AVANT-PROPOS

Le format de présentation de cette thèse correspond à une recommandation de la spécialité Maladies Infectieuses et Microbiologie, à l’intérieur du Master des Sciences de la

Vie et de la Santé qui dépend de l’Ecole Doctorale des Sciences de la Vie de Marseille.

Le candidat est amené à respecter des règles qui lui sont imposées et qui comportent un format de thèse utilisé dans le Nord de l’Europe et qui permet un meilleur rangement que les thèses traditionnelles. Par ailleurs, la partie introduction et bibliographie est remplacée par une revue envoyée dans un journal afin de permettre une évaluation extérieure de la qualité de la revue et de permettre à l’étudiant de commencer le plus tôt possible une bibliographie exhaustive sur le domaine de cette thèse. Par ailleurs, la thèse est présentée sur article publié, accepté ou soumis associé d’un bref commentaire donnant le sens général du travail. Cette forme de présentation a paru plus en adéquation avec les exigences de la compétition internationale et permet de se concentrer sur des travaux qui bénéficieront d’une diffusion internationale.

Professeur Didier Raoult REMERCIEMENTS

Au Professeur Didier Raoult, qui a accepté de diriger mon Master 2 d’abord, puis ma thèse de Science, pour m’avoir guidé tout au long de mon internat et ma spécialisation en microbiologie clinique, pour son indispensable soutien durant mon « après internat » malgré les difficultés administratives,

Au Professeur Florence Fenollar, pour avoir acceptée de présider le jury de cette thèse, pour m’avoir conseillé, orienté et suivi durant mon internat de biologie,

Aux Professeurs Jean-Philippe Lavigne et Gilbert Greub, pour m’avoir fait l’honneur de bien vouloir être rapporteurs de ma thèse et ainsi d’accepter de juger mon travail,

Au Professeur Michel Drancourt, pour son soutien sans faille, pour son énergie et son enthousiasme communicatif, pour son accessibilité de tous les jours, et sa franchise,

Au Professeur Jean-Christophe Lagier, pour sa disponibilité, ses précieux conseils et sa gentillesse, au Professeur Pierre Edouard Fournier pour votre gentillesse et votre accessibilité, au Professeur Matthieu Million, pour son aide régulière notamment sur les stat’,

À toute l’équipe du CNR des Arbovirus qui m’a chaleureusement accueilli, Christophe, Isabelle, Gilda, Mathilde, Bernard, Patrick, Pierre, Thomas, Laurent, Manon,

À Gregory Dubourg, le « grand frère », sa disponibilité sans faille, son soutien, c’était un plaisir de travailler avec toi et d’apprendre de toi, souvent première victime de mes blagues improbables, haut les cœurs !

À Sophie Baron, brillante pilier du laboratoire, j’espère que tu me pardonneras de t’avoir abandonné ; merci d’avoir toujours été là, même dans les vicissitudes de la vie professionnelle !

À mes collègues de culturomics, les « anciens » (Niokhor « président », Perrine sans qui la culturomics ne serait pas ce qu’elle est, Seck, Togo, Sory, Khoudia, Sokhna, Marion, Elodie, Pamela, Sara, Descartes, Isaac, Camille, Maryam, Gaël, Marcel, Saber, Bruno, Melhem), mais aussi les djeuns (Souad, Edmond, Rémi, Emilie, Sabrina). À Amaël parce

que Freddy c’était le meilleur et le gras c’est la vie. À Fred, le chef d’orchestre de tout ce petit monde.

Aux autres personnes que j’ai eu la chance de croiser durant mon cursus recherche, Lucie Bardet, Florent, Mordav, Erwan, Feriel, Elodie, Nawal, Ghiles, Hortense, Manolis, Maureen, Michel, Richard,

À mes anciens co-internes, PBD, Aurélie la docteure des tout-petits, Sophie Amrane, Manu, PBD, Cléa ma copilote à jamais des Rickettsies, Edouard hodor, Antonin, Raquel, Shirley, Anne Ca, P-A, Armel, Baidy, Emma, Fanny, Florent, James, Emilie, Lindsey, Robin,

À Julien Fromonot, heureusement que tu es là pour m’aider à décrypter le langage 15189 !

À la team biomol, Donia, Julie, Steph’, Emilie, Annick, Laurence, Sophie Edouard, Elsa,

À la team internat, Léa, Estelle, Baptiste, Jean Sé, Chloé, Damien, Jordan, Maude, Florence, Boby,

Aux anciens de Caen Sud, Ugo, Perlemoine, Paul, Sacha, Gros, Arnaud, Marc, Marie, Romain. À Alban et Axel pour notre trio rigolard de la P1,

À Nico, depuis la 4eme, À Alex pour les avions !

Et pour finir, un grand merci à ma compagne Clémence, merci pour ton soutien et ton amour, pour le temps passé à relire ces pages (et d’autres aussi), merci d’être entrée dans ma vie et de l’embellir,

À toute ma famille, Papa, qui m’a donné goût à la médecine et à qui je dédie cette thèse, Maman, qui a toujours cru en moi, mais aussi Amélie et Charles ainsi que leurs conjoints Marc et Satomi, merci pour votre soutien et merci de me rappeler que la famille, c’est le plus important,

À mes neveux, Quentin, Pierre, Héloïse, Adrien et Virgile

À Anne et Eric, pour leur soutien, à Nathan et Nils,

Et à tous ceux que j’ai maladroitement oublié de citer…

SOMMAIRE

Résumé 1

Abstract 3

Introduction 5

Partie I : Le microbiote digestif comme source de nouveaux antibiotiques 8

Publication n°1: Antibiotic discovery: history, methods and perspectives (review) 11

Publication n°2: Antagonisms from gut microbiota against human-pathogens 62

Partie II : Culturomics et incompatibilités de culture 113

Publication n°3: Culture of previously uncultured members of the human gut

microbiota by culturomics 116

Publication n°4: Dynamic variations of anaerobes during non-specific liquid enrichment of human fresh stool sample 152

Conclusion et perspectives 189

Annexes 192

Annexe 1 : Description de nouvelles espèces bactériennes par taxono-genomique 194

Publication n°5: Blautia massiliensis sp. nov., isolated from a fresh human fecal sample and emended description of the genus Blautia 196

Publication n°6: “Intestinimonas massiliensis” sp. nov, a new bacterium isolated from human gut 210

Publication n°7: Noncontiguous finished genome sequence and description of Intestinimonas massiliensis sp. nov strain GD2T, the second

Intestinimonas species cultured from the human gut. 212

Publication n°8: Drancourtella massiliensis gen. nov., sp. nov. isolated from fresh healthy human faecal sample from South France 223

Publication n°9: ‘Bittarella massiliensis’ gen. nov., sp. nov. isolated by culturomics from the gut of a healthy 28-year-old man 232

Publication n°10: Description of Clostridium phoceensis sp. nov., a new species within the genus Clostridium 234

Publication n°11: Description of ‘Gorbachella massiliensis’ gen. nov., sp. nov.,

‘Fenollaria timonensis’ sp. nov., ‘Intestinimonas timonensis’ sp. nov. and

‘Collinsella ihuae’ sp. nov. isolated from healthy fresh stools with culturomics 242

Publication n°12: Fournierella massiliensis gen. nov., sp. nov., a new human associated member of the family Ruminococcaceae 245

Annexe 2: autres publications 252

Publication n°13: A new highly sensitive and specific real-time PCR assay targeting the malate dehydrogenase gene of 3 Kingella kingae and application to 201 pediatric clinical specimens 253

Publication n°14: Acute Septic Arthritis of the Knee Caused by Kingella kingae in a

5-Year-Old Cameroonian Boy 274

Publication n°15: A modified multilocus sequence typing protocol to genotype Kingella kingae from oropharyngeal swabs without bacterial isolation 281

Publication n°16: Molecular tests that target the RTX locus do not distinguish between Kingella kingae and the recently described K. negevensis species 286

Publication n°17: Emergence of Clostridium difficile tcdC variant 078 in

Marseille, France 296

Références 300 RESUMÉ

L’étude du microbiote digestif est actuellement un enjeu important de recherche en microbiologie. Des altérations du microbiote digestif ont été corrélées à divers états pathologiques, comme l’obésité, les maladies inflammatoires chroniques de l’intestin ou encore des pathologies neuropsychiatriques (anxiété, dépression). Les relations entre l’hôte et le microbiote digestif sont extrêmement complexes et font intervenir des procaryotes, des virus, des champignons et des parasites. La description du microbiote intestinal apparaît donc comme la première étape dans la compréhension de la relation hôte – microbiote. Parmi les principales approches, la culturomics a permis une explosion du nombre de bactéries isolées dans le tube digestif, passant de 690 à 1705 en moins de dix ans.

La première partie de cette thèse porte sur la recherche au sein du microbiote digestif de nouvelles molécules antibactériennes. La recherche et le développement de nouvelles molécules antibiotiques sont une des pistes clés dans la lutte contre la résistance aux antibiotiques, sujet majeur de santé publique actuellement. En effet, les trois quarts des antibiotiques actuels sont des produits naturels, ou dérivés de produits naturels, sécrétés par des bactéries ou des champignons de l’environnement et découvert entre 1940 et 1960 par des tests d’antagonisme de culture. Tout comme l’environnement, le microbiote digestif représente également un environnement complexe où les bactéries vivent en compétition.

Dans cette compétition, la synthèse de molécules antibiotiques est un outil pour la survie.

Nous avons recherché des antagonismes de culture parmi les bactéries commensales du tube digestif humain. Nous nous sommes focalisés sur des antagonismes contre les six bactéries pathogènes les plus isolées en microbiologie clinique. Nous avons trouvé une inhibition de croissance de S. aureus par P. avidum, une inhibition de E. cloacae par B. fragilis, E. dispar,

L. delbruckii, P. acidipropionici, S. equinus, S. gallolyticus, et enfin une inhibition de E. aerogenes par B. vulgatus et E. dispar. De plus, des clusters de gène codant des métabolites

1 secondaires ont été trouvées dans le génome de toutes ces bactéries. Ce travail préliminaire confirme que le microbiote digestif est une source potentielle de nouveau antibactériens.

En dépit de l’explosion du nombre d’espèces bactériennes isolées dans le microbiote digestif humain grâce à la culturomics, certaines espèces fastidieuses demeurent peu ou non isolées. Nous avons effectué une analyse métagénomique et culturomics d’une selle fraichement émise avant et après incubation dans un flacon d’hémoculture anaérobie enrichie avec 5% de rumen et 5% de sang de mouton. Ce travail montre que la dynamique de croissance des bactéries anaérobies est très hétérogène. De plus le milieu d’enrichissement utilisé était efficace et permettait la culture d’un plus grand nombre d’espèces bactériennes.

En particulier, nous avons isolé des bactéries non ou peu retrouvées en métagénomique, alors qu’à l’inverse des OTUs présents en grand nombre n’étaient pas retrouvés en culture. Ce travail apporte des éléments nouveaux qui permettront d’optimiser encore le processus de culturomics. En particulier le repiquage précoce des bactéries devrait permettre d’isoler en culturomics des bactéries jusqu’à présent peu ou pas isolées.

Mots clé : culturomics, microbiote digestif, antibiotiques

2 ABSTRACT

Gut microbiota is a major health concern for microbiologists. Alterations of gut microbiota were previously related to diseases such as obesity, inflammatory bowel diseases or neuropsychiatric diseases. Relations between host and gut microbiota are extremely complexes and involve prokaryotes, viruses, fungi and parasites. The description of gut microbiota is the first step in the comprehension of host-microbiota relationship. Among the main approaches for gut studying, culturomics is of concern because it has significantly increased the human gut repertoire from 690 to 1,705 in less than ten years.

In the first step of this thesis, we have searched for new antimicrobials within the gut microbiota. Indeed, antibiotic resistance is a global health concern and research for new antibiotics is a cornerstone for fight against it, according to the WHO. Three quarter of all current antibiotics are natural products, or derived from them, synthesised by bacteria and fungi from soil. They were mainly discovered between 1940 and 1960 using culture antagonism technics. Gut microbiota is another complex ecosystem in which bacteria are living in competition. We have searched for antagonism in the gut microbiota commensals using culture inhibition methods. We focused on antagonisms against six of the most human pathogenic species. We found an inhibition of growth of S. aureus by P. avidum, of E. cloacae by B. fragilis, E. dispar, L. delbruckii, P. acidipropionici, S. equinus, S. gallolyticus, and an inhibition of E. aerogenes by B. vulgatus and E. dispar. Using Antismash, we also found BGCs for all these species. This preliminary work confirm that gut microbiota is a potential source for new antibiotics.

Despite the explosion of the number of bacterial species isolated using culturomics from gut, some fastidious species remains difficult to grow. We performed a metagenomic and culturomics analysis of a fresh stool sample before and after incubation into an anaerobic

3 blood bottle supplemented with sheep blood and rumen fluid. This medium used in culturomics for enrichment was effective, allowing the isolation of higher number of species.

This work show that the dynamic growth of bacteria is very variable. Particularly, we isolated some species that were not or lower found by metagenomic, and at the opposite some species belonging to OTUs highly present were not isolated. This work brings some precisions in the dynamic of bacterial growth that could improve the culturomics process.

Keywords : culturomics, gut microbiota, antibiotics

4 Introduction

Du fait de sa diversité et sa complexité, l’étude approfondie du microbiote digestif, constitué de dizaines de milliards de micro-organismes, est devenue un enjeu majeur de recherche en microbiologie. Les bactéries commensales du tube digestif jouent un rôle clé de régulation physiologique et pathologique chez l’hôte (1). Ainsi, des altérations de la composition du microbiote digestif ont été corrélés à diverses pathologies, comme l’obésité

(2), le syndrome du côlon irritable, les maladies inflammatoires chroniques de l’intestin

(maladie de Crohn, colite ulcéreuse) (3), ou encore des pathologies neuropsychiatriques liées au stress comme l’anxiété, et la dépression (1). Les succès de la transplantation de microbiote fécale dans le traitement de la colite à Clostridium difficile illustrent l’importance du microbiote digestif (4). Les relations entre l’hôte et le microbiote digestif sont extrêmement complexes et font intervenir des procaryotes, des virus, des champignons et des parasites (5).

La description du microbiote intestinal apparaît donc comme la première étape dans la compréhension de la relation hôte – microbiote.

De récentes avancées technologiques ont permis le développement de diverses approches complémentaires d’étude du microbiote digestif faisant appel aux techniques de séquençage haut débit (génomique, transcriptomique, protéomique) mais également de nouvelles méthodes de culture comme la culturomics. En effet les progrès de la biologie moléculaire ont permis la naissance de la métagénomique, qui consiste à séquencer « en masse » les gènes 16S ribosomaux d’un échantillon complexe afin d’identifier les bactéries qui le composent. Néanmoins cette technique présente certaines limites, comme par exemple l’extraction du matériel génétique, mais également la profondeur de séquençage, ainsi que la longueur du gène séquencé (souvent une ou deux régions du gène 16S), qui rendent difficile la comparaison inter-laboratoire des résultats (6). L’approche par culture a été révolutionnée grâce au développement de la spectrométrie de masse par Matrix Assisted Laser Desorption

5 Ionisation - Time of Flight (MALDI-TOF). En effet, l’utilisation du MALDI-TOF rend l’identification bactérienne possible rapidement et à moindre coût (7). Dans ce contexte, l’utilisation de plusieurs conditions de culture (atmosphère, nutriments, enrichissement en milieu liquide) associée à une identification en masse des bactéries par MALDI-TOF a donné naissance à la culturomics (8). La culturomics est une approche de culture novatrice qui consiste à multiplier les conditions de culture, notamment en mimant l’environnement naturel des bactéries dans le but d’en cultiver le plus grand nombre. C’est pourquoi l’essor de la culturomics repose sur le développement de différentes conditions spécifiques afin de cultiver les bactéries les plus fastidieuses, comme par exemple l’utilisation d’anti-oxydants, ou du jus de rumen, coordonnés avec une mise en culture immédiate des prélèvements. La culturomics a permis une explosion du nombre de bactéries isolées dans le microbiote digestif ce qui atteste de son efficacité. Ainsi, 690 espèces bactériennes étaient connues lors du premier projet de culturomics en 2012, alors qu’elles étaient au nombre de 1525 en 2016 (9), et de

1705 en 2018 lors de l’écriture de ce document (données non publiées). Parmi elles, plus de

87% ont été isolées par culturomics et 392 d’entre elles sont de nouvelles espèces.

La description du microbiote digestif constitue une première étape dans la recherche de nouveaux antibiotiques. En effet, on estime la concentration d’un gramme de selle entre

1011 et 1012 bactéries. Tout comme l’environnement, le microbiote digestif représente

également un environnement complexe où ces espèces vivent en compétition. Dans cette compétition, la synthèse de molécules antibiotiques est un outil pour la survie. L’historique de la découverte des antibiotiques réalisé dans la première partie de ce travail (publication n°1) montre qu’en effet la plupart des molécules encore utilisées actuellement sont issues de produits naturels synthétisés par des bactéries ou des champignons de l’environnement. Dans la deuxième publication de cette partie, nous avons recherché parmi les bactéries commensales du tube digestif humain des espèces ayant la capacité d’inhiber la croissance de

6 bactéries pathogènes pour l’homme. Pour cela, nous avons dressé une liste de candidats bactériens antagonistes avec les principales bactéries pathogènes isolées en microbiologie clinique, grâce à l’analyse de 70 métagénomes. Ces candidats ont ensuite été testé par culture ce qui constitue l’originalité de ce travail (publication n°2, en cours de rédaction)

La seconde partie de cette thèse est dédiée à l’étude de la dynamique de croissance des espèces bactériennes lors de la phase initiale du processus de culturomics, qui consiste en un enrichissement non spécifique en milieu liquide. En effet, certaines espèces comme par exemple Faecalibacterium prausnitzii ou Akkermansia muciniphila figurent parmi les espèces les plus abondamment retrouvées en métagénomique (10) alors qu’elles restent difficilement isolées encore par culturomics. Outre leur abondance, ces espèces présentent un intérêt clinique : les données de la littérature suggèrent que F. prausnitzii présente des propriétés anti-inflammatoires (11), et la présence d’A. muciniphila dans les selles de patients est associée à certaines maladies métaboliques, comme le diabète ou l’obésité (12). Dans la publication n°4, nous avons réalisé l’étude métagénomique et culturomics d’une selle fraichement émise et après différentes durées d’incubation dans un milieu d’enrichissement liquide anaérobie non spécifique, associant jus de rumen et sang de mouton, dans le but de caractériser la dynamique de croissance bactérienne et ainsi essayer de comprendre l’origine de ces incompatibilités. Cette approche innovante combine ainsi les deux gold standard de l’étude du microbiote digestif que sont la culturomics et la métagénomique.

7

Partie I :

Le microbiote digestif comme source de nouveaux antibiotiques

8 Avant-propos

La première partie de cette thèse porte sur la recherche au sein du microbiote digestif de nouvelles molécules antibactériennes. La résistance aux antibiotiques est un phénomène mondial qui interpelle de plus en plus les grandes organisations mondiales de santé (13–15).

A l’échelle européenne, on assiste à une augmentation des résistances aux carbapénèmes et aux céphalosporines de troisième génération chez les bactéries à Gram négatives (15). Dans ce contexte, la recherche et le développement de nouvelles molécules antibiotiques sont une des pistes clés selon l’organisation mondiale de la santé (16).

La première publication de cette partie est une revue de la littérature portant sur l’historique et les méthodes de découverte des antibiotiques actuellement utilisés en santé humaine. Les trois quarts des antibiotiques actuels sont des produits naturels, ou dérivés de produits naturels, sécrétés par des bactéries appartenant au phylum des Actinobacteria ou par des champignons. Ces antibiotiques ont été découvert pour la plupart entre 1940 et 1970 grâce

à des tests d’antagonisme de culture comme la méthode de diffusion en puit, des stries croisées, ou encore de « spot-on-lawn » à partir de prélèvements environnementaux. Les micro-organismes vivant dans l’environnement sont en compétition les uns avec les autres depuis des milliers d’années (17). La présence de gènes de résistances aux antibiotiques dans des prélèvements archéologiques démontre la présence de molécules antimicrobiennes dans l’environnement bien avant la présence de l’homme. Dans cette compétition, la synthèse de molécules antibiotiques est un avantage évolutif. Cet âge d’or de découverte des antibiotiques s’est ensuite essoufflé au profit de la recherche de molécules de synthèse. Depuis la commercialisation de la daptomycine en 2003, plus aucune nouvelle classe d’antibiotique n’a

été mise sur le marché.

9 Le microbiote digestif représente un environnement complexe où les bactéries vivent en compétition. De nombreux clusters de gènes de production de métabolites secondaires ont ainsi été trouvés dans le microbiome digestif (18). De nombreuses études sur les probiotiques ont également montré l’importance de la sécrétion de bactériocines par ces bactéries probiotiques (19). La deuxième publication présente les résultats préliminaires des tests d’inhibition de croissance de bactéries pathogènes par des bactéries issues de culturomics.

Les antibiotiques sont des produits naturels, essentiellement des peptides non ribosomaux (NRP) ou des polyketides (PK) codés par des clusters de gènes (BGCs) puis synthétisés par des complexes multi-enzymatiques et qui subissent finalement des modifications post-traductionnelles. La recherche de ces BGCs dans les métagénomes est appelée « genome mining » et est actuellement une approche prometteuse et largement utilisée (20). Cette méthode consiste à rechercher des BGCs dans des génomes ou des métagénomes grâce à l’utilisation d’outils bio-informatiques basés sur la reconnaissance de motifs protéiques et leur comparaison avec les BGCs connus. Bien que des algorithmes de prédictions aient été développés (20), la confirmation par des tests fonctionnels reste compliquée à mettre en œuvre, et limite donc la découverte de nouvelles molécules. Dans le travail présenté dans la publication n°2, nous avons donc directement recherché par culture des antagonismes bactériens au sein du microbiote digestif.

10 Publication n°1

Review: Antibiotic discovery: history, methods and perspectives

Guillaume André Durand, Didier Raoult, Grégory Dubourg

Soumis dans International Journal of Antimicrobial Agents

11 *Highlights

Highlights

 No new class has been discovered since daptomycin in 1986.

 Environmental resistome is ancient.

 Majority of antibiotics come from soil-living organisms, bacteria and fungi.

 Most of antibiotics are secondary metabolites non-ribosomally synthesised.

 Gut microbiota harbours thousands of BGC.

12 *Manuscript Click here to view linked References

1 Antibiotic discovery: history, methods and perspectives

2 Guillaume André Durand, Didier Raoult, Grégory Dubourg *

3 Aix Marseille Univ, IRD, AP-HM, MEPHI, IHU-Méditerranée Infection, Marseille, France

4 * Corresponding author : Dr Grégory Dubourg, MEPHI, Aix Marseille Université, IRD

5 IHU - Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13005 Marseille,

6 [email protected], Phone : (33) 413 73 24 01, Fax : (33) 413 73 24 02

7 Keywords: Antibacterial Agents, drug resistance, Actinobacteria, gastrointestinal

8 microbiome, gut

9 Text words count: 4,730

10 Abstract word count: 238

11 Tables / Figures: 2 Tables, 3 Figures

12 References: 98

13 13 Abstract

14 Antibiotic resistance is considered a major public health issue. Policies recommended by the

15 World Health Organization include research on new antibiotics. No new class was discovered

16 since daptomycin and linezolid in the 1980s and only optimization or combination of already

17 known compounds has been recently commercialized. Antibiotics are natural products of soil-

18 living organisms. Actinobacteria and fungi are the source of about two thirds of the

19 antimicrobial agents currently used in human medicine; they were mainly discovered during

20 the golden age of antibiotic discovery. This era declined after the 1970s due to the difficulty

21 of cultivating fastidious bacterial species under laboratory conditions. Various strategies such

22 as rational drug design have not led to date to the discovery of new antimicrobial agents.

23 However, new promising approaches, e.g. genome mining or CRISPR Cas9 are now being

24 developed. The recent rebirth of culture methods from complex samples has, as a matter of

25 fact, permitted the discovery of teixobactin from a new species isolated from the soil.

26 Recently, many Biosynthetic Gene Clusters were identified from Human-associated

27 microbiota, especially from the gut and oral cavity. As an example, the antimicrobial

28 lugdunine was recently discovered in the oral cavity. The repertoire of human gut microbiota

29 has recently substantially increased, with the discovery of hundreds of new species. The

30 exploration of the repertoire of prokaryotes associated with human beings using genome

31 mining or newer culture approaches could be promising strategies for discovering new classes

32 of antibiotics.

14 33 1. Introduction

34 Antibiotic resistance is considered a major public health concern by several

35 international organizations and by local agencies [1–3]. As a matter of fact, the Centers for

36 Disease Control and Prevention claim 23,000 deaths each year in the US related to antibiotic

37 resistance and some studies predict millions of deaths in the coming decades [4–6]. The

38 United Nations then created a group in order to coordinate the fight against antibiotic

39 resistance [7]. Interestingly, the global mortality related to infectious diseases is decreasing

40 every year, from 10,7 million deaths in 2005 to 8,6 million in 2015 [8]. In addition, it was

41 recently shown that the current mortality due to antibiotic resistance seems far from these

42 predictions [9]. One of the approaches used over the past decade to treat Multi-Drug Resistant

43 (MDR) and Extremely-Drug Resistant (XDR) bacteria was to break the vicious circle of Beta-

44 lactams, enlarging the panel of antimicrobial agents commonly tested. It has been

45 demonstrated that “old” antibiotics (i.e. forgotten molecules) had remarkable efficacy against

46 such isolates. For instance, minocycline, sulfadiazine and clofazimine are active against XDR

47 tuberculosis strains, as for fosfomycin, colistin and minocycline against multi-drug resistant

48 Gram-negative isolates [10,11].

49 According to the World Health Organization, antimicrobial resistance control policies

50 include the rational use of antibiotics, in particular on farms, the increase of surveillance, and

51 the research and development for new tools and molecules [12]. Indeed, despite the rising

52 number of available molecules (Figure 1), the last new class of antibiotic discovered is

53 daptomycin (1986), which was only approved in 2003 by the Food and Drug Administration

54 (FDA) [13–15]. This fact confirms that antimicrobial agents found on the market in the last 30

55 years are associations or improvements of existing molecules. Examples are given of new

56 antibiotics recently marketed that belong to an already known class, such as oxazolidinones

57 (tedizolide), lipoglycopeptides (dalbavancin), or cephalosporins (ceftaroline, ceftobiprole).

15 58 Combination of improved molecules of an already known class is another example of recently

59 commercialized new antibiotics, as for example ceftolozane plus tazobactam, or ceftazidime

60 plus avibactam. The research and development of a totally new class of antibiotic appear as a

61 major issue. Herein, we propose to recall the history of antibiotic discovery, their structural

62 nature, and the methods that were used for their discovery. Finally, we review the potential

63 new approaches for the discovery of new classes of antibiotics.

64 2. The history of antibiotics

65 2.1. Antibiotic resistance is in fact very ancient

66 Most of the antibiotics currently used in human medicine are natural secretions of

67 environmental bacteria or fungi. Indeed, the majority of antibiotics currently used are derived

68 from Streptomyces isolated from soil samples [16]. In their natural environment,

69 microorganisms have to fight against each other by producing antimicrobial substances, and

70 have to develop resistance mechanisms to other antimicrobials [17]. Secondly, the species

71 naturally producing antibacterials also have resistance genes to these antibacterials in order to

72 avoid self-toxicity, within a biosynthetic antibiotic operon [18]. D’Costa et al. demonstrated

73 the presence of antibacterial resistance genes in an environment in which there was no innate

74 antibiotic. They first proposed the existence of a “reservoir of resistance determinants that can

75 be mobilized into the microbial community” [19]. Gerard D. Wright proposed the term

76 “resistome” to design the collection of all the antibiotic resistance genes and their precursors

77 in bacteria [20].

78 Interestingly, multi-drug resistant bacterial species as well as resistance genes to

79 antibiotics currently used were also found from environmental archaeological samples. The

80 OXA genes that encode beta-lactamases have been dated to several million years [21].

81 D’Costa et al. have found resistance genes to β-lactam, tetracyclines and glycopeptides from

16 82 30,000 years old permafrost samples [22]. Kashuba et al. have found several resistance genes

83 in the genome of a Staphylococcus hominis isolated from permafrost [23]. Of the 93 strains

84 cultured by Bhullar et al. from the 4 million-year-old Lechuguilla cave (New Mexico), 65%

85 of the species were resistant in vitro to three or four antibiotics classes [24]. Also, resistance

86 genes to β-lactam and glycopeptides were also found in the 5,300-year-old gut microbiome of

87 the mummy Ötzi [25]. Recently, 177 antibiotic resistance genes belonging to 23 families (that

88 represent all the mechanisms of resistance, i.e. mutation, efflux and antibiotic inactivation)

89 were found in the antibiotic naïve Mackay Glacier region [26].

90 Orthologous genes within mobile elements known from environmental bacteria were

91 also found in bacteria isolated from clinical isolates [19]. For instance, Marshall et al. found

92 orthologous genes of the van HAX cassette from the environmental species Streptomyces

93 toyocaensis and Amycolatopsis orientalis [27]. This cassette is responsible for the

94 glycopeptide resistance of Enterococcus faecium. Some experimental studies seem to show

95 that transfer of resistance genes from environmental producers in the soil to human

96 pathogenic species is possible [28]. The horizontal gene transfer of entire clusters of

97 resistance genes from the resistome to clinical strains under selective pressure related to the

98 human use of antibiotics is actually suspected [29].

99 2.2. Historic of the antibiotic discovery

100 Antibiotics were used for a long time before the advent of modern medicine. The

101 effects of bread on which filamentous fungi grew for the treatment of wounds and burns has

102 been known since ancient Egypt [30]. In the Middle Ages, healers in China and Greece used

103 musty textures to treat various ailments. In the 19th century, Sir John Scott Burden-Sanderson

104 noticed the absence of bacteria from a liquid growth culture covered with mould. In 1871,

105 Joseph Lister discovered the inhibitory effects of Penicillium glaucum on bacterial growth,

17 106 allowing him to cure a nurse’s injury with Penicillium glaucum extracts. At the same time,

107 Louis Pasteur noticed that some bacteria could inhibit others. He discovered with his

108 colleague Jules François Joubert in 1877, while studying the growth of Bacillus anthracis in

109 urine samples, that it was inhibited when co-cultivated with "common" aerobic bacteria. In

110 1889, Jean Paul Vuillemin defined the word “antibiosis” as any biological relationship in

111 which “one living organism kills another to ensure its own existence”. Several antagonisms

112 between microorganisms, notably moulds, were published in the thesis works of Ernest

113 Duchesne in 1897. He discovered the inhibition of Escherichia coli by Penicillium glaucum

114 thirty years before Fleming. Despite several observations of antagonisms between

115 microorganisms, no antimicrobial molecule was purified. The first antimicrobial molecules

116 discovered were chemical compounds. In 1909, Paul Ehrlich discovered the arsphenamine, an

117 arsenic derivative active against Treponema pallidum, the agent of syphilis. This antibiotic

118 was commercialized in 1911 under the name Salvarsan® then Mapharsen®. In 1930, Gerhard

119 Domagk discovered the antibiotic effects of sulphanilamide, a molecule synthetized 22 years

120 before by Paul Gelmo. This antibiotic was marketed under the name Protonsil® in 1935 and

121 was used by soldiers during World War II [32].

122 In 1928, Alexander Fleming accidentally discovered in his forgotten colonies of

123 Staphylococcus aureus that a fungus was inhibiting the growth of Staphylococcus. The

124 Penicillium notatum molecule has been purified and called penicillin. But the industrial

125 production of this antibiotic was performed only in 1940 by Howard Florey et Ernst Chain,

126 using Penicillium chrysogenum [33]. Fleming also discovered the lysozyme, an antibacterial

127 enzyme [33]. In 1930, René Dubos discovered an enzyme from a soil-derived Bacillus that

128 specifically decomposes Streptococcus pneumoniae type III capsular polysaccharide. With

129 this enzyme, he was able to treat mice with pneumococcal peritonitis [34]. Ten years later, he

130 isolated from Bacillus brevis the oligopeptide gramicidin that widely inhibited Gram-positive

18 131 species [35]. Unfortunately, gramicidin showed too much toxicity for humans, except for

132 local treatment [36]. In the US, Selman Waksman was the first to perform a systematic

133 research of antimicrobial activity of soil bacteria, particularly from Streptomyces members or

134 Streptomycetes. He developed several culture techniques and strategies (“Waksman

135 platform”) in order to highlight antagonisms between bacterial species [37]. Using his

136 platform, he discovered in the 40’s several major antibiotics and antifungals, such as

137 actinomycin (from Streptomyces spp.) [38], streptomycin (from Streptomyces griseus) [39],

138 neomycin (from Streptomyces fradiae) [40], fumigacin (from Aspergillus fumigatus) and

139 clavacin (from Aspergillus clavatus) [41]. Actinomycin, neomycin and streptomycin are still

140 in use today. Moreover, streptomycin has revolutionized the treatment of tuberculosis, and is

141 still active against multi-drug resistant tuberculosis [42]. The pharmaceutical industry was

142 inspired by the Waksman platform, which led to the discovery of all current antibiotics

143 between the 1940s and the 1970s. During this golden age, 23 classes of antibiotic were

144 discovered from 19 bacterial species and 7 fungi (Table 1).

145 Despite recent commercialization for some, the last classes of antibiotic discovered are

146 from the 1980s. After 50 years of discoveries, no new classes have been found. Therefore,

147 new strategies were needed. After the culture approach through Waksman’s platform, the

148 industry turned to the in vitro synthetization of new molecules, based on knowledge of the

149 known mechanism of action of antibiotics. Unfortunately, few new classes of antibiotics have

150 been discovered, the nitrofuran in 1953, the quinolone in 1960, the sulphonamide in 1961 and

151 oxazolidinones in 1987. The modification and improvement of already known molecules has

152 also been carried out. This is reflected by the commercialization of linezolid in 2003 and

153 daptomycin in 2001, although these molecules had been known since 1955 and 1986,

154 respectively [32]. Recently, a new cephalosporin called cefiderocol was found to be active

155 against carbapenem-resistant Gram-negative bacteria [43]. Hemisynthetic compounds from

19 156 natural products were also developed, such as ketolides (derived from macrolides) or

157 metronidazole (derived from a natural product of Streptomyces sp.) (Table 1). But the lack of

158 return on investment and the emergence of resistance have led the industry to gradually let

159 down the research on antibiotics, preferring to invest in drugs for chronic diseases [32]. Of the

160 20 pharmaceutical companies that invested in antibiotic discovery in the 1980s, there were

161 only five left by 2015. [44]. More than 1,200 antimicrobial peptides were discovered from

162 various origins, from plants to invertebrates and animals, but none has been used as an

163 antibiotic [45]. In conclusion, the majority of antibiotics were discovered during the golden

164 age. Bacteria and fungi were the greatest producers. The genus Streptomyces is the source of

165 about half of the antimicrobial agents currently used in human medicine (Table 1).

166 3. Chemical nature of antimicrobial agents

167 Antimicrobial molecules are represented by a wide variety of chemical compounds.

168 Most of the time, they are natural products and secondary metabolites, implying that they are

169 not required for survival under laboratory conditions but still provide some advantages in the

170 environment [46]. Among the antimicrobial substances used in human medicine, it is possible

171 to classify antibiotics in five groups of chemical molecules. The first are derived amino acids

172 that are non-ribosomally synthesized (non-ribosomally synthetized peptides, NRP). The

173 second are acetyl coenzyme A or malonyl coenzyme A derived (polyketides, PK). NRP and

174 PK represent about 50% of all current antibiotics (Table 1). The third are hybrids between

175 NRP and PK and the fourth are composed of several carbohydrates units substituted with

176 amine groups (aminoglycosides). And finally, the last group is composed of various

177 molecules, such as terpenoids, fusidic acid or alkaloids such as metronidazole. In addition to

178 these molecules used as antibiotics, thousands of antimicrobial peptides are known from

179 insects, mammals, plants or amphibians [47]. These peptides are usually classified as

20 180 “ribosomally synthesized and post-translationally modified peptides” (RiPPs), a subgroup of

181 natural products [48].

182 3.1. Antibiotics

183 The NRP and PK are synthetized by multi-enzymatic complexes (non-ribosomal

184 peptides synthetase, NRPS and polyketides synthetase, PKS) encoded by biosynthetic gene

185 clusters [49]. These complexes are organized in modules including several domains. For

186 example, the erythromycin synthase complex consists of three proteins and seven modules

187 each containing three to six domains. [49]. The NRPS use amino acids as substrate elongated

188 by peptidic connection, whereas PKS use acyl-coenzyme A as substrate elongated by Claisen

189 condensation reaction [49]. This complex organization allows the production of a great

190 number of different products, and therefore a great diversity. For example, ciclosporin

191 belongs to the NRP category [50]. Hybrid assembly line that uses both amino acids and acetyl

192 coenzyme A, or fatty acid synthetases (FAS), are also described [51]. NRP antibiotics

193 comprise molecules such as β-lactams, daptomycin, lincomycin, polymyxins, and

194 vancomycin while macrolides, mupirocin and tetracyclines belong to polyketides antibiotics.

195 Rifampicin belongs to hybrid NRP/PK (Table 1).

196 3.2. Antimicrobial peptides

197 Antimicrobial peptides are broad spectrum antibacterial molecules that have been

198 discovered in blood cells in 1957 by Robert Skarnes [52]. They belong to natural products

199 that are small peptides (i.e. less than 100 amino acids) ribosomally synthesized with post-

200 translational modifications (RiPPs). Natural peptides have a multifunctional activity that

201 participates in innate immunity in eukaryotic and prokaryotic cells. [47]. Antimicrobial

202 peptides are evolutionary well-conserved amphipathic molecules with hydrophobic and

203 cationic amino-acids [53]. They can be categorized according to their secondary

21 204 conformation: group I (alpha helical), group II (beta sheet), group III (mixed) and IV

205 (extended) [54]. They have been isolated from almost all living organisms from prokaryotes

206 to vertebrates. Indeed, thousands of antimicrobial peptides are known and come from insects,

207 plants, or amphibians [45,47]. The main known mechanisms of action are related to their

208 cationic charge and amphipathic structure, responsible for the membrane disruption of the

209 negatively charged bacterial cell. Antimicrobial peptides can also interact with membrane-

210 associated protein targets, as well as intracellular targets after penetration into the bacterial

211 cytoplasm [54]. Finally, antimicrobial peptides can also have immunomodulatory effects on

212 the innate immune system of the host [55].

213 Bacteriocins are antimicrobial peptides that were first discovered in bacterial species.

214 They are used in the agro-alimentary industry as food preservatives and in veterinary

215 medicine [56]. Bacteriocins are separated into four groups: class I are small heat-stable post-

216 translationally modified peptides (<5kDa) which use the amino acid lanthionine, there are

217 therefore called lantibiotics; class II are not modified heat-stable small peptides (<10kDa)

218 which do not use lanthionine; class III are large heat-labile peptides (>30kDa) and class IV

219 are complex or cyclic peptides containing lipids or carbohydrates [56]. Bacteriocins inhibited

220 closely related species [57]. For example, lacticin 3147 and nisin are lantibiotics that

221 exhibited an antibiotic activity against Gram-positive bacteria, notably Meticillin-resistant

222 Staphylococcus aureus (MRSA) or Vancomycin-resistant Enterococcus [58]. Nisin is the

223 most famous lantibiotic, largely used as food preservative because of its activity against

224 Listeria monocytogenes, MRSA and Streptococcus pneumoniae [59]. Nisin is also used in

225 veterinary medicine under the name Wipe Out ® for the prevention of dairy mastitis [57]. The

226 synergistic effects of lantibiotics with antibiotics have also been demonstrated in vitro [60].

227 According to the Antimicrobial Peptide database, 2,478 antibacterial peptides are

228 already known [61]. All medical fields combined, 60 peptides have obtained the FDA

22 229 approval and 140 are in clinical trials, mainly for oncological and metabolic diseases [62].

230 Yet, when focusing on infectious diseases, only 12 naturally occurring antimicrobial peptides

231 have reached the stage of human clinical trial. (Table 2). Among them, only three (LYX-109,

232 LL-37 and nisin) demonstrated a better efficacy than placebo, but none obtained the FDA

233 approval for human application. Only gramicidin has received FDA approval for topical

234 application in association with polymyxin and neomycin for ophthalmic use. Teixobactin is a

235 new promising depsipeptide that provides an inhibition of S. aureus and M. tuberculosis [63].

236 This molecule uses notably the rare amino acid L-allo-enduracididine, which is challenging to

237 synthesize and therefore limits its use despite efforts made to develop analogues [64]. Despite

238 their narrow or broad spectrum of activity against human pathogens and easier bioengineering

239 compared to NRPS or PKS, bacteriocins are not used in humans as antibiotics due to several

240 limitations.

241 The main limitations in the use of antimicrobial peptides as antibiotics in clinical

242 practice are their instability (proteolytic digestion, oxidation), their high cost and low yield of

243 production, their short half-life, and their quick elimination [60]. Notably, their low

244 bioavailability after oral ingestion related to proteolytic degradation is a great bottleneck. To

245 solve these problems, production of analogues using rational drug design or nano-engineering

246 is used to improve the pharmacokinetic properties. For instance, nano-engineering has

247 increased the nisin spectrum to Gram-negative bacterial species [65] or allowed HPA3PHIS to

248 be highly effective against Vibrio vulnificus in a mouse model [66]. The use of nanoparticles

249 also prolonged the stay in the stomach of pexiganan after oral administration, which reduces

250 the concentration of pexiganan required in a mouse model for Helicobacter pylori eradication

251 [67]. Another approach is the combination of bacteriocins with other antimicrobials in order

252 to reduce the resistance risk and to increase the antimicrobial potency [60]. Another matter of

253 concern with AMP is the risk of development of resistance against our own immunity

23 254 peptides. For instance, pexiganan have been previously found to induce cross-resistance to

255 human-neutrophil-defensine 1 [68].

256 4. Methods for the discovery of antibiotics

257 4.1. Culture-based approaches

258 In the 1940s, Selman Waksman systematically screened the soil bacteria for

259 antagonisms. Its culture-based approach is still in use today [69]. All the methods are based

260 on the same principle: showing an inhibition of a tested strain over an indicator strain closely

261 cultivated. The tested strain is the strain suspected to produce antimicrobials targeting the

262 strain used as indicator. Several techniques exist to detect antimicrobial activity, either in

263 solid or in liquid culture. There are three main methods regarding solid culture approaches:

264 the cross-streak method, the spot-on-the-lawn and the well diffusion method (Figure 2).

265 The cross streak is the inoculation of the bacterial strain tested vertically on the agar

266 plate. The incubation time of the plate depends on the life cycle of the bacterial strain required

267 to reach the exponential phase, which is the moment where secondary metabolites are

268 excreted. Then, the indicator strain is inoculated into horizontal streaks and the plate is

269 incubated (Figure 2) [70]. This technique is easy and powerful for screening but requires that

270 both bacterial strains have the same culture conditions (e.g., atmosphere, temperature and

271 growth duration).

272 The second and the third methods are respectively called “spot-on-the-lawn” and “well

273 diffusion method”. The “spot-on-the-lawn” method consists in depositing a drop of the tested

274 strain on a lawn of the indicator strain [71]. After incubation, an inhibition zone is searched

275 around the sediment. The “well diffusion method” is based on the diffusion of antimicrobials

276 through agar which inhibits sensitive species. An agar plate is pooled with the indicator strain

277 or inoculated with a lawn of the indicator and agar holes are punched out aseptically. Then,

24 278 two main variants exist. The first is the Agar Plug Diffusion method which consists in

279 removing a cylinder of agar from a plate previously inoculated with the tested strain. This

280 cylinder of agar is then placed into the hole of the indicator plate [72]. The second variant

281 method consists in placing a liquid broth of the tested strain or a growth supernatant in the

282 hole [73]. After an optional rest time at +4°C, the agar plate is incubated and inhibition

283 growth zones are measured (Figure 2). Several variants were developed, like using stress

284 conditions or iron chelator [74,75]. The main limitation of solid culture tests concerns

285 bacterial species that have different growth conditions or fastidious species.

286 Liquid culture-based approaches can solve this problem. Liquid broth co-culture has

287 been used since the existence of the Waksman platform [69]. This is the simultaneous culture

288 of the tested species and the indicator strain separated by a filter allowing the diffusion of

289 nutrients but not the diffusion of cells. Following incubation, the bacterial growth of the

290 indicator strain is determined by numeration, coloration or optical density [69]. Another

291 method is to add the growth supernatant of the test species previously filtered and

292 concentrated to a liquid culture of the indicator strain [76]. The latter method allows the

293 bacteria tested to be grown under conditions different from those of the indicator strain

294 (Figure 3).

295 4.2. Discovering antimicrobial effects from already known compounds

296 Some antibiotics were discovered years before they were used (fidaxomycine,

297 daptomycin, or linezolid). Several million of chemical compounds are known in chemical

298 databases and could provide a potential source of antibiotics [32]. Moy et al. have tested the

299 activity against E. faecalis of more than 6,000 chemical compounds and 1,136 natural

300 products in the Caenorhabditis elegans animal model, discovering 16 molecules increasing

25 301 the survival of animals [77]. The bottleneck remains the selection and the high throughput

302 testing of these molecules.

303 4.3. Synthesis of new molecules and improvement of already known compounds

304 The « Rational Drug Design » consists in the empirical synthesis of new molecules

305 that are designed according to several rules to be well-absorbed, non-toxic, and active against

306 a specific target [32]. The most famous rules used by the industry are Lipinsks’s rules.

307 Despite more than 10 million of new molecules synthesized, only a few active molecules have

308 reached the market, notably antituberculous drugs [32,37] (Table 1). This can be explained by

309 the fact that antibiotics have generally poor economics [78].

310 The improvement of already known molecules is another strategy that can yield

311 benefits. The modification of cephalosporin lead to the development of cefiderocol, that

312 demonstrated safety and tolerability in healthy subjects, and clinical trials for the treatment of

313 urinary tract infections are ongoing [79]. Another example is the modification of the

314 aminoglycoside sisomicin that leads to the development of plazomicin [80]. This antibiotic

315 was recently approved by the FDA and a small phase 2 clinical trial found it had an efficacy

316 comparable to that of levofloxacin in the treatment of urinary tract infections and acute

317 pyelonephritis [81]. If the potential use of these antibiotics remains to determine, their

318 mechanism of action is not novel, and the apparition of resistance is expected in the same way

319 as for their related antibiotic parent.

320 4.4.Genome mining

321 Secondary metabolites are encoded by Biosynthetic Gene Cluster (BGCs). Thousands

322 of prokaryotic genomes are available in sequence databases. These data have generated

323 thousands of BGCs that potentially encode unknown molecules [82,83]. Despite the fact that

324 many of them do not have any antimicrobial activity, little is known about them. Walsh et al.

26 325 found 74 putative BGC from 59 genomes from the Human Microbiome Project [84]. These

326 BGC belonged mainly to Firmicutes, Proteobacteria and Bacteroidetes and the most

327 commonly putative bacteriocins encoded belonged to classes III and IV.

328 Several approaches of genome mining are possible. The most used are sequence-based

329 approach, ecology-based genome mining, mode-of-action-based genome mining, or function-

330 based genome mining [83,85]. For instance, lichenicidin is a bacteriocin synthesized by

331 Bacillus licheniformis that was discovered using the mode-of-action genome mining

332 approach. The authors screened the databases for LanM genes, which are involved in the

333 biosynthesis of lantibiotics [86]. The identification of putative bacteriocins encoded by BGC

334 from the genome sequence is possible using bioinformatic algorithms [87]. Bacteriocins are

335 easily found using bioinformatic tools compared to NRPS or PKS. Tools such as BAGEL,

336 antiSMASH or PRISM are widely used for this purpose. These tools exploit two main

337 approaches [83]. The first consists in finding new congeners of already known scaffolds. This

338 approach is based on the homology comparison by the research of conserved domains

339 (anchors), as for the thiotemplate domain of NRPS and PKS [88]. Small structural changes of

340 the new homologue may result in a significant change in the activity of the product. The

341 second approach is more difficult and consists in finding new scaffolds. Predicting the

342 chemical structure and biological activity of a BGC informatic sequence is a real challenge

343 [83]. The main problem that remains is proving the functional activity of BGC [89]. Both

344 approaches often require the engineering expression of BGC from the native host or from a

345 heterologous host, which represents the main bottleneck of genome mining for the discovery

346 of new antibiotics.

347 Recently, Hover et al. have screened more than 2,000 soil samples from various areas

348 of the US, searching for BGCs encoding calcium-binding motif Asp-X-Asp-Gly [90]. This

349 motif is related to calcium-dependant antibiotics such as lipopeptides, for which the

27 350 mechanism of action is not fully understood [91]. They found several clades of

351 uncharacterized BGC, of which the most abundant was present in 19% of all samples. This

352 clade was named malacidins, the molecules were then synthesised, and their antibiotic activity

353 was characterized in vitro. The malacidin A exhibited a broadly antibiotic activity against

354 Gram-positive bacteria, notably MRSA, which was successfully confirmed in a skin-wound

355 infection rat model. In conclusion, genome mining is a promising approach for new antibiotic

356 discovery, despite the fact that the method is fastidious and time-consuming.

357 4.5.CRISPR Cas9

358 Bacteria and fungi have an immune system that protects them from foreign genetic

359 material that could be inserted by phages. This immune system consists of restriction

360 enzymes, toxin-antitoxin and Clustered Regularly Interspaced Short Palindromic Repeats

361 (CRISPR) Cas system [92]. The utilization of CRISPR Cas9 to design new antimicrobials

362 with a predetermined activity spectrum has been already performed with promising results.

363 Citorik et al. have developed RNA-guided nucleases that target the resistance gene blaSHV-18

364 and blaNDM-1. After transformation by plasmids and transduction by bacteriophages, the

365 authors observed a significant reduction in the number of E. coli containing the targeted

366 resistance gene, either chromosomal or plasmid [93]. The same approach has been

367 successfully tested using phagemid targeting the S. aureus methicillin resistance gene [94].

368 The same authors tested the use of bacteriophages in a mouse model of S. aureus skin

369 infection with an efficacy comparable to that of mupirocin [94]. Other authors used genome

370 editing technology to re-sensitize MDR cells. As an example, Kim et al. used CRISPR Cas9

371 to target a conserved sequence of Extended-Spectrum β-Lactamases, thus restoring the

372 susceptibility of E.coli in their in vitro model [95].

373 5. Conclusion and perspective

28 374 More than three quarters of all antibiotics currently used in human health are natural

375 products or derived from them. The discovery of antibiotics declined after the 1970s due to

376 the difficulty of cultivating bacterial species from soil under laboratory conditions. New

377 innovative culture approaches were then created thanks to the bloom of new molecular

378 methods. In this way, genome mining searching for new BGCs as well as CRISPR Cas9

379 technology are promising new approaches. Recently, the ability to rapidly identify bacterial

380 strains using Matrix Assisted Laser Desorption Ionisation Time of Flight (MALDI Tof) has

381 permitted the rebirth of culture. Therefore, new culture approaches trying to mimic the natural

382 environment were invented in order to grow fastidious species. This led to the discovery of

383 new bacterial species. For instance, Ling et al. discovered the new antibiotic teixobactin from

384 the new species Eleftheria terrae isolated in the soil using a diffusion chamber [63].

385 The same approach is now possible for the human-associated microbiota. The nose is

386 an example of ecological niche poor in nutrients in which the microbiota is probably in strong

387 competition [75]. Using a home-made nasal synthetic medium under iron-limited condition,

388 Krismer B et al. discovered lugdunin, a new antibiotic inhibiting S. aureus growth [74]. The

389 gut microbiota is another microbiota of interest for antibiotic research. Indeed, the human gut

390 has an average load concentration ranging from 104 to 1012 CFU/mL from the duodenum to

391 the colon. These species live in extreme competition, as they did before human colonization

392 where bacteria lived in a competitive world that led them to naturally develop many

393 antimicrobial products. Metagenomic analysis from the human gut microbiota found many

394 BGCs. In 2014, Donia et al. found 3,118 BGCs including NRPS, RiPPs and PKs in the

395 human microbiome. They also found 599 BGCs in the gut. Taken together with the oral

396 cavity, it is one of the richest sites in BGCs of the human microbiota. They also found and

397 purified lactocillin, a new thiopeptide antibiotic, isolated from the vaginal microbiota [96].

29 398 The study of the gut microbiota using culture methods was recently improved by

399 culturomics, a novel approach that consists in the multiplication of growth conditions [97].

400 This has led to the discovery of previously uncultivated species. With this approach, Lagier et

401 al. significantly increased in a couple of years the gut repertoire from 690 to 1,525 species, of

402 which 247 were totally new [98]. Indeed, the new species described from the gut represent an

403 opportunity for the research for novel antibiotics. The search for new antibiotics naturally

404 synthesized by organisms living in complex ecosystems like the gut microbiota, using the

405 culture approach, seems the modern continuity of what has already worked in the past. Thus,

406 if the study of antagonisms between environmental bacteria has led to the discovery of a

407 substantial proportion of antibiotic classes, such studies were rarely performed from human-

408 derived microbiota.

409 In conclusion, the research for new antibiotic molecules is a key point among the

410 strategies for the fight against antibiotic resistance. The recent advances from both culture-

411 dependant and culture-independent methods of exploration of complex ecosystems such as

412 soil or human-associated microbiota open a new era in antimicrobial research.

30 413 Acknowledgements and financial disclosure

414 This work was supported by the French Government under the « Investissements d’avenir »

415 (Investments for the Future) a program managed by the Agence Nationale de la Recherche

416 (ANR, fr: National Agency for Research), (reference: Méditerranée Infection 10-IAHU-03).

417 The authors thank Magdalen Lardière for reviewing the English language of the manuscript.

31 418 References

419 [1] World Health Organization. Antimicrobial resistance, Factsheet 194 2016.

420 http://www.who.int/mediacentre/factsheets/fs194/en/ (accessed October 30, 2018).

421 [2] Centers for Disease Control and Prevention. Antibiotic / Antimicrobial Resistance. Biggest

422 Threats. https://www.cdc.gov/drugresistance/biggest_threats.html (accessed October 30,

423 2018).

424 [3] European Centre for Disease Prevention and Control. Antimicrobial resistance surveillance in

425 Europe. Annual report of the European Antimicrobial Resistance Surveillance Network (EARS-

426 Net). 2015. 2017.

427 [4] Centers for Disease Control and Prevention. About antimicrobial resistance 2018.

428 https://www.cdc.gov/drugresistance/about.html (accessed October 30, 2018).

429 [5] European Society of Clinical Microbiology and Infectious Diseases. ESCMID warns: Europe may

430 surpass one million deaths due to ineffective antibiotics by 2025 2015.

431 http://semmelweis.info/escmid-warns-europe-may-surpass-one-million-deaths-due-to-

432 ineffective-antibiotics-by-2025/ (accessed October 30, 2018).

433 [6] O’Neill J. Atiioial esistae: taklig a isis fo the health ad ealth of atios .

434 https://amr-review.org/sites/default/files/AMR%20Review%20Paper%20-

435 %20Tackling%20a%20crisis%20for%20the%20health%20and%20wealth%20of%20nations_1.pd

436 f (accessed October 30, 2018).

437 [7] United Nations. UN announces interagency group to coordinate global fight against

438 antimicrobial resistance n.d.

439 http://www.un.org/apps/news/story.asp?NewsId=56365#.WYlavRXyjIV (accessed October 30,

440 2018).

441 [8] GBD 2015 Mortality and Causes of Death Collaborators. Global, regional, and national life

442 expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-

32 443 2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Lond Engl

444 2016;388:1459–544. doi:10.1016/S0140-6736(16)31012-1.

445 [9] Abat C, Rolain J-M, Dubourg G, Fournier P-E, Chaudet H, Raoult D. Evaluating the Clinical

446 Burden and Mortality Attributable to Antibiotic Resistance: The Disparity of Empirical Data and

447 Simple Model Estimations. Clin Infect Dis 2017;65:S58–63. doi:10.1093/cid/cix346.

448 [10] Kaye KS, Gales AC, Dubourg G. Old antibiotics for multidrug-resistant pathogens: from in vitro

449 activity to clinical outcomes. Int J Antimicrob Agents 2017;49:542–8.

450 doi:10.1016/j.ijantimicag.2016.11.020.

451 [11] Brouqui P, Quenard F, Drancourt M. Old antibiotics for emerging multidrug-

452 resistant/extensively drug-resistant tuberculosis (MDR/XDR-TB). Int J Antimicrob Agents

453 2017;49:554–7. doi:10.1016/j.ijantimicag.2017.02.008.

454 [12] World Health Organization. World Health Day 2011: policy briefs 2011.

455 http://www.who.int/world-health-day/2011/policybriefs/en/ (accessed October 30, 2018).

456 [13] Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, et al. Bad bugs, no drugs: no

457 ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect Dis Off Publ

458 Infect Dis Soc Am 2009;48:1–12. doi:10.1086/595011.

459 [14] Slee AM, Wuonola MA, McRipley RJ, Zajac I, Zawada MJ, Bartholomew PT, et al. Oxazolidinones,

460 a new class of synthetic antibacterial agents: in vitro and in vivo activities of DuP 105 and DuP

461 721. Antimicrob Agents Chemother 1987;31:1791–7.

462 [15] First in a new class of antibiotics. FDA Consum 2003;37:4.

463 [16] Kieser T, Bibb MJ, Buttner MJ, Chater KF, Hopwood DA. Practical Streptomyces Genetics. John

464 Innes Foundation.

465 [17] Raoult D. Alie’s liig ouet theoy. Int J Antimicrob Agents 2016;47:249.

466 doi:10.1016/j.ijantimicag.2016.01.013.

467 [18] Davies J. Inactivation of antibiotics and the dissemination of resistance genes. Science

468 1994;264:375–82.

33 469 [19] D’Costa VM, MGa KM, Hughes DW, Wight GD. “aplig the atiioti esistoe. “iee

470 2006;311:374–7. doi:10.1126/science.1120800.

471 [20] Wright GD. The antibiotic resistome: the nexus of chemical and genetic diversity. Nat Rev

472 Microbiol 2007;5:175–86. doi:10.1038/nrmicro1614.

473 [21] Barlow M, Hall BG. Phylogenetic analysis shows that the OXA beta-lactamase genes have been

474 on plasmids for millions of years. J Mol Evol 2002;55:314–21. doi:10.1007/s00239-002-2328-y.

475 [22] D’Costa VM, Kig CE, Kala L, Moa M, “ug WWL, “haz C, et al. Atiioti esistae is

476 ancient. Nature 2011;477:457–61. doi:10.1038/nature10388.

477 [23] Kashuba E, Dmitriev AA, Kamal SM, Melefors O, Griva G, Römling U, et al. Ancient permafrost

478 staphylococci carry antibiotic resistance genes. Microb Ecol Health Dis 2017;28:1345574.

479 doi:10.1080/16512235.2017.1345574.

480 [24] Bhullar K, Waglechner N, Pawlowski A, Koteva K, Banks ED, Johnston MD, et al. Antibiotic

481 resistance is prevalent in an isolated cave microbiome. PloS One 2012;7:e34953.

482 doi:10.1371/journal.pone.0034953.

483 [25] Lugli GA, Milani C, Mancabelli L, Turroni F, Ferrario C, Duranti S, et al. Ancient bacteria of the

484 Ötzi’s ioioe: a geoi tale fo the Coppe Age. Mioioe ;:.

485 doi:10.1186/s40168-016-0221-y.

486 [26] Van Goethem MW, Pierneef R, Bezuidt OKI, Van De Peer Y, Cowan DA, Makhalanyane TP. A

487 eseoi of histoial atiioti esistae gees i eote pistie Atati soils.

488 Microbiome 2018;6:40. doi:10.1186/s40168-018-0424-5.

489 [27] Marshall CG, Lessard IA, Park I, Wright GD. Glycopeptide antibiotic resistance genes in

490 glycopeptide-producing organisms. Antimicrob Agents Chemother 1998;42:2215–20.

491 [28] Jiang X, Ellabaan MMH, Charusanti P, Munck C, Blin K, Tong Y, et al. Dissemination of antibiotic

492 resistance genes from antibiotic producers to pathogens. Nat Commun 2017;8:15784.

493 doi:10.1038/ncomms15784.

34 494 [29] Waglehe N, Wight GD. Atiioti esistae: it’s ad, ut hy is’t it ose? BMC Biol

495 2017;15:84. doi:10.1186/s12915-017-0423-1.

496 [30] Pećaa M, Jajić Z, Koaeić A, Pajić M, Doaoaki D, Miskoić ““. Bus teatet i

497 ancient times. Med Pregl 2013;66:263–7.

498 [31] Gelmo P. Über Sulfamide der p-Amidobenzolsulfonsäure. J Für Prakt Chem 1908;77:369–82.

499 [32] Lewis K. Platforms for antibiotic discovery. Nat Rev Drug Discov 2013;12:371–87.

500 doi:10.1038/nrd3975.

501 [33] Fleming A. Penicillin. Nobel Lecture. 1945.

502 [34] Avery OT, Dubos R. The specific action of a bacterial enzyme on Pneumococci of type III. Science

503 1930;72:151–2. doi:10.1126/science.72.1858.151.

504 [35] Dubos RJ, Hotchkiss RD. The production of bactericidal substances by aerobic sporulating

505 bacilli. J Exp Med 1941;73:629–40.

506 [36] Herrell WE, Heilman D. Experimental and clinical studies on gramicidin. J Clin Invest

507 1941;20:583–91. doi:10.1172/JCI101251.

508 [37] Lewis K. Antibiotics: Recover the lost art of drug discovery. Nature 2012;485:439–40.

509 doi:10.1038/485439a.

510 [38] Waksman SA, Geiger WB, Reynolds DM. Strain Specificity and Production of Antibiotic

511 Substances: VII. Production of Actinomycin by Different Actinomycetes. Proc Natl Acad Sci U S A

512 1946;32:117–20.

513 [39] Jones D, Metzger HJ, Schatz A, Waksman SA. Control of Gram-negative bacteria in experimental

514 animals by streptomycin. Science 1944;100:103–5. doi:10.1126/science.100.2588.103.

515 [40] Waksman SA, Lechevalier HA. Neomycin, a New Antibiotic Active against Streptomycin-

516 Resistant Bacteria, including Tuberculosis Organisms. Science 1949;109:305–7.

517 doi:10.1126/science.109.2830.305.

518 [41] Waksman SA, Horning ES, Spencer EL. The production of two antibacterial substances,

519 fumigacin and clavacin. Science 1942;96:202–3. doi:10.1126/science.96.2487.202.

35 520 [42] Streptomycin treatment of pulmonary tuberculosis. Br Med J 1948;2:769–82.

521 [43] Falagas ME, Skalidis T, Vardakas KZ, Legakis NJ, Hellenic Cefiderocol Study Group. Activity of

522 cefiderocol (S-649266) against carbapenem-resistant Gram-negative bacteria collected from

523 inpatients in Greek hospitals. J Antimicrob Chemother 2017;72:1704–8.

524 doi:10.1093/jac/dkx049.

525 [44] Blaskovich MAT, Butler MS, Cooper MA. Polishing the tarnished silver bullet: the quest for new

526 antibiotics. Essays Biochem 2017;61:103–14. doi:10.1042/EBC20160077.

527 [45] Boman HG. Antibacterial peptides: basic facts and emerging concepts. J Intern Med

528 2003;254:197–215.

529 [46] Katz L, Baltz RH. Natural product discovery: past, present, and future. J Ind Microbiol Biotechnol

530 2016;43:155–76. doi:10.1007/s10295-015-1723-5.

531 [47] Diamond G, Beckloff N, Weinberg A, Kisich KO. The roles of antimicrobial peptides in innate

532 host defense. Curr Pharm Des 2009;15:2377–92.

533 [48] Arnison PG, Bibb MJ, Bierbaum G, Bowers AA, Bugni TS, Bulaj G, et al. Ribosomally synthesized

534 and post-translationally modified peptide natural products: overview and recommendations for

535 a universal nomenclature. Nat Prod Rep 2013;30:108–60. doi:10.1039/c2np20085f.

536 [49] Walsh CT, Fischbach MA. Natural products version 2.0: connecting genes to molecules. J Am

537 Chem Soc 2010;132:2469–93. doi:10.1021/ja909118a.

538 [50] Zhang J, Liu N, Cacho RA, Gong Z, Liu Z, Qin W, et al. Structural basis of nonribosomal peptide

539 macrocyclization in fungi. Nat Chem Biol 2016;12:1001–3. doi:10.1038/nchembio.2202.

540 [51] Meier JL, Burkart MD. The chemical biology of modular biosynthetic enzymes. Chem Soc Rev

541 2009;38:2012–45. doi:10.1039/b805115c.

542 [52] Skarnes RC, Watson DW. Antimicrobial factors of normal tissues and fluids. Bacteriol Rev

543 1957;21:273–94.

544 [53] Reinholz M, Ruzicka T, Schauber J. Cathelicidin LL-37: an antimicrobial peptide with a role in

545 inflammatory skin disease. Ann Dermatol 2012;24:126–35. doi:10.5021/ad.2012.24.2.126.

36 546 [54] Steckbeck JD, Deslouches B, Montelaro RC. Antimicrobial peptides: new drugs for bad bugs?

547 Expert Opin Biol Ther 2014;14:11–4. doi:10.1517/14712598.2013.844227.

548 [55] Hancock REW, Nijnik A, Philpott DJ. Modulating immunity as a therapy for bacterial infections.

549 Nat Rev Microbiol 2012;10:243–54. doi:10.1038/nrmicro2745.

550 [56] Ahmad V, Khan MS, Jamal QMS, Alzohairy MA, Al Karaawi MA, Siddiqui MU. Antimicrobial

551 potential of bacteriocins: in therapy, agriculture and food preservation. Int J Antimicrob Agents

552 2017;49:1–11. doi:10.1016/j.ijantimicag.2016.08.016.

553 [57] Cotter PD, Ross RP, Hill C. Bacteriocins - a viable alternative to antibiotics? Nat Rev Microbiol

554 2013;11:95–105. doi:10.1038/nrmicro2937.

555 [58] Piper C, Draper LA, Cotter PD, Ross RP, Hill C. A comparison of the activities of lacticin 3147 and

556 nisin against drug-resistant Staphylococcus aureus and Enterococcus species. J Antimicrob

557 Chemother 2009;64:546–51. doi:10.1093/jac/dkp221.

558 [59] Shin JM, Gwak JW, Kamarajan P, Fenno JC, Rickard AH, Kapila YL. Biomedical applications of

559 nisin. J Appl Microbiol 2016;120:1449–65. doi:10.1111/jam.13033.

560 [60] Mathur H, Field D, Rea MC, Cotter PD, Hill C, Ross RP. Bacteriocin-Antimicrobial Synergy: A

561 Medical and Food Perspective. Front Microbiol 2017;8:1205. doi:10.3389/fmicb.2017.01205.

562 [61] Wang G, Li X, Wang Z. APD3: the antimicrobial peptide database as a tool for research and

563 education. Nucleic Acids Res 2016;44:D1087-1093. doi:10.1093/nar/gkv1278.

564 [62] Fosgerau K, Hoffmann T. Peptide therapeutics: current status and future directions. Drug Discov

565 Today 2015;20:122–8. doi:10.1016/j.drudis.2014.10.003.

566 [63] Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, et al. A new antibiotic kills

567 pathogens without detectable resistance. Nature 2015;517:455–9. doi:10.1038/nature14098.

568 [64] Parmar A, Iyer A, Prior SH, Lloyd DG, Leng Goh ET, Vincent CS, et al. Teixobactin analogues

569 reveal enduracididine to be non-essential for highly potent antibacterial activity and lipid II

570 binding. Chem Sci 2017;8:8183–92. doi:10.1039/c7sc03241b.

37 571 [65] Vukoaoić M, Žuič V, Kuej Š, Jača B, Jeerica S, Podlipec R, et al. Nano-engineering the

572 Antimicrobial Spectrum of Lantibiotics: Activity of Nisin against Gram Negative Bacteria. Sci Rep

573 2017;7:4324. doi:10.1038/s41598-017-04670-0.

574 [66] Lee B, Park J, Ryu M, Kim S, Joo M, Yeom J-H, et al. Antimicrobial peptide-loaded gold

575 nanoparticle-DNA aptamer conjugates as highly effective antibacterial therapeutics against

576 Vibrio vulnificus. Sci Rep 2017;7:13572. doi:10.1038/s41598-017-14127-z.

577 [67] Zhang X-L, Jiang A-M, Ma Z-Y, Li X-B, Xiong Y-Y, Dou J-F, et al. The synthetic antimicrobial

578 peptide pexiganan and its nanoparticles (PNPs) exhibit the anti-helicobacter pylori activity in

579 vitro and in vivo. Mol Basel Switz 2015;20:3972–85. doi:10.3390/molecules20033972.

580 [68] Habets MGJL, Brockhurst MA. Therapeutic antimicrobial peptides may compromise natural

581 immunity. Biol Lett 2012;8:416–8. doi:10.1098/rsbl.2011.1203.

582 [69] Waksman Selman. Microbial antagonisms and antibiotic substances. Oxford university press.

583 1947.

584 [70] Williston EH, Zia-Walrath P, Youmans GP. Plate Methods for Testing Antibiotic Activity of

585 Actinomycetes against Virulent Human Type Tubercle Bacilli. J Bacteriol 1947;54:563–8.

586 [71] Oliveira L de C, Silveira AMM, Monteiro A de S, Dos Santos VL, Nicoli JR, Azevedo VA de C, et al.

587 In silico Prediction, in vitro Antibacterial Spectrum, and Physicochemical Properties of a

588 Putative Bacteriocin Produced by Lactobacillus rhamnosus Strain L156.4. Front Microbiol

589 2017;8:876. doi:10.3389/fmicb.2017.00876.

590 [72] Dubourg G, Elsawi Z, Raoult D. Assessment of the in vitro antimicrobial activity of Lactobacillus

591 species for identifying new potential antibiotics. Int J Antimicrob Agents 2015;46:590–3.

592 doi:10.1016/j.ijantimicag.2015.05.011.

593 [73] Adnan M, Patel M, Hadi S. Functional and health promoting inherent attributes of Enterococcus

594 hirae F2 as a novel probiotic isolated from the digestive tract of the freshwater fish Catla catla.

595 PeerJ 2017;5:e3085. doi:10.7717/peerj.3085.

38 596 [74] Zipperer A, Konnerth MC, Laux C, Berscheid A, Janek D, Weidenmaier C, et al. Human

597 commensals producing a novel antibiotic impair pathogen colonization. Nature 2016;535:511–

598 6. doi:10.1038/nature18634.

599 [75] Krismer B, Liebeke M, Janek D, Nega M, Rautenberg M, Hornig G, et al. Nutrient limitation

600 governs Staphylococcus aureus metabolism and niche adaptation in the human nose. PLoS

601 Pathog 2014;10:e1003862. doi:10.1371/journal.ppat.1003862.

602 [76] Vijayakumar PP, Muriana PM. A Microplate Growth Inhibition Assay for Screening Bacteriocins

603 against Listeria monocytogenes to Differentiate Their Mode-of-Action. Biomolecules

604 2015;5:1178–94. doi:10.3390/biom5021178.

605 [77] Moy TI, Ball AR, Anklesaria Z, Casadei G, Lewis K, Ausubel FM. Identification of novel

606 antimicrobials using a live-animal infection model. Proc Natl Acad Sci U S A 2006;103:10414–9.

607 doi:10.1073/pnas.0604055103.

608 [78] Rex JH, Outterson K. Antibiotic reimbursement in a model delinked from sales: a benchmark-

609 based worldwide approach. Lancet Infect Dis 2016;16:500–5. doi:10.1016/S1473-

610 3099(15)00500-9.

611 [79] Saisho Y, Katsube T, White S, Fukase H, Shimada J. Pharmacokinetics, Safety, and Tolerability of

612 Cefiderocol, a Novel Siderophore Cephalosporin for Gram-Negative Bacteria, in Healthy

613 Subjects. Antimicrob Agents Chemother 2018;62. doi:10.1128/AAC.02163-17.

614 [80] Armstrong ES, Miller GH. Combating evolution with intelligent design: the neoglycoside ACHN-

615 490. Curr Opin Microbiol 2010;13:565–73. doi:10.1016/j.mib.2010.09.004.

616 [81] Connolly LE, Riddle V, Cebrik D, Armstrong ES, Miller LG. A Multicenter, Randomized, Double-

617 Blind, Phase 2 Study of the Efficacy and Safety of Plazomicin Compared with Levofloxacin in the

618 Treatment of Complicated Urinary Tract Infection and Acute Pyelonephritis. Antimicrob Agents

619 Chemother 2018;62. doi:10.1128/AAC.01989-17.

39 620 [82] Cimermancic P, Medema MH, Claesen J, Kurita K, Wieland Brown LC, Mavrommatis K, et al.

621 Insights into secondary metabolism from a global analysis of prokaryotic biosynthetic gene

622 clusters. Cell 2014;158:412–21. doi:10.1016/j.cell.2014.06.034.

623 [83] Tracanna V, de Jong A, Medema MH, Kuipers OP. Mining prokaryotes for antimicrobial

624 compounds: from diversity to function. FEMS Microbiol Rev 2017;41:417–29.

625 doi:10.1093/femsre/fux014.

626 [84] Walsh CJ, Guiae CM, Hill C, Ross RP, O’Toole PW, Cotte PD. I silio idetifiatio of

627 bacteriocin gene clusters in the gastrointestinal tract, based on the Human Microbiome

628 Pojet’s efeee geoe dataase. BMC Mioiol ;:. doi:./s-015-

629 0515-4.

630 [85] Charlop-Powers Z, Milshteyn A, Brady SF. Metagenomic small molecule discovery methods.

631 Curr Opin Microbiol 2014;19:70–5. doi:10.1016/j.mib.2014.05.021.

632 [86] Begley M, Cotter PD, Hill C, Ross RP. Identification of a novel two-peptide lantibiotic,

633 lichenicidin, following rational genome mining for LanM proteins. Appl Environ Microbiol

634 2009;75:5451–60. doi:10.1128/AEM.00730-09.

635 [87] Zheng J, Gänzle MG, Lin XB, Ruan L, Sun M. Diversity and dynamics of bacteriocins from human

636 microbiome. Environ Microbiol 2015;17:2133–43. doi:10.1111/1462-2920.12662.

637 [88] Weber T, Blin K, Duddela S, Krug D, Kim HU, Bruccoleri R, et al. antiSMASH 3.0-a comprehensive

638 resource for the genome mining of biosynthetic gene clusters. Nucleic Acids Res

639 2015;43:W237-243. doi:10.1093/nar/gkv437.

640 [89] Wilson MC, Piel J. Metagenomic approaches for exploiting uncultivated bacteria as a resource

641 for novel biosynthetic enzymology. Chem Biol 2013;20:636–47.

642 doi:10.1016/j.chembiol.2013.04.011.

643 [90] Hover BM, Kim S-H, Katz M, Charlop-Powers Z, Owen JG, Ternei MA, et al. Culture-independent

644 discovery of the malacidins as calcium-dependent antibiotics with activity against multidrug-

40 645 resistant Gram-positive pathogens. Nat Microbiol 2018;3:415–22. doi:10.1038/s41564-018-

646 0110-1.

647 [91] Taylor SD, Palmer M. The action mechanism of daptomycin. Bioorg Med Chem 2016;24:6253–

648 68. doi:10.1016/j.bmc.2016.05.052.

649 [92] Pawluk A, Davidson AR, Maxwell KL. Anti-CRISPR: discovery, mechanism and function. Nat Rev

650 Microbiol 2018;16:12–7. doi:10.1038/nrmicro.2017.120.

651 [93] Citorik RJ, Mimee M, Lu TK. Sequence-specific antimicrobials using efficiently delivered RNA-

652 guided nucleases. Nat Biotechnol 2014;32:1141–5. doi:10.1038/nbt.3011.

653 [94] Bikard D, Euler CW, Jiang W, Nussenzweig PM, Goldberg GW, Duportet X, et al. Exploiting

654 CRISPR-Cas nucleases to produce sequence-specific antimicrobials. Nat Biotechnol

655 2014;32:1146–50. doi:10.1038/nbt.3043.

656 [95] Kim J-S, Cho D-H, Park M, Chung W-J, Shin D, Ko KS, et al. CRISPR/Cas9-Mediated Re-

657 Sensitization of Antibiotic-Resistant Escherichia coli Harboring Extended-Spectrum β-

658 Lactamases. J Microbiol Biotechnol 2016;26:394–401. doi:10.4014/jmb.1508.08080.

659 [96] Donia MS, Cimermancic P, Schulze CJ, Wieland Brown LC, Martin J, Mitreva M, et al. A

660 systematic analysis of biosynthetic gene clusters in the human microbiome reveals a common

661 family of antibiotics. Cell 2014;158:1402–14. doi:10.1016/j.cell.2014.08.032.

662 [97] Lagier J-C, Dubourg G, Million M, Cadoret F, Bilen M, Fenollar F, et al. Culturing the human

663 microbiota and culturomics. Nat Rev Microbiol 2018. doi:10.1038/s41579-018-0041-0.

664 [98] Lagier J-C, Khelaifia S, Alou MT, Ndongo S, Dione N, Hugon P, et al. Culture of previously

665 uncultured members of the human gut microbiota by culturomics. Nat Microbiol 2016;1:16203.

666 doi:10.1038/nmicrobiol.2016.203.

667

41 668 Table legend

669 Table 1. Discovery date, origin of the organism and synthesis pathway of antibiotics.

670 Table 2. Clinical trials involving antimicrobial peptides.

671 Figure legend

672 Figure 1. Evolution of the number of antibiotics approved by the US Food and Drug

673 Administration (grey) compared to the total cumulative number of antibiotics available (light

674 grey). Adapted from Boucher HW et al., CID 2009.

675 Figure 2. Solid culture approaches highlight inhibition between two bacterial species.

676 Figure 3. Liquid culture approaches highlight inhibition between two bacterial species.

42 Tables

Table 1. Discovery date, origin of the organism and synthesis pathway of antibiotics.

FDA Class Antibiotic Discovery Organism Synthesis pathway References * approval

Aminoglycosides Capreomycin 1960 1969 Streptomyces capreolus Aminoglycoside [1] Framycetin 1953 1955 Streptomyces lavendulae Aminoglycoside [2] Gentamicin 1963 1979 Micromonospora purpurea Aminoglycoside [3] Kanamycin 1957 1973 Streptomyces kanamyceticus Aminoglycoside [4] Natamycin 1957 Streptomyces natalensis Aminoglycoside [5] Neomycin 1949 1954 Streptomyces fradiae Aminoglycoside [6] Plazomicin 2009 2018 semisynthetic [7] Sisomicin 1970 Micromonospora inyoensis Aminoglycoside [8]

Streptomycin 1943 1946 Streptomyces griseus Aminoglycoside [9]

Tobramycin 1967 1975 Streptomyces tenebrarius Aminoglycoside [10]

Antituberculous Ethambutol 1961 1967 Synthetic [11] drugs Ethionamide 1956 1965 Synthetic

Isoniazid 1952 1952 Synthetic [12] Pyrazinamide 1936 1952 Synthetic [13] β-lactams Carbapenem 1976 1986 Streptomyces cattleya NRPS [14] Cephalosporin 1948 1964 Cephalosporium acremonium NRPS Brotzu G, unpublished

Monobactam 1981 1987 Chromobacterium violaceum NRPS [15]

Penicillium notatum, Penicillin 1928 1938 NRPS [16] Penicillium chrysogenum

Carboxylic acid Mupirocin 1971 1987 Pseudomonas flurorescens PKS [17] Chloramphenicols Chloramphenicol 1946 1948 Streptomyces venezuelae shikimate [18]

Fosfomycin Fosfomycin 1969 1989 Streptomyces fradiae Carbon-Phosphate [19]

Glycopeptides Dalbavancin 2002 2014 semisynthetic [20] Oritavancin 1996 2014 semisynthetic [21] Teicoplanin 1978 1987 ** Actinoplanes teichomyceticus NRPS [22] Vancomycin 1953 1958 Amycolatopsis orientalis NRPS [23]

semisynthetic derived Ketolides Telithromycin 1997 2004 [24] of macrolide

Lincosamide Lincomycin 1963 1964 Streptomyces lincolnensis NRPS [25]

Lipopeptides Daptomycin 1986 2003 Streptomyces roseosporus NRPS [26]

Aguilar A & McGuire JM, Macrolides Erythromycine 1948 1951 Streptomyces erythreus PKS unpublished

Streptomyces narbonensis Josamycin 1967 PKS [27] var. josamyceticus Midecamycin 1975 Streptomyces mycarofaciens PKS [28]

43 Spiramycin 1952 1955 Streptomyces ambofaciens PKS [29] Dactylosporangium Fidaxomicine 1975 2011 aurantiacum subspecies PKS [30] hamdenesis Nitrofuran Nitrofurantoin 1952 1953 synthetic [31] Nitroimidazole Metronidazole 1960 1960 Streptomyces sp. semisynthetic [32] Ornidazole 1975 synthetic [33]

Oxazolidinones Linezolid 1987 2000 synthetic [34]

Tedizolid 2008 2014 synthetic [35]

Polypeptides Polymyxin 1947 1959 Paenibacillus polymyxa NRPS [36]

Quinolones Delafloxacin 2000 2017 synthetic [37] Norfloxacin 1961 1968 synthetic [38] Nalidixic acid 1960 1967 synthetic [39]

Rifamycins Rifampicin 1957 1958 Streptomyces mediterranei Hybrid NRPS-PKS [40]

Steroid Fusidic acid 1962 1983 Fusidium coccineum Terpene [41]

Streptogramines Streptogramin B 1953 1998 Streptomyces graminofaciens NRPS [42]

Pristinamycin 1961 Streptomyces pristinaespiralis NRPS [43] Sulfonamide Sulfamethoxazole 1961 1961 synthetic [44] Sulfamethoxazole synthetic + trimethoprim 1968 1974 [45] Streptomyces aureofaciens, Tetracyclines Chlortetracycline 1948 1952 PKS [46] Streptomyces rimosus Eravacycline 2010 synthetic [47] Minocycline 1961 1971 semisynthetic [48] Tigecycline 1999 2005 synthetic [49] * All the references are gathered in the supplementary references

** First approved in Italy, then Europe, Asia, South America. Not approved by the FDA

44 Table 2. Clinical trials involving antimicrobial peptides.

Antimicrobi Natural Spectru Identifier Ye Phas Administration Indication Results References * al peptide product of m ar e

Pexiganan Xenopus Large, NCT01594762 20 III Topical cream 0.8% Diabetic Foot No significant difference compared to [50] acetate laevis antitoxin 17 Infection placebo (=MSI 78 (African activity =Suponex clawed TM) frog)

NCT01590758 20 III Topical cream 0.8% Diabetic Foot No significant difference compared to [51] 16 Infection placebo

NCT00563433, 20 III Topical cream 1-2% Diabetic Foot No significant difference compared to [52] NCT00563394 07 Infection oral ofloxacin 400mg

Iseganan Porcine Large NCT00118781 20 II / III Oral solution 9mg Ventilator- No significant difference compared to [53] (=IB367) Neutrophils 05 Associated placebo Pneumonia

NCT00022373 20 III Oral solution 9mg Oral mucositis No significant difference compared to [54] 04 placebo

20 III Oral solution 9mg Oral mucositis No significant difference compared to [55] 03 placebo

20 III Oral solution 9mg Oral mucositis No significant difference compared to [56] 04 placebo

Omiganan Bovine Large NCT03091426 20 II Topical cream 1%, 1.75%, 2.5% Atopic dermatitis Work in progress [57] (=MBI 226) Neutrophils 17

NCT03071679 20 I Topical cream 1%, 2.5% Healthy volunteers Work in progress [58] 17

NCT02849262 20 II Topical Gel 2.5% Genital warts unknown [59] 16

NCT02456480 20 II Topical cream 1%, 2.5% Atopic dermatitis unknown [60] 15

NCT02576847 20 III Topical cream Rosacea Work in progress [61] 15

NCT02596074 20 II Topical cream 2.5% Vulval Work in progress [62] 15 Intraepithelial Neoplasia

NCT00608959 20 III Topical cream 1% Skin Antisepsis in No significant difference compared to [63] 10 Healthy Adult chlorexidine

45 NCT00231153 20 III Topical cream 1% Prevention of Significantly better than Povidone- [64] 09 infection/colonizati Iodine for Microbiologically-confirmed on of Central catheter infection, but not in clinical Catheter local catheter infection

NCT02571998 20 II Topical cream Inflammatory acne Work in progress [65] 15 vulgaris

Lytixar Large NCT01223222 20 II Topical cream 1%, 2%, 5% Skin Infection unknown [66] (=LTX-109 11 (Gram positive) =AMC 109)

NCT01158235 20 I / II Topical cream 1%, 2%, 5% Nasal MRSA Decolonization significantly better [67] 15 decolonization than placebo

NCT01803035 20 II Topical cream 1%, 2% Impetigo unknown [68] 14 hLF1-11 Human Large NCT00509834 20 I / II Intraveinous 0.5mg daily bolusfor Candidaemia unknown [69] 07 14 days

NCT00509847 20 I / II Intraveinous 0.5mg daily bolusfor Staphylococcus unknown [70] 07 10 days Epidermidis bacteremia

NCT00509938 20 I / II Intraveinous 5mg single dose Hematopoietic unknown [71] 07 Stem Cell Transplantation Bacterial Infections and Mycoses

NCT00430469 20 I / II Intraveinous 0.5mg 10 days Autologous unknown [72] 07 Haematopoietic Stem Cell Transplant Recipients

PXL01 Human NCT01022242 20 II Local 0.5 ml Flexor Tendon No significant difference compared to [73] 09 Surgery placebo

NCT00860080 20 I Local (intra abdominal injection) Healthy volunteers unknown [74] 09 10, 20, and 40 mg

PAC-113 Human Narrow NCT00659971 20 II Topical mouthrinse 0.15%; Oral Candidiasis unknown [75] saliva (fungus) 08 0.075%; 0.0375% among Seropositive Individuals

Novexatin Human Narrow NCT02343627 20 II Topicalbrush-on-treatment Onychomycosis unknown [76] (=NP-213) (fungus) 10

46 LL-37 Human Large 20 I Topical 0.5, 1.6, or 3.2 mg/mL Venous leg ulcers 0.5mg was significantly better than [77] (=CAP-18) (epithelial 13 twice weekly Placebo cells)

Gramicidin NCT00990392 20 I Topical (gramicidin, polymixin, Prevention of Withdrawn [78] 09 bacitracin) infection/colonizati on of Central Catheter

19 Topical cream (triamcinolone Infected Less effective than topical cream [79] 80 acetonide, neomycin sulphate, dermatoses triamcinolone acetonide, neomycin nystatin and gramicidin) sulphate and than undecenoic acid

19 Topical eye drops (neomycin- Bacterial No differences versus trimethoprim- [80] 82 polymyxin-gramicidin) conjunctivitis polymyxin

19 Topical ear drops Acute external No differences versus [81] 85 (framycitin/gramicidin) otitis oxytetracycline/hydrocortisone/polym yxin B

19 Topical ear spray Otitis externa Significantly less effective than [82] 90 (framycetin/gramicidin/dexamethas neomycin/dexamethasone one )

20 Topical eye drops (neomycin Hordeolum No significant difference compared to [83] 05 sulfate, polymyxin B sulfate and placebo gramicidin)

Gramicidin + NCT00400595 20 IV Topical ointment Prevention of Not superior to mupirocin [84] Polymyxin B 15 Catheter-related (Polysporin) Infections in Patients Treated with Peritoneal Dialysis

20 Intranasal Eradication of Significantly less effective than [85] 09 colonization with mupirocin MRSA

NVB-302 NVB302/001 20 I unknown Clostridium difficile unknown [86] 11 infection

Nisin Lactococcu Gram- 20 Topical cream (6 microg/mL) Staphylococcal Significantly better than Placebo [87] s lactis Positive 08 mastitis bacteria

* All the references are gathered in the supplementary references.

47 Fig. 1

60

54 50 52

47

40 40

30 30

20

16 16 14 10 10 7 5 2 0 1983-1987 1988-1992 1993-1997 1998-2002 2003-2007 2008-2012

New antibiotics approved bye the FDA Total number of antibiotics discovered

48 Fig. 2 Cross streak

Vertical streak of tested strain Incubation Incubation

Horizontal streak of indicator strain Spot-on-lawn

Incubation

Drop-off Lawn of the broth of tested strain indicator strain Culture supernatant or culture broth of the tested bacteria

Well diffusion Rest time Incubation Well diffusion (ambient temperature or +4°C)

A cylinder of agar is punched-out

Incubation Lawn or bacterial agar of the indicator strain Agar plug diffusion

A cylinder of agar is punched-out Incubation Indicator strain

Lawn or Tested strain bacterial agar of the 49 tested strain Fig. 3 Co-culture in liquid broth

Membrane with 0.22µm pore size

Incubation

Tested strain Indicator strain

Supernatant

Numeration of colonies Forming Incubation Centrifugation Units, coloration, or optical Cell-free supernatant density measurement

Tested strain

Indicator strain

Indicator strain Tested strain 50 Supplementary references

[1] Black HR, Griffith RS, Brickler JF. Preliminary laboratory studies with capreomycin.

Antimicrob Agents Chemother 1963;161:522–9.

[2] Decaris LJ. Study of a strain of Streptomyces lavendulae producing a new antibiotic:

framycetin. Rev Pathol Gen Physiol Clin 1953;53:676–9.

[3] Black J, Calesnick B, Williams D, Weinstein MJ. Pharmacology of gentamicin, a new

broad-spectrum antibiotic. Antimicrob Agents Chemother 1963;161:138–47.

[4] Takeuchi T, Hikiji T, Nitta K, Yamazaki S, Abe S, Takayama H, et al. Biological studies

on kanamycin. J Antibiot (Tokyo) 1957;10:107–14.

[5] Struyk AP, Hoette I, Drost G, Waisvisz JM, Van Eek T, Hoogerheide JC. Pimaricin, a

new antifungal antibiotic. Antibiot Annu 1957;5:878–85.

[6] Waksman SA, Lechevalier HA. Neomycin, a New Antibiotic Active against

Streptomycin-Resistant Bacteria, including Tuberculosis Organisms. Science

1949;109:305–7. https://doi:10.1126/science.109.2830.305.

[7] Endimiani A, Hujer KM, Hujer AM, Armstrong ES, Choudhary Y, Aggen JB, et al.

ACHN-490, a neoglycoside with potent in vitro activity against multidrug-resistant

Klebsiella pneumoniae isolates. Antimicrob Agents Chemother 2009;53:4504–7.

https://doi:10.1128/AAC.00556-09.

[8] Waitz JA, Moss EL, Drube CG, Weinstein MJ. Comparative activity of sisomicin,

gentamicin, kanamycin, and tobramycin. Antimicrob Agents Chemother 1972;2:431–7.

[9] Jones D, Metzger HJ, Schatz A, Waksman SA. Control of Gram-negative bacteria in

experimental animals by streptomycin. Science 1944;100:103–5.

https://doi:10.1126/science.100.2588.103.

[10] Stark WM, Hoehn MM, Knox NG. Nebramycin, a new broad-spectrum antibiotic

complex. I. Detection and biosynthesis. Antimicrob Agents Chemother 1967;7:314–23.

51 [11] Thomas JP, Baughn CO, Wilkinson RG, Shepherd RG. A new synthetic compound with

antituberculous activity in mice: ethambutol (dextro-2,2’-(ethylenediimino)-di-l-

butanol). Am Rev Respir Dis 1961;83:891–3. https://doi:10.1164/arrd.1961.83.6.891.

[12] De Carvalho P. First results of isonicotinic acid hydrazide therapy of urinary

tuberculosis; preliminary note. Gaz Medica Port 1952;5:466–8.

[13] Yeager RL, Munroe WGC, Dessau FI. Pyrazinamide (aldinamide*) in the treatment of

pulmonary tuberculosis. Trans Annu Meet Natl Tuberc Assoc 1952;48:178–201.

[14] Birnbaum J, Kahan FM, Kropp H, MacDonald JS. Carbapenems, a new class of beta-

lactam antibiotics. Discovery and development of imipenem/cilastatin. Am J Med

1985;78:3–21.

[15] Wells JS, Trejo WH, Principe PA, Bush K, Georgopapadakou N, Bonner DP, et al. SQ

26,180, a novel monobactam. I Taxonomy, fermentation and biological properties. J

Antibiot (Tokyo) 1982;35:184–8.

[16] Fleming A. On the antibacterial action of cultures of a penicillium, with special

reference to their use in the isolation of B. influenzae. 1929. Bull World Health Organ

2001;79:780–90.

[17] Sutherland R, Boon RJ, Griffin KE, Masters PJ, Slocombe B, White AR. Antibacterial

activity of mupirocin (pseudomonic acid), a new antibiotic for topical use. Antimicrob

Agents Chemother 1985;27:495–8.

[18] Ehrlich J, Bartz QR, Smith RM, Joslyn DA, Burkholder PR. Chloromycetin, a New

Antibiotic From a Soil Actinomycete. Science 1947;106:417.

https://doi:10.1126/science.106.2757.417.

[19] Hendlin D, Stapley EO, Jackson M, Wallick H, Miller AK, Wolf FJ, et al.

Phosphonomycin, a new antibiotic produced by strains of Streptomyces. Science

1969;166:122–3.

52 [20] Steiert M, Schmitz F-J. Dalbavancin (Biosearch Italia/Versicor). Curr Opin Investig

Drugs Lond Engl 2000 2002;3:229–33.

[21] Cooper RD, Snyder NJ, Zweifel MJ, Staszak MA, Wilkie SC, Nicas TI, et al. Reductive

alkylation of glycopeptide antibiotics: synthesis and antibacterial activity. J Antibiot

(Tokyo) 1996;49:575–81.

[22] Bardone MR, Paternoster M, Coronelli C. Teichomycins, new antibiotics from

Actinoplanes teichomyceticus nov. sp. II. Extraction and chemical characterization. J

Antibiot (Tokyo) 1978;31:170–7.

[23] Mccormick MH, Mcguire JM, Pittenger GE, Pittenger RC, Stark WM. Vancomycin, a

new antibiotic. I. Chemical and biologic properties. Antibiot Annu 1955;3:606–11.

[24] Jones RN, Biedenbach DJ. Antimicrobial activity of RU-66647, a new ketolide. Diagn

Microbiol Infect Dis 1997;27:7–12.

[25] Meyer CE, Lewis C. Absorption and fate of lincomycin in the rat. Antimicrob Agents

Chemother 1963;161:169–75.

[26] Eliopoulos GM, Willey S, Reiszner E, Spitzer PG, Caputo G, Moellering RC. In vitro

and in vivo activity of LY 146032, a new cyclic lipopeptide antibiotic. Antimicrob

Agents Chemother 1986;30:532–5.

