Contrôle et modulation de la réponse immunitaire par Brucella abortus

AIX-MARSEILLE UNIVERSITE Faculté des sciences de Luminy Ecole Doctorale des Sciences de la Vie et de la Santé THESE DE DOCTORAT Biologie - Spécialité : Immunologie En vue d'obtenir le titre de DOCTEUR DE L UNIVERSITÉ D AIX MARSEILLE Présentée et soutenue publiquement par: Clara DEGOS 27 Novembre 2014 Contrôle et modulation de la réponse immunitaire par Brucella abortus Directeur de thèse : Dr Jean-Pierre Gorvel Thèse soutenue devant le jury composé de: Prof Franck GALLAND Président Dr David O CALLAGHAN Rapporteur Prof Jean-Jacques LETESSON Rapporteur Dr Jean-Pierre GORVEL Directeur de thèse Prof Jean-Louis MEGE Examinateur Le travail réalisé dans cette thèse a été effectué au Centre d Immunologie de Marseille- Luminy, UM 2 Aix Marseille Université, UMR_S 1104 CNRS, UMR 7280 Inserm

SOMMAIRE REMERCIEMENTS 5 ABBREVIATIONS 7 TABLE DES FIGURES ET TABLEAUX 10 I. INTRODUCTION 12 I. A. LA BRUCELLOSE 13 I. A. 1. ORIGINE : BRUCELLA 13 I. A. 2. REPARTITION SUR LE GLOBE 13 I. A. 3. CONTAMINATION ET SYMPTOMES 13 I. A. 4. CONSEQUENCES ET PROBLEMES VACCINAUX 14 I. B. LES FACTEURS DE VIRULENCE 16 I. B. 1. LE LIPOPOLYSACCHARIDE (LPS) 16 I. B. 2. LE GLUCANE CYCLIQUE Β 1,2 (CΒG) 17 I. B. 3. LE SYSTEME DE SECRETION DE TYPE IV : VIRB 18 I. B. 4. LES PROTEINES DE MEMBRANE EXTERNE (OMP) 19 I. C. LA VIE INTRACELLULAIRE 21 I. C. 1. ENTREE DANS LES CELLULES 21 I. C. 2. TRAFIC INTRACELLULAIRE : LA BCV SUR LES TRACES DES ENDOSOMES 22 I. C. 3. LE RETICULUM ENDOPLASMIQUE, UN HAVRE DE PAIX 23 I. D. BRUCELLA ET LE SYSTEME IMMUNITAIRE 25 I. D. 1. UNE STRATEGIE D EVITEMENT : REPONSES AUX TLR 25 I. D. 2. LA RESISTANCE AUX DEFENSES INNEES 27 I. D. 3. BRUCELLA ET LES DC 31 I. D. 4. L IMMUNITE ADAPTATIVE CONTRE BRUCELLA 33 I. E. CD150, UN RECEPTEUR A LA SURFACE DES CELLULES IMMUNITAIRES 37 I. E. 1. CD150, UNE MOLECULE HOMOPHYLIQUE DE CO-STIMULATION 37 2

I. E. 2. LES PROPRIETES IMMUNOMODULATRICES DE CD150 38 I. E. 3. CD150 ET LES INFECTIONS 40 I. E. 4. CD150, UN RECEPTEUR BACTERIEN? 42 II. RESULTATS 43 II. A. RESUME DES ACTIVITES 44 II. B. OMP25 SE LIE A CD150 POUR CONTROLER L ACTIVATION DES DC DURANT L INFECTION PAR BRUCELLA 45 II. B. 1. INTRODUCTION 45 II. B. 2. RESULTATS ARTICLE EN PREPARATION 46 II. C. LE CΒG DE BRUCELLA ACTIVE LES DC ET CONTROLE LE RECRUTEMENT DES NEUTROPHILES 67 II. C. 1. INTRODUCTION 67 II. C. 2. MANUSCRIT SOUMIS 68 II. D. BTPB, UNE PROTEINE CAPABLE DE MODULER L ACTIVATION DES DC 92 II. D. 1. INTRODUCTION 92 II. D. 2. ARTICLE 93 III. DISCUSSION ET CONCLUSION GENERALE 107 IV. MATERIEL ET METHODES 113 IV. A. MATERIEL VIVANT 114 IV. B. REACTIFS 115 IV. C. BACTÉRIOLOGIE 119 IV. D. BIOLOGIE CELLULAIRE 122 IV. E. BIOLOGIE MOLECULAIRE 123 IV. F. BIOCHIMIE 127 3

V. REFERENCES 129 VI. ANNEXE 141 VI. A. ARTICLE : LIPOPOLYSACCHARIDES WITH ACYLATION DEFECTS POTENTIATE TLR4 SIGNALING AND SHAPE T CELL RESPONSES. 142 4

Remerciements La meilleure partie d une thèse selon une autre doctorante Je tenais donc à remercier tout d abord mon directeur de thèse, Jean-Pierre, pour m avoir permis de réaliser ma thèse dans le laboratoire. Aussi merci pour m avoir laissée une certaine liberté que je n aurais pas eu dans d autres laboratoires. Il était aussi très rassurant qu à chaque fois que j entrais dans le bureau, dépitée par une énième expérience ratée, vous me disiez : «Non mais je suis sûr que tu vas réussir!», et au final j ai réussi à le faire ce blot (et à reproduire 3 fois mes manips! I did it!) Et merci de m avoir aidée pour la suite, j apprécie énormément. Un grand merci à Stéphane pour m avoir tout appris ou presque en biochimie et biologie moléculaire. J ai bien compris que tu aimais ça, j ai un peu plus de mal, cela dit je vois que ça peut quand même être super cool un blot réussi. Et tu sais que je n aurais sans doute pas réussi les clonages, et prouver que slam interagit avec Omp25 sans toi! Merci aussi pour toutes les discussions sur tout et rien qu on a eu, ça faisait toujours du bien. Suzana, tu sais à quel point je t adore, et je dois te dire un grand merci pour tant de choses. Merci pour m avoir prise en M2, m avoir fait découvrir Brucella (ou BruBru de son petit nom), m avoir initiée à la recherche et m avoir transmis ta vision de la science. Merci aussi pour un tas d autres choses (le confocal le 31 décembre, les post it dans le labo, les coups de fil/mails qui remontent le moral, et plein de choses). Tu es ma chercheuse préférée en tout point tu sais :) Merci l Equipe JPG (comme dirait Alexia) pour l ambiance, les croissants du vendredi, les pots en tout genre (merci à Chantal et son fantastique gâteau au citron (le meilleur de la planète)), c était vraiment cool. Et puis on a quand même le bureau le plus cool du CIML : Papa Hugues/Huggy/Oggy, Johnny (vive la nourriture gratuite et non on n aime pas les gens), Clément, Aurélie, Raja et Alexia. Merci à tout plein de personnes aux CIML : à Olivier mon parrain de thèse, Sylvain, Marc et Atika de la cytométrie parce que vous avez toujours souri à mes blagues, et rien que pour ça, merci (et bravo. Je connais le niveau quand même). Merci à Lionel pour sa bonne humeur. Merci à Lydia (toujours vivante?!), Toufik pour les blagues (je m en suis toujours pas remise, meilleure blague du monde tu sais), Djélani pour m avoir chouchouté niveau ordi et écran, même si tu aimes Internet Explorer, je sais qu un jour tu abandonneras IE, j ai foi en toi. Merci à Carole, Franck et Fred, pour notre semaine de cours qui chaque année me permettait de respirer un peu. Merci aux autres thésards pour les discussions pas scientifiques mais tellement mieux : Yannick, Clément C, Clément G, Yaya (you speak french too well you know). Merci à Mohammed mon copain du samedi/dimanche (au choix), à Voa et tous les gens du midi. 5

Merci à mes copines du P3/bureau/team BruBru (yes!) : Alexia et Aurélie (dans l ordre alphabétique, je ne veux froisser personne hein!). Vous allez vraiment trop me manquer les filles. Bon je suis quand même heureuse de partir, maintenant je pourrais chanter quand/comme je veux et Alexia ne me menacera pas de couper la radio (super virulente (comme BruBru) la fille), et Aurélie ne fera pas de blagues plus que douteuses sur une phrase tout à fait innocente. Merci pour les fous rires, pour les coups de mains dans les manips, pour les pauses café, les quizz, danses, et autre choses bizarres (hein Aurélie) au P3 pour passer le temps. Merci à Aurélie pour ses dessins qui ornent à peu près tout (de l agrafeuse aux post-it oui oui), pour avoir un petit côté nerveux (je me sens vachement moins seule du coup) et pour être aussi petite que moi, ouf! Merci à Alexia d avoir su me rassurer/consoler quand j en avais besoin (et d être tellement plus calme et maligne que moi sur plein de choses). Je vous aime. Merci à ceux qui sont partis trop tôt (je ne veux pas dire morts voyons! juste partis vers d autres labos!) : Sandra tu m as manquée, Philippe (aka le Plombier), Irène, tes blagues (volontaires ou pas), ton rire, Aude-Agnès (heureusement que tu es revenue via Ciphe), Samira, Caroline, et tous les gens qui vont soutenir juste avant moi et ne seront pas présents (merci les gars du soutien!). Merci à Amélie d être la fille la plus coool du monde, de ne pas m en vouloir quand j oublie de te répondre, d avoir posé un jour exprès pour ma thèse, d être toujours prête pour un thé/tisane, et d être mon amie et ce depuis un bon nombre d années (ça nous rajeunit pas cette histoire!). Et puis merci de m avoir sauvé pour la bio mol. Merci à ma famille de penser que je suis hyper douée et super brillante, parce que je connais plein de mots bizarres terminant par -cytes, ça fait toujours du bien à l égo. Bon et surtout merci du soutien. Et le meilleur à la fin (et forcément dans cet ordre-là ) : Rémi, merci, merci, merci et merci, pour tout. Merci de m avoir accompagné certaines fois le week end au labo, ou m avoir ramené parce, oups, j ai oublié d éteindre/ranger/blabla quelque chose. Merci de m avoir rassurée et calmée quand j en avais besoin. Et merci parce que sans toi je serais devenue plus folle que je ne le suis déjà, parce que ça fait toujours du bien de te retrouver le soir, et parce que je t aime (bien plus que BruBru!). Pour résumer ma thèse Il n y a pas de réussite facile ni d échecs définitifs. Marcel Proust (merci les papillotes et Johnny!) 6

ABBREVIATIONS AP-1 : Activator protein 1 B. abortus : Brucella abortus B. melitensis : Brucella melitensis B. suis : Brucella suis BAL : Lavages bronchoalvérolaires BCR : B cell receptor BCV : Brucella containing-vacuole BLS : Brucella lumazine synthase Brucella wt OM : Extraits de membrane de Brucella sauvage Brucella omp25 OM : Extraits de membrane de Brucella mutant pour Omp25 Bsp : Brucella secreted protein Btp : Brucella tir protein CβG : Glucane cyclique β 1,2 CDC : Center of disease control CLR : Receptors lectines de type C CMH : Complexe majeur d histocompatibilité COX : Cyclooxygénase CPA : Cellule présentatrice d antigène CTL : LT cytotoxiques DC : Cellules dendritiques E. coli : Escherichia coli EAT2 : EWS-Fli1-activated transcript-2 ERES : RE exit sites HSC : Hematopoïetic stem cell Ig : Immunoglobuline IFN-γ : Interferon-γ IL-1β : Interleukine-1β IL-2 : Interleukine-2 IL-4 : Interleukine-4 IL-6 : Interleukine-6 IL-8 : Interleukine-8 7

IL-10 : Interleukine-10 IL-12 : Interleukine-12 IRAK : Interleukin-1 receptor-associated kinase IRF : Interferon regulatory factor IRF-1 : Interferon regulatory factor 1 IRF-8 : Interferon regulatory factor 8 ITAM : Immunoreceptor tyrosine-based activation motifs ITSM : Immunoreceptor tyrosine-based switch motifs LAMP-1 : Lysosomal-associated membrane protein 1 LB : Lymphocytes B LBP : LPS binding proteins LPS : Lipopolysaccharide LT : Lymphocytes T mabs : Anticorps monoclonaux ME : Membrane externe MI : Membrane interne MMP : Métallo-matrix protéases NF-AT : Nuclear factor of activated T-cells NF-κB : Nuclear factor-kappa B NK : Natural Killer NLR : Nod-like receptors NO : Oxyde nitrique NOS : Synthase de l oxyde nitrique NOX2 : NADPH Oxydase 2 OMV : Outer membrane vesicles PAMP : Pathogen associated molecule pattern PrpA : Proline racemase protein A PRR : Pattern recognition receptor PtdIns(3)P : Phosphatidylinositol-3-phosphate RE : Réticulum endoplasmique ROS : Réactifs oxygénés SAP : SLAM-associated protein Sbi : Ig -binding protéine A staphylococcale SHP-2 : SH2 domain-containing protein 8

SLAM : Signaling lymphocyte activation molecule SPA : Protéine staphylococale A SR-A : Scavenger receptor A S19 : Souche 19 TCR : T cell receptor Th : LT helper TIR : Toll/IL-1 receptor TLR : Toll-like receptor TNF-α : Tumor necrosis factor alpha Treg : LT régulateurs T4SS : Système de sécrétion de type IV UPR : Unfolded protein response 9

TABLE DES FIGURES ET TABLEAUX p.13 / Fig. 1 : Phylogénie des différentes espèces de Brucella et leurs hôtes naturels. p.14 / Fig. 2 : Répartition mondiale des cas de brucellose en l an 2000. p.15 / Fig. 3 : Modes de contamination humain et animal. p.16 / Fig. 4 : Structure du LPS de Brucella. p.18 / Fig. 5 : Structure schématique d un système de sécrétion de type IV. p.19 / Fig. 6 : Structure de la membrane de Brucella. p.22 / Fig. 7 : Schématisation du site d entrée de Brucella et des protéines pouvant y participer. p.23 / Fig. 8 : Trafic intracellulaire de la BCV et les protéines eucaryotes ou bactériennes requises. p.25 / Fig. 9 : Les récepteurs TLR et leurs ligands. p.26 / Fig. 10 : Voies de signalisation en aval des TLR. p.27 / Fig. 11 : Différents mécanismes de détection de Brucella et les voies de signalisation en aval des récepteurs. p.31 / Fig. 12 : Différentes étapes de maturation des DC. p.33 / Fig. 13 : Différentes populations de LT CD4 + et leurs réponses face aux infections. p.37 / Fig. 14 : Récepteurs murins de la famille SLAMF. p.48 / Fig. 15 : CD150 expression onto BMDC following stimulation with Brucella membrane extracts. p.48 / Fig. 16 : CD25 expression onto T CD4 + cells stimulated by differently activated BMDC. p.49 / Fig. 17 : Proliferation of T CD4 + cells stimulated by differently activated BMDC. p.50 / Fig. 18 : Brucella omp25 replicates as the wt strain within the ER in BMDC. p.51 / Fig. 19 : Co-stimulatory molecules and MHC-II expression onto BMDC after Brucella infection. p.52 / Fig. 20 : NF-κB translocation within infected BMDC. p.53 / Fig. 21 : mrna expression of different pro-inflammatory genes in infected BMDC. p.54 / Fig. 22 : Pro-inflammatory cytokines secretion upon BMDC infection. p.55 / Fig. 23 : Brucella replication and intracellular localization in BMDC after CD150 blockade. p.56 / Fig. 24 : NF-κB translocation within infected BMDC after CD150 blockade 10

p.57 / Fig. 25 : Pro-inflammatory cytokines secretion upon BMDC infection after CD150 blockade. p.58 / Fig. 26 : NF-κB translocation within infected CD150 KO BMDC. p.59 / Fig. 27 : Bacterial growth and weight organs in wt and omp25 infected mice at 5 days post-infection. p.60 / Fig. 28 : Competitive index proliferation between Brucella wt and omp25 mutant strains. p.60 / Fig. 29 : Bacterial growth and weight organs in wt and omp25 infected mice at 60 days post-infection. p.61 / Fig. 30 : Survival curve of IFN-γ KO mice infected with Brucella wt or omp25. p.61 / Fig. 31 : Bacterial growth and weight organs in wt and omp25 infected CD150 KO mice at 8 days post-infection. p.62 / Fig. 32 : Competitive index in CD150 KO mice. p.62 / Fig. 33 : Brucella Omp25 binds CD150. p.110 / Fig. 34 : Réplication de Brucella dans les souris CCR2 KO. p.115 / Tableau 1 : Réactifs utilisés p.116 / Tableau 2 : Anticorps p.118 / Tableau 3 : Plasmides p.119 / Tableau 4 : Souches bactériennes utilisées p.124 / Tableau 5 : Amorces 11

I. Introduction 12

Figure 1 : Phylogénie des différentes espèces de Brucella et leurs hôtes naturels. Le phylum de Brucella compte dix espèces différentes : abortus, canis, ceti, inopinata, melintensis, microti, neotomae, ovis, pinnipidialis, suis. Différents isolats peuvent avoir différents hôtes et tropismes. Ainsi certaines souches de B. ceti infectent les mammifères marins tandis qu une autre est propre à l homme. Adapté de [3].

I. A. LA BRUCELLOSE I. A. 1. Origine : Brucella La brucellose ou «Fièvre de Malte» est une zoonose, maladie transmissible de l animal à l homme. David Bruce a identifié en 1887 le pathogène responsable de cette infection, Brucella. Cette bactérie semble être un pathogène très ancien et la brucellose pourrait sévir depuis des millions d années [1, 2]. Brucella est une bactérie pathogène à gram négatif appartenant au groupe α-2 des protéobactéries. Ces bactéries sont décrites comme étant des bactéries intracellulaires facultatives. De nombreuses espèces appartiennent au genre Brucella et ont un tropisme particulier. A ce jour, nous comptons près de 10 espèces différentes (Fig. 1) [3]. Parmi ces espèces Brucella melitensis (B. melitensis), Brucella abortus (B. abortus) et Brucella suis (B. suis) sont celles qui sont les plus pathogéniques pour l homme [4]. I. A. 2. Répartition sur le globe Avec plus de 500 000 nouveaux cas d infection humaine par an, Brucella est l un des agents zoonotiques parmi les plus virulents et dont la répartition est très représentée au niveau du globe terrestre. La brucellose est endémique dans de nombreux pays, en particulier dans les pays du bassin Méditerranéen, de l Amérique Latine et du Moyen Orient (Fig. 2) [4, 5]. De nouveaux foyers apparaissent ou ré-émergent chaque année comme par exemple dans le nord de la Chine ou la Mongolie. En France, les Alpes du Sud restent une zone où B. melitensis est endémique chez les animaux sauvages comme les bouquetins [6, 7]. Récemment, Brucella a été classée par l OMS dans le top 7 des zoonoses négligées, responsables à la fois d un problème de santé humain et économiques à cause de l impact négatif que la maladie cause aux animaux d élevage [8]. I. A. 3. Contamination et symptômes Les principaux hôtes naturels de Brucella sont les bovins, ovins, caprins ou encore les mammifères marins. L homme est un hôte secondaire. La contamination se fait par 13

Figure 2 : Répartition mondiale des cas de brucellose en l an 2000. Répartition des cas déclarés de brucellose humaine en l an 2000. Les pays les plus touchés sont la Mongolie et les pays du Moyen-Orient, et des Balkans. Certains pays d Amérique Centrale et du Sud ainsi que les pays du bassin Méditerranéen comptent aussi de nombreux cas de brucellose humaine. Adapté de [4].

consommation de produits laitiers contaminés, par inhalation de poussières ou d aérosols contaminés, ou encore par contact direct avec des animaux infectés (Fig. 3) [9]. Du fait de la contamination possible par aérosol, Brucella est considérée et listée par le Center of Disease Control (CDC) comme un agent du bioterrorisme de la liste B, les agents de seconde priorité. B. suis fut même la première arme biologique développée par les Etats-Unis dans les années 1950-1960 avant que le programme ne soit abandonné en 1969. Brucella pénètre l organisme par les voies aériennes et la voie orale, elle peut aussi pénétrer par les lésions cutanées et les muqueuses. Chez les animaux, la bactérie va cibler le tractus génital et provoquer des avortements chez les femelles et des infertilités chez les mâles [10]. L infection chez l homme conduit au bout de deux à quatre semaines à une infection aigüe caractérisée par une fièvre ondulante et une asthénie générale. La bactérie se dissémine et peut toucher différents organes. Chez 30 % des patients, cette phase évolue en maladie chronique. Les foyers infectieux sont les os et articulations, le foie, et parfois le cœur ainsi que le système nerveux, ces deux derniers cas provoquent des endocardites et neuro-brucelloses qui peuvent être létales [11]. La transmission d homme à homme étant très rare [12], les contaminations humaines sont très étroitement liées à la présence d un réservoir animal infecté. Le contrôle de la présence de la bactérie dans ces réservoirs est donc critique pour la lutte contre la dissémination de la maladie. I. A. 4. Conséquences et problèmes vaccinaux Les conséquences des infections à Brucella sont premièrement d ordre économique. Des campagnes de vaccinations concernant les animaux domestiques et d élevage ont été déployées pour éradiquer la bactérie dans plusieurs pays comme la France, grâce auxquelles les contaminations humaines sont passées de 405 cas en 1983 à 44 cas en l an 2000 [4]. Parmi les souches utilisées comme vaccins vivants, nous citerons la souche 19 (S19) de B. abortus. Cette souche est éliminée plus rapidement que les souches virulentes chez les bovins vaccinés [13]. De plus, elle induit une immunité protectrice. La souche avec un LPS naturellement rugueux RB51 est aussi utilisée comme vaccin vivant chez les bovins. La 14

Figure 3 : Modes de contamination humain et animal. La contamination des hôtes naturels se produit pendant des avortements, par l allaitement, ou encore par contact génital pendant la reproduction. Les contaminations humaines sont dues à la consommation de produits contaminés (lait, fromage), au contact direct avec des animaux infectés, ou l inhalation de poussière. La plupart des cas concernent des fermiers, vétérinaires, ou des personnes travaillant en laboratoire. Adapté de [9].

protection induite est inférieure à celle induite par S19 [14]. Bien que cette souche soit moins virulente que S19, il semble qu elle provoque des avortements chez les animaux gravides [15]. De plus, cette souche ayant été obtenue après plusieurs passages sur milieu contenant de la rifampicine et de la pénicilline, elle est donc résistance à ces antibiotiques, ce qui pose un problème de traitement. Une des autres souches actuellement utilisée chez les moutons et chèvres est B. melitensis Rev1. Cette souche porte une mutation dans le gène rpsl codant pour la protéine ribosomale S12, qui confère la résistance à la streptomycine. A ce jour Rev1 est la souche la plus efficace contre la brucellose ovine et caprine. Elle confère en effet 80 à 100 % d efficacité [16]. S19 et Rev1 sont des bactéries ayant un phénotype lisse (ou «smooth» en anglais, S). Cette particularité vient de la composition de leur lipopolysaccharide (LPS) (voir partie I. B. 1). Les vaccins utilisés actuellement chez les animaux, comme Rev1, S19 et RB51 ne sont cependant pas efficaces à 100 % et ne protègent pas contre toutes les espèces de Brucella. Ils induisent en outre des effets secondaires (avortements). Une autre conséquence est l obligation d avoir une campagne de santé publique adéquate en cas de situation endémique. La prophylaxie actuelle pour éradiquer la bactérie est, chez l homme, une combinaison de plusieurs antibiotiques : doxycycline et rifampicine ou doxycycline et streptomycine pendant plusieurs semaines [17, 18]. L élaboration de nouveaux vaccins efficaces chez l homme notamment, reste donc toujours un problème majeur dans la lutte contre la brucellose. La pathogénicité de Brucella est liée à sa capacité à exprimer divers facteurs de virulence agissant à la fois sur les étapes de la vie extracellulaire et intracellulaire de la bactérie. Dans un premier temps, nous allons nous intéresser à décrire les principaux facteurs de virulence importants pour la bactérie, puis nous décrirons sa vie intracellulaire. 15

Figure 4 : Structure du LPS de Brucella. Ancré dans la membrane externe de la bactérie, le LPS de Brucella est composé d un lipide A relié au core lui-même relié à une chaîne O-polysaccharide (ou O-antigène). Le lipide A est composé de diaminoglucoses reliés à des chaînes de 18 à 28 carbones. Les LPS ne possédant pas de chaîne O-polysaccharide sont dits rugueux. Adapté de [19].

I. B. LES FACTEURS DE VIRULENCE I. B. 1. Le lipopolysaccharide (LPS) Composant majeur de la membrane externe des bactéries à gram négatif, le LPS est un motif associé au pathogène ou PAMP (Pathogen Associated Molecule Pattern) reconnu par des récepteurs immunitaires à la surface des cellules immunitaires tels que les Toll-Like Receptor (TLR) et plus particulièrement TLR4 (voir partie I. D. 1). Le LPS est composé du lipide A hydrophobe inséré dans la membrane bactérienne. Le lipide A est relié à un core polysaccharide, lui-même généralement relié à une chaîne oligosaccharide (chaîne O). On parle alors de LPS lisse. Si le core n est pas relié à la chaîne O, le LPS est défini comme étant rugueux (ou «rough» en anglais, R) (Fig. 4) [19]. Certaines espèces de Brucella ont naturellement un LPS rugueux comme B. ovis ou B. canis, les autres possèdent un LPS lisse. Le phénotype rugueux est quant à lui associé à une élimination des bactéries durant l infection chez la plupart des cas pour ces espèces [20]. Le LPS de Brucella est un LPS non canonique en raison de plusieurs changements au niveau de sa structure comparé à des LPS canoniques plus classiques comme celui d Escherichia coli (E. coli). En effet, le lipide A de Brucella contient un squelette de diaminoglucose alors que les LPS canoniques sont composés de glucosamine, et est relié à des groupes acyles plus longs que pour les LPS classiques (18, 19 ou 28 carbones contre 12 ou 14). Les liaisons au core sont aussi différentes. Dans le cas de Brucella, ce sont des liaisons amines exclusivement, ce qui le différencie des autres LPS d entérobactéries qui présentent des liaisons esters. Le LPS de Brucella a un rôle très important dans la pathogénicité en jouant à la fois sur la vie intracellulaire et le trafic de la bactérie. Mais il a aussi un rôle prépondérant dans la réponse du système immunitaire. Nous reviendrons sur cet aspect dans la partie I. D. 16

I. B. 2. Le glucane cyclique β 1,2 (CβG) Les glucanes cycliques sont des composés de l enveloppe de bactéries à gram négatif. Ils ont des rôles divers selon les organismes : par exemple, la stabilité de la membrane des bactéries, la motilité, la synthèse des exopolysaccharides. Composé de 17 à 25 glucoses reliés par des liaisons en β (1,2), le CβG de Brucella est trouvé en forte concentration (1 à 10 mm) dans le périplasme et constitue 1 à 5 % du poids sec de la bactérie. Il a aussi été montré comme crucial pour la survie intracellulaire des bactéries en contrôlant la maturation de la vacuole dans laquelle elle se trouve pour éviter qu elle ne fusionne avec les lysosomes. De plus, en interagissant avec les radeaux lipides et le cholestérol, le CβG a un rôle dans les premières étapes de la vie intracellulaire de la bactérie dans les cellules infectées [21]. Récemment, une étude menée au laboratoire a montré que ce composé avait des propriétés activatrices des cellules dendritiques (DC) murines et humaines. En effet, après stimulation avec le CβG, les DC acquièrent un phénotype mature, caractérisé par une surexpression des molécules de co-stimulation et des molécules du complexe majeur d histocompatibilité de classe II (CMH-II) à la membrane, par une production de cytokines pro-inflammatoire et une capacité à activer les lymphocytes T (LT). Cette activation est un mécanisme dépendant de TLR-4, mais ne dépend pas de CD14, molécule co-réceptrice de TLR-4 [22]. Contrairement aux macrophages, un mutant pour le CβG (cgs-) n a pas de défaut de réplication dans les DC [21, 23]. De plus, cette molécule est non toxique et non immunogénique, ce qui lui vaut d être considérée comme un nouvel adjuvant potentiel [22]. Ce facteur de virulence est donc crucial à la fois pour établir une infection, une vie intracellulaire au sein des cellules hôtes mais aussi pour déclencher une réponse immunitaire. Les différentes réponses induites, selon les cellules ciblées restent inexpliquées. Aucun récepteur reconnaissant le CβG n a été encore identifié, et le fait que cette molécule soit capable d accéder à la BCV reste encore un mécanisme inconnu. 17

Outer e ra e I er e ra e Figure 5 : Structure schématique d un système de sécrétion de type IV. Différentes protéines composent le T4SS. Dans le cas d Agrobacterium, le T4SS est composé de VirB5 et VirB2 dans le cytosol de la cellule, puis les protéines VirB7, VirB9 et VirB10 composent le cœur du T4SS qui transporte les molécules à transloquer. Les protéines VirB4 et VirB11 (des ATPases) permettent à l appareil de sécrétion de fonctionner grâce notamment à l utilisation de l ATP. VirB6 et VirB8 forment le complexe entre le cytosol de la bactérie et sa membrane interne. Adapté de [29].

I. B. 3. Le système de sécrétion de type IV : VirB Les systèmes de sécrétion ont un rôle essentiel dans la pathogénicité des bactéries. A ce jour, il y a 9 types de système de sécrétion connus [24, 25]. Ces composants sont essentiels à la translocation de protéines au sein de cellules hôtes. Brucella possède un système de sécrétion de type IV (T4SS), tout comme Agrobacterium tumefaciens, Helicobacter pylori, Bordetella pertussis, ou encore Legionella pneumophila [26-29]. Etant donné que la structure T4SS de Brucella n est pas encore complètement caractérisée, les études sont souvent basées sur des analogies avec celui d Agrobacterium. La seringue moléculaire qui traverse la double membrane des bactéries permet la translocation de protéines et d ADN. Elle a donc des fonctions dans la conjugaison, la capture d ADN et la translocation de protéines depuis et vers l environnement externe [30]. L opéron virb est composé de plusieurs gènes : le complexe cytoplasmique-membrane interne est composé de VirB6 et VirB8 qui fonctionne avec les ATPases VirB4, VirB11 [31] pour déclencher les processus de sécrétion. Ce sont les protéines VirB7, VirB9 et VirB10 qui interagissent pour former le cœur du complexe, un canal traversant la double membrane bactérienne. Enfin les protéines VirB2 et VirB5 composent le pillus. D autres protéines, VirB3 et VirB12 semblent avoir un rôle dans l assemblage du pillus (Fig. 5). L expression des gènes de cet opéron est étroitement régulée et dépend notamment de l acidité du milieu. Ainsi, lors d une infection, l expression du T4SS serait maximale après 5 heures d infection et serait réprimée dès lors que les bactéries ont rejoint leur niche réplicative [32, 33]. La régulation par le ph de cet opéron corrèle avec le fait que Brucella réside dans une vacuole (appelée BCV pour Brucella Containing-Vacuole) qui va suivre un processus de maturation au cours du temps. VjbR, un régulateur transcriptionnel liée au quorum sensing, régule l expression de l opéron virb [34]. Le T4SS est essentiel à la survie intracellulaire en permettant la maturation de la BCV [27, 35]. Un mutant virb pour cet appareil de sécrétion ne parvient pas à établir des interactions avec le réticulum endoplasmique (RE), la niche réplicative de Brucella. En absence de T4SS, la bactérie n est pas capable de sécréter les effecteurs permettant la maturation de la BCV. 18

Figure 6 : Structure de la membrane de Brucella. La double enveloppe de la bactérie est composée d une membrane externe (ME) sur laquelle le LPS est ancré. Des protéines Omp (Outer membrane protein) sont enchâssées dans la ME et régulées par le système à deux composants BvrS/BvrR, situé dans la membrane interne (MI). Le périplasme, situé entre les deux membranes, contient différents composants dont le CβG. Le T4SS VirB traverse les deux membranes pour permettre la sécrétion d effecteurs dans le cytosol de la cellule hôte.

Certaines études in vivo montrent que pendant les premiers jours d infection un mutant virb n est pas atténué, mais qu à partir du 5 ème jour sa réplication est deux fois plus faible que celle de la souche sauvage [36]. VirB serait donc requis pour le maintien de la réplication après les premiers temps d infection et la survie de la bactérie. I. B. 4. Les protéines de membrane externe (Omp) Ces protéines/lipoprotéines sont insérées au niveau de la membrane externe bactérienne (Fig. 6). Brucella spp compte 3 groupes d Omp qui sont classées selon leur poids moléculaire : le groupe 1 (43-94 kda), le groupe 2 (36-38 kda and 41-43 kda) et le groupe 3 (25-27 kda and 31-34 kda). Les protéines du premier groupe sont des composantes mineures de la membrane externe de la bactérie. Les protéines du groupe 2 quant à elles seraient pour la plupart des porines [37, 38]. Les Omp du groupe 3 sont présentes en grande quantité dans les extraits provenant des membranes externes et leurs fonctions ne sont pas toutes connues [39]. Dans le groupe 3, les deux premières protéines à avoir été identifiées sont Omp31 et Omp25 (ou Omp3a). Toutefois, Omp31 n est pas exprimée par B. abortus [40, 41]. De nombreuses Omp semblent avoir des fonctions dans la réponse immunitaire. Ainsi, la lipoprotéine Omp19 est connue pour diminuer la présentation antigénique et l expression du CMH-II dans les monocytes humains activés avec de l interferon-γ (IFN-γ) [42]. Omp16 est reconnue par les DC via TLR-4 et induit une réponse immunitaire, notamment via la sécrétion de cytokines pro-inflammatoires telles que le Tumor Necrosis Factor alpha (TNFα) et l interleukine 12 (IL-12) et la surexpression de molécules de co-stimulation comme CD80, CD86 et CD40. Cette réponse des DC va polariser la réponse immunitaire en une réponse de type Th1 [43]. Omp25, dont l expression est contrôlée par le système à deux composants BvrR/BvrS, joue aussi un rôle dans la réponse immunitaire. Une souche mutante pour Omp25 ( omp25) ne présente pas de défaut de réplication dans les cellules épithéliales (HeLa), les macrophages (Raw et THP-1), les DC humaines ou encore les polynucléaires neutrophiles [44, 45]. Il semble que l absence d Omp25 dans des cellules HeLa induise une plus forte association des bactéries avec les cellules [45]. Une autre étude propose que Omp25 et Omp22, une autre 19

protéine du groupe 3, sont essentielles à la survie de B. ovis au sein des cellules hôtes (des cellules HeLa) [46]. La souche omp25 chez B. suis induit une forte sécrétion de TNF-α et l IL-12 dans les DC et macrophages humains infectés [47, 48]. Omp25 semble donc réguler négativement la production de cytokines pro-inflammatoires ainsi que l activation de certaines cellules de l immunité innée pour réguler la réponse immunitaire adaptative via les LT [47]. Des études contradictoires ont montré in vivo que le mutant B. abortus omp25 est atténué dans des souris Balb/c à partir de 18 semaines d infection [49]. Cependant, une autre étude ne permet pas de distinguer une différence de réplication des bactéries dans la rate de souris Balb/c infectées jusqu à 24 semaines [45]. Les mécanismes liés au contrôle de l activation par Omp25 ne sont pas connus. De même, les processus à l origine de l atténuation de la virulence d une souche déficiente en Omp25 dans les infections in vivo ne sont pas encore élucidés. Les vésicules de membrane externe ou «Outer Membrane Vesicles» (OMV) de Brucella, contenant les protéines de la membrane externe, ont été très étudiées. Le processus d internalisation des OMV est dépendant de la clathrine dans les monocytes humains (THP- 1), une des voies d endocytose classique des cellules mammifères [50]. Les OMV modulent la sécrétion de cytokines par ces cellules en la diminuant pendant l infection par Brucella ou pendant une stimulation avec des agonistes des TLR. Enfin, le traitement de monocytes par les OMV précédant l infection conduit à une augmentation de l adhésion et de l internalisation de Brucella [50]. Les OMV de B. melintensis semblent avoir un effet protecteur contre l infection in vivo [51]. L ensemble de ces facteurs de virulence contribuent à établir un environnement favorable à l infection par Brucella. La caractérisation des différents facteurs influant à la fois sur la vie extracellulaire, l entrée dans les cellules, la vie intracellulaire et réplication de la bactérie sont critiques pour la compréhension des mécanismes d infection, ce qui pourra permettre de développer de nouvelles cibles thérapeutiques. 20

I. C. LA VIE INTRACELLULAIRE La capacité de Brucella à établir une infection chronique est à mettre en relation directe avec sa capacité à envahir les cellules hôtes, à y survivre, se répliquer sans toutefois déclencher une forte réponse cellulaire et immunitaire. I. C. 1. Entrée dans les cellules Le mode d invasion ou d entrée de Brucella dans les cellules hôtes n est pas encore complètement caractérisé. Dans les cellules mammifères, trois voies majeures d endocytose sont connues : la première est une endocytose dépendante de la clathrine et de récepteurs endocytiques spécifiques [52] ; la seconde voie est une invagination de la membrane des cellules (enrichie en cholestérol), grâce aux radeaux lipidiques, qui peut être dépendante de la clathrine ou non. La troisième voie est la formation d une vacuole positive pour l actine-f permettant la capture de particules depuis l espace extracellulaire. Ce phénomène est appelé phagocytose [52]. De nombreuses études ont été menées sur les voies d entrée de Brucella dans les cellules hôtes. Dans le cadre des cellules non phagocytaires, une étude a montré que l inhibition de la clathrine conduisait à une abolition de l entrée de Brucella dans les cellules HeLa [53]. De plus, les radeaux lipidiques associés à la dynamine et à la clathrine sont aussi cruciaux pour l entrée et la survie intracellulaire de Brucella. L interaction entre la clathrine et les radeaux lipidiques permettent la polymérisation de l actine, qui est requise pour l entrée de la bactérie. Par la suite, cette étude a montré que la clathrine est aussi requise pour l association de certaines protéines eucaryotes avec la BCV, comme Rab5 [53]. Ces découvertes laissent penser que le mode d entrée de Brucella se déroule ainsi : la bactérie entre dans les cellules selon un mécanisme dépendant des radeaux lipidiques et de l actine-f. La BCV interagit avec des protéines de la voie des endosomes précoces pour permettre à la BCV de suivre le trafic intracellulaire jusqu à sa niche réplicative [53]. 21

Figure 7 : Schématisation du site d entrée de Brucella et des protéines pouvant y participer. Brucella est capable de pénétrer dans les cellules hôtes grâce à des radeaux lipidiques, mais aussi via la présence de récepteurs tels PrPc, SR-A. TLR4 pourrait aussi être impliqué dans l entrée de la bactérie. SP-41, la HSP60 et le LPS de Brucella participerait à l entrée au sein des cellules. L actine, ainsi que des petites protéines G : Cdc42, Rho ou encore Rac sont requises pour ce processus. Adapté de [62].

Ces données ne sont pas très surprenantes sachant que d autres bactéries telles que Listeria [54, 55], E. coli [56], Chlamydia [57], ou encore Yersinia [58] utilisent des mécanismes d entrer dépendants de la clathrine, que ce soit d une manière active ou non. Dans le cas des phagocytes professionnels comme les macrophages, Brucella entre soit via les radeaux lipidiques, soit par opsonisation [59]. L entrée via les radeaux lipidiques est, dans ces cellules, dépendante de la PI3-kinase et de TLR4 [60, 61]. Deux molécules semblent aussi être importantes dans le mécanisme d entrée : le scavenger récepteur de classe A (SR-A) et la protéine PrPc. Ce récepteur pourrait être capable de lier la protéine heat-shock Hsp60 de Brucella, même si ce rôle est controversé [62-64], tandis que SR-A pourrait lier le LPS [65] (Fig. 7). Récemment, une autre étude a montré que TLR4 semblait aussi être impliqué dans l entrée de la bactérie dans des cellules immunitaires comme les macrophages [66]. De nouvelles protéines bactériennes ont été identifiées comme ayant potentiellement un rôle dans l adhésion ou l entrée de Brucella dans les cellules hôtes, parmi celles-ci, une adhésine (Bab1_2009) [67] et la protéine SP-41 [68]. L entrée de la bactérie semble étroitement liée à l activité des GTPases Cdc42, Rho et Rac, recrutées au niveau du site d entrée, et qui interagissent avec le cytosquelette d actine et le réseau de microtubules pour faciliter l internalisation [69]. I. C. 2. Trafic intracellulaire : la BCV sur les traces des endosomes Une fois à l intérieur des cellules, Brucella réside dans une vacuole, la BCV. Celle-ci suit la voie endocytique et devient mature au cours du temps. Elle interagit avec les endosomes précoces et acquiert certains de leurs marqueurs, comme EEA1 ou Rab5. Puis, la vacuole s acidifie, déclenchant ainsi la transcription de l opéron virb, et son expression [30]. La BCV interagit ensuite avec les endosomes tardifs et les lysosomes. Elle acquiert en effet le marqueur Lysosomal-associated membrane protein 1 (LAMP-1) ainsi que Rab-7 [70]. La fusion de la BCV avec les lysosomes pourrait expliquer l acidification de la BCV. Les auteurs de cette étude [70] pensent que la durée de cette interaction serait limitée, pour éviter un contact prolongé entre les composés antimicrobiens présents dans les lysosomes et la bactérie [70]. 22

LPS RicA Bsp Figure 8 : Trafic intracellulaire de la BCV et les protéines eucaryotes ou bactériennes requises. Après l entrée dans la cellule, Brucella réside dans une vacuole, la BCV. La BCV va suivre la voie endocytique et acquérir des marqueurs des différents compartiments endosomaux et lysosomaux. Le trafic de la BCV jusqu au RE se fait grâce à l action de différentes molécules bactériennes (en rouge) : le T4SS VirB, le LPS, RicA et le CβG. L association de la BCV avec le RE et les ERES est permise grâce à des protéines eucaryotes (en bleu). Adapté de [9].

Certains facteurs de virulence de Brucella sont requis pour éviter une fusion prolongée et permanente avec les lysosomes, dont le CβG et le LPS. En effet, un mutant cgs- n est pas capable d éviter la fusion avec les lysosomes et est dégradé. En présence de CβG purifié, ajouté avant l infection, ce mutant est alors capable de se répliquer dans le RE comme la souche sauvage dans les macrophages infectés [21, 71]. Différents effecteurs sont requis pour le trafic intracellulaire de la BCV et l établissement d une niche réplicative dans le RE. Parmi ceux-ci RicA est transloqué pendant la phase intracellulaire de l infection et interagit avec Rab2 [72]. Cette interaction est nécessaire au recrutement de Rab2 sur les BCVs. Le recrutement sur la BCV de Rab2 (ainsi que du complexe GAPDH) permet le trafic intracellulaire de la vacuole jusqu au RE et la survie de la bactérie (Fig. 8) [73]. De plus, une infection avec un mutant RicA conduit à une accélération du trafic intracellulaire qui induit une maturation plus rapide de la BCV quand on la compare à une infection par une souche sauvage [72]. La translocation de protéines effectrices par virb pendant ces processus de trafic intracellulaire a été décrite récemment dans la littérature [30, 62, 74-76]. Les protéines, BspA, BspB, BspC, BspE, BspF (Bsp : Brucella secreted protein) font partie des effecteurs de virb [74]. L expression ectopique de BspA, BspB et BspF conduit à une inhibition générale de la sécrétion de protéines dans les cellules. De façon intéressante, l infection par Brucella conduit à une diminution de sécrétion de protéines, via l action de BspA, BspB et BspF. Cette inhibition a lieu avant que la BCV ne devienne la niche réplicative de la bactérie grâce à BspB et BspF. Ce mécanisme serait indispensable pour la persistance et réplication de Brucella [74]. I. C. 3. Le Réticulum Endoplasmique, un havre de paix Tous les phénomènes décrits ci-dessus ont pour but d aboutir à l arrivée de Brucella dans le RE. Dans la plupart des types cellulaires, Brucella se réplique au sein du RE, à l exception notable des trophoblastes extravillaires infectés par B. abortus et B. suis dans lesquels la bactérie se réplique dans des inclusions positives pour LAMP-1, ou des monocytes humains dans lesquels Brucella réside dans des phagosomes positifs pour LAMP-1 [35, 77-80]. Les BCVs interagissent avec les ERES (endoplasmic reticulum exit sites), et acquièrent des marqueurs Sec61, la calnexine, la calreticuline, probablement par des échanges de membranes 23

[35, 71, 77]. Ces mécanismes sont dépendants de protéines hôtes comme Sar1 et COPII [77]. On peut relier la présence de Bsp (BspA, BspB, BspF) et l inhibition de la sécrétion de protéines avec le fait que la bactérie réside au niveau du RE et interagit avec les ERES, qui se situent au début des voies de sécrétion de la cellule. Une fois que Brucella a atteint le RE, les bactéries commencent à se répliquer sans perturber l intégrité de la cellule, ni la tuer [62]. La réplication de la bactérie dans le RE est suivie par la conversion des BCVs en vacuoles ayant des propriétés des autophagosomes [62, 77, 81]. Dans une étude récente, on note ainsi que l acquisition de certaines protéines de la voie autophagique (Beclin-1, ATG14L) sont nécessaires à la formation de cette BCV particulière. Elle est requise pour le cycle intracellulaire de Brucella et la sortie des bactéries de la cellule hôte avec pour conséquence une infection des cellules environnantes [81]. 24

Figure 9 : Les récepteurs TLR et leurs ligands. Les TLR 1, 2, 4, 5 et 6 sont membranaires, tandis que les TLR 3, 7 et 9 endosomaux. Ils sont capables de reconnaître différents types de PAMP indiqués ici. TLR2 fonctionne avec TLR1 ou TLR6 et reconnaît des ligands différents, tandis que TLR4 utilise MD-2, son co-récepteur CD14 et les LBP (LPS binding proteins) pour reconnaître le LPS. Adapté de [87].

