WO2013045863A1

WO2013045863A1 - Detection of e.coli strains of serotype 0104

Info

Publication number: WO2013045863A1
Application number: PCT/FR2012/052212
Authority: WO
Inventors: Laurent Debarbieux; Damien MAURA
Original assignee: Institut Pasteur
Priority date: 2011-09-29
Filing date: 2012-09-28
Publication date: 2013-04-04

Abstract

The subject matter of the invention is means for the detection of E. coli serotype 0104 bacteria, in particular means for the specific detection of the bacteria of this serotype. The means defined in the context of the invention can be used in the context of protocols or processes for diagnosing an infection by E. coli serotype 0104 bacteria, and in particular constitute reagents for the in vitro detection, in a biological sample, of the presence of these bacteria or of an episode of contamination by said bacteria, in particular in a mammal and especially in human beings. These means comprise gp17.1 polypeptides encoded by ORF 17.1 of a genome of a bacteriophage of the family Podoviridae.

Description

DETECTION OF E.COLI STRAINS OF SEROTYPE O104

The subject of the invention is means for the detection of E. coli bacteria of serotype 0104, in particular means allowing the specific detection of bacteria of this serotype.

The means defined in the context of the invention may be used in the context of protocols or methods for diagnosing infection with E. coli bacteria of serotype 0104, and in particular constitute reagents for in vitro detection in a biological sample of the presence of these bacteria or an episode of contamination by said bacteria, in particular in a mammal and in particular in humans.

These means comprise nucleic acid molecules, polypeptides encoded by these nucleic acids or antibodies against said polypeptides and also bacteriophages comprising these nucleic acids and expressing said polypeptides.

The subject of the invention is also a method for the in vitro detection of the presence in a biological sample of bacteria of serotype E.coli 0104, in which the above-mentioned means are used.

Escherichia coli (E. coli) is a proteobacterium belonging to the family Enterobacteriaceae. It has a bacillary form, it is facultative aero-anaerobic, negative oxidase, catalase and nitrate reductase positive, and generally mobile. The species has a large variability of surface antigenic structures (lipopolysaccharide somatic antigens O, capsular antigens K and flagellar antigens H), which allowed the establishment of a serological typing system (Orskov et al., 1992). Thus, more than 170 types of O antigens have been described in E. coli strains, potentially representing as many different receptors for bacteriophages (coliphages) interacting specifically with lipopolysaccharide (LPS) (Germani et al., 2008). .

Most E. coli strains are not pathogenic to humans. However, some strains may prove to be opportunistic, even responsible for extra-intestinal pathologies (ExPEC for Extra-intestinal Pathogenic E.coli) or intestinal pathologies (InPEC for intestinal Pathogenic E.coli) (Russo et al., 2000; 2004). In the context of the invention, the inventors have been interested in E.coli strains of EAEC (EnteroAgrégative E. coli) pathotype constituting a group of emerging pathogens capable of causing intestinal pathologies, including diarrhea in humans such as mentioned below.

The serotype O104: H4 strains recently implicated in diarrhea outbreaks in Germany, Denmark and France in May and June 201 1 are a perfect example of E.coli EAEC strains that have become highly virulent. These strains have virulence factors associated with EAEC strains (plasmid pAA, AggR regulator, AAF adhesion fimbriae, etc.) and are very close to the EAEC 55989 strain (more than 93% genome in common). However, they acquired via a prophage the Stx2 toxin characteristic of STEC pathotype strains. However, they do not have the LOCE (Locus for Enterocyte Erasure) locus, which is the second characteristic of the STEC strains (Denamur 201 1, Rasko et al., 201 1).

In the context of the invention, the inventors have used the strain EAEC 55989.

It is highly virulent in a model of sepsis in mice, and is able to colonize the intestine of mice at a high rate (about 10 ⁹ cfu / g of faeces) over several weeks (Martinez-Jéhanne et al., 2009; et al 2009).

More specifically, the inventors have been interested in the interactions of bacterial strains of serotype O104: H4, with the bacteriophages capable of infecting them, in order to determine their effects in the pathologies caused by these bacteria. E.coli bacteria, and in particular enteroaggregative bacteria (EAEC) can be infected by different viruses or bacteriophages that use them to replicate such as by diverting the cellular machinery of bacteria during a cycle to their advantage. lytic.

During the infection cycle, lytic cycle bacteriophages are able to destroy the bacterium by lysis of the infected cell. Lysogenic cycle bacteriophages behave like lytic cycle bacteriophages, but from time to time their genome is inserted into the bacterial chromosome as a prophage.

A bacteriophage isolated in the context of the invention is a caudal bacteriophage (of the order of the Caudovirales) belonging to the Podoviridae family.

Podoviridae are subdivided into 4 genera according to morphological and genomic criteria, all named after the phage type of genus. each genus: "T7-like", "P22-like", "phi29-like" and "N4-like" (Fauquet et al., 2005).

In their study of bacteriophages capable of infecting enteroaggregative E.coli bacteria of the 0104 group, in order to determine their interactions in a complex polymicrobial environment such as that of the intestine, the inventors have demonstrated the existence of a bacteriophage capable of specifically infecting serotype O104 strains, because of the presence in its genome, of specific elements encoding proteins capable of interacting specifically with these bacteria, said elements also being contained within identified prophages in the genome of bacteria.

On the basis of their observations, the inventors have defined means that can be used in particular for the detection of E. coli bacteria of serotype 0104.

Given the sometimes fatal consequences in humans, observed following infection with serotype 0104 bacteria, there is a clear need for means to detect them, in particular means specific to their specific detection, including means for discriminating against other bacteria of different serotypes.

The subject of the present application is therefore a nucleic acid molecule comprising a nucleotide sequence coding for a polypeptide domain for recognition of a bacterial receptor present specifically on E. coli bacteria of serotype 0104, or constituted by a fragment of at least 9 nucleotides of a nucleotide sequence encoding such a domain, said nucleotide sequence belonging to the ORF 17.1 said ORF 17.1 being specific to the genome of a bacteriophage of the family Podoviridae.

According to the above definition, when present in a bacteriophage genome containing it from the family Podoviridae, the nucleotide sequence of ORF 17.1 follows the sequence of the ORF 17 encoding the tail fiber protein.

The inventors have observed that the ORF 17.1 contained in the bacteriophage genome is also present specifically in E.coli bacteria of serotype 0104. The expression "specific" as regards the presence of a nucleotide sequence or means that said sequence is not found identically in other bacteriophage genomes that do not infect E. coli serotype 0104 or in strains of E. coli of another serotype. By "nucleic acid molecule" is meant a polynucleotide, regardless of its size, formed of a sequence of nucleotides, especially within a DNA, double strand or single strand. In general, a nucleic acid molecule according to the invention may be a genomic DNA, a complementary DNA (cDNA) obtained by reverse transcription, or may be an RNA, in particular a messenger RNA (mRNA). A nucleic acid molecule according to the invention may be prepared by any means known to those skilled in the art, in particular by extraction from an organism containing it and purification, by synthesis, in particular by chemical synthesis, by cloning or by amplification from a matrix.

A nucleic acid molecule is said to be "coding" or "coding for" in the context of the invention, to express the fact that it can be translated in the form of a chain of amino acids, in particular when it can be translated in the form of a polypeptide naturally expressed by a bacteriophage-type cell or virus that contains it, or a polypeptide constituting a fragment of such a polynucleotide. Thus, according to a particular embodiment of the invention, the nucleic acid molecule corresponds to an open reading frame (ORF) of a bacteriophage genome or to a fragment of an ORF. In a particular embodiment of the invention, it is a gene or the coding sequence of a gene.

The size of a nucleic acid molecule of the invention varies from a few nucleotides, in particular at least 9 nucleotides to several thousand nucleotides, in particular to coincide with the sequence of IORF17.1. The nucleic acid molecule of the invention has, for example, 9 (or 15, 21, 33, 42, 60, 102, 193, 300, 402 or 501) nucleotides at 1300 nucleotides, for example 9 (or 15, 21, 33, 42, 60, 102, 193, 300, 402 or 501) nucleotides at 1000, or 9 (or 15, 21, 33, 42, 60, 102, 193, 300, 402 or 501) nucleotides at 800 nucleotides or 9 to 30 nucleotides. In a particular embodiment of the invention, it is a fragment of IORF17.1.

The nucleic acid molecules according to the invention are useful for the preparation of the polypeptides of the invention, including peptides or epitopes.

The bacteriophage genome according to the invention has been mapped with reference to strains designated as reference bacteriophage in a given family or subgroup. The bacteriophage T7 and bacteriophage K1 F are for example reference strains. In the context of the invention, the organization of the bacteriophage T7 genome is taken as a reference for positioning the ORFs of the genome of other bacteriophages. Thus, the perfectly determined ORF 17 in the bacteriophage T7 genome can, by alignment of the genomes, locate the ORF 17 in another genome. ORF 17 is part of the genome region with the late-expressed bacteriophage structural genes and is known to express the gp17 protein.

A bacteriophage according to the invention is a bacteriophage of which the ORF 17 is followed by a ORF called ORF17.1 whose sequence has similarities with those of the ORF 17, but which differs in particular by its N-part. terminal, that is to say by the sequence included in the first 510 nucleotides of the 5 'region. The gp17.1 protein does not possess the domain of the fiber proteins called the head-tail binding domain. In contrast, gp17.1 possesses the C-terminal domain of fiber proteins called receptor recognition domain on the surface of bacteria.

The term "specific" for the ability of bacteriophage to infect E. coli strains of serotype 0104 means that the bacteriophages considered do not or do not significantly infect E. coli bacteria. coli other serotypes. This ability to infect can be determined by searching for the host spectrum of a particular bacteriophage, as illustrated in the examples.

By way of example, the bacteriophage CLB_P1 according to the invention has been identified as having a host spectrum enabling it to infect E. coli bacteria of serotype O104, in particular by examining its ability to infect strains of the serotype O104 collection. E. coli ECOR bacteria.

As an indication and in accordance with the foregoing definitions, the following examples show the location of the ORF 17.1 in the bacteriophage CLB_P1 compared to the bacteriophage T 7 genome.

By analogy with the ORF 17 sequence of bacteriophages of the Podoviridae family, the nucleotide sequence of IORF17.1 encodes a polypeptide corresponding to the C-terminal region capable of specifically interacting with receptors present on E. coli bacteria. coli serotype O104 but essentially absent from other E. coli bacteria. According to the inventors, this interaction seems to go hand in hand with the ability of a bacteriophage bearing ORF 17.1 to specifically infect E. coli bacteria of serotype 0104. It can be determined by a study of the host spectrum of the bacteriophage considered as described in the following examples.

In one embodiment of the invention, the nucleic acid molecule comprises a nucleotide sequence belonging to the ORF gp17.1 bacteriophage CLB_P1 deposited according to the rules of the Budapest Treaty to the CNCM Collection (National Collection of Culture of Microorganisms - Paris, France) under No I-4534 on 28 September 201 1.

The identification and characterization of bacteriophage CLB_P1 are described in the examples of the present application. In the bacteriophage CLB_P1 genome, ORF 17.1 is located between nucleotides 34344 and 35636 and encodes a polypeptide designated gp17.1 of 431 amino acids.

The sequence of the nucleic acid molecule (hereinafter also referred to as polynucleotide) of the bacteriophage CLB_P1 ORF 17.1 is described in FIG. 14. It is also represented by SEQ ID No. 1.

In another embodiment of the invention, the nucleic acid molecule is characterized by encoding a polypeptide having the amino acid sequence of the gp17.1 binding domain to the E. coli receptor. serotype O104, represented in FIG. 14 and also by the sequence SEQ ID No. 2.