[27] Nitta K, Yano K, Miyamoto F, Hasegawa Y, Sato T. A new antibiotic, josamycin. II.

Biological studies. J Antibiot (Tokyo) 1967;20:181–7.

[28] Rolston KV, Nguyen H, Messer M. In vitro activity of LY264826, a new glycopeptide

antibiotic, against Gram-positive bacteria isolated from patients with cancer. Antimicrob

Agents Chemother 1990;34:2137–41.

[29] Giroud P. Comparative anti-rickettsial effects of spiramycin (RP 5337) with aureomycin

and terramycin. Bull Soc Pathol Exot Filiales 1954;47:642–4.

53 [30] Swanson RN, Hardy DJ, Shipkowitz NL, Hanson CW, Ramer NC, Fernandes PB, et al.

In vitro and in vivo evaluation of tiacumicins B and C against Clostridium difficile.

Antimicrob Agents Chemother 1991;35:1108–11.

[31] Norfleet CM, Beamer PR, Carpenter HM. Furadantin in infections of the genito-urinary

tract. Trans Am Urol Assoc Southeast Sect 1952;2–5:26–34.

[32] Durel P, Couture J, Collart P, Girot C. Flagyl (metronidazole). Br J Vener Dis

1960;36:154–62.

[33] Whillans DW, Adams GE. Electron-affinic sensitization. VI. A pulse radiolysis and ESR

comparison of some 2- and 5-nitroimidazoles. Radiat Res 1975;62:407–21.

[34] Slee AM, Wuonola MA, McRipley RJ, Zajac I, Zawada MJ, Bartholomew PT, et al.

Oxazolidinones, a new class of synthetic antibacterial agents: in vitro and in vivo

activities of DuP 105 and DuP 721. Antimicrob Agents Chemother 1987;31:1791–7.

[35] Moellering RC. Tedizolid: a novel oxazolidinone for Gram-positive infections. Clin

Infect Dis Off Publ Infect Dis Soc Am 2014;58 Suppl 1:S1-3.

https://doi:10.1093/cid/cit658.

[36] Stansly PG, Shepherd RG, White HJ. Polymyxin: a new chemotherapeutic agent. Bull

Johns Hopkins Hosp 1947;81:43–54.

[37] Andrews JM, Ashby JP, Jevons G, Wise R. Determination of in vitro susceptibility to

ABT-492 by BSAC standardized methodology. J Antimicrob Chemother 2003;52:526–

7. https://doi:10.1093/jac/dkg328.

[38] Ito A, Hirai K, Inoue M, Koga H, Suzue S, Irikura T, et al. In vitro antibacterial activity

of AM-715, a new nalidixic acid analog. Antimicrob Agents Chemother 1980;17:103–8.

[39] Lesher GY, Froelich EJ, Gruett MD, Bailey JH, Brundage RP. 1,8-naphthyridine

derivatives. A new class of chemotherapeutic agents. J Med Pharm Chem

1962;91:1063–5.

54 [40] Maggi N, Pasqualucci CR, Ballotta R, Sensi P. Rifampicin: a new orally active

rifamycin. Chemotherapy 1966;11:285–92. https://doi:10.1159/000220462.

[41] Godtfredsen WO, Jahnsen S, Lorck H, Roholt K, Tybring L. Fusidic acid: a new

antibiotic. Nature 1962;193:987.

[42] Charney J, Fisher WP, Curran C, Machlowitz RA, Tytell AA. Streptogramin, a new

antibiotic. Antibiot Chemother Northfield Ill 1953;3:1283–6.

[43] Darbon A. Clinical study of pristinamycin (RP 7293) in infectious pathology and

principally in staphylococcal infections. Sem Hop Ther 1961;37:741–3.

[44] Holloman JL, Fullilove RE, Yergan CM, Barouk E. Clinical experience with a new

sulfonamide sulphamethoxazole. Appl Ther 1961;3:775–9.

[45] Rager J. Clinical study of the sulfamethoxazole-trimethoprim combination in urology.

Acta Urol Belg 1969;37:399–406.

[46] Bryer MS, Schoenbach EB. Aureomycin; experimental and clinical investigations. J Am

Med Assoc 1948;138:117–9.

[47] Sutcliffe JA, O’Brien W, Fyfe C, Grossman TH. Antibacterial activity of eravacycline

(TP-434), a novel fluorocycline, against hospital and community pathogens. Antimicrob

Agents Chemother 2013;57:5548–58. https://doi:10.1128/AAC.01288-13.

[48] Redin GS. Antibacterial activity in mice of minocycline, a new tetracycline. Antimicrob

Agents Chemother 1966;6:371–6.

[49] Petersen PJ, Jacobus NV, Weiss WJ, Sum PE, Testa RT. In vitro and in vivo

antibacterial activities of a novel glycylcycline, the 9-t-butylglycylamido derivative of

minocycline (GAR-936). Antimicrob Agents Chemother 1999;43:738–44.

[50] ClinicalTrials.gov. NCT01594762 n.d.

https://clinicaltrials.gov/ct2/show/NCT01594762?cond=NCT01594762&rank=1

(accessed October 26, 2018).

55 [51] ClinicalTrials.gov. NCT01590758 n.d.

https://clinicaltrials.gov/ct2/show/NCT01590758?cond=NCT01590758&rank=1

(accessed October 26, 2018).

[52] Lipsky BA, Holroyd KJ, Zasloff M. Topical versus systemic antimicrobial therapy for

treating mildly infected diabetic foot ulcers: a randomized, controlled, double-blinded,

multicenter trial of pexiganan cream. Clin Infect Dis Off Publ Infect Dis Soc Am

2008;47:1537–45. https://doi:10.1086/593185.

[53] Kollef M, Pittet D, Sánchez García M, Chastre J, Fagon J-Y, Bonten M, et al. A

randomized double-blind trial of iseganan in prevention of ventilator-associated

pneumonia. Am J Respir Crit Care Med 2006;173:91–7.

https://doi:10.1164/rccm.200504-656OC.

[54] Trotti A, Garden A, Warde P, Symonds P, Langer C, Redman R, et al. A multinational,

randomized phase III trial of iseganan HCl oral solution for reducing the severity of oral

mucositis in patients receiving radiotherapy for head-and-neck malignancy. Int J Radiat

Oncol Biol Phys 2004;58:674–81. https://doi:10.1016/S0360-3016(03)01627-4.

[55] Giles FJ, Miller CB, Hurd DD, Wingard JR, Fleming TR, Sonis ST, et al. A phase III,

randomized, double-blind, placebo-controlled, multinational trial of iseganan for the

prevention of oral mucositis in patients receiving stomatotoxic chemotherapy

(PROMPT-CT trial). Leuk Lymphoma 2003;44:1165–72.

https://doi:10.1080/1042819031000079159.

[56] Giles FJ, Rodriguez R, Weisdorf D, Wingard JR, Martin PJ, Fleming TR, et al. A phase

III, randomized, double-blind, placebo-controlled, study of iseganan for the reduction of

stomatitis in patients receiving stomatotoxic chemotherapy. Leuk Res 2004;28:559–65.

https://doi:10.1016/j.leukres.2003.10.021.

56 [57] ClinicalTrials.gov. NCT03091426 n.d.

https://clinicaltrials.gov/ct2/show/NCT03091426?cond=NCT03091426&rank=1

(accessed October 26, 2018).

[58] ClinicalTrials.gov. NCT03071679 n.d.

https://clinicaltrials.gov/ct2/show/NCT03071679?cond=NCT03071679&rank=1

(accessed October 26, 2018).

[59] ClinicalTrials.gov. NCT02849262 n.d.

https://clinicaltrials.gov/ct2/show/NCT02849262?term=genital&cond=omiganan&rank=

1 (accessed October 26, 2018).

[60] ClinicalTrials.gov. NCT02456480 n.d.

https://clinicaltrials.gov/ct2/show/NCT02456480?cond=NCT02456480&rank=1

(accessed October 26, 2018).

[61] ClinicalTrials.gov. NCT02576847 n.d.

https://clinicaltrials.gov/ct2/show/NCT02576847?cond=NCT02576847&rank=1

(accessed October 26, 2018).

[62] ClinicalTrials.gov. NCT02596074 n.d.

https://clinicaltrials.gov/ct2/show/NCT02596074?cond=NCT02596074&rank=1

(accessed October 26, 2018).

[63] ClinicalTrials.gov. NCT00608959 n.d.

https://clinicaltrials.gov/ct2/show/NCT00608959?cond=NCT00608959&rank=1

(accessed October 26, 2018).

[64] ClinicalTrials.gov. NCT00231153 n.d.

https://clinicaltrials.gov/ct2/show/NCT00231153?cond=NCT00231153&rank=1

(accessed October 26, 2018).

57 [65] ClinicalTrials.gov. NCT02571998 n.d.

https://clinicaltrials.gov/ct2/show/NCT02571998?cond=NCT02571998&rank=1

(accessed October 26, 2018).

[66] ClinicalTrials.gov. NCT01223222 n.d.

https://clinicaltrials.gov/ct2/show/NCT01223222?cond=NCT01223222&rank=1

(accessed October 26, 2018).

[67] Nilsson AC, Janson H, Wold H, Fugelli A, Andersson K, Håkangård C, et al. LTX-109

is a novel agent for nasal decolonization of methicillin-resistant and -sensitive

Staphylococcus aureus. Antimicrob Agents Chemother 2015;59:145–51.

https://doi:10.1128/AAC.03513-14.

[68] ClinicalTrials.gov. NCT01803035 n.d.

https://clinicaltrials.gov/ct2/show/NCT01803035?cond=NCT01803035&rank=1

(accessed October 26, 2018).

[69] ClinicalTrials.gov. NCT00509834 n.d.

https://clinicaltrials.gov/ct2/show/NCT00509834?cond=NCT00509834&rank=1

(accessed October 26, 2018).

[70] ClinicalTrials.gov. NCT00509847 n.d.

https://clinicaltrials.gov/ct2/show/NCT00509847?cond=NCT00509847&rank=1

(accessed October 26, 2018).

[71] ClinicalTrials.gov. NCT00509938 n.d.

[72] ClinicalTrials.gov. NCT00430469 n.d.

https://clinicaltrials.gov/ct2/show/NCT00430469?cond=NCT00430469&rank=1

(accessed October 26, 2018).

58 [73] ClinicalTrials.gov. NCT01022242 n.d.

https://clinicaltrials.gov/ct2/show/NCT01022242?cond=NCT01022242&rank=1

(accessed October 26, 2018).

[74] ClinicalTrials.gov. NCT00860080 n.d.

https://clinicaltrials.gov/ct2/show/NCT00860080?cond=NCT00860080&rank=1

(accessed October 26, 2018).

[75] ClinicalTrials.gov. NCT00659971 n.d.

https://clinicaltrials.gov/ct2/show/NCT00659971?cond=NCT00659971&rank=1

(accessed October 26, 2018).

[76] ClinicalTrials.gov. NCT02343627 n.d.

https://clinicaltrials.gov/ct2/show/NCT02343627?cond=nvxt&rank=1 (accessed October

26, 2018).

[77] Grönberg A, Mahlapuu M, Ståhle M, Whately-Smith C, Rollman O. Treatment with LL-

37 is safe and effective in enhancing healing of hard-to-heal venous leg ulcers: a

randomized, placebo-controlled clinical trial. Wound Repair Regen Off Publ Wound

Heal Soc Eur Tissue Repair Soc 2014;22:613–21. https://doi:10.1111/wrr.12211.

[78] ClinicalTrials.gov. NCT00990392 n.d.

https://clinicaltrials.gov/ct2/show/NCT00990392?cond=NCT00990392&rank=1

(accessed October 26, 2018).

[79] Vallé-Jones JC. Treatment of infected dermatoses: a randomized, double-blind clinical

trial in general practice. Pharmatherapeutica 1980;2:293–5.

[80] Genée E, Schlechtweg C, Bauerreiss P, Gibson JR. Trimethoprim-polymyxin eye drops

versus neomycin-polymyxin-gramicidin eye drops in the treatment of presumptive

bacterial conjunctivitis - a double-blind study. Ophtalmologica 1982;184:92–6.

https://doi:10.1159/000309190.

59 [81] Wadsten CJ, Bertilsson CA, Sieradzki H, Edström S. A randomized clinical trial of two

topical preparations (framycitin/gramicidin and oxytetracycline/hydrocortisone with

polymyxin B) in the treatment of external otitis. Arch Otorhinolaryngol 1985;242:135–9.

[82] Smith RB, Moodie J. A general practice study to compare the efficacy and tolerability of

a spray ('Otomize’) versus a standard drop formulation ('Sofradex’) in the treatment of

patients with otitis externa. Curr Med Res Opin 1990;12:12–8.

https://doi:10.1185/03007999009111486.

[83] Hirunwiwatkul P, Wachirasereechai K. Effectiveness of combined antibiotic ophthalmic

solution in the treatment of hordeolum after incision and curettage: a randomized,

placebo-controlled trial: a pilot study. J Med Assoc Thail Chotmaihet Thangphaet

2005;88:647–50.

[84] McQuillan RF, Chiu E, Nessim S, Lok CE, Roscoe JM, Tam P, et al. A randomized

controlled trial comparing mupirocin and polysporin triple ointments in peritoneal

dialysis patients: the MP3 Study. Clin J Am Soc Nephrol CJASN 2012;7:297–303.

https://doi:10.2215/CJN.07970811.

[85] O’Grady S, Hirji Z, Pejcic-Karapetrovic B, Fung S, Dedier H, Takata-Shewchuk J, et al.

A double-blind, randomized, controlled trial of topical polysporin triple compound

versus topical mupirocin for the eradication of colonization with methicillin-resistant

Staphylococcus aureus in a complex continuing care population. Can J Infect Dis Med

Microbiol J Can Mal Infect Microbiol Medicale 2009;20:e49-55.

[86] ISRCTN. NVB302 n.d. http://www.isrctn.com/ISRCTN40071144 (accessed October 26,

2018).

[87] Fernández L, Delgado S, Herrero H, Maldonado A, Rodríguez JM. The bacteriocin

nisin, an effective agent for the treatment of Staphylococcal mastitis during lactation. J

60 Hum Lact Off J Int Lact Consult Assoc 2008;24:311–6. https://doi:10.1177/0890334408317435.

61

Publication n°2

Antagonisms from gut microbiota against human-pathogens

Guillaume André Durand, Didier Raoult, Grégory Dubourg

Draft

62 1 Antagonisms from gut microbiota against human-pathogens

2 Guillaume André Durand *, Grégory Dubourg, Didier Raoult

3 Aix Marseille Univ, IRD, AP-HM, MEPHI, IHU-Méditerranée Infection, Marseille, France

4 * Corresponding author : Dr Guillaume Durand, MEPHI, Aix Marseille Université, IRD

5 IHU - Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13005 Marseille,

6 [email protected], Phone : (33) 413 73 24 01, Fax : (33) 413 73 24 02

7 Acknowledgements and financial disclosure

8 This work was supported by the French Government under the « Investissements d’avenir »

9 (Investments for the Future) a program managed by the Agence Nationale de la Recherche

10 (ANR, fr: National Agency for Research), (reference: Méditerranée Infection 10-IAHU-03).

63 11 Abstract

12 Antibiotic resistance is a public health concern. Research and development of new antibiotics

13 is a cornerstone to address this problem. Actual antimicrobials drugs come mainly from

14 natural products from bacteria and fungi. The aim of this work was to search for new

15 antagonisms against the most pathogen bacteria in human health (S. aureus, E. faecalis, K.

16 pneumoniae, E. cloacae, E. aerogenes and P. aeruginosa). We focused on human gut

17 microbiota as source of new antimicrobials. First, we have selected putative candidates from

18 the analysis of antagonisms of 70 gut metagenomes. Second, we have tested these candidates

19 using well-diffusion method. Third, we have searched for biosynthetic gene clusters (BGCs)

20 in the genomes of the antagonisms using antismash software. 60 strains isolated from gut

21 microbiota were used for 81 antagonism tests. Using liquid broth, we found an inhibition of S.

22 aureus by P. avidum, an inhibition of E. cloacae by B. fragilis, E. dispar, L. delbruckii, P.

23 acidipropionici, S. equinus, S. gallolyticus, and an inhibition of E. aerogenes by B. vulgatus

24 and E. dispar. Using cell-free supernatants, only B. fragilis was active against E. cloacae.

25 BGCs assigned to putative antimicrobial were found in all 8 antagonists, including 32 BGCs

26 of unknown function. In conclusion, this preliminary work highlights the importance of gut

27 microbiota as source of putative new antimicrobials against human pathogen bacteria.

28 Keywords: gut microbiota, culturomics, antibiotics, resistance

29 Text words count: 3,936

30 Abstract word count: 221

31 Tables / Figures: 6 Tables, 4 Figures, 12 Supplementary Figures

32 References: 28

64 33 1. INTRODUCTION

34 Antibiotic resistance appears as a major concern by international organizations and by

35 local agencies. Gram-negative multidrug-resistant bacteria are particularly of concern as some

36 studies have forecasted thousands of death in the next decades [1,2]. According to the World

37 Health Organization (WHO), one of the key strategies is the research and development of new

38 antimicrobial medicines [3]. Indeed, the current pipeline appears insufficient and need more

39 investment, more research in basic science and drug discovery. Most of the agents actually in

40 the pipeline are innovations of existing classes, that are only short term solutions, according to

41 the WHO [3]. Moreover, no new class of antibiotic were commercialized since 2003 and

42 antibiotics commercialized the last years are optimizations or combinations or already known

43 classes [4].

44 Most of current antibiotics in the market are natural products or derived from them,

45 discovered from fungi or Actinobacteria in the 1950s. Recently, important advances were

46 recorded from both culture-dependant and culture-independent methods of exploration of

47 complex ecosystems. Notably, new culture approaches mimicking the natural environment of

48 bacteria have permitted the discover of new bacterial species and new promising antibiotics

49 molecules. From soil, Ling et al., discovered the new antibiotic teixobactin from the new

50 species Eleftheria terrae isolated in the soil using a diffusion chamber [5]. From human nose,

51 Krismer B et al. discovered lugdunin, a new antibiotic inhibiting S. aureus growth [6], and

52 from vaginal microbiota Donia MS et al. found lactocillin [7].

53 The human gut microbiota is another field of interest for antibiotic research. First,

54 many biosynthetic gene clusters were previously described [7]. Secondly, the effects of

55 probiotics were determined by the natural secretion of bacteriocins from gut microbiota [8].

56 Third, new culture approaches particularly culturomics were developed increasing

57 dramatically the gut repertoire of bacteria [9]. Fourthly, the human gut account for high

65 58 number of bacteria living in extreme competition, with concentration ranging from 104 in the

59 duodenum to 1012 CFU/mL in the colon, that represent a favorable condition for the secretion

60 of natural products. Gut repertoire actually account for 1,705 species including 392 new

61 species (data unpublished). Therefore, the first step is to select the species to test among them.

62 This work aims to find bacterial species among commensals of gut microbiota that can inhibit

63 the growth of human pathogen species. We have firstly selected candidates from

64 metagenomic analysis, and then we have tested these candidates in vitro for growth inhibition

65 of a selection of pathogens.

66 2. MATERIAL AND METHODS

67 2.1. Selection of pathogenic bacteria

68 The first step was to select in silico the species that potentially inhibit the growth of

69 human pathogens. For these last, we have selected the nine most isolated bacteria from human

70 blood samples in clinical microbiology laboratory (Institut Hospitalo-Universitaire

71 Mediterranée Infection, IHU MI, Marseille, France) in 2017: Staphylococcus aureus,

72 Klebsiella pneumoniae, Enterobacter cloacae, Enterobacter aerogenes, Enterococcus

73 faecalis, Pseudomonas aeruginosa, Escherichia coli, Streptococcus pneumoniae,

74 Streptococcus pyogenes (Fig. 1, data unpublished). These species were grown from the

75 Collection de Souches de l’Urmite (CSUR, WDCM 875, http://www.mediterranee-

76 infection.com/article.php?laref=14&titre=collection-de-souches) that is the biobank of the

77 Institut Hospitalo-Universitaire of infectious disease of Marseille, France (IHU MI). For the

78 in vitro tests, we used the followed strains: Staphylococcus aureus P8402, Klebsiella

79 pneumoniae P2577, Enterobacter cloacae P8404, Enterobacter aerogenes P8401,

80 Enterococcus faecalis P2525, Pseudomonas aeruginosa P2283, Escherichia coli P2285,

81 Streptococcus pneumoniae P5700, Streptococcus pyogenes P635.

82 2.2. Screening candidates by metagenomics

66 83 Specimen included

84 The presence of the pathogenic bacteria was screened from 70 stool metagenomes

85 samples. These samples came from patients suffering from Clostridium difficile infections

86 (CDI, n=30) and from controls (n=40). Controls were constituted by 27 healthy volunteers

87 and 13 patients carrying C. difficile without clinical symptoms. All patients have signed

88 written consent and the study was approved by the ethic committee of the IHU Méditerranée-

89 Infection (agreement number 2016-011).

90 Metagenomic processing

91 All samples were sequenced and analysed within the same run. Samples were

92 sequenced for V4-V3 regions of 16S rRNA gene using MiSeq technology. The paired end

93 reads of corresponding raw fastq files were assembled into longer joined sequences using

94 FLASH. High-quality sequences were then filtered using QIIME, as previously described

95 [10]. Primers were trimmed and sequences shorter than 200 nucleotides and greater than 500

96 nucleotides were removed. The sequences were grouped into operational taxonomic units

97 (OTUs) using UCLUST. The best taxonomic assignment was defined for each OTUs, using

98 threshold of 97% for the species level, 95% for the genera level, and 80% for the phyla level.

99 OTUs were extracted de novo, without considering singletons [10]. OTUs with less than 5

100 reads were considered as contaminant and therefore excluded for analysis.

101 Selection of candidates

102 From each pathogen species, we have selected the metagenomes in which no OTU

103 were assigned to this pathogen. Then, we have selected the OTUs that were present

104 exclusively when the pathogens were absent. For each pathogen, two analysis were performed

105 : the OTUs were sorted by decrease number of reads and also by decrease number of OTUs.

106 Next, the thirty first most abundant OTUs were selected for each analysis. This strategy was

67 107 performed for both all metagenomes and only controls, in order to limit the impact of the

108 dysbiosis induced by CDI, that appeared as confounding biais. Finally, the OTUs assigned at

109 species taxonomic level were selected as candidates for further analysis. All this strategy is

110 summarized in Fig. 2.

111 2.3. Screening candidates for antimicrobial secretion

112 After growth from CSUR biobank, the identification of the species was confirmed

113 using Matrix Assisted Laser Desorption Ionisation Time of Flight Mass spectrometry

114 (MALDI-TOF MS). The first seeding was performed on Columbia supplemented with 5%

115 sheep blood (Cos, Biomerieux, Marcy l’Etoile, France) with the optimal growth temperature

116 and atmosphere. After 24 or 48h of incubation, the strains were inoculated into a Falcon tube

117 (Sarstedt AG & Co, Numbrecht, Germany) or Hungate tube for fastidious anaerobes,

118 containing liquid broth. The liquid media used was Mueller Hinton (Sigma-Aldrich,

119 Steinheim, Germany) and Anaerobe Basal Broth (CM0957, Thermo Scientific, Oxoid

120 Limited, United Kingdom) for fastidious species. These liquid broths were prepared and

121 autoclaved in the 24h before inoculation, according to the manufacturer’s instructions. After

122 24h of liquid incubation, the broth was tested for antimicrobial activity. The growth in the

123 liquid media was assessed by seeding on Cos agar and identification on MALDI-TOF MS.

124 For each experiment, a negative control was performed using non-inoculated liquid.

125 The antimicrobial activity was searched using agar well-diffusion method. The well-

126 diffusion method consists in the inoculation and spread of a calibrated volume of the

127 indicative strain (i.e. the pathogen strain) over the entire surface of an agar plate. The agar

128 used were Mueller-Hinton (MH, bioMerieux SA, Marcy l’Etoile, France). The inoculation of

129 the indicative strain was performed using a swab using calibrated McFarland solutions. Two

130 inoculation concentrations were systematically performed: 106 CFU/mL and 108 CFU/mL.

131 Then a cylinder of agar was aseptically punched out using a sterile pipette previously trimmed

68 132 in the middle. The holes were 9mm of diameter. Then each well was filled up with 150µL of

133 liquid broth of the tested strain, or with cell free supernatant (CFS). The agar plate is then let

134 rest for one hour in the biosafety cabinet at laboratory temperature prior to incubation at 37°C.

135 Finally after 18h, the inhibition diameter was measured in case of growth inhibition of the

136 indicative strain by the tested strain (Figure 3) [11]. Each test was performed in duplicate and

137 the correct identification of both tested strain and indicative strain was systematically

138 confirmed by MALDI-TOF MS. The CFS of species that exhibited an inhibition was also

139 tested using the same protocol. The pH of these CFS were also measured (Accumet AE150,

140 Fisher Scientific, US).

141 2.4. Research for biosynthetic gene cluster of natural products

142 For each candidate that exhibited an antagonism of the indicative strain, we searched

143 for cluster genes that code for natural products. The genomes of the type strain of each

144 antagonist were dropped from National Center for Biotechnology Information website (NCBI,

145 https://www.ncbi.nlm.nih.gov/genome). The research for biosynthetic gene clusters (BGCs)

146 and their classification into the natural products classes were performed using the antibiotics

147 and secondary metabolite analysis shell (Antismash 4.2.0) with ClusterFinder algorithm

148 enabled [12]. The ClusterFinder algorithm is announced to be able to predict unknown BGCs

149 [13].

150 2.5. Proof of concept

151 Before testing the candidates, we wanted to evaluate the agar well diffusion method

152 using liquid enrichment. Therefore, we have tested bacteria species known to produce

153 bacteriocin and NRP/PK. The first species tested was Paenibacillus polymyxa P3562, the

154 source organism of polymyxine, colistine, as well as several lantibiotics (paenibacilline,

155 paenicidin, paenilan) [14–16]. We also tested 18 strains of Enterococcus hirae, the source

69 156 organism of the class IIa bacteriocin hiracin [17], and the enterocin LD3 [18]. Particularly, we

157 have tested the strain IGR7 [19]. Finally, we have tested three strains of Bacillus subtilis

158 (P3433, P3435, P8403). This species is known to produce 21 bacteriocins, such as the

159 lantibiotic subtilin [20], and ericin [21]. For each species, the agar well diffusion method was

160 compared with the cross-streak and the spot-on-lawn methods. The cross-streak method was

161 performed by seeding onto a Cos agar. Then the tested strain previously grown on liquid

162 media was seeded in a line on the agar. The indicative strain at 106 CFU/mL was then seeded

163 in a perpendicular streak. After 18h of incubation at 37°C, an inhibition zone is searched at

164 the intersection of the streaks [11]. The spot-on-lawn method was performed by depositing

165 5µL of liquid broth of the tested strain on an agar plate previously seeded with the indicative

166 strain using the same protocol than for the agar well diffusion method [22]. After incubation,

167 an inhibitory zone is searched around the spot.

168 3. RESULTS

169 3.1. Screening candidates by metagenomics

170 Metagenomic analysis from 70 stool samples was performed including 40 controls and

171 30 specimens from patients suffering C. difficile infection. From the 40 control samples, there

172 were an average of 144,441 reads (24,094 – 847,692) and 358 OTUs (103 - 664). From the 30

173 colitis samples, there were an average of 102,508 reads (11,364 – 259,084) and 235 OTUs (70

174 – 497). The alpha diversity assessed using Shannon index was 2.53 (1.16 – 3.94), and 3.38

175 (1.29 – 4.59) from the colitis and controls metagenomes, respectively (Table 1).

176 Candidates for growth inhibition of Staphylococcus aureus growth

177 From all metagenomes, absence of OTUs assigned to S. aureus was observed in 64

178 metagenomes (91%) accounting for 8,233,483 reads and 1,750 OTUs. Among these, 272

179 OTUs (16%) were specifically identified in metagenomes in which no OTUs were assigned to

70 180 S. aureus. From control metagenomes, absence of OTUs assigned to S. aureus was observed

181 in 40 metagenomes (100%). Therefore, the strategy was not applicable using controls.

182 Considering all metagenomes, the 272 OTUs were sorted by decreased number of reads and

183 by decreased number of metagenomes and the first 30 of each were selected. From the 30 first

184 OTUs by number of reads, 25 were assigned at species taxonomic level from all

185 metagenomes. From the 30 first OTUs by number of metagenomes, 18 were assigned at

186 species taxonomic level (Suppl. Fig. 1). A total of 32 OTUs were assigned to species level

187 and 14 candidates were available from the CSUR biobank for in vitro testing.

188 Candidates for growth inhibition of Klebsiella pneumoniae

189 From all metagenomes, absence of OTUs assigned to K. pneumoniae was observed in

190 27 metagenomes (39%) accounting for 3,907,201 reads and 1,420 OTUs. Among these, 324

191 OTUs (23%) were specifically detected in metagenomes in which no OTUs were assigned to

192 K. pneumoniae. From the control metagenomes, absence of OTUs assigned to K. pneumoniae

193 was observed in 21 metagenomes (53%) accounting for 3,564,370 reads and 1,404 OTUs.

194 Among these, 417 OTUs (30%) were specifically detected in metagenomes in which no

195 OTUs were assigned to K. pneumoniae. These OTUs were then sort by decreased number of

196 reads and by decreased number of metagenomes and the first 30 of each were selected. From

197 the 30 first OTUs by number of reads, 12 were assigned at species taxonomic level from all

198 metagenomes, and 15 from control metagenomes only. From the 30 first OTUs by number of

199 metagenomes, 10 were assigned at species taxonomic level from all metagenomes, and 10

200 from control metagenomes only (Suppl. Fig. 2). A total of 27 OTUs were assigned to species

201 level and 9 candidates were available from the CSUR biobank for in vitro testing.

202 Candidates for growth inhibition of Enterobacter aerogenes

71 203 From all metagenomes, absence of OTUs assigned to E. aerogenes was observed in 65

204 metagenomes (93%) accounting for 8,269,274 reads and 1,869 OTUs. Among these, 1,315

205 OTUs (70%) were specifically detected in metagenomes in which no OTUs were assigned to

206 E. aerogenes. From the control metagenomes, absence of OTUs assigned to E. aerogenes was

207 observed in 40 metagenomes (100%). Considering all metagenomes, from the 30 first OTUs

208 by number of reads, 21 were assigned at species taxonomic level from all metagenomes. From

209 the 30 first OTUs by number of metagenomes, 2 were assigned at species taxonomic level

210 from all metagenomes (Suppl. Fig. 3). A total of 23 OTUs were assigned to species level and

211 14 candidates were available from the CSUR biobank for in vitro testing.

212 Candidates for growth inhibition of Enterobacter cloacae

213 From all metagenomes, absence of OTUs assigned to E. cloacae was observed in 52

214 metagenomes (74%) accounting for 6,864,703 reads and 1,764 OTUs. Among these, 679

215 OTUs (38%) were specifically detected in metagenomes in which no OTUs were assigned to

216 E. cloacae. From the control metagenomes, absence of OTUs assigned to E. cloacae was

217 observed in 35 metagenomes (88%) accounting for 5,279,495 reads and 1,581 OTUs. Among

218 these, 876 OTUs (54%) were specifically detected in metagenomes in which no OTUs were

219 assigned to E. cloacae. From the 30 first OTUs by number of reads, 21 were assigned at

220 species taxonomic level from all metagenomes, and 23 from control metagenomes only. From

221 the 30 first OTUs by number of metagenomes, 4 were assigned at species taxonomic level

222 from all metagenomes, and 9 from control metagenomes only (Suppl. Fig. 4). A total of 37

223 OTUs were assigned to species level and 15 candidates were available from the CSUR

224 biobank for in vitro testing.

225 Candidates for growth inhibition of Enterococcus faecalis

72 226 From all metagenomes, absence of OTUs assigned to E. faecalis was observed in 39

227 metagenomes (56%) accounting for 4,933,107 reads and 1,590 OTUs. Among these, 389

228 OTUs (24%) were specifically detected in metagenomes in which no OTUs were assigned to

229 E. faecalis. From the control metagenomes, absence of OTUs assigned to E. faecalis was

230 observed in 27 (68%) accounting for 3,836,824 reads and 1,419 OTUs. Among these, 445

231 OTUs (31%) were specifically detected in metagenomes in which no OTUs were assigned to

232 E. faecalis. From the 30 first OTUs by number of reads, 17 were assigned at species

233 taxonomic level from all metagenomes, and 17 from control metagenomes only. From the 30

234 first OTUs by number of metagenomes, 10 were assigned at species taxonomic level from all

235 metagenomes, and 10 from control metagenomes only (Suppl. Fig. 5). A total of 34 OTUs

236 were assigned to species level and 18 candidates were available from the CSUR biobank for

237 in vitro testing.

238 Candidates for growth inhibition of Pseudomonas aeruginosa

239 From all metagenomes, absence of OTUs assigned to P. aeruginosa was observed in

240 53 metagenomes (75%) accounting for 7,446,522 reads and 1,762 OTUs. Among these, 770

241 OTUs (44%) were specifically detected in metagenomes in which no OTUs were assigned to

242 P. aeruginosa. From the control metagenomes, absence of OTUs assigned to P. aeruginosa

243 was observed in 34 metagenomes (85%) accounting for 5,373,736 reads and 1,577 OTUs.

244 Among these, 957 OTUs (61%) were specifically detected in metagenomes in which no

245 OTUs were assigned to P. aeruginosa. From the 30 first OTUs by number of reads, 14 were

246 assigned at species taxonomic level from all metagenomes, and 16 from control metagenomes

247 only. From the 30 first OTUs by number of metagenomes, 7 were assigned at species

248 taxonomic level from all metagenomes, and 5 from control metagenomes only (Suppl. Fig.

249 6). A total of 37 OTUs were assigned to species level and 12 candidates were available from

250 the CSUR biobank for in vitro testing.

73 251 Candidates for growth inhibition of Escherichia coli, Streptococcus pneumoniae,

252 Streptococcus pyogenes

253 Because there were OTUs assigned to E. coli from all metagenomes, the strategy was

254 not applicable (Suppl. Fig. 7). At the opposite, no OTUs were assigned to S. pneumoniae

255 neither to S. pyogenes from all metagenomes, therefore the strategy was not applicable too

256 (Suppl. Fig. 7, Suppl. Fig. 8).

257 3.2. Proof of concept

258 The well diffusion method was tested using Paenibacillus polymyxa P3562 for

259 inhibition of S. aureus P8402, E. coli P2285, K. pneumoniae P2577, E. aerogenes P8401, P.

260 aeruginosa P2283, P. vulgaris P8722, E. faecalis P2525 and L. monocytogenes P825. An

261 inhibition zone was observed for all these species (Table 2, Suppl. Fig. 9). Cross-streaks and

262 spot-on-lawn were performed using S. aureus P8402. The cross-streak method failed to

263 highlight an inhibition whereas a weak diameter was observed by spot-on-lawn (Fig. 4A and

264 C).

265 The well diffusion method applied to Enterococcus hirae IGR7 highlights an

266 inhibition of E. faecalis P2525, E. faecalis P720, E. faecalis P2282, E. faecium P2282, E.

267 durans P8400, L. monocytogenes P828 (Table 3, Suppl. Fig. 10). The test was also positive

268 after incubation under aerobic condition, or after liquid incubation into anaerobe broth as well

269 as into MH broth. However, no inhibition of E. casseliflavus P3601, E. dispar P5040, E.

270 gallinarum P364, S. aureus P8402, S. carnosus P1892, S. devriesi P4716, S. kloosi P837, S.

271 schleiferi P4020, E. aerogenes P8401, E. coli P2285, P. aeruginosa P2283, P. vulgaris

272 P8722, Salmonella enteridis and K. pneumoniae P2577 was found. The strain E. hirae 13144

273 highlighted the same spectrum as IGR7. The strain E. hirae P5207 inhibited the growth of E.

274 faecalis P2525, E. faecium P2282 and L. monocytogenes P828. Inhibition of E. faecalis P2525

74 275 was negative for the strains E. hirae P1518, P1519, P1536, P1537, P3553, P557, P544,

276 P4115, P4871, P3548, P584, P4128, P5364, P4488 and P4127 (Table 3). Finally, the three

277 strains showed positive results were successfully tested by spot-on-lawn. Cross streaks were

278 positive only for the strain P5207 (Fig. 4B, D and E).

279 3.3. Screening candidates for antimicrobial secretion

280 A total of 60 species were tested through 81 inhibition tests, each at both 106 CFU/mL

281 et à 108 CFU/mL (Table 4). The species E. faecalis, K. pneumoniae and P. aeruginosa were

282 not inhibited. Nine well-diffusion tests were positive for eight species. E. aerogenes was

283 inhibited by B. vulgatus P2616, E. dispar P5040. S. aureus was inhibited by P. avidum P6585.

284 E. cloacae was inhibited by B. fragilis P2573, E. dispar P5040, L. delbrueckii P4496, P.

285 acidipropionici P2295, S. equinus P2987, and S. gallolyticus P6782 (Table 4, Suppl. Fig.

286 12). Only E. dispar P5040 inhibited the growth of more than one pathogenic bacterium (i.e. E.

287 cloacae and E. aerogenes). For all these eight strains, the CFS was also tested (Table 5). Only

288 the CFS of B. fragilis P2573 (pH 6.51) inhibited the growth of E. cloacae (Table 5). The

289 average pH of CFS was 6.37, indicating that the pH of CFS of B. fragilis is probably not the

290 origin of the observed antagonism.

291 3.4. NRP, PK, bacteriocins

292 In order to search for BGCs into the antagonist found above, we recovered the

293 available genome of each type strain : Bacteroides vulgatus ATCC 8482

294 (GCF_000012825.1), Enterococcus dispar ATCC 51266 (GCF_000406945.1),

295 Propionibacterium avidum ATCC 25577 (GCF_000227295.1), Bacteroides fragilis ATCC

296 25285 (GCF_000025985.1), Lactobacillus delbrueckii ATCC 9649 (GCF_001908495.1),

297 Propionibacterium acidipropionici ATCC 25562 (GCF_000427845.1), Streptococcus equinus

298 ATCC 9812 (GCF_000187265.1), and Streptococcus gallolyticus ATCC 43143

75 299 (GCF_000270145.1). For each antagonist, several BGCs were found (Table 6). For B.

300 vulgatus, among 18 BGCs, there were three polysaccharide B, one lipopolysaccharide, one

301 putative coelimycin, and 12 BGCs of unknown function. For E. dispar, among 7 BGCs, there

302 were one R1128 polyketide pathway BGC, four BGCs of unknown function and two other

303 BGCs. For P. avidum, among 16 BGCs, there were one bacteriocin, one kosinostatin, one

304 phosphonoglycans, one chlorizidine A, one svaricin BGCs, 8 BGCs of unknown function and

305 three other BGC. For B. fragilis, among 25 BGCs there were two exopolysaccharide, one

306 polysaccharide A, three polysaccharide B and C, 15 BGCs of unknown function and one other

307 BGC. For L. delbrueckii among six BGCs, there were two exopolysaccharide, one fusaricidin,

308 and three BGCs of unknown function. For P. acidipropionici among 24 BGCs, there were one

309 bacteriocin, one exopolysaccharide, one phosphonoglycans, one chlorizidine A, one svaricin,

310 14 BGCs of unknown function and five other BGC. For S. equinus among five BGCs, there

311 were one streptomycin, one exopolysaccharide, one dutomycin fatty acid and two BGCs of

312 unknown function. For S. gallolyticus among nine BGCs, there were one bacteriocin, were

313 one streptomycin, one exopolysaccharide, one R1128 polyketide pathway BGC, one

314 maklamicin and four BGCs of unknown function.

315 4. CONCLUSION AND DISCUSSION

316 Putative antagonists of human pathogen species were selected from metagenomic data

317 of human gut microbiota. Testing these antagonisms using culture method allow us to find

318 nine species that inhibit in vitro the growth of S. aureus, E. cloacae and E. aerogenes. These

319 spectrums of activity are probably related to the secretion of bacteriocins. Indeed, except for

320 E. dispar that exhibited a cross inhibition reaction of E. aerogenes and E. cloacae, no other

321 cross reaction were observed, even among the Enterobacteriacae. Moreover, the absence of

322 cross reaction limits the possibility of pH interaction or another artefact. All antagonisms are

323 reproductible since they were performed at least twice.

76 324 We focused in this study on the gut microbiota commensals as natural source of

325 antimicrobials. Indeed, more than one million of biosynthetic gene clusters suspected to code

326 for secondary metabolites were identified from gut microbiome [7,23]. According to Zheng et

327 al., the gut is the second most abundant source of bacteriocins of human microbiome, after the

328 oral cavity [24]. Antimicrobial testing was performed using the well-diffusion method, that is

329 one of the most appropriate for in vitro testing of antibiotic production for decades [11].

330 However, our results showed that the use of broth is more sensible than the cell free

331 supernatant (CFC) for screening purpose, since only one CFS inhibited the growth of the

332 pathogenic species.

333 We found that S. aureus growth was inhibited by P. avidum. This antagonism was

334 unknown from culture. The importance of coproporphyrin III naturally synthesized by

335 Propionibacterium sp. was previously reported as aggreging factor of S. aureus [25]. We

336 found 16 BGCs including notably one bacteriocin of unknown class and one Kosinostatin

337 BGC, a quinocycline antibiotic which was previously described as a strong Gram-positive

338 inhibitor particularly against S. aureus [26,27]. In our work, E. cloacae growth was inhibited

339 by B. fragilis, E. dispar, L. delbruckii, P. acidipropionici, S. equinus and S. gallolyticus. No

340 specific antagonism between these strains is published except for L. delbrueckii that is known

341 to inhibit the growth of E. cloacae and E. aerogenes, as well as other coliform strains [28].

342 However, many BGCs including bacteriocins and putative antibiotics were found in the

343 genome. Finally, we found two antagonisms of E. aerogenes: B. vulgatus and E. dispar. Many

344 BGCs coding for antibiotics were also found for these species and 36 (29%) of all BGCs

345 found were unknown.

346 The antimicrobial molecule of these species remains to be isolated and purified.

347 Indeed, the characterization of their biochemical nature, their effects and of their spectrum of

348 activity is crucial. The whole genome sequencing of these strains will allow us to elucidate

77 349 the mechanisms of action. In conclusion, these preliminary results highlight the great potential

350 of the gut microbiota commensals species for the new antibiotic research. This driven

351 approach using 70 metagenomes allowed us to focus on putative antagonisms using an easy and

352 low cost culture antagonism method.

78 353 References

354 [1] World Health Organization, Antimicrobial resistance, Factsheet 194, (2016).

355 http://www.who.int/mediacentre/factsheets/fs194/en/ (accessed October 30, 2018).

356 [2] European Society of Clinical Microbiology and Infectious Diseases, ESCMID warns:

357 Europe may surpass one million deaths due to ineffective antibiotics by 2025, (2015).

358 http://semmelweis.info/escmid-warns-europe-may-surpass-one-million-deaths-due-to-

359 ineffective-antibiotics-by-2025/ (accessed October 30, 2018).

360 [3] World Health Organization, Antibacterial agents in clinical development, (2017).

361 http://apps.who.int/iris/bitstream/handle/10665/258965/WHO-EMP-IAU-2017.11-

362 eng.pdf;jsessionid=1D402A86A40175305D37CE8454A1C0EB?sequence=1 (accessed

363 November 15, 2018).

364 [4] G.A. Durand, D. Raoult, G. Dubourg, Antibiotic discovery: history, methods and

365 perspectives, Int. J. Antimicrob. Agents. (Submitted).

366 [5] L.L. Ling, T. Schneider, A.J. Peoples, A.L. Spoering, I. Engels, B.P. Conlon, A. Mueller,

367 T.F. Schäberle, D.E. Hughes, S. Epstein, M. Jones, L. Lazarides, V.A. Steadman, D.R.

368 Cohen, C.R. Felix, K.A. Fetterman, W.P. Millett, A.G. Nitti, A.M. Zullo, C. Chen, K.

369 Lewis, A new antibiotic kills pathogens without detectable resistance, Nature. 517 (2015)

370 455–459. doi:10.1038/nature14098.

371 [6] A. Zipperer, M.C. Konnerth, C. Laux, A. Berscheid, D. Janek, C. Weidenmaier, M.

372 Burian, N.A. Schilling, C. Slavetinsky, M. Marschal, M. Willmann, H. Kalbacher, B.

373 Schittek, H. Brötz-Oesterhelt, S. Grond, A. Peschel, B. Krismer, Human commensals

374 producing a novel antibiotic impair pathogen colonization, Nature. 535 (2016) 511–516.

375 doi:10.1038/nature18634.

376 [7] M.S. Donia, P. Cimermancic, C.J. Schulze, L.C. Wieland Brown, J. Martin, M. Mitreva,

377 J. Clardy, R.G. Linington, M.A. Fischbach, A systematic analysis of biosynthetic gene

79 378 clusters in the human microbiome reveals a common family of antibiotics, Cell. 158

379 (2014) 1402–1414. doi:10.1016/j.cell.2014.08.032.

380 [8] E. Angelakis, V. Merhej, D. Raoult, Related actions of probiotics and antibiotics on gut

381 microbiota and weight modification, Lancet Infect. Dis. 13 (2013) 889–899.

382 doi:10.1016/S1473-3099(13)70179-8.

383 [9] J.-C. Lagier, G. Dubourg, M. Million, F. Cadoret, M. Bilen, F. Fenollar, A. Levasseur, J.-

384 M. Rolain, P.-E. Fournier, D. Raoult, Culturing the human microbiota and culturomics,

385 Nat. Rev. Microbiol. (2018). doi:10.1038/s41579-018-0041-0.

386 [10] E. Angelakis, M. Yasir, D. Bachar, E.I. Azhar, J.-C. Lagier, F. Bibi, A.A. Jiman-

387 Fatani, M. Alawi, M.A. Bakarman, C. Robert, D. Raoult, Gut microbiome and dietary

388 patterns in different Saudi populations and monkeys, Sci. Rep. 6 (2016) 32191.

389 doi:10.1038/srep32191.

390 [11] M. Balouiri, M. Sadiki, S.K. Ibnsouda, Methods for in vitro evaluating antimicrobial

391 activity: A review, J. Pharm. Anal. 6 (2016) 71–79. doi:10.1016/j.jpha.2015.11.005.

392 [12] K. Blin, T. Wolf, M.G. Chevrette, X. Lu, C.J. Schwalen, S.A. Kautsar, H.G. Suarez

393 Duran, E.L.C. de Los Santos, H.U. Kim, M. Nave, J.S. Dickschat, D.A. Mitchell, E.

394 Shelest, R. Breitling, E. Takano, S.Y. Lee, T. Weber, M.H. Medema, antiSMASH 4.0-

395 improvements in chemistry prediction and gene cluster boundary identification, Nucleic

396 Acids Res. 45 (2017) W36–W41. doi:10.1093/nar/gkx319.

397 [13] P. Cimermancic, M.H. Medema, J. Claesen, K. Kurita, L.C. Wieland Brown, K.

398 Mavrommatis, A. Pati, P.A. Godfrey, M. Koehrsen, J. Clardy, B.W. Birren, E. Takano, A.

399 Sali, R.G. Linington, M.A. Fischbach, Insights into secondary metabolism from a global

400 analysis of prokaryotic biosynthetic gene clusters, Cell. 158 (2014) 412–421.

401 doi:10.1016/j.cell.2014.06.034.

80 402 [14] E. Huang, A.E. Yousef, Biosynthesis of paenibacillin, a lantibiotic with N-terminal

403 acetylation, by Paenibacillus polymyxa, Microbiol. Res. 181 (2015) 15–21.

404 doi:10.1016/j.micres.2015.08.001.

405 [15] C.T. Lohans, Z. Huang, M.J. van Belkum, M. Giroud, C.S. Sit, E.M. Steels, J. Zheng,

406 R.M. Whittal, L.M. McMullen, J.C. Vederas, Structural characterization of the highly

407 cyclized lantibiotic paenicidin A via a partial desulfurization/reduction strategy, J. Am.

408 Chem. Soc. 134 (2012) 19540–19543. doi:10.1021/ja3089229.

409 [16] J.-E. Park, H.-R. Kim, S.-Y. Park, S.-K. Choi, S.-H. Park, Identification of the

410 biosynthesis gene cluster for the novel lantibiotic paenilan from Paenibacillus polymyxa

411 E681 and characterization of its product, J. Appl. Microbiol. 123 (2017) 1133–1147.

412 doi:10.1111/jam.13580.

413 [17] J. Sánchez, D.B. Diep, C. Herranz, I.F. Nes, L.M. Cintas, P.E. Hernández, Amino acid

414 and nucleotide sequence, adjacent genes, and heterologous expression of hiracin JM79, a

415 sec-dependent bacteriocin produced by Enterococcus hirae DCH5, isolated from Mallard

416 ducks (Anas platyrhynchos), FEMS Microbiol. Lett. 270 (2007) 227–236.

417 doi:10.1111/j.1574-6968.2007.00673.x.

418 [18] A. Gupta, S.K. Tiwari, V. Netrebov, M.L. Chikindas, Biochemical Properties and

419 Mechanism of Action of Enterocin LD3 Purified from Enterococcus hirae LD3,

420 Probiotics Antimicrob. Proteins. 8 (2016) 161–169. doi:10.1007/s12602-016-9217-y.

421 [19] R. Daillère, M. Vétizou, N. Waldschmitt, T. Yamazaki, C. Isnard, V. Poirier-Colame,

422 C.P.M. Duong, C. Flament, P. Lepage, M.P. Roberti, B. Routy, N. Jacquelot, L. Apetoh,

423 S. Becharef, S. Rusakiewicz, P. Langella, H. Sokol, G. Kroemer, D. Enot, A. Roux, A.

424 Eggermont, E. Tartour, L. Johannes, P.-L. Woerther, E. Chachaty, J.-C. Soria, E. Golden,

425 S. Formenti, M. Plebanski, M. Madondo, P. Rosenstiel, D. Raoult, V. Cattoir, I.G.

426 Boneca, M. Chamaillard, L. Zitvogel, Enterococcus hirae and Barnesiella

81 427 intestinihominis Facilitate Cyclophosphamide-Induced Therapeutic Immunomodulatory

428 Effects, Immunity. 45 (2016) 931–943. doi:10.1016/j.immuni.2016.09.009.

429 [20] A.J. Salle, G.J. Jann, Subtilin, an antibiotic produced by Bacillus subtilis; action on

430 various organisms, Proc. Soc. Exp. Biol. Med. Soc. Exp. Biol. Med. N. Y. N. 60 (1945)

431 60–64.

432 [21] T. Stein, S. Borchert, B. Conrad, J. Feesche, B. Hofemeister, J. Hofemeister, K.-D.

433 Entian, Two different lantibiotic-like peptides originate from the ericin gene cluster of

434 Bacillus subtilis A1/3, J. Bacteriol. 184 (2002) 1703–1711.

435 [22] M. de Souza Barbosa, S.D. Todorov, I. Ivanova, J.-M. Chobert, T. Haertlé, B.D.G. de

436 Melo Franco, Improving safety of salami by application of bacteriocins produced by an

437 autochthonous Lactobacillus curvatus isolate, Food Microbiol. 46 (2015) 254–262.

438 doi:10.1016/j.fm.2014.08.004.

439 [23] J.M. Luber, A.D. Kostic, Gut Microbiota: Small Molecules Modulate Host Cellular

440 Functions, Curr. Biol. CB. 27 (2017) R307–R310. doi:10.1016/j.cub.2017.03.026.

441 [24] J. Zheng, M.G. Gänzle, X.B. Lin, L. Ruan, M. Sun, Diversity and dynamics of

442 bacteriocins from human microbiome, Environ. Microbiol. 17 (2015) 2133–2143.

443 doi:10.1111/1462-2920.12662.

444 [25] M.S. Wollenberg, J. Claesen, I.F. Escapa, K.L. Aldridge, M.A. Fischbach, K.P.

445 Lemon, Propionibacterium-Produced Coproporphyrin III Induces Staphylococcus aureus

446 Aggregation and Biofilm Formation, MBio. 5 (2014). doi:10.1128/mBio.01286-14.

447 [26] T. Furumai, Y. Igarashi, H. Higuchi, N. Saito, T. Oki, Kosinostatin, a quinocycline

448 antibiotic with antitumor activity from Micromonospora sp. TP-A0468, J. Antibiot.

449 (Tokyo). 55 (2002) 128–133.

82 450 [27] V. Rambabu, S. Suba, S. Vijayakumar, Antimicrobial and antiproliferative prospective

451 of kosinostatin - a secondary metabolite isolated from Streptomyces sp, J. Pharm. Anal. 5

452 (2015) 378–382. doi:10.1016/j.jpha.2014.11.002.

453 [28] F. Savino, L. Cordisco, V. Tarasco, E. Locatelli, D. Di Gioia, R. Oggero, D.

454 Matteuzzi, Antagonistic effect of Lactobacillus strains against gas-producing coliforms

455 isolated from colicky infants, BMC Microbiol. 11 (2011) 157. doi:10.1186/1471-2180-11-

456 157.

457

83 458 Acknowledgements and financial disclosure

459 This work was supported by the French Government under the « Investissements d’avenir »

460 (Investments for the Future) a program managed by the Agence Nationale de la Recherche

461 (ANR, fr: National Agency for Research), (reference: Méditerranée Infection 10-IAHU-03).

462 Table legend

463 Table 1. Characteristics of the metagenomic samples used for the selection of putative

464 antagonisms. Control sample were healthy adult donors. Clostridium difficile infection group

465 were patients with documented C. difficile infection.

466 Table 2. Inhibition diameter (in millimeter) of Paenibacillus polymyxa and Bacillus subtilis

467 against indicative strain for the proof of concept.

468 Table 3. Inhibition diameter (in millimeter) of Enterococcus hirae strains against indicative

469 strain for the proof of concept.

470 Table 4. Results of inhibition diameter (in millimeter) of the tested strains against indicative

471 strains.

472 Table 5. Results of inhibition diameter (in millimeter) of the cell-free supernatants (CFS) of

473 the antagonisms against indicative strains.

474 Table 6. Results of the research for Biosynthetic gene clusters (BGCs) and their assignment

475 using Antismash 4.2.0 (K. Blin et al., 2017).

476 Figure legend

477 Figure 1. Bacterial and fungi isolated in 2017 from blood samples in the clinical

478 microbiology laboratory of IHU Mediterranée Infection.

479 Figure 2. Design of the metagenomic study for the selection of putative antagonisms.

84 480 Figure 3. Well-diffusion method for antagonism testing.

481 Figure 4. Cross-streak and spot-on-lawn results for Paenibacillus polymyxa and Enterococcus

482 hirae (proof of concept)

483 Supplementary Figures

484 Suppl. Figure 1. Result of the research for putative antagonists against Staphylococcus

485 aureus

486 Suppl. Figure 2. Result of the research for putative antagonists against Klebsiella

487 pneumoniae

488 Suppl. Figure 3. Result of the research for putative antagonists against Enterobacter

489 aerogenes

490 Suppl. Figure 4. Result of the research for putative antagonists against Enterobacter cloacae

491 Suppl. Figure 5. Result of the research for putative antagonists against Enterococcus faecalis

492 Suppl. Figure 6. Result of the research for putative antagonists against Pseudomonas

493 aeruginosa

494 Suppl. Figure 7. Result of the research for putative antagonists against Escherichia coli and

495 Streptococcus pneumoniae

496 Suppl. Figure 8. Result of the research for putative antagonists against Streptococcus

497 pyogenes

498 Suppl. Figure 9. Result of the research for culture antagonism of Paenibacillus polymyxa

499 P3562

500 Suppl. Figure 10. Result of the research for culture antagonism of Enterococcus hirae strains

501 IGR7, IGR4 and P3144 against E. faecalis P2525

85 502 Suppl. Figure 11. Result of the research for culture antagonism of Bacillus subtilis strains

503 P3435, P3433 and P8403 against Staphylococcus aureus P8402

504 Suppl. Figure 12. Antagonisms found using well diffusion method and the results of their

505 Cell-free supernatant

86 Table 1

Control samples CDI samples Metagenome No. reads No. OTUs Shannon Metagenome No. reads No. OTUs Shannon Control_1 60614 266 2,94 Colitis_1 11364 118 2,65 Control_2 24094 249 3,37 Colitis_2 32879 380 3,83 Control_3 33350 166 3,06 Colitis_3 60133 277 3,09 Control_4 60923 228 2,52 Colitis_4 17204 140 2,85 Control_5 102335 172 1,87 Colitis_5 41057 324 3,45 Control_6 48019 257 2,84 Colitis_6 27931 70 1,87 Control_7 86265 209 2,57 Colitis_7 29956 179 3,00 Control_8 84626 172 2,12 Colitis_8 54858 116 1,78 Control_9 139050 284 2,65 Colitis_9 45700 173 2,19 Control_10 54825 103 1,29 Colitis_10 111215 305 3,21 Control_11 150843 241 2,26 Colitis_11 101800 458 2,99 Control_12 94518 330 3,38 Colitis_12 63972 197 2,92 Control_13 88141 214 2,66 Colitis_13 66636 188 2,44 Control_14 63667 413 3,84 Colitis_14 61106 165 1,84 Control_15 33052 190 1,67 Colitis_15 44527 231 2,80 Control_16 65508 387 3,76 Colitis_16 132062 497 3,94 Control_17 59141 376 4,33 Colitis_17 44120 159 1,81 Control_18 29902 283 3,74 Colitis_18 201928 352 3,34 Control_19 84735 375 4,02 Colitis_19 105626 140 1,86 Control_20 70657 396 3,98 Colitis_20 137191 137 2,28 Control_21 82206 360 4,09 Colitis_21 128412 387 3,09 Control_22 141831 510 4,25 Colitis_22 252595 362 2,85 Control_23 156858 325 3,57 Colitis_23 211310 215 2,31 Control_24 144434 547 4,49 Colitis_24 78207 143 1,98 Control_25 212030 389 3,85 Colitis_25 98177 219 2,57 Control_26 118902 462 3,95 Colitis_26 79751 274 2,25 Control_27 200949 450 4,35 Colitis_27 211312 320 2,17 Control_28 155615 444 4,39 Colitis_28 205370 203 1,36 Control_29 157874 519 4,10 Colitis_29 159761 136 1,16 Control_30 132440 364 2,64 Colitis_30 259084 175 2,01 Control_31 80180 522 4,25 Control_32 156516 664 4,59 Control_33 847692 295 2,97 Control_34 120597 472 3,83 Control_35 99478 438 2,19 Control_36 181302 566 4,18 Control_37 111066 555 4,03 Control_38 310523 362 3,57 Control_39 602677 348 3,04 Control_40 330204 403 4,10 Mean 144440,98 358 3,38 102508,13 235 2,53 CDI: Clostridium difficile infection No. Number of.

87 Table 2

Tested strain Indicator strain DOI (mm) 106 CFU/mL Paenibacillus polymyxa P3562 Enterococcus faecalis P2525 19 Paenibacillus polymyxa P3562 Listeria monocytogenes P828 17 Paenibacillus polymyxa P3562 Staphylococcus aureus P8402 20 Paenibacillus polymyxa P3562 Escherichia coli P2585 18 Paenibacillus polymyxa P3562 Pseudomonas aeruginosa P2283 14 Paenibacillus polymyxa P3562 Proteus vulgaris P8722 10 Paenibacillus polymyxa P3562 Klebsiella pneumoniae P2577 19 Paenibacillus polymyxa P3562 Enterobacter aerogenes 8401 20 Bacillus subtilis P3433 Staphylococcus aureus P8402 12 Bacillus subtilis P3433 Staphylococcus aureus P8402 9 Bacillus subtilis P3433 Staphylococcus aureus P8402 14 DOI : Diameter of inhibition (mm)

88 Table 3

DOI (mm) 106 DOI (mm) 108 Test Indicator strain CFU/mL CFU/mL Enterococcus hirae P5207 Enterococcus faecium P2282 19 NA Enterococcus hirae P5207 Enterococcus faecalis P2525 12 NA Enterococcus hirae P5207 Listeria monocytogenes P828 25 NA Enterococcus hirae IGR7 Enterococcus faecalis P2525 11 14 Enterococcus hirae IGR7 Enterococcus faecalis P720 11 NA Enterococcus hirae IGR7 Enterococcus faecalis P2282 10 NA Enterococcus hirae IGR7 Enterococcus durans P8400 11 NA Enterococcus hirae IGR7 Listeria monocytogenes P828 15 NA Enterococcus hirae IGR7 Enterococcus casseliflavus P3601 9 NA Enterococcus hirae IGR7 Enterococcus dispar P5040 9 NA Enterococcus hirae IGR7 Enterococcus gallinarum P364 9 NA Enterococcus hirae IGR7 Staphylococcus aureus P8402 9 NA Enterococcus hirae IGR7 Staphylococcus carnosus P1892 9 NA Enterococcus hirae IGR7 Staphylococcus devriesi P4716 9 NA Enterococcus hirae IGR7 Staphylococcus kloosi P837 9 NA Enterococcus hirae IGR7 Staphylococcus schleiferi P4020 9 NA Enterococcus hirae IGR7 Enterobacter aerogenes P8401 9 NA Enterococcus hirae IGR7 Escherichia coli P2585 9 NA Enterococcus hirae IGR7 Pseudomonas aeruginosa P2283 9 NA Enterococcus hirae IGR7 Proteus vulgaris P8722 9 NA Enterococcus hirae IGR7 Salmonella enteridis 9 NA Enterococcus hirae IGR7 Klebsiella pneumoniae P2577 9 NA Enterococcus hirae P13144 Enterococcus faecalis P2525 11 14 Enterococcus hirae P13144 Enterococcus faecalis P720 10 NA Enterococcus hirae P13144 Enterococcus durans P8400 10 NA Enterococcus hirae P13144 Listeria monocytogenes P828 13 NA Enterococcus hirae P13144 Enterococcus casseliflavus P3601 9 NA Enterococcus hirae P13144 Enterococcus dispar P5040 9 NA Enterococcus hirae P13144 Enterococcus gallinarum P364 9 NA Enterococcus hirae P13144 Staphylococcus aureus P8402 9 NA Enterococcus hirae P13144 Staphylococcus carnosus P1892 9 NA Enterococcus hirae P13144 Staphylococcus devriesi P4716 9 NA Enterococcus hirae P13144 Staphylococcus kloosi P837 9 NA Enterococcus hirae P13144 Staphylococcus schleiferi P4020 9 NA Enterococcus hirae P13144 Enterobacter aerogenes P8401 9 NA Enterococcus hirae P13144 Escherichia coli P2285 9 NA Enterococcus hirae P13144 Pseudomonas aeruginosa P2283 9 NA Enterococcus hirae IGR7 Salmonella enteridis 9 NA Enterococcus hirae P13144 Proteus vulgaris P8722 9 NA Enterococcus hirae P13144 Klebsiella pneumoniae P2577 9 NA Enterococcus hirae P1518 Enterococcus faecalis P2525 9 NA Enterococcus hirae P1519 Enterococcus faecalis P2525 9 NA Enterococcus hirae P1536 Enterococcus faecalis P2525 9 NA Enterococcus hirae P1537 Enterococcus faecalis P2525 9 NA Enterococcus hirae P3553 Enterococcus faecalis P2525 9 NA Enterococcus hirae P557 Enterococcus faecalis P2525 9 NA Enterococcus hirae P4115 Enterococcus faecalis P2525 9 NA Enterococcus hirae P4871 Enterococcus faecalis P2525 9 NA Enterococcus hirae P3548 Enterococcus faecalis P2525 9 NA Enterococcus hirae P5364 Enterococcus faecalis P2525 9 NA Enterococcus hirae P4488 Enterococcus faecalis P2525 9 NA Enterococcus hirae P4127 Enterococcus faecalis P2525 9 NA DOI : Diameter of inhibition (mm)

89 Table 4

DOI (mm) 106 DOI (mm) 108 Tested strain CSUR Indicator strain CFU/mL CFU/mL Bacteroides vulgatus P2616 Enterobacter aerogenes 16 10 Bifidobacterium catenulatum P5445 Enterobacter aerogenes 9 9 Clostridium symbiosum P7105 Enterobacter aerogenes 9 9 Enterococcus dispar P5040 Enterobacter aerogenes 12 9 Hungatella hathewayi P4379 Enterobacter aerogenes 9 9 Lactobacillus johnsonii P7075 Enterobacter aerogenes 9 9 Lactobacillus rhamnosus P567 Enterobacter aerogenes 9 9 Methanobrevibacter smithii Enterobacter aerogenes 9 9 Proteus mirabilis P2540 Enterobacter aerogenes 9 9 Pseudomonas putida P4293 Enterobacter aerogenes 9 9 Solibacillus silvestris P5387 Enterobacter aerogenes 9 9 Streptococcus equinus P2987 Enterobacter aerogenes 9 9 Veillonella parvula P2874 Enterobacter aerogenes 9 9 Acidaminococcus intestini P5762 Enterococcus faecalis 9 9 Actinomyces naeslundii P2751 Enterococcus faecalis 9 9 Actinomyces oris P3221 Enterococcus faecalis 9 9 Bacteroides finegoldii P2615 Enterococcus faecalis 9 9 Bacteroides fragilis P2573 Enterococcus faecalis 9 9 Bacteroides mediterraneensis P2644 Enterococcus faecalis 9 9 Bifidobacterium pseudocatenulatum P3697 Enterococcus faecalis 9 9 Campylobacter gracilis P3167 Enterococcus faecalis 9 9 Clostridium perfringens P6247 Enterococcus faecalis 9 9 Clostridium saccharogumia P3138 Enterococcus faecalis 9 9 Gordonibacter pamelaeae Enterococcus faecalis 9 9 Lachnoclostridium edouardii P6493 Enterococcus faecalis 9 9 Lactobacillus johnsonii P7075 Enterococcus faecalis 9 9 Lactobacillus mucosae P5355 Enterococcus faecalis 9 9 Lactobacillus parabuchneri P6269 Enterococcus faecalis 9 9 Megasphaera massiliensis P245 Enterococcus faecalis 9 9 Raoultella ornithinolytica P3384 Enterococcus faecalis 9 9 Clostridium sporogenes P682 Enterococcus faecalis 9 9 Anaerococcus murdochii P4938 Klebsiella pneumoniae 9 9 Bacteroides mediterraneensis P2644 Klebsiella pneumoniae 9 9 Bifidobacterium longum subsp. infantis P4929 Klebsiella pneumoniae 9 9 Enterococcus pseudoavium P1289 Klebsiella pneumoniae 9 9 Helcococcus kunzii P4909 Klebsiella pneumoniae 9 9 Lactobacillus crispatus P6380 Klebsiella pneumoniae 9 9 Porphyromonas asaccharolytica P716 Klebsiella pneumoniae 9 9 Solibacillus silvestris P5387 Klebsiella pneumoniae 9 9 Citrobacter freundii P2486 Pseudomonas aeruginosa 9 9 Clostridium spiroforme P4351 Pseudomonas aeruginosa 9 9 Corynebacterium simulans P4763 Pseudomonas aeruginosa 9 9 Dermabacter hominis P2881 Pseudomonas aeruginosa 9 9 Enterococcus gallinarum P364 Pseudomonas aeruginosa 9 9 Fusobacterium nucleatum P6559 Pseudomonas aeruginosa 9 9 Peptoniphilus gorbachii P1050 Pseudomonas aeruginosa 9 9 Porphyromonas asaccharolytica P716 Pseudomonas aeruginosa 9 9 Shewanella algae P601 Pseudomonas aeruginosa 9 9 Solibacillus silvestris P5387 Pseudomonas aeruginosa 9 9 Solobacterium moorei P6725 Pseudomonas aeruginosa 9 9 Streptococcus equinus P2987 Pseudomonas aeruginosa 9 9 Lactobacillus rhamnosus P567 Pseudomonas aeruginosa 9 9 Acidaminococcus intestini P5762 Staphylococcus aureus 9 9 Actinomyces naeslundii P2751 Staphylococcus aureus 9 9 Bacteroides finegoldii P2615 Staphylococcus aureus 9 9 Campylobacter gracilis P3776 Staphylococcus aureus 9 9

90 Clostridium butyricum P0071 Staphylococcus aureus 9 9 Emergencia timonensis P2260 Staphylococcus aureus 9 9 Fusobacterium necrophorum P1016 Staphylococcus aureus 9 9 Lactobacillus johnsonii P7075 Staphylococcus aureus 9 9 Parabacteroides gordonii P2178 Staphylococcus aureus 9 9 Propionibacterium avidum P6585 Staphylococcus aureus 12 10 Pseudomonas putida P4293 Staphylococcus aureus 9 9 Rummeliibacillus pycnus P819 Staphylococcus aureus 9 9 Stenotrophomonas maltophilia P1004 Staphylococcus aureus 9 9 Terrisporobacter glycolicus P846 Staphylococcus aureus 9 9 Acidaminococcus intestini P5762 Enterobacter cloacae 9 9 Bacteroides fragilis P2573 Enterobacter cloacae 15 12 Bacteroides ovatus P4189 Enterobacter cloacae 9 9 Clostridium perfringens P6247 Enterobacter cloacae 9 9 Clostridium scindens P6814 Enterobacter cloacae 9 9 Enterococcus dispar P5040 Enterobacter cloacae 16 11 Holdemania filiformis Enterobacter cloacae 9 9 Lactobacillus delbrueckii P4496 Enterobacter cloacae 12 9 Lactobacillus johnsonii P7075 Enterobacter cloacae 9 9 Megasphaera massiliensis P245 Enterobacter cloacae 9 9 Porphyromonas asaccharolytica P716 Enterobacter cloacae 9 9 Propionibacterium acidipropionici P2295 Enterobacter cloacae 15 10 Streptococcus equinus P2987 Enterobacter cloacae 15 10 Streptococcus gallolyticus P6782 Enterobacter cloacae 16 10 Veillonella ratti P6972 Enterobacter cloacae 9 9 DOI: Diameter of inhibition (mm)

91 Table 5

DOI (mm) of CFS DOI (mm) of CFS Tested strain Indicator strain 106 CFU/mL 108 CFU/mL pH of CFS Bacteroides vulgatus P2616 Enterobacter aerogenes 9 9 6.35 Enterococcus dispar P5040 Enterobacter aerogenes 9 9 7.01 Propionibacterium avidum P6585 Staphylococcus aureus 9 9 7.12 Bacteroides fragilis P2573 Enterobacter cloacae 12 10 6.51 Enterococcus dispar P5040 Enterobacter cloacae 9 9 7.01 Lactobacillus delbrueckii P4496 Enterobacter cloacae 9 9 6.73 Propionibacterium acidipropionici P2295 Enterobacter cloacae 9 9 5.63 Streptococcus equinus P2987 Enterobacter cloacae 9 9 5.59 Streptococcus gallolyticus P6782 Enterobacter cloacae 9 9 5.54 DOI: Diameter of inhibition (mm) CFS: Cell free supernatant

92 Table 6

Species B. vulgatus E. dispar P. avidum B. fragilis L. delbrueckii P. acidipropionici S. equinus S. gallolyticus BGCs 18 7 16 25 6 24 5 9 NRPS 0 0 0 0 0 0 0 0 PKS 0 0 0 0 0 0 0 0 Bacteriocin 0 0 1 0 0 1 0 1 Saccharides 11 1 3 14 2 5 3 4 Streptomycin 0 0 0 0 0 0 1 1 Exopolysaccharide 0 0 0 2 2 1 1 1 Polysaccharide A 0 0 0 1 0 0 0 0 Polysaccharide B 3 0 0 3 0 0 0 0 Polysaccharide C 0 0 0 3 0 0 0 0 Lipopolysaccharide 1 0 0 0 0 0 0 0 Kosinostatin 0 0 1 0 0 0 0 0 Phosphonoglycans 0 0 1 0 0 1 0 0 Other 0 0 0 1 0 0 0 0 Unknown 7 1 1 4 0 3 1 2 Fatty acid 2 1 1 3 1 1 1 1 Dutomycin 0 0 0 0 0 0 1 0 Fusaricidin 0 0 0 0 1 0 0 0 Chlorizidine A 0 0 1 0 0 1 0 0 R1128 PK Pathway 0 1 0 0 0 0 0 1 Unknown 2 0 0 3 0 0 0 0 Arylpolyene 1 0 0 0 0 0 0 0 Putative classes 4 5 11 8 3 17 1 3 Svaricin 0 0 1 0 0 1 0 0 Coelimycin 1 0 0 0 0 0 0 0 Maklamicin 0 0 0 0 0 0 0 1 Other 0 2 3 0 0 5 0 0 Unknown 3 3 7 8 3 11 1 2

93

Fig. 3

Well diffusion

Incubation Rest time

18h Lawn of the indicator strain A cylinder of agar Culture supernatant or culture broth is punched-out of the tested bacteria

Indicator strain Reading of inhibition diameter Tested strain

96 Cross streak Spot-on-the-lawn Fig. 4 A B C

P. polymyxa

P. polymyxa E. faecalis P2525

Negative S. aureus E. hirae P5207 S. aureus 106 CFU/m D E F E. hirae IGR7

E. faecalis P2525 E. faecalis P2525

E. hirae P5207 E. hirae 13144

Negative E. hirae 13144 E. hirae97IGR7 E. faecalis P2525 106 CFU/mL

Suppl. Fig. 9

Paenibacillus polymyxa P3562

Escherichia coli P2585 Klebsiella pneumoniae P2577

Staphylococcus aureus P8402 Pseudomonas aeruginosa P2283 Proteus vulgaris P8722

Enterococcus faecalis P2525 Listeria monocytogenes P825 Enterobacter aerogenes P8401 106 Suppl. Fig. 10

E. hirae P5207

E. hirae IGR4 E. hirae IGR7 E. hirae P4115 E. hirae P3548

E. hirae P3144 E. hirae P584

E. hirae P4971

Negative Negative

E. faecalis P2525 106 CFU/mL E. faecalis P2525 106 CFU/mL 107 Suppl. Fig. 11

Bacillus subtilis Bacillus subtilis P3433 P3433

Bacillus subtilis Bacillus subtilis Bacillus subtilis Bacillus subtilis P3435 P8403 P3435 P8403

Negative Negative

S. aureus 106 CFU/mL S. aureus 108 CFU/mL

108 Suppl. Fig. 12 Well diffusion of strain Well diffusion of CFS Tested strain 6 8 10 CFU/mL 10 CFU/mL 106 CFU/mL 108 CFU/mL

EC Bacteroides vulgatus

Enterobacter aerogenes Enterobacter aerogenes

Enterococcus dispar

Enterobacter aerogenes Enterobacter aerogenes

Propionibacterium avidum

Staphylococcus aureus Staphylococcus aureus 109 Tested strain Well diffusion of strain Well diffusion of CFS 106 CFU/mL 108 CFU/mL 106 CFU/mL 108 CFU/mL

Bacteroides fragilis

Enterobacter cloacae Enterobacter cloacae

Enterococcus dispar

Enterobacter cloacae Enterobacter cloacae

Lactobacillus delbrueckii

Enterobacter cloacae Enterobacter cloacae

Propionibacterium acidipropionici

110 Enterobacter cloacae Enterobacter cloacae Tested strain Well diffusion of strain Well diffusion of CFS 106 CFU/mL 108 CFU/mL 106 CFU/mL 108 CFU/mL

Streptococcus equinus

Enterobacter cloacae Enterobacter cloacae

Streptococcus gallolyticus

Enterobacter cloacae Enterobacter cloacae

Hungatella hathewayii

Enterobacter cloacae Enterobacter cloacae

Veillonella parvula

111 Enterobacter cloacae Enterobacter cloacae Tested strain Well diffusion of strain Well diffusion of CFS 106 CFU/mL 108 CFU/mL 106 CFU/mL 108 CFU/mL

Bacteroides vulgatus EC

Enterobacter cloacae Enterobacter cloacae

Lactobacillus rhamnosus EC

Enterobacter cloacae Enterobacter cloacae

112

Partie II :

Culturomics et incompatibilités de culture

113 Avant-propos

La culturomics est une approche innovante dans l’étude du microbiote digestif. Elle consiste en la multiplication des conditions de culture à partir d’un échantillon complexe, dans le but d’isoler le maximum d’espèces bactériennes différentes. L’utilisation d’antioxydants et de jus de rumen figurent parmi les conditions les plus contributives. Les antioxydants (acide ascorbique, acide urique, glutathion) ont en effet permis de cultiver des espèces intolérantes à l’oxygène (par exemple Bacteroides fragilis ou Clostridium sporogenes) en condition aérobies, ce qui a donné naissance au milieu de culture R-medium

(21).

Durant mon travail de thèse, j’ai travaillé en particulier sur des selles fraîchement

émises. L’idée était de cultiver plus de bactéries anaérobies en réduisant au maximum le temps d’exposition à l’oxygène entre l’émission de la selle et son incubation. Les repiquages se faisaient en enceinte anaérobie. Les conditions utilisées ciblaient la croissance des bactéries anaérobies : utilisation d’antioxydants, de rumen, flacon d’hémoculture anaérobies. Ce travail a participé à l’accroissement du répertoire digestif et est présenté dans la troisième publication. Les nouvelles espèces isolées lors de ce travail ont été décrite par taxono- génomique et ont fait l’objet de publications présentées en Annexe.

Malgré l’explosion du nombre d’espèces bactériennes isolées dans le microbiote digestif humain grâce à la culturomics, certaines espèces fastidieuses demeurent peu ou non isolées. Dans la quatrième publication, nous avons effectué une analyse métagénomique et culturomics d’une selle fraichement émise avant et après incubation pendant deux et dix jours dans un flacon d’hémoculture anaérobie enrichie avec 5% de rumen et 5% de sang de mouton.

Ce travail montre que le cycle de croissance des bactéries anaérobies est très hétérogène,

114 certaines étant enrichis dès J2 d’incubation alors que d’autres nécessitent 10 jours d’incubation.

115 Publication n°3

Culture of previously uncultured members of the

human gut microbiota by culturomics

Jean-Christophe Lagier, Saber Khelaifia, Maryam Tidjani Alou, Sokhna Ndongo, Niokhor

Dione, Perrine Hugon, Aurelia Caputo, Frédéric Cadoret, Sory Ibrahima Traore, El Hadji

Seck, Gregory Dubourg, Guillaume Durand, Gaël Mourembou, Elodie Guilhot, Amadou

Togo, Sara Bellali, Dipankar Bachar, Nadim Cassir, Fadi Bittar, Jérémy Delerce, Morgane

Mailhe, Davide Ricaboni, Melhem Bilen, Nicole Prisca Makaya Dangui Nieko, NdeyeMery

Dia Badiane, Camille Valles, Donia Mouelhi, Khoudia Diop, Matthieu Million, Didier

Musso, Jônatas Abrahão, Esam Ibraheem Azhar, Fehmida Bibi, Muhammad Yasir, Aldiouma

Diallo, Cheikh Sokhna, Felix Djossou, Véronique Vitton, Catherine Robert, Jean Marc

Rolain, Bernard La Scola, Pierre-Edouard Fournier, Anthony Levasseur1 and Didier Raoult

Publié dans Nature Microbiology

116 LETTERS PUBLISHED: 7 NOVEMBER 2016 | ARTICLE NUMBER: 16203 | DOI: 10.1038/NMICROBIOL.2016.203 OPEN Culture of previously uncultured members of the human gut microbiota by culturomics Jean-Christophe Lagier1, Saber Khelaifia1, Maryam Tidjani Alou1,SokhnaNdongo1, Niokhor Dione1, Perrine Hugon1,AureliaCaputo1,FrédéricCadoret1, Sory Ibrahima Traore1,ElHadjiSeck1, Gregory Dubourg1,GuillaumeDurand1, Gaël Mourembou1,ElodieGuilhot1, Amadou Togo1, Sara Bellali1,DipankarBachar1, Nadim Cassir1, Fadi Bittar1, Jérémy Delerce1, Morgane Mailhe1, Davide Ricaboni1,MelhemBilen1,NicolePriscaMakayaDanguiNieko1,NdeyeMeryDiaBadiane1, Camille Valles1, Donia Mouelhi1, Khoudia Diop1, Matthieu Million1, Didier Musso2, Jônatas Abrahão3, Esam Ibraheem Azhar4, Fehmida Bibi4, Muhammad Yasir4, Aldiouma Diallo5,CheikhSokhna5, Felix Djossou6, Véronique Vitton7, Catherine Robert1, Jean Marc Rolain1, Bernard La Scola1, Pierre-Edouard Fournier1, Anthony Levasseur1 and Didier Raoult1*

Metagenomics revolutionized the understanding of the years, microbial culture techniques have been neglected, which relations among the human microbiome, health and diseases, explains why the known microbial community of the human gut but generated a countless number of sequences that have not is extremely low13. Before we initiated microbial culturomics13 of been assigned to a known microorganism1. The pure culture the approximately 13,410 known bacterial and archaea species, of prokaryotes, neglected in recent decades, remains essential 2,152 had been identified in humans and 688 bacteria and 2 to elucidating the role of these organisms2. We recently intro- archaea had been identified in the human gut. Culturomics consists duced microbial culturomics, a culturing approach that uses of the application of high-throughput culture conditions to the study multiple culture conditions and matrix-assisted laser desorp- of the human microbiota and uses matrix-assisted laser desorption/ tion/ionization–time of flight and 16S rRNA for identification2. ionization–time of flight (MALDI–TOF) or 16S rRNA amplification Here, we have selected the best culture conditions to increase and sequencing for the identification of growing colonies, some of the number of studied samples and have applied new protocols which have been previously unidentified2. With the prospect of iden- (fresh-sample inoculation; detection of microcolonies and tifying new genes of the human gut microbiota, we extend here the specific cultures of Proteobacteria and microaerophilic and number of recognized bacterial species and evaluate the role of this halophilic prokaryotes) to address the weaknesses of the strategy in resolving the gaps in metagenomics, detailing our strategy previous studies3–5. We identified 1,057 prokaryotic species, step by step (see Methods). To increase the diversity, we also thereby adding 531 species to the human gut repertoire: 146 obtained frozen samples from healthy individuals or patients with bacteria known in humans but not in the gut, 187 bacteria various diseases from different geographical origins. These frozen and 1 archaea not previously isolated in humans, and 197 poten- samples were collected as fresh samples (stool, small-bowel and tially new species. Genome sequencing was performed on the colonic samples; Supplementary Table 1). Furthermore, to determine new species. By comparing the results of the metagenomic appropriate culture conditions, we first reduced the number of and culturomic analyses, we show that the use of culturomics culture conditions used (Supplementary Table 2a–c) and then allows the culture of organisms corresponding to sequences focused on specific strategies for some taxa that we had previously previously not assigned. Altogether, culturomics doubles the failed to isolate (Supplementary Table 3). number of species isolated at least once from the human gut. First, we standardized the microbial culturomics for application The study of the human gut microbiota has been revived by to the sample testing (Supplementary Table 1). A refined analysis metagenomic studies6–8. However, a growing problem is the gaps of our first study, which had tested 212 culture conditions4,showed that remain in metagenomics, which correspond to unidentified that all identified bacteria were cultured at least once using one sequences that may be correlated with an identified organism9. of the 70 best culture conditions (Supplementary Table 2a). We Moreover, the exploration of relations between the microbiota and applied these 70 culture conditions (Supplementary Table 2a) to human health require—both for an experimental model and the study of 12 stool samples (Supplementary Table 1). Thanks to therapeutic strategies—the growing of microorganisms in pure the implementation of the recently published repertoire of human culture10, as recently demonstrated in elucidations of the role of bacteria13 (see Methods), we determined that the isolated bacteria Clostridium butyricum in necrotizing enterocolitis and the influence included 46 bacteria known from the gut but not recovered by of gut microbiota on cancer immunotherapy effects11,12. In recent culturomics before this work (new for culturomics), 38 that had

1Aix Marseille Université URMITE, UM63, CNRS 7278, IRD 198, INSERM 1095, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France. 2Institut Louis Malardé, Papeete, Tahiti, Polynésie Française. 3Departamento de Microbiologia Laboratorio de Virus, Universidade Federal de Minas Gerais, Belo Horizonte, Brasil. 4Special Infectious Agents Unit, King Fahd Medical Research Center, King Abdulaziz University, Jeddah 21589, Saudi Arabia. 5Institut de Recherche pour le Développement, UMR 198 (URMITE), Campus International de Hann, IRD, BP 1386, CP, 18524 Dakar, Sénégal. 6Department of Infectious and Tropical Diseases, Centre Hospitalier de Cayenne, Cayenne, French Guiana. 7Service de Gastroentérologie, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, 13915 Marseille, France. *e-mail: [email protected]

NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology117 1 LETTERS NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203

Present work Total number of 1,476 1,480 1,525 1,430 microorganisms 1,394 1,400 known in human 1,324 1,283 gut 1,170 1,071 1,099 1,103 1,051 1,012 994 997 247 New species 216 217 (NS) 857 900 206 847 199 199 159 First isolation in 149 269 human (NH) 225 259 260 205 208 572 188 175 Already known Culturomics 690 449 60 in human other 240 242 244 250 results 110 231 234 site (H) 341 50 200 218 30 81 142 Microorganisms 60 104 identified by 77 404 335 362 371 380 382 382 culturomics 260 323 174 214 (H(GUT))

286 Microorganisms identified by 355 328 319 310 308 308 430 367 other laboratories only 516 476 69 (ref. 15) 690 69 69 69 69 69 69 69 69 69

ABCDE FGH I J K

A: First project of culturomics E: Cohorts I: Halophilic Archaea B: Published culturomics studies F: Fresh stools J: Microcolonies C: 70 culture conditions G: Proteobacteria K: Duodenum D: 18 culture conditions H: Microaerophilic

Figure 1 | Number of different bacteria and archaea isolated during the culturomics studies. Columns A and B represent the results from previously published studies, and columns C to K the different projects described herein. The bacterial species are represented in five categories: NS, new species; NH, prokaryotes first isolated in humans; H, prokaryotes already known in humans but never isolated from the human gut; H (GUT), prokaryotes known in the human gut but newly isolated by culturomics; and prokaryotes isolated by other laboratories but not by culturomics. already been isolated in humans but not from the gut (non-gut Among the gut species mentioned in the literature13 and not pre- bacteria), 29 that had been isolated in humans for the first time viously recovered by culturomics, several were extremely oxygen- (non-human bacteria) and 10 that were completely new species sensitive anaerobes, several were microaerophilic and several were (unknown bacteria) (Fig. 1 and Supplementary Tables 4a and 5). Proteobacteria, and we focused on these bacteria (Supplementary Beginning in 2014, to reduce the culturomics workload and Table 3). Because delay and storage may be critical with anaerobes, extend our stool-testing capabilities, we analysed previous studies we inoculated 28 stools immediately upon collection. This enabled and selected the 18 best culture conditions2. We performed cultures the culture of 27 new gut species for culturomics, 13 non-gut bacteria, in liquid media in blood culture bottles, followed by subcultures on 17 non-human bacteria and 40 unknown bacteria (Fig. 1 and agar (Supplementary Table 2b). We designed these culture con- Supplementary Tables 3a and 4). When we specifically tested 110 ditions by analysing our first studies. The results of those studies samples for Proteobacteria, we isolated 9 bacteria new to culturomics, indicated that emphasizing three components was essential: pre- 3 non-gut bacteria and 3 non-human bacteria (Fig. 1 and incubation in a blood culture bottle (56% of the new species iso- Supplementary Tables 4a and 5). By culturing 242 stool specimens lated), the addition of rumen fluid (40% of the new species isolated) exclusively under a microaerophilic atmosphere, we isolated 9 bacteria and the addition of sheep blood (25% of the new species isolated)2–5. new to culturomics, 6 non-gut bacteria, 17 non-human bacteria and 7 We applied this strategy to 37 stool samples from healthy individ- unknown bacteria (Fig. 1 and Supplementary Tables 4a and 5). We also uals with different geographic provenances and from patients with introduced the culture of halophilic prokaryotes from the gut and different diseases (Supplementary Table 1). This new strategy microcolony detection. The culture of halophilic bacteria was per- enabled the culture of 63 organisms new to culturomics, 58 non- formed using culture media supplemented with salt for 215 stool gut bacteria, 65 non-human bacteria and 89 unknown bacteria samples, allowing the culture of 48 halophilic prokaryotic species, (Fig. 1 and Supplementary Tables 4a and 5). including one archaea (Haloferax alexandrinus), 2 new bacteria for cul- We also applied culturomic conditions (Supplementary turomics, 2 non-gut bacteria, 34 non-human bacteria, 10 unknown bac- Table 2c) to large cohorts of patients sampled for other purposes teria and one new halophilic archaea (Haloferax massiliensis sp. nov.) (premature infants with necrotizing enterocolitis, pilgrims returning (Fig. 1 and Supplementary Tables 4a and 5). Among these 48 halophilic – from the Hajj and patients before or after bariatric surgery) prokaryotic species, 7 were slight halophiles (growing with 10–50 g l 1 – (Supplementary Table 1). A total of 330 stool samples were ana- of NaCl), 39 moderate halophiles (growing with 50–200 g l 1 of NaCl) – lysed. This enabled the detection of 13 bacteria new to culturomics, and 2 extreme halophiles (growing with 200–300 g l 1 of NaCl). 18 non-gut bacteria, 13 non-human bacteria and 10 unknown We also introduced the detection of microcolonies that were species (Fig. 1 and Supplementary Tables 4a and 5). barely visible to the naked eye (diameters ranging from 100 to

2 118NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203 LETTERS

Extension of human gut repertoire Decipher metagenomic gaps by culturomics

(1) Comparison of 16S rRNA of our 247 new species (197 + 50 previously published) with HMP

125 of our species previously detected 945 dierent prokaroytes including 2 archaea as OTU by metagenomics studies 973 samples 1,200 (2) 19,980 new ORFans genes including 1,326 from 54 of our new species 1,000 197 New species (3) From 7.7 to 60.7% of our new species detected in Nielsen and Browne metagenomic studies, respectively 800 New for human 10−1 188 10−2 New for human gut 10−3 10−4 Previously known 10−5 600 146 Comparison of 84 samples 10−6 from gut but new 10−7 analysed by metagenomics 10−8 10−9 190 for culturomics −10 400 and culturomics 10 Previously known from gut 200 336 (4) Among the 200 16S rRNAs of the new 0 species: 102 recovered 827 times (average Number of species 9.8 per stool)

ATGACGTGACGGGCGGTGTGTACAAGGCCC GGG AA C GT ATT C (5) Analysis of the species with a cut o of 20 100 110 120 130 reads = 4,158 OTU and 556 species

600 1,258 16S rRNAs of Never found in 500 86 unidentified colonies Not human gut 50 culturomics 400 Previously known (136 species) 102 from gut but not by 300 61 culturomics 47 New species 2.7 million spectra 200 MALDI–TOF Culturomics New in human gut 901,364 colonies (420 species) 210 6,000 100 New in humans

4,000 Known from gut

Intensity (a.u.) 0 2,000 Number of species 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 m/z

Figure 2 | Summary of the culturomics work that has extended the gut repertoire and filled some of the gaps in metagenomics.

300 µm) and could only be viewed with magnifying glasses. These (Supplementary Table 7), we blasted these with 13,984,809 colonies were transferred into a liquid culture enrichment contigs/scaffolds from the assembly of whole metagenomic studies medium for identification by MALDI–TOF mass spectrometry by HMP, enabling the detection of 1,326 ORFans (6.6%) from 54 (MS) or 16S rRNA amplification and sequencing. By testing ten of our new bacterial species (including 45 detected also from 16S) stool samples, we detected two non-gut bacteria, one non-human (Supplementary Table 8). Therefore, at least 102 new bacterial bacterium and one unknown bacterium that only formed micro- species were found but not identified in previous metagenomic colonies (Fig. 1 and Supplementary Tables 4a and 5). Finally, by studies from the HMP. Third, we searched for our 247 new culturing 30 duodenal, small bowel intestine and colonic samples, species in the 239 human gut microbiome samples from healthy we isolated 22 bacteria new to culturomics, 6 non-gut bacteria, individuals described by Browne et al., in which 137 bacterial 9 non-human bacteria and 30 unknown bacteria (Fig. 1 and species were isolated15. We captured 150 of our new species in Supplementary Tables 4a and 5). To continue the exploration of these metagenomics data, representing 60.7% (Supplementary gut microbiota, future culturomics studies could also be applied to Table 9). Moreover, we also identified 19 of our species (7.7%) intestinal biopsies. from 396 human stool individuals described by Nielsen et al., In addition, we performed five studies to evaluate the role of cul- from which 741 metagenomic species and 238 unique metagenomic turomics for deciphering the gaps in metagenomics9. First, we com- genomes were identified16 (Supplementary Table 9). Fourth, we pared the 16S rRNA sequences of the 247 new species (the 197 new analysed the 16S rRNA metagenomic sequences of 84 stools also prokaryotic species isolated here in addition to the 50 new bacterial tested by culturomics (Supplementary Table 10). We compared the species isolated in previous culturomic studies3–5) to the 5,577,630 OTUs identified by blast with a database including the 16S rRNA reads from the 16S rRNA metagenomic studies listed by the of all species isolated by culturomics. Among the 247 16S rRNA of Human Microbiome Project (HMP) (http://www.hmpdacc.org/ the new species, 102 were recovered 827 times, with an average of catalog). We found sequences, previously termed operational taxo- 9.8 species per stool. Finally, analysis of these species using a cutoff nomic units (OTUs), for 125 of our bacterial species (50.6%). These threshold of 20 reads identified 4,158 OTUs and 556 (13.4%) identified bacterial species included Bacteroides bouchedurhonense, species (Supplementary Table 11), among which 420 species which was recovered in 44,428 reads, showing that it is a common (75.5%) were recovered by culturomics. Of these, 210 (50%) were bacterium (Supplementary Table 6). Second, because the genome previously found to be associated with the human gut, 47 were not sequencing of 168 of these new species allowed the generation of previously found in humans (11.2%), 61 were found in humans but 19,980 new genes that were previously unknown (ORFans genes) not in the gut (14.5%) and 102 (24.3%) were new species.