I. D. BRUCELLA ET LE SYSTEME IMMUNITAIRE Brucella, pour engendrer une maladie chronique et persister dans l organisme va établir une stratégie d évitement du système immunitaire. En effet, elle contrôle l inflammation trop importante dès le début de l infection, prévenant ainsi une destruction rapide de la bactérie [82-85]. I. D. 1. Une stratégie d évitement : réponses aux TLR Un des piliers de cette stratégie d évitement prônée par Brucella repose sur sa détection. En effet, sans détection, ou induction de réponse après détection, le système immunitaire ne peut réagir et monter une réponse immunitaire efficace contre le pathogène. Brucella agit donc sur les récepteurs des cellules pouvant la détecter. Ces récepteurs, appelés PRR pour «Pattern Recognition Receptor» sont les TLR, les nod-like récepteurs (NLR) ou les récepteurs lectines de type C (CLR) [86]. Les TLR vont être capables de reconnaître des PAMP de bactéries, de virus et de champignons (Fig. 9) et de déclencher des voies de signalisation conduisant à la transcription de gènes cibles pour y répondre. Les PRR sont exprimés par des cellules immunitaires telles que les macrophages, DC. La reconnaissance de ces motifs va déclencher une cascade de signalisation passant par des molécules adaptatrices comme TIRAP, Myd88, TRIF, TRAM puis par des MAPK. Ces molécules ainsi que la partie cytoplasmique des TLR contiennent des domaines Toll/IL-1 receptor (TIR) qui permettent les interactions entre les récepteurs TLR et leurs molécules adaptatrices [86, 87]. Les voies de signalisation aboutissent à la translocation dans le noyau des cellules de facteurs de transcription comme nuclear factor-kappa B (NF-κB), activator protein 1 (AP-1), interferon regulatory factor (IRF) ou nuclear factor of activated T-cells (NF- AT) (Fig. 10). Ces différents facteurs sont responsables de la transcription de gènes cibles comme les cytokines pro-inflammatoires TNF-α, l IL-12, l interleukine 6 (IL-6), l interleukine 1β (IL-1β) [88]. Des chimiokines, qui attirent les neutrophiles, comme CCL-2 (MCP-1), CXCL-12 (MIP-2), KC (analogue de l interleukine 8 (IL-8) humain) sont aussi sécrétées [89]. 25

Figure 10 : Voies de signalisation en aval des TLR. La signalisation via les TLR requièrent la présence de molécules adaptatrices comme TIRAP, MyD88, TRIF ou encore TRAM. Après activation les molécules adaptatrices vont permettre l activation de différentes kinases (IRAK (interleukin-1 receptor-associated kinase), et les MAP kinases). Après la cascade de signalisation, différents facteurs de transcription (AP-1, NF-κB) transloquent dans le noyau où ils permettent la transcription de gènes cibles pour répondre à la détection d un PAMP. Adapté d Invivogen

Les TLR sont des récepteurs cruciaux pour la détection de Brucella, via la reconnaissance de PAMP. Différentes études ont décrit l importance des TLR dans les réponses contre la bactérie, ainsi que dans la résistance conférée à l hôte. TLR2 semble être important dans la génération de cytokines comme TNF-α, l IL-6, l IL-12 et l IL-10 par les macrophages péritonéaux stimulés par les lipoprotéines Omp16 et Omp19 [90]. On peut donc supposer que TLR2 est capable de reconnaître certains composants de la membrane externe de la bactérie. Certaines études démontrent un rôle de Brucella sur la signalisation en aval de TLR2 et TLR4 et un rôle de TLR4 dans la résistance de l hôte à l infection [91, 92]. Un des ligands connus de TLR4 est le LPS. Dans le cadre de Brucella, nous avons vu précédemment que son LPS est non canonique et présente une structure particulière. Cette structure permet au LPS de Brucella d être un faible inducteur de la signalisation en aval de TLR4 [82, 93]. Alors que la liaison d un LPS classique à TLR4 déclenche une forte réponse immunitaire et inflammatoire, ici, la réponse est très atténuée. De plus, des DC stimulées avec le LPS de Brucella purifié restent immatures [94]. Si on introduit une mutation wadc dans le core oligosaccharidique (WadC est une glycosyltransferase qui transfère des mannosides formant la partie du core oligosaccharidique externe) du LPS de Brucella on augmente la liaison du LPS à MD2, co récepteur de TLR4, ce qui entraîne une forte réponse immunitaire. La capacité du LPS de Brucella à induire une faible activation des DC est donc conférée par son core [94]. Brucella utilise donc son LPS pour éviter une signalisation proinflammatoire via TLR4. Une autre étude a démontré que la lumazine synthase de Brucella spp (BLS) est reconnue par TLR4 et est capable d activer les DC après stimulation avec cette synthase [95]. De même, le CβG est reconnu par TLR4 et va induire une activation des DC [22]. Cela conduit donc à avoir une balance entre des molécules aux propriétés activatrices et d autres aux propriétés inhibitrices des voies de signalisation en aval des TLR. TLR9, récepteur endosomal, joue un rôle important dans l initiation des réponses immunitaires contre Brucella [96, 97]. L ADN de Brucella est un ligand de ce récepteur et déclenche une réponse de type Th1 [98]. Or, la production de cytokines pro-inflammatoires de type Th1 comme l IL-12 est réduite pendant l infection des DC et macrophages TLR9 KO [96, 98], suggérant un rôle protecteur de ce récepteur pour la cellule. Cependant, la production de composés oxygénés (ici l oxyde nitrique, NO) et de TNF-α n est pas impactée, suggérant 26

Figure 11 : Différents mécanismes de détection de Brucella et les voies de signalisation en aval des récepteurs. Divers composés de Brucella sont détectés via TLR4 (Omp16, LPS), TLR2 (Omp16, Omp19), ou encore TLR9 (ADN). La reconnaissance de ces molécules entraîne l activation de voies de signalisation en aval de MyD88 et TRIF/TRAM. La reconnaissance de Brucella se fait aussi par les récepteurs NOD, ou d autres récepteurs non connus qui activent RIP-2 et STING respectivement. Cela induit la translocation d IRF-3, AP-1 et NF-κB dans le noyau des cellules, provoquant la transcription de gènes pro-inflammatoires. Adapté de [97].

qu il y a d autres voies d activation telles que les voies de signalisation en aval de TLR2/6 [99]. TLR6 semble aussi important pour permettre de monter une réponse immunitaire efficace contre Brucella [100]. En effet, dans un modèle de souris KO pour TLR6, il n y a plus de contrôle de l infection et donc une réplication plus importante de Brucella. Au cours de l infection, TLR6 et TLR2 seraient requis pour l activation de BMDC via la transduction de signal des MAPK. TLR2 et TLR6 seraient de plus capables de reconnaître Brucella et activer les DC [100]. Un modèle de souris KO pour MyD88, molécule adaptatrice présente dans les voies de signalisation en aval de TLR1/6, TLR2, TLR9 et TLR4 notamment, a permis de démontrer le rôle crucial de cette protéine dans les réponses immunitaires contre Brucella [92, 96]. En effet, dans des souris KO pour MyD88, Brucella se réplique plus que dans des souris contrôles. Ce défaut de contrôle de l infection serait dû à une déficience de présentation antigénique, et donc de production d IFN-γ par les LT, ainsi qu un manque de sécrétion de cytokines pro-inflammatoires par les macrophages et les DC [92, 96]. Plusieurs TLR semblent donc être important à la fois dans la détection et reconnaissance de Brucella, mais aussi dans leur rôle de molécules en amont de voies de signalisation, cruciales pour monter une réponse immunitaire. Les molécules adaptatrices comme MyD88, ainsi que les protéines de signalisation en aval de MyD88 sont aussi requises pour le contrôle de l infection (Fig. 11). I. D. 2. La résistance aux défenses innées L une des grandes forces du système immunitaire inné repose sur la sécrétion de composants capable d éliminer les pathogènes : les ROS (réactifs oxygénés), les défensines, le complément et peptides anti-microbiens. En cas d infection, des protéines plasmatiques et des polynucléaires neutrophiles sont attirés sur le site d inflammation et s y infiltrent. Les neutrophiles s activent dès leur arrivée dans le tissu soit via la détection du pathogène/antigène, soit à travers l action de cytokines sécrétées par les macrophages et mastocytes. 27

Ils sécrètent alors des granules toxiques (phénomène appelé «dégranulation») contenant des ROS comme le NO produit par les NOS (nitric oxydase synthase) telle inos, des réactifs azotés, de la cathepsine G, protéinase 3, de l élastase [101]. La dégranulation vise à l élimination directe et rapide des pathogènes présents au site d infection. Cependant, Brucella est capable de limiter l action des molécules anti-microbiennes. En effet, des études ont permis de montrer qu à la fois le LPS et la membrane externe de la bactérie permettent d éviter la lyse de Brucella par les peptides cationiques bactéricides. Les auteurs de ces études ont testé une vingtaine de peptides, et systématiquement Brucella résiste mieux à la mort induite par ces composés que les autres bactéries testées [102-104]. Brucella résiste aussi à l action de dégranulation des neutrophiles, ainsi qu à l activation du complément qui permet l opsonisation des pathogènes, leur phagocytose et leur dégradation [82, 105]. La résistance au complément pourrait dépendre en partie d une protéine de la bactérie, WboA, qui inhiberait l activation du complément via la voie des lectines du complément [106]. D autres pathogènes ont déployé des moyens pour éviter l élimination par le complément. Ainsi, parmi les protéines de Staphylococcus aureus, la protéine staphylococcale A (SPA) et l immunoglobuline (Ig) -binding protéine A staphylococcale (Sbi) sont capables de lier les fragments Fc des anticorps IgG, et ainsi de prévenir la phagocytose dépendante de l opsonisation [107]. De même, Neisseria meningitides exprime la protéine GNA1870 capable de se lier au Facteur H, une protéine requise pour le clivage permettant la formation de C3b du complément [108]. Une explication possible à la résistance de Brucella aux molécules anti-microbiennes se trouve dans la composition de la membrane de Brucella (LPS, lipoprotéines, phospholipides, etc ). Ces éléments sont très hydrophobes et portent peu de charges négatives comparés à d autres bactéries [109]. Un autre mécanisme de résistance aux défenses innées est le contrôle et la manipulation du trafic intracellulaire de la vacuole par les bactéries, via le LPS, le CβG, virb, etc Cela permet à Brucella de résister à la lyse par les lysosomes dans les cellules comme les macrophages, qui sont l une des premières lignes de défense immunitaire. Les neutrophiles qui sont recrutés en cas d infection font aussi partie de la stratégie de Brucella pour minimiser l inflammation causée par l infection. 28

Brucella est capable d échapper à la mort induite par ces cellules et les active faiblement [110, 111]. De plus, le recrutement des neutrophiles dans les tissus à de temps précoces après infection est relativement faible, ce qui est dû à la faible sécrétion de cytokines proinflammatoires et de chimiokines par les cellules immunitaires résidentes comme les macrophages [82]. Dans un modèle de souris neutropéniques (déplétés en neutrophiles, soit par injection d un anticorps (anti RB-6) soit un modèle KO, Genista) [82, 112], on constate une plus forte activation des LT CD4 + et CD8 +, signe d activation de la réponse immunitaire adaptative. De plus, un fort recrutement de monocytes dans le sang, ainsi qu une diminution de la réplication dans la rate ont été observés à des temps d infection longs, correspondant à la phase chronique de la maladie. La sécrétion de cytokines de type Th1 est aussi plus importante en absence de neutrophiles. À des temps plus précoces (5 j après infection), la présence de neutrophiles est critique pour éliminer la bactérie. Mais à des temps plus tardifs (15 j) c est la situation inverse, la présence de neutrophiles induit une réplication de Brucella plus élevée et une réponse immunitaire plus faible [113]. Tout cela nous indique que la présence de neutrophiles a un rôle double dans la réponse à l infection, en phase aigüe elle est bénéfique pour la réponse immunitaire, et en phase chronique elle est délétère pour l hôte. Cependant ces résultats sont différents chez l homme, une étude a montré que les neutrophiles infectés étaient activés. En effet, ils vont sur-exprimer des molécules d activation comme CD25, ou diminuer l expression de CD62L (les cellules sont naïves quand elles l expriment fortement). L infection par Brucella va aussi conduire à la sécrétion d IL-8, une chimiokine requise pour l attraction et le recrutement de leucocytes dans le tissu. Les auteurs de cette étude ont établi que la lipoprotéine Omp19 était responsable de l activation des neutrophiles. L infection par des Brucella inactivées à la chaleur ou la stimulation avec Omp19 conduit les neutrophiles à monter une réponse de stress oxydatif (relarguage de ROS, NO), à leur migration, ainsi qu à prolonger leur survie [114]. Les macrophages sont parmi les cellules ciblées préférentiellement par la bactérie. Capable de phagocyter et dégrader des pathogènes, ces cellules sécrètent aussi des chimiokines pour attirer les neutrophiles sur le site d infection. Les granulomes, caractéristiques d inflammation prolongée sont souvent constitués de macrophages. Dans ce cas, les infiltrats de neutrophiles sont remplacés par ceux de macrophages et des LT. C est le cas lors d infection par Mycobacterium ou par Brucella [115, 116]. 29

Les macrophages font partie des cellules les plus étudiées, notamment au niveau du cycle de vie intracellulaire de Brucella. La capacité des macrophages à éliminer des pathogènes réside dans leur capacité de phagocytose qui conduit à la fusion de la vacuole (phagosome) avec les lysosomes et la destruction du pathogène. Dans le cas de Brucella, environ 90% des bactéries sont détruites par fusion de la BCV avec les lysosomes, mais les 10 % restants sont capables de rejoindre le RE et d y établir une niche réplicative [62]. Pendant l infection in vivo, les macrophages de la pulpe rouge de la rate (F4/80 + ) et ceux de la zone marginale (MOMA + ) sont les premières cellules spléniques à être infectées [116]. La bactérie va inhiber l apoptose des macrophages murins et humains infectés pour garder sa niche réplicative intacte [117, 118]. Les macrophages et les DC ne sont pas ou peu activés par l infection [119, 120]. Une étude a démontré que la sécrétion d IL-10 par les LT CD4 + in vivo conduisait à cette non-activation des macrophages et à la persistance de Brucella [121]. Les cellules NK (Natural Killer) sont des cellules de l immunité innée, capables de dégranulation, de lyse cytotoxique et de sécréter des cytokines pro-inflammatoires dont l IFNγ. Ces cellules jouent un rôle critique dans beaucoup d infections virales et bactériennes [122]. Dans le modèle murin, les NK ne semblent pas avoir de rôle dans la réponse immunitaire contre Brucella [123], alors qu il était établi que les NK étaient affectés par l infection et que leurs fonctions (mais pas leur nombre) étaient altérées chez l homme [124]. Différentes études se sont donc penchées sur le rôle de ces cellules et leur pertinence dans l infection. Ainsi, la réplication de B. suis au sein de macrophages est diminuée en présence de NK [125]. Cette diminution ne serait pas due à une sécrétion augmentée de cytokines comme l IFN-γ mais à un effet cytotoxique contact dépendent de la part des NK [125]. Une autre étude, menée avec des souris immunisées avec brucelles tuées à la chaleur suggère que les NK sont importantes pour l induction d une réponse anticorps par les LB [126]. Brucella va donc essayer de limiter sa reconnaissance par les cellules immunitaires via une modification de ses molécules membranaires. Elle résiste aux mécanismes classiques de défense innée comme le complément et les peptides bactéricides. La bactérie va également limiter l activation, et donc la réponse, des premières cellules immunitaires présentes ou recrutées au site d infection comme les macrophages et neutrophiles. 30

Figure 12 : Différentes étapes de maturation des DC. Les DC naïves s activent suite à la phagocytose d un antigène. Elles apprêtent l antigène pour présenter un peptide antigénique via les molécules de CMH-II aux LT. Elles augmentent leur expression de molécules de co-stimulation : CD40, CD80 et CD86. Ces molécules sont requises pour déclencher un signal d activation aux LT. A la suite de leur activation, les DC migrent dans les organes lymphoïdes secondaires (rate, ganglions notamment) pour présenter l antigène. Elles sécrètent différentes cytokines qui polariseront la réponse immunitaire induite. Adapté de [129].

I. D. 3. Brucella et les DC Les DC sont les autres cellules immunitaires innées que Brucella va cibler et tenter de contrôler. Ces cellules sont capables de s activer après la détection d un pathogène. Elles expriment, en plus du marqueur CD11c, des molécules dites de co-stimulation qui vont faciliter l activation des LT. Ces molécules sont, chez la souris, CD80, CD86 et CD40, qui se lient à CD28 et CD40L sur les LT. De même, le CMH-II est fortement surexprimé dans les DC activées, pour aider à la présentation antigénique aux LT [127-129]. Les DC vont dégrader les antigènes en peptides antigéniques qui seront ensuite présentés aux LT via le CMH. Les DC sécrètent aussi des cytokines pro-inflammatoires comme TNF-α, IL-6, IL-12 [127-129]. Par la suite et après activation, elles migreront dans les organes lymphoïdes secondaires comme la rate et les ganglions pour y trouver les cellules effectrices (Fig. 12). Dans les DC humaines, B. suis est capable de se répliquer et va causer une inflammation limitée. En effet, les molécules de co-stimulation et récepteurs aux chimiokines, CD40, CD83, CD86, CCR7 et les molécules du CMH-II (HLA chez l homme) sont plus exprimées dans des DC infectées, mais restent néanmoins à un niveau d expression intermédiaire comparé à une infection par E. coli [44, 47]. Les DC humaines infectées par Brucella montrent une activation limitée des LT comparées aux DC infectées avec E. coli [47]. Une autre étude renforce l idée de DC humaines peu activées. En effet, les auteurs constatent que B. abortus (tout comme Coxiella) induit peu la voie des interférons de type I dans les DC humaines, et limiterait l activation de ces cellules [84]. Chez la souris, l infection par Brucella induit, in vivo, l activation et la migration des DC spléniques (infectées ou non) dans la pulpe blanche de la rate, où les LT résident [116]. Des DC inflammatoires participent aussi à la formation de granulomes, et serviraient de réservoir pour la bactérie en phase chronique [116]. Dans un modèle d infection nasal chez la souris, les DC des poumons ne sont pas impactées en termes d activation ou de localisation par l infection. En absence de macrophages, les DC inflammatoires des poumons migrent dans le ganglion drainant les poumons et pourraient permettre la dissémination de la bactérie dans l organisme [130]. 31

Une étude sur les DC bovines (dérivées à partir du sang avec du GM-CSF et de l IL-4), a montré que celles-ci éliminent rapidement Brucella et il n y donc pas de réplication de la bactérie. De plus, les DC ne semblent pas être activées par l infection [131]. Dans le laboratoire, nous avons montré que BtpA (Brucella tir protein A, aussi appelé TcpB), une protéine de Brucella contenant un domaine TIR est capable de réguler l activation des DC [23]. En effet, en interagissant avec les voies de signalisation en aval des TLR, BtpA est capable de bloquer l activation via TLR2 et TLR4. Il y a plusieurs hypothèses pour expliquer l effet de BtpA. BtpA interagirait avec TIRAP et/ou avec MyD88 [132-136]. L action de BtpA sur les DC contribue à limiter leur activation en termes de sécrétion de cytokines proinflammatoires, de présence de DALIS (DC specific Aggresome Like Induced Structures) dans les cellules [23]. Une autre étude, ainsi que des données non publiées du laboratoire montrent que lorsque BtpA est exprimée ectopiquement, elle colocalise avec les microtubules et pourrait participer à leur désorganisation [134]. L interleukine-10 (IL-10) est une cytokine anti-inflammatoire sécrétée par un large panel de cellules incluant les DC, monocytes, mastocytes et certaines populations de LT et LB. Cette cytokine dite tolérogène va diminuer la réponse inflammatoire. Dans la plupart des cas, la sécrétion d IL-10 promeut la survie de l hôte comme pour les infections à Toxoplasma gondii [137], Trypanosoma cruzi [138], Plasmodium spp [139]. Cependant, dans le cadre des infections à mycobactéries, l IL-10 semble avoir un rôle délétère pour l hôte et est associée à une susceptibilité accrue ainsi qu à une réplication plus rapide des bactéries [140]. D autres pathogènes semblent aussi profiter du rôle anti-inflammatoire de l IL-10, comme Coxiella dont la virulence dépend d une production importante d IL-10 par les cellules immunitaires [141]. Dans les infections à Brucella, il semblerait que la présence d IL-10 favorise l infection et la survie de la bactérie. En effet, une absence d IL-10 (souris KO IL-10) conduit à une augmentation de la sécrétion de cytokines pro-inflammatoires et une élimination des bactéries [85, 142]. Cependant, nous n avons pas été en mesure de détecter la production d IL-10 dans les BMDC infectées par Brucella d après les différentes études menées dans notre laboratoire précédemment (données non publiées). Cela dit, il n est pas à exclure qu in vivo certaines 32

Figure 13 : Différentes populations de LT CD4 + et leurs réponses face aux infections. Lorsqu un LT CD4 + naïf reconnait un antigène présenté par une CPA via le CMH-II, celui-ci va se différencier. Selon le cocktail de cytokines auquel il est soumis (via la CPA), il peut notamment devenir un Th1, Th2 ou Th17. En présence d IL-12 et d IFN, les LT deviendront des Th1, cytotoxiques, capables de sécréter de l IFN-γ et du TNF- α. Les Th1 sont importants dans l immunité contre les bactéries et parasites intracellulaires. Si le LT naïf est en présence d IL-23 et d IL-1, il se différenciera en Th17. Les Th17 sont pro-inflammatoires et jouent un rôle dans la résistance aux bactéries extracellulaires et champignons via la sécrétion d IL-17, IL-22. Enfin, la présence d IL-4 induira la différenciation des LT en Th2. Ces cellules sécrètent de l IL-4 et de l IL-5. Elles agissent dans l immunité parasitaire, et participent à la réponse humorale. Adapté de [143].

populations de DC, ou de monocytes soient des sources d IL-10 durant l infection et contribue à établir un contexte anti-inflammatoire propice à l établissement d une pathologie chronique. In vivo, l infection par B. melitensis induit une production de TNF-α et d inos dans les DC inflammatoires (CD11b + ; Ly6C + ) de la cavité péritonéale et de la rate (pour l inos seulement) [92]. De plus, en utilisant des souris KO pour inos, on note que la réplication de Brucella est augmentée comparée à des souris sauvages. En utilisant des souris déficientes en MyD88, TLR4 et TLR9, les auteurs de cette étude ont pu observer une diminution de la production d IFN-γ et d inos par les LT CD4 + et les DC inflammatoires qui corrélait avec une croissance non contrôlée de la bactérie [92]. I. D. 4. L immunité adaptative contre Brucella La réponse immunitaire adaptative vise à développer une réponse spécifique contre un antigène donné pour éliminer un pathogène, ainsi que développer des mécanismes de mémoire immunologique. Les LT CD4 + peuvent se différencier en différentes populations, parmi elles, Th1, Th2 ou encore Th17 (Fig. 13). Les Th1 vont promouvoir une réponse de type cellulaire (même si ces cellules participent aussi à la réponse humorale en promouvant la production d anticorps), en sécrétant des cytokines comme l IFN-γ, le TNF-α mais aussi en activant les macrophages ou en aidant au recrutement de cellules sur les sites d infection [143]. Les Th2 vont quant à eux promouvoir une immunité dite humorale. Ils sécrètent des cytokines comme l IL-4, l IL-10, coopérant ainsi avec les LB ils régulent la production d anticorps et notamment d IgE et IgG1 par les LB [144]. Les Th17 ont un rôle inflammatoire via la sécrétion d IL-17 et IL-22 dans la réponse anti-microbienne [145]. Les LT CD8 + se différencient en CTL (LT cytotoxique) et lysent les cellules infectées par l induction de l apoptose via les récepteurs Fas, mais aussi en libérant des granules cytotoxiques contenant du Granzyme B et des perforines [146]. Les réponses immunitaires adaptatives à l infection par Brucella ont été étudiées dans différents modèles d infection in vivo de souris [147]. 33

Il a été montré que la réponse immunitaire de type Th1 est primordiale pour lutter contre l infection. Elle va consister en une sécrétion de cytokines pro-inflammatoires comme l IFN-γ et l IL-12, ainsi qu une réponse cytotoxique pour détruire les cellules infectées évitant ainsi la réplication des bactéries. Différentes études ont montré le rôle critique de l IFN-γ dans la survie et l élimination des bactéries [148-151]. Les gènes régulant la production d IFN-γ comme irf-1 (interferon regulatory factor 1), irf-8 (aussi appelé ICSBP) ou l IL-12 ont aussi été montré comme important dans l induction d une réponse immunitaire efficace [150, 152]. Ainsi, des souris déficientes pour l IFN-γ ou IRF-1 sont des modèles létaux de brucellose. Ils permettent d étudier la virulence des souches bactériennes utilisées. L IFN-γ est une cytokine importante qui régule la production de d inos et la synthèse de réactifs oxygénés. Les souris IFN-γ KO présentent un manque de production d inos et donc de réactifs oxygénés pendant l infection [92]. Cela peut conduire à une réponse immunitaire précoce incomplète et permettre l établissement d une infection persistance plus facilement. L IFN-γ va aussi permettre l activation des macrophages et de leur capacité à activer leur machinerie bactéricide pour éliminer les pathogènes. Une étude récente a démontré qu en utilisant différents modèles de souris KO, les réponses des LT CD8 +, les Th17, Th2 et les LB ne sont pas requises pour le contrôle de l infection par Brucella contrairement aux LT CD4 + Th1 qui sécrètent de l IFN-γ [92, 153]. Cependant, les LT CD8 + ainsi que les LB sont plus importants dans la réponse immunitaire à l infection par Brucella en présence de Th1 que les Th2 et Th17 [153]. Il est important de préciser que la production d IFN-γ par les LT CD4 + au cours de l infection est dépendante de la reconnaissance spécifique d un antigène présenté par des molécules du CMH-II. Les LT γδ, n ont pas besoin de la présentation antigénique pour s activer et sont capables de sécréter rapidement de fortes quantités d IFN-γ. Ces cellules ont été montrées comme étant importantes dans la réponse immunitaire à d autres pathogènes intracellulaires comme Listeria, Salmonella, Mycobacterium ou encore Francisella [154-157]. Il n est pas exclu que cette population de LT particulière participe aussi à la sécrétion d IFN-γ requise pour la lutte contre Brucella. 34

Dans une étude récente, les auteurs ont remarqué que les LT γδ murins sont la première source d IL-17 durant l infection et qu ils participent aussi à la sécrétion d IFN-γ [158]. Il semblerait aussi qu ils soient importants dans le contrôle de la réponse immunitaire précoce. 7 jours après infection, la réplication de Brucella est plus importante dans la rate des souris KO pour le TCR γδ que dans celle des souris sauvages. En revanche, aucune différence n est détectée à 15 ou 30j après infection. [158]. Ce rôle protecteur des LT γδ est dépendant de la sécrétion de TNF-α. Une partie de ces résultats a été confirmée dans des bovins, chez lesquels les LT γδ sont une population de LT très importante et présente, contrairement au modèle murin. Les LT γδ des bovins seraient capables d empêcher la réplication intracellulaire au sein des macrophages infectés, via notamment leur sécrétion d IFN-γ [158]. Des résultats similaires d activation de LT γδ humains par un peptide de B. suis ont été obtenus. Dans ce modèle, les LT γδ activés permettaient aussi un contrôle de la réplication de Brucella au sein de macrophages [159]. Pour finir, il est important de mentionner les lymphocytes B. Différentes études soulignent le rôle de ces cellules dans la sécrétion d anticorps et la réponse immunitaire au cours de l infection. Dans un modèle de souris déficientes pour les LB matures (souris µmt), on observe une élimination de Brucella plus rapide que dans les souris sauvages infectées, et cette élimination n est pas due à une absence d anticorps puisque des souris µmt injectées avec des anticorps étaient toujours capables d éliminer les bactéries [160]. Il semblerait que ce phénotype soit associé à une augmentation de la sécrétion d IFN-γ ainsi qu une production réduite d IL-10 dans ces souris [160]. De plus, les LB sont une niche réplicative pour Brucella, capable de se répliquer dans ces cellules, même si cela conduit probablement à l activation des LB qui sécrètent alors du TGFβ [161]. Le TFG-β est connu pour son rôle anti-inflammatoire, cela pourrait faire partie d un mécanisme déployé par la bactérie pour limiter l activation des cellules immunitaires. Brucella est capable d exprimer une protéine appelée PrpA (pour proline racemase protein A) qui induit la prolifération des LB [162]. Cette protéine se lie aux macrophages via la nonmuscular myosin IIA, NMM-IIA ce qui va conduire à leur activation (augmentation de l expression de CD86), et à la sécrétion d un facteur soluble conduisant à la prolifération des LB [163]. 35

La production d anticorps est la fonction principale des LB dans le cadre des infections à Brucella. Le rôle des anticorps semble être double : des anticorps anti-lps permettrait de protéger contre une infection [164] et des modèles de souris KO pour les LB (et donc déficients en anticorps spécifiques contre la bactérie) seraient capables d éliminer la bactérie plus rapidement, suggérant un rôle moindre des anticorps dans la réponse immunitaire à Brucella [160]. Dans le sérum des souris infectées, les deux isotypes d anticorps les plus abondants sont les IgG2a et les IgG3, qui sont souvent associées avec une réponse Th1 [165]. Dans les cas de brucellose bovine, les anticorps ne semblent pas protecteurs et aideraient même à la dissémination de la bactérie en empêchant la lyse des bactéries extracellulaires par le complément [166]. En cas de réinfection, la réponse immunitaire est de type humorale et cellulaire (via les LT CD4 + Th1). Ces deux types de réponses sont requises pour conférer une protection à l hôte [167]. Il ressort de toutes ces études que Brucella est une bactérie capable d échapper au système immunitaire en modulant la structure de ses molécules, mais aussi en produisant des protéines qui interfèrent directement avec des récepteurs, molécules eucaryotes pour bloquer la réponse immunitaire. On s aperçoit que l inflammation déclenchée par l infection est relativement faible puisque l on observe peu de splénomégalie, de sécrétion de cytokines dans le sérum (et donc de réponse systémique), de recrutement de cellules immunitaires. Brucella va induire la sécrétion de facteurs anti-inflammatoires comme l IL-10 et le TGF-β pour limiter l inflammation. La réponse immunitaire induite, qu elle soit innée ou adaptative est tardive, et donc peu efficace pour éliminer la bactérie qui se réplique déjà dans sa niche intracellulaire et est capable de se disséminer. 36

Figure 14 : Récepteurs murins de la famille SLAMF. Les récepteurs de la famille SLAMF sont composés de domaines immunoglobulines IgV like en N-terminal (V). Ils possèdent tous un domaine C2 contenant un pont di-sulfure. Les différentes protéines SLAM contiennent dans leur partie cytoplasmique des tyrosines phosphorylables, appartenant à un ITSM (Immunoreceptor tyrosine-based switch motifs). C est sur ces sites que SAP (SLAM-associated protein) et EAT-2 (EWS-Fli1-activated transcript-2) vont se lier. Ces molécules permettent de déclencher des cascades de signalisation en aval des SLAM et induire la transcription de gènes cibles. Adapté de [170].

I. E. CD150, UN RECEPTEUR A LA SURFACE DES CELLULES IMMUNITAIRES CD150 (aussi appelé SLAM (signaling lymphocyte activation molecule) ou Slamf1) est une molécule exprimée à la surface des cellules immunitaires (DC, monocytes, LT, LB, cellules souches hématopoïétiques (HSC), etc ). Décrite au milieu des années 90, CD150 permet l induction des réponses immunitaires, notamment dans les LT, LB et DC. Plus récemment, certaines études soulignent son rôle dans l inhibition de réponses immunitaires [168, 169]. Cette molécule appartient à une large famille de récepteurs membranaires aussi appelé SLAM Family comprenant 9 membres (Slamf1 à Slamf9) tous jouant un rôle plus ou moins important dans la réponse immunitaire (Fig. 14) [170]. Ces récepteurs appartiennent à la super famille CD2 des molécules contenant domaines immunoglobulines. I. E. 1. CD150, une molécule homophylique de co-stimulation CD150 a tout d abord été décrit comme un récepteur important dans l activation des LT et LB [171]. Son expression augmente à la surface des cellules activées (des LT dans cette étude), et la liaison homophylique de CD150 conduit à une sécrétion de cytokines par les LT CD4 + d IFN-γ notamment. Ces cellules vont aussi proliférer même en absence de co-stimulation via CD28 et de manière indépendante de l IL-2 [171, 172]. Ce récepteur est capable de former des homodimères entre deux cellules immunitaires comme les LT, ou encore entre un LT et une DC [173, 174]. La production d IFN-γ via les Th1 dépend de l activation de CD150. En effet, l ajout d anticorps anti-cd150 activateurs induit la production par les LT CD4 + Th1 du récepteur β2 à l IL-12, qui est nécessaire à l activation complète des LT, et donc à la production d IFN-γ par les Th1 [175]. L activation de CD150 via son association avec une autre molécule de CD150, induit le recrutement de la protéine tyrosine phosphatase SHP-2 (SH2 domain-containing protein) mais pas de SHP-1 sur un immunoreceptor tyrosine-based switch motif (ITSM) situé dans la partie intracellulaire de la molécule [176]. Une autre molécule, SAP (SLAM-associated protein, aussi appelé SH2D1A) a été identifiée comme étant capable de lier les tyrosines présentes sur la partie cytoplasmique de CD150 (humain et murin) [176, 177]. SAP interagit 37

avec FynT, une kinase, et bloque la liaison de SHP-2 à CD150 en s associant avec les mêmes motifs ITSM que SHP-2, ce qui conduit à une situation de compétition entre les deux molécules [176-180]. SAP serait requis à la sécrétion de cytokines de type Th2 par les LT CD4 + et à l activation de la réponse humorale [181, 182]. En revanche, SAP inhibe la réponse cytotoxique des CD8 +, peut-être pour éviter des situations d auto-immunité [179, 183]. L association de CD150 sur les LT CD4 + conduit à la phosphorylation de Akt mais pas à celle des MAPK ERK1/2 [175]. La phosphorylation d Akt pourrait ainsi permettre aux LT d augmenter la production d IL-2 et d IFN-γ [175, 184]. L arthrite rhumatoïde (RA) est caractérisée par une accumulation de LT, macrophages au niveau des articulations enflammées et par la grande production de cytokines proinflammatoires [185]. Une étude sur les LT de patients atteints de RA démontre une présence accrue de LT CD150 + dans leurs fluides synoviaux [186]. De plus, une activation de CD150 via des anticorps monoclonaux (mabs) conduit à une production plus élevée de cytokines pro-inflammatoires, ici l IFN-γ et le TNF-α mais aussi d IL-10, suggérant une boucle de rétro-contrôle négatif pour limiter l inflammation. Les auteurs concluent en suggérant que l association CD150-CD150 des LT dans les zones enflammées de ces patients conduit à l activation de voies de signalisation inflammatoires dans les articulations [186]. I. E. 2. Les propriétés immunomodulatrices de CD150 L expression de CD150 ne se restreint pas uniquement aux LT, en effet les LB, DC, MO, NKT, HSC et autres cellules immunitaires expriment aussi CD150 [171-173, 187, 188]. De ce fait, CD150 semble jouer un rôle dans différents aspects de la réponse immunitaire et de l activation des cellules. La liaison de deux molécules de CD150 à la surface de LB va conduire à une hausse de la sécrétion d immunoglobulines et la prolifération des cellules [173]. En effet, on observe une expression plus importante de CD150 après activation des LB. Cette expression accrue induit une prolifération des LB après association homophylique de CD150 ainsi qu une hausse de 38

sécrétion d IgM quand les LB sont traités avec la partie soluble (extracellulaire) de CD150 [173]. CD150 a aussi été montré comme étant important dans la différenciation des cellules NKT. En utilisant une souche de souris NOD déficiente en NKT, les auteurs ont identifié CD150 comme étant requis au bon développement de ces cellules. C est via des interactions homophyliques que CD150 participe à la différenciation des NKT. CD150 aiderait aussi au maintien de la tolérance au sein du système immunitaire, ces souris étant promptes à développer des maladies auto-immunes [189]. CD150 est aussi exprimé à la surface des DC et son expression augmente suite à l activation des DC via une stimulation à l IL-1β, la liaison de CD40 à son ligand (CD40L), ou la reconnaissance de divers signaux microbiens [187, 190]. En utilisant un anticorps monoclonal spécifique de CD150, Bleharski et coll. ont démontré que des DC activées via CD40L et traitées avec cet anticorps étaient plus promptes à sécréter de l IL-12 et de l IL-8, mais pas de l IL-10 [187]. Cela suggère donc un effet proinflammatoire de CD150. Une autre étude a démontré que des interactions CD150/CD150 entre deux cellules (des DC ici), ont en revanche un effet anti-inflammatoire en inhibant la sécrétion d IL-12, IL-6 et TNF-α dans des DC activées via CD40L [168]. Cela pourrait indiquer que selon la façon dont la signalisation en aval de CD150 est déclenchée (un anticorps, des interactions entre deux molécules de CD150 sur des cellules, ou un autre ligand propre à CD150), cela modifierait la réponse passant de pro-inflammatoire à anti-inflammatoire selon le contexte. Les macrophages expriment CD150. Dans un modèle de souris déficiente pour CD150 (CD150 KO), une stimulation (LPS et/ou IFN-γ) des macrophages aboutit à une sécrétion de cytokines pro-inflammatoires (IL-12, IL-6, TNF-α) et du NO moins importantes contrairement aux cellules provenant de souris contrôles [191]. CD150 ne semble pas impliqué dans la phagocytose ou la réponse au CpG et au peptidoglycane [191]. Les auteurs de cette étude sont allés plus loin en étudiant les allergies pulmonaires pour déterminer quelle pouvait être l implication de CD150 dans ce phénomène. En utilisant des souris KO pour CD150, ils ont établi un modèle d allergie (avec de l OVA) et des lavages 39

bronchoalvérolaires (BAL). Le BAL permet de récupérer toutes les cellules présentes dans les bronches, et donc de témoigner d un recrutement de cellules inflammatoires. En mimant l allergie par stimulation, le nombre d éosinophiles récupérés par BAL est diminué dans les souris CD150 KO comparé aux souris sauvages [192]. Des coupes de tissus montrent clairement une inflammation moins importante dans les souris CD150 KO stimulées. Les souris CD150 KO ne sont pas non plus capables de produire des cytokines de type Th1 (TNF-α et IL12p70 ici) et Th2 (IL-10, IL-4 et IL-13) en réponse à un allergène à la différence des souris sauvages. Toutes ces données ont amené Wang et coll. à conclure que CD150 est nécessaire à la réponse inflammatoire locale dans le cas des allergies pulmonaires, probablement à cause de ses fonctions dans les réponses inflammatoires des macrophages, ou à cause de son impact sur l activation des LT et LB [191, 192]. I. E. 3. CD150 et les infections Il a été caractérisé depuis le début des années 2000 que CD150 est le récepteur du virus de la rougeole [193]. En effet, l hémagglutinine du virus est capable d interagir avec CD150 et ainsi permettre l entrée du virus dans les cellules infectées [193]. De nombreuses études se sont donc intéressées à l impact de la liaison du virus à CD150. La liaison du virus à CD150 sur les DC va causer une immunosuppression et une incapacité des DC à activer les LT [169]. Après infection par le virus de la rougeole de DC murines exprimant le récepteur humain CD150, celles-ci ne sont plus capables de surexprimer des molécules de co-stimulation comme CD86, CD80, CD40, le CMH-II, ni d induire une prolifération des LT. Ce mécanisme n est pas dépendant de la présence de CD150 à la surface des LT, suggérant que ce défaut vient directement des DC, et donc des voies de signalisation pouvant être activées après liaison du virus à CD150 [169]. Une autre étude a démontré que la liaison de l hémagglutinine du virus de la rougeole à CD150 induit une inhibition de la sécrétion d IL-12 après activation d un signal TLR4 dans les DC [194]. Dans le cas d une infection à M. tuberculosis, il a été montré que CD150 semble promouvoir la réponse immunitaire de type Th1 via la production d IFN-γ [195]. L augmentation de la 40

production d IFN-γ est en fait dépendante de l activation de CREB (un facteur de transcription), elle-même dépendante de la liaison de l hémagglutinine du virus à CD150 [196]. Dans un modèle d infection par un parasite, Leishmania major, CD150 joue un rôle important dans le contrôle de l infection. Bien que des souris déficientes pour CD150 sur fond génétique Balb/c ne sont pas plus susceptibles à l infection que des souris contrôles, ce n est pas le cas pour des souris C57Bl/6 déficientes pour CD150. Les souris de type C57Bl/6 ont une réponse immunitaire orientée Th1, au contraire des Balb/c, qui vont promouvoir une réponse Th2 via la sécrétion d IL-4 notamment [197]. Les souris CD150 KO sur fond C57Bl/6 sont en effet incapables d éliminer L. major après infection [191]. La déficience en CD150 induit une diminution de la production de NO, d IL- 12 et de TNF-α. La diminution de l expression de ces composés pro-inflammation participe à la persistance de L. major dans les souris KO pour CD150. [191]. Plus récemment, une étude a étudié l impact d une déficience en CD150 au cours de l infection par le parasite Trypanosoma cruzi. En utilisant des souris KO pour CD150 (sur fond Balb/c), les auteurs ont démontré que ces souris étaient capables de survivre à une phase aigüe d infection au cours de laquelle les souris sauvages succombent [198]. La meilleure survie des souris KO pour CD150 s explique par une plus faible sécrétion d IFN-γ dans leur cœur. De plus, des macrophages infectés par T. cruzi provenant des souris KO pour CD150 expriment plus faiblement les ARNm de Ptgs2, Nos2, et de l arginase 1 [198]. Les DC provenant de ces mêmes souris produisent aussi moins d IL-12 et d IFN-γ après infection que des DC provenant de souris sauvages. Il est intéressant de noter que dans les macrophages et DC provenant des souris CD150 KO, le parasite se réplique moins rapidement que dans des cellules provenant de souris sauvages [198]. CD150 est donc capable dans le cas d infection virale, parasitaire, ou bactérienne de promouvoir, ou au contraire inhiber la réponse immunitaire. Dans la plupart des cas, cela concerne la réponse de type Th1 dépendante de l IL-12 et de l IFN-γ. 41

I. E. 4. CD150, un récepteur bactérien? Une dernière étude faite dans des macrophages provenant de souris CD150 KO (sur fond C57BL/6) tend à prouver le rôle de CD150 dans les réponses aux infections. En utilisant des macrophages déficients pour CD150 KO, après infection par E. coli, ceux-ci présentent un défaut d induction de NOX2 (NADPH Oxydase 2, nécessaire à la production de ROS) [199]. CD150 serait capable d interagir avec le complexe Vps34 Vps15 beclin-1, ce qui lui permettrait de contrôler la production de phosphatidylinositol-3-phosphate (PtdIns(3)P) dans la membrane du phagosome. Le PtdIns(3)P régule l activité de NOX2 dans le phagosome. CD150 pourrait ainsi permettre le bon fonctionnement du phagolysosome via la production de PtdIns(3)P et donc de ROS [199]. Le rôle de CD150 dans la maturation du phagosome est bien démontré dans des macrophages déficients pour CD150, en effet ceux-ci ne sont plus capables de devenir matures et donc d éliminer E. coli. De façon intéressante, en infectant les macrophages par Staphylococcus aureus, les auteurs n ont vu aucun impact sur la sécrétion de NO des macrophages déficients en CD150, ou sur la maturation du phagolysosome. Ces résultats suggèrent que le rôle de CD150 pourrait être restreint à certains types de bactéries à gram négatif. CD150 est aussi capable de reconnaître et lier deux protéines d E. coli, OmpC et OmpF, ainsi qu une protéine (non identifiée pour le moment) de Salmonella. La liaison de OmpC et OmpF à CD150 induit une augmentation de l expression CD150 à la membrane des cellules [199, 200]. Cette fonction est étendue à toute la famille des récepteurs SLAM. En effet, d après certaines de leurs données non publiées SLAMF6 (ou Ly108) serait aussi capable de reconnaitre E. coli (mais toujours aucune réponse à S. aureus). De plus SLAMF2, un autre membre de cette famille, est un des récepteurs à FimH, une lectine présente sur le pili d entérobactérie. La conclusion est que les récepteurs SLAMF seraient une nouvelle famille de récepteurs microbiens. 42

II. Résultats 43

II. A. RESUME DES ACTIVITES Pendant mon master 2 effectué au sein du laboratoire de Jean-Pierre Gorvel, j ai étudié les interactions entre Brucella et les DC. En poursuivant en thèse, j avais souhaité continuer à m intéresser à l interaction entre des cellules hôtes et acteurs majeurs du système immunitaire et une bactérie, Brucella. J ai travaillé sur trois projets différents pendant ma thèse : Le premier d entre eux commence avec un crible génétique qui nous a permis d identifier CD150 comme étant surexprimé dans des DC stimulées avec du CβG. En parcourant la littérature, on s est aperçu que CD150 pouvait être un récepteur bactérien, et nous avons donc décidé d étudier son rôle dans l infection à Brucella. Par la suite, j ai travaillé sur le CβG, cette molécule capable d activer les DC, de moduler le contenu en cholestérol des membranes et permettre un trafic correct de la BCV dans les cellules. Le dernier projet est en continuation avec mon projet de master 2 pour lequel j avais commencé à travailler sur BtpB. Pendant ma thèse, j ai pu finir des expériences sur ce projet et démontrer que cette protéine participe au contrôle de l activation des DC durant l infection. Le point commun et l intérêt pour moi de ces projets est le lien entre les DC et comment la bactérie, via ses facteurs de virulence (CβG), ou ses protéines (Omp25, BtpB) va moduler le système immunitaire pour favoriser l établissement d une pathologie chronique et la survie de Brucella. 44

II. B. OMP25 SE LIE A CD150 POUR CONTROLER L ACTIVATION DES DC DURANT L INFECTION PAR BRUCELLA II. B. 1. Introduction Grâce à des études de transcriptomique faites sur des DC humaines stimulées avec du CβG, nous avons identifié des gènes étant très exprimés, parmi lesquels, CD150. La littérature nous a appris que ce gène avait un rôle dans la co-stimulation des LT, dans la signalisation au sein de ces cellules et d autres cellules du système immunitaire conduisant à l activation ou inhibition de leurs fonctions immunitaires [168, 171, 187, 191]. Par ailleurs, l étude de Berger et coll. a démontré que CD150 était un récepteur pour des protéines membranaires d E. coli, potentiellement de Salmonella, mais pas des bactéries à gram positif [199]. Différentes études menées sur des infections dans des modèles de souris déficientes pour CD150 KO ont montré l importance que ce récepteur peut avoir dans les réponses immunitaires [191, 196, 199, 200]. Nous avons donc décidé d étudier le rôle de CD150 dans l infection à Brucella, et essayer d identifier une ou plusieurs protéines bactérienne qu il serait capable de reconnaître. 45

II. B. 2. Résultats Article en préparation CD150 interacts with Brucella Omp25 and controls Brucella infection in vivo. Clara Degos 1,2,3, Alexia Papadopoulos 1,2,3, Ignacio Moriyón 4, Yusuke Yanagi 5, Stéphane Méresse 1,2,3 and Jean-Pierre Gorvel 1,2,3* 1: Aix-Marseille Université UM 2, Centre d'immunologie de Marseille-Luminy, Marseille, France 2: INSERM U 1104, Marseille, France 3: CNRS UMR 7280, Marseille, France 4: Departamento de Microbiología e Instituto de Salud Tropical, Universidad de Navarra, 31008 Pamplona, Spain 5: Department of Virology, Faculty of Medicine, Kyushu University, Fukuoka 812-8582, Japan Running title: CD150 controls Brucella infection *Corresponding author: gorvel@ciml.univ-mrs.fr Keywords: Brucella, Omp25, CD150, inflammation, dendritic cells, NF-κB Abbreviations: dendritic cells (DC), endoplasmic reticulum (ER), interferon (IFN), interleukine (IL), outer membrane protein (Omp), Escherichia coli (E. coli), Brucella abortus (B. abortus), membrane extracts (OM), ovalbumin (OVA), post-infection (p.i), intraperitoneal (IP), competitive index (CI), outer membrane fragments (OMF), SLAMadaptor protein (SAP), ITSM (Immunoreceptor tyrosine-based switch motif). 46

Abstract: Brucella is an intracellular pathogenic bacterium responsible for brucellosis. One of the main strategies for establishing a chronic disease is based on the control of immune response. CD150 is a receptor for Escherichia coli outer membrane proteins and is known to regulate T cell, B cell, macrophage and dendritic cell (DC) activation. We identified CD150 in a transcriptomic assay of human DC treated with Brucella antigens and we studied the role of CD150 in Brucella infection. Using a mouse model, we discovered that CD150 is required to limit NF-κB translocation in infected DC. This membrane receptor is also involved in controlling bacterial replication in vivo in mice. Finally, we demonstrate that CD150 is a receptor for Omp25, a major Brucella outer membrane protein and we suggest that it constitutes a new receptor for Brucella involved in the inhibition of immune response. Introduction Brucella is an intracellular bacterium responsible for brucellosis, one of the most common zoonosis. In human it causes a debilitating febrile illness and in mammals it is responsible for abortion and sterility leading to economic losses [3, 4, 11]. One of the main aspects of Brucella pathogenesis is its ability to evade the immune system detection [10, 82]. DC have been widely studied in brucellosis and proven to be important for the induction of immune responses and also for providing a safe replication niche for the bacterium [23, 92, 96]. DC play a central role in the induction of both innate and adaptive immune responses by activating T cells during the course of infection [23, 92, 96]. Brucella controls DC maturation and TLR signaling by the action of at least two Brucella proteins BtpA and BtpB, to counteract the immune response [23, 201]. In vitro, infection with Brucella leads to a mild activation of DC regarding co-stimulatory molecules expression and 47

% o f C D 2 5 + T c e lls A C D 1 5 0 M e d ia n F lu o re s c e n c e 3 0 0 0 2 0 0 0 1 0 0 0 0 * * * * * * * * B P B S E.c o li L P S B. a b o rtu s w t O M B. a b o rtu s o m p 2 5 O M M e d ia n F lu o re s c e n c e 3 0 0 0 2 0 0 0 1 0 0 0 0 * ns ns C D 8 6 C D 4 0 ns ns * * * ns ns ns M H C II Figure 15: CD150 expression onto BMDC following stimulation with Brucella membrane extracts. A. BMDC have been stimulated for 16 h with PBS (negative control white bars), 100 ng/ml of E. coli LPS (activation control light grey bars), 10 µg/ml of B. abortus wt OM (dark grey bars) or 10 µg/ml of B. abortus omp25 OM (blanc bars). CD150 expression was analyzed by flow cytometry and is represented here by its median fluorescence. B. BMDC have been treated as described above and CD86, CD40 and MHC-II expression were analyzed. Their expression is represented by the median fluorescence. At least 100,000 CD11c + (DC marker) have been analyzed. This experiment has been repeated 4 times independently. p<0,005 are denoted *, p<0,01: ** and p<0,005: ***, non significant data were denoted «ns». A PBS LPS E. coli B. abortus wt OM B. abortus omp25 OM 4.7 3 % 1 8.5 % 6.5 3 % 1 7.4 % 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 C D 2 5 B 3 0 * * 2 0 1 0 P B S E.c o li L P S B. a b o rtu s w t O M B. a b o rtu s o m p 2 5 O M 0 Figure 16: CD25 expression in CD4 + T cells stimulated by differently activated BMDC. BMDC have been stimulated for 16 h with PBS (negative control white bars), 100 ng/ml of E. coli LPS (activation control light grey bars), 10 µg/ml of B. abortus wt OM (dark grey bars) or 10 µg/ml of B. abortus omp25 OM (blanc bars). Cells were then incubated with OVA (50 µg/ml) for 4 h before co-culture with T cells at a ratio of 1 : 4 (DC : T) for 3 days. CD25 expression was analyzed by flow cytometry. Results represent the percentage of CD25 + cells in graph (A) and in histogram (B). At least 20,000 CD3 + and CD4 + (CD4 + T cells markers) events were collected. This experiment has been repeated 4 times. p< 0,05: *.

cytokine secretion [23, 44, 47]. In vivo, DC seems to help Brucella dissemination in the absence of macrophages [202]. DC are also required to control bacterial growth through the secretion of interferon-γ (IFN-γ), interleukin-12 (IL-12) and and i-nos [92, 96]. Outer membrane proteins (Omp) are also Brucella components capable of modulating immune responses. Omp16 is recognized by DC and triggers an immune response by the secretion of cytokines such as tumor necrosis factor alpha (TNF-α) and IL-12, and the overexpression of co-stimulatory molecules such as CD80, CD86 and CD40 [43]. Another example is Omp19, a lipoprotein, which was shown to dampen antigen presentation and MHC-II expression in IFN-γ-activated human monocytes [42]. Omp25 (or Omp3a) also seems to play a role in the immune response against the bacterium. Indeed, a Brucella mutant for this protein ( omp25) seemed to induce human DC and macrophage activation, IL-12 and TNF-α secretion [47, 48] although it is not attenuated in human DC, epithelial cells (HeLa), macrophages (Raw, THP-1) and neutrophils [44, 45]. In vivo, the role of Omp25 needs further work to demonstrate its implication in immune responses [45, 49] even though this membrane protein has been considered as a potential vaccine candidate due to its immunogenic properties [203-206]. Recently, we identified CD150 (or signaling lymphocyte activation molecule (SLAM)) in a transcriptomic analysis as an upregulated gene in human DC stimulated with Brucella cyclic glucan [22]. CD150 belongs to a family receptor called SLAMF [207]. This receptor is a selfligand that triggers T cell activation [171, 172]. CD150 has been involved in immune response to various infections. While it has been shown that this receptor is a ligand for measles virus, CD150 can also control DC and T cell immune responses during the course of the infection [169, 193]. It has also been shown to play a role as activator or inhibitory of the 48

% o f p ro life ra tin g T c e lls PBS LPS E. coli B. abortus wt OM B. abortus omp25 OM A 0.4 % 2 7.9 % 1 7.2 % 2 7.2 % 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 10 1 10 2 10 3 10 4 10 5 C F S E B p = 0.0 6 3 0 2 0 1 0 P B S E.c o li L P S B. a b o rtu s w t O M B. a b o rtu s o m p 2 5 O M 0 Figure 17: Proliferation of CD4 + T cells stimulated by differently activated BMDC. BMDC have been stimulated for 16 h with PBS (negative control white bars), 100 ng/ml of E. coli LPS (activation control light grey bars), 10 µg/ml of B. abortus wt OM (dark grey bars) or 10 µg/ml of B. abortus omp25 OM (blanc bars). Cells were then incubated with OVA (50 µg/ml) for 4 h. T cells were stained with CFSE and then co-cultured with BMDC at a ratio of 1 : 4 (DC : T) for 3 days. CFSE fluorescence was analyzed by flow cytometry. Each cell cycle division will lead to a decrease of CFSE fluorescence. Results represent the percentage of CD4 + T cells proliferating in graph (A) and in histogram (B). At least 20,000 CD3 + and CD4 + (CD4 + T cells markers) events were collected. This experiment has been repeated 4 times. p< 0,05: *.

immune system, in other infectious models [195, 197, 198]. Recent studies showed that CD150 was involved in controlling the maturation of phagosome during Escherichia coli (E. coli) infection [208] and that its expression was upregulated following the binding of two E. coli Omp: OmpC and OmpF [200, 208]. Here, we demonstrate that CD150 expression was increased by interaction with the Brucella Omp25 membrane protein and this led to an inhibition of T cell activation by BMDC. We show a Brucella abortus omp25 mutant ( omp25) strongly activates DC showing that this outer membrane protein plays an important role in the control of immune responses. CD150 blockade by specific peptide or the use of CD150 KO leads to an increase of NF-κB activation in wt Brucella-infected BMDC suggesting a role for this receptor in controlling the immune response against Brucella infection. We also demonstrate that CD150 controls Brucella replication in vivo. Finally we showed that Omp25 binds CD150 using a pull-down assay. Results Brucella outer membrane extracts induce CD150 expression in BMDC through Omp25. CD150 expression in macrophages was shown to increase by the binding of OmpC and OmpF from E. coli [200]. We therefore investigated the impact of Brucella Omp onto CD150 expression. We used BMDC as a model since CD150 is expressed at the surface of activated DC and these cells are known to play an important role in the Brucella immune response. BMDC were incubated with Brucella wild type membrane extracts (B. abortus wt OM). 16 h after exposure, CD150 expression was assessed by flow cytometry. We observed an upregulation of CD150 expression following stimulation with B. abortus wt OM compared to 49

L o g C F U /m l A 6 B. a b o rtu s v irb B. a b o rtu s w t B. a b o rtu s o m p 2 5 4 2 B 0 2 0 h 4 0 h T im e p o s t-in fe c tio n 6 0h w t E R M erg e o m p 2 5 E R M erg e Figure 18: Brucella omp25 replicates as the wt strain within the ER in BMDC. A. Replication of wt Brucella (square), omp25 (triangle) and virb mutant ( virb, circle) within BMDC for 48 h. The mean of 5 independent experiments is represented here. B. Representatives pictures of confocal microscopy of Brucella intracellular trafficking at 24 h p.i. BMDC were stained with an anti-calnexin (ER, red), and anti-lps (green). Scale: 10µm. This experiment has been done 3 times independently.

non-stimulated cells (PBS) or cells stimulated with E. coli LPS (activation control) (Fig. 15A) [209]. Interestingly, membrane extracts (OM) from a omp25 deletion mutant (B. abortus omp25) failed to induce CD150 expression, which was then expressed at basal level (Fig. 15A). The decrease in CD150 expression between the OM from Brucella wt and omp25 was related to the absence of Omp25 rather than a BMDC activation defect (Fig. 15B). Omp25 inhibits CD4 + T cell activation. We then wonder whether the decrease in CD150 expression could impact the ability of DC to stimulate T cell responses. We used CD4 + T cells from OTII mice. These mice carry CD4 + T cells with a specific TCR for ovalbumine (OVA). In the presence of activated DC carrying MHC-II presenting antigen (here OVA), these T cells produce IL-2 receptor (CD25) and proliferate. BMDC were stimulated with PBS (negative control), E. coli LPS, B. abortus wt OM or B. abortus omp25 OM. We show that in the presence of OVA non-stimulated BMDC (white bar) and BMDC stimulated with B. abortus wt OM were poor inducers of CD25 expression in CD4 + T cells. In contrast, LPS-stimulated or omp25 OM-stimulated DC allowed 3 times more expression of CD25 by T cells than B. abortus wt OM (Fig. 16). T cell proliferation was stimulated by LPS-stimulated BMDC (27,9 % of proliferating T cells), B. abortus wt OM (17,2 % of proliferating T cells) and B. abortus omp25 OM (27,2 % of proliferating T cells). CD4 + T cells co-cultivated with PBS-stimulated BMDC were not able to proliferate (Fig. 17). Omp25 inhibited the up-regulation of CD25 expression in BMDC-stimulated CD4 + T cells (about 3 times less expression) and in a lesser extent their proliferation (1.5 less proliferation) (Fig. 17). 50

A R a tio M e d ia n F lu o re scence 1.5 * 1.0 0.5 * * B. a b o rtu s w t/w t B.a b o rtu s o m p 2 5 /w t R a tio M e d ia n B F lu o re scence 2.5 2.0 1.5 1.0 0.5 * * * * * * * 0.0 C D 4 0 C D 8 0 C D 8 6 M H C II C D 1 5 0 0.0 C D 4 0 C D 8 0 C D 8 6 M H C II C D 1 5 0 Figure 19: Co-stimulatory molecules and MHC-II expression onto BMDC after Brucella infection. BMDC were infected with wt Brucella (white bars) or omp25 (grey bars) for 8 h (A) or 24 h (B) with a MOI of 30. Cells were harvested and stained for flow cytometry analysis. Ratio of median fluorescence were shown here for the different molecules as indicated on the x axis. At least 100,000 CD11c + events were collected. This experiment has been done 3 times independently. p< 0,05: *, p< 0,01: **, p< 0,005: ***.

