interdependence makes the immune processes particularly sensitive to environmental stressors. It is commonly accepted that exposure to contaminants such as heavy metals could impair immune response of organisms and thus interfere with the host susceptibility to parasites or infectious agents (Pipe and Coles, 1995). Therefore, it is surprising that only a limited number of studies have focused on the combined effects of pathogens/diseases and contaminants (Ford et al., 2008; Sures, 2008). In our experiment, cadmium contamination had little or no effect on immune parameters while the macroparasite H. elongata did. Immunological effects of pollutants strongly differ between both nature and concentration of contaminants (Carajaville et al., 1996). Effects of cadmium on immune parameters are widely described in literature and differ between species and conditions of exposure. Positive correlations between hemocytes density and cadmium exposure were shown in vivo in Mytilus edulis (Coles et al., 1995), Crassostrea gigas (Auffret and Oubella, 1994). Under low doses of cadmium (10-8-10-9 M) bivalve phagocytosis and oxidative burst were stimulated whereas they were suppressed at high doses (10-3-10-4 M) (Boutet et al., 2002; Brousseau et al., 2000). However, in vivo experiment have also lead to no or low variations of hemocytes parameters after Cd exposure (Bouilly et al., 2006) and the strongest effects of cadmium were mainly reported within in vitro studies.

Pathogenic effects of Vibrio tapetis are also highly species dependent and this specificity is generally considered as a result of specific interaction at the molecular level involving both host‘s and pathogen‘s recognition factors (Charalambidis et al., 1996; Olafsen, 1996). For instance, an experimental study conducted on the impact of V. tapetis on susceptible and non susceptible bivalves suggested a high specificity of this bacteria to R. philippinarum reflecting specific host- pathogen interactions in contradiction to other bivalves (Ruditapes decussatus, Mercenaria mercenaria and Crassostrea virginica) which presented no modification of immune parameters after bacterial challenge (Allam et al., 2006). The lack of significant effect of V. tapetis on the immune parameters in cockles, except for hemocyte adhesion, could be related to the status of healthy carrier bivalve previously described in the literature. Indeed, Maes and Paillard (1992) demonstrated that cockles infected by V.tapetis did not develop clinical signs of disease. In addition, in the synthetic PCA, Vibrio treatment appeared close to control and seemed to confirm this paradigm. However, this hypothesis is in contradiction with the strong mortality which

206

Chapitre IV. Approche multistress en laboratoire occurred in all units challenged with V. tapetis wherein all cockles died between 7 and 14 days (no mortality was found in other conditions). As any immune parameters responded to V. tapetis at T7, the massive mortality can be explained by the fact that immune system of cockles was not efficient to fight against bacterial contamination. Accordingly, adhesion capability of hemocyte was negatively affected by these bacteria suggesting that hemocytes became unable to fight this pathogen. Thus, this result really raises the question of the status of ―healthy carrier bivalve‖ of cockles with regards V. tapetis (Paillard, 2004). This phenomenon of cockle mortalities suggest that in stress conditions (handling, environmental multistress etc….) opportunistic bacteria such as V. tapetis in cockles could dramatically affect the mollusc in a same manner that this pathogen can cause clam mortalities, i.e. only when it penetrates into tissues (Allam et al., 2002).

The most significant effect on immune parameters was due to the parasite H. elongata either alone or combined with others stresses. Oxidative burst corresponds to the production of ROS by hemocytes in association with phagocytosis. The oxidative mechanism plays an important role to eliminate many pathogens (Torreilles et al., 1996). A particular interaction between Cd and parasite was found on the oxidative burst process whereas other treatments did not lead to significant variations. Moreover, a significant increase of circulating hemocytes concentration was revealed in parasitized cockles in single and double stress conditions (H, HV, CH). The site of hematopoiesis in bivalves is not really established, and thus cell renewal and maturation process are not yet understood (Auffret, 1988). Therefore, an increase in the total number of circulating hemocytes (THC) may result from either proliferation of cells or migration of cells from tissues to circulation. Such leucocytosis, which corresponds to an increase of total hemocyte count followed by an infiltration of hemocytes in the site of tissue lesions, is a common response to pathogenic infection for bivalves (Fisher, 1986). The following infiltration of hemocytes in damaged tissues was realized in order to repair wounds and encapsulate parasites. Surprisingly at

T7, when cockles underwent parasite infection, bacterial challenge and cadmium contamination, parasitism did not lead to a significant leucocytosis in cockles. Nevertheless, it is important to take into account results of hemocyte viability at T7 to interpret the absence of high hemocytes concentration in this triple stress condition. Indeed, a significant decrease of viable hemocytes was only found in this condition. Other treatments did not have an effect on hemocytes viability. This critical mortality of cockle‘s circulating hemocytes in triple stress condition was in accordance with the following mortality of triple stressed cockles and revealed the strong impact

207

Chapitre IV. Approche multistress en laboratoire of the three combined stresses on cockles. Therefore, this typical synergistic interaction between stresses highlighted the importance of taking into account interactions between pathogens and contaminants in order to describe more realistically the scenario of perturbation since under natural conditions no organism is only affected by one or the other.

In the present study, three environmental parameters were only considered. In field situations there are many more parameters, which may be interacting at any one time. Each stressor and their interaction altered some aspect of metabolism and in combination with the effects on physiology of aquatic organisms it becomes rather complex. The immune system of cockles is particularly susceptible to trematode parasites. The accumulation of cadmium and its effects on detoxification processes vary according to other environmental variables; in this case, the presence of a potential pathogen. The results show that interpretation of sublethal effects of contaminants obtained from single stressor studies needs to be done with care as it is never the case under natural conditions, and demonstrate the need for a multi assay approach when studying defense related activities of organisms to environmental stressors.

Acknowledgments

This work was carried out with the financial support of the ―Région Aquitaine‖ and the French National Research Agency (ANR) with the funds of the ―Multistress project‖. The authors are grateful to sailor Francis Prince, to Pascal Lebleu, Henri Bouillard and the Breton team including Philippe Soudant and Fabienne Le Grand, for their valuable help on the field, for their assistance in the laboratory and for all technical support before and during the experiment.

208

C. Influence de l’espèce sur la réponse aux stress multiples

2. Interactive effects of metal contamination and pathogenic organisms on the marine bivalve Ruditapes philippinarum. Paul-Pont I., de Montaudouin X., Gonzalez P., Jude F., Raymond N., Paillard C., Baudrimont M. Submitted in: Environmental Pollution.

209

210

Chapitre IV. Approche multistress en laboratoire

Abstract

In natural environment, marine organisms are concomitantly exposed to pollutants and multiple disease agents resulting in detrimental interactions. The present study evaluated interactive effects of metal contamination (cadmium) and pathogenic organisms (trematode parasites Himasthla elongata and pathogenic bacteria Vibrio tapetis) singularly and in combination on the bivalve Ruditapes philippinarum under laboratory controlled conditions. After 7 days, metal bioaccumulation and pathogen load were analyzed as well as metallothionein response and hemocyte concentrations and activities. Results showed that infection by opportunistic pathogens affects metal accumulation, leading to maximal Cd accumulation in co-infected clams. Among stressors only V. tapetis induced significant effects on immune parameters whereas a particular interaction ―trematode-bacteria‖ was shown on MT responses. Despite low trematode infestation in agreement with resistant status of R. philippinarum, interaction with bacteria and pollutant occured, highlighting the necessity to take into account pathogen in ecotoxicological studies and suggesting potential heavier alterations with more species-specific pathogen.

Capsule: Co-infection by opportunistic pathogens affects metal accumulation and some defense- related activities in the Manila clam Ruditapes philippinarum.

Key words: Bivalve; pathogens; Cadmium; Metallothionein; Immune defense

211

212

Chapitre IV. Approche multistress en laboratoire

Introduction

Coastal marine organisms are at risk of exposure to a wide variety of both natural and anthropogenic stressors. Among those, coastal zones are endangered by anthropogenic inputs of contaminants due to their extensive use in agricultural, chemical and industrial processes, and metals such as cadmium (Cd) have long been recognized as major pollutants of the marine environment, constituting a hazard to marine organisms (Cheung et al., 2003; Roesijadi, 1994). There is growing concern on the environmental effects of metal pollution in terms of disease and disease susceptibility (Pipe and Coles, 1995). In many invertebrates, immunity is carried out by effectors molecules and proliferation and activities of non specific cells, the hemocytes. These components were contained in the hemolymph, an open circulating system in direct link with nervous and endocrine systems to maintain homeostasis in invertebrate (Pipe and Coles, 1995). This interdependence makes the immune processes particularly sensitive to environmental stressors, including metals. Pollutants could also directly affect pathogenic organisms and consequently their occurrence, distribution and virulence in a positive or negative way depending on an uncountable number of interactive variables (Sures, 2008). In contrast, infectious agents could interfere with the bioaccumulation of toxic compounds (Cross et al., 2003; Evans et al., 2001). Therefore, concomitant exposure to metal and pathogens may lead to different kind of interactions and are more complex than a simple sum of both of them (Sures, 2008). Exploited marine bivalves are often used as sentinel organisms in contaminant monitoring programs due to their ability to filter large quantities of water leading to the accumulation of contaminants from seawater and food (Beliaeff et al., 2002; Nicholson, 2003). During feeding, bivalves may ingest many species of bacteria, prividing substantial carbon and nitrogen contribution (McHenery and Birkbeck, 1986). Some species of bacteria are pathogenic. Among those, species from the genus Vibrio are an important cause of diseases in cultured fish and shellfish (Jeffries, 1982; McGladdery, 1999). Numerous reports of Vibriosis and associated mass mortalities were reported in the Manila clam Ruditapes philippinarum along Japan, Korea and the European Atlantic coastline (England, Ireland, Italy, Spain and Norway). The disease was caused by the Gram-negative, non-sporulating motile bacterium Vibrio tapetis (Paillard, 2004). In a context of stressful environmental condition, several studies demonstrated an increase of susceptibility to Vibrio spp. infections in different hosts in relation with environmental perturbation such as modification of temperature or salinity (Paillard et al., 2004; Reid et al.,

213

Chapitre IV. Approche multistress en laboratoire

2003). Finally, co-infection with parasites could also increase susceptibility of host to Vibrio spp. infections (Rajkumar et al., 2007). Thus, in a context of metal pollution, we could wonder if such portal of entry, locally increasing bacterial load, might interfere with metal accumulation in host tissues. The aim of this work was thus to assess potential interactions between cadmium contamination, bacterial challenge (Vibrio tapetis) and trematode parasite infestation (Himasthla elongata) on the clam Ruditapes philippinarum through a three factors experimental design. Previous studies showed that the introduced species Manila clam was resistant to trematodes (Dang et al., 2009a). However, there are some reports of such infestation in the native area (Lee et al., 2001; Ngo and Choi, 2004) with high prevalence of infested clams (7-42.9%) (Park et al., 2008). Thus, it is relevant to assess wether the apparent resistance of Manila clams to trematode remained the same in association with the highly pathogenic bacteria V. tapetis and/or metal exposure. Laboratory experiment was performed with these sources of stress applied in single, double and triple combinations during seven days. At the end of the experiment, Cd accumulation, bacterial load, parasite infestation and several physiological parameters were assessed. Metallothionein concentration was determined in gills and visceral mass in relation with its implications in detoxification of toxic metals and its ability to respond to inflammatory processes (Gagné et al., 2007). Finally, analysis of hemocytes concentration and activities (phagocytosis, oxidative burst, viability, adherence capability) were performed in order to assess the modulation of immune function in relation to multiple environmental stressors.

Material and methods

1. Bivalve treatments

Manila Clams Ruditapes philippinarum (38.5 ± 0.6 mm shell length, mean ± SD) were collected at Andernos, Arcachon Bay, France (44°42‘N, 1°8‘W). The sampling site consists in an intertidal Zostera noltii seagrass bed that is under the influenc of freshwater input (Leyre River) (Lassalle et al., 2007). Salinity ranges from 22 to 32 psu and temperature ranges from 1°C in winter to 25°C in summer. The median grain size is 142 μm corresponding to fine sand (Lassalle et al., 2007).

The experiment procedure followed a three-factor design with three treatments (C: Cadmium; H: Himasthla elongata; V: Vibrio tapetis) and the four possible interactions (CH; CV; HV; CHV).

214

Chapitre IV. Approche multistress en laboratoire

This experiment was conducted in laboratory controlled conditions using glass aquaria of 12x12x24 cm filled with 1L of synthetic sea water (Instant Ocean, salinity: 30.3 ± 0.2 psu, mean ± SD), aerated by a diffuser system. Plastic coverings were placed over the inside walls to avoid cadmium contamination and to minimize cadmium adsorption on walls. The temperature was fixed at 15°C (± 0.7°C, mean ± SD), pH was regularly monitored and remained stable (8.1 ± 0.3, mean ± SD) and the photoperiod was fixed at 12h using a timer and artificial light sources. No sand was added in order to minimize bacterial contamination as the bacteria may adsorb to the sediment surface and biases bacterial load calculation at the end of the experiment. Four clams were introduced in each experimental unit (EU) and three EU replicates were carried out for each experimental condition. Therefore a total of 24 EU were running simultaneously for a period of seven days.

2. Contamination protocol

Vibrio tapetis (CTC4600) was isolated from diseased R. philippinarum (Borrego et al., 1996). Bacterial contamination was achieved by adding a culture of Vibrio tapetis to the water column to 9 -1 reach a concentration of 1.10 bacteria.mL at T0. This recreation of natural conditions was selected rather than injection of the culture into the palleal cavity or directly into the muscle to minimize stress. In order to keep a relatively constant load of V. tapetis, concentration of 1.109 bacteria.mL-1 was also adjusted after 3 and 7 days.

Himasthla elongata cercariae were collected from infected periwinkles (Littorina littorea). Infected snails were kept at 15°C and fed with macro algae (Ulva spp.). For the experiment, periwinkles were placed at 20°C under artificial light source. This condition induces a cercariae release into the water from which batches of ten were immediatly transferred (cercariae age < 1 hour) into the EU until each EU contained 100 cercariae. To maintain a constant pressure of infection, cercariae were added at T0 and after 1, 3 and 4 days.

-1 A single nominal concentration of Cd at 133nM (corresponding to 15 μg Cd.L added as CdCl2) was selected. After the first contamination (T0), metal quantification in water of experimental units was performed everyday to ensure that metal concentration remains constant throughout the experiment period. Any decrease of metal concentration due to adsorption and absorption mechanisms was daily compensated by addition of aqueous Cd solution.

3. Sampling procedure

215

Chapitre IV. Approche multistress en laboratoire

At T7, one clam per EU was removed. Hemolymph was withdrawn from the adductor muscle of each individual and was observed under a stereomicroscope to ensure the quality of samples (no gamete, no dilution with seawater, no cell debris, etc…). Hemolymph was kept on ice until immunological analyses. Foot and mantle were dissected and squeezed between two sterilized glass slides under a stereomicroscope. Then, H. elongata metacercariae were rapidly counted. The remaining tissues were dissected to separate gills and visceral mass. A small section (> 50 mg) of both organs was used for metallothionein quantification and was placed into polyethylene bags (Whirl-Pak) under N2 atmosphere at -80°C to minimize metallothionein oxidation. The rest of both organs was kept for cadmium quantification.

4. Metals and metallothionein quantification

Cadmium quantification was carried out in the water and the different organs of each clam.

Sampling water was acidified at 2% with nitric acid (Fluka; Buchs, Switzerland. 65% HNO3) before Cd analysis. Tissues were digested in 1 to 1.5 mL (depending on weight of tissue) of nitric acid at 100°C for 3h to dissolve metals in the liquid for quantification. After a six-fold dilution of digestates with ultrapure water (MilliQ®, Bedford, MA, USA), Cd concentrations were determined by electrothermic atomic absorption spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. Analytical methods were simultaneously validated for each sample series by the analysis of standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Throughout our cadmium analyses, mean values of Tort-2 and Dolt-3 were 27.3 ± 0.4 µg.g-1 and 18.9 ± 0.6 µg.g-1, respectively. These values were in certified ranges of Tort-2 and Dolt-3.

To quantify metallothionein, an inorganic mercury saturation method followed by total mercury determination using flameless atomic absorption spectrometry (LECO AMA 254, ALTEC, Prague, Czech Republic) was used. This technique was first described by (Baudrimont et al., 1997b). Inorganic mercury saturation relies on mercury‘s affinity to metallothioneins and its ability to displace other heavy metals which can also fix to metallothioneins. The amount of mercury detected in each sample is proportional to MT concentration. As the exact quantity of Hg binding sites per MT molecule is unknown for R. philippinarum, MT concentration cannot be expressed in mol.g-1 (wet weight, ww). MT concentrations were expressed as nmol Hg-binding

216

Chapitre IV. Approche multistress en laboratoire sites per g (ww): [(ng Hg in sample) / (mL of supernatant)] x [(tissue dilution) / (Hg molar mass)]. The detection limit was estimated at 1 ng Hg. Three reference samples or ―blanks‖ were prepared to monitor the Hg complexation efficiency of the hemoglobin. The mean of the three blank values at each analytical run was deduced from the Hg burdens measured in each sample. A recovery percentage from purified rabbit liver MT (Alexis biochemicals ALX-202-071) was systematically determined. This ―internal standard‖ enabled us to determine the ratio between the binding sites measured after Hg saturation and the potential binding sites indicated by the supplier and previously verified by Cd and Zn determinations on purified MT solution samples. Throughout our metallothionein analyses, the mean recovery percentage was 94.3 ± 2.8 %. This value was consistently within the certified ranges (100 ± 20 %) of the method.

5. Immunological analyses

As soon as the hemolymph was withdrawn it was passed through an 80-µm mesh and stored temporarily in microtubes on ice to avoid cell clumping. Immunological parameters measured were (i) total hemocyte count (THC / number of hemocytes.mL-1); (ii) hemocyte viability (percentage of viable hemocytes); (iii) hemocyte reactive oxygen species (ROS) production that measures the potential to kill non-self particles; (iv) phagocytosis activity (ingestion of fluorescence beads by hemocytes), and (v) adhesion of hemocytes. Hemocyte analyses were adapted from Delaporte et al. (2003), Lambert et al. (2007), Marie et al. (1999), Soudant et al. (2004) and were performed using a FACScalibur flow-cytometer (BD Biosciences, San Jose, CA).

6. Statistical analyses All variables were analyzed using a three-way ANOVA in order to determine possible interactive effects between Cd contamination, parasite and bacterial infection on clams‘ responses (Sokal and Rohlf, 1981). Previously, normality was assumed and homogeneity of variance was verified with Cochran‘s test (data were log-transformed when homogeneity of variance was not achieved). Percentages of phagocytic, adhesion and dead hemocytes were arcsin-transformed to follow homogeneity requirements. Whenever ANOVA was significant, differences between treatments were separated by Fisher test of means comparison. Statistical analysis was performed using STATISTICA 7.1 software for Windows.

217

Chapitre IV. Approche multistress en laboratoire

Results

1. Trematode infestation Even if the trematode infestation remained low in clams after 7 days of experiment, the average number of metacercariae in tissues (1.33 ± 0.38, mean ± SE) of experimentally infected clams was significantly higher than those of non infected clams (Figure 1, Table 1). No effect of cadmium or V. tapetis contamination, alone or in combination, was shown on trematode infestation in clams (Table 1).

7

6

5

4

3

2 Number ofindividual cyst per Number 1

0 Control H V HV C CH CV CHV Experimental conditions

Figure 1. Average number of cyst in tissues of Ruditapes philippinarum (mean ± SE) after 7 days of experiment. H: parasitized with Himasthla elongata ; V: infected by Vibrio tapetis; C: cadmium contaminated; HV, CH, CV, CHV: interactions between stressors.

2. Bacterial challenge

No quantitative results were available for the bacterial concentration in the body tissue of individuals as count of bacterial colonies in TCBS medium cannot allow us to discriminate Vibrio tapetis from all Vibrio spp. Consequently, analyses of results were considered in a ―binary‖ sense i.e. either EUs were infected by V. tapetis or not.

3. Cadmium exposure

Cd concentrations in both organs of non-exposed clams were similar between treatments (i.e. control, H, V and HV) and remained significantly lower than Cd-exposed clams (Figure 2). In non- exposed clams, digestive gland showed higher Cd concentration (330 ± 63 ng.g-1, dry weight (dw),

218

Chapitre IV. Approche multistress en laboratoire

219

Chapitre IV. Approche multistress en laboratoire

mean ± SE) than gills (99 ± 19.7 ng.g-1, dw) (mean ± SE, p=0.0027). In contaminated clams, all treatments induced significant effects on Cd concentration in gills (Figure 2a, Table 1). In addition, the interaction Cd x bacteria x parasite (p=0.023) highlighted the higher Cd accumulation found in clams infected by both pathogens (CHV: 1229 ± 166 ng.g-1, dw) compared to clams only infected by one pathogen (CH: 408 ± 55 ng.g-1, dw and CV: 589 ± 79 ng.g-1, dw) (Figure 2a, Table 1).

In the visceral mass such interaction was also found in relation with the significantly highest Cd accumulation in triple stressed clams (CHV: 2135 ± 181 ng.g-1, dw) (Figure 2b, Table 1). Finally, when pooling all Cd conditions, visceral mass was significantly more contaminated (1473 ± 153 ng.g-1, dw) than gills (651 ± 112 ng.g-1, dw) (p=0.0004).

1600 2500 a b 1400

2000 )

1200 )

dw dw

( (

1 1 -

1000 - 1500

ng.g ng.g 800

1000 [Cd] in in [Cd] 600 in [Cd]

400 500 200

0 0 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions Figure 2. Cadmium concentrations in ng.g-1, dw (mean ± SE) in gills (a) and visceral mass (b) of clams after 7

days of experiment. H: parasitized with Himasthla elongata; V: infected by Vibrio tapetis; C: cadmium contaminated; HV, CH, CV, CHV: interactions between stressors.

4. Metallothionein response

High MT concentration was found in control individuals, as well in gills than in visceral mass (10.9 ± 4.3 and 59.5 ± 14.8 nmol sites Hg.g-1 wet weight (ww), respectively (mean ± SE)) (Figure 3a, 3b). Parasite had significant effect MT concentrations (p<0.05) both in gills and in visceral mass (Table 1). A significant interaction between bacteria and trematode on MT responses was noticed in both organs suggesting that effect of a first pathogen on MT synthesis was modulated by the presence of the second infectious agent, whatever the Cd-contamination treatment was. In fact, Vibrio treatment tended to induce a decrease of MT concentration compared to control, whereas

220

Chapitre IV. Approche multistress en laboratoire

the double infection Himasthla-Vibrio led to an increase of MT concentration. This trend was shown in both non-contaminated and Cd-exposed group, especially in gills.

a b 25 80

70

20

dw

, ,

dw

1 , ,

- 60

1 -

15 50 sites Hg.g sites

sites Hg.g sites 40

nmol 10 nmol

in in 30 in in

[MT] [MT] 20 5 [MT] 10

0 0 Control H V HV C CH CV CHV Control H V HV C CH CV CHV Experimental conditions Experimental conditions Figure 3. Metallothionein concentrations in nmol Hg-binding sites.g-1, ww (mean ± SE) in gills (a) and visceral mass (b) of cockles after 7 days of experiment. H: parasitized with Himasthla elongata; V: infected by Vibrio tapetis; C: cadmium contaminated; HV, CH, CV, CHV: interactions between stressors.

5. Immune responses

No significant effect of cadmium, parasite and bacterial challenge, combined or not, was shown on phagocytosis activity (37.7 ± 2.0%, mean value ± SE), oxidative burst (142.6 ± 14.2 A.U., mean value ± SE), total hemocyte count (THC) (711 333 ± 60 137 cell.mL-1, mean value ± SE) and hyalinocyte concentration (H) (210 190 ± 22 306 cell.mL-1, mean value ± SE) after 7 days of experiment (Table 2). However, a significant effect of bacteria on granulocyte concentration appeared after 7 days of experiment (Figure 4a, Table 1). The granulocyte concentration decreased in clams infected with Vibrio compared to non-Vibrio infected corresponding treatment (HV vs. H / CV vs. C / CHV vs. CH), except in V treatment (Figure 4a, Table 2). This bacterial challenge also led to a significant decrease of hemocytes viability (Figure 4b, Table 1,2), whereas trematode parasite had no effect on this parameter (Table 1). Finally, the percentage of hemocyte adhesion was not significantly modified in clams by Vibrio or Himasthla alone, although there was a significant interaction ―Bacteria x Parasite‖ (p=0.004, Table 1) on this immune parameter.

221

Chapitre IV. Approche multistress en laboratoire

6,E+05 a 6,E+05 a

a 1 5,E+05 -

1 5,E+05 - 1 5,E+05 - 4,E+05 4,E+05 3,E+05 3,E+05 3,E+05 2,E+052,E+05

2,E+05 [Granulocytes] cell.mL [Granulocytes] [Granulocytes] cell.mL [Granulocytes] 1,E+051,E+05 [Granulocytes] cell.mL [Granulocytes] 1,E+05 0,E+000,E+00 0,E+00 ControlControl HH VV HV C CH CVCV CHVCHV Control H V ExperimentalHV C conditionsCH CV CHV Experimental conditions 100 Experimental conditions b 100 bb 100 90 9090 80 8080 70 Viability (%) Viability 70 70

Viability (%) Viability 60 Viability (%) Viability 60 6050 T H V HV C CH CV CHV 50 Experimental conditions 50 T H V HV C CH CV CHV 110 c T H V ExperimentalHV conditionsC CH CV CHV Experimental conditions c 110100 110 c 10090 100 9080 90

8070 Adhesion (%) Adhesion 807060

Adhesion (%) Adhesion 50 7060

T H V HV C CH CV CHV Adhesion (%) Adhesion 6050 Experimental conditions T H V HV C CH CV CHV 50 Experimental conditions T H V HV C CH CV CHV Experimental conditions

Figure 4. Response of immune parameters in function of experimental conditions (mean ± SE) after 7 days of experiment: (a) granulocyte concentration (cell/mL); (b) Adhesion capability (%); (c) Hemocytes viability (%). H: parasitized with Himasthla elongata; V: infected by Vibrio tapetis; C: cadmium contaminated; HV, CH, CV, CHV: interactions between stressors. 222

Chapitre IV. Approche multistress en laboratoire

Discussion

Even though Ruditapes philippinarum constitutes a resistant species to trematode infection, this study showed that co-infection with the pathogenic bacteria Vibrio tapetis can interact on metal accumulation and some defense-related activities in the Manila clam. Although Cd exposure did not lead to strong effects on physiological parameters, interaction between trematode and bacteria were not negligible and bacterial infection showed the strongest effects on immune parameters.

In non contaminated clams, Cd concentrations were higher in visceral mass than in gills. Clams came from a site exhibiting traces of Cd in the sediment (Paul-Pont et al., in revision). Since, the digestive gland reflected metal exposure via the trophic route, acting as a storage compartment, a trophic contamination of clam population via ingested particles could occur in field. Indeed, with the application of a biokinetic model (Luoma and Rainbow, 2005; Wang and Fisher, 1999; Wang

223

Chapitre IV. Approche multistress en laboratoire et al., 1996), it is now well established that trophic route is often an important pathway for metal accumulation in marine suspension and deposit feeders. Cd concentrations of experimentally contaminated bivalves were also three fold higher in visceral mass than in gills, despite the fact that Cd contamination was applied by the directly from water (direct route) instead of by trophic route. Daily Cd quantification on the water of EU filtered or not (Ø 0.2 µm) confirmed that algae did not adsorb Cd fraction (data not shown). Digestive gland acts as a storage compartment in relation with its main detoxification roles for metals (Raspor et al., 2004). Such transfer of toxic compounds from gills to visceral mass corresponds to detoxification processes as it has been already suggested by Smaoui-Damak et al. (2004) and Bebianno et al. (1994) in relation with high MT content in this organ. Shi and Wang (2005) demonstrated a close relationship between Cd body burden, MT content and Cd assimilation efficiency. Thus, high MT concentration in visceral mass of clams may be responsible of the high Cd accumulation.

Metal accumulation was modified by V. tapetis and, in a lesser extent trematode infestation (Table 1) (Figure 5-(1)). The effect of H. elongata was probably indirect since it did not increase Cd accumulation in single stress condition (Figure 2a, 2b) whereas in association with V. tapetis it showed a synergistic effect on Cd bioaccumulation both in gills and visceral mass of bivalves. Due to the low number of cysts of H.elongata in the Manila clams, it was not intuitive to hypothesize that such low infestation level could modify metal bioaccumulation. However, V. tapetis quantification in clams tissues being unavailable, we might envisaged an increase of bacterial concentration following co-infection with parasite as it was already shown in others host-pathogen models (Pylkkö et al., 2006; Evans et al., 2007; Rajkumar et al., 2007). Such increase of V. tapetis infection could severely damaged clams to such a point where pollutant exposure and accumulation were enhanced.

Success of experimental trematode infestation in Manila clams was effective. However, only few trematodes cercariae succeeded to encyst in clams (from 1 to 7 cysts) as it is mentionned by Dang et al. (2009a) who suggested that clam flesh was too hard for cercariae perforation. The experimental infestation efficiency remained similar whatever the presence or not of other sources of stress, i.e. metal and V. tapetis contaminations (Figure 5-(2)). This result revealed that Cd did not impact the infestation capabilities of H. elongata cercariae. This is contrary to the findings of other investigations where an increase in pollutant levels has affected parasites in

224

Chapitre IV. Approche multistress en laboratoire such a way that they are less able to infect their hosts (Cross et al., 2003; Lafferty and Kuris, 1999; Morley et al., 2003a, 2003b). In the same way, the presence of V. tapetis did not seem to enhance trematode infestation in the Manila clam. Similar result was found in a study conducted by (Pylkkö et al., 2006) in which bacterial infection (Aeromonas salmonicida) did not increase parasite infestation (Diplostomum spathaceum) in the European grayling (Thymallus thymallus). Finally, the similar infestation efficiency between tretaments also indicated that there was not a weakening of bivalves due to cadmium contamination to such a point where infestation by H. elongata was promoted.

Unfortunately, in this study we could not assess the effective bacteria load in clams at the end of the experiment. No quantitative results of bacterial concentration in fluids and tissues of clams were available due to the lack of V. tapetis specific primers allowing us to quantify it by real-time quantitative PCR. Since V. tapetis constitutes one of the most pathogenic bacteria for the Manila clam, it would have been very interesting to assess potential facilitation mechanism due to an infection with trematode parasites or cadmium exposure (Figure 5-(3)). Cusack and Cone (1986) concluded that parasites acted as vector to other pathogens by creating portal of entry (tissue lesions). Recent studies have confirmed this hypothesis (Evans et al., 2007; Pylkkö et al., 2006; Shoemaker et al., 2008) and others studies have demonstrated a decreased host resistance following parasite infection linked to increase host susceptibility to other pathogens (Bowers et al., 2000; Tully and Nolan, 2002). In addition, immunotoxicity of metal contamination could also lead to an increase of bacterial infection.

One of the most important heavy metal homeostasis and detoxification mechanisms in marine invertebrate involves the sequestration of metal compounds by metallothioneins (Coyle et al., 2002; Viarengo and Nott, 1993). Cd is known to be one of the powerful inducers of MT (Stillman, 1995). However, in our experiment no effect of Cd, alone or in combination was found in MT concentrations both in gills and visceral mass (Figure 5-(4)). This lack of MT induction following Cd exposure could be explained by the presence of other efficient intracellular ligand such as high molecular mass protein, enzymes, amino acids and peptides like glutathione, which can interfere in the process of metal sequestration (Langston et al., 1998; Viarengo et al., 1989). The important role of insoluble fraction (metal-rich granules, cellular debris and organelles) in detoxification of Cd was also demonstrated in R. philippinarum (Ng and Wang, 2004). In addition, control clams displayed high basal MT concentration in visceral mas and in a lesser

225

Chapitre IV. Approche multistress en laboratoire

Cd V H + + 

MT Synthesis  +  + THC Viability Hyalinocyte - Phagocytosis ROS « H x V »  Granulocytes   -  Cd - Adhesion ? H   H V Cd ? V

Legend:   …. : References to the each step of the discussion

Cd H V : Stressors (Cadmium, Himasthla elongata, Vibrio tapetis)

« H x V » : Particular interaction between parasite and trematode

Viability : Physiological parameters

Figure 5. Synopsis of the interactive effects of cadmium exposure, trematode and bacterial infections in some defense-related activities of the Manila clam Ruditapes philippinarum. 226

Chapitre IV. Approche multistress en laboratoire extent in gills. This high baseline level could be sufficient to respond to metallic stress and could consequently mask any effect of Cd exposure. Even if the digestive gland of molluscs is often selected as a target tissue for the analysis of MT as a biomarker of metal exposure in relation with its main metabolic and storage roles for metals, it could be not enough sensitive to trace metal exposure as it has been shown in other experiements (Bebianno and Langston, 1991; Geffard et al., 2005; Roméo and Gnassia-Barelli, 1995). In fact, the basal MT pool in digestive gland of mollusc is high in relation with homeostasis of essential metals (Geffard et al., 2001) or others metabolic processes such as reproduction or nutrition. Numerous studies reported that physiological changes caused by gonadal development (Baudrimont et al., 1997a) (Baudrimont et al., 2006) and food abundance (Raspor et al., 2004) contribute more to change MT levels than bioavailable toxic metal concentrations. Thus, trace metal exposure at sublethal concentrations might not reach a level which induces additional MT synthesis.

In both organs, trematode parasites led to significant effects on MT responses (Table 1; Figure 5- (5)). It is possible that few metacercariae, or even just penetration attempt, has a significant detrimental effect on a species that is not used to trematodes attack. Penetration and encystement of parasite as well as bacteria colonization may lead to tissue damages and inflammatory state (Canesi et al., 2002; Desclaux, 2003; Jensen et al., 1999) and it is now well documented that MT were induced in inflammatory state related to wounds, infections or diseases (Gagné et al., 2007). Co-infection led to additional effect on MT responses with MT concentration twofold higher in HV treatment than in H and V conditions in gills (Figure 5-(6)). Once more such result was perhaps counterintuitive due to the low trematode infestation success in Manila clams. It suggested possible faciliation mechanism as described above or at least interaction between both parasite and bacteria. Thus, infection by opportunistic pathogen as well as co-infection may not only modify pollutant accumulation but also detoxification processes. Such interactions have to be taken into consideration in the context of ecotoxicological studies.