119 3 NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. LETTERS NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203

* Methanobrevibacter s Haloferax massiliensis

Bacteroides bouchedurhonensis

Bacte Bactero Mediterranea massiliensis nsis * se Massilioprevotell Prevotella roides congolense B liensis Metaprevotella massiliensis Bactero acteroides fragilis alen ides Bac mithii Ihup Prevotella phoceensis Bacteroides timonensis mediterraneensi massi massilia timonensisteroides p revotella massilie ides neonati Mar hense s seneg is massiliensis ensis

coccus timone llu ns

a massiPrevotella ca ssiliensis dda sei iae nsis lie iensis Parabacteroideslla massili hoceense massiliensis sil la massili * Parabacte li ensis ssiliensis s ecta timonensis Marseillo cella ma Sanguibactero Lascolabacillus massiliensisensis nsis cc ma rella massi i * Culturomica massiliensis ae Massilioamazonia e ssiliensis Massiliomal Negl

Anaerotruncus massiliense eromassilibaci Ihuella massiliensis eensis Massilioculturomica massiliensis Anaerotruncus rubiinfantis Clostridium je Soleaferrea massiliensis lense cea r massil Bittarella massiliensisGorbachelrni s phoceensis

oides di o Gabonia massiliensis Proven nimonas gabone s ma eil Butyricimonas Ruminiclostridium massiliense Ana na

Butyricimonas Ph Fou phoc ensis Oscilibacter mas Butyricimonas massiliensisides mas ien Marseillobacter lense stasonis Intesti sis ega Clostridium beduini massili 0.05 Colidextra massiliensis estinimo ensis timonensissiliensis Intestinimonas timonensis pho Int ClostridiumIntestinimona nellansenella timonensis ceens Flavonifractor plauti massili sis Alistip Flavonifractor mars is is Bedui s Tidjanibacter mas Christe rhonense es Christensenella massiliensis du phoceensis NiameyiaPolynesiastridium massiliensis massiliosen Alistipes obesihomAlistipes ihumii Eisenbergiella massiliensis sis CloCorynebacterium provencensis Alisti BlautiaBlautia massilienutia timonens phoceensi Alistipespes senegalensis jeddahsiliensis Bla is iensis Alistipes timone dium timonensesil Alis inis Ruminococcus massiliensis * Fusobtipes pr ense Lagierella massilienostri mas a timonens Lachnoclostridium bouches ac ovence Fusobacteriumterium nucleatum massiliensens LachnoclClostridium touaregenseourtella massiliensis * ns is AfricanellaDrancourtell Enterobacter cloacae is Dranc Enterobacter timonens liensis * Klebsiella pneumoniae Ruminococcus phoceensis Enterobacter massiliensis Eubacterium massilienseassiliensis Dorea massi is * Escherichia coli Tyzzerella m dahensis Halomonas massiliensis Bariatricus massiliensis Bacillus Bacillusjed saudii Pse amazoniensis udomonas mass Bacillus rubiinfantislio Xanthomonas us massi Vitreoscilla massiliensis iliensis Bacill cillus jeddahtimonensis Herbaspirillum massiliense massiliensis Ba Bacillus massilionigeriensis Bacillus massilioanorexius Sutterella massiliensis Bacillus testis Dakarella massiliensis Bacillus massiliosenegalensis Duodena massiliensis Bacillus touaregensis Bacillus timonensis Microvirga massiliensis Bacillus phoceensis Desulfomassilia massiliensis Bacillus massiliogabonensis Pacaella massiliensis Bacillus mediterraneensis Bacillus niameyensis Cellulomonas timonensis Bacillus andreraoultii Cellulomonas massiliensis Rubiinfantum massiliensis Timonella senegalensis Planococcus massiliensis Oceanobacillus jeddahense Mobilicoccus massiliensis Oceanobacillus massiliensis Brachybacterium massiliensis Virgibacillus massiliensis Nesterenkonia massiliensis Thalassobacillus massiliensis Virgibacillus senegalensis Brevibacterium senegalense Gracilibacillus massiliensis Brevibacterium phoceense Gracilib Flaviflexus massiliensis Lentibacillusacillus massiliensis timonensis s ihumii sis Ba Paraliobacillus massiliens Actinomyceces grossen cillus cer Actinomy eus * uchesdurhonensissis No is ces bo socomiicoccu Actinomy esiense En yn sis aense Enterococcterococcus sf massiliensis Actinomycesmyces provencen pol een ac Actino Kurth us ma aecium * Kur ia s ssilien ActinomycesCorynebacterium phoc p Kurththia massiliensisenegalensisStreptococcusis bouchedurhonensism jeddahense Massil ia ti um eriu Rub monen s ti olinskerse ensis eoparvulumioba massiliensis monens Corynebacteri m kar sis Numidum cteri sis is Corynebact ien BrevibacillusRisungbinell massiliensisum acteriu bacterium ihumiisil iense senega ococcus massilisil massiliensis Coryneb Corynest Gor lense Bla iliensis PaenibacillusPaenibaillibacterium mara atimonense massil Paeni Nocardioides mas Pa cillus senegalensis ssiliensis eni baci ien AeromicrobiumNigerium mas mass ma sis PaePaenibacillus ihumii sis bacil llu um acnes P nibacillus numidislus bs re aeniba amassm Tessaracoccuseptomyces massiliensis Paenibacillu ouchesdurhonensisiensis Str ngum PaenibacillusPa phoceensiscillus touaregensis siliensis ropionibacteri Paenibacilluen * P ella lenta iba nsis cillus Raoultibacter massilien s ant ggerth lie obacterium lo senegaibioticoph Eggerthella* E timonensis s rubi * Bifid Holdemani Diel lo anaerobiassiliensis Hol inf massilie onibacterArabia massiliense massi an ila ia ma Beduma fasdemania t ClostriClostridiu tis Gord ClostClostridiu Clost ini ma a massil n sella ihumii s Clostrid tidiosa sis M Hugonellalin Clostr C dium sau D rid StoqueStoquefichG imonensis Col Clostridium massi ridiumlost as uyanassiliensis mas Senegalimassil is Khel es ium terti m jed siliomicrobiota timonensis iensis ensis s Clostridium niamey m n ssiliensis nensis nu ridiumiu mediterraneense sili idium amazonitim aerofacien phoceensis iens m mass fi silioamazoniensis ClostridiumClo aifiab esiell igeriense dahtimonense iensis Maln po dii chu Clostridiu Clo

Ro Anaerosalibacter timonensis Anaerosalibacter us ella Anaerofustis mass

mas sil stridium culturomics ly Me Collinsella massiliensis timonensis Peptoniphilus Senegalia mass s

linsella Clost um * s je il Urmitella timonensis Urmitella stridium senegalense a massilien nes mbo ac ma i

orma mas massiliens Ndiopella ilioa dda ens senella ma utritionisia ma

diannikovella m diannikovella teri Olsen massiliensis Urmitella

* Col En JeddahellaOl massiliensis Peptoniph iense s is Anaerosalibacter massiliensis Anaerosalibacter

Pe si

Peptoniphilus phoceensis Peptoniphilus ridium Enorma timonensis duode Peptoniphilus utsia timo hense

Olsenella timo Emergencia timonense ma lie

ella massiliensi Peptonip

Ihubacte um mas

Olsenella provencensis ptoniphil Collinsella timonensis Micromassilia ih m bife liod nsis

zonie

Neofamilia massiliensis Neofamilia cus massiliensis massiliensis Murdochiella Ileobacteri umii gasphaera massil Mobi ielmoense

dak ense sis one

Kallipyga gabonensis Kallipyga

Me Caecum ococ senegalensis ilus ns

rmen sil

hilus obe hilus ns

Colonella massiliensis r m us grossensis us ne iliensis arense is Kallipyga massiliensis Kallipyga lob ie s il e n Massiliobacillus mas si ns

Anaerococcus rubiinfantis Anaerococcus iensis

assiliensis assiliensis acillus m Acidaminococcus timonensis Anaeroc sis ens idamin liensis tans * is Anaerococcus obesihominis

Negativicoccus massiliense um mass Anaerococcus senegalensis Anaerococcus

Ac is e

sihominis numensis

occus jeddahense assilie

iliens

nsis

e

Figure 3 | Phylogenetic tree of the 247 new prokaryote species isolated by culturomics. Bacterial species from Firmicutes are highlighted in red, Actinobacteria (light green), Proteobacteria (blue), Bacteroidetes (purple), Synergistetes (green), Fusobacteria (dark green) and Archaea (grey), respectively. The sequences of 16 prokaryotic species belonging to six phyla previously known from the human gut and more frequently isolated by culture in human gut are highlighted in bold and by an asterisk.

Interestingly, among the 136 species not previously found by culturo- to Proteobacteria (a phylum that we have under-cultured to date; mics, 50 have been found in the gut and 86 have never previously Supplementary Table 5), 88 to Bacteroidetes, 9 to Fusobacteria, 3 to been found in the human gut (Fig. 2 and Supplementary Table 11). Synergistetes, 2 to Euryarchaeota, 1 to Lentisphaerae and 1 to Overall, in this study, by testing 901,364 colonies using MALDI– Verrucomicrobia (Supplementary Table 4a). Among these 197 new TOF MS (Supplementary Table 1), we isolated 1,057 bacterial species, prokaryotes species, 106 (54%) were detected in at least two stool including 531 newly found in the human gut. Among them, 146 samples, including a species that was cultured in 13 different stools were non-gut bacteria, 187 were non-human bacteria, one was a non- (Anaerosalibacter massiliensis) (Supplementary Table 4a). In compari- human halophilic archaeon and 197 were unknown bacteria, including son with our contribution, a recent work using a single culture medium two new families (represented by Neofamilia massiliensis gen. nov., sp. was able to culture 120 bacterial species, including 51 species known nov. and Beduinella massiliensis gen. nov., sp. nov.) and one unknown from the gut, 1 non-gut bacterium, 1 non-human bacterium and 67 halophilic archaeon (Fig. 1 and Supplementary Table 4a). Among these, unknown bacteria, including two new families (Supplementary 600 bacterial species belonged to Firmicutes, 181 to Actinobacteria, 173 Table 12).

4 120 NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203 LETTERS

To obtain these significant results we tested more than 900,000 the growth of all the bacteria4. We applied these culture conditions to 12 more stool colonies, generating 2.7 million spectra, and performed 1,258 samples and tested 160,265 colonies by MALDI–TOF (Supplementary Table 1). The fi fi 18 best culture conditions were selected using liquid media enrichment in a medium molecular identi cations of bacteria not identi ed through fl – fi containing blood and rumen uid and subculturing aerobically and anaerobically in MALDI TOF, using 16S rRNA ampli cation and sequencing. The a solid medium (Supplementary Table 2b)2. Subcultures were inoculated every three new prokaryote species are available in the Collection de Souches days on solid medium, and each medium was kept for 40 days. We applied these de l’Unité des Rickettsies (CSUR) and Deutsche Sammlung von culture conditions to 40 stool samples, ultimately testing 565,242 colonies by – Mikroorganismen und Zellkulturen (DSMZ) (Supplementary MALDI TOF (Supplementary Table 1). Tables 4a and 5). All 16S sequences of the new species and the Cohorts. In parallel to these main culturomics studies, we used fewer culture species unidentified by MALDI–TOF, as well as the genome conditions to analyse a larger number of stool samples. We refer to these projects as sequences of the new species, have been deposited in GenBank cohorts. Four cohorts were analysed (pilgrims returning from the Hajj, premature (Supplementary Tables 5 and 13). In addition, thanks in part to infants with necrotizing enterocolitis, patients before and after bariatric surgery, and patients for acidophilic bacterial species detection). A total of 330 stool samples an innovative system using a simple culture for the archaea – 17 generated the 52,618 colonies tested by MALDI TOF for this project without an external source of hydrogen , among these prokaryotes (Supplementary Table 1). we isolated eight archaeal species from the human gut, including two new ones for culturomics, one non-gut archaea, four Pilgrims from the Hajj. A cohort of 127 pilgrims was included and 254 rectal swabs non-human archaea and one new halophilic species. were collected from the pilgrims: 127 samples were collected before the Hajj and 127 We believe that this work is a key step in the rebirth of the use of samples were collected after the Hajj. We inoculated 100 µl of liquid sample in an 2–5,16 8 ml bottle containing Trypticase Soy Broth (BD Diagnostics) and incubated the culturing in human microbiology and only the efforts of several sample at 37 °C for 1 day. We inoculated 100 µl of the enriched sample into four teams around the world in identifying the gut microbiota repertoire culture media: Hektoen agar (BD Diagnostics), MacConkey agar+Cefotaxime (bioMérieux), Cepacia agar (AES Chemunex) and Columbia ANC agar will allow an understanding and analysis of the relations between the − (bioMérieux). The sample was diluted 10 3 before being plated on the MacConkey microbiota and human health, which could then participate in −4 adapting Koch’s postulates to include the microbiota21. The and Hektoen agars and 10 before being plated on the ANC agar. The sample was not diluted before being inoculated on the Cepacia agar. Subcultures were performed rebirth of culture, termed culturomics here, has enabled the cultur- on Trypticase Soy Agar (BD Diagnostics) and 3,000 colonies were tested using ing of 77% of the 1,525 prokaryotes now identified in the human gut MALDI–TOF. (Fig. 1 and Supplementary Table 4b). In addition, 247 new species (197 cultured here plus 50 from previous studies) and their genomes Preterm neonates. Preterm neonates were recruited from four neonatal intensive care units (NICUs) in southern France from February 2009 to December 2012 are now available (Fig. 3). The relevance of the new species found by (ref. 12). Only patients with definite or advanced necrotizing enterocolitis culturomics is emphasized because 12 of them were isolated in our corresponding to Bell stages II and III were included. Fifteen controls were matched routine microbiology laboratory from 57 diverse clinical samples to 15 patients with necrotizing enterocolitis by sex, gestational age, birth weight, days (Supplementary Table 14). In 2016, 6 of the 374 (1.6%) different of life, type of feeding, mode of delivery and duration of previous antibiotic therapy. identifications performed in the routine laboratory were new The stool samples were inoculated into 54 preselected culture conditions (Supplementary Table 2c). The anaerobic cultures were performed in an anaerobic species isolated from culturomics. As 519 of the species found by – fi chamber (AES Chemunex). A total of 3,000 colonies were tested by MALDI TOF culturomics in the gut for the rst time (Fig. 1) were not included for this project. in the HMP (Supplementary Table 15) and because hundreds of their genomes are not yet available, the results of this study Stool analyses before and after bariatric surgery. We included 15 patients who had should prompt further genome sequencing to obtain a better bariatric surgery (sleeve gastrectomy or Roux-en-Y gastric bypass) from 2009 to fi 2014. All stool samples were frozen before and after surgery. We used two different identi cation in gut metagenomic studies. culture conditions for this project. Each stool sample was diluted in 2 ml of Dulbecco’s phosphate-buffered saline, then pre-incubated in both anaerobic (BD Methods Bactec Plus Lytic/10 Anaerobic) and aerobic (BD Bactec Plus Lytic/10 Aerobic) Samples. To obtain a larger diversity of gut microbiota, we analysed 943 different blood culture bottles, with 4 ml of sheep blood and 4 ml of sterile rumen fluid being stool samples and 30 small intestine and colonic samples from healthy individuals added as previously described4. These cultures were subcultured on days 1, 3, 7, 10, living or travelling in different geographical regions (Europe, rural and urban Africa, 15, 21 and 30 in 5% sheep blood Columbia agar (bioMérieux), and 33,650 colonies Polynesia, India and so on) and from patients with diverse diseases (for example, were tested by MALDI–TOF. anorexia nervosa, obesity, malnutrition and HIV). The main characteristics are summarized in Supplementary Table 1. Consent was obtained from each patient, Acidophilic bacteria. The pH of each stool sample was measured using a pH meter: and the study was approved by the local Ethics Committee of the IFR48 (Marseille, 1 g of each stool specimen was diluted in 10 ml of neutral distilled water (pH 7) and France; agreement no. 09–022). Except for the small intestine and stool samples that centrifuged for 10 min at 13,000g; the pH values of the supernatants were then we directly inoculated without storage (see sections ‘Fresh stool samples’ and measured. Acidophilic bacteria were cultured after stool enrichment in a liquid ‘Duodenum and other gut samples’), the faecal samples collected in France were medium consisting of Columbia Broth (Sigma-Aldrich) modified by the addition of − immediately aliquoted and frozen at 80 °C. Those collected in other countries were (per litre) 5 g MgSO4, 5 g MgCl2, 2 g KCl, 2 g glucose and 1 g CaCl2. The pH was sent to Marseille on dry ice, then aliquoted and frozen at −80 °C for between 7 days adjusted to five different values: 4, 4.5, 5, 5.5 and 6, using HCl. The bacteria were and 12 months before analysis. then subcultured on solid medium containing the same nutritional components and pH as the culture enrichment. They were inoculated after 3, 7, 10 or 15 incubation − Culturomics. Culturomics is a high-throughput method that multiplies culture days in liquid medium for each tested pH condition. Serial dilutions from 10 1 to − conditions in order to detect higher bacterial diversity. The first culturomics study 10 10 were then performed, and each dilution was plated on agar medium. Negative concerned three stool samples, 212 culture conditions (including direct inoculation controls (no inoculation of the culture medium) were included for each condition. in various culture media), and pre-incubation in blood culture bottles incubated Overall, 16 stool samples were inoculated, generating 12,968 colonies, which aerobically and anaerobically4. Overall, 352 other stool samples, including stool were tested by MALDI–TOF. samples from patients with anorexia nervosa3, patients treated with antibiotics5,or Senegalese children, both healthy and those with diarrhoea22, were previously Optimization of the culturomics strategy. In parallel with this standardization studied by culturomics, and these results have been comprehensively detailed in period, we performed an interim analysis in order to detect gaps in our strategy. previous publications3–5. In this work, we only included the genome sequences of the Analysing our previously published studies, we observed that 477 bacterial species 50 new bacterial species isolated in these previous works to contribute to our analysis previously known from the human gut were not detected. Most of these species grew of culturomics and to fill some of the gaps left by metagenomics. In addition, these in strict anaerobic (209 species, 44%) or microaerophilic (25 species, 5%) conditions, previously published data are clearly highlighted in Fig. 1, illustrating the overall and 161 of them (33%) belonged to the phylum Proteobacteria, whereas only 46 of contribution of culturomics in exploring the gut microbiota. them (9%) belonged to the phylum Bacteroidetes (Supplementary Table 3). The Bacterial species isolated from our new projects and described here were classification was performed using our own database: (http://www.mediterranee- obtained using the strategy outlined in the following sections. infection.com/article.php?laref=374&titre=list-of-prokaryotes-according-to-their- aerotolerant-or-obligate-anaerobic-metabolism). Focusing on these bacterial Standardization of culturomics for the extension of sample testing. Arefined species, we designed specific strategies with the aim of cultivating these analysis allowed the selection of 70 culture conditions (Supplementary Table 2a) for missing bacteria.

NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology121 5

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. LETTERS NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203

Fresh stool samples. As the human gut includes extremely oxygen-sensitive laboratories. With this technique, we isolated seven methanogenic archaea through bacterial species, and because frozen storage kills some bacteria10, we tested 28 stool culturomic studies as previously described25–27. In addition, we propose here an samples from healthy individuals and directly cultivated these samples on collection affordable alternative that does not require specific equipment17. Indeed, a simple and without storage. Each sample was directly cultivated on agar plates, enriched in double culture aerobic chamber separated by a microfilter (0.2 μm) was used to grow blood culture bottles (BD Bactec Plus Lytic/10 Anaerobic) and followed on days 2, 5, two types of microorganism that develop in perfect symbiosis. A pure culture of 10 and 15. Conditions tested were anaerobic Columbia with 5% sheep blood Bacteroides thetaiotaomicron was placed in the bottom chamber to produce the (bioMérieux) at 37 °C with or without thermic shock (20 min/80 °C), 28 °C, hydrogen necessary for the growth of the methanogenic archaea, which was trapped anaerobic Columbia with 5% sheep blood agar (bioMérieux) and 5% rumen fluid in the upper chamber. A culture of Methanobrevibacter smithii or other – – – and R-medium (ascorbic acid 1 g l 1, uric acid 0.4 g l 1, and glutathione 1 g l 1,pH hydrogenotrophic methanogenic archaea had previously been placed in the adjusted to 7.2), as previously described23. For this project, 59,688 colonies were chamber. In the case presented here, the methanogenic archaea were grown tested by MALDI–TOF. aerobically on an agar medium supplemented with three antioxidants (ascorbic acid, glutathione and uric acid) and without the addition of any external gas. We Proteobacteria. We inoculated 110 stool samples using pre-incubation in blood subsequently cultured four other methanogenic archaeal species for the first time culture bottles (BD Bactec Plus Lytic/10 Anaerobic) supplemented with vancomycin aerobically, and successfully isolated 13 strains of M. smithii and 9 strains of – (100 µg l 1; Sigma-Aldrich). The subcultures were performed on eight different Methanobrevibacter oralis from 100 stools and 45 oral samples. This medium allows selective solid media for the growth of Proteobacteria. We inoculated onto aerobic isolation and antibiotic susceptibility testing. This change allows the routine MacConkey agar (Biokar-Diagnostics), buffered charcoal yeast extract (BD study of methanogens, which have been neglected in clinical microbiology Diagnostic), eosine-methylene blue agar (Biokar-Diagnostics), Salmonella–Shigella laboratories and may be useful for biogas production. Finally, to culture halophilic agar (Biokar-Diagnostics), Drigalski agar (Biokar-Diagnostics), Hektoen agar archaea, we designed specific culture conditions (described in the ‘Halophilic (Biokar-Diagnostics), thiosulfate-citrate-bile-sucrose (BioRad) and Yersinia agar bacteria’ section). (BD Diagnostic) and incubated at 37 °C, aerobically and anaerobically. For this project, 18,036 colonies were tested by MALDI–TOF. Identification methods. The colonies were identified using MALDI–TOF MS. Each deposit was covered with 2 ml of a matrix solution (saturated α-cyano acid-4- Microaerophilic conditions. We inoculated 198 different stool samples directly hydroxycinnamic in 50% acetonitrile and 2.5% trifluoroacetic acid). This analysis onto agar or after pre-incubation in blood culture bottles (BD Bactec Plus Lytic/10 was performed using a Microflex LT system (Bruker Daltonics). For each spectrum, a Anaerobic bottles, BD). Fifteen different culture conditions were tested using Pylori maximum of 100 peaks was used and these peaks were compared with those of agar (bioMérieux), Campylobacter agar (BD), Gardnerella agar (bioMérieux), 5% previous samples in the computer database of the Bruker Base and our homemade sheep blood agar (bioMérieux) and our own R-medium as previously described23. database, including the spectra of the bacterial species identified in previous We incubated Petri dishes only in microaerophilic conditions using GENbag works28,29. An isolate was labelled as correctly identified at the species level when at microaer systems (bioMérieux) or CampyGen agar (bioMérieux), except the least one of the colonies’ spectra had a score ≥1.9 and another of the colonies’ R-medium, which was incubated aerobically at 37 °C. These culture conditions spectra had a score ≥1.7 (refs 28,29). generated 41,392 colonies, which were tested by MALDI–TOF. Protein profiles are regularly updated based on the results of clinical diagnoses and on new species providing new spectra. If, after three attempts, the species could Halophilic bacteria. In addition, we used new culture conditions to culture not be accurately identified by MALDI–TOF, the isolate was identified by 16S rRNA halophilic prokaryotes. The culture enrichment and isolation procedures for the sequencing as previously described. A threshold similarity value of >98.7% was culture of halophilic prokaryotes were performed in a Columbia broth medium chosen for identification at the species level. Below this value, a new species was fi 30 (Sigma-Aldrich), modi ed by adding (per litre): MgCl2·6H2O, 5 g; MgSO4·7H2O, suspected, and the isolate was described using taxonogenomics . 5 g; KCl, 2 g; CaCl2·2H2O, 1 g; NaBr, 0.5 g; NaHCO3, 0.5 g and 2 g of glucose. The pH was adjusted to 7.5 with 10 M NaOH before autoclaving. All additives Classification of the prokaryotes species cultured. We used our own online were purchased from Sigma-Aldrich. Four concentrations of NaCl were used prokaryotic repertoire13 (http://hpr.mediterranee-infection.com/arkotheque/client/ – – – – (100 g l 1, 150 g l 1, 200 g l 1 and 250 g l 1). ihu_bacteries/recherche/index.php) to classify all isolated prokaryotes into four A total of 215 different stool samples were tested. One gram of each stool categories: new prokaryote species, previously known prokaryote species in the specimen was inoculated aerobically into 100 ml of liquid medium in flasks at 37 °C human gut, known species from the environment but first isolated in humans, and while stirring at 150 r.p.m. Subcultures were inoculated after 3, 10, 15 and 30 known species from humans but first isolated in the human gut. Briefly, to complete − − incubation days for each culture condition. Serial dilutions from 10 1 to 10 10 were the recent work identifying all the prokaryotes isolated in humans13, we examined then performed in the culture medium and then plated on agar medium. Negative methods by conducting a literature search, which included PubMed and books on controls (no inoculation of the culture medium) were included for each culture infectious diseases. We examined the Medical Subject Headings (MeSH) indexing condition. After three days of incubation at 37 °C, different types of colonies provided by Medline for bacteria isolated from the human gut and we then appeared: yellow, cream, white and clear. Red and pink colonies began to appear established two different queries to automatically obtain all articles indexed by after the 15th day. All colonies were picked and re-streaked several times to obtain Medline dealing with human gut isolation sites. These queries were applied to all pure cultures, which were subcultured on a solid medium consisting of Colombia bacterial species previously isolated from humans as previously described, and we agar medium (Sigma-Aldrich) NaCl. The negative controls remained sterile in all obtained one or more articles for each species, confirming that the bacterium had culture conditions, supporting the authenticity of our data. been isolated from the human gut13.

Detection of microcolonies. Finally, we began to focus on microcolonies detected International deposition of the strains, 16S rRNA accession numbers and using a magnifying glass (Leica). These microcolonies, which were not visualized genome sequencing accession number. Most of the strains isolated in this study with the naked eye and ranged from 100 to 300 µm, did not allow direct were deposited in CSUR (WDCM 875) and are easily available at http://www. identification by MALDI–TOF. We subcultured these bacteria in a liquid medium mediterranee-infection.com/article.php?laref=14&titre=collection-de- (Columbia broth, Sigma-Aldrich) to allow identification by MALDI–TOF after souches&PHPSESSID=cncregk417fl97gheb8k7u7t07 (Supplementary Tables 4a and centrifugation. Ten stool samples were inoculated and then observed using this b). All the new prokaryote species were deposited into two international collections: magnifying glass for this project, generating the 9,620 colonies tested. CSUR and DSMZ (Supplementary Table 5). Importantly, among the 247 new prokaryotes species (197 in the present study and 50 in previous studies), we failed to Duodenum and other gut samples. Most of the study was designed to explore the subculture 9 species that were not deposited, of which 5 were nevertheless genome gut microbiota using stool samples. Nevertheless, as the small intestine microbiota sequenced. Apart from these species, all CSUR accession numbers are available in are located where the nutrients are digested24, which means there are greater Supplementary Table 5. Among these viable new species, 189 already have a DSMZ difficulties in accessing samples than when using stool specimens, we analysed number. For the other 49 species, the accession number is not yet assigned but the different levels of sampling, including duodenum samples (Supplementary Table 1). strain is deposited. The 16S rRNA accession numbers of the 247 new prokaryotes First, we tested five duodenum samples previously frozen at −80 °C. A total of species are available in Supplementary Table 5, along with the accession number of 25,000 colonies were tested by MALDI–TOF. In addition, we tested samples from the known species needing 16S rRNA amplification and sequencing for the different gut levels (gastric, duodenum, ileum and left and right colon) of other identification (Supplementary Table 14). Finally, the 168 draft genomes used for our patients. We tested 25,048 colonies by MALDI–TOF for this project. We tested analysis have already been deposited with an available GenBank accession 15 culture conditions, including pre-incubation in blood culture bottles with sterile number (Supplementary Table 5) and all other genome sequencing is still in rumen fluid and sheep blood (BD Bactec Plus Lytic/10 Anaerobic), 5% sheep blood progress, as the culturomics are still running in our laboratory. agar (bioMérieux), and incubation in both microaerophilic and anaerobic conditions, R-medium23 and Pylori agar (bioMérieux). Overall, we tested New prokaryotes. All new prokaryote species have been or will be comprehensively 50,048 colonies by MALDI–TOF for this project. described by taxonogenomics, including their metabolic properties, MALDI–TOF spectra and genome sequencing30. Among these 247 new prokaryote species, 95 have Archaea. The culture of methanogenic archaea is a fastidious process, and the already been published (PMID available in Supplementary Table 5), including 70 necessary equipment for this purpose is expensive and reserved for specialized full descriptions and 25 ‘new species announcements’. In addition, 20 are under

6 122 NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203 LETTERS review and the 132 others are ongoing (Supplementary Table 5). This includes 37 Sensitivity Bioanalyzer LabChip (Agilent Technologies). The libraries were bacterial species already officially recognized (as detailed in Supplementary Table 5). normalized at 2 nM and pooled. After a denaturation step and dilution at 15 pM, All were sequenced successively with a paired-end strategy for high-throughput the pool of libraries was loaded onto the reagent cartridge and then onto the pyrosequencing on the 454-Titanium instrument from 2011 to 2013 and using instrument along with the flow cell. To prepare the paired-end library, 1 ng of MiSeq Technology (Illumina) with the mate pair strategy since 2013. genome as input was required. DNA was fragmented and tagged during the tagmentation step, with an optimal size distribution at 1 kb. Limited-cycle PCR Metagenome sequencing. Total DNA was extracted from the samples using a amplification (12 cycles) completed the tag adapters and introduced dual-index method modified from the Qiagen stool procedure (QIAamp DNA Stool Mini Kit). barcodes. After purification on Ampure XP beads (Beckman Coulter), the library For the first 24 metagenomes, we used GS FLX Titanium (Roche Applied Science). was normalized and loaded onto the reagent cartridge and then onto the instrument Primers were designed to produce an amplicon length (576 bp) that was along with the flow cell. For the 2 Illumina applications, automated cluster approximately equivalent to the average length of reads produced by GS FLX generation and paired-end sequencing with index reads of 2 × 250 bp were Titanium (Roche Applied Science), as previously described. The primer pairs performed in single 39-hour runs. commonly used for gut microbiota were assessed in silico for sensitivity to sequences from all phyla of bacteria in the complete Ribosomal Database Project (RDP) ORFans identification. Open reading frames (ORFs) were predicted using Prodigal database. Based on this assessment, the bacterial primers 917F and 1391R were with default parameters for each of the bacterial genomes. However, the predicted selected. The V6 region of 16S rRNA was pyrosequenced with unidirectional ORFs were excluded if they spanned a sequencing gap region. The predicted sequencing from the forward primer with one-half of a GS FLX Titanium bacterial sequences were searched against the non-redundant protein sequence (NR) PicoTiterPlate Kit 70×75 per patient with the GS Titanium Sequencing Kit XLR70 database (59,642,736 sequences, available from NCBI in 2015) using BLASTP. after clonal amplification with the GS FLX Titanium LV emPCR Kit (Lib-L). ORFans were identified if their BLASTP E-value was lower than 1e-03 for an Sixty other metagenomes were sequenced for 16S rRNA sequencing using MiSeq alignment length greater than 80 amino acids. We used an E-value of 1e-05 if the technology. PCR-amplified templates of genomic DNA were produced using the alignment length was <80 amino acids. These threshold parameters have been used surrounding conserved regions’ V3–V4 primers with overhang adapters in previous studies to define ORFans (refs 12–14). The 168 genomes considered in (FwOvAd_341F TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGG this study are listed in Supplementary Table 7. These genomes represent 615.99 Mb NGGCWGCAG; ReOvAd_785RGTCTCGTGGGCTCGGAGATG TGTATAAGA and contain a total of 19,980 ORFans. Some of the ORFans from 30 genomes were GACAGGACTACHVGGGTATCTAATCC). Samples were amplified individually calculated in a previous study4 with the non-redundant protein sequence database for the 16S V3–V4 regions by Phusion High Fidelity DNA Polymerase (Thermo containing 14,124,377 sequences available from NCBI in June 2011. Fisher Scientific) and visualized on the Caliper Labchip II device (Illumina) by a DNA 1K LabChip at 561 bp. Phusion High Fidelity DNA Polymerase was chosen for Metagenomic 16S sequences. We collected 325 runs of metagenomic 16S rRNA PCR amplifications in this biodiversity approach and deep sequencing: a sequences available in the HMP data sets that correspond to stool samples from thermostable DNA polymerase characterized by the greatest accuracy, robust healthy human subjects. All samples were submitted to Illumina deep sequencing, reactions and high tolerance for inhibitors, and finally by an error rate that is resulting in 761,123 Mo per sample on average, and a total of 5,970,465 high-quality approximately 50-fold lower than that of DNA polymerase and sixfold lower than sequencing reads after trimming. These trimmed data sets were filtered using CLC that of Pfu DNA polymerase. After purification on Ampure beads (Thermo Fisher Genomics Workbench 7.5, and reads shorter than 100 bp were discarded. We Scientific), the concentrations were measured using high-sensitivity Qbit technology performed an alignment of 247 16S rRNA sequences against the 5,577,630 reads (Thermo Fisher Scientific). Using a subsequent limited-cycle PCR on 1 ng of each remaining using BLASTN. We used a 1e-03 e-value, 100% coverage and 98.7% PCR product, Illumina sequencing adapters and dual-index barcodes were added to cutoff, corresponding to the threshold for defining a species, as previously described. each amplicon. After purification on Ampure beads, the libraries were then Finally, we reported the total number of aligned reads for each 16S rRNA sequence normalized according to the Nextera XT (Illumina) protocol. The 96 multiplexed (Supplementary Table 8). samples were pooled into a single library for sequencing on the MiSeq. The pooled We collected the sequences of the 3,871,657 gene non-redundant gene catalogue library containing indexed amplicons was loaded onto the reagent cartridge and from the 396 human gut microbiome samples (https://www.cbs.dtu.dk/projects/ then onto the instrument along with the flow cell. Automated cluster generation and CAG/)15. We performed an alignment of 247 16S rRNA sequences against the paired-end sequencing with dual index reads of 2 × 250 bp were performed in a 3,871,657 gene non-redundant gene catalogue using BLASTN with a threshold of single 39-hour run. On the instrument, the global cluster density and the global 1e-03 e-value, 100% coverage and 98.7% cutoff. The new species identified in these passed filter per flow cell were generated. The MiSeq Reporter software (Illumina) data are reported in Supplementary Table 9. We collected the raw data sets of 239 determined the percentage indexed and the clusters passing the filter for each runs deposited at EBI (ERP012217)16. We used the PEAR software (PMID amplicon or library. The raw data were configured in fasta files for R1 and R2 reads. 24142950) for merging raw Illumina paired-end reads using default parameters. We performed an alignment of 247 16S rRNA sequences against the 265,864,518 Genome sequencing. The genomes were sequenced using, successively, two high- merged reads using BLASTN. We used a 1e-03 e-value, 100% coverage and 98.7% throughput NGS technologies: Roche 454 and MiSeq Technology (Illumina) with cutoff. The list of the new species identified in these data is included in paired-end application. Each project on the 454 sequencing technology was loaded Supplementary Table 9. on a quarter region of the GS Titanium PicoTiterPlate and sequenced with the GS FLX Titanium Sequencer (Roche). For the construction of the 454 library, 5 μg DNA Whole metagenomic shotgun sequences. We collected the contigs/scaffolds from was mechanically fragmented on the Covaris device (KBioScience-LGC Genomics) the assembly of 148 runs available in the HMP data sets. The initial reads of these through miniTUBE-Red 5Kb. The DNA fragmentation was visualized through the samples were assembled using SOAPdenovo v.1.04 (PMID 23587118). These Agilent 2100 BioAnalyser on a DNA LabChip7500. Circularization and assemblies correspond to stool samples from healthy human subjects and generated fragmentation were performed on 100 ng. The library was then quantified on Quant- 13,984,809 contigs/scaffolds with a minimum length of 200 bp and a maximum it Ribogreen kit (Invitrogen) using a Genios Tecan fluorometer. The library was length of 371,412 bp. We aligned the 19,980 ORFans found previously against these clonally amplified at 0.5 and 1 cpb in 2 emPCR reactions according to the conditions data sets using BLASTN. We used a 1e-05 e-value, 80% coverage and 80% identity for the GS Titanium SV emPCR Kit (Lib-L) v2 (Roche). These two enriched clonal cutoff. Finally, we reported the total number of unique aligned ORFans for each amplifications were loaded onto the GS Titanium PicoTiterPlates and sequenced species (Supplementary Table 8). with the GS Titanium Sequencing Kit XLR70. The run was performed overnight and then analysed on the cluster through gsRunBrowser and gsAssembler_Roche. Study of the gaps in metagenomics. The raw fastq files of paired-end reads from an Sequences obtained with Roche were assembled on gsAssembler with 90% identity Illumina Miseq of 84 metagenomes analysed concomitantly by culturomics were and 40 bp of overlap. The library for Illumina was prepared using the Mate Pair filtered and analysed in the following steps (accession no. PRJEB13171). technology. To improve the assembly, the second application in was sometimes performed with paired ends. The paired-end and the mate-pair strategies were Data processing: filtering the reads, dereplication and clustering. The paired-end barcoded in order to be mixed, respectively, with 11 other genomic projects prepared reads of the corresponding raw fastq files were assembled into contigs using with the Nextera XT DNA sample prep kit (Illumina) and 11 others projects with Pandaseq31. The high-quality sequences were then selected for the next steps of the Nextera Mate Pair sample prep kit (Illumina). The DNA was quantified by a Qbit analysis by considering only those sequences that contained both primers (forward assay with high-sensitivity kit (Life Technologies). In the first approach, the mate and reverse). In the following filtering steps, the sequences containing N were pair library was prepared with 1.5 µg genomic DNA using the Nextera mate pair removed. Sequences with length shorter than 200 nt were removed, and sequences Illumina guide. The genomic DNA sample was simultaneously fragmented and longer than 500 nt were trimmed. Both forward and reverse primers were also tagged with a mate-pair junction adapter. The profile of the fragmentation was removed from each of the sequences. An additional filtering step was applied to validated on an Agilent 2100 Bioanalyzer (Agilent Technologies) with a DNA 7500 remove the chimaeric sequences using UCHIME (ref. 32) of USEARCH (ref. 33). LabChip. The DNA fragments, which ranged in size, had an optimal size of 5 kb. No The filtering steps were performed using the QIIME pipeline34. Strict dereplication size selection was performed, and 600 ng of ‘tagmented’ fragments measured on the (clustering of duplicate sequences) was performed on the filtered sequences, and Qbit assay with the high-sensitivity kit were circularized. The circularized DNA was they were then sorted by decreasing number of abundance35–37. For each mechanically sheared to small fragments, with optimal fragments being 700 bp, on a metagenome, the clustering of OTUs was performed with 97% identity. Total OTUs Covaris S2 device in microtubes. The library profile was visualized on a High from the 84 metagenomes (Supplementary Table 10) clustered with 93% identity.

123 7 NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. LETTERS NATURE MICROBIOLOGY DOI: 10.1038/NMICROBIOL.2016.203

Building reference databases. We downloaded the Silva SSU and LSU database1 22. Samb-Ba, B. et al. MALDI–TOF identification of the human gut microbiome in and release 123 from the Silva website and, from this, a local database of predicted people with and without diarrhea in Senegal. PLoS ONE 9, e87419 (2014). amplicon sequences was built by extracting the sequences containing both primers. 23. Dione, N., Khelaifia, S., La Scola, B., Lagier, J.C. & Raoult D. A quasi-universal Finally, we had our local reference database containing a total of 536,714 well- medium to break the aerobic/anaerobic bacterial culture dichotomy in clinical annotated sequences separated into two subdatabases according to their gut or non- microbiology. Clin. Microbiol. Infect. 22, 53–58 (2016). gut origin. We created four other databases containing 16S rRNA of new species 24. Raoult, D. & Henrissat, B. Are stool samples suitable for studying the link sequences and species isolated by culturomics separated into three groups (human between gut microbiota and obesity? Eur. J. Epidemiol. 29, 307–309 (2014). gut, non-human gut, and human not reported in gut). The new species database 25. Khelaifia, S., Raoult, D. & Drancourt, M. A versatile medium for cultivating contains 247 sequences, the human gut species database 374 sequences, the non- methanogenic archaea. PLoS ONE 8, e61563 (2013). human gut species database 256 sequences and the human species not reported in 26. Khelaifia, S. et al. Draft genome sequence of a human-associated isolate of gut database 237 sequences. methanobrevibacter arboriphilicus, the lowest-G+C-content archaeon. Genome Announc. 2, e01181 (2014). Taxonomic assignments. For taxonomic assignments, we applied at least 20 reads 27. Dridi, B., Fardeau, M.-L., Ollivier, B., Raoult, D. & Drancourt, M. per OTU. The OTUs were then searched against each database using BLASTN Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon (ref. 38). The best match of ≥97% identity and 100% coverage for each of the OTUs isolated from human faeces. Int. J. Syst. Evol. Microbiol. 62, 1902–1907 (2012). was extracted from the reference database, and taxonomy was assigned up to the 28. Seng, P. et al. Identification of rare pathogenic bacteria in a clinical microbiology species level. Finally, we counted the number of OTUs assigned to unique species. laboratory: impact of matrix-assisted laser desorption ionization-time of flight mass spectrometry. J. Clin. Microbiol. 51, 2182–2194 (2013). Data availability. The GenBank accession numbers for the sequences of 29. Seng, P. et al. Ongoing revolution in bacteriology: routine identification of the16SrRNA genes of the new bacterial species as well as their accession numbers in bacteria by matrix-assisted laser desorption ionization time-of-flight mass both Collection de Souches de l’Unité des Rickettsies (CSUR, WDCM 875) and the spectrometry. Clin. Infect. Dis. 49, 543–551 (2009). Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ) are listed in 30. Ramasamy, D. et al. A polyphasic strategy incorporating genomic data for the Supplementary Table 5. Sequencing metagenomics data have been deposited in taxonomic description of novel bacterial species. Int. J. Syst. Evol. Microbiol. 64, NCBI under Bioproject PRJEB13171. 384–391 (2014). 31. Masella, A. P., Bartram, A. K., Truszkowski, J. M., Brown, D. G. & Neufeld, J. D. Received 20 April 2016; accepted 14 September 2016; PANDAseq: paired-end assembler for Illumina sequences. BMC Bioinformatics published 7 November 2016 13, 31 (2012). 32. Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C. & Knight, R. UCHIME References improves sensitivity and speed of chimera detection. Bioinformatics 27, – 1. Lagier, J. C., Million, M., Hugon, P., Armougom, F. & Raoult, D. Human gut 2194 2200 (2011). microbiota: repertoire and variations. Front. Cell Infect. Microbiol. 2, 136 (2012). 33. Edgar, R. C. Search and clustering orders of magnitude faster than BLAST. – 2. Lagier, J. C. et al. The rebirth of culture in microbiology through the example Bioinformatics 26, 2460 2461 (2010). of culturomics to study human gut microbiota. Clin Microbiol Rev. 28, 34. Caporaso, J. G. et al. QIIME allows analysis of high-throughput community – 237–264 (2015). sequencing data. Nat. Methods 7, 335 336 (2010). 3. Pfleiderer, A. et al. Culturomics identified 11 new bacterial species from a 35. Stoeck, T. et al. Massively parallel tag sequencing reveals the complexity of – single anorexia nervosa stool sample. Eur. J. Clin. Microbiol. Infect. Dis. 32, anaerobic marine protistan communities. BMC Biol. 7, 72 77 (2009). 1471–1481 (2013). 36. Mondani, L. et al. Microbacterium lemovicicum sp. nov., a bacterium 4. Lagier, J. C. et al. Microbial culturomics: paradigm shift in the human gut isolated from a natural uranium-rich soil. Int. J. Syst. Evol. Microbiol. 63, – microbiome study. Clin. Microbiol. Infect. 18, 1185–1193 (2012). 2600 2606 (2013). 5. Dubourg, G. et al. Culturomics and pyrosequencing evidence of the reduction in 37. Boissiere, A. et al. Midgut microbiota of the malaria mosquito vector Anopheles gut microbiota diversity in patients with broad-spectrum antibiotics. Int. J. gambiae and interactions with Plasmodium falciparum infection. PLoS Pathog. Antimicrob. Agents 44, 117–124 (2014). 8, e1002742 (2012). 6. Ley, R. E., Turnbaugh, P. J., Klein, S. & Gordon, J. I. Microbial ecology: human 38. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., Lipman, D. J. Basic local – gut microbes associated with obesity. Nature 444, 1022–1023 (2006). alignment search tool. J. Mol. Biol. 215, 403 410 (1990). 7. Ley, R. E. et al. Obesity alters gut microbial ecology. Proc. Natl Acad. Sci. USA 102, 11070–11075 (2005). Acknowledgements 8. Gill, S. R. et al. Metagenomic analysis of the human distal gut microbiome. The authors thank R. Valero, A.A. Jiman-Fatani, B. Ali Diallo, J.-B. Lekana-Douki, Science 312, 1355–1359 (2006). B. Senghor, A. Derand, L. Gandois, F. Tanguy, S. Strouk, C. Tamet, F. Lunet, M. Kaddouri, 9. Rinke, C. et al. Insights into the phylogeny and coding potential of microbial L. Ayoub, L. Frégère, N. Garrigou, A. Pfleiderer, A. Farina and V. Ligonnet for technical dark matter. Nature 499, 431–437 (2013). support. This work was funded by IHU Méditerranée Infection as a part of a Foundation 10. Lagier, J. C. et al. Current and past strategies for bacterial culture in clinical Louis D grant and by the Deanship of Scientific Research (DSR), King Abdulaziz microbiology. Clin. Microbiol. Rev. 28, 208–236 (2015). University, under grant no. 1–141/1433 HiCi. 11. Vetizou, M. et al. Anticancer immunotherapy by CTLA-4 blockade relies on the gut microbiota. Science 350, 1079–1084 (2015). Author contributions 12. Cassir, N. et al. Clostridium butyricum strains and dysbiosis linked to necrotizing D.R. conceived and designed the experiments. J.-C.L., S.K., M.T.A., S.N., N.D., P.H., A.C., enterocolitis in preterm neonates. Clin. Infect. Dis. 61, 1107–1115. F.C., S.I.T., E.H.S., G.Dub., G.Dur., G.M., E.G. A.T., S.B., D.B., N.C., F.B., J.D., M.Ma., D.R., 13. Hugon, P. et al. A comprehensive repertoire of prokaryotic species identified in M.B., N.P.M.D.N., N.M.D.B., C.V., D.M., K.D., M.Mi., C.R., J.M.R., B.L.S., P.-E.F. and A.L. human beings. Lancet Infect. Dis. 15, 1211–1219 (2015). performed the experiments. D.M., J.A., E.I.A., F.B., M.Y., A.D., C.S., F.D. and V.V. 14. The Human Microbiome Project Consortium. A framework for human contributed materials/analysis tools. J.-C.L., A.C., A.L. and D.R. analysed the data. J.-C.L., microbiome research. Nature 486, 215–221 (2012). A.L. and D.R. wrote the manuscript. All authors read and approved the final manuscript. 15. Browne, H. P. et al. Culturing of ‘unculturable’ human microbiota reveals novel taxa and extensive sporulation. Nature 533, 543–546 (2016). Additional information 16. Nielsen, H. B. et al. Identification and assembly of genomes and genetic elements Supplementary information is available for this paper. Reprints and permissions information in complex metagenomic samples without using reference genomes. is available at www.nature.com/reprints. Correspondence and requests for materials should be Nat. Biotechnol. 32, 822–828 (2014). addressed to D.R. 17. Khelaifia, S. et al. Aerobic culture of methanogenic archaea without an external source of hydrogen. Eur. J. Clin. Microbiol. Infect. Dis. 35, 985–991 (2016). Competing interests 18. Rettedal, E. A., Gumpert, H. & Sommer, M. O. Cultivation-based multiplex The authors declare no competing financial interests. phenotyping of human gut microbiota allows targeted recovery of previously uncultured bacteria. Nat. Commun. 5, 4714 (2014). This work is licensed under a Creative Commons Attribution 4.0 19. Hiergeist, A., Gläsner, J., Reischl, U. & Gessner, A. Analyses of intestinal International License. The images or other third party material in microbiota: culture versus sequencing. ILAR J. 56, 228–240 (2015). this article are included in the article’s Creative Commons license, 20. Rajilic-Stojanovic, M. & de Vos, W. M. The first 1000 cultured species of the unless indicated otherwise in the credit line; if the material is not included under the human gastrointestinal microbiota. FEMS Microbiol. Rev. 38, 996–1047 (2014). Creative Commons license, users will need to obtain permission from the license holder to 21. Byrd, A. L. & Segre, J. A. Infectious disease. Adapting Koch’s postulates. Science reproduce the material. To view a copy of this license, visit http://creativecommons.org/ 351, 224–226 (2016). licenses/by/4.0/

8 124 NATURE MICROBIOLOGY | VOL 1 | DECEMBER 2016 | www.nature.com/naturemicrobiology © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION ARTICLE NUMBER: 16203 | DOI: 10.1038/NMICROBIOL.2016.203 1

2 Culture of previously uncultured members of the human gut microbiota by culturomics

3 Jean-Christophe Lagier1, Saber Khelaifia1, Maryam Tidjani Alou1, Sokhna Ndongo1, Niokhor

4 Dione1, Perrine Hugon1, Aurelia Caputo1, Frédéric Cadoret1, Sory Ibrahima Traore1, El Hadji

5 Seck1, Gregory Dubourg1, Guillaume Durand1, Gaël Mourembou1, Elodie Guilhot1, Amadou

6 Togo 1, Sara Bellali1, Dipankar Bachar1, Nadim Cassir1, Fadi Bittar1, Jérémy Delerce1,

7 Morgane Mailhe1, Davide Ricaboni1, Melhem Bilen1, Nicole Prisca Makaya Dangui Nieko1,

8 Ndeye Mery Dia Badiane1, Camille Valles1, Donia Mouelhi1, Khoudia Diop1, Matthieu

9 Million1, Didier Musso2, Jõnatas Abrahao3, Esam Ibraheem Azhar4, Fehmida Bibi4,

10 Muhammad Yasir4, Aldiouma Diallo5, Cheikh Sokhna5, Felix Djossou6, Véronique Vitton7,

11 Catherine Robert1, Jean Marc Rolain1, Bernard La Scola1, Pierre-Edouard Fournier1, Anthony

12 Levasseur1 and Didier Raoult1*

13 1. Aix Marseille Université URMITE, UM63, CNRS 7278, IRD 198, INSERM 1095, 27 Bd 14 Jean Moulin, 13385 Marseille Cedex 5, France. 15 2. Institut Louis Malardé – Papeete – Tahiti – Polynésie Française. 16 3. Departamento de Microbiologia Laboratorio de Virus Universidade Federal de Minas Gerais, 17 Brasil. 18 4. Special Infectious Agents Unit, King Fahd Medical Research Center, King Abdulaziz 19 University, Jeddah, 21589 Saudi Arabia. 20 5. Institut de Recherche pour le Développement, UMR 198 (URMITE), Campus International 21 de Hann, IRD, BP 1386, CP 18524, Dakar, Sénégal 22 6. Department of Infectious and Tropical Diseases, Centre Hospitalier de Cayenne, Cayenne, 23 French Guiana. 24 7. Service de Gastroentérologie, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille, 25 13915 Marseille, France 26 *To whom correspondence should be addressed 27 Prof. Didier Raoult (corresponding author) 28 Email: [email protected] 29 Phone: 33 4 91 38 55 17 30 Fax: 33 4 91 38 77 72

31 Supplementary Tables 1 to 15

125 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 1

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 1: Main characteristics of the samples tested in the culturomics studies

N° N° Category samples Sample Origin Clinical status Age Sex BMI colonies Name tested Previous studies Data previously A First study 3 described 32500 Data previously B Previously published 352 described 47453 Total 355 79953 Current study French C 70 conditions 12 Raiatea Polynesia Healthy 39 M 24 12735 Inde India Healthy 33 M 21 14482 French Polynésie8 Polynesia Obese 49 M 27 10674 Amaz1 Amazonia Healthy 10 F A 17338 Amaz2 Amazonia Healthy 16 F A 34700 AIDS-infected VIH 1 France patient 32 M 23 10290 AIDS-infected VIH3 France patient 46 M 21 16090 AIDS-infected HIV11 France patient 36 M 25 10630 AIDS-infected HIV12 France patient 21 M 19 9326 AIDS-infected HIV13 France patient 31 M 20 10136 AIDS-infected HIV4 France patient 72 F 22 8064 AIDS-infected HIV8 France patient 32 M 24 5800 D 18 conditions 40 Gab3 Gabon Healthy 27 M 22 14868 Gab5 Gabon Healthy 16 M 19 14532 SAOB1 Saudi Arabia Obese individual 25 M 51 8538 SA OB3 Saudi Arabia Obese individual 24 M 52 23476 SA OB6 Saudi Arabia Obese individual 25 M 46 8896 SAN1 Saudi Arabia Healthy 25 M 20 16584 SAN7 Saudi Arabia Healthy 27 M 23 9732 Touareg 8 Niger Healthy na M na 7600 Marasme 03 Malnutrition 5700 Marasme 06 Malnutrition 6800 Marasme 9 Niger Malnutrition na M na 12753 BD8 Saudi Arabia Healthy 29 M 51,6 10254 BD2 Saudi Arabia Healthy 19 M 14 9200 BD4 Saudi Arabia Healthy 34 M 18.4 7200 BD10 Saudi Arabia Healthy 19 M 24.4 12564 BD11 Saudi Arabia Healthy 31 M 25.5 9897 BD9 Saudi Arabia Healthy 42 F 20.8 13024 BD3 Saudi Arabia Healthy 50 F 25.7 8200 Control N6 Niger Healthy 7 m F 17 11474 13 Control N2 Niger Healthy m F 15 11352 Control N12 Healthy 11686 Control S07 Senegal Healthy 3 F 15 11750 Control S05 Senegal Healthy 3 M 14 12540

126 2 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Control S500 Healthy 11008 Kwashiorkor 5004 Senegal Malnutrition 7 m F na 11402 Kwashiorkor 1 (Niger) Niger Malnutrition na na na 4895 Kwashiorkor 01 (Senegal) Senegal Malnutrition 8 m F 15 12422 Kwashiorkor 02 Senegal Malnutrition 4 F 15 17210 Kwashiorkor 4 Senegal Malnutrition 4 m F 11 9312 Kwashiorkor 5 (Senegal) Senegal Malnutrition 1 M 12 8402 Kwashiorkor 6 Senegal Malnutrition 6 m M 12 8358 Kwashiorkor 8 (Niger) Niger Malnutrition na na na 9587 Kwashiorkor 10 (Senegal) Senegal Malnutrition 2 m M 13 9820 Kwashiorkor 12 Niger Malnutrition na na na 7854 Kwashiorkor 14 Niger Malnutrition na na na 8654 Amaz10 Amazonia Healthy 1020 Amaz15 Amazonia Healthy 10 M na 1020 Amaz14 Amazonia Healthy 15693 Amaz20 Amazonia Healthy 15700 Pygmies 1 Congo Healthy 35 F na 6500 E Cohortes Pilgrims 254 Nd nd nd nd nd nd 3000 Bariatric surgery 30 Nd nd nd nd nd nd 33650 Preterm 30 Nd nd nd nd nd nd 3000 Acidophiles 16 Nd nd nd nd nd nd 12968 F Fresh stools 28 Nd France Healthy nd nd nd 59688 G Proteobacteria 110 Nd nd nd nd nd nd 18036 H Microaerophilie 198 Nd nd nd nd nd nd 41392 I Microcolonies 10 Nd France Healthy nd nd nd 9620 J Halophiles 215 Nd nd nd nd nd nd 91220 Small bowel and K Colonic 19 Nd nd nd nd nd nd 31048 Total 973 901,364 nd= not detailed ; na : not available, m= months

127 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 3

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 2A: The 70 most effective culture conditions selected after the pioneer study. The 70 most effective culture conditions 1. 5% sheep blood agar 0,45µm, Aerobic 37°C 2. 5% sheep blood agar Anaerobic, 37°C (Anaerobic cabinet) 3. 5% sheep blood agar, Aerobic 28°C 4. 5% sheep blood agar, Aerobic 37°C 5. 5% sheep blood agar, Anaerobic 28°C 6. 5% sheep blood agar, Anaerobic 37°C 7. 5% sheep blood agar, microaerophilic, 37°C 8. BCP, Aerobic 37°C 9. BCYE, Aerobe 37°C, 2,5% CO2 10. Bordetella, Aerobic 37°C 2,5% C02 11. Brain Heart Infusion + 5% 0.2µm filtered stool, Aerobic 37°C 12. Brain Heart Infusion + NaCl 15g/l, 37°C Aerobic 13. Brain Heart Infusion + NaCl 1g/l, Aerobic 37°C 14. Brain Heart Infusion + NaCl 3g/l, Aerobic 37°C 15. Brain Heart Infusion + sheep blood 5%, Aerobic 37°C 16. Brain Heart Infusion + sheep blood 5%, Anaerobic, 37°C 17. Brain Heart Infusion + Vanco µg/l, Aerobic 37°C 18. Brain Heart Infusion 3% NaCl, Anaerobic 37°C 19. Brain Heart Infusion, 57°C Ae 20. Brain Heart Infusion, Anaerobic 55°C 21. Brucella, Aerobic 37° C 22. CaCO3, aerobe 37°C 23. Co culture amoeba 24. EMB, Aerobic 37°C 25. Filtration 5% sheep blood agar 5µm Anaerobic , 37°C 26. Filtration Brain Heart Infusion 0,45 µm, Aerobic 37°C 27. Filtration Brain Heart Infusion 0.8 µm, Aerobic 37°C 28. Filtration Brain Heart Infusion 0.8µm, Anaerobic 37°C 29. Filtration Brain Heart Infusion 5µm, Aerobic 37°C 30. Filtration Brain Heart Infusion 5µm, Anaerobic 37°C 31. Glu Asp Aerobic 37°C 32. HTM Aerobic 37°C, 2,5% CO2 33. M17, Aerobic 37°C 34. Marine medium, Aerobic 28°C 35. MOD-2, Aerobic 37°C 36. MRS, Anaerobic 37°C 37. Mueller Hinton vancomycine 50µg/l, Aerobic 38. Mueller Hinton Ae 37°C 39. Orange Ae 37°C 40. Passive filtration with BSKH, 5% sheep blood agar, Aerobic 37°C 41. Passive filtration with Leptospira broth, 5% sheep blood agar, aerobic 37°C

128 4 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

42. Passive filtration with Leptospira broth, 5% sheep blood agar, microaerophilic 37°C 43. Phage T1 + T4, then 5% sheep blood agar, aerobe 37°C 44. Preincubation in blood culture bottle 11 days with rumen, 5% sheep blood agar, Anaerobic 37°C 45. Preincubation in blood culture bottle 2 days with rumen + filtered stool, 5% sheep blood agar, Anaerobic 37°C 46. Preincubation in blood culture bottle 26 days with rumen and sheep blood, 5% sheep blood agar Anaerobic 37°C 47. Preincubation in blood culture bottle 3 days with 5 ml sheep blood, 5% sheep blood agar Anaerobic 37°C 48. Preincubation in blood culture bottle 3 days with rumen and sheep blood, 5% sheep blood agar, Anaerobic 37°C 49. Preincubation in blood culture bottle 4 days after thermic shock, 5% sheep blood agar, Aerobic 45°C 50. Preincubation in blood culture bottle Anaerobic 10 days, Brain Heart Infusion, Anaerobic 37°C 51. Preincubation in blood culture bottle Anaerobic 10 days, Schadler Kana-Vanco, Anaerobic 37°C 52. Preincubation in blood culture bottle Anaerobic 14 days with 8ml rumen fluid, 5% sheep blood agar, Anaerobic 37°C 53. Preincubation in blood culture bottle Anaerobic 3 days with 8ml rumen fluid, 5% sheep blood agar, Anaerobic 37°C 54. Preincubation in blood culture bottle Anaerobic 5 days, 5% sheep blood agar Anaerobic 37°C 55. Preincubation in blood culture bottle Anaerobic 5 days, Schaedler kana vanco 37°C 56. Preincubation in blood culture bottle with thioglycolate, 4 days, 5% sheep blood agar Anaerobicerobe 37°C 57. Preincubation in in blood culture bottle aerobic 3 days with 8ml rumen fluid, 5% sheep blood agar, Aerobic 37°C 58. Preincubation in in blood culture bottle Anaerobic 30 days with 8ml rumen fluid, 5% sheep blood agar, Anaerobic 37°C 59. PVX Ae 37°C 5% CO2 60. R2A, Aerobic 37°C 61. Sabouraud 37°C 62. Schaedler Kana Vanco, Anaerobic 37°C 63. Schaedler, Anaerobic 37°C 64. Thermic shock 65°C 5% sheep blood agar Ae 37°C 65. Thermic shock, Preincubation in blood culture Anaerobic 15 days, 5% sheep blood agar, Anaerobic 37°C 66. Thermic shock, Preincubation in blood culture Anaerobic 40 days, 5% sheep blood agar, Anaerobic 37°C 67. Thermic shock, Preincubation in blood culture Anaerobic 50days, 5% sheep blood agar, Anaerobic 37°C 68. Thioglycolate + agar + 5% human blood, microaerophilic 37°C 69. TSA, Aerobic 37°C 70. Wilkins Chalgren, Aerobic 37°C

129 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 5

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 2B: The 18 standardized liquid culture conditions ranked in ascending number of species isolated.

Culturomics standardization by 18 culture conditions

1. Preincubation in aerobic blood culture bottle with stool filtered at 5µm, then 5% sheep blood agar, Aerobic condition, 37°C 2. Preincubation in anaerobic blood culture bottle after thermic shock at 80°C during 20min, then 5% sheep blood agar, Anaerobic condition, 37°C 3. Preincubation in anaerobic condition in marine broth, then 5% sheep blood agar, Anaerobic condition, 37°C 4. Preincubation in aerobic blood culture bottle with 5ml sheep blood, then 5% sheep blood agar, Aerobic condition, 37°C 5. Preincubation in aerobic marine broth, then 5% sheep blood agar, Aerobic condition 37°C 6. Preincubation in aerobic blood culture bottle with rumen fluid and sheep blood, then 5% sheep blood agar, Aerobic condition, 37°C 7. Preincubation in aerobic condition in trypticase soy broth, then 5% sheep blood agar, Aerobic condition, 37°C 8. Preincubation in aerobic condition in 5% sheep blood broth, then 5% sheep blood agar, Aerobic condition, 37°C 9. Preincubation in anaerobic blood culture bottle with stool filtered at 5µm, then 5% sheep blood agar, Anaerobic condition, 37°C 10. Preincubation in aerobic condition in Brain Heart Infusion broth with 5% sheep blood, then 5% sheep blood agar, Aerobic condition, 37°C 11. Preincubation in anaerobic condition in 5% sheep blood broth, then 5% sheep blood agar , Anaerobic condition, 28°C 12. Preincubation in aerobic blood culture bottle with rumen fluid, then 5% sheep blood agar , Aerobic condition, 37°C 13. Preincubation in aerobic condition in 5% sheep blood broth, then 5% sheep blood agar, Aerobic condition, 28°C 14. Preincubation in anaerobic blood culture bottle with 5ml sheep blood, then 5% sheep blood agar , Anaerobic condition, 37°C 15. Preincubation in anaerobic blood culture bottle, then 5% sheep blood agar , Anaerobic condition, 37°C 16. Preincubation in anaerobic blood culture bottle with rumen fluid and sheep blood, then 5% sheep blood agar, Anaerobic condition, 37°C 17. Preincubation in anaerobic condition in 5% sheep blood broth, then 5% sheep blood agar, Anaerobic condition, 37°C 18. Preincubation in anaerobic blood culture bottle with rumen fluid, then 5% sheep blood agar , Anaerobic condition, 37°C

130 6 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 2C: Culture conditions used during the cohort that included 15 preterm neonates with necrotizing enterocolitis and the 15 controls.

54 culture conditions performed for the NEC cohort

1. 5% Columbia sheep blood agar, aerobe, 37°C 2. 5% Columbia sheep blood agar, anaerobe, 37°C 3. Chocolat Polyvitex, aerobe, 37°C 4. Chocolat Polyvitex, anaerobe, 37°C 5. Mueller Hinton, aerobe, 37°C 6. Mueller Hinton, anaerobe, 37°C 7. Schaedler Kana-Vanco + 5% sheep blood, aerobe, 37°C 8. Schaedler Kana-Vanco + 5% sheep blood, anaerobe, 37°C 9. Brain Heart Infusion Agar, aerobe, 37°C 10. Brain Heart Infusion Agar, anaerobe, 37°C 11. Brain Heart Infusion + sheep blood 5% + vancomycin 10 µg/ml, aerobe, 37°C 12. Brain Heart Infusion + sheep blood 5% + vancomycin 10 µg/ml, anaerobe, 37°C 13. Orange aerobe, aerobe, 37°C 14. Bromocresol Purple, 37°C 15. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, aerobe, 37°C 16. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 17. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, aerobe, 37°C 18. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 19. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, aerobe, 37°C 20. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 21. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, aerobe, 37°C 22. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 23. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, aerobe, 37°C 24. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 25. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, aerobe, 37°C 26. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, anaerobe, 37°C 27. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, aerobe, 37°C 28. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, anaerobe, 37°C 29. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, aerobe, 37°C 30. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then

131 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 7

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Chocolat Polyvitex agar, anaerobe, 37°C 31. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, aerobe, 37°C 32. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, anaerobe, 37°C 33. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, aerobe, 37°C 34. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Chocolat Polyvitex agar, anaerobe, 37°C 35. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, aerobe, 37°C 36. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, anaerobe, 37°C 37. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, aerobe, 37°C 38. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, thenSchaedler Kana-Vanco + 5% sheep blood agar, anaerobe, 37°C 39. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, aerobe, 37°C 40. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, anaerobe, 37°C 41. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, aerobe, 37°C 42. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, anaerobe, 37°C 43. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, aerobe, 37°C 44. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Schaedler Kana-Vanco + 5% sheep blood agar, anaerobe, 37°C 45. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, aerobe, 37°C 46. Inoculation blood culture bottle day 2 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, anaerobe, 37°C 47. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, aerobe, 37°C 48. Inoculation blood culture bottle day 3 with 5 ml sheep blood, 10 mL Rumen, then 5% Columbia sheep blood agar, anaerobe, 37°C 49. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, aerobe, 37°C 50. Inoculation blood culture bottle day 7 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, anaerobe, 37°C 51. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, aerobe, 37°C 52. Inoculation blood culture bottle day 14 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, anaerobe, 37°C 53. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, aerobe, 37°C 54. Inoculation blood culture bottle day 30 with 5 ml sheep blood, 10 mL Rumen, then Brain Heart Infusion agar, anaerobe, 37°C

132 8 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 3: List of the 477 species isolated from gut by other laboratories but not by the previously published works of culturomics, classified by phyla and by preferred atmosphere of growth.

133 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 9

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 4A: List of the 1,057 bacterial species isolated in this project; new species are highlighted in bold. We used the main categories than in Fig. 1 and in main manuscript. C= 70 culture conditions; D= 18 culture conditions; E= cohorts; F= Fresh stools; G= Proteobacteria project; H= Microaerophilic project; I= Halophilic bacteria project; J= Microcolonies detection; K= Small bowel and colonic samples. H (GUT) = previously known from human gut; NH : 1st isolation in human; H = 1st isolation in human gut but previously known from human; newspe= new species. X= species isolated in culturomics for the first time in this project.

134 10 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 4B: List of the 1,170 bacterial species isolated by culturomics including previously published studies (columns A and B). A= First culturomics study; B= other previously published studies; C= 70 culture conditions; D= 18 culture conditions; E= cohorts; F= Fresh stools; G= Proteobacteria project; H= Microaerophilic project; I= Halophilic bacteria project; J= Microcolonies detection; K= Small intestine and colonic samples. H(GUT) = previously known from human gut; NH : 1st isolation in human; H = 1st isolation in human gut but previously known from human; newspe= new species. X= species isolated in culturomics for the first time in this project.

135 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 11

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 5: List of the 247 new bacterial species isolated by culturomics (197 in the present study plus 50 previously reported species) with CSUR accession number, DSMZ accession number, and GenBank accession number of the 16SrRNA. For the bacterial species already described in the literature, the PMID is available, as well as the list number of the species already officially recognized by the International Committee of Taxonomy. Bacteria that we failed to subculture are highlighted by a star (*). New bacterial species published as ‘new species announcement’ are highlighted by a pound (#), The Archaea species are highlighted by two stars (**). Finally, the 50 new bacterial species are underlined in grey. *subculture failure; # archaea; NA : not available

136 12 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 6: Reads recovered by metagenomics from the HMP data. The 247 16S rRNA of new bacterial species are reported in the table. *subculture failure, # archaea

137 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 13

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table7: Main characteristics of the genomes of the new bacterial species isolated by culturomics (genome size, ORFans number and percentage).*subculture failure, # archaea

138 14 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 8: Number of ORFans identified from our new species and retrieved from the assembly of whole metagenomic studies by HMP.

ORFans Species finding Ihuella massiliensis 136 Butyricimonas phoceensis 132 Intestinimonas massiliensis 112 Clostridium phoceense 99 Bacteroides timonense 95 Alistipes ihumii 95 Blautia phoceensis 83 Clostridium saudii 76 Clostridium bouchedurhonense 63 Gabonia massiliensis 55 Polynesia massiliensis 43 Alistipes jeddahensis 39 Neglecta timonensis 36 Alistipes obesihominis 35 Blautia massiliensis 25 Sanguibacteroides massiliense 22 Megasphaera massiliensis 21 Guyana massiliensis* 19 Alistipes provencensis 14 Bacteroides neonati 11 Dakarella massiliensis 9 Ruminococcus phoceensis 9 Fournierella massiliensis 8 Alistipes senegalensis 7 Drancourtella massiliensis 7 Emergencia timonensis 6 Holdemania massiliensis 6 Alistipes timonensis 6 Stoquefichus massiliensis 6 Clostridium dakarense 5 Collinsella massilioamazoniensis 5 Intestinimonas gabonensis 5 Gorbachella massiliensis 5 Stoquefichus jeddahensis 3 Senegalimassilia anaerobia 3 Christensenella massiliensis 2 Ruminococcus massiliensis* 2 Eubacterium massiliense 2 Gorillibacterium timonense 2 Clostridium massiliosenegalense 2 Prevotella phoceensis 2

139 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 15

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Christensenella timonensis 1 Clostridium nigeriense 1 Dielma fastidiosa 1 Peptoniphilus grossensis 1 Peptoniphilus obesihominis 1 Clostridium amazonitimonense 1 Bacillus massiliosenegalensis 1 Clostridium senegalense 1 Enterobacter massiliensis 1 Enorma massiliensis 1 Niameyia massiliensis 1 Anaerococcus rubiinfantis 1 Anaerotruncus rubiinfantis 1 Total 1,326 *subculture failure

140 16 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 9: List of the new species identified in the 396 and 239 human gut microbiome samples from Nielsen et al. and Browne et al., respectively15,16.*subculture failure

141 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 17

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 10: List of the 84 metagenomes used for the study exploring the dark matter. We used the main categories than in Fig. 1 and in main manuscript : B= other previously published studies; C= 70 culture conditions; D= 18 culture conditions; E= cohorts; F= Fresh stools; I= Halophilic bacteria project;

Metagenomic Category of code Stool specimen project A315 Amazonie 2 C M1034 BLA bariatric (before) E M1030 MC bariatric (before) E M1015 BMI bariatric (after) E A309 Amazonie 20 D M1019 GMT bariatric E A31 VIH8 C A150 SAN-7 D M933 LL bariatric (after) E A420 Kw 1 D KW01 KW01 Senegal D M1046 CCH bariatric (before) E M631 BD4 D M1037 DN bariatric (after) E M407 Senegalese Stool NDIOP 6 I M932 LL bariatric (before) E M406 Senegalese Stool NDIOP 5 I M638 BD11 Stool human bédouin D M1018 GMT bariatric (before) E M927 Senegalese Stool NDIOP 4 I KW04 KW04 D M405 Senegalese Stool NDIOP 4 I M400 VIH 12_Protocole 5 C M394 VIH 3_Protocole 5 C M1016 MS bariatric (before) E M1017 MS bariatric (after) E A430 Kw 14 D A144 SAN-1 D KW06 KW 06 D M404 Senegalese Stool NDIOP 3 J M926 CC bariatric (before) E M1033 HA bariatric (after) E M901 N12 control Niger D M401 VIH 13_Protocole 5 C A312 Amazonie 15 D M1035 BLA bariatric (after) E S07 Control S07 Sénégal D M893 N2 control Niger D

142 18 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

A428 Kw 12 D M1032 HA bariatric (before) E M1047 CCH bariatric (after) E A310 Amazonie 1 C M629 BD2 Stool human bédouin D A8 Touareg D A140 OB-6 D GD2 Fresh stool 2 F M1036 DN bariatric (before) E M1029 FE bariatric (after) E M1021 BC bariatric (after) E KW02 KW02 D A307 Amazonie 14 D M935 MM bariatric (after) E A418 MR9 D M636 BD9 Stool human bédouin D A2 Réa 2 B A39 Polynésie 8 C A9 VIH1 C A425 Kwashiorkor 8 D M930 BM bariatric (before) E M1020 BC bariatric (before) E KW10 Kw 10 D M1014 BMI bariatric (before) E A302 Amazonie 10 D M637 BD10 Stool human bédouin D M897 N6 control Niger E M1031 MC bariatric (after) E A10 Réa 1 B MR09 MR09 D M929 ACZ bariatric (after) E A21 Inde C S05 Control S05 Sénégal D M931 BM bariatric (after) E A3 Raiatea C M928 ACZ bariatric (before) E A17 VIH4 C A137 OB-3 D M402 Senegalese Stool NDIOP 1 I M635 BD8 D M1028 FE bariatric (before) E M403 Senegalese Stool NDIOP 2 I M934 MM bariatric (before) E A294 Touareg 8 D A46 VIH11 C A135 OB1 D

143 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 19

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 11: Identification of Operational Taxonomic Units (OTUs) and assigned species from the gut metagenomes: The information concerning these samples is available in Supplementary Table 10.

N° of Sample Total OTU's without Total of Non-Human Non-gut Human New Gut bacteria Non-gut bacteria sample Name OTUs identified bacterial bacteria bacteria bacteria species not found not found number species species bacteria by culturomics by culturomics 1 A315 800 751 49 2 1 17 12 13 4 2 M1034 450 379 71 3 6 37 9 13 3 3 M1030 499 445 54 4 4 20 12 14 0 4 M1015 432 350 82 4 6 49 9 8 6 5 A309 656 600 56 2 3 21 12 13 5 6 M1019 358 304 54 1 4 32 11 6 0 7 A31 853 801 52 3 3 17 13 14 2 8 A150 64 21 43 1 1 22 11 8 0 9 M933 1311 1237 74 3 5 33 14 12 7 10 A420 431 389 42 2 3 22 6 6 3 11 KW01 484 443 41 1 4 28 1 4 3 12 M1046 362 316 46 1 2 23 7 10 3 13 M631 541 483 58 4 4 24 11 13 2 14 M1037 238 194 44 1 3 24 7 6 3 15 M407 349 270 79 6 6 31 14 15 7 16 M932 1140 1087 53 1 5 26 7 12 2 17 M406 296 249 47 4 2 23 5 10 3 18 M638 554 495 59 3 2 31 6 13 4 19 M1018 342 305 37 1 0 20 8 8 0 20 M927 997 934 63 1 3 35 9 12 3 21 KW04 200 157 43 2 6 24 4 3 4 22 M405 374 309 65 11 1 18 14 14 7 23 M400 287 261 26 0 0 21 2 1 2 24 M394 430 389 41 1 2 18 11 7 2 25 M1016 301 265 36 2 0 15 7 11 1 26 M1017 454 399 55 4 0 26 8 16 1 27 A430 540 491 49 6 7 22 6 3 5 28 A144 83 20 63 3 4 28 12 9 7 29 KW06 125 92 33 0 5 21 3 3 1 30 M404 185 123 62 11 5 17 10 14 5 31 M926 951 876 75 0 5 43 10 9 8 32 M1033 380 326 54 1 3 32 6 9 3 33 M901 956 888 68 4 3 36 12 9 4 34 M401 286 261 25 1 1 12 5 5 1 35 A312 462 406 56 4 1 19 12 18 2 36 M1035 424 352 72 4 4 40 14 7 3 37 S07 807 750 57 5 6 30 6 5 5 38 M893 728 670 58 2 1 29 12 12 2 39 A428 527 486 41 2 4 29 2 2 2 40 M1032 426 368 58 1 3 30 13 11 0 41 M1047 459 399 60 6 3 26 8 16 1 42 A310 596 539 57 4 1 22 10 14 6 43 M629 692 589 103 4 8 48 13 13 17 44 A8 439 366 73 6 1 31 25 10 0 45 A140 94 39 55 1 2 24 10 16 2 46 GD2 119 105 14 0 0 7 6 0 1 47 M1036 274 236 38 0 2 20 8 7 1 48 M1029 302 249 53 1 3 25 14 10 0 49 M1021 275 236 39 1 1 20 7 10 0 50 KW02 1084 1021 63 3 1 43 7 4 5 51 A307 638 590 48 3 0 16 11 14 4 52 M935 1217 1126 91 4 9 42 18 14 4 53 A418 248 210 38 2 5 25 4 1 1 54 M636 815 759 56 3 2 27 8 14 2 55 A2 135 96 39 0 1 21 13 2 2 56 A39 276 234 42 2 0 22 9 8 1 57 A9 625 567 58 0 1 32 15 9 1 58 A425 189 143 46 2 3 24 8 7 2 59 M930 1202 1161 41 2 1 18 8 11 1 60 M1020 313 275 38 2 0 22 6 6 2 61 KW10 1429 1354 75 2 2 45 17 8 1 62 M1014 416 364 52 0 1 27 12 6 6 63 A302 616 570 46 2 0 18 10 14 2 64 M637 611 559 52 2 1 27 9 11 2 65 M897 1320 1231 89 5 0 45 18 16 5 66 M1031 605 544 61 3 2 29 13 13 1

144 20 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

67 A10 478 435 43 2 4 26 8 2 1 68 MR09 484 457 27 0 2 17 3 3 2 69 M929 1445 1345 100 2 6 51 20 19 2 70 A21 407 371 36 0 1 18 7 10 0 71 S05 1925 1843 82 2 1 38 16 18 7 72 M931 844 767 77 5 4 33 15 10 10 73 A3 131 110 21 0 0 17 2 1 1 74 M928 1242 1167 75 5 6 39 7 6 12 75 A17 812 743 69 3 2 36 13 15 0 76 A137 91 12 79 4 5 39 10 13 8 77 M402 318 275 43 9 0 12 7 11 4 78 M635 522 467 55 3 1 20 11 14 6 79 M1028 378 326 52 2 7 28 6 8 1 80 M403 275 220 55 7 0 15 10 17 6 81 M934 723 641 82 5 10 37 14 13 3 82 A294 868 802 66 5 0 23 14 20 4 83 A46 430 384 46 1 1 22 12 8 2 84 A135 73 15 58 3 2 28 12 11 2 Total unique 4158 3602 556 47 61 210 102 50 86 Culturomics (420) Not Culturomics (136)

145 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 21

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 12: List of the 120 species isolated by Browne et al.15

closest 16S BLAST match to named Phylum bacterium-99 or 100% match unless Novelty Type of species otherwise stated 1 Bacteroidetes Alistipes finegoldii characterised H(Gut) 2 Firmicutes Anaerostipes hadrus characterised H(Gut) 3 Firmicutes Anaerostipes hadrus_98% novel species NS 4 Firmicutes Anaerotruncus colihominis characterised H(Gut) 5 Firmicutes Anaerotruncus colihominis_91% novel genus NS 6 Bacteroidetes Bacteroides caccae characterised H(Gut) 7 Bacteroidetes Bacteroides coprocola_94% novel species NS 8 Bacteroidetes Bacteroides finegoldi characterised H(Gut) 9 Bacteroidetes Bacteroides finegoldii_98% novel species NS 10 Bacteroidetes Bacteroides intestinalis_98% novel species NS 11 Bacteroidetes Bacteroides ovatus characterised H(Gut) 12 Bacteroidetes Bacteroides plebius_95% novel species NS 13 Bacteroidetes Bacteroides salyersiae characterised H(Gut) 14 Bacteroidetes Bacteroides thetaiotaomicron characterised H(Gut) 15 Bacteroidetes Bacteroides uniformis characterised H(Gut) 16 Bacteroidetes Bacteroides vulgatus characterised H(Gut) 17 Bacteroidetes Bacteroides xylanisolvens characterised H(Gut) 18 Firmicutes Balutia luti_96% novel species NS 19 Actinobacteria Bifidobacterium adolescentis characterised H(Gut) 20 Actinobacteria Bifidobacterium bifidum characterised H(Gut) 21 Actinobacteria Bifidobacterium pseudocatenulatum characterised H(Gut) 22 Firmicutes Blautia hydrogenotrophica characterised H(Gut) 23 Firmicutes Blautia hydrogenotrophica_96% novel genus NS 24 Firmicutes Blautia luti_95% novel species NS 25 Firmicutes Blautia luti_96% novel species NS 26 Firmicutes Blautia luti_98% novel species NS 27 Firmicutes Blautia obesum characterised H(Gut) 28 Firmicutes Blautia producta_94% novel species NS 29 Firmicutes Blautia wexlerae characterised H(Gut) 30 Firmicutes Butyricicoccus pullicaecorum_94% novel species NS 31 Firmicutes Butyricicoccus pullicaecorum_94% novel species NS 32 Firmicutes Catenibacterium mitsuokai characterised H(Gut) 33 Firmicutes Clostridium baratti characterised H(Gut) 34 Firmicutes Clostridium bartlettii characterised H(Gut) 35 Firmicutes Clostridium boltae_94% novel species NS 36 Firmicutes Clostridium celerecrescens_93% novel genus NS 37 Firmicutes Clostridium celerescens_93% novel genus NS 38 Firmicutes Clostridium clostridioforme characterised H(Gut) 39 Firmicutes Clostridium clostridioforme_93% novel genus NS 40 Firmicutes Clostridium clostridioforme_98% novel species NS 41 Firmicutes Clostridium cocleatum_93% novel species NS 42 Firmicutes Clostridium disporicum characterised H(Gut) 43 Firmicutes Clostridium disporicum_98% novel species NS

146 22 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

44 Firmicutes Clostridium ghonii_98% novel species NS 45 Firmicutes Clostridium hathewayi characterised H(Gut) 46 Firmicutes Clostridium hathewayi_92% novel genus NS 47 Firmicutes Clostridium innocuum characterised H(Gut) 48 Firmicutes Clostridium innocuum_95% novel species NS 49 Firmicutes Clostridium lituseburense_98% novel species NS 50 Firmicutes Clostridium methylpentosum_92% novel species NS 51 Firmicutes Clostridium nexile_94% novel species NS 52 Firmicutes Clostridium oroticum_95% novel genus NS 53 Firmicutes Clostridium oroticum_95% novel genus NS 54 Firmicutes Clostridium oroticum_96% novel species NS 55 Firmicutes Clostridium paraputrificum characterised H(Gut) 56 Firmicutes Clostridium perfringens characterised H(Gut) 57 Firmicutes Clostridium saccharogumia_93% novel species NS 58 Firmicutes Clostridium saccharolyticum_94% novel species NS 59 Firmicutes Clostridium thermocellum_86% novel family 2 NS 60 Firmicutes Clostridium thermocellum_87% novel genus NS 61 Firmicutes Clostridium xylanolyticum_95% novel genus NS 62 Firmicutes Clostridium xylanolyticum_96% novel species NS 63 Actinobacteria Collinsella aerofaciens characterised H(Gut) 64 Actinobacteria Collinsella aerofaciens_92% novel species NS 65 Firmicutes Coprococcus comes characterised H(Gut) 66 Firmicutes Coprococcus eutactus characterised H(Gut) 67 Firmicutes Coprococcus eutactus_97% novel species NS 68 Firmicutes Dorea formicigenerans_98% novel species NS 69 Firmicutes Dorea longicatena characterised H(Gut) 70 Firmicutes Eubacterium contortum characterised H 71 Firmicutes Eubacterium contortum_97% novel genus NS 72 Firmicutes Eubacterium eligens characterised H(Gut) 73 Firmicutes Eubacterium fissicatens_95% novel species NS 74 Firmicutes Eubacterium hallii characterised H(Gut) 75 Firmicutes Eubacterium hallii_97% novel species NS 76 Firmicutes Eubacterium infirmum_91% novel family 1 NS 77 Firmicutes Eubacterium ramulus characterised H(Gut) 78 Firmicutes Eubacterium rectale characterised H(Gut) 79 Firmicutes Eubacterium siraeum characterised H(Gut) 80 Firmicutes Faecalibacterium prausnitzii characterised H(Gut) 81 Firmicutes Faecalibacterium prausnitzii_98% novel species NS 82 Firmicutes Flavonifractor plautii characterised H(Gut) 83 Firmicutes Flavonifractor plautii_94% novel species NS 84 Firmicutes Flavonifractor plautii_95% novel species NS 85 Firmicutes Flavonifractor plautii_95% novel genus NS 86 Firmicutes Flavonifractor plautii_95% novel species NS 87 Firmicutes Flavonifractor plautii_96% novel genus NS 88 Firmicutes Flavonifractor plautii_97% novel genus NS 89 Firmicutes Fusicatenibacter saccharivorans novel genus NS 90 Firmicutes Fusicatenibacter saccharivorans_93% novel genus NS 91 Firmicutes Lachnospira pectinoschiza characterised NH

147 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 23

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

92 Firmicutes Lachnospira pectinoschiza_91% novel genus NS 93 Firmicutes Megasphaera elsdenii_95% novel species NS 94 Firmicutes Mitsuokella jalaludinii characterised H(Gut) 95 Firmicutes Oscillibacter valericigenes_96% novel species NS 96 Bacteroidetes Parabacteroides distasonis characterised H(Gut) 97 Bacteroidetes Parabacteroides merdae characterised H(Gut) 98 Bacteroidetes Prevotella copri characterised H(Gut) 99 Firmicutes Roseburia faecis characterised H(Gut) 100 Firmicutes Roseburia faecis_95% novel genus NS 101 Firmicutes Roseburia hominis characterised H(Gut) 102 Firmicutes Roseburia intestinalis characterised H(Gut) 103 Firmicutes Roseburia inulinivorans_94% novel species NS 104 Firmicutes Roseburia inulinvorans characterised H(Gut) 105 Firmicutes Ruminococcus albus_95% novel species NS 106 Firmicutes Ruminococcus albus_98% novel species NS 107 Firmicutes Ruminococcus bromii characterised H(Gut) 108 Firmicutes Ruminococcus bromii_93% novel species NS 109 Firmicutes Ruminococcus bromii_94% novel species NS 110 Firmicutes Ruminococcus flavefaciens_93% novel genus NS 111 Firmicutes Ruminococcus flavefaciens_95% novel species NS 112 Firmicutes Ruminococcus gnavus characterised H(Gut) 113 Firmicutes Ruminococcus gnavus_98% novel species NS 114 Firmicutes Ruminococcus obeum characterised H(Gut) 115 Firmicutes Ruminococcus obeum_96% novel species NS 116 Firmicutes Ruminococcus obeum_98% novel species NS 117 Firmicutes Ruminococcus torques characterised H(Gut) 118 Firmicutes Ruminococcus torques_97% novel genus NS 119 Firmicutes Sarcina ventriculi characterised H(Gut) 120 Firmicutes Turicibacter sanguinis characterised H(Gut)

148 24 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 13: List of the bacterial species formerly isolated by culturomics and then identified in our clinical microbiology laboratory in La Timone, Marseille, France.

CSUR accession Type of N° of time Strain number samples isolated Actinomyces grossensis CSURP242 Puncture fluid 4 Abcesses 1 Actinomyces ihumii CSURP2006 Abcesses 6 Anaerococcus senegalensis CSURP156 Puncture fluid 1 Anaerosalibacter massiliensis CSURP762 Diarrhea 1 Bacteroides timonense CSURP194 Blood culture 3 Butycirimonas phoceencis CSURP2478 Diarrhea 1 Clostridium jeddahtimonense CSURP1230 Diarrhea 7 Clostridium massilioamazoniensis CSURP1360 Diarrhea 1 Clostridium saudii CSURP697 Diarrhea 12 Peptoniphilus grossensis CSURP184 Abcesses 5 Biopsies 6 Material 1 Puncture fluid 5 Polynesia massiliensis CSURP1280 Peritoneal fluid 1 Pseudomonas massiliensis CSURP1334 Urine 1 Skin 1 Total 57

149 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 25

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 14: Bacterial species isolated by culturomics, unidentified by MALDI-TOF and requiring 16SrRNA amplification and sequencing for identification.

150 26 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. SUPPLEMENTARY INFORMATION

Supplementary Table 15: Genome of the bacterial species isolated from the gastrointestinal tract and referenced by the Human Microbiome Project (available 29 March 2016). The bacterial species officially recognized by the International Committee on Systematic Bacteriology are highlighted in green.

151 NATURE MICROBIOLOGY | www.nature.com/naturemicrobiology 27

© 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

Article 4

Dynamic variations of anaerobes during non-specific liquid

enrichment of human fresh stool sample

Guillaume André Durand, Dipankar Bachar, Didier Raoult, Emmanouil Angelakis

Soumis dans Current Microbiology

152 1 Dynamic variations of anaerobes during non-specific liquid enrichment of human fresh stool sample

2 Guillaume André Durand a*, Dipankar Bachar a, Didier Raoult a, Emmanouil Angelakis b

3 a Aix Marseille Univ, IRD, AP-HM, MEPHI, IHU Méditerranée Infection, Marseille, France

4 b Aix Marseille Univ, IRD, AP-HM, SSA, VITROME, IHU Méditerranée Infection, Marseille, France

5 * Corresponding author : Dr Guillaume Durand, Aix Marseille Univ, IRD, AP-HM, MEPHI, IHU

6 Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13005 Marseille, [email protected], Phone:

7 (33) 413 73 24 01, Fax: (33) 413 73 24 02

8 Text words count: 1,520

9 Tables / Figures: 3

10 References: 21

11 Keywords: culturomics, metagenomic, gut microbiota

153 12 Abstract

13 Some bacterial species are fastidious to grow under standard conditions, notably anaerobes. The last represents

14 the dominant gut microbiota flora. The aim of this study was to evaluate the modifications that occur in the

15 oxygen sensitive stool bacteria populations after liquid incubation in presence of blood and rumen fluid. Using

16 pyrosequencing and culturomics, we analysed a fresh stool sample incubated at 37°C during ten days into an

17 anaerobic blood bottle with 5% sheep blood and 5% rumen fluid. Our enrichment medium increased the number

18 of isolated bacteria from six to 41 species while Faecalibacterium and Akkermansia genera decreased. We

19 observe a decrease of biodiversity from Shannon 4.36 to 3.66, and a decrease at day two of the number of reads

20 and OTUs assigned whereas it increases after ten days of incubation. Some OTUs were increased at all time

21 (Escherichia coli, increase of 25,294 reads, Gammaproteobacteria, increase of 3,881 reads and Neglecta

22 phocaeensis increase of 1,385 reads) whereas others are decreased at all time (Akkermansia muciniphila,

23 decreased of 6,489 reads, Faecalibacterium prausnitzii, decreased of 4,867 reads and Bacteroides uniformis

24 decreased of 4,793 reads). Interestingly, some OTUs were increased at days two prior to decreased after ten days

25 of incubation (Oscilibacter massiliensis, Ruminococcus gnavus and Blautia wexlerae) and the reverse

26 (Selenomonadales, Tidjanibacter massiliensis, and Negativicutes). In conclusion, we provide evidence that non-

27 selective liquid medium using blood and rumen fluid increased the number of isolated bacterial species even

28 after only two days of incubation and can plays a critical role in the efforts to increase the repertory of gut

29 microbiota.

154 30 Acknowledgements

31 This work was supported by the French Government under the « Investissements d’avenir » (Investments for the

32 Future) program managed by the Agence Nationale de la Recherche (ANR, fr: National Agency for Research),

33 (reference: Méditerranée Infection 10-IAHU-03). The study was approved by the ethic committee of the IHU

34 Méditerranée-Infection (agreement number 2016-011).

35 Conflict of Interest

36 Nothing to declare.

155 1 Dynamic variations of anaerobes during non-specific liquid enrichment of human fresh stool

2 sample

3 Introduction

4 Bacterial culture is the cornerstone for both clinical microbiology and investigation of the human gut

5 flora [1, 2]. However in many cases, culture appears difficult as some bacterial species are fastidious to grow

6 under standard conditions. To resolve this, various approaches have been proposed, including the enrichment of

7 the growth medium (cooked meat, blood, antioxidants, …), the inhibition of other bacterial species using various

8 molecules (antibiotics, green malachite, bile, etc…), the cellular culture and the adjustment of growth conditions

9 (temperature, time of incubation and oxygen exposure time) [3–7]. For example, blood use have previously

10 allowed the axenic growth of Borrelia recurrentis and Mycobacterium tuberculosis [8, 9]. Moreover, blood

11 liquid culture systems has improved the isolation of low-counted or of tedious bacteria in clinical microbiology,

12 e.g. by Kingella Kingae [10].

13 Culturomics consists in isolating the maximum number of species using various conditions of media

14 and atmosphere. This approach has expanded the exploration of gut microbiota, increasing its repertory from 690

15 to 1,525 bacterial species the last few years [2]. Metagenomic studies have found that anaerobes represent the

16 dominant gut microbiota flora [11]. The anaerobic preincubation using rumen fluid and sheep blood was

17 previously considered as the better condition of culturomics for this purpose [2]. In this work, our objective was

18 to evaluate using both culture and metagenomic, the modifications that occur in the oxygen sensitive stool

19 bacteria populations after incubation of our enrichment medium.

20 Material and Methods

21 We tested a stool sample from a healthy 28-year-old French volunteer (Body Mass Index 22 m².kg-1),

22 who had not received antibiotics the year before the sample was taken. The sample was incubated for 8 minutes

23 after release in a homemade liquid medium consisting of an anaerobic blood bottle (Becton-Dickinson

24 Diagnostics, Le Pont-de-Claix, France) previously enriched with 5% sheep blood (Eurobio SA, Montpellier,

25 France) and 5% autoclaved-supernatant of rumen at 37°C. The fresh sample, as well as the content of the blood

26 bottles after two and ten days of incubation were tested (Supplementary Fig. 1). Culturomics under anaerobic

27 conditions was performed as previously described [2].

156 28 The fresh stool sample, the medium before inoculation and the content of medium after two and ten

29 days of incubation were all sequenced for V4-V3 regions of 16S rRNA gene using MiSeq technology. 16S

30 rRNA results of the blood bottle without stools sample were considered DNA contamination and were therefore

31 removed from further metagenomic analysis. The paired end reads of corresponding raw fastq files were

32 assembled into longer joined sequences using FLASH. High-quality sequences were then filtered using QIIME,

33 as previously described [12]. Primers were trimmed and sequences shorter than 200 nucleotides and greater than

34 500 nucleotides were removed. The sequences were grouped into operational taxonomic units (OTUs) using

35 UCLUST. The best taxonomic assignment was performed for each OTUs, using threshold of 97% for species

36 level, 95% for genera level, and 80% for phyla level. OTUs were extracted de novo, without considering

37 singletons [12].