Brucella omp25 replicates within the ER of BMDC. We first analyzed omp25 replication within BMDC in comparison with the wt strain (Fig. 18A). We observed that the mutant replicated as much as the wt strain while a virb mutant was not able to persist for more than 48 h (Fig. 18A). omp25 replicated within the ER as shown at 24 h post-infection (p.i) and as previously described for the wt strain [23] (Fig. 18B). The mutant also followed the same intracellular trafficking as the wt strain in BMDC at early time point after infection (data not shown). Therefore, no difference in the intracellular trafficking was observed between the wt Brucella strain and the omp25 mutant. Omp25 controls BMDC activation upon infection. It has previously been published that Brucella infection in BMDC leads to an intermediate level of activation compared to other bacterial pathogen infections [23]. Therefore, we checked omp25-infected BMDC phenotype at both 8 h and 24 h p.i. Costimulatory molecule expression (CD80, CD86, CD40 and CD150) and MHC-II expression were analyzed by flow cytometry (Fig. 19). At 8 h p.i, omp25-infected BMDC exhibited a higher expression of CD80, CD40 and CD86 compared to wt-infected BMDC (Fig. 19A). This difference increased at 24 h p.i with almost twice more expression of co-stimulatory molecules and CD150. However, MHC-II expression was not statistically different between omp25-infected BMDC and wt-infected BMDC (Fig. 19B). Another aspect of BMDC activation is the ability of NF-κB to translocate within the nucleus to engage pro-inflammatory gene transcription. At 2 h p.i., we observed twice more NF-κB translocation in the nucleus of omp25 infected BMDC than that of Brucella wt infected 51

% o f B M D C w ith N F - B tra n s lo c a tio n in th e n u c le u s A T O P R O -3 N F - B C D 1 1 c w t o m p 2 5 B 7 0 6 0 5 0 4 0 3 0 2 0 1 0 0 * U n in fe c te d B. a b o rtu s w t B.a b o rtu s o m p 2 5 Figure 20: NF-κB translocation within infected BMDC. A. Representatives pictures from confocal microscopy of infected BMDC with wt Brucella (upper panel) or omp25 (lower panel) for 2 h with MOI of 30. Cells were fixed and labeled with TOPRO-3 for the nucleus (yellow), anti-p65 (NF-κB, red), anti-lps (green) and anti-cd11c for DC (cyan). Scale: 10 µm. B. Quantification of NF-κB translocation in the nucleus of infected cells. BMDC were treated as described above. The percentage of cells containing NF-κB within the nucleus is indicated on the graph. Uninfected cells are represented in white while wt infected cells are in light grey and mutant infected cells in dark grey. At least 50 cells in 4 independent experiments were counted each time. p< 0,05: *.

BMDC or the non-infected cells (Fig. 20). As previously shown, Brucella wt infection did not induce a high level of translocation of NF-κB within the nucleus (Fig. 20B) [201]. Considering the different levels of NF-κB translocation using the two strains we investigated the gene expression profile induced by either the wt strain or the mutant strain (Fig. 21). We showed that BMDC infection with wt Brucella induced at both 6 h and at 24 h the overexpression of IL-6, IL-12β, CCL-2, IL-1β, TNF-α, KC, NOS-2, PTGS-2, and IFN-β mrna (Fig. 21). Interestingly, omp25 infection induced the over-expression of IL-6, IL-12β, CCL- 2, IL-1β, NOS-2, PTGS-2 mrna but in a higher amount than the wt strain. We then analyzed the cytokine and chemokine secretion induced by wt Brucella and omp25 mutant at 8 h, 24 h, and 32 h p.i (Fig. 22). omp25- BMDC produced more IL-12, IL-1β, TNF-α, CCL-2, IL-6 and IFN-γ than non-infected cells or wt-infected cells as shown at 32 h p.i (Fig. 22). In conclusion, we show that Omp25 controls BMDC activation during infection by inhibiting co-stimulatory molecules expression, NF-κB translocation in the nucleus, pro-inflammatory mrna expression and pro-inflammatory cytokine secretion. CD150 blockade does not affect replication, intracellular trafficking of Brucella within BMDC and DC co-stimulatory molecule expression. To characterize the role of CD150 in Brucella infection we used control and blocking peptides prior to BMDC infection. CD150 blockade affected neither the ability of Brucella to invade nor to replicate within BMDC (Fig. 23A) nor its ER intracellular localization (Fig. 23B). Bacteria (wt or omp25) replicated with the same efficiency within the ER of infected cells at 24 p.i (Fig. 23B). 52

F o ld in c re a s e F o ld in c re a s e F o ld in c re a s e 5 0 IL 6 1 2 5 IL 1 2 1 5 C C L 2 4 0 3 0 2 0 1 0 1 0 0 7 5 5 0 2 5 F o ld in c re a s e 1 0 5 0 6 h IL 1 2 4 h 0 6 h T N F 2 4 h 0 6 h K C 2 4 h 4 0 8 5 0 F o ld in c re a s e 3 0 2 0 1 0 F o ld in c re a s e 6 4 2 F o ld in c re a s e 4 0 3 0 2 0 1 0 0 6 h 2 4 h 0 6 h 2 4 h 0 6 h 2 4 h 2 0 0 N O S 2 1 4 0 P tg s 2 2 5 0 IF N F o ld in c re a s e 1 7 5 1 5 0 1 2 5 1 0 0 7 5 5 0 2 5 0 6 h 2 4 h 1 2 0 1 0 0 8 0 6 0 4 0 2 0 0 6 h 2 4 h F o ld in c re a s e 2 0 0 1 5 0 1 0 0 5 0 0 6 h 2 4 h B. a b o rtu s w t B. a b o rtu s o m p 2 5 Figure 21: mrna expression of different pro-inflammatory genes in infected BMDC. BMDC were infected (MOI: 30) with wt (black bars) or mutant for omp25 (grey bars) or with PBS as negative control for 6 h or 24 h. Cells were then harvested and RNA was extracted. mrna expression of IL- 6, IL-12β, CCL2, IL-1β, TNF-α, KC, NOS2, PTGS2, and IFN-β was assessed. Data were normalized onto housekeeping gene (HPRT) and fold increase was calculated with the uninfected cells. Basal expression level is indicated with a dashed line. An induction of 2 or more is considered as significant. Mean ± standard deviation of 3 independent experiment is represented here.

We also assessed the expression of co-stimulatory molecules and MHC-II by infected BMDC stimulated with blocking or control peptides. No difference was observed between BMDC stimulated with the control peptide or with the blocking peptide at 8 h and 24 h p.i (data not shown). CD150 blockade enhances NF-κB translocation in the nucleus of wt Brucella-infected BMDC. BMDC incubated with CD150 blocking peptide (Fig. 24A left panel Fig. 24B hatched bars) and infected with Brucella wt (Fig. 24A upper panel Fig. 24B grey bars) were more prone to induce NF-κB translocation than BMDC treated with control peptide (Fig. 24A right panel Fig. 24B full bars). Interestingly, the percentage of NF-κB translocation within the nucleus of BMDC incubated with CD150 blocking peptide reached the level of BMDC infected with the omp25 mutant (Fig. 24B). CD150 blockade did not affect NF-κB translocation within cells infected with omp25. We also assessed the secretion of pro-inflammatory cytokines upon CD150 blockade. Pro-inflammatory cytokine secretion was not impacted by the blockade of CD150 at 24 h or 48 h p.i in infected BMDC no matter what Brucella strain was used (Fig 25). CD150 controls NF-κB translocation in the nucleus of Brucella wt-infected BMDC. Using CD150 -/- mice (CD150 KO) we confirmed the results obtained above. Indeed, in BMDC harvested from CD150 KO mice and infected by Brucella wt and omp25, both strains induced an efficient NF-κB translocation in the nucleus, which reached the same percentage as the BMDC from normal mice infected with omp25 showing that CD150 presence or absence did not affect NF-κB translocation in omp25 infected DC (Fig. 26). 53

IL -1 (p g /m l) 1 0 0 0 * 3 0 0 0 * * T N F (p g /m l) 7 5 0 5 0 0 2 5 0 C C L -2 (p g /m l) 2 5 0 0 2 0 0 0 1 5 0 0 1 0 0 0 5 0 0 * * 0 2 0 0 0 8 h 2 4 h 3 2 h * 0 3 0 8 h 2 4 h 3 2 h U n in fe c te d B. a b o rtu s w t B.a b o rtu s o m p 2 5 IL 6 (p g /m l) 1 5 0 0 1 0 0 0 5 0 0 IF N (p g /m l) 2 0 1 0 0 8 h 2 4 h 3 2 h 0 2 4 h 3 2 h IL 1 2 p 7 0 + p 4 0 (p g /m l) 6 0 0 4 0 0 2 0 0 0 8 h ** 2 4 h ** *** 3 2 h 1 0 0 0 8 0 0 6 0 0 4 0 0 2 0 0 0 8 h 2 4 h *** 3 2 h Figure 22: Pro-inflammatory cytokine secretion upon BMDC infection. BMDC were infected at MOI of 30 with Brucella wt (light grey), omp25 (dark grey) or PBS as negative control (white) for 8 h, 24 h or 32 h. Culture supernatants were harvested and cytokine secretion was assessed by cytometric bead assay (CBA, IL-6, TNF-α, CCL-2 et IFN-γ) or ELISA (IL-12 and IL-1β). Mean ± standard deviation of 3 independent experiment is represented here. p< 0,05: *, p< 0,01: **, p<0,001: ***.

Role of Omp25 during the infection process. Considering the discrepancy in literature on Omp25 role in vivo [45, 49], we decided to study the replication of this mutant strain in various mouse models such as C57BL/6 mice, known the be brucellosis resistant and Balb/c mice, which are susceptible to the infection [147]. Mice were infected with 1.10 6 CFU intraperitoneally (IP). We measured bacteria growth in the spleen and liver of infected animals and weight the organs at 5 days p.i (acute phase of brucellosis) and at 60 days p.i corresponding to the chronic phase. At 5 days p.i no statistically difference in the replication or organ weights were observed between wt and omp25 infected mice (Fig. 27). Bacterial growth and organ weights were higher in Balb/c mice compared to C57BL/6 mice (Fig. 27), suggesting a less controlled infection in Balb/c mice as previously published [92, 210]. We also performed a competitive index (CI) experiment to assess the virulence of Omp25 [211]. A CI below 1.0 means a slower and lesser growth of the mutant strain compared to the wt. In both C57BL/6 and Balb/c mice, omp25 strain was attenuated in the liver and mesenteric lymph node after IP infection (Fig. 28). Interestingly, in resistant mice (C57BL/6), omp25 was not attenuated in the spleen (Fig. 28). We then assessed whether the omp25 mutant would be attenuated in the chronic phase (at 60 days p.i.). Although we did not observe any difference in term of growth between wt and omp25 strains (Fig. 29A), organs weight was increased in omp25 infected Balb/c mice, suggesting a high level of inflammation (Fig. 29B). No difference in replication between C57BL/6 and Balb/c was observed, while organ weight was higher in Balb/c mice than C57BL/6 (Fig. 29B). 54

A 8 L o g C F U /m l 6 4 2 B M D C tre a te d w ith c o n tro l p e p tid e B. a b o rtu s w t B.a b o rtu s o m p 2 5 B M D C tre a te d w ith C D 1 5 0 b lo c k in g p e p tid e B. a b o rtu s w t B.a b o rtu s o m p 2 5 0 2h 8h 2 4h 3 2h B wt La p ER Merge o p 5 Figure 23: Brucella replication and intracellular localization in BMDC after CD150 blockade. A. BMDC were treated with 10 µg/ml of control peptide (full line) or the blocking peptide for CD150 (dashed line) for 3 h prior to infection with the wt strain (square) or omp25 (triangle). CFU counts were enumerated at 2 h, 8 h, 24 h and 32 h p.i. The mean of 2 independent experiments is represented. B. Representative pictures of confocal microscopy of infected BMDC with wt Brucella (upper panel) or omp25 (lower panel) for 24 h and treated with a blocking peptide for CD150. Cells were stained for calnexin (ER, blue), Lamp-1 (lysosomes, red), Brucella (green). Scale: 10 µm. This experiment has been repeated 3 times independently.

We finally used IFN-γ KO mice as a lethal model to assess the virulence of the mutant. We infected IP mice with 1.10 6 CFU and checked the survival of the mice. omp25 infected mice died faster than wt infected mice (25 days versus 31 days) (Fig. 30). All these data suggest that omp25 induce a higher inflammation in vivo and could participate in Brucella virulence, but not in its survival and replication. CD150 controls Brucella replication in vivo. To determine the role of CD150 in vivo infection, wt mice and CD150 KO mice were infected and checked at 8 days p.i. In the spleen of CD150 KO mice infected with Brucella wt bacterial replication was higher than in normal mice whereas no difference was observed in the replication of Brucella within the liver. The spleen weight of CD150 KO mice Brucella wt infected with Brucella wt strain was also higher than in control mice (Fig. 30B). No difference in bacterial replication was detected when we infected C57BL/6 or CD150 KO mice with the omp25 mutant (Fig. 30A). We also performed CI experiments using CD150 KO mice and observed that omp25 strain was not attenuated in the liver and in the spleen (Fig. 32). In contrast, the ratio was much higher than 1.0 (respectively 3.5 and 2.5 for the spleen and liver), meaning that within the organs the omp25 mutant strain was more efficient to replicate inside CD150 KO mice compared to the wt Brucella strain (Fig. 32). Brucella Omp25 binds CD150. In order to determine whether Omp25 was a ligand for CD150 we constructed a plasmid coding for myc in N-terminal fused to the second and third exons of murine CD150 (which represents the extracellular part of CD150). A plasmid containing a Salmonella protein (SifA) fused to myc was used as a control. After transfection in COS-7, proteins were extracted and a 55

C e lls w ith N F - B in th e n u c le u s (% ) A BMDC treated with o trol peptide BMDC treated with CD 5 lo ki g peptide TOPRO-3 NF-κB CD11c wt wt omp25 omp25 B * n s 5 0 4 0 * n s T re a te d w ith c o n tro l p e p tid e 3 0 2 0 1 0 T re a te d w ith C D 1 5 0 b lo c k in g p e p tid e 0 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 Figure 24: NF-κB translocation within infected BMDC after CD150 blockade. A. Representatives pictures from confocal microscopy of infected BMDC with wt Brucella (upper panel) or omp25 (lower panel) for 2 h with MOI of 30. Prior to infection cells were treated with 10 µg/ml of control peptide (left panel) or CD150 blocking peptide (right panel). Cells were fixed and labeled with TOPRO-3 for the nucleus (yellow), anti-p65 (NF-κB, red), anti-lps (green) and anti-cd11c for DC (cyan). Scale: 10 µm. B. Quantification of NF-κB translocation in the nucleus of infected cells. BMDC were treated as described above. The percentage of cells containing NF-κB within the nucleus is indicated on the graph. Uninfected cells are represented in white while wt infected cells are in light grey and mutant infected cells in dark grey. Cells treated with the control peptide are indicated in full bars. At least 50 cells in 4 independent experiments are counted each time. p< 0,05: *.

pull-down was performed with an anti-myc antibody after incubation with Brucella wt OM or Brucella omp25. We show that Omp25 could be specifically pulled-down by CD150 (Fig. 33). Therefore, we concluded CD150 is a receptor able to interact with Brucella Omp25. Discussion Here we show for the first time the role of CD150 in Brucella infection and its association with Omp25, a major outer membrane protein of the Brucella envelope. A role of Omp25 regulating immune responses in monocyte-derived cells was previously proposed [47, 48]. Here, we show that Omp25 inhibits co-stimulatory molecule expression in Brucella-infected DC, NF-κB translocation within DC nucleus, pro-inflammatory gene expression and cytokine and chemokine secretion in infected DC. Interestingly, Omp25 seems to inhibit T cell activation induced by primed DC stimulated with Brucella membrane extracts, supporting an anti-inflammatory role of Omp25 during the infection process. This is in agreement with the fact that pro-inflammatory cytokines seem to be released in a higher amount in omp25-infected mice. Omp25 appears as a new antigen synthesized by Brucella to control immune responses. Omp25 can be added to the list of already identified Brucella proteins inhibiting DC immune response in addition to BtpA, BtpB, PrpA or wboa [23, 83, 106, 201]. For this reason we consider Omp25 as a virulence factor. Indeed, although we did not observe any difference in term of bacterial replication in wt- or omp25-infected mice, we show a clear attenuation of the mutant strain in CI experiment using wild type mice, suggesting that Omp25 plays a role in virulence in vivo in association with CD150. Indeed, co-infection with both wt Brucella and omp25 mutant in CD150 KO did not result in an attenuation of 56

IL -6 (p g /m l) M C P -1 (p g /m l) IL -12p70+p40 (pg/m l) A 1 5 0 0 8 0 0 0 4 0 0 0 T N F (p g /m l) 1 0 0 0 5 0 0 IL -6 (p g /m l) 6 0 0 0 4 0 0 0 2 0 0 0 M C P -1 (p g /m l) 3 0 0 0 2 0 0 0 1 0 0 0 0 0 0 IL -12p70+p40 (pg/m l) 8 0 0 6 0 0 4 0 0 2 0 0 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 U n in fe c te d B. a b o rtu s w t Treated w ith c o n tro l p e p tid e B. a b o rtu s o m p 2 5 Treated w ith C D 1 5 0 b lo c k in g p e p tid e U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 0 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 B 3 0 0 0 1 0 0 0 0 6 0 0 0 8 0 0 0 5 0 0 0 T N F (p g /m l) 2 0 0 0 1 0 0 0 6 0 0 0 4 0 0 0 2 0 0 0 4 0 0 0 3 0 0 0 2 0 0 0 1 0 0 0 0 0 0 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 8 0 0 Treated w ith c o n tro l p e p tid e Treated w ith C D 1 5 0 b lo c k in g p e p tid e 6 0 0 4 0 0 2 0 0 0 U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 Figure 25 : Pro-inflammatory cytokine secretion upon BMDC infection after CD150 blockade. BMDC were infected at MOI of 30 with Brucella wt (grey), omp25 (black) or PBS as negative control (white) for 24 h (A) or 48 h (B). Prior to infection, cells were incubated with 10 µg/ml of control peptide (full bars) or CD150 blocking peptide (hatched bars). Culture supernatants were harvested and cytokine secretion was assessed by cytometric bead assay (CBA, IL-6, TNF-α, CCL-2 et IFN-γ) or ELISA (IL-12). This experiment has been reproduced 4 times independently.

omp25 strain also suggesting a strong relationship between the bacterial and the host protein. This hypothesis is confirmed by showing that Omp25 binds CD150 in a pull-down assay. CD150 acts as a co-stimulatory molecule between DC and T cells [171]. In the absence of Omp25, Brucella outer membrane extracts were more potent in inducing T cell activation through primed DC. A possible mechanism would be that the binding of Omp25 to CD150 would prevent CD150 to form dimers and to transduce danger signals to the host cell. Therefore, Omp25 could be a tool expressed by the bacterium to control CD150 signaling and consequently T cell activation. Another question raised by this study is Omp25 accessibility to CD150. Omp25 is part of the outer membrane of Brucella and so should not be free outside the bacterium. However, release and shedding of outer membrane fragments (OMF) by Brucella are well described [212, 213]. During the course of infection, OMF are likely to be released from infected cells for CD150 encounter. Interestingly, Pollak et al. showed that OMF pre-treatment of human macrophages (THP-1) inhibits pro-inflammatory cytokine secretion after TLR agonist exposure or Brucella infection [50]. In addition, OMF are internalized by phagocytic and nonphagocytic cells leading to a down-regulation of MHC-II expression in THP-1, thereby impacting antigen presentation [50]. This mechanism could be used by the bacterium to both limit the inflammation and antigen presentation, but also to deliver Brucella proteins including Omps to surrounding cells. In the context of Brucella infection, CD150 might play a dual role. While CD150 seems to control in vivo bacterial replication, it also participates to the control and inhibition of inflammation by DC through Omp25 binding. It has previously been reported that IFN-γ is 57

C e lls w ith N F - B in th e n u c le u s (% ) A T O P R O -3 N F - B C D 1 1 c w t o m p 2 5 B * ns ns 5 0 * 4 0 3 0 B M D C fro m C 5 7 B L /6 m ic e 2 0 1 0 0 B M D C fro m C D 1 5 0 -/- m ic e U n in fe c te d B. a b o rtu s w t B. a b o rtu s o m p 2 5 Figure 26 : NF-κB translocation within infected CD150 KO BMDC. A. Representatives pictures from confocal microscopy of infected CD150 KO BMDC with wt Brucella (upper panel) or omp25 (lower panel) for 2 h with MOI of 30. Cells were fixed and labeled with TOPRO-3 for the nucleus (yellow), anti-p65 (NF-κB, red), anti-lps (green) and anti-cd11c for DC (cyan). Scale: 10 µm. B. Quantification of NF-κB translocation in the nucleus of infected cells. BMDC were treated as described above. The percentage of cells containing NF-κB within the nucleus is indicated on the graph. Uninfected cells are represented in white while wt infected cells are in light grey and mutant infected cells in dark grey. BMDC coming from CD150 KO mice are represented with hatched bars. At least 50 cells in 3 independent experiments are counted each time. p< 0,05: *.

crucial to control Brucella growth in vivo [149, 214]. Since CD150 regulates IFN-γ production by T cells in various infectious models [196, 198], we hypothesize that the higher bacterial replication observed in vivo in CD150 KO mice is due to a lack of IFN-γ production leading to a lack of Brucella growth control. It is known that CD150 binds several Omp from E. coli. Further work is necessary to determine whether it is the case for Brucella. In addition, it has recently been described that several receptors from the SLAM family are capable of binding and sensing various microbial components: CD150 binds OmpC, OmpF and the measles virus, slamf6 binds E. coli and Slamf2 binds FimH, a lectin from E. coli [207, 215]. Further investigations need to be done regarding the role of other Slamf receptors in brucellosis. This could help us to better understand the Brucella detection by the immune cells, the immune response to infection and could provide new therapeutic targets. CD150 and other SLAMF receptors are known to interact with SAP (SLAM adaptor protein). Indeed, SLAMF receptor cytoplasmic tail contains ITSM (Immunoreceptor tyrosine-based switch motif), which are used by both phosphatase (such as SHP-2) and kinase (as Fyn, recruited through SAP to CD150 tail for example) to control T cell activation. SAP association with Fyn would be required for the induction of a humoral response and cytokine production by Th2 cells [216]. SAP would also lead to an inhibition of CD8 + T cells and NK cytotoxicity by interacting with 2B4 (Slamf4) or CD150 [217-219]. Considering the role of SLAM receptors and SAP in controlling the immune response, further studies need to assess their potential role in brucellosis. 58

S p le e n L iv e r L o g C F U /o rg a n L o g C F U /o rg a n m g S p le e n m g L iv e r A 8 7 6 5 4 * * 8 6 4 * 3 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C B. a b o rtu s w t B. a b o rtu s o m p 2 5 2 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C B 3 0 0 S p le e n 1 8 0 0 L iv e r 1 6 0 0 2 0 0 1 4 0 0 1 0 0 1 2 0 0 1 0 0 0 0 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C 8 0 0 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C B. a b o rtu s w t B. a b o rtu s o m p 2 5 Figure 27: Bacterial growth and weight organs in wt and omp25 infected mice at 5 days postinfection. A. C57BL/6 and Balb/c mice (n=5) have been infected IP with 1.10 6 CFU. 5 days later organs are removed and CFU are counted. B. abortus wt infected mice are indicated with a circle and the ones infected with omp25 are indicated with a square. Mouse strain used is indicated in the x axis. B. Organs weight was measured for infected mice. Each symbol represents an animal and the median values are marked by horizontal bold lines. P < 0,05 : *.

The binding of OmpC and OmpF to CD150 leads to macrophage activation and efficient killing of E. coli during infection, through phagolysosome maturation for example [199]. Studies have pointed out a versatile role of CD150 after various engagement signals therefore leading to reverse effects either pro- or anti-inflammatory [168, 187]. This explains why E. coli Omp binding to CD150 triggers an efficient immune response while Brucella Omp25- CD150 interaction triggers an anti-inflammatory response. This may also have consequences on intracellular trafficking: we show that CD150 deletion does not affect Brucella intracellular trafficking and phagosome escape from endocytic pathway while intracellular trafficking is affected upon E. coli infection in CD150 KO cells [199]. Despite its inhibitory and anti-inflammatory properties, Omp25 has also been studied for its ability to induce a response against Brucella. Intradermal immunization of mice with the recombinant protein leads to antibodies production against Omp25 [204, 205]. After infection of immunized mice with Brucella, intradermal immunized mice are more potent in producing pro-inflammatory cytokines such as IL-12 and IFN-γ than non-immunized mice [204, 205]. These studies underline the potential role of Omp25 as a target for the immune system. Although in the context of infection with no prior immunization, Omp25 does not seem to enhance immunity since a deletion mutant does not show any inflammation decrease. We provide here the evidence that Brucella also target host cell membrane receptors to dampen the immune response thereby contributing to the establishment of a chronic infection. 59

2.0 n s S p le e n w e ig h t (m g ) L iv e r w e ig h t (m g ) L o g C F U /o rg a n L o g C F U /o rg a n A B 2.0 C I 1.5 1.0 0.5 * * * * * * * C I 1.5 1.0 0.5 * * * * * * * * * * 0.0 S p le e n L iv e r M L N 0.0 S p le e n L iv e r M L N Figure 28: Competitive index proliferation between Brucella wt and omp25 mutant strains. C57BL/6 (A) and Balb/C (B) mice have been infected IP with 1.10 6 CFU of a mixture containing 50 % of Brucella wt and 50 % of omp25. 5 days later organs are removed and CFU are enumerated. Each symbol represents an animal and data represent means ± standard deviations. CI statistically different from 1.0 was indicated as follow: P < 0,01: **, P<0,005: ***, P<0,001: ****. A S p le e n L iv e r 5 4 4 3 3 2 2 1 1 B 0 1 5 0 0 C 5 7 B L /6 B a lb /C C 5 7 B L /6 S p le e n B a lb /C * B. a b o rtu s w t B. a b o rtu s o m p 2 5 0 2 5 0 0 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C L iv e r p = 0.0 6 1 0 0 0 **** **** 2 0 0 0 ** **** 5 0 0 1 5 0 0 0 1 0 0 0 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C P B S B. a b o rtu s w t B. a b o rtu s o m p 2 5 C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C C 5 7 B L /6 B a lb /C Figure 29: Bacterial growth and weight organs in wt and omp25 infected mice at 60 days postinfection. A. Mice have been infected IP with 1.10 6 CFU. 60 days later organs are removed and CFU are counted. B. abortus wt infected mice are indicated with a circle and the ones infected with omp25 are indicated with a square. Mouse strain used is indicated in the x axis. B. Organs were weighted for infected mice. Each symbol represents an animal and the median values are marked by horizontal bold lines. P < 0,05 : *, P< 0,01: **, P < 0,001: ****.

Material and Methods Bacterial strains In this study we used B. abortus smooth virulent strain 2308, omp25 strain was a gift from Ignacio Moriyón and has been described previously [45]. Brucella strains were grown onto TSA plates (Sigma Aldrich) containing Kanamycin for omp25. For infection, strains were grown overnight at 16h, 37 C under shaking in TSB (Sigma Aldrich) with kanamycin for omp25 until the OD (OD at 600nm) reached 1.8. All experiments with Brucella were carried out in our BSL3 facility. E. coli DH5-α thermocompetent bacteria were used to amplify CD150 constructs. Liquid cultures were done in LB during overnight incubation at 37 C under shaking. Solid cultures were made onto LB-Agar. Mice Wild type Balb/c mice, wild type C57BL/6 mice, C57BL/6 OTII mice were obtained from Charles River. IFN-γ KO mice were obtained from the Jackson Laboratory. CD150 KO (on C57BL/6 background) mice were provided by Yusuke Yanagi and the method to obtain them is described in the ref. [220]. Animal experimentation was conducted in strict accordance with good animal practice as defined by the French animal welfare bodies (Law 87 848 dated 19 October 1987 modified by Decree 2001-464 and Decree 2001-131 relative to European Convention, EEC Directive 86/609). INSERM guidelines have been followed regarding animal experimentation (authorization No. 02875 for mouse experimentation). All animal work was approved by the Direction Départementale Des Services Vétérinaires des Bouches du Rhône (authorization number 13.118). All the in vivo work protocols have been submitted to the Regional Ethic Committee for evaluation. 60

S p le e n w e ig h t (m g ) L iv e r w e ig h t (m g ) P e rc e n t s u rv iv a l 1 0 0 7 5 5 0 2 5 * * B. a b o rtu s o m p 2 5 B. a b o rtu s w t 0 0 5 1 0 1 5 2 0 2 5 3 0 3 5 D a y s p o s t-in fe c tio n Figure 30: Survival curve of IFN-γ KO mice infected with Brucella wt or omp25. IFN-γ KO mice have been infected IP with 1.10 6 CFU. Mice weight was assessed each two days and after a loss of more than 30 % of the original weight, mice were considered as dead. B. abortus wt infected mice are indicated in blue, and omp25 infected mice are indicated in red. Each black dot represents an animal. P < 0,01 : **. A L o g C F U /o rg a n 5.5 5.0 4.5 4.0 3.5 * L o g C F U /o rg a n 4 3 2 1 0 w t m ic e C D 1 5 0 K O m ic e w t m ic e B. a b o rtu s w t C D 1 5 0 K O m ic e B B. a b o rtu s o m p 2 5 1 0 0 0 p = 0.0 5 7 3 0 0 0 8 0 0 6 0 0 4 0 0 2 0 0 0 2 0 0 0 1 0 0 0 0 w t m ic e C D 1 5 0 K O m ic e B. a b o rtu s w t w t m ic e C D 1 5 0 K O m ic e B. a b o rtu s o m p 2 5 Figure 31: Bacterial growth and weight organs in wt and omp25 infected CD150 KO mice at 8 days post-infection. A. Mice have been infected IP with 1.10 6 CFU. 8 days later organs are removed and CFU are counted. B. abortus wt infected mice are indicated with a circle and the ones infected with omp25 are indicated with a triangle. Mouse strain used is indicated in the x axis. B. Organs were weighted for infected mice. Each symbol represents an animal, data represent means ± standard deviations. P < 0,05 : *.

Reagents Antibodies used in flow cytometry are the following, CD11c-APC Cy7, CD80-PE Cy5, CD40-Alexa 647, CD150-PE Cy7, MHC II (I-A/I-E) PE, CD25 FITC, CD86 FITC, CD62L PE, CD4 and CD8-PE Cy5 were all purchased at BioLegend. CD3 efluor450 was purchased at ebioscience. CD44-Alexa700 was purchased at BD Biosciences. E. coli LPS used for this study has been purchased at Sigma. Brucella membrane extracts are a gift of I. Moriyón. Blocking and control peptides to CD150 were synthetized by Thermo Scientific and the sequences used were described there [189]. Cell Culture BMDCs were prepared from 6 8 week-old female C57BL/6 mice as previously described [23]. OTII CD4 + T cells are prepared from OTII mice. Spleen and lymph nodes of mice were harvested, and then cells were extracted and purified using magnetic beads (Dynal, Invitrogen). T cells were cultivated in RPMI 1640 (Gibco, Life Technologies) supplemented with 5% of FCS, 1% HEPES (Gibco, Life Technologies), Penicilline/Streptomycine, 1% of Sodium Pyruvate. For co-culture of BMDC with OTII cells, we put a BMDC/T cell ratio of 1:4, and cells were co-cultivated for 3 days before labeling for flow cytometry. Construction of myc-cd150 exon 2 3: cdna fragment of mouse CD150 was obtained from Origene and the two first exons were amplified by PCR using the following primers CD150-Fw: 5 GGGGACAAGTTTGTACAAAAAAGCAGGCTTCACAGGTGGAGGTGTGATGGAT- 3 and CD150-Rv 5 - GGGGACCACTTTGTACAAGAAAGCTGGGTCCTACTGAGGAGGATTCCTGCTTGC- 3 and cloned into a pcmv-myc and using Gateway Technology (Invitrogen, Life 61

5 * 4 ns C I 3 2 * * * 1 0 S p le e n L iv e r M L N Figure 32: Competitive index in CD150 KO mice. CD150 KO mice have been infected IP with 1.10 6 CFU of a mixture containing 50 % of Brucella wt and 50 % of omp25. 5 days later organs (as indicated on the x axis) are removed and CFU are enumerated. Each symbol represents an animal and data represent means ± standard deviations. CI statistically different from 1.0 was indicated as follow: P < 0,05: *, P<0,005: ***. -m y c B. a b o rtu s w t O M B. a b o rtu s o m p 2 5 O M m y c ::C D 1 5 0 m y c ::C D 1 5 0 m y c ::S ifa + - + - - + - + m y c ::S ifa P u ll dow n 2 5 k D a -o m p 2 5 P u ll dow n 3 7 k D a 2 5 k D a -m y c w t O M o m p 2 5 O M B. a b o rtu s O M : In p u t 1 0 0 % 2 5 k D a -o m p 2 5 Figure 33: Brucella Omp25 binds CD150. COS-7 cells were transfected with 1,5 µg of plasmid containing myc::cd150(2-3) or myc::sifa. 48 h after transfection, cells were harvested and proteins extracted. After myc pull-down and incubation (+) with OM extracts (1µg) from either Brucella wt or omp25, the binding of Omp25 to CD150 was assessed by western blot. Pulled myc tagged proteins were indicated as well as the input (100 %) for Brucella membrane extracts. One representative experiment out of 3 independent is shown here.

Technologies). Plasmids were transformed into E. coli thermocompetent strain DH5-α for amplification and then purified with MaxiPrep Plasmid Kit (Qiagen). Expression and purification of myc-cd150 exon 2 3: 1.5 µg of myc-cd150 was transfected into COS-7 cells using Fugene (Promega) technology according to the manufacturer s instructions. 48 h after transfection cells were harvested and lysed into PBS, NP-40 0.5 % containing Proteases Inhibitor (Roche). Expression of the protein was confirmed by western blot against myc. Myc-SifA plasmid was provided by Stéphane Méresse. Immunoprecipitation of myc-cd150 and myc-sifa Protein G beads were coupled to 1 µg of myc (9E10) antibody during 1 h at 4 C and then incubated with the whole cell extracts for 1 h at 4 C. After washes with PBS containing 0.1 % NP-40 and proteases inhibitors, myc-cd150 and myc-sifa were incubated 1 h with 1 µg of B. abortus wt OM or B. abortus omp25 OM at 4 C. Washes with 0.5M NaCl and 0.001 % SDS were then performed to eliminate all non-specific interactions. Samples were then boiled at 95 C for 5 minutes, and centrifuge for 10 min at 14,000 g prior to supernatants analysis. Western blot against Omp25 (antibody A595F1C9 was a gift from A. Cloeckaert) was then perfom using Mouse IgG True Blot HRP (ebiosciences) as secondary antibody to avoid unspecific bands. Infection assays BMDC infections were performed at a multiplicity of infection (MOI) of 30:1. Bacteria were centrifuged onto cells at 400 g for 10 min at 4 ºC and then incubated for 30 min at 37 ºC with 5% CO 2. Cells were washed twice with medium and then incubated for 1 h in medium 62

containing 100 µg/ml gentamicin (Sigma Aldrich) to kill extracellular bacteria. Thereafter, the antibiotic concentration was decreased to 20 µg/ml. To monitor bacterial intracellular survival, infected cells were washed 3 times in PBS and lysed with 0.1% Triton X-100 in H 2 O and serial dilutions plated in triplicates onto TSB agar to enumerated CFUs after 3 days at 37 C. RNA extraction and RT Total RNAs were extracted from infected BMDC using RNeasy Mini Kit (Qiagen) and following manufacturer s instructions. cdnas were generated by using Quantitech Reverse Transcription Kit (Qiagen) following manufacturer s instructions and using 300 ng of RNA as matrix. qpcr 2 µl of cdna were used as matrix for qpcr, which was performed with SYBR Green (Takara) following the manufacturer s instructions in 7500 Fast Real-time PCR (Applied Biosystem). Primers used are listed into Table 1. HPRT was used as a housekeeping gene to determine ΔCt. Fold increase was compared between the control and the treated cells. mrnas which were expressed more than 2 fold more were considered as significantly upregulated. Cytokines measurement Culture supernatants or sera from mice were analyzed to determine the cytokine profiles were whether by cytometric beads assay (BD, Mouse Inflammation kit) or by Sandwich enzymelinked immunosorbent assays (ELISA) from ebiosciences for total IL12, and IL1β. 63

Flow cytometry Cells were harvested and stained for 20 min at 4 C with the antibodies cited above. Cells were then washed once in 2 % FCS in PBS and once in PBS. Infected cells were then fixed for 20 min in 3 % PFA at room temperature (RT). Events were collected on flow cytometry using a FACSCantoII (Becton Dickinson) or FACSLSRII UV and analysis was performed on FlowJo software (TreeStar) and FACS DIVA (BD). Immunofluorescence microscopy Cells were fixed in 3 % paraformaldehyde, ph 7.4, at room temperature for 20 min. For NFκB studies, cells were then permeabilized for 10 min with 0.1 % saponin in PBS, followed by a blocking for 1 h with 2 % BSA in PBS. Primary antibodies were incubated for 1 h followed by 2 washes in PBS, 45 min incubation for secondary antibodies, 2 washes in PBS and 1 wash in water before mounting with Prolong Gold (Life technologies). Primary antibodies used: rabbit anti-p65 from Santa Cruz at 1/200, hamster anti-cd11c from BioLegend at 1/100 and cow anti-brucella LPS antibody at 1/2000. Secondary antibodies used: goat anti-hamster Alexa 594, donkey anti-rabbit Alexa-546, goat anti-cow FITC, all from Jackson Immunoresearch. Nuclei were stained with TOPRO-3. For all the others immunofluorescence labelling, we used 2% BSA in PBS for 1 h to block non-specific interaction, then we incubate the primary antibodies for 30 min in PBS containing 0,1 % saponine and of horse serum. Coverslips with cells are then washed twice in PBS 0,1 % saponine before 30 min incubation with secondary antibodies. Coverslips are then mounted in Prolong Gold. We used as primary antibodies: Rabbit anti-mouse calnexin (Abcam) at 1/200, rat anti-mouse Lamp1 (clone 1D4B) from Santa Cruz at 1/200, phalloïdine coupled to Alexa 546 (Invitrogen) at 1/1000 and a cow anti-brucella LPS antibody at 1/2000. Secondary antibodies used were purchased at Jackson Immunoresearch and Life Technologies (Invitrogen): anti-rabbit Pacific Blue, anti- 64

goat Alexa 546, anti-rat 647. Samples were examined on a Leica SP5 laser scanning confocal microscope for image acquisition. Images of 1024x1024 pixels were then assembled using Adobe Photoshop 7.0 or ImageJ. In all experiments we used an anti-cd11c antibody confirming analysis of DCs only. Quantification was always done by counting at least 50 cells in 5 independent experiments, for a total of at least 250 host cells analyzed. Mice infection Balb/c mice or C57BL/6 mice were infected in our BSL3 facility by intra peritoneal injection. 1x10 6 CFU were injected into 200 µl of sterile endotoxin-free PBS for each mouse. Organs were harvested 5 or 60 days post-injection and then scratched into sterile Triton X-100 0.1 % diluted in H 2 O, serial dilutions in sterile PBS were used to count CFU. Blood was collected into EDTA tubes at 5 or 60 days post infection and spin at 3500 rpm for 5 min at RT to collect sera. For histology studies, organs were harvested and placed into 10 % of formalin for 24 h at RT before inclusion in paraffine. The slides were then stained with hematoxylin and eosin. For competitive index experiment a mixture of 50 % of wt Brucella and 50 % of omp25 were injected at a final concentration of 1.10 6 CFU per mouse. Acknowledgements CD and AP held fellowships from Aix-Marseille University. This work was supported by the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, Aix-Marseille University. Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed. 65

Table 1: Primers used for cdna amplification. Gene Sens Sequence HPRT Fw 5'-3' AGCCCTCTGTGTGCTCAAGG HPRT Rv 5'-3' CTGATAAAATCTACAGTCATAGGAATGGA PTGS2 Fw 5'-3' ACCTCTGCGATGCTCTTCC PTGS2 Rv 5'-3' TCATACATTCCCCACGGTTT TNF-α Fw 5'-3' CATCTTCTCAAAATTCGAGTGACAA TNF-α Rv 5'-3' TGGGAGTAGACAAGGTACAACCC NOS-2 Fw 5'-3' CAGCTGGGCTGTACAAACCTT NOS-2 Rv 5'-3' CATTGGAAGTGAAGCGTTTCG IL-12b Fw 5'-3' AAATTACTCCGGACGGTTCA IL-12b Rv 5'-3' ACAGAGACGCCATTCCACAT IL-6 Fw 5'-3' GAGGATACCACTCCCAACAGACC IL-6 Rv 5'-3' AAGTGCATCATCGTTGTTCATACA IFN-β Fw 5'-3' GAAAAGCAAGAGGAAAGATT IFN-β Rv 5'-3' AAGTCTTCGAATGATGAGAA IL1-β Fw 5'-3' TCCAGGATGAGGACATGAGCAC IL1-β Rv 5'-3' GAACGTCACACACCAGCAGGTTA KC Fw 5'-3' CAGCCACCCGCTCGCTTCTC KC Rv 5'-3' TCAAGGCAAGCCTCGCGACCAT CCL-2 Fw 5'-3' GCCTGCTGTTCACAGTTGC CCL-2 Rv 5'-3' ATTGGGATCATCTTGCTGGT 66

II. C. LE CΒG DE BRUCELLA ACTIVE LES DC ET CONTROLE LE RECRUTEMENT DES NEUTROPHILES II. C. 1. Introduction Le CβG est un facteur de virulence essentiel à Brucella pour permettre sa réplication au sein des macrophages [21], mais pas des DC [23]. Il est capable de contrôler le niveau de cholestérol au niveau de la BCV [21]. Le CβG est un activateur des DC humaines et murines. En effet, après ajout sur des DC de CβG purifié, celles-ci vont s activer, exprimer des molécules de co-stimulation comme CD86, CD80, CD40, du CMH-II, sécréter des cytokines pro-inflammatoires comme l IL-6, le TNF-α ou encore l IL-12. Ces DC sont aussi d efficaces inducteurs de l activation des LT [22]. Contrairement au LPS d E. coli, cette molécule est non immunogénique et non toxique, ce qui en fait un excellent candidat pour être un adjuvant. Dans cet article nous décrivons le rôle du CβG dans l induction de l inflammation in vivo. Grâce à des analyses transcriptomiques, nous montrons qu en effet le CβG induit un profil d activation des DC, mais aussi des gènes liés à un rôle plus anti-inflammatoire comme les SOCS ou TNFAIP6. Nous avons étudié aussi le recrutement de cellules immunitaire au site d injection in vivo du CβG, pour essayer de déterminer comment cette molécule peut promouvoir une réponse immunitaire innée et adaptative, mais ne pas être toxique comme le LPS. 67