In our experiment neither cadmium exposure nor trematode infestation induced significant changes in hemocytes counts and activities (Figure 5-(7)). Even if the low efficiency of trematode infestation in clams tallied with the lack of any significant effect, modulation of immune system following the important Cd accumulation in exposed clams should have been expected. Cadmium toxicity on bivalve immune parameters has been widely reported in literature in numerous species. (Auffret and Oubella, 1994; Boutet et al., 2002; Brousseau et al., 2000; Coles, et al.,

227

Chapitre IV. Approche multistress en laboratoire

1995). However, strongest effects of cadmium were mainly reported within in vitro studies although in vivo experiment have lead to no or low variations of hemocytes parameters after Cd exposure (Bouilly et al., 2006). In vitro studies working on cellular defenses of R. philippinarum after experimental exposure to V. tapetis mainly revealed that this bacterium caused a decrease of hemocytes viability, phagocytic activity (Allam and Ford, 2006), adhesion (Choquet et al., 2003) and lysosome concentrations (Allam et al., 2000). In vivo experiments, realized by injection of V. tapetis suspensions in adductor muscle, palleal and extrapalleal cavities, confirmed these findings and evidenced the positive correlation between the percentage of granulocytes and phagocytosis with a consistent resistance to Brown Ring Disease, whereas total hemocyte counts showed no consistent associations (Allam et al., 2001). To our knowledge the present experiment is the first report of V. tapetis infection by addition of bacterial suspension directly in water column. Bacterial suspension used in this experiment (109 cfu.mL-1) was higher than concentrations of suspensions used in palleal/extrapalleal injections (5.107 to 5.108 cfu.mL-1) due to the indirect way of infection. Clams infection by balneation was chosen to minimize stress in animals and to approach as far as possible more realistic conditions of infection in natural environment. This kind of experimental infection might explain the lower intensity of hemocyte responses to V. tapetis exposures, alone or in combination. No significant effect of this bacterium was shown on total hemocyte count, phagocytosis and oxidative burst after 7 days of infection (Figure 5-(7)). No changes in THC and phagocytic activity were also reported in the hemolymph after infection by V. tapetis in a study conducted by (Reid, et al., 2003). Authors suggested no direct role of such increase of total hemocytes densities in reduced disease level. Despite similar THC, a negative effect of bacteria was revealed on granulocyte concentration (Figure 5-(8); Table 2). Changes in proportion of cell type in infected individuals without any modification of THC were already reported in oysters Ostrea edulis infected by Bonamia ostreae (Cochennec-Laureau et al., 2003). Authors correlated this modification of granulocytes concentration to both degranulation or degradation of granulocytes and specific hemocyte type recruitment from hemolymph to site of infections. Specific recruitment to infection site was suspected in the present experiment. Since bacteria colonized firstly mantle and periostracal lamina (Paillard, 2004), cellular components of extrapalleal fluids are known to play a crucial role in defense processes during the first step of infection (Allam et al., 2000). Granulocytes are considered to be the most phagocytic hemocytes in Ruditapes spp. in relation with their high

228

Chapitre IV. Approche multistress en laboratoire phagocytic activities (López et al., 1997) and their wide production of hydrolytic enzymes and antimicrobial substances (López-Cortes et al., 1999), and are consequently considered as the main mobilized hemocytes following bacterial challenge (Allam et al., 2006). Finally, it was accepted that hemocytes can be mobilized toward hemolymph and extrapalleal fluid in relation with the origin of infection (Allam et al., 2006). Infiltration of hemocytes to damaged mantle or extrapalleal compartment has been already reported (Bricelj et al., 1992; Allam et al., 2000). Therefore, specific recruitment of granulocytes from hemolymph to extrapalleal fluid following bacterial infection was suggested to explain the decrease of such hemocyte type in the hemolymph. No direct effect of bacteria was revealed on hemocyte adhesion (Table 2) in contrary with findigs of Choquet et al. (2003). However, a significant interaction between bacteria and parasite on this parameter was observed (Figure 5-(9)) and might be linked with an increase of bacterial concentration due to co-infection with trematodes as described above. An important decrease of hemocyte viability by V. tapetis infection was demonstrated in this study (Figure 5-(10); Table 2) in accordance with previous works (Paillard, 2004). V. tapetis possesses thermosensitive cytotoxic factors that are able to kill in vitro hemocytes of R. philippinarum (Allam et al., 2006).

Conclusion Co-infection by opportunistic pathogens affects metal accumulation and some defense-related activities in the Manila clam Ruditapes philippinarum. Such modulation was highlighted with an apparent non deleterious pathogen (i.e. trematode) for Manila clams. Thus, heavy alterations would have been expected with more pathogenic organisms. The main perspective of this work will consist to determine valid method to accurately quantify Vibrio tapetis in clams fluids and tissues. Such further work constitutes an essential step to assess potential enhancement of bacterial infection following trematode penetration. Overall, particular features appeared in triple stress conditions, in terms of metal detoxification and immune processes. Therefore, it demonstrated the necessity of considering the multiplicity of perturbations sources in coastal ecosystems to assess the health status of organisms.

Acknowledgments

This work was carried out with the financial support of the ―Région Aquitaine‖ and the French National Research Agency (ANR - ―Multistress project‖). Authors are grateful to Francis Prince,

229

Chapitre IV. Approche multistress en laboratoire

Pascal Lebleu, Henri Bouillard and Frances Haynes for their help and their technical support before and during the experiment.

230

Chapitre IV. Approche multistress en laboratoire

Conclusion chapitre IV-C

Différence interspécifique dans la réponse au multistress

Les coques et les palourdes utilisées lors des deux expériences triple stress proviennent des populations de bivalves qui ont été suivies lors de l‘approche in situ : coques du banc d‘Arguin et palourdes d‘Andernos. Ces populations subissent donc un climat similaire au niveau du bassin d‘Arcachon, même s‘il est vrai que cette comparaison souffre du fait qu‘elles ne vivent pas en sympatrie et sont donc soumises à des stress locaux différents. La notion de différence interspécifique est donc à prendre avec précaution et constitue ici une approche très préliminaire. Il est cependant intéressant de comparer les niveaux de base observés in situ en termes de stress et de réponses physiologiques, avec les réponses observées dans le cadre d‘expériences en laboratoire dans lesquelles les sources de stress sont appliquées de manière distincte et contrôlée. Au travers de ces expériences triple stress, des différences interspécifiques marquées ont bien été révélées et concernent les différents points synthétisés dans le tableau IV.1et détaillés ci-dessous.

a. Les niveaux de base de charge en polluant et en pathogènes chez les individus témoins Aucune métacercaire d‘Himasthla elongata n‘est présente chez les individus témoins, en accord avec l‘absence de ce parasite trématode au niveau du banc d‘Arguin (coque), et avec le caractère résistant des palourdes aux parasites trématodes. En ce qui concerne la bactérie Vibrio tapetis, les bivalves du bassin d‘Arcachon ne sont que très peu affectés par cette bactérie. Aucun signe de la maladie (MAB) n‘a pu être observé sur les coquilles des individus témoins. Paradoxalement, les concentrations en cadmium dans les différents tissus des bivalves témoins sont inférieures aux valeurs observées in situ : les concentrations moyennes de Cd sont de 116 ± 31 ng.g-1 et de 42 ± 5 ng.g-1 dans les branchies et la masse viscérale des coques non contaminées. Les palourdes non contaminées quant à elles montrent des valeurs moyennes de 99 ± 20 ng.g-1 et 330 ± 64 ng.g-1 dans les branchies et la masse viscérale, respectivement (les concentrations sont exprimées en poids sec (ps), moyenne (m) ± erreur standard (es), n=12). Ces résultats ont été obtenus par dosage du métal dans les tissus des bivalves après 7 jours d‘expérimentation. Les bivalves étant maintenus dans de l‘eau de mer artificielle préparée de manière à éviter toute contamination métallique (valeur de concentration de Cd dans l‘eau

231

Chapitre IV. Approche multistress en laboratoire

C - ; MT: ; MT: Unités palourde V. tapetis V. en en méthodologique viscérale différentes du parasite duparasite / résistance duparasite / résistance gamétogénèse Commentaires Commentaires Commentaires Impact Lien Cd, Perkinsus, Cd,Perkinsus, Lien Résistance Résistance Contamination locale Contamination l'oxygène Différence interspécifique Différence Problème méthodologique Problème : : Masse de Prise en charge détoxication charge en / Prise Voir conclusion conclusion chapitre IV Voir Absence Absence Mv ; " réactives et du Cdduet H x V Branchies Vt 39 4,3 9,9 5 2,1 7,2 29,7 127 14,7 2,4 ± ± ± ± ± ± ± ? ± ± ± nul Espèces stress faible 34 d'effet 102 8,0 Palourde Palourde Palourde 233 10,9 54,5 91,6 88,2 paramètres 59,5 172,3 infection: effetinfection: additif - Infection unique: pas pas Infectionunique: l'ensemble de ces sur Chute viabilité triple en viabilité Chute Interaction forte " forteInteraction Co

Indépendantde réponse aux stress multiples chez la coque et la palourde. et la la coque chez multiples stress aux réponse ? ? = < ≈ < < < ≈ < ≈ ≈ ≈ < > > < > = ≠ ≠ ≠ = ; ; Cadmium; Cd:Br: " Brun de granulocytes; deERO: granulocytes; et du Cdduet Cd x H Vt 5 4,2 0,4 4,1 3,7 0,1 2,2 21 0,6 2,2 ± ± ± ± ± ± ± ?

± ± ± sensibilité et la et sensibilité nul nul l'Anneau stress 32 Coque Coque Coque 51 1,7 6,2 de 18,4 21,1 25,1 89,5 98,2 13,7 infection: effetinfection: pathogènes - pourcentage diminution [Cd] diminution Infection unique: Infectionunique: : Co antagoniste des deuxdes antagoniste Gr Effet du parasite sur les les sur Effetparasite du Chute viabilité triple en viabilité Chute interaction forte " forteinteraction paramètres immunitaires immunitaires et paramètres Indépendantde Maladie ; ; % , pf) , , pf) , 1 1 - Br - Mv ) - Totale - 1 - ; ; MAB: , ps) , , ps) , (%) 1 He He 1 - - (%) Vt (%) cell.mL (U.A.) 5 (ng.g (ng.g MAB Gr % (10 Hémocytaire Mortalité Br Mv Charge Vibrio tapetis Vibrio - Paramètres - Paramètres Paramètres ERO Viabilité Viabilité V. tapetis (Vt) tapetis V. sites (nmol Hg.g : (nmol Hg.gsites (nmol Adhésion Infestation Infestation Infestation H. elongata elongata (He) H. Phagocytose Phagocytose Cd Vt Cd THC Br ; Mv Mv Bioaccumulation Cd Bioaccumulation - - Bioaccumulation CdBioaccumulation Bioaccumulation CdBioaccumulation Paramètres immunitaires immunitaires Paramètres Réponses protéique (MT) (MT) protéique et Réponses MT

MT

Synthèse des différences et des similitudes dans la et des similitudes différences des Synthèse

Himasthla Himasthla elongata

:

1. .

. V He Efficacité : Interactions Témoins, n=3)* Témoins, Niveaux de base base de Niveaux ( Comparaison de effectuée avec testComparaison moyenne student,le de0,05 = seuil infestation/contamination

Légende Concentration THC: métallothionéines; Arbitraires * Tableau Tableau I

232

Chapitre IV. Approche multistress en laboratoire inférieure à la limite de détection de l‘appareil), une possible décontamination des bivalves durant la semaine d‘expérimentation peut être envisagée pour expliquer ces faibles concentrations de Cd dans les individus non contaminés expérimentalement. Malgré des concentrations en Cd plus faibles que celles observées lors du suivi in situ, les palourdes d‘Andernos présentent des valeurs dans leur masse viscérale bien supérieures à celles obtenues dans les tissus des coques, au niveau des individus non contaminés expérimentalement. Ce résultat est en accord avec les observations apportées dans le chapitre III et confirme une probable contamination trophique des palourdes provenant du site d‘Andernos.

b. Les niveaux des paramètres moléculaire et cellulaire chez les individus témoins : concentrations en métallothionéines et paramètres hémocytaires Les analyses du suivi in situ ont révélé de fortes concentrations en MT dans les tissus des palourdes d‘Andernos, en comparaison avec les autres populations de bivalves étudiées. Ces fortes teneurs en MT observées chez cette population de palourde étaient mises en lien avec la contamination métallique du sédiment, mais aussi avec d‘autres paramètres externes (Perkinsus olseni) et internes (reproduction efficace) pouvant influencer leur synthèse. Lors de nos expériences, ces fortes valeurs sont en conséquence retrouvées chez les individus témoins, avec des palourdes présentant des concentrations nettement supérieures aux coques à la fois dans les branchies (x 6,4) et la masse viscérale (x 3,2). Il est à noter toutefois que les concentrations en MT des palourdes témoins utilisées pour cette expérience présentent des niveaux comparables aux valeurs maximales obtenues lors des deux années de suivi in situ. Dès lors, il est fortement envisagé que la réalisation de cette expérience à un moment différent de l‘année aurait amené à des niveaux relativement plus faibles en valeur absolue. Ce point particulier doit être pris en considération puisque des concentrations élevées de MT chez des individus témoins peut expliquer l‘absence d‘induction de cette protéine lors des conditions stressantes, le niveau de base pouvant être suffisant pour répondre à une faible exposition métallique expérimentale et/ou une inflammation des tissus suite à l‘attaque d‘agents pathogènes. En ce qui concerne les paramètres hémocytaires pris en compte lors de l‘analyse du suivi in situ (THC et % granulocytes), aucune différence interspécifique n‘est observée dans cette expérience sur le THC des individus témoins (coques : 8.105 ± 2.105 cell.mL-1 vs. Palourdes : 6,2.105 ± 2,2.105 cell.mL-1, m ± es). Malgré la forte variabilité retrouvée au niveau de ce paramètre, il est à noter que ces valeurs moyennes se situent au-dessus des valeurs médianes et

233

Chapitre IV. Approche multistress en laboratoire moyennes observées lors du suivi. Un biais méthodologique peut être la cause d‘une telle différence. Tout d‘abord, les analyses hémocytaires n‘ont pas été effectuées sur le même cytomètre entre les analyses de terrain (cytomètre LEMAR) et les expérimentations en laboratoire (cytomètre Bordeaux 2) ce qui peut engendrer de légères variations lors de l‘analyse des concentrations et des activités hémocytaires. De plus, alors que les analyses hémocytaires étaient effectuées dans les heures suivant les prélèvements des individus lors des expérimentations en laboratoire (max. 10h), les échantillons d‘hémolymphe issus du suivi saisonnier étaient conservés à 4°C pendant une plus longue période. Ce décalage d‘analyse, même si les échantillons d‘hémolymphe sont fixés, peut être la cause d‘une plus faible concentration dans les échantillons issus du suivi. En ce qui concerne le pourcentage de granulocytes, le suivi in situ n‘a pas révélé de différences entre les populations de coque et de palourde du bassin d‘Arcachon en raison d‘une forte variabilité de ce paramètre. Lors des expériences triple stress, une différence notable est mise en évidence entre les individus témoins avec des pourcentages de granulocytes nettement plus élevés chez les palourdes (54,5 ± 9,9 %, m ± es) que chez les coques (21,1 ± 0,4 %, m ± es). La phagocytose étant principalement réalisée par ce type hémocytaire chez ces deux espèces de bivalves, des différences d‘activité phagocytaire peuvent être raisonnablement attendues. Cela dit, même si un pourcentage plus élevé de phagocytose est observé chez les palourdes (34 ± 5 %, m ± es) comparativement aux coques (25,1 ± 4,1 %, m ± es), cette différence n‘apparaît pas significative dans notre expérience. De même, aucune différence n‘est observée sur la viabilité des hémocytes entre les deux espèces dans la condition témoin (89,5 ± 3,7 % vs. 91.6 ± 2,1 %, pour les coques et les palourdes témoins respectivement). La capacité d‘adhésion des hémocytes est plus faible chez les palourdes que chez les coques, toutefois cette différence peut être imputée à une forte variabilité des résultats chez la palourde. Enfin, la production d‘ERO est beaucoup plus importante chez les palourdes (x 18) que chez les coques de la condition témoin et constitue la différence interspécifique la plus marquée parmi l‘ensemble des paramètres hémocytaires observés. c. La sensibilité aux sources de stress en termes de capacité de bioaccumulation métallique et de succès d’infestation Le succès d‘infestation des parasites trématodes est en accord avec ce qui a été observé in situ, puisque les coques sont significativement plus infestées (8,3 ± 0,7 métacercaires / individus, moyenne ± erreur standard) que les palourdes (1,3 ± 0,4 métacercaires / individus, moyenne ±

234

Chapitre IV. Approche multistress en laboratoire erreur standard) pour une pression parasitaire et un protocole d‘infestation identique. Ce résultat est en accord avec des sensibilités différentes des deux espèces aux parasites trématodes, et à l‘espèce H. elongata en particulier (Dang et al., 2009a). Concernant Vibrio tapetis, les expériences en laboratoire ont été conduites par balnéation et non par injection afin de ne pas risquer de créer des lésions dans la chair du bivalve, ce qui aurait facilité l‘infection par V. tapetis, et de rester dans dans un mode de contamination physiologiquement réaliste. Le manque de technique quantitative précise nous a empêchés de déterminer a posteriori la charge en bactérie pathogène dans les tissus et les fluides des deux espèces de bivalves. Cette lacune constitue l‘un des plus importants verrous rencontrés lors de ces expérimentations, puisqu‘il est alors impossible de : - Déterminer des différences interspécifiques en termes de développement de l‘infection bactérienne sur 7 jours (progression de la bactérie dans les différents compartiments de l‘hôte) dans les différentes conditions d‘expérimentations. - Déterminer l‘effet des autres sources de stress (Cd, trématode) directement sur la bactérie pathogène (viabilité, succès d‘infestation, etc.) - D‘évaluer de possibles mécanismes de favorisation en lien avec une co-infection par les parasites trématodes digènes ou avec la présence d‘un contaminant métallique dans le milieu. Enfin, les deux espèces de bivalves ont démontré d‘importantes différences en termes d‘accumulation et d‘organotropisme du cadmium dans les conditions contaminées. Alors que les coques ont bioaccumulé préférentiellement le Cd dans leurs branchies, le schéma inverse est retrouvé chez les palourdes contaminées qui présentent des concentrations en Cd dans la masse viscérale supérieure à celle des branchies. Cette prise en charge différente du métal entre les deux espèces de bivalves est à mettre en relation à la fois avec (i) les paramètres biocinétiques (taux de filtration, taux d‘assimilation, taux d‘excrétion) espèce-dépendante (Wang et Rainbow, 2008) ; mais aussi avec (ii) la synthèse de ligands intracellulaires, tel que les métallothionéines, qui influence de manière importante l‘assimilation et la répartition cellulaire et tissulaire des métaux (Marigómez et al., 2002). Les palourdes semblent en effet transférer de manière plus rapide le Cd dissous accumulé dans les branchies à la masse viscérale, à des fins de détoxication.

d. Les interactions entre polluants et pathogènes : Différence des effets interactifs des stress et des réponses moléculaires et cellulaires.

235

Chapitre IV. Approche multistress en laboratoire

Chez les deux espèces de bivalves, la présence d‘autres sources de stress (Cd, V. tapetis) n‘a pas modifié le succès d‘infestation du parasite H. elongata chez l‘espèce hôte, qu‘elle soit sensible (coque) ou résistante (palourde) à ce parasite, dans le cadre de nos expériences. La bioaccumulation de Cd quant à elle est fortement modifiée par la présence d‘un ou deux pathogènes, et ce de manière différente selon l‘espèce de bivalve concernée. Alors que la présence d‘un pathogène (trématode ou bactérie) entraîne une diminution de la bioaccumulation chez la coque dans les conditions contaminées au Cd, aucun effet n‘est observé chez la palourde exposée au Cd lorsque les pathogènes sont appliqués isolément (condition double stress CH et CV). Certes, une légère augmentation de l‘accumulation de Cd est notée dans les branchies des palourdes contaminées et infectées par V. tapetis (condition CV). Toutefois, cette dernière n‘est pas retrouvée dans la masse viscérale des bivalves et l‘effet de V. tapetis seul sur la bioaccumulation de Cd par les palourdes reste imprécis. La différence de sensibilité aux trématodes entre les deux espèces de bivalve peut aisément expliquer l‘absence d‘effet du parasite seul sur la bioaccumulation des palourdes : ces bivalves n‘étant que très peu infestés ils ne voient pas leur bioaccumulation de Cd modifiée, à l‘inverse de l‘espèce beaucoup plus sensible que constitue la coque. La coque, présentée dans la littérature comme espèce porteuse saine de V. tapetis, voit son accumulation de Cd modifiée par la présence de la bactérie à l‘inverse de la palourde qui constitue une des espèces les plus sensibles à cette bactérie. Ce point particulier reste obscur et peut être éventuellement mis en relation avec une mortalité précoce des coques dans les conditions « Vibrio » (V, HV, CV, CHV) qui remet en cause le statut de porteur sain de cette espèce. Certes, la coque est connue pour ne pas développer de signe de la maladie, i.e. formation d‘un anneau brun, mais semble, du moins dans nos conditions expérimentales, sensible à cette bactérie. Enfin, la présence des deux pathogènes en condition contaminée (condition CHV) entraîne sur la bioaccumulation de Cd des effets interactifs différents selon l‘espèce. Alors que la présence d‘un pathogène diminue l‘accumulation de Cd chez les coques, un effet antagoniste des deux pathogènes sur la bioaccumulation de Cd est révélé puisque les individus exposés au Cd et co-infectés (CHV) présentent des concentrations en Cd dans leurs tissus similaires à celles retrouvées chez les coques contaminées mais indemne de tout pathogène (condition C). Au contraire, la présence des deux pathogènes entraîne un effet additif sur la bioaccumulation de Cd puisque les palourdes co-infectées (CHV) présentent des concentrations supérieures à celle retrouvées chez les palourdes contaminées et infectées par un seul pathogène (CH, CV). Les interactions « polluant x pathogène » sur la bioaccumulation métallique diffèrent

236

Chapitre IV. Approche multistress en laboratoire donc selon l‘espèce en relation avec la sensibilité de chaque hôte aux pathogènes étudiés. Ce résultat met en lumière la difficulté de généraliser de tels mécanismes à l‘échelle d‘un même groupe taxonomique. L‘analyse des concentrations en métallothionéines dans les différents tissus des bivalves révèle des différences flagrantes en termes de réponse au stress métallique, à l‘infestation parasitaire et microbienne ainsi qu‘à leurs interactions. Chez les coques, il est important de noter que la réponse protéique dans les branchies semble être majoritairement guidée par les concentrations en Cd, comme le révèle les résultats de l‘analyse de variance, masquant ainsi les effets des autres facteurs de stress. Aucune interaction significative n‘apparaît dans la masse viscérale en dépit d‘une concentration en MT globalement plus élevée dans l‘ensemble des conditions de stress. Ainsi chez cette espèce, les MT semblent répondre de manière efficace à l‘exposition métallique dans les branchies, alors que la réponse de cette protéine de la masse viscérale semble être influencée par d‘autres processus métaboliques. L‘observation plus fine des mécanismes mis en jeu chez les coques non exposées au Cd met en évidence que la présence d‘un pathogène isolé (parasite ou bactérie) entraîne une augmentation de la concentration en MT dans les deux tissus, probablement en relation avec la mise en place d‘une réponse inflammatoire suite à l‘infection par le pathogène. De plus, un effet antagoniste des deux pathogènes est observé chez les coques co-infestées et entraîne une légère diminution de cette concentration en MT par rapport aux conditions de stress simple. Chez les palourdes des tendances différentes sont mises en évidence. En premier lieu, le Cd ne semble avoir aucun effet sur la synthèse de ces protéines dans l‘ensemble des conditions contaminées. Ceci peut être mis en relation avec un fort niveau basal de cette protéine qui peut être suffisant pour répondre à une faible contamination métallique. Chez les palourdes non exposées au Cd, un effet antagoniste des pathogènes est observé sur la réponse des MT mais suit un schéma inverse de celui observé chez les coques : alors que la présence d‘un pathogène entraîne une diminution des concentrations en MT, la co-infection entraîne une augmentation importante de cette concentration. Une interaction significative « parasite x bactérie » sur la réponse des MT est révélée par l‘analyse de variance dans les deux tissus. Toutefois les effets les plus nets sont observés dans les branchies des palourdes. Enfin, la réponse immunitaire des bivalves semble être fortement liée à la sensibilité propre de chaque espèce par rapport aux organismes pathogènes étudiés. Seule la présence de la bactérie V. tapetis semble influencer les paramètres hémocytaires des palourdes, avec toutefois

237

Chapitre IV. Approche multistress en laboratoire l‘apparition d‘une interaction « trématode x bactérie » sur le pourcentage d‘adhésion des hémocytes. Aucun effet du Cd n‘est démontré sur ces paramètres. A l‘inverse, chez la coque les plus forts effets observés sur les paramètres hémocytaires sont dus aux parasites trématodes, avec notamment la mise en lumière d‘une interaction particulière « trématode x cadmium ». Un des seuls points communs entre les deux espèces correspond à une chute de la viabilité des hémocytes chez les bivalves subissant les trois sources de perturbations (CHV), soulignant ainsi le caractère délétère de ce triple stress sur les bivalves.

En résumé, et avec les restrictions d‘interprétation rappelés au début de cette discussion, cette expérience sur les interactions « polluant x pathogènes » a révélé certains points communs entre les deux espèces : l‘infestation par le trématode n‘est pas perturbée par les autres stress alors que la bioaccumulation métallique est modifiée par la présence de pathogènes. La co- infection engendre un effet antagoniste sur la réponse des MT en dehors de toute contamination métallique. Enfin, les paramètres immunitaires sont principalement affectés par le pathogène spécifique de l‘espèce. Les différences quant à elles sont liées à la sensibilité propre de chaque espèce par rapport aux sources de stress et concernent principalement : la bioaccumulation et l‘organotropisme du Cd, ainsi que la manière dont va être modulée la bioaccumulation métallique par les organismes pathogènes. De même, la réponse au stress métallique diffère entre les deux espèces de bivalves en termes de réponse des MT. Finalement, des interactions particulières sont mises en évidence au niveau des paramètres observés (MT et hémocytes) : « bactérie x parasite » chez la palourde et « cadmium x parasite » chez la coque.

238

Chapitre V. Appronfondissement de la réponse des métallothionéines chez la coque

Structure tridimensionnelle d'une métallothionéine de mammifère (β-α). Les résidus cystéines (sphères jaunes) permettent la séquestration de 7 atomes métalliques (sphères bleues) dans les clusters α et β.

239

240

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Introduction

Les résultats obtenus lors des expériences multifactorielles sont complexes et soulèvent de nombreuses questions qui seront discutées dans la conclusion générale. Toutefois, certains verrous rencontrés lors de ces travaux ont motivé des expériences complémentaires.

Une attention toute particulière a été donnée aux observations des réponses à l‘échelle moléculaire, à travers l‘expression de certains gènes ; protéique, à travers la réponse des métallothionéines ; et cellulaire, via l‘analyse des concentrations et activités hémocytaires. A travers ces trois types de réponses, l‘expression et la régulation des gènes apparaît comme la réponse la plus précoce mise en jeu par la cellule à la suite d‘une exposition métallique, bactérienne ou parasitaire, et détermine l‘ensemble des réponses aux autres niveaux (protéique, cellulaire, physiologique) par un jeu de régulation traductionnel et transcriptionnel.

Une comparaison de la réponse des métallothionéines entre expression du gène Cemt1 et réponse de la protéine a été réalisée chez la coque lors de plusieurs expériences (double stress « Cd x parasite » et « Cu x parasite », données non présentées ; triple stress « Cd x parasite x bactérie »). Ces expériences ont chacune démontré un manque flagrant de corrélation à un temps donné entre la réponse génétique, i.e. expression du gène Cemt1, et la réponse protéique en termes de concentration [MT]. Chacun des niveaux d‘observation induisant des interprétations souvent contradictoires, il est apparu essentiel de réaliser des expériences complémentaires nécessaires à la compréhension des divergences observées.

Ainsi, la caractérisation du gène de métallothionéine, comprenant les parties 5‘ et 3‘ non codantes ainsi que son organisation génomique a été réalisé chez la coque. Par la suite, des expériences en stress simple (exposition métallique) ont été réalisées afin de caractériser le schéma de réponse génétique/protéique chez cette espèce suite à une contamination au cadmium et au mercure. Ces deux éléments font partie des métaux ayant la plus forte affinité pour les métallothionéines et leur potentiel fortement inducteur est de plus largement reconnu à ce jour. Enfin, la coque montrant des capacités de bioaccumulation très différentes entre ces deux métaux, il apparaît intéressant de s‘interroger sur les mécanismes de séquestration mis en place et sur le rôle relatif des métallothionéines. Ces expériences ont été réalisées à différents points cinétiques, de l‘heure à la semaine, de manière à estimer le délai de réponse entre l‘exposition au stress

241

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque métallique, la réponse génétique (expression de mt1) et la réponse protéique (concentration de la protéine MT).

Ces expériences sont présentées dans ce chapitre sous la forme de deux articles, soumis (Biometals) et paru (Aquatic Toxicology).

242

A. Caractérisation moléculaire d’une isoforme du gène de métallothionéine – Cemt1 Cloning, characterization and gene expression of a metallothionein isoform in the edible cockle Cerastoderma edule after cadmium or mercury exposure.

Paul-Pont I., Gonzalez P., Montero N., de Montaudouin X. and Baudrimont M. Submitted in: Biometals

243

244

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Abstract Metallothionein (MT) genes encode crucial metal-binding proteins ubiquitously expressed in living organisms, which play important roles in homeostasis of essential metals and detoxification processes. The molecular organization of the first metallothionein gene of the edible cockle Cerastoderma edule and its expression after cadmium (Cd) or mercury (Hg) exposures were determined. The resulting sequence (Cemt1) exhibits unusual features. The full length cDNA encodes a protein of 73 amino acids with nine classical Cys-X(1-3)-Cys motifs, but also one Cys-Cys not generally found in molluscan MT. Moreover, characterization of the molecular organization of the Cemt1 gene revealed the occurrence of two different alleles (A1 and A2). Length differences between these alleles were due to large deletion events in their intronic sequences involving direct Short Interspersed repeated Element (SINE), while the exonic sequences were identical. To our knowledge, such large excision mechanisms have never been reported in other bivalves‘ gene sequences to date. Analysis, in gills of C. edule, of the MT gene expression by quantitative real-time PCR and MT protein induction after Cd or Hg exposures at environmental relevant doses revealed a strong induction of Cemt1 by Cd after 10 days of exposure. Surprisingly, no induction of Cemt1 gene nor MT protein was shown after Hg exposure, while this organism is able to bioaccumulate a high amount of this trace metal. The lack of response of both Cemt1 gene and MT protein by mercury, which is one of the most theoretical powerful inducer of MT biosynthesis, was widely discussed.

Keywords: Cerastoderma edule / metallothionein / gene structure / gene expression / cadmium / mercury

245

246

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Introduction Metallothioneins (MT) are well-known low molecular mass (6-7 KDa), cytoplasmic metal- binding proteins ubiquitous in eukaryotes. They are distinguished by highly conserved cysteine residues (18-20), no or few aromatic amino acids or histidine and high metal content (predominantly Cu, Zn and Cd) forming characteristic metal-thiolate clusters through the sulphur atoms of all cysteine groups (Hamer, 1986). Based on sequence similarities, it is recognized that multiple forms of MT constitute a gene superfamily divided in three groups: Class I which includes MT of mammals, crustacean and bivalves, Class II including MT of fungi, insects, nematodes and echinoderms and Class III including MT of plants and yeasts (Kägi, 1993). The mean feature of MT is the distribution of conserved Cys residues in the recurring structural motifs Cys-Cys and Cys-X(1-3)-Cys. The arrangement of the cystein residues plays a role in the formation of two domains ( and β) within the molecule and constitutes metal thiolate clusters. Nevertheless, MT isolated from invertebrate species are much more variable in their cysteine residue alignment, domain and central domain structure (Park et al. 2007). Even if the primary biological function of this protein remains elusive, the structural characteristics of MT give them metal-binding and redox capabilities (Hamer, 1986). Thus, they are known to play important biological roles such as essential metal homeostasis, detoxification of trace metal ions, protection against free radicals and intracellular oxidative damage and act as a metal reservoir that can be supply to other metalloproteins. In addition, MT is a multi-regulated protein and a variety of factors cans induce its expression in tissues (hormones, physical and physiological stress, inflammation processes…) (Amiard et al. 2006). Therefore, MT is also a stress indicator in organisms, but most studies concern the relationship between metal accumulation and MT expression. The edible cockle Cerastoderma edule is an infaunal suspension feeder, living in the upper few centimeters of the sediment. This exploited bivalve has been previously used as a bioindicator of metal pollution because it accumulates large amount of environmental particles due to its filtering activity related to nutritional and respiratory needs (Cheggour et al. 2001). Indeed, such mollusk exhibits great spatial sensitivity to pollutant and therefore, is the most reliable tool to detect biologically available heavy metal contamination. A field study conducted in the Nord Médoc salt marshes (Gironde estuary, France) on metal bioaccumulation of three bivalve species showed a strong and specific accumulation capacity of C. edule for mercury (Hg) compared to Cd, Cu, Zn and to others bivalve species (C.gigas, R. philippinarum) (Baudrimont et al. 2005). Moreover,

247

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque analysis on macrobenthic community structure along a Hg contamination in an estuarine system revealed the disappearance of cockles at sites with high sediment Hg concentration, probably due to the Hg-specific sensitivity of this species (Nunes et al. 2008). Previous studies on this species were already carried out on Cd contamination, coupled or not with other stresses, and associated MT responses (Desclaux-Marchand et al. 2007). Very few studies treated on bioaccumulation and detoxification processes of Hg in the cockle. According to the fact that Hg is considered as the most potent inducer, as well as Cd, for MT synthesis, we decided to study MT responses in the cockle after experimental Cd or Hg contaminations. Hitherto, a small partial sequence of the MT gene of C. edule has been described (Desclaux-Marchand et al. 2007). It was also necessary to obtain the complete MT sequence and to find potential other MT isoforms. Indeed, studies of MT induction both at the gene and protein levels, beside the determination of Cd and Hg accumulation, would be of great interest to better understanding the molecular mechanisms used by C. edule during exposure to these metals. The present study was firstly designed to isolate and identify the full-length cDNA of metallothionein and the genomic organization of this gene with its 5‘ and 3‘ flanking regions in the cockle C. edule. In a second step, we examined the expression of MT gene and MT protein induction in gills of C. edule after Cd or Hg exposures.

Materials and methods 1. Sampling site and bivalve contamination Cockles were collected from Banc d‘Arguin, a moderately sheltered sandflat, in Arcachon Bay (France, 44°40‘N 1°10‘W). This site was particularly selected because of the lack of heavy metals contamination and therefore the possibility to obtain uncontaminated control specimens (Baudrimont et al. 2006). Cockles were maintained in laboratory for 3 days prior to experiments; they were fed with a concentrated diatom culture (Thalassiosira weissflogii) (≈ 105 algae.ml-1 per experimental unit) during both acclimatization and exposure periods in experimental units (EU). The culture of diatoms was carried out at the laboratory under controlled conditions allowing us to obtain an axenic culture without trace metals contamination. The experiment was conducted in laboratory using glass aquaria of 12x12x24 cm filled with 3L of synthetic sea water (Instant Ocean, salinity = 30‰) aerated by a diffuser system. A 3-cm layer of ultra pure sand allowed the cockles to bury. Plastic coverings (polyethylene) were placed over inside walls to avoid metal

248

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque contamination and to minimize metal adsorption. The temperature was fixed at 18°C (± 1°C), pH was regularly monitored and remained stable (8.2 ± 0.4) and the photoperiod was fixed at 12h using a timer and artificial light sources. For each exposure condition, EU were carried out in triplicate with 10 cockles (33-37mm length) per EU exposed to metal contamination during 10 days. At T0 the three following conditions were applied to EU: control condition without metals, -1 mercury (Hg) contamination at 7.5 nM (corresponding to 1.5 µg Hg.L added as HgCl2) and cadmium (Cd) contamination, with a nominal concentration of 44 nM (corresponding to 5 μg -1 Cd.L added as CdCl2).

2. Sampling procedure At each sampling time (T2 = 2 days and T10 = 10 days) two cockles were taken from each experimental unit corresponding to n=6 individuals per contamination treatment at one time. As soon as cockles were removed, gills were separated from visceral mass and remaining tissues (including muscles, mantle and foot). 50 mg of gills were put in 500 µL of RNA later (Qiagen) and kept at -80°C in order to realize genetic analyses. The rest of gills was divided for metallothionein quantification and metal bioaccumulation. Tissues were placed into polyethylene bags (Whirl-Pak) under N2 atmosphere at -80°C to minimize metallothionein oxidation until the analysis.

3. Metallothionein quantification The concentration of total MT protein was determined in gills by the mercury-saturation assay, using cold inorganic mercury (Baudrimont et al. 1997). Metallothionein analysis was conducted on 6 replicates per exposure condition, the saturation assay being repeated twice per sample. This technique is based on the quantification of Hg bound to the saturated MT. The denaturation of non-MT proteins was performed using trichloroacetic acid and the excess Hg not bound to the MT was removed by scavenging with lyophilized beef hemoglobin (Sigma) prepared in 30 mM Tris-HCl buffer (pH 8.2 at 20°C). The final supernatant was then quantitatively recovered and used for Hg determination by flameless atomic absorption spectrometry (AMA 254, Altec, Prague, Czech Republic). The detection limit was estimated at 0.01 ng Hg. The exact quantity of Hg binding sites per MT molecule being unknown for this species, MT concentrations were expressed in nmol sites Hg.g−1 (wet weight).