38 Results

39 Using culturomics, six bacterial species belonging to the four genera Enterococcus, Escherichia,

40 Bacteroides and Lactococcus were isolated from the fresh stool sample before inoculation. In contrast, we

41 isolated 41 different bacterial species, corresponding to 19 genera after two days incubation and 33 different

42 bacterial species belonging to 13 genera after ten days incubation. Our medium allowed us to isolate 44 bacterial

43 species that were not isolated directly from the stool sample, which 16 of them were only after two days of

44 incubation, only nine after ten days and 19 after both two and ten days of incubation (Fig. 1, Supplementary

45 Table 1).

46 16S rRNA sequencing revealed that the fresh stool sample contained 99,428 high quality reads that

47 were assigned to 435 OTUs, mostly belonging to Firmicutes and Bacteroidetes (Supplementary Table 2). The

48 analysis of sample after two days of incubation revealed 30,773 high quality trimmed reads assigned to 269

49 OTUs, mostly belonging to Proteobacteria followed by Bacteroidetes. Similarly, following ten days of

50 incubation, we amplified 85,627 high quality trimmed reads, assigned to 376 OTUs, mostly belonging to

51 Firmicutes followed by Proteobacteria and Bacteroidetes. At the phyla level, our medium increased the

52 abundance of Proteobacteria and Actinobacteria and decreased the abundance of Bacteroidetes, Firmicutes and

53 Verrucomicrobia comparing to the fresh stool. The diversity, assessed by the Shannon index, decreased between

54 fresh sample (4.36) and two (4.02) or ten days (3.66) (Fig. 2).

55 A total of 69 OTUs were increased after both two and ten days of liquid incubation (Supplementary

56 Table 3). The most enriched OTUs were assigned to Escherichia coli (increase of 25,294 reads),

157 57 Gammaproteobacteria, (increase of 3,881 reads) and Neglecta phocaeensis (increase of 1,385 reads) (Fig. 3). At

58 the opposite, there were 295 OTUs decreased after both two and ten days. The most decreased were assigned to

59 Akkermansia muciniphila (decreased of 6,489 reads), Faecalibacterium prausnitzii (decreased of 4,867 reads)

60 and Bacteroides uniformis (decreased of 4,793 reads) (Fig. 3).

61 A total of 35 OTUs were increased after two days but decreased after ten days (Supplementary Table

62 4). The most OTUs firstly enriched and decreased after 10 days were assigned to Oscilibacter massiliensis,

63 Ruminococcus gnavus and Blautia wexlerae (Fig. 3). For these species, the liquid enrichment for two days

64 appeared optimal. A total of 66 OTUs were decreased firstly after two days but increased at ten days of

65 incubation (Supplementary Table 5). The most OTUs recovered after ten days were assigned to

66 Selenomonadales, Tidjanibacter massiliensis, and Negativicutes (Fig. 3). For these species, ten days of

67 incubation appeared as the optimal method for their enrichment.

68 Comparing culturomics and metagenomic, there were 36 species isolated for which less than 30 reads

69 were found by metagenomic. Moreover, 31 of them were not found by metagenomic (Supplementary Table 1).

70 At the opposite, despite un great number of reads, some species were not isolated. There were 26 OTUs present

71 with more than 1,000 reads from fresh sample that were not isolated as for example Bacteroides uniformis (5,629

72 reads, 5.7 %) or Gemmiger formicilis (4,486 reads, 4.5 %). Particularly, Faecalibacterium sp. were sequenced

73 with 8,698 reads (8.8 %) but not cultivated, as well as Akkermansia muciniphila for which 6,891 reads (6.9 %)

74 were sequenced.

75 Discussion

76 Incubation of a stool sample in our media was helpful because it increased the number of isolated

77 species compared to culturing a fresh sample without enrichment. Although we did not isolate new bacterial

78 species, this medium has previously permitted to isolate 247 new species [2]. Non-specific media was proposed

79 as critical for the isolation of fastidious species, particularly obligates anaerobes [13, 14]. To confirm our result,

80 we also analysed our samples by 16S rRNA sequencing which revealed that our medium decreased the stool’s

81 biodiversity. 16S rRNA pyrosequencing, as well as clonal, Sanger sequencing and studies of gut microbiota were

82 not able to detect bacterial concentrations that were below 107 CFU/mL [15]. It was previously found that when

83 the bacterial concentration is below 104 CFU/mL, which is the detection threshold limit for molecular assays,

84 diagnosis can only be achieved by culture [16, 17]. Moreover, given the short longer of reads generated, the

85 taxonomic assignment to species level is difficult [18]. Finally, our results confirm that culture is critical for the

158 86 exploration of the gut microbiota because for many isolated agents, 16S rRNA sequencing has been negative or

87 amplified very few reads (Supplementary Table 1).

88 Several bacterial species isolated were found in a low quantity of read or were not found from the

89 metagenomic results at the species taxonomy level, indicating the complementarity of culturomics with

90 metagenomic. The overlap between culturomics and metagenomic was 20.5 %, that is comparable to previous

91 studies [19]. At the opposite, some OTUs were largely present but not isolated, indicating that culturomics failed

92 to cultivate them, or that these sequences concern dead species at the time of culture. In this study we failed in

93 isolating Akkermansia sp. and Faecalibacterium prausnitzii. Moreover, we detected an important decrease of the

94 sequences of these species, indicating that they were probably not alive at the time of inoculation to our medium.

95 Their culture remains challenging as they are extremely sensitive to oxygen [20]. Akkermansia sp. have been

96 previously associated with metabolic diseases, like diabetes and obesity [21]. Although the use of non-selective

97 liquid media was proposed as critical for the isolation of fastidious species particularly obligates anaerobes [14],

98 we believe that the oxygen exposure of our stool sample before medium incubation killed these bacteria. To

99 solve this problem, we are now striving to reduce the exposure time of the fresh stool sample to oxygen before

100 incubation to our medium. Finally, this work pointed out that some species (OTUs assigned to Selenomonadales,

101 Tidjanibacter massiliensis, and Negativicutes) needed ten days to grow up, whereas other decreased after 48h of

102 incubation (OTUs assigned to Oscilibacter massiliensis, Ruminococcus gnavus and Blautia wexlerae).

103 In conclusion, we provide evidence that the use of our non-selective liquid medium significantly

104 increased the number of isolated bacterial species even after only two days of incubation. The use of blood and

105 rumen fluid allowed us to isolate many bacteria that were not isolated without enrichment. We believe that this

106 non-selective liquid medium plays a critical role in our efforts to increase the repertory on new bacterial species.

159 107 References

108 1. Editorial (2013) The cultural revolution. Nat Rev Microbiol 11:1

109 2. Lagier J-C, Khelaifia S, Alou MT, et al (2016) Culture of previously uncultured members of the human 110 gut microbiota by culturomics. Nat Microbiol 1:16203. https://doi.org/10.1038/nmicrobiol.2016.203

111 3. Fournier P-E, Drancourt M, Colson P, et al (2013) Modern clinical microbiology: new challenges and 112 solutions. Nat Rev Microbiol 11:574–585

113 4. Freney J, Renaud F, Leclercq R, Riegel P (2007) Précis de Bactériologie Clinique, 2ème édition. 114 ESKA/LACASSAGNE

115 5. Gouriet F, Fenollar F, Patrice J-Y, et al (2005) Use of shell-vial cell culture assay for isolation of bacteria 116 from clinical specimens: 13 years of experience. J Clin Microbiol 43:4993–5002. 117 https://doi.org/10.1128/JCM.43.10.4993-5002.2005

118 6. Singh S, Eldin C, Kowalczewska M, Raoult D (2013) Axenic culture of fastidious and intracellular 119 bacteria. Trends Microbiol 21:92–99. https://doi.org/10.1016/j.tim.2012.10.007

120 7. Dione N, Khelaifia S, La Scola B, et al (2016) A quasi-universal medium to break the aerobic/anaerobic 121 bacterial culture dichotomy in clinical microbiology. Clin Microbiol Infect Off Publ Eur Soc Clin 122 Microbiol Infect Dis 22:53–58. https://doi.org/10.1016/j.cmi.2015.10.032

123 8. Cutler SJ, Fekade D, Hussein K, et al (1994) Successful in-vitro cultivation of Borrelia recurrentis. Lancet 124 Lond Engl 343:242

125 9. Drancourt M, Raoult D (2007) Cost-effectiveness of blood agar for isolation of mycobacteria. PLoS Negl 126 Trop Dis 1:e83. https://doi.org/10.1371/journal.pntd.0000083

127 10. Yagupsky P (1999) Use of blood culture systems for isolation of Kingella kingae from synovial fluid. J 128 Clin Microbiol 37:3785

129 11. Lagier J-C, Dubourg G, Million M, et al (2018) Culturing the human microbiota and culturomics. Nat Rev 130 Microbiol. https://doi.org/10.1038/s41579-018-0041-0

131 12. Angelakis E, Yasir M, Bachar D, et al (2016) Gut microbiome and dietary patterns in different Saudi 132 populations and monkeys. Sci Rep 6:32191. https://doi.org/10.1038/srep32191

133 13. Duffy G, Walsh D, Sheridan JJ, et al (2001) Comparison of selective and non-selective enrichment media 134 in the detection of Listeria monocytogenes from meat containing Listeria innocua. J Appl Microbiol 135 90:994–999

136 14. Lagier J-C, Edouard S, Pagnier I, et al (2015) Current and past strategies for bacterial culture in clinical 137 microbiology. Clin Microbiol Rev 28:208–236. https://doi.org/10.1128/CMR.00110-14

138 15. Armougom F, Henry M, Vialettes B, et al (2009) Monitoring bacterial community of human gut 139 microbiota reveals an increase in Lactobacillus in obese patients and Methanogens in anorexic patients. 140 PloS One 4:e7125. https://doi.org/10.1371/journal.pone.0007125

141 16. Lagier J-C, Armougom F, Million M, et al (2012) Microbial culturomics: paradigm shift in the human gut 142 microbiome study. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis 18:1185–1193. 143 https://doi.org/10.1111/1469-0691.12023

144 17. Lynch MDJ, Neufeld JD (2015) Ecology and exploration of the rare biosphere. Nat Rev Microbiol 145 13:217–229. https://doi.org/10.1038/nrmicro3400

146 18. Jovel J, Patterson J, Wang W, et al (2016) Characterization of the Gut Microbiome Using 16S or Shotgun 147 Metagenomics. Front Microbiol 7:459. https://doi.org/10.3389/fmicb.2016.00459

160 148 19. Lagier J-C, Hugon P, Khelaifia S, et al (2015) The rebirth of culture in microbiology through the example 149 of culturomics to study human gut microbiota. Clin Microbiol Rev 28:237–264. 150 https://doi.org/10.1128/CMR.00014-14

151 20. Walker AW, Ince J, Duncan SH, et al (2011) Dominant and diet-responsive groups of bacteria within the 152 human colonic microbiota. ISME J 5:220–230. https://doi.org/10.1038/ismej.2010.118

153 21. Shen W, Shen M, Zhao X, et al (2017) Anti-obesity Effect of Capsaicin in Mice Fed with High-Fat Diet Is 154 Associated with an Increase in Population of the Gut BacteriumAkkermansia muciniphila. Front Microbiol 155 8:272. https://doi.org/10.3389/fmicb.2017.00272

156

161 157 Figure legend

158 Fig. 1. Bacterial species isolated directly from the fresh sample and after two and ten days of incubation into a

159 liquid medium previously enriched with 5% sheep blood and 5% rumen fluid.

160 Fig. 2. Number of reads, OTUs, Shannon index and phyla repartition of OTUs found after the metagenomic

161 analysis (A) and number of bacterial species and the phylum repartition after culturomics (B).

162 Fig. 3. OTUs and their best taxonomic assignment. The most abundant OTUs found from the fresh sample are in

163 the table A. The table B show the most enriched OTUs after both two and ten days. The table C show the OTUs

164 that were enriched after two days but that decreased after ten days. The table D show the OTUs the most

165 decreased after both two and ten days. The table E show the OTUs that were decreased after two days prior to

166 increased after ten days.

167 Supplementary Figures

168 Supplementary Fig. 1. Design of the study.

162 Fig. 1 Bacteroides fragilis Alistipes ihumii Escherichia coli Anaerotruncus colihominis Enterococcus faecium Clostridium aldenense Enterococcus faecalis Fresh stool Clostridium bolteae Enterococcus durans Clostridium clostridioforme Clostridium symbiosum Drancourtella massiliensis Odoribacter splanchnicus 0 Streptococcus anginosus Lactococcus garvieae

Acidaminococcus intestini 1 0 Alistipes finegoldii Alistipes onderdonkii Bacteroides vulgatus 5 Bacteroides cellulosilyticus Bifidobacterium adolescentis Bacteroides ovatus Clostridium anorexicus Bacteroides salyersiae Clostridium bartlettii Bacteroides thetaiotaomicron Clostridium innocuum 16 9 19 Bacteroides timonensis Enterobacter cloacae Bacteroides uniformis Gordonibacter pamelaeae Barnesiella intestinihominis Intestinimonas butyriciproducens Blautia coccoides Klebsiella oxytoca Clostridium hathewayi Lactobacillus parabuchneri Clostridium scindens Lactobacillus paracasei After two After ten Eggerthella lenta Lactobacillus pentosus days of days of Enterococcus avium Micrococcus luteus Enterococcus dispar Peptoniphilus grossensis incubation incubation Enterococcus gallinarum Staphylococcus epidermidis Parabacteroides distasonis Staphylococcus hominis 163 Parabacteroides merdae Fig. 2

A. Metagenomic B. Culturomics

No reads 99,428 30,773 85,627 No 6 41 33 species No OTUs 435 269 376 No 1 25 25 Shannon 4.36 4.02 3.66 oxygen intolerant

100% 0,4 0,6 100% 3 8,8 7 17 90% 1,8 90% 28,5 80% 35,7 80%

70% 70% 52 Synergistetes 51 51,1 Fusobacteria 60% 60% Verrucomicrobia 28,8 50% 50% 67 Proteobacteria 40% 37,7 40% Firmicutes 30% Bacteroidetes 30% 32 32,0 42 Actinobacteria 20% 37,9 20%

10% 17,2 10% 17 4,9 3,3 10 0% 0,4 0% 0 3 Fresh stool After two days of After ten days of Fresh sample After two days of After 10 days of incubation incubation 164 incubation incubation Fig. 3

A B C OTUs No Taxonomic assignement No of reads % OTUs No Taxonomic assignement Difference of reads OTUs No Taxonomic assignement Day two versus fresh Day ten versus fresh OTU_958 Akkermansia muciniphila 6891 6,93 OTU_170 Escherichia coli 25294 OTU_389 Eisenbergiella tayi 361 -6 OTU_365 Bacteroides uniformis 5629 5,66 OTU_419 Gammaproteobacteria 3881 OTU_376 Ruminococcus gnavus 203 -244 OTU_192 Faecalibacterium prausnitzii 5233 5,26 OTU_440 Neglecta phocaeensis 1385 OTU_45 Blautia wexlerae 191 -213 OTU_139 Lachnospiraceae 4641 4,67 OTU_870 Salmonella 782 OTU_251 Ruminococcaceae 142 -16 OTU_386 Gemmiger formicilis 4486 4,51 OTU_104 Blautia coccoides 738 OTU_377 Blautia massiliensis 122 -183 OTU_549 Faecalibacterium 3465 3,48 OTU_949 Clostridium citroniae 724 OTU_842 Lachnospiraceae 98 -67 OTU_1268 Bacteroides thetaiotaomicron 3353 3,37 OTU_153 Collinsella aerofaciens 630 OTU_983 Bacteroides 95 -13 OTU_800 Bacteroides vulgatus 2747 2,76 OTU_150 Gammaproteobacteria 599 OTU_568 Clostridia 77 -37 OTU_315 Selenomonadales 2746 2,76 OTU_300 Clostridiales 593 OTU_146 Lachnospiraceae 60 -43 OTU_550 Clostridiales 2208 2,22 OTU_895 Acidaminococcus intestini 549 OTU_1220 Clostridiales 46 -44 OTU_885 Clostridia 2186 2,20 OTU_369 Ruminococcaceae 508 OTU_603 Clostridiales 40 -26 OTU_34 Bacteroidaceae 2144 2,16 OTU_1223 Enterobacteriales 478 OTU_231 Blautia 36 -30 OTU_112 Ruminococcus bromii 2036 2,05 OTU_46 Bifidobacterium adolescentis 471 OTU_93 Clostridia 36 -6 OTU_1021 Bacteroides cellulosilyticus 1953 1,96 OTU_3 Clostridiales 402 OTU_900 Bacteroidales 30 -1 OTU_337 Bacteroidaceae 1853 1,86 OTU_92 Hungatella hathewayi 380 OTU_487 Clostridia 24 -7 OTU_335 Bacteroidia 1713 1,72 OTU_767 Bifidobacterium longum 333 OTU_467 Parabacteroides merdae 21 -91 OTU_269 Clostridiales 1625 1,63 OTU_981 Acidaminococcaceae 285 OTU_648 Oscilibacter massiliensis 19 -595 OTU_551 Clostridia 1405 1,41 OTU_960 Coriobacteriia 264 OTU_384 Bacteroidales 19 -40 OTU_828 Bacteroides finegoldii 1390 1,40 OTU_846 Bifidobacterium bifidum 256 OTU_1007 Ruminococcus 18 -13 OTU_1133 Bacteroidales 1276 1,28 OTU_873 Blautia hydrogenotrophica 178 OTU_201 Bacteroidaceae 18 -9

OTU_946 Bacteroides ovatus 1248 1,26 D E OTU_827 Bacteroidales 1177 1,18 OTUs No Taxonomic assignement Difference of reads OTUs No Taxonomic assignement Day two versus fresh Day ten versus fresh OTU_36 Bacteroidia 1166 1,17 OTU_637 Bacteroides ovatus -1152 OTU_494 Selenomonadales -1487 4992 OTU_1130 Bacteroidia 1089 1,10 OTU_125 Bacteroidales -1184 OTU_927 Tidjanibacter massiliensis -838 2277 OTU_1069 Negativicutes 1079 1,09 OTU_407 Clostridia -1229 OTU_235 Negativicutes -791 531 OTU_42 Roseburia faecis 1051 1,06 OTU_1103 Bacteroides finegoldii -1345 OTU_48 Bacteroidales -329 76 OTU_1123 Clostridiales 948 0,95 OTU_743 Bacteroidia -1353 OTU_541 Ruminococcaceae -274 36 OTU_400 Tidjanibacter massiliensis 924 0,93 OTU_942 Clostridiales -1395 OTU_618 Negativicutes -225 179 OTU_380 Bacteroides 896 0,90 OTU_800 Bacteroides vulgatus -1437 OTU_509 Clostridia -180 108 OTU_706 Verrucomicrobiales 794 0,80 OTU_1049 Bacteroidaceae -1733 OTU_87 Flavonifractor -169 118 OTU_907 Alistipes onderdonkii 789 0,79 OTU_1217 Bacteroidaceae -1806 OTU_457 Rikenellaceae -163 54 OTU_780 Parasutterella excrementihominis 758 0,76 OTU_1021 Bacteroides cellulosilyticus -1863 OTU_490 Clostridiales -159 835 OTU_1090 Clostridia 743 0,75 OTU_1112 Ruminococcus bromii -1963 OTU_350 Barnesiella intestinihominis -158 168 OTU_381 Bacteroidia 742 0,75 OTU_1162 Clostridiales -2071 OTU_55 Clostridia -127 65 OTU_1132 Bacteroides 723 0,73 OTU_359 Clostridia -2186 OTU_639 Intestinimonas butyriciproducens -119 526 OTU_580 Ruminococcus torques 711 0,72 OTU_50 Bacteroides thetaiotaomicron -3229 OTU_130 Clostridia -115 38 OTU_68 Clostridia 703 0,71 OTU_1088 Faecalibacterium -3403 OTU_280 Clostridiales -113 9 OTU_541 Oscilibacter massiliensis 661 0,66 OTU_1089 Gemmiger formicilis -3573 OTU_1136 Bacteroidia -71 501 OTU_705 Akkermansia 647 0,65 OTU_161 Lachnospiraceae -4641 OTU_917 Ruminococcaceae -66 636 OTU_281 Clostridiales 646 0,65 OTU_365 Bacteroides uniformis -4793 OTU_40 Clostridiales -65 3162 OTU_903 Clostridiales 635 0,64 OTU_976 Faecalibacterium prausnitzii -4867 OTU_1265 Clostridiales -64 8 OTU_180 Bacteroides fragilis 633 0,64 OTU_262 Akkermansia muciniphila -6489 OTU_805 Odoribacter splanchnicus -60 82

165 Culturomics Suppl. Fig. 1 Fresh stool Metagenomic (99,428 reads)

Inoculation (1g of stool)

Incubation 37°C

Anaerobic After two days After ten days blood bottle of incubation of incubation + sheep blood 5% + rumen 5%

Culturomics Culturomics Culturomics Metagenomic Metagenomic Metagenomic (29,009 reads) (30,773 reads) (85,627 reads) 166 Sup. Table 1. Bacterial species that were isolated from the fresh sample and after two and ten days of incubation, and the associated number of reads sequenced.

Fresh Day 2 Day 10 Species Culture No reads Culture No reads Culture No reads Acidaminococcus intestini NO 150 YES 273 YES 699 Alistipes finegoldii NO 8 YES 0 YES 5 Alistipes ihumii NO 0 NO 0 YES 0 Alistipes onderdonkii NO 789 YES 75 YES 90 Anaerotruncus colihominis NO 14 NO 0 YES 0 Bacteroides cellulosilyticus NO 1953 YES 201 YES 90 Bacteroides fragilis YES 633 YES 211 YES 108 Bacteroides ovatus NO 414 YES 5 YES 0 Bacteroides salyersiae NO 263 YES 116 YES 13 Bacteroides thetaiotaomicron NO 9 YES 0 YES 0 Bacteroides timonensis NO 0 YES 0 YES 0 Bacteroides uniformis NO 5629 YES 1175 YES 836 Bacteroides vulgatus NO 2747 YES 1090 NO 1310 Barnesiella intestinihominis NO 457 YES 299 YES 625 Bifidobacterium adolescentis NO 123 YES 244 NO 594 Blautia coccoides NO 0 YES 30 YES 738 Clostridium aldenense NO 10 NO 0 YES 23 Clostridium anorexicus NO 0 YES 0 NO 0 Clostridium bolteae NO 0 NO 0 YES 0 Clostridium clostridioforme NO 110 NO 0 YES 0 Clostridium hathewayi NO 0 YES 0 YES 0 Clostridium innocuum NO 0 YES 0 NO 96 Clostridium scindens NO 10 YES 186 YES 77 Clostridium symbiosum NO 10 NO 7 YES 41 Drancourtella massiliensis NO 0 NO 4 YES 64 Eggerthella lenta NO 5 YES 0 YES 493 Enterobacter cloacae NO 0 YES 0 NO 0 Enterococcus avium NO 0 YES 0 YES 0 Enterococcus dispar NO 0 YES 0 YES 0 Enterococcus durans YES 0 YES 0 YES 0 Enterococcus faecalis YES 0 YES 38 YES 10 Enterococcus faecium YES 0 YES 0 YES 0 Enterococcus gallinarum NO 0 YES 0 YES 0 Escherichia coli YES 116 YES 6505 YES 25410 Gordonibacter pamelaeae NO 0 YES 0 NO 61 Intestinibacter bartlettii NO 124 YES 0 NO 121 Intestinimonas butyriciproducens NO 134 YES 15 NO 660 Klebsiella oxytoca NO 0 YES 0 NO 28 Lactobacillus parabuchneri NO 0 YES 0 NO 0 Lactobacillus paracasei NO 0 YES 0 NO 0 Lactobacillus pentosus NO 0 YES 0 NO 0 Lactococcus garvieae YES 0 YES 0 NO 0 Micrococcus luteus NO 0 YES 0 NO 0 Odoribacter splanchnicus NO 82 NO 22 YES 164 Parabacteroides distasonis NO 165 YES 221 YES 295 Parabacteroides merdae NO 99 YES 120 YES 8

167 Peptoniphilus grossensis NO 8 YES 0 NO 0 Staphylococcus epidermidis NO 0 YES 0 NO 0 Staphylococcus hominis NO 0 YES 0 NO 0 Streptococcus anginosus NO 0 NO 0 YES 0 No: number of

168 Sup. Table 2. OTUs and their best taxonomic assignment found from the metagenomic analysis of the fresh sample. OTUs No Taxonomic assignement No reads Culture OTU_958 Akkermansia muciniphila 6891 OTU_365 Bacteroides uniformis 5629 OTU_192 Faecalibacterium prausnitzii 5233 OTU_139 Lachnospiraceae 4641 OTU_386 Gemmiger formicilis 4486 OTU_549 Faecalibacterium 3465 OTU_1268 Bacteroides thetaiotaomicron 3353 OTU_800 Bacteroides vulgatus 2747 OTU_315 Selenomonadales 2746 OTU_550 Clostridiales 2208 OTU_885 Clostridia 2186 OTU_34 Bacteroidaceae 2144 OTU_112 Ruminococcus bromii 2036 OTU_1021 Bacteroides cellulosilyticus 1953 OTU_337 Bacteroidaceae 1853 OTU_335 Bacteroidia 1713 OTU_269 Clostridiales 1625 OTU_551 Clostridia 1405 OTU_828 Bacteroides finegoldii 1390 OTU_1133 Bacteroidales 1276 OTU_946 Bacteroides ovatus 1248 OTU_827 Bacteroidales 1177 OTU_36 Bacteroidia 1166 OTU_1130 Bacteroidia 1089 OTU_1069 Negativicutes 1079 OTU_42 Roseburia faecis 1051 OTU_1123 Clostridiales 948 OTU_400 Tidjanibacter massiliensis 924 OTU_380 Bacteroides 896 OTU_706 Verrucomicrobiales 794 OTU_907 Alistipes onderdonkii 789 OTU_780 Parasutterella excrementihominis 758 OTU_1090 Clostridia 743 OTU_381 Bacteroidia 742 OTU_1132 Bacteroides 723 OTU_580 Ruminococcus torques 711 OTU_68 Clostridia 703 OTU_541 Oscilibacter massiliensis 661 OTU_705 Akkermansia 647 OTU_281 Clostridiales 646 OTU_903 Clostridiales 635 OTU_180 Bacteroides fragilis 633 YES OTU_638 Clostridiales 569 OTU_35 Bacteroidales 498 OTU_926 Clostridia 485 OTU_350 Barnesiella intestinihominis 457 OTU_268 Subdoligranulum 449 OTU_707 Bacteroides ovatus 414

169 OTU_703 Verrucomicrobiae 402 OTU_558 Bacteroidales 365 OTU_901 Clostridia 364 OTU_1067 Bacteroides 330 OTU_1057 Negativicutes 320 OTU_1 Lachnospiraceae 303 OTU_627 Ruminococcus 287 OTU_640 Flavonifractor plautii 284 OTU_174 Ruminococcaceae 279 OTU_902 Roseburia 269 OTU_589 Lachnoclostridium bouchesdurhonense 267 OTU_366 Bacteroides salyersiae 263 OTU_379 Bacteroidales 261 OTU_941 Ruminococcaceae 260 OTU_464 Clostridia 245 OTU_494 Ruminococcus gnavus 244 OTU_847 Bacteroides 243 OTU_743 Blautia massiliensis 238 OTU_1162 Blautia wexlerae 231 OTU_605 Bacteroidia 225 OTU_138 Clostridiales 219 OTU_723 Bacteroidaceae 216 OTU_463 Flavonifractor 210 OTU_462 Clostridiales 195 OTU_234 Bacteroidales 180 OTU_1121 Clostridia 179 OTU_270 Clostridia 178 OTU_27 Clostridiales 173 OTU_177 Clostridia 173 OTU_479 Erysipelotrichales 171 OTU_559 Rikenellaceae 168 OTU_1111 Alistipes 168 OTU_712 Bacteroidaceae 166 OTU_721 Clostridiales 166 OTU_1261 Parabacteroides distasonis 165 OTU_370 Bacteroidales 162 OTU_610 Intestinimonas timonensis 153 OTU_1030 Bacteroides 151 OTU_895 Acidaminococcus intestini 150 OTU_955 Ruminococcaceae 149 OTU_143 Clostridia 143 OTU_458 Bilophila wadsworthia 142 OTU_324 Clostridia 140 OTU_865 Clostridia 139 OTU_1124 Bacteroidia 138 OTU_66 Lachnospiraceae 135 OTU_639 Intestinimonas butyriciproducens 134 OTU_264 Clostridia 131 OTU_1116 Intestinibacter bartlettii 124 OTU_46 Bifidobacterium adolescentis 123

170 OTU_357 Ruminococcus callidus 120 OTU_307 Alistipes shahii 119 OTU_535 Clostridia 117 OTU_170 Escherichia coli 116 YES OTU_530 Lachnospiraceae 115 OTU_686 Lachnospiraceae 115 OTU_688 Ruminococcaceae 113 OTU_748 [Clostridium] clostridioforme 110 OTU_202 Clostridiales 104 OTU_467 Parabacteroides merdae 99 OTU_332 Parasutterella 98 OTU_652 Proteobacteria 97 OTU_466 Barnesiellaceae 95 OTU_702 Clostridia 94 OTU_606 Bacteroidales 93 OTU_497 Parasutterella excrementihominis 88 OTU_687 Clostridiales 87 OTU_237 Bacteroidia 83 OTU_805 Odoribacter splanchnicus 82 OTU_1049 Ruminococcaceae 82 OTU_501 Murimonas 81 OTU_787 Ruminococcaceae 80 OTU_772 Clostridiales 80 OTU_970 Parasutterella excrementihominis 79 OTU_398 Clostridiales 77 OTU_1112 Clostridiales 77 OTU_637 Lachnospiraceae 77 OTU_685 Clostridiales 75 OTU_722 Bacteroidales 75 OTU_1243 Anaerostipes hadrus 74 OTU_788 Clostridia 74 OTU_1192 Bacteroidia 71 OTU_1097 Anaerostipes caccae 69 OTU_852 Bacteroidales 66 OTU_260 Clostridia 65 OTU_961 Erysipelotrichia 63 OTU_851 Bacteroidia 63 OTU_401 Acidaminococcaceae 63 OTU_438 Clostridiales 63 OTU_1074 Negativicutes 62 OTU_647 Clostridiales 61 OTU_262 Gammaproteobacteria 60 OTU_1191 Bifidobacterium longum subsp. longum 60 OTU_28 Clostridia 58 OTU_243 Bacteroidales 55 OTU_980 Bacteroidales 54 OTU_801 Proteobacteria 53 OTU_150 Bacteroidales 52 OTU_725 Bacteroidia 52 OTU_47 Ruthenibacterium lactatiformans 51

171 OTU_51 Clostridia 51 OTU_928 Clostridiales 48 OTU_63 Bacteroidales 47 OTU_924 Clostridiales 45 OTU_690 Clostridia 45 OTU_151 Bacteroidaceae 45 OTU_875 Clostridiales 44 OTU_1110 Bacteroidia 44 OTU_927 Lachnospiraceae 43 OTU_142 Bacteroides clarus 43 OTU_313 Negativicutes 43 OTU_718 Clostridia 42 OTU_678 Neglecta phocaeensis 41 OTU_85 Ruminococcaceae 41 OTU_439 Blautia massiliensis 41 OTU_1015 Clostridiales 41 OTU_990 Clostridiales 40 OTU_369 Fusicatenibacter saccharivorans 39 OTU_1092 Lachnospiraceae 39 OTU_770 Streptococcus oralis subsp. oralis 38 OTU_731 Lachnospiraceae 38 OTU_1087 Lactobacillus rogosae 37 OTU_929 Clostridia 37 OTU_592 Clostridiales 37 OTU_325 Clostridiales 37 OTU_649 Clostridiales 36 OTU_257 Clostridiales 36 OTU_930 Eubacterium hallii 36 OTU_648 Blautia 36 OTU_258 Lachnospiraceae 35 OTU_710 Bacteroidia 35 OTU_1240 Colidextribacter massiliensis 35 OTU_1262 Clostridiales 35 OTU_90 Ruminococcaceae 35 OTU_1125 Bacteroidales 35 OTU_56 Clostridiales 35 OTU_625 Negativibacillus massiliensis 34 OTU_108 Blautia faecis 33 OTU_803 Coriobacteriales 32 OTU_465 Bacteroidales 32 OTU_334 Gammaproteobacteria 32 OTU_210 Proteobacteria 32 OTU_1053 Clostridiales 31 OTU_443 Selenomonadales 31 OTU_203 Bacteroidia 31 OTU_442 Intestinimonas massiliensis 30 OTU_791 Clostridiales 30 OTU_266 Tannerellaceae 30 OTU_591 Ruminococcaceae 29 OTU_1048 Intestinibacillus massiliensis 29

172 OTU_786 Clostridiales 29 OTU_515 Clostridiales 29 OTU_2 Clostridiales 28 OTU_484 Gammaproteobacteria 28 OTU_10 Clostridiales 28 OTU_297 Clostridiales 27 OTU_956 Clostridiales 26 OTU_190 Lachnospiraceae 25 OTU_1075 Clostridia 25 OTU_655 Clostridiales 24 OTU_512 Clostridiales 24 OTU_1072 Clostridiales 24 OTU_570 Pseudoflavonifractor capillosus 24 OTU_1155 Streptococcus salivarius 24 OTU_230 Clostridia 24 OTU_403 Lachnospiraceae 23 OTU_873 Clostridiales 22 OTU_949 Bifidobacteriales 22 OTU_1258 Proteobacteria 22 OTU_22 Bacteroidales 22 OTU_566 Bacteroides 22 OTU_244 Alistipes 22 OTU_1073 Clostridiales 21 OTU_405 Clostridia 20 OTU_232 Bacteroidaceae 19 OTU_692 Erysipelotrichia 19 OTU_50 Clostridium glycyrrhizinilyticum 19 OTU_960 Staphylococcus warneri 18 OTU_490 Dorea formicigenerans 18 OTU_665 Clostridia 18 OTU_1081 Romboutsia timonensis 17 OTU_1122 Ruminococcaceae 17 OTU_1218 Bifidobacteriales 17 OTU_948 Actinobacteria 17 OTU_1086 Selenomonadales 17 OTU_113 Firmicutes 16 OTU_537 [Clostridium] lactatifermentans 16 OTU_1045 Klebsiella pneumoniae 16 OTU_760 Betaproteobacteriales 16 OTU_1016 Lachnospiraceae 16 OTU_106 Actinomyces odontolyticus 15 OTU_656 Ruminococcaceae 15 OTU_790 Lachnoclostridium 15 OTU_602 Clostridia 15 OTU_1070 Clostridia 15 OTU_413 Granulicatella adiacens 14 OTU_459 Bacteroidales 14 OTU_478 Deltaproteobacteria 14 OTU_634 Clostridiales 14 OTU_1152 Anaerotruncus colihominis 14

173 OTU_1263 Clostridiales 14 OTU_1039 Clostridiales 14 OTU_547 Clostridia 14 OTU_393 Proteobacteria 13 OTU_771 Lachnospiraceae 13 OTU_909 Clostridia 13 OTU_925 Veillonella parvula 13 OTU_39 Bacteroides 13 OTU_509 Ruminococcus 13 OTU_343 Ihubacter massiliensis 13 OTU_976 Collinsella aerofaciens 13 OTU_1033 Clostridiales 13 OTU_730 Bifidobacteriaceae 13 OTU_233 Bacteroidales 12 OTU_432 Prevotella timonensis 12 OTU_511 Pseudomonas taiwanensis 12 OTU_957 Clostridia 12 OTU_992 Proteobacteria 12 OTU_1231 Firmicutes oral clone BX005 12 OTU_635 Clostridia 12 OTU_1089 Enterobacteriales 12 OTU_271 Peptostreptococcaceae 12 OTU_414 Firmicutes 12 OTU_37 Brochothrix thermosphacta 11 OTU_238 Clostridiales 11 OTU_404 Clostridiales 11 OTU_436 Clostridia 11 OTU_779 Carnobacterium maltaromaticum 11 OTU_843 Tannerellaceae 11 OTU_918 Clostridia 11 OTU_1149 Negativicutes 11 OTU_600 Hungatella 11 OTU_480 Clostridiales 11 OTU_1100 Lachnoclostridium 11 OTU_812 Eubacterium siraeum 11 OTU_79 Clostridia 10 OTU_317 Clostridia 10 OTU_412 Clostridia 10 OTU_505 Clostridiales 10 OTU_769 Clostridia 10 OTU_822 Corynebacterium tuberculostearicum 10 OTU_998 Clostridiales 10 OTU_267 Bacteroidales 10 OTU_1190 Bacteroides 10 OTU_646 Clostridia 10 OTU_359 Bifidobacterium bifidum 10 OTU_1107 Holdemania filiformis 10 OTU_1164 Lachnospiraceae 10 OTU_597 [Clostridium] scindens 10 OTU_164 [Clostridium] symbiosum 10

174 OTU_993 [Clostridium] aldenense 10 OTU_418 Marinifilaceae 10 OTU_1029 Bacteroidales 10 OTU_621 Blautia obeum 10 OTU_121 Firmicutes 9 OTU_563 Clostridiales 9 OTU_608 Bacteroides thetaiotaomicron 9 OTU_870 Ruminococcaceae 9 OTU_931 Bifidobacteriales 9 OTU_1208 Clostridiales 9 OTU_697 Bacteroidaceae 9 OTU_329 Lachnospiraceae 9 OTU_1044 Lachnospiraceae 9 OTU_1172 Clostridia 9 OTU_361 Bacilli 9 OTU_176 Alistipes senegalensis 9 OTU_308 Clostridia 8 OTU_451 Bacteroidales 8 OTU_555 Peptoniphilus grossensis 8 OTU_684 Clostridia 8 OTU_836 Clostridiales 8 OTU_864 Firmicutes 8 OTU_954 Clostridia 8 OTU_967 Clostridia 8 OTU_1004 Clostridiales 8 OTU_153 Bacteroidia 8 OTU_7 Bacteroidaceae 8 OTU_729 Clostridia 8 OTU_1007 Clostridia 8 OTU_1217 Actinobacteria 8 OTU_595 Clostridiales 8 OTU_485 Alistipes finegoldii 8 OTU_3 Clostridia 7 OTU_92 Clostridiales 7 OTU_503 Clostridiales 7 OTU_526 Roseburia intestinalis 7 OTU_588 Firmicutes 7 OTU_741 Psychromonas 7 OTU_1052 Ensifer fredii 7 OTU_1215 Clostridia 7 OTU_536 Phocea massiliensis 7 OTU_981 Bacteroidaceae 7 OTU_247 Bacteroidetes 7 OTU_1183 Clostridia 7 OTU_1170 Sellimonas 7 OTU_389 Ruminococcaceae 7 OTU_93 Clostridia 7 OTU_19 Geobacillus 6 OTU_178 Cutibacterium acnes 6 OTU_241 Ruminococcaceae 6

175 OTU_328 Clostridiales 6 OTU_341 Dialister invisus 6 OTU_474 Corynebacterium kroppenstedtii 6 OTU_487 Lachnospiraceae 6 OTU_594 Clostridia 6 OTU_761 Clostridiales 6 OTU_814 Coriobacteriia 6 OTU_889 Curtobacterium luteum 6 OTU_920 Clostridia 6 OTU_968 Bacilli 6 OTU_984 Actinomyces odontolyticus 6 OTU_1051 Lachnospiraceae 6 OTU_1056 Clostridiales 6 OTU_1241 Bacteroides faecis 6 OTU_1255 Clostridia 6 OTU_161 Eisenbergiella tayi 6 OTU_809 Clostridia 6 OTU_679 Clostridiales 6 OTU_205 Hungatella hathewayi 6 OTU_983 Lachnospiraceae 6 OTU_181 Lachnospiraceae 6 OTU_137 Clostridia 6 OTU_348 Clostridia 6 OTU_157 Clostridiales 6 OTU_58 Bilophila wadsworthia 6 OTU_54 Leuconostoc carnosum 5 OTU_160 Clostridiales 5 OTU_172 Ruminococcaceae 5 OTU_201 Blautia 5 OTU_274 Bifidobacterium thermophilum 5 OTU_330 Bacteroidetes 5 OTU_430 Roseburia inulinivorans 5 OTU_444 Clostridia 5 OTU_659 Lachnospiraceae 5 OTU_681 Clostridia 5 OTU_717 Turicibacter sanguinis 5 OTU_855 Clostridiales 5 OTU_893 Clostridiales 5 OTU_906 Cloacibacterium normanense 5 OTU_908 Gemella haemolysans 5 OTU_914 Roseburia 5 OTU_1095 Lachnospiraceae 5 OTU_1136 Clostridiales 5 OTU_562 Ruminiclostridium 5 OTU_578 Bacteroides ovatus 5 OTU_1088 Salmonella 5 OTU_355 Eggerthella lenta 5 OTU_376 Actinobacteria 5 OTU_989 Deltaproteobacteria 5 OTU_1000 Clostridia 5

176 OTU_242 Clostridiales 5 OTU_55 Ruminococcaceae 5 OTU_74 Clostridia 4 OTU_156 Ruminococcaceae 4 OTU_173 Proteobacteria 4 OTU_204 Ruminococcaceae 4 OTU_214 Firmicutes 4 OTU_224 Clostridia 4 OTU_316 Clostridiales 4 OTU_340 Clostridiales 4 OTU_392 Lachnoanaerobaculum umeaense 4 OTU_397 Clostridia 4 OTU_502 Prevotella bivia 4 OTU_552 Bacteroidales 4 OTU_586 Clostridiales 4 OTU_590 Clostridiales 4 OTU_599 Parasutterella excrementihominis 4 OTU_711 Clostridiales 4 OTU_713 Bacilli 4 OTU_724 Clostridia 4 OTU_751 Anaerococcus prevotii 4 OTU_799 Gammaproteobacteria 4 OTU_841 Pseudomonas psychrophila 4 OTU_846 Bacteroidales 4 OTU_900 Bacilli 4 OTU_975 Parasutterella excrementihominis 4 OTU_1041 Desulfovibrionales 4 OTU_1050 Clostridiales 4 OTU_1239 Gammaproteobacteria 4 OTU_1260 Neisseria perflava 4 OTU_300 Clostridiales 4 OTU_842 Bacteroidia 4 OTU_235 Tannerellaceae 4 OTU_538 Bittarella massiliensis 4 OTU_557 Desulfovibrionaceae 4 OTU_211 Odoribacter splanchnicus 4 OTU_1223 Clostridia 4 No: Number of

177 Sup. Table 3. OTUs that were increased or decreased after both two and ten days of incubation compared with fresh sample. OTUs No Taxonomic assignement No reads day 2 No reads day 10 Culture Difference of reads OTU_170 Escherichia coli 6505 25410 YES 25294 OTU_419 Gammaproteobacteria 1747 3941 3881 OTU_440 Neglecta phocaeensis 61 1426 1385 OTU_870 Salmonella 300 787 782 OTU_104 Blautia coccoides 30 738 YES 738 OTU_949 Clostridium citroniae 30 724 724 OTU_153 Collinsella aerofaciens 388 643 630 OTU_150 Gammaproteobacteria 99 599 599 OTU_300 Clostridiales 210 670 593 OTU_895 Acidaminococcus intestini 273 699 YES 549 OTU_369 Ruminococcaceae 9 508 508 OTU_1223 Enterobacteriales 328 490 478 OTU_46 Bifidobacterium adolescentis 244 594 YES 471 OTU_3 Clostridiales 12 408 402 OTU_92 Hungatella hathewayi 10 386 380 OTU_767 Bifidobacterium longum subsp. longum 146 393 333 OTU_981 Acidaminococcaceae 109 348 285 OTU_960 Coriobacteriia 116 264 264 OTU_846 Bifidobacterium bifidum 183 266 256 OTU_873 Blautia hydrogenotrophica 36 178 178 OTU_142 Coriobacteriia 4 171 171 OTU_518 Negativicutes 96 224 162 OTU_1261 Parabacteroides distasonis 221 295 YES 130 OTU_257 Clostridiales 7 123 123 OTU_334 Coriobacteriales 83 114 114 OTU_931 Actinobacteria 22 117 112 OTU_1073 Coriobacteriaceae 58 110 110 OTU_613 Selenomonadales 55 135 104 OTU_1258 Erysipelatoclostridium ramosum 27 94 94 OTU_975 Holdemania filiformis 34 93 83 OTU_597 Clostridium scindens 186 77 YES 67 OTU_671 Bacteroidales 75 98 66 OTU_493 Drancourtella massiliensis 4 64 YES 64 OTU_319 Bifidobacteriales 21 77 60 OTU_397 Negativicutes 18 70 59 OTU_10 Massilimaliae massiliensis 12 54 54 OTU_908 Actinobacteria 136 54 46 OTU_210 Bifidobacteriales 7 44 44 OTU_279 Actinobacteria 50 59 42 OTU_318 Selenomonadales 36 56 39 OTU_62 Clostridium glycyrrhizinilyticum 301 57 38 OTU_244 Actinobacteria 47 32 32 OTU_511 Tannerellaceae 81 30 26 OTU_1048 Bifidobacterium pseudocatenulatum 11 25 25 OTU_232 Ruminococcaceae 107 21 21 OTU_502 Lachnoclostridium 13 31 20 OTU_1231 Bittarella massiliensis 6 24 20

178 OTU_258 Adlercreutzia equolifaciens 5 20 20 OTU_875 Clostridia 4 20 20 OTU_655 Clostridiales 18 17 17 OTU_484 Blautia 18 17 17 OTU_592 Clostridiales 4 16 16 OTU_803 Firmicutes 7 15 15 OTU_710 Erysipelotrichales 5 13 13 OTU_503 Clostridia 9 18 12 OTU_90 Bifidobacteriaceae 6 12 12 OTU_1240 Erysipelotrichia 5 11 11 OTU_619 Enterococcus faecalis 38 10 YES 10 OTU_167 Bacteroidaceae 55 55 10 OTU_108 Actinobacteria 24 10 10 OTU_1015 Erysipelotrichia 4 9 9 OTU_2 Clostridiales 12 7 7 OTU_1087 Erysipelotrichaceae 4 7 7 OTU_770 Blautia 4 7 7 OTU_1155 Lachnospiraceae 21 6 6 OTU_625 Desulfovibrionales 9 6 6 OTU_649 Firmicutes 7 6 6 OTU_403 Lachnospiraceae 26 5 5 OTU_990 Bacteroidia 4 5 5 OTU_239 Bacteroidales 0 46 -1 OTU_127 Clostridiales 0 36 -1 OTU_430 Clostridia 0 8 -1 OTU_968 Ruminococcaceae 0 6 -1 OTU_984 Clostridia 0 5 -2 OTU_523 Alistipes shahii 15 116 -3 OTU_444 Bacilli 0 6 -3 OTU_485 Alistipes finegoldii 0 5 YES -3 OTU_1116 Intestinibacter bartlettii 0 121 YES -3 OTU_763 Bilophila wadsworthia 45 138 -4 OTU_711 Eubacterium siraeum 0 7 -4 OTU_1072 Clostridia 0 0 -4 OTU_393 Ruminococcaceae 0 0 -4 OTU_771 Proteobacteria 0 0 -4 OTU_909 Ruminococcaceae 0 0 -4 OTU_925 Firmicutes 0 0 -4 OTU_1039 Clostridia 0 0 -4 OTU_547 Clostridiales 0 0 -4 OTU_413 Clostridiales 0 0 -4 OTU_459 Lachnoanaerobaculum umeaense 0 0 -4 OTU_478 Clostridia 0 0 -4 OTU_634 Prevotella bivia 0 0 -4 OTU_1263 Bacteroidales 0 0 -4 OTU_203 Clostridiales 0 0 -4 OTU_801 Clostridiales 0 0 -4 OTU_602 Parasutterella excrementihominis 0 0 -4 OTU_1070 Clostridiales 0 0 -4 OTU_106 Bacilli 0 0 -4

179 OTU_656 Clostridia 0 0 -4 OTU_1016 Anaerococcus prevotii 0 0 -4 OTU_113 Gammaproteobacteria 0 0 -4 OTU_537 Pseudomonas psychrophila 0 0 -4 OTU_1045 Bacteroidales 0 0 -4 OTU_566 Bacilli 0 0 -4 OTU_647 Parasutterella excrementihominis 0 0 -4 OTU_1081 Desulfovibrionales 0 0 -4 OTU_1122 Clostridiales 0 0 -4 OTU_786 Gammaproteobacteria 0 0 -4 OTU_570 Neisseria perflava 0 0 -4 OTU_1174 Clostridiales 0 16 -5 OTU_659 Alistipes senegalensis 0 4 -5 OTU_79 Bacteroides ovatus 4 0 -5 OTU_769 Leuconostoc carnosum JB16 0 0 -5 OTU_822 Clostridiales 0 0 -5 OTU_998 Ruminococcaceae 0 0 -5 OTU_600 Blautia 0 0 -5 OTU_480 Bifidobacterium thermophilum 0 0 -5 OTU_812 Bacteroidetes 0 0 -5 OTU_37 Roseburia inulinivorans 0 0 -5 OTU_238 Clostridia 0 0 -5 OTU_404 Lachnospiraceae 0 0 -5 OTU_436 Clostridia 0 0 -5 OTU_779 Turicibacter sanguinis 0 0 -5 OTU_702 Clostridiales 0 0 -5 OTU_515 Clostridiales 0 0 -5 OTU_271 Cloacibacterium normanense 0 0 -5 OTU_414 Gemella haemolysans 0 0 -5 OTU_233 Roseburia 0 0 -5 OTU_432 Lachnospiraceae 0 0 -5 OTU_173 Firmicutes 0 6 -6 OTU_692 Geobacillus 0 0 -6 OTU_595 Cutibacterium acnes 0 0 -6 OTU_308 Ruminococcaceae 0 0 -6 OTU_451 Clostridiales 0 0 -6 OTU_684 Dialister invisus 0 0 -6 OTU_836 Corynebacterium kroppenstedtii 0 0 -6 OTU_864 Lachnospiraceae 0 0 -6 OTU_954 Clostridia 0 0 -6 OTU_967 Clostridiales 0 0 -6 OTU_1004 Coriobacteriia 0 0 -6 OTU_439 Curtobacterium luteum 0 0 -6 OTU_297 Clostridia 0 0 -6 OTU_1044 Bacilli 0 0 -6 OTU_1172 Actinomyces odontolyticus 0 0 -6 OTU_361 Lachnospiraceae 0 0 -6 OTU_176 Clostridiales 0 0 -6 OTU_121 Bacteroides faecis 0 0 -6 OTU_563 Clostridia 0 0 -6

180 OTU_1051 Clostridia 0 0 -7 OTU_1056 Clostridiales 0 0 -7 OTU_1241 Clostridiales 0 0 -7 OTU_1255 Roseburia intestinalis 0 0 -7 OTU_790 Firmicutes 0 0 -7 OTU_512 Psychromonas 0 0 -7 OTU_1183 Ensifer fredii 0 0 -7 OTU_914 Bacteroidia 7 0 -8 OTU_329 Bacteroidaceae 5 0 -8 OTU_918 Clostridia 4 0 -8 OTU_137 Clostridia 0 0 -8 OTU_58 Bacteroidales 0 0 -8 OTU_19 Clostridia 0 0 -8 OTU_178 Clostridiales 0 0 -8 OTU_241 Firmicutes 0 0 -8 OTU_328 Clostridia 0 0 -8 OTU_341 Clostridia 0 0 -8 OTU_474 Clostridiales 0 0 -8 OTU_555 Peptoniphilus grossensis 0 0 YES -8 OTU_136 Clostridia 0 9 -9 OTU_274 Lachnospiraceae 4 0 -9 OTU_681 Firmicutes 0 0 -9 OTU_717 Clostridiales 0 0 -9 OTU_855 Ruminococcaceae 0 0 -9 OTU_893 Bifidobacteriales 0 0 -9 OTU_906 Clostridiales 0 0 -9 OTU_608 Bacteroides thetaiotaomicron 0 0 YES -9 OTU_1260 Clostridia 6 0 -10 OTU_690 Clostridia 0 0 -10 OTU_989 Clostridia 0 0 -10 OTU_1000 Clostridia 0 0 -10 OTU_242 Clostridiales 0 0 -10 OTU_54 Clostridia 0 0 -10 OTU_160 Corynebacterium tuberculostearicum 0 0 -10 OTU_172 Clostridiales 0 0 -10 OTU_581 Clostridia 0 13 -11 OTU_586 Clostridia 6 0 -11 OTU_713 Brochothrix thermosphacta 0 0 -11 OTU_724 Clostridiales 0 0 -11 OTU_751 Clostridiales 0 0 -11 OTU_799 Clostridia 0 0 -11 OTU_841 Carnobacterium maltaromaticum 0 0 -11 OTU_584 Lachnospiraceae 0 4 -12 OTU_74 Clostridia 5 0 -12 OTU_204 Bacteroidales 0 0 -12 OTU_214 Prevotella timonensis 0 0 -12 OTU_224 Pseudomonas taiwanensis 0 0 -12 OTU_316 Clostridia 0 0 -12 OTU_340 Proteobacteria 0 0 -12 OTU_392 Firmicutes oral clone 0 0 -12

181 OTU_1170 Ihubacter massiliensis 10 0 -13 OTU_22 Proteobacteria 0 0 -13 OTU_1097 Lachnospiraceae 0 0 -13 OTU_557 Clostridia 0 0 -13 OTU_211 Veillonella parvula 0 0 -13 OTU_82 Granulicatella adiacens 0 0 -14 OTU_299 Bacteroidales 0 0 -14 OTU_247 Deltaproteobacteria 0 0 -14 OTU_843 Clostridiales 0 0 -14 OTU_578 Clostridiales 0 0 -14 OTU_1152 Anaerotruncus colihominis 0 0 YES -14 OTU_179 Ruminococcaceae 9 98 -15 OTU_894 Lachnoclostridium 9 0 -15 OTU_1037 Actinomyces odontolyticus 0 0 -15 OTU_1137 Ruminococcaceae 0 0 -15 OTU_680 Firmicutes 0 0 -16 OTU_726 Clostridium lactatifermentans 0 0 -16 OTU_742 Klebsiella pneumoniae 0 0 -16 OTU_342 Romboutsia timonensis 0 0 -17 OTU_391 Ruminococcaceae 0 0 -17 OTU_24 Blautia massiliensis 33 23 -18 OTU_312 Clostridiales 4 23 -18 OTU_163 Staphylococcus warneri 0 0 -18 OTU_660 Clostridiales 4 17 -19 OTU_80 Erysipelotrichia 12 0 -19 OTU_89 Bacteroidaceae 0 0 -19 OTU_548 Clostridiales 0 6 -20 OTU_1216 Clostridia 8 0 -20 OTU_426 Clostridiales 0 13 -22 OTU_209 Gammaproteobacteria 12 10 -22 OTU_823 Bacteroidales 19 0 -22 OTU_987 Bacteroides 6 0 -22 OTU_1068 Clostridiales 0 0 -22 OTU_1108 Bifidobacteriales 0 0 -22 OTU_1151 Proteobacteria 0 0 -22 OTU_819 Lachnospiraceae 0 0 -23 OTU_428 Blautia faecis 10 9 -24 OTU_694 Bacteroidia 17 7 -24 OTU_1059 Clostridiales 18 0 -24 OTU_632 Clostridiales 0 0 -24 OTU_157 Lachnospiraceae 0 0 -25 OTU_363 Bacteroidia 5 199 -26 OTU_374 Clostridia 0 25 -26 OTU_295 Proteobacteria 0 5 -27 OTU_506 Clostridiales 19 0 -27 OTU_132 Gammaproteobacteria 4 0 -28 OTU_388 Clostridiales 0 0 -28 OTU_1205 Ruminococcaceae 0 0 -29 OTU_32 Intestinibacillus massiliensis 0 0 -29 OTU_1006 Bacteroidia 0 14 -30

182 OTU_117 Intestinimonas massiliensis 0 0 -30 OTU_123 Clostridiales 0 0 -30 OTU_383 Lachnospiraceae 5 8 -31 OTU_866 Clostridiales 6 0 -31 OTU_1203 Negativicutes 0 11 -32 OTU_368 Coriobacteriales 0 0 -32 OTU_1042 Bacteroidales 33 18 -34 OTU_1250 Lachnospiraceae 0 0 -35 OTU_1252 Bacteroidia 0 0 -35 OTU_21 Colidextribacter massiliensis 0 0 -35 OTU_336 Clostridiales 0 0 -35 OTU_1218 Barnesiellaceae 8 59 -36 OTU_816 Clostridiales 4 0 -36 OTU_853 Clostridiales 30 43 -37 OTU_470 Lactobacillus rogosae 0 0 -37 OTU_676 Lachnospiraceae 6 0 -38 OTU_185 Streptococcus oralis subsp. oralis 0 0 -38 OTU_1091 Clostridiales 0 0 -40 OTU_1022 Clostridia 8 0 -42 OTU_849 Bacteroides clarus 7 0 -43 OTU_267 Bacteroidales 70 136 -44 OTU_189 Clostridiales 0 0 -45 OTU_657 Bacteroidia 0 4 -48 OTU_162 Proteobacteria 39 4 -49 OTU_260 Clostridiales 34 10 -53 OTU_155 Bacteroidales 0 0 -54 OTU_514 Clostridiales 45 0 -61 OTU_1190 Bacteroidales 0 12 -63 OTU_348 Clostridia 20 11 -63 OTU_449 Erysipelotrichia 28 0 -63 OTU_496 Bacteroidia 18 0 -63 OTU_538 Bacteroidales 4 0 -66 OTU_1207 Parasutterella excrementihominis 13 5 -74 OTU_760 Clostridiales 0 0 -75 OTU_569 Murimonas 22 5 -76 OTU_536 Clostridiales 0 0 -77 OTU_679 Bacteroidia 42 53 -85 OTU_641 Bacteroidales 7 7 -86 OTU_205 Parasutterella 48 11 -87 OTU_469 Parasutterella excrementihominis 0 0 -88 OTU_1063 Proteobacteria 46 8 -89 OTU_953 Clostridia 0 80 -98 OTU_243 Flavonifractor plautii 38 185 -99 OTU_1010 Intestinimonas timonensis 15 52 -101 OTU_250 Ruminococcus callidus 0 14 -106 OTU_272 Bacteroides 21 41 -110 OTU_748 Clostridium clostridioforme 0 0 YES -110 OTU_183 Lachnospiraceae 0 4 -111 OTU_128 Lachnospiraceae 75 22 -113 OTU_839 Clostridia 0 4 -113

183 OTU_1214 Clostridia 95 25 -115 OTU_818 Lachnospiraceae 0 0 -115 OTU_644 Clostridia 0 44 -129 OTU_621 Alistipes 5 34 -134 OTU_353 Clostridia 0 0 -143 OTU_371 Ruminococcaceae 0 0 -149 OTU_1149 Bacteroidales 14 11 -151 OTU_1169 Bacteroidaceae 41 7 -159 OTU_165 Clostridiales 65 6 -160 OTU_556 Erysipelotrichales 126 0 -171 OTU_151 Clostridia 4 6 -173 OTU_1038 Bacteroidales 40 64 -197 OTU_296 Bacteroidales 142 300 -198 OTU_373 Bacteroidaceae 19 7 -209 OTU_825 Clostridiales 4 0 -219 OTU_697 Ruminococcaceae 160 32 -228 OTU_930 Bacteroides 18 0 -243 OTU_366 Bacteroides salyersiae 116 13 YES -250 OTU_669 Lachnoclostridium bouchesdurhonense 0 0 -267 OTU_668 Roseburia 0 0 -269 OTU_1086 Ruminococcus 0 13 -274 OTU_678 Lachnospiraceae 0 0 -303 OTU_326 Bacteroides 0 0 -330 OTU_1107 Verrucomicrobiae 6 58 -344 OTU_1215 Clostridia 12 4 -360 OTU_443 Subdoligranulum 0 85 -364 OTU_86 Clostridia 0 99 -386 OTU_1191 Bacteroidia 439 759 -407 OTU_707 Bacteroides ovatus 5 0 YES -414 OTU_180 Bacteroides fragilis 211 108 YES -525 OTU_1029 Clostridiales 11 4 -565 OTU_401 Parasutterella excrementihominis 403 166 -592 OTU_1074 Akkermansia 7 55 -592 OTU_809 Clostridia 81 89 -614 OTU_465 Bacteroidia 200 116 -626 OTU_187 Clostridiales 4 0 -635 OTU_948 Clostridiales 0 4 -642 OTU_907 Alistipes onderdonkii 75 90 YES -699 OTU_325 Bacteroides 288 14 -709 OTU_29 Ruminococcus torques 37 0 -711 OTU_928 Clostridia 6 23 -720 OTU_1209 Verrucomicrobiales 0 27 -767 OTU_409 Bacteroides 51 19 -877 OTU_39 Bacteroidales 249 263 -914 OTU_888 Clostridiales 11 18 -930 OTU_929 Roseburia faecis 0 0 -1051 OTU_410 Bacteroidia 434 27 -1062 OTU_637 Bacteroides ovatus 481 96 -1152 OTU_125 Bacteroidales 393 92 -1184 OTU_407 Clostridia 5 176 -1229

184 OTU_1103 Bacteroides finegoldii 284 45 -1345 OTU_743 Bacteroidia 279 360 -1353 OTU_942 Clostridiales 0 230 -1395 OTU_800 Bacteroides vulgatus 1090 1310 YES -1437 OTU_1049 Bacteroidaceae 905 411 -1733 OTU_1217 Bacteroidaceae 405 47 -1806 OTU_1021 Bacteroides cellulosilyticus 201 90 YES -1863 OTU_1112 Ruminococcus bromii 45 73 -1963 OTU_1162 Clostridiales 6 137 -2071 OTU_359 Clostridia 0 0 -2186 OTU_50 Bacteroides thetaiotaomicron 1015 124 -3229 OTU_1088 Faecalibacterium 23 62 -3403 OTU_1089 Gemmiger formicilis 20 913 -3573 OTU_161 Lachnospiraceae 0 0 -4641 OTU_365 Bacteroides uniformis 1175 836 YES -4793 OTU_976 Faecalibacterium prausnitzii 0 366 -4867 OTU_262 Akkermansia muciniphila 102 402 -6489

185 Sup. Table 4. OTUs that were increased after two days and decreased after ten days of incubation compared with fresh sample

No reads Difference of reads day 2 No reads Difference of reads day 10 OTUs No Taxonomic assignement day 2 versus fresh day 10 versus fresh Culture

OTU_389 Eisenbergiella tayi 367 361 0 -6 OTU_376 Ruminococcus gnavus 447 203 0 -244 YES OTU_45 Blautia wexlerae 422 191 18 -213 OTU_251 Ruminococcaceae 224 142 66 -16 OTU_377 Blautia massiliensis 360 122 55 -183 OTU_842 Lachnospiraceae 175 98 10 -67 OTU_983 Bacteroides 108 95 0 -13 OTU_568 Clostridia 114 77 0 -37 OTU_146 Lachnospiraceae 103 60 0 -43 OTU_1220 Clostridiales 94 46 4 -44 OTU_603 Clostridiales 77 40 11 -26 OTU_231 Blautia 72 36 6 -30 OTU_93 Clostridia 42 36 0 -6 OTU_900 Bacteroidales 40 30 9 -1 OTU_487 Clostridia 31 24 0 -7 OTU_467 Parabacteroides merdae 120 21 8 -91 OTU_648 Oscilibacter massiliensis 680 19 66 -595 OTU_384 Bacteroidales 74 19 15 -40 OTU_1007 Ruminococcus 31 18 0 -13 OTU_201 Bacteroidaceae 27 18 0 -9 OTU_488 Eubacterium hallii 52 16 9 -27 OTU_1041 Bacteroidales 21 11 0 -10 OTU_133 Dorea formicigenerans 28 10 16 -2 OTU_1050 Blautia obeum 19 9 4 -6 OTU_1208 Ruminococcaceae 12 7 4 -1 OTU_886 Tannerellaceae 34 4 0 -30 OTU_468 Betaproteobacteriales 20 4 0 -16 OTU_1239 Bacteroides 14 4 0 -10 OTU_594 Phocea massiliensis 11 4 0 -7 OTU_957 Clostridiales 8 4 0 -4 OTU_992 Bacteroidia 8 4 0 -4 OTU_1164 Clostridiales 8 3 0 -5 OTU_562 Clostridia 67 2 49 -16 OTU_761 Bacteroidaceae 9 2 0 -7 OTU_418 Ruminiclostridium 7 2 0 -5 No: Number of

186 Sup. Table 5. OTUs that were decreased after two days and increased after ten days of incubation compared with fresh sample Difference of reads day 2 versus No reads Difference of reads day 10 versus OTUs No Taxonomic assignement No reads day 2 Culture fresh day 10 fresh OTU_494 Selenomonadales 1259 -1487 7738 4992 OTU_927 Tidjanibacter massiliensis 86 -838 3201 2277 OTU_235 Negativicutes 288 -791 1610 531 OTU_48 Bacteroidales 36 -329 441 76 OTU_541 Ruminococcaceae 5 -274 315 36 OTU_618 Negativicutes 95 -225 499 179 OTU_509 Clostridia 65 -180 353 108 OTU_87 Flavonifractor 41 -169 328 118 OTU_457 Rikenellaceae 5 -163 222 54 OTU_490 Clostridiales 14 -159 1008 835 OTU_350 Barnesiella intestinihominis 299 -158 625 168 YES OTU_55 Clostridia 12 -127 204 65 OTU_639 Intestinimonas butyriciproducens 15 -119 660 526 YES OTU_130 Clostridia 16 -115 169 38 OTU_280 Clostridiales 82 -113 204 9 OTU_1136 Bacteroidia 0 -71 572 501 OTU_917 Ruminococcaceae 14 -66 716 636 OTU_40 Clostridiales 22 -65 3249 3162 OTU_1265 Clostridiales 40 -64 112 8 OTU_805 Odoribacter splanchnicus 22 -60 164 82 YES OTU_916 Clostridia 0 -58 130 72 OTU_944 Bacteroidia 42 -41 172 89 OTU_1219 Ruminococcaceae 0 -41 196 155 OTU_387 Clostridiales 7 -37 88 44 OTU_120 Ruthenibacterium lactatiformans 16 -35 66 15 OTU_919 Ruminococcaceae 0 -35 141 106 OTU_416 Bacteroidales 0 -35 80 45 OTU_148 Fusicatenibacter saccharivorans 9 -30 95 56 OTU_766 Negativibacillus massiliensis 4 -30 44 10 OTU_253 Clostridiales 0 -28 87 59 OTU_611 Clostridia 0 -25 41 16 OTU_266 Anaerostipes hadrus 53 -21 470 396 OTU_995 Clostridiales 11 -18 120 91 OTU_49 Pseudoflavonifractor capillosus 6 -18 44 20 OTU_1001 Clostridia 0 -15 105 90 OTU_1012 Clostridia 0 -15 54 39 OTU_1138 Clostridiales 0 -14 91 77 OTU_1146 Clostridia 0 -14 18 4 OTU_1120 Clostridiales 11 -13 54 30 OTU_7 Clostridiales 0 -13 149 136 OTU_343 Bifidobacteriaceae 0 -13 47 34

187 OTU_1043 Clostridia 82 -12 140 46 OTU_561 Clostridiales 17 -12 61 32 OTU_156 Peptostreptococcaceae 0 -12 14 2 OTU_590 Hungatella 0 -11 63 52 OTU_599 Clostridiales 0 -11 34 23 OTU_729 Lachnospiraceae 0 -10 81 71 OTU_646 Marinifilaceae 0 -10 12 2 OTU_993 Clostridium aldenense 0 -10 23 13 YES OTU_330 Lachnospiraceae 0 -9 123 114 OTU_181 Clostridiales 0 -8 28 20 OTU_920 Clostridia 0 -7 41 34 OTU_741 Lachnospiraceae 0 -6 49 43 OTU_1052 Clostridia 0 -6 21 15 OTU_635 Bilophila wadsworthia 0 -6 8 2 OTU_182 Clostridia 40 -5 302 257 OTU_355 Eggerthella lenta 0 -5 493 488 YES OTU_317 Deltaproteobacteria 0 -5 16 11 OTU_412 Clostridia 0 -5 13 8 OTU_1100 Anaerostipes caccae 65 -4 279 210 OTU_405 Desulfovibrionaceae 0 -4 17 13 OTU_1033 Odoribacter splanchnicus 0 -4 10 6 OTU_730 Clostridia 0 -4 5 1 OTU_164 Clostridium symbiosum 7 -3 41 31 YES OTU_1095 Clostridia 6 -2 55 47 OTU_588 Lachnospiraceae 4 -2 112 106 No: Number of

188 CONCLUSION ET PERSPECTIVES

Les trois quarts des antibiotiques actuellement utilisés en santé humaine sont des produits naturels, ou des dérivés de produits naturels. Selon l’OMS, sur les 42 molécules actuellement « dans le pipeline », huit représentent une nouvelle classe et seulement un est un produit naturel (lefamulin, groupe des pleuromutilin, naturellement synthétisé par Clitopilus passeckerianus) (22). Ce dernier est d’ailleurs le plus avancé (étude clinique de phase 3 pour le traitement des pneumopathies aigues communautaires). Il appartient à une famille de molécules connues depuis 1950 ayant déjà un représentant commercialisé, la retapamulin, un antibiotique topique (16). La recherche de nouvelles molécules antibiotiques a connu un nouvel essor avec les progrès de la biologie moléculaire. Des approches innovantes, telles que le « genome mining » pour la recherche de BGCs, ou encore CRISPR Cas9 sont très prometteuses. Dans les approches par « genome mining », la confirmation de la fonctionnalité des BGCs détectés passe nécessairement par des test in vitro qui sont souvent le facteur limitant (23). Sur 3013 peptides antimicrobiens connus, 2530 possèdent une activité antibactérienne, et 11% sont produits par des bactéries (24).

Avec le développement de nouvelles méthodes de culture, le répertoire bactérien du tube digestif humain a connu un accroissement sans précédent. L’imitation de l’environnement naturel des bactéries a permis la découverte de nombreuses nouvelles espèces. Les microbiologistes environnementalistes ont été les premiers à essayer, avec succès, d’imiter l’environnement naturel des bactéries. Ling et al. ont découvert Eleftheria terrae, cultivé à partir du sol grâce à l’utilisation d’une chambre de diffusion. L’étude de cette espèce leur a permis de découvrir une nouvelle molécule antibiotique active sur S. aureus, la teixobactine (25).

189 Cette approche est également réalisable dans le microbiote humain. Par exemple, le nez constitue un environnement pauvre en nutriments dans lequel les bactéries vivent en forte compétition (26). C’est dans cet environnement que la lugdunine, une nouvelle molécule active sur S. aureus, a été découvert (27). Le tube digestif est également un milieu où réside une forte compétition entre les bactéries. Donia et al. ont listé 3118 BGCs incluant des NRPS,

RiPPs et des PKs dans le microbiome digestif, en faisant le deuxième site le plus abondant après le microbiome oral chez l’homme (18). Dans notre travail, 13% des bactéries testées présentaient un antagonisme non connu jusqu’à présent, et un tiers des BGCs identifiés étaient inconnus. Cette étude est préliminaire et nécessite d’aller plus loin. En effet, il reste à isoler et purifier les molécules antimicrobiennes afin de caractériser leurs propriétés. Ensuite, la délimitation du spectre d’activité de ces molécules, large ou étroit, est à définir. Nous avons notamment identifié six antagonistes de E. cloacae et deux de E. aerogenes, des entérobactéries du groupe 3 souvent isolés en clinique et posant régulièrement des problèmes en termes de résistances aux antibiotiques. Il serait donc intéressant pour poursuivre ce travail de tester ces antagonismes contre des entérobactéries résistantes aux céphalosporines de troisième génération ou encore résistantes aux carbapénèmes. Ce travail ouvre la voie à d’autres travaux de recherche d’antagonisme de pathogènes dans le microbiote digestif. De plus, les nouvelles espèces découvertes en culturomics représentent également une source potentielle de nouvelles molécules à explorer.

Dans la deuxième partie de cette thèse, nous avons montré que le milieu d’enrichissement utilisé en culturomics était efficace et permettait la culture d’un plus grand nombre d’espèces bactériennes bien que la biodiversité alpha soit diminuée. En particulier, des espèces bactériennes ont été isolées alors qu’elles étaient absentes, ou présentes avec très peu de reads en métagénomique. A l’inverse, des OTUs présents en grand nombre n’étaient pas retrouvé en culture. Ces OTUs correspondaient à des bactéries intolérantes à l’oxygènes.

190 Ils étaient également le plus diminué après deux et dix jours d’incubation, ce qui laisse supposer que ces bactéries étaient soit déjà mortes au moment de l’incubation, soit sont mortes précocement. En effet, il a été découvert récemment le phénomène de « suicide

écologique », dans lequel l’activité métabolique de bactéries peut entrainer leur extinction après seulement quelques heures (28). Les auteurs montrent que ce phénomène n’est pas rare puisque 25% des bactéries capables de modifier le pH environnemental étaient concernées.

De plus, ce phénomène est accentué lorsque les bactéries sont incubées dans un milieu enrichi en glucose. Si ce phénomène existe dans le microbiote digestif, le repiquage précoce pourrait conduire à l’isolement de nouvelles bactéries.

191

Annexes

192 ANNEXES

Mon travail en culturomics sur les bactéries anaérobies m’a amené à découvrir cinq nouveaux genres et cinq nouvelles espèces. Ces bactéries n’étaient pas connues de notre base donnée de spectres de référence MALDI-TOF, ce qui nous a amené à séquencer leur gène ribosomal 16S. Dans tous les cas celui-ci présentait une homologie de séquence inférieure à

98.65% avec les séquences de référence des espèces les plus proches « standing in nomenclature », permettant ainsi de proposer ces espèces comme nouvelles, avec un risque d’erreur de 0.5% pour les Firmicutes (29,30). La confirmation du statut de nouvelle espèce a ensuite été déterminé par une étude taxonogénomique, consistant en la comparaison phénotypique, biochimique et génomique avec les espèces les plus proches (31). Ces descriptions sont détaillées dans les publications de l’annexe 1.

Médecin biologiste, j’ai effectué ma thèse à mi-temps avec mon cursus de spécialisation en microbiologie au sein du laboratoire de l’IHU. Durant ces stages, je me suis intéressé à diverses thématiques. En particulier, Kingella kingae est un agent responsable d’arthrites et parfois d’endocardites chez le nourrisson, parfois également d’épidémies dans des crèches. La physiopathologie est en train d’être ré-inventée depuis la découverte de sa probable entrée dans l’organisme par voie orale à l’occasion d’une infection virale par virus

Coxsackie. J’ai ainsi pu participer à améliorer le diagnostic moléculaire de cette infection en aidant au travail de thèse de Nawal El Houmami. J’ai également eu l’occasion de suivre l’émergence du clone 078 de Clostridium difficile lors de mes semestres en biologie moléculaire. Ces publications sont présentées en annexe 2.

193

Annexe 1

Description de nouvelles espèces bactériennes par taxono-genomique

Publication n° 5: Blautia massiliensis sp. nov., isolated from a fresh human fecal sample and emended description of the genus Blautia

G. A. Durand, T. Pham, S. Ndongo, S. Traore, G. Dubourg, J-C. Lagier, C. Michelle, N. Armstrong, P-E. Fournier, D. Raoult, M. Million

Publié dans Anaerobe

Publication n° 6: “Intestinimonas massiliensis” sp. nov, a new bacterium isolated from human gut

G. A. Durand, P. Afouda, D. Raoult, G. Dubourg

Publié dans New Microbes and New Infections

Publication n° 7: Noncontiguous finished genome sequence and description of Intestinimonas massiliensis sp. nov strain GD2T, the second Intestinimonas species cultured from the human gut.

P. Afouda, G. A. Durand, J-C. Lagier, N. Labas, F. Cadoret, N. Armstrong, D. Raoult, G. Dubourg.

Publié dans Microbiology Open

Publication n° 8: Drancourtella massiliensis gen. nov., sp. nov. isolated from fresh healthy human faecal sample from South France

G. A. Durand, J-C. Lagier, S. Khelaifia, N. Armstrong, C. Robert, J. Rathored, P-E. Fournier, D. Raoult

Publié dans New Microbes and New Infections

Publication n° 9: ‘Bittarella massiliensis’ gen. nov., sp. nov. isolated by culturomics from the gut of a healthy 28-year-old man

G. A. Durand, P-E. Fournier, D. Raoult, S. Edouard

Publié dans New Microbes and New Infections

194 Publication n° 10: Description of Clostridium phoceensis sp. nov., a new species within the genus Clostridium

M. Hosny, S. Benamar, G. A. Durand, N. Armstrong, C. Michelle, F. Cadoret, B. La Scola, N. Cassir

Publié dans New Microbes and New Infections

Publication n° 11: Description of ‘Gorbachella massiliensis’ gen. nov., sp. nov., ‘Fenollaria timonensis’ sp. nov., ‘Intestinimonas timonensis’ sp. nov. and ‘Collinsella ihuae’ sp. nov. isolated from healthy fresh stools with culturomics

G. A. Durand, F. Cadoret, J. C. Lagier, P. E. Fournier, D. Raoult

Publié dans New Microbes and New Infections

Publication n° 12: Fournierella massiliensis gen. nov., sp. nov., a new human associated member of the family Ruminococcaceae

A. H. Togo, G. A. Durand, S. Khelaifia, N. Armstrong, C. Robert, F. Cadoret, F. Di Pinto, J. Delerce, A. Levasseur, D. Raoult, M. Million

Publié dans International Journal of Systematic and Taxonomy

195 Anaerobe 43 (2017) 47e55

Contents lists available at ScienceDirect

Anaerobe

journal homepage: www.elsevier.com/locate/anaerobe

Anaerobes in the microbiome Blautia massiliensis sp. nov., isolated from a fresh human fecal sample and emended description of the genus Blautia

Guillaume A. Durand a, b, Thao Pham a, Sokhna Ndongo a, Sory Ibrahima Traore a, Gregory Dubourg a, b, Jean-Christophe Lagier a, b, Caroline Michelle a, Nicholas Armstrong a, Pierre-Edouard Fournier a, b, Didier Raoult a, b, c, * Matthieu Million a, b, a URMITE UM63, CNRS7278, IRD198, INSERM1095, FacultedeMedecine, Aix Marseille Universite, 27 Boulevard Jean Moulin, 13385, Marseille Cedex 5, France b Pole^ des Maladies Infectieuses, Hopital^ La Timone, Assistance Publique-Hopitaux^ de Marseille, Marseille, France c Special Infectious Agents Unit, King Fahd Medical Research Center, King Abdulaziz University, Jeddah, Saudi Arabia article info abstract

Article history: The strain GD9T is the type strain of the newly proposed species Blautia massiliensis sp. nov., belonging to Received 15 October 2016 the family Lachnospiraceae. It was isolated from a fresh stool sample collected from a healthy human Received in revised form using the culturomics strategy. Cells are Gram-negative rods, oxygen intolerant, non-motile and non- 1 December 2016 spore forming. The 16S rRNA gene sequencing showed that strain GD9T was closely related to Blautia Accepted 2 December 2016 luti, with a 97.8% sequence similarity. Major fatty acids were C14:0 (19.8%) and C16:0 (53.2%). Strain GD9T Available online 5 December 2016 exhibits a genome of 3,717,339 bp that contains 3,346 protein-coding genes and 81 RNAs genes including Handling Editor: Elisabeth Nagy 63 tRNAs. The features of this organism are described here, with its complete genome sequence and annotation. Compared with other Blautia species which are Gram positive, the strain was Gram negative Keywords: justifying an emended description of the genus Blautia. Blautia massiliensis sp. nov. © 2016 Elsevier Ltd. All rights reserved. Taxono-genomics Culturomics Anaerobe Gut microbiota

1. Introduction demonstrated to be successful in the isolation of novel taxa [7,8]. The species described here was isolated from fresh human stool of a The description of the human gut microbiota as detailed in the healthy volunteer using the anaerobic conditions of the culturo- Human Microbiome Project [1] is an important concern for mi- mics approach. The species belongs to Blautia, a genus including 13 crobiologists, because of its importance to humans (53% of human- species, proposed by Liu et al. [9]. Million et al. [10] had previously linked species were cultivated from the gut [2]) and its potential shown that malnutrition was linked to a decrease of the relative link to diseases, such as chronic inflammatory bowel disease [3e6]. abundance of the Blautia genus in the human gut. Here, we describe The use of molecular tools with pyrosequencing is a widespread the type strain GD9T of the species for which we propose the name approach, but suffers from several biases including an extraction Blautia massiliensis (¼CSUR P2132, ¼ DSM 101187). bias or the detection of the most abundant species, thus neglecting minority but important species (the ‘depth’ bias). The combination of cultivation methods in tandem with pyrosequencing has been 2. Materials and methods

2.1. Ethics and sample collection

* Corresponding author. URMITE UM63, CNRS7278, IRD198, INSERM1095, Aix The stool specimen was collected at La Timone hospital in Marseille Universite, Fondation Mediterran ee Infection, 19-21 boulevard Jean Moulin, 13385, Marseille Cedex 5, France. Marseille (France) in April 2015 from a 28-year-old healthy French 2 E-mail address: [email protected] (M. Million). male (BMI 23.2 kg/m ), without current treatment. Informed and http://dx.doi.org/10.1016/j.anaerobe.2016.12.001 1075-9964/© 2016 Elsevier Ltd. All rights reserved.

196 48 G.A. Durand et al. / Anaerobe 43 (2017) 47e55

Courtaboeuf, France). Spectral database search was performed us- Abbreviations ing MS Search 2.0 operated with the Standard Reference Database 1A (NIST, Gaithersburg, USA) and the FAMEs mass spectral database fi URMITE Unite de Recherche sur les Maladies Infectieuses et (Wiley, Chichester, UK). The pro le obtained was compared to Blautia faecis T Blautia stercoris T Blautia coccoides Tropicales Emergentes M25 , GAM6-1 , T Blautia hansenii T Blautia schinkii CSUR Collection de Souches de l’Unite des Rickettsies DSM 935 , DSM 20583 , DSM 10518T, Blautia obeum DSM 25238T, and Blautia glucerasea DSM DSM Leibniz Institute DSMZ-German Collection of T Microorganisms and Cell Cultures, Braunschweig, 22028 [18]. Germany 2.4. Antibiotic susceptibility ATCC American Type Culture Collection MALDI-TOF MS Matrix-assisted laser-desorption/ionization time-of-flight mass spectrometry The susceptibility to classical antibiotics was tested with the FAME Fatty Acid Methyl Ester diffusion method according to CASFM/Eucast 2015 recommenda- GC/MS Gas Chromatography/Mass Spectrometry tions for fastidious anaerobes [19]. A suspension of 1 McFarland of TE buffer Tris-EDTA buffer species was grown on Wilkins-Chalgren agar (Sigma Aldrich, AGIOS Average Genomic Identity Of the gene Sequence Steinheim, Germany) supplemented with 5% sheep blood. In- cubations were performed under anaerobic conditions at 37 C and GGDC Genome-to-Genome Distance Calculator © dDDH digital DNA:DNA hybridization reading was done after 48 h using Sirscan system (i2a, Mont- pellier, France). Inhibition diameters were controlled by manual measurement using a ruler. signed consent, approved by the Institut Fed eratif de Recherche 48 2.5. Genome sequencing, annotation and comparison (Faculty of Medicine, Marseille, France), was obtained under T  agreement number 09-022. The strain GD9 was cultivated on Cos at 37 C under anaerobic atmosphere and then resuspended in 400 mL of TE buffer. Then, 200 mL of this suspension was diluted in 1 ml TE buffer for lysis 2.2. Isolation, identification and growth conditions of the strain treatment that included a 30 min incubation with 2.5 mg/mL lyso- zyme at 37 C, followed by an overnight incubation with 20 mg/mL The stool specimen was directly inoculated on Columbia agar proteinase K at 37 C [20]. Extracted DNA was then purified using supplemented with sheep blood (Cos) after dilutions, and then three successive phenol-chloroform extractions and ethanol pre- incubated at 37 C in anaerobic conditions. Subculturing was per- cipitations at À20 C overnight. After centrifugation, the DNA was formed on days one, two, five and ten on Cos at 37 C. Identification resuspended in 160 mL TE buffer. Quantification and sequencing of was performed using MALDI-TOF MS, comparing the spectrum the whole genome was done on the MiSeq Technology (IlluminaInc, with our database (which includes the Bruker database and our San Diego, CA, USA) with the mate pair strategy as previously own collection), as previously described [11]. When the identifi- described [21]. The open reading frames (ORFs) were predicted cation failed (score < 1.7), the 16S rRNA gene was amplified and using Prodigal [22] with default parameters but the predicted ORFs sequenced, as previously described [12]. The 16S rRNA sequence were excluded if they were spanning a sequencing gap region was compared to the nucleotide database using the BLAST simi- (contain N). The predicted bacterial protein sequences were larities web-service. In case of a sequence similarity value lower searched against the Clusters of Orthologous Groups (COG) using than 98.65%, the species was suspected to be novel, albeit without BLASTP (E-value 1e-03, coverage 0.7 and identity percent 30%). If no performing DNA-DNA hybridization, as previously suggested hit was found, a search was performed against the NR database [13,14]. Characterization of growth conditions was tested as pre- À using BLASTP with an E-value of 1e 03 coverage 0.7 and an identity viously described [11]. Sporulation and different culture conditions percent of 30%. If the sequence lengths were smaller than 80 amino were tested in order to determine the best culture conditions [15]. À acids, we used an E-value of 1e 05. The tRNAScanSE tool [23] was used to find tRNA genes, whereas ribosomal RNAs were found using 2.3. Morphological and biochemical characterization RNAmmer [24]. Lipoprotein signal peptides and the number of transmembrane helices were predicted using Phobius [25]. ORFans Morphological characterization was first performed by micro- were identified if all the performed BLASTP did not give positive scopic observation of Gram staining and motility of the fresh results (E-value smaller than 1eÀ03 for ORFs with sequence size À sample. Negative staining was performed after bacterial fixation in greater than 80 aa or E-value smaller than 1e 05 for ORFs with glutaraldehyde 2.5%. This solution was deposited on carbon for- sequence length smaller than 80 aa). Such parameter thresholds mvar film incubated for one second on ammonium molybdate 1%, have already been used in previous works to define ORFans [26,27]. dried on blotting paper and finally observed using TECNAI G20 For each selected genome, the complete genome sequence, prote- transmission electron microscope (FEI Company, Limeil-Brevannes, ome genome sequence and ORFeome genome sequence were France) at an operating voltage of 200 keV. Biochemical features, retrieved from the FTP of NCBI. All proteomes were analyzed with such as oxidase, catalase, API 20A, API ZYM and 50CH galleries proteinOrtho [28]. Then, for each genomes pair, a similarity score (Biomerieux, Marcy l’Etoile, France) were investigated, according to was computed. This score is the mean value of nucleotide similarity the manufacturer's instructions. Cellular FAME analysis was per- between all pairs of orthologues between the two genomes studied formed by GC/MS. Two samples were prepared with approximately (AGIOS) [29]. An annotation of the entire proteome was performed 30 mg of bacterial biomass per tube harvested from several culture to define the distribution of the functional classes of predicted plates, grown under anaerobic conditions on Cos agar for 48 h. genes according to the clusters of orthologous groups of proteins FAME were prepared as described previously [16]. GC/MS analyses (using the same method as for the genome annotation). Annotation were carried out as described before [17]. Briefly, fatty acid methyl and comparison of genome size, G þ C content, and gene content esters were separated using an Elite 5-MS column and monitored with other close species were performed in the Multi-Agent soft- by mass spectrometry (Clarus 500 - SQ 8 S, Perkin Elmer, ware system DAGOBAH [30], that include Figenix libraries. To

197 G.A. Durand et al. / Anaerobe 43 (2017) 47e55 49 assess the affiliation of our novel strain to the type strains of known microscopic observation showed bacilli with Gram-negative species with available genome, the Genome-to-Genome Distance staining (Supplementary Fig. 2). Electronic microscopy showed Calculator web service was used to calculate digital DNA:DNA hy- small rods of about 1 mm(Supplementary Fig. 3). Classification of bridization estimates (dDDH) with confidence intervals under the strain and main characteristics are presented in Table 2. recommended settings (Formula 2, BLASTþ) [31,32]. We observed no production reaction for catalase and oxidase. Using API 20A, positive reactions were found with D-Glucose, D- 3. Results and discussion Mannitol, D-Lactose, D-Saccharose, D-Maltose, Salicine, D-Xylose, L- Arabinose, Esculine, Glycerol, D-Cellobiose, D-Mannose, D-Melezi- fi 3.1. Classification and features tose, D-Raf nose, D-Sorbitol, L-Rhamnose and D-Trehalose. Using API ZYM strips, positive reactions were observed with a-Galacto- The first isolation of type strain GD9T occurred after direct sidase, b-Galactosidase, a-Glucosidase and b-Glucosidase. Using inoculation of fresh stool on Cos agar, without enrichment into API 50CH strips, positive reactions were found with Erythritol, D- blood bottles. The MALDI-TOF spectrum neither matched against Arabinose, D-Ribose, L-Xylose, L-Rhamnose, Dulcitol, N-Acetylglu- our database nor Brucker's one (Supplementary Fig. 1). The 16S cosamine, Amygdaline, Arbutine, Inuline, Amidon, Glycogene, rRNA gene is 1,493bp long (accession number: AA00076 from 16S Gentiobiose, D-Lyxose, D-Tagatose, D-Fucose, L-Fucose, Potassium IHU bank, LN890282 from EBI Sequence Database), with BLASTN Gluconate and Potassium 5-Cetogluconate. A comparison of search against reference sequences indicating Blautia luti DSM phenotypic and biochemical characteristics was made with other 14534T (NR_041960) as the most closely cultured species at 97.8% representatives of the family Lachnospiraceae (Table 3). The major (Fig. 1) [33]. The pairwise comparison of our strain with all type fatty acids found for this strain were C16:0 (53%) and C14:0 (20%). strains of the genus Blautia is represented in Table 1. Close species Saturated fatty acids were the most abundant and represented 83% on the basis of 16S rRNA tree and their presence into our MALDI- of the fatty acids found (Table 4). The cellular fatty acids of Blautia T fi TOF spectrum database were compared at the protein level with massiliensis GD9 were compared with the pro les of 7 other spe- B. massiliensis GD9T and represented in a gel view (Fig. 2). cies of the genus Blautia retrieved from the literature: B. faecis T T T Optimal growth was at 37 C after 48 h under anaerobic con- M25 , B. stercoris GAM6-1 , B. coccoides DSM 935 , B. hansenii DSM T T T ditions. Colonies appeared to be smooth, white, non-hemolytic, 20583 , B. schinkii DSM 10518 , B. obeum ATCC 29174 , and T fi non-motile, non spore-forming and 1 mm in size. Optical B. glucerasea DSM 22028 (pro les described by Park et al., 2013

Fig. 1. Phylogenetic tree highlighting the position of B. massiliensis strain GD9T relative to other phylogenetically close type strains. Genbank accession numbers of the 16S rRNA gene reference sequences are indicated in parenthesis. Sequences were aligned using CLUSTALW, and phylogenetic inferences were obtained with kimura two parameter model using neighbor-joining method with 1000 bootstrap replicates, within MEGA6 software. The scale bar represents a 1% nucleotide sequence divergence.

198 50 G.A. Durand et al. / Anaerobe 43 (2017) 47e55

Table 1 Pairwise comparison between Blautia massiliensis and type strains within the genus Blautia.

Type strains (sequence accession number)a BLAST similarity compared with B. massiliensis (LN890282)

B. coccoides (AB571656) 1409/1498 (94.1%) B. faecis (HM626178) 1216/1264 (96.2%) B. glucerasea (AB439724) 1392/1470 (94.7%) B. hansenii (AB534168) 1395/1493 (93.4%) B. hydrogenitrophica (X95624) 1357/1464 (92.7%) B. luti (AJ133124) 1307/1336 (97.8%) B. obeum (X85101) 1398/1460 (95.7%) B. producta (X94966) 1385/1478 (93.7%) B. schinkii (X94964) 1390/1475 (94.2%) B. stercoris (HM626177) 1256/1317 (95.4%) B. wexlerae (EF036467) 1391/1440 (96.6%)

a The sequences were those proposed as reference by LPSN.

Fig. 2. Gel view comparing B. massiliensis strain GD9T to other phylogenetically close species. The gel view displays the raw spectra of loaded spectrum files arranged in a pseudo-gel like look. The x-axis records the m/z value. The left y-axis displays the running spectrum number originating from subsequent spectra loading. The peak intensity is expressed by a gray scale scheme code. The color bar and the right y-axis indicate the relation between the color of the peak and its intensity, in arbitrary units. Displayed species are indicated on the left. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 2 showing higher relative abundance of C12:0 and C16:1n7. More- T Classification and general features of Blautia massiliensis strain GD9 . over, Liu et al. also described C16:0 and C14:0 as the most abundant Property Term fatty acids for Blautia species [9]. Antibiotic susceptibility was observed for the class of beta- Current classification Domain: Bacteria Phylum: Firmicutes lactam compounds that included amoxicillin, amoxicillin with Class: Clostridia clavulanic acid, ticarcillin with clavulanic acid, tazocillin, ceftriax- Order: Clostridiales one, cefepime and imipenem. For aminoglycoside antibiotics, Family: Lachnospiraceae resistance was observed for gentamicin, tobramycin and amikacin. Genus: Blautia T Species: Blautia massiliensis The GD9 strain was also susceptible to tigecyclin, rifampicin, Type strain: GD9T nitrofurantoin, but resistant to sulfamethoxazole, aztreonam and Gram stain Negative ciprofloxacin. Cell shape Rod Motility Non-motile Sporulation Non-sporulating Temperature range Mesophilic 3.2. Genomic characterization and comparison Optimum temperature 37 C The genome is 3,717,339 bp long with a 43.99% GC content (Table 5). It is composed of five scaffolds (composed of nine con- [18]). Regarding others species among the Blautia genus, fatty acid tigs). Of the 3,427 predicted genes, 3,346 were protein-coding profiles are comparable with C16:0 and C14:0 as the most prevalent genes, and 81 were RNAs (eight genes are 5S rRNA, six genes are fatty acids, except for B. schinkii, B. glucerasea and B. coccoides 16S rRNA, four genes are 23S rRNA, 63 genes are TRNA genes). A total of 2,570 genes (76.81%) were assigned as putative function (by

199 G.A. Durand et al. / Anaerobe 43 (2017) 47e55 51

Table 3 Comparison of B. massiliensis strain GD9T with close species Blautia luti DSM 14534T, Blautia coccoides DSM 935T, Blautia hensenii DSM 20583T, Blautia schinkii DSM 10518T, Blautia obeum DSM 25238T, na: non-available data.