II. C. 2. Manuscrit soumis Title: Brucella cyclic glucan tunes up inflammation in human and mouse dendritic cells Running title: Brucella cyclic glucan controls inflammation Clara Degos 1,2,3, Aurélie Gagnaire 1,2,3, Romain Banchereau 4 Ignacio Moriyón 5 and Jean- Pierre Gorvel 1,2,3* 1 Centre d'immunologie de Marseille-Luminy (CIML), Aix-Marseille University, UM2, Marseille, France, 2 Institut National de la Santé et de la Recherche Médicale (INSERM), U1104, Marseille, France; 3 Centre National de la Recherche Scientifique (CNRS), UMR7280, Marseille, France; 4 Baylor Institute for Immunology Research, Dallas TX, USA ; 5 Instituto de Salud Tropical y Depto. Microbiología y Parasitología, Universidad de Navarra, Pamplona, Spain. *To whom correspondence should be addressed: gorvel@ciml.univ-mrs.fr Keywords: Brucella, cyclic beta glucan, inflammation, neutrophil, lipopolysaccharide, SOCS, PTGS2 Abbreviations: Beta-1,2 cyclic glucan: CβG, lipopolysaccharide: LPS, dendritic cells: DC, outer membrane protein: omp 68

Abstract Brucella is the causing agent of a chronic zoonosis called brucellosis. Brucella follows a stealthy strategy that relies on specific pathogen-associated molecular patterns and on virulence factors that dampen immune responses. The Brucella beta-1,2 cyclic glucan (CβG) has been described as a potent immune stimulator, albeit with no toxicity for cells and animals. Here we used a genome-wide approach to characterize human mdc responses to CβG and compared them to LPS. We found 34 differently regulated genes related to inflammation (IL-6, IL2RA, PTGS2), chemokine (CXCR7, CXCL2) and anti-inflammatory pathways (TNFAIP6, SOCS2). We validated these results in mouse BMDC and characterized the inflammatory infiltrates at the level of mouse ear inflammatory sites when injected with CβG or LPS. CβG yielded a lower and transient recruitment of neutrophils compared to LPS. The consequence of these dual pro- and anti-inflammatory signals triggered by CβG is the induction of local inflammation without alerting neutrophils. Introduction Brucellosis is a zoonosis affecting humans, farm animals and livestock, which represents a significant economic burden in developing countries. Brucella, the agent of brucellosis, is a pathogenic Gram-negative bacterium belonging to the alpha-2 proteobacteria group. The World Health Organization (WHO) has classified brucellosis among the top seven neglected zoonoses, a group of diseases that are simultaneously a threat to human health and a cause of poverty. 1 It is now recognized that in countries such as Mongolia and northern China, brucellosis is becoming a threat for populations living in close contact with animals. Pathogenic brucellae can efficiently replicate within the endoplasmic reticulum of infected macrophages and dendritic cells, a safe intracellular niche located at the crossroad of many vital host cell functions. The three main species leading to brucellosis in humans are B. 69

melitensis, B. suis and B. abortus. B melitensis is the most potent among the Brucella species, since 1-5 bacteria are enough to cause disease. In addition, the existing vaccine against B. melitensis (vaccine Rev 1) is virulent for humans, resistant to some antibiotics used to treat human brucellosis, and not stable (spontaneous change from smooth (S) full potency to a more attenuated rough (R) phenotype), which leads to reduced protection efficacy. Over the last decade, most studies have focused on understanding the role of virulence factors expressed by pathogenic brucellae. However, their role in virulence and the characterization of the mechanisms involved, including their associated host cell counterparts, have only been described in a few of them and reviewed in refs. 2, 3. In addition to their role in Brucella replication and intracellular trafficking, some virulence factors were shown to play an important role in dampening both innate and adaptive immunity. For example, BtpA and BtpB act as inhibitors of TLR signaling 4, 5, while the products of the wadb and wadc genes encoding mannosyl transferases are known to bring mannosyl residues in the core region of LPS to avoid recognition via TLR4. 6 Mutant strains lacking these proteins lead to enhanced recognition by innate receptors, increased inflammation and a strong adaptive immune response. WadC and wadb mutants can even induce protection in mice. 6 Recently, we described the β-1,2 cyclic glucan (CβG), an additional virulence factor synthesized by Brucella which is concentrated in the periplasm of the bacterium. This polysaccharide is composed of a cyclic backbone of 17 to 25 glucose units in β-1,2 linkages and can harbor substitutions such as succinyl, mevalonyl and methyl groups. Brucella CβG was demonstrated to modulate lipid raft organization both at the plasma membrane of infected cells and intracellularly at the site of the Brucella-containing vacuole. 7 CβG is expressed in large amounts, representing 1-5% of the bacteria dry weight. When bacteria are killed by the host immune system, CβG is therefore released in the surrounding inflammatory environment in µm concentrations. This may have important consequences for the modulation of 70

intracellular trafficking of the bacterium by shaping the lipid microdomain composition of the Brucella-containing vacuole and by modulating host immune responses. Brucella CβG signals through TLR4, without the contribution of CD14. 8 In contrast to Btps and WadC virulence factors, which are involved in the inhibition of immune responses, Brucella CβG is a strong activator of both human and mouse dendritic cells, promoting pro-inflammatory cytokine expression, antigen cross-priming and cross-presentation to specific CD8 + T cells. However, unlike E. coli LPS and its derivatives, CβG does not display endotoxicity both in vitro and in vivo, and has recently been referred as a new class of adjuvants. We therefore asked how the lack of endotoxicity and the strong immune response generated by CβG could be reconciled. Herein, we identify differential expression profiles between LPS and CβG-stimulated mdc, highlighting specific anti-inflammatory networks in response to CβG that may lead to dampening of neutrophil recruitment at the site of injection, thus leading to a reduced inflammation. Results Brucella CβG differentially modulates transcriptomic responses in human blood mdc. We previously described a number of immune pathways modulated by the Brucella CβG in human blood mdc. 8 Herein, we compared the transcriptional profiles of mdc stimulated with CβG or LPS. We identified 34 genes that were expressed at a level at least 4.5-fold higher in CβG-stimulated mdc than in LPS-stimulated cells (Table 1). As previously described, 8 some of these genes are related to inflammation (IL-6, BATF, IL2RA, PTGS2), or chemokines pathways (CXCR7, CXCL2) while other are related to anti-inflammatory pathways such as TNFAIP6, SOCS2. 9 71

We show that Brucella CβG triggers the transcription of regulatory genes, particularly those involved in the inhibition of pro-inflammatory cytokine secretion to stop the inflammatory process (Fig. 1A). Several transcripts encoding chemokines were over-expressed in CβGstimulated mdc. CXCL2, IL8, CCL5, and CCL20 were highly expressed upon stimulation with either Brucella CβG or E. coli LPS (Fig. 1B). These genes encode chemoattractants for immune cells. 10 E. coli LPS but to a larger extent CβG were capable of inducing the transcription of SOCS genes known to encode regulatory elements that were described to control inflammation (Fig. 1C). 11 We focused on the expression levels of CXCL2, TNFAIP6, PTGS2, SOCS3, and IL8 (Fig. 1C) since these transcripts were over-expressed in CbGstimulated compared to E. coli LPS-stimulated mdc. CXCL2 and IL8 are chemokines that were previously described to promote neutrophil recruitment, while SOCS3 negatively 12, 13 regulates pro-inflammatory cytokine secretion. PTGS2 is involved in prostaglandins 14, 15 metabolism and TNFAIP6 has been shown to be an anti-inflammatory molecule. Whilst compared to E. coli LPS, Brucella CβG seems to enhance inflammatory pathways by upregulating a selection of genes related to chemotaxis (Fig. 1) and pro-inflammatory cytokines 8, it also preferentially induces anti-inflammatory genes such as SOCS, TNFAIP6, LILRB and IDO2 (Figs. 1A,C). 16-18 Over-expressed genes in Brucella CβG- stimulated human mdc are expressed in mouse DC. We confirmed the expression of CXCL2, KC, PTGS2, SOCS3 and TNFAIP6 mrnas in Brucella CβG or E. coli LPS-stimulated murine BMDC after stimulation by RT-PCR (Figs. 2A,B). We observed a strong induction of all these genes at 8 h post-stimulation, which declined at 24 h post-stimulation. CβG- and E. coli LPS-treated BMDC expressed similar levels of CXCL2 and TNFAIP6 transcripts at 8 h post-stimulation (Figs. 2A,B), in contrast to 72

human mdc, in which higher amounts of both transcripts were observed in CβG-treated cells (Fig. 1C) thus reflecting differences between human and mouse DC subsets. We then validated transcriptional profiles at the protein level. To this end, we measured the expression level of the tsg-6 protein, the translated product of the TNFAIP6 gene (Fig. 2C). tsg-6 was highly expressed in murine BMDC after 8 h stimulation and, as observed at the transcriptional level, its expression was decreased at 24 h post-injection in all conditions (Fig. 2C). In murine DC, LPS and CβG both act on similar inflammatory pathways through the modulation of chemokines and anti-inflammatory gene transcription. The induction of both inflammatory and anti-inflammatory genes was higher at early post-stimulation time (8 h) suggesting an early response to stimulation induced by both LPS and CβG. Injection of Brucella CβG into mouse ear induces the recruitment of innate immune cells in vivo. Since we observed a high expression of chemokines transcripts in DC stimulated with E. coli LPS and Brucella CβG, we measured the levels of inflammation triggered at the site of injection in mice that had been intradermally immunized in the ear. At 48 h post-injection ear tissue was recovered and stained for hematoxylin and eosin. Injection with CβG or LPS led to the formation of an edema with cell infiltrates, which was not observed in PBS-injected control mice (Fig. 3A). However, in CβG-injected ears, the edema was significantly smaller than in LPS-injected ears, suggesting a lower level of inflammation in response to CβG as compared to LPS. To determine which cells were recruited at the inflammatory site, we immuno-stained ear sections to detect monocytes (CD11b +, Gr1 + ) and neutrophils (CD11b +, Gr1 + and Ly6G + ) by confocal microscopy (Fig. 3B). Although these three markers were localized in both LPS- and 73

CβG-treated ears, LPS-injected ears showed a higher recruitment of neutrophils (white labeling indicating the presence of neutrophils positive for CD11b +, Gr1 + Ly6G + ) (Fig.3.B). Brucella CβG induces a transient recruitment of neutrophils. We then characterized the kinetics of neutrophil recruitment at the site of injection. To this end, we injected PBS, LPS or CβG and harvested ears at 2 h, 6 h, 12 h, 24 h or 48 h postinjection. Cells were stained for F4/80 and Ly6G and analyzed by flow cytometry (Fig. 4). In LPS-injected ears, neutrophils (F4/80 -, Ly6G + ) were recruited as soon as 2 h post-injection and increased in number until 24 h when almost 4,000 neutrophils were recruited to finally reach about 2,000 cells at 48 h post-injection (Figs. 4A,C). In CβG-injected ears, neutrophil recruitment was first detected at 6 h post-injection, reached a maximum at 12 h post-injection with more than 2,000 neutrophils recruited (Fig. 4B), and then strongly decreased from 24 h onwards (Figs. 4A,C), thereby corroborating histology experiments (Fig. 3B). In the case of CβG, neutrophils were transiently recruited between 6 h and 12 h and disappeared at 24 h post-injection, indicating a reduced inflammation in comparison to LPS treatment. Discussion Here, we show that CβG induced the transcription of both pro- and anti-inflammatory genes. Injection of CβG into mouse ear led to a local inflammation characterized by an edema and a fast and transient neutrophil recruitment. In LPS-injected ears, the edema was larger and neutrophil stayed longer in the tissue suggesting a higher and prolonged inflammation. Brucella is considered a stealthy pathogen that aims at keeping the host immune response under control and avoiding inflammatory processes detrimental to the survival of the bacterium. Up until now, most virulence factors expressed by pathogenic Brucella were found 74

to be involved in dampening host cell functions. At the entry site, different virulence factors have been shown to be essential for the bacterium such as inva, BtaE, BtaF and BmaC 19-22 as well as the LPS. 23 When the bacterium enters in its Brucella-containing vacuole, other proteins participate in the infection process to allow Brucella survival; Later on VirB 3 and RicA 24 are required for establishing a safe replication niche. Finally, through evolution and design, virulence factors have been selected to reduce its exposure with the host by interacting with signaling pathways that may promote immune responses. For instance, while PrpA can activate macrophages and this leads to the proliferation of B cells 25, 26, Brucella also expresses Btps, DC inhibitory molecules that can serve to limit inflammation in infected DC. 5, 8 WadC has been shown to protect Brucella LPS from detection and recognition by TLR4 and so modulate the immune response.[94] Elsewhere, outer membrane proteins (Omps) seem to also play a role either in the induction or inhibition of the immune response. 23 Interestingly, CβG, which has been described as a virulence factor in macrophages but not in dendritic cells, has recently been shown to be a strong activator of the immune system, possibly by signaling through TLR4. The consequence of this interaction between DC and CβG is the secretion of pro-inflammatory cytokines. 4, 8 DC infected with null-mutants in the cgs synthase does not show any sign of DC activation, 8 yielding a phenotype similar to PBStreated cells. In contrast, E. coli LPS has been shown to induce a strong activation and maturation of DC. 23, 27 A recent study has shown that upon infection with different intracellular bacteria, including Brucella different intermediate levels of maturation can be observed in human mdc. Brucella, in this case B. abortus, induced a significant but lower activation profile compared to DC treated with LPS or infected with Coxiella burnetii and Orienta tsutsugamushi infection, but at a higher level than when infections were performed with T. whipplei. 28 This intermediate level of activation has also been observed in mouse DC 75

4, 8 and may have consequences on further T and B cell activation. In vitro Brucella infections of mouse DC have shown infected DC to exhibit impaired antigen presentation and T cell activation properties. 4 In another infection system, it has been demonstrated that intermediate level of DC maturation give rise to Th1 response instead of Th2. 29 Whereas most of virulence factors act as inhibitors of host cell function, CβG may be bringing DC maturation to an intermediate level. Moreover, CβG does not show any toxicity, a characteristic that fits very well with Brucella and its goal to bring the host to a chronic disease state whereby the pathogen can be present but silent, and better placed to survive and persist. In this context therefore, Brucella may have adapted the CβG molecule to allow a limited activation of host immune pathways in the absence of toxicity. As we demonstrate herein, when compared with LPS, CβG induces a combination of inflammatory and anti-inflammatory pathways that directs a transient inflammation to better facilitate the control of neutrophil recruitment, corroborating previous studies on neutrophil/brucella interactions. 30 By inducing a limited/controlled response, Brucella can better control disease progression in a manner that is both tolerated by the host and amenable to the ensuring survival and persistence of the pathogen/development of chronic disease. Further work is necessary to understand at the level of the infected host, how CβG and other virulence factors impact the activation status of Brucella-infected DC sub-types and their subsequent role in developing chronic disease. Anyway our study provides evidence of a better understanding of the complex interplay between inflammatory and anti-inflammatory molecular networks in response to bacterial PAMPs. 76

Materials and Methods Cell Culture BMDCs were prepared from 6 10 week-old female C57BL/6 mice as previously described. 4 Briefly, tibias and femur were removed from mice; bone marrow was harvested by flushing with RPMI-1640 (Gibco, Life Technologies). Red blood cells are then lysed, and after three washes cells were seeded onto 6 well plates at 0.6x10 6 cells/ml. Cells were grown in RPMI- 1640, 5% FCS, 50 µm β-mercaptoethanol. Human monocyte-derived DC were purified from blood using Ficoll (Ge Healthcare), cells were cultivated in serum-free Cellgro DC culture media supplemented with 100 ng/ml GM-CSF and 20 ng/ml IL-4. Reagents and antibodies Purified cyclic glucan was obtained from Brucella abortus 2308 as previously described. 8 E. coli LPS (055:B5) was purchased from Sigma Aldrich. Cells were stimulated with 10 µg/ml of CβG or 100 ng/ml or LPS corresponding to the same molarity (0.25µM). Flow cytometry antibodies were: anti Ly6G-V450 (clone 1A8), anti CD45.2-PerCP-Cy5.5 (clone 104), anti Ly6C-PE-CF594 (clone AL-21), anti CD64-AF647 (clone X54-5/7.1) from BD Biosciences ; anti CD24-AF488 (clone M1/69), anti F4/80-PE (clone CI :A3-1), anti CD11b-APC/Cy7 (clone M1/70), anti CD150-PECy7 (clone TC15-12F12.2) from BioLegend, anti I-A/I-E- A700 (clone M5/114-15.2) from ebiosciences. Antibodies used for confocal microscopy were: anti Gr1-PE (clone RB6-8C5), anti CD11b-AF647 (clone M1/70) from BioLegend; anti-ly6g (clone 1A8) from BD Biosciences; anti-rat AF555 (clone A21434) from Invitrogen. Human/Mouse TSG-6 MAb (RD systems) was used to detect the tsg-6 protein in western blots using anti-mouse HRP (Invitrogen) as a secondary antibody. Western blots were revealed using Amersham ECL Detection system (GE Healthcare). 77

Mice Immunization 6-10 weeks old C57BL/6 females were intradermally injected into the internal face of the ear with 200 µg of CβG, 10 µg of LPS (Sigma Aldrich) or PBS as a negative control. At 48 h post-injection mice were sacrificed and ears recovered for further analyses. mrna extraction and hybridization for transcriptomic analysis RNA was extracted and purified from purified blood human mdc. RNA hybridization performed using Illumina HT12 v4 Beadchip arrays was performed as previously described. 8 Bioinformatic analysis of microarrays was performed as previously described. 8 Briefly, pathway analysis was carried out using the Ingenuity Pathway Analysis (IPA) software. A non-parametric test was applied to the different samples from 4 donors with a false discovery rate of 0.01. Only significant transcripts were considered, transcripts presented were normalized to their control, the transcripts showed here have an absolute fold changes equal or superior to 2. Genespring 7.3 was used for analysis and generating heatmaps and GraphPad Prism 5 was used for barcharts. mrna extraction and RT Total RNAs were extracted from infected BMDC using RNeasy Mini Kit (Qiagen) following the manufacturer s instructions.cdnas were generated by using Quantitech Reverse Transcription Kit (Qiagen) following the manufacturer s instructions and using 300 ng of RNA as matrix. qpcr Recovered cdna (2 µl) was used as template for qpcr, was performed with SYBR Green (Takara) following the manufacturer s instructions by 7500 Fast Real-time PCR (Applied 78

Biosystem). Primers used are listed into Table 2. HPRT was used as a housekeeping gene to determine Ct. Fold increases were determined by comparing control (non-treated) and treated cells. mrnas which were expressed more than 2 fold more were considered as significantly upregulated. Protein extraction BMDC were harvested, washed once in PBS, and cell pellets frozen at - 80 C. To purify protein, cell pellets were resuspended in lysis buffer (PBS containing 0.5 % of NP-40 and proteases inhibitor (Roche Diagnosis)) and the resulting cell lysate was centrifuged at 10,000 g for 10 min at 4 C. The supernatants were recovered and analyzed by SDS-PAGE and western blot. Immunohistochemistry staining Ear tissue was recovered and fixed with 10 % formalin for 48 h and embedded in paraffin. A Leica RM2245 microtome was used to prepare 5 µm slides, which were then stained with hematoxylin and eosin. Images were subsequently acquired using a Nikon Eclispe Ci. Confocal microscopy Ear tissues were embedded in tissue-tec OCT (Sakura Finetek). 15 µm cryosections were then saturated in PBS containing 2 % BSA for 30 min, incubated overnight with primary antibodies, and thereafter for 1 h with secondary antibodies. Slides were mounted with ProLong Gold containing DAPI (Invitrogen). Images were acquired using a LSM 510 confocal microscope (Carl Zeiss, Inc.), before analyzed and assembled using ImageJ software (ImageJ). 79

Ear mouse skin cell isolation Ear skin tissues were split in two sheets (dorsal and ventral) and incubated overnight in PBS containing 2.5 mg/ml dispase II (Roche) at 4 C to separate the dermal and epidermal sheets. The separated epidermal and dermal sheets were then cut in to pieces and incubated for 90 min at 37 C with RPMI containing 1 mg/ml DNase (Sigma Aldrich) and 1 mg/ml collagenase type IV (Worthington Biochemical) to obtain a homogeneous cell suspension. Flow cytometry Skin cells were incubated for 10 min with 2.4G2 antibody to block non-specific signal before staining for 20 min at 4 C with the antibodies cited above. Cells were then washed once in 2% FCS in PBS and once in PBS before fixing in 3% PFA for 20 min at room temperature (RT). At least 400,000 events were collected by flow cytometry using a FACSCantoII (Becton Dickinson) or FACSLSRII UV. Analyses were performed using FlowJo software (TreeStar) and FACS DIVA (BD). Acknowledgements CD and AG held fellowships from Aix-Marseille University. This work was supported by the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, Aix-Marseille University, the Baylor Institute for Immunology Research (NIH/NIAID-U19 grant N AI057234). We thank Sean Hanniffy for critical reading and suggestions. We also thank the CIML histology core platform and especially Lionel Chasson. Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed. 80

References 1. Ian Maudlin SW-M. The Control of Neglected Zoonotic Diseases - A route to poverty alleviation. Geneva: WHO, 2006. 2. de Jong MF, Tsolis RM. Brucellosis and type IV secretion. Future Microbiol 2012; 7:47-58. 3. Lacerda TL, Salcedo SP, Gorvel JP. Brucella T4SS: the VIP pass inside host cells. Current opinion in microbiology 2013; 16:45-51. 4. Salcedo SP, Marchesini MI, Lelouard H, Fugier E, Jolly G, Balor S, Muller A, Lapaque N, Demaria O, Alexopoulou L, et al. Brucella control of dendritic cell maturation is dependent on the TIR-containing protein Btp1. PLoS pathogens 2008; 4:e21. 5. Salcedo SP, Marchesini MI, Degos C, Terwagne M, Von Bargen K, Lepidi H, Herrmann CK, Santos Lacerda TL, Imbert PR, Pierre P, et al. BtpB, a novel Brucella TIRcontaining effector protein with immune modulatory functions. Front Cell Infect Microbiol 2013; 3:28. 6. Conde-Alvarez R, Arce-Gorvel V, Iriarte M, Mancek-Keber M, Barquero-Calvo E, Palacios-Chaves L, Chacon-Diaz C, Chaves-Olarte E, Martirosyan A, von Bargen K, et al. The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS pathogens 2012; 8:e1002675. 7. Arellano-Reynoso B, Lapaque N, Salcedo S, Briones G, Ciocchini AE, Ugalde R, Moreno E, Moriyon I, Gorvel JP. Cyclic beta-1,2-glucan is a Brucella virulence factor required for intracellular survival. Nat Immunol 2005; 6:618-25. 8. Martirosyan A, Perez-Gutierrez C, Banchereau R, Dutartre H, Lecine P, Dullaers M, Mello M, Salcedo SP, Muller A, Leserman L, et al. Brucella beta 1,2 cyclic glucan is an activator of human and mouse dendritic cells. PLoS pathogens 2012; 8:e1002983. 9. Posselt G, Schwarz H, Duschl A, Horejs-Hoeck J. Suppressor of cytokine signaling 2 is a feedback inhibitor of TLR-induced activation in human monocyte-derived dendritic cells. Journal of immunology 2011; 187:2875-84. 10. McColl SR, Clark-Lewis I. Inhibition of murine neutrophil recruitment in vivo by CXC chemokine receptor antagonists. Journal of immunology 1999; 163:2829-35. 11. Krebs DL, Hilton DJ. SOCS: physiological suppressors of cytokine signaling. Journal of cell science 2000; 113 ( Pt 16):2813-9. 12. Carow B, Rottenberg ME. SOCS3, a Major Regulator of Infection and Inflammation. Front Immunol 2014; 5:58. 13. Babon JJ, Nicola NA. The biology and mechanism of action of suppressor of cytokine signaling 3. Growth Factors 2012; 30:207-19. 14. Danchuk S, Ylostalo JH, Hossain F, Sorge R, Ramsey A, Bonvillain RW, Lasky JA, Bunnell BA, Welsh DA, Prockop DJ, et al. Human multipotent stromal cells attenuate lipopolysaccharide-induced acute lung injury in mice via secretion of tumor necrosis factoralpha-induced protein 6. Stem cell research & therapy 2011; 2:27. 15. Choi H, Lee RH, Bazhanov N, Oh JY, Prockop DJ. Anti-inflammatory protein TSG-6 secreted by activated MSCs attenuates zymosan-induced mouse peritonitis by decreasing TLR2/NF-kappaB signaling in resident macrophages. Blood 2011; 118:330-8. 81

16. Anderson KJ, Allen RL. Regulation of T-cell immunity by leucocyte immunoglobulinlike receptors: innate immune receptors for self on antigen-presenting cells. Immunology 2009; 127:8-17. 17. Katz HR. Inhibition of inflammatory responses by leukocyte Ig-like receptors. Advances in immunology 2006; 91:251-72. 18. Trabanelli S, Ocadlikova D, Ciciarello M, Salvestrini V, Lecciso M, Jandus C, Metz R, Evangelisti C, Laury-Kleintop L, Romero P, et al. The SOCS3-independent expression of IDO2 supports the homeostatic generation of T regulatory cells by human dendritic cells. Journal of immunology 2014; 192:1231-40. 19. Alva-Perez J, Arellano-Reynoso B, Hernandez-Castro R, Suarez-Guemes F. The inva gene of Brucella melitensis is involved in intracellular invasion and is required to establish infection in a mouse model. Virulence 2014; 5:563-74. 20. Ruiz-Ranwez V, Posadas DM, Estein SM, Abdian PL, Martin FA, Zorreguieta A. The BtaF trimeric autotransporter of Brucella suis is involved in attachment to various surfaces, resistance to serum and virulence. PLoS One 2013; 8:e79770. 21. Ruiz-Ranwez V, Posadas DM, Van der Henst C, Estein SM, Arocena GM, Abdian PL, Martin FA, Sieira R, De Bolle X, Zorreguieta A. BtaE, an adhesin that belongs to the trimeric autotransporter family, is required for full virulence and defines a specific adhesive pole of Brucella suis. Infection and immunity 2013; 81:996-1007. 22. Posadas DM, Ruiz-Ranwez V, Bonomi HR, Martin FA, Zorreguieta A. BmaC, a novel autotransporter of Brucella suis, is involved in bacterial adhesion to host cells. Cellular microbiology 2012; 14:965-82. 23. Martirosyan A, Moreno E, Gorvel JP. An evolutionary strategy for a stealthy intracellular Brucella pathogen. Immunol Rev 2011; 240:211-34. 24. de Barsy M, Jamet A, Filopon D, Nicolas C, Laloux G, Rual JF, Muller A, Twizere JC, Nkengfac B, Vandenhaute J, et al. Identification of a Brucella spp. secreted effector specifically interacting with human small GTPase Rab2. Cell Microbiol 2011; 13:1044-58. 25. Spera JM, Herrmann CK, Roset MS, Comerci DJ, Ugalde JE. A Brucella virulence factor targets macrophages to trigger B-cell proliferation. The Journal of biological chemistry 2013; 288:20208-16. 26. Spera JM, Ugalde JE, Mucci J, Comerci DJ, Ugalde RA. A B lymphocyte mitogen is a Brucella abortus virulence factor required for persistent infection. Proceedings of the National Academy of Sciences of the United States of America 2006; 103:16514-9. 27. Banchereau J, Steinman RM. Dendritic cells and the control of immunity. Nature 1998; 392:245-52. 28. Gorvel L, Textoris J, Banchereau R, Ben Amara A, Tantibhedhyangkul W, von Bargen K, Ka MB, Capo C, Ghigo E, Gorvel JP, et al. Intracellular bacteria interfere with dendritic cell functions: role of the type I interferon pathway. PLoS One 2014; 9:e99420. 29. Dirix V, Mielcarek N, Debrie AS, Willery E, Alonso S, Versheure V, Mascart F, Locht C. Human dendritic cell maturation and cytokine secretion upon stimulation with Bordetella pertussis filamentous haemagglutinin. Microbes and infection / Institut Pasteur 2014; 16:562-70. 30. Barquero-Calvo E, Martirosyan A, Ordonez-Rueda D, Arce-Gorvel V, Alfaro-Alarcon A, Lepidi H, Malissen B, Malissen M, Gorvel JP, Moreno E. Neutrophils exert a suppressive 82

effect on Th1 responses to intracellular pathogen Brucella abortus. PLoS pathogens 2013; 9:e1003167. Legends Table 1: Fold changes in gene expression in CβG- versus LPS-treated human blood mdc. Table 1 lists 34 genes over-expressed at least 4.5-fold higher in CβG-stimulated cells compared to LPS-stimulated cells (at 6 h post-treatment). Data were normalized against cells treated for 6 h with culture media only. Table 2: Sequence of the qpcr primers used. Figure 1: Transcriptional profiling of human blood mdc from healthy donors stimulated with CβG (n=5) or LPS (n=4), respectively. A. Heatmap representing the transcription expression levels of 15 regulatory genes differentially expressed in mdc stimulated with either CβG, LPS or cell culture medium (control) (Welch T-test, p<0.05). Data were normalized against cells from each donor that had been treated for 6 h with cell culture medium alone. B. Heatmap representing the transcriptional expression levels of 18 chemokines differentially expressed in mdc treated with CβG, LPS or cell culture medium. C. Bar charts representing the mean raw expression values of CXCL2, IL8, PTGS2, SOCS3 and TNFAIP6. Error bars represent the standard deviation. Figure 2: Induction of gene expression in murine DC stimulated with CβG or LPS. A. Murine BMDC were stimulated for 8 h and 24 h with E. coli LPS (black bars) or Brucella CβG (grey bars). mrna was extracted from stimulated cells and qpcr performed to determine transcript expression levels of CXCL2, KC and PTGS2. Fold-increases were 83

estimated by comparing with cells that had been stimulated with PBS as a negative control. Basal expression levels for each gene are indicted by a dashed line. HPRT was used as a housekeeping gene to normalize the data. Three independent experiments were carried out. B. Expression levels of SOCS3 and TNFAIP6 mrna were assessed. Experiments were processed as described above. Three independent experiments were carried out. C. BMDC stimulated for 8 h or 24 h with PBS (control), E. coli LPS or Brucella CβG were lysed and protein purified. The expression of tsg-6 protein was assessed by western blot using 10 µg of recombinant tsg-6 as a positive control. β-actin expression was used as control. At least 3 independent experiments were carried out and one representative is shown here. Figure 3: Brucella CβG induces the recruitment of CD11b +, LyG6 +, and Gr1 + cells at the site on injection. A. Mouse ears intradermally injected with PBS, E. coli LPS or Brucella CβG were recovered at 48 h post-injection and stained for hematoxylin and eosin. B. Mouse ears injected with PBS, E. coli LPS or Brucella CβG were recovered at 48 h post-injection, embedded in tissueteck OCT compound and frozen in isopentan. 15 µm thick cryosections were then stained with CD11b (blue), Gr1 (red), Ly6G (green) and DAPI (grey) before observation under a Zeiss LSM 510. Bar: 0.25mm. Figure 4: Brucella CβG induces a transient neutrophil recruitment at 12 h post-injection. A. Cells were extracted from mouse ears at 2 h, 6 h, 12 h, 24 h and 48 h following injection with PBS (white bars), E. coli LPS (black bars) or Brucella CβG (grey bars) (see figure 3) and neutrophils were quantified by flow cytometry. Mean ± SD of 3 independent experiments is represented here. B. Dot-plots of neutrophil recruitment to the ear at 12 h post-injection with PBS, LPS or CβG. Cells were gated on CD45 +, MHC II -, CD11b +, Ly6C +. Neutrophils 84

represented in blue population are negative for F4/80 and positive for Ly6G. C. Dot-plots of neutrophil recruitment to the ear at 24 h post-injection with PBS, LPS or CβG. Cells were gated on CD45 +, MHC II -, CD11b + and Ly6C +. Neutrophils (in blue) are negative for F4/80 and positive for Ly6G. Three independent experiments were carried out, and one representative is shown here. Figure 4: Brucella CβG induces a transient neutrophil recruitment at 12 h post-injection. A. Cells were extracted from ears injected with PBS (white bars), E. coli LPS (black bars), or Brucella CβG (grey bars) as shown in figure 3. Neutrophils were quantified by flow cytometry at 2 h, 6 h, 12 h, 24 h and 48 h post-injection. Mean ± SD of 3 independent experiments is represented here. B. Dot-plots of neutrophil recruitment into the ear at 12 h PBS, LPS or CβG post-injection. Cells were gated onto CD45 +, MHC II -, CD11b +, Ly6C +. Neutrophils are the blue population negative for F4/80 and positive for Ly6G. C. Dot-plots of neutrophil recruitment into the ear at 24 h PBS, LPS or CβG post-injection. Cells were gated onto CD45 +, MHC II -, CD11b + and Ly6C +. Neutrophils are the blue population negative for F4/80 and positive for Ly6G. Three independent experiments were carried out, and one representative is shown here. Table 1: 34 transcripts upregulated in CβG-stimulated DC versus LPS-stimulated DC Fold change Gene Symbol Gene Description Genbank number 16,2960457 DFNA5 deafness, autosomal dominant 5 NM_004403.2 14,0900781 CFB complement factor B NM_001710.4 13,60613 MAOA monoamine oxidase A, nuclear gene encoding mitochondrial protein NM_000240.2 13,1097147 IL6 interleukin 6 NM_000600.1 12,580871 ITGA9 integrin, alpha 9 NM_002207.2 11,9337847 TNIP3 TNFAIP3 interacting protein 3 NM_024873.3 11,8534761 GJB2 gap junction protein, beta 2, 26kDa NM_004004.4 85

9,83164123 PTGS2 prostaglandin-endoperoxide synthase 2 (prostaglandin G/H synthase and NM_000963.1 cyclooxygenase) 9,7861406 BATF basic leucine zipper ion factor, ATF-like NM_006399.2 8,4711145 IL1F9 interleukin 1 family, member 9 NM_019618.2 8,45774549 IL2RA interleukin 2 receptor, alpha NM_000417.1 8,32768687 CXCR7 chemokine (C-X-C motif) receptor 7 NM_020311.2 8,239224 PLAT plasminogen activator, tissue NM_000930.2 8,18870443 ADARB1 adenosine deaminase, RNA-specific, B1 NM_001033049.1 7,96050487 CLGN calmegin NM_004362.1 6,77065691 KCNJ2 potassium inwardly-rectifying channel, subfamily J, member 2 NM_000891.2 6,63411098 TNFAIP6 tumor necrosis factor, alpha-induced protein 6 NM_007115.2 6,56358736 HEY1 hairy/enhancer-of-split related with YRPW motif 1 NM_001040708.1 6,22630244 PLAC8 placenta-specific 8 NM_016619.1 6,21326365 TFRC transferrin receptor (p90, CD71) NM_003234.1 5,7156289 UPB1 ureidopropionase, beta NM_016327.2 5,6420305 AQP9 aquaporin 9 NM_020980.2 5,63400352 ZP3 zona pellucida glycoprotein 3 (sperm receptor) NM_007155.4 5,6153173 TNFRSF21 tumor necrosis factor receptor superfamily, member 21 NM_014452.3 5,38003118 LILRA3 leukocyte immunoglobulin-like receptor, subfamily A (without TM domain), member 3 NM_006865.2 5,29230455 ITM2C integral membrane protein 2C NM_001012516.1 5,22805328 SMPDL3A sphingomyelin phosphodiesterase, acid-like 3A NM_006714.2 5,16244294 RASGRP1 RAS guanyl releasing protein 1 (calcium and DAG-regulated) NM_005739.2 5,0282775 CXCL2 chemokine (C-X-C motif) ligand 2 NM_002089.3 4,92399264 SLC11A1 solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1 NM_000578.3 4,89471027 TRAF3IP2 TRAF3 interacting protein 2 NM_147686.1 4,85717514 SOCS2 suppressor of cytokine signaling 2 NM_003877.3 4,84210367 PTX3 pentraxin-related gene, rapidly induced by IL-1 beta NM_002852.2 4,50411755 ADAMDEC1 ADAM-like, decysin 1 NM_014479.2 Table 2: Primers used for qpcr experiments Name Sens Sequence HPRT Forward 5'-3' AGCCCTCTGTGTGCTCAAGG HPRT Reverse 5'-3' CTGATAAAATCTACAGTCATAGGAATGGA Ptgs2 Forward 5'-3' ACCTCTGCGATGCTCTTCC 86

Ptgs2 Reverse 5'-3' TCATACATTCCCCACGGTTT SOCS3 Forward 5'-3' CCTTCAGCTCCAAAAGCGAGTAC SOCS3 Reverse 5'-3' GCTCTCCTGCAGCTTGCG CXCL2 Forward 5'-3' GCGGTCAAAAAGTTTGCCTTG CXCL2 Reverse 5'-3' CTCCTCCTTTCCAGGTCAGTT KC Forward 5'-3' CAGCCACCCGCTCGCTTCTC KC Reverse 5'-3' TCAAGGCAAGCCTCGCGACCAT tnfaip6 Forward 5'-3' TTCCATGTCTGTGCTGCTGGATGG tnfaip6 Reverse 5'-3' AGCCTGGATCATGTTCAAGGTCAAA Figures 87

88

89

90

91

II. D. DES DC BTPB, UNE PROTEINE CAPABLE DE MODULER L ACTIVATION II. D. 1. Introduction Nous avons décrit précédemment dans le laboratoire une protéine de Brucella, BtpA, qui est capable d interférer avec les voies de signalisation en aval des TLR et de moduler l activation des DC. BtpA pourrait interagir avec des molécules adaptatrices de TLR comme TIRAP ou MyD88 pour inhiber la réponse TLR [133, 134]. De plus, cette protéine pourrait jouer un rôle dans la régulation de la réponse adaptative en jouant sur les LT CD8 +. Ces dernières années, de nombreuses études ont montré que les protéines bactériennes à domaines TIR existent chez Salmonella, TlpA [221], un homologue à TlpA est PdTlp chez Paracoccus denitrificans [222], chez E. coli TcpC [132] ou encore chez Yersinia pestis YpTdp [223]. Certaines de ces protéines ont été montrées comme interagissant avec des acteurs majeurs des voies de signalisation des TLR comme MyD88, TLR4 ou encore NF-κB [221, 222]. Cependant, il n est pas exclu que ces protéines à domaine TIR puissent jouer d autres rôles au sein des cellules que la régulation et le contrôle d une partie du système immunitaire. Ici, nous décrivons une seconde protéine de Brucella, BtpB, possédant aussi un domaine TIR. Nous avons cherché à caractériser le rôle de cette nouvelle protéine dans la réponse immunitaire à l infection ainsi que dans la régulation de l inflammation. 92

II. D. 2. Article 93

CELLULAR AND INFECTION MICROBIOLOGY ORIGINAL RESEARCH ARTICLE published: 08 July 2013 doi: 10.3389/fcimb.2013.00028 BtpB, a novel Brucella TIR-containing effector protein with immune modulatory functions Suzana P. Salcedo 1,2,3,4, María I. Marchesini 5,ClaraDegos 1,2,3, Matthieu Terwagne 6, Kristine Von Bargen 1,2,3, Hubert Lepidi 7, Claudia K. Herrmann 5, Thais L. Santos Lacerda 1,2,3,4, Paul R. C. Imbert 4, Philippe Pierre 1,2,3, Lena Alexopoulou 1,2,3, Jean-Jacques Letesson 6, Diego J. Comerci 5 and Jean-Pierre Gorvel 1,2,3 * 1 Aix-Marseille Univ UM 2, Centre d Immunologie de Marseille-Luminy, Marseille, France 2 INSERM U 1104, Marseille, France 3 CNRS UMR 7280, Marseille, France 4 Bases Moléculaires et Structurales des Systèmes Infectieux, CNRS UMR 5086, Institute of Biology and Chemistry of Proteins, Université Lyon 1, Lyon, France 5 Instituto de Investigaciones Biotecnológicas Dr. Rodolfo A. Ugalde (IIB-INTECH), Universidad Nacional de San Martín, Consejo Nacional de Investigaciones Científicas y Técnicas, San Martín, Buenos Aires, Argentina 6 URBM, NARILIS, University of Namur (FUNDP), Namur, Belgium 7 Laboratoire d anatomie pathologique-neuropathologique, Aix-Marseille Université, Marseille, France Edited by: Rey Carabeo, University of Aberdeen, UK Reviewed by: Jean Celli, NIAID, NIH, USA Renee M. Tsolis, University of California-Davis, USA Nelson Gekara, Umea University, Sweden *Correspondence: Jean-Pierre Gorvel, Centre d Immunologie de Marseille-Luminy, Parc Scientifique et Technologique de Luminy, Case 906, 13288 Marseille Cedex 09, France e-mail: gorvel@ciml.univ-mrs.fr Joint-first authors. Several bacterial pathogens have TIR domain-containing proteins that contribute to their pathogenesis. We identified a second TIR-containing protein in Brucella spp. that we have designated BtpB. We show it is a potent inhibitor of TLR signaling, probably via MyD88. BtpB is a novel Brucella effector that is translocated into host cells and interferes with activation of dendritic cells. In vivo mouse studies revealed that BtpB is contributing to virulence and control of local inflammatory responses with relevance in the establishment of chronic brucellosis. Together, our results show that BtpB is a novel Brucella effector that plays a major role in the modulation of host innate immune response during infection. Keywords: Brucella, TIR domain, Btp1/BtpA, TLR, DC, NF-κB INTRODUCTION Innate immune recognition of microbial components is critical for the onset of an appropriate immune response against invading pathogens. Key contributors include the toll-like receptor (TLR)/IL-1R superfamily characterized by the presence of a conserved region designated TIR domain located in the cytosolic part of each TLR. The TIR domain is critical for protein-protein interactions between TLRs with the corresponding TIR-containing adaptors, which couple downstream protein kinases. This signaling cascade ultimately leads to activation of specific transcription factors such as nuclear factor-κb (NF-κB) and production of inflammatory mediators. Although a variety of TLR receptors have been described, in humans the most relevant for recognition of bacterial molecules are TLR2, TLR4, TLR5, and TLR9. In addition to TLRs and their adaptors, TIR domains are present in plant resistance proteins that mediate hypersensitive responses to pathogens, as well as in a variety of bacteria, including species present in the human gut microbiota, soil bacteria and human pathogens (Spear et al., 2009; Zhang et al., 2011). Their evolutionary history is complex and their role in interaction with eukaryotic hosts remains mostly uncharacterized. Nevertheless, in a number of bacterial pathogens, bacterial TIR-containing proteins have been implicated in virulence or control of cellular responses. Salmonella enterica serovar Enteritidis TlpA is capable of reducing NF-κBactivationbyTLR4, IL-1R and MyD88-dependent pathways and to contribute to control of IL-1β secretion during infection (Newman et al., 2006). In the case of uropathogenic E. coli CFT073, the TIR-containing protein TcpC is able to interfere with TLR4 and TLR2 signaling by targeting MyD88 (Cirl et al., 2008) butalsotoinhibit TRIF- and IL-6/IL-1 dependent pathways (Yadav et al., 2010). During infection, TcpC is implicated in the control of secretion of TNF-α and IL-6 and tcpc mutants show a defect in intracellular replication in a mouse model of pyelonephritis. The Yersinia pestis TIR-containing protein YpTdp interacts with MyD88 to reduce IL-1β- and LPS-dependent signaling and to contribute to modulation of cytokine secretion during infection (Spear et al., 2012). In the case of Brucella spp. two groups independently reported on the role of a TIR domain containing protein in control of TLR signaling (Cirl et al., 2008; Salcedo et al., 2008). Naming of this protein as Btp1 or TcpB, respectively, by two distinct laboratories has led to some confusion in the literature and has been misinterpreted by some as two independent proteins. Since neither Btp1 Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 1

Salcedo et al. Brucella TIR protein B nor TcpB conforms to the international guidelines for bacterial nomenclature we will hereafter designate Btp1/TcpB as BtpA. BtpA is present in B. abortus 2308 (BAB1_0279), B. abortus 9-941 (BruAb1_0274) and B. melitensis 16 M (BMEI1674) but is absent from B. suis 1330. Cirl et al. described that ectopically expressed BtpA cloned from B. melitensis 16 M is able to interfere with TLR4 and TLR2 signaling by directly interacting with MyD88. Several reports have proposed that BtpA targets the adaptor protein MAL/TIRAP (Radhakrishnan et al., 2009; Sengupta et al., 2010). Direct comparison of the in vitro interaction between BtpA and either MyD88 or TIRAP shows a stronger interaction with MyD88 (Chaudhary et al., 2011). BtpA has been shown to bind phosphoinositides at the plasma membrane (Radhakrishnan et al., 2009) but also to induce ubiquitination of TIRAP (Sengupta et al., 2010). In accordance to its modulation of TLR function, previous work from our laboratory described the role of the BtpA from B. abortus in the control of dendritic cell (DC) activation during infection (Salcedo et al., 2008). Purified BtpA was also shown to inhibit CD8 + T cell-mediated killing suggesting it may also control adaptive immune responses (Durward et al., 2012). Here we present a novel Brucella effector with a TIR domain that we designated as BtpB. We show that BtpB efficiently inhibits TLR signaling and contributes to control of DC activation. Together, Brucella TIR-containing proteins BtpA and BtpB modulate host inflammatory responses during infection. RESULTS IDENTIFICATION OF A SECOND Brucella TIR DOMAIN-CONTAINING PROTEIN Analysis of the Brucella genome revealed the presence of a second TIR domain-containing protein (BAB1_0756) that we have designated BtpB (Figure 1A). We choose to continue with the Btp nomenclature to avoid any confusion with the tcpb gene necessary for conjugative transfer in Clostridium perfringens (Parsons et al., 2007). Search for conserved domains in BtpB revealed the presence of a C terminal TIR domain (aa 144-256) that belongs to the Pfam family TIR_2 (E value 1.9 e 11 ), a family of bacterial Toll-like receptors. TIR domains share conserved motifs called box 1 (F/Y-DAFISY), box2 (GYKLC-RD-PG) and box 3 (W residue surrounded by basic amino acids). Sequence comparison of BtpB TIR domain with the human TIR-containing proteins MAL, MyD88, TLR2 and TLR4 showed sequence similarity and conservation of box 1, essential for signaling (Rana et al., 2013) (Figure 1A). Unlike BtpA, BtpB is present in all sequenced Brucella strains, including B. suis 1330 (BR0735), B. abortus 2308 (BAB1_0756), B. abortus 9-941 (BruAb1_0752) and B. melitensis 16 M (BME1216). Alignment of the btpb sequences derived from different Brucella strains revealed 4 different annotations for the start codon (highlighted in red in Figure 1B). Analysis of the 18 to +18 nucleotides around the ATG/GTG (Kolaskar and Reddy, 1985) predicted as the most likely start codon the second methionine highlighted in Figure 1B. This open reading frame has been annotated for B. abortus 9-941 and encodes a 292 amino acid protein, BtpB (1-292). The additional annotated start codons include the first highlighted methionine, the valine (GTG) and the B. melitensis methionine resulting in proteins of either 325, 277 or 178 amino acids. None of them scored high enough to be considered as likely start codons. Comparison of all Brucella sequences available revealed only one BtpB (1-178), in B. melitensis 16 M, whereas the majority correspond to BtpB (1-292). In consequence, we decided to use in this study the BtpB (1-292). We first investigated the ability of BtpB to interfere with TLR signaling using an in vitro NF-κB-dependent luciferase reporter system. BtpB was able to inhibit TLR2, TLR4 and TLR9 signaling (Figure 2A) evenmoreefficientlythanbtpa(salcedo et al., 2008). This inhibition was independent on the first 114 amino acids as both BtpB (1-178) (Figure 2A), as well as, BtpB (1-292) (Figure 2B) strongly inhibited TLR signaling. BtpB was also able to inhibit flagellin-induced TLR5 signaling (Figure 2B). These results suggest that BtpB may interfere with a common molecule of these TLR pathways, such as MyD88. Consistently, BtpB did not reduce TLR3-dependent signaling which does not involve the adaptormyd88 (Figure 2C). In addition, we observed by directed yeast two-hybrid that BtpB was able to interact with MyD88 (Figure 2D). BtpA was also able to interact with MyD88 by yeast-two hybrid as previously shown by pull-down and proteinfragment complementation assays (Cirl et al., 2008; Chaudhary et al., 2011). Neither BtpA nor BtpB interacted with any of the TLR1 to TLR10 TIR domains nor with the adaptors TIRAP or TRAM. As BtpA was previously shown to reduce TLR2 and TLR4 signaling but not TLR9 (Cirl et al., 2008; Salcedo et al., 2008) we investigated its ability to interfere with TLR5, which is also dependent on MyD88. BtpA was able to significantly reduce TLR5 signaling, following stimulation with S. typhimurium flagellin (Figure 2E). Overall, our results show that BtpB is a potent inhibitor of TLR signaling in vitro, which may result from binding to MyD88. BtpB IS TRANSLOCATED INTO HOST CELLS In order for BtpA and BtpB to target TLR pathways they would have to be exported across the bacterial membranes and the vacuolar membrane into the host cell cytosol. To test this hypothesis we analysed the translocation of BtpA and BtpB fused at their N- terminus with the TEM-1 β-lactamase during infection of RAW macrophage-like cells. This method has been successfully used to establish translocation of several Brucella effectors, namely VceA, VceC and RicA (de Jong et al., 2008; de Barsy et al., 2011) and is traditionally carried out in live cells. RAW cells were used in order to achieve high rates of infection. VceC and VceA were included as positive controls. We could detect BtpA translocation into host cells at 4 h and 24 h after inoculation (Figures 3A,B). Translocated BtpB was detected in less than 0.5% of infected cells at 4 h and did not significantly increase at 24 h. In an attempt to try to enhance the sensitivity of this assay, we carried out the same experiments in fixed samples with observation of FRET within 15 min of fixation, which enhances the shift to 450 nm (Nothelfer et al., 2011). As in live cells, BtpA and to a lower extent BtpB were translocated into host cells at 24 h after infection (Figures 3C,D). Since the overall percentage of cells showing translocated effectors is very low with this assay, even following fixation, we analysed translocation of BtpA and BtpB fused to the adenylate cyclase CyaA (Figure 3E). In addition, we used a constitutive promoter of B. abortus bcsp31 gene to enhance expression, since this alternative Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 2