The inventors have shown that such a polynucleotide corresponding to the ORF 17.1 or a part of this ORF coding for the gp17.1 binding domain to the E. coli serotype 0104 receptor is present specifically in bacteriophages that specifically infect bacteria. E. coli serotype 0104 and that it is also present in said bacteria in particular in the form of a prophage. In bacteriophage CLB_P1, the sequence of ORF 17.1 is present in a form in which, relative to gp17 or prophage sequences, the N-terminal region encoding the 170 amino acid domain involved in attachment to bacteriophage is absent. In contrast, in the prophage contained in E. coli bacteria of serotype 0104, a sequence having an identity with the sequence of ORF 17.1 is present under the name YP_002403617.1 (SEQ ID No. 3), the latter sequence comprising in addition to the sequence similar to IORF17.1, in the N-terminal position, the bacteriophage binding domain. In the context of the invention, a "bacteriophage" encompasses the various forms likely to be isolated from the virus, including virions and prophages.

In the context of the invention, and taking into account the observations above, supplemented by the findings made on bacteriophage CLB_P1, a polynucleotide provides a means for defining detection reagents specific to E. coli bacteria of serotype 0104. detection is said to be "specific" insofar as the polynucleotide of the invention does not lead to the detection of other E. coli bacteria belonging to different serotypes. In particular, these reagents can be used to detect the presence of, or infection with, E. coli bacteria of serotype O104 in a biological sample taken from humans.

In a particular embodiment of the invention, the nucleic acid molecule is characterized in that it responds to one of the following sequences:

(i) a nucleotide sequence encoding a polypeptide having the amino acid sequence SEQ ID NO: 2 of the gp17.1 binding domain to the E. coli serotype 0104 receptor;

(ii) the nucleotide sequence shown in Figure 14 or also by SEQ ID No. 1;

(iii) a sequence comprising said sequence (ii) or a sequence included in sequence (ii) and constituting a fragment of IORF17.1 of at least 9 nucleotides;

(iv) a variant sequence of the sequence (ii) or of the sequence (iii) having an identity of at least 75%, in particular at least 78%, for example at least 80%, in particular at least 85%, by at least 90% or at least 95% with the sequence (ii) said identity being measured along the length of the shortest sequence of the so-called variant sequences on the one hand and (ii) or (iii) on the other hand, respectively ;

(v) A sequence according to one of sequences (ii), (iii) or (iv) encoding a gp17.1 polypeptide.

Thus, according to the above definitions, the invention relates to variants of the polynucleotide 17.1, which are in particular variants resulting from a modification of the size of the polynucleotide compared with that of the ORF 17.1, in particular variants having a smaller size (so-called fragments) and / or variants substitution, deletion or addition of nucleotides, having an identity with the sequence of ORF 17.1 defined by the above percentages.

The identity between two polynucleotide sequences is determined by performing an optimal alignment of these sequences, relative to the nucleotides of the shortest sequence, followed by the determination of the number of positions showing an identical nucleotide in both sequences ( corresponding to identical positions or "matched"), dividing the number of identical positions by the total number of positions considered and multiplying this result by 100 to arrive at the percentage of identity. This identity determination can be done using comparison algorithms such as the Needieman-Wunsch Global Alignment Algorithm or the Smith-Waterman Alignment Algorithm. The same procedures are applicable for the comparison of two amino acid sequences and the determination of their percentage of identity.

In this context, a variant polynucleotide may furthermore or alternatively be defined by its ability to hybridize with the sequence of ORF 17.1, in particular with the sequence of this ORF coding for the binding domain of gp17.1 to the E receptor. .coli serotype O104. The hybridization conditions required to identify a polynucleotide of the invention are so-called stringent conditions. By way of illustration, stringent conditions are described below. They can naturally be adapted by those skilled in the art, for example depending on the context of the reaction carried out.

In a particular embodiment of the invention, the nucleic acid molecule is characterized in that it is the bacteriophage CLB_P1 ORF 17.1 gene.

In another particular embodiment of the invention, the nucleic acid molecule is characterized in that it is a fragment comprising at least 9 consecutive nucleotides of the sequence of the nucleic acid molecule according to the definitions (i), (ii), (iii), (iv) given in the present application, in particular a fragment having from 9 to 27 consecutive nucleotides of said molecule.

In a particular embodiment of the invention, a polynucleotide has a nucleotide sequence encoding a polypeptide having an amino acid sequence belonging to a polypeptide gp17.1, selected from the following sequences, shown in FIG. 13: the amino acid sequence from residue 6 to residue 430 of gp17.1 of CLB_P1;

the amino acid sequence from residue 234 to residue 659 of the EGR74047.1 protein (prophage of E. coli strain LB226692).

the amino acid sequence from residue 234 to residue 659 of EGT66234.1 protein (prophage of E. coli strain C227-1 1).

the amino acid sequence from residue 234 to residue 659 of protein EGR63661 .1 (prophage strain E.coli 01 -09591)

the amino acid sequence from residue 234 to residue 659 of protein YP_002403617.1 (prophage strain E. coli 55989).

In general, a gp17.1 polypeptide is a polypeptide expressed by a bacteriophage according to the invention, which is in particular recognized by antibodies prepared against the bacteriophage CLB_P1 gp17.1 polypeptide sequence or a serotype E.coli strain. O104 comprising a prophage as defined in the present application, or against a truncated polypeptide expressed by such a prophage and comprising amino acids 171 to 430 of the amino acid sequence of the prophages polypeptides of FIG. 13, or against a polypeptide corresponding to another strain of E. coli.

The present application also relates to a polypeptide characterized in that it is encoded by a nucleic acid molecule according to the invention.

A particular polypeptide is characterized in that it has an amino acid sequence selected from:

(i) the amino acid sequence shown in Figure 14 or SEQ ID No. 2;

(ii) an amino acid sequence comprising the sequence

(i) an amino acid sequence constituting a fragment, in particular a fragment of at least five amino acids of the sequence of the gp17.1 protein of (i), in particular a fragment whose amino acid sequence from residue 6 to residue 430 of gp17.1 of CLB_P1 (residues 6 to 430 in the sequence SEQ ID No. 2);

(iii) an amino acid sequence variant of the sequence (i) or of the sequence (ii) having an identity of at least 75%, in particular at least 78%, for example at least 80%, especially at least 85%, for example at least 90% or at least 95% with the sequence (i) said identity being measured along the length of the shortest sequence of so-called variant sequences on the one hand and (i) or (ii) on the other hand.

Particular polypeptide fragments of the invention are, for example, fragments having from 3 to 9 amino acids or fragments having 3 (or 5, 7, 1, 14, 20, 34, 64, 100, 134, 167) acids. amino acids with 433 amino acids or 333 or 266 amino acids.

The subject of the invention is in particular a polypeptide characterized in that it comprises an epitope recognized by antibodies induced against gp17.1 or against the region of recognition of bacterial receptors interacting with the bacteriophage CLB_P1 gp17.1 fiber protein, characterized in that it is deposited in the CNCM Collection under No. I-4534 and designated CLB-P1.

In a particular embodiment of the invention, a polypeptide has an amino acid sequence belonging to a polypeptide gp17.1, chosen from the following sequences, represented in FIG. 13:

the amino acid sequence from residue 6 to residue 430 of gp17.1 of CLB_P1;

The peptides or polypeptides according to the invention may be natural peptides or polypeptides, or fragments of a natural protein or polypeptide, or they may be recombinant peptides or polypeptides, or they may be synthetic peptides or polypeptides.

Any synthetic technique known to those skilled in the art can be used. Examples of synthetic techniques, such as Merrifield-type solid phase synthesis, can be found, for example, in the book "Solid Phase Peptide Synthesis" (JM Steward & Young JD, 1969, Ed. WH Freeman Co., San Francisco), or "Peptide synthesis" (M. Bodansky et al., 1976, John Wiley & Sons, 2 ^nd Edition).

The invention also relates to a virulent bacteriophage of the family Podoviridae capable of specifically infecting a bacterial strain of E. coli serotype 0104, characterized by the presence in its genome of an open reading frame (ORF) designated gp17.1 comprising a gene encoding a gp17.1 fiber protein.

In a particular embodiment of the invention, the bacteriophage is characterized in that it is the bacteriophage deposited in the CNCM Collection under NO 1-4534 on September 28, 201 1.

In another embodiment of the invention, the bacteriophage is a bacteriophage deleted for the ORF 17 or the gene 17 or not including the entire sequence of the gp17 gene, ie, a bacteriophage that does not express the protein functional gp17. Such a bacteriophage can be prepared from bacteriophage CLB_P1 deposited at the CNCM under No. I-4534.

The invention thus relates to a mutated bacteriophage in which the gp17 gene is no longer expressed or a bacteriophage that does not include the gp17 gene sequence. Alternatively, a mutated bacteriophage is characterized in that it comprises a fusion polynucleotide between the sequence coding for the N-terminal domain of gp17 and the sequence encoding the receptor recognition domain on the surface of the bacteria, of the gp17.1 protein. to allow the binding of the expressed recombinant protein on the tail of the bacteriophage.

The bacteriophage can be in the form of a prophage.

These genes can be introduced into the bacteriophage of the invention by any technique known per se, for example by homologous recombination, by transposon mutagenesis.

Alternatively, a bacteriophage of the invention may be labeled, for example by a fluorescent molecule, including GFP, incorporated into the capsid protein.

The invention also relates to the genome of a bacteriophage as defined in the application, in particular to the bacteriophage genome CLB_P1 or to the bacteriophage CLB_P1 genome mutated for IORF17 or the gp17 gene.

The invention also relates to antibodies induced against a polypeptide according to any one of the definitions given above. Antibodies according to the invention may be polyclonal antibodies or monoclonal antibodies.

Antibodies may, for example, be produced by immunization of a non-human mammal (such as a rabbit) with a polypeptide according to the invention, or with an antigenic fragment of such a polypeptide, optionally associated with or coupled with an adjuvant. immunization (such as Freund's adjuvant or KLH -keyhole limpet hemocyanin), for example by intraperitoneal or subcutaneous injection, and by collecting antibodies thus obtained in the serum of said mammal.

Monoclonal antibodies can be produced according to a lymphocyte hybridization technique (hybridomas) such as the Kohler and Milstein 1975 (see also US 4,376,10), the human B-cell hybridoma technique (Kosbor et al. 1983, Cole et al., 1983), or the technique of immortalizing lymphocytes using the Epstein-Barr-EBV-virus (Cole et al., 1985). Such antibodies may for example be IgG, IgM, IgE, IgA, IgD or any subclass of these immunoglobulins.

The invention also relates to a method for the in vitro detection of an infection with an E. coli bacterium serotype 0104 comprising culturing a biological sample under conditions of growth and, where appropriate, counting of E. coli bacteria. , with a bacteriophage according to the invention and the determination of the lysis of E. coli bacteria.

The antibodies used in the invention may be antibody fragments or artificial derivatives of such fragments, insofar as these fragments or derivatives have said specific binding property with gp17.1 or with a fragment of this protein in particular. with the recognition domain of the receptor of an E. coli bacterium serotype O104. Such fragments may for example be Fab, F (ab ') 2, Fv, Fab / c, scFv (single chain Fragment variable) fragments.

Also within the scope of the invention, a kit for the in vitro detection of an infection with a bacterium E.coli serotype 0104 comprising a bacteriophage according to one of the definitions given herein or a polypeptide according to the invention, or antibodies according to the invention.

The subject of the invention is also a method for the in vitro detection of an infection with an E. coli bacterium of serotype 0104 comprising bringing into contact with a biological sample, in particular a biological fluid sample, a reagent comprising a bacteriophage according to the invention or a polypeptide according to the invention, or antibodies of the invention and the detection of a reaction between the reagent and the sample.