Properties B. massiliensis B. luti B. coccoides B. hensenii B. schinkii B. obeum

Cell diameter (mm) 0.5 0.7e0.9 1e1.5 1e1.5 1e1.5 1e1.5 Oxygen requirement anaerobic anaerobic anaerobic anaerobic anaerobic anaerobic Gram stain e þþ þþþ Motility eeeeee Endospore formation eeeeee Indole eeeeee Production of Catalase eeeeee Oxidase e NA eeee Urease eeNA NA NA NA Gelatinase eeeeee a-Galactosidase þþþe þþ b-Galactosidase þþþþee a -Glucosidase þþþe þþ b -Glucosidase þþþe þ e Acid from L-Arabinose þþþe þþ Mannose þþþe þþ Mannitol þ e þ e NA NA Cellobiose þþþe þþ Raffinose þþþþþþ Saccharose þþþe þþ D-maltose þþþþþþ D-lactose þþþþNA NA Habitat human gut human gut human gut human gut human gut human gut

Table 4 Cellular fatty acid profiles of Blautia massiliensis strain GD9T compared with other Blautia type strains of closely related species.

Fatty acid NAME 1 2 3 4 5 6 7 8

C5:0 anteiso 2-methyl-Butanoic acid ND ND ND ND ND ND ND ND C12:0 Dodecanoic acid ND ND 2.3 3.4 2.6 26.1 2.1 50.2 C14:0 Tetradecanoic acid 19.8 39.2 47.3 35.9 19.2 31.6 29.3 15.7 C15:0 Pentadecanoic acid 4.3 1.3 0.7 1.1 0.6 0.6 0.9 ND C15:0 iso 13-methyl-tetradecanoic ND ND ND ND ND ND ND ND C16:0 Hexadecanoic acid 53.5 44.3 37.4 34.1 41.4 27.0 32.3 17.6 C16:1n5 11-Hexadecenoic acid ND 0.6 0.6 0.6 ND 0.5 ND ND C16:1n7 9-Hexadecenoic acid 2.6 2.4 3.0 15.4 3.2 3.7 6.8 ND C16:1n9 7-Hexadecenoic acid ND ND 0.4 ND 1.5 1.0 0.4 ND C17:0 Heptadecanoic acid 1.3 ND ND ND ND ND ND ND C17:1n7 10-Heptadecenoic acid 0.7 ND ND ND ND ND ND ND C18:0 Octadecanoic acid 4.7 2.3 3.5 1.8 10.4 2.1 5.4 1.4 C18:1n7 11-Octadecenoic acid 3.3 ND ND ND ND ND ND ND C18:1n9 9-Octadecenoic acid 6.1 8.8 18.3 7.7 20.8 7.0 22.9 13.9 C18:2n6 9,12-Octadecadienoic acid 1.7 0.7 0.8 ND 0.3 0.5 ND 1.3 C20:4n6 5,8,11,14-Eicosatetraenoic acid 2.1 ND ND ND ND ND ND ND

Strains: 1, B. massiliensis GD9T;2,B. faecis M25T;3,B. stercoris GAM6-1T;4,B. coccoides DSM 935T;5,B. hansenii DSM 20583T;6,B. schinkii DSM 10518T;7,B. obeum DSM 25238T;8,B. glucerasea DSM 22028T. Strains 2 to 8 were described by Park, 2013 [18]. Values represent the percentage of total identified fatty acid methyl esters only (al- dehydes, dimethyl acetals and unidentified “summed features” described previously were not included). ND, Not detected.

Table 5 cogs or by NR blast). 101 genes were identified as ORFans (3.02%). Nucleotide content and gene count levels of the genome. The remaining genes were annotated as hypothetical proteins (581 Attribute Genome (total) genes, 17.36%, Supplementary Fig. 4). The genome sequence has

Value % of totala been deposited in EMBL-EBI under accession number PRJEB11857. The draft genome sequence of B. massiliensis is smaller than that Size (bp) 3,717,339 100 þ of Blautia schinkii and Blautia obeum (3.72, 6.68 and 3.76 MB G C content (bp) 1,634,069 43.98 T Coding region (bp) 3,269,367 87.94 respectively), but larger than that of Blautia wexlerae DSM 19850 , Total genes 3427 100 Blautia hydrogenotrophica, and Blautia hansenii (3.58, 3.35 and RNA genes 81 2.36 3.05 MB respectively). The G þ C content of B. massiliensis (44.0%) is Protein-coding genes 3346 97.63 smaller than that of B. schinkii and B. hydrogenotrophica (44, 46 and Proteins with function prediction 2570 76.8 B. wexlerae B. hansenii Genes assigned to COGs 1978 59.11 45 respectively), but larger than that of , and Genes with peptide signals 361 10.78 B. obeum (42, 39 and 42% respectively). The gene content of Genes with transmembrane helices 742 22.17 B. massiliensis is smaller than that of B. schinkii (3346 and 5851 Genes with Pfam domains 3091 90 respectively), but larger than that of B. wexlerae (3,297), a The total is based on either the size of the genome in base pairs or the total B. hydrogenotrophica (3,087), B. hansenii (3,171) and B. obeum number of protein coding genes in the annotated genome. (3,155).

200 52 G.A. Durand et al. / Anaerobe 43 (2017) 47e55

Table 6 Pairwise comparison of B. massiliensis GD9T with other species using the GGDC web-service, formula 2 (upper right)a.

B. massiliensis B. obeum DSM B. schinkii B. hansenii DSM B. hydrogenotrophica DSM B. wexlerae DSM GD9T 25238T DSM 20583T 10507T 19850T 10518T

B. massiliensis GD9T 100% ± 00 31.8 ± 2.5 20.8% ± 2.4 36.9% ± 2.6 21.5% ± 2.4 31.1% ± 2.5 B. obeum DSM 25238T 100% ± 00 20.5% ± 2.4 27.1% ± 2.5 20.8%±±2.4 27.9% ± 2.5 B. schinkii DSM 10518T 100% ± 00 23.9% ± 2.5 20.4% ± 2.4 23.2% ± 2.5 B. hansenii DSM 20583T 100% ± 00 22.4% ± 2.5 38.0% ± 2.5 B. hydrogenotrophica DSM 100% ± 00 19.9% ± 2.4 10507T B. wexlerae DSM 19850T 100% ± 00

a dDDH values are DDH estimates based on identities/HSP length. The confidence intervals indicate the inherent uncertainty in estimating DDH values from intergenomic distances based on models derived from empirical test datasets (which are always limited in size).

Table 7 Numbers of orthologous protein shared between genomes (upper right), AGIOS values (lower left) and numbers of proteins per genome (bold numbers).

Blautia obeum DSM Blautia schinkii DSM Blautia hansenii DSM Blautia hydrogenotrophica Blautia wexlerae DSM Blautia massiliensis 25238T 10518T 20583T DSM 10507T 19850T GD9T

Blautia obeum DSM 25238T 3155 1389 1086 1111 1393 1450 Blautia schinkii DSM 10518T 65.32 5851 1313 1371 1556 1586 Blautia hansenii DSM 20583T 63.63 66.87 3171 1128 1224 1225 Blautia hydrogenotrophica 63.25 68.34 67.33 3087 1227 1232 DSM 10507T Blautia wexlerae DSM 19850T 67.75 72.33 68.37 68.41 3297 1600 Blautia massiliensis GD9T 68.08 71.80 68.44 68.18 74.75 3346

DNA-DNA hybridization (DDH) is currently considered as the we used two parameters: digital DDH (dDDH) that exhibits a high “gold standard” criterion for species delineation of prokaryotes. correlation with DDH [32,37] and AGIOS [29] that was designed to However, this tool suffers from limitations, notably the 70% cutoff be independent from DDH. When considering only Blautia species value that is not applicable to all prokaryotic genera [34], and the with available genome and standing in nomenclature, dDDH values need of special facilities that are available in a limited number of ranged from 19.9 ± 2.4% between B. hydrogenotrophica and laboratories. Moreover, determining DDH is a labor-intensive B. wexlerae to 38.0 ± 2.5% between B. hansenii and B. wexlerae. method that lacks reproducibility and cannot be used to establish When comparing strain GD9T, dDDH values ranged from a comparative reference database incrementally [35,36]. Therefore, 20.8% ± 2.4 with B. schinkii strain DSM10518T to 36.9% ± 2.6 with in order to evaluate the genomic similarity among studied strains, B. hansenii strain DSM 20583T with a probability of error of 1.19%

Fig. 3. Distribution of functional classes of predicted genes according to the clusters of orthologous groups of proteins. The functional classes of predicted genes was assessed using the NCBI COGs database updated in 2014 [43] and the Blastp [44] tool (Evalue 1e-03, coverage 0,7 and identity percent 30%). The tRNA genes were predicted with the tRNAscan-SE tool [23] and the RNA genes were predicted with the Rnammer [24] tool. The figure was made using the DNAPlotter [45] tool.

201 G.A. Durand et al. / Anaerobe 43 (2017) 47e55 53

Table 8 Number of genes associated with the 25 general COG functional categories.

Code Value % of totala Description

[J] 195 5.83 Translation [A] 0 0 RNA processing and modification [K] 182 5.44 Transcription [L] 89 2.66 Replication, recombination and repair [B] 0 0 Chromatin structure and dynamics [D] 32 0.96 Cell cycle control, mitosis and meiosis [Y] 0 0 Nuclear structure [V] 86 2.57 Defense mechanisms [T] 111 3.32 Signal transduction mechanisms [M] 95 2.84 Cellwall/membrane biogenesis [N] 13 0.39 Cellmotility [Z] 0 0 Cytoskeleton [W] 4 0.12 Extracellular structures [U] 25 0.75 Intracellular trafficking and secretion [O] 81 2.42 Post translational modification, protein turnover, chaperones [X] 58 1.73 Mobilome: prophages, transposons [C] 125 3.74 Energy production and conversion [G] 229 6.84 Carbohydrate transport and metabolism [E] 212 6.34 Amino acid transport and metabolism [F] 82 2.45 Nucleotide transport and metabolism [H] 139 4.15 Coenzyme transport and metabolism [I] 65 1.94 Lipid transport and metabolism [P] 94 2.81 Inorganic ion transport and metabolism [Q] 29 0.87 Secondary metabolites biosynthesis, transport and catabolism [R] 165 4.93 General function prediction only [S] 94 2.81 Function unknown _ 1368 40.89 Not in COGs

a The total is based on the total number of protein coding genes in the annotated genome. according to GGDC (Table 6). Among compared Blautia species, strain GD9 accounted for as much as 20% of all reads in 4 human gut AGIOS values ranged from 63.25 between B. hydrogenotrophica and metagenomes (SRR2143746, 27%; SRR2658700, 27%; SRR515348, B. obeum to 72.33% between B. schinkii and B. wexlerae among 24%; SRR578392, 22%). compared species except B. massiliensis and from 68.08 with The specific role of Blautia species in human health remains to B. obeum to 74.75% with B. wexlerae (Table 7). Fig. 3 demonstrates be determined. Jenq et al. found a lower mortality due to a graft- that strain GD9T exhibits a similar distribution of genes into COG versus-host disease after allogenic blood/marrow transplantation categories when compared to other Blautia species. B. massiliensis among patients with high abundance of Blautia sequences obtained also shared 1450; 1586; 1225; 1232 and 1600 orthologous proteins by pyrosequencing [39]. Chen et al. have shown that colorectal with B. obeum, B. schinkii DSM 10518, B. hansenii DSM 20583, cancer was associated with lower Blautia in their digestive tract B. hydrogenotrophica DSM 10507T and B. wexlerae DSM 19850 [40]. Touyama et al. calculated that B. wexlerae and B. luti were respectively (Table 8). found at the concentration of 5.109 bacteria per gram of stool [41], suggesting that Blautia are important members of the healthy hu- 3.3. Importance of B. massiliensis in human gut and possible role in man mature anaerobic gut microbiota. human health

Blautia massiliensis is likely to be an important species for hu- 4. Conclusion man health because it was isolated and detected frequently and abundantly in human gut culturomes and metagenomes. Anaerobic culturomics conditions applied to fresh human stool B. massiliensis was isolated from the fecal samples from three other have permitted the culture of a new species belonging to Blautia. healthy individuals by three other researchers (TP, SD, ST) in our This taxogenomic study confirmed the new species Blautia massi- culturomics team. These strains were identified by MALDI-TOF MS liensis sp. nov., which appeared close to B. hensenii strain DSM and confirmed by sequencing of the 16S rRNA (unpublished data). 20583 on genome-based analysis. These species are both anaerobes B. massiliensis corresponded to 4.25% of all reads (200,239/ inhabitants of the human gut, indole and catalase negative, sharing 4,716,269 reads) in an ongoing study on gut microbiota and bar- the same repartition of cellular fatty acids C14:0 and C16:0. How- iatric surgery in our lab, ranking 6th in relative abundance. In ever, B. massiliensis has lower C18:0 than B. hensenii, but higher addition, we investigated the presence of 16S rRNA from G þ C content. B. massiliensis is the only species with Gram-negative B. massiliensis in the high-throughput DNA and RNA sequence read staining among the Blautia genus. The closest species on the 16S archive (SRA) using an open resource online [38]. We found se- rRNA sequence identity was B. luti, for which no genome was quences with a similarity greater than 99% with B. massiliensis in available. However, beyond the 16S rRNA divergence (2.2%), several gut metagenomes (human, mouse and primate), skin and B. massiliensis GD9T differed from B. luti by several phenotypic vaginal metagenomes; as well as from environmental samples characteristics including a lower diameter (0.5 versus 0.7e0.9 mm), (wastewater, bioreactor, coral). Sequences corresponding to a negative Gram-staining (verified several times with positive B. massiliensis were found in 11% (9752/88,579) of all metagenomes controls), an absence of bacterial chains (chains up to 10 cells for and 41% (6839/16,667) of human gut metagenomes in this data- B. luti [42]), and the use of mannitol, melezitose, rhamnose and base. Among the 6839 positive samples from human gut, the mean salicin (not found for B. luti [42]). This led us to propose relative abundance was 0.0088 but sequences corresponding to B. massiliensis as a new species.

202 54 G.A. Durand et al. / Anaerobe 43 (2017) 47e55

4.1. Taxonomic and nomenclatural proposals [3] R.H. Siggers, J. Siggers, M. Boye, T. Thymann, L. Mølbak, T. Leser, B.B. Jensen, P.T. Sangild, Early administration of probiotics alters bacterial colonization and limits diet-induced gut dysfunction and severity of necrotizing entero- 4.1.1. Emended description of the genus Blautia colitis in preterm pigs, J. Nutr. 138 (2008) 1437e1444. Blautia (Blauti.a. N.L. fem. n. Blautia in honour of Michael Blaut, a [4] G. De Hertogh, J. Aerssens, R. de Hoogt, P. Peeters, P. Verhasselt, P. Van Eyken, German microbiologist, in recognition of his many contributions to N. Ectors, S. Vermeire, P. Rutgeerts, B. Coulie, K. Geboes, Validation of 16S rDNA sequencing in microdissected bowel biopsies from Crohn's disease pa- human gastrointestinal microbiology). Gram-positive or Gram- tients to assess bacterial flora diversity, J. Pathol. 209 (2006) 532e539. negative staining, non-motile. Coccoid or ovalshaped, pointed [5] C. Manichanh, L. Rigottier-Gois, E. Bonnaud, K. Gloux, E. Pelletier, L. Frangeul, ends are often observed. Spores are not normally observed, but may R. Nalin, C. Jarrin, P. Chardon, P. Marteau, J. Roca, J. Dore, Reduced diversity of faecal microbiota in Crohn's disease revealed by a metagenomic approach, Gut be produced by some strains. Chemo-organotrophic and obligately 55 (2006) 205e211. anaerobic having a fermentative type of catabolism. Some species [6] P.D. Scanlan, F. Shanahan, C. O'Mahony, J.R. Marchesi, Culture-independent use H2/CO2 as major energy sources. The major end products of analyses of temporal variation of the dominant fecal microbiota and targeted bacterial subgroups in Crohn's disease, J. Clin. Microbiol. 44 (2006) glucose metabolism are acetate, ethanol, hydrogen, lactate and 3980e3988. succinate. The G þ C content of the DNA is 37e47 mol%. Isolated [7] J.-C. Lagier, P. Hugon, S. Khelaifia, P.-E. Fournier, B. La Scola, D. Raoult, The from animal and human faeces. The type species of the genus is rebirth of culture in microbiology through the example of culturomics to e Blautia coccoides (Kaneuchi, Benno & Mitsuoka, 1976). study human gut microbiota, Clin. Microbiol. Rev. 28 (2015) 237 264. [8] J.-C. Lagier, S. Khelaifia, M.T. Alou, S. Ndongo, N. Dione, P. Hugon, A. Caputo, F. Cadoret, S.I. Traore, E.H. Seck, G. Dubourg, G. Durand, G. Mourembou, 4.1.2. Description of Blautia massiliensis, sp. nov E. Guilhot, A. Togo, S. Bellali, D. Bachar, N. Cassir, F. Bittar, J. Delerce, M. Mailhe, Blautia massiliensis (ma.si.li.en'sis. L. fem. adj. Massiliensis, from D. Ricaboni, M. Bilen, N.P.M. Dangui Nieko, N.M. Dia Badiane, C. Valles, D. Mouelhi, K. Diop, M. Million, D. Musso, J. Abrahao,~ E.I. Azhar, F. Bibi, the Latin Massilia, the city where the bacteria was isolated, Mar- M. Yasir, A. Diallo, C. Sokhna, F. Djossou, V. Vitton, C. Robert, J.M. Rolain, B. La seille) presented white smooth colonies of 1 mm diameter. The Scola, P.-E. Fournier, A. Levasseur, D. Raoult, Culture of previously uncultured bacteria appear to be 1 mm long and with Gram-negative staining. members of the human gut microbiota by culturomics, Nat. Microbiol. 1 (2016) 16203, http://dx.doi.org/10.1038/nmicrobiol.2016.203. Their metabolism is oxygen intolerant, and growth is optimal at [9] C. Liu, S.M. Finegold, Y. Song, P.A. Lawson, Reclassification of Clostridium 37 C. B. massiliensis is non spore-forming and non-motile. Oxidase coccoides, Ruminococcus hansenii, Ruminococcus hydrogenotrophicus, Rumino- and catalase are negative. Positive reactions observed using API 20A coccus luti, Ruminococcus productus and Ruminococcus schinkii as Blautia coc- coides gen. nov., comb. nov., Blautia hansenii comb. nov., Blautia are D-Glucose, D-Mannitol, D-Lactose, D-Sucrose, D-Maltose, Sali- hydrogenotrophica comb. nov., Blautia luti comb. nov., Blautia producta comb. cine, D-Xylose, L-Arabinose, Esculine, Glycerol, D-Cellobiose, D- nov., Blautia schinkii comb. nov. and description of Blautia wexlerae sp. nov., e Mannose, D-Melezitose, D-Raffinose, D-Sorbitol, L-Rhamnose, and D- isolated from human faeces, Int. J. Syst. Evol. Microbiol. 58 (2008) 1896 1902. [10] M. Million, A. Diallo, D. Raoult, Gut microbiota and malnutrition, Microb. Trehalose. Using API 50CH strips, positive reactions are found with Pathog. (2016 Feb 4), http://dx.doi.org/10.1016/j.micpath.2016.02.003 pii: Erythitol, D-Arabinose, D-Ribose, L-Xylose, L-Rhamnose, Dulcitol, N- S0882-4010(15)30212-6. [Epub ahead of print]. Acetylglucosamine, Amygdaline, Arbutine, Inuline, Amidon, Gly- [11] J.-C. Lagier, K. Elkarkouri, R. Rivet, C. Couderc, D. Raoult, P.-E. Fournier, Non contiguous-finished genome sequence and description of Senegalemassilia cogene, Gentiobiose, D-Lyxose, D-Tagatose, D-Fucose, L-Fucose, Po- anaerobia gen. nov., sp. nov, Stand. Genomic. Sci. 7 (2013) 343e356. tassium Gluconate and Potassium 5-Cetogluconate. Antibiotic [12] G. Mourembou, J. Rathored, J.B. Lekana-Douki, A. Ndjoyi-Mbiguino, F. Fenollar, resistance for aminoglycosides, sulfamethoxazole, aztreonam, and C. Michelle, P.-E. Fournier, D. Raoult, J.-C. Lagier, Noncontiguous finished ciprofloxacin are observed. The major fatty acids are C16:0 and genome sequence and description of Kallipyga gabonensis sp. nov, New Mi- crobes. New Infect. 9 (2016) 15e23. C14:0. This strain exhibits a G þ C content of 44.0%. Its 16S rRNA [13] E. Stackebrandt, J. Ebers, Taxonomic parameters revisited: tarnished gold sequence is deposited in EMBL-EBI under the accession number standards, Microbiol. Today 33 (2006) 152e155. LN890282, and the whole genome shotgun sequence is deposited [14] M. Kim, H.-S. Oh, S.-C. Park, J. Chun, Towards a taxonomic coherence between average nucleotide identity and 16S rRNA gene sequence similarity for species in EMBL-EBI under accession number PRJEB11857. The type species demarcation of prokaryotes, Int. J. Syst. Evol. Microbiol. 64 (2014) 346e351. is Blautia massiliensis strain GD9T (¼ CSUR P2132, ¼ DSM 101187) [15] R. Aghnatios, C. Cayrou, M. Garibal, C. Robert, S. Azza, D. Raoult, M. Drancourt, isolated from the stool of a healthy person from Marseille, France. Draft genome of Gemmata massiliana sp. nov, a water-borne Planctomycetes species exhibiting two variants, Stand. Genomic. Sci. 10 (2015). [16] Myron Sasser, Bacterial Identification by Gas Chromatographic Analysis of Conflict of interest Fatty Acids Methyl Esters (GC-FAME), MIDI, 2006. [17] N. Dione, et al., Genome sequence and description of Anaerosalibacter massi- e fl liensis sp. nov. New Microbes. New Infect 10 (2016) 66 76. The authors declare no con ict of interest. [18] S.-K. Park, M.-S. Kim, J.-W. Bae, Blautia faecis sp. nov., isolated from human faeces, Int. J. Syst. Evol. Microbiol. 63 (2013) 599e603.  Acknowledgements [19] European Committee on Antimicrobial Susceptibility Testing, Comitede l’Antibiogramme de la Societ e Française de Microbiologie, CASFM-EUCAST recommendations (n.d.), http://www.sfm-microbiologie.org/UserFiles/files/ The authors thank the Xegen Company (www.xegen.fr) for casfm/CASFM_EUCAST_V1_2015.pdf, 2015. Last access 2016/11/28. automating the genomic annotation process. This study was funded [20] B. Sengüven, E. Baris, T. Oygur, M. Berktas, Comparison of methods for the fi fi “   ” extraction of DNA from formalin- xed, paraf n-embedded archival tissues, by the Fondation Mediterranee Infection . We also thank Karolina Int. J. Med. Sci. 11 (2014) 494e499. Griffiths and Magdalen Lardiere for English reviewing, Nicholas [21] J.-C. Lagier, F. Bibi, D. Ramasamy, E.I. Azhar, C. Robert, M. Yasir, A.A. Jiman- Armstrong and Magali Richez who performed the fatty acids Fatani, K.Z. Alshali, P.-E. Fournier, D. Raoult, Non contiguous-finished genome sequence and description of Clostridium jeddahense sp. nov, Stand. Genomic. analyses. Sci. 9 (2014) 1003e1019. [22] Prodigual, (n.d.). http://prodigal.ornl.gov/. Last access 2016/11/28. Appendix A. Supplementary data [23] T.M. Lowe, S.R. Eddy, tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence, Nucleic Acids Res. 25 (1997) 955e964. Supplementary data related to this article can be found at http:// [24] K. Lagesen, P. Hallin, E.A. Rødland, H.-H. Staerfeldt, T. Rognes, D.W. Ussery, dx.doi.org/10.1016/j.anaerobe.2016.12.001. RNAmmer: consistent and rapid annotation of ribosomal RNA genes, Nucleic Acids Res. 35 (2007) 3100e3108. [25] J.D. Bendtsen, H. Nielsen, G. von Heijne, S. Brunak, Improved prediction of References signal peptides: SignalP 3.0, J. Mol. Biol. 340 (2004) 783e795. [26] J.-C. Lagier, F. Armougom, M. Million, P. Hugon, I. Pagnier, C. Robert, F. Bittar, [1] P.J. Turnbaugh, R.E. Ley, M. Hamady, C.M. Fraser-Liggett, R. Knight, J.I. Gordon, G. Fournous, G. Gimenez, M. Maraninchi, J.-F. Trape, E.V. Koonin, B. La Scola, The human microbiome project, Nature 449 (2007) 804e810. D. Raoult, Microbial culturomics: paradigm shift in the human gut micro- [2] P. Hugon, J.-C. Dufour, P. Colson, P.-E. Fournier, K. Sallah, D. Raoult, biome study, Clin. Microbiol. Infect. Off. Publ. Eur. Soc. Clin. Microbiol. Infect. A comprehensive repertoire of prokaryotic species identified in human beings, Dis. 18 (2012) 1185e1193. Lancet Infect. Dis. 15 (2015) 1211e1219. [27] A.H. Togo, S. Khelaifia, J.-C. Lagier, A. Caputo, C. Robert, P.-E. Fournier,

203 G.A. Durand et al. / Anaerobe 43 (2017) 47e55 55

M. Maraninchi, R. Valero, D. Raoult, M. Million, Noncontiguous finished Syst. Evol. Microbiol. 64 (2014) 384e391. genome sequence and description of Paenibacillus ihumii sp. nov. strain AT5, [37] A.F. Auch, M. von Jan, H.-P. Klenk, M. Goker,€ Digital DNA-DNA hybridization New Microbes. New Infect. 10 (2016) 142e150. for microbial species delineation by means of genome-to-genome sequence [28] M. Lechner, S. Findeiss, L. Steiner, M. Marz, P.F. Stadler, S.J. Prohaska, Protei- comparison, Stand. Genomic. Sci. 2 (2010) 117e134. northo: detection of (co-)orthologs in large-scale analysis, BMC Bioinforma. [38] I. Lagkouvardos, D. Joseph, M. Kapfhammer, S. Giritli, M. Horn, D. Haller, 12 (2011) 124. T. Clavel, IMNGS: a comprehensive open resource of processed 16S rRNA [29] D. Ramasamy, A.K. Mishra, J.-C. Lagier, R. Padhmanabhan, M. Rossi, microbial profiles for ecology and diversity studies, Sci. Rep. 6 (2016) 33721. E. Sentausa, D. Raoult, P.-E. Fournier, A polyphasic strategy incorporating [39] R.R. Jenq, Y. Taur, S.M. Devlin, D.M. Ponce, J.D. Goldberg, K.F. Ahr, genomic data for the taxonomic description of novel bacterial species, Int. J. E.R. Littmann, L. Ling, A.C. Gobourne, L.C. Miller, M.D. Docampo, J.U. Peled, Syst. Evol. Microbiol. 64 (2014) 384e391. N. Arpaia, J.R. Cross, T.K. Peets, M.A. Lumish, Y. Shono, J.A. Dudakov, H. Poeck, [30] P. Gouret, J. Paganini, J. Dainat, D. Louati, E. Darbo, P. Pontarotti, P. Pontarotti, A.M. Hanash, J.N. Barker, M.-A. Perales, S.A. Giralt, E.G. Pamer, M.R.M. van den Integration of Evolutionary Biology Concepts for Functional Annotation and Brink, Intestinal Blautia is associated with reduced death from graft-versus- Automation of Complex Research in Evolution: the Multi-agent Software host disease, biol. Blood marrow transplant, J. Am. Soc. Blood Marrow System DAGOBAH, in: Evol. Biol. e Concepts Biodivers, Macroevolution Transpl. 21 (2015) 1373e1383. Genome Evol., Springer, Berlin Heidelberg, 2011, pp. 71e87. [40] W. Chen, F. Liu, Z. Ling, X. Tong, C. Xiang, Human intestinal lumen and [31] A.F. Auch, H.-P. Klenk, M. Goker,€ Standard operating procedure for calculating mucosa-associated microbiota in patients with colorectal cancer, PloS One 7 genome-to-genome distances based on high-scoring segment pairs, Stand. (2012) e39743. Genomic. Sci. 2 (2010) 142e148. [41] M. Touyama, J.S. Jin, R. Kibe, H. Hayashi, Y. Benno, Quantification of Blautia [32] J.P. Meier-Kolthoff, A.F. Auch, H.-P. Klenk, M. Goker,€ Genome sequence-based wexlerae and Blautia luti in human faeces by real-time PCR using specific species delimitation with confidence intervals and improved distance func- primers, Benef. Microbes. 6 (2015) 583e590. tions, BMC Bioinforma. 14 (2013) 60. [42] R. Simmering, D. Taras, A. Schwiertz, G, Le Blay, B. Gruhl, P.A. Lawson, [33] P. Yarza, M. Richter, J. Peplies, J. Euzeby, R. Amann, K.-H. Schleifer, W. Ludwig, M.D. Collins, M. Blaut, Ruminococcus luti sp. nov., isolated from a human faecal F.O. Glockner,€ R. Rossello-M ora, The All-Species Living Tree project: a 16S sample, Syst. Appl. Microbiol. 25 (2002) 189e193. rRNA-based phylogenetic tree of all sequenced type strains, Syst. Appl. [43] M.Y. Galperin, K.S. Makarova, Y.I. Wolf, E.V. Koonin, Expanded microbial Microbiol. 31 (2008) 241e250. genome coverage and improved protein family annotation in the COG data- [34] P.-E. Fournier, D. Raoult, Current knowledge on phylogeny and taxonomy of base, Nucleic Acids Res. 43 (2015) D261eD269. Rickettsia spp. Ann. N. Y. Acad. Sci. 1166 (2009) 1e11. [44] C. Camacho, G. Coulouris, V. Avagyan, N. Ma, J. Papadopoulos, K. Bealer, [35] H.-J. Busse, B.J. Tindall, W. Ludwig, R. Rossello-M ora, P. Kampfer,€ Notes on the T.L. Madden, BLASTþ: architecture and applications, BMC Bioinforma. 10 characterization of prokaryote strains for taxonomic purposes, Int. J. Syst. (2009) 421. Evol. Microbiol. 60 (2010) 249e266. [45] T. Carver, N. Thomson, A. Bleasby, M. Berriman, J. Parkhill, DNAPlotter: circular [36] D. Ramasamy, A.K. Mishra, J.-C. Lagier, R. Padhmanabhan, M. Rossi, and linear interactive genome visualization, Bioinforma. Oxf, Engl 25 (2009) E. Sentausa, D. Raoult, P.-E. Fournier, A polyphasic strategy incorporating 119e120. genomic data for the taxonomic description of novel bacterial species, Int. J.

204 1

2 Supplementary Figure 1. Reference mass spectrum from B. massiliensis strain GD9T.

3 Spectra from 12 individual colonies were compared and a reference spectrum was generated.

205 4

5 Supplementary Figure 2. Gram-staining of B.massiliensis strain GD9T.

206 6

7 Supplementary Figure 3. Transmission electron microscopy of B.massiliensisstrain GD9T,

8 using a TECNAI G20 (FEI) at an operating voltage of 200keV. The scale bar represents 500

9 nm.

207

60,0

50,0

40,0 B. massiliensis P2132 Blautia faecis M25T (1) 30,0 B. stercoris GAM6-1T (2) B. coccoides DSM 935T (3) 20,0 B. hansenii DSM 20583T (4) B. schinkii DSM 10518T (5) 10,0 R. obeum ATCC 29174T (6) B. glucerasea DSM 22028T (7) 0,0

Supplementary Figure 4.

208

Supplementary Figure 5. Graphical circular map of the chromosome of B.massiliensis strainGD9T. From outside to the center: Genes on the forward strand colored by COG categories (only genes assigned to COG), genes on the reverse strand colored by COG categories (only genes assigned to COG), RNA genes (tRNAs green, rRNAs red), G+C content and G+C skew.

209 NEW SPECIES

“Intestinimonas massiliensis” sp. nov, a new bacterium isolated from human gut

G. Durand1,3, P. Afouda1, D. Raoult1,2,3 and G. Dubourg1,3 1) Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes, UM 63, CNRS 7278, IRD 198, Inserm 1095, Institut Hospitalo-Universitaire Méditerranée-Infection, Faculté de médecine, Aix-Marseille Université, 2) Institut Hospitalo-Universitaire (IHU) Méditerranée Infection, Assistance Publique-Hôpitaux de Marseille and 3) Pôle des Maladies Infectieuses et Tropicales Clinique et Biologique, Fédération de Bactériologie-Hygiène–Virologie, University Hospital Centre Timone, Marseille, France

Abstract

Here we report the main features of the proposed new bacterial species “Intestinimonas massiliensis” sp. nov. The type strain GD2T (CSUR = P1930) was isolated from the gut microbiota of a healthy patient using a culturomics approach combined with taxonogenomics. © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases.

Keywords: Anaerobe, culturomics, gut microbiota, “Intestinimonas massiliensis” sp. nov., taxonogenomics Original Submission: 21 September 2016; Accepted: 28 September 2016 Article published online: 3 October 2016

Colonies appeared white and regular with a mean diameter Corresponding author: G. Dubourg, URMITE UM63, CNRS7278, of 1 to 2 mm on blood agar–enriched Colombia. “Intestinimonas IRD198, INSERM1095, Faculté de Médecine, Aix-Marseille Université, 27 boulevard Jean Moulin, 13385 Marseille cedex 5, France massiliensis” is a nonmotile, Gram-negative rod with a mean E-mail: [email protected] diameter of 0.5 μm and 1.8 μm in length, without spore- forming activity. Catalase and oxidase were also negative. The 16S rRNA gene was completely sequenced as previously As a part of our study of the human microbiome by culturomics described [4]. It shared 94.4% sequence identity with Intestini- [1], we isolated in the stool of a healthy 28-year-old French monas butyriciproducens DSM 26588T (NR_118554). The bac- donor the Gram-negative rod and strictly anaerobe strain terium was therefore putatively classified as a new species T GD2 . The written consent of the donor was obtained, and the belonging to the Intestinimonas genus. study was validated by the ethics committee of the Federative Because of the 16S identity percentage was lower than 98.65% Research Institute IFR48 under number 09-022. The stool was to the species closest with a validly published name standing in − stored at 20°C for 10 days, then inoculated on agar enriched nomenclature [5], we propose the new strain “Intestinimonas ’ fl fi T with sheep s blood (5%) and rumen uid (5%) previously lter massiliensis” GD2 (mas.il.i.en’sis, L. gen. masc. n. massiliensis, “of fi T sterilized through a 0.2 μm pore lter (Thermo Fisher Scien- Massilia,” the Latin name for Marseille, where the strain GD2 was fi ti c, Villebon sur Yvette, France). The plates were then incu- first isolated) belonging to the genus Intestinimonas (Fig. 1). bated under anaerobic condition into an anaerobic cabinet for 72 hours. The subculture of colonies using the same protocol MALDI-TOF MS spectrum accession number allowed the isolation of the GD2T strain. The strain GD2T could not be identified by matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry (MALDI-TOF MS) The MALDI-TOF MS spectrum of “Intestinimonas massiliensis” is screening (score <1.7) using a Microflex spectrometer (Bruker available at http://mediterranee-infection.com/article.php? Daltonics, Bremen, Germany) [1–3]. laref=256&titre=urms-database.

New Microbe and New Infect 2017; 15: 1–2 © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 210 http://dx.doi.org/10.1016/j.nmni.2016.09.014 2 New Microbes and New Infections, Volume 15 Number C, January 2017 NMNI

FIG. 1. Phylogenetic tree based on 16S rRNA gene sequence showing position of “Intestinimonas massiliensis” sp. nov., strain GD2T with other close relative species among Firmicutes phylum. European Molecular Biology Laboratory (EMBL) database accession numbers are indicated in parentheses. Sequences were aligned using CLUSTALW, and phylogenetic inferences were obtained with Kimura two-parameter model using neighbour-joining method with 1000 bootstrap replicates within MEGA6 software. Scale bar represents 1% nucleotide sequence divergence.

Nucleotide sequence accession number References

The 16S rRNA gene sequence was deposited in GenBank under [1] Lagier JC, Hugon P, KhelaifiaS,FournierPE,LaScolaB,RaoultD. accession number LN866996. The rebirth of culture in microbiology through the example of cul- turomics to study human gut microbiota. Clin Microbiol Rev 2015;28: 237–64. [2] Lagier JC, Armougom F, Million M, Hugon P, Pagnier I, Robert C, et al. Deposit in a culture collection Microbial culturomics: paradigm shift in the human gut microbiome study. Clin Microbiol Infect 2012;18:1185–93. [3] SengP,AbatC,RolainJM,ColsonP,LagierJC,GourietF,etal. T Strain GD2 was deposited in the collection de Souches de Identification of rare pathogenic bacteria in a clinical microbiology l’Unités des Rickettsies (CSUR, WDCM 875) under number laboratory: impact of matrix-assisted laser desorption ionization– time of flight mass spectrometry. J Clin Microbiol 2013;51: P1930. 2182–94. [4] Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 16S ribosomal DNA sequence analysis of a large collection of environmental Acknowledgement and clinical unidentifiable bacterial isolates. J Clin Microbiol 2000;38: 3623–30. [5] Kim M, Oh HS, Park SC, Chun J. Towards a taxonomic coherence This study was funded by the Fondation Méditerranée Infection. between average nucleotide identity and 16S rRNA gene sequence similarity for species demarcation of prokaryotes. Int J Syst Evol Microbiol 2014;64(Pt 2):346–51. Conflict of Interest

None declared.

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 15,1–2 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

211

Received: 21 November 2017 | Revised: 15 February 2018 | Accepted: 19 February 2018 DOI: 10.1002/mbo3.621

ORIGINAL RESEARCH

Noncontiguous finished genome sequence and description of Intestinimonas massiliensis sp. nov strain GD2T, the second Intestinimonas species cultured from the human gut

Pamela Afouda | Guillaume A. Durand | Jean-Christophe Lagier | Noémie Labas | Fréderic Cadoret | Nicholas Armstrong | Didier Raoult | Grégory Dubourg

Microbes, Evolution, Phylogeny and Infection, Aix-Marseille Université, UM Abstract 63, CNRS 7278, IRD 198, Inserm 1095, IHU Intestinimonas massiliensis sp. nov strain GD2T is a new species of the genus - Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13005 Marseille, France Intestinimonas (the second, following Intestinimonas butyriciproducens gen. nov., sp. nov). First isolated from the gut microbiota of a healthy subject of French origin using Correspondence Grégory Dubourg, IHU Méditerranée a culturomics approach combined with taxono-genomics, it is strictly anaerobic, Infection, Marseille, France. nonspore-forming, rod-shaped, with catalase- and oxidase- negative reactions. Its Email: [email protected] growth was observed after preincubation in an anaerobic blood culture enriched Funding information with sheep blood (5%) and rumen fluid (5%), incubated at 37°C. Its phenotypic and This study was supported by Méditerranée Infection and the National Research Agency genotypic descriptions are presented in this paper with a full annotation of its ge- under the program « Investissements d’ave- nome sequence. This genome consists of 3,104,261 bp in length and contains 3,074 nir », reference ANR-10-IAHU-03. predicted genes, including 3,012 protein-coding genes and 62 RNA-coding genes. Strain GD2T significantly produces butyrate and is frequently found among available 16S rRNA gene amplicon datasets, which leads consideration of Intestinimonas massil- iensis as an important human gut commensal.

KEYWORDS anaerobe, butyrate, culturomics, new species, taxono-genomics

1 | INTRODUCTION several limitations of these methods have been extensively dis- cussed (Poretsky, Rodriguez-R, Luo, Tsementzi, & Konstantinidis, The description of the human microbiome has become one the most 2014). Among these, 16S rRNA gene sequences may not match to exciting challenges of the 21st century in the field of microbiology, a corresponding species in the database, which can potentially lead as reflected by the Human Microbiome Project (HMP) (Turnbaugh to missed and unknown taxa of great interest. Recently, Lagier et al. et al., 2007). In particular, alterations in the composition of the human (2012, 2016) have shown that extensive bacterial culture, referred gut microbiota have been associated with several diseases, including to as culturomics, can fill in the blanks of metagenomic data through obesity and inflammatory bowel disease. More recently, specific the discovery of hundreds of new bacterial species associated with microbial signatures were predictive of the response to anticancer humans. therapy in lung cancer (Vétizou et al., 2015). While high- throughput Considering the limitations of the traditional combination of sequencing techniques have enabled substantial advances in under- phenotypic and genotypic characteristics to describe these new standing the role exerted by the gut microbiota in human health, species (Kim, Oh, Park, & Chun, 2014; Rosselló- Mora, 2006; Tindall,

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2018 The Authors. MicrobiologyOpen published by John Wiley & Sons Ltd.

MicrobiologyOpen. 2018;e621. 212 www.MicrobiologyOpen.com | 1 of 11 https://doi.org/10.1002/mbo3.621 2 of 11 | AFOUDA et Al.

Rosselló- Móra, Busse, Ludwig, & Kämpfer, 2010; Wayne et al., 1987), the sequences of different species (Thompson, Higgins, & Gibson, we also proposed using genomic information to help define and de- 1994) and the Kimura two- parameter model was used to calculate scribe new bacterial species (Fournier, Lagier, Dubourg, & Raoult, evolutionary distance (Kimura, 1980). 2015). We isolated a species belonging to the Intestinimonas genus as part of a culturomics study, using an anaerobic culture applied to a 2.3 | Physiological and phenotypic characteristics stool sample from a healthy subject. The genus Intestinimonas, which belongs to the Firmicutes phylum, was created in 2013. To date, this The strain was tested for growth in anaerobic conditions at varying genus contains only Intestinimonas butyriciproducens gen. nov, sp. nov, temperatures: 28°C, 37°C, 45°C, and 56°C. Growth under aerobic which was first isolated from mice (Kläring et al., 2013). It has also and microaerophilic conditions was also assessed. To determine the been cultured from the human gut (Bui et al., 2015). Furthermore, biochemical characteristics of the strain, API ZYM (bioMérieux), it has recently been abundantly detected in human colonic samples API Rapid ID 20 NE (bioMérieux), and API 50 CH (bioMérieux) (Bui et al., 2015), with a particular focus on butyrate production. In were used, following the instructions of the manufacturer. Catalase this paper, we present a summary of the classification and set of fea- and oxidase activities were also tested. Gram staining and motil- tures for Intestinimonas massiliensis sp. nov. strain GD2T, together ity were determined using the light microscope DM1000 (Leica with a description of its complete genomic sequencing and annota- Microsystems, Nanterre, France). Cell morphology was determined tion. These characteristics enable the creation of the Intestinimonas using Tecnai G20 transmission electron microscopy (FEI Company, massiliensis species, which represents the second Intestinimonas spe- Limeil- Brévannes, France), after negative staining of the bacteria cies and the first cultured from the human gut microbiota. and elements determining the gram- stain characteristics of the bacteria were evaluated using the Morgagni 268D TEM (Philips). For preparation for transmission electron microscopy (TEM), bac- 2 | MATERIAL AND METHODS teria were recovered and pelleted for 10 min at 5,000 g. The pel- let was resuspended in 1 ml of phosphate- buffered saline (PBS) 2.1 | Sample information with 2.5% glutaraldehyde in a 0.1mol/L sodium cacodylate buffer The specimen was sampled from a healthy 28- year- old male of and incubated for at least 1 hr at 4°C. The pellet was then washed French origin, with a body mass index of 23.4 kg/m2. Consent was three times with 0.1mol/L cacodylate-saccharose and resuspended obtained, and the study was approved by the Institut Fédératif de in the same buffer. After repelleting, the sample was embedded in Recherche 48 (Faculty of Medicine, Marseille, France), under agree- Epon resin using a standard method, as follows: 1 hr of fixation in ment Number 09- 022. 1% osmium tetroxide, two washes in distilled water, dehydration in increasing ethanol concentrations (30%, 50%, 70%, 96%, and 100% ethanol), and embedding in Epon-812. Ultrathin sections of 70 nm 2.2 | Strain identification and phylogenetic were poststained with 5% uranyl acetate and lead citrate following classification the Reynolds method (Reynolds, 1963). Samples were then observed Strain GD2T was isolated in February 2015 from a stool stored using a Morgagni 268D TEM (Philips) operating at 60 keV. To deter- 10 days at −20°C after preincubation 72 hr and subculture under mine sporulation, thermal shock was carried out on the bacteria at strict anaerobic conditions in the presence of sheep blood (5%) 80°C for 20 min, which were then seeded on Colombia blood agar. and rumen fluid (5%). Identification was performed using MALDI- Plates were then incubated for 48 hr under anaerobic conditions. TOF mass spectrometry and by sequencing of the 16S rRNA gene. We determined antibiotic susceptibility using the E- test gradient DNA extraction was realized using an EZ1 DNA Tissue Kit (Qiagen, strip method (bioMérieux) to define the minimal inhibitory concen- Courtaboeuf, France). The DNA extract was amplified using PCR tration (MIC) of each tested antibiotic. After culture of strain GD2T technology and universal primers FD1 and RP2 (Eurogentec, Angers, on 5% sheep blood- enriched Columbia agar (bioMérieux), the bacte- France). The amplifications and sequencing of the amplified prod- rial inoculum of 0.5 McFarland turbidity was prepared by suspend- ucts were performed as previously described (Dubourg et al., 2013). ing the culture in sterile saline (0.85% NaCl). Due to the inability of Afterward, 16S rRNA gene sequences were compared with those Intestinimonas massiliensis to grow on the medium recommended by available in GenBank (http://www.ncbi.nlm.nih.gov/genbank/). EUCAST (Citron, Ostovari, Karlsson, & Goldstein, 1991; Matuschek, When the percentage of identity of the entire 16S sequence was Brown, & Kahlmeter, 2014) (i.e., MH- F agar), the bacterial suspen- below the generally accepted threshold of 98.65%, the studied sion was swabbed on Columbia agar (bioMérieux). Then, each of the strain was considered as a new species (Kim et al., 2014). E-test strips (amoxicillin, ceftriaxone, ofloxacin, penicillin G, imipe- Phylogenetic analysis based on 16S rRNA of our isolate was nem, and vancomycin) were separately placed in culture plates and performed to identify its phylogenetic affiliations with other near incubated under anaerobic conditions for 72 hr. The test was done isolates, including other members of the genus Intestinimonas. The in duplicate and a quality control was done with the Escherichia coli MEGA 6 (Molecular Evolutionary Genetics Analysis) software en- strain DSM 1103. The MIC was determined by measuring the inter- abled us to build a phylogenetic tree (Tamura, Stecher, Peterson, section of the E- test strips with the elliptic zones of inhibition (Citron Filipski, & Kumar, 2013). The use of CLUSTALW permitted us to align et al., 1991).

213 AFOUDA et Al. | 3 of 11

chromatograms. Coefficients of determination were all above 0.999. 2.4 | Fatty acid methyl ester analysis Back- calculated standards and calculated quality controls (0.5 and Cellular fatty acid methyl ester (FAME) analyses of Intestinimonas 5 mmol/L) all showed good accuracy, with deviations below 15%. massiliensis strain GD2T (=CSUR P1930) and Intestinimonas butyric- SCFA quantities in samples were presented after subtraction of the iproducens (=CSUR P1453 = DSM 103501) were performed using quantities measured in the blank samples. GC/MS. Two bacterial biomass sample tubes of about 4 mg each obtained from cultures plates were prepared after 72 hr of culture 2.6 | Genomic sequencing of the bacteria on 5% sheep blood- enriched Columbia agar (bioMé- rieux) in anaerobic conditions. Then, fatty acid methyl esters were We used MiSeq Technology (Illumina Inc, San Diego, CA, USA) to se- prepared according to the description of Sasser (2006). GC/MS quence genome DNA (gDNA) of the Intestinimonas massiliensis strain analyses were carried out as previously stated (Dione et al., 2016). GD2T, along with the mate pair strategy by Nextera Mate Pair sam- Mass spectrometry (Clarus 500 - SQ 8 S, Perkin Elmer, Courtaboeuf, ple prep kit (Illumina), as previously described (Lagier et al., 2014). France) allowed us to separate fatty acid methyl esters by utiliza- Using a Qubit assay with broad range kit (Life Technologies, tion of an Elite 5- MS column. Utilization of the Standard Reference Carlsbad, CA, USA) allowed to us to quantify genomic DNA to Database 1A (NIST, Gaithersburg, USA) and the FAMEs mass spec- 137 ng/μl. Then, we prepared a mate pair library with 1.5 μg of tral database (Wiley, Chichester, UK), permitted us to search a spec- gDNA using the Nextera mate pair Illumina guideline as per man- tral database with MS Search 2.0. ufacturer’s instructions. Afterward, we simultaneously splintered and tagged the gDNA sample with a mate pair junction adapter. Subsequently, to validate the splitting pattern, we used a DNA 7500 2.5 | Short- chain fatty acids analysis LabChip on the Agilent 2100 Bioanalyzer (Agilent Technologies Inc, Short- chain fatty acids (SCFA) were measured with a Clarus 500 Santa Clara, CA, USA). The fragments obtained had the required size chromatography system connected to a SQ8s mass spectrometer of 6.01 kb. No size selection was performed, and tagged fragments (Perkin Elmer, Courtaboeuf, France), as previously detailed (Zhao, 428.4 ng were circularized. Next, small fragments were obtained by Nyman, & Åke, 2006), with modifications. As a prelude to this, mechanical shearing from the circularized DNA on the Covaris de- 500 μg of bacterial suspension were placed in Lytic/10 anaerobic/F vice S2 in T6 tubes (Covaris, Woburn, MA, USA). The optimal size (BD ™ Bactec ™ Media) medium and incubated at 37°C for 72 hr. of these small fragments was 950 bp. Following visualization of the Acetic, propanoic, isobutanoic, butanoic, isopentanoic, pentanoic, library profile on the High Sensitivity Bioanalyzer LabChip (Agilent hexanoic, and heptanoic acids were purchased from Sigma Aldrich Technologies Inc, Santa Clara, CA, USA), the final concentration li- (Lyon, France). A stock solution was prepared in water/methanol brary obtained was 4.593 nmol/L. (50% v/v) at a final concentration of 50 mmol/L and then stored This library was then combined with the other 11 projects and at −20°C. Calibration standards were freshly prepared in acidified finally normalized to 2 nmol/L, which was further denatured and di- water (pH 2- 3 with HCl 37%) from the stock solution at the follow- luted to 15 pM. The automated cluster was generated after loading ing concentrations: 0.5; 1; 5; 10 mmol/L. SCFA were analyzed from in the reactant cartridge along with the flow cell instrument, and a three independent culture bottles (both blanks and samples). Culture 39- hour long sequencing run was carried out. medium was collected, then centrifuged for 5 min at 16,000 g to re- With a cluster density of 653 K/mm2, the information acquired move bacteria and debris. The clear supernatant was adjusted to pH represented a total of 6.1 Gb; this contains a group pass quality con- 2-3 and spiked with 2-ethylbutyric acid as the internal standard (IS) trol filter estimated at 96.1% (12,031,000 pairs of pass filters). The at a final concentration of 1 mmol/L (Sigma Aldrich). The solution index representation of the Intestinimonas massiliensis, correspond- was once again centrifuged before injection. Aqueous samples were ing to the proportion of reads attributed to this project among the directly injected (0.5 μl) in a splitless liner heated at 200°C. The in- total number of number of reads-, was of 8.06%. The 1,208,418 jection carry-over was decreased with 10 syringe washes in water/ paired reads were trimmed and afterward assembled into seven methanol (50:50 v/v). Compounds were then separated on an Elite- scaffolds. FFAP column (30 m, 0.25 mm id 0.25 mm film thickness) using a lin- ear temperature gradient from 100°C to 200°C at 8°C/min. Helium 2.7 | Genome annotation and comparison at a flow rate of 1 mL/min was utilized. The MS inlet line and electron ionization source were set at 200°C. To insure compound selectivity, We predicted open reading frames (ORFs) utilizing Prodigal with selected ion recording (SIR) was performed after a 4.5 min solvent default settings (http://prodigal.ornl.gov/) (Hyatt et al., 2010). All delay with the following masses: 43 m/z (isobutanoic acid), 60 m/z predicted ORFs not covering a region of the standard sequence (acetic, butanoic, pentanoic, isopentanoic, hexanoic, and hepta- were excluded. We searched predicted bacterial protein sequences noic acids) 74 m/z (propanoic acid), 88 m/z (2-ethylbutyric acid, IS). against GenBank and Clusters of Orthologous Groups (COG) All data were collected and processed using TurboMass 6.1 (Perkin (Benson et al., 2012) using BLASTP. We then used the tRNAScan- SE, Elmer, Courtaboeuf, France). Quadratic internal calibration was cal- RNAmmer tools (Lagesen et al., 2007; Lowe & Eddy, 1997), SignalP culated for each acid using the peak areas from the associated SIR and TMHMM, (Bendtsen, Nielsen, Heijne, & Brunak, 2004; Krogh,

214 4 of 11 | AFOUDA et Al.

Larsson, von Heijne, & Sonnhammer, 2001) for prediction of tRNAs, which only one sequence was detected were not considered as posi- rRNAs, signal peptides, and numbers of transmembrane helices, re- tive. To assess relative abundance of these two species in the human spectively. PHAST and RAST were used to predict mobile genetic gut, only datasets labeled “human gut metagenome” were consid- elements (Aziz et al., 2008; Zhou, Liang, Lynch, Dennis, & Wishart, ered. Number of sequences attributed to the bacteria was divided 2011). Identification of ORF without homologues in other lineages by the sample size to estimate relative abundance. (ORFans) depended on parameter thresholds of their BLASTP E- value. So, identification was possible for BLASTP E- values lower 3 | RESULTS than 1e−03 for an alignment length greater than 80 amino acids, but if they were smaller than 80 amino acids we used an E-value of 1e−5. 3.1 | Phylogenetic classification Data management, visualization of genomic characteristics, and multiple genomic sequence alignment were performed by utilization Intestinimonas massiliensis strain GD2T was first isolated in February of the Artemis DNAPlotter (Carver, Thomson, Bleasby, Berriman, & 2015 on agar enriched with sheep blood (5%) and rumen fluid (5%) Parkhill, 2009; Rutherford et al., 2000) and alignment tools (version at 37°C under anaerobic conditions (Table 1). The MALDI- TOF MS 2.3.1), respectively (Darling, Mau, Blattner, & Perna, 2004). spectrum was subsequently added to our database (Figure S1). We took the complete sequence of the genome, the genome The gel view highlights marked spectral differences with other sequence of the proteome and genome sequence of the ORFeome members of the Firmicutes phylum (Figure S2), in particular with from the FTP of NCBI. Proteomes were analyzed using Proteinortho the Intestinimonas butyriciproducens spectrum.Identification of our (Lechner et al., 2011). The average similarity of orthologous pro- strain by MALDI- TOF yielded no reliable identification despite reg- teins was evaluated using the Average Genomic Identity Of gene ular database updates. Strain GD2T exhibited a 94.96% 16S rRNA Sequences (AGIOS) software (Ramasamy et al., 2014). This allowed sequence identity with the type strain Intestinimonas butyricip- us to compare the pairwise orthologous proteins in combination with roducens S R B - 5 2 1 - 5 - I T (GenBank accession number KC311367), the Proteinortho software (Lechner et al., 2011). The corresponding the phylogenetically closest bacterial species with standing in the genes were recovered and the percentage nucleotide identity among nomenclature (Figure 1). Its 16S rRNA sequence was deposited in ORF orthologs was calculated using the Needleman–Wunsch global GenBank under number LN866996. This value was lower than the alignment algorithm. Finally, the Multi- Agent Software System 98.65% 16S rRNA gene sequence threshold recommended by Kim DAGOBAH was used to achieve all annotation and comparison pro- et al. (2014) to delineate a new species without carrying out DNA– cesses (Gouret et al., 2011), including Figenix libraries that provide DNA hybridization. pipeline analysis (Gouret et al., 2005). The 16S rRNA sequence of Intestinimonas massiliensis strain GD2T 3.2 | Phenotypic description was compared to those of other close species belonging to the Firmicutes phylum, such as Intestinimonas butyriciproducens, Pseudoflavonifractor The growth of Intestinimonas massiliensis strain GD2T was ob- capillosus, Oscillibacter valericigenes, Flavonifractor plautii, Clostridium served at 37°C after 72 hr of incubation in anaerobic conditions, cellulosi, Clostridium viride, Ethanoligenens harbinense, Clostridium lep- whereas no growth was observed at 28°C, 45°C, and 56°C. No tum, and Eubacterium siraeum. growth occurred under aerobic conditions. The Intestinimonas We performed the gel view for protein profile comparisons for massiliensis strain GD2T is thus strictly anaerobic and grows up Intestinimonas massiliensis strain GD2T with the following Firmicutes to 37°C. Its pH range for growth was 6-8.5 and it tolerated NaCl species: Intestinimonas butyriciproducens, Flavonifractor plautii, Clostridium papyrosolvens, and Clostridium cellobioparum. TABLE 1 Classification and General Features of Intestinimonas massiliensis strains GD2T

2.8 | Frequency and relative Property Term abundance of Intestinimonas species among 16S Current classification Domain: Bacteria rRNA sequence databases Phylum: Firmicutes Class: Clostridia To investigate the relative abundance and frequency of I. massiliensis Order: Clostridiales and I. butyriciproducens we used the IMNGS open resource platform Family: unclassified clostridiales Genus: Intestinimonas that provides a research of abundance of our 16S rRNA sequence into Species: Intestinimonas massiliensis 16S rRNA gene amplicon datasets from the Sequence Read Archive Type strain: strain GD2T (Lagkouvardos et al., 2016). For this purpose, the entire sequence Cell shape Rod of I. massiliensis (Genbank accession number LN866996) and that Temperature range Mesophilic of I. butyriciproducens (Genbank accession number KC311367) was Optimum temperature 37°C used for search with a similarity threshold of 99% and a minimum pH 6-8.5 size of 200 bp. Results were then manually filtered, and frequency Salinity 0-5 g/L was calculated according to the origin of the sample. Samples for

215 AFOUDA et Al. | 5 of 11

FIGURE 1 Phylogenetic tree highlighting the position of Intestinimonas massiliensis strain GD2 T relative to other Firmicutes. Numbers at the nodes are percentages of bootstrap values obtained by repeating the analysis 1,000 times to generate a majority consensus tree. The scale bar represents a 5% nucleotide sequence divergence

concentrations ranging from 0 to 5 g/L. Cells were immotile and nonsporulating. Colonies were regular, white, with a mean diam- eter of 1–2 mm on sheep blood- enriched Colombia agar. Gram staining (Figure 2) showed gram- negative rods. Using electron mi- croscopy, the rods had a mean diameter of 0.5 μm and a length of 1.8 μm (Figure S3). Catalase and oxidase activities were negative for Intestinimonas massiliensis strain GD2T. Using API ZYM, positive reactions were observed for naphthol- AS- BI- phosphohydrolase and acid phosphatase. Negative reactions were observed for alka- line phosphatase, esterase (C4), esterase lipase (C8), lipase (C14), leucine arylamidase, valine arylamidase, trypsin, α- chymotrypsin, β- galactosidase, N- acetyl- β- glucosaminidase, α- galactosidase, β- glucuronidase, α- glucosidase, β- glucosidase, α- fucosidase, and α-mannosidase. An API 50 CH strip showed positive fermentation reactions for Gram staining of the Intestinimonas massiliensis strain D-arabinose, D-ribose, D-xylose, L-xylose, D-galactose, L-sorbose, FIGURE 2 GD2T amygdalin, esculin ferric citrate, D-melibiose, D-trehalose, inulin, D-melezitose, D-raffinose, starch, glycogen, xylitol, gentiobiose, D-lyxose, D-tagatose, D-fucose, and potassium 5-ketogluco nate, negative assimilation reactions for L- arabinose, D- mannose, D- but a faint positive reaction was observed for D-fructose. Negative mannitol, N- acetylglucosamine, D-maltose, potassium gluconate, fermentation reactions were recorded for glycerol, erythritol, L- capric acid, adipic acid, malic acid, trisodium citrate, and pheny- arabinose, D- adonitol, methyl-β D-xylopyranoside, D-glucose, D- lacetic acid. mannose, L-rhamnose, dulcitol, inositol, D-mannitol, D-sorbito l, When compared with its phylogenetically closest neighbor (i.e., methyl- αD-mannopyranoside, methyl-α D-glucopyranoside, N- Intestinimonas butyriciproducens s t r a i n S R B - 5 2 1 - 5 - I T), Intestinimonas acetyl- glucosamine, arbutin, salicin, D- cellobiose, D- maltose, D- massiliensis strain CD2T differed in endospore formation, nitrate re- lactose, D- sucrose, D- turanose, L- fucose, D- arabitol, L- arabitol, ductase, fermentation of L- arabinose, and D- glucose (Table 2). potassium gluconate, and potassium 2-ketogluconate. Using API MICs for the GD2T strain were distributed as follows: vanco- 20 NE demonstrated a positive reaction for gelatin hydrolysis, but mycin (MIC 0.50 μg/ml), penicillin G (MIC 0.19 μg/ml), imipenem negative reactions for β- galactosidase, potassium nitrate (nitrate (MIC 0.25 μg/ml), ceftriaxone (MIC 1 μg/ml), and amoxicillin (MIC reductase), L-tryptophan (indole formation), D-glucose (fermenta- 0.125 μg/ml). A high level of resistance to ofloxacin was observed tion and assimilation), L- arginine, urease, esculin ferric citrate, and (MIC > 32 μg/ml).

216 6 of 11 | AFOUDA et Al.

TABLE 2 Differential characteristics of Intestinimonas massiliensis strain GD2T (1) and other strains: Intestinimonas butyriciproducens S R B - 5 2 1 - 5 - I T (2); Flavonifractor plautii DSM 4000T (3); Pseudoflavonifractor capillosus CCUG 15402AT (4); and Clostridium cellulosi AS 1.1777 (5). +, positive; - , negative; na, not available; V, variable; No., number

Properties 1 2 3 4 5

Cell diameter (μm) 1.8 × 0.5 2- 5 2-7 na 0.3-0.6 Oxygen requirement strictly anaerobic strictly anaerobic strictly anaerobic strictly anaerobic strictly anaerobic Gram stain − + V − − Motility − − V − + Endospore formation − + V − + Indole − Na na − - Major Fatty acids FAME 16: 0; 18: 1n9 14: 0; 12: 0 16: 0; 14: 0 16: 0; 14: 0 na DNA G+C content (mol %) 60.68 58.4 61.6 60 35 Genome size (bp) 3,104,261 3,376,475 3,818,478 4,241,076 5,680,000 Gene content (No.) 3,074 3,529 4,278 4,829 5,171 Production of Catalase − Na na − − Oxidase − Na na na na Nitrate reductase − + na na − Urease − − na na na β-galactosidase − na na + na N- acetyl- glucosamine − na na na na Acid from L-arabinose − + + na − D-mannose − na na − + D-mannitol − − − + + D- glucose − + − + + D-fructose +/− − − + + D- maltose − − − + + D- lactose − na na na + Habitat Human gut Mouse gut Human gut Human and animal Cow manure Human gut gut compost References This study (Bui et al., 2015; (Carlier, Bedora- Faure, (Kläring et al., 2013; (He, Ding, & Long, Kläring et al., 2013), K’ouas, Alauzet, & Madsen & Justesen, 1991), This study This study Mory, 2010; Kläring 2011), This study et al., 2013), This study

(6.4 ± 0.7 mmol/L) and minor production of acetic (0.7 ± 0.1 mmol/L), 3.3 | Fatty acid methyl ester analysis propanoic (0.4 ± 0.4 mmol/L), and pentanoic (0.1 ± 0.1 mmol/L) Cellular fatty acid composition showed that the two most abundant acids. Isobutanoic, isopentanoic, hexanoic, and heptanoic acids were fatty acids are unsaturated 9- octadecenoic acid (35%) and saturated not produced. hexadecanoic acid (30%) (Table 3). Table 3 also demonstrates the comparison of cellular fatty acid composition (%) of Intestinimonas 3.5 | Genome properties massiliensis strain GD2T with Intestinimonas butyriciproducens CSUR P1453- DSM 103501; a significant difference is observed with 1 tet- The genome of Intestinimonas massiliensis strain GD2T is radecanoic acid, 2- methyl- tridecanoic acid, and hexadecanoic acid. 3,104,261 bp long with 60.66% GC content. This noncontiguous finished genome is composed of seven scaffolds accounting for nine contigs. Of the 3,074 predicted genes, 3,012 were protein- 3.4 | Short-chain fatty acids analysis coding genes and 62 were RNAs (two genes were 5S rRNA, two Production of SCFA by Intestinimonas massiliensis strain GD2T genes were 16S rRNA, two genes were 23S rRNA, and 56 genes was positively detected, with a major production of butanoic acid were TRNA genes). A total of 1,933 genes (64.18%) were assigned

217 AFOUDA et Al. | 7 of 11

TABLE 3 Cellular fatty acid methyl ester composition (%) of TABLE 4 Nucleotide content and gene count levels of the Intestinimonas massiliensis strain GD2T genome

Mean Genome (total) Fatty acids IUPAC name Relative %a Attribute Value % of totala 12:0 Dodecanoic acid TR Size (bp) 3,104,261 100 13:0 Tridecanoic acid TR G+C content (%) 1,882,912 60.66 14:0 Tetradecanoic acid 4.8 ± 1.1 Coding region (bp) 2,769,278 89.21 14:0 iso 12-methyl- Tridecanoic acid TR Total genes 3,074 100 15:0 Pentadecanoic acid 1.2 ± 0.1 RNA genes 62 2.02 15:0 iso 13-methyl- tetradecanoic acid TR Protein-coding genes 3,012 100 15:0 anteiso 12-methyl- tetradecanoic acid TR Number of proteins associated with 1,933 64.18 16:0 Hexadecanoic acid 30.3 ± 5.5 function prediction (nr+cogs not [S]) 16:0 TR 9,10- methylene- Hexanoic acid Number of proteins associated with 763 25.33 9,10- methylene hypothetical protein 16:1n7 3.3 ± 0.2 9- Hexadecenoic acid Genes with function prediction 413 13.71 17:0 Heptadecanoic acid 1.7 ± 1.4 Genes assigned to COGs 134 4.45 17:1n7 10-Heptadecenoic acid 1.4 ± 0.4 Genes with peptide signals 375 12.45 17:0 anteiso 14-methyl- Hexadecanoic acid 1.7 ± 0.3 Gene associated with resistance 1 0.03 18:0 Octadecanoic acid 7.7 ± 1.1 genes 18:1n9 9- Octadecenoic acid 34.6 ± 1.6 Gene associated with bacteriocin 22 0.73 18:2n6 9,12- Octadecadienoic acid 11.5 ± 2.8 genes Proteins associated with ORFans 182 6.04 aMean peak area percentage ± standard deviation; TR: trace amounts < 1. Genes associated with PKS or NRPS 9 0.29

aThe total is based on either the size of the genome in base pairs or the a putative function (by cogs or by NR blast) and the ORFans are total number of protein- coding genes in the annotated genome. represented by 182 genes (6.04%). Finally, the rest of the genes were considered and annotated as hypothetical proteins (763 genes, 25.33%) (Table 4). Tables 5 and Table S1 summarize the correlation with DDH (Auch, von Jan, Klenk, & Göker, 2010; Meier- properties and statistics of the genome. Figure S4 shows a graph- Kolthoff, Auch, Klenk, & Göker, 2013) and AGIOS (Ramasamy et al., ical circular map of the genome and Figure 3 shows the distribu- 2014) that was designed to be independent from DDH. When consid- tion of functional classes of predicted genes on the chromosomes ering only the closest species with standing in nomenclature for which of strain CD2T. Intestinimonas massiliensis and its closest species a genome is available, dDDH values ranged from 17.70 ± 2.25 be- seem for the most part associated with the same Clusters of tween Flavonifractor plautii and Sporobacter termitidis to 29.50 ± 2.45 Orthologous Groups (COG) genes. Nevertheless, COGs functional between Intestinimonas butyriciproducens and Butyricicoccus pulli- categories “Inorganic ion transport and metabolism, carbohy- caecorum. When we include the strain GD2T in the comparison, the drate transport and metabolism, posttranslational modification, dDDH values ranged from 18.40 ± 2.52 with Sporobacter termitidis to protein turnover, chaperones, cell wall/membrane biogenesis” 28.10 ± 2.40 with Butyricicoccus pullicaecorum (Table 5). are more represented in Clostridium inocuum, while that concern- Regarding AGIOS, values ranged from 52.31 between ing “carbohydrate transport and metabolism” are more present Pseudoflavonifractor capillosus and Clostridium cellulosi to 73.57% in Clostridium leptum. In addition, Pseudoflafonifractor capillosus between Flavonifractor plautii and Intestinimonas butyriciprodu- is enriched in genes belonging to COGs functional categories cens among compared species. Including Intestinimonas massilien- concerning “inorganic ion transport and metabolism, posttrans- sis, AGIOS ranged from 57.10 with Ethanoligenens harbinense to lational modification, protein turnover, chaperones,” whereas 76.46% with Intestinimonas butyriciproducens (Table S2). As the ob- COGs functional categories concerning “inorganic ion transport tained dDDH values were lower than 70%, and because dDDH and and metabolism” are more represented in Flavonifractor plautii. AGIOS values were close to the range of those obtained among compared species with standing in nomenclature, and because of the production of butyrate and acetate, and finally because 3.6 | Comparison of genome properties the difference of G+C content with other Intestinimonas species The genome size, the G+C content, and the gene content of I. massil- was greater than 1 with Intestinimonas butyriciproducens (Table 2) iensis and among the closest species are summarized in Table 2. (Meier- Kolthoff, Klenk, & Göker, 2014), we are confident that In order to evaluate the genomic similarity among studied strains, strain GD2T is the representative strain of a new species within the we used two parameters: digital DDH (dDDH) that exhibits a high genus Intestinimonas.

218 8 of 11 | AFOUDA et Al.

TABLE 5 Pairwise comparison of Intestinimonas massiliensis GD2T with other species using GGDC, formula 2 (DDH estimates based on identities/HSP length)* upper right. (1) Intestinimonas massiliensis GD2T; (2) Pseudoflavonifractor capillosus strain ATCC 29799; (3) Flavonifractor plautii strain Prevot S1; (4) Intestinimonas butyriciproducens strain SRB- 521- 5- I; (5) Clostridium viride strain T2- 7; (6) Oscillibacter valericigenes strain Sjm18- 20; (7) Sporobacter termitidis strain SYR; (8) Oscillibacter ruminantium strain GH1; (9) Butyricicoccus pullicaecorum strain 25-3

1 2 3 4 5 6 7 8 9

1 100% 22.70 ± 2.4 21.70 ± 2.35 21.50 ± 2.35 26.40 ± 2.45 20.30 ± 2.30 18.40 ± 2.52 19.70 ± 2.30 28.10 ± 2.40 2 100% 22.20 ± 2.10 22.10 ± 2.35 20.50 ± 2.30 19.50 ± 2.30 19.40 ± 2.30 19.40 ± 2.30 25.00 ± 2.40 3 100% 22.00 ± 2.35 21.00 ± 2.35 21.10 ± 2.35 17.70 ± 2.25 19.00 ± 2.30 25.60 ± 2.40 4 100% 23.30 ± 2.40 20.40 ± 2.35 19.60 ± 2.30 20.30 ± 2.30 29.50 ± 2.45 5 100% 24.30 ± 2.40 24.20 ± 2.40 21.80 ± 2.35 24.60 ± 2.40 6 100% 22.20 ± 2.30 25.30 ± 2.40 26.90 ± 2.45 7 100% 26.10 ± 2.40 29.00 ± 2.40 8 100% 25.40 ± 2.45 9 100%

*Confidence intervals indicate inherent uncertainty in estimating DDH values from intergenomic distances based on models derived from empirical test data sets (which are always limited in size).

FIGURE 3 Distribution of functional classes of predicted genes on the chromosomes of strain CD2T and related taxa Clostridium cellulosi, Clostridium leptum, Clostridium viride, Ethanoligenens harbinense, Eubacterium siraeum, Flavonifractor plautii, Intestinimonas butyriciproducens, Intestinimonas massiliensis, Oscillibacter valericigenes, Pseudoflavonifractor capillosus, according to the clusters of orthologous groups of protein

219 AFOUDA et Al. | 9 of 11

phylogenetic characteristics, we have isolated a new species, named 3.7 | Frequency and relative abundance of Intestinimonas massiliensis sp. nov strain GD2T, isolated for the first I. massiliensis and I. butyriciproducens time in the human gut microbiota. The 16S rRNA gene sequence With a similarity threshold of 99%, I. massiliensis was detected in 4.40% and whole- genome shotgun sequence of Intestinimonas massilien- of all datasets, mainly involving the human gut, but also in animals sis strain GD2T has been deposited in GenBank with the accession and the environment. In comparison, I. butyriciproducens was present Number LN866996. in 1.98% of all available 16S rRNA amplicon datasets. Interestingly, I. massiliensis was detected more frequently than I. butyriciproducens 6 | DESCRIPTION OF INTESTINIMONAS in the human gut, as they were present in 19.8% of and in 8.1% of the MASSILIENSIS SP. NOV STRAIN GD2 T (= 16,950 datasets, respectively (Chi- squared test <10−7) (Table S3). The CSUR P1930, = DSM100417) mean relative abundances from these datasets were of 0.079% and 0.087% for I. massiliensis and I. butyriciproducens, respectively. 6.1 | Intestinimonas massiliensis (mas.si.li.en’sis. L. masc. adj. massiliensis of Massilia, the ancient Roman name for Marseille, where the strain was isolated) 4 | DISCUSSION Strictly anaerobic, gram- negative, oxidase and catalase negative, Herein, we describe a new species belonging to the genus nonendospore forming, and nonmotile rods, the colonies are circu- Intestinimonas. The strain GD2T was isolated for the first time in the lar, small, and glossy with a diameter of approximately 0.5–1 mm on stool of a healthy 28- year- old French male using a “culturomics” ap- Columbia agar + 5% sheep blood. Growth was noticed at 37°C after proach. Based on different biochemical, phylogenetic, and genomic 3–4 days of incubation and with a pH between 6 and 8.5. Cells meas- properties when compared with the phylogenetically closest species ure about 1–1.5 μm in length and 0.5 μm in diameter. (i.e., Intestinimonas butyriciproducens S R B - 5 2 1 - 5 - I T) (Fournier et al., Using API 50 CH and ZYM strips, positive reactions were 2015), we proposed the creation of the second bacterial species, observed for: arabinose, D-ribose, D-xylose, L-xylose, D- strain GD2T, belonging to the genus Intestinimonas. galactose, D-fructose, L- sorbose, amygdalin, esculin fer- Like Kläring et al. (2013) with I. butyriciproducens, we experi- ric citrate, D- melibiose, D-trehalose, inulin, D-melez itose, enced difficulties in determining if the strain GD2T was gram pos- D- raffinose, starch, glycogen, xylitol, gentiobiose, D- lyxose, itive or negative. Indeed, gram staining combined with optical D- tagatose, D- fucose, potassium 5-ketogluconate, naphthol- microscopy revealed the presence of gram- negative bacilli. In addi- AS- BI- phosphohydrolase, and acid phosphatase. The API 20 tion, the susceptibility to vancomycin as well as its classification with NE strip showed a positive reaction for gelatin hydrolysis and gram- positive microbes led us to assume that I. massiliensis should negative reaction for other biochemical tests. Intestinimonas be considered as a gram-positive microorganism, according to the massiliensis sp. nov strain GD2T is susceptible to amoxicillin, cef- genus formal description (Kläring et al., 2013). However, we did not triaxone, penicillin G, imipenem, and vancomycin. With regard observe by transmission electron microscopy (in ultrathin sections to fatty acids, an abundance of unsaturated 9- octadecenoic of resin- embedded cells) a clear membrane arrangement of the cells acid (35%) and saturated hexadecanoic acid (30%) was ob- resembling a gram+ ultrastructure (Figure S5). served. This bacterium produces acetic (0.7 ± 0.1 mmol/L), Intestinimonas massiliensis significantly produces butyrate, which propanoic (0.4 ± 0.4 mmol/L), butanoic (6.4 ± 0.7 mmol/L), and is an SCFA of potential medical importance. Butyrate is known to pentanoic (0.1 ± 0.1 mmol/L) acids. be an energy source for epithelial cells and plays a key role in main- The G+C content of the genome is 60.68%. Accession numbers taining homeostasis of colonic cells. In addition, several works have of the sequences of 16S rRNA and genome deposited in EMBL- EBI shown its inhibiting role in inflammation and oxidative stress (Hamer are LN866996 and CWJP00000000, respectively. The microorgan- et al., 2008), whereas its contribution to improving insulin sensitivity ism was isolated within the human gut microbiota. The type strain and glucose homeostasis has been reported, as with other SCFAs GD2T (= CSUR P1930 = DSM100417) was isolated from a stool (Canfora, Jocken, & Blaak, 2015). specimen of a healthy 28- year- old French male. Also, being detected more frequently in 16S rRNA amplicon datasets than Intestinimonas butyriciproducens, Intestinimonas mas- ACKNOWLEDGMENTS siliensis appears to be a common human gut commensal that may contribute to the gut microbiota homeostasis. The authors thank the Xegen Company (www.xegen.fr) for automat- ing the genomic annotation process and Jean- Pierre BAUDOIN for the work carried out in electron microscopy. 5 | CONCLUSION

CONFLICT OF INTEREST With a similarity level of 94.96% to the strain Intestinimonas butyric- iproducens gen. nov., sp. nov and based on phenotypic, genomic, and The authors declare no financial conflict of interest.

220 10 of 11 | AFOUDA et Al.

ORCID prokaryotes in clinical microbiology. Anaerobe, 36, 73–78. https:// doi.org/10.1016/j.anaerobe.2015.10.011 Jean-Christophe Lagier http://orcid.org/0000-0003-0771-8245 Gouret, P., Paganini, J., Dainat, J., Louati, D., Darbo, E., Pontarotti, P., & Levasseur, A. (2011). Integration of evolutionary biology concepts for Fréderic Cadoret http://orcid.org/0000-0002-9553-8959 functional annotation and automation of complex research in evolu- Nicholas Armstrong http://orcid.org/0000-0002-8365-2244 tion: The multi-agent software system DAGOBAH. In Pontarotti, P., (Eds.), Evolutionary biology – concepts, biodiversity, macroevolution and Grégory Dubourg http://orcid.org/0000-0002-5772-7628 genome evolution [Internet]. (pp. 71–87). Berlin, Heidelberg: Springer [cited 2017 Oct 11]. Available from: https://link.springer.com/ chapter/10.1007/978-3-642-20763-1_5 REFERENCES Gouret, P., Vitiello, V., Balandraud, N., Gilles, A., Pontarotti, P., & Danchin, E. G. J. (2005). FIGENIX: Intelligent automation of genomic Auch, A. F., von Jan, M., Klenk, H.-P., & Göker, M. (2010). Digital DNA- annotation: Expertise integration in a new software platform. BMC DNA hybridization for microbial species delineation by means of Bioinformatics, 5(6), 198. https://doi.org/10.1186/1471-2105-6-198 genome-to- genome sequence comparison. Standards in Genomic Hamer, H. M., Jonkers, D., Venema, K., Vanhoutvin, S., Troost, F. J., & Sciences, 2(1), 117–134. https://doi.org/10.4056/sigs.531120 Brummer, R.-J. (2008). Review article: The role of butyrate on colonic Aziz, R. K., Bartels, D., Best, A. A., DeJongh, M., Disz, T., Edwards, R. function. Alimentary Pharmacology & Therapeutics, 27(2), 104–119. A., … Meyer, F. (2008). The RAST server: Rapid annotations using He, Y. L., Ding, Y. F., & Long, Y. Q. (1991). Two cellulolytic Clostridium spe- subsystems technology. BMC Genomics., 8(9), 75. https://doi. cies: Clostridium cellulosi sp. nov. and Clostridium cellulofermentans sp. org/10.1186/1471-2164-9-75 nov. International Journal of Systematic Bacteriology, 41(2), 306–309. Bendtsen, J. D., Nielsen, H., vonHeijne, G., & Brunak, S. (2004). Improved Hyatt, D., Chen, G.-L., Locascio, P. F., Land, M. L., Larimer, F. W., & Hauser, prediction of signal peptides: SignalP 3.0. Journal of Molecular Biology, L. J. (2010). Prodigal: Prokaryotic gene recognition and translation 340(4), 783–795. https://doi.org/10.1016/j.jmb.2004.05.028 initiation site identification. BMC Bioinformatics, 8(11), 119. https:// Benson, D. A., Karsch-Mizrachi, I., Clark, K., Lipman, D. J., Ostell, J., & doi.org/10.1186/1471-2105-11-119 Sayers, E. W. (2012). GenBank. Nucleic Acids Research, 40(Database Kim, M., Oh, H.-S., Park, S.-C., & Chun, J. (2014). Towards a taxonomic issue), D48–D53. https://doi.org/10.1093/nar/gkr1202 coherence between average nucleotide identity and 16S rRNA Bui, T. P. N., Ritari, J., Boeren, S., de Waard, P., Plugge, C. M., & de gene sequence similarity for species demarcation of prokaryotes. Vos, W. M. (2015). Production of butyrate from lysine and the International Journal of Systematic and Evolutionary Microbiology, Amadori product fructoselysine by a human gut commensal. Nature 64(Pt 2), 346–351. https://doi.org/10.1099/ijs.0.059774-0 Communications, 1(6), 651. Kimura, M. (1980). A simple method for estimating evolutionary rates Canfora, E. E., Jocken, J. W., & Blaak, E. E. (2015). Short- chain fatty of base substitutions through comparative studies of nucleotide se- acids in control of body weight and insulin sensitivity. Nature quences. Journal of Molecular Evolution, 16(2), 111–120. https://doi. Reviews Endocrinology, 11(10), 577–591. https://doi.org/10.1038/ org/10.1007/BF01731581 nrendo.2015.128 Kläring, K., Hanske, L., Bui, N., Charrier, C., Blaut, M., Haller, D., … Carlier, J.-P., Bedora-Faure, M., K’ouas, G., Alauzet, C., & Mory, F. (2010). Clavel, T. (2013). Intestinimonas butyriciproducens gen. nov., sp. Proposal to unify Clostridium orbiscindens Winter et al. 1991 and nov., a butyrate- producing bacterium from the mouse intestine. Eubacterium plautii (Séguin 1928) Hofstad and Aasjord 1982, with de- International Journal of Systematic and Evolutionary Microbiology, scription of Flavonifractor plautii gen. nov., comb. nov., and reassign- 63(Pt 12), 4606–4612. https://doi.org/10.1099/ijs.0.051441-0 ment of Bacteroides capillosus to Pseudoflavonifractor capillosus gen. Krogh, A., Larsson, B., von Heijne, G., & Sonnhammer, E. L. (2001). nov., comb. nov. International Journal of Systematic and Evolutionary Predicting transmembrane protein topology with a hidden Markov Microbiology, 60(Pt 3), 585–590. https://doi.org/10.1099/ model: Application to complete genomes. Journal of Molecular ijs.0.016725-0 Biology, 305(3), 567–580. https://doi.org/10.1006/jmbi.2000.4315 Carver, T., Thomson, N., Bleasby, A., Berriman, M., & Parkhill, J. (2009). Lagesen, K., Hallin, P., Rødland, E. A., Staerfeldt, H.-H., Rognes, T., & DNAPlotter: Circular and linear interactive genome visualiza- Ussery, D. W. (2007). RNAmmer: Consistent and rapid annotation tion. Bioinformatics (Oxford, England), 25(1), 119–120. https://doi. of ribosomal RNA genes. Nucleic Acids Research, 35(9), 3100–3108. org/10.1093/bioinformatics/btn578 https://doi.org/10.1093/nar/gkm160 Citron, D. M., Ostovari, M. I., Karlsson, A., & Goldstein, E. J. (1991). Lagier, J.-C., Armougom, F., Million, M., Hugon, P., Pagnier, I., Robert, Evaluation of the E test for susceptibility testing of anaerobic bacte- C., … Trape, J. F. (2012). Microbial culturomics: Paradigm shift in ria. Journal of Clinical Microbiology, 29(10), 2197–2203. the human gut microbiome study. Clinical Microbiology and Infection, Darling, A. C. E., Mau, B., Blattner, F. R., & Perna, N. T. (2004). Mauve: 18(12), 1185–1193. Multiple alignment of conserved genomic sequence with rearrange- Lagier, J.-C., Bibi, F., Ramasamy, D., Azhar, E. I., Robert, C., Yasir, M, … ments. Genome Research, 14(7), 1394–1403. https://doi.org/10.1101/ Raoult, D. (2014). Non contiguous- finished genome sequence and gr.2289704 description of Clostridium jeddahense sp. nov. Standards in Genomic Dione, N., Sankar, S. A., Lagier, J. –C., Khelaifia, S., Michele, C., Armstrong, Sciences, 9(3), 1003–1019. https://doi.org/10.4056/sigs.5571026 N., … Fournier, P. E. (2016). Genome sequence and description of Lagier, J.-C., Khelaifia, S., Alou, M. T., Ndongo, S., Dione, N., Hugon, P., Anaerosalibacter massiliensis sp. nov. New Microbes and New Infections, … Durand, G.. (2016). Culture of previously uncultured members of 10, 66–76. https://doi.org/10.1016/j.nmni.2016.01.002 the human gut microbiota by culturomics. Nature Microbiology, 1(12), Dubourg, G., Lagier, J. C., Armougom, F., Robert, C., Hamad, I., Brouqui, 16203. https://doi.org/10.1038/nmicrobiol.2016.203 P., Raoult, D. (2013). The gut microbiota of a patient with resis- Lagkouvardos, I., Joseph, D., Kapfhammer, M., Giritli, S., Horn, M., tant tuberculosis is more comprehensively studied by culturomics Haller, D., Haller, D., … Clavel, T. (2016). IMNGS: A comprehensive than by metagenomics. European Journal of Clinical Microbiology open resource of processed 16S rRNA microbial profiles for ecology & Infectious Diseases, 32(5), 637–645. https://doi.org/10.1007/ and diversity studies. Scientific Reports, 23(6), 33721. https://doi. s10096-012-1787-3 org/10.1038/srep33721 Fournier, P.-E., Lagier, J.-C., Dubourg, G., & Raoult, D. (2015). From cul- Lechner, M., Findeiss, S., Steiner, L., Marz, M., Stadler, P. F., & turomics to taxonomogenomics: A need to change the taxonomy of Prohaska, S. J. (2011). Proteinortho: Detection of (co- )orthologs in

221 AFOUDA et Al. | 11 of 11

large-scale analysis. BMC Bioinformatics, 28(12), 124. https://doi. Thompson, J. D., Higgins, D. G., & Gibson, T. J. (1994). CLUSTAL W: org/10.1186/1471-2105-12-124 Improving the sensitivity of progressive multiple sequence align- Lowe, T. M., & Eddy, S. R. (1997). tRNAscan- SE: A program for improved ment through sequence weighting, position-specific gap penalties detection of transfer RNA genes in genomic sequence. Nucleic Acids and weight matrix choice. Nucleic Acids Research, 22(22), 4673–4680. Research, 25(5), 955–964. https://doi.org/10.1093/nar/25.5.0955 https://doi.org/10.1093/nar/22.22.4673 Madsen, I. R., & Justesen, U. S. (2011). Bacteremia with Bacteroides pyo- Tindall, B. J., Rosselló-Móra, R., Busse, H.-J., Ludwig, W., & Kämpfer, P. genes after a cat bite. Journal of Clinical Microbiology, 49(8), 3092– (2010). Notes on the characterization of prokaryote strains for tax- 3093. https://doi.org/10.1128/JCM.00250-11 onomic purposes. International Journal of Systematic and Evolutionary Matuschek, E., Brown, D. F. J., & Kahlmeter, G. (2014). Development of the Microbiology, 60(Pt 1), 249–266. https://doi.org/10.1099/ EUCAST disk diffusion antimicrobial susceptibility testing method and ijs.0.016949-0 its implementation in routine microbiology laboratories. Clin Microbiol Turnbaugh, P. J., Ley, R. E., Hamady, M., Fraser-Liggett, C. M., Knight, Infect Off Publ Eur Soc Clin Microbiol Infect Dis., 20(4), O255–O266. R., & Gordon, J. I. (2007). The human microbiome project. Nature, Meier-Kolthoff, J. P., Auch, A. F., Klenk, H.-P., & Göker, M. (2013). Genome 449(7164), 804–810. https://doi.org/10.1038/nature06244 sequence- based species delimitation with confidence intervals and Vétizou, M., Pitt, J. M., Daillère, R., Lepage, P., Waldschmitt, N., Flament, improved distance functions. BMC Bioinformatics, 14, 60. https://doi. C., … Poirier-Colame, V. (2015). Anticancer immunotherapy by org/10.1186/1471-2105-14-60 CTLA-4 blockade relies on the gut microbiota. Science, 350(6264), Meier-Kolthoff, J. P., Klenk, H.-P., & Göker, M. (2014). Taxonomic use 1079–1084.https://doi.org/10.1126/science.aad1329 of DNA G+C content and DNA-DNA hybridization in the genomic Wayne, L. G., Brenner, D. J., Colwell, R. R., Grimont, P. A. D., age. International Journal of Systematic and Evolutionary Microbiology, Kandler, O., Krichevsky, M. I., … Starr, M. P. (1987). Report 64(Pt 2), 352–356. https://doi.org/10.1099/ijs.0.056994-0 of the ad hoc committee on reconciliation of approaches Poretsky, R., Rodriguez-R, L. M., Luo, C., Tsementzi, D., & Konstantinidis, to bacterial systematics. International Journal of Systematic K. T. (2014). Strengths and limitations of 16S rRNA gene amplicon se- and Evolutionary Microbiology, 37(4), 463–464. https://doi. quencing in revealing temporal microbial community dynamics. PLoS org/10.1099/00207713-37-4-463 ONE, 9(4), e93827. https://doi.org/10.1371/journal.pone.0093827 Zhao, G., Nyman, M., & Åke, Jönsson J. (2006). Rapid determination of Ramasamy, D., Mishra, A. K., Lagier, J.-C., Padhmanabhan, R., Rossi, M., short-chain fatty acids in colonic contents and faeces of humans and Sentausa, E., … Fournier, P. E. (2014). A polyphasic strategy incorporat- rats by acidified water- extraction and direct- injection gas chroma- ing genomic data for the taxonomic description of novel bacterial spe- tography. Biomedical Chromatography, 20(8), 674–682. https://doi. cies. International Journal of Systematic and Evolutionary Microbiology, org/10.1002/(ISSN)1099-0801 64(Pt 2), 384–391. https://doi.org/10.1099/ijs.0.057091-0 Zhou, Y., Liang, Y., Lynch, K. H., Dennis, J. J., & Wishart, D. S. (2011). Reynolds, E. S. (1963). The use of lead citrate at high pH as an electron- PHAST: A fast phage search tool. Nucleic Acids Research, 39(Web opaque stain in electron microscopy. Journal of Cell Biology, 17, 208– Server issue), W347–W352. https://doi.org/10.1093/nar/gkr485 212. https://doi.org/10.1083/jcb.17.1.208 Rosselló-Mora, R. (2006). DNA-DNA reassociation methods applied to microbial taxonomy and their critical evaluation. In Stackebrandt, E., SUPPORTING INFORMATION Mol identif syst popul struct prokaryotes (pp. 23–50). Berlin: Springer. https://doi.org/10.1007/978-3-540-31292-5 Additional Supporting Information may be found online in the sup- Rutherford, K., Parkhill, J., Crook, J., Horsnell, T., Rice, P., Rajandream, porting information tab for this article. M. A., Barrell, B. (2000). Artemis: Sequence visualization and anno- tation. Bioinformatics (Oxford England), 16(10), 944–945. https://doi. org/10.1093/bioinformatics/16.10.944 How to cite this article: Afouda P, Durand GA, Lagier J-C, Sasser, M. Bacterial identification by gas chromatographic analysis of fatty acids methyl esters (GC-FAME). Newark NY Microb ID et al. Noncontiguous finished genome sequence and [Internet]. 2006 [cited 2017 Oct 11]; Available from: http://youngin. description of Intestinimonas massiliensis sp. nov strain GD2T, com/application/AN-0702-0013EN.pdf the second Intestinimonas species cultured from the human Tamura, K., Stecher, G., Peterson, D., Filipski, A., & Kumar, S. (2013). gut. MicrobiologyOpen. 2018;e621. https://doi.org/10.1002/ MEGA6: Molecular evolutionary genetics analysis version 6.0. Molecular Biology and Evolution, 30(12), 2725–2729. https://doi. mbo3.621 org/10.1093/molbev/mst197

222 TAXONOGENOMICS: GENOME OF A NEW ORGANISM

Drancourtella massiliensis gen. nov., sp. nov. isolated from fresh healthy human faecal sample from South France

G. A. Durand1,2, J.-C. Lagier1,2, S. Khelaifia1, N. Armstrong1, C. Robert1, J. Rathored1, P.-E. Fournier1,2 and D. Raoult1,2,3 1) URMITE UM63, CNRS7278, IRD198, INSERM1085, Faculté de Médecine, Aix Marseille Université, 2) Pôle des Maladies Infectieuses, Hôpital La Timone, Assistance Publique-Hôpitaux de Marseille, Marseille, France and 3) Special Infectious Agents Unit, King Fahd Medical Research Center, King Abdulaziz University, Jeddah, Saudi Arabia

Abstract

Strain GD1T gen. nov., sp. nov., is the type strain of the newly proposed genus and species Drancourtella massiliensis, belonging to the Clostridiales order. This strain, isolated from the stool of a healthy person, is a Gram-positive rod, oxygen intolerant and nonmotile, with spore-forming activity. The features of this organism and its genome sequence are described. The draft genome is 3 057 334 bp long with 45.24% G + C content; it contains 2861 protein-coding genes and 64 RNA genes. New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases.

Keywords: Anaerobe, culturomics, Drancourtella massiliensisgen. nov. et sp. nov., gut microbiota, taxonogenomics Original Submission: 14 November 2015; Revised Submission: 30 January 2016; Accepted: 3 February 2016 Article published online: 10 February 2016

Clostridiales order in Firmicutes phylum, which was separated to Corresponding author: D. Raoult, URMITE UM63, CNRS7278, a different clade according to 16S rRNA analysis [5].Cluster IRD198, INSERM1085, Faculté de Médecine, Aix Marseille Université, 27 boulevard Jean Moulin, 13385 Marseille cedex 5, France XIV notably contains Ruminococcus, Coprococcus and Blautia E-mail: [email protected] genera. Ruminococcus was first described from rumen fluid. Human infections caused by these bacteria have been re- ported [6–9]. We propose the novel genus and species Drancourtella Introduction massiliensis for a strain isolated from the fresh stool of a healthy French person, which is phylogenetically close to T The gut microbiota is an important part of the Human Micro- Ruminococcus torques.TypestrainisGD1 (= CSUR biome Project. Indeed, 53% of cultivated bacterial species from P1506 = DSM 100357). human are isolated from the gut [1], and 80% of gut phylotypes found using the metagenomic method, which are mainly anaer- Material and Methods obic strains [2], are not yet cultivated [3]. Those phylotypes belong especially to Firmicutes and Proteobacteria. The culturo- mics project consists in the use of many different culture Ethics approval and sample collection methods to cultivate the most bacterial species possible from a After receiving signed informed consent, approved by the single plurimicrobial sample—notably the use of multiple Institut Fédératif de Recherche 48 (Faculty of Medicine, Mar- anaerobic conditions [4]. With the aim of cultivating the most seille, France) under agreement 09-022, a stool specimen was anaerobic species possible from human stool, we used culturo- collected at La Timone Hospital Marseille (France) in January mics methods to study the fresh stool of a healthy French person. 2015. The specimen was from a healthy French man (body mass ThenewgenusweproposehereisclosetoRuminococcus index 23.2 kg/m2), 28 years old, with no current treatment, torques, according to 16S rRNA phylogenetic analysis, part of especially no antibiotics.