Salcedo et al. Brucella TIR protein B FIGURE 1 Identification of BtpB. (A) Identification of BtpB as bacterial member of TLR/IL-1R (TIR) family. Comparison of the predicted amino acid sequences of the TIR domain of BtpB with BtpA and the human members of the TIR family: MAL, MyD88, TLR2 and TLR4. The alignment was constructed with T-Coffee::advanced server from EMBnet (http://www.ch.embnet.org) and coloring scheme corresponds to standard ClustalX in which each residue in the alignment is assigned a color if the amino acid profile at each position meets a minimum criteria specific for the residue type. Box 1 corresponds to the signature sequence of the TLR family. (B) Alignment of BtpB amino acid sequences for B. abortus 2308 (BAB1_0756), B. suis 1330 (BR0735), B. abortus 9-941 (BruAb1_0752) and B. melitensis 16M (BME1216). The annotated starting codons (Methionine/Valine) are highlighted in red. Amino acid differences are shaded in red. approach was successfully used with the Brucella effector protein BPE123 (Marchesini et al., 2011), which was includedas a positive control in our experiments. In this system, any value of camp bellow 1500 2000 fmol/ml corresponds to background (dotted line in Figure 3E) and is not indicative of translocation as determined after performing an exhaustive screening for the identification of Brucella abortus type IV secretion system (T4SS) substrates (Marchesini et al., 2011). We found that at 4 h after infection both BtpA and BtpB were translocated into J774.A1 macrophage-like cells (Figure 3E). Interestingly, translocation of BtpA fused with the CyaA seems to depend on the position of the tag as only BtpA with C-terminal CyaA was efficiently translocated into host cells at early stages of the infection. In contrast, for BtpB, the presence of the CyaA tag on the C-terminus reduced translocation (Figure 3E). To determine if the translocation of BtpA and BtpB was dependent on the Brucella VirB T4SS, cells were infected with the virb mutant carrying either TEM- or CyaA-fused Btp proteins. We could not detect any differences between wild type and virb mutant using the TEM-1 β-lactamase assay (Figures 3A D). In sharp contrast, translocation of BtpA-CyaA and CyaA-BtpB was clearly reduced in a virb genetic background, indicating that Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 3

Salcedo et al. Brucella TIR protein B FIGURE 2 BtpB interferes with TLR signaling. (A) HEK293 cells were transiently transfected for 24 h with the luciferase reporter vector and either TLR2, TLR4 and TLR9, in the presence or the absence of the 178 amino acid BtpB (50 ng). Cells were then stimulated with the appropriate ligand (PAM, LPS and CpG) for 6 h before measurement of luciferase activity. White bars correspond to negative control, black bars to cells stimulated with the appropriate ligand and grey bars to cells transfected with BtpB and stimulated with the ligand. Data represent the means ± standard errors of relative luciferase activity obtained from triplicates of a representative experiment. (B) Luciferase activity in the presence or absence of the BtpB (1-292) (red bars). TLR5 was also included and stimulated with Flagellin from S. typhimurium (Fl-ST) and (C) TLR3 following stimulation with poly(i:c). (D) Yeast containing Gal4 BD- and Gal4 AD-fusion proteins were selected on synthetic medium lacking leucine (Leu) and tryptophan (Trp) (left panel). Protein interactions were identified on synthetic medium lacking histidine (His) and supplemented with 20 mm 3AT (middle panel). Growth on this medium indicates interaction between fusion proteins. The blue yeast colonies observed in the β-galactosidase expression filter assay indicate interaction between the fusion proteins (right panel). BD and AD indicate empty vectors and were used as negative controls, while MyD88 homodimerization was used as positive control. (E) Luciferase activity in cells transfected with TLR5 in the absence or presence of 100 ng and 50 ng of BtpA (275 aa). P 0.001 are denoted with ; P 0.01 are denoted with and P between 0.01 and 0.05 are denoted with. delivery of both proteins is dependent on the T4SS. We conclude that BtpA and BtpB are translocated into host cells and may constitute substrates for the VirB T4SS. BtpB REPLICATION WITHIN MURINE BONE MARROW-DERIVED DCs To determine the role of BtpB during infection we infected murine bone marrow-derived DCs with wild type Brucella,aswell as, with a btpa, btpb or btpabtpb mutant strains. No attenuation wasobservedasthebtpabtpb replicated to equivalent levels of the wild type B. abortus strain (Figure 4A). The survival curves for thesinglemutantsoverlapwiththatofthewildtype(figure 4A, right panel). As previously described for BtpA, murine DCs infected with the btpb mutant showed higher level of MHC class II surface Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 4

Salcedo et al. Brucella TIR protein B FIGURE 3 BtpB is translocated into host cells during infection. (A) RAW macrophages were infected with wild type (wt) or virb9 B. abortus strains carrying N-terminal TEM-1 fused VceA, VceC, BtpA, and BtpB for 4 h and 24 h. Data represents the means ± standard errors of the percentage of cells with coumarin fluorescence from 5 independent experiments. (B) Representative confocal images of RAW cells infected with either wilt-type B. abortus (wt) or virb9 mutant carrying TEM-fused BtpA, at 24 h after inoculation. Appearance of blue cells is indicative of translocated TEM lactamase. (C) and (D) Analysis of TEM-1 translocation assay for fixed samples of VceA, VceC, BtpA, and BtpB 24 h after infection. (E) Intracellular camp levels in J774.A1 cells infected for 4 h with isogenic strains with a functional (wt) or non-functional VirB system (virb10) expressing Btp proteins fused to CyaA. Non-infected cells and a wild type strain expressing the CyaA domain alone (pcyaa) were included as negative controls. A wild type strain expressing BPE123-CyaA was included as a positive control. Means and SD are shown for one representative out of three independent experiments. Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 5

Salcedo et al. Brucella TIR protein B FIGURE 4 Role of BtpB in control of DC activation. (A) BMDCs infected with wild type B. abortus or the btpabtpb mutant (left panel) and the single mutants (right panel) were lysed and intracellular CFUs enumerated at different times after inoculation. (B) Representative images of BMDCs infected with either the wild type, btpb or btpabtpb mutants for 24 h. Cells were labeled for MHC class II (red) and surface expression is of a representative area is shown in zoom inlets. (C) Quantification of the percentage of DCs containing DALIS after 24 h of infection with wild type B. abortus (wt), btpb or btpabtpb mutant. (D) Flow cytometry of the surface expression of MHC class II, CD40, CD80 and CD86 at 24 h post-infection. Data are normalized to wt values. (E) Analysis of TNF-α and (F) IL-12 (p40/p70) secretion measured by ELISA from the supernatant of DCs 24 h after inoculation. All the results correspond to the means ± standard errors of 4 independent experiments. P 0.001 are denoted with ; P 0.01 are denoted with and P between 0.01 and 0.05 are denoted with. expression and higher percentage of formation of aggresomelike induced structures (DALIS) that transiently appear during the process of activation of these immune cells (Figures 4B,C) (Lelouard et al., 2002). However, there was no additive effect of depletion of both btpa and btpb as the btpabtpb mutant did not show an increased phenotype compared to single mutant. Flow cytometry analysis of infected cells did not reveal a statistically significant increase in CD40, CD80 and CD86 surface Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 6

Salcedo et al. Brucella TIR protein B expression in DCs infected with btpb mutant when compared to the wild type at 8 h post-infection. At 24 h post-infection there was a significant increase in MHC class II surface expression in DCs infected with btpb mutant relative to those infected with the wild type (Figure 4D, white bars) consistent with our microscopy observations (Figure 4B). In the case of DCs infected with btpabtpb mutant, CD40 and CD80 co-stimulation markers were up-regulated (Figure 4D,blackbars). In terms of cytokine secretion, BtpB did not seem to be involved in the control of TNF-α secretion during infection of murine DCs (Figure 4E).The increase in TNF-α secretion observed for btpabtpb is probably due to a lack of BtpA, previously shown to be involved in the control of secretion of this cytokine (Salcedo et al., 2008). However, an increase in the level of total IL-12 (p40/p70) secreted during infection was observed in the case of the btpb mutant compared to the wild type 24 h after infection (Figure 4F). The difference between btpb and btpabtpb mutants is not significant. These results suggest that BtpB is contributing to the control of the inflammatory response induced in infected DCs in vitro. BtpB CONTROLS NF-κB TRANSLOCATION IN DCs In order to analyse the effect of the BtpB effector on the early stages of DC activation, translocation of NF-κBwas monitored by immunofluorescence microscopy during the course of the infection. As early as 2 h post-infection, bone marrow-derived DCs infected with the btpb mutant showed an increased translocation of NF-κB into the nucleus compared to those infected with wild type B. abortus (Figure 5). The btpb mutant phenotype was rescued by expression of BtpB from a plasmid confirming the role of BtpB in the control of NF-κB translocation into the nucleus. These results confirm that BtpB has an effect on the induction of inflammatory responses during Brucella infection. ROLE OF BtpB IN THE MOUSE MODEL OF BRUCELLOSIS To further investigate the role of BtpB during infection we carried out in vivo studies. BtpA B. melitensis mutants were previously shown to have enhanced survival in immuno-compromised Interferon Regulatory Factor-1 (IRF-1) / mice (Radhakrishnan et al., 2009) inoculated intra-peritoneally (i.p.), a lethality model that has been used for studying Brucella virulence (Ko et al., FIGURE 5 Modulation of NF-κB translocation to the nucleus during Brucella infection. Bone marrow-derived DCs were infected with wild type (wt) B. abortus, btpa, btpb and btpabtpb mutants as well as btpb mutant carrying the complementing plasmid (pbtpb) for 2h and processed for immunofluorescence confocal microscopy. Cells were labeled for CD11c (cyan) and p65 NF-κB (red). Bacteria were labeled with anti-lps antibody followed by FITC secondary and nuclei with TOPRO3. Salmonella infected cells were used as a positive control. (A) Data corresponds to means ± standard errors of 4 independent experiments. (B) Representative images obtained by confocal microscopy are shown for DCs infected with wild type, btpb mutant, btpbpbtpb complemented strain and btpabtpb mutant. Scale bars correspond to 5µm. P 0.001 are denoted with ; P 0.01 are denoted with and P between 0.01 and 0.05 are denoted with. Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 7

Salcedo et al. Brucella TIR protein B 2002). We therefore, inoculated IRF / i.p. with either btpa, btpb or btpabtpb B. abortus mutants. Although the btpb mutant had no defect in intracellular replication in cultured cells in vitro, it showed an attenuation phenotype in IRF-1 / mice. Mice infected with the btpb mutant survived longer than those infected with the wild type B. abortus and synergistic effect was observed for the double btpabtpb mutant (Figure 6A, P < 0.005), despite equivalent bacterial CFU counts in the spleen at each sampling date after infection (median of 1.05 10 8 CFU/spleen for wild type versus 6.5 10 7 CFU/spleen for the btpabtpb mutant). To better study the role of BtpB in brucellosis in vivo, we inoculated wild type BALB/c mice i.p. and enumerated the bacterial load at 30, 60, 90, and 130 days post-infection. No significant differences in bacterial CFU counts between the wild type Brucella and the btp mutants were observed at different stages of the infection (30, 60, 90, and 130 days). Data at 60 days post-infection is shown as an example (Figure 6B). We then performed histological examination of spleens obtained from wild type BALB/c mice infected with the btpa, btpb, btpabtpb mutants or the wild type B. abortus strains to quantify granuloma formation, which usually reflects the host s ability to develop a protective immune response. No granuloma was seen in the spleen of non-infected mice. In infected mice, granulomas were detected in splenic red pulp. A significantly higher number of granulomas was observed following btpb and btpabtpb infection (after 60 days) compared to the wild type Brucella (Figure 6C). Inflammatory granulomas showed a similar organization in all populations of infected mice and were composed mainly of macrophages and a few lymphocytes (Figure 6D). Bacteria were detected by immunohistochemistry in the spleen of mice infected with wild type B. abortus or with the btpabtpb mutant (Figure 6E). They were seen as coarse granular immune-positive FIGURE 6 Role of BtpB during Brucella infection in the mouse model of brucellosis. (A) Susceptibility of IRF-1 / to B. abortus 2308(wt), btpa, btpb and btpabtpb mutant (n = 9 per group). Infected mice were monitored daily for survival. Mice infected with btpb and tpabtpb survived longer than wild type Brucella infected mice (P = 0.0433 and P = 0.0152, respectively). (B) Persistence of B. abortus 2308(wt), btpa, btpb or btpabtpb mutants in spleens of wild type BALB/c infected mice at 60 days p.i. Each symbol represents an animal and the median values are marked by horizontal bold lines. (C) Analysis of granuloma formation in the spleens of wild type BALB/c mice infected for 60 days with wild type B. abortus, btpa, btpb or btpabtpb mutants. Data represent means ± standard deviations of 4 or 5 mice. (D) Representative image from the spleen of a mouse infected with btpabtpb mutant (hematoxylin-eosin, original magnification 400). (E) Bacteria were revealed by immunostaining in the spleen of wild type BALB/c mice infected by btpabtpb mutant of B. abortus. Macrophages present in inflammatory granulomas in the red pulp are packed with coarse immunopositive material (hemalun counterstain, original magnification 400). P 0.01 are denoted with. Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 8

Salcedo et al. Brucella TIR protein B material associated with cells, which had the morphology of macrophages. These results are consistent with a role for BtpB in the control of inflammatory response during Brucella infection in vivo. DISCUSSION Previous work from our laboratory demonstrated a role for BtpA in control of DC activation. Here we show that Brucella contains a second TIR-domain protein called BtpB that is translocated into host cells and which participates in the control of the inflammatory response during Brucella infection. In vitro, BtpB is a potent inhibitor of TLR2, TLR4, TLR5 and TLR9. Together with BtpA, BtpB contributes to the control of DC activation during infection. Using the TEM-1 lactamase and CyaA assays we were able to detect BtpA and BtpB translocated into the cytosol as early as 4 h after infection of RAW and J774 macrophages. This is an essential step to enable BtpA and BtpB to cross the bacterial and vacuolar membranes to reach their host cellular targets during infection. It would be interesting to localize the translocated proteins during infection. We have not been able to detect neither 2HA- nor 3FLAG-tagged BtpA and BtpB by immunofluorescence microscopy. It is possible that the amounts of translocated BtpA and BtpB are too low or perhaps these proteins are quickly degraded once they reach the host cytosol. It is important to note that using the TEM-1 lactamase assay alone we were unable to detect translocation of BtpB and any differences between wild type and virb mutant in BtpA translocation. It is possible that the low sensitivity of the TEM-1 lactamase compared to the CyaA assay makes this methodology inappropriate to assess VirB dependency in the case of effectors translocated at low levels. We conclude from our results that BtpA and BtpB are likely substrates of the VirB T4SS. In this study, we found that B. abortus lacking btpa and btpabtpb mutants showed an increased survival time in the IRF-1 / mouse model, highlighting the importance of these TIRcontaining proteins in virulence. Similar results were obtained for B. melitensis lacking BtpA, which is defective in systemic spread at early stages of infection (Radhakrishnan et al., 2009). However, the use of such a severely immune-compromised mouse model hampers detailed analysis of the role of these proteins in control of inflammatory responses during infection. Therefore, we proceeded with our in vivo studies using immune-competent mice. We found that absence of BtpA and/or BtpB leads to increased granuloma formation in wild type mice, probably restricting bacterial dissemination as a consequence of the inability of the mutants to modulate the inflammatory response. Infection of DCs with B. abortus lacking BtpB revealed that this effector protein is contributing to the modulation of the inflammatory response during infection. Interestingly, significant differences were observed between BtpA and BtpB. For example, BtpA had an impact on TNF-α secretion (Salcedo et al., 2008) but not BtpB, which affected surface expression of MHC class II and co-stimulatory molecules that we had not previously seen with BtpA. These differences may be due to different kinetics of translocation and time of action of each effector or perhaps the kinetics of the cellular processes affected. These differences could also be explained by specific targeting of host pathways. Interestingly, translocation of VceC results in enhanced pro-inflammatory responses as a result of the induction of the unfolded protein response by this T4SS effector (de Jong et al., 2012). This suggests that VirB effectors can have opposing effects, resulting in either activation of host immune responses or specific inhibition of inflammatory pathways. These differences may represent host cell or tissue specificity or simply reflect different stages of disease. It is now crucial to undertake a more global analysis of the specific contribution of these effectors during infection and a better characterization of host immune responses elicited in vivo. It is also possible that some of the phenotypes observed with effectors are simply an indirect or secondary effect of their action on eukaryotic cells during infection. Defining at the molecular level the effector cellular targets and analysing their contribution during infection will hopefully shed some light on these issues. The host interacting partner of BtpA remains controversial. BtpA has been shown to induce degradation of phosphorylated TIRAP by enhancing its poly-ubiquitination (Sengupta et al., 2010) and to efficiently block TIRAP-induced NF-κB activation (Radhakrishnan et al., 2009). Together, these studies present TIRAP as the main target of BtpA whereas other groups have shown a direct interaction with MyD88 (Cirl et al., 2008; Chaudhary et al., 2011). Although comparison of the ability of BtpA to interact with TIRAP and MyD88 revealed a stronger binding to MyD88 (Chaudhary et al., 2011), surprisingly this interaction was dependent on the Death Domain of MyD88 and not the TIR domain. By yeast-two hybrid we found that both BtpA and BtpB can interact with MyD88. In the case of BtpB this result could explain its ability to block TLRs that are dependent on MyD88 signaling but not TLR3, which is dependent on the adaptor TRIF. Although inhibition of TLR2 and TLR4 by BtpA has been described, we did not detect any inhibition of TLR9 (Salcedo et al., 2008), which would be expected if BtpA was blocking MyD88. It is possible that inhibition of TLR9 by BtpA requires higher levels of expression of BtpA and could not be detected with our assay. Consistently, BtpA interfered with TLR5 signaling which is dependent on MyD88. Further work is now required to understand the molecular mechanism by which BtpA controls TLR activation, which may involve interaction and/or competition with both MyD88 and TIRAP. In addition to control of inflammatory responses, BtpA has been shown to interact with phosphoinositides at the plasma membrane and modulate microtubule dynamics (Radhakrishnan et al., 2009, 2011). Ectopically expressed BtpA localizes to microtubules. These could constitute important activities that may also have a consequence on control of the inflammatory response, for example by misplacing specific adaptor molecules within the cell. In addition, these data indicate that BtpA may have additional eukaryotic targets yet to be identified. It will be interesting to evaluate during infection the contribution of these different functions of BtpA described in vitro and determine if they are dependent on the TIR domain or if other domains are contributing to assigning multiple functions to this effector. Our results strongly implicate BtpB in the control of host inflammatory responses during Brucella infection. However, it is possible BtpB has additional functions as it has been described for BtpA. We are currently determining if multiple pathways are targeted by Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 9

Salcedo et al. Brucella TIR protein B BtpB to better understand the role of this novel effector during Brucella infection. MATERIALS AND METHODS BACTERIAL STRAINS The bacterial strains used in this study were S. enterica serovar Typhimurium strain 12023, smooth virulent B. abortus strain 2308 (Pizarro-Cerdá et al., 1998) and the isogenic mutants virb9 (Celli et al., 2005), virb10 (Sieira et al., 2000), btpa (Salcedo et al., 2008), btpb (this study) and btpa btpb (this study). In the case of Brucella, green fluorescent protein (GFP)-expressing derivatives contain a pbbr1mcs-2 (Kovach et al., 1995) derivative expressing the gfp-mut3 gene under the control of the lac promoter. Brucella strains were grown in tryptic soy broth (TSB; Sigma-Aldrich) and Salmonella in Luria Bertani (LB) medium. For infection, we inoculated 2 ml of media for 16 h at 37 Cup to an optical density (OD 600 nm ) of approximately 2.0 (Celli et al., 2003). Salmonella strains were cultured 16 h at 37 C with aeration to obtain stationary phase cultures. CONSTRUCTION OF btpb AND btpabtpb MUTANTS The btpb gene (BAB1_0756) was amplified from B. abortus genomic DNA using primers 5 -acgcgacctttccggctccctt-3 and 5 -ttcggctagacaggaatgcatg-3 and ligated to pgem-t-easy vector (Promega) to generate pgem-tbtpb. The plasmid was linearized with EcoRV. Linearized pgem-btpb was ligated to a fragment containing a kanamycin resistance cassette to generate pgem- TbtpB::Kan. This plasmid was electroporated into B. abortus 2308 where it is incapable of autonomous replication. Homologous recombination events were selected using kanamycin resistance (50 µg/ml) and carbenicillin sensitivity (50 µg/ml) in tryptic soy agar plates. PCR and sequencing analyses showed that the btpb wild type gene was replaced by the disrupted one. The mutant strain obtained was called btpb. btpabtpb double mutant was obtained after electropration of pgem-tbtpb::kan into btpa mutant (Salcedo et al., 2008). Homologous recombination events were selected as previously described for btpb single mutant. PCR and sequencing analyses confirmed that btpa and btpb wild type genes were replaced by the disrupted ones in the double mutant strain. CONSTRUCTION OF THE btpb COMPLEMENTED STRAIN A DNA fragment coding for BtpB (325 aa) was amplified by PCR using primers 5 -atggatccgtggcgaatgaaccaatccgc-3 and 5 - gcactagtctaggtgatgagggcgacgcg-3.thepcrproductwasinserted by the flanking BamHI/SpeI sites (underlined) in the corresponding sites of pbbr1 MCS-4 (Kovach et al., 1995). The integrity of the construct was confirmed by sequence analysis. The plasmid was introduced into btpb mutant by biparental mating. CONSTRUCTION OF TEM-1 AND CyaA FUSIONS DNA fragments coding for VceC, VceA, BtpA (1-275), and BtpB (1-292) were amplified by PCR, digested with XbaI and PstI and cloned into pflagtem1 (Raffatellu et al., 2005). Primers sequences with XbaI and PstI sites (underlined) are: VceC-Fw: 5 -tcctctagagaacgttcagagcgtccagaa-3 ; VceC-Rv: 5 -aaactgcagctaattgcgggtttctcccttg-3 ;VceA-Fw:5 -tcctctagaaaaat catcatcacggcagca-3 ;VceA-Rv:5 -aaactgcagctagttcttgggcgcgtggcc- 3 ; BtpA-Fw: 5 - tcctctagaagttcgtactcttctaatatt-3 ; BtpA-Rv: 5 - aaactgcagtcagataagggaatgcagttc-3 ; BtpB-Fw: 5 - tcctctagatacaa tttatttgtttcgggc-3 ;BtpB-Rv:5 - aaactgcagctaggtgatgagggcgacgcg- 3. The integrity of all constructs was confirmed by sequence analysis. Plasmids were introduced into B. abortus 2308 or virb9/virb9 by electroporation. pflagtem1 encodes a copy of TEM1 β-lactamase, in which the Sec-dependent signal sequence has been deleted and replaced with a 3 FLAG tag at the N-terminus (Raffatellu et al., 2005). Expression of the fusion proteins in Brucella was confirmed by Western blot using a mouse anti-flag M2 antibody (Sigma-Aldrich). To generate plasmids coding for fusions to the N-terminus of CyaA, BamHI/SpeI DNA fragments coding for BtpA (1-275) and BtpB (1-292) were obtained by PCR amplification with primers carrying BamHI/SpeI sites and ligated into the corresponding sites of pcyaa (Marchesini et al., 2011). Primers sequences with BamHI and SpeI sites (underlined) are: BtpA-Fw: 5 -atggatccatgagttc gtactcttctaata-3 ; BtpA-Rv: 5 -ggactagtgataagggaatgcagttcttt-3 ; BtpB-Fw 5 - atggatccatgtacaatttatttgtttcgggc-3 ; BtpB-Rv: 5 - ggactagtggtgatgagggcgacgcgctc-3. To generate plasmids coding for fusions to the C-terminus of CyaA, the genes coding for BtpA (1-275) and BtpB (1-292) with flanking XbaI and SacII sites (underlined) were amplified using primers BtpA-Fw: 5 -tatctagaatgagttc gtactcttctaatattg-3 /BtpA-Rv: 5 -tccccgcggtcagataagggaatgcagttc- 3 and BtpB-Fw: 5 -tatctagaatgtacaatttatttgtttcgggct-3 /BtpB-Rv: 5 -tccccgcggctaggtgatgagggcgacgcg-3.thednafragmentcoding for CyaA was amplified with flanking BamHI and SpeI sites (underlined) using primers 5 -cgggatccatgcagcaatcgcatcaggct-3 and 5 -cgactagtaaggctgtcatagccggaatcctggc-3. DNA fragments coding for CyaA and BtpA or BtpB were ligated in the corresponding sites of pdk51 under bcsp31 gene promoter as described in (Marchesini et al., 2011). The integrity of all constructs was confirmed by sequence analysis. Plasmids expressing CyaA fusion proteins were introduced in B. abortus strains by biparental mating. Expression of the fusion proteins in Brucella was confirmed by Western blot using a mouse serum raised against CyaA. BACTERIAL INFECTION AND REPLICATION ASSAYS BMDCs were prepared from 6-8 week-old female C57BL/6 mice (Lelouard et al., 2002). Infections were performed at a multiplicity of infection of 30:1. Bacteria were centrifuged onto BMDCs at 400 g for 10 min at 4 Candthenincubatedfor30minat37 C with 5% CO 2 atmosphere. Cells were gently washed twice with medium and then incubated for 1 h in medium supplemented with 100 µg/ml streptomycin to kill extracellular bacteria (or gentamicin for Salmonella). Thereafter, the antibiotic concentration was decreased to 20 µg/ml. Control samples were always performed by incubating cells with media only and following the exact same procedure for infection. To monitor bacterial intracellular survival, infected cells were lysed with 0.1% Triton X-100 in H 2 O and serial dilutions plated onto TSB agar to enumerated CFUs. IMMUNOFLUORESCENCE MICROSCOPY NF-kB Cells were fixed in 3% paraformaldehyde, ph 7.4, at room temperature for 20 min. Cells were then permeabilized for 10 min Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 10

Salcedo et al. Brucella TIR protein B with 0.1% saponin in PBS, followed by a blocking for 1 h with 2% BSA in PBS. Primary antibodies were incubated for 1 h followed by 3 washes in PBS, 1 h incubation for secondary antibodies, 2 washes in PBS and 1 wash in water before mounting with Prolong Gold (Life technologies). Primary antibodies used: rabbit anti-p65 from Santa Cruz at 1/250, hamster anti-cd11c from BioLegend at 1/100 and cow anti-brucella LPS antibody at 1/2000. Secondary antibodies used: goat antihamster Alexa 594, donkey anti-rabbit Cy3, goat anti-cow FITC, all from Jackson Immunoresearch. Nuclei were stained with TOPRO-3. Samples were examined on a Leica SP5 laser scanning confocal microscope for image acquisition. Images of 1024 1024 pixels were then assembled using Adobe Photoshop 7.0. In all experiments we used an anti-cd11c antibody confirming analysis of DCs only. Quantification was always done by counting at least 100 cells in 4 independent experiments, for a total of at least 400 host cells analysed. FLOW CYTOMETRY OF INFECTED CELLS BMDCs were harvested 8 h or 24 h after infection and stained for 20 min at 4 C with anti CD11c APC-Cy7, anti CD40 Alexa 647, anti CD80 Pe-Cy5, anti CD86 FITC and anti MHC class II PE (all purchased at BioLegend). Cells were then washed once in 1% FCS in PBS and once in PBS. Cells were fixed for 20 min in 3% PFA at room temperature. At least 100,000 CD11c+ events were collected on flow cytometry using a FACS Canto II (Becton Dickinson) and analysis was done on FlowJo software (TreeStar). TEM TRANSLOCATION ASSAY RAW cells were seeded in a 96 well plates at 1 10 4 cells/well overnight. Cells were then infected with an MOI of 500:1 by centrifugation at 4 C, 400 g for 5 min and 20 min at 37 C5%CO 2. Cells were then washed twice with DMEM and 200 µl ofcomplete media, with gentamicin (50 µg/ml) and 1 mm of IPTG was added for 1 h. Media was replaced by 200 µl of complete DMEM, with gentamicin (10µg/ml) and 1 mm of IPTG. At 4 or 24 h after infection cells were washed with 100 µl HBSS. 20 µl of CFF2 mix (as described by Life Technologies protocol) was then added to each well, and plate incubated for 1.5 h at room temperature in the dark. Cells were finally washed with 100µl PBSandanalysed immediately by microscopy. A total of 5000 cells were counted from 5 independent experiments in an automated manner using imagej. CyaA ASSAYS Translocation of BtpA and BtpB into host cells was assayed using the CyaA fusion approach. After infection of J774.A1 cells (MOI 250:1) for 4 h in 96-wells plates (10 5 cells/well), cells were gently washed five times with PBS and lysed. Intracellular camp levels were determined by Direct camp Enzyme Immunoassay Kit (Sigma, CA200) as described by the manufacturer. CYTOKINE MEASUREMENT Sandwich enzyme-linked immunosorbent assays (ELISA) from ebioscience were used to detect IL-12 (p40/p70) and TNFα from supernatants of BMDCs infected with different Brucella strains. LUCIFERASE ACTIVITY ASSAY HEK 293 T cells were transiently transfected using Fugene (Roche) for 24 h, according to manufacturer s instructions, for a total of 0.4 µg of DNA consisting of 50 ng TLR plasmids, 200 ng of pbiixluc reporter plasmid, 5 ng of control Renilla luciferase (prl-null, Promega) and 50 ng of myc-btpa or myc- BtpB expression vectors. The total amount of DNA was kept constant by adding empty vector. Where indicated, cells were treated with E. coli LPS (1 µg/ml), Pam 2 CSK4 (100 ng/ml), CpG ODN1826 (1 µm), Flagellin Fl-ST (1 µg/ml) and poly(i:c) (25 µg/ml), all obtained from Invivogen, for 6 h and then cells were lysed and luciferase activity measured using Dual-Glo Luciferase Assay System (Promega). The BtpB constructs were obtained by cloning in the gateway (Life Technologies) entry vector and then cloned in pmyc. The following primers were used for BtpB (178 aa) ggggacaagtttgtacaaaaaagcaggcttcatgaatcgtacgca ctgggcg and as reverse primer ggggaccactttgtacaagaaagctgggtcc taggtgatgagggcgacgcg. For BtpB (1-292) the forward primer was ggggacaagtttgtacaaaaaagcaggcttctacaatttatttgtttcgggc. The BtpA (1-275) primers were: ggggacaagtttgtacaaaaaagcaggcttcatgagttcgta ctcttctaatatt and the reverse primer was ggggaccactttgtacaagaaa gctgggtctcagataagggaatgcagttc. YEAST TWO-HYBRID ASSAY The plasmids used for the Y2H interaction test were obtained by using the Gateway technique, except the pact2 vector encoding human MyD88-Gal4 activation domain (AD) fusion that was provided by L. O Neill. Briefly, human MyD88 and B. melitensis 16 M BtpA (BMEI1674) were amplified by PCR respectively from the pact2-myd88 vector and from genomic DNA with Gateway primers (GWMyD88F and GWMyD88R; GWbtpAF and GWbtpAR). PCR products were then separately cloned into the entry vector pdonr201 (Invitrogen Life-technologies) as previously described (Dricot et al., 2004). For btpb, the corresponding entry vector pdonr201-bmei1216 from the ORFeome was used (Dricot et al., 2004). LR reactions were then performed as recommended by the manufacturer (Invitrogen Life-technologies) in order to clone MyD88 into pvv212 (Van Mullem et al., 2003) downstream of the Gal4 DNA-binding domain (BD), and BtpA and BtpB into pvv213 (Van Mullem et al., 2003) downstream of the Gal4 AD. Haploïd Saccharomyces cerevisiae strains Mav103 and Mav203 (Walhout and Vidal, 2001) were transformed with BD and AD fusion protein vectors respectively. Diploid yeasts carrying both plasmids were obtained by mating and selected on synthetic dextrose medium (SD) lacking leucine (leu) and tryptophan (trp) as previously described (Hallez et al., 2007). Protein interactions were assessed on medium lacking histidine (his) supplemented with 20 mm triaminotriazole (3AT). The β-galactosidase expression filter assay using the LacZ reporter gene was performed as described previously (Dozot et al., 2010). The primers used for two-hybrid constructs were: GWMyD88Fggggacaagtttgtacaaaaaagcaggctcgcgatggctgcagg aggtcccg ;GWMyD88Rggggaccactttgtacaagaaagctgggtaagggcaggga caaggccttg ;GWbtpAFggggacaagtttgtacaaaaaagcaggctcgatgagttcgta Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 11

Salcedo et al. Brucella TIR protein B ctcttctaat ;GWbtpARggggaccactttgtacaagaaagctgggtagataagggaatg cagttc. MOUSE INFECTION STUDIES Groups of 7- to 9-week-old female IRF 1 / or BALB/c mice were intraperitoneally inoculated with 10 6 CFU of B. abortus strains in 0.2 ml PBS. The infected mice were housed in cages within a biosafety level 3 facility and IRF 1 / mice were monitored daily for survival. At the indicated times post-infection, spleens from infected mice were removed and homogenized in 2 ml of PBS. Tissue homogenates were serially diluted and plated in duplicate on TSA with the appropriate antibiotic. CFU were counted after 3 4 days of incubation at 37 C. HISTOLOGICAL AND IMMUNOHISTOLOGICAL ANALYSIS For each mouse, the spleen was removed, fixed with buffered formalin 4%, and embedded in paraffin. Serial sections (3 µm) of these specimens were obtained for routine hematoxylin-eosin and immunohistochemical investigations to assess the presence of granulomas and bacterial antigens, respectively. Granulomas were defined as collections of ten or more macrophages within the organs. The inflammatory granulomas present in each tissue section of the spleens were counted during microscopic examination, and the total area of tissue sections was determined by quantitative image analysis as described previously (Stein et al., 2005). The results were expressed as the number of granulomas found per surface unit (i.e., square centimeters). Counts of granulomas were expressed as the mean ± the standard deviation per square centimeter and compared by using the Student t test. Immunohistochemical analysis was performed with a rabbit anti-b. abortus antibody used at a 1:1000 dilution with hemalun counterstain. The immunohistological procedure, in which an immunoperoxidase kit was used, has been described elsewhere (Leone et al., 2004). For each section, a negative control was performed with normal rabbit serum. STATISTICAL ANALYSIS Unpaired two-tailed Student s t test was carried out to determine the statistical differences between experimental data sets. P 0.05 were not considered significant; P 0.001 are denoted with ; P 0.01 are denoted with and P between 0.01 and 0.05 are denoted with. Statistical differences between IRF-1 / mice survival curves were determined with Mantel-Cox test. ACKNOWLEDGMENTS We are grateful to R. Tsolis for the pflag-tem1, L. O Neill for the pact2 vector encoding human MyD88-Gal4 activation domain (AD) fusion and R. Jerala for the MD2 plasmid. This work was supported by the Agence National Recherche (ANR BruTir), the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale and the Aix-Marseille Université. Matthieu Terwagne held a Ph.D. fellowship from the Fonds National de la Recherche Scientifique and his work was supported by an ARC convention from the French-Speaking Community of Belgium (N 08/13 015). Paul Roger Claude Imbert held a Ph.D. fellowship from FINOVI and TLSL was funded by the ANR grant CELLPATH, awarded under the ERA-NET PathoGenoMics scheme. Diego José Comerci and María Ines Marchesini are members of the Scientific Research Career from Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET, Argentina); Claudia Karina Herrmann is a Ph.D. fellow from CONICET and this work was supported by ANPCyT PICT 2011-0253 and 2011-1485. Kristine Von Bargen was funded by the ANR BruTir and Clara Degos held a fellowship from the Aix-Marseille Université. REFERENCES Celli, J., de Chastellier, C., Franchini, D.-M., Pizarro-Cerda, J., Moreno, E., and Gorvel, J.-P. (2003). Brucella evades macrophage killing via VirB-dependent sustained interactions with the endoplasmic reticulum. J. Exp. Med. 198, 545 556. doi: 10.1084/jem.2003 0088 Celli, J., Salcedo, S. P., and Gorvel, J.-P. (2005). Brucella coopts the small GTPase Sar1 for intracellular replication. Proc. Natl. Acad. Sci. U.S.A. 102, 1673 1678. doi: 10.1073/pnas.0406873102 Chaudhary, A., Ganguly, K., Cabantous, S., Waldo, G. S., Micheva-Viteva, S. N., Nag, K., et al. (2011). The Brucella TIR-like protein TcpB interacts with the death domain of MyD88. Biochem. Biophys. Res. Commun. 417, 1 6. Cirl, C., Wieser, A., Yadav, M., Duerr, S., Schubert, S., Fischer, H., et al. (2008). Subversion of Toll-like receptor signaling by a unique family of bacterial Toll/interleukin- 1 receptor domain-containing proteins. Nat. Med. 14, 399 406. doi: 10.1038/nm1734 de Barsy, M., Jamet, A., Filopon, D., Nicolas, C., Laloux, G., Rual, J.- F., et al. (2011). Identification of a Brucella spp. secreted effector specifically interacting with human small GTPase Rab2. Cell Microbiol. 13, 1044 1058. doi: 10.1111/j.1462-5822.2011.01601.x de Jong, M. F., Starr, T., Winter, M. G., den Hartigh, A. B., Child, R., Knodler, L. A., et al. (2012). Sensing of bacterial type IV secretion via the unfolded protein response. mbio 4:e00418-12. doi: 10.1128/mBio.e00418 12 de Jong, M. F., Sun, Y.-H., den Hartigh, A. B., van Dijl, J. M., and Tsolis, R. M. (2008). Identification of VceA and VceC, two members of the VjbR regulon that are translocated into macrophages by the Brucella type IV secretion system. Mol. Microbiol. 70, 1378 1396. doi: 10.1111/j.1365-2958.2008.06487.x Dozot, M., Poncet, S., Nicolas, C., Copin, R., Bouraoui, H., Mazé, A., et al. (2010). Functional characterization of the incomplete phosphotransferase system (PTS) of the intracellular pathogen Brucella melitensis. PLoS ONE 5:e12679. doi: 10.1371/journal.pone.0012679 Dricot, A., Rual, J.-F., Lamesch, P., Bertin,N.,Dupuy,D.,Hao,T.,etal. (2004). Generation of the Brucella melitensis ORFeome version 1.1. Genome Res. 14, 2201 2206. doi: 10.1101/gr.2456204 Durward, M., Radhakrishnan, G., Harms, J., Bareiss, C., Magnani, D., and Splitter, G. A. (2012). Active evasion of CTL mediated killing and low quality responding CD8+ T cells contribute to persistence of brucellosis. PLoS ONE 7:e34925. doi: 10.1371/journal.pone.0034925 Hallez, R., Letesson, J.-J., Vandenhaute, J., and De Bolle, X. (2007). Gatewaybased destination vectors for functional analyses of bacterial ORFeomes: application to the Min system in Brucella abortus. Appl. Environ. Microbiol. 73, 1375 1379. doi: 10.1128/AEM.01873-06 Ko, J., Gendron-Fitzpatrick, A., Ficht, T. A., and Splitter, G. A. (2002). Virulence criteria for Brucella abortus strains as determined by interferon regulatory factor 1-deficient mice. Infect. Immun. 70, 7004 7012. doi: 10.1128/IAI.70.12.7004-7012.2002 Kolaskar, A. S., and Reddy, B. V. (1985). A method to locate protein coding sequences in DNA of prokaryotic systems. Nucleic Acids Res. 13, 185 194. doi: 10.1093/nar/13. 1.185 Kovach, M. E., Elzer, P. H., Hill, D. S., Robertson, G. T., Farris, M. A., Roop, R. M. 2nd., et al. (1995). Four new derivatives of the broad-host-range cloning vector pbbr1mcs, carrying different antibiotic-resistance cassettes. Gene 166, 175 176. Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 12

Salcedo et al. Brucella TIR protein B Lelouard, H., Gatti, E., Cappello, F., Gresser, O., Camosseto, V., and Pierre, P. (2002). Transient aggregation of ubiquitinated proteins during dendritic cell maturation. Nature 417, 177 182. doi: 10.1038/417177a Leone, M., Honstettre, A., Lepidi, H., Capo, C., Bayard, F., Raoult, D., et al. (2004). Effect of sex on Coxiella burnetii infection: protective role of 17beta-estradiol. J. Infect Dis. 189, 339 345. doi: 10.1086/380798 Marchesini, M. I., Herrmann, C. K., Salcedo, S. P., Gorvel, J.-P., and Comerci, D. J. (2011). In search of Brucella abortus type IV secretion substrates: screening and identification of four proteins translocated into host cells through VirB system. Cell Microbiol. 13, 1261 1274. doi: 10.1111/j.1462-5822.2011.01618.x Newman,R.M.,Salunkhe,P.,Godzik, A., and Reed, J. C. (2006). Identification and characterization of a novel bacterial virulence factor that shares homology with mammalian Toll/interleukin- 1 receptor family proteins. Infect Immun. 74, 594 601. doi: 10.1128/IAI.74.1.594-601.2006 Nothelfer, K., Dias Rodrigues, C., Bobard, A., Phalipon, A., and Enninga, J. (2011). Monitoring Shigella flexnerivacuolar escape by flow cytometry. Virulence 2, 54 57. doi: 10.4161/viru.2.1.14666 Parsons, J. A., Bannam, T. L., Devenish, R. J., and Rood, J. I. (2007). TcpA, an FtsK/SpoIIIE homolog, is essential for transfer of the conjugative plasmid pcw3 in Clostridium perfringens. J. Bacteriol. 189, 7782 7790. doi: 10.1128/JB.00783-07 Pizarro-Cerdá, J., Méresse, S., Parton, R. G., van der Goot, G., Sola-Landa, A., Lopez-Goñi, I., et al. (1998). Brucella abortus transits through the autophagic pathway and replicates in the endoplasmic reticulum of nonprofessional phagocytes. Infect Immun. 66, 5711 5724. Radhakrishnan, G. K., Harms, J. S., and Splitter, G. A. (2011). Modulation of microtubule dynamics by a TIR domain protein from the intracellular pathogen Brucella melitensis. Biochem. J. 439, 79 83. doi: 10.1042/BJ20110577 Radhakrishnan, G. K., Yu, Q., Harms, J. S., and Splitter, G. A. (2009). Brucella TIR Domain-containing protein mimics properties of the toll-like receptor adaptor protein TIRAP. J. Biol. Chem. 284, 9892 9898. doi: 10.1074/jbc.M805458200 Raffatellu, M., Sun, Y.-H., Wilson, R. P., Tran, Q. T., Chessa, D., Andrews- Polymenis, H. L., et al. (2005). Host restriction of Salmonella enterica serotype Typhi is not caused by functional alteration of SipA, SopB, or SopD. Infect Immun. 73, 7817 7826. doi: 10.1128/IAI.73.12.7817-7826. 2005 Rana,R.R.,Zhang,M.,Spear,A.M., Atkins, H. S., and Byrne, B. (2013). Bacterial TIR-containing proteins and host innate immune system evasion. Med. Microbiol. Immunol. 202, 1 10. doi: 10.1007/s00430-012- 0253-2 Salcedo, S. P., Marchesini, M. I., Lelouard, H., Fugier, E., Jolly, G., Balor, S., et al. (2008). Brucella control of dendritic cell maturation is dependent on the TIR-containing protein Btp1. PLoS Pathog. 4:e21. doi: 10.1371/journal.ppat. 0040021 Sengupta,D.,Koblansky,A.,Gaines,J., Brown, T., West, A. P., Zhang, D., et al. (2010). Subversion of innate immune responses by Brucella through the targeted degradation of the TLR signaling adapter, MAL. J. Immunol. 184, 956 964. doi: 10.4049/jimmunol.0902008 Sieira, R., Comerci, D. J., Sánchez, D. O., and Ugalde, R. A. (2000). A homologue of an operon required for DNA transfer in Agrobacterium is required in Brucella abortus for virulence and intracellular multiplication. J. Bacteriol. 182, 4849 4855. doi: 10.1128/JB.182.17.4849-4855.2000 Spear, A. M., Loman, N. J., Atkins, H. S., and Pallen, M. J. (2009). Microbial TIR domains: not necessarily agents of subversion? Trends Microbiol. 17, 393 398. doi: 10.1016/j.tim.2009.06.005 Spear, A. M., Rana, R. R., Jenner, D. C.,Flick-Smith,H.C.,Oyston,P. C. F., Simpson, P., et al. (2012). A Toll/interleukin (IL)-1 receptor domain protein from Yersinia pestis interacts with mammalian IL-1/Toll-like receptor pathways but does not play a central role in the virulence of Y. pestis in a mouse model of bubonic plague. Microbiology 158, 1593 1606. doi: 10.1099/mic.0.055012-0 Stein, A., Louveau, C., Lepidi, H., Ricci, F., Baylac, P., Davoust, B., et al. (2005). Q fever pneumonia: virulence of Coxiella burnetii pathovars in a murine model of aerosol infection. Infect Immun. 73, 2469 2477. doi: 10.1128/IAI.73.4.2469-2477.2005 Van Mullem, V., Wery, M., De Bolle, X., and Vandenhaute, J. (2003). Construction of a set of Saccharomyces cerevisiae vectors designed for recombinational cloning. Yeast 20, 739 746. doi: 10.1002/yea.999 Walhout, A. J., and Vidal, M. (2001). High-throughput yeast twohybrid assays for large-scale protein interaction mapping. Methods 24, 297 306. doi: 10.1006/meth.2001.1190 Yadav, M., Zhang, J., Fischer, H., Huang, W., Lutay, N., Cirl, C., et al. (2010). Inhibition of TIR domain signaling by TcpC: MyD88- dependent and independent effects on Escherichia coli virulence. PLoS Pathog. 6:e1001120. doi: 10.1371/journal.ppat.1001120 Zhang, Q., Zmasek, C. M., Cai, X., and Godzik, A. (2011). TIR domain-containing adaptor SARM is a late addition to the ongoing microbe-host dialog. Developmental and Comparative Immunology 35, 461 468. doi: 10.1016/j.dci.2010.11.013 Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Received: 08 March 2013; accepted: 18 June 2013; published online: 08 July 2013. Citation: Salcedo SP, Marchesini MI, Degos C, Terwagne M, Von Bargen K, Lepidi H, Herrmann CK, Santos Lacerda TL, Imbert PRC, Pierre P, Alexopoulou L, Letesson J-J, Comerci DJ and Gorvel J-P (2013) BtpB, a novel Brucella TIR-containing effector protein with immune modulatory functions. Front. Cell. Infect. Microbiol. 3:28. doi: 10.3389/fcimb.2013.00028 Copyright 2013 Salcedo, Marchesini, Degos, Terwagne, Von Bargen, Lepidi, Herrmann, Santos Lacerda, Imbert, Pierre, Alexopoulou, Letesson, Comerci and Gorvel. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc. Frontiers in Cellular and Infection Microbiology www.frontiersin.org July 2013 Volume 3 Article 28 13

III. Discussion et conclusion générale 107

La survie de Brucella dépend de sa capacité à déployer des facteurs de virulence (virb, CβG, LPS) tant pour assurer sa vie intracellulaire que pour limiter l activation du système immunitaire. Pendant ma thèse, je me suis attachée à essayer de comprendre la relation entre les cellules immunitaires, particulièrement les DC, et Brucella. Le CβG, qui a été montré comme étant un puissant activateur des DC, est aussi capable d induire des signaux anti-inflammatoires comme SOCS2, LILRA ou tsg-6 dans les DC humaines. De plus, nous avons montré qu une injection de CβG dans le derme de l oreille provoque un œdème moins important que dans le derme de souris injectées avec du LPS d E. coli. Cet œdème s accompagne d un recrutement de neutrophiles, qui est transitoire avec un pic à 12 h et diminue ensuite jusqu à 48h, alors que dans le cas d une injection avec du LPS, le recrutement persiste jusqu à 24 h puis commence à diminuer. La différence de cinétique entre les deux types d injection (LPS versus CβG) est intéressante. Les neutrophiles jouent un rôle dans les premières réponses à une infection grâce à leur fonction de dégranulation notamment [224]. Le recrutement plus transitoire de ces cellules indique une inflammation qui sera plus rapidement résolue, des dommages tissulaires moins importants. Une étude récente d immunisation de souris a démontré que le recrutement de neutrophiles depuis le ganglion conduit à une compétition avec les DC et macrophages pour présenter des peptides antigéniques aux LT, et donc le recrutement de neutrophiles aboutit, dans ce cas, à une inhibition de la réponse des LT [225]. Une autre étude d infection à Mycobaterium indique un rôle inverse des neutrophiles : ils aideraient à l activation des DC et des LT CD4 +, cruciaux pour induire une réponse immunitaire anti-mycobactéries [226]. Il est intéressant de constater qu une molécule de Brucella comme le CβG impacte le recrutement des neutrophiles dans le tissu. De plus, différentes études ont montré que les neutrophiles avaient des rôles différents durant l infection [82, 113]. Il serait intéressant de voir en quoi la modulation du recrutement de neutrophiles au cours de la brucellose impacte la présentation antigénique par les DC et macrophages. La résolution rapide de l inflammation en cas d injection de CβG pourrait indiquer la présence plus importante de molécules anti-inflammatoires. Les protéines Btp sont aussi un formidable exemple de la capacité de la bactérie à s adapter à son hôte. En effet, les TLR et les DC permettent à la fois de détecter Brucella, et d activer le système immunitaire. Or, lors de sa vie intracellulaire, Brucella est capable de sécréter BtpA et BtpB. Ces protéines vont alors interagir avec les molécules adaptatrices TIRAP et MyD88 108

en aval des TLR pour inhiber les voies de signalisation [132-136]. Ici, nous avons mis en évidence le rôle de BtpB dans le contrôle de l activation des DC, et son potentiel rôle dans la virulence in vivo. Cependant de nombreux points sont à éclaircir quant à son rôle dans l inhibition de la signalisation en aval des TLR. Nous ne pouvons exclure que BtpB soit impliquée dans la régulation d autres fonctions cellulaires au cours de l infection. En effet, récemment, une étude a montré l implication de BtpA dans la réponse UPR (unfolded protein response, réponse au stress cellulaire) [227]. Au vu des similitudes entre les voies de signalisation inhibées par BtpA et BtpB, il serait intéressant de vérifier l implication de BtpB dans l induction de l UPR ; et si comme c est le cas dans les macrophages, la réponse UPR est nécessaire à la réplication de Brucella dans les DC [227]. BtpA semble aussi être impliquée dans la restructuration du RE, indépendamment de la réponse UPR [227]. Deux autres études ont souligné l association de BtpA avec les microtubules, et son rôle de stabilisateur [134, 228]. Les auteurs expliquent que via la modulation des microtubules et des phosphoinositides BtpA peut impacter les signaux en aval de TLR (en plus de sa capacité à lier MyD88 et à provoquer la dégradation de TIRAP) [229]. Aucune étude semblable n a été menée avec BtpB, et il serait intéressant de voir si c est grâce à un mécanisme semblable que BtpB inhibe les signaux TLR. Dans le cas d autres infections, les bactéries expriment des molécules leur permettant de moduler le cytosquelette d actine ou les microtubules pour envahir dans les cellules, ou encore modifier le trafic intracellulaire des vacuoles et ainsi éviter une fusion avec les lysosomes [230, 231]. Considérant le rôle de BtpA dans la modulation et stabilisation des microtubules et du RE, il serait intéressant d étudier le possible rôle de cette protéine, ainsi que de BtpB dans ce processus et dans le maintien de la stabilité de la BCV. Une autre question intéressante est celle du rôle joué par les DC au cours de l infection. Ces cellules peuvent clairement être infectées par Brucella, qui y survit en établissant sa niche réplicative dans le RE [23, 232]. Brucella utilise des mécanismes pour limiter leur détection via les TLR (LPS, Btp), pour diminuer l activation en elle-même des DC (Btp, Omp25). On peut donc se poser la question de l importance des DC dans les réponses contre la bactérie au vu des mécanismes déployés par Brucella pour limiter leur activation. 109

S p le e n L iv e r * * 8 * 7 L o g C F U /o rg a n 7 6 5 4 L o g C F U /o rg a n 6 5 4 3 3 C 5 7 B L /6 C C R 2 K O B. a b o rtu s w t B. a b o rtu s o m p 2 5 2 C 5 7 B L /6 C C R 2 K O Figure 34 : Réplication de Brucella dans la rate de souris sauvages ou CCR2 KO infectées durant 5 j. Des souris sauvages (C57BL/6) ou CCR2 KO ont été infectées en IP avec 1.10 6 CFU/souris. 5 jours p.i les organes sont prélevés et la réplication bactérienne est analysée par dénombrement de CFU. Les souris infectées par la souche sauvage sont indiquées par un cercle bleu, et celles infectées par le mutant Δomp25 par des carrés rouge. Chaque symbole représente un animal et la médiane des résultats est indiquée par une ligne horizontale. P < 0,05 : * et P < 0,01 : **.