Many methods can be used to achieve detection of target bacteria. Thus, the invention relates to an in vitro detection method characterized in that it comprises an ELISA immunoenzymatic reaction carried out with antibodies according to the invention.

For the detection of bacteria, the reagents (bacteriophages, antibodies, polypeptide, etc.) can be fixed on a solid support. Said reagents may be attached to a solid support, for example a support made of polymer, plastic, in particular polystyrene, glass or silicon.

Said reagents may be attached directly or indirectly to said solid support, for example via a binding or capture agent which is attached to the solid support. This binding or capture agent may comprise a portion attached to said solid support and a portion that comprises a ligand that specifically binds to one of said selected reagents. Such a ligand may for example be an antibody, a monoclonal antibody, or a fragment of such an antibody having retained the binding specificity.

Said solid support may for example be a plastic plate, in particular polystyrene, comprising several analysis wells, such as a titration plate or protein microtitre, for example, an ELISA plate.

Said solid support can also be magnetic microbeads or not, for micro-titrations, for example according to the technique described by Luminex.

Said solid support may for example be a nucleic acid, protein or peptide chip, for example a chip made of plastic, glass or silicon.

The invention also relates to the use of the nucleic acid molecules described, where appropriate labeled, for the detection of E. coli bacteria serotype 0104.

Detection can be done using the nucleic acid molecules as probes or as primers. In particular, the detection can be carried out by PCR or by other techniques based on DNA amplification. In a particular embodiment of the invention, the detection using nucleic acids of the invention comprises a step of fixing these nucleic acids on a support, for example on a nucleic acid chip.

The invention also relates to a therapeutic composition for the treatment of an infection with E. coli 0104: H4 bacteria, comprising as active ingredient at least one bacteriophage CLB_P1, CLB_P2 or CLB_P3, and a pharmaceutically acceptable carrier.

The bacteriophages CLB_P2 and CLB_P3 were deposited at the CNCM (Paris, France) respectively under numbers I-4675 and I-4676, on September 27, 2012.

LEGEND OF FIGURES

FIGURE 1: Host and EFP spectrum of twelve bacteriophages active on strain 55989

The efficiency to form lysis ranges (EFP) of the twelve bacteriophages was determined from the phage titer using the double layer technique with 7 different strains of E.coli: JM221, 042, E2348 / 69, AL51 1, 536 and MG1655, as well as with the indicator strain 55989Str. In the name of the phages, the letters represent the source of the water used (C = Clichy, Co = Corbeil, E = Evry, L = Lagny and S = Sèvres). The bacteriophages selected in each of groups A, B and C are respectively C1A1A renamed CLB_P1, E2A1A renamed CLB_P2 and S1A1A renamed CLB_P3. +++: EFP> 80%; ++: 80%> VET> 0.1%; +: 0.1%> EFP> 0; /: EFP = 0

FIGURE 2: Host spectrum of the three bacteriophages selected from the ECOR collection

+++: EFP>80%; ++: 80%>VET>0.1%; +: 0.1%>EFP>0; -: EFP = 0. The strains not indicated in the table (59 in total) are not sensitive to any of these 3 bacteriophages. FIGURE 3: Photographs of CLB_P1, CLB_P2 and CLB_P3 virions in electron microscopy.

The scale bar represents 100 nm

FIGURE 4: Analysis of the majority proteins of CLB_P1, CLB_P2 and

CLB_P3.

(A) SDS-PAGE gel separation of the proteins of the three virions. The proteins analyzed in mass spectrometry are indicated with a black star. (B) Protein sequences of K1 F (gp10A = SEQ ID No. 1), JS98 (gp23 = SEQ ID No. 12) and T1 (T1 p47 = SEQ ID No. 13) closest to proteins sequenced. The amino acids represented in red are those corresponding to the peptides obtained by mass spectrometry.

FIGURE 5: identification and functional annotation of the ORFs of CLB_P1.

FIGURE 6: Genetic map of bacteriophage CLB_P1

FIGURE 7: Alignment between the CLB P1 gp17.1 protein (SEQ ID No. 2) and the YP_002403617 protein (SEQ ID No. 3) of a prophage located in the chromosome of E. coli strain 55989.

Alignment made with ClustalW 2 via Geneious software. The Head Binding Domain is indicated with a gray arrow.

FIGURE 8: Alignment between the CLB_P1 gp17 protein (SEQ ID No. 7) and bacteriophage of the T7-like genus EcoDSI (SEQ ID No. 9), K1F (SEQ ID No. 8) and T7 (SEQ ID No. ID No. 10).

Alignment made with ClustalW 2 via Geneious software. FIGURE 9: Genomic comparison of CLB P1 with K1 F and T7

Comparison of the ORFs of the 3 bacteriophages by BLASTP. For the ORFs present in the 3 genomes (indicated as white arrows), the percentage of identity in protein sequences is indicated according to a color code of purple variations. The ORFs indicated in red are specific to CLB_P1, those in yellow are specific to K1 F and those in green are specific to T7. ORFs in orange are common to CLB_P1 and K1 F but absent in T7, the ORFs in blue are common to CLB_P1 and T7 but absent in K1 F.

FIGURE 10: Kinetics of lysis of the bacteriophage-infected strain 5589Str.

55989Str uninfected (black circles), 55989Str infected with CLB_P1 (white squares), 55989Str infected with CLB_P2 (white triangles), 55989Str infected with CLB_P3 (white circles), 55989Str infected with the cocktail of 3 bacteriophages (white diamonds).

FIGURE 11: Infection of biofilms of strain 55989Str by the 3 bacteriophages.

Twenty-four hour old 55989Str Biofilms exposed to LB (white) or at different concentrations (3 to 10 ³ , 5 to 10 ⁵ , and 7 to 10 ⁷ pfu / ml) of bacteriophage CLB_P1 (\), CLB_P2 (gray), CLB_P3 (/) or cocktail (black). The total biofilm is quantified by crystal violet staining (A), viability by XTT staining (B) and biofilm matrix by DMMB staining (C). The values are expressed as the ratio of the OD of the sample divided by the OD of the control. The experiments were performed in three replicates, n = 15 per condition. ^* : p <0.05, ^*** : p <0.001.

FIGURE 12: The bacteriophage cocktail reduces aggregates formed on the surface of human epithelial cells.

HeLa cells (A, B, C, D, E, F) and T84 (G, H, I, J, K, L) were incubated with the bacterium 55989Str for 2 h before adding either MEM medium without bacteriophage (A, B, C, G, H, I) or MEM medium with the bacteriophage cocktail (2 x 10 ⁶ pfu / ml; D, E, F, J, K, L). After 2 hours of incubation, the samples were observed by scanning electron microscopy. The scale bars represent 10 μηη (A, D, G, J) and 1 μηη (B, C, E, F, H, I, K, L). The arrows indicate aggregates of 55989Str (white) or bacteriophages (black). FIG. 13: Alignment of the CLB_P1 gp17.1 protein (SEQ ID No.

2) and its homologues in the genomes of strains of serotype 0104: 1-14: the strains LB226692 prophage (SEQ ID No. 5), C227-11 prophage (SEQ ID No. 6), 01 -09591 are shown; prophage (SEQ ID No. 4) and 55989 prophage (SEQ ID No. 3)

The domain common to these 5 proteins is framed.

FIGURE 14: Nucleic acid sequence (SEQ ID No. 1) and amino acid sequence (SEQ ID No. 2) of bacteriophage CLB_P1 gp17.1 protein

FIGURE 15: The bacteriophage cocktail reduces the concentration of bacteria 55989 in the mouse ileum

The concentrations of bacteria (A) and bacteriophage (B) were determined on days 4 and 7 for 3 groups of mice colonized with E. coli O104: H4 55989 on day 0, and having received on day 3 a drinking water without bacteriophages (white circles), or drinking water containing a cocktail of bacteriophages at 3 10 ⁸ pfu / ml (gray circles) or 3 10 ¹⁰ pfu / ml (black circles) for 24h only. The experiments were performed in triplicate, n = 15 to 29 per condition. ^** = p <0.01; ^*** = p <0.001. The E. coli 55989 concentrations in the ileum were determined by qPCR (C). Equivalent cfu ratios of E. coli normalized to the total number of bacteria are given. The experiments were performed in duplicate with n = 6 to 8 for each condition.

FIGURE 16: The bacteriophage cocktail has a limited impact on strain 55989 in the colon and in the excretions of colonized mice

3 groups of mice colonized with E. coli O104: H4 55989 at day 0, and having received at day 3 a drinking water without bacteriophages (white circles), or drinking water containing a cocktail of bacteriophages at 3 10 ⁸ pfu / ml (gray circles) or 3 10 ¹⁰ pfu / ml (black circles) for 24h only. Bacteria concentrations (A), and bacteriophage (B) were determined at day 4 and day 7. The experiments were performed in duplicate or triplicate, n = 10-15 by condition ^** = p <0.01; ^*** = p <0.001. FIGURE 17: E. coli 55989 infection with bacteriophages is not limited by intestinal microbiota

2 groups of axenic mice (n = 4) colonized with E. coli 0104: H4 55989 at day 0, and having received at day 3 a drinking water without bacteriophage (white), or drinking water containing a cocktail of bacteriophage at 3 10 ⁸ pfu / ml (gray) for 24h only. Concentrations of bacteria (A) and bacteriophage (B) in the ileum (circles) and droppings (triangles) were determined on days 4 and 7.

FIGURE 18: The location of 55989 bacteria in the digestive system affects the permissiveness of infection by bacteriophages.

A: E.coli 55989 cells in exponential phase (white) and in stationary phase (black) growth were mixed with bacteriophages CLB_P1, CLB_P2 or CLB_P3 individually (MOI 2 10 ² ) and incubated for 5 hours before quantification. The results are expressed as multiplication n times compared to the initial number of bacteriophages added. B: LB medium containing 55989 exponentially growing bacteria (Ct) as well as portions of ileum (i) and excreta (f) of the intestine collected from mice colonized with strain 55989 for 3 days were mixed with each bacteriophage individually at a 1 10 ^-2 MOI and incubated for 5 hours The results are expressed as multiplication n times compared to the initial number of bacteriophages added.

EXAMPLES

In order to study the interactions between bacteriophages and a pathogenic bacterium in a complex polymicrobial environment such as the intestine, the inventors chose a murine model in which a bacterial strain was able to colonize this organ for several weeks without any apparent signs. infection. The bacterium chosen is part of the E. coli species, widely studied both as a model bacterium and as an intestinal pathogen. More particularly, strain 55989 belongs to the group of enteroaggregative E. coli (EAEC), an emergent pathotype for which experimental data in relation to bacteriophages are desirable.

MATERIAL AND METHODS 1. Bacterial strains and growing conditions

E. coli strain 55989 was originally isolated from feces of an adult patient with persistent diarrhea (Mossoro et al., 2002). A spontaneous variant of streptomycin-resistant strain 55989 has been isolated (55989Str) and has been shown to be able to stably colonize the mouse intestine (Martinez-Jéhanne et al., 2009). The other E. coli strains used in this study are listed in the Table below. The ECOR collection, which contains 72 commensal and pathogenic E. coli strains isolated from human or animal hosts on different continents and representative of the genetic diversity of the E. coli species, was provided by the Biological Resources Center of the Pasteur Institute (Ochman et al., 1984).

The strains were routinely cultured with liquid Luria-Bertani (LB) medium, agar or on plates of Drigalski medium, at 37 ° C. The anaerobiosis cultures were carried out in Anoxomat anaerobic chamber (Mart Microbiology BV, The Netherlands), under an atmosphere composed of 90% N ₂ , 5% H ₂ and 5% CO 2. Streptomycin (100 mg / l) was added when necessary.