249

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

At the same time, three references samples or ―blanks‖ were prepared to monitor the Hg complexation efficiency of the hemoglobin. The mean of the three blank values at each analytical run was deducted from the Hg burdens measured in each sample. A recovery percentage from purified rabbit liver MT (Alexis biochemicals ALX-202-071) was systematically determined. This ―internal standard‖ enabled us to determine the ratio between the binding sites measured after Hg saturation and the potential binding sites indicated by the supplier and previously verified by Cd and Zn determinations on purified MT solution samples.

4. Metal quantification Metal quantification was carried out for the water in the EU and the gills of cockles. Tissues were broken down in 1 to 1.5 mL (depending on weight of tissue) of nitric acid (Fluka; Buchs,

Switzerland. 65% HNO3) at 100°C under high pressure for 3h to dissolve metals in the liquid for quantification. After a six-fold dilution of digestates with ultrapure water (MilliQ®, Bedford, MA, USA) Hg and Cd concentrations were analyzed. Cd concentration was determined by electrothermic atomic absorption spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. For Cd determination in EUs, water samples were diluted and acidified at 2% HNO3. Total Hg concentrations were determined by flameless atomic absorption spectrometry. Analyses were carried out automatically after thermal decomposition at 750 °C under an oxygen flow (AMA 254, Prague, Czech Republic). The detection limit was 0.01 ng Hg.

The analytical methods were simultaneously validated for each sample series by the analysis of standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Values were consistently within the certified ranges (data not shown).

5. Total RNA and DNA extraction Total RNA were extracted from 20 to 40 mg of tissue using the Absolutely RNA RT-PCR Miniprep kit (Stratagene), according to the manufacturer‘s instructions. The concentration of all RNAs was determined by spectrophotometry at 260 nm. The genomic DNA was extracted in order to determine the molecular organization of the Cemt1 gene. Gills were crushed and digested overnight at 50°C with 10 ml per gram of tissue of DNA

250

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque extraction buffer: 10 mM Tris pH 8, 100 mM EDTA pH 8, 0.5 % SDS, and 200 µg.ml-1 proteinase K (Promega). The DNA solution was treated with 100 µg.ml-1 RNAse A (Quiagen) during 2 hours at 37°C. Genomic DNA was then extracted two times with one volume of equilibrated phenol and mixed gently until an emulsion has been formed. Phases were separated by centrifugation at 3000-5000xg for 10 min and aqueous phase was carefully recovered using a wide-bore pipette. A third extraction with phenol:chloroform:isoamyl alcohol (25:24:1) was done and aqueous phase was then transferred into fresh tube with NaCl to a final concentration of 300 mM. Solution was overlaid with 2 volumes of absolute ethanol and mixed gently. The DNA precipitates immediately and should be easily visible. The precipitated DNA was recovered using a Pasteur pipette and transferred into a fresh tube containing 70% ethanol. DNA was recovered by centrifugation (15000g, 10 min, 4°C) , air dry for 5 min and was resuspended in TE (usually 1-2 ml per g of starting material). The DNA concentration was then determined by measuring OD at 260 nm in a spectrophotometer (1 OD unit is 50 µg/ml).

6. Reverse transcription of RNAs First-strand cDNA was synthesized from 5 μg total RNA using the Stratascript First-Strand Synthesis System (Stratagene) according to the manufacturer‘s instructions. The cDNA mixture was conserved at −20 ◦C until using it in a real-time PCR reaction.

7. Primer design MT gene sequences from Crassostrea angulata (Accession number: AF349907), C. gigas (AJ242657), C. virginica (AF506977), Ostrea edulis (AJ306366), Mytilus galloprovincialis (AF199020), Perna viridis (AF092971), available in databases, were aligned using the ClustalW software. From this alignment, one primer named MtF and one reverse primer named MtR were deduced in conserved regions. All primer pairs used in this study were listed in Table 1:

Table 1. Specific primer pairs used in our study Name Sequence 5'-3' Role MT-F ATGTCTGATCCTTGTAACTG cDNA amplification MT-R AACTACTTTACAGCCTGAACA CeMTcpF CGAATCTCAACAACCATCCAG PCR on gDNA CeMTcpR CCACCGACAAGTACTGGGCTCG CeMTrF GCAGCGATACTCCGTGTCGATG 3'RACE CeMTrR CATCGACACGGAGTATCGCTGC 5'RACE

CeMTqF CGATGTGGATCAGGCTGTGGG Quantitative PCR CeMTqR CGACAGATACTCCCACCGAC 251

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

8. PCR and molecular cloning The cDNAs were amplified using primers MtF and MtR. The amplification program consisted of 30 cycles of 95°C for 30 s, 47°C for 30 s, 72°C for 30 s and a final elongation step of 72°C for 5 min. Genomic DNA was also amplified using specific primers CeMTcpF and CeMTcpR designed in the 5‘ and 3‘untranslated regions of complete cDNA, respectively (Table 1). PCR was performed with 40 cycles of 94°C for 1 min, 52°C for 1 min, 72°C for 1 min 40s, then a final elongation step at 72°C for 5 min. Amplified products were analyzed on 1% agarose gels and fragments of expected size were excised and purified using QIAquick Gel extraction kit according to manufacturer‘s instructions (Qiagen). These PCR products were then cloned into the pGEMT cloning vector (Promega) and sequenced (MilleGen). The resulting sequence called Cemt1 has been submitted to the GenBank with accession number EF520705.

9. RT- RACE Single-stranded cDNA for rapid amplification of cDNA ends (RACE) were prepared from total RNA of gills of cockle. 5‘-RACE and 3‘-RACE were conducted according to the manufacturer‘s instructions using the SMARTTM RACE cDNA Amplification Kit (Clontech) and Advantage® 2 PCR Kit (Clontech). The specific primers CeMTrF for 3‘RACE and CeMTrR for 5‘-RACE (Table 1) were designed from the complete coding sequence of Cemt1. The RACE-PCR was conducted at the following settings: 30 cycles at 94°C for 1 min, 68°C for 1 min, 72°C for 3 min and then elongation at 72°C for 10 min. Purification, cloning and sequencing of RACE PCR products were performed as described above.

10. Real-time quantitative PCR Real-time quantitative PCR reactions were performed in a LightCycler (Roche) following the manufacturer‘s instructions (one cycle at 95°C for 10 min, and 50 amplification cycles at 95°C for 5 s, 60°C for 5 s and 72°C for 20 s). Each 20 μl reaction contained 1 μl of cDNA or genomic DNA, 17 μl of master mix including the SyberGreen I fluorescent dye (Roche), enabling the monitoring of the PCR amplification, and 2 µl of the gene specific primer pair at a final concentration of 300 nM for each primer. Gene-specific primer pairs CeMTqF and CeMTqR (Table 1) were determined using the LightCycler probe design software (version 1.0, Roche). Reaction specificity was determined for each reaction from the dissociation curve of the PCR product. This dissociation curve was obtained by following the SyberGreen fluorescence level

252

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque during a gradual heating of the PCR products from 60 to 95 ◦C. Relative quantification of each gene expression level was normalized according to a housekeeping gene expression, the β-actin. Relative gene expression was generated using the 2-ΔCT method as described by (Livak and Schmittgen, 2001) and ΔCT represents the difference between the cycle threshold (cycle at which the fluorescence from a sample crosses the threshold) of specific gene and the cycle threshold of the β-actin. Therefore, induction factor (IF) of Cemt1 gene in comparison with control corresponds to the following equation: IF = 2-ΔCT (Treatment) / 2-ΔCT (control)

11. Statistical analyses The effect of time and metal exposure (independant variables) on MT gene and protein expressions (dependant variables) was tested using a crossed two-ways ANOVA. Prior to analysis, normality was assumed, homogeneity of variance was checked and data were log(x + 1) transformed when necessary. A posteriori test of Tukey was applied to compare means which are significantly different from another (p<0.05). In the real time quantitative PCR, for each gene expression level, the mean value and the associated standard deviation (n = 6) were determined. The mean values for each exposure level were statistically compared to those of the control condition using the non-parametric Mann–Whitney U-test. From this comparison, induction factors of MT gene were obtained by comparing each mean value observed in the contaminated water with that of the control water (p < 0.05). Analyses were performed using Statistica software.

Results 1. Analysis of MT cDNA sequence A partial cDNA sequence of 230 pb was isolated using primers (MtS and MtdegAS) derived from conserved regions in the MT sequence alignment. In order to obtain the complete cDNA sequence, 5‘ and 3‘ RACE were performed using primers CeMtrF and CeMtrR designed from this partial sequence. The complete cDNA sequence (748 pb) named Cemt1 included an open ready frame (ORF) of 222 pb surrounded on both side by 5‘ and 3‘ UTR of 65 pb and 461 pb, respectively (Figure 1). Searching for promoter regions revealed one putative TATA-box (TAATAAA) and one CAT-box (CGAAT) in the 5‘ region at position -15 to -18 pb and -37 to - 42 pb, respectively. A polyadenylation signal (AATTAAA) was recognized from position 698 to

253

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

703 pb upstream the poly-A tail. This cDNA encoded a protein of 73 amino acid residues (Figure 2). This coding sequence revealed a high content of cysteine residues (28%) arranged in nine characteristic Cys-X(1-3)-Cys motifs and one Cys-Cys motif. This sequence presented high similarities with several MT from others bivalves which ranged from 76% with M. edulis, M. galloprovincialis, O. edulis and P. Viridis to more than 83% with C. fluminea, T. agranosa, C. gigas and R. phillipinarum.

GATTAACCAGAATAAGAATTCTTCGAATCTCAACAACCACTCCAGATAATAAAACAACAGCAA

CTATGGGTGATCCATGCAACTGCGCTCAAACCGGTGGTGACTGTAAATGTGCTGCTGGGAAC M G D P C N C A Q T G G D C K C A A G N TGTTGTAGCAGCGATACTCCGTGTCGATGTGGATCAGGCTGTGGGTGCGGATCCGAATGT C C S S D T P C R C G S G C G C G S E C ACCTGTCACGTCAAATGTACCTGTTCCGGATCGTGCGCATGTGGAAACAACTGTACGGGA T C H V K C T C S G S C A C G N N C T G CCGGCCAACTGTACATGTGGAGCAGGGTGCTCCTGCAAGTGACAAGTTAACGGTACATGGTC P A N C T C G A G C S C K GAGGGAACTGACGGATTCGAGCCCAGTACTTGTCGGTGGGAGTATCTGTCGCATCAAAAGCC

GAGCATCGGACAACAGATGGTGGTTTCGAAATAATTCTTTCATCCGTAATGTCCCGATAGAT

ATAGGACTTTAATTTGAATACACAGTATTGTCTTTGGTTAAGACTTTGTATACTTACAATTT

TCCTAATTTTCTTCAACAGATTATTAGTCTCGAAATGCTGAAATGGCTTTTTGTCATGCTTA

CTTGACATGTTACAAGAGTATAACCACAAATATACTTTAAATAGTCTCCATGTAGGAAGGCC

TGAATATTTATTTCAGCCACAAGCCAAATGGAGTTATACAATTTCCTTGACCATGTATGTTT

TGTAATACATTATACTTTATTAAACTGTTCATTTCTAATAAAAAAAAAAAAAAAAAAAAAAA

AAAAAAA

Figure 1. Complete cDNA and amino acid sequence of metallothionein gene of Cerastoderma edule (748 pb). Coding sequence is in bold and italic. CAT-box, TATA-box and polyadenilation signal are underlined in the sequence. In amino acid sequence, region highlighted in grey corresponds to mollusk MTs consensus sequence.

2. Organization of the Cemt1 gene In order to determine the molecular organization of the Cemt1 gene, the primer pair CeMtcpF and CeMtcpR was determined in the 5‘ and 3‘ UTR of the cDNA sequence, respectively. PCR on genomic DNA using this primer-pair revealed two different fragments of 1740 and 1940 pb (Accession numbers GQ259211 and GQ259212, respectively) (Figure 2). Comparison of the resulting sequences suggested that these two fragments could constitute two distinct alleles of the Cemt1 gene. Indeed, no differences existed in the three exonic sequences of these fragments.

254

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

These exons, called E1, E2 and E3 have 28, 120 and 74 pb length, respectively. However, length differences were evidenced between the two large non coding intronic sequences present in the two alleles (A1 and A2), while they were located at the same position in the gene sequence. Indeed, between A1 and A2, intron 1 (I1) varied from 830 to 1133 pb and intron 2 (I2) varied from 586 to 493 pb. Alignment of the two I1 and the two I2 sequences showed that length differences between introns from A1 and A2 were due to the deletion of 303 pb for I1 and 88 pb for I2, while the remaining sequences were highly homologous. At both extremities of these deleted fragments, direct SINE (Short Interspersed repeated Element) appeared. For example, on I2 the SINE sequence TTGCCAC was clearly visible on both extremities of deleted fragment. The intron/exon boundaries of Cemt1 showed the consensus GT…AG splicing signals.

b Size differences between Intron 2.1 et Intron 2.2 : I2-1 TGCGGATCCGAATGTACCTGTCACGTCAAATGTACCTGTTCCGGTAAGTCCGTTTAAACA 120 I2-2 TGCGGATCCGAATGTACCTGTCACGTCAAATGTACCTGTTCCGGTAAGTCCGTTTAAACA 99 ************************************************************

I2-1 CATTACAAGTTGCCACATTTTTAAGCCTTGCAACAGTCGACTTTCAAAAAGCATCGTGGC 180 I2-2 CATTACAAATTGCCAC------115 ******** *******

I2-1 AAAAGGAGAAATTGCTTTTAACAAGCAATTTCTCTTTTTGCCACGATGCTTTATCAATCT 240 I2-2 ------GATGCTTTATCAAACT 131 ************* **

I2-1 ATTCACATATCACAACAGACACTGAAAGATTTTCACATGCTTACAAAGCTTTCCTAATGT 300 I2-2 ATTCACATATCACAACAGACACTGAAAGATTTTCACATGTTTACAAAGCTTTCCTAATGT 191 *************************************** ********************

Figure 2. Genomic organization of Cemt1 gene (a) and example of deletion events in intron 2 between A1 and A2 (b) involving direct Short Interspersed repeated Element (SINE) represented in bold characters.

255

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

3. Metal bioaccumulation Cd concentration in gills of control group remained low and no significant difference appeared

between T2 and T10 (Figure 3a). Contaminated cockles showed a significant accumulation of Cd -1 both at T2 where the concentration reached 5.47 µg Cd.g , dw and at T10 where Cd concentration was 5 times higher than control cockles at this time and 3 times higher than contaminated cockles

at T2. A stronger accumulation of Hg in gills occurred in Hg-contaminated cockles (Figure 3b).

The highest Hg concentration value, which was shown at T10, was 68 times higher than control value. Moreover it was significantly higher than Hg concentration in gills of contaminated

cockles at T2. a b 25 Control 40 Control Contaminated * 20 35 Contaminated *

30

, , dw

, dw ,

1 1

- 15 - 25 20 * 10

15 [Cd] in µg.g in [Cd]

* in [Hg] µg.g 10 5 5

0 0 T2 T10 T2 T10 Time of exposure Time of exposure

Figure 3. Cadmium (a) and mercury (b) concentrations in gills of C. edule according to the time of exposure. Results correspond to average values and standard errors (n=6). * : significant difference compared to controls (p<0.05).

4. Protein response and expression of CeMt1 gene No induction of either Cemt1 gene or MT protein was found at T2 both in Cd and Hg contaminated cockles whereas they significantly accumulated metals (Figure 4). At T10 bivalves experimentally contaminated by Cd presented a significant induction of Cemt1 gene (x 11) compared to controls. Cd exposure also resulted in a 1.6 fold increase of MT protein

concentration at T10 (Figure 4). Neither Cemt1 gene nor MT protein increased significantly at T10 in Hg-contaminated cockles despite a strong accumulation of Hg in gills. Concentration of MT in Hg-treated cockles was significantly lower than in control group at T10.

256

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

8 control

7 Cd *

6 Hg

1 -

5

4

3

[MT] nmolin [MT] sites Hg.g 2 *

1

0 T2 T10

Time of exposure

Figure 4. Metallothionein concentrations after 2 and 10 days of cadmium and mercury exposures. Results correspond to average values and standard errors (n=6). * : significant difference compared to controls (p<0.05).

Discussion Metallothioneins are involved in the adaptative cell mechanisms of protection against the heavy metals toxicity. Therefore, these proteins have been commonly used as metal pollution biomarkers to assess water quality. However a wide range of other factors such as hormones, cytotoxic agents, physical and chemical stress could interfere with the MT biosynthesis (Amiard et al. 2006). Thus, molecular cloning is an important tool for understanding the regulation and the functions of MT. In this study, we characterized the first full-length cDNA and genomic sequence of metallothionein in the common cockle Cerastoderma edule. The Cemt1 gene is characterized by three coding exons separated by two large non coding introns. This tripartite organization in which two introns interrupted three coding exons corresponded to MT-I class family type and has been classically described both for mammalian and invertebrate MTs (Hamer, 1986). Length and position of exons in the gene are very similar in mollusk species

257

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque suggesting similar function and/or regulation of these MT genes (Ceratto et al. 2002). In the mussel Mytilus edulis the length of ORF and exons 1, 2 and 3 of MT-10 were exactly identical to those of Cemt1 (Leignel and Laulier, 2006). Despite a high homology between coding regions, the non coding sequences are more divergent with the exception of short homologous regions in both the 5‘ and 3‘ UTR. The length of intronic sequences was very heterogeneous between species. Indeed, in mouse and human MT intron lengths ranged from 250 to 290 bp for intron 1 to 140-350 bp for intron 2 (Hamer, 1986). This heterogeneity was also found in mollusk MTs such as oysters C. gigas which have very short introns similar to several bivalves (Tanguy et al. 2001), whereas unusual longer intronic sequences were found in the mussel Mytilus galloprovincialis (Intron 1 = 1177 bp ; Intron 2= 1014 bp) (Ceratto et al. 2002). The cockle presented long size introns similar to those of M. galloprovincialis, and they were quite different between allele 1 and allele 2. Exons/introns junctions showed GT-AG splicing rules formulated in the literature (Breathnach and Chambon, 1981), and long introns 1 and 2 presented high AT content (65-69%). In superior eukaryotes, a positive relationship exists between AT content and intron length (Leignel and Laulier, 2006) and this relation was confirmed in C. edule. Length differences between introns from A1 and A2 were due to the deletion of 303 pb for I1 and 88 pb for I2 and at both extremities of these deleted fragments, direct SINE (Short INterspersed repeated Element) appeared. This suggested a possible length evolution of these introns by deletions events involving the SINE sequences. To date, the general function and evolution of introns remain unclear and controversial (Parsch, 2003). However, several observations suggest that intron size is subject to natural selection involving insertion or deletion (indels) mechanisms. Indeed, over evolutionary time, length of intronic sequences tend to decrease, while transition from short to long intron appear to be rare events (Moriyama et al. 1998). However, these indels events seem to be classically reduced in size. In this way, more than 90% of such indels were <10 bp when intron lengths were compared from nine different genes across species of the Drosophila melanogaster species subgroup (Comeron and Kreitman, 2000; Parsch, 2003). Consequently, excision of large nucleotide sequences in the two introns of the Cemt1 alleles A1 and A2 constituted unusual events. To our knowledge, such large excision mechanisms have never been reported in other bivalves‘ gene sequences to date. However, in mammals, recent reports evidenced the occurrence of one long (313 bp) indels in intron 1 of the porcine pou1F1 gene alleles (Cheng-Yi et al. 2007), and a 1550 bp indels in intron 4 of the Kappa-casein gene (csn3) in European rabbits (Carneiro and Ferrand, 2007).

258

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

From our observation A1 and A2 could constitute two different alleles of the same gene. However, it could not be excluded that these two sequences resulted from the duplication of a single copy. In this way, the lack of sequence variations in the coding sequences of these two genes suggested that this duplication event occurred recently. Indeed, it is known that, once gene duplication take place, the two initially identical copies of the gene immediately start to diverge by accumulating mutations (Wagner, 2002). Usually, mollusk MTs comprise 72 to 74 amino acid residues, while smaller chains (61-62 amino acids) have been found in vertebrate MTs (Park et al. 2007). The deduced primary structure of C.edule MT yields a cysteine-rich protein of 73 amino acids. It contained 21 cysteins organized in nine characteristic Cys-X(1-3)-Cys and one Cys-Cys motif arrangements. This Cys-Cys pair was classically found in the domains of mammals MT (Hamer, 1986) or in other invertebrate (Cancer pagurus and Carcinus maenas ; Overnell et al. 1988 ; Pedersen et al. 1994) MTs. However, it is interesting to note that this Cys-Cys motif was not generally found in molluscan MTs (C. gigas, C. virginica, C. fluminea, M. edulis, M. galloprovincialis) except for Argopecten irradians which possessed a quite long MT (110 amino acids) with two Cys-Cys motifs (Wang et al. 2008). Cemt1 did not present Cys-Cys-Cys cluster but this motif is infrequently found in classical MTs except for some mollusk (Crassostrea virginica), arthropod (Callinectes sapidus), or annelid (Eisenia fetida) organisms (Amaro et al. 2008). In addition, Cemt1 sequence included the specific mollusk conserved C-terminal structural pattern C-x-C-x(3)-C-Y-G-x(3)-C-x-C-x(3)-C- x-C-K early described in family 2 of class I mollusk MTs (Tanguy et al. 2001). Despite a small modification of this consensus sequence (Tyr(Y) instead of Thr(T)) for Cemt1, this confirms that C. edule belongs to this gene family. A general feature of classical MT is that they contained two distinct polynuclear clusters, and β. The domain is constituted by 12 cysteine residues organized in Cys-X(1,3)-Cys and Cys-Cys motifs while the β domain contains only 9 cysteines organized only in Cys-X(1,3)-Cys motifs (Kägi and Kojima, 1987). From this point, the number of cysteins and the presence of Cys-Cys motif in Cemt1 allowed us to distinguish the domain in N-terminal from the β domain in C-terminal (Figure 1). MT protein of C. edule also showed typical features such as a lack of aromatic residues and histidine, but also peculiar features i.e. the presence of 13 glycine residues, randomly distributed throughout the sequence, rather than 5-6 as compared to mammalian MT proteins (Ceratto et al. 2002). Additionally, Cemt1 contained little number of lysine residues (4%) compared to other bivalve mollusks such as mussels (6.8%) or Laternula elliptica (11%) (Park et al. 2007). The preference to use mainly lysine instead of

259

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque arginine seems to be a general rule among MTs with a common high K/R ratio in many invertebrate MTs (Amaro et al. 2008). In the cockle, this was not the same, even a higher quantity of lysine rather than arginine, the ratio remains quite small (3/1). A high content of lysine is known to decrease the ability of the protein to release metals due to stronger metal-thiolate interaction (Park et al. 2007). Indeed, Pan et al. (1994) showed that substitution of lysine with glutamate in MT modified the metal-binding of MTs. Therefore, lysine residues of MT were not critical for protein structure but seemed to play a role in stability of both metal-binding clusters. Thus, a strong ability to release metals was suspected for the MT of the cockle C. edule due to its little number of lysine residues. Bivalves have shown a high capacity for metal accumulation, even when found as trace levels in water environments, due to the high volume of water that they filter for feeding and respiratory purposes. The accumulation pattern of many bivalve species has been widely studied. However, few studies have been directed towards the understanding of C. edule physiological processes, which are the clue for understanding the ability of this organism to handle with metal pollution. In gills of C. edule, exposure to Cd showed a progressive accumulation of the metal. Regarding to the bioconcentration of metals, it is important to highlight the specific accumulation capacity that cockles display for Hg. This is in accordance with the results obtained by Baudrimont et al. (2005) in a geochemical survey (Cd, Zn, Cu and Hg) during which cockles exhibited the highest bioaccumulation of Hg, in comparison with other bivalves (Crassostrea gigas and Ruditapes philippinarum) and metal species.

Since MT was first described by Margoshes and Vallee (1957), studies related to Cd as inducer of MT synthesis were over represented and this metal was effectively one of very powerful inducers for MT transcription (Amiard et al. 2006). This is in agreement with the present results which showed an induction of either Cemt1 gene or MT protein at T10 in Cd-contaminated cockles. The lack of gene and protein induction by Cd-exposure at T2 could be explained both by the short time of exposure and by the relatively low exposure concentration. Indeed, in a study conducted by Géret and Cosson (2002) no variation of MT concentration was observed in gills of mussels after 4 days of Cd-exposure at a higher concentration of 200 µg.L-1. A significant increase of MT concentration only appeared after 21 days of contamination. In gills of Hemibarbus mylodon, Cho et al. (2008) showed a significant increase of MT mRNA after 2 days of Cd contamination but only at highest tested dose (10 µM). In our conditions, gills of Cd-

260

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque contaminated cockles unexpectedly showed a quite low responsiveness to Cd contamination, which could be explained by the relatively high basal level of MT mRNA and MT protein which would be sufficient to respond to the low metal contamination in a first time. Marie et al. (2006) also showed no induction of MT in the whole soft body of two bivalves D. polymorpha and C. fluminea after 1 and 3 days of Cd-exposure whereas the concentration was higher (15 µg.L-1). A significant increase of MT concentration only appeared after 10 days of contamination for D. polymorpha and C. fluminea (factor 2 and 1.7 compared to controls, respectively). The induction factor of MT mRNA expression was 20 at the same time for D. polymorpha. These results were in accordance with the response of C. edule to Cd contamination in our experiment with an induction factor at T10 of 1.6 and 11 for MT protein and Cemt1 gene, respectively. Induction of Cemt1 gene was higher than the induction of MT protein at this time indicating a potential time- lag between gene and protein regulation. Indeed, an important induction of gene could be resulting in a strong increase of protein concentration at a later time. This discrepancy suggested that MT synthesis in cockles was regulated both at transcriptional and translational levels. A time delay has also been reported for MT of zebrafish Danio rerio (Gonzalez et al. 2006) and mussels Mytilus edulis (Lemoine and Laulier, 2003). Consequently, MT gene expression and MT protein concentrations have to be investigated together since they give complementary results. Despite a strong accumulation of mercury in gills of Hg-contaminated cockles, neither induction of Cemt1 gene nor MT protein induction was found both at T2 and T10. At T10, a significant decrease of MT concentration was even shown in Hg-treated cockles compared to controls. These results were surprising with regard to the strong affinity of Hg(II) for the cysteinyl thiolate complexes in metallothionein both in vivo and in vitro. Considering the hierarchical sequence of in vitro MT affinity for metals, inorganic Hg is placed in first position and could displace from the MT protein a large number of metals such as Cd, Zn or Cu (Amiard et al. 2006). In spite of the fact that Hg was overall considered as one of the most effective inducer of MT biosynthesis, a few number of studies was really conducted with this metal, and literature data presented many contradictions and inconsistencies in MT induction by Hg. Regala and Rice (2004) showed an induction of MT protein in response to Hg contamination in cells of catfish but not in striped bass. They explained this result by a difference in metal homeostasis between cells of the two fish species. In crustacean Scylla serrata, MT was induced by Cd and Zn but not by Cu and Hg (Olafson et al. 1979). Recently, Verlecar et al. (2008) reported a significant induction of MT in Perna viridis exposed to inorganic Hg. Induction by Hg is highly variable according to species,

261

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque tissues, concentrations and time of exposure (Amiard et al. 2006). In our case, various hypotheses could be advanced in order to understand the lack of MT gene and protein induction by Hg in C. edule. Firstly, if we consider the induction of MT at gene level, in the cockle C. edule only one isoform of MT was sequenced and analyzed by real time PCR. Most bivalves presented at least two isoforms of MT such as C. Virginica (Jenny et al. 2004), M. edulis (Géret and Cosson, 2002) and M. galloprovincialis (Zorita et al. 2007). C. gigas presented three different isoforms (Tanguy and Moraga, 2001). It was also suspected that Cemt1 was not the unique isoform of MT in C. edule and other potential isoforms may respond more specifically to Hg contamination than Cemt1. Indeed, in the zebrafish Danio rerio, (Gonzalez et al. 2006) analyzed the expression of two isoforms of MT gene (mt1 and mt2) in response to metal contamination and showed that mt1 gene was specific to Cd contamination whereas mt2 appears to be common to Cd and Hg contamination. However, in the cockle C. edule even if other isoforms could exist and could be induced more specifically by Hg, an increase of total MT protein should be expected at least at T10. Secondly, a competition between sequestration mechanisms (non-MT cytosolic ligands, biomineralised granules, membranes…) can also interfere with the response of MT to metal exposure (George and Olsson, 1994). Among cytosolic ligands, GSH has protective action, which is known to bind heavy metals through its –SH group (Rabestein et al. 1985). Induction of GSH in bivalves exposed to Hg has been reported in literature (Géret et al. 2002). Due to the high affinity of Hg for this ligand, it should be very interesting to analyze expression or activity of enzymes implied on GSH synthesis such as Glutathion-S-transferase (GST) or Gluthation peroxidase (GPX). In the Asiatic clam C. fluminea contaminated by inorganic mercury during 30 days at a higher concentration (5 µg.L-1 instead of 1.5 µg.L-1), Baudrimont et al. (1997) showed that approximately 83% of total Hg was accumulated in insoluble fraction whereas only 17% remained in cytosol where MTs are localized. Moreover, in Hg-contaminated mussels (M. edulis), Géret and Cosson (2002) showed that Hg concentration of soluble fraction is higher than insoluble one the first four days, but from 4 to 21 days of contamination, Hg concentration only increased in insoluble fraction up to 72% of total Hg at the end of experiment. X-ray analysis and autometallography could be used to determine intracellular localization of heavy metals in aquatic molluscs and several studies confirmed the insoluble localization of Hg since this metal was mainly found in mineralized granules and in lysosomal vacuolar system (Marigómez et al. 2002). Moreover, immunochemistry and ultrastructural studies at the transmission electron microscope revealed that MTs were mainly located in the cytoplasm of chloride cells in gills

262

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

(Alvarado et al. 2007). Thus, the cellular distribution of metal between insoluble and soluble fraction could modulate the amount of metal available for MT sequestration in the cytosol and therefore the induction capacity of MT. In the cockle C. edule, further analyses will be necessary in order to determine the cellular distribution of Hg and metallothioneins between the different intracellular compartments. Such details could help us to understand the lack of induction of both Cemt1 gene and MT protein found in this study after Hg exposure.

Acknowledgments

This work was carried out with the financial support of the ―Région Aquitaine‖ and the French National Research Agency with the funds of the ―multistress project‖.

263

264

B. Etude cinétique de l’expression de Cemt1 au regard de la réponse protéique suivant une exposition au Cd et au Hg

Short-term metallothionein inductions in the edible cockle Cerastoderma edule after cadmium or mercury exposure: discrepancy between mRNA and protein responses.

Paul-Pont I., Gonzalez P., Baudrimont M., Nili H. and de Montaudouin X. Accepted in: Aquatic Toxicology (in press)

265

266

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Abstract Metallothionein (MT) are essential metal binding proteins involved in metallic homeostasis and detoxification in living organisms. Numerous studies have focused on MT response to metal exposure and showed an important variability according to species, metal, concentration and time of exposure. In this study, the expression of one isoform of MT gene (Cemt1) and associated MT protein synthesis were determined after 1, 3, 9, 24, 72 and 168 hours of cadmium (Cd) or mercury (Hg) exposures in gills of the cockle Cerastoderma edule. This experiment, carried out in laboratory conditions, revealed that in Cd-exposed cockles, induction of Cemt1 is time- dependent following a ―pulse-scheme‖ with significant up regulations at 24h and 168h intersected by time point (72h) with significant down regulation. MT protein concentration increases with time in gills of exposed cockles in relation with the progressive accumulation of Cd in soluble fraction. On contrary, Hg exposure does not lead to any induction of Cemt1 mRNA expression or MT protein synthesis compared to control, despite a higher accumulation of this metal in gills of cockles compared to Cd. The localization of Hg (85-90 %) in insoluble fraction, whereas MT were located in the cytoplasm of cells, gives us a first clue to understand the inability of Hg to activate MT synthesis. However, other biochemical processes probably occur in gills of C. edule since the remaining soluble fraction of Hg exceeds MT sequestration abilities. Finally, since one of the first main targets of metal toxicity in cells was the mitochondria, some genes involved in mitochondria metabolism were also analyzed in order to assess potential differences in cellular damages between two metal exposures. Indeed, until T168, no impact on mitochondrial genes was shown following Hg exposure, despite the complete lack of MT response. This result indicated the presence of other effective cellular ligands which sequester the cytosolic fraction of this metal and consequently inhibit metal reactivity. Such competition mechanisms with other cytosolic ligands more sensitive to Hg were particularly argued in the discussion.

Key words: Cerastoderma edule / metallothionein / gills / cadmium / mercury / mRNA expression

267

268

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

1. Introduction Metallothioneins (MT) are low molecular mass cysteine-rich metal-binding proteins found in a large variety of organisms. MT have been suggested to play an important role in zinc and copper homeostasis, detoxification of non essential trace elements such as cadmium (Cd) and mercury (Hg) (Langston et al. 1998) and in the protection against the oxidative stress by the intracellular scavenging of free radicals (Andrews, 2000). Due to its highly inducible expression during exposures to various trace metals, MT have been given much attention as a potential molecular bioindicator to monitor metal pollution of aquatic environment, a major recipient of pollutants. However, a variety of factors can induce its expression in tissues (hormones, physical and physiological stress, inflammation processes…) leading to the difficulty to interpret correctly MT variation in the field. Therefore, MT are now considered as a stress indicator in organisms, but most studies concern the primary relationship between metal accumulation and MT expression (for review see Amiard et al. 2006; Coyle et al. 2002). The edible cockle Cerastoderma edule is an infaunal suspension feeder previously used as a bioindicator of metal pollution because it accumulates large amount of environmental pollutants due to its high filtering activity related to nutritional and respiratory needs (Cheggour et al. 2001). Indeed, such molluscs exhibit great spatial sensitivity to pollutants and are the most reliable tool to detect biologically available heavy metals exposure in environment. With regards to overall toxic effects of heavy metals to bivalve molluscs, studies on mechanisms of metal accumulation, associated with detoxification processes such as MT induction would contribute to better understand physiological response to cellular stress and adaptation in presence of elevated metal bioavailability. Indeed, metal bioavailability, but not metal concentration, can often increase as a function of natural phenomenon, leading to a difficult evaluation of MT responses in term of biomarker in field conditions. The complete amino acid sequence of one isoform of MT (Cemt1) has been determined in the cockle C. edule and a preliminary study showed an important induction of Cemt1 expression and associated MT protein synthesis after 2 and 10 days of Cd-exposure in C. edule whereas no induction both at gene and protein levels was shown after Hg exposure (Paul-Pont et al. in revision). The lack of MT response after Hg-exposure in the cockle was particularly interesting in regard to the strongest accumulation of Hg compared to other metals and to other bivalve species (Baudrimont et al. 2005). The present work aims to (i) understand the apparent differences in MT gene and protein responses between Cd and Hg, known as powerful inducers of MT and to (ii)

269

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

assess potential differences in cellular damages between both metal exposures by the analyze of some mitochondrial genes. Indeed, numerous studies have shown that one of the main targets of cellular metal toxicity was mitochondria (Cambier et al. 2009; de Oliveira Ribeiro et al;, 2008). Since MT protein synthesis differs strongly between both metals exposures, cell protection against metal toxicity may also be very different. Particularly, in the context of Hg exposure, the lack of MT response may lead to several impacts in mitochondria metabolism and integrity, except if other efficient detoxification mechanisms take place in gills of C. edule. Gills of aquatic molluscs constitute a key interface for the uptake of dissolved metal ions from water, therefore gills were considered as the main target organ in metal exposure studies. First, we focused on the repartition of metals between soluble and insoluble fractions of gills in order to determine the amount of metal really available for MT sequestration which can therefore modulate the induction capacity of MT. Secondly, MT mRNA expression and protein synthesis were observed in short period of exposure at particularly early kinetic points (1, 3, 9, 24, 72 and 168 hours). Indeed, most studies working on MT response in cockles have been realized only after 7 or 14 days of metal exposure without taking into account potential earlier responses of MT (Baudrimont and de Montaudouin, 2007; Desclaux-Marchand et al. 2007). However, among physiological processes involved in response to metal exposure, molecular response and genes regulation set up precociously from the first hours and occurred upstream from other biochemical pathways like protein synthesis. Finally, by studying one gene allowing us to quantify number of mitochondria in cells (12S), another involved in mitochondrial electron transfer chain (coxI), and a last one implied in oxidative stress response (sod), we could assess potential damages of Cd and/or Hg on the mitochondria integrity and metabolism. Indeed, metals such as Cd or Hg can severely affect mitochondrial integrity and metabolism by inhibiting the mitochondrial electron transfer chain and inducing reactive oxygen species (ROS) production (Wang et al. 2004).