Beaucoup d études sur les DC, dont celles menées pendant ma thèse, proviennent de données in vitro [23, 85, 100, 232, 233]. L importance des DC in vivo reste à démontrer. Dans le cas d une infection nasale, l immunité pulmonaire dépend des macrophages et non pas des DC [202]. De plus, la localisation des DC et macrophages au sein de la rate ainsi que leur activation ne sont pas impactées par l infection. En revanche, en absence de macrophages, les DC inflammatoires des poumons sont recrutées, et migrent dans le ganglion médiastinal, ce qui pourrait permettre une dissémination de la bactérie [202]. Une autre étude d infection in vivo révèle que les DC spléniques (B220 - CD11b + LY-6C + NK1.1 - inos + ) sont requises pour l induction d une réponse immunitaire de type Th1 contre la bactérie. Cette étude renforce l hypothèse de l importance des DC in vivo, et montre qu il y a une différence de comportement selon l organe infecté et le type d infection (aérosol/nasal ou en IP) [92, 202]. Au cours d une infection in vivo, les DC spléniques s activent et migrent dans la pulpe blanche de la rate où se concentrent les LT [116]. Les DC inflammatoires qui sont recrutées au cours de l infection participent à la formation des granulomes au sein de la rate et du foie [116]. Ces études montrent donc que les DC ont différents rôles et capacité à s activer selon qu elles soient résidentes dans le tissu (DC spléniques ou pulmonaires) ou recrutées (DC inflammatoires) au cours de l infection. Pour étudier le rôle des DC dans l infection par Brucella, j ai utilisé un modèle de souris KO pour CCR2. CCR2 est un récepteur à la chimiokine CCL2. En plus de ses propriétés chimioattractives, il permet la sortie des DC dérivées de monocytes, ainsi que des monocytes, de la moelle osseuse vers le sang. Dans de nombreux modèles d infection, les monocytes CCR2 + ont été montrés comme étant critiques dans l établissement d une réponse immunitaire [234]. Ces souris ont un défaut de recrutement des DC inflammatoires et monocytes dans les tissus enflammés [235]. 5 j après infection (en IP) de ces souris, (Fig. 34). Ces résultats, préliminaires, indiqueraient que le contrôle de la réplication de Brucella in vivo dépend des monocytes ainsi que des DC inflammatoires. Le problème de ce genre de modèle est que la délétion de CCR2 ne va pas impacter uniquement ces cellules. Il y a par exemple un manque de production d IFN-γ dans les splenocytes de ces souris activés d une manière Th1 dépendante [236]. L utilisation d autres modèles (comme les modèles de délétion de CD11c induit par la toxine diphtérique par exemple [237]), pourrait confirmer l importance de ces cellules in vivo. 110

D après l étude menée sur CD150 ici, nous pouvons affirmer que la liaison de Omp25 à CD150 fait partie des mécanismes mis en place par Brucella pour inhiber l activation des DC et ainsi contrôler une partie de la réponse immunitaire. Il est probable que ce type de mécanisme soit retrouvé dans différents types cellulaires au sein du système immunitaire pour limiter l inflammation. D autres récepteurs inhibiteurs peuvent aussi jouer un rôle. CTLA-4, PD-1 sont des récepteurs importants pour l activation des LT. Parfois, dans les cancers par exemple, ces récepteurs sont exprimés fortement et induisent une situation anti-inflammatoire [238, 239]. Pourtant aucune étude à notre connaissance n a été menée pour voir si ces protéines jouaient un rôle dans la réponse immunitaire contre Brucella. Du fait de l importance des réponses des LT CD4 + IFN-γ pour éliminer la bactérie, et des mécanismes déployés par Brucella pour diminuer la réponse immunitaire, il serait intéressant de d étudier ce type de récepteur [153]. Il est intéressant de constater que Brucella joue parfois un double rôle en produisant également des facteurs qui activent le système immunitaire. En effet, PrpA est capable d activer la prolifération des LB et leur production d anticorps anti-brucella (via l activation de macrophages) [83, 162, 163]. Mais, ce facteur permet aussi d inhiber la réponse immunitaire. En effet, la production d anticorps déclenchée par l activation des LB permet l opsonisation de Brucella dans les macrophages et donc un taux d infection plus important [83]. La production de cytokines in vivo dans les souris infectées est aussi inhibée par PrpA [83]. De plus, cette protéine de Brucella est requise pour la persistance bactérienne [162]. Cet exemple permet de constater que Brucella est à la fois capable d activer et inhiber le système immunitaire pour permettre sa survie. Cela donne un autre regard sur le fait que CD150 puisse aussi jouer un double rôle dans l activation ou l inhibition du système immunitaire lors de l infection par Brucella. Pour finir, les DC et macrophages sont activés par des composants de Brucella (CβG, PrpA, Omp16, etc ). Considérant que les DC participeraient à la dissémination de la bactérie au sein de l organisme (en absence de macrophages) [130], on peut imaginer que cette balance entre des composés pro-inflammatoires (CβG sur les DC) et anti-inflammatoires (Btp, Omp25) a un rôle. En effet, pour migrer dans les organes lymphoïdes, les DC ont besoin d être activées, seulement trop d activation est délétère pour Brucella. D où cette balance entre inflammation et anti-inflammation pour permettre à la fois aux DC de migrer et 111

disséminer la bactérie, mais limiter en même temps l activation pour établir une niche réplicative et éviter une réponse immunitaire trop forte. Les études menées durant ma thèse permettent d avoir une meilleure compréhension de l activation des DC au cours de l infection par Brucella. Cela nous a notamment permis d identifier un nouveau mécanisme d inhibition de l activation des DC par Brucella via la liaison de Omp25 à CD150. 112

IV. Matériel et Méthodes 113

IV. A. MATERIEL VIVANT Souris Des souris C57BL/6J ou Balb/c provenant de chez Charles River âgées de 6 à 12 semaines ont été utilisées. Les souris CD150 KO sur fond C57BL/6 ont été générées dans le laboratoire de l Université de Kyushu en remplaçant l exon 2 du gène CD150 par une cassette neo et sont décrites dans l article suivant [220]. Les souris IFN-γ sur fond C57BL/6 et proviennent du Jackson Laboratory. Les souris OTII C57BL/6 proviennent de Charles River. Cellules BMDC : Les cellules dendritiques dérivées de moelle osseuse (BMDC) ont été produites à partir de souris C57BL/6. Après sacrifice, les fémurs et tibias des souris sont prélevés, puis, lavés dans de l éthanol à 70 %, et ensuite disposés dans du RPMI (Gibco) contenant 5 % de sérum de veau fœtal (FCS - Eurobio) et 50 µm de 2-mercaptoéthanol (Sigma Aldrich). Les os sont coupés à leurs extrémités, avec une seringue contenant du milieu les os sont vidés de leur moelle. Après homogénéisation et filtration sur un filtre 70µm, Une première centrifugation est effectuée à 300g, 4 C durant 5 minutes. Le culot est resuspendu dans du RBC lysis Buffer (ebiosciences) pour lyser les globules rouges durant 1 minute avant de compléter avec du milieu et de centrifuger une nouvelle fois. Les cellules sont resuspendues dans du milieu puis filtrées sur un filtre 70µm. Après une centrifugation, les cellules sons resuspendues dans du milieu contenant du GM-CSF à 0. 6x 10 6 cellules / ml. Les BMDC sont cultivées 5 jours avant utilisation à 37 C, 5% CO 2. Le milieu est changé tous les 2 j. JL558-GMCSF : Des hybridomes contenant sont cultivés pour produire du GM-CSF. Les cellules une fois décongelées sont cultivées dans de l IMDM (Gibco) contenant 20 % de FCS durant 7 j. Le FCS est ensuite diminué à 5 % et l ajout de G418 (Gibco) permet la sélection du des cellules exprimant le GM-CSF. Les cellules sont ensuite cultivées jusqu à confluence et épuisement du milieu (environ 7 j). Le surnageant est filtré puis titré. BMDM : Les macrophages dérivés de moelle osseuse (BMDM) sont produits de la même façon que les cellules dendritiques exceptée concernant le milieu utilisé : DMEM (Gibco) supplémenté avec 10 % de FCS, 1 % de L-Glutamine (Gibco), 10 % de L-CSF. Les cellules sont diluées à 2 x 10 5 cellules / ml et cultivées 7 j avant utilisation et le milieu est changé le 5 ème et 6 ème jour de culture. 114

Cos-7 : Les cellules Cos-7 sont cultivées en DMEM supplémenté avec 10 % de FCS, 1 % de L-Glutamine. Les cellules sont trypsinisées puis diluées tous les 2 jours. Lymphocytes T : Les LT purifiés sont maintenus en culture en RPMI, 10 % FCS, 1 % L- glutamine, 1 % HEPES, 1 % Sodium Pyruvate. Souches bactériennes Pour ces études nous avons utilisées différentes souches bactériennes, regroupées dans le tableau X : Toutes les expériences avec Brucella ont été réalisées dans notre laboratoire de niveau de sécurité biologie 3 (BSL3). Les souches de Brucella sont isolées à partir de stock en glycérol et cultivées sur du Tryptic Soy Agar (TSA - Sigma Aldrich) durant 5 j avant utilisation. Pour les infections 4 à 6 colonies sont mises en culture en Tryptic Soy Broth (TSB Sigma Aldrich) 16 h à 37 C et 200 rpm jusqu à ce qu elles atteignent environ 2.0 de densité optique (DO) à 600nm. Salmonella et E. coli sont cultivés en LB-Agar et les sous-cultures sont faites en LB durant 16h à 37 C avec aération et sous 200 rpm agitation. Extraction des lymphocytes T (LT) des organes de souris Les rates et ganglions des souris OTI et OTII ont été extraits stérilement et placés dans des tubes contenant du RPMI. Les organes sont écrasés à l aide d un piston d une seringue sur un tamis cellulaire 70µm (BD Biosciences) puis centrifugés à 350 g pendant 8 minutes à 4 C. Les globules rouges sont ensuite lysés avec le RBC lysis Buffer durant 2 min à température ambiante. Les cellules sont ensuite lavées une fois avec du RPMI. La concentration cellulaire est ajustée à 1.108 cellules / ml et LT CD4 ou CD8 sont purifiés avec les kits Dynabeads Untouched Mouse CD4 Cells et Dynabeads Untouched mouse CD8 Cells (Invitrogen, Life Technologies), selon les instructions du fabricant. IV. B. REACTIFS Tableau 1 : Réactifs Utilisés Produit Utilisation Concentration finale Provenance LPS E.coli 055:B5 Activation de BMDC 100ng/ml Sigma Aldrich 115

Extrait de membranes de Brucella abortus sauvage - ba183 Activation de BMDC 10µg/ml I. Moriyón Extrait de membranes de Brucella abortus omp25 - ba135 Activation de BMDC 10µg/ml I. Moriyón Ovalbumine Activation LT (co-culture) spécifique des OTI et OTII 50µg/ml Hyglos OVA 323-339 Activation LT (co-culture) Peptide spécifique des LT CD4 OTII 0,12µg/ml Tebu Bio OVA 257-264 Activation LT (co-culture) Peptide spécifique des LT CD8 OTI 1µg/ml Invivogen BrdU Proliferation BMDC 10µM BD Biosciences Cell Trace Violet Proliferation LT 1µM Life Technologies CFSE Proliferation LT 2,7µM Life Technologies Peptide bloquant CD150 Peptide contrôle Blocage de CD150 100µg/ml Thermo Science Blocage de CD150 : contrôle négatif 100 µg/ml Thermo Science Tableau 2 : Anticorps Antigène Couplage Fournisseur Dilutio n Réactivité Hôte et Isotype Clône Cytométrie de Flux CD3 e-fluor450 ebiosciences 1/200 Souris Rat IgG2b, κ 17A2 CD3 APC-Cy7 BioLegend 1/100 Souris Rat IgG2b, κ 17A2 CD4 Alexa 647 BioLegend 1/200 Souris Rat IgG2b, κ GK1.5 CD8 PE-Cy5 BioLegend 1/400 Souris Rat IgG2a, κ 53-6.7 CD11c APC BioLegend 1/200 Souris CD11c APC-Cy7 BioLegend 1/200 Souris Hamster Arménien IgG Hamster Arménien IgG N418 N418 CD25 FITC Biolegend 1/800 Souris Rat IgG1, λ PC61 CD40 PE-Cy5 BioLegend 1/400 Souris CD40 Alexa 647 BioLegend 1/200 Souris Hamster Arménien IgM Hamster Arménien IgM CD44 Alexa 700 BioLegend 1/200 Souris Rat IgG2b, κ IM7 HM40-3 HM40-3 116

CD62 L PE BD Biosciences 1/300 Souris Rat IgG2a, κ MEL-14 CD80 APC BioLegend 1/100 Souris CD80 PE-Cy5 BioLegend 1/200 Souris Hamster Arménien IgG Hamster Arménien IgG 16-10A1 16-10A1 CD86 FITC BioLegend 1/500 Souris Rat IgG2a, κ GL-1 CD86 PE-Cy5 BioLegend 1/1500 Souris Rat IgG2a, κ GL-1 CD86 PE-Cy7 BioLegend 1/800 Souris Rat IgG2a, κ GL-1 CD150 PE-Cy7 BioLegend 1/400 Souris Rat IgG2a, λ TC15-12F12.2 CMH-II (I-A/I-E) PE BioLegend 1/4000 Souris Rat IgG2b, κ M5/114.15.2 BrdU FITC BD Biosciences 1/100 Microscopie Anticorps primaires Calnexin Non couplé Abcam 1/200 Souris Lapin CD11c Non couplé BioLegend 1/100 Souris Hamster Arménien EEA-1 Non couplé 1/200 Souris Chèvre Mono- and polyubiquitinylated conjugates FK2 Non couplé Enzo Life Science 1/1000 Souris Souris I-A/I-E Non couplé BioLegend 1/300 Souris Rat IgG2b, κ M5/114.15.2 Lamp-1 Non couplé Developmental Studies Hybridoma Bank, University of Iowa N418 1/100 Souris Rat 1D4B NF-κB p65 Non couplé Santa Cruz 1/200 Souris Lapin LPS lisse Non couplé Fait maison 1/2000 Brucella abortus 1E6 Non couplé Fait maison 1/2000 Salmonella Souris Anticorps secondaires Vache 546 IgG (chaines légères + lourdes) FITC Jackson ImmunoResearch 1/100 Vache Chèvre IgG (chaines légères + lourdes) Alexa-488 Jackson ImmunoResearch 1/500 Vache Chèvre IgG (chaines légères + lourdes) Alexa-555 Jackson ImmunoResearch 1/500 Lapin Âne 117

IgG (chaines légères + lourdes) Alexa-647 Jackson ImmunoResearch 1/500 Lapin Chèvre IgG (chaines légères + lourdes) Pacific Blue Invitrogen 1/500 Lapin Chèvre IgG (chaines légères + lourdes) Alexa-594 Jackson ImmunoResearch 1/500 Hamster Arménien Chèvre IgG (chaines légères + lourdes) Alexa-647 Jackson ImmunoResearch 1/500 Rat Poulet IgG (chaines légères + lourdes) Alexa-546 Jackson ImmunoResearch 1/500 Chèvre Âne IgG (chaines légères + lourdes) Alexa-647 Jackson ImmunoResearch 1/500 Souris IgG1 Chèvre Autres Noyau : TOPRO-3 TOPRO-3 Invitrogen 1/1000 Phalloïdine Alexa-546 Invitrogen 1/1000 Biochimie myc Fait maison 1/1000 Humain Souris 9E10 Omp25 A. Cloeckaert 1/3000 Brucella abortus Souris IgG (chaines légères + lourdes) HRP Sigma Aldrich 1/1000 0 Souris Mouse IgG TrueBlot HRP Ebiosciences 1/1000 Souris Tableau 3 : Plasmides Plasmide pdonr Zeo pdonr Zeo::CD150 (2-3) Description Vecteur de clonage BP - Confère une résistance à la zéocine Vecteur de clonage BP contant les exons 2 et 3 de CD150 - Confère une résistance à la zéocine Référence Source ou Life Technologies Cette étude 118

pcmv::myccd150 pgem::omp25 pbbr-mcs4::omp25 Vecteur de destination de clonage Gateway contenant un tag myc en N-terminal et contenant les exons 2 et 3 de CD150 - Expression cellules eucaryotes - Confère une résistance à l'ampicilline Vecteur de clonage contenant omp25 - Confère une résistance à l'ampicilline Vecteur de destination, replicatif dans Brucella contenant omp25 pour complémenter la souche mutante omp25 - Confère une résistance à l'ampicilline Cette étude Cette étude Cette étude Peptides bloquants et contrôles CD150 CD150 a été bloqué avec un peptide bloquant, ou avec un peptide contrôle (ne bloquant pas CD150). Ces peptides ont été synthétisés par Thermo Science. Les séquences des peptides sont les suivantes : FCKQLKLYEQVSPPE pour le peptide bloquant, et pour le peptide non bloquant : DLSKGSYPDHLEDGY. Ils sont resuspendus en PBS stérile sans endotoxine à 10mg/ml. A cause de la cystéine contenue dans le peptide bloquant, les deux peptides sont traités au N-Ethylmaleimide (NEM) (Pierce, Thermo Scientific). Une fois les peptides resuspendus à 10mg/ml, 10mg de NEM est ajouté, puis les peptides sont incubés durant 2 h à température ambiante sous agitation. Les peptides sont ensuite dialysés pour enlever le surplus de NEM dans le milieu. IV. C. BACTÉRIOLOGIE Tableau 4 : Souches bactériennes utilisées Souche Description B.abortus 2308 souche sauvage btpa btpa dans la souche 2308 btpb btpb dans la souche 2308 Résistance Antibiotique Acide Nalidixique (Nal) Nal Gentamicine (Gm) Nal Kanamycine (Km) Référence ou Source D. Comerci Salcedo et al 2008 Salcedo et al 2012 btpa/btpb btpa btpb dans la souche 2308 Nal - Gm - Km Salcedo et al 2012 119

btpbpbtpb btpb complémenté avec btpb dans le plasmide pbbr1mcs-4 dans la souche 2308 Nal - Ampicilline (Amp) - Km omp25 omp25 dans la souche 2308 Nal - Km Salcedo et al 2012 Manterola et al 2007 virb9 virb9 dans la souche 2308 Celli et al 2005 omp25pomp25 omp25 dans la souche 2308 complémenté avec un plasmide Nal - Km - Amp Cette étude contenant omp25 E. coli DH5α Souche E. coli permettant l amplification de plasmide, DH5α pcmv contenant un plasmide pcmv myc::cd150 avec un tag myc en N terminal et les 2 premiers exons de CD150 Amp Cette étude JMH109 pgem::omp25 S17 λpir S17 λpir pbbr- MCS4::omp25 Souche E. coli pour le clonage du pgem et permettant un screen bleu/blanc contenant le pgem avec omp25 souche E. coli capable de conjugaison avec Brucella souche E. coli capable de conjugaison avec Brucella contenant le plasmide pour complémenté omp25 Amp Amp Cette étude X. De Bolle Cette étude Salmonella enterica typhymurium 12023 souche sauvage S. Méresse Infection des cellules Les cellules sont infectées à différents multiplicity of infection (MOI) : 30 pour les BMDC, 50 pour les BMDM, et 500 pour les HeLa. Les bactéries sont ajoutées aux cellules dans du milieu cellulaire, puis les cellules sont centrifugées à 400g 4 C pendant 10 min, les cellules sont ensuite placées à 37 C /5 % CO 2 durant 10 min pour les BMDM et Raw, 30 min pour les BMDC, 1 h pour les HeLa. Les cellules sont alors lavées 2 fois avec du milieu puis l on rajoute du milieu contenant 50µg/ml de gentamicine (ou 100µg/ml streptomycine quand il y a une bactérie résistante à la gentamicine) pour tuer les bactéries extracellulaires et sont incubées durant 1 h à 37 C / 5 % CO 2, ensuite le milieu est remplacé par un milieu contenant seulement 10µg/ml de gentamicine (ou 20µg/ml streptomycine). Les cellules sont incubées à 37 C / 5 % CO 2 le temps voulu. 120

Infection des souris Des cultures liquides de Brucella sont faites sur 16 h, les bactéries sont ensuite diluées dans du PBS stérile sans endotoxine à la concentration voulue. Les souris de 6 à 8 semaines sont ensuite infectées par injection intra-péritonéale. 1.10 6 CFU/souris sont injectées dans 200 µl de PBS stérile sans endotoxine. Dénombrement des bactéries Les cellules sont lavées 5 fois en PBS, puis lysées dans 1ml de 0.1 % Triton X-100 (Sigma- Aldrich) en H 2 O stérile, le lysat est ensuite dilué en série dans du PBS stérile avant d être répartis sur des boîtes de TSA, laissées à 37 C 3 j avant le dénombrement. Les organes sont écrasés avec le piston d une seringue dans 1 ml (pour la rate, le MLN) ou 2 ml (pour le foie) de 0.1% Triton X-100 en H 2 O, le lysat est ensuite dilué en série dans du PBS avant d être répartis sur des boîtes de TSA, laissées à 37 C 3 j avant le dénombrement. Préparation de stock glycérol 1 ml de culture de bactérie (phase stationnaire) est ajouté à 300µl de glycérol 80 %. Les tubes sont ensuite congelés à -80 C. Préparation de bactéries thermo-compétentes Les bactéries thermo-compétentes sont faites à partir d une culture liquide sur la nuit. 500µl de cette culture est diluée dans 100 ml de LB et les bactéries sont laissées à 37 C sous agitation jusqu à atteindre une DO de 0.5 à 600 nm. Les bactéries sont alors centrifugées 15 min à 5000 rpm à 4 C. Elles sont ensuite lavées avec 10 ml de MgCl 2 0.1M sur glace, re centrifugées 15 min à 5000 rpm à 4 C, resuspendues et lavées en CaCl 2 0.1 M sur glace. Après une dernière centrifugation, les bactéries sont resuspendues dans 5 ml de 15 % glycérol dans 0.1 M de CaCl 2 et congelées à 80 C. Transformation des bactéries Les bactéries thermo-compétentes sont décongelées sur glace, 50µl de bactéries sont ajoutées à 2 µl à 5 µl d ADN et incubées durant 30 minutes sur glace. Le mélange bactérie / ADN est ensuite mis à 42 C durant 45 secondes, puis 2 min sur glace. 500µl de SOC (contenant du glucose) sont ajoutées avant une incubation à 37 C sous agitation 1 h. Les bactéries sont 121

ensuite étalées sur boite LB agar contenant l antibiotique approprié pour la sélection selon le plasmide transformé. IV. D. BIOLOGIE CELLULAIRE Marquage pour cytométrie de flux Les cellules sont récoltées puis centrifugées à 400 g durant 5 min, les culots cellulaires sont resuspendus dans le mix d anticorps et incubés pendant 20 min protégés de la lumière. Un premier lavage en tampon FACS est réalisé, puis un second en PBS avant une fixation en Paraformaldéhyde (PFA) 3.2 % (EMS) durant 20 min à température ambiante, protégé de la lumière. Les échantillons sont ensuite dilués au 1/2 dans du PBS et stockés à 4 C avant le passage au FACS. Pour les marquages BrdU, le protocole suivi est celui indiqués par les fabricants dans le kit BD BrdU FITC Assay (BD Biosciences). Les échantillons sont analysés soit sur un FACS CantoII (BD Biosciences) ou un FACS LSRII (BD Biosciences). Les résultats sont analysés sur BD FACSDiva (BD Biosciences) et FlowJO (TreeStar). Marquage immunofluorescent Les cellules sont réparties sur coverslips, infectées puis au temps voulu, les coverslips sont lavés deux fois dans du PBS, puis fixés en PFA 3.2% pendant 20 min avant d être lavés en PBS deux fois. Pour le marquage NF-κB les cellules sont perméabilisées en Saponine 0.1 % en PBS durant 10 min à température ambiante, puis les interactions non spécifiques sont bloquées pendant 1 h à température ambiante dans du PBS / 2% BSA. Les anticorps primaires sont incubés durant 1 h, les coverslips sont ensuite lavés deux fois en PBS avant l incubation avec les anticorps secondaires durant 45 min. Les coverslips sont montés sur lame avec du Prolong Gold (Life Technologies). Pour les autres marquages, les cellules sont incubées pendant 1 h dans un mélange de Saponine 0.1 % / 1 % Sérum de Cheval / PBS pour perméabiliser les cellules et bloquer les interactions spécifiques. Les anticorps primaires sont incubés 1 h à température ambiante, et les anticorps secondaires sont incubés durant 30 minutes. Les lamelles sont ensuite montées sur lame soit en Prolong Gold, soit en Prolong Gold + DAPI (Life Technologies). 122

Les marquages sont analysés sur un microscope confocal Leica SP5X avec le logiciel LAS de Leica. Au moins 50 cellules sont comptées dans 4 expériences indépendantes pour les quantifications, des images de 1024 x 1024 pixels sont ensuite assemblées dans Image J ou Adobe Photoshop CS5 (Adobe). Quantification de cytokines par Cytometric Beads Assay (CBA) Les cytokines et sérum des souris sont analysés par CBA (BD Biosciences) soit avec le kit Mouse Inflammation pour les cytokines TNF-α, IL-6, IL-12p70, IL-10, MCP-1, et IFN-γ soit avec le kit Th1/Th2/Th17 pour les cytokines IL-2, IL-4 et IL-17A. Les échantillons sont traités suivant le protocole fourni par le fabricant. Les échantillons sont acquis sur un FACS CantoII (BD Biosciences). Les résultats sont ensuite analysés sur le logiciel FCAP Array (BD Biosciences). Co-culture lymphocytes T BMDC Les BMDC ont été préalablement réparties en plaque 96 puits à fond rond (BD Falcon) pour un ratio DC : T de 1 : 4. Les cellules ont été soit stimulées avec les LPS et les extraits de Brucella 16 h, soit infectées et incubées avec de l ovalbumine ou les peptides correspondants durant 4 h. Les cellules sont ensuite lavées pour se débarrasser de l ovalbumine et des peptides. Pour analyser la prolifération des LT dans les différentes conditions, ceux-ci sont marqués, après purification, avec du CFSE (Life Technologies), selon les instructions du fabricant. Après le marquage, les LT sont ajoutés aux BMDC dans du milieu LT contenant du GM-CSF pour permettre la survie des BMDC, et sont cultivés durant 3 j à 37 C 5 % CO 2. IV. E. BIOLOGIE MOLECULAIRE Purification de plasmides : «mini et maxi prep» Les plasmides utilisés ont été amplifiés par mini et maxi prep selon les besoins avec les kits «Wizard Plus SV Minipreps DNA Purification» de Promega en suivant le protocole du fabricant, «Plasmid Mini Kit» (Qiagen) pour le clonage de CD150 en gateway, en suivant le protocole du fabricant, et le kit «Plasmid Maxi Kit» (Qiagen) pour les maxiprep de plasmides, en suivant le protocole du fabricant. 123

Extraction d ARN, rétro-transcription en cdna et qpcr Les cellules sont récoltées, lavées une fois en PBS puis lysées à l aide de tampon RLT présent dans le Rneasy Mini Kit (Qiagen), l extraction est effectuée selon le protocole fourni par le fabricant. Les rétro-transcription sont faites à partir de 300ng d ARN, et avec le kit QuantiTect Reverse Transcription (Qiagen) selon les instructions du fabricant. Les qpcr sont faites avec du SYBR Green (Takara) selon le protocole fourni par le fabricant à partir de 2 µl de cdna. Les échantillons sont utilisés sur une machine à qpcr 7300 Real- Time PCR System (Applied Biosystems). Les échantillons sont normalisés en fonction de l expression de l HPRT. Amorces Les différentes amorces utilisées en qpcr et pour les clonages sont listées dans le tableau cidessous : Tableau 5 : Amorces Nom Sens Séquence Utilisation Pour qpcr HPRT sens 5'-3' AGCCCTCTGTGTGCTCAAGG qpcr HPRT antisens 5'-3' CTGATAAAATCTACAGTCATAGGAA TGGA qpcr ATF6 sens 5'-3' CCACCAGAAGTATGGGTTCG qpcr ATF6 antisens 5'-3' TGGCCTCCAGTCCTAGCATA qpcr IRE1 sens 5'-3' CGGCCTTTGCTGATAGTCTC qpcr IRE1 antisens 5'-3' GGCAGTGAGGCTGCATAGTC qpcr SLAM sens 5'-3' CGGGAGAGTGAAGGATGGTA qpcr SLAM antisens 5'-3' TCTCGTTCTCCTGGGTTTTG qpcr Mincle sens 5'-3' TGGCAATGGGTGGATGATA qpcr Mincle antisens 5'-3' AGTCCCTTATGGTGGCACAG qpcr Ptgs2 sens 5'-3' ACCTCTGCGATGCTCTTCC qpcr PTGS2 antisens 5'-3' TCATACATTCCCCACGGTTT qpcr IDO1 sens 5'-3' CCCTGGGGTACATCACCAT qpcr IDO1 antisens 5'-3' GAGAGCTCGCAGTAGGGAAC qpcr IL-4 sens 5'-3' ACTCTTTCGGGCTTTTCGAT qpcr IL-4 antisens 5'-3' TTGCATGATGCTCTTTAGGC qpcr SOCS1 sens 5'-3' CACCTTCTTGGTGCGCGAC qpcr SOCS1 antisens 5'-3' AAGCCATCTTCACGCTGAGC qpcr SOCS2 sens 5'-3' GTTGCCGGAGGAACAGTC qpcr SOCS2 antisens 5'-3' TCTCTTTGGCTTCATTAACAGTCA qpcr SOCS3 sens 5'-3' CCTTCAGCTCCAAAAGCGAGTAC qpcr SOCS3 antisens 5'-3' GCTCTCCTGCAGCTTGCG qpcr TNFα sens 5'-3' CATCTTCTCAAAATTCGAGTGACAA qpcr 124

TNFα antisens 5'-3' TGGGAGTAGACAAGGTACAACCC qpcr MIP2 sens 5'-3' GCGGTCAAAAAGTTTGCCTTG qpcr MIP2 antisens 5'-3' CTCCTCCTTTCCAGGTCAGTT qpcr IFNγ sens 5'-3' TCAAGTGGCATAGATGTGGAAGAA qpcr IFNγ antisens 5'-3' TGGCTCTGCAGGATTTTCATG qpcr NOS2 sens 5'-3' CAGCTGGGCTGTACAAACCTT qpcr NOS2 antisens 5'-3' CATTGGAAGTGAAGCGTTTCG qpcr IL-12b sens 5'-3' AAATTACTCCGGACGGTTCA qpcr IL-12b antisens 5'-3' ACAGAGACGCCATTCCACAT qpcr IL-6 sens 5'-3' GAGGATACCACTCCCAACAGACC qpcr IL-6 antisens 5'-3' AAGTGCATCATCGTTGTTCATACA qpcr IFNβ sens 5'-3' GAAAAGCAAGAGGAAAGATT qpcr IFNβ antisens 5'-3' AAGTCTTCGAATGATGAGAA qpcr IL1β sens 5'-3' TCCAGGATGAGGACATGAGCAC qpcr IL1β antisens 5'-3' GAACGTCACACACCAGCAGGTTA qpcr IFN-α sens 5'-3' GAGGAAATACTTCCACAGGATCACT GT qpcr IFN-α antisens 5'-3' GACAGGGCTCTCCAGACTTCTGCTC TG qpcr KC sens 5'-3' CAGCCACCCGCTCGCTTCTC qpcr KC antisens 5'-3' TCAAGGCAAGCCTCGCGACCAT qpcr Pour clonage SLAM2-3 sens SLAM2-3 antisens Omp25 sens 5'-3' 5'-3' 5'-3' GGGGACAAGTTTGTACAAAAAAGC AGGCTTCACAGGTGGAGGTGTGATG GAT GGGGACCACTTTGTACAAGAAAGCT GGGTCCTACTGAGGAGGATTCCTGC TTGC CCCGAATTCATGCGCACTCTTAAGT CTCTCG Omp25 antisens 5'-3' CCCGGATCCTTAGAACTTGTAGCCG Clonage Gateway Tag en N terminal de CD150/SLAM Clonage Gateway Tag en N terminal de CD150/SLAM Clonage pour complémentation (site EcoRI ajouté + codon initiation ATG) Clonage pour complémentation (site BamHI ajouté + codon STOP) Clonage de SLAM : Système Gateway Nous avons voulu construire une protéine de fusion entre un tag (HA ou myc) et les exons 2 et 3 de SLAM (correspondant à l ectodomaine de la protéine). Pour cela nous avons utilisé le système Gateway Complémentation de omp25 Pour complémenter le mutant omp25, le gène omp25 a été amplifié par PCR (25 cycles) à partir d un extrait de Brucella heat killed (1/10), avec de la HiFi Taq polymérase (invitrogen) 125

selon les instructions du fabricant. Après purification à partir du gel d agarose 1 %, 7 µl du produit de PCR, 2 µl de Taq Buffer (5X, Promega), 0,2mM de ATP, 5 unités de Taq (Promega) sont incubés à 70 C durant 30 minutes pour du «A-tailing», permettant le clonage dans le vecteur pgem-t (Promega). 3 µl de cette réaction sont utilisés pour la ligation dans le pgem-t selon les instructions du fabricant. Après transformation dans des E. coli JMH109, les clones sélectionnés sont séquencés, puis 1 µg de plasmide est digéré avec EcoRI HF et BamHI (NEB). 1 µg du vecteur de destination, pbbr-mcs4 (Ampicilline résistant) a aussi été digéré par les mêmes enzymes. Après la digestion enzymatique, les fragments sont chargés sur un gel, et purifiés à partir du gel (en utilisant qiaquick gel extraction de Qiagen selon les instructions du fabricant. 7,5 µl d insert (omp25 provenant du pgem digéré) et 2,5 µl de pbbr-mcs4 sont utilisés pour une ligation avec la T4 DNA ligase (Promega) sur la nuit à 16 C. 5 µl de la ligation est transformé dans des DH5α. A partir de cette culture, des minipreps sont faites et le plasmide pbbr-mcs4 omp25 est transformé dans des S17 λpir. Conjugaison pour échange de plasmide 50µl de S17 λpir contenant le plasmide recombiné pbbr-mcs4 omp25 sont ajoutés à 1 ml de culture liquide (phase stationnaire) de la souche omp25. Après 2 lavages en TSB, le culot est resuspendu dans 100µl de TSB et mis sur une boite durant 4h à 37 C ou 16h à température ambiante. Les bactéries sont ensuite ré-isolées sur une boite contenant de l ampicilline, de l acide nalidixique et de la kanamycine et mises à 37 C durant 3-4 jours. Génotypage des souris Knock-Out (KO) Les queues des souris à génotyper sont coupées puis digérées dans 500 µl de tampon de lyse à 55 C durant une nuit avec agitation (750rpm). Le lendemain, les échantillons sont centrifugés, à vitesse maximale, 15 min pour éliminer les déchets, le surnageant est récupéré et dilué volume à volume avec de l isopropanol 100 %, une fois l ADN précipité, les échantillons sont centrifugés à vitesse maximale 5 min, puis les culots sont séchés, avant d être resolubilisés en H 2 O sans nucléases (Qiagen) à 55 C durant 1 h sous agitation. La quantité d ADN est mesurée au nanodrop (Applied Biosystem) 200 ng de cet ADN sera utilisé pour la PCR. La PCR est effectuée avec un kit Taq Polymérase de Takara selon les instructions du vendeur, les amorces utilisées pour détecter l allèle sauvage sont : 5 GAAGGATGGTACTTGGTG 3 et 5 CTCAGAACCTCTGCTGTAGC 3, pour détecter l allèle KO les amorces utilisées sont 5 CTCAGAACCTCTGCTGTAGC 3 et 5 126

TCCGGCAGTTGGGAAGCAAAG 3. Après migration sur gel d agarose 1 % en TAE. L allèle sauvage migrera à 500 bp et l allèle KO à 800 bp. Transfection Les cellules Cos-7 sont trypsinisées et diluées dans des boîtes de culture appropriées, 8 h après les cellules sont transfectées comme décrit ci-dessous. Du Fugène (Promega) est ajouté aux 4/5 du volume de DMEM final et incubé à température ambiante 5 min. Durant ce temps l ADN est dilué dans 1/5 volume de DMEM final. L ADN est ensuite ajouté au Fugène et incubé durant 15 min à température ambiante. Le mélange ADN / Fugène est ensuite ajouté aux cellules. 1.5µg d ADN est utilisé pour la transfection. Les cellules sont récoltées 48 h post-transfection. I. A. BIOCHIMIE ELISA Les cytokines et sérum quantifiés par ELISA l ont été en utilisant les kits ebiosciences correspondants : TNF-α, IFN-β, IL12p70, IL23/IL12p40 et IL1-β et selon les instructions du fabricant. Extraction de protéines Les cellules sont récoltées, lavées une fois en PBS, puis lysées dans du tampon de lyse composé de PBS 1X, Triton 0.1 %, et d inhibiteur de protéases (PMSF 20 µg/ml, LECK 20 µg/ml, TPCK 20 µg/ml dilué en éthanol 100%). Le lysat est congelé à -80 C. Après décongélation, le lysat est centrifugé à vitesse maximale durant 10 min à 4 C. Le surnageant est récupéré et congelé à -80 C ou -20 C selon l utilisation. Pull-down myc::cd150(2-3) et myc::sifa Les billes de protéines G (GE Healthcare) sont lavées en PBS, 0.1 % NP-40 et inhibiteur de protéases (tampon A). 50 µl de billes est ajouté à 10 µl d anticorps anti-myc 9E10 (à 10 mg/ml stock) dans 500 µl de tampon et incubé 1 h à 4 C sous agitation. Après des lavages, les billes couplées à l anticorps anti-myc (1/10) sont incubées durant 1 h à 4 C sous agitation avec les extraits protéiques provenant des COS-7 transfectés avec myc::cd150(2-3) ou myc:: SifA. Après 3 lavages en tampon A, les billes, liées aux protéines tagguées myc sont incubées 127

durant 1 h à 4 C sous agitation avec 1 µg d extraits membranaires de Brucella sauvage ou mutant pour omp25. Après 5 lavages en tampon A contenant du NaCl 0.5 M, et du SDS à 0.001 % les échantillons, resuspendus en tampon de charge SDS 5X (Bleu de bromophénol à 0.05 %, 50 % glycérol, 10 % SDS, 0.25 M de Tris-HCl ph 6.8, 0.5 M de DTT (ou 5% de β mercaptoéthanol)) sont bouillis à 95 C durant 10 minutes, puis centrifugés 5 min à vitesse maximale. Le surnageant est analysé par western blot pour détecter la présence de Omp25. Gel SDS-PAGE et western blot Les échantillons sont chargés sur un gel SDS-PAGE 12 %, après migration, les protéines sont transférées sur une membrane PVDF (pore 0.45 µm, Millipore) durant 35 min. Après 1 h de blocage dans du PBS / Tween 0.1 % / 4 % lait (tampon de blocage), la membrane est incubée toute la nuit à 4 C avec l anticorps primaire dilué dans du tampon de blocage. Après 3 lavages en PBS / Tween 0.1 %, la membrane est incubée durant 1 h à température ambiante avec l anticorps secondaire couplé à l HRP (dans le cas du pull-down, un anticorps Mouse IgG TrueBlot (ebioscience) est utilisé pour minimiser les bandes non spécifiques). Après 3 nouveaux lavages, la membrane est incubée 1 min en ECL (Pierce) avant d être exposée au film (ECL Amersham, Pierce). 128

V. Références 129

1. D'Anastasio, R., et al., Possible brucellosis in an early hominin skeleton from sterkfontein, South Africa. PLoS One, 2009. 4(7): p. e6439. 2. D'Anastasio, R., et al., Origin, evolution and paleoepidemiology of brucellosis. Epidemiology and infection, 2011. 139(1): p. 149-56. 3. Moreno, E., Retrospective and prospective perspectives on zoonotic brucellosis. Frontiers in microbiology, 2014. 4. Pappas, G., et al., The new global map of human brucellosis. The Lancet infectious diseases, 2006. 6(2): p. 91-9. 5. Pappas, G., The changing Brucella ecology: novel reservoirs, new threats. International journal of antimicrobial agents, 2010. 36 Suppl 1: p. S8-11. 6. Mick, V., et al., Brucella melitensis in France: persistence in wildlife and probable spillover from Alpine ibex to domestic animals. PLoS One, 2014. 9(4): p. e94168. 7. Garin-Bastuji, B., et al., Reemergence of Brucella melitensis Infection in Wildlife, France. Emerging infectious diseases, 2014. 20(9): p. 1570-1. 8. Ian Maudlin, S.W.-M., The Control of Neglected Zoonotic Diseases - A route to poverty alleviation, 2006, WHO: Geneva. 9. Atluri, V.L., et al., Interactions of the Human Pathogenic Brucella Species with Their Hosts. Annu Rev Microbiol, 2011. 65: p. 523-541. 10. Martirosyan, A., E. Moreno, and J.P. Gorvel, An evolutionary strategy for a stealthy intracellular Brucella pathogen. Immunological reviews, 2011. 240(1): p. 211-34. 11. Corbel, M.J., Brucellosis: an overview. Emerging infectious diseases, 1997. 3(2): p. 213-21. 12. Kato, Y., et al., Brucellosis in a returned traveler and his wife: probable person-to-person transmission of Brucella melitensis. Journal of travel medicine, 2007. 14(5): p. 343-5. 13. Godfroid, J., et al., Brucellosis at the animal/ecosystem/human interface at the beginning of the 21st century. Preventive veterinary medicine, 2011. 102(2): p. 118-31. 14. Moriyon, I., et al., Rough vaccines in animal brucellosis: structural and genetic basis and present status. Veterinary research, 2004. 35(1): p. 1-38. 15. Yazdi, H.S., et al., Abortions in pregnant dairy cows after vaccination with Brucella abortus strain RB51. The Veterinary record, 2009. 165(19): p. 570-1. 16. Barrio, M.B., et al., Rough mutants defective in core and O-polysaccharide synthesis and export induce antibodies reacting in an indirect ELISA with smooth lipopolysaccharide and are less effective than Rev 1 vaccine against Brucella melitensis infection of sheep. Vaccine, 2009. 27(11): p. 1741-9. 17. Young, E.J., et al., Thrombocytopenic purpura associated with brucellosis: report of 2 cases and literature review. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 2000. 31(4): p. 904-9. 18. Pappas, G., et al., Doxycycline-rifampicin: physicians' inferior choice in brucellosis or how convenience reigns over science. The Journal of infection, 2007. 54(5): p. 459-62. 19. Cardoso, P.G., et al., Brucella spp noncanonical LPS: structure, biosynthesis, and interaction with host immune system. Microbial cell factories, 2006. 5: p. 13. 20. Ferrero, M.C., C.A. Fossati, and P.C. Baldi, Smooth Brucella strains invade and replicate in human lung epithelial cells without inducing cell death. Microbes and infection / Institut Pasteur, 2009. 11(4): p. 476-83. 21. Arellano-Reynoso, B., et al., Cyclic beta-1,2-glucan is a Brucella virulence factor required for intracellular survival. Nat Immunol, 2005. 6(6): p. 618-25. 22. Martirosyan, A., et al., Brucella beta 1,2 cyclic glucan is an activator of human and mouse dendritic cells. PLoS pathogens, 2012. 8(11): p. e1002983. 23. Salcedo, S.P., et al., Brucella control of dendritic cell maturation is dependent on the TIRcontaining protein Btp1. PLoS pathogens, 2008. 4(2): p. e21. 24. Desvaux, M., et al., Secretion and subcellular localizations of bacterial proteins: a semantic awareness issue. Trends in microbiology, 2009. 17(4): p. 139-45. 25. McBride, M.J. and Y. Zhu, Gliding motility and Por secretion system genes are widespread among members of the phylum bacteroidetes. Journal of bacteriology, 2013. 195(2): p. 270-8. 26. Burns, D.L., Type IV transporters of pathogenic bacteria. Current opinion in microbiology, 2003. 6(1): p. 29-34. 130