Table: List of strains of bacteria and bacteriophages used in this study

2. Isolation, preparation and characterization of bacteriophages

Bacteriophages specific for strain 55989Str were isolated from wastewater by an enrichment technique, and preparations at high viral concentration were obtained as described in (Morello et al., 201 1). Initially, 12 different bacteriophages were isolated according to the morphology of their lysis range. The three bacteriophages that were selected for this study and are listed in the Table above.

The host spectrum of each bacteriophage was determined according to the following technique. 1 ml of an exponential growth phase culture was distributed on LB agar medium. The excess liquid was removed and the box allowed to dry at room temperature for 30 to 60 minutes. 4 μl of each serial dilution (from 10 to 10) of the bacteriophage suspension in PBS were deposited on the surface of each dish. These dishes were incubated at 37 ° C overnight. Lysis rate (EFP) efficiency was determined by calculating the ratio of the bacteriophage solution titer obtained on the test strain divided by the titer obtained on the 55989Str reference strain.

The samples for the analysis of the morphology of the virions in electron microscopy were prepared as described in (Debarbieux et al., 2010).

The virion proteins were obtained by boiling a suspension of ¹¹ pfu / ml of each bacteriophage at 100 ° C for 10 minutes. 20 μl of this suspension was put on a 4-12% precursed polyacrylamide gel (Invitrogen). The gel was then stained with Coomassie blue and the majority bands excised, subjected to trypsin digestion and analyzed by mass spectrometry. The resulting masses were then compared to the protein databases for identification.

3. Sequencing, annotation and comparison of viral genomes

Sequencing of the bacteriophage genome CLB_P1 (CNCM I-4534), CLB_P2 (CNCM I-4675) and CLB_P3 (CNCM I-4676) was performed by Eurofins and Beckman Coulter Genomics using 454 technology (20 - 25x coverage) with extracted DNA according to the procedure described in (Sambrook et al., 2001).

The annotation of the genome of CLB_P1 was carried out. The identification of the putative ORFs of the genome of CLB_P1 was performed through the RAST software. The results were then confirmed by BLASTP using the bank of non-redundant NCBI data. Each start and stop codon was manually checked and the position of each putative ORF was compared with that of the bacteriophage genomes K1 F, T7 and EcoDSI to validate the annotations through the Geneious software and alignment tool. of Mauve genomes. The presence of promoters was detected using the PHIRE1 .0 software and the search for tRNA was performed with the tRNA-Scan SE tool.

The comparison of the putative ORFs of CLB_P1 with those of K1 F and T7 was performed with the BLASTP tool of the CLC Genomics program.

4. Kinetics of in vitro lysis

Growth phase exponential culture of strain 55989Str was diluted in LB to a DOeoonm of 0.1 and then 50 μl of this suspension were distributed in each well of a 96-well plate (Microtest 96 plates, vial). 50 μΙ LB medium or bacteriophage suspension (diluted in LB to obtain a multiplicity of infection of 10 ^"3) were added to each well. The plate was then incubated in a microplate reader at 37 ° C and orbital shaking at medium control (Glomax MultiDetection System, Promega, USA) OD at 600nm was recorded at 15-minute intervals over a 12-hour period, after 12 hours an aliquot of each sample was taken and centrifuged at 8000 g for 10 minutes The pellets were washed twice with 1X PBS and plated on LB plate to isolate colonies The sensitivity / resistance was determined for 30 colonies in each bacteriophage-bacteria pair.

5. Formation and quantification of biofilms

The formation and quantification of biofilms was performed with a modified version of the protocol described by Tré-Hardy et al. (Tré-Hardy et al., 2008). A culture of 55989Str incubated overnight at 37 ° C without agitation was diluted 1: 100 ^e in LB medium and then distributed in a 96-well plate (100 μΙ per well, Nunc MicroWell Plates, Nunc, Wiesbaden, Germany). The plates were then covered with a lid containing pins on which a biofilm can be formed (Nunc TSP Screening System). The microplate was then incubated in a static wet chamber at 37 ° C for 24 hours. The lid was removed and introduced into a new 96-well plate containing 100 μl of PBS per well for 10 seconds to remove non-adherent plankton cells. The lid containing the Biofilms was then introduced into a new 96-well plate containing 100 μl of LB medium alone or containing bacteriophage and incubated as previously described. After 24 hours, the lids were rinsed twice as described above. The biofilms attached to the pins were stained by introducing the lid into a 96-well plate containing 100 μl of crystal violet per well (0.5% w / v) for 20 minutes. The excess crystal violet was removed by rinsing the pins twice in a 96-well plate containing 100 μl of distilled water per well. The attached dye was then solubilized by incubating the nibs in a 96-well plate containing 200 μl of ethanol: acetone in 4: 1 ratio for 15 minutes. 50 μl of the resulting solution was transferred to a new 96-well plate to measure OD at 560nm (Glomax MultiDetection System, Promega, USA). Quantification of the biofilm matrix was performed with DMMB (diMethyl-Methylene Blue) as described by Toté et al. (Toté et al., 2008). Bacterial viability was quantified with XTT (2,3-bis (2-methoxy-4-nitro-5-sulfo-phenyl) -2H-tetrazolium-5-carboxanilide) as described by Ramage et al. (Ramage et al., 2001). Each experiment was performed 3 times independently, with n = 15 samples for each condition.

6. Aggregate formation on epithelial cells

The HeLa and T84 human cells were maintained as described in (Bernier et al., 2002, Sampaio et al., 2009). HeLa cells were incubated in humid chambers containing 95% air and 5% CO ₂ in MEM (Gibco) medium supplemented with heat inactivated fetal calf serum (FCS). The T84 cells were maintained in a humid chamber containing 95% air and 5% CO 2 in a medium consisting of a 1: 1 mixture of DMEM-F12 and GlutaMAX ™ (Gibco) supplemented with 10% FCS. The cells were incubated with glass slides at the bottom of the culture dishes 24 hours before the addition of bacteria. E.coli 55989Str incubated statically overnight at 37 ° C in LB medium was added to subconfluent HeLa cells or differentiated T84 single-layered 15-day-old cells at a final concentration of 10 ⁶ cfu / ml. ml. Cells and bacteria were incubated together for two hours to allow aggregation formation, washed three times with 1X PBS to remove unattached bacteria, and incubated for one hour with fresh cell culture medium. Three hours after the addition of the bacteria, the cells were washed in PBS and incubated with medium of fresh cell culture or medium supplemented with the cocktail of 3 bacteriophages at a final concentration of 10 ⁸ pfu / ml. Two hours after the addition of bacteriophages, the cells were washed in PBS and fixed on the glass slides by incubation with 2.5% buffered glutaraldehyde. Post-fixation in 1% osmium tetroxide was carried out, followed by dehydration with a series of ethanol solutions (25% to 100%) and then drying by the critical point technique using the Leica EM CDP030 device. The glass slides were then covered with palladium gold with a GATAN Ion Beam Coater before examination with a JEOL JSM-6700F scanning microscope used at 5 kV. The images were acquired with the SE detector.

7. Murine model of intestinal colonization

Mice (7 week old BALB / c YJ females) acquired from Charles River Laboratories were kept in acclimation for at least one week before the start of the experiments. Before the start of the experiments, the drinking water of the mice was replaced for 48 hours by a solution of streptomycin sulfate at a concentration of 5 g / l (Sigma, St. Louis, MO). This treatment aims to reduce the number of facultative aero / anaerobic resident bacteria that prevent the implantation of E. coli in mice (Croswell et al., 2009). Between the end of the streptomycin treatment and the administration of the bacteria, no colony morphologically close to E. coli was detected in the feces of the mice. After 2 days of streptomycin treatment and overnight fasting, the mice were force-fed with strain 55989Str contained in 200 μl of a sterile solution of 20% sucrose-2.6% bicarbonate (w / v) at pH 8. The animals then recovered their usual food and drinking water containing 5 g / 1 of streptomycin. On the following days, freshly produced feces were collected, homogenized, diluted in 1 X PBS and 4 μl of 10-fold serial dilutions were spotted on LB medium dishes to quantify E.coli or LB dishes containing layer of strain 55989Str to quantify bacteriophages. These dishes were then incubated overnight at 37 ° C, and the colonies were then counted. The detection limit of this technique is 3.10 ³ cfu / g or pfu / g of feces.

The count of bacteria and bacteriophages from the intestine samples was performed after euthanasia of the mice by CO2 asphyxiation. aseptic dissection to collect the duodenum-jejunum, ileum, cecum and colon. Each was then placed in 5 ml of 1X PBS at 4 ° C. After weighing the samples, each was ground with an Ultra-Thurax T25 homogenizer (LaboModerne, France), then a serial dilution of 10 in 10 was carried out and 4 μΙ of each dilution were spotted on LB medium containing or not a layer of strain 55989Str to quantify strain 55989Str or bacteriophages.

The mice received bacteriophages alone or in a cocktail at a concentration of 3.10 ⁸ pfu / ml (1 .10 ⁸ pfu / ml of each bacteriophage) or 3.10 ¹⁰ pfu / ml in the drinking water supplemented with 5 g / l of streptomycin for 24 hours. The mice then received a drinking water without bacteriophages composed of only 5g / l of streptomycin.

Experiments conducted with C3H axenic mice (provided by the central animal facility of the Institut Pasteur) were carried out following the same protocol as that with conventional flora mice, but in the absence of treatment with streptomycin.

8. Quantification of bacteria by quantitative PCR

Bacterial DNA was extracted with QIAGEN Stool DNA Minikit Kit 50 (QIAGEN, USA), with the option of incubating at 95 ° C from crushed feces or intestine samples.

The qPCR reactions were performed in a Biorad iCycler iQ device

(Biorad) with the MyiQ software.

The amplification and detection were carried out in 96-well plates with the SYBR Green Supermix iQ mixture (Biorad). Each reaction was performed in duplicate in a final volume of 25 μί with a final concentration of 0.25 μΜ for each primer and 5 μί of extracted DNA. The amplifications were carried out using the following program: 1 cycle at 95 ° C for 10 minutes, followed by 45 cycles of 95 ° C 15 seconds and 60 ° C 1 minute, then 1 cycle of 72 ° C for 1, 5 minutes and ending 80 incubation cycles at 55 ° C with an increment of 0.5 ° C each cycle to determine the denaturation curves.

Each qPCR reaction was performed with E.coli 16S rDNA specific primers (F = CAT GCC GCG TGT ATG AAA AA, R = CGG GTA ACG TCA ATG AGC AAA, (Furet et al 2009)) and in parallel with specific primers of the universal bacterial 16S rDNA to quantify total bacteria (F = CGG TGA ATA CGT TCC CGG, R = TAC GGC TAC CTT GTT ACG ACT T, (Furet et al 2009)). The data were expressed as the ratio between the amount of E. coli divided by the amount of total bacteria, in order to normalize each sample by the amount of total bacteria as described in (Furet et al., 2009). 9. Determination of the sensitivity / resistance of cells to a bacteriophage

The sensitivity / resistance of cells to a bacteriophage was determined by streaking an isolated colony on a box of LB agar medium, and depositing a drop of each bacteriophage solution to be tested on the bacterial streak. After incubation at 37 ° C for 16 hours, a resistant colony gives a streak that grows at the site of drop deposition, a sensitive colony will present a growth inhibition zone at the location of the bacteriophage drop.