2. Materials and methods 2.1. Sampling site and bivalve exposure Cockles were collected in January 2009 at Banc d‘Arguin, a moderately sheltered sand flat, in Arcachon Bay (France, 44°40‘N 1°10‘W). This site was selected because of the lack of heavy metals exposure and therefore the possibility to obtain uncontaminated control specimens (Baudrimont et al. 2006). Indeed, Banc d‘Arguin is a National Natural Reserve with a large part dedicated to oyster farming and tourism. However, this nature reserve also includes a large

270

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque prohibited area where there is strictly no activity except scientific sampling. Our cockles were coming from this area, free of human disturbance. The only identified stressor is digenean parasites but we checked that our individuals were free of the most deleterious species (Bucephalus minimus). Cockles were maintained in laboratory for 3 days prior to experiments; this relatively short delay was chosen because we considered that the stress that was undergone by cockles was minimized due to the utilisation of the same water than in the field (with sediment) at similar field temperature. Cockles were fed with a concentrated diatom culture (Thalassiosira weissflogii) (≈ 105 algae.ml-1 per experimental unit) every two days during both acclimatization and exposure periods in experimental units (EU). The culture of diatoms was carried out at the laboratory under controlled conditions allowing us to obtain an axenic culture without trace metal exposure. The experiment was conducted in laboratory using glass aquaria of 25x25x30 cm filled with 10 L of sea water sampled at Banc d‘Arguin (salinity = 30 psu) aerated by a diffuser system. A 3-cm layer of ultra pure sand allowed the cockles to bury. Plastic coverings (polyethylene) were placed over inside walls to avoid metal contamination and to minimize metal adsorption on walls. The temperature was fixed at 15.0°C (± 0.2°C), pH was regularly monitored and remained stable (8.0 ± 0.1) and the photoperiod was fixed at 12h using a timer and artificial light sources. Five EU were carried out for each exposure condition with 16 cockles (31-35 mm length) per EU. At T0 the three following conditions were applied to EU: control condition without metals, mercury -1 (Hg) exposure at 7.5 nM (corresponding to 1.5 µg Hg.L added as HgCl2) and cadmium (Cd) -1 exposure, with a nominal concentration of 44 nM (corresponding to 5 μg Cd.L added as CdCl2). Metal quantification in water of experimental units was performed everyday to ensure that metal concentration remains constant throughout the experiment period. Any decrease of metal concentration due to adsorption and absorption mechanisms was daily compensated by addition of metal solution. Finally, the metal concentrations used in this laboratory experiment were pertinent in terms of environmental relevance since same range of concentrations have been recorded in severely impacted rivers or estuaries (Audry et al. 2004; Feng, 2005; Ferrara et al. 1998; Mohamed et al. 1998; Nascimento et al. 2006). In addition, it is important to precise that the Hg concentration was also chosen in relation with previous studies conducted on cockles C. edule in order to assess the potential toxicity of Hg contents found in cockles from the Gironde estuary (Baudrimont et al. 2005).

271

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

2.2. Sampling procedure

At each sampling time, T0 = 0h, T1 = 1h, T3 = 3h, T9 = 9h, T24 = 24h, T72 = 72h and T168 = 168h = 7 days, two cockles were taken from each experimental unit leading to n=10 individuals per exposure treatment at one time. As soon as cockles were removed, gills were separated from visceral mass and remaining tissues (including muscles, mantle and foot tissues). Foot and mantle of cockles were squeezed between two sterilized glass slides and observed under a stereomicroscope in order to verify the absence of parasites and particularly the macroparasite Bucephalus minimus in relation to their strong negative impacts on host physiology. Gills from two individuals were pooled and homogenized. Therefore, five replicates of two homogenized individuals were treated for analyzes. Gills were chosen for this study instead of digestive gland because this organ represents the first barrier involved in dissolved metal uptake via the direct exposure route, due to its large exchange area in direct contact with the surrounding environment. Since we aim to observe response of metallothionein to dissolved metal exposures in a very short time course (four of the six kinetic points were done during the first 24h), it seems more appropriate to focus on gills as a target tissue for short term dissolved metal exposure and MT synthesis than the digestive gland which reflects a more long term metal exposure.

A minimum of 50 mg of gills were dissected for metallothionein quantification and were kept into polyethylene bags (Whirl-Pak) under N2 atmosphere at -80°C in order to minimize metallothionein oxidation until the analysis. Gills (20-40 mg) were put in 500 µL of RNA later (Qiagen) and kept at -80°C in order to realize genetic analyses. The rest of gills was divided for total and subcellular (soluble / insoluble fractions) metal determination.

2.3. Metallothionein quantification The concentration of total MT protein was determined in gills by the mercury-saturation assay, using cold inorganic mercury (Baudrimont et al. 2003; Dutton et al. 1993). Metallothionein analysis was conducted on 5 replicates per exposure condition, the saturation assay being repeated twice per sample. This technique is based on the quantification of Hg bound to the saturated MT. The denaturation of non-MT proteins was performed using trichloroacetic acid and the excess Hg not bound to the MT was removed by scavenging with lyophilized beef hemoglobin (Sigma) prepared in 30 mM Tris-HCl buffer (pH 8.2 at 20°C). The final supernatant, containing the metallothionein fraction, was then quantitatively recovered and used for Hg determination by flameless atomic absorption spectrometry (AMA 254, Altec, Prague, Czech

272

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Republic). The detection limit was estimated at 0.01 ng Hg. The exact quantity of Hg binding sites per MT molecule being unknown for this species, MT concentrations were expressed in nmol sites Hg.g−1 (wet weight). The measure of MT concentration in relation to specific tissue weight allowed us to directly compare MT concentration with bioaccumulation results expressed in relation to tissue weight too. Moreover, this calculation method in relation with tissue weight is classically used in the metal saturation technique and allowed us to compare our results with numerous data available in the literature (Campbell et al. 2005; Couillard et al. 1993; Perceval et al. 2006).

At the same time, three references samples or ―blanks‖ were prepared to monitor the Hg complexation efficiency of the hemoglobin. The mean of the three blank values at each analytical run was deducted from the Hg burdens measured in each sample.

A recovery percentage from purified rabbit liver MT (Alexis biochemicals ALX-202-071) was systematically determined. This ―internal standard‖ enabled us to determine the ratio between the binding sites measured after Hg saturation and the potential binding sites indicated by the supplier and previously verified by Cd and Zn determinations on purified MT solution samples. Throughout our metallothionein analyses, the mean recovery percentage was 92.7 ± 5.6 %. This value was consistently within the certified ranges (100 ± 20 %) of the method.

2.4. Metal quantification and cellular fractionation Metal quantification was carried out for the water in the EU and the gills of cockles. For metal determination in EUs, water samples were acidified at 2 % HNO3 before analyses. Gills of cockles were separated on two aliquots for total metal quantification and for cellular fractionation. For total metal quantification in gills of cockles, tissues were digested in 1 to 1.5 mL (depending on weight of tissue) of nitric acid (Fluka; Buchs, Switzerland. 65 % HNO3) at 100°C under for 3h in a drying oven in order to dissolve metals in the liquid for quantification. After a six-fold dilution of digestates with ultrapure water (MilliQ®, Bedford, MA, USA) Hg and Cd concentrations were analyzed. Cd concentration was determined by electrothermic atomic absorption Spectrophotometry with Zeeman correction, using a graphite furnace (M6 Solar AA spectrometer, Thermoptec). The detection limit was 0.1 μg Cd.L-1. Total Hg concentrations were determined by flameless atomic absorption spectrometry. Analyses were carried out

273

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque automatically after thermal decomposition at 750 °C under an oxygen flow (AMA 254, Prague, Czech Republic). The detection limit was 0.01 ng Hg.

Metal repartition in gills was only determined at T24, T72 and T168, corresponding to times from which accumulation of both metals was significantly different from control. In order to determine metal concentration and repartition in soluble and insoluble fractions, gills were homogenized in 1 mL of 25 mM Tris HCl buffer (pH = 7.2 at 20°C) with an electric crusher (Ultra-Turrax T-25). Homogenate was centrifugated at 4°C and 20 000g during 60 minutes. Supernatant was recovered and acidified at 2 % for metal quantification in soluble fraction. The remaining pellet was digested in 1.5 mL of nitric acid (Fluka; Buchs, Switzerland. 65 % HNO3) at 100°C for 3h in a drying oven in order to determine metal concentration in insoluble fraction. Metal concentrations in both fractions (supernatant and pellet) were analyzed as described above.

The analytical methods of cadmium and mercury quantification were simultaneously validated for each sample series by the analysis of standard biological reference materials (Tort-2: Lobster hepatopancreas and Dolt-3: Dogfisher liver from National Research Council of Canada, Ottawa). Throughout our cadmium analyses, mean values (n=60) of Tort-2 and Dolt-3 were 27.2 ± 0.5 µg/g and 19.2 ± 0.7 µg/g , respectively. Throughout mercury analyses, mean values (n=60) of Tort-2 and Dolt-3 were 27.1 ± 0.3 µg/g and 19.9 ± 0.2 µg/g, respectively. These values were consistently in certified ranges of Tort-2 and Dolt-3.

2.5. Total RNA extraction and reverse transcription of RNAs Total RNA were extracted from 20 to 40 mg of tissue using the ―Absolutely RNA Miniprep kit‖ (Agilent), according to the manufacturer‘s instructions. The concentration of all RNAs was determined by Spectrophotometry at 260 nm. First-strand cDNA was synthesized from 5 μg of total RNA using the Stratascript First-Strand Synthesis System (Agilent) according to the manufacturer‘s instructions. The cDNA mixture was conserved at −20 °C until using it in a real- time PCR reaction.

2.6. Real time quantitative PCR The amplification of cDNA was monitored using the DNA intercalating dye SybrGreen I. Real- time PCR reactions were performed in a Light Cycler (Roche) following the manufacturer‘s instructions. The amplification program consisted of one cycle at 95 °C for 10 min and 50 amplification cycles at 95 °C for 5 s, 60 °C for 5 s, and 72 °C for 20 s. Each 20 μL reaction

274

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque contained 1 μL of cDNA, 17 μL of master mix including the SybrGreen I fluorescent dye (Roche), enabling the monitoring of the PCR amplification, and 2 µl of the gene specific primer pair at a final concentration of 300 nM for each primer. Primer pairs used to quantify genes of interest are listed in Table 1. Reaction specificity was determined for each reaction from the dissociation curve of the PCR product. This dissociation curve was obtained by following the SybrGreen fluorescence level during a gradual heating of the PCR products from 60 to 95 ◦C. Relative quantification of each gene expression level was normalized according to a housekeeping gene expression, the β-actin. Relative gene expression was generated using the 2- ΔCT method as described by (Livak and Schmittgen, 2001) and ΔCT represents the difference between the cycle threshold (cycle at which the fluorescence from a sample crosses the threshold) of specific gene and the cycle threshold of the β-actin. Therefore, induction factor (IF) of each gene in comparison with control corresponds to the following equation: IF = 2-ΔCT (Treatment) / 2-ΔCT (control)

Table 1 Nucleotide sequences of specific primer pairs used in this study. Gene Function Sequence 5'-3' CTGTTGTATGTGGTCTCATGGATb β-actin Cytoskeletal gene (house keeping gene) GCTACGTTGCCCTTGACTa sub unit 1 of the cytochrome C oxidase CTCGCTAATACGACTCCAGTCAb coxI (complex 4 of the mitochondrial respiratory channel) GGTCGGCTTGGACGTAGAa AATACGGAAGTGTTGGGCGa 12S Small subunit 12S of the ribosomal RNA AGAAGAATGGCGAAGCTCTTTb GGCGTGCTCCCAGACATCb sod Mitochondrial superoxide dismutase (Mn) CAGGGATCAGGATGGGGa CGATGTGGATCAGGCTGTGGGa Cemt1 Metallothionein gene (isoform 1) CGACAGATACTCCCACCGACb

Abbreviations: a: upstream primer / b: forward primer

2.7. Statistical analyses The effect of time and metal exposure (independent variables) on metal accumulation and MT protein expression (dependent variables) was tested using a crossed two-ways ANOVA. Prior to analysis, normality was assumed, homogeneity of variance was checked and data were log(x + 1) transformed when necessary. Whenever ANOVA was significant, differences between treatments were separated by Tukey means comparison test. In the real time quantitative PCR, for each mRNA expression level, the mean value and the associated standard deviation (n = 5) were determined. The mean values for each exposure level were statistically compared to those of the control condition using the non-parametric Mann–Whitney U-test. From this comparison, induction factors of specific gene were obtained by comparing each mean value observed in the

275

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

exposed bivalves with that of the control ones. Analyses were performed using Statistica 7.1 software for Windows.

3. Results 3.1. Cd and Hg bioaccumulation Cd and Hg concentrations remained low and constant in gills of control bivalves (660 ± 59 and 428 ± 50 ng.g-1 (dw), mean ± SE, respectively) throughout the experiment (Figure 1, 2). In exposed cockles, Cd concentration increased significantly with time (p < 0.0001) (Figure 1). Cadmium accumulation, not relevant during the twelve first hours, became statistically

significant from T24 (value of Tukey means comparison test, p = 0.0180) and reached 12 436 ± -1 3160 ng.g ,dw (mean ± SE), in gills of cockles at T168 (Figure 1). In Hg-exposed cockles,

bioaccumulation was statistically significant from the first hour (T1) (value of Tukey means comparison test, p = 0.0421) and Hg concentration in gills of cockles increased significantly with

time (p < 0.0001) up to 19127 ± 2508 ng.g-1, dw (mean ± SE) at T168 (Figure 2). Moreover, Hg concentrations were significantly higher than Cd concentrations in gills of exposed cockles (p < 0.0001) throughout the experiment, although concentration of Cd in water was 3.3 fold higher than those of Hg.

18000 Control Cd d 25000 Control 16000 c Hg

14000 20000

, dw , 1

- 12000

, dw , 1 10000 - 15000

[Cd] ng.g [Cd] 8000 10000

6000 ng.g [Hg] cd b 4000 bcd 5000 ab ab ab abc 2000 abc a a a a a aa a a a a 0 0 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180 Time (hours) Time (hours) Figure 1. Concentration of cadmium in gills of Figure 2. Concentration of mercury in gills of control ( ) and exposed ( ) cockles in function of control ( ) and exposed ( ) cockles in function time (mean ± SE, n=5). Letters in bold (a,b,c,d) of time (mean ± SE, n=5). Letters in bold (a,b,c,d) gather similar values (p>0.05). gather similar values (p>0.05).

3.2. Cellular repartition of metals

Metal repartition in gills of cockles was analyzed at T24, T72 and T168 corresponding to times from which accumulation of both metals was significantly higher from control (Figure 1, 2). Metal

276

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque repartition in soluble and insoluble fractions was formulated as a percentage of total metal content in whole gills of cockles (Table 2). Metal repartition differed significantly in function of the nature of metal (p = 0.0211) both in control and exposed cockles. While Cd was equally found in soluble and insoluble fraction, Hg was located especially in insoluble fraction (85-90 %). Repartition of Cd differed between control and exposed bivalves since 67 to 74 % of Cd was found in soluble fraction of gills from control bivalves, whereas this percentage decreased to 49- 53 % in soluble fraction of Cd-exposed bivalves. On the contrary, repartition of Hg between soluble and insoluble fractions was equivalent between control and exposed cockles with an average of 89 % of Hg associated with insoluble fraction.

Table 2. Metal repartition in soluble (supernatant) and insoluble (pellet) fractions expressed as a percentage (mean ± SE, n=5) of total metal content in whole gills of control and contaminated cockles after 24h, 72h and 168h of metal exposure.

Control Contaminated

Supernatant Pellet Supernatant Pellet Cd T24 66.8 ± 2.6 33.2 ± 2.6 53.0 ± 3.2 46.9 ± 3.2 T72 74.3 ± 1.5 25.7 ± 1.5 49.6 ± 6.9 50.4 ± 6.9 T168 73.6 ±1.2 26.4 ± 1.2 48.5 ± 4.5 51.5 ± 4.5 Hg

T24 12.1 ± 2.6 87.9 ± 2.6 13.9 ± 1.3 86.1 ± 1.3 T72 9.3 ± 1.4 90.7 ± 1.4 15.4 ± 3.1 84.6 ± 3.1 T 8.3 ± 0.9 91.7 ± 0.9 9.7 ± 0.8 90.4 ± 0.8 168

3.3. Metallothionein response No significant effect of time (p = 0.1491) was shown on MT response in control cockles, MT concentrations remained low and constant in gills of control bivalves throughout the experiment (Figure 3, 4). In Cd-exposed cockles a slight increase of MT concentration in gills was found at

T72. However, this increase was statistically significant only after 168 hours of Cd exposure and reached 16.7 ± 5.0 nmol sites Hg.g-1, ww at this time (p = 0.0002) (Figure 3). In Hg-exposed cockles, MT concentration remained low and no significant difference appeared in comparison with control bivalves throughout the experiment (p = 0.1523) (Figure 4).

277

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

20 c 10 control control 18 Cd Hg

16 8 ww

1 14

- 1 dw 1 12 - 6 10 b 8 4 6 ab ab nmol sites Hg.g sites nmol 4 ab 2 ab ab Hg.g sites nmol 2 ab ab 0 ab ab 0 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180 time (hours) Time (hours) Figure 3. Metallothionein concentrations in gills of control Figure 4. Metallothionein concentrations in gills of control ( ) and exposed ( ) cockles after Cd exposure (mean ± ( ) and exposed ( ) cockles after Hg exposure (mean ± SE, n=5). Letters in bold (a,b,c,d) gather similar values SE, n=5). (p>0.05).

3.4. Genetic expression The expression of few genes involved in mitochondrial metabolism was analyzed by real time quantitative PCR: the subunit 1 of the cytochrome C oxidase which constitutes the complex 4 of the mitochondrial respiratory channel, the mitochondrial rRNA 12S which gives us an idea of the number of mitochondria in cells and the mitochondrial superoxide dismutase catalyzing the dismutation of superoxide into hydrogen peroxide. Table 3 presented only induction factors statistically different from control (p < 0.05) according to the non-parametric Mann–Whitney U- test.

Cd exposure led to a significant induction of mitochondrial genes such as CoxI (x4), 12S (x2) and

sod (Mn) (x3) after one hour of metal exposure (T1) (Table 3). From T3 to T168, the expression of

these genes was equivalent to control (T3 and T72) or significantly down regulated (T9, T24 and

T168). While no significant modulation of mitochondrial genes appeared at T3 and T72, Cemt1 gene was significantly down-regulated at these times. Finally, inductions of Cemt1 gene by a factor 4.5 and 2 were shown after 24 and 168 hours of Cd exposure, respectively.

Pattern of mRNA expression was significantly different after Hg-exposure. Indeed, no significant variation of mitochondrial genes transcription was shown during the first 72 hours (Table 3) except for 12S and sod (Mn) which were down-regulated at T1. Cemt1 gene was down-regulated

at T1, T3, T24 and T72. At T168, an opposite pattern was found since significant inductions of

278

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque mitochondrial genes (x 7, x 2.5, and x 61.5 for CoxI, 12S and sod (Mn), respectively) were observed in gills of Hg-exposed bivalves in comparison to control, as well as for Cemt1 gene (x6).

Table 3. Differential gene expression observed in gills of cockles after direct contamination. Results are given as significant induction (>1) or repression (<1) factors as compared to control cockles. ― / ‖ indicates no significant variation compared to control cockles.

Cd Hg

Cox1 12S Sod (Mn) Cemt1 Cox1 12S Sod Cemt1

T1 4 2.5 2.5 / / 0.2 0.1 0.3

T3 / / / 0.2 / / / 0.2

T9 0.2 0.1 / / / / / /

T24 0.4 0.1 0.3 4.5 / / / 0.3

T72 / / / 0.4 / / / 0.4

T 0.3 0.4 0.3 2 7 2.5 61.5 5.5 168

4. Discussion Marine molluscs are known to accumulate and tolerate high concentrations of metallic and organic pollutants (Kim et al. 2008). The gut and the gills are associated with the absorption of dissolved compounds, usually observing higher concentrations in the gills (Lagadic et al., 1997), due to the high proportion of metals which are sequestered by MTs in that tissue (Bebianno & Serafim, 1998; Lecoeur et al., 2004). This is supported by the literature, on which gills are represented as a short time storage organ, while in the digestive gland absorption will lead to an accumulation of toxic metals for a longer time (Amiard et al., 1989). Regarding to cadmium and mercury accumulation and elimination, Marigómez et al. (1995) concluded that those processes were associated with lysosomes of the digestive cells. Therefore, visceral mass of bivalves may play significant role on Cd and Hg detoxification processes. However, in this study we only

279

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque focused on gills because preliminary study conducted on both tissues (gills and visceral mass) of cockles demonstrated similar results in terms of MT responses (mRNA and protein) after 2 and 10 days of Cd and Hg exposures (personnal communication). In addition, gills were more appropriated for the present purpose which aims to understand metal toxicity linked to MT responses in a very short time of dissolved metal exposure (four of the six kinetic points were done during the first 24h).

Gills of aquatic molluscs constitute a key interface for the uptake of dissolved metal ions from water which enter via facilitated diffusion or by energy-dependent processes leading to a rapid transport across the epithelium. In addition, the rate of metal uptake in bivalves tissues can be enhanced by increasing formation of mineralized granules (Domouhtsidou and Dimitriadis, 2000) or by the synthesis of intracellular ligands, such as metallothionein (Marigómez et al. 2002). Moreover, reactive ligands such as MT may be induced at a gene level, synthesized in the form of protein and finally excreted and degraded; thus, the turnover and availability of these ligands may ultimately determine the final concentration of the metal in the cells (Simkiss and Mason, 1984). In our experiment, cellular fractionation revealed an average of 50.4 % of Cd associated with soluble fraction in gills of exposed cockles. In the soluble fraction, containing MT, Cd concentration increased from 1.64 nmol.g-1, ww at T24 to 3.66 nmol.g-1, ww at T168 and was followed by a simultaneous increase of MT concentration (1.97 and 16.71 nmol sites Hg.g-1, ww, at T24 and T168, respectively). These concomitant increases suggested an effective sequestration and detoxification of Cd by MT. The significant induction of Cemt1 gene at T24 was in accordance with the accumulation of Cd in gills which became significant from this time. Indeed, the lack of gene induction before 24h could be consequently related to the slight accumulation of Cd before this time, in relation to a basal content of MT sufficient for metals homeostasis and detoxification in gills of cockles. This regulation at gene level could explain the increase of MT protein concentration between 24 and 72h following the earlier induction of Cemt1 at 24h. Globally, Cemt1 gene was induced by pulse in gills of C. edule following Cd exposure, since significant inductions highlighted at T24 and T168 were intersected by time point with significant depression (T72). Such ―pulse-scheme‖ is characterized by intermittent occurrences of change of mRNA expression corresponding to rhythmical induction and repression series. This particular mechanism has been already described by Desclaux-Marchand et al. (2007) and Lemoine and Laulier (2003). Both in vitro and in vivo studies focused on MT

280

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque mRNA expression in function of time and associated results were in accordance with the present work since MT-mRNA expression following metal exposure appeared to be largely dependent of time (Ren et al. 2003a, 2003b; Wang et al. 1994). Ren et al. 2003b showed an early increase of MT gene (3h) followed by a important decline (6h) in liver of rats exposed to Cd exposure. Therefore, the time selected to analyze mRNA levels could be another source of discrepancy and had a particular importance in order to avoid misinterpretation of MT responses to metal exposure. Several studies reported an unchanged MT protein level after Cd exposure whereas MT genes were induced. Therefore, authors warn against the use of MT protein level as indicator of environmental pollution since MT induction at gene level seemed to be more relevant (Bourdineaud et al. 2006). In addition, the present work called attention to the use of only one scale of analysis (protein or mRNA expression) since Cd-induced transcription of MT gene is not only tissue-dependent, but also time-dependent. Consequently, MT mRNA expression and MT protein concentrations have to be investigated together since they give complementary results. Finally, in terms of mitochondrial impact of Cd exposure the precocious induction of mitochondrial genes from T1 in exposed cockles corresponded to a rapid adaptative response of cells with a concomitant induction of coxI and 12S. Induction of these genes sign an increase of energetic supply in response to metal exposure and associated oxidative stress as it is revealed by the induction of sod (Mn) at the same time.

Regarding to the bioaccumulation of metals, it is important to highlight the specific accumulation capacity that cockles display for Hg in comparison with Cd. This is in accordance with the results obtained by Baudrimont et al. (2005) in a geochemical survey (Cd, Zn, Cu and Hg) during which cockles exhibited the highest bioaccumulation of Hg, in comparison with other bivalves (Crassostrea gigas and Ruditapes philippinarum) and metal species (Cu, Zn, Cd). A preliminary study on cockle exposed to Cd and Hg showed a two times higher accumulation of Hg than Cd in the soft tissues after 10 days of experiment whereas exposure concentration of Hg was more than 3 fold lower than Cd (1.5 µg/L-1 vs. 5 µg.L-1 for Hg and Cd, respectively) (Paul-Pont et al. in revision). In this previous work, neither induction of Cemt1 gene nor MT protein induction was found both after 2 and 10 days of Hg exposure despite a strong accumulation of mercury in gills of Hg-exposed cockles. The present study deals with earlier kinetic points and showed similar results. Various hypotheses could be advanced in order to understand the lack of MT response after Hg exposure in C. edule. First, only one isoform of MT was sequenced and analyzed by real

281

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque time PCR in C. edule. Most bivalves presented at least two isoforms of MT such as C. Virginica (Jenny et al. 2004), M. edulis (Géret and Cosson, 2002) and M. galloprovincialis (Zorita et al. 2007). C. gigas presented three different isoforms (Tanguy and Moraga, 2001). It was also suspected that Cemt1 was not the unique isoform of MT in C. edule. Therefore, the fact that mRNA expression has been assessed on only one isoform of MT gene might partly explain discrepancies between mRNA and MT protein responses. Particularly, in the case of Hg exposure, it could be possible that other potential isoforms may respond more specifically to Hg contamination than Cemt1. Indeed, in the zebrafish Danio rerio, (Gonzalez et al. 2006) analyzed the expression of two isoforms of MT gene (mt1 and mt2) in response to metal contamination and showed that mt1 gene was specific to Cd contamination whereas mt2 appears to be common to Cd and Hg contamination. However, in the cockle C. edule even if other isoforms could exist and could be induced more specifically by Hg, an increase of total MT protein should be expected at least at T168. In addition, others hypotheses might explain the lack of MT response after Hg exposure such as (i) the existence of competition between sequestration mechanisms (non-MT cytosolic ligands, biomineralized granules, membranes…) which could interfere with the response of MT to metal exposure, or (ii) an insoluble repartition of Hg unavailable for MT protein. These points are particularly detailed in a previous work (Paul-Pont et al. in revision).

In the present work, cellular fractionation revealed an average of 87 % of Hg associated with insoluble fraction in gills of exposed cockles. Several studies using X-ray analysis and autometallography confirmed the insoluble location of Hg since this metal was mainly found in mineralized granules and in lysosomal vacuolar system (Dimitriadis et al. 2003; Domouhtsidou and Dimitriadis, 2000; Marigómez et al. 2002). Moreover, immunochemistry and ultra structural studies at the transmission electron microscope revealed that MTs were mainly located in the cytoplasm of chloride cells in gills (Alvarado et al. 2007). Hg and MT were localized in different compartments and such result gives us a first clue to understand the non induction of MT by inorganic Hg since the cellular distribution of metal between insoluble and soluble fraction could modulate the amount of metal available for MT sequestration in the cytosol and therefore the induction capacity of MT. However, despite the large insoluble Hg repartition, after 168h of exposure to a concentration of mercury as low as 1.5 µg.L-1, the soluble fraction of gills contained a non negligible Hg concentration which was largely higher than the sequestration ability of MT proteins (2.46 nmol.g-1 Hg against 1.95 nmol sites Hg.g-1, ww). It was not the case

282

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque in Cd-exposed cockles for which the increase of Cd concentration was followed by a concomitant increase of MT concentration in soluble fraction of gills from T24 to T168. Higher Hg concentration compared to MT concentration in soluble fraction indicated that MT had reached their detoxification capacity. Therefore it was consistent to suppose that other cytosolic ligands, such as GSH which is known to bind heavy metals through its –SH group (Rabestein et al. 1985) were involved in Hg detoxification in gills of C. edule. Induction of GSH in bivalves exposed to Hg has been reported in literature (Géret and Cosson, 2002; Yan et al. 1997). Due to the high affinity of Hg for this ligand, it should be very interesting to analyze expression or activity of enzymes implied on GSH biosynthesis pathways such as glutathione synthetase (GSS), glutathione-S-transferase (GST) or glutathione peroxidase (GPX) in further works. Almost at every time point, Cemt1 gene was significantly down regulated in Hg-exposed cockles compared to control. To date, one of the best characterized molecular mechanisms of MT gene activation is the binding of the zinc finger protein Metal Transcription Factor-1 (MTF-1) to the Metal Responsive Element (MRE) in the MT proximal promoter (Saydam et al. 2002). A model has been proposed postulating that inducing concentrations of heavy metals or oxidative stress can fix to MT and thus displace Zn from MT, increasing the pool of free Zn available for activating MTF-1 and stimulating its DNA binding and transactivation capabilities (Zhang et al. 2003). With such model, the non activation of Cemt1 gene following Hg-exposure in the cockle could be explained by the incapacity of Hg to bind to MT protein, displace Zn and thus activate MTF-1. However, the hierarchical sequence of in vitro MT affinity for metals showed that inorganic Hg is placed in first position and could displace from the MT protein a large number of metals such as Cd, Zn or Cu (Amiard et al. 2006). Moreover, several experiments are not matching with this model and many contradictions occurred suggesting that other ways of MT gene activation could exist (Bourdineaud et al. 2006; Marie et al. 2006). In this context, further analyses will be necessary to discriminate whether inorganic Hg was unable to bind to MT protein in vivo or whether more complex pathways were responsible of the lack of MT gene activation by Hg. This lack of induction at gene level despite an expected fixation of Hg to MT protein in soluble fraction was also demonstrated by (Leiva-Presa et al. 2004) which revealed that MeHg+ was efficiently sequestrated by MT in vitro without any concomitant induction of MT mRNA expression. In our experiment, an exception was found at T168 wherein Cemt1 was significantly up regulated (x 5.5) in comparison with control. At this time, significant induction of mitochondrial genes and especially sod (Mn) revealed a toxic impact of Hg on mitochondria

283

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque activity and / or integrity as it has been already demonstrated in both in vitro and in vivo studies (Burlando et al. 2004; Cambier et al. 2009; Limke et al. 2003). It is now well documented that MT have a role in the protection of tissues and organs from oxidative stress (Andrews, 2000; Viarengo et al. 2000), then it was consistent to relate the Cemt1 induction more to the apparition of an important oxidative stress following Hg-exposure highlighted by strong induction of sod

(Mn) gene (x 61.5) than to a direct effect of Hg. Finally, until T168, no impact on mitochondrial genes was shown following Hg exposure, despite the lack of MT response in regard with the precocious and important mercury accumulation in gills of C. edule. This result supported the hypothesis of the effective implication of other cellular ligands in the sequestration of cytosolic fraction of mercury as it was discussed above. 5. Conclusion Exposure to Cd and Hg led to important differences between metals in terms of accumulation, cellular distribution and MT protein response and genes expression from the first hours. Induction of Cemt1 gene by Cd was realized by pulse and was consequently time dependent since significant inductions were intersected by down-regulations or absence of significant variations from control. Following Cemt1 gene induction, results showed a gradual increase of MT protein which was well correlated to the increase of Cd concentration in soluble fraction throughout the experiment. On contrary, Hg which was localized mainly in insoluble fraction (85-90 %) did not lead to direct induction of MT both at gene and protein levels and generated a strong toxicity on mitochondria after 7 days of experiment. Further analyses, particularly on glutathione dependent pathways, will be necessary to understand more specifically the lack of Hg-induction of Cemt1 gene and the kind of cellular ligands involved in Hg sequestration and detoxification in the cockle C. edule.

Acknowledgments

This work was carried out with the financial support of the ―Région Aquitaine‖ and the French National Research Agency with the funds of the ―multistress project‖. The authors are grateful to sailor Francis Prince and Henri Bouillard for their valuable help on the field.

284

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque

Conclusion chapitre V

La caractérisation de la séquence codante complète du gène de métallothionéine de coque (Cemt1), ses régions 3‘-5‘ non codantes ainsi que son organisation au niveau génomique ont révélé de nombreuses particularités telles que la présence de deux allèles différents dus à des phénomènes de délétions dans les séquences introniques. De même, la présence de motif Cys- Cys a été révélée alors qu‘il est généralement absent des MT de bivalves, ce qui permet de distinguer le domaine α en amont du domaine β présent dans la partie C-terminal. Toutefois, la détermination précise du nombre d‘atomes métalliques pouvant être fixés par cette isoforme ne peut pas être déduite directement de cette séquence. Une analyse de la protéine par cristallographie ou rayon-X serait nécessaire pour déterminer la capacité de séquestration de l‘isoforme 1 de manière rigoureuse. Malheureusement, peu d‘éléments de régulation dans la partie 5‘ non codante ont pu être identifiés. Ces éléments de régulation peuvent être activés par les métaux, les glucocorticoïdes (réponse hormonale et inflammatoire) ou encore les interférons (système immunitaire) afin de réguler l‘expression du gène de MT (Daniels et Andrews, 2003). L‘identification de telles séquences régulatrices sur Cemt1 serait en outre extrêmement utile à la compréhension de la réponse de ces protéines dans des situations de multistress « polluants x pathogènes ». Enfin, comme il a été démontré chez de nombreuses autres espèces, l‘existence d‘autres isoformes de ce gène, en relation avec des affinités particulières par rapport à certains métaux / tissus / voies métaboliques est suspectée chez C. edule. L‘observation de la réponse d‘un organisme soumis à un stress, une contamination métallique dans le cas présent, à l‘échelle moléculaire, i.e. expression du gène, requiert certaines précautions comme le soulignent nos travaux. Plusieurs études ont déjà rapporté des divergences entre l‘expression de l‘ARNm et la synthèse effective de la protéine qui peuvent être dues à la fois : - A un décalage temporel évident entre l‘expression de l‘ARNm et la traduction en protéine. - Aux mécanismes de régulation post-transcriptionnel : défaut de maturation de l‘ARNm dû à des phénomènes d‘épissage aberrants ; dégradation de l‘ARNm avant traduction ou arrêt de la traduction. En plus de son caractère précoce (en amont des autres voies métaboliques) et transitoire, l‘expression de certains gènes apparaît de plus fortement dépendante du temps, comme le

285

Chapitre V. Approfondissement de la réponse des métallothionéines chez la coque souligne le schéma pulsatile d‘induction du gène Cemt1 suivant une exposition au Cd. Ainsi, l‘expression d‘un gène à un temps t n‘a que peu de pertinence, s‘il n‘est pas observé dans un cadre cinétique et/ou rapporté à la réponse protéique à ce même temps t. La comparaison des réponses génétiques et protéiques réalisée dans ces travaux à court et moyen termes, met donc en lumière le caractère complémentaire de ces approches, et la nécessité de coupler les échelles d‘observation afin d‘interpréter de manière juste la réponse d‘un organisme. De la même manière lorsqu‘on s‘intéresse à la réponse d‘une protéine enzymatique, l‘analyse de son activité apparaît nécessaire à l‘approche génétique et protéique. En effet, les divergences observées entre expression génétique et protéique s‘observent de la même manière entre expression protéique et activité enzymatique (Miyamoto et al., 2001). L‘analyse des réponses des organismes à la fois à l‘échelle génétique, protéique et enzymatique constitue l‘approche la plus rigoureuse et la plus fiable. Enfin, ces travaux révèlent une absence totale de réponse des MT au niveau génétique et protéique suite à la contamination par le mercure, et ce malgré une très forte bioaccumulation de ce métal dans les tissus du bivalve. Dès lors, afin d‘améliorer la compréhension des processus de détoxication impliqués dans la prise en charge du mercure chez la coque, l‘exploration de la voie glutathion-dépendante à ces différents niveaux (gène, protéine, activité enzymatique) pourrait être envisagée.