27. O'Callaghan, D., et al., A homologue of the Agrobacterium tumefaciens VirB and Bordetella pertussis Ptl type IV secretion systems is essential for intracellular survival of Brucella suis. Molecular microbiology, 1999. 33(6): p. 1210-20. 28. Llosa, M., C. Roy, and C. Dehio, Bacterial type IV secretion systems in human disease. Mol Microbiol, 2009. 73(2): p. 141-51. 29. Fronzes, R., H. Remaut, and G. Waksman, Architectures and biogenesis of non-flagellar protein appendages in Gram-negative bacteria. The EMBO journal, 2008. 27(17): p. 2271-80. 30. Lacerda, T.L., S.P. Salcedo, and J.P. Gorvel, Brucella T4SS: the VIP pass inside host cells. Current opinion in microbiology, 2013. 16(1): p. 45-51. 31. Watarai, M., S. Makino, and T. Shirahata, An essential virulence protein of Brucella abortus, VirB4, requires an intact nucleoside-triphosphate-binding domain. Microbiology, 2002. 148(Pt 5): p. 1439-46. 32. Sieira, R., et al., Integration host factor is involved in transcriptional regulation of the Brucella abortus virb operon. Molecular microbiology, 2004. 54(3): p. 808-22. 33. Boschiroli, M.L., et al., The Brucella suis virb operon is induced intracellularly in macrophages. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(3): p. 1544-9. 34. Delrue, R.M., et al., A quorum-sensing regulator controls expression of both the type IV secretion system and the flagellar apparatus of Brucella melitensis. Cellular microbiology, 2005. 7(8): p. 1151-61. 35. Celli, J., et al., Brucella evades macrophage killing via VirB-dependent sustained interactions with the endoplasmic reticulum. The Journal of experimental medicine, 2003. 198(4): p. 545-56. 36. Rolan, H.G. and R.M. Tsolis, Mice lacking components of adaptive immunity show increased Brucella abortus virb mutant colonization. Infection and immunity, 2007. 75(6): p. 2965-73. 37. Verstreate, D.R., et al., Outer membrane proteins of Brucella abortus: isolation and characterization. Infection and immunity, 1982. 35(3): p. 979-89. 38. Douglas, J.T., et al., Porins of Brucella species. Infection and immunity, 1984. 44(1): p. 16-21. 39. Cloeckaert, A., et al., Major outer membrane proteins of Brucella spp.: past, present and future. Veterinary microbiology, 2002. 90(1-4): p. 229-47. 40. Cloeckaert, A., et al., Molecular and immunological characterization of the major outer membrane proteins of Brucella. FEMS microbiology letters, 1996. 145(1): p. 1-8. 41. Vizcaino, N., et al., Molecular characterization of a Brucella species large DNA fragment deleted in Brucella abortus strains: evidence for a locus involved in the synthesis of a polysaccharide. Infection and immunity, 1999. 67(6): p. 2700-12. 42. Barrionuevo, P., et al., Brucella abortus inhibits major histocompatibility complex class II expression and antigen processing through interleukin-6 secretion via Toll-like receptor 2. Infection and immunity, 2008. 76(1): p. 250-62. 43. Pasquevich, K.A., et al., Immunization with recombinant Brucella species outer membrane protein Omp16 or Omp19 in adjuvant induces specific CD4+ and CD8+ T cells as well as systemic and oral protection against Brucella abortus infection. Infection and immunity, 2009. 77(1): p. 436-45. 44. Billard, E., J. Dornand, and A. Gross, Interaction of Brucella suis and Brucella abortus rough strains with human dendritic cells. Infection and immunity, 2007. 75(12): p. 5916-23. 45. Manterola, L., et al., BvrR/BvrS-controlled outer membrane proteins Omp3a and Omp3b are not essential for Brucella abortus virulence. Infection and immunity, 2007. 75(10): p. 4867-74. 46. Martin-Martin, A.I., et al., Importance of the Omp25/Omp31 family in the internalization and intracellular replication of virulent B. ovis in murine macrophages and HeLa cells. Microbes and infection / Institut Pasteur, 2008. 10(6): p. 706-10. 47. Billard, E., J. Dornand, and A. Gross, Brucella suis prevents human dendritic cell maturation and antigen presentation through regulation of tumor necrosis factor alpha secretion. Infection and immunity, 2007. 75(10): p. 4980-9. 131

48. Jubier-Maurin, V., et al., Major outer membrane protein Omp25 of Brucella suis is involved in inhibition of tumor necrosis factor alpha production during infection of human macrophages. Infection and immunity, 2001. 69(8): p. 4823-30. 49. Edmonds, M.D., A. Cloeckaert, and P.H. Elzer, Brucella species lacking the major outer membrane protein Omp25 are attenuated in mice and protect against Brucella melitensis and Brucella ovis. Veterinary microbiology, 2002. 88(3): p. 205-21. 50. Pollak, C.N., et al., Outer membrane vesicles from Brucella abortus promote bacterial internalization by human monocytes and modulate their innate immune response. PLoS One, 2012. 7(11): p. e50214. 51. Avila-Calderon, E.D., et al., Characterization of outer membrane vesicles from Brucella melitensis and protection induced in mice. Clinical & developmental immunology, 2012. 2012: p. 352493. 52. Grant, B.D. and J.G. Donaldson, Pathways and mechanisms of endocytic recycling. Nature reviews. Molecular cell biology, 2009. 10(9): p. 597-608. 53. Lee, J.J., et al., Interplay between clathrin and Rab5 controls the early phagocytic trafficking and intracellular survival of Brucella abortus within HeLa cells. The Journal of biological chemistry, 2013. 288(39): p. 28049-57. 54. Veiga, E., et al., Invasive and adherent bacterial pathogens co-opt host clathrin for infection. Cell host & microbe, 2007. 2(5): p. 340-51. 55. Veiga, E. and P. Cossart, Listeria hijacks the clathrin-dependent endocytic machinery to invade mammalian cells. Nature cell biology, 2005. 7(9): p. 894-900. 56. Van Nhieu, G.T., et al., Mutations in the cytoplasmic domain of the integrin beta1 chain indicate a role for endocytosis factors in bacterial internalization. The Journal of biological chemistry, 1996. 271(13): p. 7665-72. 57. Wyrick, P.B., et al., Entry of genital Chlamydia trachomatis into polarized human epithelial cells. Infection and immunity, 1989. 57(8): p. 2378-89. 58. Veiga, E. and P. Cossart, The role of clathrin-dependent endocytosis in bacterial internalization. Trends in cell biology, 2006. 16(10): p. 499-504. 59. Baldwin, C.L. and R. Goenka, Host immune responses to the intracellular bacteria Brucella: does the bacteria instruct the host to facilitate chronic infection? Critical reviews in immunology, 2006. 26(5): p. 407-42. 60. Guzman-Verri, C., et al., GTPases of the Rho subfamily are required for Brucella abortus internalization in nonprofessional phagocytes: direct activation of Cdc42. The Journal of biological chemistry, 2001. 276(48): p. 44435-43. 61. Pei, J., J.E. Turse, and T.A. Ficht, Evidence of Brucella abortus OPS dictating uptake and restricting NF-kappaB activation in murine macrophages. Microbes and infection / Institut Pasteur, 2008. 10(6): p. 582-90. 62. von Bargen, K., J.P. Gorvel, and S.P. Salcedo, Internal affairs: investigating the Brucella intracellular lifestyle. FEMS microbiology reviews, 2012. 36(3): p. 533-62. 63. Watarai, M., et al., Cellular prion protein promotes Brucella infection into macrophages. The Journal of experimental medicine, 2003. 198(1): p. 5-17. 64. Fontes, P., et al., Absence of evidence for the participation of the macrophage cellular prion protein in infection with Brucella suis. Infect Immun, 2005. 73(10): p. 6229-36. 65. Kim, S., et al., Lipid raft microdomains mediate class A scavenger receptor-dependent infection of Brucella abortus. Microb Pathog, 2004. 37(1): p. 11-9. 66. Lee, J.J., et al., Toll-like receptor 4-linked Janus kinase 2 signaling contributes to internalization of Brucella abortus by macrophages. Infection and immunity, 2013. 81(7): p. 2448-58. 67. Czibener, C. and J.E. Ugalde, Identification of a unique gene cluster of Brucella spp. that mediates adhesion to host cells. Microbes Infect, 2012. 14(1): p. 79-85. 68. Castaneda-Roldan, E.I., et al., Characterization of SP41, a surface protein of Brucella associated with adherence and invasion of host epithelial cells. Cell Microbiol, 2006. 8(12): p. 1877-87. 132

69. Guzman-Verri, C., et al., GTPases of the Rho subfamily are required for Brucella abortus internalization in nonprofessional phagocytes: direct activation of Cdc42. J Biol Chem, 2001. 276(48): p. 44435-43. 70. Starr, T., et al., Brucella intracellular replication requires trafficking through the late endosomal/lysosomal compartment. Traffic, 2008. 9(5): p. 678-94. 71. Celli, J., Surviving inside a macrophage: the many ways of Brucella. Res Microbiol, 2006. 157(2): p. 93-8. 72. de Barsy, M., et al., Identification of a Brucella spp. secreted effector specifically interacting with human small GTPase Rab2. Cell Microbiol, 2011. 13(7): p. 1044-58. 73. Fugier, E., et al., The glyceraldehyde-3-phosphate dehydrogenase and the small GTPase Rab 2 are crucial for Brucella replication. PLoS Pathog, 2009. 5(6): p. e1000487. 74. Myeni, S., et al., Brucella modulates secretory trafficking via multiple type IV secretion effector proteins. PLoS pathogens, 2013. 9(8): p. e1003556. 75. Dohmer, P.H., et al., Identification of a type IV secretion substrate of Brucella abortus that participates in the early stages of intracellular survival. Cellular microbiology, 2014. 16(3): p. 396-410. 76. de Jong, M.F., et al., Identification of VceA and VceC, two members of the VjbR regulon that are translocated into macrophages by the Brucella type IV secretion system. Molecular microbiology, 2008. 70(6): p. 1378-96. 77. Celli, J., S.P. Salcedo, and J.P. Gorvel, Brucella coopts the small GTPase Sar1 for intracellular replication. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(5): p. 1673-8. 78. Pizarro-Cerda, J., et al., Virulent Brucella abortus prevents lysosome fusion and is distributed within autophagosome-like compartments. Infection and immunity, 1998. 66(5): p. 2387-92. 79. Pizarro-Cerda, J., et al., Brucella abortus transits through the autophagic pathway and replicates in the endoplasmic reticulum of nonprofessional phagocytes. Infection and immunity, 1998. 66(12): p. 5711-24. 80. Bellaire, B.H., R.M. Roop, 2nd, and J.A. Cardelli, Opsonized virulent Brucella abortus replicates within nonacidic, endoplasmic reticulum-negative, LAMP-1-positive phagosomes in human monocytes. Infection and immunity, 2005. 73(6): p. 3702-13. 81. Starr, T., et al., Selective subversion of autophagy complexes facilitates completion of the Brucella intracellular cycle. Cell host & microbe, 2012. 11(1): p. 33-45. 82. Barquero-Calvo, E., et al., Brucella abortus uses a stealthy strategy to avoid activation of the innate immune system during the onset of infection. PLoS One, 2007. 2(7): p. e631. 83. Spera, J.M., D.J. Comerci, and J.E. Ugalde, Brucella alters the immune response in a prpadependent manner. Microbial pathogenesis, 2014. 67-68: p. 8-13. 84. Gorvel, L., et al., Intracellular bacteria interfere with dendritic cell functions: role of the type I interferon pathway. PLoS One, 2014. 9(6): p. e99420. 85. Corsetti, P.P., et al., Lack of endogenous IL-10 enhances production of proinflammatory cytokines and leads to Brucella abortus clearance in mice. PLoS One, 2013. 8(9): p. e74729. 86. Takeuchi, O. and S. Akira, Pattern recognition receptors and inflammation. Cell, 2010. 140(6): p. 805-20. 87. Takeda, K. and S. Akira, TLR signaling pathways. Semin Immunol, 2004. 16(1): p. 3-9. 88. Dinarello, C.A., Proinflammatory cytokines. Chest, 2000. 118(2): p. 503-8. 89. Parkin, J. and B. Cohen, An overview of the immune system. Lancet, 2001. 357(9270): p. 1777-89. 90. Giambartolomei, G.H., et al., Lipoproteins, not lipopolysaccharide, are the key mediators of the proinflammatory response elicited by heat-killed Brucella abortus. Journal of immunology, 2004. 173(7): p. 4635-42. 91. Campos, M.A., et al., Role of Toll-like receptor 4 in induction of cell-mediated immunity and resistance to Brucella abortus infection in mice. Infection and immunity, 2004. 72(1): p. 176-86. 92. Copin, R., et al., MyD88-dependent activation of B220-CD11b+LY-6C+ dendritic cells during Brucella melitensis infection. Journal of immunology, 2007. 178(8): p. 5182-91. 133

93. Lapaque, N., et al., Differential inductions of TNF-alpha and IGTP, IIGP by structurally diverse classic and non-classic lipopolysaccharides. Cellular microbiology, 2006. 8(3): p. 401-13. 94. Conde-Alvarez, R., et al., The lipopolysaccharide core of Brucella abortus acts as a shield against innate immunity recognition. PLoS pathogens, 2012. 8(5): p. e1002675. 95. Berguer, P.M., et al., A polymeric bacterial protein activates dendritic cells via TLR4. Journal of immunology, 2006. 176(4): p. 2366-72. 96. Macedo, G.C., et al., Central role of MyD88-dependent dendritic cell maturation and proinflammatory cytokine production to control Brucella abortus infection. Journal of immunology, 2008. 180(2): p. 1080-7. 97. Gomes, M.T., et al., The role of innate immune signals in immunity to Brucella abortus. Frontiers in cellular and infection microbiology, 2012. 2: p. 130. 98. Huang, L.Y., et al., Th1-like cytokine induction by heat-killed Brucella abortus is dependent on triggering of TLR9. Journal of immunology, 2005. 175(6): p. 3964-70. 99. Macedo, G.C., et al., Central role of MyD88-dependent dendritic cell maturation and proinflammatory cytokine production to control Brucella abortus infection. J Immunol, 2008. 180(2): p. 1080-7. 100. de Almeida, L.A., et al., Toll-like receptor 6 plays an important role in host innate resistance to Brucella abortus infection in mice. Infection and immunity, 2013. 81(5): p. 1654-62. 101. Nathan, C., Neutrophils and immunity: challenges and opportunities. Nature reviews. Immunology, 2006. 6(3): p. 173-82. 102. Martinez de Tejada, G., et al., The outer membranes of Brucella spp. are resistant to bactericidal cationic peptides. Infection and immunity, 1995. 63(8): p. 3054-61. 103. Freer, E., et al., Brucella-Salmonella lipopolysaccharide chimeras are less permeable to hydrophobic probes and more sensitive to cationic peptides and EDTA than are their native Brucella sp. counterparts. Journal of bacteriology, 1996. 178(20): p. 5867-76. 104. Moreno, E., D.T. Berman, and L.A. Boettcher, Biological activities of Brucella abortus lipopolysaccharides. Infection and immunity, 1981. 31(1): p. 362-70. 105. Sarma, J.V. and P.A. Ward, The complement system. Cell and tissue research, 2011. 343(1): p. 227-35. 106. Fernandez-Prada, C.M., et al., Deletion of wboa enhances activation of the lectin pathway of complement in Brucella abortus and Brucella melitensis. Infection and immunity, 2001. 69(7): p. 4407-16. 107. Zhang, L., et al., A second IgG-binding protein in Staphylococcus aureus. Microbiology, 1998. 144 ( Pt 4): p. 985-91. 108. Blom, A.M., T. Hallstrom, and K. Riesbeck, Complement evasion strategies of pathogensacquisition of inhibitors and beyond. Molecular immunology, 2009. 46(14): p. 2808-17. 109. Maite Iriarte, D.G., Rose M.Delrue, Daniel Monreal,, I.L.-G. Raquel Conde, Jean-Jacques Letesson, and, and I. Moriyón, Brucella Lipopolysaccharide: Structure, Biosynthesis and Genetics - Brucella Molecular and Cellular Biology, I. López-Goñi and I. Moriyón, Editors. 2004: Norfolk. 110. Kreutzer, D.L., L.A. Dreyfus, and D.C. Robertson, Interaction of polymorphonuclear leukocytes with smooth and rough strains of Brucella abortus. Infection and immunity, 1979. 23(3): p. 737-42. 111. Canning, P.C., et al., Isolation of components of Brucella abortus responsible for inhibition of function in bovine neutrophils. The Journal of infectious diseases, 1985. 152(5): p. 913-21. 112. Jaeger, B.N., et al., Neutrophil depletion impairs natural killer cell maturation, function, and homeostasis. The Journal of experimental medicine, 2012. 209(3): p. 565-80. 113. Barquero-Calvo, E., et al., Neutrophils exert a suppressive effect on Th1 responses to intracellular pathogen Brucella abortus. PLoS pathogens, 2013. 9(2): p. e1003167. 114. Zwerdling, A., et al., Brucella abortus activates human neutrophils. Microbes and infection / Institut Pasteur, 2009. 11(6-7): p. 689-97. 115. Co, D.O., et al., Mycobacterial granulomas: keys to a long-lasting host-pathogen relationship. Clinical immunology, 2004. 113(2): p. 130-6. 134

116. Copin, R., et al., In situ microscopy analysis reveals local innate immune response developed around Brucella infected cells in resistant and susceptible mice. PLoS pathogens, 2012. 8(3): p. e1002575. 117. Gross, A., et al., In vitro Brucella suis infection prevents the programmed cell death of human monocytic cells. Infection and immunity, 2000. 68(1): p. 342-51. 118. Wang, M., et al., High levels of nitric oxide production decrease early but increase late survival of Brucella abortus in macrophages. Microbial pathogenesis, 2001. 31(5): p. 221-30. 119. Caron, E., et al., Brucella species release a specific, protease-sensitive, inhibitor of TNFalpha expression, active on human macrophage-like cells. Journal of immunology, 1996. 156(8): p. 2885-93. 120. Barrionuevo, P., et al., Brucella abortus inhibits IFN-gamma-induced FcgammaRI expression and FcgammaRI-restricted phagocytosis via toll-like receptor 2 on human monocytes/macrophages. Microbes and infection / Institut Pasteur, 2011. 13(3): p. 239-50. 121. Xavier, M.N., et al., CD4+ T cell-derived IL-10 promotes Brucella abortus persistence via modulation of macrophage function. PLoS pathogens, 2013. 9(6): p. e1003454. 122. Horowitz, A., K.A. Stegmann, and E.M. Riley, Activation of natural killer cells during microbial infections. Frontiers in immunology, 2011. 2: p. 88. 123. Fernandes, D.M., R. Benson, and C.L. Baldwin, Lack of a role for natural killer cells in early control of Brucella abortus 2308 infections in mice. Infection and immunity, 1995. 63(10): p. 4029-33. 124. Salmeron, I., et al., Impaired activity of natural killer cells in patients with acute brucellosis. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, 1992. 15(5): p. 764-70. 125. Dornand, J., et al., Impairment of intramacrophagic Brucella suis multiplication by human natural killer cells through a contact-dependent mechanism. Infection and immunity, 2004. 72(4): p. 2303-11. 126. Gao, N., et al., Regulatory role of natural killer (NK) cells on antibody responses to Brucella abortus. Innate immunity, 2011. 17(2): p. 152-63. 127. Banchereau, J., et al., Immunobiology of dendritic cells. Annu Rev Immunol, 2000. 18: p. 767-811. 128. Joffre, O., et al., Inflammatory signals in dendritic cell activation and the induction of adaptive immunity. Immunol Rev, 2009. 227(1): p. 234-47. 129. Hackstein, H. and A.W. Thomson, Dendritic cells: emerging pharmacological targets of immunosuppressive drugs. Nature reviews. Immunology, 2004. 4(1): p. 24-34. 130. Archambaud, C., et al., Contrasting roles of macrophages and dendritic cells in controlling initial pulmonary Brucella infection. Eur J Immunol, 2010. 40(12): p. 3458-71. 131. Heller, M.C., et al., Characterization of Brucella abortus infection of bovine monocytederived dendritic cells. Veterinary immunology and immunopathology, 2012. 149(3-4): p. 255-61. 132. Cirl, C., et al., Subversion of Toll-like receptor signaling by a unique family of bacterial Toll/interleukin-1 receptor domain-containing proteins. Nature medicine, 2008. 14(4): p. 399-406. 133. Sengupta, D., et al., Subversion of innate immune responses by Brucella through the targeted degradation of the TLR signaling adapter, MAL. Journal of immunology, 2010. 184(2): p. 956-64. 134. Radhakrishnan, G.K., et al., Brucella TIR Domain-containing Protein Mimics Properties of the Toll-like Receptor Adaptor Protein TIRAP. The Journal of biological chemistry, 2009. 284(15): p. 9892-8. 135. Chaudhary, A., et al., The Brucella TIR-like protein TcpB interacts with the death domain of MyD88. Biochemical and biophysical research communications, 2012. 417(1): p. 299-304. 136. Alaidarous, M., et al., Mechanism of bacterial interference with TLR4 signaling by Brucella Toll/interleukin-1 receptor domain-containing protein TcpB. The Journal of biological chemistry, 2014. 289(2): p. 654-68. 137. Gazzinelli, R.T., et al., In the absence of endogenous IL-10, mice acutely infected with Toxoplasma gondii succumb to a lethal immune response dependent on CD4+ T cells and 135

accompanied by overproduction of IL-12, IFN-gamma and TNF-alpha. Journal of immunology, 1996. 157(2): p. 798-805. 138. Hunter, C.A., et al., IL-10 is required to prevent immune hyperactivity during infection with Trypanosoma cruzi. Journal of immunology, 1997. 158(7): p. 3311-6. 139. Li, C., I. Corraliza, and J. Langhorne, A defect in interleukin-10 leads to enhanced malarial disease in Plasmodium chabaudi chabaudi infection in mice. Infection and immunity, 1999. 67(9): p. 4435-42. 140. Jacobs, M., et al., Increased resistance to mycobacterial infection in the absence of interleukin-10. Immunology, 2000. 100(4): p. 494-501. 141. Capo, C., et al., Production of interleukin-10 and transforming growth factor beta by peripheral blood mononuclear cells in Q fever endocarditis. Infection and immunity, 1996. 64(10): p. 4143-7. 142. Fernandes, D.M. and C.L. Baldwin, Interleukin-10 downregulates protective immunity to Brucella abortus. Infection and immunity, 1995. 63(3): p. 1130-3. 143. D'Elios, M.M., et al., T-cell response to bacterial agents. Journal of infection in developing countries, 2011. 5(9): p. 640-5. 144. Snapper, C.M., F.D. Finkelman, and W.E. Paul, Regulation of IgG1 and IgE production by interleukin 4. Immunological reviews, 1988. 102: p. 51-75. 145. Ouyang, W., J.K. Kolls, and Y. Zheng, The biological functions of T helper 17 cell effector cytokines in inflammation. Immunity, 2008. 28(4): p. 454-67. 146. Kagi, D., et al., Fas and perforin pathways as major mechanisms of T cell-mediated cytotoxicity. Science, 1994. 265(5171): p. 528-30. 147. Grillo, M.J., et al., What have we learned from brucellosis in the mouse model? Veterinary research, 2012. 43: p. 29. 148. Stevens, M.G., G.W. Pugh, Jr., and L.B. Tabatabai, Effects of gamma interferon and indomethacin in preventing Brucella abortus infections in mice. Infection and immunity, 1992. 60(10): p. 4407-9. 149. Brandao, A.P., et al., Host susceptibility to Brucella abortus infection is more pronounced in IFN-gamma knockout than IL-12/beta2-microglobulin double-deficient mice. Clinical & developmental immunology, 2012. 2012: p. 589494. 150. Ko, J., A. Gendron-Fitzpatrick, and G.A. Splitter, Susceptibility of IFN regulatory factor-1 and IFN consensus sequence binding protein-deficient mice to brucellosis. Journal of immunology, 2002. 168(5): p. 2433-40. 151. Murphy, E.A., et al., Interferon-gamma is crucial for surviving a Brucella abortus infection in both resistant C57BL/6 and susceptible BALB/c mice. Immunology, 2001. 103(4): p. 511-8. 152. Ko, J. and G.A. Splitter, Molecular host-pathogen interaction in brucellosis: current understanding and future approaches to vaccine development for mice and humans. Clinical microbiology reviews, 2003. 16(1): p. 65-78. 153. Vitry, M.A., et al., Crucial role of gamma interferon-producing CD4+ Th1 cells but dispensable function of CD8+ T cell, B cell, Th2, and Th17 responses in the control of Brucella melitensis infection in mice. Infection and immunity, 2012. 80(12): p. 4271-80. 154. Ladel, C.H., C. Blum, and S.H. Kaufmann, Control of natural killer cell-mediated innate resistance against the intracellular pathogen Listeria monocytogenes by gamma/delta T lymphocytes. Infection and immunity, 1996. 64(5): p. 1744-9. 155. Hara, T., et al., Predominant activation and expansion of V gamma 9-bearing gamma delta T cells in vivo as well as in vitro in Salmonella infection. The Journal of clinical investigation, 1992. 90(1): p. 204-10. 156. Rojas, R.E., et al., Regulation of human CD4(+) alphabeta T-cell-receptor-positive (TCR(+)) and gammadelta TCR(+) T-cell responses to Mycobacterium tuberculosis by interleukin-10 and transforming growth factor beta. Infection and immunity, 1999. 67(12): p. 6461-72. 157. Poquet, Y., et al., Expansion of Vgamma9 Vdelta2 T cells is triggered by Francisella tularensis-derived phosphoantigens in tularemia but not after tularemia vaccination. Infection and immunity, 1998. 66(5): p. 2107-14. 158. Skyberg, J.A., et al., Murine and bovine gammadelta T cells enhance innate immunity against Brucella abortus infections. PLoS One, 2011. 6(7): p. e21978. 136

159. Ottones, F., et al., Activation of human Vgamma9Vdelta2 T cells by a Brucella suis nonpeptidic fraction impairs bacterial intracellular multiplication in monocytic infected cells. Immunology, 2000. 100(2): p. 252-8. 160. Goenka, R., et al., B cell-deficient mice display markedly enhanced resistance to the intracellular bacterium Brucella abortus. The Journal of infectious diseases, 2011. 203(8): p. 1136-46. 161. Goenka, R., et al., B Lymphocytes provide an infection niche for intracellular bacterium Brucella abortus. The Journal of infectious diseases, 2012. 206(1): p. 91-8. 162. Spera, J.M., et al., A B lymphocyte mitogen is a Brucella abortus virulence factor required for persistent infection. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(44): p. 16514-9. 163. Spera, J.M., et al., A Brucella virulence factor targets macrophages to trigger B-cell proliferation. The Journal of biological chemistry, 2013. 288(28): p. 20208-16. 164. Winter, A.J., et al., Capacity of passively administered antibody to prevent establishment of Brucella abortus infection in mice. Infection and immunity, 1989. 57(11): p. 3438-44. 165. Elzer, P.H., et al., BALB/c mice infected with Brucella abortus express protracted polyclonal responses of both IgG2a and IgG3 isotypes. Immunology letters, 1994. 42(3): p. 145-50. 166. Hoffmann, E.M. and J.J. Houle, Contradictory roles for antibody and complement in the interaction of Brucella abortus with its host. Critical reviews in microbiology, 1995. 21(3): p. 153-63. 167. Vitry, M.A., et al., Humoral immunity and CD4+ Th1 cells are both necessary for a fully protective immune response upon secondary infection with Brucella melitensis. Journal of immunology, 2014. 192(8): p. 3740-52. 168. Rethi, B., et al., SLAM/SLAM interactions inhibit CD40-induced production of inflammatory cytokines in monocyte-derived dendritic cells. Blood, 2006. 107(7): p. 2821-9. 169. Hahm, B., N. Arbour, and M.B. Oldstone, Measles virus interacts with human SLAM receptor on dendritic cells to cause immunosuppression. Virology, 2004. 323(2): p. 292-302. 170. Cannons, J.L., S.G. Tangye, and P.L. Schwartzberg, SLAM family receptors and SAP adaptors in immunity. Annual review of immunology, 2011. 29: p. 665-705. 171. Cocks, B.G., et al., A novel receptor involved in T-cell activation. Nature, 1995. 376(6537): p. 260-3. 172. Aversa, G., et al., Engagement of the signaling lymphocytic activation molecule (SLAM) on activated T cells results in IL-2-independent, cyclosporin A-sensitive T cell proliferation and IFN-gamma production. Journal of immunology, 1997. 158(9): p. 4036-44. 173. Punnonen, J., et al., Soluble and membrane-bound forms of signaling lymphocytic activation molecule (SLAM) induce proliferation and Ig synthesis by activated human B lymphocytes. The Journal of experimental medicine, 1997. 185(6): p. 993-1004. 174. Mavaddat, N., et al., Signaling lymphocytic activation molecule (CDw150) is homophilic but self-associates with very low affinity. The Journal of biological chemistry, 2000. 275(36): p. 28100-9. 175. Howie, D., et al., The role of SAP in murine CD150 (SLAM)-mediated T-cell proliferation and interferon gamma production. Blood, 2002. 100(8): p. 2899-907. 176. Castro, A.G., et al., Molecular and functional characterization of mouse signaling lymphocytic activation molecule (SLAM): differential expression and responsiveness in Th1 and Th2 cells. Journal of immunology, 1999. 163(11): p. 5860-70. 177. Sayos, J., et al., The X-linked lymphoproliferative-disease gene product SAP regulates signals induced through the co-receptor SLAM. Nature, 1998. 395(6701): p. 462-9. 178. Le Borgne, M. and A.S. Shaw, SAP signaling: a dual mechanism of action. Immunity, 2012. 36(6): p. 899-901. 179. Latour, S., et al., Regulation of SLAM-mediated signal transduction by SAP, the X-linked lymphoproliferative gene product. Nature immunology, 2001. 2(8): p. 681-90. 180. Chan, B., et al., SAP couples Fyn to SLAM immune receptors. Nature cell biology, 2003. 5(2): p. 155-60. 181. Cannons, J.L., et al., Biochemical and genetic evidence for a SAP-PKC-theta interaction contributing to IL-4 regulation. Journal of immunology, 2010. 185(5): p. 2819-27. 137

182. Cannons, J.L., et al., SAP regulates T cell-mediated help for humoral immunity by a mechanism distinct from cytokine regulation. The Journal of experimental medicine, 2006. 203(6): p. 1551-65. 183. Nichols, K.E., G.A. Koretzky, and C.H. June, SAP: natural inhibitor or grand SLAM of T cell activation? Nature immunology, 2001. 2(8): p. 665-6. 184. Kane, L.P., et al., Akt provides the CD28 costimulatory signal for up-regulation of IL-2 and IFN-gamma but not TH2 cytokines. Nature immunology, 2001. 2(1): p. 37-44. 185. McInnes, I.B. and G. Schett, The pathogenesis of rheumatoid arthritis. The New England journal of medicine, 2011. 365(23): p. 2205-19. 186. Isomaki, P., et al., Increased expression of signaling lymphocytic activation molecule in patients with rheumatoid arthritis and its role in the regulation of cytokine production in rheumatoid synovium. Journal of immunology, 1997. 159(6): p. 2986-93. 187. Bleharski, J.R., et al., Signaling lymphocytic activation molecule is expressed on CD40 ligand-activated dendritic cells and directly augments production of inflammatory cytokines. Journal of immunology, 2001. 167(6): p. 3174-81. 188. Kiel, M.J., et al., SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell, 2005. 121(7): p. 1109-21. 189. Jordan, M.A., et al., Slamf1, the NKT cell control gene Nkt1. Journal of immunology, 2007. 178(3): p. 1618-27. 190. Kruse, M., et al., Signaling lymphocytic activation molecule is expressed on mature CD83+ dendritic cells and is up-regulated by IL-1 beta. Journal of immunology, 2001. 167(4): p. 1989-95. 191. Wang, N., et al., The cell surface receptor SLAM controls T cell and macrophage functions. The Journal of experimental medicine, 2004. 199(9): p. 1255-64. 192. Wang, N., et al., The costimulatory molecule SLAM is critical for pulmonary allergic responses. American journal of respiratory cell and molecular biology, 2006. 35(2): p. 206-10. 193. Tatsuo, H., et al., SLAM (CDw150) is a cellular receptor for measles virus. Nature, 2000. 406(6798): p. 893-7. 194. Hahm, B., J.H. Cho, and M.B. Oldstone, Measles virus-dendritic cell interaction via SLAM inhibits innate immunity: selective signaling through TLR4 but not other TLRs mediates suppression of IL-12 synthesis. Virology, 2007. 358(2): p. 251-7. 195. Pasquinelli, V., et al., Expression of signaling lymphocytic activation molecule-associated protein interrupts IFN-gamma production in human tuberculosis. Journal of immunology, 2004. 172(2): p. 1177-85. 196. Pasquinelli, V., et al., IFN-gamma production during active tuberculosis is regulated by mechanisms that involve IL-17, SLAM, and CREB. The Journal of infectious diseases, 2009. 199(5): p. 661-5. 197. Watanabe, H., et al., Innate immune response in Th1- and Th2-dominant mouse strains. Shock, 2004. 22(5): p. 460-6. 198. Calderon, J., et al., The receptor Slamf1 on the surface of myeloid lineage cells controls susceptibility to infection by Trypanosoma cruzi. PLoS pathogens, 2012. 8(7): p. e1002799. 199. Berger, S.B., et al., SLAM is a microbial sensor that regulates bacterial phagosome functions in macrophages. Nature immunology, 2010. 11(10): p. 920-7. 200. Ma, C., et al., Receptor signaling lymphocyte-activation molecule family 1 (Slamf1) regulates membrane fusion and NADPH oxidase 2 (NOX2) activity by recruiting a Beclin- 1/Vps34/ultraviolet radiation resistance-associated gene (UVRAG) complex. The Journal of biological chemistry, 2012. 287(22): p. 18359-65. 201. Salcedo, S.P., et al., BtpB, a novel Brucella TIR-containing effector protein with immune modulatory functions. Frontiers in cellular and infection microbiology, 2013. 3: p. 28. 202. Archambaud, C., et al., Contrasting roles of macrophages and dendritic cells in controlling initial pulmonary Brucella infection. European journal of immunology, 2010. 40(12): p. 3458-71. 203. Bowden, R.A., et al., Evaluation of immunogenicity and protective activity in BALB/c mice of the 25-kDa major outer-membrane protein of Brucella melitensis (Omp25) expressed in Escherichia coli. Journal of medical microbiology, 1998. 47(1): p. 39-48. 138

204. Goel, D., et al., Cell mediated immune response after challenge in Omp25 liposome immunized mice contributes to protection against virulent Brucella abortus 544. Vaccine, 2013. 31(8): p. 1231-7. 205. Goel, D. and R. Bhatnagar, Intradermal immunization with outer membrane protein 25 protects Balb/c mice from virulent B. abortus 544. Molecular immunology, 2012. 51(2): p. 159-68. 206. Commander, N.J., et al., Liposomal delivery of p-ialb and p-omp25 DNA vaccines improves immunogenicity but fails to provide full protection against B. melitensis challenge. Genetic vaccines and therapy, 2010. 8: p. 5. 207. Cannons, J.L., S.G. Tangye, and P.L. Schwartzberg, SLAM family receptors and SAP adaptors in immunity. Annu Rev Immunol, 2011. 29: p. 665-705. 208. Berger, S.B., et al., SLAM is a microbial sensor that regulates bacterial phagosome functions in macrophages. Nat Immunol, 2010. 11(10): p. 920-7. 209. Banchereau, J. and R.M. Steinman, Dendritic cells and the control of immunity. Nature, 1998. 392(6673): p. 245-52. 210. Fernandes, D.M., et al., Comparison of T cell cytokines in resistant and susceptible mice infected with virulent Brucella abortus strain 2308. FEMS immunology and medical microbiology, 1996. 16(3-4): p. 193-203. 211. Beuzon, C.R. and D.W. Holden, Use of mixed infections with Salmonella strains to study virulence genes and their interactions in vivo. Microbes and infection / Institut Pasteur, 2001. 3(14-15): p. 1345-52. 212. Lamontagne, J., et al., Extensive cell envelope modulation is associated with virulence in Brucella abortus. Journal of proteome research, 2007. 6(4): p. 1519-29. 213. Gamazo, C. and I. Moriyon, Release of outer membrane fragments by exponentially growing Brucella melitensis cells. Infection and immunity, 1987. 55(3): p. 609-15. 214. Ritchie, J.A., et al., Host interferon-gamma inducible protein contributes to Brucella survival. Frontiers in cellular and infection microbiology, 2012. 2: p. 55. 215. Malaviya, R., et al., The mast cell tumor necrosis factor alpha response to FimH-expressing Escherichia coli is mediated by the glycosylphosphatidylinositol-anchored molecule CD48. Proceedings of the National Academy of Sciences of the United States of America, 1999. 96(14): p. 8110-5. 216. Veillette, A., et al., SAP expression in T cells, not in B cells, is required for humoral immunity. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(4): p. 1273-8. 217. Henning, G., et al., Signaling lymphocytic activation molecule (SLAM) regulates T cellular cytotoxicity. European journal of immunology, 2001. 31(9): p. 2741-50. 218. Mehrle, S., et al., Enhancement of anti-tumor activity in vitro and in vivo by CD150 and SAP. Molecular immunology, 2008. 45(3): p. 796-804. 219. Kambayashi, T., et al., Cutting edge: Regulation of CD8(+) T cell proliferation by 2B4/CD48 interactions. Journal of immunology, 2001. 167(12): p. 6706-10. 220. Davidson, D., et al., Genetic evidence linking SAP, the X-linked lymphoproliferative gene product, to Src-related kinase FynT in T(H)2 cytokine regulation. Immunity, 2004. 21(5): p. 707-17. 221. Newman, R.M., et al., Identification and characterization of a novel bacterial virulence factor that shares homology with mammalian Toll/interleukin-1 receptor family proteins. Infection and immunity, 2006. 74(1): p. 594-601. 222. Low, L.Y., et al., Characterization of a TIR-like protein from Paracoccus denitrificans. Biochemical and biophysical research communications, 2007. 356(2): p. 481-6. 223. Spear, A.M., et al., A Toll/interleukin (IL)-1 receptor domain protein from Yersinia pestis interacts with mammalian IL-1/Toll-like receptor pathways but does not play a central role in the virulence of Y. pestis in a mouse model of bubonic plague. Microbiology, 2012. 158(Pt 6): p. 1593-606. 224. Amulic, B., et al., Neutrophil function: from mechanisms to disease. Annual review of immunology, 2012. 30: p. 459-89. 139

225. Yang, C.W., et al., Neutrophils influence the level of antigen presentation during the immune response to protein antigens in adjuvants. Journal of immunology, 2010. 185(5): p. 2927-34. 226. Blomgran, R. and J.D. Ernst, Lung neutrophils facilitate activation of naive antigen-specific CD4+ T cells during Mycobacterium tuberculosis infection. Journal of immunology, 2011. 186(12): p. 7110-9. 227. Smith, J.A., et al., Brucella induces an unfolded protein response via TcpB that supports intracellular replication in macrophages. PLoS pathogens, 2013. 9(12): p. e1003785. 228. Radhakrishnan, G.K., J.S. Harms, and G.A. Splitter, Modulation of microtubule dynamics by a TIR domain protein from the intracellular pathogen Brucella melitensis. The Biochemical journal, 2011. 439(1): p. 79-83. 229. Radhakrishnan, G.K. and G.A. Splitter, Biochemical and functional analysis of TIR domain containing protein from Brucella melitensis. Biochemical and biophysical research communications, 2010. 397(1): p. 59-63. 230. Galan, J.E. and D. Zhou, Striking a balance: modulation of the actin cytoskeleton by Salmonella. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(16): p. 8754-61. 231. Leone, P. and S. Meresse, Kinesin regulation by Salmonella. Virulence, 2011. 2(1): p. 63-6. 232. Billard, E., et al., High susceptibility of human dendritic cells to invasion by the intracellular pathogens Brucella suis, B. abortus, and B. melitensis. Infection and immunity, 2005. 73(12): p. 8418-24. 233. Weinhold, M., et al., The Attenuated Strain 19 Invades, Persists in, and Activates Human Dendritic Cells, and Induces the Secretion of IL-12p70 but Not IL-23. PLoS One, 2013. 8(6): p. e65934. 234. Serbina, N.V., et al., Monocyte-mediated defense against microbial pathogens. Annual review of immunology, 2008. 26: p. 421-52. 235. Serbina, N.V. and E.G. Pamer, Monocyte emigration from bone marrow during bacterial infection requires signals mediated by chemokine receptor CCR2. Nature immunology, 2006. 7(3): p. 311-7. 236. Peters, W., M. Dupuis, and I.F. Charo, A mechanism for the impaired IFN-gamma production in C-C chemokine receptor 2 (CCR2) knockout mice: role of CCR2 in linking the innate and adaptive immune responses. Journal of immunology, 2000. 165(12): p. 7072-7. 237. Bar-On, L. and S. Jung, Defining in vivo dendritic cell functions using CD11c-DTR transgenic mice. Methods in molecular biology, 2010. 595: p. 429-42. 238. Grosso, J.F. and M.N. Jure-Kunkel, CTLA-4 blockade in tumor models: an overview of preclinical and translational research. Cancer immunity, 2013. 13: p. 5. 239. Callahan, M.K. and J.D. Wolchok, At the bedside: CTLA-4- and PD-1-blocking antibodies in cancer immunotherapy. Journal of leukocyte biology, 2013. 94(1): p. 41-53. 140