10. Histological analyzes

Mouse ileum samples were fixed in 4% formaldehyde for 24 hours at 4 ° C and then paraffin embedded. Serial sections of 4 μιτι thick were cut and stained with hematoxylin-eosin (HE) to assess the integrity of the ileal epithelium. A marking E. coli was performed with a primary antibody rabbit anti-LPS of E. coli (Biodesign, USA) diluted ^1/1000 before use, then a revelation with a secondary antibody of goat anti-rabbit coupled to peroxidase. 1 1. Inhibition of 55989Str infection by bacteriophages

The stability of bacteriophage and strain 55989Str was studied in the presence of different concentrations of EDTA (0 to 100mM), MgCl ₂ (0 to 2.5M) or ascorbic acid mixture (17 to 1700 mg / L) - copper chloride (0.17 to 17 mg / L) in PBS with a contact time of 10 minutes at 37 ° C. In order to evaluate their stability, serial dilutions of 10 to 10 were made, then spotted (4 L) on LB medium in order to enumerate viable bacteria and infectious viral particles. In parallel, the inhibition efficiency of these molecules on the infection of 55989Str by bacteriophages was measured by listing the bacteria present in a mixture bacteriophages - bacteria capable of growing on LB medium.

In order to eliminate extracellular bacteriophages, a mixture of the strain

55989Str and bacteriophage cocktail was centrifuged at 8000g for 10 minutes and then washed with a volume of 1 X PBS, this cycle having been repeated 3 times in order to eliminate as many bacteriophages as possible.

In order to compete for the bacteriophage-cell receptor interaction, a stationary phase culture of strain 55989WT (streptomycin sensitive) was incubated for 40 minutes in LB medium with a concentration of 100 mg / l streptomycin, then contacted for 20 minutes with a bacteriophage - 55989Str mixture before the mixture is diluted and then spot to perform a viable bacteria count as previously described. 12. Bacteriophage replication test in intestinal morsels

Ileum, cecum and colon samples were taken from mice colonized or not for 3 days with strain 55989Str and then ground in 1 X PBS as previously described (paragraph No. 7). In non-colonized samples, strain 55989Str was added from an exponential growth phase culture (OD = 0.3) in biological samples at a final concentration of 10 ⁶ cfu / ml. Each bacteriophage was then added to samples colonized with strain 55989Str and in samples to which strain 55989Str was exogenously added as well as in non-colonized samples with strain 55989Str at a final concentration of 10 ⁶ pfu / ml. . After 5 hours of incubation at 37 ° C, the amount of bacteriophage and strain 55989Str were measured in each sample and compared to the sample at t = 0.

13. Ethical considerations

The mice used during this work were kept in a pet shop according to the European recommendations and those of the Institut Pasteur. Food and drink were provided ad libitum. The protocols used have been approved by the veterinary service of the Institut Pasteur pet shop (authorization n ° 10.565).

14. Statistical analyzes

During this work, ANOVA and Mann-Whitney one-way statistical tests with Graphpad software (Graphpad software Inc., USA) were used. RESULTS

1 Isolation and characterization of bacteriophages

1 .1 Isolation and determination of the host spectrum

The specific bacteriophages of the strain EAEC 55989Str identified in the context of the invention were isolated from wastewater samples from 5 purification plants in the Paris region. After 4 successive steps of purification of the lysis plaques on agar medium, 12 bacteriophages supposed to be different in respect of their origin and the morphology of their lysis ranges were selected.

These bacteriophages were then differentiated according to their host spectrum using 7 different E. coli strains: 3 EAEC strains (55989Str, JM221 and 042), one EPEC strain (E2348 / 69), 2 UPEC strains ( Uropathogenic E. coli) (AL51 1 and 536) and 1 commensal strain (MG1655). The results obtained with this host spectrum made it possible to classify these 12 bacteriophages into 3 distinct groups (A, B and C) according to their efficiency in forming lysis plaques (EFP) on these 7 strains (FIG. 1). A bacteriophage for each of the 3 groups was randomly selected for further characterization. They have been renamed respectively CLB_P1, CLB_P2 and CLB_P3.

The host spectrum of these 3 bacteriophages was determined using the ECOR collection representative of the genetic diversity of the E. coli species (Ochman et al., 1984). CLB_P1 appears to be the most specific bacteriophage of strain 55989Str since it is highly effective on only two strains (ECOR 27 and 28, with an EFP greater than 80%). CLB_P2 is the one that has the widest host spectrum with activity on 1 1 strains, but with very low efficiency (ECOR 16, 19, 25, 26, 27, 28, 29, 39, 56, 61 and 70, with an EFP greater than 0.1%). CLB_P3 is the most effective bacteriophage because it is able to infect 6 strains (ECOR 4, 13, 16, 19, 26 and 39) with an efficiency higher than 80% and 3 other strains (ECOR 25, 27 and 28) with an efficiency greater than 0.1% (Figure 2). 1 .2 Morphology of virions

Observation by electron microscopy of the virions made it possible to determine that the 3 selected bacteriophages belong to the order of the Caudoviruses. Each belongs to a different family: CLB_P1 has a short tail (I = 15 ± 2 nm, L = 25 ± 2 nm) and an icosahedral capsid (I = 60 ± 5 nm, L = 55 ± 4 nm) typical of the family Podoviridae; CLB_P2 has a contractile tail (I = 105 ± 5 nm, L = 20 ± 3 nm) and an elongated icosahedral capsid (I = 109 ± 3 nm, L = 80 ± 4 nm) typical of the family Myoviridae; CLB_P3 has a long, non-contractile flexible tail (I = 145 ± 8 nm, L = 12 ± 2 nm) and an icosahedral capsid (I = 65 ± 3nm, L = 66 ± 2nm) typical of the Siphoviridae family (Figure 20) .

1 .3 Structural proteins of virions

The total proteins of each of the three virions were separated by SDS-PAGE and their respective major proteins were analyzed by mass spectrometry. The peptides of the major protein of CLB_P1 are identical to peptides of the gp10A protein, the major bacteriophage K1 F capsid protein (NC_007456) belonging to the genus T7-like. The peptides of the majority proteins 1 and 2 of CLB_P2 are respectively identical to peptides of the proteins gp23 (major capsid protein) and gp18 (major protein of the caudal sheath) of bacteriophage JS98 (NC_010105) belonging to the genus T4-like. The peptides of the majority protein of CLB_P3 are identical to peptides of the T1 T1-like T1 bacteriophage T1 protein (NC_005833) (Figure 4) (Roberts et al., 2004, Scholl et al., 2005, Zuber et al. et al 2007).

1 .4 Sequencing of the genomes of the 3 bacteriophages

In order to obtain more precise information on these 3 bacteriophages and to know how close they are genetically bacteriophages K1 F, JS98 and T1, was sequenced.

1.4.1 Genomes of CLB_P2 and CLB_P3

The genomes of CLB_P2 and CLB_P3 are composed of linear double-stranded DNA of respectively 171, 787 kb and 50,444 kb, with percentages in G + C of 39% and 46%. The genome of CLB_P2 is 92% identical to the genome of JS98, while the genetically closest sequenced bacteriophage of CLB_P3 is T1 with which it has 84% of the identical genome. These results therefore confirm those obtained previously in mass spectrometry during the identification of the major capsid proteins. 1.4.2 Genome of CLB_P1

1.4.2.1 General characteristics of the genome

The genome of CLB_P1 is composed of linear double-stranded DNA of 40.218 kb. It has a percentage of guanine plus cytosine (% G + C) of 50.1%. The search for coding sequences made it possible to identify 59 putative ORFs, between 54 and 3888 nt in size. The initiation codon of 93% of the ORFs is ATG. Only 3 ORFs (gp4.3, gp7.7 and gp10.1) start with a GTG and 1 ORF (gp19) starts with a TTG. The genes are physically close to each other so as to occupy 91.8% of the total genomic sequence,

The genome of CLB_P1 is identical to 82% genome of K1 F and 80% with EcoDSI, against only 55% with T7. Given the strong genomic proximity of CLB_P1 with these bacteriophages of the genus T7-like, the genetic nomenclature of this genus has been adopted to annotate the genes of CLB_P1.

1.4.2.2 Functional Analysis of the CLB P1 Genome and Comparative Genomics

The detection of the ORFs contained in the genome of CLB_P1 was carried out using the RAST program, and then comparing the results obtained with the bacteriophage genomes K1 F and EcoDSI. Of the 59 ORFs of CLB_P1, 32 (54%) have a known function, 25 (43%) are similar to unknown function proteins present in non-redundant databases, and 2 (3%) encode putative proteins unique to this bacteriophage (Figure 5). While the vast majority of CLB_P1 proteins (80%) are homologous to T7-like K1 F or EcoDSI proteins, some possess sequence similarities to rv5 (gp0.15) and 13a (gp7.7) coliphage proteins. ), bacteriophages of Erwinia L1 (gp5.5) and Ea104 (gp5.8), bacteriophage Yersinia Ye03-12 (gp1 .45) and bacteriophage Klebsiella KP32 (gp1 .7). Apart from the bacteriophages rv5 and Ea104 which are Myoviridae (Niu et al 2009, Muller et al 201 1), the other viruses (13a, L1, KP32 and ye03-12) are all Podoviridae of the T7- genus. like (Pajunen et al 2001, Born et al 201 1). Early genes and hijacking of the host's machinery In CLB_P1, there are 13 putative ORFs among the class I genes. In particular, we find the essential genes 0.3, 1, 1.2 and 1.3 which respectively encode a protein that inhibits the type I restriction system, the viral RNA polymerase, an inhibitor of bacterial dGTPase and DNA ligase. The other putative ORFs (gp0.1, 0.15, 0.35, 0.351, 0.36, 0.37, 0.38, 1 .1 and 1 .15) are of unknown function. Of these, 5 have homologs in the K1 F genome (gp0.1, 0.35, 0.36, 0.37 and 0.38), the latter 2 (gp0.15 and 0.351) being specific for CLB_P1. In addition, CLB_P1 does not have a homologue of non-essential gp0.4, 0.5 and 0.6AB proteins found in the T7 genome. Interestingly, CLB_P1 and K1 F do not possess the T7 protein kinase (gp0.7) involved in the inhibition of host RNA polymerase (Figures 5, 6 and 9). DNA metabolism

The genome of CLB_P1 comprises 21 putative ORFs of class II. As in T7 and K1 F, we find the essential viral genes encoding proteins such as the gp2 transcription inhibitor, but also the gp3 and gp6 nucleases that degrade the genome of the host and provide nucleotides necessary for the replication of the viral genome (Figures 5, 6 and 9). The proteins involved in replication such as gp5 DNA polymerase, gp4 / 4B primase / helicase, gp2.5 SSB protein and viral RNA polymerase are also present in CLB_P1. In CLB_P1 and K1 F we find a putative nucleotide kinase (gp1 .7) which could also participate in the replication by regulating the concentration of the different nucleotides.

Lysozyme gp3.5 interacting with viral RNA polymerase to initiate transcription of late genes is also present. CLB_P1 also contains a homing endonuclease (gene 5.35) inserted at 174 nt from the end of gene 5, constituting a group I intron, as in some bacteriophages of the T7-like genus such as K1 F (Scholl et al., 2005). The other 9 putative ORFs (gp1.6, 3.2, 3.7, 4.2, 4.3, 7.7, 5.5, 5.7 and 5.8) encode hypothetical proteins of unknown function. Of these, 2 have homologs in the K1 F (gp1.6 and 3.7) genome and 1 (gp7.7) genome in the T7 genome. Furthermore, CLB_P1 does not have a homologue of the non-essential gp1 .4, 1 .5, 1 .6, 1 .8, 4.3, 4.5, 4.7, 5.3, 5.9, 6.3 proteins found in the T7 genome. Structure, assembly and maturation and release of virions The class III region encodes the proteins involved in the structure of virions, as well as their assembly, maturation and release in the extracellular medium. This highly conserved region between CLB_P1, K1 F and T7 encodes the assembly proteins of the tail (gp7.3, 1 1 and 12), capsid assembly (gp9 and 10) as well as the head-tail connector ( gp8) and the inner core proteins gp13, 14, 15 and 16. We also find the terminase (gp18 and 19) which is used to encapsidate the DNA and the holine (gp17.5) - endolysin pair (gp18.5 and 18.7). ) that degrade the bacterial wall and lyse the host cell at the end of the viral cycle. CLB_P1 also contains 6 putative ORFs (gp10.1, 1 1 .1, 17.1, 18.8, 19.2 and 19.4) (Figures 22, 23 and 26). Of these, the first 4 are specific to CLB_P1. Gp17.1 has 77% identity on about two-thirds of the sequence with a putative putative fiber protein found in the genome of E. coli 55989 (Figure 8).