286

Chapitre VI. Synthèse générale et perspectives de recherche

287

288

Chapitre VI. Synthèse générale et perspectives de recherche

Synthèse générale et conclusions

Les écosystèmes littoraux semi-abrités, de type lagune, estuaire ou baie, sont à la fois le siège d‘une intense production biologique, le terrain d‘une exploitation de ressources vivantes, et le réceptacle de divers contaminants d‘origine anthropique. Ces diverses applications favorisent l‘occurrence de perturbations multiples dans ces milieux et notamment la prolifération d‘organismes pathogènes, l‘apparition de maladies, et le rejet de contaminants d‘origine anthropique. C‘est dans ce contexte particulier que s‘inscrit cette problématique de thèse dont les travaux avaient pour objectif d‘appréhender les interactions entre polluants métalliques, parasites trématodes et bactéries pathogènes sur deux espèces de mollusques bivalves exploités, la coque Cerastoderma edule et la palourde japonaise Ruditapes philippinarum. Ceci a été réalisé à travers le développement de plusieurs approches faisant appel à différents niveaux d'intégration (moléculaire, protéique, cellulaire, populationnel) et à différents champs disciplinaires, s'intéressant aussi bien aux réponses immunitaires, aux processus de bioaccumulation et détoxication métallique ou encore à la régulation transcriptionnelle de certains gènes impliqués dans le métabolisme mitochondrial, la réponse au stress oxydant et la détoxication. L‘ensemble des travaux réalisés souligne à l‘évidence la complexité des interactions existantes entre les organismes pathogènes, bactéries et trématodes, et les polluants métalliques chez les mollusques bivalves. Néanmoins, nous allons tenter, au cours de cette synthèse générale, de dégager les résultats les plus marquants. Un effort de synthèse sur l‘ensemble des paramètres révélés jouant un rôle clé sur le jeu d‘interactions « polluant-hôte-pathogène » apparaît indispensable. Ces différents mécanismes sont illustrés dans la figure synthétique VI.1 et seront discutés tout au long de cette conclusion générale.

Les différences interspécifiques de réponse aux stress ont été observées à travers deux expériences triple stress (Cd x trématode x bactérie) conduites en laboratoire sur la coque et la palourde. (Chapitre IV-C). Incontestablement, les interactions découlant de ces trois sources de perturbation sont dépendantes de l’espèce hôte, d‘une part au niveau de la nature des interactions engendrées (i.e. antagonisme, neutralisme, synergisme) et d‘autre part en termes de sensibilité aux stress et de réponse adaptative mise en place par les organismes. Ces différences, nombreuses, ont été précédemment discutées en conclusion du Chapitre IV-C et illustrent parfaitement la difficulté de dégager de ce type de contexte multifactoriel des règles simples, généralisables et applicables ne serait-ce qu‘au sein d‘un même groupe taxonomique. Toutefois,

289

Chapitre VI. Synthèse générale et perspectives de recherche

Environnement Facteurs environnementaux T°C, Salinité, ressources trophiques, etc.

Hôte -Espèce -Statut: Chronologie - native / introduite Chronologie des stress - sauvage / cultivée des stress 1 - sensible /résistante 2 -Histoire de vie 2 1

Immunotoxicité Immunotoxicité Etat de santé Polluant Pathogène Exposition - Nature hôte amont, larve - Nature -Exposition: - Espèce/Stade - Aigu / chronique - Intensité - Favorisation?

Biodisponibilité Viabilité Dispersion Occurrence Dispersion Facteurs environnementaux T°C, Salinité, ressources trophiques, etc.

Figure VI.1 Synthèse des différents mécanismes prépondérants dans les interactions « polluant-pathogène »

290

Chapitre VI. Synthèse générale et perspectives de recherche certains mécanismes sont retrouvés quel que soit l‘espèce hôte, pouvant constituer des éléments de base, si ce n‘est extrapolables directement à d‘autres modèles, du moins à prendre en considération lorsque qu‘on se place dans une telle problématique.

Au travers des diverses expérimentations effectuées, il est apparu que l’infestation par le trématode chez une espèce hôte, sensible (C. edule) ou résistante (R. philippinarum), n’est pas perturbée par la présence d’autres stress : (i) une contamination au cadmium (Cd), à des concentrations chroniques et représentatives des niveaux de contaminations retrouvées dans de nombreuses rivières et estuaires particulièrement impactés par des pollutions métalliques (Audry et al. 2004; Feng, 2005; Ferrara et al. 1998; Mohamed et al. 1998; Nascimento et al. 2006) ; (ii) la présence de la bactérie Vibrio tapetis. Un tel résultat révèle en premier lieu l‘absence d‘effets néfastes de ces stress sur la sensibilité de l‘hôte à l‘infestation parasitaire. Il aurait en effet pu être envisagé une possible immunomodulation de l‘organisme hôte suite à l‘exposition métallique ou bactérienne, le rendant ainsi plus susceptible à l‘infestation parasitaire, phénomène qui n‘a pas été révélé dans nos conditions d‘expérimentation. De plus, ce résultat met aussi en évidence l‘absence d‘effets délétères de ces sources de stress, seules ou en combinaison, sur l‘ensemble des paramètres permettant au parasite d‘infester son hôte, à savoir son comportement général, et plus précisément sa capacité de nage, d‘orientation et de reconnaissance, ainsi que sa capacité à pénétrer les tissus de l‘hôte et à s‘y enkyster (Morley et al., 2003a,b, 2004). Il convient néanmoins de noter que nos expériences réalisées impliquaient l‘exposition directe de la larve de trématode et n‘ont cependant pas permis d‘appréhender les effets toxiques de telles sources de stress sur l‘hôte amont qui peut être lui aussi affecté par ces perturbations biotiques et abiotiques. Dans les milieux naturels la présence de l‘ensemble des hôtes, définitif, premier et deuxième intermédiaire, dans un même biotope est indispensable à l‘accomplissement du cycle parasitaire des digènes. Dès lors, une pollution métallique et/ou la présence d‘autres agents pathogènes peuvent entraîner des impacts non négligeables sur l‘ensemble des acteurs du cycle, ce qui pourra interférer sur la réussite de la transmission du parasite d‘un hôte à l‘autre. Si l‘on s‘intéresse à l‘autre espèce de pathogène étudié, la bactérie Vibrio tapetis, une fois encore le manque de technique quantitative précise nous a empêchés de déterminer de manière fiable la charge en bactérie dans les tissus et les fluides des deux espèces de bivalves. Cette lacune constitue l‘un des plus importants verrous rencontrés lors de ces expérimentations, puisqu‘il est alors impossible de déterminer l‘effet des autres sources de stress (Cd, trématode)

291

Chapitre VI. Synthèse générale et perspectives de recherche directement sur la bactérie pathogène (viabilité, développement, avancée de l‘infection, etc.) et d‘évaluer de possibles interactions antagonistes ou synergiques sur l‘infection bactérienne, en lien avec des réponses particulières dans les conditions de stress multiples. Concernant l‘exposition au Cd de bivalves « sains », « infestés » ou « co-infestés », les expériences triple stress ont mis en évidence une forte modification de la bioaccumulation métallique en présence d’un ou plusieurs pathogènes et ce, chez les deux espèces de bivalve. Les mécanismes sous-jacents à ces différences de bioaccumulation métallique demeurent néanmoins complexes puisque la nature des effets différe selon (i) le protocole d‘expérimentation, (ii) le(s) pathogène(s) concernés et (iii) l‘espèce hôte. (i) Au sein de la même espèce (C. edule), la manière dont va être pratiquée l‘infestation expérimentale semble influencer l‘impact de cette dernière sur l‘accumulation du polluant par l‘organisme hôte. Dans l‘expérience du Chapitre IV-A, l‘infestation expérimentale par les trématodes a été effectuée en une seule fois en début d‘expérimentation (T0) 4h avant la contamination des unités par le Cd, alors que lors des expériences du Chapitre IV-C l‘infestation expérimentale a été réalisée au T0 de manière synchrone à la contamination métallique (et/ou bactérienne) des unités, et a été maintenue sur plusieurs jours (T0, T1, T3, T4). Cette pression d‘infestation maintenue sur plusieurs jours a pu entraîner des réactions de défense plus marquées du bivalve conduisant à une modification de la physiologie et/ou du comportement, et amenant in fine à une perturbation de l‘accumulation du Cd plus prononcée. La pression d‘infestation (ponctuelle vs. continue) et l‘ordre d‘application des stress apparaissent comme des facteurs importants. (ii) Par ailleurs, la présence d‘un ou deux types de pathogènes dans le milieu n‘est pas sans effet puisque les bivalves co-infectés montrent des accumulations différentes des individus infectés par un seul pathogène. Dans le cas de la coque, la bioaccumulation était réduite en présence d‘un pathogène (trématode ou bactérie) mais rétablie à un niveau similaire aux individus non infectés en condition de triple stress (Cd x trématode x bactérie). A l‘inverse chez la palourde, la présence d‘un seul pathogène semble n‘avoir que peu d‘effet alors que la co- infection entraînait un effet fortement synergique sur l‘accumulation de Cd dans les tissus de l‘hôte. Ces accumulations différentes en présence de plusieurs pathogènes (triple stress) sont vraisemblablement reliées à une modification de la physiologie de l‘organisme hôte par rapport aux situations moins stressantes (i.e. un pathogène unique). Ces différences peuvent être imputées à une augmentation des besoins énergétiques de l‘hôte pour mettre en place des

292

Chapitre VI. Synthèse générale et perspectives de recherche mécanismes de défense effectifs, impliquant ainsi une augmentation de l‘activité de filtration et du métabolisme général. Toutefois, un affaiblissement des individus en condition de triple stress peut également être envisagé et expliquerait une exposition accrue des organismes par inhibition du réflexe de fermeture des valves en présence d‘un contaminant par exemple. En effet, en milieu naturel les coques parasitées ont tendance à rester plus souvent à la surface du sédiment, les valves ouvertes, facilitant à la fois une contamination potentielle via le manteau et le pied ainsi que la prédation de l‘individu infesté par l‘hôte définitif (Desclaux, 2003). Il nous est pour l‘heure impossible de discriminer de manière sûre ces deux mécanismes dans les niveaux d‘accumulation particuliers observés en conditions de triple stress. Des études bioénergétiques (mesure de l‘intégrité et du fonctionnement des complexes de la chaîne respiratoire mitochondriale, de la production/consommation d‘ATP ou encore analyses par micro- respirométrie) permettrait de statuer quant à l‘état énergétique des animaux dans ces conditions particulièrement stressantes. Ce type même de résultat obtenu dans des conditions extrêmement contrôlées souligne la complexité des schémas pouvant se profiler in situ et par conséquent la nécessité de prendre en compte la présence des organismes pathogènes et des maladies affectant les populations naturelles dans le cadre des réseaux de surveillance utilisant les niveaux de bioaccumulation dans les mollusques bivalves comme indicateurs de pollution. Ces phénomènes peuvent alors conduire à des interprétations erronnées de la contamination du milieu ambiant.

En plus de perturber les niveaux d‘accumulation métallique, la présence d’organismes pathogènes peut aussi interférer dans les mécanismes de détoxication cellulaire. En effet, en ce qui concerne l‘analyse des métallothionéines en condition de multiple stress, les diverses expérimentations ont donné lieu à des interactions complexes, avec toutefois certains processus similaires entre les différents couples hôte-pathogène observés. En effet, en dehors de toute contamination métallique, la co-infection trématodes-bactéries a engendré un effet antagoniste sur la réponse des MT, et ce chez nos deux espèces. Cette inhibition de la synthèse des MT en condition de double stress constitue un processus précédemment observé dans d‘autres études utilisant la coque comme second hôte intermédiaire (Baudrimont et al., 2006 ; Baudrimont et de Montaudouin, 2007). Ce type d‘effet confirme le caractère délétère de chaque pathogène chez des espèces réputées résistantes (palourde vs. H. elongata) ou uniquement porteuse saine (coque vs. V. tapetis). Un tel effet antagoniste peut d‘autre part exacerber la

293

Chapitre VI. Synthèse générale et perspectives de recherche mortalité des populations de bivalves vivant dans des sites pollués, à travers l‘altération des mécanismes de détoxication comme les MT. Ainsi, l‘utilisation de ces protéines dans le cadre d‘études écotoxicologiques sur le terrain implique impérativement la prise en compte des organismes pathogènes et des maladies qui peuvent fortement moduler cette réponse protéique. L‘étude multistress chez la coque a de plus mis en évidence de fortes divergences de réponse des MT entre l’échelle génétique et protéique (Chapitre IV-C). Des expériences de simple stress menées ultérieurement à la fois sur le gène et la protéine ont amené des éléments de réflexion quant à l‘échelle d‘observation à adopter afin d‘interpréter de manière juste la réponse de cette protéine (Chapitre V). En effet, la comparaison de ces deux types de réponse réalisée dans ces travaux à plus ou moins court termes, met en lumière le caractère complémentaire de ces approches, et la nécessité de coupler les échelles d‘observation afin d‘interpréter de manière juste la réponse d‘un organisme. Lorsqu‘il s‘agit d‘une enzyme, la mesure de l‘activité en parallèle aux autres approches est nécessaire puisque des décalages apparaissent de la même manière entre la synthèse protéique et l‘activité réelle de l‘enzyme. L‘importance et la pertinence des études cinétiques est aussi révélée par ces travaux tant au niveau de l‘expression d‘un gène, qu‘au niveau de la réponse protéique. L‘interprétation et l‘extrapolation de réponse observées à un temps t doivent être effectuées avec une grande vigilance, ce qui est aussi le cas lorsqu‘on s‘intéresse aux paramètres immunitaires. En effet, une étude conduite par Hégaret et al. (2007) montre d‘importantes réponses immunitaires chez la palourde R. philippinarum après 3 jours d‘exposition à une algue toxique alors qu‘aucun effet n‘apparaît sur ces mêmes paramètres après 6 jours d‘exposition. Des résultats similaires ont été observés dans le cadre de certaines expériences en laboratoire (résultats non présentés). Les auteurs suggèrent un retour rapide de ces paramètres à l‘homéostasie après la mise en place d‘une réponse effective. Ce genre d‘étude souligne typiquement l’importance du temps d’observation et de la cinétique. Si l‘on revient aux réponses immunitaires des espèces hôtes dans un contexte multistress, là encore des interactions variées ont été révélées. Cette approche expérimentale multistress a néanmoins pu mettre en évidence une implication particulière de ces paramètres immunitaires dans les couples « hôte-pathogène » spécifiques, en accord avec la sensibilité de chaque espèce à son pathogène propre (H. elongata vs. coque ; V. tapetis vs. palourde). Toutefois, l‘interprétation fiable des réponses immunitaires observées au sein de chaque espèce peut demeurer difficile. Une augmentation de la concentration et des proportions hémocytaires dans l‘hémolymphe après infection par un pathogène peut être analysée comme une réponse efficace

294

Chapitre VI. Synthèse générale et perspectives de recherche de l‘organisme dans la lutte contre un pathogène, aussi bien qu‘une diminution de ces deux paramètres qui peut être interprétée comme une migration des cellules sur le site de l‘infection. Un autre exemple illustrant la complexité d‘interpréter correctement ce type de réponse correspond à l‘absence d‘effets marqués de l‘infection par la bactérie V. tapetis sur l‘ensemble des paramètres hémocytaires chez la coque et ce malgré une forte mortalité (100%) des individus dans ces unités expérimentales uniquement (V, HV, CV, CHV) à partir du neuvième jour d‘expérimentation. On pourrait alors relier la mortalité des coques à cette apparente absence de réponse effective mise en place par le système immunitaire, sans toutefois noter de forte pathogénicité de la bactérie sur les hémocytes, phénomène classiquement observé suite à l‘infection par V. tapetis (notamment perte de l‘adhésion et diminution de la viabilité des hémocytes). Ce type même d‘ambiguïtés révèle le caractère complexe de la compréhension des changements de concentrations et activités cellulaires, et ce particulièrement dans des contextes de multistress. Ainsi, les approches multiparamétriques constituent des atouts importants dans l‘interprétation correcte de la capacité de défense immunitaire du bivalve aussi bien en laboratoire que sur le terrain. Dans le cadre de cette thèse, l‘implication des facteurs humoraux via leurs activités antimicrobiennes n‘a pas été analysée mais peut constituer un élément de compréhension supplémentaire quand à la réponse ou non-réponse des bivalves en condition de multistress. Ceci est particulièrement vrai pour la coque dont l‘hémolymphe possède un large et puissant spectre d‘activités antimicrobiennes comparées à d‘autres espèces de mollusques (Defer, 2009), même si ces facteurs humoraux demeurent plus mineurs chez les mollusques que les processus cellulaires (Roch, 1999). Malgré cette approche uniquement cellulaire, les travaux menés en laboratoire ont tout de même révélé des interactions fortes et propres à chaque espèce de bivalves. L‘interaction « Cd x trématode » est par exemple particulièrement prononcée chez la coque sur certains paramètres immunitaires (concentration hémocytaire maximale et flambée oxydative 15 fois plus forte) mais aussi génétiques (métabolisme mitochondrial). Ceci est en accord avec les résultats du Mussel Watch Program (Kim et al., 2008) qui dénote in situ l‘importance de l‘interaction « métal x trématode » parmi de nombreux autre polluants (24 dont métaux, HAP, PCB, pesticides) et pathogènes (8 genres différents) étudiés. Ces résultats démontrent de plus l‘importance de prendre en compte l‘impact des infestations métacercariales, souvent moins considérées que les infestations par le stade sporocyste du fait de leur caractère apparemment peu délétère, dans des contextes environnementaux soumis à des pollutions métalliques. De même, malgré une faible

295

Chapitre VI. Synthèse générale et perspectives de recherche infestation chez la palourde, espèce résistante, les parasites trématodes peuvent influencer les mécanismes de bioaccumulation et de défense, ce qui confirme l‘importance de prendre en compte des pathogènes apparemment peu ou pas délétères dans le cadre d‘un contexte multistress. L‘expérience triple stress a ainsi mis en lumière une forte interaction « trématode x bactérie » sur l‘accumulation métallique, les paramètres protéiques et immunitaires de R. philippinarum. Dans ce cadre particulier, le manque d‘outil pour quantifier précisément les bactéries dans les fluides et les tissus de l‘hôte nous a empêchés de clarifier les raisons sous- jacentes d‘une telle interaction. Néanmoins, des mécanismes d‘augmentation des infections bactériennes avec la présence concomitante de macroparasites ont précédemment été décrits (Cusack et Cone, 1986), et sont envisagés dans nos études, soulignant l‘influence de la présence dans le milieu d‘autre pathogènes, même peu délétères, sur la sensibilité d‘un organisme à une maladie ou un organisme pathogène particulier. Ces expériences multistress ont été menées chez nos deux espèces modèles à travers l‘exposition à un ou deux organismes pathogènes présentant des hôtes spécifiques différents, et l‘exposition à un seul élément métallique, le cadmium. Aucun autre métal n‘a été testé dans le cadre d‘expériences en stress multiples bien que des sensibilités différentes de l‘hôte et du pathogène puissent être incontestablement envisagées. En effet, les expériences en stress simple du Chapitre V révèlent de fortes différences en termes d‘accumulation et de mécanismes de détoxication mis en place chez des coques exposées au Cd ou au mercure (Hg). Chez cette espèce, l‘accumulation du Hg est bien plus élevée que celle du Cd et les mécanismes de détoxication à une contamination mercurielle emprunte des voies métaboliques différentes chez la coque, avec notamment une absence de réponse des MT (au niveau protéique et génétique). Dans ce contexte, les effets des organismes pathogènes sur les mécanismes de détoxication mis en place chez la coque seront différents et non intuitifs. Le Hg est largement reconnu pour ses effets immunotoxiques (Fournier et al., 2001 ; Gagnaire, 2005) mais l‘impact du Hg sur les pathogènes, ainsi que la nature de leurs interactions demeurent inconnues. De même, un mélange de plusieurs contaminants peut interférer sur l‘ensemble des paramètres observés. Un exemple peut être avancé avec l‘exposition concomittante au Cd et au Zn. De nombreuses études éctoxicologiques se sont intéressées à l‘exposition simultanée à ces deux métaux, phénomène connu pour interférer dans la bioaccumulation du Cd chez de nombreux bivalves (Marie et al., 2006). Une étude conduite par Morley et al. (2005) a de plus mis en évidence une toxicité

296

Chapitre VI. Synthèse générale et perspectives de recherche particulière du mélange Cd-Zn sur la durée de vie de cercaires (Diplostomum spathaceum) en comparaison avec l‘exposition au Cd ou au Zn de manière isolée. Un des intérêts de ces expériences en laboratoire était d‘améliorer la compréhension des processus observés lors de l‘étude in situ. La sensibilité de chaque espèce de bivalve aux pathogènes et aux métaux, particulièrement le Cd, ainsi que leur niveau de base sur certains sites ont pu être confirmés en laboratoire, comme il a été précédemment discuté dans la conclusion du Chapitre IV. Ces expériences avaient aussi pour objetcif de mettre à jour des interactions particulières difficilement identifiables sur le terrain du fait de la multitude de paramètres mis en jeu. Ceci a été réalisé à travers quatre expériences tenant compte de la saison, de l‘espèce et de son origine. La mise en évidence en laboratoire d‘interactions particulières engendrant des réponses fortes des organismes, telles que l‘interaction « Cd x trématode » chez la coque, permet de poursuivre la réflexion autour de la notion de « hot spots ». En effet, alors que Landéda apparaît déjà comme un site « critique » dans lequel la présence de la BRD, mais aussi d‘autres paramètres environnementaux posent un sérieux problème quant à la survie et au maintien des populations, la situation d‘Arguin est considérée comme bien plus favorable en termes de dynamique de population des bivalves. Toutefois, dans un contexte riche en parasites trématodes digènes (13 espèces différentes, prévalence de 100% chez les individus adultes, de Montaudouin et al., 2000), l‘apparition, même ponctuelle et de faible niveau, d‘une contamination métallique pourrait perturber de façon importante les populations de bivalves y résidant en raison de la présence d‘un stress parasitaire majeur touchant déjà ces populations. L‘interaction « métal x trématode », affectant notamment les mécanismes de détoxication (MT) et la production d‘ERO par le système immunitaire, peut ainsi engendrer de lourds dommages cellulaires soit par une toxicité directe (métal) ou indirecte (apparition d‘un stress oxydant majeur). Dans cette logique, le banc d‘Arguin pourrait être considéré comme un « hot spot » dans lequel la présence d‘un stress parasitaire, même modéré, peut être la cause d‘une aggravation non négligeable de l‘état de santé des populations dans le cas d‘une contamination métallique du milieu. Andernos constitue quant à lui un site critique en termes de stress identifiés (perkinsus, métaux) et de dynamique de populations (densité, croissance, recrutement faibles) même si une adaptation locale des populations de palourdes au stress métallique semble être mise en place. En effet, l‘expérience du Chapitre IV-B même si elle n‘est pas directement reliée aux populations étudiées lors du suivi sur le terrain (palourdes d‘Arguin) met en évidence que le contexte environnemental et l’histoire de vie d‘une espèce peut fortement influencer sa sensibilité aux sources de perturbations

297

Chapitre VI. Synthèse générale et perspectives de recherche biotiques et abiotiques. De même, cette influence environnementale peut, au-delà de l‘hôte seul, affecter le couple « hôte-parasite » comme le suggère l‘étude menée par Morley et al. (2003a) qui montre des abondances parasitaires différentes selon l‘origine géographique des souches « hôte- parasite » après exposition à un même polluant métallique. Enfin, l’influence du cycle de vie de l’espèce semble dominer l‘impact d‘infestation parasitaire ou d‘expositions sub-léthales à des contaminants métalliques sur la bioaccumulation de Cd et la synthèse des MT (Chapitre IV-A). Cela souligne la nécessité de standardiser la période d‘échantillonnage in situ lorsque cette protéine est utilisée à des fins de marqueurs d‘exposition métallique. De même puisque l‘influence du cycle de vie constitue un paramètre majeur, l‘analyse des résultats issus du suivi doit être envisagée saison par saison et site par site au sein de chaque espèce.

A l‘heure actuelle, l‘extrapolation des résultats issus d‘expériences en laboratoire aux mécanismes observés in situ n‘est que peu concluante dans ce type de problématique multistress (Morley, 2009). Les conditions sur le terrain dans des sites pollués présentent un tableau bien plus complexe que les études en laboratoire du fait de la quantité de facteurs confondants influençant ces interactions. En effet, les facteurs environnementaux (température, salinité, oxygène, etc.) jouent un rôle prépondérant sur :

- le métabolisme général des organismes : cycle de reproduction, statut énergétique, etc. - l‘occurrence, la dispersion, le développement et la viabilité des organismes pathogènes ainsi que la transmission des maladies ; - l‘occurrence, la dispersion, la biodisponibilité et les effets toxiques des contaminants.

Dans un tel contexte, interpréter et anticiper les risques des interactions polluants- pathogènes dans un environnement en évolution modérée ou rapide risque de s'avérer difficile. Par conséquent, le rôle significatif que jouent les pathogènes / maladies dans la structuration de la dynamique des populations des organismes aquatiques doit être reconnu par les écotoxicologues, notamment dans le cadre des réseaux de surveillance des milieux littoraux.

En conclusion, notre étude témoigne à l‘évidence que la réponse des bivalves aux sources de stress dépend des espèces et des combinaisons subies. La réponse des bivalves à plusieurs stress ne se déduit pas intuitivement à partir des réponses « mono-stress » ce qui souligne la notion d‘interactions. La complexité de ces dernières rend difficile la mise en évidence de

298

Chapitre VI. Synthèse générale et perspectives de recherche processus généraux et extrapolables, même si certains mécanismes semblent prédominer. Enfin, ces travaux confirment la nécessité d‘observer les réponses des organismes à différentes échelles dans une telle problématique, à travers des approches à la fois in situ et expérimentales, la mise en place d‘outils multiparamétriques et la complémentarité des observations au niveau génétique et protéique afin d‘avoir une vision globale des processus interactifs existants entre polluants et pathogènes.

Perspectives de Recherche

Notre travail a mis en évidence certains mécanismes importants interférant sur les relations hôte-pathogène-polluant qui méritent d‘être pris en considération tant dans le cadre d‘études mesurant l‘état de santé des populations que dans le cadre d‘approches plus écotoxicologiques sur le suivi des contaminants et de leurs effets dans l‘environnement. Il y a encore beaucoup à apprendre sur le parasitisme et les maladies infectieuses chez les populations sauvages exposées à des polluants. Les perspectives qui se dégagent de ce travail sont donc nombreuses et concernent aussi bien les interactions particulières entre polluants et pathogènes mises en lumière chez nos deux mollusques dans le cadre de ces travaux en laboratoire, que l‘extrapolation des effets marquants dans un contexte plus général.

Comme nous l'avons évoqué précédemment, des interactions spécifiques sont apparues chez nos espèces hôtes. Afin d‘aller plus loin dans l‘interprétation et la compréhension de tels mécanismes, le développement d‘outils supplémentaires apparaît essentiel, et particulièrement :

- Le développement d‘outils moléculaires de quantification de la bactérie V. tapetis, comme la synthèse d‘amorces spécifiques permettant de quantifier cette espèce précise par PCR quantitative. En effet, l‘interaction « trématode x bactérie » entraîne de forts impacts tant au niveau de la bioaccumulation de Cd, que sur les processus de détoxication et de réponse immunitaire chez la palourde R. philippinarum, espèce particulièrement sensible à Vibrio tapetis. Dès lors, la mise en place de tels outils est en effet indispensable au suivi de l‘infection bactérienne dans les fluides et les tissus de l‘hôte en amont de l‘apparition des premiers symptômes, et à la détermination de l‘effet des autres sources de stress (Cd, trématode) directement sur la bactérie pathogène (viabilité, développement, avancée de l‘infection, etc.). Cette nécessité est entre autres révélée par la forte interaction « trématode x

299

Chapitre VI. Synthèse générale et perspectives de recherche

bactérie » chez une espèce réputée pourtant résistante aux parasites trématodes digènes. De même, ces outils pourraient permettre de comprendre la forte mortalité de coques infectées par cette bactérie (Chapitre IV-C) et de statuer plus précisément sur le mélange de ces différents pathogènes dans des populations déjà stressées.

- En ce qui concerne la coque, l‘interaction « Cd x trématode » donne lieu à des effets conséquents en termes de réponse immunitaire, en contradiction avec une accumulation de Cd plus faible que les individus non parasités. Il convient notamment de s‘intéresser à la production d‘ERO très importante dans les hémocytes de ces individus, en contradiction avec ce qu‘il est observé chez les individus uniquement contaminés ou parasités. Au vu de l‘implication particulière de ces intermédiaires réactifs dans la lutte contre un pathogène, des mécanismes de protection anti radicalaires efficaces doivent être mis en place de manière concomitante par l‘hôte infesté afin d‘éviter qu‘une partie de ces radicaux libres n‘échappent aux systèmes de protection et ne causent des dégâts importants. De même, les métaux peuvent engendrer un stress oxydant dans la cellule par génération de molécule oxygénées réactives ou par atteinte des systèmes enzymatiques de protection contre le stress oxydant. Dans le cadre de l‘interaction « métal x trématode » un dépassement possible des mécanismes de réponse au stress oxydant peut être la cause de dommages cellulaires importants. Ceci peut constituer une piste pour comprendre la forte interaction entre ces deux facteurs, observée chez C. edule mais aussi chez d‘autres espèces comme M. edulis et C. gigas. Dans ce contexte, il serait très intéressant d‘élargir l‘analyse génétique à d‘autres marqueurs de réponse au stress oxydant et de compléter cette analyse par les dosages enzymatiques correspondants reflétant de manière plus fine ce qu‘il se passe dans la cellule (catalase, superoxyde dismutase cytoplasmique et mitochondriale, enzymes glutathion dépendantes). Actuellement, nous avons réussi à déterminer une séquence codante de 249 pb correspondant au gène de la glutathion-S-transférase chez la coque, malheureusement, aucun couple d‘amorce robuste pour la PCR quantitative n‘a encore pu être obtenu. La voie glutathion dépendante apparaît particulièrement importante chez cette espèce puisqu‘elle pourrait être en outre impliquée dans les mécanismes de séquestration et de détoxication du mercure.

- Enfin, la détermination de l‘état énergétique des bivalves soumis à de multiples stress permettrait de comprendre les différences observées entre simple, double et triple stress en

300

Chapitre VI. Synthèse générale et perspectives de recherche

termes de bioaccumulation métallique, de réponse physiologique mais aussi de mortalité. Des études bioénergétiques peuvent être envisagées par mesure de l‘intégrité et du fonctionnement des complexes de la chaîne respiratoire mitochondriale, de la production/consommation d‘ATP ou encore par micro-respirométrie (Cambier et al., 2009 ; Goulletquer et al., 2004 ; Le Moullac et al., 2007 ; Tschischka et al., 2000) et permettrait de statuer plus précisement quant à l‘état énergétique des organismes dans des conditions particulièrement stressantes.

En replaçant maintenant cette problématique dans un contexte plus général, avec la volonté d‘extrapoler les mécanismes mis en évidence dans nos expériences en laboratoire, plusieurs points essentiels méritent d‘être pris en compte et évalués.

Tout d‘abord, la nature du polluant implique à l‘évidence d‘importantes différences d‘effets sur les couples « hôte-pathogène », et ce même à l‘intérieur d‘une classe spécifique de contaminants, à savoir ici, les polluants métalliques. Les expériences multistress conduites lors de cette thèse (Chapitre IV) concernaient uniquement un polluant métallique, le Cd. Toutefois, les expériences du Chapitre V montrent une forte spécificité de la coque pour le mercure comparé au cadmium, qui peut être la cause d‘effets particulièrement délétères dans le cadre de multiples stress. Une étude in situ relie notamment la disparition des coques le long d‘un système estuarien à la présence de mercure dans le milieu en possible interaction avec d‘autres facteurs biotiques et abiotiques (Nunes et al., 2008). Dès lors, l‘élargissement de telles expériences à d‘autres contaminants métalliques comme le mercure pour la coque, mais aussi à des mélanges de contaminants afin de se rapprocher des conditions de l‘environnement pourrait apporter de nouvelles indications quant aux sensibilités des couples « hôte-pathogène » étudiés.

Il convient également de noter que les expériences multistress conduites lors de cette thèse impliquaient l‘exposition d‘individus adultes. Toutefois, l‘occurrence de pollution métallique en parallèle à des infestations parasitaires et/ou bactériennes lors des périodes de recrutement ou lors des phases de croissance des juvéniles peut gravement affecter la dynamique de population d‘un bivalve. Ainsi, la détermination de l'étape de vie la plus sensible des mollusques à ces sources de stress multiples apparaît nécessaire dans le but d‘améliorer la compréhension des mécanismes (épizooties, mortalités, etc.) observés in situ. Cependant, une

301

Chapitre VI. Synthèse générale et perspectives de recherche barrière méthodologique liée à la taille des organismes peut être envisagée, notamment pour les prélèvements d‘hémolymphe.

De la même manière, les expériences multistress impliquaient l‘exposition de la larve de parasite et ne nous ont pas permis d‘appréhender les effets toxiques de telles sources de stress sur l‘hôte amont. Dans le cas du parasite H. elongata, l‘exposition préalable du premier hôte intermédiaire d‘où sont issues les cercaires (Littorina littorina) à ces stress multiples serait à réaliser de manière à se placer dans des conditions représentatives de ce qu‘il peut subvenir dans l‘environnement, et de statuer ainsi plus largement sur l‘impact de tels stress (Cd, V. tapetis) sur la réussite de la transmission parasitaire de son premier à son second hôte intermédiaire.

Un autre point semble important dans l‘étude des interactions « polluant-hôte-pathogène » et concerne la chronologie de l‘apparition des sources de perturbations. Très peu d‘études en laboratoire ont examiné l‘effet de l‘ordre d‘application des stress sur la sensibilité de l‘espèce hôte à des stress multiples. Une expérience conduite par Chou et al. (1998) sur les palourdes (Meretrix lusoria) met pourtant en évidence une différence notable dans les taux de mortalité entre les individus subissant une infection virale suivi d‘une exposition métallique (52%) et les individus subissant l‘exposition métallique avant l‘infection virale (90%). Ce genre de résultat soulève de nombreuses questions quant à l‘importance de l‘apparition de maladies et pollutions dans le milieu naturel sur la capacité de résistance des populations. Des expériences de cinétique en laboratoire mais aussi des essais de transplantation croisées in situ (entre des sites caractérisés par la présence de certains pathogènes et des sites particulièrement contaminés) pourraient apporter des éléments de compréhension quant à l‘importance de ce mécanisme.

Enfin, l‘approche terrain développée dans le Chapitre III avait pour objectif d‘estimer la santé des populations (paramètres de dynamique de population), de la mettre en rapport avec les niveaux de base des stress subis (métaux, parasites) et d‘évaluer les réponses adaptatives mises en place par les bivalves (échelle moléculaire, protéique et cellulaire). Ce dernier point n‘a pas pu être pleinement satisfait du fait de la variabilité des mécanismes liée au cycle de vie des espèces et aux variables climatiques. La mise en évidence de relations ou d‘interactions particulières n‘a pas pu être réalisée in situ et l‘interprétation des réponses protéiques et cellulaires observées de manière « globale » et moyennée sur 2 ans s‘est retrouvée rapidement périlleuse et limitée. Ainsi, une analyse saison par saison, dans chaque site, pour chaque espèce et à l‘échelle individuelle est indispensable à la discrimination de l‘influence du cycle de vie sur les réponses observées et la

302

Chapitre VI. Synthèse générale et perspectives de recherche détermination des corrélations avec les sources de stress identifiées. De plus, la mise en place d‘indices multiparamétriques se basant sur les variations de chaque paramètre par rapport à une valeur référence (dont la détermination constitue une des étapes les plus délicates) permettrait d‘augmenter la pertinence du jeu de paramètres observés in situ (Auffret et al., 2006 ; Hagger et al., 2009). L‘analyse approfondie d‘une telle base de données englobant les variations spatio- temporelles et les mécanismes liés au cycle de vie des espèces, en complémentarité avec les approches expérimentales indispensables à la compréhension des processus cellulaires, moléculaires et physiologiques sous-jacents, semble constituer une approche des plus efficaces pour mettre en évidence des relations entre certaines maladies ou pathogènes et certains contaminants.