New Microbe and New Infect 2016; 11: 34–42 New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) http://dx.doi.org/10.1016/j.nmni.2016.02.002 223 NMNI Durand et al. Drancourtella massiliensis gen. nov., sp. nov. genome 35

Isolation and growth conditions of strain using a bacterial dilution with a turbidity equivalent to the The stool specimen was incubated at 37°C into an anaerobic McFarland 1.0 standard. Wilkins Chalgren agar (Sigma, Aldrich, blood bottle Bactec Lytic/10 Anaerobic/F (Becton Dickinson, Steinheim, Germany) supplemented with 5% sheep blood was Le Pont de Claix, France) supplemented with 5% sheep’s blood, used for the experiment.. Incubations were done in anaerobic after a thermal shock of 20 minutes at 80°C. Then dilution conditions at 37°C, and reading was done at 48 hours using the cultures were performed, and characterization of growth Sirscan system (i2a, Montpellier, France) and eye controlled. conditions was tested as previously described [10]. Finally, sporulation, different pH levels and NaCl concentrations were Genome sequencing, annotation and comparison tested in the agar plate under the best culture conditions [11]. DNA extraction, in view of whole-genome sequencing, con- sisted of growing the species on Columbia agar supplemented Strain identification with 5% sheep’s blood (bioMérieux) at 37°C in an anaerobic Identification of colonies was performed by matrix-assisted atmosphere. Bacteria grown on three petri dishes were laser desorption/ionization time-of-flight mass spectrometry collected and resuspended in 4 × 100 μL Tris-EDTA (TE) (MALDI-TOF) Microflex spectrometer (Bruker Daltonics, buffer. Then 200 μL of this suspension was diluted in 1 mL TE Leipzig, Germany) as previously described, and the spectrum buffer for lysis treatment that included a 30-minute incubation was compared with our database (which includes the Bruker with 2.5 μg/μL lysozyme at 37°C, followed by an overnight database and our own collection) [10,12]. In case of non- incubation with 20 μg/μL proteinase K at 37°C [16]. As pre- identification, i.e. if the spectrum did not find a match in the viously described [10,15,17], our genomic platform uses a database (score <1.7), we proceeded to spectrum verification. protocol including proteinase K incubated overnight in order to If the spectrum was without background noise, and after digest contaminating proteins. Extracted DNA was then puri- exclusion of culture contamination, 16S rRNA was sequenced fied using three successive phenol–chloroform extractions and as previously described [10]. In case of a sequence similarity ethanol precipitations at −20°C overnight. After centrifugation, value lower than 96%, the species is considered to be a new the DNA was resuspended in 160 μL TE buffer. The whole genera without performing DNA-DNA hybridization, as sug- genome was then sequenced using the MiSeq Technology gested by Stackebrandt and Ebers [13]. (Illumina, San Diego, CA, USA) with the mate-pair strategy as previously described [17]. Open reading frames (ORFs) were Morphologic and biochemical characterization predicted using Prodigal [18] with default parameters, but the Morphologic characterization was first performed by observa- predicted ORFs were excluded if they were spanning a tion of Gram staining and motility of the fresh sample. Negative sequencing gap region (contain N). The predicted bacterial staining was then performed using bacteria fixed with 2.5% protein sequences were searched against the Clusters of glutaraldehyde, deposited on carbon formvar film and then Orthologous Groups (COGs) database using BLASTP (E value incubated for 1 second on ammonium molybdate 1%, dried on 1e-03, coverage 0.7 and identity percentage 30%). If no hit was blotting paper and finally observed using a TECNAI G20 found, a search was performed against the NR database using transmission electron microscope (FEI, Limeil-Brevannes, BLASTP with an E value of 1e-03, coverage 0.7 and identity France) at an operating voltage of 200 keV. Biochemical fea- percentage of 30%, If the sequence lengths were smaller than tures, such as oxidase, catalase, API 50CH, 20A and ZYM strips 80 aa, we used an E value of 1e-05. The tRNAScanSE tool [19] (bioMérieux, Marcy l’Étoile, France), were investigated ac- was used to find tRNA genes, whereas ribosomal RNAs were cording to the manufacturer’s instructions. Cellular fatty acids found by using RNAmmer [20]. Lipoprotein signal peptides and were analysed from two samples prepared with approximately the number of transmembrane helices were predicted using 10 mg of bacterial biomass each collected from several culture Phobius [21]. ORFans, sequences that did not blast in the plates. Fatty acid methyl esters were prepared as described [14] BLASTP program to a known sequence, have been defined by and gas chromatography mass spectrometry (GC/MS) analyses sequences with an E value smaller than 1e-3 in case of a were carried out as described previously [15]. sequence length higher than 80 aa, and an E value smaller than 1e-5 in case of a sequence length smaller than 80 aa. Such Antibiotic susceptibility parameter thresholds have already been used in previous works Antibiotic susceptibility was tested with the diffusion method to define ORFans [22]. For each selected genome, the complete according to the CASFM/EUCAST 2015 recommendations for genome sequence, proteome genome sequence and Orfeome fastidious anaerobes (http://www.sfm-microbiologie.org/ genome sequence were retrieved from the National Center for UserFiles/files/casfm/CASFM_EUCAST_V1_2015.pdf) using a Biotechnology Information (http://www.ncbi.nlm.nih.gov/news/ suspension of 1 McFarland on Wilkins Chalgren agar (Sigma 08-26-2014-new-genomes-FTP-live/). All proteomes were

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

224 36 New Microbes and New Infections, Volume 11 Number C, May 2016 NMNI

analysed with proteinOrtho [23]. Then for each couple of ge- conditions. MALDI-TOF spectrum of GD1T did not match nomes a similarity score was computed. This score is the mean anything in our database or Brucker’s database, even though value of nucleotide similarity between all couples of ortho- there was no background noise (Figs. 1 and 2). 16S sequencing logues between the two genomes studied (AGIOS) [24].An of the subunit of rRNA shows complete sequence of 1476 pb, annotation of the entire proteome was performed to define the with nucleotide BLAST results indicating Ruminococcus torques distribution of the functional classes of predicted genes ac- JCM6553 as the most closely cultured species, at 93.8% (Fig. 3). cording to the COGs of proteins (using the same method as for This result allowed us to define a new genus according to the the genome annotation). Annotation and comparison processes thresholds delimited by Stackebrandt and Ebers [13]. The were performed in the Multi-Agent software system DAG- GD1T 16S rRNA accession number from the EBI Sequence OBAH [25], which include Figenix libraries [26] that provide Database is LN828944. A gel view was performed in order to pipeline analysis. observe the spectra differences of Drancourtella massiliensis with other close bacteria (Fig. 2). Tested culture conditions have identified optimal growth at Results 37°C after 48 hours under anaerobic conditions, but a little growth appeared in microaerophilic conditions, suggesting a Classification and features relative oxygen tolerance. The pH range for growth is 6.5 to Type strain GD1T was first isolated after 10 days’ anaerobic 7.0, and NaCl concentration needs to be lower than 10 g/L. The T incubation of the stool sample in the presence of sheep’s blood, GD1 strain appears to be approximately 2 mm in size, ho- after thermal shock, and then cultivated on Columbia agar mogeneous, translucent and smooth, nonhemolytic, nonmotile, supplemented with 5% sheep’s blood under anaerobic with spore-forming activity colonies. Gram colouration was

FIG. 1. Reference mass spectrum from Drancourtella massiliensis strain GD1T. Spectra from 12 individual colonies were compared and reference spectrum generated.

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

225 NMNI Durand et al. Drancourtella massiliensis gen. nov., sp. nov. genome 37

FIG. 2. Gel view comparing Drancourtella massiliensis to other phylogenetically close species. Gel view displays raw spectra of loaded spectrum files arranged in pseudo-gel-like look. x-axis records m/z value. Left y-axis displays running spectrum number originating from subsequent spectra loading. Peak intensity is expressed by greyscale scheme code. The colour bar and right axis indicate the intensity each MALDI-TOF MS peak is displayed with and peak intensity in arbitrary units. Displayed species are indicated at left. positive, and bacteria demonstrated a bacilli aspect under the susceptible to gentamicin and resistant to tobramycin and microscope (Fig. 4). Electronic microscopy revealed small rods amikacin. GD1T strain was resistant to sulfamethoxazole, about 500 nm in size (Fig. 5). Classification and principal aztreonam, ciprofloxacin and ofloxacin, and susceptible to phenotypic features are listed in Table 1. imipenem, fosfomycin, clindamycin, metronidazole, rifampicin Catalase and oxidase production reactions were negatives. and colistin. Using an API 50CH strip, positive reactions were observed for D-ribose, methyl-αD-glucopyranoside and D-turanose and Genomic characterization and comparison for potassium 5-ketogluconate. Negative reactions were The genome is 3 057 334 bp long with 45.24% G + C content observed for all others. Using an API 20A strip, all reactions (Table 4). It is composed of seven scaffolds (composed of seven were negative. Using an API ZYM strip, reactions were pos- contigs). On the 2925 predicted genes, 2861 were protein-coding itive for leucine arylamidase, valine arylamidase, cysteine genes, and 64 were RNAs (two genes are 5S rRNA, two genes are arylamidase, naphthol-AS-BI-phosphohydrolase, β-galactosi- 16S rRNA, three genes are 23S rRNA and 57 genes are tRNA dase and N-acetyl-β-glucosaminidase and were negative for genes). A total of 1969 genes (68.82%) were assigned as putative others. The differences of characteristics compared with function (by COGs or by NR BLAST). Seventy-eight genes other representatives of the family Ruminococcaceae are (2.73%) were identified as ORFans. The remaining genes (703 detailed in Table 2. The major fatty acids detected (16:0 and genes, 24.57%) were annotated as hypothetical proteins. Table 5 14:0) are saturated species. The GC/MS results indicated summarizes the distribution of genes into COGs functional cat- lower amounts of unsaturated acids and other saturated egories. The genome sequence has been deposited in GenBank compounds (Table 3). under accession number CVPG00000000. Concerning susceptibility on β-lactam, the GD1T strain was The draft genome sequence of Drancourtella massiliensis (3.05 susceptible to amoxicillin, ticarcillin, cefepime, vancomycin, Mb) is smaller than those of Ruminococcus gnavus, Clostridium teicoplanin and linezolid but resistant to ceftriaxone and cefta- scindens, Coprococcus comes and Dorea formicigenerans (3.72, zidime. Concerning aminoglycoside antibiotics, cells were 3.62, 3.24 and 3.19 Mb respectively) but larger than those of

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

226 38 New Microbes and New Infections, Volume 11 Number C, May 2016 NMNI

FIG. 3. Phylogenetic tree highlighting position of Drancourtella massiliensis gen. nov., sp. nov., strain GD1T relative to other phylogenetically close type strains. Clusters are made according to Collins et al. [5]. Sequences were aligned using CLUSTALW, and phylogenetic in- ferences were obtained with Kimura two-parameter model using neighbour- joining method with 1000 bootstrap replicates within MEGA6. Scale bar = 1% nucleotide sequence divergence.

Eubacterium ventriosum (2.87 Mb). The G + C content of Drancourtella massiliensis is smaller than those of Clostridium scindens (45.24 and 46.35% respectively), but larger than those of Ruminococcus gnavus, Coprococcus comes, Dorea for- micigenerans and Eubacterium ventriosum (42.52, 42.49, 40.97 and 34.92% respectively). The gene content of Drancourtella massiliensis is smaller than those of Ruminococcus gnavus, Clos- tridium scindens, Coprococcus comes and Dorea formicigenerans (2861, 3762, 3995, 3913 and 3277 respectively) but larger than those of Eubacterium ventriosum (2802). Figures 6 and 7

FIG. 5. Transmission electron microscopy of Drancourtella massiliensis strain GD1T using TECNAI G20 (FEI) at operating voltage of 200 keV. FIG. 4. Gram staining of Drancourtella massiliensis strain GD1T. Scale bar = 200 nm.

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

227 NMNI Durand et al. Drancourtella massiliensis gen. nov., sp. nov. genome 39

TABLE 1. Classification and general features of Drancourtella TABLE 4. Nucleotide content and gene count levels of massiliensis strain GD1T genome

Property Term Genome (total)

Current classification Domain: Bacteria Attribute Value % of totala Phylum: Firmicutes Class: Clostridia Size (bp) 3 057 334 100 Order: Clostridiales G + C content (bp) 1 383 137 45.24 Family: Ruminococcaceae Coding region (bp) 2 772 730 90.7 Genus: Drancourtella Total genes 2925 100 Species: Drancourtella massiliensis T RNA genes 64 2.18 Type strain: GD1 Protein-coding genes 2861 97.81 Gram stain Positive Genes with function prediction 1969 67.32 Cell shape Rod Genes assigned to COGs 1763 61.62 Motility Nonmotile Genes with peptide signals 245 8.56 Sporulation Sporulating Genes with transmembrane helices 664 23.2 Temperature range Mesophilic Genes with Pfam domains 2693 92 Optimum temperature 37°C COGs, Clusters of Orthologous Groups database. aTotal is based on either the size of the genome in base pairs or the total number of protein coding genes in the annotated genome. TABLE 2. Differential characteristics of Drancourtella massiliensis strain GD1T, Ruminococcus faecis Eg2T, Ruminococcus torques ATCC27756 and Ruminococcus lactaris ATCC29176 demonstrate that the distribution of genes into COGs cate- gories was similar in all compared genomes. To evaluate the Property D. massiliensis R. faecis R. torques R. lactaris genomic similarity among the studied strains, we determined Cell diameter (μm) 0.5 1.0 NA NA two parameters, digital DNA-DNA hybridization (dDDH), Oxygen requirement Anaerobic Anaerobic Anaerobic Anaerobic Gram stain + + + + which exhibits a high correlation with DDH (Table 6) [27,28], Salt requirement <10 g/L NA NA NA Motility −−NA NA and AGIOS (Table 7) [24], which was designed to be inde- Endospore formation + −−− Indole −−+ − pendent from DDH. Drancourtella massiliensis shared 1235, 922, Production of: Alkaline phosphatase − +NANA 1082, 1202 and 1266 orthologous proteins with Dorea for- Catalase − +NANA Oxidase −−NA NA micigenerans, Eubacterium ventriosum, Coprococcus comes, Rumi- Nitrate reductase −−+ − nococcus gnavus and Clostridium scindens respectively (Table 7). Urease −−NA NA β-Galactosidase −−−+ N-acetyl-glucosamine −−++ Acid from: L-Arabinose −−+ − Ribose + NA NA NA Mannose −−+ − Mannitol −−NA NA TABLE 5. Number of genes associated with 25 general COGs Saccharose −−+ − D-Glucose − + −− functional categories D-Fructose − NA NA NA − − D-Maltose ++ Code Value % of totala Description D-Lactose − ++− Habitat Human gut Human gut Human gut Human gut J 152 5.31 Translation +, positive result; −, negative result; v, variable result; w, weakly positive result; NA, A 0 0 RNA processing and modification data not available. K 177 6.18 Transcription L 115 4.02 Replication, recombination and repair B 0 0 Chromatin structure and dynamics D 23 0.8 Cell cycle control, mitosis and meiosis TABLE 3. Total cellular fatty acid composition Y 0 0 Nuclear structure V 71 2.48 Defense mechanisms Fatty acids IUPAC name Mean relative %a T 56 1.96 Signal transduction mechanisms M 95 3.32 Cell wall/membrane biogenesis N 3 0.10 Cell motility 16:0 Hexadecanoic acid 39.5 ± 4.1 Z 0 0 Cytoskeleton 14:0 Tetradecanoic acid 21.4 ± 9.6 W 0 0 Extracellular structures 18:1n9 9-Octadecenoic acid 8.2 ± 5.2 U 22 0.77 Intracellular trafficking and secretion 16:1n7 9-Hexadecenoic acid 6.3 ± 0.9 O 59 2.06 Posttranslational modification, protein 18:0 Octadecanoic acid 6.2 ± 1.6 turnover, chaperones 18:1n7 11-Octadecenoic acid 3.8 ± 0.5 C 98 3.43 Energy production and conversion 13:0 Tridecanoic acid 3.3 ± 1.8 G 202 7.06 Carbohydrate transport and metabolism 18:2n6 9,12-Octadecadienoic acid 3.2 ± 0.8 E 227 7.93 Amino acid transport and metabolism 14:1n5 9-Tetradecenoic acid 2.7 ± 0.4 F 62 2.16 Nucleotide transport and metabolism 12:0 Dodecanoic acid 1.4 ± 0.8 H 67 2.34 Coenzyme transport and metabolism 15:0 Pentadecanoic acid 1.4 ± 0.8 I 44 1.53 Lipid transport and metabolism 5:0 anteiso 2-methyl-butanoic acid TR P 104 3.64 Inorganic ion transport and metabolism 15:0 anteiso 12-methyl-tetradecanoic acid TR Q 28 0.98 Secondary metabolites biosynthesis, 15:1n5 10-Pentadecenoic acid TR transport and catabolism C14:03OH 3-hydroxy-Tridecanoic acid TR R 237 8.28 General function prediction only 17:0 Heptadecanoic acid TR S 111 3.88 Function unknown — 1098 38.38 Not in COGs TR, trace amounts (<1%). aMean peak area percentage calculated from analysis of FAMEs in three sample COGs, Clusters of Orthologous Groups database. preparations ± standard deviation (n = 3). aTotal is based on total number of protein-coding genes in annotated genome.

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

228 40 New Microbes and New Infections, Volume 11 Number C, May 2016 NMNI

FIG. 6. Graphical circular map of chromosome. From outside to center: Genes on forward strand coloured by COGs categories (only genes assigned to COGs), genes on reverse strand coloured by COGs categories (only genes assigned to COGs), RNA genes (tRNAs green, rRNAs red), GC content and GC skew.

Conclusion Gram-positive bacilli. Spore forming and nonmotile. Dran- courtella massiliensis is anaerobic. Oxidase and catalase nega- tive. Indole negative, nitrate reductase negative. Habitat is the Anaerobic conditions of culturomics have permitted the cul- human gut. Type species is Drancourtella massiliensis. ture of the first strain of the new genus Drancourtella. Tax- fi ogenomics studies con rmed this species to be Drancourtella Description of Drancourtella massiliensis gen. nov., sp. nov. massiliensis gen. nov., sp. nov. Drancourtella massiliensis (ma.si.li.en’sis, L. fem. adj. Massiliensis, from Latin Massilia, the city where the bacteria was isolated, Taxonomic and nomenclatural proposals Marseille). D. massiliensis strain GD1T (= CSUR P1506 = DSM 100357) presented as translucent colonies 2 mm in diameter, Description of Drancourtella gen. nov. isolated from the human stool of a healthy French person. The Drancourtella (dran.cour.tel’la, N.L. gen. fem.) name was Bacteria appeared as Gram-positive bacilli 0.5 μmlong. chosen in honor of the French microbiologist Michel Drancourt Metabolism was strictly anaerobic, and growth was optimal (Université de la Méditerranée, Marseille, France). at 37°C. D. massiliensis was spore forming and nonmotile.

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

229 NMNI Durand et al. Drancourtella massiliensis gen. nov., sp. nov. genome 41

FIG. 7. Distribution of functional classes of predicted genes according to Clusters of Orthologous Groups (COGs) database of proteins.

TABLE 6. Pairwise comparison of Drancourtella massiliensis (upper right) with eight other species using GGDC, formula 2 (DDH estimates based on identities/HSP length)a

Dorea formicigenerans Eubacterium ventriosum Coprococcus comes Ruminococcus gnavus Clostridium scindens Drancourtella massiliensis

D. formicigenerans 100% ± 00 32.8% ± 2.56 39.4% ± 2.70 25.6% ± 2.59 22.8% ± 2.62 22.9% ± 2.56 E. ventriosum 100% ± 00 38.9% ± 2.56 32.2% ± 2.55 28% ± 2.54 25.5% ± 2.53 C. comes 100% ± 00 23.1% ± 2.58 22.5% ± 2.56 21.5% ± 2.56 R. gnavus 100% ± 00 25.7% ± 2.58 22.2% ± 2.56 C. scindens 100% ± 00 22.2% ± 2.57 D. massiliensis 100% ± 00

DDH, DNA-DNA hybridization; GGDC, Genome-to-Genome Distance Calculator; HSP, high-scoring segment pairs. aConfidence intervals indicate inherent uncertainty in estimating DDH values from intergenomic distances based on models derived from empirical test data sets (which are always limited in size). These results are in accordance with 16S rRNA (Fig. 3) and phylogenomic analyses as well as GGDC results.

Oxidase and catalase were negative. Positive reactions were BI-phosphohydrolase, β-galactosidase and N-acetyl-β-gluco- observed for D-ribose, methyl-αD-glucopyranoside, D-tur- saminidase. Resistance for ceftriaxone, ceftazidime, tobra- anose and for potassium 5-ketogluconate, leucine arylami- mycin, amikacin, sulfamethoxazole, aztreonam, ciprofloxacin dase, valine arylamidase, cysteine arylamidase, naphthol-AS- and ofloxacin was observed.

TABLE 7. Numbers of orthologous protein shared between genomes (upper right) and average percentage similarity of nucleotides corresponding to orthologous protein shared between genomes (lower left) and numbers of proteins per genome (bold)

Dorea formicigenerans Eubacterium ventriosum Coprococcus comes Ruminococcus gnavus Clostridium scindens Drancourtella massiliensis

D. formicigenerans 3277 987 1194 1234 1337 1235 E. ventriosum 66.46 2802 870 954 946 922 C. comes 71.29 65.98 3913 1075 1144 1082 R. gnavus 69.73 65.47 70.00 3762 1269 1202 C. scindens 70.87 63.52 68.76 69.15 3995 1266 D. massiliensis 68.07 63.73 68.96 69.31 68.68 2861

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

230 42 New Microbes and New Infections, Volume 11 Number C, May 2016 NMNI

This strain exhibited a G + C content of 45.24%. Its 16S rRNA [10] Lagier JC, Elkarkouri K, Rivet R, Couderc C, Raoult D, Fournier PE. fi sequence was deposited in GenBank under accession number Non contiguous- nished genome sequence and description of Sene- galemassilia anaerobia gen. nov., sp. nov. Stand Genomic Sci 2013;7: LN828944, and the whole genome shotgun sequence was 343–56. deposited in GenBank under accession number CVPG00000000. [11] Aghnatios R, Cayrou C, Garibal M, Robert C, Azza S, Raoult D, et al. The type strain GD1T (= CSUR P1506 = DSM 100357) was Draft genome of Gemmata massiliana sp. nov, a water-borne Plancto- fl mycetes species exhibiting two variants. Stand Genomic Sci 2015;10:120. isolated from the fecal ora of a healthy patient in France. [12] Seng P, Abat C, Rolain JM, Colson P, Lagier JC, Gouriet F, et al. Identification of rare pathogenic bacteria in a clinical microbiology laboratory: impact of matrix-assisted laser desorption ionization-time Acknowledgements of flight mass spectrometry. J Clin Microbiol 2013;51:2182–94. [13] Stackebrandt E, Ebers J. Taxonomic parameters revisited: tarnished gold standards. Microbiol Today 2006;33:152–5. The authors thank the Xegen Company (http://www.xegen.fr/) [14] Myron Sasser. Bacterial identification by gas chromatographic analysis of for automating the genomic annotation process. This study was fatty acids methyl esters (GC-FAME). MIDI 2006; (Technical Note 101). [15] Dione N, Sankar SA, Lagier JC, Khelaifia S, Michele C, Armstrong N, funded by the Fondation Méditerranée Infection. We thank K. et al. Genome sequence and description of Anaerosalibacter massiliensis Griffiths for English-language review and C. Andrieu for sp. nov. New Microbes New Infect 2016;10:66–76. administrative assistance. [16] Sengüven B, Baris E, Oygur T, Berktas M. Comparison of methods for the extraction of DNA from formalin-fixed, paraffin-embedded archival tissues. Int J Med Sci 2014;11:494–9. fl [17] Lagier JC, Bibi F, Ramasamy D, Azhar EI, Robert C, Yasir M, et al. Non Con ict of Interest contiguous-finished genome sequence and description of Clostridium jeddahense sp. nov. Stand Genomic Sci 2014;9:1003–19. [18] Hyatt D, Chen GL, Locascio PF, Land ML, Larimer FW, Hauser LJ. None declared. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 2010;11:119. [19] Lowe TM, Eddy SR. tRNAscan-SE: a program for improved detection References of transfer RNA genes in genomic sequence. Nucleic Acids Res 1997;25:955–64. [20] Lagesen K, Hallin P, Rødland EA, Staerfeldt HH, Rognes T, Ussery DW. RNAmmer: consistent and rapid annotation of ribosomal RNA genes. [1] Hugon P, Dufour JC, Colson P, Fournier PE, Sallah K, Raoult D. Nucleic Acids Res 2007;35:3100–8. A comprehensive repertoire of prokaryotic species identified in human [21] Bendtsen JD, Nielsen H, von Heijne G, Brunak S. Improved prediction beings. Lancet Infect Dis 2015;15:1211–9. of signal peptides: SignalP 3.0. J Mol Biol 2004;340:783–95. [2] Faith JJ, Guruge JL, Charbonneau M, Subramanian S, Seedorf H, [22] Daubin V, Ochman H. Bacterial genomes as new gene homes: the Goodman AL, et al. The long-term stability of the human gut micro- genealogy of ORFans in E. coli. Genome Res 2004;14:1036–42. biota. Science 2013;341(6141):1237439. [23] Lechner M, Findeiss S, Steiner L, Marz M, Stadler PF, Prohaska SJ. [3] Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, Proteinortho: detection of (co-)orthologs in large-scale analysis. BMC Gordon JI. The human microbiome project. Nature 2007;449(7164): Bioinformatics 2011;12:124. 804–10. [24] Ramasamy D, Mishra AK, Lagier JC, Padhmanabhan R, Rossi M, [4] Lagier JC, Hugon P, Khelaifia S, Fournier PE, La Scola B, Raoult D. The Sentausa E, et al. A polyphasic strategy incorporating genomic data for rebirth of culture in microbiology through the example of culturomics the taxonomic description of novel bacterial species. Int J Syst Evol to study human gut microbiota. Clin Microbiol Rev 2015;28:237–64. Microbiol 2014;64(Pt 2):384–91. [5] Collins MD, Lawson PA, Willems A, Cordoba JJ, Fernandez- [25] Gouret P, Paganini J, Dainat J, Louati D, Darbo E, Pontarotti P, et al. Garayzabal J, Garcia P, et al. The phylogeny of the genus Clostridium: Integration of evolutionary biology concepts for functional annotation proposal of five new genera and eleven new species combinations. Int J and automation of complex research in evolution: the multi-agent Syst Bacteriol 1994;44:812–26. software system DAGOBAH. In: Evolutionary biology—concepts, [6] Hansen SGK, Skov MN, Justesen US. Two cases of Ruminococcus gnavus biodiversity, macroevolution and genome evolution. Berlin: Springer; bacteremia associated with diverticulitis. J Clin Microbiol 2013;51: 2011. p. 71–87. 1334–6. [26] Gouret P, Vitiello V, Balandraud N, Gilles A, Pontarotti P, [7] Livaoglu M, Yilmaz G, Kerimoglu S, Aydin K, Karacal N. Necrotizing Danchin EGJ. FIGENIX: intelligent automation of genomic annotation: fasciitis with ruminococcus. J Med Microbiol 2008;57(Pt 2):246–8. expertise integration in a new software platform. BMC Bioinformatics [8] Sucu N, Köksal I, Yilmaz G, Aydin K, Caylan R, Aktoz Boz G. [Liver 2005;6:198. abscess and infective endocarditis cases caused by Ruminococcus pro- [27] Auch AF, von Jan M, Klenk HP, Göker M. Digital DNA-DNA hy- ductus]. Mikrobiyoloji Bül 2006;40:389–95. bridization for microbial species delineation by means of genome-to- [9] Wang L, Christophersen CT, Sorich MJ, Gerber JP, Angley MT, genome sequence comparison. Stand Genomic Sci 2010;2:117–34. Conlon MA. Increased abundance of Sutterella spp. and Ruminococcus [28] Meier-Kolthoff JP, Auch AF, Klenk HP, Göker M. Genome sequence- torques in feces of children with autism spectrum disorder. Mol Autism based species delimitation with confidence intervals and improved 2013;4:42. distance functions. BMC Bioinformatics 2013;14:60.

New Microbes and New Infections © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 11,34–42 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

231 NEW SPECIES

‘Bittarella massiliensis’ gen. nov., sp. nov. isolated by culturomics from the gut of a healthy 28-year-old man

G. A. Durand1, P.-E. Fournier1, D. Raoult1,2 and S. Edouard1 1) Aix-Marseille Université, Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE), UM63, CNRS 7278, IRD 198, INSERM U1095, Institut Hospitalo-Universitaire Méditerranée Infection, Marseille, France and 2) Special Infectious Agents Unit, King Fahd Medical Research Centre, King Abdulaziz University, Jeddah, Saudi Arabia

Abstract

We report here the main features of the proposed new bacterial genus Bittarella. The type strain ‘Bittarella massiliensis’ GD6T (CSUR P2149) was isolated from a stool sample from a healthy French man. © 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases.

Keywords: Anaerobe, ‘Bittarella massiliensis’ gen. nov., sp. nov., culturomics, gut microbiota, taxono-genomics Original Submission: 30 November 2016; Revised Submission: 14 December 2016; Accepted: 14 December 2016 Article published online: 19 December 2016

motile, Gram-negative and rod-shaped bacteria with a mean Corresponding author: S. Edouard, URMITE UM63, CNRS7278, diameter of 250 nm and a length of 1 μm without spore-forming IRD198, INSERM1095, Faculté de Médecine, Aix Marseille Université, 27 boulevard Jean Moulin, 13385 Marseille cedex 5, France activity. Catalase and oxidase were negative. The complete 16S E-mail: [email protected] rRNA gene of the bacterium was sequenced as previously described [4] and shared 89.5% of identity with Anaerotruncus massiliensis strain AT3 (GenBank Accession number LN866995) We isolated in April 2015, as part of the culturomics study of [5]. The phylogenetically closest species standing in nomencla- the Human Microbiome [1], an oxygen-intolerant species that ture was Hydrogenoanaerobacterium saccharovorans with 89.2% of could not be identified by matrix-assisted laser desorption/ similarity (Fig. 1). The phylogenetic analysis confirms the bac- ionization time-of-flight mass spectrometry (MALDI-TOF MS) terium as a member within the Ruminococcaceae family belonging screening (score <1.7) using a Microflex spectrometer (Bruker to the phylum Firmicutes (Fig. 1). Hydrogenoanaerobacterium sac- Daltonics, Bremen, Germany) [1–3]. The species was isolated charovorans is an anaerobic, hydrogenogenic, rod-shaped bac- fl from the faeces of a healthy, 28-year-old French man. The stool terium isolated from a laboratory-scale H2-producing up- ow was inoculated without delay in different culture conditions anaerobic sludge blanket reactor [6]. used for culturomics [1]. The initial growth of the GD6 strain Strain GD6 exhibits a 16S rRNA sequence divergence >5% occurred after 48 h of anaerobic incubation in a 5% sheep with its phylogenetically closest species with a validly published blood-enriched Columbia agar (bioMérieux, Marcy l’Etoile, name standing in nomenclature [7], so we propose the creation France). The stool was pre-incubated for 10 days at 37°C in an of the genus ‘Bittarella’ whose type strain is ‘Bittarella massiliensis’ anaerobic atmosphere in a culture bottle in the presence of 5% GD6T (Bit.ar. Masc. Adj., in honour of Dr Bittar, a French sheep blood and 5% rumen fluid filter-sterilized through a 0.2- microbiologist, and mas.il.i.en’sis. L. gen. masc. n. massiliensis,of μm-pore filter (Thermo Fisher Scientific, Villebon sur Yvette, Massilia, the Latin name of Marseille where the strain GD6T was France). The donor gave signed informed consent and the study first isolated). was validated by the ethics committee of the Institut Fédératif MALDI-TOF-MS spectrum accession number. The de Recherche IFR48 under number 09-022. MALDI-TOF-MS spectrum of ‘Bittarella massiliensis’ is available The colonies appeared to be 0.5 mm in size, homogeneous, at http://mediterranee-infection.com/article.php? laref=256& translucent, smooth and non-haemolytic. Strain GD6 was a non- titre=urms-database. (Last accessed 28 November 2016).

New Microbe and New Infect 2017; 16: 28–29 © 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) http://dx.doi.org/10.1016/j.nmni.2016.12.014 232 NMNI Durand et al. ‘Bittarella massiliensis’ gen. nov., sp. nov. 29

FIG. 1. Phylogenetic tree based on the 16S rRNA gene sequence showing the position of Bittarella massiliensis GD6T (bold) with other closely related species among the Ruminococcaceae. The EMBL database accession numbers are indicated in parentheses. Sequences were aligned using CLUSTALW, and phylogenetic inferences were obtained with a Kimura two-parameter model using the neighbour-joining method with 1000 bootstrap replicates, within MEGA6 software. The scale bar represents a 2% nucleotide sequence divergence.

Nucleotide sequence accession number. The 16S culturomics to study human gut microbiota. Clin Microbiol Rev – rRNA gene sequence was deposited in GenBank under 2015;28:237 64. [2] Lagier J-C, Armougom F, Million M, Hugon P, Pagnier I, Robert C, et al. Accession number LN881596. Microbial culturomics: paradigm shift in the human gut microbiome Deposit in a culture collection. Strain GD6T was study. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis deposited in the collection de Souches de l’Unités des Rick- 2012;18:1185–93. [3] Seng P, Abat C, Rolain JM, Colson P, Lagier J-C, Gouriet F, et al. ettsies (CSUR, WDCM 875) under number P2149. Identification of rare pathogenic bacteria in a clinical microbiology laboratory: impact of matrix-assisted laser desorption ionization- time of flight mass spectrometry. J Clin Microbiol 2013;51: Transparency declaration 2182–94. [4] Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 16S ribosomal DNA sequence analysis of a large collection of environmental No conflicts of interest are declared. and clinical unidentifiable bacterial isolates. J Clin Microbiol 2000;38: 3623–30. [5] Togo AH, Valero R, Delerce J, Raoult D, Million M. “Anaerotruncus ” fi Acknowledgements massiliensis , a new species identi ed from human stool from an obese patient after bariatric surgery. New Microbes New Infect 2016;14: 56–7. [6] Song L, Dong X. Hydrogenoanaerobacterium saccharovorans gen. nov., sp. This work was funded by the Mediterrannée-Infection Foun- nov., isolated from H2-producing UASB granules. Int J Syst Evol dation. We thank Magdalen Lardière for English revision. Microbiol 2009;59:295–9. [7] Kim M, Oh H-S, Park S-C, Chun J. Towards a taxonomic coherence between average nucleotide identity and 16S rRNA gene sequence References similarity for species demarcation of prokaryotes. Int J Syst Evol Microbiol 2014;64:346–51.

[1] Lagier J-C, Hugon P, Khelaifia S, Fournier P-E, La Scola B, Raoult D. The rebirth of culture in microbiology through the example of

© 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 16,28–29 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

233 TAXONOGENOMICS: GENOME OF A NEW ORGANISM

Description of Clostridium phoceensis sp. nov., a new species within the genus Clostridium

M. Hosny, S. Benamar, G. Durand, N. Armstrong, C. Michelle, F. Cadoret, B. La Scola and N. Cassir Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE), UM63 CNRS 7278 IRD 198 INSERM U1095, IHU Méditerranée Infection, Pôle des Maladies Infectieuses, Assistance Publique-Hôpitaux de Marseille, Faculté de Médecine, Marseille, France

Abstract

Clostridium phoceensis sp. nov., strain GD3T (= CSUR P1929 = DSM 100334) is the type strain of C. phoceensis sp. nov., a new species within the genus Clostridium. This strain was isolated from the gut microbiota of a 28-year-old healthy French man. C. phoceensis is a Gram-negative, spore-forming, nonmotile, strictly anaerobic bacterium. We describe its complete genome sequence and annotation, together with its phenotypic characteristics. © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases.

Keywords: Anaerobe, Clostridium phoceensis, culturomics, gut microbiota, taxono-genomics Original Submission: 17 June 2016; Revised Submission: 9 September 2016; Accepted: 16 September 2016 Article published online: 25 September 2016

and the analysis and comparison of the complete genome Corresponding author: N. Cassir, Unité de Recherche sur les sequence. Maladies Infectieuses et Tropicales Emergentes, URMITE, UM63, CNRS 7278, IRD 198, INSERM 1095, Faculté de médecine, Aix- Since the creation of the genus Clostridium in 1880, more Marseille Université, 27 Boulevard Jean Moulin, 13385 Marseille than 200 species have been described [5]. While species Cedex 05, France belonging to the family Clostridiaceae are mainly associated with E-mail: [email protected] fl M. Hosny and S. Benamar contributed equally to this article, and the commensal digestive ora of mammals and can be both should be considered first author. commonly found in the environment, some are major human pathogens, including toxigenic Clostridium botulinum, Peptoclos- tridium difficile, Clostridium tetani and Clostridium perfringens [6]. T Introduction We propose Clostridium phoceensis sp. nov., strain GD3 (= CSUR P1929 = DSM 100334) as the type strain of C. phoceensis sp. nov., a new species within the genus Clostridium. This strain Human adult gut microbiota is estimated to consist of up to 100 was isolated from the gut microbiota of a 28-year-old healthy trillion microorganisms, comprising at least 500 different spe- French man as part of a culturomics study aiming at individually cies, mostly anaerobic bacteria [1]. Although the advent of cultivating all bacterial species from a stool sample. Here we modern molecular microbiological methods has expanded the describe the characteristics of C. phoceensis sp. nov. strain degree of bacterial detection from stool samples, it does not GD3T, including its phenotype and genome sequence. allow for the phenotypic description of new living species [2]. Consequently, there has been renewed interest in culture Materials and Methods methods [3]. A new taxonomic approach known as taxonogenomics has recently been proposed to describe new isolated bacterial Sample collection species [4]. This polyphasic strategy combines phenotypic The stool sample was taken from a 28-year-old healthy French characteristics, the matrix-assisted laser desorption/ionization man. The sample was collected as part of a research study on fl time-of- ight mass spectrometry (MALDI-TOF MS) spectrum, the human gut microbiota. The study was approved by the

New Microbe and New Infect 2016; 14: 85–92 © 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 234 http://dx.doi.org/10.1016/j.nmni.2016.09.008 86 New Microbes and New Infections, Volume 14 Number C, November 2016 NMNI

Institut Fédératif de Recherche 48 (agreement no. 09-022, carried out as described by Dione et al. [11]. Briefly, FAMEs Marseille, France), and the patient’s consent was obtained. The were separated using an Elite 5-MS column and monitored by sample was stored at −80°C in La Timone Hospital (Marseille, mass spectrometry (Clarus 500-SQ8S; PerkinElmer, Courta- France). boeuf, France). A spectral database search was performed using MS Search 2.0 operated using the Standard Reference Database Strain isolation and identification (MALDI-TOF MS and 1A (National Institute of Standards and Technology, Gaithers- 16S rRNA sequencing) burg, MD, USA) and the FAME mass spectral database (Wiley, The faecal sample was treated using the concept of culturomics Chichester, UK). [7]. The colonies obtained were identified using MALDI-TOF MS [8,9] (Bruker Daltonics, Leipzig, Germany) and analysed Microscopy using a Microflex spectrometer (Bruker), leading to the protein Individual cells of C. phoceensis strain GD3T were captured spectrum being obtained. A score under 1.7 did not enable any using a Tecnai G20 electron microscopy (FEI Company, Limeil- identification. Subsequently, 16S rRNA was sequenced and the Brevannes, France). A Color Gram 2 Kit (bioMérieux) was used sequence was matched using BLAST (Basic Local Alignment to perform the Gram coloration observed with a 100× oil- Search Tool) against the National Center for Biotechnology immersion objective lens using the DM1000 photonic micro- Information database [10]. DNA was extracted using the EZ1 scope (Leica, Wetzlar, Germany) [12,13]. Tissue Extraction Kit (Qiagen, Hilden, Germany), and se- quences were aligned using ChromasPro 1.6 (Technelysium, Genome project history South Brisbane, Queensland, Australia). The organism was selected for sequencing because it had been isolated from a healthy person for the first time and on the basis Growth conditions of its 16S rRNA similarity, phylogenetic position and phenotypic The growth condition of the strain was determined by testing differences with other members of the genus Clostridia. The different temperatures and atmospheres. Five growth temper- GenBank accession number is CVUG01000000 and consists of atures (ambient, 28, 37, 45 and 56°C) were tested under 10 scaffolds with a total of 16 contigs [14]. anaerobic (GENbag anaer) microaerophilic atmospheres (GENbag microer) (bioMérieux, Marcy l’Étoile, France) and Genome sequencing and assembly aerobic condition on 5% sheep’s blood agar (bioMérieux). The genomic DNA of Clostridium phoceensis GD3T was Colonies were obtained after thermal shock for 20 minutes at sequenced on the MiSeq sequencer (Illumina, San Diego, CA, 80°C in an anaerobic blood culture bottle (Bactec Lytic/10 USA) using the mate-pair strategy. The gDNA was barcoded in Anaerobic/F) supplemented with 5% sheep’s blood at 37°C. order to be mixed with 11 other projects using the Nextera mate-pair sample prep kit (Illumina). The mate-pair library was Phenotypic, biochemical and antibiotic susceptibility prepared with 1.5 μg of genomic DNA using the Nextera Mate- tests Pair Illumina guide. The gDNA sample was simultaneously Gram staining, motility, catalase and oxidase were determined fragmented and tagged using a mate-pair junction adapter. The as described by Lagier et al. [3]. Sporulation was tested using a pattern of the fragmentation was validated on an Agilent 2100 thermal shock on bacterial colonies (diluted in phosphate- BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA) with buffered saline) for 10 minutes at 80°C. The biochemical a DNA 7500 lab chip. The DNA fragments ranged in size from characteristics were tested using API 50CH, API ZYM and API 1 to 10 kb, with an optimal size of 4.490 kb. No size selection 20A strips (bioMérieux). Antibiotic susceptibility referred to was performed, and only 600 ng of tagmented fragments were European Committee on Antimicrobial Susceptibility Testing circularized. The circularized DNA was mechanically sheared 2015 recommendations. to small fragments with an optimal size of 938 bp on the Covaris S2 device in microtubes (Covaris, Woburn, MA, USA). Fatty acid methyl ester (FAME) analysis by gas The library profile was visualized on a High Sensitivity Bio- chromatography/mass spectrometry (GC/MS) analyzer LabChip (Agilent Technologies), and the final library FAME analysis was performed by GC/MS. Two samples were concentration was measured at 4.457 nmol/L. The libraries prepared with 2 mg of bacterial biomass from several culture were normalized at 2 nM and pooled. After a denaturation step plates. Two samples were prepared with approximately 2 mg of and dilution at 15 pM, the pool of libraries was loaded onto the bacterial biomass per tube collected from several culture plates. reagent cartridge and then onto the instrument along with the FAMEs were prepared as described by Sasser (http://www.midi- flow cell. Automated cluster generation and sequencing run inc.com/pdf/MIS_Technote_101.pdf). GC/MS analyses were were performed in a single 39-hour run in a 2 × 251 bp; 6.1 Gb

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

235 NMNI Hosny et al. Clostridium phoceensis sp. nov. 87

FIG. 1. Reference mass spectrum (matrix-assisted desorption ioniza- tion–time of flight mass spectrom- etry) from Clostridium phoceensis strain GD3T.

of total information was obtained from a 653K/mm2 cluster Results density, with a cluster passing quality control filters of 96.1% (12 031 000 clusters). Within this run, the index representation fi for C. phoceensis GD3T was determined to 9.32%. The Strain isolation, identi cation and phylogeny T 1 121 200 paired reads were filtered according to the read MALDI-TOF MS failed to identify the strain GD3 , so its mass qualities. These reads were trimmed, then assembled by SPDES spectrum was added to the Bruker database (Fig. 1). To fi software. Finally, the draft genome of C. phoceensis GD3T improve identi cation, 16S rRNA sequencing was performed, consisted of ten scaffolds with 16 contigs and generated a and the access number in 16S rRNA EMBL-EBI (European – genome size of 3.4 Mb with 59.32% G+C content. Molecular Biology Laboratory European Bioinformatics Insti- tute) was assigned as LN846907. The highest value of nucleo- Genome annotation tide sequence similarity was observed with Flavonifractor plautii Open reading frames (ORFs) were predicted using Prodigal (97%), the phylogenetically closest species. [15] with default parameters, although the predicted ORFs were excluded if they spanned a sequencing gap region. The predicted bacterial protein sequences were searched against the GenBank database [16] and the Clusters of Orthologous Groups (COGs) databases using BLASTP. The tRNA genes were found using the tRNAScanSE tool [17], while RNAmmer [18] was used to find ribosomal RNAs and BLASTn against the GenBank database. Lipoprotein signal peptides and the number of transmembrane helices were predicted using SignalP [19] and TMHMM [20] respectively. Artemis [21] was used for data management, and DNA Plotter [22] was used to visualize genomic features. To estimate the mean level of nucleotide sequence similarity at the genome level, we used homemade average genomic identity of orthologous gene sequences (AGIOS) software [4]. FIG. 2. Gram staining of Clostridium phoceensis strain GD3T.

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

236 88 New Microbes and New Infections, Volume 14 Number C, November 2016 NMNI

Phenotypic and biochemical characterization TABLE 1. Classification, general features and biochemical C. phoceensis strain GD3T is a Gram-negative (Fig. 2), spore- tests of Clostridium phoceensis strain GD3T forming, nonmotile, strictly anaerobic bacterium that has no Property Term catalase and oxidase activities, measuring 1.8 μm in length and Current classification Domain: Bacteria 0.5 μm in diameter (Fig. 3). The sporulation test was positive; Phylum: Firmicutes Class: Clostridia the organism grows at 45°C in anaerobic conditions. Using the Order: Clostridiales Family: Clostriadaceae API ZYM gallery, C. phoceensis exhibits esterase (C4), phos- Genus: Clostridium phatase acid, naphthol-AS-BI-phosphohydrolase and β-glucosi- Species: Clostridium phoceensis Type strain: GD3 dase activities. When using the API 20A gallery, a positive Gram stain Negative Cell shape Bacillus reaction was only observed for D-glucose (Table 1). Cell diameter (μm) 0.5 μm Cell length 1.8 μm Of the antibiotics tested, C. phoceensis was found to be Motility No Sporulation Yes sensitive to amoxicillin, amoxicillin–clavulanate, ceftriaxone, Temperature range Mesophilic fl Production of: ceftazidime, imipenem, doripenem and cipro oxacin. Alkaline phosphatase No Catalase No Oxidase No Predominant cellular fatty acids Nitrate reductase No Urease No The predominant cellular fatty acids of C. phoceensis strain β-Galactosidase No T N-acetyl-glucosamine No GD3 are hexadecanoic acid (16:0; 33.1 ± 1.7%), 9- Esterase Yes Acid from: octadecenoic acid (18:1n9; 24.3 ± 0.6%), octadecanoic acid L-Arabinose No Ribose No (18:0; 20.0 ± 0.1%), 9,12-octadecadienoic acid (18:2n6; Mannose No Mannitol No 8.6 ± 0.1%), 11-octadecenoic acid (18:1n7; 5.5 ± 0.5%), tetra- Sucrose No decanoic acid (14:0; 4.7 ± 0.8%) and trace amounts (less than D-Glucose Yes D-Fructose No 1%) of heptadecanoic acid, 9-hexadecenoic acid, pentadecanoic D-Maltose No D-Lactose No acid, 15-methyl-hexadecanoic acid, 14-methyl-hexadecanoic Habitat Human acid, 10-heptadecenoic acid, 13-methyl-tetradecanoic acid and 12-methyl-tetradecanoic acid (Table 2).

Genome properties TABLE 2. Cellular fatty acid composition (%) of Clostridium A phylogenetic tree highlighting the position of Clostridium phoceensis strain GD3T T phoceensis GD3 relative to other type strains within the order Fatty acid Name Mean relative %a Clostridiales is provided in Fig. 4. T 16:0 Hexadecanoic acid 33.1 ± 1.7 The genome of the C. phoceensis strain GD3 is 18:1n9 9-Octadecenoic acid 24.3 ± 0.6 18:0 Octadecanoic acid 20.0 ± 0.1 3 453 562 bp long with 59.32% G+C content. A total of 3320 18:2n6 9,12-Octadecadienoic acid 8.6 ± 0.1 18:1n7 11-Octadecenoic acid 5.5 ± 0.5 genes were predicted, of which 3264 were protein-coding 14:0 Tetradecanoic acid 4.7 ± 0.8 17:0 Heptadecanoic acid TR 16:1n7 9-Hexadecenoic acid TR 15:0 Pentadecanoic acid TR 17:0 iso 15-methyl-Hexadecanoic acid TR 17:0 anteiso 14-methyl-Hexadecanoic acid TR 17:1n7 10-Heptadecenoic acid TR 15:0 iso 13-Methyl-tetradecanoic acid TR 15:0 anteiso 12-Methyl-tetradecanoic acid TR

genes and 56 were RNA genes (three genes are 5S rRNA, one gene is 16S rRNA, one gene is 23S rRNA and 51 genes are TRNA genes). A total of 1967 genes (60.26%) were assigned a putative function. A total of 227 genes (6.95%) were identified as ORFans. The remaining genes (28.92%) were annotated as hypothetical proteins. The properties and statistics of the genome are summarized in Table 3 and Fig. 5. The distribution of genes into COGs functional categories is presented in FIG. 3. Electron microscopy of Clostridium phoceensis strain GD3T. Table 4 and Fig. 6.

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

237 NMNI Hosny et al. Clostridium phoceensis sp. nov. 89

FIG. 4. Phylogenetic tree high- lighting position of Clostridium pho- ceensis GD3T relative to other type strains within Clostridiales order. GenBank accession numbers are indicated. Sequences were aligned using CLUSTALW, and phylogenetic inferences were obtained using maximum-likelihood method within MEGA 4 software [23]. Numbers at nodes are bootstrap values obtained by repeating analysis 500 times to generate majority consensus tree. Scale bar represents 5% nucleotide sequence divergence. Rubidus massi- liensis was used as outgroup.

Genome comparison capillosus, Oscillibacter valericigenes, Flavonifractor plautii, Clos- We made some brief comparisons against nine genomes: tridium cellulosi and Intestinimonas butyriciproducens (3.45, 4.24, Intestinimonas butyriciproducens strain ER1 (GenBank accession 4.47, 3.81, 5.68 and 3.57 MB respectively), but larger than those no. JPJD01000000), Flavonifractor plautii ATCC 29863 of Clostridium viride, Ethanoligenens harbinense, Clostridium leptum (AGCK01000000), Clostridium leptum DSM 753 and Eubacterium siraeum (2.41, 3.01, 2.82 and 2.84 MB (ABCB02000000), Clostridium cellulosi DG5 (NZ_LM995447), respectively). The G+C content of C. phoceensis is smaller than Ethanoligenens harbinense YUAN-3 (NC_01482), Oscillibacter those of Flavonifractor plautii (59.32 and 61.07% respectively) valericigenes Sjm18-20 (NC_016048), Clostridium viride DSM but larger than those of Pseudoflavonifractor capillosus, Oscil- 6836 (NZ_JHZO01000000), Eubacterium siraeum V10Sc8a libacter valericigenes, Clostridium viride, Ethanoligenens harbinense, (FP929059) and Pseudoflavonifractor capillosus ATCC 29799 Clostridium leptum, Eubacterium siraeum, Clostridium cellulosi and (NZ_AAXG02000000). The draft genome sequence of Intestinimonas butyriciproducens (59.11, 53.19, 49.28, 55.56, C. phoceensis is smaller than those of Pseudoflavonifractor 50.25, 45.13, 42.07 and 58.44% respectively). The gene content of C. phoceensis is smaller than those of Pseudoflavonifractor capillosus, Oscillibacter valericigenes, Flavonifractor plautii, Clos- TABLE 3. Nucleotide content and gene count levels of tridium cellulosi and Intestinimonas butyriciproducens (3264, 4829, genome 4723, 4278, 5171 and 3529 respectively) but larger than those Genome (total) of Clostridium viride, Ethanoligenens harbinense, Clostridium leptum Attribute Value % of totala and Eubacterium siraeum (2321, 2701, 2482 and 2211 respec-

Genome size (bp) 3 453 562 100 tively). C. phoceensis has a similar distribution of genes into DNA coding region (bp) 2 924 785 84.68 COGs categories with the most of the compared species DNA G+C content (bp) 2 044 366 59.31 Total genes 3320 100 (Fig. 4). However, Clostridium cellulosi was overrepresented for rRNA 5 0.136 tRNA 51 0.116 all the categories, and Clostridium viride was overrepresented for Protein-coding genes 3264 98.31 Genes with function prediction 3111 17.02 category Z (cytoskeleton). Genes assigned to COGs 2534 57.39 Pseudo genes 62 1.86 In addition, Clostridium phoceensis shared 846, 884, 492, 562, Genes in internal clusters 1060 31.92 Genes with Pfam domains 2817 84 482, 717, 634, 407 and 842 orthologous genes respectively with Genes with signal peptides 414 12.68 Genes with transmembrane helices 722 22.12 Intestinimonas butyriciproducens, Flavonifractor plautii, Clostridium ORFan genes 227 6.95 leptum, Clostridium cellulosi, Ethanoligenens harbinense, Oscil- CRISPR repeats 14 0.02 libacter valericigenes, Clostridium viride, Eubacterium siraeum and COGs, Clusters of Orthologous Groups database; CRISPR, clustered regularly interspaced short palindromic repeat. Pseudoflavonifractor capillosus (Table 5). Of these species, the aTotal is based on either size of genome in base pairs or total number of protein- coding genes in annotated genome. orthologous genes shared ranged from 365 between Eubacte-

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

238 90 New Microbes and New Infections, Volume 14 Number C, November 2016 NMNI

FIG. 5. Graphical circular map of Clostridium phoceensis GD3T genome. From outside in, outer two circles show ORFs oriented in forward (colored by COGs categories) and reverse (colored by COGs cate- gories) directions, respectively. Third circle marks tRNA genes (green). Fourth circle shows G+C% content plot. Innermost circle shows GC skew, with purple indicating negative values and olive indicating positive values. COGs, Clusters of Ortholo- gous Groups database; ORF, open reading frame.

rium siraeum and Clostridium viride to 1079 between Flavoni- fractor plautii and Intestinimonas butyriciproducens. Compared to other species, C. phoceensis exhibited AGIOS values ranging TABLE 4. Number of genes associated with 25 general COGs from 56.64 with Ethanoligenens harbinense to 71.58 with Fla- functional categories vonifractor plautii. %of Code Description Value totala

J Translation 150 4.59 Conclusion A RNA processing and modification 0 0 K Transcription 212 6.49 L Replication, recombination and repair 106 3.24 B Chromatin structure and dynamics 0 0 D Cell cycle control, mitosis and meiosis 22 0.67 On the basis of taxonogenomic analyses, we propose Clos- Y Nuclear structure 0 0 T V Defense mechanisms 75 2.29 tridium phoceensis sp. nov., strain GD3 (= CSUR P1929 = DSM T Signal transduction mechanisms 94 2.87 100334), as the type strain of C. phoceensis sp. nov., a new M Cell wall/membrane biogenesis 54 1.65 N Cell motility 27 0.82 species within the genus Clostridium (Fig. 4). This strain was Z Cytoskeleton 0 0 W Extracellular structures 0 0 isolated from the gut microbiota of a 28-year-old healthy U Intracellular trafficking and secretion 25 0.76 O Posttranslational modification, protein turnover, 51 1.56 French man. chaperones C Energy production and conversion 94 2.87 G Carbohydrate transport and metabolism 106 3.24 E Amino acid transport and metabolism 217 6.54 Taxonomic and nomenclatural proposals F Nucleotide transport and metabolism 49 1.50 H Coenzyme transport and metabolism 63 1.93 I Lipid transport and metabolism 53 1.62 P Inorganic ion transport and metabolism 94 2.87 Q Secondary metabolites biosynthesis, transport and 29 0.88 Description of Clostridium phoceensis sp. nov. catabolism T R General function prediction only 231 7.07 Clostridium phoceensis strain GD3 (pho.ce.en’sis, L. fem. adj., S Function unknown 126 3.86 — Not in COGs 1684 51.59 from phoceensis referring to Phocea, the Greek name of the city

COGS, Clusters of Orthologous Groups database. which founded Marseille, where it was isolated) is a Gram- a Total is based on total number of protein-coding genes in annotated genome. negative nonmotile bacillus whose individual cell size is

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

239 NMNI Hosny et al. Clostridium phoceensis sp. nov. 91

FIG. 6. Distribution of functional classes of predicted genes in genomes from Clostridium phoceensis (Cp), Intestinimonas butyriciproducens strain ER1 (Ib), Flavonifractor plautii ATCC 29863 (Fp), Clostridium leptum DSM 753 (Cl), Clostridium cellulosi DG5 (Cc), Ethanoligenens harbinense YUAN-3 (Eh), Oscillibacter valericigenes Sjm18-20 (Ov), Clostridium viride DSM 6836 (Cv), Eubacterium siraeum V10Sc8a (Es) and Pseudoflavonifractor capillosus ATCC 29799 (PC) genomes according to clusters of orthologous groups of proteins. ATCC, American Type Culture Collection (Manassas, VA, USA).

TABLE 5. Orthologous genes shared (upper right) and AGIOS values obtained (lower left)a

Cp Ib Fp Cl Cc Eh Ov Cv Es Pc

Cp 3264 846 884 492 562 482 717 634 407 842 Ib 70.10 3529 1079 508 627 501 771 671 407 1016 Fp 71.58 73.64 4278 523 649 518 818 686 433 1039 Cl 61.83 61.60 61.65 2482 587 495 475 432 405 523 Cc 57.17 57.91 56.82 61.62 5171 619 604 562 465 616 Eh 56.64 56.86 56.92 56.13 55.34 2701 515 449 415 499 Ov 65.72 65.34 65.55 61.60 58.40 56.54 4723 653 409 710 Cv 63.84 64.57 64.21 60.85 59.13 55.02 62.58 2321 364 614 Es 57.83 58.77 57.72 60.78 61.38 54.37 59.13 58.93 2211 427 Pc 61.55 61.95 62.89 55.43 52.37 57.21 58.02 56.73 53.24 4829

AGIOS, average genomic identity of orthologous gene sequences; ATCC, American Type Culture Collection (Manassas, VA, USA). aValues in bold are gene numbers. Ten genomes were used for this study: Clostridium phoceensis (Cp), Intestinimonas butyriciproducens strain ER1 (Ib), Flavonifractor plautii ATCC 29863 (Fp), Clostridium leptum DSM 753 (Cl), Clostridium cellulosi DG5 (Cc), Ethanoligenens harbinense YUAN-3 (Eh), Oscillibacter valericigenes Sjm18-20 (Ov), Clostridium viride DSM 6836 (Cv), Eubacterium siraeum V10Sc8a (Es) and Pseudoflavonifractor capillosus ATCC 29799 (PC) genomes according to the clusters of orthologous groups of proteins.

1.8 μm in length and 0.5 μm in diameter. It is a strictly anaer- and DSMZ numbers are respectively CSUR P1929 and DSM obic and endospore-forming bacterium. Strain GD3T is catalase 100334, was identified from the stool sample of a 28-year-old and oxidase negative, and its optimal growth temperature is 45° healthy French man. C, but it also grows weakly at 37°C. Biochemical analyses showed positive reactions of C. phoceensis for D-glucose and Acknowledgements produced esterase (C4), phosphatase acid, naphthol-AS-BI- phosphohydrolase and β-glucosidase enzymes. C. phoceensis was sensitive to amoxicillin, amoxicillin–clavulanate, ceftriax- The authors thank the Xegen Company (www.xegen.fr) for one, ceftazidime, imipenem, doripenem and ciprofloxacin. Its automating the genomic analysis process and culturomics team predominant cellular fatty acids are hexadecanoic acid (16:0; in URMITE for the bacteria isolation. This project was funded 33.1 ± 1.7%), 9-octadecenoic acid (18:1n9; 24.3 ± 0.6%), octa- by the Fondation Méditerranée Infection. decanoic acid (18:0; 20.0 ± 0.1%), 9,12-octadecadienoic acid (18:2n6; 8.6 ± 0.1%), 11-octadecenoic acid (18:1n7; 5.5 ± 0.5%) Conflict of Interest and tetradecanoic acid (14:0; 4.7 ± 0.8%). Its 16S rRNA se- quences were deposited in GenBank under accession numbers CVUG01000000 and LN846907. Strain GD3T, whose CSUR None declared.

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

240 92 New Microbes and New Infections, Volume 14 Number C, November 2016 NMNI

References [12] Mourembou G, Yasir M, Azhar EI, Lagier JC, Bibi F, Jiman-Fatani AA, et al. Rise of microbial culturomics: noncontiguous finished genome sequence and description of Beduini massiliensis gen. nov., sp. nov. OMICS J Integr Biol 2015;19:766–76. [1] Palmer C, Bik EM, DiGiulio DB, Relman DA, Brown PO. Development [13] Fournier PE, Raoult D, Drancourt M. New species announcement: a of the human infant intestinal microbiota. PLoS Biol 2007;5:e177. new format to prompt the description of new human microbial spe- [2] Lagier JC, Million M, Hugon P, Armougom F, Raoult D. Human gut cies. New Microbes New Infect 2016. In press. microbiota: repertoire and variations. Front Cell Infect Microbiol [14] Field D, Garrity G, Gray T, Morrison N, Selengut J, Sterk P, et al. The 2012;2:136. minimum information about a genome sequence (MIGS) specification. [3] Lagier JC, Edouard S, Pagnier I, Mediannikov O, Drancourt M, Nat Biotechnol 2008;26:541–7. Raoult D. Current and past strategies for bacterial culture in clinical [15] Hyatt D, Chen GL, LoCascio PF, Land ML, Larimer FW, Hauser LJ. microbiology. Clin Microbiol Rev 2015;28:208–36. Prodigal: prokaryotic gene recognition and translation initiation site [4] Ramasamy D, Mishra AK, Lagier JC, Padhmanabhan R, Rossi M, identification. BMC Bioinformatics 2010;11:119. Sentausa E, et al. A polyphasic strategy incorporating genomic data for [16] Benson DA, Karsch-Mizrachi I, Clark K, Lipman DJ, Ostell J, Sayers EW. the taxonomic description of novel bacterial species. Int J Syst Evol GenBank. Nucleic Acids Res 2012;40(Database issue):D48–53. Microbiol 2014;64:384–91. [17] Lowe TM, Eddy SR. tRNAscan-SE: a program for improved detection [5] Lawson PA, Rainey FA. Proposal to restrict the genus Clostridium of transfer RNA genes in genomic sequence. Nucleic Acids Res (Prazmowski) to Clostridium butyricum and related species. Int J Syst 1997;25:955–64. Evol Microbiol 2015. [18] Lagesen K, Hallin P, Rødland EA, Stærfeldt HH, Rognes T, Ussery DW. [6] Popoff MR, Bouvet P. Clostridial toxins. Future Microbiol 2009;4: RNAmmer: consistent and rapid annotation of ribosomal RNA genes. 1021–64. Nucleic Acids Res 2007;35:3100–8. [7] Lagier JC, Armougom F, Million M, Hugon P, Pagnier I, Robert C, et al. [19] Dyrløv Bendtsen J, Nielsen H, von Heijne G, Brunak S. Improved Microbial culturomics: paradigm shift in the human gut microbiome prediction of signal peptides: SignalP 3.0. J Mol Biol 2004;340: study. Clin Microbiol Infect 2012;18:1185–93. 783–95. [8] Grosse-Herrenthey A, Maier T, Gessler F, Schaumann R, Böhnel H, [20] Krogh A, Larsson B, von Heijne G, Sonnhammer ELL. Predicting Kostrzewa M, et al. Challenging the problem of clostridial identification transmembrane protein topology with a hidden markov model: with matrix-assisted laser desorption and ionization–time-of-flight application to complete genomes. J Mol Biol 2001;305:567–80. mass spectrometry (MALDI-TOF MS). Anaerobe 2008;14:242–9. [21] Rutherford K, Parkhill J, Crook J, Horsnell T, Rice P, Rajandream MA, [9] Seng P, Abat C, Rolain JM, Colson P, Lagier JC, Gouriet F, et al. et al. Artemis: sequence visualization and annotation. Bioinformatics Identification of rare pathogenic bacteria in a clinical microbiology 2000;16:944–5. laboratory: impact of matrix-assisted laser desorption ionization–time [22] Carver T, Thomson N, Bleasby A, Berriman M, Parkhill J. DNAPlotter: of flight mass spectrometry. J Clin Microbiol 2013;51:2182–94. circular and linear interactive genome visualization. Bioinformatics [10] Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 16S 2009;25:119–20. ribosomal DNA sequence analysis of a large collection of environ- [23] Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular Evolutionary mental and clinical unidentifiable bacterial isolates. J Clin Microbiol Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 2000;38:3623–30. 2007;24:1596–9. [11] Dione N, Sankar SA, Lagier JC, Khelaifia S, Michele C, Armstrong N, et al. Genome sequence and description of Anaerosalibacter massiliensis sp. nov. New Microbes New Infect 2016;10:66–76.

© 2016 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 14,85–92 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

241 NEW SPECIES

Description of ‘Gorbachella massiliensis’ gen. nov., sp. nov., ‘Fenollaria timonensis’ sp. nov., ‘Intestinimonas timonensis’ sp. nov. and ‘Collinsella ihuae’ sp. nov. isolated from healthy fresh stools with culturomics

G. A. Durand, F. Cadoret, J. C. Lagier, P. E. Fournier and D. Raoult Aix-Marseille Université, URMITE, UM63, CNRS7278, IRD198, Inserm 1095, Institut Hospitalo-Universitaire Méditerranée-Infection, Faculté de médecine, Marseille, France

Abstract

We report here the main characteristics of ‘Gorbachella massiliensis’ GD7T gen. nov., sp. nov., ‘Fenollaria timonensis’ GD5T sp. nov., ‘Intestinimonas timonensis’ GD4T sp. nov., and ‘Collinsella ihuae’ sp. nov. GD8T isolated from one fresh stool of a French volunteer. We used a bacterial culturomics approach combined with taxono-genomics. © 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases.

Keywords: Collinsella ihuae, fenollaria timonensis, gorbachella massiliensis, gut microbiota, intestinimonas timonensis’ Original Submission: 9 December 2016; Revised Submission: 5 January 2017; Accepted: 9 January 2017 Article published online: 16 January 2017

catalase or oxidase activity. The 16S rRNA gene was sequenced Corresponding author: D. Raoult, Aix-Marseille Université, using fD1-rP2 primers as described previously using a 3130-XL URMITE, UM63, CNRS7278, IRD198, Inserm 1095, Institut Hospitalo-Universitaire Méditerranée-Infection, Faculté de médecine, sequencer (Applied Biosciences, Saint Aubin, France) [3]. The 27 Boulevard Jean Moulin, 13385, Marseille cedex 05, France strain GD7T had a 16S rRNA gene sequence identity of 93.4% E-mail: [email protected] with Subdoligranulum variabile strain BI 114T (NR_028997), the phylogenetically closest species with standing in nomenclature (Fig. 1). This similarity <98.65% leads us to putatively classify In our study concerning the intolerant oxygen species from the GD7T as a new member in the Ruminococcaceae family of Fir- human gut microbiota, we isolated four bacteria in 2015 using a micutes [4]. Therefore we propose the creation of the new bacterial culturomics approach. These bacteria could not be genus ‘Gorbachella’ (Gor.ba.chel’la. NL gen fem, in honour of identified by matrix-assisted laser desorption/ionization mass the microbiologist Sherwood Gorbach of the Tufts University spectrometry (MALDI-TOF MS; http://www.mediterranee- School of Medicine, Boston, MA, USA). GD7T is the type strain infection.com/article.php?laref=256&titre=urms-database (last of the species Gorbachella massiliensis (ma.ssi.li.en’sis L. adj. fem access 01/30/17)) on a Microflex spectrometer (Bruker Dal- to Massilia, the Latin name of Marseille, France, where this tonics, Bremen, Germany) [1,2]. These species were isolated strain was isolated). from the same fresh stool from a healthy volunteer. The indi- Strain GD5T was isolated from the fresh sample after 48 h vidual has signed informed consent and the study has been anaerobic growth on Columbia agar supplemented with 5% validated by the Ethics Committee of the IFR48 Federative sheep blood at 37°C. The colonies appeared to be translucent, Research Institute under the number 09-022. rough, non-haemolytic, motile, non-spore-forming, and 1 mm in Strain GD7T was isolated from a dilution of the fresh sample. size. The cells were rod-shaped with Gram-negative staining. The species was grown after 48 h on Columbia agar supple- Oxidase and catalase activities were negative. Strain GD5T mented with 5% sheep blood at 37°C under strict anaerobic showed 97.4% sequence homology with the 16S RNA of conditions. The colonies appeared translucent and rough, non- Fenollaria massiliensis strain 9401234T (NR_133038) (Fig. 2) [6]. haemolytic, non-motile, nonspore-forming, and 1 mm size. The So we propose to classify GD5T as a new species within the cells were Gram-negative, rod-shaped. The strain did not show genus Fenollaria in the phylum Firmicutes [4]. GD5T is the type

New Microbe and New Infect 2017; 16: 60–62 © 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) http://dx.doi.org/10.1016/j.nmni.2017.01.005 242 NMNI Durand et al. G. massiliensis, F. massiliensis, I. timonensis, and C. ihuae 61

FIG. 1. Phylogenetic tree showing the position of ‘Intestinimonas timonensis’ GD4T and ‘Gorbachella massiliensis’ GD7T relative to other phylogenetically close neighbours. Sequences were aligned using CLUSTALW, and phylogenetic inferences were obtained with Kimura two-parameter models using the maximum-likelihood method within the MEGA software. Numbers at the nodes are percentages of bootstrap values obtained by repeating the analysis 1000 times to generate a majority consensus tree. Only bootstrap values >95% are displayed. The scale bar indicates a 2% nucleotide sequence divergence.

FIG. 2. Phylogenetic tree showing the position of ‘Fenollaria timonensis’ GD5T relative to other phylogenetically close neighbours. Alignment and phylogenetic inferences were made as described for Fig. 1.

© 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 16,60–62 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

243 62 New Microbes and New Infections, Volume 16 Number C, March 2017 NMNI

FIG. 3. Phylogenetic tree showing the po- sition of ‘Collinsella ihuae’ GD8T relative to other phylogenetically close neighbours. Alignment and phylogenetic inferences were made as described for Fig. 1. The scale bar indicates a 1% nucleotide sequence divergence.

strain of the species ‘Fenollaria timonensis’ (ti.mo.nen’sis L. adj. timonensis’ GD4T (LN870298) and ‘Collinsella ihuae’ GD8T fem to Timone, the name of the main hospital of Marseille, (LN881598). France, where this strain was isolated). Deposit in a culture collection. The strains were Strain GD4T was isolated after 48 h anaerobic growth on deposited in the Collection de Souches de l’Unité des Rick- Columbia agar supplemented with 5% sheep blood and 5% ettsies (CSUR, WDCM 875) under the following numbers: rumen fluid at 37°C. The colonies appeared translucent, rough, ‘Gorbachella massiliensis’ GD7T (P2021), ‘Fenollaria timonensis’ non-haemolytic, motile, non-spore-forming, and 1 mm in size. GD5T (P2133), ‘Intestinimonas timonensis’ GD4T (P2010) and The cells were Gram-negative. Catalase and oxidase activities ‘Collinsella ihuae’ GD8T (P2019). were negative. Strain GD4T presented a sequence identity of 97.08% with 16S rRNA sequence of Intestinimonas butyr- Transparency declaration iciproducens DSM 26588T (NR_118554), the closest species with a valid name (Fig. 1). We propose to putatively classify GD4T as a new member of the genus Intestinimonas in the The authors have no conflicts of interest to declare. phylum Firmicutes [4]. GD4T is the type strain of the species ‘Intestinimonas timonensis’ (ti.mo.nen’sis L. adj. fem to Timone, the name of the main hospital of Marseille, France, where this Funding strain was isolated). T ‘Collinsella ihuae’ strain GD8 was isolated after 48 h of This work was funded by the Fondation Mediterrannée- anaerobic growth on Columbia agar supplemented with 5% Infection. sheep blood and 5% rumen fluid at 37°C. Colonies appeared as microcolonies rough, non-haemolytic, motile, non-spore- References forming, and 0.5 mm in size. The cells were rod-shaped with Gram-positive staining. Catalase and oxidase activities were negative. The 16S rRNA sequence of the strain GD8T pre- [1] Lagier J-C, Hugon P, Khelaifia S, Fournier P-E, La Scola B, Raoult D. The rebirth of culture in microbiology through the example of culturomics sented an identity of 96.2% with the 16S rRNA sequence of to study human gut microbiota. Clin Microbiol Rev 2015;28:237–64. Collinsella tanakaei strain JCM 16071 (NR_113273), the closest [2] Lagier J-C, Khelaifia S, Alou MT, Ndongo S, Dione N, Hugon P, et al. phylogenetic species with nomenclature (Fig. 3) [5]. Then we Culture of previously uncultured members of the human gut microbiota propose the creation of the new species ‘Collinsella ihuae’ within by culturomics. Nat Microbiol 2016;1:16203. T [3] Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP, Raoult D. 16S the phylum Actinobacteria [4]. GD8 is the type strain of the ribosomal DNA sequence analysis of a large collection of environmental species ‘Collinsella ihuae’ (i.hu.ae L. adj. fem to Institut Hospitalo- and clinical unidentifiable bacterial isolates. J Clin Microbiol 2000;38: Universitaire (IHU), the name of the name of the laboratory 3623–30. [4] Kim M, Oh H-S, Park S-C, Chun J. Towards a taxonomic coherence (Marseille, France) where this strain was isolated). between average nucleotide identity and 16S rRNA gene sequence MALDI-TOF-MS spectra accession numbers. The similarity for species demarcation of prokaryotes. Int J Syst Evol MALDI-TOF-MS spectra of these species are available at http:// Microbiol 2014;64:346–51. [5] Pagnier I, Croce O, Robert C, Raoult D, La Scola B. Non-contiguous mediterranee-infection.com/article.php? laref=256&titre=urms- finished genome sequence and description of Fenollaria massiliensis gen. database. Last access 7 December 2016. nov., sp. nov., a new genus of anaerobic bacterium. Stand Genomic Sci Nucleotide sequence accession number. The 16S r 2014;9:704–17. RNA gene sequences were deposited in GenBank under [6] Nagai F, Watanabe Y, Morotomi M. Slackia piriformis sp. nov. and Collinsella tanakaei sp. nov., new members of the family Coriobacter- ‘ ’ T accession numbers: Gorbachella massiliensis GD7 (LN870316), iaceae, isolated from human faeces. Int J Syst Evol Microbiol 2010;60: ‘Fenollaria timonensis’ GD5T (LN881613), ‘Intestinimonas 2639–46.

© 2017 The Authors. Published by Elsevier Ltd on behalf of European Society of Clinical Microbiology and Infectious Diseases, NMNI, 16,60–62 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

244 NOTE Togo et al., Int J Syst Evol Microbiol DOI 10.1099/ijsem.0.001826

Fournierella massiliensis gen. nov., sp. nov., a new human- associated member of the family Ruminococcaceae

Amadou Hamidou Togo,1 Guillaume Durand,1 Saber Khelaifia,1 Nicholas Armstrong,1 Catherine Robert,1 Fred eric Cadoret,1 Fabrizio Di Pinto,1 Jer emy Delerce,1 Anthony Levasseur,1 Didier Raoult1,2 and Matthieu Million1,*

Abstract An anaerobic bacterium, strain AT2T, was isolated from the fresh stool sample of a healthy French man using the culturomics approach. The 16S rRNA gene sequence analysis showed that strain AT2T had 95.2 % nucleotide sequence similarity with Gemmiger formicilis ATCC 27749T, the phylogenetically closest species with standing in nomenclature. Cells are Gram-stain-negative, catalase- and oxidase-negative, obligately anaerobic, non-motile, non-spore-forming, rod-shaped,

and the bacilli were mesothermophilic. The major fatty acids were C16 : 0 (43.8 %) and C18 : 1n9 (20 %). The DNA G+C content of the strain based on its genome sequence was 56.8 mol%. Based on the phenotypic, biochemical and phylogenetic analysis, we propose the creation of the genus Fournierella gen. nov., which contains strain AT2T (=CSUR P2014T=DSM 100451T) as the type strain of the type species Fournierella massiliensis gen. nov., sp. nov.

Culturomics is a new approach for the characterization of extremely oxygen-sensitive bacterium that is difficult to living microbial diversity in any environmental or human cultivate, even in anaerobic conditions [6]. F. prausnitzii sample [1]. With the development of new technologies such sustains growth in the presence of low partial pressure as high-throughput sequencing enabling public access to the of oxygen, in presence of antioxidants [7], and showed complete genome sequences of many bacterial species, we mutualism with epithelial cells, possibly through mucin proposed the inclusion of the complete genome sequence [8]. F. prausnitzii is one of the leading representatives of analysis in a polyphasic approach to describe new bacterial the human healthy mature anaerobic gut microbiota taxa [2]. This strategy, which we named taxono-genomics, (HMAGM), suggesting the link between dietary antioxi- combines phenotypic characteristics, notably the matrix dants and maintenance of the HMAGM [9, 10]. It con- assisted laser desorption/ionization time-of-flight (MALDI- tributes to maintaining host–microbial homeostasis by TOF) MS spectrum and genomic properties [3, 4]. secreting a microbial anti-inflammatory molecule that inhibits cellular NF-kB signalling and inflammation [11]. During an exploratory study of fresh stool by culturo- Changes in the abundance of F. prausnitzii have been mics [1], an isolate was obtained and a new genus was linked to dysbiosis in several human disorders [12]. proposed to accommodate this strain as a member of Here, we propose the main phenotypic, phylogenetic and the family Ruminococcaceae [5]. At the time of writing, genomic properties of strain AT2T (=CSUR P2014T the family Ruminococcaceae contained 17 genera, includ- =DSM 100451T), that is close to but substantially differs ing Acetanaerobacterium, Acetivibrio, Anaerobacterium, from Gemmiger formicilis. Anaerofilum, Anaerotruncus, Ercella, Ethanoligenens, Fae- calibacterium, Fastidiosipila, Gemmiger, Hydrogenoanaero- Strain AT2T was isolated from a fresh stool sample collected bacterium, Oscillibacter, Oscillospira, Papillibacter, from a healthy 28-year-old French man in January 2015.  Ruminococcus, Sporobacter and Subdoligranulum (www. The stool sample was immediately stored at 4 C after col- bacterio.net/ruminococcaceae.html). Among members of lection until being used for culture. The donor gave a writ- the family, Faecalibacterium prausnitzii is one of the ten informed consent and the study was validated by the most abundant bacteria of the human gut. It is an ethics committee of the Institut Federatif de Recherche

Author affiliations: 1URMITE, UM63, CNRS7278, IRD198, INSERM1095, Institut Hospitalier Universitaire Mediterran ee-Infection, Aix Marseille Universite, Marseille, France; 2Special Infectious Agents Unit, King Fahd Medical Research Center, King Abdulaziz University, Jeddah, Saudi Arabia. *Correspondence: Matthieu Million, [email protected] Keywords: ; Fournierella massiliensis; ; taxonogenomics; ; culturomics; ; gut microbiota; ; human microbiome; ; anaerobic bacteria. Abbreviations: dDDH, digital DNA-DNA hybridization; FAME, fatty acid methyl esters; MALDI-TOF, matrix-assisted laser desorption/ionization time-of- flight; ML, maximum-likelihood; SCFA, short-chain fatty acid. The GenBank/EMBL/DDBJ accesion numbers for the genome sequence and 16S rRNA gene sequence of strain AT2T are FAUK00000000 and LN846908, respectively. Four supplementary figures and five supplementary tables are available with the online Supplementary Material.

001826 ã 2017 IUMS 245 1 Togo et al., Int J Syst Evol Microbiol

 IFR48 under agreement number 09–022. To isolate the magnification. A thermic shock at 80 C for 20 min on fresh novel strain, 1 g stool sample was injected in an anaerobic colonies of the strain was carried out in order to test sporu- blood culture bottle (BACTEC Lytic/10 Anaerobic/F Cul- lation. The viability of cells was checked by subculturing ture Vials) supplemented with 4 ml filter-sterilized rumen them on the same media before heating, while the motility  fluid and 5 % sheep blood, and then incubated at 37 C. of strain AT2T was tested observing fresh colonies using a After incubation for 3 days, 100 µl culture suspension was DM1000 photonic microscope (Leica Microsystems) with a collected, plated on 5 % sheep blood-enriched Columbia Â100 oil-immersion objective lens. Catalase (BioMerieux)  agar (BioMerieux) and incubated at 37 C in an anaerobic activity was determined in 3 % hydrogen peroxide solution, atmosphere for 48 h. Emerging colonies were subcultured and oxidase activity was assessed using an oxidase reagent individually for purification using the same conditions, and (Becton Dickinson). Biochemical properties of the strain identified by MALDI-TOF MS as described by were investigated using API ZYM, 20A and 50CH strips Seng et al. [13]. Isolates were re-streaked three times and (BioMerieux) according to the manufacturer’s instructions. purity was confirmed by direct examination and MALDI- The antibiotic susceptibility of strain AT2T was tested using TOF MS analysis. Purity was confirmed when all 12 spots the disc diffusion method [25]; the results are shown in the from 12 different colonies yielded 12 perfectly matching supplementary data. Cellular fatty acid methyl ester MALDI-TOF MS spectra. When the strain was not identi- (FAME) analysis was performed by GC/MS. Two samples fied by MALDI-TOF MS, its 16S rRNA gene sequence was were prepared with approximately 5 mg bacterial biomass assessed as previously reported [14] using the fD1-rP2 pri- per tube harvested from several culture plates. FAMEs were mers, a GeneAmp PCR System 2720 thermal cycler prepared as described by Sasser [26]. GC/MS analyses were (Applied Biosystems) and an ABI Prism 3130-XL capillary carried out as described before [27]. Briefly, FAMEs were sequencer (Applied Biosciences). separated using an Elite 5 MS column and monitored by For taxonomic assignment, the Chromas Pro 1.34 software mass spectrometry (Clarus 500 - SQ 8 s; Perkin Elmer). (Technelysium Pty.) was used to correct sequences. Pairwise Spectral database search was performed using MS Search sequence similarities were calculated using the method rec- 2.0 operated with the Standard Reference Database 1A ommended by Meier-Kolthoff [15] for the 16S rRNA gene (NIST) and the FAMEs mass spectral database (Wiley). sequence, available via the GGDC web server [16] and at For the analysis of short-chain fatty acid (SCFA) produc- http://ggdc.dsmz.de/. Phylogenies were inferred by the tion, a Wilkins-Chalgren-Anaerobe broth supplemented GGDC web server using the DSMZ phylogenomics pipeline with cysteine and DTT (reduced WCA broth described pre- [17] adapted to single genes. A multiple sequence alignment viously by Kl€aring et al. and Pfeiffer et al. [28, 29]) and pre- was created with MUSCLE [18]. A maximum-likelihood (ML) pared using strictly anaerobic techniques (100 % N ) was tree was inferred from the alignment with RAxML [19]. 2 used. Samples were collected at 24, 48 and 72 h after inocu- Rapid bootstrapping in conjunction with the autoMRE boot lation. Acetic, propanoic, butanoic, isobutanoic, pentanoic, stopping criterion [20] and subsequent search for the best hexanoic and heptanoic acids were purchased from Sigma tree was used. The sequences were checked for a composi- 2 Aldrich. A stock solution was prepared in water/methanol tional bias using the X test as implemented in PAUP* [21]. (50 %, v/v) at 0.1 M for each SCFA and then stored at –   Different growth temperatures (25, 28, 37, 45 and 55 C) 20 C. Calibration standards were freshly prepared in water: were tested. Growth of the strain was tested under anaerobic 0.05, 0.5, 1, 5 and 10 mM. Culture samples, prepared in and microaerophilic conditions using GENbag Anaer and duplicate, were centrifuged 5 min at 13 000 r.p.m. and the GENbag Microaer systems, respectively (BioMerieux), and supernatants were collected. All solutions were adjusted to under aerobic conditions, with or without 5 % CO2. Trans- pH 2–3 with HCl before injection. SCFAs were measured mission electron microscopy of the strain, using a Tecnai with a Clarus 500 chromatography system connected to a G20 (FEI company) at an operating voltage of 60kV, was SQ8s mass spectrometer (Perkin Elmer). Analysis was per- performed after negative staining. The pH for growth of the formed with an Elite FFAP column (30 m, 0.25 mm id, strain was tested at a range from pH 6 to 8.5. Tolerance to 0.25 µm film thickness) such as detailed previously [30]. – À1  NaCl was tested using a range of 5 100 g l NaCl on Injection volume was 0.5 µl (split less, 200 C). Helium was  À Schaedler agar with 5 % sheep blood (BioMerieux) in supplied at 1 ml min 1 as the carrier gas. Compounds were an anaerobic atmosphere. Gram staining was performed separated according to a linear temperature gradient from   À and observed using a Leica DM 2500 photonic microscope 100 to 200 C at 8 C min 1. Selected ion recording mass  (Leica Microsystems) with a 100 oil immersion lens. In spectrometry SCFA analysis by GC/MS was performed addition to Gram staining, the KOH test was carried out to using the following masses: 43 m/z for isobutanoic acid; 60 confirm the cell-wall type according to the procedures m/z for acetic, butanoic, pentanoic, hexanoic and heptanoic described elsewhere [22, 23]. Staining was also performed acids; 74 m/z for propanoic acid. The transfer line and the  under anaerobic conditions as differences can occur with electron impact source were set at 200 C. Quadratic calibra- exposure to oxygen [24]. tion curves were automatically fitted with an acceptable Motility of the bacterium was assessed using a Leica DM coefficient of determination above 0.999 and deviation 1000 photonic microscope (Leica Microsystems) at Â100 below 20 % (Turbo mass 6.1, Perkin Elmer). SCFA

246 2 Togo et al., Int J Syst Evol Microbiol

 quantities in samples were presented after subtraction of the 28, 37 and 45 C but optimal growth was observed in an  quantities found in the blank samples. anaerobic atmosphere at 37 C after incubation for 48 h. No  growth was obtained at 55 C or in an aerobic atmosphere. Genomic DNA (gDNA) of the strain was sequenced using a Cells of strain AT2T were Gram-stain-negative (confirmed MiSeq sequencer (Illumina) with the mate-pair strategy. by the KOH test and under anaerobic conditions), rod- The gDNA was barcoded in order to be mixed with 11 other shaped, measured 0.5 µm in diameter and 2 µm in length projects with the Nextera Mate Pair sample prep kit [31]. (Fig. 2), were non-motile, non-spore-forming and without The gDNA was quantified by a Qubit assay with the high À catalase and oxidase activities. The strain grew at a pH rang- sensitivity kit (Life technologies) to 72.2 ng µl 1. The mate- ing from 6 to 8.5, with optimal growth at pH 7.0 to 7.3. No pair library was prepared with 1.5 µg genomic DNA using growth was observed on Schaedler agar enriched with 5 % À the Nextera mate-pair Illumina guide. The gDNA sample sheep blood at 10 g l 1 NaCl. The main characteristics of was simultaneously fragmented and tagged with a mate-pair strain AT2T compared to the closest species are shown in junction adapter. The fragmentation pattern was validated Table 1. The classification and general features of strain on an Agilent 2100 BioAnalyzer (Agilent Technologies) AT2T are summarized in Table S1. Analysis of the total cel- with a DNA 7500 labchip. The DNA fragments ranged in lular fatty acid composition demonstrated that the major size from 1.5 kb up to 11 kb with a maximum at 6.7 kb. No fatty acid was the saturated acid C16 : 0 (43.8 %) followed by size selection was performed and 412 ng of tagmented frag- the unsaturated acid C18 : 1n9 (20 %). Values represent the ments were circularized. The circularized DNA was GC area percentage from total identified fatty acid methyl mechanically sheared to small fragments with a maximum esters only (aldehydes, dimethyl acetates and unidentified at 1033 bp on the Covaris device S2 in T6 tubes (Covaris). ‘summed features’ described previously were not included). The library profile was visualized on a High Sensitivity Bioa- Cellular fatty acid profiles of strain AT2T compared with nalyzer LabChip (Agilent Technologies) and the final con- À other closely related species are summarized in Table 2. centration library was measured at 24.1 nmol l 1. The T libraries were normalized at 4 nM and pooled. After a dena- Strain AT2 produced SCFAs after 24, 48 and 72 h of cul- turation step and dilution, the pool of libraries was loaded ture in reduced WCA broth. After 72 h, the production of onto the reagent cartridge and then onto the instrument acetic acid was predominant (>10 mM), higher than buta- along with the flow cell. Automated cluster generation and noic (6.0±0.3 mM), isobutanoic (2.4±0.1 mM), propanoic sequencing run were performed in a single 2Â251 bp run. (0.6±0.1 mM), pentanoic (0.1±0.1 mM), isopentanoic and isohexanoic acids (the last two were not quantified). Hexa- T MALDI-TOF MS failed to identify strain AT2 at the genus noic and heptanoic acids were not detected. and species levels. Therefore, its spectrum was added to our T database to improve its content. For phylogenetic analysis, The draft genome of strain AT2 was deposited in EMBL- the input nucleotide matrix comprised 11 operational taxo- EBI under accession number FAUK00000000 (Fig. S4), it is nomic units and 1572 characters, 405 of which were variable 3 829 842 bp long with a G+C content of 56.8 %. It is com- and 267 of which were parsimony-informative. The base- posed of 19 scaffolds (27 contigs). Of the 3632 predicted frequency check indicated no compositional bias (P=0.99, genes, 3553 were protein-coding genes, and 79 were RNAs a=0.05). ML analysis under the GTR+Gamma model (one 16S rRNA, four 23S rRNA, six 5S rRNA, 68 tRNAs). A yielded a highest log likelihood of À5973.03, whereas the total of 2514 genes (70.7 %) were assigned a putative func- estimated alpha parameter was of 0.17. The ML bootstrap- tion by COGs or NR BLAST comparison. A total of 298 genes ping converged after 100 replicates; the average support was (8.4 %) were identified as ORFans. Using ARG-ANNOT of 87.75 %. [33], two genes (0.06 %) associated with resistance were detected and seven genes (0.20 %) associated to PKS or The 16S rRNA gene sequencing showed that the strain NRPS [34] were discovered through genome analysis. Using T AT2 exhibited 95.2 % nucleotide sequence similarity with PHAST and RAST, 1799 genes (50.6 %) were associated to Gemmiger formicilis, the phylogenetically closest species mobilome elements. The remaining genes (616) were anno- with standing in nomenclature [32]. The resulting phyloge- tated as hypothetical proteins (Table S2). The distribution netic tree highlighting the position of strain AT2T with the of genes into COGs functional categories is shown in phylogenetically closest species with a validly published Table S3. Considering closest species with available genome, name is shown in Fig. 1 (see also Figs S1 and S2, available in the digital DNA–DNA hybridization (dDDH) values ranged the online Supplementary Material), and strain AT2T was from 23.4 % with Subdoligranulum variabile BI 114T to thus classified in the family Ruminococcaceae [5]. Differen- 36.6 % with F. prausnitzii ATCC 27768T (Table S4). ces in MALDI-TOF MS spectra between strain AT2T and Compared with the closest phylogenetic species (Fig. 1), the other closely related species with available spectrum are pre- phylogenetic distance between strain AT2T and its closest sented in Fig. S3. neighbour was superior to the distance between G. formicilis Colonies of strain AT2T obtained on 5 % sheep blood- X2-56T and S. variabile BI 114T, between G. formicilis X2- enriched Columbia agar (BioMerieux) were translucent 56T and F. prausnitzii ATCC 27768T and between S. varia- with a diameter of 0.3 to 1 mm. Growth of the strain was bile BI 114T and F. prausnitzii ATCC 27768T. Phylogenom- observed in anaerobic and microaerophilic atmospheres at ics analysis was not possible given the unavailability of the

247 3 Togo et al., Int J Syst Evol Microbiol

Ruminococcus callidus VPI 57-31T (NR_029160)

99/100 Butyricicoccus pullicaecorum 25-3T (EU410376)

Clostridium methylpentosum R2T (NR_029355)

Ethanoligenens harbinense YUAN-3T (NR_042828) 96/79

Acetanaerobacterium elongatum Z7T (NR_042930)

Anaerofilum agile FT (NR_029315) 100/100

Anaerofilum pentosovorans FaeT (NR_029313) 99/100 Fournierella massiliensis AT2T (LN846908)

98/100 Faecalibacterium prausnitzii ATCC 27768T (NR_028961) 99/94 Gemmiger formicilis X2-56T (NR_104846) 0.02 99/100

Subdoligranulum variabile BI 114T (NR_028997)

Fig. 1. ML tree inferred under the GTR+GAMMA model and rooted by midpoint-rooting. The branches are scaled in terms of the expected number of substitutions per site. umbers above the branches are support values when larger than 60 % from ML (left) and MP (right) bootstrapping.

closest species’ genomes, but genomic comparisons (dDDH and AGIOS, reported in Tables S4 and S5, respectively) con- firmed that the similarities between strain AT2T and the closest species are in accordance with the proposition of a new genus in the family Ruminococcaceae. Moreover, the G+C% difference exceeded 1 % (À2.2 % compared with G. formicilis X2-56T, +4.6 % compared with S. variabile BI 114T). According to Qin et al. [35], a strain from a new genus will have less than 50 % pairwise percentage of con- served proteins with its closest phylogenetic neighbours. This percentage was 25.4 % (902/3553) with F. prausnitzii ATCC 27768T and 29.4 % (1045/3553) with S. variabile BI 114T, confirming strain AT2T as a member of a new genus (Table S5). The genome of G. formicilis was not available but the 16S rRNA gene phylogenetic distance between strain AT2T and G. formicilis X2-56T was very similar to that of F. prausnitzii ATCC 27768T (Fig. 1). The 16S rRNA gene simi- larity values further support the proposal of a novel genus. Indeed, Yarza et al. [36] reported a median sequence iden- tity of 96.4 % (95 % confidence interval 96.2 to 96.55) to dis- tinguish two genera. This confirms our strain as a new genus (95.2 % 16S rRNA gene sequence similarity with G. formicilis X2-56T, its closest phylogenetic neighbour). Kim T et al. [37] also confirmed a taxonomic coherence between Fig. 2. Transmission electron micrograph of strain AT2 , obtained using a Tecnai G20 (FEI company) at an operating voltage of 60kV. genomic and 16S rRNA gene sequence similarity for taxo- Bar, 100 nm. nomic demarcation of prokaryotes.