Interestingly, the CLB_P1 gp17 protein encoding the fibers that serve to recognize the bacterial receptor has a very low homology with the gp17 proteins of K1 F and T7 (respectively 18% and 14% identical amino acids). The 170 amino acids located at the N-terminus of gp17 are conserved in many T7-like including K1 F, EcoDSI and T7 (T7 tail fiber domain, pfam03906) (Figure 7). In contrast to K1 F, CLB_P1 does not contain domains corresponding to an endosialidase within gp17, whose enzymatic activity degrades capsule polysaccharides (Scholl et al., 2005).

2 In vitro activity of bacteriophages

2.1 Activity of bacteriophages on E. coli 55989Str in planktonic culture

The realization of test spots on petri dishes incubated under aerobic atmosphere and anaerobiosis revealed no difference in terms of efficiency to form lysis ranges. The ability of bacteriophages to lyse 55989Str strain in the exponential growth phase liquid culture was then determined. CLB_P2 induces the fastest cell lysis, with a decrease in optical density at 600 nm (ϋθβοο) starting 75 minutes after viral infection, against 105 minutes with CLB_P1 and CLB_P3. Ϋθβοο increases after cell lysis, 3 h after the addition of CLB_P3 and 6 h after the addition of CLB_P1, suggesting the appearance of bacteriophage-resistant bacterial cells. This is confirmed by the analysis isolated colonies, which shows that 100% of the 30 colonies tested are resistant to the bacteriophage added to the culture medium during the kinetics. Interestingly, no resistant cell was detected after 12 h infection with CLB_P2 (Figure 10). 2.2 Bacteriophages reduce a biofilm formed by E.coli 55989Str in vitro

The capacity of the 3 bacteriophages to infect a biofilm formed by strain 55989Str was determined by adapting a biofilm formation model to 96-well plates (see material and methods and (Tré-Hardy et al., 2008)).

The amount of total biofilm was determined by crystal violet (CV) staining in the presence of bacteriophage concentrations ranging from 10 ^{3 to} 10 ⁷ pfu / ml (FIG. 11A). The presence of CLB_P1, even at the highest concentration tested, does not significantly affect the amount of total biofilm. CLB_P3 reduced CV staining in a dose-dependent manner by up to 40% for the dose of 10 ⁷ pfu / ml compared with uninfected control (p <0.001). CLB_P2 is the most effective bacteriophage because it reduces the amount of total biofilm in a dose-dependent manner up to 60% for the highest concentration (p <0.001).

Quantification of the viable cells (XTT staining) and the matrix (DMMB staining) composing the biofilm confirmed the results obtained for CLB_P2 and CLB_P3 (FIGS. 11 B and C). Each of the 3 bacteriophages induced a dose-dependent reduction in cell viability for concentrations ranging from 10 ³ to 10 ⁷ pfu / ml (p <0.001). The biofilm matrix is also significantly reduced in a dose-dependent manner by up to 40% following the action of bacteriophage CLB_P2 or CLB_P3 (p <0.001 for each), whereas CLB_P1 only reduces the amount of matrix by 20% to the highest concentration (p <0.05). Therefore, CLB_P2 is the most effective bacteriophage to reduce the total biofilm formed by strain 55989Str, as well as the number of viable cells and the biofilm matrix.

2.3 A cocktail is more effective than individual bacteriophages

The results obtained with individual bacteriophages having revealed significant differences according to the virus used, a combination of these three bacteriophages mixed in equivalent proportions to form a cocktail was evaluated. The kinetics of lysis of a plankton-infected 55989Str culture infected with this cocktail is similar to that obtained with CLB_P2 alone, without growth of resistant bacteria detected after 12 hours of incubation (Figure 10). In contrast, the use of this cocktail on a 55989Str biofilm in vitro reveals a cumulative effect, with a reduction in total biofilm, viable cells and biofilm matrix for all concentrations used except the lowest (10 ³ pfu ml) (Figures 11A, B and C). The highest concentration of this cocktail is the most effective, reducing the total biofilm amount by 75%, the viable cells by 95% and the biofilm matrix by 70%.

2.4 The bacteriophage cocktail reduces the formation of aggregates on the surface of epithelial cells

The inventors then investigated whether the bacteriophage cocktail was capable of infecting 55989Str aggregates on the surface of human cells. Scanning electron microscope observations shown in Figure 12 indicate that bacterial aggregates are largely present on the surface of HeLa cells in the absence of bacteriophages, while only isolated bacteria are observed in the condition with the bacteriophage cocktail. Moreover, bacteriophages are visible on the surface of the remaining bacterial cells. The same type of observation was obtained with a model of aggregation formation on the surface of T84 epithelial cells (Figure 12).

During the study of interactions between virulent bacteriophages and their bacterial host, a pathogenic human EAEC strain was studied in the mouse intestine. The manner in which bacteriophages are able to target their host in this complex polymicrobial environment has been characterized, and the factors of this environment that influence the course of infection by these viruses have been determined. The scope of this work has made it possible both to obtain information on the ecological aspect of the replication of these bacteriophages at the expense of their bacterial host in the intestinal environment, but also to better understand the phenomena involved at the time. the use of bacteriophages to target a pathogenic bacterial strain in the intestine. The specific bacteriophages of the isolated strain EAEC 55989Str were characterized.

3. In vivo activity of bacteriophages. Targeting of enteroaggregative E. coli bacteria was evaluated using a mouse intestinal model for colonization with O104: H4 55989 strain and infection with a cocktail of three virulent bacteriophages CLB_P1, CLB_P2 and CLB_P3. It has been shown beforehand that the colonization model mimics the asymptomatic intestinal presence of bacteria observed in humans. 3.1 The strain 55989 can colonize both the large intestine and the small intestine

To characterize the colonization of the mouse intestine by strain EAEC 55989 0104: H4, the bacteria were quantified at different sections of the digestive system, from the duodenum to the colon and in the excrements, 3 and 7 days later. force-feeding Cells of the strain 55989 have been found in the faeces and in the large intestine, with an average concentration August ^¹⁰ to September 10 cfu / g, and in the small intestine at an average concentration of 10 ³ to 10 ⁸ CFU / in the tissues of the duodenum-jejunum and the ileum with no major difference between days 3 and 7. The histological and immunohistochemical analyzes on the tissues of the ileum and colon showed the presence of bacteria not only in the lumen but also between the villi of the ileum and in the crypts, where they formed aggregates on the surface of the epithelial cells. Aggregates on the epithelial cells of the intestine in the colon have also been observed.

3.2 the bacteriophage cocktail reduces the concentration of strain 55989 in the ileum

The viral and bacterial concentrations in the intestinal ileum section, 4 and 7 days after colonization, were determined in two groups of mice colonized by strain 55989. One group received drinking water without bacteriophages, the other received a drinking water containing 3x10 ⁸ pfu / ml of a cocktail of bacteriophages for 24 hours (days 3 to 4), then a drinking water without bacteriophages. The bacteriophage cocktail led to concentrations of strain 55989 in the ileum corresponding to only 1/8 of that of the controls after 24h treatment (p <0.01) (Figure 15A). No cells of strain 55989 were detected in the ileum in 40% of these mice. However, by day 7, the concentrations in the ileum had reached the level of the control group concentrations. The bacteriophage concentration in the ileum did not change significantly between days 4 and 7, remaining relatively high (10 ⁸ pfu / g) (FIG. 15B). It was then determined whether an increase in the concentration of bacteriophage cocktail could increase the amplitude and duration of the decrease of the 55989 strain in the ileum. The use of a 100-fold higher concentration of the bacteriophage cocktail led to a decrease in the concentration of bacteria 55989 in the ileum 70 times greater than that obtained with a cocktail of lower concentration, as well as to bacterial levels 470 lower than those of the untreated control (p <0.001). In addition, for 14 of the 15 mice (93%), no cells of strain 55989 could be detected in the ileum sections after 24h treatment. However, by day 7, the concentration in the 55989 strain ileum had returned to untreated control animal levels, as previously observed with the lower concentration cocktail (Figure 15A). In both experiments, 100% of the 55989 cells isolated at day 7 from the ileum samples were susceptible to the phage cocktail. 3.3 Impact of the bacteriophage cocktail in the colon and the droppings.

Under the same experimental conditions as those described above, the impact of bacteriophages in the colon, on the flora of the light and the adherent flora, as well as on the droppings, was observed. The bacteria concentrations 55989 in the adherent flora and the flora of the light remained relatively stable regardless of the number of bacteriophages added to the drinking water (Figure 16A). the bacteria concentrations 55989 in the faeces of the mice receiving the highest dose cocktail accounted for 1/3 (p <0.001) of those of the untreated controls (Figure 16A). In comparison, the highest concentration of bacteriophages led to a decrease in the concentration of strain 55989 by a factor of 430 (p <0.001), suggesting that one or more factors (physical and / or physiological) could affect the infection by bacteriophages differentially at different locations in the intestine.

3.4 The infection of strain 55989 by the bacteriophages in the intestine is not limited by the microbiota of the mouse.

Using 2 different groups of gnobiotic mice colonized exclusively with the bacterium 55989, experiments similar to those described above were carried out, a group receiving the cocktail at a concentration of 3 × 10 ⁸ pfu / ml and the control group receiving the water of drink lacking bacteriophage. In ileum and excreta samples, no significant changes in bacteriophage concentration were observed between days 4 and 7 (Figure 17A). Thus, the gut microbiota did not have a major effect on the infection of strain 55989 with bacteriophages.

3.5 Permissivity of strain 55989 to infection by bacteriophages is not uniform along the digestive system

The role of the physiological state of strain 55989 on the limitation of infection by bacteriophages has been observed. First, the in vitro amplification of each bacteriophage on strain 55989 was examined in an exponential growth phase and in a stationary growth phase of the bacteria on LB medium. At an MOI of 0.2 (but also at an MOI of 100, results not shown), the amplification of CLB_P1 and CLB_P3 was weaker in stationary phase than in exponential phase, whereas CLB_P2 was much less affected (FIG. 18A).

It has also been shown that pH can affect the permissiveness of the host to bacteriophage infection since bacteriophages and bacteria do not survive at pH 2 and do not multiply at pH4, whereas at pH5 and 7 bacteria and bacteriophages multiplied, showing that the bacteriophage infection cycle and the physiology of the bacteria could be affected in vitro. .

The ex vivo bacteriophage amplification in an intestinal context by adding the bacteriophages separately to homogenates of ileum and excretions obtained from non-colonized mice or from mice colonized with the strain 55989 was evaluated (FIG. 18B). Strain 55989 was added to samples taken from non-colonized mice and mixed before adding each bacteriophage separately. Under these conditions, each bacteriophage showed amplification levels in ileum and excretion samples that were slightly higher or similar to those obtained in LB medium. The intestinal environment therefore had no effect on the infection. After addition of bacteriophages individually in samples colonized with strain 55989, all these bacteriophages were amplified as efficiently in the ileum while CLB_P1 and CLB_P3 were amplified less efficiently in the droppings than CLB_P2 (Figures 17B and Figure 18B). The strain had the same growth in ileum and excreta samples (Figure 18B). Replication of CLB_P1 and CLB_P3 along the in vivo digestive system is not uniform due to a change in permissiveness of strain 55989.