303

304

References bibliographiques

305

306

Références bibliographiques

A

Abd Allah, A.T., Wanas, M.Q., Thompson, S.N., 1996. The effects of lead, cadmium, and mercury on the mortality and infectivity of Schistosoma mansoni cercariae. Journal of Parasitol. 82, 1024- 1026.

Adams, S.M., 2001. Biomarker/bioindicator response profiles of organisms can help differentiate between sources of anthropogenic stressors in aquatic ecosystems. Biomarkers 6, 33-44.

Allam, B., Paillard, C., Auffret, M., 2000. Alterations in hemolymph and extrapallial fluid parameters in the Manila Clam, Ruditapes philippinarum, challenged with the pathogen Vibrio tapetis. Journal of Invertebrate Pathology 76, 63-69.

Allam, B., Ashton-Alcox, K.A., Ford, S.E., 2001. Haemocyte parameters associated with resistance to brown ring disease in Ruditapes spp. clams. Developmental and Comparative Immunology 25, 365-375.

Allam, B., Paillard, C., Ford, S.E., 2002. Pathogenicity of Vibrio tapetis, the etiological agent of brown ring disease. Diseases of Aquatic Organisms 48, 221-231.

Allam, B., Ford, S.E., 2006. Effects of the pathogenic Vibrio tapetis on defence factors of susceptible and non-susceptible bivalve species: I. Haemocyte changes following in vitro challenge. Fish & Shellfish Immunology 20, 374-383.

Allam, B., Paillard, C., Auffret, J.P., Ford, S.E., 2006. Effects of the pathogenic Vibrio tapetis on defence factors of susceptible and non-susceptible bivalve species: II. Cellular and biochemical changes following in vivo challenge. Fish & Shellfish Immunology 20, 384-397.

Alvarado, N.E., Cancio, I., Hylland, K., Marigómez, I., Soto, M., 2007. Immunolocalization of metallothioneins in different tissues of turbot (Scophthalmus maximus) exposed to Cd. Histology and Histopathology 22, 719-728.

Amaro, F., del Pilar de Lucas, M., Martín-González, A., Gutiérrez, J.C., 2008. Two new members of the Tetrahymena multi-stress-inducible metallothionein family: Characterization and expression analysis of T. rostrata Cd/Cu metallothionein genes. Gene. 423, 85-91.

Amiard, J.C., Amiard-Triquet, C., Ballan-Dufrançais, C., Berthet, B., Jeantet, A.Y., Martoja, R. and Truchet, M., 1989. Study of the bioaccumulation at the molecular, cellular and organism levels of lead and copper transferred to the oyster Crassostrea gigas Thunberg directly from water or via food. European Marine Biology Symposium, 20e EMBS.

Amiard, J.C., Amiard-Triquet, C., Barka, S., Pellerin, J., Rainbow, P.S., 2006. Metallothioneins in aquatic invertebrates: Their role in metal detoxification and their use as biomarkers. Aquatic Toxicology 76, 160-202.

Anajjar, E.M., Chiffoleau, J.F., Bergayou, H., Moukrim, A., Burgeot, T., Cheggour, M., 2008. Monitoring of trace metal contamination in the souss estuary (South Morocco) using the clams Cerastoderma edule and Scrobicularia plana. Bulletin of Environmental Contamination and Toxicology 80, 283-288.

Anderson, R.S., Burreson, E.M., Paynter, K.T., 1995. Defense responses of hemocytes withdrawn from Crassostrea virginica infected with Perkinsus marinus. Journal of Invertebrate Pathology 66, 82-89.

307

Références bibliographiques

Anderson, R.S., Patel, K.M., Roesijadi, G., 1999. Oyster metallothionein as an oxyradical scavenger: Implications for hemocyte defense responses. Developmental and Comparative Immunology 23, 443-449.

Andrews, G.K., 2000. Regulation of metallothionein gene expression by oxidative stress and metal ions. Biochemistry and Pharmacology 59, 95-104.

Audry, S., Schäfer, J., Blanc, G., Jouanneau, J.-M., 2004. Fifty-year sedimentary record of heavy metal pollution (Cd, Zn, Cu, Pb) in the Lot River reservoirs (France). Environmental Pollution 132, 413-426.

Auffret, M., 1988. Bivalve hemocyte morphology. In American Fisheries Society Special Publication 18, 169–177.

Auffret, M.R., S. Boutet, I. Tanguy, A. Baron, J. Moraga, D. Duchemin, M., 2006. A multiparametric approach for monitoring immunotoxic responses in mussels from contaminated sites in Western Mediterranea. Ecotoxicology and Environmental Safety 63, 393-405.

Auffret, M., Oubella, R., 1994. Cytometric parameters of bivalve molluscs : effect of environmental factors. In: Modulators of Fish Immune Responses, Stolen J. S., T. C. Fletcher eds. Models for environmental toxicology, biomarkers, immunostimulators SOS Publications, Fair Haven U.S.A. 1:23-32.

B

Bachelet, G., Guillou, J., Labourg, P.-J., 1992. Adult-larval and juvenile interactions in the suspension-feeding bivalve, Cerastoderma edule: field observations and experiments. in: Colombo, G., Ferrari, I., Ceccherelli, V., Rossi, R. (Eds.), Marine Eutrophication and Population Dynamics. Olsen & Olsen, Fredensborg, pp. 175-182.

Bachelet, G., Simon-Bouhet, B., Desclaux, C., Garcia-Meunier, P., Mairesse, G., de Montaudouin, X., Raigné, H., Randriambao, K., Sauriau, P.-G., Viard, F., 2004. Invasion of the eastern Bay of Biscay by the nassariid gastropod Cyclope neritea: origin and effects on resident fauna. Marine Ecology Progress Series 276, 147-159.

Bachère, E., Boulo, V., Godin, P., Goggin, L., Hervio, D., Le Gall, G., Morvan, A., Mialhe, E., 1991. In vitro chemiluminescence studies of marine bivalve defence mechanisms and responses against specific pathogens. Developmental and Comparative Immunology 15.

Bartoli, P., 1974. Recherche sur les Gymnophallidae F. N. Morozov, 1955 (Digenea), parasites d'oiseaux des côtes de Camargue : Systématique, biologie et écologie. Aix-Marseille.

Baudrimont, M., Andres, S., Durrieu, G., Boudou, A., 2003. The key role of metallothioneins in the bivalve Corbicula fluminea during the depuration phase, after in situ exposure to Cd and Zn. Aquatic Toxicology 63, 89-102.

Baudrimont, M., de Montaudouin, X., 2007. Evidence of altered protective effect of metallothioneins after cadmium exposure in the digenean parasite-infected cockle (Cerastoderma edule). Parasitology 134, 237-245.

Baudrimont, M., de Montaudouin, X., Palvadeau, A., 2006. Impact of digenean parasite infection on metallothionein synthesis by the cockle (Cerastoderma edule): A multivariate field monitoring. Marine Pollution Bulletin 52, 494-502.

308

Références bibliographiques

Baudrimont M., Marie V, de Montaudouin X., Gonzalez P., Bourdineaud J.-P. and Boudou A. 2007. Confounding factors acting on metal-induced metallothionein synthesis in bivalve molluscs: reproductive cycle, parasitism and triploid status. Proceedings of the International Conference on Environmental Management, Engineering, Planning and Economics, 1, 217-223.

Baudrimont, M., Métivaud, J., Maury-Brachet, R., Ribeyre, F., Boudou, A., 1997a. Bioaccumulation and metallothionein response in the Asiatic clam (Corbicula fluminea) after experimental exposure to cadmium and inorganic mercury. Environmental Toxicology and Chemistry 16, 2096-2105.

Baudrimont, M., Lemaire-Gony, S., Ribeyre, F., Métivaud, J. & Boudou, A., 1997b. Seasonal variations of metallothioein concentrations in the asiatic clam (Corbicula fluminea). Comparative Biochemistry and Physiology, 118 (C), 361-367.

Baudrimont, M., Schäfer, J., Marie, V., Maury-Brachet, R., Bossy, C., Boudou, A., Blanc, G., 2005. Geochemical survey and metal bioaccumulation of three bivalve species (Crassostrea gigas, Cerastoderma edule and Ruditapes philippinarum) in the Nord Médoc salt marshes (Gironde estuary, France). Science of The Total Environment 337, 265-280.

Bebianno, M.J., Langston, W.J., 1991. Metallothionein induction in Mytilus edulis exposed to cadmium. Marine Biology 108, 91-96.

Bebianno, M.J., Serafim, M.A., Rita, M.F., 1994. Involvement of metallothionein in cadmium accumulation and elimination in the clam Ruditapes decussata. Bulletin of environmental Contamination and Toxicology, 53, 726-732.

Bebianno M.J., Serafim M.A., 1998. Comparison of metallothionein induction in response to cadmium in the gills of the bivalve molluscs Mytilus galloprovincialis and Ruditapes decussates. Science of the Total Environment 214, 123-131.

Beliaeff, B., O'Connor, T.P., Munschy, C., Raffin, B., Claisse, D., 2002. Comparison of polycyclic aromatic hydrocarbon levels in mussels and oysters in France and the United States. Environmental Toxicology and Chemistry 21, 1783-1787.

Bergayou, H., Mouneyrac, C., Pellerin, J., Moukrim, A., 2009. Oxydative stress responses in bivalves (Scrobicularia plana, Cerastoderma edule) from the Oued Souss estuary (Morocco). Ecotoxicology and Environmental Safety 72, 765-769.

Bervoets, L., Voets, J., Chu, S., Covaci, A., Schepens, P., Blust, R., 2004. Comparison of accumulation of micropollutants between indigenous and transplanted zebra mussels (Dreissena polymorpha). Environmental Toxicology and Chemistry 23, 1973-1983.

Bietz, H., Mattig, F.R., Becker, P.H., 1999. Spatial and temporal variation in heavy metal contamination of Nereis and Cerastodema. Senckenbergiana Maritima 29, 33-36.

Blackmore, G., Wang, W.X., 2002. Uptake and efflux of Cd and Zn by the green mussel Perna viridis after metal preexposure. Environmental Science and Technology 36, 989-995.

Blackmore, G., Wang, W.X., 2003. Comparison of metal accumulation in mussels at different local and global scales. Environmental Toxicology and Chemistry 22, 388-395.

Blanar, C.A., Munkittrick, K.R., Houlahan, J., MacLatchy, D.L., Marcogliese, D.J., 2009. Pollution and parasitism in aquatic animals: A meta-analysis of effect size. Aquatic Toxicology 93, 18-28.

309

Références bibliographiques

Blanchet, H., Raymond, N., de Montaudouin, X., Capdepuy, M., Bachelet, G., 2003. Effects of digenean trematodes and heterotrophic bacteria on mortality and burying capability of the common cockle Cerastoderma edule (L.). Journal of Experimental Marine Biology and Ecology 293, 89-105.

Blanchet, H., de Montaudouin, X., Lucas, A., Chardy, P., 2004. Heterogeneity of macrozoobenthic assemblages within a Zostera noltii seagrass bed: diversity, abundance, biomass and structuring factors. 61, 111-123.

Blanchet, H., de Montaudouin, X., Chardy, P., Bachelet, G., 2005. Structuring factors and recent changes in subtidal macrozoobenthic communities of a coastal lagoon, Arcachon Bay (France). Estuarine and Coastal Shelf Science 64, 561-576.

Boisson, F., Hartl, M.G.J., Fowler, S.W., Amiard-Triquet, C., 1998. Influence of chronic exposure to silver and mercury in the field on the bioaccumulation potential of the bivalve Macoma balthica. Marine Environmental Research 45, 325-340.

Bordin, G., McCourt, J., Cordeiro Raposo, F., Rodriguez, A.R., 1997. Metallothionein-like metallothioneins in the Baltic clam Macoma baltica: Seasonal variations and induction upon metal exposure. Marine Biology 129, 453-463.

Borrego, J.J., Castro, D., Luque, A., Paillard, C., Maes, P., Garcia, M.T., Ventosa, A., 1996. Vibrio tapetis sp. nov., the causative agent of the brown ring disease affecting cultured clams. International Journal of Systematic Bacteriology 46, 480-484.

Boudou, A., Delnomdedieu, M., Georgescauld, D., Ribeyre, F., Saouter, E., 1991. Fundamental roles of biological barriers in mercury accumulation and transfer in freshwater ecosystems (analysis at organism, organ, cell and molecular levels. Water, Air, and Soil Pollution 56, 807-821.

Bouilly, K., Gagnaire, B., Bonnard, M., Thomas-Guyon, H., Renault, T., Miramand, P., Lapègue, S., 2006. Effects of cadmium on aneuploidy and hemocyte parameters in the Pacific oyster, Crassostrea gigas. Aquatic Toxicology 78, 149-156.

Bourdais J.L., 1998. Agrobiologie et environnement : une comparaison des systèmes de production agrobiologiques et conventionnels en Aquitaine sur la base d'indicateurs : rapport de synthèse. CEMAGREF - 98-0700.

Bourdineaud, J.P., Baudrimont, M., Gonzalez, P., Moreau, J.L., 2006. Challenging the model for induction of metallothionein gene expression. Biochimie 88, 1787-1792.

Bourget, E., Brock, V., 1990. Short-term shell growth in bivalves: individual, regional, and age-related variations in the rhythm of deposition of Cerastoderma (=Cardium) edule. Marine Biology 106, 103-108.

Bourne, N., 1982. Distribution, reproduction, and growth of the Manila clam, Tapes philippinarum (Adams and Reeves), in British Columbia. Journal of Shellfish Research 2, 47-54.

Boutet, I., Tanguy, A., Auffret, M., Riso, R., Moraga, D., 2002. Immunochemical quantification of metallothioneins in marine mollusks: Characterization of a metal exposure bioindicator. Environmental Toxicology and Chemistry 21, 1009-1014.

Boutier, B., Chiffoleau, J.F., Jouanneau, J.M., Latouche, C., Philipps, I., 1989. La contamination de la Gironde par le cadmium : origine, extension, importance. IFREMER, Plouzané, 105 pp.

310

Références bibliographiques

Bowers, E.A., Bartoli, P., Russell-Pinto, F., James, B.L., 1996. The metacercariae of sibling species of Meiogymnophallus, including M. rebecqui comb. nov. (Digenea: Gymnophallidae), and their effects on closely related Cerastoderma host species (Mollusca: Bivalvia). Parasitology Research 82, 505-510.

Breathnach, R., Chambon, P., 1981. Organization and expression of eucaryotic split genes coding for proteins. Annual Review of Biochemistry 50, 349-383.

Bryan-Walker, K., Leung, T.L.F., Poulin, R., 2007. Local adaptation of immunity against a trematode parasite in marine amphipod populations. Marine Biology 152, 687-695.

Bricelj, V.M., Ford, S.E., Borrero, F.J., Perkins, F.O., Rivara, G., Hillman, R.E., Elston, R.A., Chang, J., 1992. Unexplained mortalities of hatchery-reared, juvenile oysters, Crassostrea virginica (Gmelin). Journal of Shellfish Research 11, 331-347.

Breber, P., 2002. Introduction and acclimatisation of the Pacific carpet clam, Tapes philippinarum, to italian waters, in: Leppäkoski, E., Gollasch, S., Olenin, S. (Eds.), Invasive aquatic species of Europe, distribution, impacts and management. Kluwer Academic Publishers, Dordrecht, pp. 120-126.

Brousseau, P., Pellerin, J., Morin, Y., Cyr, D., Blakley, B., Boermans, H.F., Fournier, M., 2000. Flow cytometry as a tool to monitor the disturbance of phagocytosis in the clam Mya arenaria hemocytes following in vitro exposure to heavy metals. Toxicology 142, 145-156.

Brown, P.L., Markich, S.J., 2000. Evaluation of the free ion activity model of metal-organism interaction: Extension of the conceptual model. Aquatic Toxicology 51, 177-194.

Buggé, D.A., Hégaret, H., Wikfors, G.H., Allam, B., 2007. Oxidative burst in hard clam (Mercenaria mercenaria) haemocytes. Fish & Shellfish Immunology 23, 188-196.

Burlando, B., Bonomo, M., Caprí, F., Mancinelli, G., Pons, G., Viarengo, A., 2004. Different effects of Hg2+ and Cu2+ on mussel (Mytilus galloprovincialis) plasma membrane Ca2+-ATPase: Hg2+ induction of protein expression. Comparative Biochemistry and Physiology 139 (C), 201-207.

Byrne, P.A., O'Halloran, J., 2001. The role of bivalve molluscs as tools in estuarine sediment toxicity testing: a review. Hydrobiologia 465, 209-217.

C

Caira, J.N., Littlewood, D.T.J., 2001. Worms, Plathyhelminthes, in: Levin, S.A. (Ed.), Encyclopedia of Biodiversity. Academic Press, London, pp. 863-899.

Cajaraville, M.P., Bebianno, M.J., Blasco, J., Porte, C., Sarasquete, C., Viarengo, A., 2000. The use of biomarkers to assess the impact of pollution in coastal environments of the Iberian Peninsula: A practical approach. Science of the Total Environment 247, 295-311.

Cajaraville, M.P., Olabarrieta, I., Marigómez, I., 1996. In vitro activities in mussel hemocytes as biomarkers of environmental quality: A case study in the Abra Estuary (Biscay Bay). Ecotoxicology and Environmental Safety 35, 253-260.

Calvo-Ugarteburu, G., McQuaid, C.D., 1998a. Parasitism and introduced species: epidemiology of trematodes in the intertidal mussels Perna perna and Mytilus galloprovincialis. Journal of Experimental Marine Biology and Ecology 220, 47-65.

311

Références bibliographiques

Calvo-Ugarteburu, G., McQuaid, C.D., 1998b. Parasitism and invasive species: effects of digenetic trematodes on mussels. Marine Ecology Progress Series 169, 149-163.

Cambier, S., Bénard, G., Mesmer-Dudons, N., Gonzalez, P., Rossignol, R., Brèthes, D., Bourdineaud, J.P., 2009. At environmental doses, dietary methylmercury inhibits mitochondrial energy metabolism in skeletal muscles of the zebra fish (Danio rerio). International Journal of Biochemistry and Cell Biology 41, 791-799.

Campbell H.A., Couillard Y., 2004. Prise en charge et détoxication des métaux chez les organismes aquatiques. In: Écotoxicologie moléculaire : Principes fondamentaux et perspectives de développement. Eds.: E Pelletier, PGC Campbell and F Denizeau. Presses de l'Université du Québec, Sainte-Foy (Quebec), Canada, pp. 9-61.

Campbell, P.G.C., Giguère, A., Bonneris, E., Hare, L., 2005. Cadmium-handling strategies in two chronically exposed indigenous freshwater organisms - The yellow perch (Perca flavescens) and the floater mollusc (Pyganodon grandis). Aquatic Toxicology 72, 83-97.

Canesi, L., Gallo, G., Gavioli, M., Pruzzo, C., 2002. Bacteria-hemocyte interactions and phagocytosis in marine bivalves. Microscopy Research and Technique 57, 469-476.

Carballal, M.J., Iglesias, D., Santamarina, J., Ferro-Soto, B., Villalba, A., 2001. Parasites and pathologic conditions of the cockle Cerastoderma edule populations of the coast of Galicia (NW Spain). Journal of Invertebrate Pathology 78, 87-97.

Carnegie, R.B., 2005. Effects in mollusc culture, in: Rohde, K. (Ed.), Marine parasitology. CSIRO & CABI, Collingwood, Australia, pp. 391-398.

Carneiro, M. and Ferrand, N., 2007. Extensive intragenic recombination and patterns of linkage disequilibrium at the CSN3 locus in European rabbits. Genetics Selection Evolution 39, 341- 352.

Carré, J.E., Singer, M., 2008. Cellular energetic metabolism in sepsis: The need for a systems approach. Biochimica and Biophysica Acta, Bioenergetics. 1777, 763-771.

Ceratto, N., Dondero, F., Van de Loo, J.W., Burlando, B., Viarengo, A., 2002. Cloning and sequencing of a novel metallothionein gene in Mytilus galloprovincialis Lam. Comparative Biochemistry and Physiology 131(C), 217-222.

Chan, M.K., Othman, R., Zubir, D., Salmijah, S., 2002. Induction of a putative metallothionein gene in the blood cockle, Anadara granosa, exposed to cadmium. Comparative Biochemistry and Physiology 131, 123-132.

Charalambidis, N.D., Foukas, L.C., Marmaras, V.J., 1996. Covalent association of lipopolysaccharide at the hemocyte surface of insects is an initial step for its internalization - Protein-tyrosine phosphorylation requirement. European Journal of Biochemistry 236, 200-206.

Cheggour, M., Chafik, A., Langston, W.J., Burt, G.R., Benbrahim, S., Texier, H., 2001. Metals in sediments and the edible cockle Cerastoderma edule from two Moroccan Atlantic lagoons: Moulay Bou Selham and Sidi Moussa. Environmental Pollution 115, 149-160.

Cheng, T.C., 1981. Bivalves, in: Ratcliffe, N.A.R., A.F. (eds) (Ed.), Invertebrate Blood Cells I. Academic Press, London, pp. 233-299.

Cheng, T.C., 1983. The role of lysosomes in molluscan inflammation. American 23, 129-144.

312

Références bibliographiques

Cheng, T.C., 1988. In vivo effects of heavy metals on cellular defense mechanisms of Crassostrea virginica: Phagocytic and endocytotic indices. Journal of Invertebrate Pathology 51, 215-220.

Cheng, T.C., 1996. Hemocytes: forms and functions. In: V. S. Kennedy, R. I. E. Newell & A. F. Eble, editors The eastern oyster Crassostrea virginica by Maryland, College Park: Maryland Sea Grant 299-333.

Cheng, T., Allen, H.E., 2001. Prediction of uptake of copper from solution by lettuce (Lactuca sativa romance). Environmental Toxicology and Chemistry 20, 2544-2551.

Cheng-Hua, L., Jian-Min, Z., Lin-Sheng, S., 2009. A review of advances in research on marine molluscan antimicrobial peptides and their potential application in aquaculture. Molluscan Research 29, 17-26.

Cheng-Yi, S., Bo, G., Shang-Hui, T., Xiao-Yang, W., Fei, X., Guo-Hong, C., Zhi-Yue, W., Rong-Bin, J., Jiu-De, M., 2007. Polymorphisms in intron 1 of the porcine POU1F1 gene. Journal of Applied Genetics 49, 371-374.

Cheung, K.C., Poon, B.H.T., Lan, C.Y., Wong, M.H., 2003. Assessment of metal and nutrient concentrations in river water and sediment collected from the cities in the Pearl River Delta, South China. Chemosphere 52, 1431-1440.

Cho, Y.S., Lee, S.Y., Kim, K.Y., Bang, I.C., Kim, D.S., Nam, Y.K., 2008. Gene structure and expression of metallothionein during metal exposures in Hemibarbus mylodon. Ecotoxicology and Environmental Safety 71, 125-137.

Choi, K.S., Park, K.I., Lee, K.W., Matsuoka, K., 2002. Infection intensity, prevalence, and histopathology of Perkinsus sp. in the Manila clam, Ruditapes philippinarum, in Isahaya Bay, Japan. Journal of Shellfish Research 21, 119-125.

Choi, K.S., Wilson, E.A., Lewis, D.H., Powell, E.N., Ray, S.M., 1989. The energic cost of Perkinsus marinus parasitism in oysters: quantification of the thioglycollate method. Journal of Shellfish Research 8, 125-131.

Choquet, G., Soudant, P., Lambert, C., Nicolas, J.L., Paillard, C., 2003. Reduction of adhesion properties of Ruditapes philippinarum hemocytes exposed to Vibrio tapetis. Diseases of Aquatic Organisms 57, 109-116.

Chu, F.L.E., 1999. Effects of field-contaminated sediments and related water soluble components on haemocyte function and Perkinsus marinus susceptibility and expression in oysters. Biomarkers 4, 537-548.

Chu, F.-L.E., 2000. Defense mechanisms of marine bivalves. Recent Advances in Marine Biotechnology 5, 1-42.

Cima, F., Matozzo, V., Marin, M.G., Ballarin, L., 2000. Haemocytes of the clam Tapes philippinarum (Adams & Reeve, 1850): Morphofunctional characterisation. Fish and Shellfish Immunology 10, 677-693.

Clarkson, T.W., 1994. The toxicology of mercury and its compounds. Mercury Pollution: Integration and Synthesis, 631-641.

Cochennec-Laureau, N., Auffret, M., Renault, T., Langlade, A., 2003. Changes in circulating and tissue-infiltrating hemocyte parameters of European flat oysters, Ostrea edulis, naturally infected with Bonamia ostreae. Journal of Invertebrate Pathology 83, 23-30.

313

Références bibliographiques

Cole, H.A., 1956. A preliminary study of growth-rate in cockles (Cardium edule L.) in relation to commercial exploitation. Journal du Conseil Permanent International pour l'Exploration de la Mer 22, 77-90.

Coles, J.A., Farley, S.R., Pipe, R.K., 1994. Effects of fluoranthene on the immunocompetence of the common marine mussel, Mytilus edulis. Aquatic Toxicology 30, 367-379.

Coles, J.A., Farley, S.R., Pipe, R.K., 1995. Alteration of the immune response of the common marine mussel Mytilus edulis resulting from exposure to cadmium. Diseases of Aquatic Organisms 22, 59-65. Coles, J.A., Pipe, R.K., 1994. Phenoloxidase activity in the haemolymph and haemocytes of the marine mussel Mytilus edulis. Fish and Shellfish Immunology 4, 337-352.

Colwell, R.R., Liston, J.L., 1962. The natural bacterial flora of certain marine invertebrates. Journal of Insect Pathology 4, 23-33.

Comeron, J.M., Kreitman, M., 2000. The correlation between intron length and recombination in Drosophila: dynamic equilibrium between mutational and selective forces. Genetics 156, 1175- 1190.

Coombs, T.L., George, S.G., 1977. Mechanisms of immobilization and detoxication of metals in marine organisms, 12th European Symposium on Marine Biology, Stirling, Scotland, pp. 179- 187.

Cossa, D., 1985. Le cadmium et le mercure en milieu côtier : biogéochimie et utilisation du genre Mytilus comme indicateur quantitatif. Université de Bretagne Occidentale, Brest, p. 387.

Cossa, D., Noël, J., 1987. Concentrations of mercury in near shore surface waters of the Bay of Biscay and in the Gironde Estuary. Marine Chemistry 20, 389-396.

Costanza, R., D'Arge, R., De Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O'Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., Van Den Belt, M., 1997. The value of the world's ecosystem services and natural capital. Nature 387, 253-260.

Couillard Y. (1997). Technical evaluation of metallothionein as a biomarker for the mining industry. In: Aquatic Effects Technology Evaluation Program Report 2.2.1. Report Prepared for Natural Resources Canada, Ottawa, pp. 191.

Couillard, Y., Campbell, P.G.C., Tessier, A., 1993. Response of metallothionein concentrations in a freshwater bivalve (Anodonta grandis) along an environmental cadmium gradient. Molecular Marine Biology and Biotechnology 2, 299-313.

Coyle, P., Philcox, J.C., Carey, L.C., Rofe, A.M., 2002. Metallothionein: The multipurpose protein. Cellular and Molecular Life Sciences 59, 627-647.

Cram, L.S., 2002. Flow cytometry, an overview. Methods in Cell Science 24, 1-9.

Cravo, A., Lopes, B., Serafim, A., Company, R., Barreira, L., Gomes, T., Bebianno, M.J., 2009. A multibiomarker approach in Mytilus galloprovincialis to assess environmental quality. Journal of Environmental Monitoring 11, 1673-1686.

Creek, G.A., 1960. The development of the Lamellibranch Cardium edule L. Proceedings of the Zoological Society of London 135, 243-260.

314

Références bibliographiques

Cross, M.A., Irwin, S.W.B., Fitzpatrick, S., Manga, N., 2003. Trematode parasite influence on copper, iron and zinc content of polluted Littorina littorea: Infection, host sex and time effects. Journal of the Marine Biological Association of the United Kingdom 83, 1269-1272.

Curtis, L.A., Hurd, L.E., 1983. Age, sex, and parasites: spatial heterogeneity in a sandflat population of Ilyanassa obsoleta. Ecology 64, 818-828.

Curtis, L.A., 1995. Growth, trematode parasitism, and longevity of a long-lived marine gastropod (Ilyanassa obsoleta). Journal of the Marine Biological Association of the United Kingdom 75, 913-925.

Curtis, L.A., Kinley, J.L., Tanner, N.L., 2000. Longevity of oversized individuals: growth, parasitism, and history in an estuarine snail population. Journal of the Marine Biological Association of the United Kingdom 80, 811-820.

Cusack, R., Cone, D.K., 1986 A review of parasites as vectors of viral and bacterial diseases of fish. Journal of Fish Disease 9, 169–171.

D

Da Silva, P.M., Hégaret, H., Lambert, C., Wikfors, G.H., Le Goïc, N., Shumway, S.E., Soudant, P., 2008. Immunological responses of the Manila clam (Ruditapes philippinarum) with varying parasite (Perkinsus olseni) burden, during a long-term exposure to the harmful alga, Karenia selliformis, and possible interactions. Toxicon 51, 563-573.

Dang, C., 2009. Dynamique des populations de palourdes japonaises (Ruditapes philippinarum) dans le bassin d'Arcachon - Conséquences sur la gestion des stocks exploités. Ph.D. thesis, University Bordeaux 1.

Dang, C., de Montaudouin, X., Bald, J., Jude, F., Raymond, N., Lanceleur, L., Paul-Pont, I., Caill- Milly, N., 2009a. Testing the enemy release hypothesis: Trematode parasites in the non- indigenous Manila clam Ruditapes philippinarum. Hydrobiologia 630, 139-148.

Dang, C., de Montaudouin, X., Gam, M., Paroissin, C., Caill-Milly, N., 2009b. The Manila clam population in Arcachon Bay (SW France): can it be kept sustainable? Journal of Sea Research In press.

Dang, C., de Montaudouin, X., 2009. Brown Muscle Disease and Manila clam Ruditapes philippinarum dynamics in Arcachon Bay, France. Journal of Sea Research 28, In press.

Dang, C., de Montaudouin, X., Gonzalez, P., Mesmer-Dudons, N., Caill-Milly, N., 2008. Brown Muscle Disease (BMD), an emergent pathology affecting Manila clam Ruditapes philippinarum in Arcachon bay (SW France). Diseases of Aquatic Organisms 80, 219-228.

Dang, C., Sauriau, P.-G., Savoye, N., Caill-Milly, N., Martinez, P., Millaret, C., Haure, J., de Montaudouin, X., 2009c. Determination of diet in Manila clams by spatial analysis of stable isotopes. Marine Ecology Progress Series 387, 167-177.

Daniels, P.J., Andrews, G.K., 2003. Dynamics of the metal-dependent transcription factor complex in vivo at the mouse metallothionein-I promoter. Nucleic Acids Research 31, 6710-6721. de Montaudouin, X., 1997. Potential of bivalves' secondary settlement differs with species: a comparison between cockle (Cerastoderma edule) and clam (Ruditapes philippinarum) juvenile resuspension. Marine Biology 128, 639-648.

315

Références bibliographiques

de Montaudouin, X., Bachelet, G., 1996. Experimental evidence of complex interactions between biotic and abiotic factors in the dynamics of an intertidal population of the bivalve Cerastoderma edule. Oceanologica Acta 19, 449-463. de Montaudouin, X., Blanchet, H., Kisielewski, I., Desclaux, C., Bachelet, G., 2003. Digenean trematodes moderately alter Hydrobia ulvae population size structure. Journal of the Marine Biological Association of the United Kingdom, 297-305. de Montaudouin, X., Jensen, K.T., Desclaux, C., Wegeberg, A.M., Sajus, M.C., 2005. Effect of intermediate host size (Cerastoderma edule) on infectivity of cercariae of Himasthla quissetensis (Echinostomatidae, ). Journal of the Marine Biological Association of the United Kingdom 85, 809-812. de Montaudouin, X., Kisielewski, I., Bachelet, G., Desclaux, C., 2000. A census of macroparasites in an intertidal bivalve community, Arcachon Bay, France. Oceanologica Acta 23, 453-468. de Montaudouin, X., Wegeberg, A.M., Jensen, K.T., Sauriau, P.G., 1998. Infection characteristics of Himasthla elongata cercariae in cockles as a function of water current. Diseases of Aquatic Organisms. 34, 63-70. de Montaudouin, X., Thieltges, D.W., Gam, M., Krakau, M., Pina, S., Bazaïri, H., Dabouineau, L., Russell-Pinto, F., Jensen, K.T., 2009. Review - Digenean trematode species in the cockle Cerastoderma edule: identification key and distribution along the North-East Atlantic shoreline. Journal of the Marine Biological Association of the United Kingdom 80, 543-556. de Oliveira Ribeiro, C.A., Nathalie, M.D., Gonzalez, P., Yannick, D., Jean-Paul, B., Boudou, A., Massabuau, J.C., 2008. Effects of dietary methylmercury on zebrafish skeletal muscle fibres. Environmental Toxicology and Pharmacology 25, 304-309.

Delaporte, M., Soudant, P., Moal, J., Lambert, C., Quéré, C., Miner, P., Choquet, G., Paillard, C., Samain, J.-F., 2003. Effect of a mono-specific algal diet on immune functions in two bivalve species - Crassostrea gigas and Ruditapes philippinarum. The Journal of Experimental Biology 206, 3053-3064.

Defer, D., Bourgougnon, N., Fleury, Y., 2009. Screening for antibacterial and antiviral activities in three bivalve and two gastropod marine molluscs. Aquaculture 293, 1-7.

Desclaux, C., 2003. Interactions hôtes-parasites : diversité, mécanismes d'infestation et impact des trématodes digènes sur les coques Cerastoderma edule (mollusque bivalve) en milieu lagunaire macrotidal. Ph.D. thesis, University Bordeaux 1.

Desclaux, C., de Montaudouin, X., Bachelet, G., 2002. Cockle emergence at the sediment surface: 'favourization' mechanism by digenean parasites? Diseases of Aquatic Organisms, 52, 137-149.

Desclaux, C., de Montaudouin, X., Bachelet, G., 2004. Cockle (Cerastoderma edule) population mortality: the role of the digenean parasite Himasthla quissetensis. Marine Ecology Progress Series 279, 141-150.

Desclaux-Marchand, C., Paul-Pont, I., Gonzalez, P., Baudrimont, M., de Montaudouin, X., 2007. Metallothionein gene identification and expression in the cockle (Cerastoderma edule) under parasitism (trematodes) and cadmium contaminations. Aquatic Living Resource 20, 43-49.

316

Références bibliographiques

Deltreil, J.-P., His, E., 1970. Sur la présence d‘une cercaire de trématode chez Cardium edule L. dans le Bassin d‘Arcachon. Revue des Travaux de l‘Institut scientifique des Pêches Maritimes 34 (2), 225-232.

Di Toro, D.M., Allen, H.E., Bergman, H.L., Meyer, J.S., Paquin, P.R., Santore, R.C., 2001. Biotic ligand model of the acute toxicity of metals. 1. Technical basis. Environmental Toxicology and Chemistry 20, 2383-2396.

Dimitriadis, V.K., Domouhtsidou, G.P., Raftopoulou, E., 2003. Localization of Hg and Pb in the palps, the digestive gland and the gills in Mytilus galloprovincialis (L.) using autometallography and X-ray microanalysis. Environmental Pollution 125, 345-353.

Domouhtsidou, G.P. and Dimitriadis, V.K., 2000. Ultrastructural localization of heavy metals (Hg, Ag, Pb, and Cu) in gills and digestive gland of mussels, Mytilus galloprovincialis (L.). Archives of Environmental Contamination and Toxicology 38, 472-478.