VI. Annexe 141

VI. A. ARTICLE : LIPOPOLYSACCHARIDES WITH ACYLATION DEFECTS POTENTIATE TLR4 SIGNALING AND SHAPE T CELL RESPONSES. 142

Lipopolysaccharides with Acylation Defects Potentiate TLR4 Signaling and Shape T Cell Responses Anna Martirosyan 1,2,3, Yoichiro Ohne 4, Clara Degos 1,2,3, Laurent Gorvel 1,2,3, Ignacio Moriyón 5, Sangkon Oh 4, Jean-Pierre Gorvel 1,2,3 * 1 Aix-Marseille University UM 2, Centre d Immunologie de Marseille-Luminy, Marseille, France, 2 INSERM U 1104, Marseille, France, 3 CNRS UMR 7280, Marseille, France, 4 Baylor Institute for Immunology Research, Dallas, Texas, United States of America, 5 Institute for Tropical Health and Departamento de Microbiología y Parasitología, Universidad de Navarra, Pamplona, Spain Abstract Lipopolysaccharides or endotoxins are components of Gram-negative enterobacteria that cause septic shock in mammals. However, a LPS carrying hexa-acyl lipid A moieties is highly endotoxic compared to a tetra-acyl LPS and the latter has been considered as an antagonist of hexa-acyl LPS-mediated TLR4 signaling. We investigated the relationship between the structure and the function of bacterial LPS in the context of human and mouse dendritic cell activation. Strikingly, LPS with acylation defects were capable of triggering a strong and early TLR4-dependent DC activation, which in turn led to the activation of the proteasome machinery dampening the pro-inflammatory cytokine secretion. Upon activation with tetraacyl LPS both mouse and human dendritic cells triggered CD4 + T and CD8 + T cell responses and, importantly, human myeloid dendritic cells favored the induction of regulatory T cells. Altogether, our data suggest that LPS acylation controlled by pathogenic bacteria might be an important strategy to subvert adaptive immunity. Citation: Martirosyan A, Ohne Y, Degos C, Gorvel L, Moriyón I, et al. (2013) Lipopolysaccharides with Acylation Defects Potentiate TLR4 Signaling and Shape T Cell Responses. PLoS ONE 8(2): e55117. doi:10.1371/journal.pone.0055117 Editor: Edgardo Moreno, National University, Costa Rica Received August 23, 2012; Accepted December 19, 2012; Published February 4, 2013 Copyright: ß 2013 Martirosyan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Centre National de la Recherche Scientifique and Institut National de la Santé et de la Recherche Médicale, France. AM was a recipient of the Foundation de la Recherche Médicale (FRM), France. This work was also supported by National Institutes of Health (NIH)/National Institute of Allergy and Infectious Diseases (NIAID) - U19 AI057234. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: gorvel@ciml.univ-mrs.fr Introduction Dendritic cells (DC) play a key role in initiating and controlling the magnitude and the quality of adaptive immune responses [1,2]. Upon exposure to microbial stimuli, DC undergo a maturation process characterized by an increased formation of MHC peptide complexes, the up-regulation of co-stimulatory molecules, chemokine receptors and cytokine production [1,2,3]. Cytokines produced by DC play a key role in determining the type of generated CD4 + helper T cell (T H ) responses leading to T H1,T H2 or T H17 responses [1,2]. Moreover, DC play a pivotal role in the control of central tolerance and the induction of immune tolerance in the periphery. The ability of DC to induce tolerance depends on several parameters such as their maturation stage, anti-inflammatory and immunosuppressive agents, the nature of microbial stimuli, and the tissue microenvironment. In addition to deleting T cells, tolerogenic DC induce the differentiation and proliferation of T cells with regulatory/suppressive functions known as regulatory T cells (T reg ) [4]. Lipopolysaccharide (LPS) is an important virulence factor of Gram-negative bacteria responsible for septic shock in mammals. LPS is the major molecule of the bacterial outer membrane and can be massively released into the host during the course of infection [5,6]. LPS consists of the O-polysaccharide chain, the oligosaccharide core region and the lipid A. Typical LPS such as those of E. coli and most enteric bacteria express a lipid A composed of a bisphosphorylated glucosamine disaccharide carrying two amide- and two ester-linked acyl and hydroxyacyl chains. Additional acyloxyacyl chains are commonly present, resulting in penta or hexa-acyl lipid A, the dominant molecular lipid A species in most wild type enterobacteria [7,8]. It has been shown that variations of structural arrangements of lipid A such as a reduction in the number of charges or the number of acyl chains or a change in their distribution or saturation degree result in a dramatic reduction in endotoxicity. For instance, the synthetic precursor tetracyl lipid IVa has been described as a non-endotoxic molecule and proposed as an antagonist of hexa-acyl endotoxic LPS [9,10]. Moreover, some pathogens like the yersiniae modulate the degree of acylation of the lipid A depending upon the environmental conditions. Most notably, growth at 37uC causes Yersinia pestis to synthesize tri- and dominant tetra-acyl lipid A, with no hexa-acyl and only small amounts of penta-acyl molecules. Since these bacteria move from 20 25uC to 37uC when transmitted from the flea to the mammal host, Y. pestis express tetra-acyl lipid A which displays low immunostimulatory properties in mammals. This change has been described as a mark of pathogen adaptation to the host environment [7]. In this study, we investigated the relationship between lipid A acylation and the immunostimulatory properties of LPS in the context of mouse and human DC activation. We show that LPS with acylation defects described as not endotoxic are capable of inducing a strong and early TLR4-dependent cell activation. This leads to the activation of the proteasome machinery and the PLOS ONE www.plosone.org 1 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling degradation of newly synthetized pro-inflammatory cytokines. Mouse and human DC activated by tetra-acyl LPS trigger CD4 + and CD8 + T cell responses. Moreover, human DC activated by LPS with acylation defects display a semi-mature phenotype and induce high levels of regulatory T cells (T reg ). Materials and Methods Ethics Statement Animal experimentation was conducted in strict accordance with good animal practice as defined by the French animal welfare bodies (Law 87 848 dated 19 October 1987 modified by Decree 2001 464 and Decree 2001 131 relative to European Convention, EEC Directive 86/609). All animal work was approved by the Direction Départmentale des Services Vétérinaires des Bouches du Rhône (authorization number 13.118). INSERM guidelines have been followed regarding animal experimentation (authorization No. 02875 for mouse experimentation). Blood from healthy adult donors were collected at the Baylor Hospital Liver Transplant Clinic (Dallas, TX) after obtaining written informed consent. This study, including the consent form, was approved by the Institutional Review Board (IRB) of the Baylor Research Institute (BRI) (Dallas, TX). Any medical issue during blood collection from healthy donors was written and reported to the IRB at BRI. Lipopolysaccharides The methods used in the extraction, purification and characterization of the LPS used in this study have been described previously (Lapaque et al, 2006). Briefly, Y. pestis KIM6, E. coli MLK3 and its lipid A mutants MLK53 htrb 2 (lauroyl-transferase), MLK 1067 msbb 2 (miristoyl-transferase) and MLK 986 htrb 2/ msbb 2 were grown at the appropriate temperature, crude LPS obtained by the phenol-water method and then purified to remove traces of contaminant lipids and lipoproteins. The degree of lipid A acylation was determined by nanoelectrospray ionization time-of-flight mass spectrometry (ESI- TOF-MS) (Lapaque et al, 2006). For all experiments, LPS variants have been used at the concentration of 100 ng/ml. Lipid Iva was purchased from PeptaNova. Antibodies and Reagents The primary antibodies used for immunofluorecence microscopy were: mouse FK2 antibody (anti-mono- and polyubiquitinylated conjugates) (Enzo Life Science), affinity purified rabbit Rivoli antibody against murine I-A, NF-kB subunit p65/reia (Santa Cruz), CD11c (Bolegend). Pam2CSK4 was purchased from InvivoGen to activate DC. Antibodies used for flow cytometry included APC-CD11c (1 in 100), FITC-CD40 (1 in 50), FITC- CD80 (1 in 50), PE-CD86 (1 in 400), PE-IA-IE (MHC class II) (Pharmingen) (1 in 800), as well as PB-CD8 (1 in 200), A700- CD45.2 (1 in 300), APC-CD44 (1 in 400), PE-Cy7-CD25 (1 in 1500), APC-CD62L (1 in 400) (BD Biosciences and ebiosciences). For intracellular labeling of cytokines, IL-12 (p40/p70)-pe and TNF-a PE monoclonal antibodies (1 in 100)(Pharmingen) were used. The Aqua Dead Cell Stain (Invitrogen) was used to eliminate dead cells. Ovalbumine (OVA) was purchased from EndoGrade with purity.98% and endotoxin concentration,1eu/mg. SIINFEKL peptide was purchased from Schafer-N. Human mdc were sorted from PBMC of blood from healthy donors using lineage cocktail-fitc (BD Biosciences), CD123-PE (BD Biosciences), CD11c-APC (Biolegend), HLA-DR-Quantum Red (Sigma). Human mdc were stained with CD86-PE, CD83-FITC, CD40-APC and HLA-DR-PB (ebiosciences or Biolegends). 7- AAD was used to exclude dead cells. For intracellular labelling IL13-APC, INF-c-PE-Cy7, IL-17-PE and Granzyme B-APC antibodies were used. Isotype matched controls were used appropriately. Alexa Fluor 647 conjugated phospho-specific antibodies were used for Phospho flow experiments on human IL-4 DC and were all from BD Biosciences. Akt(S478), Btk(Y557)/ Itk(Y511), CREB(S133)/ATF1(S63), ERK1/2(T202/Y204), IRF- 7(S477/S479), Lck(Y505), NF-kB p65(s529), PLC-c1 (Y783), PLC-c2 (Y759), p38 MAPK(T180/Y182), b-catenin (S45), SHP- 2(Y542), Src(Y418), SLP-76(Y128), S6(S235/S236), STAT1(Y701), STAT1(S727), STAT3(Y705), STAT3(S727), STAT4(S693), STAT5(S694), STAT6(Y641), 4EBP1(T36/T45), Zap70(Y319)/Syk(Y352), JNK(T183/Y185). Mice and Cells C57Bl/6 mice from Jackson Laboratory and OT-I, OT II TCR transgenic mice on C57Bl/6 background were used. C57BL/6, Tlr4 2/2 and Tlr2 2/2 mice were maintained at the CIML animal house, France. Mouse bone marrow-derived DC (BMDC) and macrophages (BMDM) were prepared from 7 8 week-old female C57BL/6 mice as previously described (Lapaque et al, 2006). Human DC Human IL-4 monocyte-derived DC were generated from Ficollseparated PBMC from healthy volunteers. Monocytes were enriched from the leukopheresis according to cellular density and size by elutriation as per manufacturer s recommendations. For DC generation, monocytes were resuspended in serum-free Cellgro DC culture supplemented with GM-CSF and IL-4. Blood myeloid DC (HLA-DR + CD11c + CD123 2 Lin 2 ) were sorted from fresh PBMC using FACSAria (BD Biosciences). Naïve CD4 + and CD8 + T cells (CD45RA + CD45RO 2 ) (purity.99.2%) were purified by FACS-sorting. Immunofluorescence Microscopy For immunofluorescence microscopy, 2610 5 stimulated BMDCs on coverslips were fixed in 3% paraformaldehyde at RT for 15 min, washed twice in PBS 1X and processed for immunofluorescence labelling. To stain NF-kB, mouse BMDCs and BMDMs were permeabilized with PBS 1X 1% saponin (for 10 min at RT) and then saturated with PBS 1X 2% BSA (for 1 h at RT). CD11c (1 in 100), NF-kB subunit p65/reia (1 in 250) and MHC II (1 in 300) were used as primary antibodies. After staining, samples were examined on a Zeiss LSM 510 laser scanning confocal microscope for image acquisition. Images were then assembled using Adobe Photoshop 7.0. Quantifications were done by counting at least 300 cells in 3 independent experiments. Flow Cytometry To analyse mouse BMDC maturation, 2610 5 cells were stimulated and stained with antibodies for classical activation markers. Appropriate isotype antibodies were used as controls. After staining, cells were washed with PBS 2% FCS, then PBS 1X and fixed in 1.5% paraformaldehyde before analysis on a FACScalibur cytometer (Becton Dickinson). Cells were always gated on CD11c for analysis and 100,000 CD11c+ events were collected from each sample. For the intracellular staining of IL-12 and TNF-a in mouse BMDCs, BD Cytofix/Cytoperm and BD Perm/ Wash buffers were used. At least 100.000 events were collected on FACSCanto II (BDBiosciences). For mouse CD4 and CD8 T cell assays, viable cells were analyzed for the decrease of CFSE (proliferation) and the expression of CD25, CD44 and CD62L (diluted in PBS 1X EDTA 2 mm). Human mdc or IL4-DC PLOS ONE www.plosone.org 2 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling activation was analyzed by checking the surface expression of maturation markers CD40, CD83, CD86, HLA-DR after 16 h or 72 h of cell treatment with LPS variants, respectively. Flow cytometry analysis was performed using the FlowJo software. Histograms were drawn from and median fluorescence intensity values were determined on gated populations. At least 100,000 events were collected on FACSCanto II (BDBiosciences) or FACSAria (BDBiosciences). Cytokine Measurement Murine IL-12 and TNF-a were quantified in culture supernatants of stimulated DC by sandwich enzyme-linked immunosorbent assays (ELISA) according to the manufacturer s protocol (Abcys). Human cytokine (IL-6, TNF-a, and IL-12p40) were determined using the BeadLyte cytokine assay kit (Upstate, MA). Immunoblotting 30 mg of cell lysates were subjected to SDS-PAGE PAGE and, after transfer to nitrocellulose, the membrane was probed with a polyclonal antibody against phospho-s6 or S6 (Cell Signaling Technology) or an anti-actin antibody. Blots were subjected to enhanced chemiluminescence detection (ECL, PIERCE). Quantitative RT-PCR Total RNA was isolated with Trisol reagent, was reverse transcribed and analyzed by real-time quantitative PCR using the Power SYBR Green PCR Master Mix (Applied Biosystems). All reactions were performed in triplets. Data were acquired on a 7500 Fast Real-Time PCR system (Applied Biosystems) and were normalized to the expression of actin mrna transcripts in individual samples. For a given real-time qrt-pcr sample, the RNA expression level was calculated from cycle threshold (Ct). In our analysis, given gene expression is shown as mean normalized expression (MNE) relative to the expression of b-actin. The following primers were used for qpcr amplification:rt b-actin (sense): GACGGCCAGGTCATCACTATTG, RT b-actin (antisense): CAAGAAGGAAGGCTGGAAAAGA, p35 sense : 59ctcctgtgggagaagcagac39, p35 anti-sense: 59acagggtgatgggctatctga39, p40 sense:59cacactggaccaaagggact39p40 anti-sensereverse: 59ATTATTCTGCTGCCGTGCT39, TNFa sense: 59CATCTTCTCAAAATTCGAGTGACAA39TNF-a, anti-sense : 59TGGGAGTAGACAAGGTACAACCC39. 3 independent experiments were done and one representative is shown. In vitro Antigen Presentation Assays BMDC (3000 cells) were incubated overnight in 96-well culture plates either with media or OVA. T cells obtained from the lymph nodes and the spleen of OT-I and OT-II Rag-2 2/2 mice were purified with the T cell enrichment kit from Dynal following manufacturer s instructions. For CD4 and CD8 T cell proliferation assays, purified T cells were labeled with 10 mm carboxyfluorescein diacetate succinimidyl ester (CFSE from Invitrogen) for 10 min at 37uC. OT-II and OT-I cells (20000 cells) were added to BMDC that had been stimulated for 8 h with different LPS and then washed. The proliferation of OT-I and OT-II T cells was assessed after 3 days of co-culture by flow cytometry. The cells were washed and stained with anti-cd4 and anti-cd8 antibodies for identification. For CD4 and CD8 T cell activation assays, purified T cells were co-cultured with BMDC previously stimulated for 8 h with different LPS. After 3 days, the expression of surface markers such as CD25, CD44 and CD62L was analyzed by flow cytometry to study the cellular activation level. Co-culture of OT-II T cells with BMDC CD4 + T cells were isolated from the spleen of OT-II Rag-2 2/2 mice using a CD4 + T cell isolation kit (Dynal; Invitrogen). Purity was determined by staining with CD4, CD5, and TCR Va2. A total of 3610 3 BMDC stimulated for 8 h with different LPS were co-cultured with 26 10 4 OT-II Rag-2 2/2 T cells in the presence of ovalbumin, ovalbumin (257 264) peptide (0.06 mg/ml) and of TGF-b (1 ng/ml) as indicated. After 5 days of culture, the expression of Foxp3 and CD25 was evaluated. Human CD4+ and CD8+ T cell Responses 5610 3 blood mdc were co-cultured with CFSE-labeled allogeneic naïve CD4+ T and CD8+ T cells (1 2610 5 ). The DC/T ratio was 1:1000 and 1:20, respectively. Cell proliferation was tested by measuring CFSE-dilution on day 6. On day 7, the production of intracellular cytokines (INF-c, IL-17, IL-13) and Granzyme B were analyzed after 6 h of T cell stimulation by PMA and Ionomycine, in the presence of Brefeldin A. Cells were stained for analysis by flow cytometry using different fluorochromeconjugated antibodies. Phospho-flow Analysis with Fluorescent Cell Barcoding (FCB) Monocyte-derived IL4 DC were generated as previously described. Briefly, human monocyte were enriched with human monocyte enrichment kit without CD16 depletion (Stemcell Technologies, Canada) and suspended in CellGro DC medium (CellGenix, Germany) with GM-CSF and IL-4. On day 6, cells were washed and resuspended at 1 million/ml in RPMI supplemented with 2 mm L-Glutamine, 1 mm Sodium pyruvate, 1X non essential amino acid, 50 mm b-me, and 10 mm HEPES +10% FBS, and then cultured for 2 h in a CO2 incubator. Cells were stimulated with different LPS (100 ng/ml) for 2, 5, 10, 30, 60, and 180 min. Equal amount of medium was used for stimulation control. All samples were immediately fixed by adding PFA (final 1.6%) for 10 min at RT. Fixed cells were centrifuged and washed once with PBS, and then permeabilized with ice-cold Methanol (500 ml/1 million cells) for 10 min at 4uC. Two dimension FCB was performed according to the previous report [11]. Pacific Blue-NHS and Alexa Fluor 488-NHS (Invitrogen, Carlsbad, CA) were added to each condition of cells at 0.02, 0.08, 0.32, 1.0, 3.0 mg/ml or 0.05, 0.2, 0.8, 3.0 mg/ml, respectively. Each sample has a unique combination of dyes with different concentrations. After 30 min on ice, barcoded cells were washed three times with PBS+0.5% BSA and combined into one tube. Combined barcorded cells were stained with Alexa Fluor 647 conjugated phospho-specific antibodies for 30 min at RT. Cells were washed two times with PBS+0.5% BSA. For purified antiphospho-jnk antibody, cells were stained with secondary antirabbit DyLight 649 (Jackson Immunoresearch, West Grove, PA) for 30 min at RT and washed two times. Samples were immediately analyzed with FACS CantoII (BD Biosciences, San Jose, CA). Fold changes of phosphorylation were visualized as a Heatmap. The MFI of LPS-stimulated samples were normalized with medium-stimulated samples. Statistical Analysis All experiments were carried out at least 3 independent times and all the results correspond to the means 6 standard errors. PLOS ONE www.plosone.org 3 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Statistical analysis was done using two-tailed unpaired Student s t test. Significance was defined when P values were,0.05. Results Structural Modifications of LPS Affect Cytokine Secretion by DC We used an array of LPS (Table 1) differing in lipid A acylation to study their activation properties in mouse bone marrow-derived dendritic cells (BMDC) and bone marrow-derived macrophages (BMDM). In addition to the classical wild type hexa-acyl LPS purified from E. coli MLK strain, we used LPS from E. coli MLK mutants (msbb-, htrb- and msbb2/htrb- double mutant) that produce mostly penta-acyl and tetra-acyl lipid A (Table 1) or LPS purified from Y. pestis KIM grown at 37uC (mainly composed of tri- and tetra-acyl lipid A with small amounts of penta-acyl and hexa-acyl molecules, Table 1). All LPS variants induced a BMDC maturation characterized by an up-regulation of the surface expression of major histocompatibility complex MHC-II and costimulatory molecules (CD40, CD86) (Figure 1A). However, significant lower levels of secreted TNF-a and IL-12 were detected in DC stimulated by tetra-acyl LPS purified from E. coli MLK (msbb2/htrb-) double mutants or LPS purified from Y. pestis compared to DC stimulated with wild type E. coli hexa-acyl LPS (Figure 1B). Moreover, the LPS variants did not induce any IFNa secretion (not shown). While comparing the activities of LPS variants, we have also performed a dose-response study (not shown). Cell treatment by 1 ng/ml of LPS triggered DC activation, which reached a plateau at the highest concentration (100 ng/ml). The same differences in terms of cytokine secretion were observed when cells were treated both with 100 ng/ml and 10 ng/ml of different LPS (not shown). Similarly, in BMDM activated by tetra-acyl LPS, TNF-a secretion was strongly decreased compared to BMDM incubated with hexa-acyl LPS (Figure S1) as previously observed in macrophage cell lines [8,9,10]. We then tested the ability of tetra-acyl LPS (referred as purified either from E. coli MLK msbb2/htrb- double mutant or Y. pestis grown at 37uC) to induce human blood myeloid DC (mdc) activation (Figure 1C and D). Hexa-acyl and tetra-acyl LPS induced a similar up-regulation of classical cell surface activation markers (HLA-DR, CD40, CD86, and CD83) (Figure 1C). However, mdc treated with tetra-acyl LPS secreted lower levels of IL-12, IL-6 and TNF-a than those stimulated by hexa-acyl LPS (Figure 1D). Tetra-acyl LPS from Y. pestis, which contains small amounts of hexa-acyl LPS had a stronger capacity to trigger IL-12, IL-6 and TNF-a secretion (p,0.01) than LPS purified from E. coli (msbb-, htrb-) double mutant (devoid of hexa-acyl LPS) (Figure 1D, Table 1). Together, our data show that structural modifications of LPS induce an intermediate phenotype of maturation in mouse and human DC characterized by high levels of MHC-II and costimulatory molecule expression, but low levels of pro-inflammatory cytokine secretion. Tetra-acyl LPS Induce a TLR4-dependent DC Activation LPS recognition by host cells is mediated through the Toll-like receptor 4 (TLR4/MD2/CD14) receptor complex [12]. To determine the contribution of TLR4 in the cell activation induced by LPS with acylation defects, BMDC derived from Tlr4 2/2, Tlr2 2/2 and wild type mice were treated with the LPS variants. No activation was observed in Tlr4 2/2 mice-derived BMDC stimulated either by hexa-acyl or tetra-acyl LPS (p,0.001), as measured by the secretion of TNF-a (Figure S2A). In addition, TLR2 was not implicated in DC activation induced by the different LPS (Figure S2B), showing that LPS preparations were not contaminated by lipoproteins. The measurement of DC viability following treatment with different LPS showed that both hexa-acyl and tetra-acyl LPS induce a very low percentage of dead cells (0.93%) (not shown). We next tried to understand if the decrease of pro-inflammatory cytokine secretion in BMDC activated by tetra-acyl LPS was related to a defect in signal transduction. It has been shown that NF-kB translocation is a key event in LPS-induced TLR4 signalling [13]. Under unstimulated conditions, NF-kB is kept in the cytosol as an inactive form. Under hexa-acyl LPS stimulation NF-kB is translocated into the nucleus where it can bind to several gene promoters [13,14]. After 15 and 30 min of cell stimulation, tetra-acyl LPS induced a significant (p,0.01) stronger NF-kB translocation than hexa-acyl LPS (Figure 2A and B). Similar results were observed in macrophages (Figure S3A and B). Since the activation of the mammalian target of rapamycin (mtor) pathway has been implicated in DC maturation [16], we then analyzed the phosphorylation of the ribosomal protein S6, one of downstream elements of the TLR4 pathway. Compared to hexa-acyl LPS, tetra-acyl LPS induced a stronger S6 phosphorylation at 30 min post-cell activation (Figure 2C). No difference for S6 phosphorylation was observed at later time points either by hexa-acyl or tetra-acyl LPS (Figure 2C). These data show for the first time that LPS with acylation defects induce an early and strong activation of the TLR4-dependent signalling pathway in mouse DC and macrophages. We extended this study to human monocyte-derived IL-4 DC (Figure 3) by using the phospho-flow technology. Fluorescent cell barcoding (FCB) was applied to analyze many conditions simultaneously, using a collection of several anti-phosphorylated proteins [11]. All LPS variants LPS were equally able to increase the phosphorylation levels of several signaling molecules including MAPKs (ERK, p38, JNK), Akt-mTOR pathway molecules (Akt, 4EBP1, S6), and some transcription factors (CREB, NFkB p65) (Figure 3). Interestingly, although the patterns of phosphorylated molecules were same between LPS variants, the kinetics and strength of the phosphorylation changes were slightly different with several molecules (Figure 3). Y. pestis LPS could induce phosphorylation more rapidly, while LPS mutant caused phosphorylation more slowly and weakly than E. coli LPS in some molecules, especially in Akt, p38 and NFkB (Figure 3). These results suggest that as E. coli hexaacyl LPS, Y. pestis LPS and E. coli LPS mutant could act as an agonist to TLR4 pathway. However, structural differences in lipid A region may modify the LPS binding capacity to the receptor, leading to changes in activation potential. It should be also noted, E. coli LPS mutant enhanced tyrosine phosphorylation in STAT1, 3, 5 at later time point more potently than others (Figure 3). Taken together, LPS variants seem to activate the same signaling pathway with different activation potential that may affect the output and quality of immune responses induced by DC. Thus, LPS purified from E. coli MLK (msbb-, htrb-) double mutant and Y. pestis were able to trigger TLR4-dependent signalling in human DC, in agreement with data obtained on mouse BMDC (Figure 2). Altogether these data show that LPS with acylation defects act as agonists to the TLR4 pathway and efficiently induce signal transduction in mouse and human DC. PLOS ONE www.plosone.org 4 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling PLOS ONE www.plosone.org 5 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Figure 1. LPS with acylation defects induce semi-mature mouse and human dendritic cells. Mouse BMDC were stimulated for 8 h (in grey) and 24 h (in black) with medium, E. coli LPS (either hexa-acyl, penta-acyl or tetra-acyl) and Y. pestis tetra-acyl LPS. All LPS were used at the concentration of 100 ng/ml. MHC II and co-stimulatory molecules up-regulation on the cell surface was measured by flow cytometry (A) and cytokine secretion was determined by ELISA (B). Data represent means 6 standard errors of at least 5 independent experiments, **p,0.01, *p = 0.01 to 0.05. Human blood mdc were stimulated overnight with medium (in grey), hexa-acyl E. coli LPS (in red), tetra-acyl E. coli LPS (in blue) and Y. pestis tetra-acyl LPS (in orange). Surface expression of HLA-DR, CD83, CD40 and CD86 was analyzed by flow cytometry (C) and cytokine levels in the culture supernatants were measured by Luminex (D). Experiments were performed on 4 different donors. The data for one representative are shown. ***p,0.001, **p,0.01, *p = 0.01 to 0.05. doi:10.1371/journal.pone.0055117.g001 Tetra-acyl LPS Induce an Early Synthesis of Proinflammatory Cytokines followed by their Proteasomedependent Degradation We then investigated whether the decrease of pro-inflammatory cytokine secretion in BMDC activated by tetra-acyl LPS was due to a defect in cytokine synthesis (transcription/translation). BMDC were activated with different LPS and quantitative RT-PCR used to analyse gene expression. In BMDC treated by tetra-acyl LPS an earlier and stronger transcription of tnf-a, p35 and p40 genes was observed (Figure 4A) compared to BMDC treated by hexa-acyl LPS. Therefore, the decrease of pro-inflammatory cytokine secretion observed in Figure 4B cannot be attributed to transcriptional defects. We next investigated whether the defect in cytokine secretion by DC stimulated with tetra-acyl LPS was due to a change in protein translation (Figure 4C and D). BMDC were incubated with the different LPS in the presence of brefeldin A to block the secretion of newly synthesized cytokines. Intracellular levels of IL-12 and TNF-a were analysed by flow cytometry. LPS with acylation defects induced significant higher TNF-a and IL-12 synthesis at 2 h and 4 h post-stimulation compared to hexa-acyl LPS (Figure 4C and D). However, at 8 h post-stimulation, the level of intracellular cytokines was lower in DC treated with tetra-acyl LPS than in DC treated by hexa-acyl LPS (Figure 4E). It has been shown that glucose or energy deprivation, calcium homeostasis perturbation or elevated synthesis of secretory proteins induce an alteration of the Endoplasmic Reticulum (ER) homeostasis [15,16]. This leads to the disruption of protein folding, the accumulation of unfolded proteins and ER stress response or unfolded protein response (UPR) to restore ER normal function. One of the major components of UPR is the degradation of misfolded proteins by the proteasome (ER associated degradation, ERAD) [15,16]. We therefore determined if the decrease of cytokine secretion observed in DC activated by tetra-acyl LPS could be due to a proteasome-mediated degradation of newly synthesized cytokines (Figure 5). Epoxomycine (Figure 5A) or Mg132 (Figure 5B) proteasome inhibitors were used in BMDC treated by the different LPS for 8 h and intracellular the IL-12 expression was analysed. As expected, in the absence of proteasome inhibitors the level of intracellular IL-12 expression was lower in tetra-acyl LPS-treated DC than in hexa-acyl LPStreated DC. However, in the presence of proteasome inhibitors DC treated with tetra-acyl LPS levels of intracellular IL-12 were similar to those expressed by DC treated with hexa-acyl LPS (Figure 5A and B). We then studied the ubiquitinylation of proteins following DC activation by different LPS. It has been shown that upon inflammatory stimulation, DC accumulate newly synthesized ubiquitinylated proteins in large cytosolic structures. These DC aggresome-like induced structures (DALIS) are transient and require continuous protein synthesis [16]. Mouse DC treated with LPS variants underwent maturation and displayed MHC II surface localization as well as DALIS formation (Figure 5C). However, after 4 h of tetra-acyl LPS treatment, the percentage of DALIS-containing cells was significantly higher as compared to cell stimulated by hexa-acyl LPS (Figure 5C). At 24 h, the number of DALIS decreased, consistent with the transient DALIS expression previously demonstrated in the process of DC maturation (not shown) [16]. These data strongly suggest that tetra-acyl LPS induce a degradation of IL-12 by the proteasome machinery in DC. It is therefore tempting to hypothesize that LPS with acylation defects could induce an ER stress in DC activating the proteasome machinery. This will lead to the down-regulation of cytokine intracellular levels and consequently to a decrease of their secretion. LPS with Acylation Defects Induce Antigen-specific CD8 + and CD4 + T cell Responses We next studied the antigen presentation capacity of tetra-acyl LPS-treated DC and their ability to promote T cell responses (Figure 6). We used transgenic mice that express either a TCR specific for the MHC class-i restricted OVA (OT-I Rag-2 2/2 )or a TCR specific for the MHC class-ii restricted OVA (OT-II Rag- 2 2/2 ). BMDC incubated in either medium alone or medium containing ovalbumin (OVA) were activated by different LPS and co-cultured with OTI (CD8 + ) and OTII (CD4 + ) T cells for 3 days (Figure 6A). Basal level of T cell responses was determined. Table 1. Characteristics of LPS. Bacterial strain (relevant genetic features) a Proportions of lipid A species (molecular mass) E.coli MLK3 E.coli MLK53 (htrb-) E.coli MLK 1067 (msbb-).90% hexaacyl (1823.3 Da); traces of penta and tetraacyl. rough-lps; pentaacyl lipid A deficient in C12 oxyacyl of 3-OH-C14 acyl at GlcN C29 (1615.1 Da) rough-lps;.90% pentaacyl (1587.0 Da); tetraacyl traces E.coli MLK986 (msbb-, htrb-) rough-lps; 29% pentaacyl (1643.0 Da); 54% tetraacyl (1404.8 Da); and 17% triacyl (1178.6 Da) Y. pestis KIM rough-lps, 9% hexaacyl (1797.2 Da); 10% pentaacyl; 40% tetraacyl (1404.8 Da); 7% arabinosamine- tetraacyl (1535.9 Da); 30% triacyl (1178.6 Da) a All are rough-type LPSs. doi:10.1371/journal.pone.0055117.t001 PLOS ONE www.plosone.org 6 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling PLOS ONE www.plosone.org 7 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Figure 2. Tetra-acyl LPS induce the activation of TLR4-dependent molecular pathways involved in mouse DC maturation. BMDC were activated with medium (grey), E. coli hexa-acyl LPS (dark blue), E. coli tetra-acyl LPS (purple) or Y. pestis tetra-acyl LPS (light blue) for 15 min, 30 min, 1 h and 2 h. NF-kB translocation was analyzed by confocal microscopy(a). Cells were fixed and stained for CD11c (in blue), MHC-II (in green) and NF-kB subunit p65/rela (in red). The percentage of BMDC with translocated NF-kB into the nucleus was quantified (B). BMDC were stimulated for 30 min, 1 h, 4 h and 6 h with medium or different LPS. Cell lysates were subjected to SDS-PAGE and, after transfer to nitrocellulose, the membrane was probed with the antibodies against phospho-s6 (Ser235/236), S6 and an anti-actin antibody (C). Data represent means 6 standard errors of at least 4 independent experiments, **p,0.01. doi:10.1371/journal.pone.0055117.g002 BMDC incubated with LPS alone or OVA alone could not induce any T cell response (data not shown). However, BMDC incubated with OVA and activated by different LPS efficiently induced antigen-specific CD8 + and CD4 + T cell responses (Figure 6A). DC activated by tetra-acyl LPS induced a higher OTI and OTII T cell proliferation than cells treated by hexa-acyl LPS (Figure 6A). DC stimulated by tetra-acyl and hexa-acyl LPS were able to trigger T cell activation characterized by a CD25 up-regulation and a CD62L down-regulation. However hexa-acyl LPS-treated BMDC led to a higher down-regulation of CD62L by OT II T cells than those treated with tetra-acyl LPS (Figure 6A). Altogether, these data show that BMDC induced by LPS with acylation defects are able to efficiently promote antigen presentation and induce CD8 + and CD4 + T cell responses. We then investigated the functional properties of human DC stimulated with LPS variants (Figure 6B). Human blood myeloid DC (mdc) activated by the different LPS were able to induce the proliferation of allogeneic naïve CD4 + and CD8 + T cells, although to a lower level for E. coli tetra-acyl LPS compared to other LPS (Figure 6B). Tetra-acyl LPS from Y. pestis, which contains small amounts of hexa-acyl LPS had a stronger capacity to trigger T cell responses than LPS purified from E. coli (msbb-, htrb-) double mutant (devoid of hexa-acyl LPS) (Figure 6B, Table 1). These results show that tetra-acyl LPS-treated DC are able to promote CD4 + and CD8 + T cell responses both in mouse and human models. We then characterized the effector T cells induced by LPStreated mdc (Figure 7). Cells were stimulated with PMA/ Ionomycin and stained for intracellular IFN-c (T H1 response), IL- 13 (T H2 response) and IL-17 (T H17 response). mdc stimulated either by hexa- or tetra-acyl LPS polarized allogeneic naïve CD4 + T cells into IFN-c-expressing T H1 cells (Figure 7A). CD4 + T cells co-cultured with either hexa-acyl LPS-activated mdc or tetraacyl-activated mdc did not express IL-13 or IL-17 (Figure 7A). mdc stimulated by tetra-acyl LPS were also able to induce IFN-c and Granzyme B synthesis in CD8 + T cells (Figure 7B). However, we observed lower levels of IFN-c and Granzyme B production with LPS purified from E. coli MLK (msbb-, htrb-) double mutant compared to other LPS (Figure 7). These data indicate that DC activated by either hexa-acyl or tetra-acyl LPS induce T H1 responses and activate CD8 + T cells. Figure 3. Phospho-flow analysis of human IL-4 DC stimulated by LPS. Human IL-4 DC were activated by different LPS for 2 min, 5 min, 10 min, 30 min, 60 min and 180 min. A phospho-flow analysis using fluorescent cell barcoading was performed in order to assess the phosphorylation levels of molecules involved in TLR4 signaling. The heatmap visualization of phosphorylation changes is shown. The median fluorescent intensity (MFI) of stimulated cells is normalized by MFI of medium stimulated cells. Colored bar on the right shows the levels of fold changes. Experiments were performed on 4 different donors. The data for one representative are shown. doi:10.1371/journal.pone.0055117.g003 PLOS ONE www.plosone.org 8 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Figure 4. Kinetics of synthesis of pro-inflammatory cytokines. (A) BMDC were stimulated for 2 h, 4 h, 8 h or 24 h with medium (grey), E. coli hexa-acyl LPS (dark blue), E. coli tetra-acyl LPS (purple) or Y. pestis tetra-acyl LPS (light blue). Total RNA was purified from cell lysates, reverse transcribed and the amount determined by real-time quantitative PCR. Primers were used for qpcr amplification of actin (control), p35, p40 and TNFa genes. 3 independent experiments were done and one representative is shown, **p,0.01. (B) The secretion levels of IL-12p70, IL-12p40 and TNF- PLOS ONE www.plosone.org 9 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling a were determined by ELISA. Data represent means6standard errors of at least 4 independent experiments, **p,0.01. (C, D) BMDC were treated for 2 h and 4 h with medium, E. coli LPS (either hexa-acyl or tetra-acyl LPS) and Y. pestis tetra-acyl LPS. The intracellular synthesis of IL-12 (p40+p70) in (C) and TNF-a in (D) was analysed by flow cytometry. (E) The intracellular IL-12 and TNF-a production was studied in BMDC activated for 8 h with LPS variants. At least 3 independent experiments were performed and one representative is shown. doi:10.1371/journal.pone.0055117.g004 In Contrast to Murine BMDC, Tetra-acyl LPS Activate Human DC to Induce T reg cells DC with MHC II high, co-stimulation high, pro-inflammatory cytokines low phenotype are referred in the literature as semimature. It has been shown that these cells are able to trigger the differentiation of regulatory T cells (T reg ) [17]. We thus evaluated whether mouse BMDC activated by tetra-acyl LPS displaying a semi-mature phenotype were capable of generating T reg cells characterized by the expression of the transcriptional factor Foxp3 and a high CD25 expression at their cell surface. When maintained on a Rag-2 2/2 background, transgenic mice that express a TCR specific for I-A b -OVA complexes (OT-II Rag-2 2/2 mice) contain only conventional (Foxp3 2 ) CD4 + T cells in their periphery, a situation that facilitates the measurement of their conversion into T reg cells [18]. Such conversion requires I-A b+ DC and the presence of the OVA-derived peptide specifically recognized by OT-II CD4 + T cells (Figure S4). It also depends on the secretion by the antigen-presenting DC of TGF-b [18]. Accordingly, BMDC stimulated with different LPS variants were incubated with OT-II Rag-2 2/2 T cells in the presence of the OVA or OVA 257 264 peptide (0.06 mg/ml), with or without TGFb (Figure S4). We could observe that OVA and peptide-pulsed BMDC were both capable of inducing the activation of OT-II Rag-2 2/2 CD4 + T cells as measured by CD25 expression (Figure S4). However, DC stimulation either by tetra-acyl or hexa-acyl LPS did not trigger T reg responses in mouse BMDC (Figure S4A). The addition of exogenous TGF-b to the culture did not confer to LPS-activated DC the ability to generate T reg cells (Figure S4B). We then studied the capacity of human mdc activated by tetraacyl LPS to induce T reg cells. Human DC activated by LPS variants were co-cultured with allogeneic naïve CD4 + T cells and T reg population was analysed by flow cytometry (Figure 8). We could observe that mdc activated by tetra-acyl LPS induced a higher T reg population characterized by the expression of Foxp3 and a high CD25 expression at the cell surface (Figure 8). This activation profile could be due to the fact that human DC activated by different forms of tetraacyl LPS, including the synthetic Lipid IVa display an intermediate profile of DC maturation (as shown here for IL-4 DC in Figure S5) then leading to T reg proliferation. Discussion The innate immune system possesses various mechanisms to detect and facilitate host responses to microbial components such as LPS [19]. It has been described that each change in chemical composition of LPS causes a dramatic decrease of its activity down to a complete loss of endotoxicity [6]. Different cell types, mainly human and mouse monocytes/macrophages have been used to study LPS structural requirements for its immunostimulatory properties. However, to determine the endotoxic activity of enterobacterial LPS, previous studies have mainly concentrated on cytokine production. Consequently, a decrease in IL-8, IL-6 and TNF-a secretion by cells stimulated with LPS harboring acylation defects has been considered as a lack of immunogenicity or a defect of pro-inflammatory signaling [9,10,20]. In contrast, we show here that LPS with acylation defects efficiently induce a potent activation of TLR4-dependent signaling in mouse and human DC that leads to a strong cytokine synthesis, which in turn triggers the activation of the proteasome machinery. The consequence is the degradation of intracellular pro-inflammatory cytokines and consequently the decrease of their secretion. This hypothesis corroborates previous results, which showed a decrease of cytokine secretion in tetra-acyl LPS-treated macrophages [8,9,10,20]. The difference in the activation potential of LPS variants in terms of cytokine secretion could affect the output of the DC immune response. DC activated by tetra-acyl LPS triggered CD4 + T and CD8 + T cell responses both in mouse and human DC. However, human DC activated by LPS with acylation defects displayed a semi-mature phenotype and induced T reg responses. There could be several mechanisms by which tetra-acyl LPS interact with human DC to elicit distinct types of T H responses. Functional differences between the different subsets of human myeloid DC could be one possible explanation. Two main populations of circulating DC termed myeloid (mdc) and plasmacytoid (pdc) were identified in the blood of healthy donors. Additional distinctions can be made within the mdc subset with CD1c + CD141 2 mdc1, CD1c 2 CD141 + mdc2 and CD16 + mdc [21]. It has been shown that mdc1 and mdc2 differ for the expression of surface markers, cytokine production profile and the differentiation of T H responses. When co-cultured with purified human peripheral blood cells, mdc1 produce IL-12 and favor T H1 differentiation, while mdc2 produce high levels of IL-10 and direct the differentiation of T H2. Moreover, the identification of numerous phenotypic and functional differences among pulmonary mdc1 and mdc2 suggests a possible preferential role for mdc2 in regulating immunity and disease pathogenesis in the respiratory tract distinct from that of mdc1. Distinct roles in host immunity for each human DC were previously shown [21,22,23,24]. For instance, the human CD1c 2 CD141 + mdc2 subset is the functional equivalent of mouse CD8a + DC, capable of cross presentation of exogenous antigens. Regarding their capacity to secrete IL-10, mdc2 might also induce T reg populations. T reg are key players in the immune regulation, particularly in tolerance. This cell population plays a crucial role in suppressing immune responses to self-antigens and in preventing autoimmune diseases [25,26]. Evidence is emerging that T reg can control immune responses to pathogens. They are beneficial to the host through limiting the immunopathology associated with antipathogen immune responses and enabling the development of immune memory. However, pathogens can exploit T reg to subvert the protective immune responses of the host in order to survive and establish a chronic infection [27,28]. Microbes have evolved strategies for programming DC to induce T reg in order to maintain immune homeostasis that controls unbridled host immunity [4,27]. For example, filamentous hemagglutinin (FHA) from the bacteria Bordetella pertusis induces DC to provide IL-10 and prime T reg. Moreover, Yersinia pestis is known to activate DC by means of the dimer of TLR2 and TLR6 to induce T reg [29]. There is growing evidence that the induction of tolerance is not restricted to immature DC. Within the tolerogenic pool of DC, a third population is proposed, called semi-mature [17]. This new subset or developmental stage of DC is distinguished as mature by their surface marker analysis (MHC II high and co-stimulation high ). PLOS ONE www.plosone.org 10 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling PLOS ONE www.plosone.org 11 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Figure 5. Tetra-acyl LPS induce a degradation of IL-12 by the proteasome machinery in DC. BMDC were activated for 8 h with LPS variants in the presence or the absence of proteasome inhibitors such as epoxomycine (A) and Mg132 (B). The intracellular IL-12 (p40+p70) synthesis was then analysed. At least 3 independent experiments were performed and one representative is shown. (C) BMDC were activated for 2 h, 4 h, 8 h and 24 h with LPS variants and labelled with anti-mhc II(green), anti-cd11c (blue) and FK2 (red) antibodies to detect DALIS (white arrows). Quantification of the percentage of DC with DALIS at 2 h, 4 h and 8 h post-incubation with medium or post-stimulation with the different LPS. Quantifications were done by counting at least 300 cells in 3 independent experiments. Data represent means 6 standard errors of at least 3 independent experiments, *p = 0.01 to 0.05. doi:10.1371/journal.pone.0055117.g005 Figure 6. LPS with acylation defects induce functional mouse and human dendritic cells. BMDC were incubated overnight with OVA and activated for 8 h with different LPS. Stimulated DC were co-cultured with T cells from OT-I and OT-II Rag-2 2/2 mice (A). The proliferation of OT-I and OT-II T cells was assessed after 3 days of co-culture by CFSE decrease. For T cell activation assays, the expression of surface markers such as CD25 and CD62L was analyzed by flow cytometry. At least 3 independent experiments were performed and one representative is shown. (B) CFSE-labeled allogeneic naïve CD4 + T and CD8 + T cells were co-cultured with activated mdc for 7 days. Cell division was tested by measuring CFSE-dilution Experiments were performed on 4 different donors. Data for one representative experiment are shown. doi:10.1371/journal.pone.0055117.g006 PLOS ONE www.plosone.org 12 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling PLOS ONE www.plosone.org 13 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling Figure 7. Tetra-acyl LPS induce effector molecules synthesis by human T cells. Human blood mdc were activated overnight either by medium or LPS variants and co-cultured with allogeneic naïve CD4 + T and CD8 + T cells. After 7 days, cells were incubated 6 h with PMA/Ionomycine in the presence of Brefeldin A. The intracellular levels of IFN-c, IL-13 and IL-17 in CD4 + T (A) and IFN-c and Granzyme B in CD8 + T cells (B) were analysed by flow cytometry. Experiments were performed on 4 different donors. Data for one representative experiment are shown. doi:10.1371/journal.pone.0055117.g007 Figure 8. LPS with acylation defects activate human mdc to induce regulatory T cells. Human blood mdc were activated overnight either by medium or different LPS and co-cultured with allogeneic naïve CD4 + T cells. After 7 days, cells were incubated 6 h with PMA/Ionomycine in the presence of Brefeldin A. Foxp3 and CD25 expression was analysed in CD4 + T cell population. Experiments were performed on 4 different donors. Data for 2 representatives are shown. doi:10.1371/journal.pone.0055117.g008 PLOS ONE www.plosone.org 14 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling However, semi-mature DC do not release high level of proinflammatory cytokines, such as IL-1b, IL-6, TNF-a or IL-12p40 or IL-12p70. IL-10 production by semi-mature DC has been described, but it is not an absolute requirement for T reg differentiation [17]. Inducers of DC semi-maturation can be lactobacilli from the gut flora [30], intranasally applied OVA [31], apoptotic cells [32], Bordetella pertussis FHA [33] or TNF-a [34]. Here we show that, structural modifications of LPS are able to induce semi-mature human and mouse DC characterized by MHC-II high, co-stimulation high, pro-inflammatory cytokines low phenotype. In the human model, these semi-mature DC induce high levels of T reg cells. In conclusion, we describe a new mechanism, which regulates the pro-inflammatory cytokine decrease in cells activated by LPS with acylation defects. We propose that cell stimulation by tetraacyl LPS trigger the activation of the proteasome machinery. This leads to the degradation of intracellular pro-inflammatory cytokine levels and consequently to a decrease of their secretion. Our results provide new insights into the understanding of early steps of endotoxin action and suggest that structural modifications of LPS could represent an important strategy for pathogens to subvert adaptive immunity by T reg cell induction in order to survive. Supporting Information Figure S1 LPS structure effect on mouse BMDM activation. Mouse BMDM were incubated with medium, E. coli hexa-acyl LPS (dark blue), E. coli tetra-acyl LPS (purple) or Y. pestis tetra-acyl LPS (light blue). Secretion levels of TNF-a were determined by ELISA after 8 h and 24 h of cell activation. Data represent means 6 standard errors of at least 4 independent experiments. **p,0.01. (EPS) Figure S2 Tetra-acyl LPS induce a TLR4-dependent DC activation. BMDC from wild type and Tlr4 2/2 mice (A) or Tlr2 2/2 mice (B) were stimulated for 8 h and 24 h with medium (grey) or E. coli hexa-acyl LPS (dark blue), E. coli tetra-acyl LPS (purple) or Y. pestis tetra-acyl LPS (light blue) or Pam2CSK4 (brown). TNF-a secretion was measured by ELISA. Data represent means 6 standard errors of at least 3 independent experiments, ***p,0.001, **p,0.01. (EPS) References 1. Kapsenberg ML (2003) Dendritic-cell control of pathogen-driven T-cell polarization. Nature reviews Immunology 3: 984 993. 2. Mellman I, Steinman RM (2001) Dendritic cells: specialized and regulated antigen processing machines. Cell 106: 255 258. 3. Banchereau J, Steinman RM (1998) Dendritic cells and the control of immunity. Nature 392: 245 252. 4. Pulendran B, Tang H, Manicassamy S (2010) Programming dendritic cells to induce T(H)2 and tolerogenic responses. Nature immunology 11: 647 655. 5. Seydel U, Hawkins L, Schromm AB, Heine H, Scheel O, et al. (2003) The generalized endotoxic principle. European journal of immunology 33: 1586 1592. 6. Raetz CR, Reynolds CM, Trent MS, Bishop RE (2007) Lipid A modification systems in gram-negative bacteria. Annual review of biochemistry 76: 295 329. 7. Dixon DR, Darveau RP (2005) Lipopolysaccharide heterogeneity: innate host responses to bacterial modification of lipid a structure. Journal of dental research 84: 584 595. 8. Lapaque N, Takeuchi O, Corrales F, Akira S, Moriyon I, et al. (2006) Differential inductions of TNF-alpha and IGTP, IIGP by structurally diverse classic and non-classic lipopolysaccharides. Cellular microbiology 8: 401 413. 9. Schromm AB, Brandenburg K, Loppnow H, Moran AP, Koch MH, et al. (2000) Biological activities of lipopolysaccharides are determined by the shape of their lipid A portion. European journal of biochemistry/febs 267: 2008 2013. 10. Mueller M, Lindner B, Kusumoto S, Fukase K, Schromm AB, et al. (2004) Aggregates are the biologically active units of endotoxin. The Journal of biological chemistry 279: 26307 26313. Figure S3 LPS effect on mouse NF-kB translocation in mouse BMDM. Mouse BMDM were incubated with medium (grey), E. coli hexa-acyl LPS (dark blue), E. coli tetra-acyl LPS (purple) or Y. pestis tetra-acyl LPS (light blue). (A) NF-kB translocation was analyzed by confocal microscopy in cells activated with different LPS for 15 min, 30 min, 1 h and 2 h. Cells were fixed and stained for NF-kB subunit p65/rela (in red). The percentage of BMDM with translocated NF-kB into the nucleus was quantified (B). Data represent means 6 standard errors of at least 4 independent experiments, **p,0.01. (EPS) Figure S4 BMDC capacity to trigger T reg cell differentiation. BMDC stimulated with different LPS variants were incubated with OT-II Rag-2 2/2 T cells in the presence of the OVA, OVA 257 264 peptide (0.06 mg/ml) with or without TGF-b. After 5 days of culture, T cells were analyzed for the expression of Foxp3 and of CD25. Numbers in outlined areas indicate percentage of cells. Results for hexa-acyl and tetra-acyl E. coli LPS are shown. Data similar to tetra-acyl E. coli LPS are observed while BMDC are stimulated with tetra-acyl Y. pestis LPS. Data are representative of 3 independent experiments. (EPS) Figure S5 Human IL-4 DC stimulation properties in the presence of E. coli LPS analogs and Y. pestis LPS. IL- 4 DC were stimulated for 72 h with medium, hexa-acyl E. coli LPS, tetra-acyl E. coli LPS, synthetic Lipid IVa and Y. pestis at 20 ng/ml. Cell culture supernatants were kept for cytokine measurement (IL-6, IL-10 and TNFa) by Luminex (A). Surface expression of HLA-DR, CD80 and CD86 was analyzed by flow cytometry (B) Experiments were performed on 4 different donors. Data for one representative donor are shown. (TIF) Acknowledgments We thank Dr. Hugues Lelouard for critical advice on the manuscript. Author Contributions Conceived and designed the experiments: JPG AM SO. Performed the experiments: AM YO CD LG. Analyzed the data: JPG AM YO CD SO IM LG. Contributed reagents/materials/analysis tools: IM SO. Wrote the paper: JPG AM SO IM. 11. Krutzik PO, Nolan GP (2006) Fluorescent cell barcoding in flow cytometry allows high-throughput drug screening and signaling profiling. Nature methods 3: 361 368. 12. Ulevitch RJ, Tobias PS (1999) Recognition of gram-negative bacteria and endotoxin by the innate immune system. Current opinion in immunology 11: 19 22. 13. Kawai T, Akira S (2006) TLR signaling. Cell death and differentiation 13: 816 825. 14. Hayden MS, Ghosh S (2008) Shared principles in NF-kappaB signaling. Cell 132: 344 362. 15. Naidoo N (2009) ER and aging-protein folding and the ER stress response. Ageing research reviews 8: 150 159. 16. Pierre P (2009) Immunity and the regulation of protein synthesis: surprising connections. Current opinion in immunology 21: 70 77. 17. Lutz MB, Schuler G (2002) Immature, semi-mature and fully mature dendritic cells: which signals induce tolerance or immunity? Trends in immunology 23: 445 449. 18. Guilliams M, Crozat K, Henri S, Tamoutounour S, Grenot P, et al. (2010) Skindraining lymph nodes contain dermis-derived CD103(2) dendritic cells that constitutively produce retinoic acid and induce Foxp3(+) regulatory T cells. Blood 115: 1958 1968. 19. Kawai T, Akira S (2010) The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nature immunology 11: 373 384. PLOS ONE www.plosone.org 15 February 2013 Volume 8 Issue 2 e55117

Tetraacyl LPS Potentiate Intracellular Signalling 20. Backhed F, Normark S, Schweda EK, Oscarson S, Richter-Dahlfors A (2003) Structural requirements for TLR4-mediated LPS signalling: a biological role for LPS modifications. Microbes and infection/institut Pasteur 5: 1057 1063. 21. Piccioli D, Tavarini S, Borgogni E, Steri V, Nuti S, et al. (2007) Functional specialization of human circulating CD16 and CD1c myeloid dendritic-cell subsets. Blood 109: 5371 5379. 22. Dzionek A, Fuchs A, Schmidt P, Cremer S, Zysk M, et al. (2000) BDCA-2, BDCA-3, and BDCA-4: three markers for distinct subsets of dendritic cells in human peripheral blood. Journal of immunology 165: 6037 6046. 23. Jongbloed SL, Kassianos AJ, McDonald KJ, Clark GJ, Ju X, et al. (2010) Human CD141+ (BDCA-3)+ dendritic cells (DCs) represent a unique myeloid DC subset that cross-presents necrotic cell antigens. The Journal of experimental medicine 207: 1247 1260. 24. MacDonald KP, Munster DJ, Clark GJ, Dzionek A, Schmitz J, et al. (2002) Characterization of human blood dendritic cell subsets. Blood 100: 4512 4520. 25. Vignali DA, Collison LW, Workman CJ (2008) How regulatory T cells work. Nature reviews Immunology 8: 523 532. 26. Mills KH, McGuirk P (2004) Antigen-specific regulatory T cells their induction and role in infection. Seminars in immunology 16: 107 117. 27. Belkaid Y (2007) Regulatory T cells and infection: a dangerous necessity. Nature reviews Immunology 7: 875 888. 28. Josefowicz SZ, Rudensky A (2009) Control of regulatory T cell lineage commitment and maintenance. Immunity 30: 616 625. 29. Depaolo RW, Tang F, Kim I, Han M, Levin N, et al. (2008) Toll-like receptor 6 drives differentiation of tolerogenic dendritic cells and contributes to LcrVmediated plague pathogenesis. Cell host & microbe 4: 350 361. 30. Christensen HR, Frokiaer H, Pestka JJ (2002) Lactobacilli differentially modulate expression of cytokines and maturation surface markers in murine dendritic cells. Journal of immunology 168: 171 178. 31. Akbari O, DeKruyff RH, Umetsu DT (2001) Pulmonary dendritic cells producing IL-10 mediate tolerance induced by respiratory exposure to antigen. Nature immunology 2: 725 731. 32. Menges M, Rossner S, Voigtlander C, Schindler H, Kukutsch NA, et al. (2002) Repetitive injections of dendritic cells matured with tumor necrosis factor alpha induce antigen-specific protection of mice from autoimmunity. The Journal of experimental medicine 195: 15 21. 33. McGuirk P, McCann C, Mills KH (2002) Pathogen-specific T regulatory 1 cells induced in the respiratory tract by a bacterial molecule that stimulates interleukin 10 production by dendritic cells: a novel strategy for evasion of protective T helper type 1 responses by Bordetella pertussis. The Journal of experimental medicine 195: 221 231. 34. Huang FP, Platt N, Wykes M, Major JR, Powell TJ, et al. (2000) A discrete subpopulation of dendritic cells transports apoptotic intestinal epithelial cells to T cell areas of mesenteric lymph nodes. The Journal of experimental medicine 191: 435 444. PLOS ONE www.plosone.org 16 February 2013 Volume 8 Issue 2 e55117