Discussion

Among the bacteriophages isolated, three (CLB_P1, CLB_P2 and CLB_P3) were the subject of a thorough study. The morphology of these bacteriophages determined by electron microscopy indicates that all three are caudate bacteriophages of the family Caudovirales.

Myoviridae CLB_P2 has a morphology close to that of T4 and a 172 kb genome identical to 92% with T4-like JS98 (Zuber et al., 2007). Siphoviridae CLB_P3 has a morphology close to T1 and a genome identical to 84% with it (Roberts et al., 2004). The genomic proximity of these two viruses to virulent model bacteriophages suggests that they are themselves virulent.

Podoviridae CLB_P1 has a morphology close to the bacteriophage model T7. The mass spectrometry analysis coupled with the complete sequencing of the genome of CLB_P1 confirms that this bacteriophage belongs to the genus T7-like belonging to the subfamily of Autographivirinae. CLB_P1 has more than 80% genomic identity with T7-like K1 F and EcoDSI, but only 55% identity with T7. However, the genomic organization of CLB_P1 is identical to that of these 3 bacteriophages, an organization conserved in all T7-like members known to allow temporal control of gene expression during the infection cycle (Molineux 2006). Thus, the genome of CLB_P1 is organized into genes expressed early (class I), intermediate (class II) and late (class III). All proteins essential for the multiplication of T7-like on their host under laboratory conditions are present in CLB_P1. However, CLB_P1 just like K1 F does not have the protein kinase gp0.7 present in T7. This protein plays a role in the inhibition of transcription via the host RNA polymerase as well as in the stabilization of viral early mRNA and the termination of transcription of early genes.

The major difference between CLB_P1 and the other bacteriophages of the genus T7-like concerns the region encoding the gp17 fiber protein. In fact, gp17 of CLB_P1 has at most only 30% homology with the gp17 proteins of other T7-like bacteriophages. Only one domain of 170 residues in N-terminal of gp17, corresponding to the head-binding domain (HDB), is common to CLB_P1 and other T7-like (Steven et al., 1988). Interestingly, the gp17 protein of CLB_P1 does not possess in its C-terminal part the endosialidase domain present in K1 F and whose role is to degrade the polysaccharides of the host capsule (Scholl et al., 2005). The inventors have demonstrated that CLB_P1 is capable of infecting viable cells within a 55989Str biofilm, but without significantly degrading the biofilm matrix. It is therefore possible that the absence of degradation of the matrix is related to the absence of the endosialidase domain of the fiber protein.

The analysis of the viral genome has furthermore demonstrated that CLB_P1 possesses, in a region encoding the virion structure genes, an ORF encoding a putative protein called gp17.1 which is absent from the bacteriophage genomes K1 F and T7. Interestingly, this gp17.1 protein has a strong sequence similarity to a putative fiber protein (YP_002403617.1) of one of the 6 prophages integrated into the E. coli 55989 genome (Touchon et al., 2009). These data suggest that gp17.1 could be a fiber protein involved in the recognition of E. coli 55989Str by CLB_P1. However, the putative prophage protein has an HBD domain that is absent from the gp17.1 sequence, suggesting that gp17.1 would not interact directly with the CLB_P1 capsid (Steinbacher et al., 1997). We can therefore hypothesize that gp17.1 could interact with gp17 and thus participate in the recognition of the CLB_P1 receptor on the surface of strain 55989Str.

The use of the ECOR collection to determine the bacteriophage CLB_P1 host spectrum strongly suggested a host range restricted to serotype O104 strains. Indeed, this bacteriophage was able to infect the three O104 strains of the ECOR collection (ECOR-26, ECOR-27 and ECOR-28). Strains of serotype O104 have recently been implicated in diarrheal epidemics affecting more than 4,000 patients in Germany, Denmark and France (Denamur 201 1, Rasko et al 201 1). Interestingly, the putative fiber proteins of CLB_P1 and the 55989 prophage have a common domain that exhibits strong sequence similarity with a protein domain of serotype O104: H4 strains (01 -09591, C227-1 1 and LB226692) recently sequenced (Figure 13). This suggests that this protein domain might be involved in the specific recognition of a common receptor for strains of serotype 0104. As stated above, the O antigens correspond to particular motifs of the polysaccharide chain of LPS. Knowing that T7-like bacteriophages are known to use LPS as a receptor, the link between the 0104 antigen and bacteriophage CLB_P1 will be examined, and the implication of the gp17.1 protein in this recognition evaluated. The invention therefore proposes the use of this protein domain in the detection of 0104 serotype strains (Hagens et al., 2007). Moreover, the coupling of this protein to bacteriolytic agents could allow the development of specific treatments of E. coli O104: H4 (Damasko et al., 2005) if the lysis of these bacteria does not cause a large diffusion of toxins of which the effects would be counterproductive.

The inventors have shown that CLB_P2 and CLB_P3 are capable of significantly reducing a biofilm formed by strain 55989Str in vitro, by acting on both the matrix and the number of viable cells. This reduction of the matrix suggests that these two bacteriophages could produce enzymes capable of hydrolyzing exopolysaccharides (EPS), proteins or extracellular DNA, the three major components of the extracellular matrix of a biofilm (Donlan 2009; Karatan et al 2009, Abedon 201 1). This hypothesis is supported by the observation of a halo surrounding the lysis plaques of CLB_P3, indicating the production of EPS depolymerases (data not shown) (Cornelissen et al 201 1). The analysis of the genomic sequences of CLB_P2 and CLB_P3 could allow the identification of this type of enzymes.

Interestingly, the inventors have observed that the cocktail of the 3 bacteriophages allows a greater reduction of the 55989Str biofilm than the individual bacteriophages, suggesting a cooperative action of the different bacteriophages on this microbial structure. One hypothesis is that the access of CLB_P1 to cells located inside the biofilm can be facilitated by the hydrolysis of the matrix by the enzymes produced by the bacteriophages CLB_P2 and / or CLB_P3. Thus, cells potentially resistant to a given bacteriophage would still be physically accessible to another of the bacteriophages of the cocktail. These results obtained on the 55989Str biofilm are particularly interesting in the case of the EAEC pathotype because the formation of biofilms on the surface of the intestinal epithelial cells is linked to the persistent colonization and the induction of diarrhea by these strains (Kaur et al. 2010). Having shown that the Bacteriophage cocktail is able to reduce the aggregates of 55989Str formed on the surface of T84 differentiated epithelial cells, it is reasonable to think that such a cocktail would be more effective in vivo than a bacteriophage alone.

Experiments show that strain 55989 does not colonize only the large and small intestines of the mouse over a period of at least 7 days. Based on the mouse intestinal model, the behavior of the EAEC strain on an indigenous intestinal microbiota was observed. In particular we have studied the impact of a cocktail of 3 virulent bacteriophages (CLB_P1, CLB_P2 and CLB_P3) infecting the strain 55989 in the mouse with the aim of reducing the intestinal content. Bacteriophages induce a significant and dose-dependent decrease in the concentration of strain 55989 in the ileum. A complete disappearance of the strain is however not observed if qPCR quantification is carried out.

The fact that the bacterium did not completely disappear after the addition of bacteriophages at day 4 allowed the bacterium to recolonize this section in the same way as for the untreated mice at day 7.

The results also showed that infection with the bacteriophage was less effective in the large intestine and in the droppings than in the ileum. The permissiveness of strain 55989 to CLB_P1 and CLB_P3 was related to the localization of the bacterium in the digestive system. Among the 3 bacteriophages studied, belonging to different families, CLB_P2 (family Myoviridae) showed the most favorable characteristics for a therapeutic application because it was effective in all sections of the intestine while CLB_P1 (family Podoviridae) had the least favorable characteristics. Nevertheless, the results obtained showed that a single administration of bacteriophages induced a decrease in the level of E. coli O104: H4 demonstrating the potential of these bacteriophages to target intestinal pathogens.

Specificity of bacteriophage CLB P1.

The inventors established the specificity of bacteriophage CLB_P1 by demonstrating that it only infects Escherichia coli strain O104: H4 in the presence of two other strains of Escherichia coli. The bacteriophage-sensitive strain CLB-P1 (55989Str) 0104: H4 was mixed with two other strains of Escherichia coli (AL46 and LF82SK). AL46 is a clinical strain from a patient with urinary tract infection and LF82SK is a strain isolated from a patient with Crohn's disease. Each of these three strains carries different resistance to antibiotics, which allows them to be selectively

The three strains were individually cultured in LB medium with stirring at 37 ° C. until reaching an exponential phase of growth (5 × 10 ⁸ cfu / ml), then they were diluted to obtain for each a concentration of 1 × 10 ⁶ cfu / ml. ml. Then they were mixed. This mixture was introduced into 2 tubes, one of which contained a quantity of 1 10 ⁵ pfu / ml bacteriophage CLB_P1 and the other served as control without bacteriophage. After incubation for 5 hours at 37 ° C., the bacterial cultures were serially diluted and an aliquot of each dilution was deposited on agar media containing different antibiotics.

The number of bacteria 55989Str was determined on a medium containing nalidixic acid.

The number of AL46 bacteria was determined on a medium containing spectinomycin.

The number of LF82SK bacteria was determined on a medium containing kanamycin.

Results:

Values are given in cfu / ml

55989Str AL46 LF82SK

CLB_P1 no yes no yes no yes

T = 0 5.0 10 ^b 5.0 10 ^b 1, 0 10 ^B 1, 0 10 ^B 1, 3 10 ^B 1, 3 10 ^B

T = 5h 7.5 10 ° 2.5 10 7.0 10 ° 7.5 10 ° 6.0 10 ° 6.5 10 ° In 5h, the amount of bacteriophage in the presence of the 3 strains increased from 1 10 ⁵ to 5 10 ⁸ pfu / ml.

Conclusion:

The ability of bacteriophage CLB_P1 to specifically lyse E. coli strain 55989Str in the presence of two other strains of Escherichia coli is demonstrated by the fact that when all three strains are in its presence, only strain 55989 is lysed while the other two strains are are not affected in their growth.

PUBLICATIONS

Abedon, S. (2011). "Facilitation of CRISPR adaptation." Bacteriophage 1 (3).

Abedon, S., Kuhl S., Blasdel B. and Kutter E. (2011). "Phage treatment of human infections."

Bacteriophage 1 (2): 66-85.

Abedon, S.T. (2011). Bacteriophages and Biofilms: Ecology, Phage Therapy, Plaques Nova

Biomedical.

Ackermann, H. W. (2009). "Phage classification and characterization." Methods Mol Biol

501: 127-140.

Archambaud, M., P. Courcoux and A. Labigne-Roussel (1988). "Detection by molecular hybridization of pap, afa, and sfa adhesion systems in Escherichia coli strains associated with urinary and enterai infections." Ann Inst Pasteur Microbiol 139 (5): 575-588.

Baldini, M. M., J. B. Kaper, M. Levine M., D.C. Candy and H. W. Moon (1983). "Plasmid-mediated adhesion in enteropathogenic Escherichia coli." J Pediatr Gastroenterol Nutr 2 (3): 534-538.

Bernier, C., P. Gounon and C. Le Bouguenec (2002). "Identification of an aggregative

Fibramidal adhesion (AAF) type III-encoding operon in enteroaggregative Escherichia coli as a sensitive probe for detecting the AAF-encoding operon family. Infect Immun 70 (8): 4302-4311.

Blattner, F. R., G. Plunkett, 3rd, Bloch, C., N., Perna, T., Burland, M., Riley, R .; . . Y. Shao (1997). "The complete genome sequence of Escherichia coli K-12." Science

277 (5331): 1453-1462.