Donaghy, L., Lambert, C., Choi, K.-S., Soudant, P., 2009. Hemocytes of the carpet shell clam (Ruditapes decussatus) and the Manila clam (Ruditapes philippinarum): Current knowledge and future prospects. Aquaculture 297 (1-4), 10-24.

Drummond, L., O'Reilly, S.L., Mulcahy, M., Culloty, S., 2006. Comparison of techniques for diagnosis of brown ring disease and detection of Vibrio tapetis in the Manila clam, Venerupis (Ruditapes) philippinarum. Journal of Shellfish Research 25, 1043-1049.

Ducrotoy, J.P., Rybarczyk, H., Souprayen, J., Bachelet, G., Beukema, J.J., Desprez, M., Dörjes, J., Essink, K., Guillou, J., Michaelis, H., Sylvand, B., Wilson, J.G., Elkaïm, B., Ibanez, F., 1991. A comparison of the population dynamics of the cockle (Cerastoderma edule, L.) in North- Western Europe, in: Elliott, M., Ducrotoy, J.-P. (Eds.), Estuaries and Coasts: spatial and temporal intercomparisons. Olsen & Olsen, pp. 173-184.

Dutton, M.D., Stephenson, M., Klaverkamp, J.F., 1993. A mercury saturation assay for measuring metallothioneins in fish. Environmental Toxicology and Chemistry 12, 1193-1202.

E

Elandalloussi, L.M., Carrasco, N., Roque, A., Fernández-Tejedor, M., Furones, D., 2008. Occurence of Perkinsus sp. in two clam species (Ruditapes philippinarum and R. decussatus from the Ebro delta, Spain. Bulletin of the European Association of Fish Pathologists 28, 1-9.

Elston, R.A., Dungan, C.F., Meyers, T.R., Reece, K.S., 2004. Perkinsus sp. infection risk for Manila clams, Venerupis philippinarum (A. Adams and Reeve, 1850) on the Pacific coast of North and Central America. Journal of Shellfish Research 23, 101-105.

Eränen, J.K., 2006. Local adaptation of mountain birch to heavy metals in subarctic industrial barrens. For. Snow Land. Res. 80, 161-167.

Eufemia, N.A., Epel, D., 2000. Induction of the multixenobiotic defense mechanism (MXR), P- glycoprotein, in the mussel Mytilus californianus as a general cellular response to environmental stresses. Aquatic Toxicology 49, 89-100.

Eufemia, N., Girshick, S., Epel, D., 2000. Algal products as naturally occurring modulators for the multidrug resistance (MDR) transporter. Marine Environmental Research 50, 332-333.

317

Références bibliographiques

Evans, D.W., Irwin, S.W.B., Fitzpatrick, S.M., 2001. The effect of digenean (Platyhelminthes) infections on heavy metal concentrations in Littorina littorea. Journal of the Marine Biological Association of the United Kingdom 81, 349-350.

Evans, J.J., Klesius, P.H., Pasnik, D.J., Shoemaker, C.A., 2007. Influence of natural Trichodina sp. parasitism on experimental Streptococcus iniae or Streptococcus agalactiae infection and survival of young channel catfish Ictalurus punctatus (Rafinesque): Short communication. Aquaculture Research 38, 664-667.

F

Fan, W., Wang, W.-X., 2001. Sediment geochemical controls on Cd, Cr, and Zn assimilation by the clam Ruditapes philippinarum. Environmental Toxicology and Chemistry 20, 2309-2317.

Feng, S.Y., 1988. Cellular defense mechanisms of oysters and mussels. American Fisheries Society Special Publication 18, 153-168.

Feng, X., 2005. Mercury Pollution in China - An overview. In: Pirrone, N.,Mahaffey, K.R. (Eds.), Dynamics of mercury pollution on regional and global scales: Atmospheric Processes and Human Exposures Around The World, pp. 657-678.

Fernandes, J.M.O., Kemp, G.D., Molle, M.G., Smith, V.J., 2002. Anti-microbial properties of histone H2A from skin secretions of rainbow trout, Oncorhynchus mykiss. Biochemical Journal 368, 611-620.

Ferrara, R., Maserti, B.E., Andersson, M., Edner, H., Ragnarson, P., Svanberg, S., Hernandez, A., 1998. Atmospheric mercury concentrations and fluxes in the Almaden district (Spain). Atmospheric Environment 32, 3897-3904.

Fischer, H., 1986. Influence of temperature, salinity and oxygen on the cadmium balance of mussels Mytilus edulis. Marine Ecology Progress Series 32, 265-278.

Fisher, W. S., 1986. Structure and functions of oyster hemocytes. Brehélin, M. (ed) Immun. Invert. 3, 25-35.

Fisher, W.S., 2004. Antimicrobial activity of copper and zinc accumulated in eastern oyster amebocytes. Journal of Shellfish Research 23, 321-351.

Fisher, W.S., Oliver, L.M., Walker, W.W., Manning, C.S., Lytle, T.F., 1999. Decreased resistance of eastern oysters (Crassostrea virginica) to a protozoan pathogen (Perkinsus marinus) after sublethal exposure to tributyltin oxide. Marine and Environmental Research 47, 185-201.

Fisher, W.S., Oliver, L.M., Winstead, J.T., Long, E.R., 2000. A survey of oysters Crassostrea virginica from Tampa Bay, Florida: Associations of internal defense measurements with contaminant burdens. Aquatic Toxicology 51, 115-138.

Flye-Sainte-Marie, J., 2007. Ecophysiology of Brown Ring Disease in the Manila clam Ruditapes philippinarum, experimental and modelling approaches. Université de Bretagne Occidentale, p. 187.

Flye-Sainte-Marie, J., 2009. A quantitative estimation of the energetic cost of brown ring disease in the Manila clam using Dynamic Energy Budget theory. Journal of Sea Research 62 (2-3), 114- 123.

318

Références bibliographiques

Flye-Sainte-Marie, J., Soudant, P., Lambert, C., Le Goïc, N., Goncalvez, M., Travers, M.-A., Paillard, C., Jean, F., 2009. Variability of the hemocyte parameters of Ruditapes philippinarum in the field during an annual cycle. Journal of Experimental Marine Biology and Ecology 377, 1-11.

Ford, S.E., Bricelj, V.M., Lambert, C., Paillard, C., 2008. Deleterious effects of a nonPST bioactive compound(s) from Alexandrium tamarense on bivalve hemocytes. Marine Biology 154, 241- 253.

Fournier, M., Pellerin, J., Clermont, Y., Morin, Y., Brousseau, P., 2001. Effects of in vivo exposure of Mya arenaria to organic and inorganic mercury on phagocytic activity of hemocytes. Toxicology 161, 201-211.

Fournier, M., Pellerin, J., Lebeuf, M., Brousseau, P., Morin, Y., Cyr, D., 2002. Effects of exposure of Mya arenaria and Mactromeris polynyma to contaminated marine sediments on phagocytic activity of hemocytes. Aquatic Toxicology 59, 83-92.

Franchini, A., Conte, A., Ottaviani, E., 1995. Nitric oxide: An ancestral immunocyte effector molecule. Advances in Neuroimmunology 5, 463-478.

Fraysse, B., Geffard, O., Berthet, B., Quéau, H., Biagianti-Risbourg, S., Geffard, A., 2006. Importance of metallothioneins in the cadmium detoxification process in Daphnia magna. Comparative Biochemistry and Physiology 144(C), 286-293.

G

Gagnaire, B., 2005. Étude des effets de polluants sur les paramètres hémocytaires de l‘huître creuse, Crassostrea gigas – Interactions entre environnement, mécanismes de défense et maladies infectieuses. Université de la Rochelle, p. 412.

Gagné, F., Blaise, C., Pellerin, J., André, C., 2007. Neuroendocrine disruption in Mya arenaria clams during gametogenesis at sites under pollution stress. Marine Environmental Research 64, 87- 107.

Galloway, T.S., Depledge, M.H., 2001. Immunotoxicity in invertebrates: Measurement and ecotoxicological relevance. Ecotoxicology 10, 5-23.

Galloway, T.S., Brown, R.J., Browne, M.A., Dissanayake, A., Lowe, D., Jones, M.B., Depledge, M.H., 2004. A Multibiomarker Approach to Environmental Assessment. Environmental Science and Technology 38, 1723-1731.

Gam, M., 2008. Dynamique des systèmes parasites-hôte, entre trématodes digènes et coque Cerastoderma edule: comparaison de la lagune Merja Zerga avec le Bassin d'Arcachon. Université Hassan II Aïn Chock & Bordeaux 1.

Gam, M., de Montaudouin, X., Bazaïri, H., 2009. Do trematode parasites affect cockle (Cerastoderma edule) secondary production and elimination? Journal of the Marine Biological Association of the United Kingdom 89, 1395-1402.

Gam, M., de Montaudouin, X., Bazaïri, H., 2010. Population dynamics and secondary production of the cockle Cerastoderma edule: comparison between Merja Zerga (Moroccan Atlantic Coast) and Arcachon (French Atlantic coast). J. Sea Res. Sous presse.

319

Références bibliographiques

Gauthier, J.D., Vasta, G.R., 2002. Effects of plasma from bivalve mollusk species on the in vitro proliferation of the protistan parasite Perkinsus marinus. Journal of Experimental Zoology 292, 221-230.

Geffard, A., Amiard-Triquet, C., Amiard, J.C., Mouneyrac, C., 2001. Temporal variations of metallothionein and metal concentrations in the digestive gland of oysters (Crassostrea gigas) from a clean and a metal-rich site. Biomarkers 6, 91-107.

Geffard, A., Amiard-Triquet, C., Amiard, J.C., 2005. Do seasonal changes affect metallothionein induction by metals in mussels, Mytilus edulis? Ecotoxicology and Environmental Safety 61, 209-220.

Géret, F., Cosson, R.P., 2002. Induction of specific isoforms of metallothionein in mussel tissues after exposure to cadmium or mercury. Archives of Environmental Contamination and Toxicology 42, 36-42.

Géret, F., Serafim, A., Barreira, L., Bebianno, M.J., 2002. Effect of cadmium on antioxidant enzyme activities and lipid peroxidation in the gills of the clam Ruditapes decussatus. Biomarkers 7, 242-256.

Géret, F., Serafim, A., Bebianno, M.J., 2003. Antioxidant Enzyme Activities, Metallothioneins and Lipid Peroxidation as Biomarkers in Ruditapes decussatus? Ecotoxicology 12, 417-426.

Gestal, C., Roch, P., Renault, T., Pallavicini, A., Paillard, C., Novoa, B., Oubella, R., Venier, P., Figueras, A., 2008. Study of Diseases and the Immune System of Bivalves Using Molecular Biology and Genomics. Reviews in Fisheries Science 16, 1–24.

Glé, C., 2007. Structure et dynamique des communautés microbiennes autotrophes et production primaire planctonique dans une lagune côtière macrotidale, le Bassin d'Arcachon. Facteurs de contrôle de type bottom-up. University Bordeaux 1, p. 309.

Gonzalez, P., Baudrimont, M., Boudou, A. and Bourdineaud J.P., 2006. Comparative effects of direct cadmium contamination on gene expression in gills, liver, skeletal muscles and brain of the zebrafish (Danio rerio). BioMetals, 19, 225–235.

Goulletquer, P., 1989. Mortalité hivernale chez la palourde japonaise Ruditapes philippinarum sur le littoral atlantique: aspects biochimique et écophysiologique. Haliotis 19, 215-226.

Goulletquer, P., Héral, M., Bechemin, C., Richard, P., 1989. Abnormal calcification in the shell of the Manila clam, Ruditapes philippinarum: characterization and comparison of amino acid content in different shell parts analysed by HPLC. Aquaculture 81 (2), 169-183.

Goulletquer, P., Wolowicz, M., Latala, A., Brown, C., Cragg, S., 2004. Application of a micro- respirometric volumetric method to respiratory measurements of larvae of the Pacific oyster Crassostrea gigas. Aquatic Living Resources 17, 195-200.

Griscom, S.B., Fisher, N.S., 2004. Bioavailability of sediment-bound metals to marine bivalve molluscs: An overview. Estuaries 27, 826-838.

Guevara, J.M., Niell, F.X., 1989. Growth rates in a continuously immersed population of Cerastoderma edule L. Scientia Marina 53, 483-489.

Guillou, J., Bachelet, G., Desprez, M., Ducrotoy, J.P., Madani, I., Rybarczyk, H., Sauriau, P.-G., Sylvand, B., Elkaim, B., Glémarec, M., 1990. Les modalités de la reproduction de la coque

320

Références bibliographiques

(Cerastoderma edule) sur le littoral français de la Manche et de l'Atlantique. Aquatic Living Resource 3, 29-41.

Guillou, J., Bachelet, G., Glermarec, M., 1992. Influence des fluctuations de temperature sur la reproduction et le recrutement de la coque, Cerastoderma edule. Annales de l‘Institut Océanographiques de Paris 68 (1-2), 65-74.

H Haas, W., 2003. Parasitic worms: strategies of host finding, recognition and invasion. Zoology 106, 349-364.

Hagger, J.A., Galloway, T.S., Langston, W.J., Jones, M.B., 2009. Application of biomarkers to assess the condition of European Marine Sites. Environmental Pollution 157, 2003-2010.

Hamaguchi, M., Suzuki, N., Usuki, H., Ishioka, H., 1998. Perkinsus protozoan infection in short- necked clam Tapes (=Ruditapes) philippinarum in Japan. Fish Pathology 33, 473-480.

Hamer, D.H., 1986. Metallothionein. Annual Review of Biochemistry 55, 913-951.

Hayward, P.J., Ryland, J.S., 1990. The Marine Fauna of the British Isles and Nort-West Europe. Clarendon Press, Oxford.

Hégaret, H., da Silva, P.M., Sunila, I., Shumway, S.E., Dixon, M.S., Alix, J., Wikfors, G.H., Soudant, P., 2009. Perkinsosis in the Manila clam Ruditapes philippinarum affects responses to the harmful-alga, Prorocentrum minimum. Journal of Experimental Marine Biology and Ecology 371, 112-120.

Hégaret, H., da Silva, P.M., Wikfors, G.H., Lambert, C., De Bettignies, T., Shumway, S.E., Soudant, P., 2007. Hemocyte responses of Manila clams, Ruditapes philippinarum, with varying parasite, Perkinsus olseni, severity to toxic-algal exposures. Aquatic Toxicology 84, 469-479.

Hégaret, H., Wikfors, G.H., 2005. Effects of natural and field-simulated blooms of the dinoflagellate Prorocentrum minimum upon hemocytes of eastern oysters, Crassostrea virginica, from two different populations. Harmful Algae 4, 201-209.

Hégaret, H., Wikfors, G.H., Soudant, P., 2003. Flow cytometric analysis of haemocytes from eastern oysters, Crassostrea virginica, subjected to a sudden temperature elevation - II Haemocyte functions: agregation, viability, phagocytosis, and respiratory burst. Journal of Experimental Marine Biology and Ecology 293, 249-265.

Helluy, S., 1982. Relations hôtes-parasite du trématode Microphallus papillorobustus (Rankin, 1940). I. Pénétration des cercaires et rapport des métacercaires avec le tissu nerveux des Gammarus, hôtes intermédiaires. Annales de Parasitologie Humaine Comparée 57, 263-270.

Héral, M., 1993. Why carrying capacity models are useful tools for management of bivalve molluscs culture, in: Dame, R.F. (Ed.), Bivalve filter feeders. Springer-Verlag, Berlin, pp. 455-477.

Hine, P.M., 1999. The inter-relationships of bivalve haemocytes. Fish & Shellfish Immunology 9, 367-385.

Hooper, C., Day, R., Slocombe, R., Handlinger, J., Benkendorff, K., 2007. Stress and immune response in abalone: Limitations in current knowledge and investigative methods based on other models. Fish and Shellfish Immunology 22, 363-379.

321

Références bibliographiques

Humbert, E., Coustau, C., 2001. Refractoriness of host haemocytes to parasite immunosuppressive factors as a putative resistance mechanism in the Biomphalaria glabrata-Echinostoma caproni system. Parasitology 122, 651-660.

Hurd, H., 1990. Physiological and behavioural interactions between parasites and invertebrate hosts. Advances in Parasitology 29, 271-318.

Huxham, M., Raffaelli, D., Pike, A.W., 1993. The influence of Cryptocotyle lingua (Digenea: Platyhelminthes) infections on the survival and fecundity of Littorina littorea (Gastropoda: Prosobranchia); an ecological approach. Journal of Experimental Marine Biology and Ecology 168, 223-238.

Huxham, M., Raffaelli, D., Pike, A.W., 1995. The effect of larval trematodes on the growth and burrowing behaviour of Hydrobia ulvae (Gastropoda: Prosobranchiata) in the Ythan estuary, north-east Scotland. Journal of Experimental Marine Biology and Ecology 185, 1-17.

I

Iakovleva, N.V., Shaposhnikova, T.G., Gorbushin, A.M., 2006. Rediae of echinostomatid and heterophyid trematodes suppress phagocytosis of haemocytes in Littorina littorea (Gastropoda: Prosobranchia). Experimental Parasitology 113, 24-29.

Ibarrola, I., Navarro, E., Iglesias J.I.P., 1998. Short-term adaptation of digestive processes in the cockle Cerastoderma edule exposed to different food quantity and quality. Journal of Comparative Physiology 168(B), 32-40.

Islam, M.S., Tanaka, M., 2004. Impacts of pollution on coastal and marine ecosystems including coastal and marine fisheries and approach for management: A review and synthesis. Marine Pollution Bulletin 48, 624-649.

J Janeway, C.A., Travers, P., Walport, M., Shlomchik, M., 2001. Immunobiology. The immune System in Health and Disease, Fifth Edition ed. Garand Publishing pp. 928

Jeffries, V.E., 1982. Three Vibrio strains pathogenic to larvae of Crassostrea gigas and Ostrea edulis. Aquaculture 29, 201-226.

Jenny, M.J., Ringwood, A.H., Schey, K., Warr, G.W., Chapman, R.W., 2004. Diversity of metallothioneins in the American oyster, Crassostrea virginica, revealed by transcriptomic and proteomic approaches. European Journal of Biochemistry 271, 1702-1712.

Jensen, K.T., 1992. Dynamics and growth of the cockle, Cerastoderma edule, on an intertidal mud-flat in the Danish Wadden Sea: effects of submersion time and density. Journal of Sea Research 28, 335-345.

Jensen, K.T., 1993. Density-dependent growth in cockles (Cerastoderma edule): evidence from interannual comparisons. Journal of the Marine Biological Association of the United Kingdom 73, 333-342.

Jensen, K.T., Fernández Castro, N., Bachelet, G., 1999. Infectivity of Himasthla spp. (Trematoda) in cockle (Cerastoderma edule) spat. Journal of the Marine Biological Association of the United Kingdom 79, 265-271.

322

Références bibliographiques

Jensen, K.T., Latama, G., Mouritsen, K.N., 1996. The effect of larval trematodes on the survival rates of two species of mud snails (Hydrobiidae) experimentally exposed to desiccation, freezing and anoxia. Helgoländer Meeresuntersuchungen 46 50, 327-335.

Jensen, K.T., Mouritsen, K.N., 1992. Mass mortality in two common soft-bottom invertebrates, Hydrobia ulvae and Corophium volutator - the possible role of trematodes. Helgoländer Meeresuntersuchungen 46, 329-339.

Ji, J., Choi, H.J., Ahn, I.Y., 2006. Evaluation of Manila clam Ruditapes philippinarum as a sentinel species for metal pollution monitoring in estuarine tidal flats of Korea: Effects of size, sex, and spawning on baseline accumulation. Marine Pollution Bulletin 52, 447-453.

Jung, K., Stelzenmëller, V., Zauke, G.P., 2006. Spatial distribution of heavy metal concentrations and biomass indices in Cerastoderma edule Linnaeus (1758) from the German Wadden Sea: An integrated biomonitoring approach. Journal of Experimental Marine Biology and Ecology 338, 81-95.

Justine J.L., 1998. Systématique des grands groupes de Plathelminthes parasites : quoi de neuf? Bulletin de la Société Francaise de Parasitologie 16, 34-52.

K Kägi, J.H.R., 1991. Overview of metallothionein. Methods in Enzymology 205, 613-626.

Kägi, J.H.R., 1993. Evolution, structure and chemical activity of class I metallothioneins: an overview. Metallothioneins III - Biological roles and medical implications 29-55.

Kägi, J.H.R. and Kojima, Y., 1987. Chemistry and biochemistry of metallothionein. Experientia Supplementia 52, 25-61.

Kim, S.H., Sharma, R.P., 2004. Mercury-induced apoptosis and necrosis in murine macrophages: role of calcium induced reactive oxygen species and p38 mitogen-activated protein kinase signaling. Toxicology and Applied Pharmacology 196, 47-57.

Kim, Y., Powell, E.N., 2007. Distribution of parasites and pathologies in sentinel bivalves: NOAA status and trends "Mussel Watch" program. Journal of Shellfish Research 26, 1115-1151.

Kim, Y., Powell, E.N., Wade, T.L., Presley, B.J., 2008. Relationship of parasites and pathologies to contaminant body burden in sentinel bivalves: NOAA Status and Trends 'Mussel Watch' Program. Marine Environmental Research 65, 101-127.

Kingston, P.F., 1974. Studies on the reproductive cycles of Cardium edule and C. glaucum. Marine Biology 28, 317-323.

Klar, B., Sures, B., 2004. A nonlinear model of stress hormone levels in rats - The interaction between pollution and parasites. Ecotoxicology and Environmental Safety 59, 23-30.

Klerks, P.L., Weis, J.S., 1987. Genetic adaptation to heavy metals in aquatic organisms: A review. Environmental Pollution 45, 173-205.

Krist, A.C., 2001. Variation in fecundity among populations of snails is predicted by prevalence of castrating parasites. Evolutionary Ecology Research 3, 191-197.

323

Références bibliographiques

Krzystyniak, K., Tryphonas, H., Fournier, M., 1995. Approaches to the evaluation of chemical- induced immunotoxicity. Environmental Health Perspectives 103, 17-22.

Komatsu, T., 1997. Long-term changes in the Zostera bed area in the Seto Inland Sea (Japan), especially along the coast of the Okayama Prefecture. Oceanologica Acta 20, 209-216.

Kurelec, B., Pivcevic, B., 1991. Evidence for a multixenobiotic resistance mechanism in the mussel Mytilus galloprovincialis. Aquatic Toxicology 19, 291-301.

L La Peyre, M.K., Casas, S.M., Villalba, A., La Peyre, J.F., 2008. Determination of the effects of temperature on viability, metabolic activity and proliferation of two Perkinsus species, and its significance to understanding seasonal cycles of perkinsosis. Parasitology 135, 505-519.

Lacoste, A., Cueff, A., Poulet, S.A., 2001. P35-sensitive caspases, MAP kinases and Rho modulate - adrenergic induction of apoptosis in mollusc immune cells. Journal of Cell Science 115, 761- 768.

Lafferty, K.D., 1997. Environmental parasitology: what can parasites tell us about human impacts on the environment? Parasitology Today 13, 251-255.

Lafferty, K.D., Kuris, A.M., 1999. How environmental stress affects the impacts of parasites. Limnology and Oceanography 44, 925-931.

Lagadic L., Caquet T., Amiard J-C., Ramade F., 1997. Biomarqueurs en écotoxicologie. Aspects fondamentaux. Eds. Collection Écologie, Paris, Masson, pp. 419.

Lambert, C., Nicolas, J.L., 1998. Specific Inhibition of Chemiluminescent Activity by Pathogenic Vibrios in Hemocytes of Two Marine Bivalves: Pecten maximus and Crassostrea gigas. Journal of Invertebrate Pathology 71, 53-63.

Lambert, C., Soudant, P., Choquet, G., Paillard, C., 2003. Measurement of Crassostrea gigas hemocyte oxydative metabolism by flow cytometry and the inhibiting capacity of pathogenic vibrios. Fish and Shellfish Immunology 15, 225-240.

Lambert, C., Soudant, P., Dégremont, L., Delaporte, M., Moal, J., Boudry, P., Jean, F., Huvet, A., Samain, J.-F., 2007. Hemocyte characteristics in families of oysters, Crassostrea gigas, selected for differential survival during summer and reared in three sites. Aquaculture 270, 276-288.

Langston, W.J., Bebianno, M.J. and Burt, G.R., 1998. Metal handling strategies in molluscs. In: Langston, W.J. and Bebianno, M.J., Editors, 1998. Metal Metabolism in Aquatic Environments, Chapman and Hall, London, pp. 219–284.

Lapanje, A., Drobne, D., Nolde, N., Valant, J., Muscet, B., Leser, V., Rupnik, M., 2008. Long-term Hg pollution induced Hg tolerance in the terrestrial isopod Porcellio scaber (Isopoda, Crustacea). Environmental Pollution 153, 537-547.

Laporte, J.M., Andres, S., Mason, R.P., 2002. Effect of ligands and other metals on the uptake of mercury and methylmercury across the gills and the intestine of the blue crab (Callinectes sapidus). Comparative Biochemistry and Physiology 131(C), 185-196.

Laruelle, F., Guillou, J., Paulet, Y.M., 1994. Reproductive pattern of the clams, Ruditapes decussatus and R. philippinarum on intertidal flats in Brittany. Journal of the Marine Biological Association of the United Kingdom 74, 351-366.

324

Références bibliographiques

Lassalle, G., de Montaudouin, X., Soudant, P., Paillard, C., 2007. Parasite co-infection of two sympatric bivalves, the Manila clam (Ruditapes philippinarum) and the cockle (Cerastoderma edule) along a latitudinal gradient. Aquatic Living Resources 20, 33-42.

Lauckner, G., 1983. Diseases of Mollusca: Bivalvia, in: Kinne, O. (Ed.), Diseases of Marine Animals. Biologische Anstalt Helgoland, Hamburg, pp. 477-961.

Lauckner, G., 1987. Ecological effects of larval trematode infestation on littoral marine invertebrate populations. International Journal for Parasitology 17, 391-398.

Le Moullac, G., Fleury, P.G., Le Coz, J.R., Moal, J., Samain, J.F., 2007. Effect of sediment nearness on the metabolic enzyme activity and energy state of the oyster Crassostrea gigas. Aquatic Living Resources 20, 279-286.

Le Treut, Y., 1986. La palourde. Anatomie - Biologie - Elevage - Pêche - Consommation - Inspection sanitaire. Ecole Nationale Vétérinaire - Université de Nantes, Nantes, p. 162.

Lee, M.-K., Cho, B.-Y., Lee, S.J., Kang, J.-Y., Jeong, H.D., Huh, S.H., Huh, M.-D., 2001. Histopathological lesions of Manila clam, Tapes philippinarum, from Hadong and Namhae coastal areas of Korea. Aquaculture 201, 199-209.

Lecoeur, S., Videmann, B., Berny, P., 2004. Evaluation of metallothionein as a biomarker of single and combined Cd/Cu exposure in Dreissena polymorpha. Environmental Research 94, 184-191.

Legeay, A., Achard-Joris, M., Baudrimont, M., Massabuau, J.C., Bourdineaud, J.P., 2005. Impact of cadmium contamination and oxygenation levels on biochemical responses in the Asiatic clam Corbicula fluminea. Aquatic Toxicology 74, 242-253.

Leignel, V. and Laulier, M., 2006. Isolation and characterization of Mytilus edulis metallothionein genes. Comparative Biochemistry and Physiology 142(C), 12-18.

Leinio, S., Lehtonen, K.K., 2005. Seasonal variability in biomarkers in the bivalves Mytilus edulis and Macoma balthica from the northern Baltic Sea. Comparative Biochemistry and Physiology – 140(C), 408-421.

Leiva-Presa, Ã., Capdevila, M., Cols, N., Atrian, S., Gonzãlez-Duarte, P., 2004. Chemical foundation of the attenuation of methylmercury(II) cytotoxicity by metallothioneins. European Journal of Biochemistry 271, 1323-1328.

Lemoine, S., Laulier, M., 2003. Potential use of the levels of the mRNA of a specific metallothionein isoform (MT-20) in mussel (Mytilus edulis) as a biomarker of cadmium contamination. Marine Pollution Bulletin 46, 1450-1455.

Limke, T.L., Otero-Montanez, J.K.L., Atchison, W.D., 2003. Evidence for interactions between intracellular calcium stores during methylmercury-induced intracellular calcium dysregulation in rat cerebellar granule neurons. Journal of Pharmacology and Experimental Therapeutics. 304, 949-958.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔCT method. Methods 25, 402-408.

Lorteau, C., Auffret, M., Lebris, H., 1995. The Immunodefense System of Bivalves. 1. Structure and Working. Recueil De Medecine Veterinaire 171, 415-422.

325

Références bibliographiques

López, C., Carballal, M.J., Azevedo, C., Villalba, A., 1997. Differential phagocytic ability of the circulating haemocyte types of the carpet shell clam Ruditapes decussatus (Mollusca: Bivalvia). Diseases of Aquatic Organisms 30, 209-215.

López-Cortes, L., Castro, D., Navas, J.I., Borrego, J.J., 1999. Phagocytic and chemotactic responses of manila and carpet shell clam haemocytes against Vibrio tapetis, the causative agent of brown ring disease. Fish and Shellfish Immunology 9, 543-555.

Lucia, M., 2009. Analyse de la contamination polymétallique des oiseaux aquatiques ? Étude expérimentale de la bioaccumulation du cadmium, réponses adaptatives et effets toxiques associés. Ph.D. Thesis, University de Pau et des Pays de l‘Adour.

Luoma, S.N., Rainbow, P.S., 2005. Why is metal bioaccumulation so variable? Biodynamics as a unifying concept. Environmental Science and Technology 39, 1921-1931.

M Mackay, E.A., Overnell, J., Dunbar, B., Davidson, I., Hunziker, P.E., Kägi, J.H.R., Fothergill, J.E., 1993. Complete amino acid sequences of five dimeric and four monomeric forms of metallothionein from the edible mussel Mytilus edulis. European Journal of Biochemistry 218, 183-194.

MacKenzie, K., Williams, H.H., Williams, B., McVicar, A.H., Siddall, R., 1995. Parasites as indicators of water quality and the potential use of helminth transmission in marine pollution studies. Advances in Parasitology 35, 85-144.

Maes, P., Paillard, C., 1992. Effect du Vibrio P1, pathogène de Ruditapes philippinarum, sur d'autres espèces de bivalves. In, Les Mollusques marins, Biologie et Aquaculture, Vol. 14, Actes de Colloques Ifremer pp. 141-148.

Maillard, C., 1976. Distomatoses de poissons en milieu lagunaire. University Sciences et Techniques du Languedoc, p. 383.

Margoshes, M. and Vallee, B.L., 1957. A cadmium protein from equine kidney cortex. Journal of the American Chemical Society 79, 4813-4814.

Marie, D., Partensky, F., Vaulot, D., Brussaard, C.P.D., 1999. Enumeration of phytoplankton, bacteria, and viruses in marine samples. J.P. Robinson, Z. Darzynkiewicz, P.N. Dean, A. Orfao, P. Rabinovitch, C. Stewart, H.J. Tanke and L. Wheeless, Editors, Current Protocols in Cytometry, John Wiley & Sons, Inc., New York. N.Y., 11.11.11–11.11.15.

Marie, V., Gonzalez, P., Baudrimont, M., Bourdineaud, J.P., Boudou, A., 2006. Metallothionein response to cadmium and zinc exposures compared in two freshwater bivalves, Dreissena polymorpha and Corbicula fluminea. BioMetals 19, 399-407.

Marigómez I., Soto M., Cajaraville M.P., 1995. Morphofunctional patterns of cell and tissue systems involved in metal handling and metabolism‖. In: Soto M., Cajaraville M.P., 89-134.

Marigómez, I., Soto, M., Cajaraville, M.P., Angulo, E., Giamberini, L., 2002. Cellular and subcellular distribution of metals in molluscs. Microscopy Research and Technique 56, 358-392.

Markich, S.J., Brown, P.L., Jeffree, R.A., Lim, R.P., 2000. Valve movement responses of Velesunio angasi (Bivalvia: Hyriidae) to manganese and uranium: An exception to the free ion activity model. Aquatic Toxicology 51, 155-175.

326

Références bibliographiques

Marmaras, V.J., Lampropoulou, M., 2009. Regulators and signalling in insect haemocyte immunity. Cellular Signalling 21, 186-195.

Martín-Díaz, M.L., Jiménez-Tenorio, N., Sales, D., DelValls, T.A., 2008. Accumulation and histopathological damage in the clam Ruditapes philippinarum and the crab Carcinus maenas to assess sediment toxicity in Spanish ports. Chemosphere 71, 1916-1927.

Mason, A.Z., Jenkins, K.D., 1995. Metal detoxification in aquatic organisms. In: Tessier, A.., Turner, D.R. (Eds.), Metal Speciation and Bioavailability in Aquatic Systems. Wiley, Chichester, UK, 479607.

Matozzo, V., Rova, G., Marin, M.G., 2007. Haemocytes of the cockle Cerastoderma glaucum: morphological characterisation and involvement in immune responses. Fish and Shellfish Immunology 23, 732-746.

McGladdery, S.E., 1999. Shellfish Diseases (Viral, Bacterial and Fungal). Fish Disease and Disorder 3, 723–842.

McHenery, J.G., Birkbeck, T.H., 1986. Inhibition of filtration in Mytilus edulis L. by marine vibrios. Journal of Fish Disease 9, 257–261.

Meyhöfer, E., Morse, M.P., 1996. Characterization of the bivalve ultrafiltration system in mytilus edulis, chlamys hastata, and mercenaria mercenaria. Invertebrate Biology 115, 20-29.

Minier, C., Moore, M.N., 1996. Induction of multixenobiotic resistance in mussel blood cells. Marine Environmental Research 42, 389-392.

Miquel, J.C., 2001. Environment and biology of the Kara sea: A general view for contamination studies. Marine Pollution Bulletin 43, 19-27.

Mix, M.C., 1976. A general model for leucocyte cell renewal in bivalve molluscs. Marine Fisheries Review, 37-41.

Mohamed, M.A.M., Osman, M.A., Potter, T.L., Levin, R.E., 1998. Lead and cadmium in Nile River water and finished drinking water in Greater Cairo, Egypte Environmental International 24, 767-772.

Morel, F.M., 1983. Complexation: Trace metals and microorganisms. Principles of Aquatic Chemistry, 301-308.

Moriyama, E.N., Petrov, D.A., Hartl, D.L., 1998. Genome size and intron size in Drosophila. Molecular Biology and Evolution 15, 770-773.

Morley, N.J., 2006. Parasitism as a source of potential distortion in studies on endocrine disrupting chemicals in molluscs. Marine Pollution Bulletin 52, 1330-1332.

Morley, N.J., 2009. Interactive effects of infectious diseases and pollution in aquatic molluscs. Aquatic Toxicology. In press.

Morley, N.J., Crane, M., Lewis, J.W., 2001. Toxicity of cadmium and zinc to encystment and in vitro excystment of Parorchis acanthus (Digenea: Philophthalmidae). Parasitology 122, 75-79.

Morley, N.J., Crane, M., Lewis, J.W., 2003a. Toxicity of cadmium and zinc to cercarial activity of Diplostomum spathaceum (Trematoda: Diplostomatidae). Folia Parasitology 50, 57-60.

327

Références bibliographiques

Morley, N.J., Crane, M., Lewis, J.W., 2003b. Effects of cadmium and zinc toxicity on orientation behaviour of Echinoparyphium recurvatum (Trematoda: Echinostomatidae) cercariae. Disease of Aquatic Organisms 56, 89-92.

Morley, N.J., Crane, M., Lewis, J.W., 2004. Influence of cadmium exposure on the incidence of first intermediate host encystment by Echinoparyphium recurvatum cercariae in Lymnaea peregra. Journal of Helminthology 78, 329-332.

Morley, N.J., Crane, M., Lewis, J.W., 2005. Changes in survival characteristics of Diplostomum spathaceum cercariae emerged from cadmium-exposed Lymnaea stagnalis. Journal of Helminthology 79, 55-59.

Morley, N.J., Lewis, J.W., Hoole, D., 2006. Pollutant-induced effects on immunological and physiological interactions in aquatic host-trematode systems: Implications for parasite transmission. Journal of Helminthology 80, 137-149.