248 4 Togo et al., Int J Syst Evol Microbiol

By comparison with reference strains of other closely related Fournier for his contribution to the taxono-genomic species (Table 1), strain AT2T differed in the combination description of the bacteria). of nitrate reductase activity (presence), use of L-arabinose Cells are Gram-negative-staining and the non-motile bacilli (absence), and production of SCFA (acetic acid was the are 0.5 µm in diameter and 2 µm in length, and anaerobic. major SCFA produced, only a small amount of butyric acid  Optimal growth is observed at 37 C and pH tolerance ranges was produced). Interestingly, strain AT2T produced acetic from pH6–8.5. Cells do not produce catalase and oxidase. acid while the reference strain of F. prausnitzii (ATCC 27768T) consumed it. Moreover, MALDI-TOF analysis did The type species is Fournierella massiliensis. not allow identifying previously known species. These phe- notypic differences along with genomic and phylogenetic findings led us to propose that strain AT2T (=CSUR P2014T DESCRIPTION OF FOURNIERELLA =DSM 100451T) is the representative strain of a novel spe- MASSILIENSIS SP. NOV. cies of a new genus within the family Ruminococcaceae for Fournierella massiliensis (mas.si.li.en¢sis. L. fem. adj. massi- which we propose the name Fournierella massiliensis gen. liensis of Massilia, the Latin name of Marseilles). nov., sp. nov. In addition to the characteristics given in the genus This bacterium was isolated from the faeces of a 28-year-old description, colonies grown on 5 % sheep blood-enriched healthy French man living in Marseilles, France and may Columbia agar are white with a diameter of 1 mm. Unable have a beneficial role in the gut through butyrate produc- to produce indole. Using an API 50CH strip (BioMerieux), tion. Butyrate is the preferred energy source for colonic epi- positive reactions are observed for glycerol, D-galactose, D- thelial cells and is thought to play an important role in glucose, D-fructose, D-mannose, methyl a-D-glucopyrano- maintaining colonic health in humans [38]. Moreover, the side, aesculin ferric citrate, salicin, maltose, lactose, meli- production of a significant amount of acetate promotes fur- biose, sucrose, D-melezitose, rafinose, turanose and ther the butyrate production in the gut since fifty percent of potassium 5-ketogluconate. Negative reactions are observed the butyrate-producing isolates are net acetate consumers for erythritol, D-arabinose, L-arabinose, D-ribose, D-xylose, during growth, probably because they employ the butyryl L-xylose, D-adonitol, methyl b-D-xylopyranoside, L-sorbose, coenzyme A-acetyl coenzyme A1 transferase pathway for L-rhamnose, dulcitol, inositol, D-mannitol, D-sorbitol, butyrate production [39]. methyl a-D-mannopyranoside, N-acetylglucosamine, amyg- FOURNIERELLA dalin, arbutin, cellobiose, trehalose, inulin, starch, glycogen, DESCRIPTION OF GEN. NOV. xylitol, gentiobiose, D-xylose, D-tagatose, D-fucose, L-fucose, Fournierella (Four.nier.el¢la. N.L. fem. n. Fournierella named D-arabitol, L-arabitol, potassium gluconate and potassium after the French clinical microbiologist Pierre-Edouard 2-ketogluconate.

Table 1. Differential characteristics of strain AT2T compared with other cloesly related species Strains: 1, AT2T (data from this study); 2, Gemmiger formicilis X2-56T (=ATCC 27749T) [40 – Salanitro et al., 1976); 3, Subdoligranulum variabile BI 114T (=DSM 15176T) [41]; 4, Faecalibacterium prausnitzii ATCC 27768T [6]; 5, Anaerofilum pentosovorans FaeT (=DSM 7168T) [42]; 6, Anaerofilum agile strain T T F (=DSM 4272 ) [42]. All strains were strict anaerobes. +, Positive; À, negative; NA, data not available; v, variable; W, weak.

1 2 3 4 5 6

Cell diameter (µm) 0.5–2 0.3–1 0.6–2.5 0.5–0.9/2–14 0.2–0.6/3–6 0.2–0.6/3–6 Gram stain À v ÀÀ v v Motility ÀÀÀÀ v +

Production of nitrate reductase + ÀÀÀ NA NA Utilization of: L-Arabinose À NA NA NA + + D-Mannose + NA + NA + + D-Mannitol À NA À NA + + Maltose + + + W + + Short chain fatty acid production Acetate Major Minor Minor Utilization Major Major Formate NA Major NA Major Major Major Lactate NA NA Major Major Major Major Propionate Minor NA NA NA ÀÀ Butyrate Minor Major Major Major ÀÀ Isolation source Human faeces Human faeces Human faeces Human faeces Industrial wastewater bioreactor Sewage sludge

249 5 Togo et al., Int J Syst Evol Microbiol

Table 2. Cellular fatty acid profile of strain AT2T compared with other References related species 1. Lagier JC, Khelaifia S, Alou MT, Ndongo S, Dione N et al. Culture T of previously uncultured members of the human gut microbiota Strains: 1, AT2 (data from this study); 2, Subdoligranulum variabile BI by culturomics. Nat Microbiol 2016;1:16203. 114T [41]; 3, Intestinimonas butyriciproducens SRB-521–5-IT [28]; 4, T T T 2. Ramasamy D, Mishra AK, Lagier JC, Padhmanabhan R, Rossi M Ethanoligenens harbinense YUAN-3 (=JCM 12961 =CGMCC 1.5033 ) et al. A polyphasic strategy incorporating genomic data for the (Xing, et al., 2004); 5, Hydrogenoanaerobacterium saccharovorans taxonomic description of novel bacterial species. Int J Syst Evol T T T SW512 (=AS 1.5070 =JCM 14861 ) [43]; 6, Acetanaerobacterium elon- Microbiol 2014;64:384–391. T T T gatum Z7 (=JCM 12359 =AS 1.5012 ) [44]. Values are % of total fatty et al. T 3. Ramasamy D, Kokcha S, Lagier JC, Nguyen TT, Raoult D acids. Only cellular fatty acid >1 % in strain AT2 are included. NA, Not Genome sequence and description of Aeromicrobium massiliense available. sp. nov. Stand Genomic Sci 2012;7:246–257. et al. Fatty acid 1 2 3 4 5 6 4. Lagier JC, Elkarkouri K, Rivet R, Couderc C, Raoult D Non contiguous-finished genome sequence and description of Senegal- emassilia anaerobia gen. nov., sp. nov. Stand Genomic Sci 2013;7: C 11.8 6.2 67.4 NA 15. 6 NA 14 : 0 343–356. C 1.0 <1 NA NA NA NA 15 : 0 5. Rainey FA. Family VIII. Ruminococcaceae fam. nov. In: De Vos P, C16 : 0 43.8 33.0 3.9 4.98 29.1 NA Garrity GM, Jones D, Krieg NR, Ludwig W et al. (editors). Bergey’s C18 : 0 10.0 11.6 3.7 NA NA NA Manual of Systematic Bacteriology, 2nd ed., vol. 3, (The Firmicutes). Dordrecht, Heidelberg, London, New York: Springer; 2009. pp. C18 : 1n9 20.3 38.5 5.7 NA NA NA 1016. C18 : 2n6 1.2 NA <1 NA NA NA 6. Duncan SH, Hold GL, Harmsen HJ, Stewart CS, Flint HJ. Growth requirements and fermentation products of Fusobacterium praus-  nitzii, and a proposal to reclassify it as Faecalibacterium prausnitzii Using an API 20A strip (BioMerieux), positive reactions are gen. nov., comb. nov. Int J Syst Evol Microbiol 2002;52:2141–2146. observed for D-glucose, lactose, sucrose, D-mannose, aescu- 7. Khan MT, Duncan SH, Stams AJ, van Dijl JM, Flint HJ et al. The lin ferric citrate, salicin, maltose, D-melezitose, glycerol and gut anaerobe Faecalibacterium prausnitzii uses an extracellular raffinose but no reaction is obtained for urease, D-mannitol, electron shuttle to grow at oxic-anoxic interphases. Isme J 2012; 6:1578–1585. D-xylose, gelatin, D-cellulose, D-sorbitol, L-arabinose, L- rhamnose and trehalose. Using an API ZYM strip (Bio- 8. Sadaghian Sadabad M, von Martels JZ, Khan MT, Blokzijl T,  Paglia G et al. A simple coculture system shows mutualism Merieux), exhibits esterase (C4), esterase lipase (C8), acid between anaerobic Faecalibacteria and epithelial Caco-2 cells. Sci phosphatase, naphtol phosphohydrolase, a-galactosidase Rep 2015;5:17906. (melibase), a-glucosidase (maltase), b-glucosidase (cellu- 9. Million M, Diallo A, Raoult D. Gut microbiota and malnutrition. lose), and N-acetyl-b-glucosaminidase (chitinase) activities. Microb Pathog 2016:30212–30216. No alkaline phosphatase, lipase, leucine arylamidase, valine 10. Million M, Tidjani Alou M, Khelaifia S, Bachar D, Lagier JC et al. arylamidase, cystine arylamidase, trypsin, a-chymotrypsin, Increased gut redox and depletion of anaerobic and methanogenic prokaryotes in severe acute malnutrition. Sci Rep 2016;6:26051. b-galactosidase, b-glucuronidase, a-mannosidase and a- 11. Hornef MW, Pabst O. Real friends: Faecalibacterium prausnitzii fucosidase activities are observed. The major SCFA pro- supports mucosal immune homeostasis. Gut 2016;65:365–367. duced is acetic acid but butyric acid, isobutyric acid and 12. Miquel S, Martín R, Rossi O, Bermúdez-Humaran LG, Chatel JM propionic acid are also produced in small amounts. The pre- et al. Faecalibacterium prausnitzii and human intestinal health. dominant cellular fatty acids are the saturated acid C16 : 0 Curr Opin Microbiol 2013;16:255–261. and the unsaturated acid C18 : 1n9. 13. Seng P, Drancourt M, Gouriet F, La Scola B, Fournier PE et al. T T T Ongoing revolution in bacteriology: routine identification of bacter- The type strain is AT2 (=CSUR P2014 =DSM 100451 ), ia by matrix-assisted laser desorption ionization time-of-flight isolated from the faeces of a healthy 28-year-old French mass spectrometry. Clin Infect Dis 2009;49:543–551. male. The DNA G+C content of the type strain is 56.8 % 14. Drancourt M, Bollet C, Carlioz A, Martelin R, Gayral JP et al. 16S (genome sequence). The type strain is susceptible to amoxi- ribosomal DNA sequence analysis of a large collection of environ- mental and clinical unidentifiable bacterial isolates. J Clin cillin, amoxicillin/clavulanic acid, cefalexin, ciprofloxacin, Microbiol 2000;38:3623–3630. doxycycline, erythromycin, gentamicin, imipenem, metro- 15. Meier-Kolthoff JP, Göker M, Spröer C, Klenk HP. When should a nidazole, nitrofurantoin, oxacillin, penicillin G, rifampicin, DDH experiment be mandatory in microbial taxonomy? Arch trimethoprim-sulfamethoxazole, tobramycin and vancomy- Microbiol 2013;195:413–418. cin, but resistant to ceftriaxone. 16. Meier-Kolthoff JP, Auch AF, Klenk HP, Göker M. Genome sequence-based species delimitation with confidence intervals and improved distance functions. BMC Bioinformatics 2013;14:60. Funding information 17. Meier-Kolthoff JP, Klenk HP, Göker M. Taxonomic use of DNA G This study was funded by the Fondation Mediterran ee Infection. +C content and DNA-DNA hybridization in the genomic age. Int J Syst Evol Microbiol 2014;64:352–356. Acknowledgements 18. Edgar RC. MUSCLE: multiple sequence alignment with high accu- – The authors thank the Xegen Company (www.xegen.fr) for automating racy and high throughput. Nucleic Acids Res 2004;32:1792 1797. the genomic annotation process, Magali Richez for short chain fatty 19. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis acid analysis, Claudia Andrieu for administrative assistance and Mag- and post-analysis of large phylogenies. Bioinformatics 2014;30: dalen Lardiere for English correction. 1312–1313. 20. Pattengale ND, Alipour M, Bininda-Emonds OR, Moret BM, Conflicts of interest Stamatakis A. How many bootstrap replicates are necessary? The authors declare no conflicts of interest. J Comput Biol 2010;17:337–354.

250 6 Togo et al., Int J Syst Evol Microbiol

21. Swofford DL. PAUP*: Phylogenetic Analysis Using Parsimony (*and 35. Qin QL, Xie BB, Zhang XY, Chen XL, Zhou BC et al. A proposed Other Methods), Version 4.0 B10. Sunderland: Sinauer Associates; genus boundary for the prokaryotes based on genomic insights. 2002. J Bacteriol 2014;196:2210–2215. 22. Halebian S, Harris B, Finegold SM, Rolfe RD. Rapid method that 36. Yarza P, Yilmaz P, Pruesse E, Glöckner FO, Ludwig W et al. Unit- aids in distinguishing Gram-positive from Gram-negative anaero- ing the classification of cultured and uncultured bacteria and bic bacteria. J Clin Microbiol 1981;13:444–448. archaea using 16S rRNA gene sequences. Nat Rev Microbiol 2014; 12:635–645. 23. Powers EM. Efficacy of the Ryu nonstaining KOH technique for rapidly determining Gram reactions of food-borne and waterborne 37. Kim M, Oh HS, Park SC, Chun J. Towards a taxonomic coherence bacteria and yeasts. Appl Environ Microbiol 1995;61:3756–3758. between average nucleotide identity and 16S rRNA gene sequence similarity for species demarcation of prokaryotes. Int J Syst Evol 24. Johnson MJ, Thatcher E, Cox ME. Techniques for controlling vari- Microbiol 2014;64:346–351. ability in Gram staining of obligate anaerobes. J Clin Microbiol et al. 1995;33:755–758. 38. Barcenilla A, Pryde SE, Martin JC, Duncan SH, Stewart CS Phylogenetic relationships of butyrate-producing bacteria from 25. Le Page S, van Belkum A, Fulchiron C, Huguet R, Raoult D et al. the human gut. Appl Environ Microbiol 2000;66:1654–1661. Evaluation of the PREVI® isola automated seeder system com- 39. Duncan SH, Barcenilla A, Stewart CS, Pryde SE, Flint HJ. Acetate pared to reference manual inoculation for antibiotic susceptibility utilization and butyryl coenzyme a (CoA):acetate-CoA transferase testing by the disk diffusion method. Eur J Clin Microbiol Infect Dis – in butyrate-producing bacteria from the human large intestine. 2015;34:1859 1869. Appl Environ Microbiol 2002;68:5186–5190. 26. Sasser M. Bacterial Identification by Gas Chromatographic Analysis 40. Gossling J, Moore WEC. Gemmiger formicilis, n.gen., n.sp., an of Fatty Acids Methyl Esters (GC-FAME), Technical Note 101. anaerobic budding bacterium from intestines. Int J Syst Bacteriol Newark, DE: MIDI Inc; 2006. 1975;25:202–207. et al. 27. Dione N, Sankar SA, Lagier JC, Khelaifia S, Michele C 41. Holmstrøm K, Collins MD, Møller T, Falsen E, Lawson PA. Subdoli- Genome sequence and description of Anaerosalibacter massiliensis granulum variabile gen. nov., sp. nov. from human feces. Anaerobe sp. nov. New Microbes New Infect 2016;10:66–76. 2004;10:197–203. € 28. Klaring K, Hanske L, Bui N, Charrier C, Blaut M et al. Intestinimo- 42. Zellner G, Stackebrandt E, Nagel D, Messner P, Weiss N et al. nas butyriciproducens gen. nov., sp. nov., a butyrate-producing Anaerofilum pentosovorans gen. nov., sp. nov., and Anaerofilum bacterium from the mouse intestine. Int J Syst Evol Microbiol agile sp. nov., two new, strictly anaerobic, mesophilic, acidogenic 2013;63:4606–4612. bacteria from anaerobic bioreactors. Int J Syst Bacteriol 1996;46: 29. Pfeiffer N, Desmarchelier C, Blaut M, Daniel H, Haller D et al. Aceta- 871–875. tifactor muris gen. nov., sp. nov., a novel bacterium isolated from the 43. Song L, Dong X. Hydrogenoanaerobacterium saccharovorans gen. intestine of an obese mouse. Arch Microbiol 2012;194:901–907. nov., sp. nov., isolated from H2-producing UASB granules. Int J – 30. Zhao G, Nyman M, Jönsson JA. Rapid determination of short- Syst Evol Microbiol 2009;59:295 299. chain fatty acids in colonic contents and faeces of humans and 44. Chen S, Dong X. Acetanaerobacterium elongatum gen. nov., sp. rats by acidified water-extraction and direct-injection gas chroma- nov., from paper mill waste water. Int J Syst Evol Microbiol 2004; tography. Biomed Chromatogr 2006;20:674–682. 54:2257–2262. 31. Illumina. 2012. Illumina data processing of nextera® mate pair 45. Carlier JP, Bedora-Faure M, K’ouas G, Alauzet C, Mory F. Pro- reads on illumina sequencing platforms. 2012. www.illumina. posal to unify Clostridium orbiscindens Winter et al. 1991 and com/documents/products/technotes/technote_nextera_matepair_ Eubacterium plautii (Seguin 1928) Hofstad and Aasjord 1982, with data_processing.pdf. [accessed 05 January 2017]. description of Flavonifractor plautii gen. nov., comb. nov., and reas- signment of Bacteroides capillosus to Pseudoflavonifractor capillo- 32. Nomenclatural changes validly published in the 2002 issues of sus gen. nov., comb. nov. Int J Syst Evol Microbiol 2010;60:585– the IJSEM. www.bacterio.net/-changes-2002.html. 590. 33. Gupta SK, Padmanabhan BR, Diene SM, Lopez-Rojas R, Kempf M 46. Xing D, Ren N, Li Q, Lin M, Wang A et al. Ethanoligenens harbinense et al. ARG-ANNOT, a new bioinformatic tool to discover antibiotic gen. nov., sp. nov., isolated from molasses wastewater. Int J Syst resistance genes in bacterial genomes. Antimicrob Agents Evol Microbiol 2006;56:755–760. Chemother 2014;58:212–220. 47. Zhang K, Song L, Dong X. Proteiniclasticum ruminis gen. nov., sp. 34. Conway KR, Boddy CN. ClusterMine360: a database of microbial nov., a strictly anaerobic proteolytic bacterium isolated from yak PKS/NRPS biosynthesis. Nucleic Acids Res 2013;41:D402–D407. rumen. Int J Syst Evol Microbiol 2010;60:2221–2225.

Five reasons to publish your next article with a Microbiology Society journal 1. The Microbiology Society is a not-for-profit organization. 2. We offer fast and rigorous peer review – average time to first decision is 4–6 weeks. 3. Our journals have a global readership with subscriptions held in research institutions around the world. 4. 80% of our authors rate our submission process as ‘excellent’ or ‘very good’. 5. Your article will be published on an interactive journal platform with advanced metrics.

Find out more and submit your article at microbiologyresearch.org.

251 7

View publication stats Annexe 2

Autres publications

Publication n° 13: A new highly sensitive and specific real-time PCR assay targeting the malate dehydrogenase gene of 3 Kingella kingae and application to 201 pediatric clinical specimens

N. El Houmami, G. A. Durand, J. Bzdrenga, A. Darmon, P. Minodier, H. Selgmann, D. Raoult, P-E. Fournier

Publié dans Journal of Clinical Microbiology

Publication n° 14: Acute Septic Arthritis of the Knee Caused by Kingella kingae in a 5-Year-Old Cameroonian Boy N. El Houmami, D. Ceroni, K. Codjo Seignon, J. C. Pons, C. Lambert, G. A. Durand, P. Minodier, L. Lamah, P. Bidet, J. Schrenzel, D. Raoult, P-E. Fournier

Publié dans Frontiers in Pediatrics

Publication n° 15: A modified multilocus sequence typing protocol to genotype Kingella kingae from oropharyngeal swabs without bacterial isolation N. El Houmami, J. Bzdrenga, J. C. Pons, P. Minodier, G. A. Durand, A. Oubraham, D. Ceroni, P. Yagupsky, D. Raoult, P. Bidet, P-E. Fournier

Publié dans BMC Microbiology

Publication n° 16: Molecular tests that target the RTX locus do not distinguish between Kingella kingae and the recently described K. negevensis species N. El Houmami, J. Bzdrenga, G. A. Durand, P. Minodier, H. Seligmann, E. Prudent, S. Bakour, S. Bonacorsi, D. Raoult, P. Yagupsky, P-E. Fournier

Publié dans Journal of Clinical Microbiology

Publication n° 17: Emergence of Clostridium difficile tcdC variant 078 in Marseille, France N. Cassir, N. Fahsi, G. A. Durand, J. C. Lagier, D. Raoult, P-E. Fournier

Publié dans European Journal of Clinical Microbiology and Infectious Diseases

252 1 Title:

2 A new highly sensitive and specific real-time PCR assay targeting the malate dehydrogenase gene of

3 Kingella kingae and application to 201 pediatric clinical specimens

4

5 Running title (52<54 characters and spaces):

6 A new real-time PCR targeting the K. kingae mdh gene

7

8 Authors:

9 Nawal El Houmami1#, MD, PhD; Guillaume André Durand1, MD, MSc; Janek Bzdrenga2, PhD; Anne

10 Darmon1, MD, MSc; Philippe Minodier3, MD, MSc; Hervé Seligmann1, PhD; Didier Raoult1, MD, PhD;

11 Pierre-Edouard Fournier1#, MD, PhD

12

13 Authors’ affiliations :

14 1UMR VITROME, Aix-Marseille Univ, IRD, Service de Santé des Armées, Assistance Publique-Hôpitaux

15 de Marseille, Institut Hospitalo-Universitaire Méditerranée-Infection, Marseille, France.

16 2Univ. Grenoble-Alpes, CEA, CNRS, IBS, F-38000, Grenoble, France.

17 3Department of Pediatric Emergency Medicine, North Hospital, Marseille, France.

18

19 Corresponding authors :

20 #Dr. Nawal El Houmami, UMR VITROME, Institut Hospitalo-Universitaire Méditerranée Infection,

21 Assistance Publique-Hôpitaux de Marseille, 19-21 Boulevard Jean Moulin, 13005 Marseille, France.

22 [[email protected]].

23 #Prof. Pierre-Edouard Fournier, UMR VITROME, Institut Hospitalo-Universitaire Méditerranée-

24 Infection, 19-21 Boulevard Jean Moulin, 13005 Marseille, France. [pierre-edouard.fournier@univ-

25 amu.fr].

26 Telephone: + (33) 413 732 401, Fax: + (33) 413 732 402.

1 253 27 ABSTRACT

28 Kingella kingae is a significant pediatric pathogen responsible for bone and joint infections, occult

29 bacteremia, and endocarditis in early childhood. Past efforts to detect this bacterium by culture and

30 broad-range 16S rRNA gene polymerase chain reaction (PCR) assays from clinical specimens have

31 proven unsatisfactory and were gradually let out for the benefit of specific real-time

32 PCR tests targeting the groEL gene and RTX locus of K. kingae by the late 2000s. However, recent

33 studies showed that real-time PCR (RT-PCR) assays targeting the Kingella sp. RTX locus that are

34 currently available for the diagnosis of K. kingae infection lack of specificity because they could not

35 distinguish between K. kingae and the recently described K. negevensis species. Furthermore, in

36 silico analysis of the groEL gene from a large collection of 45 K. kingae strains showed that primers

37 and probes from K. kingae groEL-based RT-PCR assays display a few mismatches with K. kingae

38 groEL variations that may result in a decreased detection sensitivity, especially in paucibacillary

39 clinical specimens. In order to provide an alternative to groEL- and RTX-targeting RT-PCR assays that

40 may suffer from suboptimal specificity and sensitivity, a K. kingae-specific RT-PCR assay targeting the

41 malate dehydrogenase (mdh) gene was developed for predicting no mismatch against 18 variants

42 of the K. kingae mdh gene from 20 distinct sequences types of K. kingae. This novel K. kingae-specific

43 RT-PCR assay demonstrated a high specificity and sensitivity and was successfully used to diagnose K.

44 kingae infections and carriage in 104 clinical specimens from children aged between 7 months and 7

45 years old.

46

47

48

49

50

51

52

2 254 53 INTRODUCTION

54 Kingella kingae is a significant pediatric pathogen responsible for bone and joint infections, occult

55 bacteremia and, more rarely, endocarditis that may occur either sporadically, or in the context of

56 outbreaks in daycare centers [1,2]. Past efforts to detect this organism by culture have proven

57 unsatisfactory and molecular diagnostics were gradually used in the 2000s to successfully diagnose K.

58 kingae disease [3-5]. Consequently, the increasing number of molecularly-confirmed K. kingae

59 infections in infants led to this organism being recognized as the primary agent of septic arthritis,

60 osteomyelitis, and tenosynovitis in children aged between six and 36 months in countries where K.

61 kingae-specific real-time PCR assays are routinely employed [4-7]. This contributed to significantly

62 improve our knowledge of the etiology of infantile bone and joint infections [5, 8, 9]. In addition to

63 increasing the detection yield of microorganisms from osteoarticular samples, molecular assays have

64 contributed to a better understanding of the epidemiology of K. kingae carriage among healthy

65 carriers and ill children [10-12]. Although polymerase chain reaction (PCR) assays targeting the 16S

66 rRNA gene made it possible to moderately enhance the detection of the organism from

67 osteoarticular samples [13], the development of K. kingae-specific RT-PCR assays allowed a

68 substantial increase in the diagnosis of K. kingae infections and oropharyngeal carriage [5,14,15].

69 To date, only the groEL (also known as cpn60) gene and those located in the RTX locus, namely rtxA

70 and rtxB, have been targeted for the development of K. kingae-specific RT-PCR assays [4,5,16,17].

71 Comprehensively studied, the groEL gene encodes a chaperone protein that is considered as a

72 universal bacterial marker [18], and PCR assays targeting this gene are widely used for the molecular

73 diagnosis of infectious diseases [19-21]. However, although recent studies have confirmed that

74 targeting the groEL gene from K. kingae (KkigroEL) is a reliable strategy for the molecular detection

75 of this bacterium in clinical specimens [15,22], primers and probes from groEL-based RT-PCR assays

76 that were reported by Ilharreborde et al. [4] and Levy et al. [5] display a few mismatches with

77 KkigroEL variations that may result in a decreased detection sensitivity [15, 23].

3 255 78 In contrast, RTX-targeting RT-PCR assays gained great popularity worldwide because they were

79 initially believed to be highly specific for K. kingae [16,17]. However, the RTX locus of K. kingae is

80 flanked by mobile genetic elements that are present in genomic regions of decreased GC content

81 (30% versus an average of 46.6% for the whole genome of K. kingae). Because such a GC content

82 difference is a meaningful genetic marker of mobilome, Kehl-Fie et al. suggested that this RTX locus

83 was horizontally-acquired [24]. This assumption was recently confirmed by the presence of an

84 identical RTX locus in the genome of K. negevensis [15,25], a newly described Kingella species

85 isolated from the oropharynxes of Israeli and Swiss children [26,27], and from the vaginal discharge

86 of a young woman [25]. Furthermore, in vitro studies indicated that RT-PCR assays targeting the RTX

87 locus of K. kingae were also positive for K. negevensis and, hence, could not formally discriminate

88 both species when used alone [15]. Kingella negevensis has also been identified in the hip of an 8-

89 month-old boy with a specific qPCR targeting the K. negevensis groEL gene, indicating that this novel

90 described Kingella species may occasionally be a pediatric pathogen [15].

91 In order to provide an alternative to groEL- and RTX-targeting RT-PCR assays that may suffer from

92 suboptimal specificity and sensitivity, a K. kingae-specific RT-PCR assay targeting the malate

93 dehydrogenase (mdh) gene, a housekeeping gene, was developed. This novel RT-PCR assay targeting

94 the mdh gene from K. kingae (Kkimdh) demonstrated a high specificity and sensitivity and was

95 successfully used to diagnose K. kingae infections and carriage in 104 clinical specimens from young

96 children.

97

98

99 MATERIALS AND METHODS

100 Bacterial isolates. In the 2000s, epidemiological studies were conducted in southern Israel on 7,217

101 healthy children, from whom K. kingae strains were isolated at the Soroka University Medical Center,

102 Beer-Sheva, Israel [26]. Forty of these K. kingae strains, cultivated from children aged between six

103 and 48 months suffering from osteoarticular infections (n=12), occult bacteremia (n=4), endocarditis

4 256 104 (n=3), or asymptomatic oropharyngeal colonization (n=21), were used in this study (Table S1).

105 Oropharyngeal swabs from healthy children were first inoculated onto a selective vancomycin-

106 containing agar (also named BAV medium) to inhibit the competing Gram-positive flora and facilitate

107 the recognition of hemolytic K. kingae colonies [28]. All K. kingae isolates were then subcultured on

108 5% sheep blood-enriched Columbia agar for between 24 and 36 hours at 37°C in a 5% CO2-enriched

109 atmosphere. Additionally, 86 other bacterial strains, including all other Kingella species, namely K.

110 negevensis strain Sch538T, K. oralis CIP103803T, K. denitrificans CIP103473T and K. potus CIP108935T,

111 as well as members of the Neisseria, Haemophilus, Staphylococcus, Streptococcus and

112 Mycobacterium genera, were used to determine the specificity of the RT-PCR primers and probes

113 targeting the Kkimdh gene (Table S2).

114 Clinical specimens of K. kingae infection and carriage. Between December 2013 and December

115 2017, 106 children with K. kingae infections and/or carriage originally from Europe, South America,

116 Africa, and the South Pacific, were diagnosed by using KkigroEL-specific RT-PCR [5] at the IHU

117 Mediterranée-Infection, Marseille, France. Of those, 96 clinical specimens or extracted DNA (from 95

118 children) that were stored at -80°C were retrieved, which were derived from joint fluid (n=64), bone

119 tissue (n= 6), tenosynovial fluid (n=2), soft tissue (n=2), endocardial cushion (n=1), and pharyngeal

120 swab (n=25) (Table S3). The efficiency of DNA extraction and the possible presence of inhibitors were

121 evaluated in all clinical specimens using the RS42-Km primer pair that targets a fragment of the

122 human -globin gene [5].

123 Genomic DNA extraction. Genomic DNA from all bacterial strains and clinical samples was extracted

124 with a BioRobot EZ1 workstation and an EZ1 DNA tissue kit (Qiagen, Courtaboeuf, France) according

125 to the aufacturer’s recoedatios [15]. DNA was stored at -80°C prior to molecular assays. To

126 limit the effects of PCR inhibitors, all extracted DNAs were tested both undiluted and diluted 1:10.

127 Selection of the mdh gene from K. kingae genomes. In order to select a relevant target gene for the

128 development of a K. kingae-specific RT-PCR assay, a comparison of five K. kingae genomes available

129 in GenBank, namely K. kingae ATCC 23330T (FOJK01000000) [26], K. kingae KKWG1 (LN869922) [29],

5 257 130 K. kingae PYKK081 (NZ_JH621344), K. kingae 11220434 (JH768595) [30] and K. kingae KK247

131 (CCJT01000000) [31], was performed using the Geneious R11.0.5 software

132 (http://www.geneious.com) [32]. Genes and their flanking regions belonging to the core genome of

133 K. kingae and exhibiting a GC content close to 50% were screened to facilitate the primer and probe

134 design according to the Takyon polymerase protocol (Eurogentec, Seraing, Belgium). The mdh gene

135 of K. kingae (Kkimdh) coding the malate dehydrogenase met the above criteria and was thus

136 selected. Thereafter, paired-end sequencing of the Kkimdh gene and its flanking regions from 40 K.

137 kingae strains using a MiSeq sequencer (Illumina Inc., San Diego, CA, USA), as well as genome

138 assembly, were performed as previously described [26].

139 Characterization of the Kingella kingae mdh gene. A MAFFT alignment of the Kkimdh nucleotide

140 sequences and its flanking regions from the 45 studied K. kingae strains was performed using

141 Geneious R11.0.5 [32,33]. The related distance matrix of the 45 distinct Kkimdh genes was obtained

142 using Geneious R11.0.5 (Table S4). To detect possible lateral gene transfer from or within the

143 genome of other bacterial species, a megablastn search with default parameters

144 (http://blast.ncbi.nlm.nih.gov) was then conducted by comparing the obtained mdh orthologous

145 sequences and their genomic environment to public databases. A neighbor-joining tree of Kkimdh

146 gene sequences was then created using MEGA7 with default parameters [34].

147 RT-PCR assay targeting the Kkimdh gene. (a) Design of primers and probe. To design specific

148 primers and probes, a MAFFT alignment of mdh nucleotide sequences from the 45 studied K. kingae

149 strains was first performed. Thereafter, the primers Fwd_Kkimdh (5- TGTTCCGCATTGCTTCTG -3) and

150 Rev_Kkimdh (5- TCATGCCGTCCAACAATG -3) amplifying a 144-bp fragment, and the probe P_Kkimdh

151 (5- 6-carboxyfluorescein [FAM]- CATCATCACGCCCTGAACGGCTT -3) were manually designed.

152 Particular care was taken in order to avoid nucleotide mismatches between all K. kingae strains, and

153 to maximize mismatches with mdh orthologous detected from other bacterial species. Primers and

154 probe specificity was confirmed in silico using the BLAST tool (http://blast.ncbi.nlm.nih.gov).

6 258 155 (b) Kkimdh RT-PCR protocol. Real-time PCR amplification reactions were carried out in a final volume

156 of 20 µl of reaction mixture containing 10 µl of Takyon NoRox Probe MasterMix dTTP (Eurogentec),

157 0.45 µM (each) primers, 0.45 µM labeled probe and 5 µl of purified DNA. Amplification was

158 performed using a Bio-Rad CFX96 platform and the following cycling parameters: heating at 50°C for

159 two minutes and 95°C for three minutes, followed by 45 cycles of a two-stage temperature profile of

160 95°C for three seconds and 60°C for 30 seconds.

161 Evaluation of the sensitivity and specificity of the Kkimdh RT-PCR assay. Twenty K. kingae strains

162 belonging to 20 distinct sequence types (STs) (Table S2) which had previously tested positive for

163 KkigroEL, rtxA, and rtxB [15], and 96 specimens which had previously tested positive for KkigroEL

164 were tested using the Kkimdh-RT-PCR assay. In addition, 105 various KkigroEL-negative clinical

165 specimens derived from children with suspected osteoarticular infections were added to the analysis.

166 Furthermore, 87 other bacterial strains, including strains from K. negevensis, K. oralis, K. denitrificans

167 and K. potus, as well as members of the Neisseria, Haemophilus, Staphylococcus, Streptococcus and

168 Mycobacterium genera were tested to assess the specificity of the assay (Table S2). To determine the

169 detection limit of the method, 12-fold serial dilutions of a bacterial suspension of K. kingae strain

170 ATCC 23330T at an initial concentration of 108 bacteria ml-1 in phosphate-buffered saline were

171 evaluated and further quantified by culture on 5% sheep blood-enriched Columbia agar (bioMérieux)

172 and colony counting.

173 Ethics statement. This study was approved by the Ethics Committee of the IHU Méditerranée-

174 Infection under reference number 2017-006. Epidemiological studies performed in the 2000s were

175 approved by the Ethics Committee of the Soroka University Medical Center, as well as by the Israel

176 Ministry of Health.

177 Accession number(s). The GenBank accession numbers for the mdh genes from the 45 studied K.

178 kingae strains analyzed in this study are LT985480 to LT985524 (Table S1).

179

180

7 259 181 RESULTS

182 Genomic analysis of the mdh gene of K. kingae and its environment. A 978-bp Kkimdh gene was

183 identified in all 45 K. kingae genomes. The chromosomal region carrying the Kkimdh gene is

184 surrounded by the ribosomal small subunit-dependent GTPase gene located 218 bp upstream, and

185 downstream by a locus containing genes coding for the GTP cyclohydrolase FolE2 and sensor

186 histidine kinase (Fig. 1). The synteny of this genomic architecture was conserved in all 45 K. kingae

187 strains. The distance matrix calculated from the 45 Kkimdh DNA sequences displayed 18 distinct

188 variants (Fig. 2), with a maximum distance of 98.4% between K. kingae strains ATCC 23330T and

189 D2363 (Table S4).

190 In silico analysis of the Kkimdh gene and design of the Kkimdh-specific RT-PCR assay. The

191 Megablastn search indicated the presence of a mdh gene within the genomes of Acinetobacter

192 iwoffii ZS207 (CP019143), Acidovorax sp. RAC01 (CP016447), Neisseria sp. KEM232 (CP022527.1),

193 Neisseria elongata subsp. glycolytica ATCC 29315 (CP007726.1), Neisseria weaveri NCTC13585

194 (LT90643.1), Neisseria zoodegmatis NCTC12230 (LN869922.1), and Polaromonas naphthalenivorans

195 CJ2 (CP000529). The MAFFT alignment of the nucleotide sequences of these mdh orthologous genes

196 showed that the Kkimdh-F and Kkimdh-R primers, and the Kkimdh-P probe displayed a total of 13 to

197 15 mismatches between the mdh genes from K. kingae ATCC 23330T and those from the Neisseria

198 species, which are the closest orthologous genes related to Kkimdh by showing nucleotide sequences

199 identities ranging from 77.1 to 79.1% (Fig. S1, Table S5). These data thereby demonstrate a high

200 index of in silico specificity of Kkimdh-RT-PCR assay.

201 Validation of the Kkimdh-specific RT-PCR assay. The detection threshold of the Kkimdh RT-PCR assay

202 was determined to be 10 CFU/ml. The assay was positive for all 20 K. kingae strains belonging to 20

203 distinct STs, whereas no amplification was obtained for K. negevensis, K. denitrificans, K. oralis and K.

204 potus strains (Table 1). Similarly, no amplification was obtained from any of the other 82 tested

205 bacterial species. As expected, Kkimdh-specific RT-PCR testing was positive in the 96 KkigroEL-

206 positive specimens. Unexpectedly, eight joint fluid samples initially tested KkigroEL-negative were

8 260 207 detected Kkimdh-positive, whereas the remaining 97 KkigroEL-negative pediatric specimens were

208 both Kkimdh-and KkigroEL-negative. Triplicate assays were carried out on the eight discrepant

209 samples. All eight were repeatedly Kkimdh-positive, with cycle threshold values ranging from 30 to

210 34. Of these eight Kkimdh-positive clinical specimens, six were derived from children with septic

211 arthritis aged between seven and 29 months, one was sampled from the oropharynx of a 13-month-

212 old boy, and one was identified in the joint fluid of a seven-year-old boy (Table S3).

213

214

215 DISCUSSION

216 This study reports a novel K. kingae-specific RT-PCR assay targeting the Kkimdh gene, a housekeeping

217 gene coding the malate dehydrogenase that was identified in 45 distinct K. kingae genomes, and

218 firmly detected in 20 various clinical isolates and 104 clinical specimens from infants and young

219 children originally from Europe, South America, Africa and the South Pacific. The high specificity of

220 this PCR system was demonstrated by the absence of any mdh gene in the genomes of all other

221 Kingella species. In addition, orthologous mdh genes were found in only a few bacterial species,

222 which exhibited low levels of nucleotide identity with Kkimdh, as demonstrated by the presence of

223 13 to 15 mismatches between the nucleotide sequences of primers and probe Fwd-Kkimdh, Rev-

224 Kkimdh, P-Kkimdh, and orthologous mdh genes from Neisseria species.

225 In contrast to the Kingella sp. RTX locus, Kkimdh is only present in K. kingae, is located in a genomic

226 region that presents a conserved synteny and a GC content of only 2 to 3% greater than that the

227 whole genome of K. kingae, and is not surrounded by transposable elements. For all these reasons, it

228 appeared particularly pertinent to target Kkimdh for the development of a new K. kingae-specific

229 molecular tool in clinical microbiology.

230 To the best of our knowledge, this new K. kingae-specific RT-PCR assay is currently the only

231 molecular tool showing optimal sensitivity and specificity for the diagnosis of K. kingae infection and

232 carriage when compared to all those previously reported. Although K. kingae-specific RT-PCR assays

9 261 233 targeting the groEL gene may be considered as the gold standard for the detection of K. kingae

234 [15,22], primers and probes reported by Ilharreborde et al. and Levy et al. may present between one

235 and three mismatches with KkigroEL nucleotide sequences from K. kingae isolates belonging to STc-6

236 and -35 (data not shown), which are two STcs previously shown to be responsible for invasive

237 infections in pediatrics [35], and which may impact the degree of sensitivity of these RT-PCR tests

238 [15,23]. Although mismatches in primer regions may have a limited effect on the quality of the

239 amplification curves, they may result in a significant increase in cycle threshold values [36,37]. Two

240 mismatches can delay amplification by three to five cycles, while three mismatches can delay

241 amplification by seven to 13 cycles [37]. In addition, single mismatches in the minor groove binding-

242 modified probes may result in no or weak amplification curves leading to the risk of being

243 interpreted as negative [38]. Because K. kingae-positive clinical specimens commonly contain a low

244 bacterial load, a highly sensitive molecular tool is, therefore, of significant importance to maximize

245 the detection yield of the organism from paucibacillary specimens.

246 Regarding the RT-PCR assays targeting the RTX locus, it was recently demonstrated that such

247 molecular tools are not valid to formally confirm the diagnosis of K. kingae infection because of the

248 cross-detection with K. negevensis [15]. In addition, given the numerous uncharacterized microbes

249 colonizing humans [39], and the multiple genomic factors indicating that the Kingella sp. RTX locus

250 was horizontally acquired, such as ISKne1, a Kingella sp. RTX locus-related transposable element

251 found in multiple copies in both K. negevensis and K. kingae genomes [15], it cannot be entirely ruled

252 out that a similar Kingella sp. RTX locus may have been transferred to other as yet uncharacterized

253 Kingella species. Consequently, this implies that the numerous studies conducted over the past

254 decade to calculate the prevalence rate of K. kingae infection and carriage using Kingella sp. RTX-

255 related molecular methods are likely to have, unintentionally, overestimated the results.

256 To overcome the lack of specificity of RT-PCRs assays targeting the Kingella sp. RTX locus and to

257 distinguish K. kingae from K. negevensis in clinical samples, Opota et al. recently proposed a strategy

258 which consists in targeting both KkigroEL and Kingella sp. rtxA by using a duplex RT-PCR assay for

10 262 259 diagnosing K. kingae infection [25]. Such a diagnostic strategy is strongly debatable and comprises

260 serious limitations for its use in the clinical diagnostic setting. Indeed, the genomic nature of the RTX

261 locus from Kingella sp. makes its lateral transfer in uncharacterized Kingella species possible. More

262 importantly, while K. kingae and K. negevensis share the same oropharyngeal niche and may

263 potentially be involved in pediatric osteoarticular infections, this duplex RT-PCR assay does not make

264 it possible to diagnose potential dual infections, or carriage caused by both K. kingae and K.

265 negevensis [Pablo Yagupsky, unpublished data].

266 Recently, de Knegt et al. developed a similar approach to diagnose osteoarticular infections caused

267 by K. kingae in a Danish pediatric population, after designing new primers and probes against the

268 rtxA gene to maximize sensitivity and optimizing the KkigroEL-specific RT-PCR assay reported by

269 Ilharreborde et al. [23]. Interestingly, 12 specimens were Kingella sp. rtxA-positive, and only 10 were

270 positive for both rtxA and groEL. The authors suggested that the two rtxA-positive and KkigroEL-

271 negative specimens may be explained by a positivity near the limit of detection of their two RT-PCR

272 tests, a sampling error, or decreased sensitivity due to mismatches of either primers or probes in the

273 KkigroEL gene. Nevertheless, because it was previously demonstrated that K. negevensis, which is

274 rtxA- and rtxB-positive and KkigroEL-negative, may occasionally induce joint infections in infancy,

275 such findings in Denmark may be consistent with infections caused by K. negevensis, as previously

276 observed in France [15].

277 Therefore, as there is currently a lack of clinical data regarding K. negevensis, being able to formally

278 discriminate K. kingae from K. negevensis is important. As a consequence, the development of highly

279 species-specific and sensitive RT-PCR assays emerges as the most effective and reliable diagnostic

280 strategy in clinical microbiology. Recently, the development of a K. negevensis-specific RT-PCR testing

281 targeting the K. negevensis groEL gene enabled identification of the first arthritis caused by K.

282 negevensis in an eight-month boy [15]. To diagnose infections and carriage caused by K. kingae, the

283 new Kkimdh-specific RT-PCR assay that we describe herein thereby appears as an optimal molecular

284 tool that could be used either alone, or in combination with K. negevensis-specific RT-PCR assays.

11 263 285 However, it should be emphasized that such RT-PCR assays remain costly, and that dual target PCR is

286 particularly advantageous to compensate for the potentially decreased sensitivity of assays applying

287 minor groove-binding probes [38]. Given that no target variation in the Kkimdh gene was detected in

288 a large and diverse collection of K. kingae strains, the Kkimdh-related primers and probe designed in

289 the present study appear to be robust enough to be applied alone in the clinical diagnostic setting.

290

291 ACKNOWLEDGEMENTS

292 The study was funded by the Foundation Mediterranée Infection and the French National Research

293 Agecy uder the progra Iestisseets d’aeir, referece ANR-10-IAHU-03. The authors

294 would like to gratefully acknowledge Prof. Pablo Yagupsky, microbiologist at the Clinical Microbiology

295 Laboratory, Soroka Medical Center, Beer-Sheva, Israel, for providing numerous K. kingae isolates,

296 and for having critically reviewed this manuscript.

297

298

299

12 264 300 REFERENCES

301 1. Yagupsky P. 2015. Kingella kingae: carriage, transmission, and disease. Clin Microbiol Rev

302 28:5479.

303 2. Yagupsky P, El Houmami N, Fournier PE. 2017. Outbreaks of Invasive Kingella kingae Infections in

304 Daycare Facilities: Approach to Investigation and Management. J Pediatr 182:1420.

305 3. Fournier PE, Drancourt M, Colson P, Rolain JM, La Scola B, Raoult D. 2013. Modern clinical

306 microbiology: new challenges and solutions. Nat Rev Microbiol 11:574585.

307 4. Ilharreborde B, Bidet P, Lorrot M, Even J, Mariani-Kurkdjian P, Liguori S, Vitoux C, Lefevre Y, Doit C,

308 Fitoussi F, Penneçot G, Bingen E, Mazda K, Bonacorsi S. 2009. New real-time PCR-based method

309 for Kingella kingae DNA detection: application to samples collected from 89 children with acute

310 arthritis. J Clin Microbiol 47:18371841. Erratum in: 2009. J Clin Microbiol 47:3071.

311 5. Levy PY, Fournier PE, Fenollar F, Raoult D. 2013. Systematic PCR detection in culture-negative

312 osteoarticular infections. Am J Med 126:1143.e2533.

313 6. Chometon S, Benito Y, Chaker M, Boisset S, Ploton C, Bérard J, Vandenesch F, Freydiere AM. 2007.

314 Specific real-time polymerase chain reaction places Kingella kingae as the most common cause of

315 osteoarticular infections in young children. Pediatr Infect Dis J 26:377381.

316 7. El Houmami N, Yagupsky P, Ceroni D. 2018. Kingella kingae hand and wrist tenosynovitis in young

317 children. J Hand Surg Eur Vol. Jan 1:1753193418764818.

318 8. Juchler C, Spyropoulou V, Wagner N, Merlini L, Dhouib A, Manzano S, Tabard-Fougère A, Samara E,

319 Ceroni D. 2018. The Contemporary Bacteriologic Epidemiology of Osteoarticular Infections in

320 Children in Switzerland. J Pediatr 194:190-196.e1.

321 9. Ceroni D, Belaieff W, Cherkaoui A, Lascombes P, Schrenzel J, de Coulon G, Dubois-Ferrière V, Dayer

322 R. 2014. Primary epiphyseal or apophyseal subacute osteomyelitis in the pediatric population: a

323 report of fourteen cases and a systematic review of the literature. J Bone Joint Surg Am

324 96:15701575.

13 265 325 10. Ceroni D, Dubois-Ferrière V, Anderson R, Combescure C, Lamah L, Cherkaoui A, Schrenzel J. 2012.

326 Small risk of osteoarticular infections in children with asymptomatic oropharyngeal carriage of

327 Kingella kingae. Pediatr Infect Dis J 31:983985.

328 11. Ceroni D, Dubois-Ferriere V, Cherkaoui A, Gesuele R, Combescure C, Lamah L, Manzano S, Hibbs J,

329 Schrenzel J. 2013. Detection of Kingella kingae osteoarticular infections in children by oropharyngeal

330 swab PCR. Pediatrics 131:e230235.

331 12. Ceroni D, Llana RA, Kherad O, Dubois-Ferriere V, Lascombes P, Renzi G, Lamah L, Manzano S,

332 Cherkaoui A, Schrenzel J. 2013. Comparing the oropharyngeal colonization density of Kingella kingae

333 between asymptomatic carriers and children with invasive osteoarticular infections. Pediatr Infect

334 Dis J 32:412414.

335 13. Verdier I, Gayet-Ageron A, Ploton C, Taylor P, Benito Y, Freydiere AM, Chotel F, Bérard

336 J, Vanhems P, Vandenesch F. 2005. Contribution of a broad range polymerase chain reaction to the

337 diagnosis of osteoarticular infections caused by Kingella kingae: description of twenty-four recent

338 pediatric diagnoses. Pediatr Infect Dis J 24:692696.

339 14. Morel AS, Dubourg G, Prudent E, Edouard S, Gouriet F, Casalta JP, Fenollar F, Fournier PE,

340 Drancourt M, Raoult D. 2015. Complementarity between targeted real-time specific PCR and

341 conventional broad-range 16S rDNA PCR in the syndrome-driven diagnosis of infectious diseases. Eur

342 J Clin Microbiol Infect Dis 34:561570.

343 15. El Houmami N, Bzdrenga J, Durand GA, Minodier P, Seligmann H, Prudent E, Bakour S, Bonacorsi

344 S, Raoult D, Yagupsky P, Fournier PE. 2017. Molecular Tests That Target the RTX Locus Do Not

345 Distinguish between Kingella kingae and the Recently Described Kingella negevensis Species. J Clin

346 Microbiol 55:31133122.

347 16. Cherkaoui A, Ceroni D, Emonet S, Lefevre Y, Schrenzel J. 2009. Molecular diagnosis

348 of Kingella kingae osteoarticular infections by specific real-time PCR assay. J Med Microbiol

349 58:6568.

14 266 350 17. Lehours P, Freydière AM, Richer O, Burucoa C, Boisset S, Lanotte P, Prère MF, Ferroni A, Lafuente

351 C, Vandenesch F, Mégraud F, Ménard A. 2011. The rtxA toxin gene of Kingella kingae: a pertinent

352 target for molecular diagnosis of osteoarticular infections. J Clin Microbiol 49:12451250.

353 18. Links MG, Dumonceaux TJ, Hemmingsen SM, Hill JE. 2012. The chaperonin-60 universal target is a

354 barcode for bacteria that enables de novo assembly of metagenomics data. Plos One 7:e49755.

355 19. Chaban B, Hill JE. 2012. A 'universal' type II chaperonin PCR detection system for the investigation

356 of Archae in complex microbial communities. ISME J 6:430439.

357 20. Olson AB, Sibley CD, Schmidt L, Wilcox MA, Surette MG, Corbett CR. 2010. Development of real-

358 time PCR assays for detection of the Streptococcus milleri group from cystic fibrosis clinical

359 specimens by targeting the cpn60 and 16S rRNA genes. J Clin Microbiol 48:11501160.

360 21. Chaban B, Musil KM, Himsworth CG, Hill JE. 2009. Development of cpn60-based real-time

361 quantitative PCR assays for the detection of 14 Campylobacter species and application to screening

362 of canine fecal samples. Appl Environ Microbiol 75:30553061.

363 22. El Houmami N, Fournier PE, Ceroni D. 2017. Targeting the Kingella Kingae groEL gene is a reliable

364 method for the molecular diagnosis of K. Kingae infection and carriage. J Paediatr Child Health

365 53:10301031.

366 23. de Knegt VE, Kristiansen GQ, Schønning K. 2017. Evaluation of dual target-specific real-time PCR

367 for the detection of Kingella kingae in a Danish paediatric population. Infect Dis (Lond) 15:17.

368 24. Kehl-Fie TE, St Geme JW 3rd. 2007. Identification and characterization of an RTX toxin in the

369 emerging pathogen Kingella kingae. J Bacteriol 189:430436.

370 25. Opota O, Laurent S, Pillonel T, Léger M, Trachsel S, Prod'hom G, Jaton K, Greub G. 2017.

371 Genomics of the new species Kingella negevensis: diagnostic issues and identification of a locus

372 encoding a RTX toxin. Microbes Infect 19:546552.

373 26. El Houmami N, Bakour S, Bzdrenga J, Rathored J, Seligmann H, Robert C, Armstrong N, Schrenzel

374 J, Raoult D, Yagupsky P, Fournier PE. 2017. Isolation and characterization of Kingella negevensis sp.

15 267 375 nov., a novel Kingella species detected in a healthy paediatric population. Int J Syst Evol Microbiol

376 67:23702376.

377 27. El Houmami N, Schrenzel J, Yagupsky P, Robert C, Ceroni D, Raoult D, Fournier PE. 2017. Draft

378 Genome Sequence of Kingella negevensis SW7208426, the First European Strain of K.

379 negevensis Isolated from a Healthy Child in Switzerland. Genome Announc 5(26).

380 28. Yagupsky P, Merires M, Bahar J, Dagan R. 1995. Evaluation of novel vancomycin-containing

381 medium for primary isolation of Kingella kingae from upper respiratory tract specimens. J Clin

382 Microbiol 33:14261427.

383 29. Bidet P, Basmaci R, Guglielmini J, Doit C, Jost C, Birgy A, Bonacorsi S. 2015. Genome Analysis

384 of Kingella kingae “trai KWG1 Reeals Ho a β-Lactamase Gene Inserted in the Chromosome of

385 This Species. Antimicrob Agents Chemother 60:703708.

386 30. Fournier PE, Rouli L, El Karkouri K, Nguyen TT, Yagupsky P, Raoult D. 2012. Genomic comparison

387 of Kingella kingae strains. J Bacteriol 194:5972.

388 31. Rouli L, Robert C, Raoult D, Yagupsky P. 2014. Kingella kingae KK247, an Atypical Pulsed-Field Gel

389 Electrophoresis Clone A Strain. Genome Announc 2(6).

390 32. Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper A,

391 Markowitz S, Duran C, Thierer T, Ashton B, Mentjies P, Drummond A. 2012. Geneious Basic: an

392 integrated and extendable desktop software platform for the organization and analysis of sequence

393 data. Bioinformatics 28:16471649.

394 33. Katoh K, Misawa K, Kuma K, Miyata T. 2002. MAFFT: a novel method for rapid multiple sequence

395 alignment based on fast Fourier transform. Nucleic Acids Res 30:30593066.

396 34. Kumar S, Stecher G, Tamura K. 2016. MEGA7: Molecular Evolutionary Genetics Analysis version

397 7.0 for bigger datasets. Mol Biol Evol 33:1870–1874.

398 35. Amit U, Porat N, Basmaci R, Bidet P, Bonacorsi S, Dagan R, Yagupsky P. 2012. Genotyping of

399 invasive Kingella kingae isolates reveals predominant clones and association with specific clinical

400 syndromes. Clin Infect Dis 55:10741079.

16 268 401 36. Bourhy P, Bremont S, Zinini F, Giry C, Picardeau M. 2011. Comparison of real-time PCR assays for

402 detection of pathogenic Leptospira spp. in blood and identification of variations in target sequences.

403 J Clin Microbiol 49:21542160.

404 37. Whiley DM, Sloots TP. 2005. Sequence variation in primer targets affects the accuracy of viral

405 quantitative PCR. J Clin Virol 34:104107.

406 38. Kamau E, Agoti CN, Lewa CS, Oketch J, Owor BE, Otieno GP, Bett A, Cane PA, Nokes DJ. 2017.

407 Recent sequence variation in probe binding site affected detection of respiratory syncytial

408 virus group B by real-time RT-PCR. J Clin Virol 88:2125.

409 39. Kowarsky M, Camunas-Soler J, Kertesz M, De Vlaminck I, Koh W, Pan W, Martin L, Neff

410 NF, Okamoto J, Wong RJ, Kharbanda S, El-Sayed Y, Blumenfeld Y, Stevenson DK, Shaw GM, Wolfe

411 ND, Quake SR. 2017. Numerous uncharacterized and highly divergent microbes which colonize

412 humans are revealed by circulating cell-free DNA. Proc Natl Acad Sci USA 114:96239628.

413 40. Saitou N, Nei M. 1987. The neighbor-joining method: A new method for reconstructing

414 phylogenetic trees. Mol Biol and Evol 4:406-425.

415 41. Tamura K, Nei M, and Kumar S. 2004. Prospects for inferring very large phylogenies by using the

416 neighbor-joining method. Proc Natl Acad Sci USA 101:1103011035.

417

418 FIGURE LEGENDS

419 Figure 1. Genomic architecture of the region carrying the mdh gene of Kingella kingae (Kkimdh). The

420 genomic elements are not drawn to scale. The dashed lines indicate that the sensor histidine kinase

421 may be coded by either a single gene or two-component genes.

422 Figure 2. Neighbor-joining tree [40] based on the comparison of mdh nucleotide sequences from 18

423 genetic variants of Kingella kingae and its closest orthologs in Acinetobacter, Acidovorax,

424 Polaromonas, and Neisseria species. The tree is drawn to scale, with branch lengths in the same units

425 as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances

426 were computed using the Maximum Composite Likelihood method [41] and are in the units of the

17 269 427 number of base substitutions per site. The scale bar indicates a 10% nucleotide sequence divergence.

428 Bootstrap values (expressed as percentages of 1,000 replications) are showed next to the branches.

429 Only bootstrap values greater than or equal to 70% were displayed. All positions containing gaps and

430 missing data were eliminated. There was a total of 978 positions in the final dataset.

431

432 TABLE LEGENDS

433 Table 1. Results of the real-time PCR assays targeting the groEL and mdh genes of K. kingae (KkigroEL

434 and Kkimdh) that were tested on 20 genotypically distinct K. kingae isolates, and 201 paediatric

435 specimens.

18 270 (1) Kkimdh (2) (3)

(1) ribosomal small subunit-dependent GTPase (3) GTP cyclohydrolase FolE2 (2) sensor histidine kinase

271 KingellaKingella kingae PYKK081PYKK081 (LT985490)(NZ(LT985490) JH621344) KingellaKingella kingae KK253KK253KK253 (LT985504)(LT985504)(ST-23) KingellaKingella kingae KK238KK238KK238 (LT985483)(LT985483)(ST-23) KingellaKingella kingaekingae ALALIALALI (LT985492)(LT985492)(ST-7) KingellaKingella kingae 112204341122043411220434 (LT985496)(LT985496)(JH768595) 70 KingellaKingella kingae KK107KK107 (LT985499)(LT985499)(ST-6) KingellaKingella kingaekingae ATCCATCC 2333023330 23330TT (LT985503)(LT985503)(ST-1) (FOJK01000000) KingellaKingella kingaekingae AA574AA574 (LT985480) (L(ST-28)T985480) KingellaKingella kingaekingae KK113KK113 (LT985511)(LT985511)(ST-31) KingellaKingella kingae AA068AA068 (LT985507)(LT985507)(ST-35)

82 KingellaKingella kingaekingae AA528AA528 (LT985512)(LT985512) KingellaKingella kingae KK171KK171 (LT985505)(LT985505)(ST-5) KingellaKingella kingae KK247KK247 (LT985513)(LT985513)(ST-38) (CCJT01000000) KingellaKingella kingaekingae KK242KK242 (LT985524)(LT985524)(ST-27) 79 KingellaKingella kingae D2363D2363 (LT985518)(LT985518)(ST-8) 99 KingellaKingella kingaekingae B9853B9853 (LT985517)(LT985517)(ST-12) 80 KingellaKingella kingaekingae B10615B10615 (LT985516)(LT985516)(ST-11) KingellaKingella kingaekingae B10615KK416 (ST-24)(LT985520)(LT985520)

99 PolaromonasPolaromonas naphthalenivoransnaphthalenivorans CJ2 CJ2 (CP000529) (CP000529) AcidovoraxAcidovorax sp.sp. RAC01RAC01 (CP016447) (CP016447)(CP016447) - malate dehydrogenase

98 NeisseriaNeisseria elongata subsp.subsp. glycolyticaglycolyticaglycolytica ATCC ATCCATCC 29315 2931529315 (CP007726) (CP007726)(CP007726) NeisseriaNeisseria sp.sp. KEM232KEM232 (CP022527)(CP022527)(CP022527) 95 NeisseriaNeisseria zoodegmatiszoodegmatis NCTC12230NCTC12230strain NCTC12230 (LT906434)(LT906434) (NZ LT906434) 92 NeisseriaNeisseria weaveriweaveri NCTC13585NCTC13585strain NCTC13585 (LT571436)(LT571436) (LT571436) AcinetobacterAcinetobacter lwoffii iwoffiiiwoffii strainZS207 ZS207 ZS207 (CP019143) (CP019143) (CP019143)

0.1

272 Table 1. Results of the real-time PCR assays targeting the groEL and mdh genes of K. kingae (KkigroEL and Kkimdh) that were tested on 20 genotypically distinct K. kingae isolates, and 201 paediatric specimens.

Kkimdh-positive KkigroEL-positive K. kingae isolates (n=20) 20 20 Other bacterial species (n=86) 0 0 Clinical specimens (n=201)* Initially KkigroEL-positive (n=96) 96 96 Initially KkigroEL-negative (n=105) 8 0

* Patients were tested for a suspected K. kingae invasive infection.

273 CASE REPORT published: xx November 2017 doi: 10.3389/fped.2017.00230

Acute Septic Arthritis of the Knee 001 051 002 Caused by Kingella kingae in a 052 003 053 004 5-Year-Old Cameroonian Boy 054 005 055 056 006 Nawal El Houmami1*, Dimitri Ceroni 2, Karine Codjo Seignon1, Jean-Christophe Pons1, 057 007 Cédric Lambert3, Guillaume André Durand1, Philippe Minodier 4, Léopold Lamah 5, 058 008 Philippe Bidet 6, Jacques Schrenzel 7, Didier Raoult1 and Pierre-Edouard Fournier 1 009 059 010 1 Research Unit on Infectious and Emerging Tropical Diseases (URMITE), UM63, CNRS 7278, IRD 198, INSERM 1095, 060 011 Aix-Marseille Université, IHU Méditerranée Infection, Marseille, France, 2 Département de l’enfant et de l’adolescent, Hôpital 061 3 012 des Enfants, Hôpitaux Universitaires de Genève (HUG), Geneva, Switzerland, Department of Pediatrics, Dracénie Hospital, 062 Draguignan, France, 4 Department of Pediatric Emergency Medicine, North Hospital, Aix-Marseille Université, Marseille, 013 063 France, 5 Department of Orthopedics and Traumatology, Donka University Hospital, University of Conakry Gamal Abdel 014 064 Nasser, Conakry, Guinea, 6 Laboratoire de Microbiologie, Hôpital Robert Debré, Assistance Publique – Hôpitaux de Paris, 015 Université Paris Diderot, Sorbonne Paris Cité, INSERM, IAME, UMR 1137, Paris, France, 7 Bacteriology and Genomic 065 016 Research Laboratories, Geneva University Hospitals (HUG) and Geneva University, Geneva, Switzerland 066 017 067 Edited by: 018 068 Frederick Robert Carrick, Kingella kingae is an important cause of invasive infections in young children from 019 069 University of Cambridge, Western countries. Although increasing reports indicate that this organism is the leading 020 United Kingdom 070 021 agent of bone and joint infections in early childhood, data on K. kingae infections from 071 Reviewed by: 022 John Bernard Ziegler, resource-limited settings are scarce, and none has yet been reported in Africa. We herein 072 023 Sydney Children’s Hospital, Australia report the diagnostic and epidemiological investigations of the irst case of K. kingae 073 024 Enrique Medina-Acosta, 074 025 State University of Norte arthritis identiied in a child from sub-Saharan Africa. A 5-year-old Cameroonian boy pre- 075 026 Fluminense, Brazil sented with a sudden painful limp which appeared in the course of a mild rhinopharyngi- 076 Christian T. K.-H. Stadtlander, 027 tis. He lived in Cameroon where he had been vaccinated with BCG at birth and moved 077 Independent Researcher, 028 078 United States to France for holidays 4 days before consultation. There was no history of trauma and he 029 079 *Correspondence: did not have any underlying medical condition. Upon admission, he had a temperature 030 080 Nawal El Houmami 031 of 36.7°C, and clinical examination revealed right-sided knee tenderness and effusion 081 [email protected] 032 that was conirmed by ultrasound imaging. Laboratory results showed a white blood cell 082 033 3 083 Specialty section: count of 5,700 cells/mm , C-reactive protein level of 174 mg/L, and platelet count of 034 This article was submitted 495,000 cells/mm3. He underwent an arthrocentesis and was immediately given intra- 084 035 085 to Child Health and venous amoxicillin-clavulanate. Conventional cultures from blood samples and synovial 036 Human Development, 086 037 a section of the journal luids were negative. Polymerase chain reaction (PCR) assay targeting the broad-range 087 038 Frontiers in Pediatrics 16S rRNA gene and real-time quantitative PCR assays targeting Mycobacterium species 088 039 Received: 25 August 2017 were negative. Surprisingly, real-time PCR assays targeting the cpn60, rtxA, and rtxB 089 Accepted: 13 October 2017 040 genes of K. kingae were positive. Multicolor luorescence in situ hybridization speciic for 090 041 Published: xx November 2017 091 042 Citation: K. kingae identiied the presence of numerous coccobacilli located within the synovial 092 043 El Houmami N, Ceroni D, luid. Finally, multilocus sequence typing analysis performed on deoxyribonucleic acid 093 044 Codjo Seignon K, Pons J-C, directly extracted from joint luid disclosed a novel K. kingae sequence-type complex. 094 Lambert C, Durand GA, Minodier P, 045 095 Lamah L, Bidet P, Schrenzel J, This case report demonstrates that K. kingae may be considered as a potential cause of 046 096 Raoult D and Fournier P-E (2017) septic arthritis in children living in sub-Saharan Africa, and hence the burden of K. kingae 047 Acute Septic Arthritis of the Knee 097 048 Caused by Kingella kingae in a infection may be not limited to the Western countries. Further studies are required to 098 049 5-Year-Old Cameroonian Boy. determine the prevalence of K. kingae infection and carriage in Africa. 099 050 Front. Pediatr. 5:230. 100 doi: 10.3389/fped.2017.00230 Keywords: Kingella kingae, pediatrics, arthritis, infectious, multilocus sequence typing, Africa South of the Sahara

1

Frontiers in Pediatrics | www.frontiersin.org 274 November 2017 | Volume 5 | Article 230 El Houmami et al. Kingella kingae Arthritis in Cameroon

101 BACKGROUND a septic arthritis of the right knee. Consequently, the child was 158 102 immediately given intravenous amoxicillin-clavulanate 100 mg/kg 159 103 Kingella kingae is an emerging pathogen recognized as the three doses daily during 3 days. 160 104 primary etiology of bone and joint infections in young children 161 105 from Western countries (1, 2). Asymptomatically harbored in 162 106 the oropharynx of children aged 6–48 months, the prevalence DESCRIPTION OF LABORATORY 163 107 of K. kingae oropharyngeal carriage ranges from 8 to 23% from INVESTIGATIONS AND DIAGNOSTIC 164 108 studies carried out in Israel, Switzerland, and New Zealand (3–6). TESTS 165 109 Because this Gram-negative bacterium is usually responsible for 166 110 a mild to moderate inlammatory response, and its detection Because conventional cultures applied for Gram-positive, Gram- 167 111 is notoriously diicult by conventional culture, diagnosis of negative, mycobacterial species, and fungi from the joint luid 168 112 K. kingae infection requires a high index of suspicion and the and blood samples were negative, joint specimens were sent in 169 113 use of adequate detection methods such as real-time quantitative dry ice to the molecular diagnosis laboratory of the URMITE unit 170 114 polymerase chain reaction (qPCR) assays (6, 7). hese molecular in Marseille, where bacterial deoxyribonucleic acid (DNA) was 171 115 diagnostic tools exhibit higher sensitivity compared with culture extracted directly from the joint luid. Polymerase chain reaction 172 116 methods, shorten the time of detection from days to a few hours, (PCR) assay targeting the broad-range bacterial 16S rRNA gene 173 117 and enable the identiication of the organism among healthy (11) and qPCR assays targeting both Mycobacterium species and 174 118 carriers (4–6). Mycobacterium tuberculosis complex (12) were negative. Given 175 119 Large-scale epidemiological studies based on multilocus the age of the patient, K. kingae was also sought by using speciic 176 120 sequence typing (MLST) analysis of K. kingae showed that domi- qPCR assay targeting the K. kingae cpn60 (groEL) gene (11). 177 121 nant clones belonging to sequence-type complexes 6 (STc-6), Surprisingly, this speciic K. kingae assay was positive, as well as 178 122 -14, -23, and -25 accounted for 72% of strains disseminated qPCR assays targeting the Kingella-speciic rtxA and rtxB genes 179 123 worldwide, mainly in the USA, Europe, and Israel, with ST-14 and (7, 13), thus conirming the diagnosis of septic arthritis caused by 180 124 ST-25 being positively associated with osteoarticular infections K. kingae. he organism was also identiied by multicolor luores- 181 125 (8). To date, K. kingae infection and carriage have been studied cence in situ hybridization speciic for K. kingae (Figures S1 and 182 126 in Israel, Europe, North and South America, Australia, New S2 in Supplementary Material), which revealed the presence of 183 127 Zealand, and Japan (5, 8–10), but none have yet been reported large numbers of viable coccobacilli located within the synovial 184 128 in Africa. We herein report the diagnostic and epidemiological luid (Figure 1). Cardiac investigations ruled out endocarditis. 185 129 investigations of K. kingae arthritis in a young, previously healthy A switch to oral amoxicillin-clavulanate 100 mg/kg three doses 186 130 child from Cameroon, and we discuss the clinical implications of daily was then undertaken on 15 July 2016 and was planned for 187 131 these indings. a total duration of 2 weeks. Despite these recommendations, the 188 132 treatment was continued for another 2 months in Cameroon. 189 133 CASE PRESENTATION During the inal follow-up 3 months postoperatively, clinical 190 134 examination revealed a normal knee status with a normal range 191 135 On 11 July 2016, a 5-year-old Cameroonian boy was admitted of motion. 192 136 to the emergency department at the Dracénie Hospital in the hereater, MLST studies using a modiied protocol speciic 193 137 region Provences-Alpes-Côte d’Azur, France, due to a pain- for K. kingae was performed on bacterial DNA extracted 194 138 ful limp that appeared in the morning. He lived in Cameroon directly from the joint luid as previously described (14). Five 195 139 where he had been vaccinated with BCG at birth, and moved alleles were unambiguously identiied, namely, adk-2, aroE-2, 196 140 to Southeastern France for holidays 4 days before consultation. cpn60-2, zwf-13, and recA-2. Unexpectedly, 14 single nucleotide 197 141 A mild rhinopharyngitis had occurred the previous week, but variants of the abcZ allele were identiied from nucleotides 6–447 198 142 as the symptoms were mild, no treatment had been undertaken. (Figure S3 in Supplementary Material; Table 1). To estimate 199 143 here was no history of trauma, and he did not have any underly- the between-strain relatedness and deine an MLST scheme for 200 144 ing medical condition. Upon admission to hospital, the child had a K. kingae, a diferent allele number was given to each distinct 201 145 temperature of 36.7°C and refused to walk. Clinical examination sequence within a locus, and a distinct sequence-type (ST) 202 146 revealed right-sided knee tenderness and efusion. Neither skin number was attributed to each distinct allele combination (15). 203 147 rash nor oral ulcerations were noted. Laboratory results showed K. kingae isolates were then grouped into ST-complexes (STcs) if 204 148 an elevated C-reactive protein (CRP) level at 174 mg/L, with nor- they difered at no more than one locus from at least one other 205 3 149 mal white blood cell count of 5,700 cells/mm and platelet count member of the group. Among the 70 STs of K. kingae that are 206 3 150 of 495,000 cells/mm . Ultrasound imaging conirmed efusion documented in the multilocus sequence database (MLST) of 207 151 of the right knee, whereas conventional radiograph showed no the Institut Pasteur database (http://bigsdb.pasteur.fr/perl/ 208 152 signiicant abnormality. he child underwent an arthrocentesis, bigsdb/bigsdb.pl?db=pubmlst_kingella_seqdef_public&page= 209 153 and mildly opaque and yellowish liquid was extracted, suggesting downloadProiles&scheme_id=1), ST-26, which belongs to 210 154 the highly invasive STc-25, was the closest ST by sharing four 211 155 212 Abbreviations: CRP, C-reactive protein; DNA, deoxyribonucleic acid; MLST, alleles, namely, adk-2, cpn60-2, gdh/zwf-13, and recA-2 with the 156 multilocus sequence typing; PCR, polymerase chain reaction; ST, sequence type; causative strains that were herein identiied (Table 2). Although 213 157 STc, sequence-type complex. analysis of the combination produced by the ive unambiguous 214

2 Frontiers in Pediatrics | www.frontiersin.org 275 November 2017 | Volume 5 | Article 230 El Houmami et al. Kingella kingae Arthritis in Cameroon

215 272 216 273 217 274 218 275 219 276 220 277 221 278 222 279 223 280 224 281 225 282 226 283 227 284 228 285 229 286 230 287 231 288 232 289 233 290 234 291 235 292 236 293 237 294 238 295 239 296 240 297 241 298 242 299 243 300 244 301 245 302

246 FIGURE 1 | Multicolor luorescence in situ hybridization assays were performed on the pathogenic synovial luid after this latter was formalin ixed and parafin 303 247 embedded. Large numbers of viable coccobacilli representing the causative Kingella kingae strains are visualized in red using a rhodamine-labeled probe targeting 304 248 a consensus sequence of the bacterial 16S rRNA gene (EUB388 III, top left), in green using an FITC-labeled probe targeting the K. kingae-speciic V1 region of the 305 249 16S rRNA gene (top right), and in blue using a DAPI probe to label deoxyribonucleic acid (middle left). An internal negative control was performed by using a 306 bacterial non EUB388 probe (middle right). The merged image was obtained summing the four abovementioned images (bottom left). 250 307 251 308 252 309 253 alleles indicated that the causative K. kingae strains belongs to a however, it is largely recognized that Staphylococcus aureus is the 310 254 novel ST, the presence of multiple abcZ alleles does not allow to most common pathogen cultured in children with septic arthritis 311 255 precisely deine it. Moreover, in the MLST scheme of K. kingae, in resource-limited settings (10, 16). Nevertheless, septic arthritis 312 256 founder genotypes of STcs were deined as the ST of the STc caused by S. aureus afects most frequently older children and is 313 257 with the highest number of neighboring STs [(15), Table 3]. more prone to result in a higher systemic inlammatory response 314 258 Consequently, although analysis of the combination produced when compared with K. kingae infections, and the organism is 315 259 by the ive unambiguous alleles indicated that the causative recovered without diiculty by culture of blood and synovial luid 316 260 K. kingae strains belong also to a novel STc, no speciic denomina- aspirates (10, 16, 17). Although K. kingae arthritis is characterized 317 261 tion is yet possible. Moreover, since each of these housekeeping by normal to moderate increase in inlammatory markers, we 318 262 genes is present in one copy in the whole genome of K. kingae, point out that the patient had a markedly elevated CRP level upon 319 263 these indings suggested co-infection by strains belonging to admission, consistent with invasive infection caused by K. kingae 320 264 distinct STs. of at least several days duration. Despite this, K. kingae infection 321 265 was highly suspected because this pathogen is recognized as the 322 266 DISCUSSION irst cause of culture-negative, acute septic arthritis in young 323 267 children and afects most commonly the knee (1). In addition, 324 268 To the best of our knowledge, we herein report the irst case it was also demonstrated that viral respiratory infections may 325 269 of laboratory-conirmed invasive infection due to K. kingae in play a role in the pathogenesis of the disease by damaging the 326 270 a child living in Africa. Little is known of the epidemiology of mucosal lining of the oral cavity, thus facilitating the spread of 327 271 pediatric bone and joint infections in the African continent; the organism from blood to distant anatomic sites (2). 328

3 Frontiers in Pediatrics | www.frontiersin.org 276 November 2017 | Volume 5 | Article 230

El Houmami et al. Kingella kingae Arthritis in Cameroon 329 abcZ_21 386 330 387

331 abcZ_20 388

332 389 333 abcZ_19 390

334 391 335 abcZ_18 392 336 393 337 394

338 abcZ_17 395

339 396 340 abcZ_16 397

341 398 342 abcZ_15 399 343 400

344 401 abcZ_14 345 402 346 403

347 abcZ_13 404

348 405 349 abcZ_12 406 350 407

408 351 abcZ_11 352 409 353 410

354 abcZ_10 411

355 412 356 abcZ_9 413

357 414 358 abcZ_8 415 359 416 360 417

361 abcZ_7 418 97.57 97.35 84.11 95.58 98.45 97.57 96.47 98.45 83.89 97.57 88.08 99.12 97.13 88.08 362 419

363 abcZ_6 420

364 421 365 abcZ_4 422 366 423

367 424 abcZ_3 368 425 369 426

370 abcZ_2 427 371 428

abcZ allele corresponding to the MAFFT alignment displayed in Figure S3 in Supplementary Material. to the MAFFT alignment displayed in Figure abcZ allele corresponding 372 abcZ_1 429

373 430 374 abcZ_5 431 375 432 376 Kingella kingae 433

377 abcZ-R_1574363 434 378 435

379 abcZ-F_1574363 436 380 437 381 438 382 439 383 | Distance matrices of the 440 384 441 abcZ_2abcZ_3abcZ_4abcZ_6 96.69 96.72 97.35 95.58 97.4 96.03 97.57 96.27 88.3 96.03 95.81 98.9 88.12 96.47 87.86 96.47 96.69 96.47 88.96 95.81 96.69 90.29 90.29 87.86 95.81 99.78 88.96 87.86 96.91 88.96 96.69 84.11 96.47 96.47 96.47 94.92 90.07 83.89 97.57 84.33 87.86 95.58 98.01 99.78 91.39 98.01 96.91 96.03 88.96 97.13 97.57 97.35 88.52 96.03 97.35 97.13 89.4 98.45 83.89 96.47 98.68 88.3 97.13 95.81 83.66 90.07 88.96 84.33 96.69 97.35 88.3 95.14 87.64 99.78 88.74 91.17 90.07 98.68 96.25 88.3 96.69 87.64 96.91 88.74 99.78 88.74 90.51 99.78 TABLE 1 TABLE abcZ-F_1574363abcZ-R_1574363abcZ_5 98.53abcZ_1 98.53 98.45 98.23 98.53 98.45 96.69 98.3 98.53 98.23 97.35 98.3 96.72abcZ_7 96.03 97.4 96.69 88.3abcZ_8 96.69 96.27abcZ_9 96.69 95.58 88.12abcZ_10 96.8 96.03 96.72 96.69 97.57abcZ_11 95.81 83.89 96.83 96.72 98.9 96.8 87.86abcZ_12 95.81 84.5 95.36 83.89 96.47 96.83 95.36 97.68abcZ_13 97.79 95.81 96.04 84.5 88.96 95.81 95.81 96.8abcZ_14 99.78 96.04 97.68 97.74 97.79 97.57 83.22 83.22 96.69 95.58abcZ_15 96.58 97.74 96.83 96.8 97.57 96.91 84.55 95.58 96.47 96.25 97.68 96.47 96.61 96.58 84.55 96.47 84.11 96.83 96.47 90.07 96.47 98.01 98.68 97.74 96.61 96.25 94.92 97.68 83.89 95.58 95.58 95.58 83.66 97.57 98.08 96.25 98.68 87.86 97.74 84.33 97.79 98.01 96.25 99.78 96.91 98.01 84.28 97.13 97.79 97.13 96.47 91.39 98.01 98.08 96.47 88.96 88.08 96.03 96.95 97.13 96.91 97.13 99.12 84.33 97.13 96.47 96.47 96.25 98.01 88.52 87.9 84.33 96.03 99.12 97.57 84.77 97.35 99.34 94.7 83 96.47 97.79 96.25 97.13 99.34 98.45 98.08 89.4 97.35 88.08 84.77 97.13 84.33 97.79 88.3 84.33 95.81 96.47 96.49 98.68 94.7 98.23 87.9 84.33 98.23 97.79 87.64 88.96 95.81 96.69 97.13 84.33 97.13 95.81 96.69 88.74 96.47 97.79 88.3 97.35 84.55 84.33 97.13 97.79 99.78 95.36 97.13 83.22 95.81 97.13 84.55 87.64 97.35 97.57 97.79 84.11 96.91 97.13 98.23 83.22 88.74 97.13 84.11 96.25 98.23 96.25 96.91 84.11 97.35 97.35 99.56 96.25 97.13 96.25 84.11 84.55 89.85 98.23 99.56 95.58 99.78 99.34 97.13 97.57 96.69 98.45 84.33 84.77 87.64 98.23 99.56 84.11 94.92 91.39 89.85 96.69 97.35 97.57 96.91 88.74 84.33 96.47 96.69 84.33 88.3 87.64 96.03 98.45 84.11 96.91 83 91.17 96.69 98.45 83.89 89.18 88.74 96.69 97.35 96.91 97.57 88.3 88.08 88.74 97.79 89.18 96.91 98.9 97.13 97.79 88.08 88.74 The 453 nucleotides composing the nucleotide sequence of abcZ allele sequenced from specimen no. 1574363 (abcZ-F_1574363 and abcZ-R_1574363) range between 98.45% with abcZ_5 83.66% ab cZ_16. This table was performed by using Geneious 10.2.3 (Biomatters). The degree of allele similarity is expressed by a Blue scale scheme code, with the most divergent allelles being displayed in dark blue and similar light blue. abcZ_16 83.66 84.28 83 84.33 83.89 83.66 84.33 91.17 84.11 84.55 99.78 84.33 84.11 84.33 83 83.89 83.89 84.55 91.17 84.11 83.89 90.95 abcZ_17 96.91 96.95 95.81 97.79 97.13 96.69 95.14 88.3 97.35 99.34 84.77 94.92 96.91 96.91 96.69 96.91 97.57 84.55 88.08 97.57 96.91 88.08 abcZ_18 88.08 87.9 87.64 88.74 90.07 87.64 88.74 99.78 89.85 87.64 91.39 88.74 88.3 89.18 88.08 88.74 88.08 91.17 88.08 88.52 90.29 99.56 abcZ_19abcZ_20 98.01 98.08 96.47 96.47 96.49 99.78 95.36 97.35 97.79 98.68 99.78 96.25 96.69 88.74 96.91 97.57 90.51 97.35 99.56 84.33 96.69 96.03 84.11 98.45 96.69 97.57 97.35 96.91 97.79 98.9 97.13 97.79 99.12 97.13 84.11 83.89 97.57 96.91 88.52 90.29 97.57 97.57 88.52 90.29 385 abcZ_21 88.08 87.9 87.64 88.74 90.07442 87.64 88.74 99.78 89.85 87.64 91.17 88.74 88.3 89.18 88.08 88.74 88.08 90.95 88.08 99.56 88.52 90.29

4

Frontiers in Pediatrics | www.frontiersin.org 277 November 2017 | Volume 5 | Article 230 El Houmami et al. Kingella kingae Arthritis in Cameroon

443 TABLE 2 | Among the 70 sequence types (STs) of Kingella kingae that are TABLE 3 | Multilocus sequence typing (MLST) scheme of Kingella kingae shows 500 documented in the multilocus sequence database (MLST) of the Institut Pasteur the combination of the six alleles used to deine the sequence types (STs) and 444 501 database (http://bigsdb.pasteur.fr/perl/bigsdb/bigsdb.pl?db=pubmlst_kingella_ sequence-type complexes (STcs) of K. kingae. 445 seqdef_public&page=downloadProiles&scheme_id=1), no. 1574363 shares 502 446 four alleles, namely, adk-2, cpn60-2, gdh/zwf-13, and recA-2, with ST-26, which STc ST abcZ adk aroE cpn60 gdh/zwf recA 503 447 belongs to the ST complex (STc)-25; ST-26 is therefore the closest ST to the 504 1 1 1 1 1 1 1 1 448 causative strains no. 1574363. 505 1 2 1 1 1 1 1 3 449 Reference STc ST abcZ adk aroE cpn60 gdh/zwf recA 506 450 3 3 14 9 14 1 7 4 507 No. 1574363 NA NA NA 2 2 2 13 2 451 NA 4 3 3 9 3 7 3 508 ST-26 25 26 7 2 6 2 13 2 452 NA 5 4 2 9 3 7 3 509 NA indicates data not available. 453 6 6 5 2 4 5 5 1 510 454 511 6 7 5 2 13 5 5 1 455 512 NA 8 11 2 3 7 7 2 456 he detection of K. kingae is currently improved by sensitive 513 457 culture methods such as Bactec/Alert vials, and above all by NA 9 11 2 4 3 4 3 514 458 speciic qPCR assays (2, 7). However, these diagnostic methods NA 10 1 8 3 6 1 3 515 459 are costly and not yet available in developing countries in which 11 11 13 2 4 2 8 6 516 460 diagnostic resources such as blood culture or molecular assays are 11 12 15 2 4 2 8 6 517 461 scarce, and hence the recognition of K. kingae as a possible cause 518 NA 13 3 3 3 3 10 4 462 of acute septic arthritis in pediatrics is particularly challenging. 519 14 14 3 3 3 3 3 3 463 In low-income, high-burden settings of tuberculosis, antibiotics 520 464 with appropriate coverage against S. aureus and classical pyogenic 14 15 3 3 3 3 12 3 521 465 bacteria may be frequently administered without any cultures and 14 16 3 3 12 3 3 3 522 466 in the case of non-response to antibiotic treatment, antitubercu- 14 17 3 2 3 3 3 3 523 467 524 lous drugs may be given empirically for several weeks or months. 14 18 8 3 3 3 3 3 468 Although the child presented with an arthritis caused by 525 NA 19 4 4 4 4 1 3 469 K. kingae 4 days ater arrival in Southeastern France, we highlight 526 NA 20 4 2 3 4 1 3 470 that K. kingae infection usually develop in several days to weeks 527 471 following oropharyngeal K. kingae carriage and viral infections 23 21 10 2 2 2 2 2 528 472 (18). Moreover, MLST analysis of invasive K. kingae strains from 23 22 4 2 2 2 2 2 529 473 Southeastern France in 2016 demonstrated that strains causing 23 23 2 2 2 2 2 2 530 474 osteoarticular infections belonged to ST-6 and ST-25 in the large 531 23 24 2 2 8 2 2 2 475 majority of cases (14). Taken together with the novel K. kingae 532 25 25 7 2 6 2 2 2 476 STc herein described, these indings are consistent with the fact 533 477 that the child acquired causative K. kingae strains in Cameroon. 25 26 7 2 6 2 13 2 534 478 Notably, in an unpublished pilot study, K. kingae has been NA 27 12 6 10 3 9 2 535 479 identiied in the oropharynx of young children from Western 29 28 9 2 7 3 4 3 536 480 Africa. his study was carried out at the Donka University 29 29 9 2 4 3 4 3 537 481 Hospital in Conakry, Guinea, from 2012 to 2013 (Ceroni and 538 NA 30 16 10 7 3 4 3 482 Lamah, unpublished data). To deine the prevalence rate of 539 NA 31 6 1 4 3 1 5 483 oropharyngeal K. kingae carriage, 45 healthy children aged from 540 484 6 to 48 months were enrolled in this study. Children admitted 32 32 6 5 5 3 6 5 541 485 for either elective surgery or attending the orthopedic outpatient NA 33 6 7 11 3 11 5 542 486 clinic or visiting the emergency department for non-infectious NA 34 6 7 11 3 2 5 543 487 disease were included, whereas those presenting an invasive 35 35 1 8 15 8 1 3 544 488 infectious disease, or administration of antimicrobial drugs 545 NA 36 1 11 15 8 1 3 489 the two preceding months were excluded. Recent travel abroad 546 NA 37 3 3 3 3 2 3 490 was not reported in any child. Oropharyngeal specimens were 547 491 obtained by rubbing a cotton swab on the child’s tonsils, which NA 38 6 7 11 3 2 2 548 492 were subsequently tested by molecular assays described earlier 3 39 14 9 14 1 7 10 549 493 (13). hree children tested positive for K. kingae, thus indicat- NA 40 9 2 7 10 4 3 550 494 551 ing a prevalence rate of 6.7%, which is roughly similar to that 14 41 3 3 9 3 3 3 495 observed in Europe (4). Despite the small size of this pilot study, 552 14 42 3 3 3 3 14 3 496 these preliminary results provide evidence that K. kingae is circu- 553 NA 43 3 2 3 3 15 11 497 lating in Western Africa as well, and as a result, K. kingae might 554 498 be considered as a potential pathogen responsible for septic (Continued) 555 499 556

5

Frontiers in Pediatrics | www.frontiersin.org 278 November 2017 | Volume 5 | Article 230 El Houmami et al. Kingella kingae Arthritis in Cameroon

557 TABLE 3 | Continued CONCLUDING REMARKS 614 558 STc ST abcZ adk aroE cpn60 gdh/zwf recA 615 559 his case report demonstrates that K. kingae might be consid- 616 23 44 4 2 2 2 2 7 560 ered as a potential cause of acute septic arthritis in children 617 561 6 45 5 2 4 5 5 9 living in sub-Saharan Africa. Together with the evidence of 618 562 6 46 5 2 6 5 5 1 K. kingae carriage among healthy children from Western Africa, 619 563 NA 47 6 7 11 3 17 12 these indings suggest that K. kingae might contribute to an 620 564 underestimated burden of septic arthritis in this geographical 621 NA 48 1 1 17 2 16 2 565 area. Moreover, MLST analysis disclosed the irst K. kingae STc 622 NA 49 1 1 17 9 16 2 566 in Africa that is a novel STc close to ST-26. Further prospective 623 NA 50 17 1 1 11 1 8 567 studies to specify the prevalence of K. kingae infection and car- 624 568 NA 51 4 6 9 4 1 3 riage in sub-Saharan Africa are required to better help guiding 625 569 35 52 1 8 3 8 1 3 rational diagnostic and therapeutic strategies. 626 570 NA 53 18 2 4 2 9 3 627 571 628 NA 54 3 2 3 3 1 11 CONSENT FOR PUBLICATION 572 629 23 55 19 2 2 2 2 2 573 he written consent for publication was obtained from the par- 630 574 23 56 20 2 6 2 2 2 ents’ child. 631 575 14 57 3 3 3 3 18 3 632 576 35 58 1 8 15 8 19 3 ETHICS STATEMENT 633 577 6 59 5 2 18 5 5 1 634 578 he study was approved by the Ethics committee of the IHU 635 14 60 8 3 3 3 3 13 579 Mediterranee-Infection under reference number 2016-024. 636 6 61 5 2 4 5 20 1 580 637 581 23 62 2 2 6 2 2 2 AUTHOR CONTRIBUTIONS 638 582 NA 63 14 2 19 1 7 10 639 583 NA 64 3 2 16 3 3 3 All the authors provided a substantial contribution to the 640 conception and design of the work, and acquisition, analysis, 584 NA 65 21 7 11 3 17 5 641 585 and interpretation of data for the work. NEH and DC drated 642 32 66 6 5 10 3 6 5 586 the initial version of the manuscript, and all the authors revised 643 NA 67 5 2 3 2 2 2 587 it critically for important intellectual content. All the authors 644 588 NA 68 5 2 6 11 9 1 approved the present version to be published. 645 589 NA 69 1 2 6 2 1 14 646 590 23 70 2 2 20 2 2 2 ACKNOWLEDGMENTS 647 591 648 NA NA NA 2 2 2 13 2 592 he authors are grateful to the patient’s parents, as well as children 649 NA indicates data not yet available. 593 and their family from Conakry, Guinea, for participating to this 650 In the present case, MLST sequencing data from the joint luid specimen no. 1574363 594 indicated that the K. kingae causative strains shared four alleles with ST-26/STc- work. 651 595 25, namely, adk-2, cpn60-2, gdh/zwf-13, and recA-2. Therefore, ST-26/STc-25 is 652 596 the closest ST with the K. kingae strains that were identiied in the synovial luid no. FUNDING 653 1574363 (boxes designed on blue background in the bottom row of the table). 597 654 Please note that among the 70 STs identiied, 31 K. kingae isolates have not 598 yet an STc determined (boxes designed on orange background). his work was supported by the Mediterranee Infection 655 599 foundation (http://www.mediterranee-infection.com/article.php? 656 600 larub=126&titer=la-fondation-recrute) through a PhD grant 657 601 arthritis in young children living in this geographical area. Early awarded to NEH from 2015 to 2017. 658 602 microbiologically proven diagnosis of K. kingae infection would 659 603 enable to provide appropriate antibiotic therapy by amoxicillin, SUPPLEMENTARY MATERIAL 660 604 or amoxicillin-clavulanate, and to drastically reduce the total 661 605 duration of treatment to a few days or weeks (2, 6). his would he Supplementary Material for this article can be found online 662 606 also make it possible to avoid the administration of potentially at http://www.frontiersin.org/article/10.3389/fped.2017.00230/ 663 607 harmful antituberculous regimens. full#supplementary-material. 664 608 665 609 666 610 REFERENCES 2. Yagupsky P, Porsch E, St Geme JW. Kingella kingae: an emerging pathogen in 667 young children. Pediatrics (2011) 127:557–65. doi:10.1542/peds.2010-1867 611 668 1. Yagupsky P. Kingella kingae: from medical rarity to an emerging paedi- 3. Amit U, Dagan R, Yagupsky P. Prevalence of pharyngeal carriage of Kingella 612 atric pathogen. Lancet Infect Dis (2004) 4:358–67. doi:10.1016/S1473- kingae in young children and risk factors for colonization. Pediatr Infect Dis 669 613 3099(04)01046-1 J (2013) 32:191–3. doi:10.1097/INF.0b013e3182755779 670

6

Frontiers in Pediatrics | www.frontiersin.org 279 November 2017 | Volume 5 | Article 230 El Houmami et al. Kingella kingae Arthritis in Cameroon