Moschino, V., Meneghetti, F., Da Ros, L., 2009. Use of biomarkers to assess the welfare of the edible clam, Ruditapes philippinarum: may it be a tool for proving areas of origin? Aquaculture International, 1-11.

Mouneyrac, C., Mastain, O., Amiard, J.C., Amiard-Triquet, C., Beaunier, P., Jeantet, A.Y., Smith, B.D., Rainbow, P.S., 2003. Trace-metal detoxification and tolerance of the estuarine worm Hediste diversicolor chronically exposed in their environment. Marine Biology 143, 731-744.

Mourez, M., Jéhanno, M., Hofnung, M., Dassa, E., 2000. Role, functional mechanism and structure of ABC (ATP-binding cassette) transporters. Role, fonctionnement et structure des transporteurs a ATP binding cassette (ABC) 16, 386-394.

Mouritsen, K.N., Jensen, K.T., 1994. The enigma of gigantism: effect of larval trematodes on growth, fecundity, egestion and locomotion in Hydrobia ulvae (Pennant) (Gastropoda: Prosobranchia). Journal of Experimental Marine Biology and Ecology 181, 53-66.

Mubiana, V.K., Blust, R., 2007. Effects of temperature on scope for growth and accumulation of Cd, Co, Cu and Pb by the marine bivalve Mytilus edulis. Marine Environmental Research 63 (3), 219- 235.

Muñoz, P., Meseguer, J., Esteban, M.A., 2006. Phenoloxidase activity in three commercial bivalve species. Changes due to natural infestation with Perkinsus atlanticus. Fish and Shellfish Immunology 20, 12-19.

N Nacci, D.E., Champlin, D., Coiro, L., McKinney, R., Jayaraman, S., 2002. Predicting the occurrence of genetic adaptation to dioxinlike compounds in populations of the estuarine fish Fundulus heteroclitus. Environmental Toxicology and Chemistry 21, 1525-1532.

Nascimento, S.F., Kurzweil, H., Wruss, W., Fenzl, N., 2006. Cadmium in the Amazonian Guajará Estuary: Distribution and remobilization. Environmantal Pollution 140, 29-42.

Ng, T.Y.-T., Wang, W.-X., 2004. Detoxification and effects of Ag, Cd, and Zn pre-exposure on metal uptake kinetics in the clams Ruditapes philippinarum. Marine Ecology Progress Series 268, 161-172.

Ngo, T.T.T., Choi, K.-S., 2004. Seasonal changes of Perkinsus and Cercaria infections in the Manila clam Ruditapes philippinarum from Jeju, Korea. Aquaculture 239, 57-68.

328

Références bibliographiques

Nicholson, S., 2003. Cardiac and branchial physiology associated with copper accumulation and detoxication in the mytilid mussel Perna viridis (L.). Journal of Experimental Marine Biology and Ecology 295, 157-171.

Nordberg, M., 1998. Metallothioneins: Historical review and state of knowledge. Talanta 46, 243-254.

Nosho, T.Y., Chew, K.K., 1972. The setting and growth of the Manila clam, Venerupis japonica (Deshayes), in Hood Canal, Washington. Proceedings of National Shellfish Association 62, 50- 58.

Nunes, M., Coelho, J.P., Cardoso, P.G., Pereira, M.E., Duarte, A.C., Pardal, M.A., 2008. The macrobenthic community along a mercury contamination in a temperate estuarine system (Ria de Aveiro, Portugal). Science of the Total Environment 405, 186-194.

O Ohba, S., 1959. Ecological studies in the natural population of a clam, Tapes japonica, with special reference to seasonal variations in the size and structure of the population and to individual growth. Biological Journal of Oyakama University 5, 13-42.

OIE, 2006 (World Organization for Animal Health), 2006. Manual of Diagnostic Tests for Aquatic Animals. Fifth Edition, Paris, France.

Olafsen, J.A., 1996. Lectins: models of natural and induced molecules in invertebrates. In: Cooper EL, editor. Invertebrate immune responses: cells and molecular products. Heidelberg: Springer- Verlag 24, 49-76.

Olafson, R.W., Kearns, A., Sim, R.G., 1979. Heavy metal induction of metallothionein synthesis in the hepatopancreas of the crab Scylla serrata. Comparative Biochemistry and Physiology 62, 417- 424.

Oliver, L.M., Fisher, W.S., Volety, A.K., Malaeb, Z., 2003. Greater hemocyte bactericidal activity in oysters (Crassostrea virginica) from a relatively contaminated site in Pensacola Bay, Florida. Aquatic Toxicology 64, 363-373.

Ottaviani, E., Paemen, L.R., Cadet, P., Stefano, G.B., 1993. Evidence for Nitric-Oxide Production and Utilization as a Bacteriocidal Agent by Invertebrate Immunocytes. European Journal of Pharmacology-Environmental Toxicology and Pharmacology Section 248, 319-324.

Oubella, R., Maes, P., Paillard, C., Auffret, M., 1993. Experimentally induced variation in hemocyte density for Ruditapes philippinarum and R. decussatus (Mollusca, Bivalvia). Disease Aquatic Organisms 15, 193-197.

Oubella, R., Auffret, M., 1995. Immuno-modulation in populations of mollusc bivalves from the Brest Roadsted. Acte de Rencontre Scientifique International. Communaute Urbaine, Brest (France), p 307-319.

Overnell, J., Good, M., Vasák, M., 1988. Spectroscopic studies on cadmium (II)- and cobalt(II)- substituted metallothionein from the crab Cancer pagurus. Evidence for one additional low- affinity metal-binding site. European Journal of Biochemistry 172, 171-177.

Ownby, D.R., Newman, M.C., Mulvey, M., Vogelbein, W.K., Unger, M.A., Arzayus, L.F., 2002. Fish (Fundulus heteroclitus) populations with different exposure histories differ in tolerance of creosote-contaminated sediments. Environmental Toxicology and Chemistry21: 1897-1902.

329

Références bibliographiques

P Paillard, C., 2004. A short-review of brown ring disease, a vibriosis affecting clams, Ruditapes philippinarum and Ruditapes decussatus. Aquatic Living Resources 17, 467-475.

Paillard, C., Allam, B., Oubella, R., 2004. Effect of temperature on defense parameters in Manila clam Ruditapes philippinarum challenged with Vibrio tapetis. Diseases of Aquatic Organisms 59, 249-262.

Paillard, C., Ford, S., 2002. Repeated hemolymph sampling of individual clams: what does it tell us about sampling procedures? In, National Shellfish Association Workshop, Vol. 21 1, J. Shell. Res., Mystic, USA, pp. 406.

Paillard, C., Gausson, S., Nicolas, J.-L., le Pennec, J.-P., Haras, 2006. Molecular identification of Vibrio tapetis, the causative agent of the brown ring disease of Ruditapes philippinarum. Aquaculture 253, 25-38.

Paillard, C., Korsnes, K., Le Chevalier, P., Le Boulay, C., Harkestad, L., Eriksen, A.G., Willassen, E., Bergh, Ø., Bovo, C., Skår, C., Mortensen, S., 2008. Vibrio tapetis-like strain isolated from introduced Manila clams Ruditapes philippinarum showing symptoms of brown ring disease in Norway. Diseases of Aquatic Organisms 81, 153-161.

Paillard, C., Maes, P., 1990. Etiology of the Brown Ring Disease in Tapes-Philippinarum - Pathogenicity of a Vibrio Sp. Comptes Rendus De L Academie Des Sciences Serie Iii-Sciences De La Vie-Life Sciences 310, 15-20.

Paillard, C., Maes, P., 1994. The brown ring disease in the Manila clam, Ruditapes philippinarum. II. Microscopic study of the brown ring syndrome. Journal of Invertebrate Pathology 65, 101-110.

Paillard, C., Maes, P., Oubella, R., 1994. Brown ring disease in clams. Annual Review of Fish Diseases 4, 219-240.

Paillard, C., Maes, P., Mazurié, J., Claude, S., Marhic, A., Le Pennec, M., 1997. Epidemiological survey of the brown ring disease in clams of Atlantic coast: role of temperature in variations of prevalence, 8th Symposium of the International Society for Veterinary Epidemiology and Economics. Association pour l'Etude de l'Epidémiologie et des Maladies animales, pp. 8-11.

Paillard C., Percelay L., Le Pennec M., Picard D. 1989. Origine pathogène de l'anneau brun chez Tapes philippinarum (Mollusque bivalve). Comptes Rendus de l'Académie des Sciences. Paris 309, 235-241.

Pan, K., Wang, W.X., 2009. Biodynamics to explain the difference of copper body concentrations in five marine bivalve species. Environmental Science and Technology 43, 2137-2143.

Pan, P.K., Hou, F.Y., Cody, C.W., Huang, P.C., 1994. Substitution of glutamic acids for the conserved lysines in the domain affects metal binding in both the and β domains of mammalian metallothionein. Biochemical and Biophysical Research Communications 202, 621-628.

Park, H., Ahn, I.Y., Choi, H.J., Pyo, S.H., Lee, H.E., 2007. Cloning, expression and characterization of metallothionein from the Antarctic clam Laternula elliptica. Protein Expression and Purification 52, 82-88.

Park, K.-I., Paillard, C., Le Chevalier, P., Choi, K.-S., 2006. Report on the occurrence of brown ring disease (BRD) in Manila clam, Ruditapes philippinarum, on the west coast of Korea. Aquaculture 255, 610-613.

330

Références bibliographiques

Park, K.-I., Tsutsumi, H., Hong, J.-S., Choi, K.-S., 2008. Pathology survey of the short-neck Ruditapes philippinarum occurring on sandy tidal flats along the coast of Ariake Bay, Kyushu, Japan. Journal of Invertebrate Pathology 99, 212-219.

Parry, H.E., Pipe, R.K., 2004. Interactive effects of temperature and copper on immunocompetence and disease susceptibility in mussels (Mytilus edulis). Aquatic Toxicology 69, 311-325.

Parsch, J., 2003. Selective constraints on intron evolution in Drosophila. Genetics 165, 1843-1851.

Paul-Pont, I., Baudrimont, M., Gonzalez, P., de Montaudouin, X., 2007. Reproductive cycle modulates metallothionein synthesis in the cockle (Cerastoderma edule) after parasite and cadmium contamination, in: Kungolos, A., Aravossis, K., Karagiannidis, A., Samaras, P. (Eds.), Proceedings of the International Conference on Environmental Management, Engineering, Planning and Economics, Skiathos (Greece), pp. 89-95.

Paul-Pont, I., de Montaudouin, X., Gonzalez, P., Soudant, P., Baudrimont, M., 2009. How life-history contributes to stress response in the Manila clam Ruditapes philippinarum. Environmental Science and Pollution Research In press.

Paul-Pont, I., Gonzalez, P., Montero, N., de Montaudouin, X., Baudrimont, M., In revision. Cloning, characterization and gene expression of a metallothionein isoform in the edible cockle Cerastoderma edule after cadmium or mercury exposure. BioMetals.

Pedersen, K.L., Pedersen, S.N., Hojrup, P., Andersen, J.S., Roepstorff, P., Knudsen, J., Depledge, M.H., 1994. Purification and characterization of a cadmium-induced metallothionein from the shore crab Carcinus maenas (L.). Biochemistry Journal 297, 609-614.

Perceval, O., Couillard, Y., Pinel-Alloul, B., Bonneris, E., Campbell, P.G.C., 2006. Long-term trends in accumulated metals (Cd, Cu and Zn) and metallothionein in bivalves from lakes within a smelter-impacted region. Science of the Total Environment 369, 403-418.

Phillips, D.J.H., 1977. The use of biological indicator organisms to monitor trace metal pollution in marine and estuarine environments: a review. Environmental Pollution 13, 281-317.

Pierron, F., 2007. Etudes des mécanismes de bioaccumulation du cadmium et des effets toxiques associés à différents stades de vie de l'anguille européenne (Anguilla anguilla L.). Ph.D. Thesis, University Bordeaux 1.

Piola, R.F., Johnston, E.L., 2006. Differential tolerance to metals among populations of the introduced bryozoan Bugula neritina. Marine Biology 148, 997-1010.

Pipe, R.K., 1990. Differential binding of lectins to haemocytes of the mussel Mytilus edulis. Cell and Tissue Research 261, 261-268.

Pipe, R.K., Coles, J.A., 1995. Environmental contaminants influencing immune function in marine bivalve molluscs. Fish and Shellfish Immunology 5, 581-595.

Pipe, R.K., Coles, J.A., Carissan, F.M.M., Ramanathan, K., 1999. Copper induced immunomodulation in the marine mussel, Mytilus edulis. Aquatic Toxicology 46, 43-54.

Ponsero, A., Dabouineau, L., Allain, J., 2009. Modeling of common European cockle Cerastoderma edule fishing grounds aimed at sustainable management of traditional harvesting. Fisheries Science, 1-12.

331

Références bibliographiques

Ponurovsky, S.K., Selin, I.N., 1988. Distribution, population structure, and growth of the bivalve mollusk Ruditapes phillipinarum in Vostok Bay, Sea of Japan. 11-15.

Ponurovsky, S.K., Yakovlev, Y.M., 1992. The reproductive biology of the japanese littleneck, Tapes phillipinarum (A. Adams and Reeve, 1850) (Bivalvia: Veneridae). Journal of Shellfish Research 11, 265-277.

Potasman, I., Paz, A., Odeh, M., 2002. Infectious outbreaks associated with bivalve shellfish consumption: A worldwide perspective. Clinical Infectious Diseases 35, 921-928.

Pylkkö, P., Suomalainen, L.R., Tiirola, M., Valtonen, E.T., 2006. Evidence of enhanced bacterial invasion during Diplostomum spathaceum infection in European grayling, Thymallus thymallus (L.). Journal of Fish Diseases 29, 79-86.

R Rabestein, D.L., Young, J.D., Wolowyk, M.W., Tahir Razi, M., Arnold, A.P., Tucker, E.M., 1985. Characterization of normal, glutathione-deficient and arginase-deficient sheep erythrocytes by 1H-NMR spectroscopy. Biochimica and Biophysica Acta 846, 200-207.

Rajkumar, M., Thavasi, R., Perumal, P., Trilles, J.P., 2007. Parasite induced vibriosis in Stolephorus commersonii. Research Journal of Microbiology 2, 972-977.

Raspor, B., Dragun, Z., Erk, M., Ivanković, D., Pavičić, J., 2004. Is the digestive gland of Mytilus galloprovincialis a tissue of choice for estimating cadmium exposure by means of metallothioneins? Science of the Total Environment 333, 99-108.

Ray, S.M., 1966. Effects of various antibiotics on the fungus Dermocystidium marinum in thioglycollate cultures of oyster tissues. Journal of Invertebrate Pathology 8, 433-438.

Regala, R.P., Rice, C.D., 2004. Mycobacteria, but not mercury, induces metallothionein (MT) protein in striped bass, Morone saxitilis, phagocytes, while both stimuli induce MT in channel catfish, Ictalurus punctatus, phagocytes. Marine Environmental Research 58, 719-723.

Reid, H.I., Soudant, P., Lambert, C., Paillard, C., Birkbeck, T.H., 2003. Salinity effects on immune parameters of Ruditapes philippinarum challenged with Vibrio tapetis. Diseases of Aquatic Organisms 56, 249-258.

Ren, X.Y., Zhou, Y., Zhang, J.P., Feng, W.H., Jiao, B.H., 2003a. Expression of metallothionein gene at different time in testicular interstitial cells and liver of rats treated with cadmium. World Journal of Gastroenterology 9, 1554-1558.

Ren, X.Y., Zhou, Y., Zhang, J.P., Zhong, Y., Feng, W.H., Jiao, B.H., 2003b. Metallothionein Gene Expression under Different Time in Testicular Sertoli Cells of Rats Treated with Cadmium. Progress in Biochemistry and Biophysic 30, 965-968.

Renault, T., Le Deuff, R.M., Chollet, B., Cochennec, N., Gerard, A., 2000. Concomitant herpes-like virus infections in hatchery-reared larvae and nursery-cultured spat Crassostrea gigas and Ostrea edulis. Diseases of Aquatic Organisms 42, 173-183.

Rittschof, D., McClellan-Green, P., 2005. Molluscs as multidisciplinary models in environment toxicology. Marine Pollution Bulletin 50, 369-373.

Roch, P., 1999. Defense mechanisms and disease prevention in farmed marine invertebrates. Aquaculture 172, 125-145.

332

Références bibliographiques

Rodríguez, J.M., López-Romalde, S., Beaz, R., Alonso, M.C., Castro, D., Romalde, J.L., 2006. Molecular fingerprinting of Vibrio tapetis strains using three PCR-based methods: ERIC-PCR, REP-PCR and RADP. Diseases of Aquatic Organisms 69, 175-183.

Roesijadi, G., 1992. Metallothioneins in metal regulation and toxicity in aquatic animals. Aquatic Toxicology 22, 81-114.

Roesijadi, G., 1994. Metallothionein induction as a measure of response to metal exposure in aquatic animals. Environmental Health Perspectives 102, 91-95.

Roesijadi, G., 1996. Metallothionein and its role in toxic metal regulation. Comparative Biochemistry and Physiology 113(C), 117-123.

Roméo, M., Gnassia-Barelli, M., 1995. Metal distribution in different tissues and in subcellular fractions of the Mediterranean clam Ruditapes decussatus treated with cadmium, copper, or zinc. Comparative Biochemistry and Physiology 111(C), 457-463.

Roméo, M., Hoarau, P., Garello, G., Gnassia-Barelli, M., Girard, J.P., 2003. Mussel transplantation and biomarkers as useful tools for assessing water quality in the NW Mediterranean. Environmental Pollution 122, 369-378.

Ross, K., Cooper, N., Bidwell, J.R., Elder, J., 2002. Genetic diversity and metal tolerance of two marine species: A comparison between populations from contaminated and reference sites. Marine Pollution Bulletin 44, 671-679.

Russell-Pinto, F., Reimao, R., de Sousa, M., 1994. Haemocytes in Cerastoderma edule (Mollusca, Bivalvia): distinct cell types engage in different responses to sheep erythrocytes. Fish and Shellfish Immunology 4, 383-397.

Rygg, B., 1970. Studies of Cerastoderma edule (L.) and Cerastoderma galucum (Poiret). Sarsia 43, 65-80.

S Saavedra, Y., González, A., Blanco, J., 2009. Inter-individual distribution of metal concentrations in four marine bioindicator organisms and its use for optimal sampling design of a monitoring system. International Journal of Environmental Analytical Chemistry 89, 559-568.

Saavedra, Y., González, A., Fernández, P., Blanco, J., 2004. Interspecific variation of metal concentrations in three bivalve mollusks from Galicia. Archives of Environmental Contamination and Toxicology 47, 341-351.

Saydam, N., Adams, T.K., Steiner, F., Schaffner, W., Freedman, J.H., 2002. Regulation of metallothionein transcription by the metal-responsive transcription factor MTF-1: Identification of signal transduction cascades that control metal-inducible transcription. Journal of Biology and Chemistry 277, 20438-20445.

Serafim, M.A., Bebianno, M.J., 2001. Variation of metallothionein and metal concentrations in the digestive gland of the clam Ruditapes decussatus: sex and seasonal effects. Environmental Toxicology and Chemistry 20, 544-552.

Serafim M.A., Company R.M., Bebianno M.J., Langston W.J., 2002. Effect of temperature and size on metallothionein synthesis in the gill of Mytilus galloprovincialis exposed to cadmium. Marine Environmental Research 54, 361-365.

333

Références bibliographiques

Sfriso, A., Argese, E., Bettiol, C., Facca, C., 2008. Tapes philippinarum seed exposure to metals in polluted areas of the Venice lagoon. Estuarine, Coastal and Shelf Science 79, 581-590.

Shi, D., Blackmore, G., Wang, W.X., 2003. Effects of aqueous and dietary preexposure and resulting body burden on silver biokinetics in the green mussel Perna viridis. Environmental Science and Technology 37, 936-943.

Shi, D., Wang, W.X., 2004. Understanding the Differences in Cd and Zn Bioaccumulation and Subcellular Storage among Different Populations of Marine Clams. Environmental Science and Technology 38, 449-456.

Shi, D., Wang, W.-X., 2005. Uptake of aqueous and dietary metals by mussel Perna viridis with different Cd exposure histories. Environmental Science and Technology 39, 9363–9369.

Shirley, M.D.F., Sibly, R.M., 1999. Genetic basis of a between-environment trade-off involving resistance to cadmium in Drosophila melanogaster. Evolution 53, 826-836.

Shoemaker, C.A., Xu, D., Klesius, P.H., Evans, J.J., 2008. Concurrent Infections (Parasitism and Bacterial Disease) in Tilapia, Proceedings of the 8th International Symposium on Talipia in Aquaculture, Cairo, Egypt, pp. 1365-1375.

Simkiss, K., Mason, A.Z., 1984. Cellular responses of molluscan tissues to environmental metals. Marine Environmental Research 14, 103-118.

Simkiss, K., Taylor, M., 1989. Metal fluxes across the membranes of aquatic organisms. CRC Critical Reviews of Aquatic Sciences 1, 173–188.

Smaal, A.C., 1997. Food supply and demand of bivalve suspension feeders in a tidal system. Groningen, p. 235.

Smaoui-Damak, W., Hamza-Chaffai, A., Bebianno, M.J., Amiard, J.C., 2004. Variation of metallothioneins in gills of the clam Ruditapes decussatus from the Gulf of Gabès (Tunisia). Comparative Biochemistry and Physiology 139(C), 181-188.

Smith, V.J., Söderhäll, K., 1991. A comparison of phenoloxidase activity in the blood of marine invertebrates. Developmental and Comparative Immunology 15, 251-261.

Snape, J.R., Maund, S.J., Pickford, D.B., Hutchinson, T.H., 2004. Ecotoxicogenomics: the challenge of integrating genomics into aquatic and terrestrial ecotoxicology. Aquatic Toxicology 67, 143- 154.

Snieszko, S.F., 1974. The effects of environmental stress on outbreaks of infectious diseases of fishes. Journal of Fish Biology 6, 197-208.

Söderhäll, K., Cerenius, L., 1998. Role of the prophenoloxidase-activating system in invertebrate immunity. Current Opinion in Immunology 10, 23-28.

Sokal, R., Rohlf, F., 1981. Biometry. WH Freeman, New York.

Sokolova, I.M., Ringwood, A.H., Johnson, C., 2005. Tissue-specific accumulation of cadmium in subcellular compartments of eastern oysters Crassostrea virginica Gmelin (Bivalvia: Ostreidae). Aquatic Toxicology 74, 218-228.

334

Références bibliographiques

Soudant, P., Paillard, C., Choquet, G., Lambert, C., Reid, H.I., Marhic, A., Donaghy, L., Birkbeck, T.H., 2004. Impact of season and rearing site on the physiological and immunological parameters of the Manila clam Venerupis (=Tapes, =Ruditapes) philippinarum. Aquaculture 229 (1-4), 401-418.

Sousa, W.P., 1983. Host life history and effect of parasitic castration on growth: a field study of Cerithidea californica Haldeman (Gastropoda: Prosobranchia) and its trematode parasites. Journal of Experimental Marine Biology and Ecology 73, 273-296.

Sparks, A.K., Morado, J.F., 1988. Inflammation and wound repair in bivalve molluscs. American Fisheries Society Special Publication 18, 139-152.

Steiner, H., Hultmark, D., Engström, A., Bennich, H., Boman, H.G., 1981. Sequence and specificity of two antibacterial proteins involved in insect immunity. Nature 292, 246-248.

Stillman, M.J., 1995. Metallothioneins. Coordination Chemistry Reviews 144, 461-511.

Sures, B., 2004. Environmental parasitology: relevancy of parasites in monitoring environmental pollution. Trends in Parasitology 20, 170-177.

Sures, B., 2006. How parasitism and pollution affect the physiological homeostasis of aquatic hosts. Journal of Helminthology 80, 151-157.

Sures, B., 2008. Environmental parasitology. Interactions between parasites and pollutants in the aquatic environment. Parasite 15 (3), 434-438.

Sures, B., Radszuweit, H., 2007. Pollution-induced heat shock protein expression in the amphipod Gammarus roeseli is affected by larvae of Polymorphus minutus (Acanthocephala). Journal of Helminthology 81, 191-197.

T Tanguy, A., Boutet, I., Riso, R., Boudry, P., Auffret M., Moraga D., 2003. Metallothionein genes in the European flat oyster Ostrea edulis: a potential ecological tool for environmental monitoring. Marine Ecology Progress Series 257, 87-97.

Tanguy, A., Moraga, D., 2001. Cloning and characterization of a gene coding for a novel metallothionein in the Pacific oyster Crassostrea gigas (CgMT2): A case of adaptive response to metal-induced stress? Gene 273, 123-130.

Tanguy, A., Mura, C., Moraga, D., 2001. Cloning of a metallothionein gene and characterization of two other cDNA sequences in the Pacific oyster Crassostrea gigas (CgMT1). Aquatic Toxicology 55, 35-47.

Tebble, N., 1966. British Bivalve seashells. British Museum of Natural History, London.

Templeton, D.M., Ariese, F., Cornelis, R., Danielsson, L.G., Muntau, H., Van Leeuwen, H.P., Łobiński, R., 2000. Guidelines for terms related to chemical speciation and fractionation of elements. Definitions, structural aspects, and methodological approaches (IUPAC recommendations 2000). Pure and Applied Chemistry 72, 1453-1470.

Tessier, A., Couillard, Y., Campbell, P.G.C., Auclair, J.C., 1993. Modeling Cd partitioning in oxic lake sediments and Cd concentrations in the freshwater bivalve Anodonta grandis. Limnology & Oceanography 38, 1-17.

335

Références bibliographiques

Tétreault, F., Himmelman, J.H., Measures, L., 2000. Impact of castrating trematode, Neophasis sp., on the common whelk, Buccinum undatum, in the northern Gulf of St. Lawrence. Biological Bulletin 198, 261-271.

Tett, P., Gilpin, L., Svendsen, H., Erlandsson, C.P., Larsson, U., Kratzer, S., Fouilland, E., Janzen, C., Lee, J.Y., Grenz, C., Newton, A., Ferreira, J.G., Fernandes, T., Scory, S., 2003. Eutrophication and some European waters of restricted exchange. Continental Shelf Research 23, 1635-1671.

Tiscar, P.G., Mosca, F., 2004. Defense mechanisms in farmed marine molluscs. Veterinary Research Communications 28, 57-62.

Thieltges, D.W., Reise, K., 2006. Metazoan parasites in intertidal cockles Cerastoderma edule from the northern Wadden Sea. Journal of Sea Research 56, 284-293.

Toba, M., 2004. The decline of Manila clam stock in Tokyo Bay. Bulletin of the Fisheries Research Agency 1, 13-18.

Torreilles, J., Guerin, M.C., Roch, P., 1996. Reactive oxygen species and defense mechanisms in marine bivalves. Espèces oxygénées réactives et systèmes de défense des bivalves marins 319, 209-218. Comptes Rendus de l'Academie des Sciences - Serie III 319 (3), 209-218

Tschischka, K., Abele, D., Portner, H.O., 2000. Mitochondrial oxyconformity and cold adaptation in the polychaete Nereis pelagica and the bivalve Arctica islandica from the baltic and white seas. Journal of Experimental Biology 203, 3355-3368.

Tseng, C.M., Amouroux, D., Abril, G., Tessier, E., Etcheber, H., Donard, O.F.X., 2001. Speciation of mercury in a fluid mud profile of a highly turbid macrotidal estuary (Gironde, France). Environmental Science and Technology 35, 2627-2633.

Tully, O., Nolan, D.T., 2002. A review of the population biology and host-parasite interactions of the sea louse Lepeophtheirus salmonis (Copepoda: Caligidae). Parasitology 124 (Suppl.), 165-182.

V Vallee, B.L., 1995. The function of metallothionein. Neurochemistry International 27, 23-33.

Valko, M., Morris, H., Cronin, M.T.D., 2005. Metals, toxicity and oxidative stress. Current Medicinal Chemistry 12, 1161-1208.

Valtonen, E.T., Holmes, J.C., Koskivaara, M., 1997. Eutrophication, pollution, and fragmentation: effects on parasite communities in roach (Rutilus rutilus) and perch (Perca fluviatilis) in four lakes in central Finland. Canadian Journal of Fisheries and Aquatic Sciences 54, 572-585.

Van Der Knaap, W.P.W., Loker, E.S., 1990. Immune mechanisms in trematode-snail interactions. Parasitology Today 6, 175-182.

Van Der Oost, R., Beyer, J., Vermeulen, N.P.E., 2003. Fish bioaccumulation and biomarkers in environmental risk assessment: A review. Environmental Toxicology and Pharmacology 13, 57- 149.

Vasák, M., 1991. Metal removal and substitution in vertebrate and invertebrate metallothioneins. Methods in Enzymology 205, 452-458.

336

Références bibliographiques

Vasseur, P., Cossu-Leguille, C., 2003. Biomarkers and community indices as complementary tools for environmental safety. Environment International 28, 711-717.

Veltman, K., Huijbregts, M.A.J., Vijver, M.G., Peijnenburg, W.J.G.M., Hobbelen, P.H.F., Koolhaas, J.E., van Gestel, C.A.M., van Vliet, P.C.J., Jan Hendriks, A., 2007. Metal accumulation in the earthworm Lumbricus rubellus. Model predictions compared to field data. Environmental Pollution 146, 428-436.

Veltman, K., Huijbregts, M.A.J., Wang, W.-X., Hendriks, A.J., 2008. Metal bioaccumulation in aquatic species: quantification of uptake and elimination rate constants using physiochemical properties of metals and physiological characteristics of species. Environmental Science and Technology 42, 852-858.

Verlecar, X.N., Pereira, N., Desai, S.R., Jena, K.B., Snigdha, 2006. Marine pollution detection through biomarkers in marine bivalves. Current Science 91, 1153-1157.

Verlecar, X.N., Jena, K.B., Chainy, G.B.N., 2008. Modulation of antioxidant defences in digestive gland of Perna viridis (L.), on mercury exposures. Chemosphere 71, 1977-1985.

Viarengo, A., Burlando, B., Ceratto, N., Panfoli, I., 2000. Antioxidant role of metallothioneins: a comparative overview. Cellular and Molecular Biology 46, 407-417.

Viarengo, A., Nott, J.A., 1993. Mechanisms of heavy metal cation homeostasis in marine invertebrates. Comparative Biochemistry and Physiology 104(C), 355-372.

Viarengo, A., Pertica, M., Canesi, L., Mazzucotelli, A., Orunesu, M., Bouquegneau, J.M., 1989. Purification and biochemical characterization of a lysosomal copper-rich thionein-like protein involved in metal detoxification in the digestive gland of mussels. Comparative Biochemistry and Physiology 93(C), 389-395.

Viarengo, A., Zanicchi, G., Moore, M.N., Orunesu, M., 1981. Accumulation and detoxication of copper by the mussel Mytilus galloprovincialis Lam.: A study of the subcellular distribution in the digestive gland cells. Aquatic Toxicology 1, 147-157.

Villalba, A., Carballal, M.J., López, C., 2001. Disseminated neoplasia and large foci indicating heavy haemocytic infiltration in cockles Cerastoderma edule from Galicia (NW Spain). Diseases of Aquatic Organisms 46, 213-216.

Villalba, A., Reece, K.S., Ordás, M.C., Casas, S.M., Figueras, A., 2004. Perkinsosis in molluscs: A review. Aquatic Living Resource 17, 411-432.

Volety, A.K., 2008. Effects of salinity, heavy metals and pesticides on health and physiology of oysters in the Caloosahatchee Estuary, Florida. Ecotoxicology 17, 579-590.

W Wagner, A., 2002. Asymetric functional divergence of duplicate genes in yeast. Molecular Biology and Evolution 19, 1760-1768.

Walne, P.R., Mann, R., 1975. Growth and biogeochemical composition in Ostrea edulis and Crassostrea gigas, in: Barnes, H. (Ed.), 9th European Marine Biology Symposium. Aberdeen University Press, 587-607.

Wang, L., Song, L., Ni, D., Zhang, H., Liu, W., 2008. Alteration of metallothionein mRNA in bay scallop Argopecten irradians under cadmium exposure and bacteria challenge. Comparative Biochemistry and Physiology 149(C) (1), 50-57.

337

Références bibliographiques

Wang, S.H., Chen, J.H., Lin, L.Y., 1994. Functional integrity of metallothionein genes in testicular cell lines. Journal of Cellular Biochemistry 55, 486-495.

Wang, Y., Fang, J., Leonard, S.S., Rao, K.M.K., 2004. Cadmium inhibits the electron transfer chain and induces reactive oxygen species. Free Radical Biology and Medecine 36, 1434-1443.

Wang, W.X., Fisher, N.S., 1999. Assimilation efficiencies of chemical contaminants in aquatic invertebrates: A synthesis. Environmental Toxicology and Chemistry 18, 2034-2045.

Wang, W.X., Fisher, N.S., Luoma, S.N., 1996. Kinetic determinations of trace element bioaccumulation in the mussel Mytilus edulis. Marine Ecology Progress Series 140, 91-113.

Wang, W.X., Rainbow, P.S., 2005. Influence of metal exposure history on trace metal uptake and accumulation by marine invertebrates. Ecotoxicology and Environmental Safety 61, 145-159.

Wang, W.X., Rainbow, P.S., 2006. Subcellular partitioning and the prediction of cadmium toxicity to aquatic organisms. Environmental Chemistry 3, 395-399.

Wang, W.X., Rainbow, P.S., 2008. Comparative approaches to understand metal bioaccumulation in aquatic animals. Comparative Biochemistry and Physiology 148(C), 315-323.

Wegeberg, A.M., 1998. Digene trematoders (Echinostomatidae) infektionsøkologi og effekt pa Cerastoderma edule. University of Aarhus.

Wegeberg, A.M., de Montaudouin, X. & Jensen, K.T., 1999. Effect of intermediate host size (Cerastoderma edule) on infectivity of cercariae of three Himasthla species (Echinostomatidae, Trematoda). Journal of Experimental Marine Biology and Ecology 238, 259-269.

Wegeberg, A.M., Jensen, K.T., 2003. In situ growth of juvenile cockles, Cerastoderma edule, experimentally infected with larval trematodes (Himasthla interrupta). Journal of Sea Research 50, 37-43.

Werding, B., 1969. Morphologie, Entwicklung und Ökologie digener Trematoden-Larven der Strandschnecke Littorina littorea. Mar. Biol. 3: 306-333.

Wotton, E.C., Dyrynda, E.A., Ratcliffe, N.A., 2003. Bivalve immunity: comparisons between the marine mussel (Mytilus edulis), the edible cockle (Cerastoderma edule) and the razor-shell (Ensis siliqua). Fish and Shellfish Immunology 15, 195-210.

X Xue, Q., Renault, T., 2000. Enzymatic activities in European flat oyster, Ostrea edulis, and pacific oyster, Crassostrea gigas, hemolymphe. Journal of Invertebrate Pathology 76, 155-163.

Y Yan, T., Teo, L.H., Sin, Y.M., 1997. Effects of mercury and lead on tissue glutathione of the green mussel, Perna viridis L. Bulletin of Environmental Contamination and Toxicology 58, 845-850.

Yu, J.H., Song, J.H., Choi, M.C., Park, S.W., 2009. Effects of water temperature change on immune function in surf clams, Mactra veneriformis (Bivalvia: Mactridae). Journal of Invertebrate Pathology 102, 30-35.

Z

338

Références bibliographiques

Zasloff, M., 2002. Innate immunity, antimicrobial peptides, and protection of the oral cavity. Lancet 360, 1116-1117.

Zhang, B., Georgiev, O., Hagmann, M., Günes, Ã., Cramer, M., Faller, P., Vasák, M., Schaffner, W., 2003. Activity of Metal-Responsive Transcription Factor 1 by Toxic Heavy Metals and H2O2 in Vitro Is Modulated by Metallothionein. Molecular and Cellular Biology 23, 8471-8485.

Zorita, I., Bilbao, E., Schad, A., Cancio, I., Soto, M., Cajaraville, M.P., 2007. Tissue- and cell-specific expression of metallothionein genes in cadmium- and copper-exposed mussels analyzed by in situ hybridization and RT-PCR. Toxicology and Applied Pharmacology 220, 186-196.

339

Références bibliographiques