671 4. Ceroni D, Llana RA, Kherad O, Dubois-Ferriere V, Lascombes P, Renzi G, 13. Cherkaoui A, Ceroni D, Emonet S, Lefevre Y, Schrenzel J. Molecular diagnosis 728 672 et al. Comparing the oropharyngeal colonization density of Kingella of Kingella kingae osteoarticular infections by speciic real-time PCR assay. 729 kingae between asymptomatic carriers and children with invasive osteo- J Med Microbiol (2009) 58:65–8. doi:10.1099/jmm.0.47707-0 673 730 articular infections. Pediatr Infect Dis J (2013) 32:412–4. doi:10.1097/INF. 14. El Houmami N, Bzdrenga J, Pons JC, Minodier P, Durand GA, Oubraham A, 674 0b013e3182846e8f et al. A modiied multilocus sequence typing protocol to genotype Kingella 731 675 5. Olijve L, Podmore R, Anderson T, Walls T. High rate of oropharyngeal kingae from oropharyngeal swabs without bacterial isolation. BMC Microbiol 732 676 Kingella kingae carriage in New Zealand children. J Paediatr Child Health (2017) 17:200. doi:10.1186/s12866-017-1104-5 733 (2016) 52:1081–5. doi:10.1111/jpc.13287 15. Basmaci R, Yagupsky P, Ilharreborde B, Guyot K, Porat N, Chomton M, 677 734 6. Yagupsky P. Kingella kingae: carriage, transmission, and disease. Clin Micro- et al. Multilocus sequence typing and rtxA toxin gene sequencing analysis 678 biol Rev (2015) 28:54–79. doi:10.1128/CMR.00028-14 of Kingella kingae isolates demonstrates genetic diversity and international 735 679 7. El Houmami N, Bzdrenga J, Durand GA, Minodier P, Seligmann H, clones. PLoS One (2012) 7:e38078. doi:10.1371/journal.pone.0038078 736 680 Prudent E, et al. Molecular tests that target the RTX locus do not dis- 16. Stoesser N, Pocock J, Moore CE, Soeng S, Hor P, Sar P, et al. he epidemi- 737 681 tinguish between Kingella kingae and the recently described Kingella ology of pediatric bone and joint infections in Cambodia, 2007–11. J Trop 738 negevensis species. J Clin Microbiol (2017) 55:3113–22. doi:10.1128/ Pediatr (2013) 59:36–42. doi:10.1093/tropej/fms044 682 739 JCM.00736-17 17. Ceroni D, Cherkaoui A, Combescure C, François P, Kaelin A, Schrenzel J. 683 8. Basmaci R, Bidet P, Yagupsky P, Muñoz-Almagro C, Balashova NV, Doit C, Diferentiating osteoarticular infections caused by Kingella kingae from 740 684 et al. Major intercontinentally distributed sequence types of Kingella kingae those due to typical pathogens in young children. Pediatr Infect Dis J (2011) 741 685 and development of a rapid molecular typing tool. J Clin Microbiol (2014) 30:906–9. doi:10.1097/INF.0b013e31821c3aee 742 686 52:3890–7. doi:10.1128/JCM.01609-14 18. El Houmami N, Minodier P, Dubourg G, Mirand A, Jouve JL, Basmaci R, 743 9. Kuzumoto K, Kubota N, Saito Y, Fujioka F, Yumoto K, Hidaka R, et al. et al. Patterns of Kingella kingae disease outbreaks. Pediatr Infect Dis J (2016) 687 744 A case of osteomyelitis due to Kingella kingae. Kansenshogaku Zasshi (2013) 35:340–6. doi:10.1097/INF.0000000000001010 688 87:207–10. doi:10.11150/kansenshogakuzasshi.87.207 745 689 10. Osei L, El Houmami N, Minodier P, Sika A, Basset T, Seligmann H, et al. Conlict of Interest Statement: he authors declare that the research was con- 746 690 Paediatric bone and joint infections in French Guiana: a 6 year retrospective ducted in the absence of any commercial or inancial relationships that could be 747 691 review. J Trop Pediatr (2017) 63:380–8. doi:10.1093/tropej/fmw102 construed as a potential conlict of interest. 748 11. Levy PY, Fournier PE, Fenollar F, Raoult D. Systematic PCR detection in 692 749 culture-negative osteoarticular infections. Am J Med (2013) 126:1143.e25–33. Copyright © 2017 El Houmami, Ceroni, Codjo Seignon, Pons, Lambert, Durand, 693 doi:10.1016/j.amjmed.2013.04.027 Minodier, Lamah, Bidet, Schrenzel, Raoult and Fournier. his is an open-access 750 694 12. Bruijnesteijn van Coppenraet ES, Lindeboom JA, Prins JM, Peeters MF, article distributed under the terms of the Creative Commons Attribution License (CC 751 695 Claas EC, Kuijper EJ. Real-time PCR assay using ine-needle aspirates and BY). he use, distribution or reproduction in other forums is permitted, provided the 752 tissue biopsy specimens for rapid diagnosis of mycobacterial lymphade- original author(s) or licensor are credited and that the original publication in this 696 753 nitis in children. J Clin Microbiol (2004) 42:2644–50. doi:10.1128/JCM.42. journal is cited, in accordance with accepted academic practice. No use, distribution 697 6.2644-2650.2004 or reproduction is permitted which does not comply with these terms. 754 698 755 699 756 700 757 701 758 702 759 703 760 704 761 705 762 706 763 707 764 708 765 709 766 710 767 711 768 712 769 713 770 714 771 715 772 716 773 717 774 718 775 719 776 720 777 721 778 722 779 723 780 724 781 725 782 726 783 727 784

7

Frontiers in Pediatrics | www.frontiersin.org 280 November 2017 | Volume 5 | Article 230 El Houmami et al. BMC Microbiology (2017) 17:200 DOI 10.1186/s12866-017-1104-5

METHODOLOGYARTICLE Open Access A modified multilocus sequence typing protocol to genotype Kingella kingae from oropharyngeal swabs without bacterial isolation Nawal El Houmami1* , Janek Bzdrenga2, Jean-Christophe Pons1, Philippe Minodier3, Guillaume André Durand1, Anis Oubraham1, Dimitri Ceroni4, Pablo Yagupsky5, Didier Raoult1, Philippe Bidet6 and Pierre-Edouard Fournier1

Abstract Background: Outbreaks of Kingella kingae infection are an emerging public health concern among daycare attendees carrying epidemic clones in the oropharynx. However, genotyping of such epidemic clones from affected cases is limited by the low performance of current methods to detect K. kingae from blood samples and lack of specimens available from infected sites. We aimed at developing a modified multilocus sequence typing (MLST) method to genotype K. kingae strains from oropharyngeal samples without prior culture. We designed in silico MLST primers specific for K. kingae by aligning whole nucleotide sequences of abcZ, adk, aroE, cpn60, recA,andgdh/zwf genes from closely related species belonging to the Kingella and Neisseria genera. We tested our modified MLST protocol on all Kingella species and N. meningitidis, as well as 11 oropharyngeal samples from young children with sporadic (n =10)or epidemic (n = 1) K. kingae infection. Results: We detected K. kingae-specific amplicons in the 11 oropharyngeal samples, corresponding to sequence-type 6 (ST-6) in 6 children including the epidemic cases, ST-25 in 2 children, and 3 possible novel STs (ST-67, ST-68, and ST-69). No amplicon was obtained from other Kingella species and N. meningitidis. Conclusions: We herein developed a specific MLST protocol that enables genotyping of K. kingae by MLST directly from oropharyngeal samples. This discriminatory tool, with which we identified the first K. kingae outbreak caused by ST-6 in Europe, may be used in further epidemiological investigations. Keywords: Kingella kingae, MLST, Pediatrics, Outbreaks, Bone and joint infections

Background soft tissue infections, and occasionally endocarditis Outbreaks of Kingella kingae infections are emerging [1–3]. Epidemiological investigation of these events as a public health issue in daycare facilities [1–3]. implies isolation and genotypic characterization of the Defined as the occurrence of at least two epidemio- strain causing the outbreak. At the same time, asymp- logically connected cases of K. kingae infections tomatic daycare center attendees and staff may be within a 1 month-period, they are characterized by a colonized by this virulent strain and, thus, deemed to high attack rate and spread of a virulent clone among be at risk to develop an invasive infection and/or to children aged from 6 to 36 months sharing the same serve as reservoirs and sources of further dissemin- classroom, and causing a variety of osteoarticular and ation of the disease [1, 2]. However, not all colonizing strains are capable of penetrating the epithelial layer * Correspondence: [email protected]; [email protected] and invading the bloodstream, and it is currently rec- amu.fr 1 ognized that worldwide outbreaks are caused by a Aix-Marseille Univ, UM63, CNRS 7278, IRD 198, Inserm 1095, Assistance – Publique – Hôpitaux de Marseille, URMITE, Institut Hospitalo-Universitaire limited number of particularly invasive clones [2 4]. Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13385 Marseille, France Epidemiological investigations revealed that only K. Full list of author information is available at the end of the article

© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. 281 El Houmami et al. BMC Microbiology (2017) 17:200 Page 2 of 5

kingae clones belonging to the hypervirulent sequence 2) maximizing consensus between distinct K. kingae types 6 (ST-6), ST-14, ST-23, ST-25, and ST-66 have sequence types; caused in the past few years outbreaks in the USA, 3) selecting hybridization temperatures close to 58 °C Israel and France [2–4]. (Additional file 1: Figure S1). We designed thereafter Since K. kingae is notoriously difficult to recover in a modified MLST method for K. kingae, consisting culture, real-time polymerase chain reaction (PCR) in PCR amplification and sequencing of 6 assays have been developed during the last 10 years and housekeeping genes, namely abcZ, adk, aroE, cpn60, gained increasing acceptance for the diagnosis of K. recA, and gdh/zwf (Table 1). We first aligned the kingae infections [1, 3, 5]. These culture-independent whole nucleotide sequences of the 6 above- methods exhibit higher sensitivity compared to conven- mentioned housekeeping genes from fourty K. kingae tional cultures, shorten the time of detection from days strains, as well as those from closely-related species to a few hours, enable the diagnosis in patients being including, K. negevensis Sch538T [9], K. denitrificans administered antibiotics, as well as identification of ATCC 33394T,K.oralisATCC 51147T, N. meningitidis asymptomatic K. kingae carriers [2, 5].When no surgical Z2491,N.lactamica020–06, and N. elongata subs. N. specimen is available and blood cultures are negative, al- elongata subs. glycolytica ATCC 29315T. ternative strategies have been developed [1, 5, 6]. Notably, the presence of an oropharyngeal K. kingae car- Finally, we tested this novel K. kingae MLST protocol riage in children under the age of four with sporadic on 11 oropharyngeal samples that had previously been osteoarticular infection was demonstrated to have a tested positive for K. kingae by specific real-time PCR 90.5% positive predictive value for K. kingae infection targeting the cpn60 gene [10], and on DNA from K. [6]. On this point, it was demonstrated that K. kingae denitrificans CIP 103803, K. oralis CIP 103473, K. potus clones carried in the oropharynx of children with K. CIP 108935, K. negevensis Sch538T, and N. meningitidis kingae infection are genotypically identical to those CSUR P782. detected within infected sites [7]. Although the apparent increase in reported cases of Results K. kingae infections can be partly explained by im- K. kingae-specific amplicons were detected by Sanger se- proved isolation methods and better recognition of quencing in all tested oropharyngeal specimens corre- this emerging pathogen, the drawback of molecular sponding to ST-6 in 6, ST-25 in 2, and possible novel detection tests is that, until now, they did not enable STs in 3, but in none of the strains from others Kingella typing of the colonizing organisms and, thus, did not species and N. meningitidis (Table 2). A few single nu- distinguish between individuals carrying non-invasive cleotide polymorphisms (SNPs) were detected in some K. kingae strains and those colonized by the strain alleles for 7 specimens. In these cases, the highest peak which caused the outbreak. We herein report the de- of the chromatogram was selected to determinate the velopment of a modified multilocus sequence typing dominant sequence (Fig. 1). Given that only one copy of protocol (MLST) which enables to genotype K. kingae each reference housekeeping gene was found in the K. in oropharyngeal samples with no prior culture. This kingae KWG1 genome, the only strain for which the method was applied in clinics and successfully used whole genome was sequenced using the highly reliable to investigate an outbreak of invasive K. kingae infec- Pacific Biosciences SMRT technology [11], we postulate tion that occurred in a daycare facility in 2016 in the that K. kingae clones belonging to different STs may co- Marseille area (France). exist in the oropharynx of these individuals where one clone dominated. This method was then applied in clinics in 2016. Methods The study was approved by the Ethics committee of Development of a novel specific MLST typing tool for K. the IHU Mediterranee-Infection under reference num- Kingae ber 2016–024. From June to July 2016, an outbreak We started with the analysis of MLST primers previ- of K. kingae osteoarticular infection involving two ously described in the Institut Pasteur MLST K. kingae infants (aged 17 and 19 months) who shared the database [8], and we observed a lack of in silico primer same classroom was identified in a daycare facility in specificity between the Kingella and Neisseria genera. southern France. The first patient sustained a left Therefore, we designed specific MLST primers for K. ankle arthritis and the second a first metatarsophalan- kingae by using the following criteria: geal joint’s arthritis. Both had presented with herpan- gina,fever,andperi-oralrashinthe2previous 1) maximizing mismatches against other Kingella and weeks. Blood cultures were negative and no joint fluid Neisseria species, especially at the 3′ end; was surgically collected in either case. Both children

282 El Houmami et al. BMC Microbiology (2017) 17:200 Page 3 of 5

Table 1 PCR protocol for specific Kingella kingae multilocus sequence typing Primer design Gene Primers name Primers Primer length (bp) Amplicon length (bp) abcZ abcZ_Kki_Fwd CGCAAGAAAGCGTGTTTGAC 20 532 abcZ_Kki_Rev CAATTCCTGCGCCTTTTTCTC 21 adk adk_Kki_Fwd CACACAAGCGCAATTTATTACG 22 491 adk_Kki_Rev AAACTTCGGTTTGTTCGTGATAT 23 aroE aroE_Kki_Fwd CAAATCCCCACAAATTCATCAATG 24 621 aroE_Kki_Rev AACGCGGTGGGCTGGTTC 18 cpn60 cpn60_Kki_Fwd CATGGGCGCACAAATGGTT 19 467 cpn60_Kki_Rev CAAACAACAACACAAATGGGC 21 recA recA_Kki_Fwd GACGGCAGCCACCAAGAC 21 456 recA_Kki_Rev TCCTGCCAGTTTACGCAAG 19 gdh/zwf gdh/zwf_Kki_Fwd GAGCGCGGCGAGTTTTAT 18 671 gdh/ zwf_Kki_Rev CAGTTGTCCAAAATTGGCATG 21 10× PCR Buffer 1×

25 mM MgCl2 2.0 mM dNTP mix (10 mM of each) 200 μM of each dNTP Forward primer 0,1 μM Reverse primer 0,1 μM HotStarTaq DNA Polymerase 2.5 units/ reaction Distilled water variable Template DNA < 0,5 μg Total volume 50 μl

PCR protocol Cycle step 3 step-protocol Cycles Temperature Time Initial denaturation 95 °C 15 min 1 Denaturation 95 °C 1 min Annealing 58 °C 30 s 35 Elongation 72 °C 1 min 30 s Final elongation 72 °C 10 min 1 recovered with no sequelae after receiving intravenous sporadic infections, such an improved genotyping tool cefamandole followed by oral amoxicillin. An oropha- is relevant. Indeed, it was previoulsy demonstrated ryngeal sample from the second case was collected that the presence of an oropharyngeal invasive K. prior to antibiotic therapy. Detection of K. kingae kingae carriage in children under four with sporadic using specific real-time PCR was positive in this spe- osteoarticular infections had a 90.5% positive predict- cimen. By using our modified MLST typing tool, we ive value for K. kingae infection [6]. Regarding this unambiguously identified K. kingae belonging to ST-6 matter, it is important to note that, of the eleven K. composed of abcZ-5, adk-2, aroE-4, cpn60–5, gdh/ kingae outbreaks in daycare centers that have been zwf-5, and recA-1 alleles. reported to date [2, 3], only 30% of children (10/33) underwent surgical procedures to obtain synovial fluid Discussion or tissue samples. This may be explained by the fact We herein developed a specific MLST method enab- that most infected sites during K. kingae outbreaks ling to genotype K. kingae in oropharyngeal samples were located within small joints located in hands, without requiring prior strain isolation. Given the fas- wrists and feet, ankles, as well as bony sites rich in tidious nature of the species and the increasing use of growth cartilage such as epiphysis of long bones and molecular techniques for investigating epidemics or spine [2]. Since these are regions where joint fluids

283 El Houmami et al. BMC Microbiology (2017) 17:200 Page 4 of 5

Table 2 Specific multilocus sequence typing (MLST) for Kingella kingae performed by Sanger sequencing method on DNA directly extracted from 11 oropharyngeal specimens (=11 children) with no prior bacterial isolation allowed to detect K. kingae clones belonging to ST-6 in 6 children, ST-25 in 2, and possible new STs in 3, namely ST-67, ST-68, and ST-69 No. Age (mo) Year Syndrome Country/region abcZ adk aroE cpn60 gdh/zwf recA ST STc 1572468 17 2016 OAI France 5 2 4 5 5 1 6 6 1980738 16 2016 OAI France 5 2 4 5* 5166 1956884 18 2016 OAI France 5 2 4 5 5 1 6 6 1882247 16 2016 OAI France 5 2 4 5 5 1 6 6 1815589 12 2016 OAI France 5* 24* 55* 1* 66 6847254 8 2016 OAI France 5 2 4* 5* 51* 66 1541670 11 2016 OAI France 7 2 6 2 2 2 25 25 0990626 28 2015 OAI France 7* 26 2 2 2 2525 1822057 7 2016 OAI France 1 2 6 2* 1* 14* 69 … 1730798 33 2013 AC French Guiana 5 2 3* 22 268… 1746575 8 2013 AC French Guiana 5* 26* 11* 9* 1* 67 … *: detection of SNPs OAI osteoarticular infections, AC asymptomatical carriage, mo: months …: not defined Data referring to the second epidemic case of the Châteauneuf-Grasse K. kingae outbreak 2016 are indicated in bold are uncommonly sampled and epiphyseal bone, verte- Moreover, given that 60% of epidemic cases are not brae, or intervertebral disks specimens are rarely microbiologically confirmed during K. kingae outbreaks, obtained, many cases remain unconfirmed [1–3] but and that K. kingae may be difficult to isolate from polymi- are still treated since K. kingae clones carried in the crobial samples even on appropriate culture media [2], oropharynx of children with K. kingae infection are this specific K. kingae MLST tool may be helpful when genotypically identical to those detected within oropharyngeal swabs are the only biological samples avail- infected sites [7]. In two K. kingae outbreaks involv- able for genotyping, as was the case in this report. Clones ing four children in Israel, no suspected cases could belonging to ST-6 are among the most invasive and be formally confirmed [1, 2, 12]. In this peculiar disseminated worldwide and the main cause of K. kingae context, the genotype of epidemic clones was outbreaks in Israel [2, 4]. To the best of our knowledge, obtained from K. kingae oropharyngeal isolates culti- we here report the first K. kingae outbreak caused by ST-6 vated from either presumed cases or from healthy in Europe. Therefore, this genotype appears to be respon- classmates sharing the same classroom [1, 12]. sible for 50% of outbreaks worldwide. Blood cultures and even PCR on blood specimens are really disappointing; although skeletal system infections re- Conclusions sult from the blood-borne dissemination of the bacterium, This modified, specific K. kingae MLST tool demon- the prerequisite bacteremic episode is short and most of strated a high discriminatory power and may be used in the time, when a localized infection has been established, further epidemiological investigations for sporadic and the pathogen has usually been cleared from the blood. epidemic K. kingae infections.

Fig. 1 Chromatograms illustrating single nucleotide polymorphisms that were detected in nucleotide position 150 of the allele abcZ-5 and those identified in nucleotide 336 of the allele gdh/zwf-5 from oropharyngeal sample No. 1815589. In these cases, the highest peak was selected to determinate the dominant clone. The nucleotide positions refer to the corresponding allele reference numbers provided in the Institut Pasteur database (http://bigsdb.pasteur.fr/perl/bigsdb/bigsdb.pl?db=pubmlst_kingella_seqdef_public&page=downloadAlleles)

284 El Houmami et al. BMC Microbiology (2017) 17:200 Page 5 of 5

Additional file 4. Basmaci R, Bidet P, Yagupsky P, Muñoz-Almagro C, Balashova NV, Doit C, et al. Major intercontinentally distributed sequence types of Kingella kingae and development of a rapid molecular typing tool. J Additional file 1: Figure S1. MAFFT alignment of MLST genomic Clin Microbiol. 2014;52:3890–7. regions of the abcZ, adk, aroE, cpn60, gdh/zwf, recA genes from the 40 5. El Houmami N, Minodier P, Dubourg G, Martin-Laval E, Lafont E, Jouve JL, et Kingella kingae strains that were used in this study, and those from 6 al. An outbreak of Kingella kingae infections associated with hand, foot and closely related Kingella and Neisseria species. Only each distinct variant of mouth disease/herpangina virus outbreak in Marseille, France, 2013. Pediatr K. kingae sequence types is represented. MAFFT alignment and figures Infect Dis J. 2015;34:246–50. were performed by using Geneious 10.2.3 (Biomatters). (PPTX 24674 kb) 6. Ceroni D, Dubois-Ferriere V, Cherkaoui A, Gesuele R, Combescure C, Lamah L, et al. Detection of Kingella kingae osteoarticular infections in children by Abbreviations oropharyngeal swab PCR. Pediatrics. 2013;131:e230–5. CIP: Collection de l’institut pasteur; CSUR: Collection de souches de l’unité 7. Basmaci R, Ilharreborde B, Bidet P, Doit C, Lorrot M, Mazda K, et al. Isolation des rickettsies; DNA: Deoxyribonucleic acid; IHU: Institut hospitalo- of Kingella kingae in the oropharynx during K. kingae arthritis in children. universitaire; MLST: Multilocus sequence typing; OAI: Osteoarticular infection; Clin Microbiol Infect. 2012;18:E134–6. PCR: Polymerase chain reaction; SMRT: Single molecule real time; SNP: Single 8. Basmaci R, Yagupsky P, Ilharreborde B, Guyot K, Porat N, Chomton M, et al. nucleotide polymorphism; ST: Sequence type; STc: Sequence type complex Multilocus sequence typing and rtxA toxin gene sequencing analysis of Kingella kingae isolates demonstrates genetic diversity and international Acknowledgements clones. PLoS One. 2012;7:e38078. Not applicable. 9. El Houmami N, Bakour S, Bzdrenga J, Rathored J, Seligmann H, Robert C, et al. Isolation and characterization of Kingella negevensis sp. nov., anovelKingella Funding species detected in a healthy paediatric population. Int J Syst Evol Microbiol. – This work was supported by the Méditerranée Infection foundation. 2017;67:2370 6. 10. Levy PY, Fournier PE, Fenollar F, Raoult D. Systematic PCR detection in – Availability of data and materials culture-negative osteoarticular infections. Am J Med. 2013;126:1143.e25 33. The datasets supporting the conclusions of this article are indicated within 11. Bidet P, Basmaci R, Guglielmini J, Doit C, Jost C, Birgy A, et al. Genome the article and its additional files. analysis of Kingella kingae strain KWG1 reveals how a β-Lactamase gene inserted in the chromosome of this species. Antimicrob Agents Chemother. – Authors’ contributions 2015;60:703 8. NEH, PM, and PEF conceptualized the study. NEH and JB designed the modified 12. Yagupsky P, Ben-Ami Y, Trefler R, Porat N. Outbreaks of invasive Kingella – MLST protocol. NEH and PM collected clinical samples. NEH, JCP, AO, GD, and kingae infections in closed communities. J Pediatr. 2016;169:135 9. JB collected data and carried out the initial analyses. NEH drafted the initial manuscript that was critically revised by PEF, PY, PM, DR, and DC. All authors approved the final manuscript as submitted.

Ethics approval and consent to participate The study was approved by the Ethics committee of the IHU Mediterranee- Infection under reference number 2016–024.

Consent for publication Not applicable.

Competing interests The authors declare that they have no competing interests.

Author details 1Aix-Marseille Univ, UM63, CNRS 7278, IRD 198, Inserm 1095, Assistance Publique – Hôpitaux de Marseille, URMITE, Institut Hospitalo-Universitaire Méditerranée Infection, 19-21 Boulevard Jean Moulin, 13385 Marseille, France. 2University of Grenoble Alpes, CEA, CNRS, IBS, F-38000, Grenoble, France. 3Department of Pediatric Emergency Medicine, North Hospital, Aix-Marseille University, Marseille, France. 4Département de l’enfant et de l’adolescent, HUG-Hôpital des Enfants, Geneva, Switzerland. 5Clinical Microbiology Laboratory, Soroka University Medical Center, Beer-Sheva, Israel. 6Laboratoire de Microbiologie, Hôpital Robert Debré, Assistance Publique – Hôpitaux de Paris, Université Paris Diderot, Sorbonne Paris Cité, Inserm, IAME, UMR 1137, Paris, France.

Received: 26 May 2017 Accepted: 5 September 2017 Submit your next manuscript to BioMed Central and we will help you at every step:

References • We accept pre-submission inquiries 1. Yagupsky P, El Houmami N, Fournier PE. Outbreaks of invasive Kingella • Our selector tool helps you to find the most relevant journal kingae infections in daycare facilities: approach to investigation and • We provide round the clock customer support management. J Pediatr. 2017;182:14–20. 2. El Houmami N, Minodier P, Dubourg G, Mirand A, Jouve JL, Basmaci R, • Convenient online submission et al. Patterns of Kingella kingae disease outbreaks. Pediatr Infect Dis J. • Thorough peer review 2016;35:340–6. • Inclusion in PubMed and all major indexing services 3. El Houmami N, Cointat V, Mirand A, Fouilloux V, Bakour S, Minodier P, et al. An outbreak of Kingella kingae infections complicating a severe • Maximum visibility for your research hand, foot, and mouth disease outbreak in Nice, France, 2016. Pediatr Infect Dis J. 2017;36:530–2. Submit your manuscript at www.biomedcentral.com/submit

285 BACTERIOLOGY crossm

Molecular Tests That Target the RTX Locus Do Not Distinguish between Kingella kingae and the Recently Described Kingella negevensis Species

Nawal El Houmami,a Janek Bzdrenga,a,b Guillaume André Durand,a Philippe Minodier,c Hervé Seligmann,a Elsa Prudent,a Sofiane Bakour,a Stéphane Bonacorsi,d Didier Raoult,a Pablo Yagupsky,e Pierre-Edouard Fourniera Aix-Marseille Université, Research Unit on Infectious and Emerging Tropical Diseases (URMITE), UM63, CNRS 7278, IRD 198, Inserm 1095, Institut Hospitalo-Universitaire Méditerranée Infection, Assistance Publique- Hôpitaux de Marseille, Marseille, Francea; Université Grenoble Alpes, CEA, CNRS, IBS, F-38000, Grenoble, Franceb; Department of Pediatric Emergency, North Hospital, Marseille, Francec; Inserm, IAME, UMR 1137, Université Paris-Diderot, Sorbonne Paris Cité, Laboratoire de Microbiologie, Hôpital Robert-Debré, Assistance Publique-Hôpitaux de Paris, Paris, Franced; Clinical Microbiology Laboratory, Soroka University Medical Center, Beer-Sheva, Israele

ABSTRACT Kingella kingae is an important invasive pathogen in early childhood. The organism elaborates an RTX toxin presumably restricted to this species. Conse- Received 5 May 2017 Returned for quently, real-time quantitative PCR (qPCR) assays targeting the RTX locus have been modification 2 June 2017 Accepted 31 July 2017 developed in recent years and are gaining increasing use for the molecular diagno- Accepted manuscript posted online 9 sis of K. kingae infections. However, the present study shows that Kingella negeven- August 2017 sis,aKingella species newly identified in young children, harbors an identical Kin- Citation El Houmami N, Bzdrenga J, Durand gella RTX locus, raising the question of whether K. negevensis can be misidentified as GA, Minodier P, Seligmann H, Prudent E, Bakour S, Bonacorsi S, Raoult D, Yagupsky P, Fournier K. kingae by clinical microbiology laboratories. In silico comparison of Kingella sp. P-E. 2017. Molecular tests that target the RTX RTX and groEL genes and in vitro studies provided evidence that targeting the rtxA locus do not distinguish between Kingella and rtxB genes could not differentiate between strains of K. kingae and K. negeven- kingae and the recently described Kingella negevensis species. J Clin Microbiol sis, whereas targeting the groEL gene could. This prompted the design of a highly 55:3113–3122. https://doi.org/10.1128/JCM specific and sensitive qPCR assay targeting K. negevensis groEL (kngroEL). Ninety-nine .00736-17. culture-negative osteoarticular specimens from 99 children younger than 4 years of Editor Alexander J. McAdam, Boston Children's age were tested with a conventional 16S rRNA gene-based broad-range PCR assay Hospital Copyright © 2017 American Society for and Kingella-specific rtxB, K. kingae-specific groEL (kkgroEL), and kngroEL qPCR assays. Microbiology. All Rights Reserved. Forty-two specimens were rtxB positive, including 41 that were also kkgroEL positive Address correspondence to Nawal El and 1 (the remaining one) that was kngroEL positive. Thus, this study discloses an Houmami, [email protected]. invasive infection caused by K. negevensis in humans and demonstrates that target- ing the RTX locus cannot be used for the formal diagnosis of K. kingae infections. These findings stress the need for further studies on the epidemiology of asymp- tomatic carriage and invasive infections caused by K. negevensis in humans.

KEYWORDS IS1 family, Kingella kingae, Kingella negevensis, RTX toxins, osteoarticular infections, pediatrics, qPCR, real-time PCR

or the past 5 decades, since the original characterization of Kingella kingae in the F1960s (1, 2), four additional species have been included in the Kingella genus, namely, K. denitrificans (3), K. oralis (4), K. potus (5), and (most recently) K. negevensis (6). Kingella organisms are asymptomatically harbored in the oropharynx in humans and animals, and all members of the genus Kingella but K. negevensis have been incrimi- nated so far in invasive infections affecting, peculiarly, the musculoskeletal system, and occasionally the cardiac and central nervous systems, in humans (3–7). Kingella bacteria belong to the large Neisseriaceae family and are notoriously fastidious, and their

October 2017 Volume 55 Issue 10 Journal of Clinical286 Microbiology jcm.asm.org 3113 El Houmami et al. Journal of Clinical Microbiology recovery in routine culture media is suboptimal, which initially made their recognition as human pathogens difficult (7). In the early 1990s, the serendipitous discovery that inoculation of skeletal system exudates into blood culture vials improved the isolation of K. kingae revealed that this organism was a common etiology of joint and bone infections in young children (8). Subsequently, the advent of molecular diagnostic tools further improved the detection of K. kingae and established the species as the leading cause of skeletal system infections in children aged 6 to 48 months in countries where these modern detection methods are routinely employed (9–15). Initially, PCR assays targeting the small-subunit 16S rRNA gene followed by sequencing of the resulting amplicons enabled improvement of the detection of the organism from osteoarticular samples (9, 10). Thereafter, the development of a K. kingae-specific real-time quantita- tive PCR (qPCR) assay targeting the groEL gene (also known as cpn60, hsp60,ormopA), a housekeeping gene encoding a chaperone protein recognized as a universal bacterial marker (16, 17), allowed a further increase in the diagnostic capability for pediatric K. kingae arthritis compared to that of traditional PCR or culture methods (11–15). Subsequently, numerous in-house qPCR assays were optimized in an attempt to achieve maximum levels of specificity, sensitivity, and rapid detection of K. kingae for a wide array of clinical specimens, including joint fluids, bone (12–15), oropharyngeal specimens (18, 19), and occasionally blood samples (19–21), cerebrospinal fluid (22), or cardiac tissues (20). In 2007, Kehl-Fie and St. Geme, III, identified within the K. kingae genome a locus belonging to the RTX toxin superfamily that comprises 5 genes, namely, rtxB, rtxD, rtxC, rtxA, and tolC, encoding a cytotoxic and hemolytic toxin that plays a pivotal role in the tissue invasiveness of the bacterium (23, 24). As determined by lactic dehydrogenase (LDH) release assays, Southern blotting (23), and qPCR targeting rtxA (15), the less invasive K. oralis, K. denitrificans, and K. potus species were found to lack this RTX locus, paving the way for the development and implementation of qPCR assays targeting rtxA and rtxB to diagnose K. kingae infection (14, 15). Nevertheless, extended analysis unveiled that this K. kingae RTX locus is flanked by insertion elements and possesses a reduced GC content, strongly suggesting acquisition by horizontal gene transfer (23). However, the type strain K. negevensis Sch538, which shares close phenotypic and genomic relatedness with K. kingae (6), harbors an identical Kingella RTX locus. Because the Kingella RTX locus was previously considered to be restricted to K. kingae, this raised the question of whether qPCRs targeting rtxA, rtxB, and groEL, which are in widespread use for the diagnosis of K. kingae infection, can differentiate K. kingae from K. negeven- sis. The present study provides in silico and in vitro evidence that qPCR assays targeting the RTX locus cannot discriminate K. kingae from K. negevensis but that those targeting K. kingae groEL (kkgroEL) can. Therefore, a specific qPCR targeting K. negevensis groEL (kngroEL) was designed to discriminate K. negevensis from all other members of the Kingella genus. This novel and highly sensitive and specific K. negevensis qPCR test was then performed on 99 culture-negative osteoarticular samples from children less than 4 years of age, enabling us to disclose the first case of K. negevensis arthritis in children (Fig. 1). Finally, the DDER transposase element ISKne1, located in the 3= region of the K. negevensis RTX locus and presumably involved in this interspecies gene transfer, was further identified and characterized.

RESULTS Validation of Kingella negevensis-specific qPCR. The qPCR assay targeting kn- groEL was positive for all K. negevensis strains, while no amplification was observed for K. kingae ATCC 23330T or for K. denitrificans, K. oralis, and K. potus strains (Table 1). Similarly, no amplification was obtained for the other 115 bacterial species tested. To determine the detection limit of the method, 13-fold serial dilutions of a bacterial suspension of strain 538T at an initial concentration of 108 bacteria mlϪ1 in phosphate- buffered saline were evaluated and further quantified by culture on Columbia blood agar (bioMérieux) and colony counting. This method showed a detection threshold of 50 CFU/ml.

October 2017 Volume 55 Issue 10 287 jcm.asm.org 3114 Kingella negevensis Harbors a Kingella sp. RTX Locus Journal of Clinical Microbiology

FIG 1 Workflow diagram for this study.

Targeting the RTX locus is not specific to Kingella kingae. In silico analysis of primers and probes targeting the rtxA and rtxB genes failed to discriminate between K. negevensis and K. kingae, whereas those targeting groEL showed high discriminative power for both Kingella species. As expected, all 20 K. negevensis strains were positive for rtxA and rtxB, whereas no amplification was observed with kkgroEL (Table 1), therefore suggesting that K. negevensis harbors a constitutional RTX locus. Patient characteristics and identified pathogens. The median age of the 99 children with culture-negative bone and joint fluids included in this study was 23.1 Ϯ 5.1 months, and the male-to-female ratio was 1.3:1. PCRs were positive for 45 (45%) children (Table 2). The molecular assay targeting the 16S rRNA gene was positive for 8 (8%) patients, in whom K. kingae was further identified in 5, Streptococcus spp. in 2, and S. aureus in 1. Mycobacterium tuberculosis was not detected in any specimen. Overall, the kkgroEL PCR assay was positive for 41 children, including the 5 cases in which the 16S rRNA gene of K. kingae was also detected by conventional PCR and sequencing. One child had a dual infection caused by both K. kingae detected by qPCR and Streptococcus pyogenes identified by 16S rRNA gene PCR, which was further confirmed by a specific qPCR targeting the S. pyogenes mipB gene. Overall, 42 children were

TABLE 1 Primer and probe specificities in qPCR assays targeting the rtxA, rtxB, K. negevensis groEL, and K. kingae groEL genes (42 ؍ K. kingae (n (20 ؍ K. negevensis (n qPCR qPCR Gene target Reference In silico specificity result In silico specificity resulta rtxA 15 No mismatch ϩ 0 or 1 mismatch ϩ 14 1 mismatch ϩ 0 to 2 mismatches ϩ rtxB 14 No mismatch ϩ 0 to 2 mismatches ϩ K. negevensis groEL No mismatch ϩ 15 or 16 mismatches Ϫ K. kingae groEL 12 9 or 10 mismatches Ϫ 0 to 2 mismatches ϩ 13 11 or 12 mismatches Ϫ 0 or 1 mismatch ϩ aTested on 10 distinct K. kingae strains.

October 2017 Volume 55 Issue 10 288 jcm.asm.org 3115 El Houmami et al. Journal of Clinical Microbiology

TABLE 2 Molecular assays used to identify the 45 pathogens causing culture-negative osteoarticular infections in 99 children between 6 and 48 months of age No. of isolates Bacterial species 16S rRNA genea Kingella rtxBb kkgroELa kngroEL S. aureus nucAa S. pyogenes mipBc Kingella kingae 54141000 Kingella negevensis 010100 Staphylococcus aureus 100020 Streptococcus pyogenes 200002 Mycobacterium tuberculosisd 000000 Total (no. [%]) 8 (8) 42 (42) 41 (41) 1 (1) 2 (2) 2 (2) aData from reference 13. bData from reference 14. cDetermined by use of an in-house real-time PCR assay targeting mipB of Streptococcus pyogenes, using the forward primer Spyo_mipB_F (CCATACGGTTATAGTAAGGA GCCAAA), the reverse primer Spyo_mipB_R (GGCTATCACATCACAGCAACC), and the probe Spyo_mipB_P (FAM-TCAGCGCCAGCTTCAATGGC). dThere were no positive results for detection of the M. tuberculosis ITS (36). positive for rtxB, including the 41 previously mentioned kkgroEL-positive children and 1 child who was positive for kngroEL. Both rtxB and kngroEL were detected at high threshold cycle (CT) values in the joint fluid from the right hip of an 8-month-old boy, indicating a low concentration of K. negevensis. The DNA content extracted from this one K. negevensis-positive specimen was not sufficient for sequencing of the groEL gene. Genomic analysis of the RTX locus in Kingella spp. The RTX locus was found in all 42 K. kingae strains and, unexpectedly, in K. negevensis Sch538T and SW7208426. The Kingella negevensis RTX locus comprises the 5 genes required for the production and secretion of an active toxin, organized in a way similar to that in K. kingae (Fig. 2). Nucleotide sequence identities for rtxA between strain Sch538T and K. kingae ranged from 97.53% to 99.58% with K. kingae AA574 and KK113, respectively, and those for rtxB ranged from 99.01% to 99.58% with K. kingae KKWG1 and ATCC 23330T, respectively. At the protein level, K. negevensis Sch538T RtxA exhibited 99.48% identity with K. kingae RtxA, whereas RtxB showed 100% identity. Characterization of a novel ISKne1 transposase element. A 705-bp mobile element (ISKne1), located 1,121 bp downstream of the tolC gene and flanked by two nearly perfect terminal inverted repeats of 29 bp, was identified within K. negevensis strains Sch538T and SW7204826. Spread in multiple copies within K. negevensis ge- nomes and sometimes bordered by an 8-bp direct repeat (TAGCTGTT), ISKne1 was present in DNA regions with GC contents ranging from 30.0 to 36.0%, contrasting with the 45.5% GC content of the K. negevensis genome (6, 25). The presumed ISKne1 protein contains 234 amino acids, including D106, D165, E191, and R198, which are catalytic amino acid residues highly conserved in IS1 transposases with a DDER motif (26)(Fig. 2). ISKne1 was deposited in the IS Finder database (http://www-is.biotoul.fr). Phylogeny of groEL in K. negevensis. All groEL genes of K. negevensis were 1,638 bp long, and that of strain Sch538T showed nucleotide sequence identities ranging from 78.33% with Neisseria meningitidis Z2491 to 82.97% with K. kingae ATCC 23330T. A neighbor-joining tree created in MEGA7 (27) indicated 5 clusters of K. negevensis groEL genes, paralleling those previously described by using pulsed-field gel electro- phoresis (PFGE) (Fig. 3)(6).

DISCUSSION Overall, the measured sensitivity of qPCR assays is 1 order of magnitude higher than that of conventional PCR tests that target the universal 16S rRNA gene (28–30). Because the RTX toxin is elaborated by all K. kingae strains examined so far, the encoding RTX locus genes have been used repeatedly as species-specific assay targets for detecting the organism in normally sterile body fluids and tissues in numerous clinical microbi- ology laboratories worldwide (14, 15, 30–33). K. kingae strains exhibiting RTX locus polymorphisms, including nonsynonymous ones, have also been detected (15). Assum- ing optimal specificity, molecular detection of the rtxA and rtxB genes in oropharyngeal

October 2017 Volume 55 Issue 10 289 jcm.asm.org 3116 Kingella negevensis Harbors a Kingella sp. RTX Locus Journal of Clinical Microbiology

A RTX locus

(1) ISL rtxB rtxD rtxC rtxA tolC ISR (2)

Kingella kingae *

Kingella negevensis

(3) (4)

IRR ISKne1 IRL

B

D106 D165 E198 R203 N-terminal C-terminal

Zinc finger HNH Catalytic domain

ISL: Insertion sequences left (1) signal recognition particle-docking protein Fts Y ISR: Insertion sequences right (2) protocatechuate 3,4-dioxygenase IRL: inverted repeat left GGTGGTATCCTGAAAAGTGTGCTGTAAAA (3) NAD-dependant deacetylase IRR: inverted repeat right GGTGGTGTCCTGAAAAGTGTGCTGTAAAA (4) Lytic tranglycosylase HNH: Helix-turn-helix * Pseudogenized ISKne1

FIG 2 (A) Genomic organization of the RTX locus in Kingella kingae strain KKWG1 and K. negevensis strain SW7208426 and of the ISKne1 transposase element. (B) Schematic representation of the functional domains and catalytic amino acid residues of the DDER transposase ISKne1. specimens has also been proposed as a strategy to confirm K. kingae as the etiology of pediatric skeletal system infections in cases where clinical specimens from joints, bones, or intervertebral disks were not obtained or gave negative culture results (18, 34). The present study provides in silico and in vitro evidence that genes encoding the RTX toxin are not exclusive to K. kingae and shows that qPCR assays designed to detect the RTX locus of K. kingae fail to distinguish K. kingae from K. negevensis. Therefore, the presence of rtxA and rtxB cannot be considered to formally confirm K. kingae infection, raising the pivotal question of whether K. negevensis may have been misidentified as K. kingae in bone tissue, joint fluid, or oropharyngeal samples. In contrast, in silico analysis of a large panel of orthologous groEL genes from closely related Kingella and Neisseria species, including 43 K. kingae and 20 K. negevensis strains, showed high discriminative power between all bacterial species for testing of primers and probes routinely used for the molecular identification of K. kingae. This was successfully confirmed in vitro, as a qPCR targeting kkgroEL failed to detect all 20 of the K. negevensis strains as well as strains of other Kingella spp. Taken together, these results confirm that targeting groEL is a valuable strategy for the molecular diagnosis of invasive K. kingae infections and carriage (12, 13). Among the 45 children with positive PCR results in our series, K. kingae was identified in 91% (41/45 cases) of cases. Thus, K. kingae was the leading cause of culture-negative osteoarticular infections in children aged 6 to 48 months, paralleling the results of numerous prior studies (11–13). Notably, these findings also confirmed the superiority of qPCR targeting groEL for the molecular diagnosis of K. kingae infection, with an increase in the detection yield of the organism, to 88% (36/41 cases), compared to the results of conventional 16S rRNA gene PCR (28–30). As expected, all

October 2017 Volume 55 Issue 10 290 jcm.asm.org 3117 El Houmami et al. Journal of Clinical Microbiology

FIG 3 Neighbor-joining tree based on groEL nucleotide sequences of 20 strains of Kingella negevensis and other Kingella and Neisseria members. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Kimura 2-parameter method and are presented as numbers of base substitutions per site. The rate variation among sites was modeled with a gamma distribution (shape parameter ϭ 8). Bootstrap values (expressed as percentages of 1,000 replications) are displayed at the nodes. All positions containing gaps and missing data were eliminated. There was a total of 1,635 positions in the final data set. Clones related by PFGE pattern are indicated on the right and include clones ␩ (n ϭ 3),T(n ϭ 6), T= (n ϭ 1), unique (n ϭ 2), and b (composed of b1 [n ϭ 2], b2 [n ϭ 2], b3 [n ϭ 1], and b5 [n ϭ 1]). the kkgroEL-positive samples were also positive for rtxB, which is consistent with studies reporting joint fluids positive for both K. kingae kkgroEL and rtxA (12, 15). To further fill the gap in knowledge for cases where culture-negative osteoarticular specimens from children younger than 4 years old are also negative for kkgroEL and other pediatric pathogens, a highly sensitive and specific qPCR targeting kngroEL that enabled discrimination of K. negevensis from K. kingae was developed. Note that one joint fluid aspirate from the hip of an 8-month-old boy was positive for both kngroEL and rtxB, thus providing the first clinical evidence of K. negevensis arthritis in humans, which highlights the fact that the molecular detection of the RTX locus alone cannot be considered irrefutable proof of K. kingae infection. Finally, we found a conserved synteny of the chromosomal region carrying the RTX locus in Kingella spp., in which we identified a novel DDER transposase element (ISKne1) present in multiple copies within the whole genome of K. negevensis, suggesting a transposase activity (26, 35). ISKne1 has presumably acted with neighboring insertion sequences to enable RTX locus accretion, thus facilitating its inter-Kingella-species transfer, though the precise mechanism of transposition remains to be elucidated. Functional genomics studies previously showed that rtxA is required for the cytotoxic activity of K. kingae on human epithelial, synovial, and macrophage-like cells (23). Given the percentage of amino acid sequence identity of the RTX toxin of K. negevensis and that of K. kingae, their close genomic characteristics, and their shared habitat in the

October 2017 Volume 55 Issue 10 291 jcm.asm.org 3118 Kingella negevensis Harbors a Kingella sp. RTX Locus Journal of Clinical Microbiology oropharynx of healthy children (6), it is possible that K. negevensis invades human tissues in a manner similar to that of K. kingae, with oropharyngeal colonization being a required step prior to invasive disease. In conclusion, K. negevensis harbors a constitutional RTX locus similar to that of K. kingae, and as a result, positive assays targeting rtxA and rtxB only cannot be consid- ered definitive evidence of K. kingae infection. Similarly, there is a high risk that previous epidemiological data on K. kingae oropharyngeal carriage in children aged 6 to 48 months originating from studies using assays targeting rtxA and rtxB may be slightly overestimated because of the detection of both Kingella species. However, this study also emphasizes that the predictive value of a positive rtxB assay remains high (98%) for the diagnosis of osteoarticular infections in children younger than 4 years old. Additionally, the first invasive infection caused by K. negevensis in a young patient suggests that this organism may be an occasional player in bone and joint infections in early childhood. Taken together, these findings highlight the need for further studies to better understand the epidemiology of K. negevensis carriage and disease, as well as the clinical presentation and pathogenesis of infections in humans, for which the highly sensitive and specific qPCR assay described herein may be highly relevant.

MATERIALS AND METHODS A workflow diagram for this study is provided in Fig. 1. Bacterial strains. In the 2000s, epidemiological studies were conducted in the Negev Desert region of southern Israel on 7,217 healthy children younger than 8 years old, from whom K. kingae and K. negevensis strains were isolated at the clinical microbiology laboratory in Beer-Sheva, Israel (6). From these isolates, 20 strains of K. negevensis and 42 strains of K. kingae were used in the present study. All 20 K. negevensis strains were isolated from the oropharynxes of healthy children aged between 6 months and 8 years, after pharyngeal specimens were inoculated onto BAV medium, a selective vancomycin- containing medium, to inhibit the competing Gram-positive flora and facilitate the recognition of K. kingae colonies (6). K. negevensis isolates showed an atypical phenotype consisting of unusually long chains of coccobacilli, early autolysis, poor growth as pinpoint ␤-hemolytic colonies on blood-agar plates, and excellent growth on GC-base medium. The 42 K. kingae strains were derived from children aged 6 to 48 months with osteoarticular infections (n ϭ 12), occult bacteremia (n ϭ 4), endocarditis (n ϭ 5), or asymptomatic oropharyngeal colonization (n ϭ 21). Additionally, 115 other bacterial strains belonging to 43 genera, including other Kingella species and members of the Neisseria, Haemophilus, Staphylococcus, Streptococcus, and Mycobacterium genera, were used to determine the specificity of the qPCR primers and probes targeting K. negevensis groEL (see Table S1 in the supplemental material). Kingella species strains were cultured on 5% sheep blood-enriched Columbia agar for 24 to 36 h at 37°C in a 5% enriched

CO2 atmosphere. Culture-negative osteoarticular samples. From 2014 to 2016, 99 culture-negative bone and joint samples from children aged 6 to 48 months with suspected osteoarticular infection were retrieved at the URMITE Laboratory, Marseille, France, where they were stored at Ϫ20°C. By applying a diagnostic approach consisting of systematic PCR for the diagnosis of osteoarticular infections in culture-negative patients (13), samples were tested by conventional broad-range PCR targeting the 16S rRNA gene (13) and specific qPCRs targeting K. kingae groEL (kkgroEL), Staphylococcus aureus nucA (13), and the M. tuberculosis internal transcribed spacer (ITS) (36)(Fig. 1). When the conventional broad-range PCR targeting the 16S rRNA gene was positive for Streptococcus spp., a second control was obtained by an in-house specific qPCR assay targeting the mipB gene of Streptococcus pyogenes (Table 2). In cases where negative controls gave a positive result, the PCR was considered invalid and was repeated. The efficiency of DNA extraction and the possible presence of inhibitors in the samples were evaluated using the RS42-Km primer pair, targeting a fragment of the human ␤-globin gene, as previously described (13). Genomic DNA extraction. Genomic DNAs of all bacterial strains and clinical samples were extracted with a BioRobot EZ1 workstation and an EZ1 DNA tissue kit (Qiagen, Courtaboeuf, France) according to the manufacturer’s recommendations. DNA was stored at Ϫ80°C until required for analysis. To negate the effects of PCR inhibitors, extracted DNAs of Kingella species strains were tested both undiluted and diluted 1:10, 1:100, and 1:1,000, and those of clinical samples were tested both undiluted and diluted 1:10. Sequencing of the groEL genes and RTX loci of Kingella kingae strains. Paired-end sequencing of the complete groEL genes and RTX loci of the 42 K. kingae strains by use of an Illumina MiSeq instrument was performed as previously described (6). In addition, groEL and RTX locus nucleotide sequences from K. kingae KKWG1 (LN869922), for which the genome is available in one scaffold (37), were included in this study and used as references for the comparative analysis of Kingella spp. Sequencing of the groEL genes of Kingella negevensis strains. The set of K. negevensis primers for conventional PCR, namely, groEL_Knegev_F1 (5=-CTGGTGATGCGTGAAGAAG-3=) and groEL_Knegev_R3 (5=-TCCTATATGAAACAATTGCC-3=), located 106 bp upstream and 155 bp downstream of the groEL gene of K. negevensis Sch538T (accession no. CCNJ01000030) and SW7208426 (accession no. FXBH01000000), were manually designed. For the 18 remaining K. negevensis strains, an 1,899-bp PCR product was

October 2017 Volume 55 Issue 10 292 jcm.asm.org 3119 El Houmami et al. Journal of Clinical Microbiology

amplified by using HotStarTaq DNA polymerase (Qiagen). Amplification parameters consisted of 1 cycle of 95°C for 5 min followed by 39 cycles of 94°C for 1 min, 55°C for 30 s, and 72°C for 2 min 30 s, with 1 final extension cycle of 72°C for 10 min. The amplicons were purified using a NucleoFast 96 PCR kit (Macherey-Nagel, Düren, Germany) according to the manufacturer’s recommendations. Sequencing reactions were carried out by using a BigDye Terminator v1.1 cycle sequencing kit (Perkin-Elmer, Shelton, CT) according to the manufacturer’s instructions. Sequencing of the products was achieved on both strands by using the initial set of PCR primers and the internal primers groEL_Knegev_F2 (5=-CGCAAG TAGGTTCTATCTCTGC-3=), groEL_Knegev_F3 (5=-GGCTGTGATTAAAGTGGG-3=), groEL_Knegev_R1 (5=-CCA ATGATTTGCCGTCTTCC-3=), and groEL_Knegev_R2 (5=-CACCACCTGCAACGATACC-3=), using an ABI Prism 3130 genetic analyzer. Sequence assembly was performed in Geneious R9.1.8 (Biomatters). qPCR targeting the Kingella negevensis groEL gene (kngroEL). (i) Design of primers and probe. To design specific primers and a probe for K. negevensis qPCR, an alignment of the groEL nucleotide sequences obtained from 20 K. negevensis strains, 42 K. kingae strains, K. kingae KKWG1 (LN869922), K. denitrificans ATCC 33394T (accession no. AEWV01000000), K. oralis ATCC 51147T (accession no. ACJW02000000), N. meningitidis Z2491 (accession no. AL157959), Neisseria elongata subsp. glycolytica ATCC 29315T (accession no. NZ_CP007726), and Neisseria lactamica 020-06 (accession no. NC_014752) was performed by using MAFFT, version 7.22.2, in Geneious R9.1.8 (38). Thereafter, the primers groEL_F_Knegev (5=-CACGTTCTGCATTGAAATCTG-3=) and groEL_R_Knegev (5=-GTTCACTACTACAGACGC TTC-5=), located at nucleotides 1265 and 1401, respectively, within the groEL gene of K. negevensis, and the probe groEL_P_Knegev (5=-6-carboxyfluorescein [FAM]-CGCTGACCAAGAAGCTGGCGTG), located at nucleotide 1299, were manually designed. Particular care was taken in order to (i) minimize mismatches in the primers and probe between K. negevensis strains, especially at the 3= end, and (ii) maximize mismatches with other Kingella and Neisseria species. The primer and probe specificity was confirmed in silico by using the BLAST tool (http://blast.ncbi.nlm.nih.gov). (ii) Kingella negevensis-specific qPCR. TaqMan real-time PCR amplification and hybridization reactions were carried out in a final volume of 20 ␮l of reaction mixture containing 10 ␮l of Takyon No Rox Probe MasterMix dTTP (Eurogentec), 0.25 ␮M (each) primers, 0.25 ␮M labeled probe, and 5 ␮lof purified DNA. Amplification of a 137-bp product was performed on a Bio-Rad CFX96 platform in a C1000 Touch thermal cycler, using the following cycling parameters: heating at 50°C for 2 min and 95°C for 5 min, followed by 40 cycles of a two-stage temperature profile of 95°C for 10 s and 55°C for 45 s. Kingella sp. qPCR assays. In silico analysis of the specificities of primers and probes targeting rtxA (14, 15), rtxB (14), and K. kingae groEL (kkgroEL)(12, 13), all used routinely for the molecular diagnosis of K. kingae infection, was first performed. All 20 K. negevensis strains were then tested using qPCR assays targeting the kngroEL, kkgroEL, rtxA, and rtxB genes, as previously described (12–15). In addition to the above-mentioned qPCR tests, all 99 clinical samples were tested for kngroEL and rtxB. Genomic analysis of the RTX loci of Kingella spp. MAFFT alignment of the RTX loci of K. negevensis Sch538T (6), K. negevensis SW7208426 (25), and K. kingae KKWG1 (37) was performed in Geneious R9.1.8. Ethics statement. This study was approved by the Ethics Committee of the IHU Méditerranée Infection under reference number 2016-024. Epidemiological studies performed in the 2000s were approved by the Ethics Committee of the Soroka University Medical Center, as well as by the Israel Ministry of Health. Accession number(s). The GenBank accession numbers for the groEL genes from the 43 K. kingae strains analyzed in this study are LT838407 to LT838420, and those for the groEL genes from the 20 K. negevensis strains are LT631522 to LT631541. The GenBank accession numbers for the rtxA genes from the 43 K. kingae strains are LT841363 to LT841378, CCNJ01000000, FXBH01000000, FOJK01000000, CCJT01000000, and NZ_LN869922, and those for the rtxB genes are LT841334 to LT841346, CCNJ01000000, FXBH01000000, FOJK01000000, CCJT01000000, and NZ_LN869922.

SUPPLEMENTAL MATERIAL Supplemental material for this article may be found at https://doi.org/10.1128/JCM .00736-17. SUPPLEMENTAL FILE 1, PDF file, 0.1 MB. SUPPLEMENTAL FILE 2, PDF file, 0.2 MB.

ACKNOWLEDGMENTS We acknowledge Patricia Siguier, Laboratoire de Microbiologie et Génétique Moléculaires, CNRS UMR 5100, Université Paul Sabatier, Toulouse, France, for her helpful comments. This work was supported by the Méditerranée Infection foundation and was carried out via the A*MIDEX project (project ANR-11-IDEX-0001-02), funded by the Investisse- ments d’Avenir French government program, which is managed by the French National Research Agency (ANR). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. We have no conflicts of interest to disclose.

October 2017 Volume 55 Issue 10 293 jcm.asm.org 3120 Kingella negevensis Harbors a Kingella sp. RTX Locus Journal of Clinical Microbiology

REFERENCES 1. Henriksen SD, Bøvre K. 1968. Moraxella kingii sp.nov., a haemolytic, Lamah L, Manzano S, Hibbs J, Schrenzel J. 2013. Detection of Kingella saccharolytic species of the genus Moraxella. J Gen Microbiol 51: kingae osteoarticular infections in children by oropharyngeal swab PCR. 377–385. https://doi.org/10.1099/00221287-51-3-377. Pediatrics 131:e230–e235. https://doi.org/10.1542/peds.2012-0810. 2. Henriksen SD, Bøvre K. 1976. Transfer of Moraxella kingae Henriksen and 19. El Houmami N, Minodier P, Dubourg G, Martin-Laval A, Lafont E, Jouve Bøvre to the genus Kingella gen. nov. in the family Neisseriaceae. Int J JL, Charrel R, Raoult D, Fournier PE. 2015. An outbreak of Kingella kingae Syst Evol Microbiol 26:447–450. infections associated with hand, foot and mouth disease/herpangina 3. Snell JJS, Lapage SP. 1976. Transfer of some saccharolytic Moraxella virus outbreak in Marseille, France, 2013. Pediatr Infect Dis J 34:246–250. species to Kingella Henriksen and Bøvre 1976, with descriptions of https://doi.org/10.1097/INF.0000000000000572. Kingella indologenes sp. nov. and Kingella denitrificans sp. nov. Int J Syst 20. El Houmami N, Cointat V, Mirand A, Fouilloux V, Bzdrenga J, Bakour S, Evol Microbiol 26:451–458. Minodier P, Dubois MA, Anave-Frapech F, Charrel R, Raoult D, Fournier 4. Dewhirst FE, Chen CK, Paster BJ, Zambon JJ. 1993. Phylogeny of species PE. 2017. An outbreak of Kingella kingae infections complicating a severe in the family Neisseriaceae isolated from human dental plaque and hand, foot, and mouth disease outbreak in Nice, France, 2016. Pediatr description of Kingella oralis sp. nov. Int J Syst Bacteriol 43:490–499. Infect Dis J 36:530–532. https://doi.org/10.1097/INF.0000000000001487. https://doi.org/10.1099/00207713-43-3-490. 21. Ceroni D, Cherkaoui A, Kaelin A, Schrenzel J. 2010. Kingella kingae 5. Lawson PA, Malnick H, Collins MD, Shah JJ, Chattaway MA, Bendall R, spondylodiscitis in young children: toward a new approach for bacteri- Hartley JW. 2005. Description of Kingella potus sp. nov., an organism ological investigations? A preliminary report. J Child Orthop 4:173–175. isolated from a wound caused by an animal bite. J Clin Microbiol https://doi.org/10.1007/s11832-009-0233-2. 43:3526–3529. https://doi.org/10.1128/JCM.43.7.3526-3529.2005. 22. Le Bourgeois F, Germanaud D, Bendavid M, Bonnefoy R, Desnous B, 6. El Houmami N, Bakour S, Bzdrenga J, Rathored J, Seligmann H, Robert C, Beyler C, Blauwblomme T, Elmaleh M, Pierron C, Lorrot M, Bonacorsi S, Armstrong N, Schrenzel J, Raoult D, Yagupsky P, Fournier PE. 12 July Basmaci R. 2016. Kingella kingae sequence type 25 causing endocarditis 2017. Isolation and characterization of Kingella negevensis sp. nov., a with multiple and severe cerebral complications. J Pediatr 169: novel Kingella species detected in a healthy paediatric population. Int J 326–326.e1. https://doi.org/10.1016/j.jpeds.2015.10.091. Syst Evol Microbiol https://doi.org/10.1099/ijsem.0.001957. 23. Kehl-Fie TE, St Geme JW, III. 2007. Identification and characterization of 7. Yagupsky P. 2004. Kingella kingae: from medical rarity to an emerging an RTX toxin in the emerging pathogen Kingella kingae. J Bacteriol paediatric pathogen. Lancet Infect Dis 4:358–367. https://doi.org/10 189:430–436. https://doi.org/10.1128/JB.01319-06. .1016/S1473-3099(04)01046-1. 24. Chang DW, Nudell YA, Lau J, Zakharian E, Balashova NV. 2014. RTX toxin 8. Yagupsky P, Dagan R, Howard CW, Einhorn M, Kassis I, Simu A. 1992. plays a key role in Kingella kingae virulence in an infant rat model. Infect High prevalence of Kingella kingae in joint fluid from children with septic Immun 82:2318–2328. https://doi.org/10.1128/IAI.01636-14. arthritis revealed by the BACTEC blood culture system. J Clin Microbiol 25. El Houmami N, Schrenzel J, Yagupsky P, Robert C, Ceroni D, Raoult D, 30:1278–1281. Fournier PE. 2017. Draft genome sequence of Kingella negevensis 9. Moumile K, Merckx J, Glorion C, Berche P, Ferroni A. 2003. Osteoarticular SW7208426, the first European strain of K. negevensis isolated from a infections caused by Kingella kingae in children: contribution of poly- healthy child in Switzerland. Genome Announc 5:e00571-17. https://doi merase chain reaction to the microbiologic diagnosis. Pediatr Infect Dis .org/10.1128/genomeA.00571-17. J 22:837–839. 26. Ton-Hoang B, Turlan C, Chandler M. 2004. Functional domains of the IS1 10. Verdier I, Gayet-Ageron A, Ploton C, Taylor P, Benito Y, Freydiere AM, transposase: analysis in vivo and in vitro. Mol Microbiol 53:1529–1543. Chotel F, Bérard J, Vanhems P, Vandenesch F. 2005. Contribution of a https://doi.org/10.1111/j.1365-2958.2004.04223.x. broad range polymerase chain reaction to the diagnosis of osteoarticu- 27. Kumar S, Stecher G, Tamura K. 2016. MEGA7: Molecular Evolutionary lar infections caused by Kingella kingae: description of twenty-four Genetics Analysis version 7.0 for bigger datasets. Mol Biol Evol 33: recent pediatric diagnoses. Pediatr Infect Dis J 24:692–696. https://doi 1870–1874. https://doi.org/10.1093/molbev/msw054. .org/10.1097/01.inf.0000172153.10569.dc. 28. Fournier PE, Drancourt M, Colson P, Rolain JM, La Scola B, Raoult D. 2013. 11. Chometon S, Benito Y, Chaker M, Boisset S, Ploton C, Bérard J, Vanden- Modern clinical microbiology: new challenges and solutions. Nat Rev esch F, Freydiere AM. 2007. Specific real-time polymerase chain reaction Microbiol 11:574–585. https://doi.org/10.1038/nrmicro3068. places Kingella kingae as the most common cause of osteoarticular 29. Morel AS, Dubourg G, Prudent E, Edouard S, Gouriet F, Casalta JP, infections in young children. Pediatr Infect Dis J 26:377–381. https://doi Fenollar F, Fournier PE, Drancourt M, Raoult D. 2015. Complementarity .org/10.1097/01.inf.0000259954.88139.f4. between targeted real-time specific PCR and conventional broad-range 12. Ilharreborde B, Bidet P, Lorrot M, Even J, Mariani-Kurkdjian P, Liguori S, 16S rDNA PCR in the syndrome-driven diagnosis of infectious diseases. Vitoux C, Lefevre Y, Doit C, Fitoussi F, Penneçot G, Bingen E, Mazda K, Eur J Clin Microbiol Infect Dis 34:561–570. https://doi.org/10.1007/ Bonacorsi S. 2009. New real-time PCR-based method for Kingella kingae s10096-014-2263-z. DNA detection: application to samples collected from 89 children with 30. Filleron A, Michon AL, Jumas-Bilak Jeziorski E, Zorngiotti I, Tran TA, acute arthritis. J Clin Microbiol 47:1837–1841. https://doi.org/10.1128/ Filleron T, Rodière M, Marchandin H. 2015. What to expect from JCM.00144-09. molecular tools for non-documented pediatric infectious diseases. 13. Levy PY, Fournier PE, Fenollar F, Raoult D. 2013. Systematic PCR detec- Expert Rev Mol Diagn 15:1645–1656. https://doi.org/10.1586/ tion in culture-negative osteoarticular infections. Am J Med 126: 14737159.2015.1105132. 1143.e25–1143.e33. https://doi.org/10.1016/j.amjmed.2013.04.027. 31. Ceroni D, Cherkaoui A, Ferey S, Kaelin A, Schrenzel J. 2010. Kingella 14. Cherkaoui A, Ceroni D, Emonet S, Lefevre Y, Schrenzel J. 2009. Molecular kingae osteoarticular infections in young children: clinical features diagnosis of Kingella kingae osteoarticular infections by specific real-time and contribution of a new specific real-time PCR assay to the diag- PCR assay. J Med Microbiol 58:65–68. https://doi.org/10.1099/jmm.0 nosis. J Pediatr Orthop 30:301–304. https://doi.org/10.1097/BPO .47707-0. .0b013e3181d4732f. 15. Lehours P, Freydière AM, Richer O, Burucoa C, Boisset S, Lanotte P, Prère 32. Williams N, Cooper C, Cundy P. 2014. Kingella kingae septic arthritis in MF, Ferroni A, Lafuente C, Vandenesch F, Mégraud F, Ménard A. 2011. children: recognizing an elusive pathogen. J Child Orthop 8:91–95. The rtxA toxin gene of Kingella kingae: a pertinent target for molecular https://doi.org/10.1007/s11832-014-0549-4. diagnosis of osteoarticular infections. J Clin Microbiol 49:1245–1250. 33. Slinger R, Moldovan I, Bowes J, Chan F. 2016. Polymerase chain reaction https://doi.org/10.1128/JCM.01657-10. detection of Kingella kingae in children with culture-negative septic 16. Goh SH, Potter S, Wood JO, Hemmingsen SM, Reynolds RP, Chow AW. arthritis in eastern Ontario. Paediatr Child Health 21:79–82. https://doi 1996. HSP60 gene sequences as universal targets for microbial species .org/10.1093/pch/21.2.79. identification: studies with coagulase-negative staphylococci. J Clin Mi- 34. Ceroni D, Belaieff W, Kanavaki A, Anderson Della Llana R, Lascombes P, crobiol 34:818–823. Dubois-Ferrière V, Dayer R. 2013. Possible association of Kingella kingae 17. Hill JE, Penny SL, Crowell KG, Goh SH, Hemmingsen SM. 2004. cpn60DB: with infantile spondylodiscitis. Pediatr Infect Dis J 32:1296–1298. https:// a chaperonin sequence database. Genome Res 14:1669–1675. https:// doi.org/10.1097/INF.0b013e3182a6df50. doi.org/10.1101/gr.2649204. 35. Siguier P, Gagnevin L, Chandler M. 2009. The new IS1595 family, its 18. Ceroni D, Dubois-Ferriere V, Cherkaoui A, Gesuele R, Combescure C, relation to IS1 and the frontier between insertion sequences and trans-

October 2017 Volume 55 Issue 10 294 jcm.asm.org 3121 El Houmami et al. Journal of Clinical Microbiology

posons. Res Microbiol 160:232–241. https://doi.org/10.1016/j.resmic 2015. Genome analysis of Kingella kingae strain KWG1 reveals how a .2009.02.003. ␤-lactamase gene inserted in the chromosome of this species. Anti- 36. Bruijnesteijn van Coppenraet ES, Lindeboom JA, Prins JM, Peeters MF, microb Agents Chemother 60:703–708. https://doi.org/10.1128/AAC Claas EC, Kuijper EJ. 2004. Real-time PCR assay using fine-needle aspi- .02192-15. rates and tissue biopsy specimens for rapid diagnosis of mycobacterial 38. Katoh K, Misawa K, Kuma K, Miyata T. 2002. MAFFT: a novel method lymphadenitis in children. J Clin Microbiol 42:2644–2650. https://doi for rapid multiple sequence alignment based on fast Fourier trans- .org/10.1128/JCM.42.6.2644-2650.2004. form. Nucleic Acids Res 30:3059–3066. https://doi.org/10.1093/nar/ 37. Bidet P, Basmaci R, Guglielmini J, Doit C, Jost C, Birgy A, Bonacorsi S. gkf436.

October 2017 Volume 55 Issue 10 295 jcm.asm.org 3122 Eur J Clin Microbiol Infect Dis (2017) 36:1971–1974 DOI 10.1007/s10096-017-3022-8

ORIGINAL ARTICLE

Emergence of Clostridium difficile tcdC variant 078 in Marseille, France

N. Cassir1 & N. Fahsi1 & G. Durand1 & J.-C. Lagier1 & D. Raoult1 & P.-E. Fournier1

Received: 7 March 2017 /Accepted: 22 May 2017 /Published online: 1 June 2017 # Springer-Verlag Berlin Heidelberg 2017

Abstract The purpose of this investigation was to evaluate the such as RT027 and tV078. This emphasises the need for an epidemiology of hypervirulent Clostridium difficile ribotypes efficient surveillance system for CDI with ribotyping and an from January 2013 to February 2017 in the Marseille area of optimised management of CDI caused by hypervirulent strains. southern France. By using the Xpert Clostridium difficile Epi polymerase chain reaction (PCR) assay and sequencing the tcdC gene, we characterised C. difficile isolates from symptom- Introduction atic patients diagnosed with C. difficile infection (CDI) in Marseille university hospitals. We first tested retrospectively Clostridium difficile is the leading cause of healthcare- 278 C. difficile samples isolated from January 2013 to associated diarrhoea in developed countries, representing a December 2014 and observed a high prevalence of isolates with great clinical and economical burden [1]. This infection usu- tcdC mutations and deletions previously described in both hy- ally occurs in elderly patients with co-morbidities in whom the pervirulent ribotypes RT027 and RT078 (16.4% and 10.7%, gut microbiota has been disrupted by a previous antibiotic respectively). We highlighted the co-circulation of these two therapy [2]. Over the past decade, the hypervirulent hypervirulent C. difficile tcdC variants (tV) with distinct epide- fluoroquinolone-resistant polymerase chain reaction (PCR) miological characteristics. While an RT027 outbreak occurred ribotype NAP1/027 C. difficile (RT027), associated with an mainly as healthcare-associated infection in the elderly, CDI increased severity and mortality, emerged worldwide [3]. caused by tV078 occurred mainly in a younger population as From January 2013 to October 2014, an RT027 outbreak oc- community-associated infection. From January 2016, a system- curred in the Marseille area, southeastern France [4, 5]. In atic survey of these two hypervirulent C. difficile ribotypes total, 19 different healthcare facilities reported 112 RT027 revealed the emergence of CDI caused by tV078, currently infections, with a mortality rate of 31.2%. Successful interven- being more prevalent than RT027 in the Marseille area. The tions included the monitoring of the regional RT027 burden present study is the first report of the emergence of CDI caused and the cohorting of RT027-positive patients in a specialised by tV078 in southern France. We showed the simultaneous infectious diseases ward. In addition, an optimised manage- circulation and sequential spread of hypervirulent ribotypes, ment including faecal microbiota transplantation (FMT) led to a significant reduction of mortality [6]. However, other hypervirulent C. difficile ribotypes (RT), * N. Cassir including RT017, RT078 and RT244, have also emerged re- [email protected] cently [3, 7], and some studies reported an increase in * P.-E. Fournier community-associated cases caused by these RTs, affecting [email protected] younger patients [8, 9]. In France, data from the National Reference Laboratory for C. difficile indicate that toxinotype 1 Unité de Recherche sur les Maladies Infectieuses Tropicales et V/ribotype 078 significantly increased (3.25% vs. 11.1%) in Emergentes (URMITE), Aix Marseille Université, UM63, CNRS northern France from 2006 to 2007 (comparison period: July 7278, IRD 198, INSERM 1095, Institut Hospitalier Universitaire (IHU)—Méditerranée Infection, 19–21 Boulevard Jean Moulin, to December) [10]. In the present study, we retrospectively 13385 Marseillle Cedex 05, France sequenced the tcdC gene for genotyping 278 C. difficile

296 1972 Eur J Clin Microbiol Infect Dis (2017) 36:1971–1974 strains collected in our laboratory during the RT027 outbreak tcdC variant tV078 has a 39-base-pair deletion and a mutation that developed from January 2013 to December 2014 in the at position 184 (C instead of T). Marseille area. The aim was to retrospectively evaluate wheth- In a second aim, from January to January 2017 (inclusive), er more than one hypervirulent ribotype circulated during this all stool specimens positive for C. difficile in Marseille public outbreak. In a second aim, we prospectively and specifically hospitals were systematically tested using the same method in searched in 2016 the most common hypervirulent RTs (i.e. order to differentiate RT027 and tV078 from other C. difficile RT027 and RT078) in all patients diagnosed with C. difficile ribotypes. infection (CDI) in Marseille university hospitals. Statistical analysis

Materials and methods PASW Statistics version 17.0 was used for the statisti- cal analysis. Mean ± standard deviation was used to Definitions describe continuous variables. Percentages and numbers of events were used for quantitative variables. Student’s Patients were considered to have CDI if they had diarrhoea t-test or the Mann–Whitney test, when appropriate, (≥3 unformed stools per 24 h) and a stool sample positive for were used to perform two-group comparisons for quan- C. difficile toxin by laboratory assay. Severe CDI was defined titative variables. The Chi-squared test or Fisher’sexact if at least one of the following signs was present in the absence test, when appropriate, were used for qualitative vari- of another explanation: white blood count (WBC) >15 × 109/ ables. A significance threshold of 0.05 was adopted for L; fever (core body temperature >38.5 °C); colectomy; ileus; all of the statistical analyses. megacolon; peritonitis; septic shock requiring intensive care unit (ICU) admission; serum creatinine concentration >50% above the baseline; serum albumin concentration <30 g/L; Results death. Community-associated CDI was defined as a case that either had a diagnosis of CDI in the outpatient setting with no We first tested retrospectively 278 C. difficile samples history of hospital discharge in the 12 weeks before diagnosis, isolated from symptomatic patients with CDI from or a primary diagnosis upon hospital admission and no history January 2013 to December 2014 and observed a high of hospital discharge in the 12 weeks before diagnosis. prevalence of isolates with tcdC mutations and deletions Patients not fulfilling these criteria were diagnosed as having previously described in both hypervirulent ribotypes healthcare-associated CDI. An outbreak in a healthcare facil- RT027 and RT078 (16.4% and 10.7%, respectively). We ity was defined as the occurrence of ≥2 epidemiologically highlighted the co-circulation of these two hypervirulent linked cases within one week. Mortality was considered to C. difficile tcdC variants (tV) with distinct epidemiologi- be attributable to CDI when a patient died of the consequences cal characteristics. Indeed, the in-hospital mortality rates of CDI during hospitalisation. of infections caused by tV078 and RT027 during the study period were 23% (7/30) and 39% (18/46), respec- Detection and genotyping of C. difficile isolates tively (p = 0.16). Compared with patients with RT027, patients with CDI due to tV078 were younger (69.8 years Of the 614 stool specimens from patients with CDI diagnosed ± 18.1 vs. 82.3 years ± 12.3; p <0.001)andmorefre- from January 2013 to December 2014 using the Xpert quently had community-associated CDI (9/30 vs. 3/46; C. difficile Epi PCR assay (Cepheid, la Serre, France) in the p = 0.03). Using the European Centre for Disease two point-of-care routine laboratories located in the Timone Prevention and Control (ECDC) criteria, 58% of tV078 and North university hospitals in Marseille [11], we retrospec- patients and 70% of RT027 patients met the definition tively studied the 278 (45.2%) samples for which at least of severe disease (17/30 vs. 32/46; p = 0.26). Patients 20 μL of DNA remained. The Xpert C. difficile Epi assay with RT027 CDI were more likely to receive vancomycin detects the genes encoding toxin B (tcdB) and the binary toxin than metronidazole, in line with treatment recommenda- (cdt), as well as the tcdC gene deletion at nt117, allowing to tions favouring oral vancomycin for severe CDI [14]. detect PCR ribotype 027 C. difficile strains [12]. We also Among the 30 patients with tV078 CDI, 15 were treated amplified by PCR and sequenced a fragment of the tcdC gene with vancomycin and 11 with metronidazole, three were coding for a protein regulating the C. difficile toxin secretion, initially started on metronidazole and then switched to as previously described [13]. This fragment has the advantage vancomycin, and one received FMT. Two cases required of being variable according to the ribotype. For example, in ICU admission. addition to the above-described 18-base-pair deletion at posi- Compared with patients with CDI due to ribotypes other tion 117, RT027 has a mutation at the same position, while the than tV078 and RT027, patients with CDI due to tV078 had a

297 Eur J Clin Microbiol Infect Dis (2017) 36:1971–1974 1973

Fig. 1 Evolution of the number of samples positive for Clostridium difficile ribotype (RT) 027, tcdC variant (tV) 078 and other RTs from January 2016 to February 2017 in Marseille, France

higher in-hospital mortality rate [23% (7/30) vs. 15% (30/ [5], CDI caused by tV078 occurred mainly in a younger pop- 202); p = 0.47], were younger (69.8 years ± 18.1 vs. 73.2 years ulation as community-associated infection. ± 10.4; p < 0.001) and more frequently had community- On the basis of these preliminary findings, we performed, associated CDI (9/30 vs. 20/202; p =0.02). from January 2016, a systematic survey of C. difficile tV078 From January 2016 to February 2017, among all positive and RT027, which revealed the emergence of CDI caused by C. difficile strains detected in the laboratory, we identified 10 the former ribotype, currently representing the most prevalent (1.2%) RT027, 67 (7.7%) tV078 and 788 (91.1%) other encountered hypervirulent ribotype in the Marseille area. ribotypes (Fig. 1). During the second half of 2016 (from Access to C. difficile ribotyping is crucial to detect and control July to December), we observed a significant increase in the CDI outbreaks according to the strains’ ribotype specificities rate of tV078 when compared to all C. difficile-positive sam- [19], hypervirulent CD ribotypes requiring optimised manage- ples [39/342 (11.4%) vs. 10/416 (2.4%); p < 0.0001]. In 2016, ment and rigorous infection control measures. As we have no healthcare-associated outbreaks caused by this hyperviru- developed specific control measures for RT027-infected pa- lent ribotype have been notified in the hospitals of the tients, including cohorting in a specialised infectious diseases Marseille area. ward and systematic FMT, resulting in a significantly de- creased mortality [6], the identification of other co- circulating hypervirulent C. difficile clones requiring a similar Discussion management highlights the need to prospectively and system- atically detect these clones. In the present study, we observed that both hypervirulent We acknowledge that this study has some limitations. First, ribotypes RT027 and tV078 co-circulated from January the number of samples tested is low and only stools for which 2013 to December 2014 in the Marseille area. sufficient DNA quantity remained for genotyping procedures Although CDI has been considered a hospital-acquired dis- were analysed. Second, detailed clinical characteristics and ease mainly transmitted by symptomatic patients, recent stud- antimicrobial susceptibility testing were not assessed. Third, ies have indicated the existence of other sources of C. difficile, the methodology used in this study allowed us to distinguish showing that some CDI cases cannot be linked to a previous only the C. difficile isolates with tcdC deletions and mutations case of CDI [15, 16]. Indeed, C. difficile is ubiquitous in the previously found in ribotypes RT027 and RT078. Moreover, environment, and most RTs are shared among humans, ani- as mentioned before, other hypervirulent ribotypes such as mals and the environment [17]. In particular, RT078, a recent- RT017 or RT244 have recently been described. We intend, ly described genotype that was demonstrated to be associated in a further study, to analyse, by multi-locus sequence typing with an increased morbidity and mortality and to occur in (MLST), all of our C. difficile strains in order to better distin- younger patients than RT027, has been detected in farm ani- guish all the known PCR ribotypes. mals, especially pigs, with human and animal strains being genetically related [18]. In this study, we showed a co- circulation of two hypervirulent RTs with distinct epidemio- Conclusion logical characteristics. While an RT027 outbreak occurred mainly as healthcare-associated infection in elderly patients, This is the first report of the emergence of Clostridium difficile the most probable source being a single long-term care facility infection (CDI) caused by the tcdC variant tV078 in

298 1974 Eur J Clin Microbiol Infect Dis (2017) 36:1971–1974 southeastern France. The emergence of so-called hyperviru- nasogastric route: a preliminary report. Eur J Clin Microbiol – lent C. difficile types has intensified the challenge of CDI Infect 34:1597 1601. doi:10.1007/s10096-015-2394-x 7. Lim SK, Stuart RL, Mackin KE, Carter GP,Kotsanas D, Francis MJ epidemiology. The present study provides a comprehensive et al (2014) Emergence of a ribotype 244 strain of Clostridium view of the current paradigm of CDI, revealing the simulta- difficile associated with severe disease and related to the epidemic neous circulation and sequential spread of hypervirulent ribotype 027 strain. Clin Infect Dis 58:1723–1730. doi:10.1093/cid/ ribotypes. This emphasises the need for an efficient surveil- ciu203 8. Goorhuis A, Bakker D, Corver J, Debast SB, Harmanus C, lance system for CDI with ribotyping and an optimised man- Notermans DW et al (2008) Emergence of Clostridium difficile agement of CDI caused by hypervirulent strains. infection due to a new hypervirulent strain, polymerase chain reac- tion ribotype 078. Clin Infect Dis 47:1162–1170. doi:10.1086/ Acknowledgements The authors acknowledge the members of the mo- 592257 lecular biosciences team and the staff of the infectious diseases ward at 9. Eyre DW, Tracey L, Elliott B, Slimings C, Huntington PG, Stuart Nord hospital, Marseille, France. RL et al (2015) Emergence and spread of predominantly community-onset Clostridium difficile PCR ribotype 244 infection in Australia, 2010 to 2012. Euro Surveill 20:21059 Compliance with ethical standards 10. RupnikM,WidmerA,ZimmermannO,EckertC,BarbutF(2008) Clostridium difficile toxinotype V, ribotype 078, in animals and Financial support No financial support was provided relevant to this humans. J Clin Microbiol 46:2146. doi:10.1128/JCM.00598-08 article. 11. Cohen-Bacrie S, Ninove L, Nougairède A, Charrel R, Richet H, Minodier P et al (2011) Revolutionizing clinical microbiology lab- Conflict of interest All authors report no conflicts of interest relevant oratory organization in hospitals with in situ point-of-care. PLoS to this article. One 6:e22403. doi:10.1371/journal.pone.0022403 12. Babady NE, Stiles J, Ruggiero P, Khosa P, Huang D, Shuptar S et al Ethical approval The institutional board approved the study protocol. (2010) Evaluation of the Cepheid Xpert Clostridium difficile Epi assay for diagnosis of Clostridium difficile infection and typing of the NAP1 strain at a cancer hospital. J Clin Microbiol 48:4519– Informed consent Informed consent was not obtained from the pa- 4524. doi:10.1128/JCM.01648-10 tients, as the study design was retrospective and did not modify standard 13. Baldan R, Trovato A, Bianchini V, Biancardi A, Cichero P, care. Mazzotti M et al (2015) Clostridium difficile PCR ribotype 018, a successful epidemic genotype. J Clin Microbiol 53:2575–2580. doi:10.1128/JCM.00533-15 References 14. Debast SB, Bauer MP, Kuijper EJ; European Society of Clinical Microbiology and Infectious Diseases (2014) European Society of Clinical Microbiology and Infectious Diseases: update of the treat- 1. Wiegand PN, Nathwani D, Wilcox MH, Stephens J, Shelbaya A, ment guidance document for Clostridium difficile infection. Clin Haider S (2012) Clinical and economic burden of Clostridium Microbiol Infect 20(Suppl 2):1–26. doi:10.1111/1469-0691.12418 difficile infection in Europe: a systematic review of healthcare- 15. Eyre DW, Cule ML, Wilson DJ, Griffiths D, Vaughan A, O’Connor – facility-acquired infection. J Hosp Infect 81:1 14. doi:10.1016/j. L et al (2013) Diverse sources of C. difficile infection identified on jhin.2012.02.004 whole-genome sequencing. N Engl J Med 369:1195–1205. doi:10. 2. Loo VG, Bourgault A-M, Poirier L, Lamothe F, Michaud S, 1056/NEJMoa1216064 Turgeon N et al (2011) Host and pathogen factors for Clostridium 16. Walker AS, Eyre DW, Wyllie DH, Dingle KE, Harding RM, – difficile infection and colonization. N Engl J Med 365:1693 1703. O’Connor L et al (2012) Characterisation of Clostridium difficile doi:10.1056/NEJMoa1012413 hospital ward-based transmission using extensive epidemiological 3. Freeman J, Bauer MP, Baines SD, Corver J, Fawley WN, Goorhuis data and molecular typing. PLoS Med 9:e1001172. doi: 10.1371/ B et al (2010) The changing epidemiology of Clostridium difficile journal.pmed.1001172. http://www.ncbi.nlm.nih.gov/pmc/articles/ infections. Clin Microbiol Rev 23:529–549. doi:10.1128/CMR. PMC3274560 00082-09 17. Janezic S, Zidaric V, Pardon B, Indra A, Kokotovic B, Blanco JL 4. Lagier J-C, Dubourg G, Cassir N, Fournier P-E, Colson P, Richet H et al (2014) International Clostridium difficile animal strain collec- et al (2013) Clostridium difficile 027 emerging outbreak in tion and large diversity of animal associated strains. BMC Marseille, France. Infect Control Hosp Epidemiol 34:1339–1341. Microbiol 14:173. doi:10.1186/1471-2180-14-173 doi:10.1086/673995 18. Knetsch CW, Connor TR, Mutreja A, van Dorp SM, Sanders IM, 5. Cassir N, Delarozière J-C, Dubourg G, Delord M, Lagier J-C, Browne HP et al (2014) Whole genome sequencing reveals poten- Brouqui P et al (2016) A regional outbreak of Clostridium difficile tial spread of Clostridium difficile between humans and farm ani- PCR-ribotype 027 infections in southeastern France from a single mals in the Netherlands, 2002 to 2011. Euro Surveill 19:20954 long-term care facility. Infect Control Hosp Epidemiol 37:1337– 19. Barbut F, Mastrantonio P, Delmée M, Brazier J, Kuijper E, Poxton I 1341. doi:10.1017/ice.2016.164 et al (2007) Prospective study of Clostridium difficile infections in 6. Lagier J-C, Delord M, Million M, Parola P, Stein A, Brouqui P et al Europe with phenotypic and genotypic characterisation of the iso- (2015) Dramatic reduction in Clostridium difficile ribotype 027- lates. Clin Microbiol Infect 13:1048–1057. doi:10.1111/j.1469- associated mortality with early fecal transplantation by the 0691.2007.01824.x

299

REFERENCES

1. Torres-Fuentes C, Schellekens H, Dinan TG, Cryan JF. The microbiota-gut-brain axis in

obesity. Lancet Gastroenterol Hepatol. 2017;2(10):747‑56.

2. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, Ley RE, et al. A

core gut microbiome in obese and lean twins. Nature. 22 janv 2009;457(7228):480‑4.

3. Lopez-Siles M, Martinez-Medina M, Abellà C, Busquets D, Sabat-Mir M, Duncan SH,

et al. Mucosa-associated Faecalibacterium prausnitzii phylotype richness is reduced in

patients with inflammatory bowel disease. Appl Environ Microbiol. nov

2015;81(21):7582‑92.

4. Hocquart M, Lagier J-C, Cassir N, Saidani N, Eldin C, Kerbaj J, et al. Early Fecal

Microbiota Transplantation Improves Survival in Severe Clostridium difficile Infections.

Clin Infect Dis Off Publ Infect Dis Soc Am. 10 févr 2018;66(5):645‑50.

5. Hugon P, Lagier J-C, Colson P, Bittar F, Raoult D. Repertoire of human gut microbes.

Microb Pathog. 16 juin 2016;

6. Angelakis E, Bachar D, Henrissat B, Armougom F, Audoly G, Lagier J-C, et al. Glycans

affect DNA extraction and induce substantial differences in gut metagenomic studies.

Sci Rep. 18 mai 2016;6:26276.

7. Seng P, Rolain J-M, Fournier PE, La Scola B, Drancourt M, Raoult D. MALDI-TOF-

mass spectrometry applications in clinical microbiology. Future Microbiol. nov

2010;5(11):1733‑54.

300 8. Lagier J-C, Hugon P, Khelaifia S, Fournier P-E, La Scola B, Raoult D. The rebirth of

culture in microbiology through the example of culturomics to study human gut

microbiota. Clin Microbiol Rev. janv 2015;28(1):237‑64.

9. Lagier J-C, Khelaifia S, Alou MT, Ndongo S, Dione N, Hugon P, et al. Culture of

previously uncultured members of the human gut microbiota by culturomics. Nat

Microbiol. 7 nov 2016;1:16203.

10. Fujimoto T, Imaeda H, Takahashi K, Kasumi E, Bamba S, Fujiyama Y, et al. Decreased

abundance of Faecalibacterium prausnitzii in the gut microbiota of Crohn’s disease. J

Gastroenterol Hepatol. avr 2013;28(4):613‑9.

11. Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermúdez-Humarán LG, Gratadoux J-J, et

al. Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified

by gut microbiota analysis of Crohn disease patients. Proc Natl Acad Sci U S A. 28 oct

2008;105(43):16731‑6.

12. Shen W, Shen M, Zhao X, Zhu H, Yang Y, Lu S, et al. Anti-obesity Effect of Capsaicin

in Mice Fed with High-Fat Diet Is Associated with an Increase in Population of the Gut

Bacterium Akkermansia muciniphila. Front Microbiol. 2017;8:272.

13. World Health Organization. Antimicrobial resistance, Factsheet 194 [Internet]. 2016

[cité 30 oct 2018]. Disponible sur: http://www.who.int/mediacentre/factsheets/fs194/en/

14. Centers for Disease Control and Prevention. Antibiotic / Antimicrobial Resistance.

Biggest Threats. [Internet]. [cité 30 oct 2018]. Disponible sur:

https://www.cdc.gov/drugresistance/biggest_threats.html

301 15. European Centre for Disease Prevention and Control. Antimicrobial resistance

surveillance in Europe. Annual report of the European Antimicrobial Resistance

Surveillance Network (EARS-Net). 2015. [Internet]. 2017 [cité 30 oct 2018]. Disponible

sur:

https://ecdc.europa.eu/sites/portal/files/media/en/publications/Publications/antimicrobial

-resistance-europe-2015.pdf

16. World Health Organization. Antibacterial agents in clinical development [Internet]. 2017

[cité 15 nov 2018]. Disponible sur:

http://apps.who.int/iris/bitstream/handle/10665/258965/WHO-EMP-IAU-2017.11-

eng.pdf;jsessionid=1D402A86A40175305D37CE8454A1C0EB?sequence=1

17. D’Costa VM, McGrann KM, Hughes DW, Wright GD. Sampling the antibiotic

resistome. Science. 20 janv 2006;311(5759):374‑7.

18. Donia MS, Cimermancic P, Schulze CJ, Wieland Brown LC, Martin J, Mitreva M, et al.

A systematic analysis of biosynthetic gene clusters in the human microbiome reveals a

common family of antibiotics. Cell. 11 sept 2014;158(6):1402‑14.

19. Angelakis E, Merhej V, Raoult D. Related actions of probiotics and antibiotics on gut

microbiota and weight modification. Lancet Infect Dis. oct 2013;13(10):889‑99.

20. Cimermancic P, Medema MH, Claesen J, Kurita K, Wieland Brown LC, Mavrommatis

K, et al. Insights into secondary metabolism from a global analysis of prokaryotic

biosynthetic gene clusters. Cell. 17 juill 2014;158(2):412‑21.

21. Dione N, Khelaifia S, La Scola B, Lagier JC, Raoult D. A quasi-universal medium to

break the aerobic/anaerobic bacterial culture dichotomy in clinical microbiology. Clin

Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis. janv 2016;22(1):53‑8.

302 22. Veve MP, Wagner JL. Lefamulin: Review of a Promising Novel Pleuromutilin

Antibiotic. Pharmacotherapy. sept 2018;38(9):935‑46.

23. Tracanna V, de Jong A, Medema MH, Kuipers OP. Mining prokaryotes for

antimicrobial compounds: from diversity to function. FEMS Microbiol Rev. 1 mai

2017;41(3):417‑29.

24. Wang G, Li X, Wang Z. APD3: the antimicrobial peptide database as a tool for research

and education. Nucleic Acids Res. 4 janv 2016;44(D1):D1087-1093.

25. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, et al. A new

antibiotic kills pathogens without detectable resistance. Nature. 22 janv

2015;517(7535):455‑9.

26. Krismer B, Liebeke M, Janek D, Nega M, Rautenberg M, Hornig G, et al. Nutrient

limitation governs Staphylococcus aureus metabolism and niche adaptation in the human

nose. PLoS Pathog. janv 2014;10(1):e1003862.

27. Zipperer A, Konnerth MC, Laux C, Berscheid A, Janek D, Weidenmaier C, et al. Human

commensals producing a novel antibiotic impair pathogen colonization. Nature. 28 juill

2016;535(7613):511‑6.

28. Ratzke C, Denk J, Gore J. Ecological suicide in microbes. Nat Ecol Evol. mai

2018;2(5):867‑72.

29. Kim M, Oh H-S, Park S-C, Chun J. Towards a taxonomic coherence between average

nucleotide identity and 16S rRNA gene sequence similarity for species demarcation of

prokaryotes. Int J Syst Evol Microbiol. févr 2014;64(Pt 2):346‑51.

303 30. Meier-Kolthoff JP, Göker M, Spröer C, Klenk H-P. When should a DDH experiment be

mandatory in microbial taxonomy? Arch Microbiol. juin 2013;195(6):413‑8.

31. Fournier P-E, Lagier J-C, Dubourg G, Raoult D. From culturomics to

taxonomogenomics: A need to change the taxonomy of prokaryotes in clinical

microbiology. Anaerobe. déc 2015;36:73‑8.

304