Born, Y., Fieseler L., Marazzi J., Lurz R., Duffy B. and Loessner M. (2011). "Novel

Virulent and Broad Range Host Erwinia amylovora Bacteriophages Reveal a High Degree of Mosaicism and Relationship to Enterobacteriaceae Phages. "Appl Environ Microbiol.

Brzuszkiewicz, E., H. Bruggemann, H. Liesegang, M. Emmerth, T. Olschlager, G. Nagy,. . .

U. Dobrindt (2006). "How to become a uropathogen: comparative genomic analysis of extraintestinal pathogenic Escherichia coli strains." Proc Natl Acad Sci U S A

103 (34): 12879-12884.

Croswell, A., Amir, P., Teggatz, M., Barman, N., and Salzman, H. (2009). "Prolonged impact of antibiotics on intestinal microbial ecology and susceptibility to enteric Salmonella infection." Infection and Immunity 77 (7): 2741-2753.

Damasko, C., A. Konietzny, H. Kaspar, B. Appel, P. Dersch and E. Strauch (2005). "Studies of the efficacy of Enterocoliticin, a phage-tail like bacteriocin, an antimicrobial agent against Yersinia enterocolitica serotype 03 in a cell culture System and in mice." J Vet

Med B Disfect Dis Vet Public Health 52 (4): 171-179.

Debarbieux, L., D. Leduc, D. Maura, E. Morello, A. Criscuolo, O. Grossi,. . . L. Touqui

(2010). "Bacteriophages can treat and prevent Pseudomonas aeruginosa lung infections." J Infect Dis 201 (7): 1096-1104.

Denamur, E. (2011). "The 2011 Shiga toxin-producing Escherichia coli O104: German outbreak H4: a lesson in genomic plasticity." Clin Microbiol Infect 17 (8): 1124-1125. Donlan, R. M. (2009). "Preventing biofilm of clinically relevant organisms using

bacteriophage. "Trends in Microbiology 17 (2): 66-72.

Fauquet, C. M., M. A. Mayo, J. Maniloff, U. Desselberger and L. A. Bail (2005). Virus

Taxonomy: VlIIth Report of the International Committee on Taxonomy of Viruses,

Academy Press / Elsevier, London. Furet, J.-P., O. Firmesse, M. Gourmelon, C. Bridonneau, J. Tap, S. Mondot,. . . G. Corthier

(2009). "Comparative assessment of human and animal faecal microbiota using real-time quantitative PCR." FEMS Microbiolo v Ecologv 68 (3): 351-362.

Germani, Y., C. Bouguenec and P. J. Sansonetti (2008). "Escherichia coli in digestive pathology." Infectious diseases.

Hagens, S. and M. J. Loessner (2007). "Application of bacteriophages for detection and

control of foodborne pathogens. "Appl Microbiol Biotechnol 76 (3): 513-519.

Kaper, J.B., J.P. Nataro and H. L. Mobley (2004). "Pathogenic Escherichia coli." Nat Rev

Microbiol 2 (2): 123-140.

Karatan, E. and P. Watnick (2009). "Signals, regulatory networks, and materials that build and break bacterial biofilms." Microbiol Mol Biol Rev 73 (2): 310-347.

Kaur, P., A. Chakraborti and A. Asea (2010). "Enteroaggregative Escherichia coli: An

Emerging Enteric Food Borne Pathogen. "Interdiscip Perspect Infect Dis 2010:

254159.

Kindberg, C., W. W. Suttie, K. Uchida, K. Hirauchi and H. Nakao (1987). "Menaquinone production and utilization in germ-free rats after inoculation with specified organisms." J Nutr 117 (6): 1032-1035.

Lavigne, R., D. Seto, P. Mahadevan, H. W. Ackermann and A. M. Kropinski (2008).

"Unifying classical and molecular taxonomic classification: analysis of the

Podoviridae using BLASTP-based tools. "Res Microbiol 159 (5): 406-414.

Martinez-Jehanne, V., L. du Merle, C. Bernier-Febreau, C. Usein, A. Gassama-Sow, A.-A.

Wane,. . . C. The Bouguénec (2009). "Role of deoxyribose catabolism in colonization of the murine intestine by pathogenic Escherichia coli strains." Infection and

Immunity 77 (4): 1442-1450.

Mathewson, J. J., R. A. Oberhelman, H. Dupont L., F. Javier de la Cabada and E. V. Garibay (1987). "Enteroadherent Escherichia coli as a cause of diarrhea among children in Mexico." J Clin Microbiol 25 (10): 1917-1919.

Morello, E., E. Saussereau, D. Maura, M. Huerre, L. Touqui and L. Debarbieux (2011).

"Pulmonary Bacteriophage Therapy on Pseudomonas Aeruginosa Cystic Fibrosis Strains: First Steps Towards Treatment and Prevention." PLoS One 6 (2): e 16963.

Mossoro, C., P. Glaziou, S. Yassibanda, N. T. P. Lan, C. Bekondi, P. Minssart,. . . Y.

Germani (2002). "Chronic diarrhea, hemorrhagic colitis, and hemolytic-uremic syndrome associated with HEp-2 adhere to Escherichia coli in adults infected with human immunodeficiency virus in Bangui, Central African Republic." Journal of Clinical Microbiology 40 (8): 3086-3088.

Nataro, J.P., M.M. Baldini, J.B. Kaper, R.E. Black, N. Bravo and M.M. Levine (1985).

"Detection of an adherence factor of enteropathogenic Escherichia coli with a DNA probe." J Infect Dis 152 (3): 560-565.

Nataro, J.P. and J.B. Kaper (1998). "Diarrheagenic Escherichia coli." Clin Microbiol Rev

11 (1): 142-201.

Niu, Y. D., R. P. Johnson, Y. Xu, T. McAllister, R. Sharma, Louie, and K. Stanford (2009). "Host range and lytic capability of bacteriophage against bovine and clinical human isolates of Shiga toxin-producing Escherichia coli 0157: H7." J Appl Microbiol 107 (2): 646-656.

Ochman, H. and R. K. Selander (1984). "Standard reference strains of Escherichia coli from natural populations." J Bacteriol 157 (2): 690-693.

Orskov, F. and I. Orskov (1992). "Escherichia coli serotyping and disease in man

animais. "Can J Microbiol 38 (7): 699-704. Pajunen, M. L, SJ Kiljunen, ME Soderholm and M. Skurnik (2001). "Complete genomic sequence of the lytic bacteriophage phiYe03-12 of Yersinia enterocolitica serotype

0: 3. J Bacteriol 183 (6): 1928-1937.

Pichon, C., C. Héchard, L. du Merle, C. Chaudray, I. Bonne, S. Guadagnini,. . . C. The

Bouguénec (2009). "Uropathogenic Escherichia coli AL511 requires flagellum to enter renal collecting duct cells." Cellular Microbiology 11 (4): 616-628.

Puck, T.T., A. Garen and J. Cline (1951). "The mechanism of virus attachment to host cells.

I. The role of ions in the primary reaction. J Exp Med 93 (1): 65-88.

Ramage, G., K. Vandewalle, B. L. Wickes and J. Lopez-Ribot L. (2001). "Characteristics of biofilm formation by Candida albicans." Rev Iberoam Micol 18 (4): 163-170.

Rasko, D. A., D. Webster, R. W. Sahl, A. Bashir, N. Boisen, F. Scheutz,. . . Mr. K. Waldor

(2011). "Origins of the E.coli Strain Causing an Outbreak of Hemolytic-Uremic

Syndrome in Germany. "N Engl J Med.

Roberts, M. D., N. L. Martin and A. M. Kropinski (2004). "The genome and proteome of coliphage Tl." Virologv 318 (1): 245-266.

Russo, T. A. and J. R. Johnson (2000). "Proposai for a new inclusive designation for

Extraintestinal pathogenic isolates of Escherichia coli: ExPEC. J Infect Dis 181 (5):

1753-1754.

Sambrook, J. and D. W. Russel (2001). Protocol 11: Extraction of bacteriophage lambda DNA from large-scale cultures using proteinase K and SDS. Molecular Cloning, a laboratory manual. C. S. H. L. Press. 1: 2.56-52.58.

Sampaio, S. C. F., T. A. T. Gomes, C. Pichon, L. du Merle, S. Guadagnini, C. M. Abe,. . . vs.

The Bouguénec (2009). "The flagella of an atypical enteropathogenic Escherichia coli strain is required for efficient interaction and stimulation of interleukin-8 production by enterocytes in vitro." Infection and Immunity 77 (10): 4406-4413.

Savageau, M. A. (1983). "Escherichia coli habitats, cell types, and molecular mechanisms of gene control." The American Naturalist (122): 732-744.

Scholl, D., S. Adhya and C. Merril (2005). "Escherichia coli Kl 's capsule is a barrier to

bacteriophage T7. Appl. Microbiol 71 (8): 4872-4874.

Scholl, D. and C. Merril (2005). "The genome of bacteriophage K1F, has T7-like phage that has acquired the ability to replicate strains of Escherichia coli." Journal of

Bacteriology 187 (24): 8499-8503.

Schubert, S., P. Darlu, O. Clermont, A. Wieser, G. Magistro, C. Hoffmann,. . . E. Denamur

(2009). "Role of intraspecies recombination in the spread of pathogenicity in Escherichia coli species." PLoS Pathog 5 (1): el000257.

Schulze, J., M. Schiemann and U. Sonnenborn (2006). "120 years of E. coli.

importance in research and medicine. "Alfred-Nissle-Gesellschaft.

Steven, A. C, Trus B., Maizel J., Unser M, Parry D, Wall, J.S. . . F. W. Studier

(1988). "Molecular substructure of a viral receptor-recognition protein." The gpl7 tail-fiber of bacteriophage T7. J Mol Biol 200 (2): 351-365.

Toté, K., Vanden Berghe D, Maes L., and Cos P. (2008). "A new colorimetric microtitre model for the detection of Staphylococcus aureus bio films." Letters in Applied

Microbiology 46 (2): 249-254.

Touchon, M., C. Hoede, O. Tenaillon, V. Barbe, S. Baeriswyl, P. Bidet,. . . E. Denamur (2009). "Organized genome dynamics in the Escherichia coli species results in highly diverse adaptive paths." PLoS Genetics 5 (1): el000344-el000344.

Tré-Hardy, M., F. Vanderbist, H. Traore and MJ Devleeschouwer (2008). "In vitro activity of antibiotic combinations against Pseudomonas aeruginosa biofilm and planktonic cultures." International Journal of Antimicrobial Agents 31 (4): 329-336. Zuber, S., Ngom-Bru C., Barretto C., Bruttin A., Brussow H. and Denou E. (2007). "Genome analysis of phage JS98 defaults to fourth major subgroup of T4-like phages in

Escherichia coli. Journal of Bacteriology 189 (22): 8206-8214.

Claims

Nucleic acid molecule characterized in that it is chosen from molecules having one of the following sequences:

(ii) the nucleotide sequence shown in Figure 14 or also by SEQ ID No. 1;

(iv) a variant sequence of the sequence (ii) or of the sequence (iii) having an identity of at least 75%, in particular at least 78%, for example at least 80%, in particular at least 85%, by example at least 90% or at least 95% with the sequence

(ii) said identity being measured along the length of the shortest sequence of the so-called variant sequences on the one hand and (ii) or

(iii) respectively, respectively;

Nucleic acid molecule according to claim 1, characterized in that it is the ORF 17.1 bacteriophage CLB_P1 gene.

Nucleic acid molecule, characterized in that it is a fragment comprising at least 9 consecutive nucleotides of the sequence of the nucleic acid molecule according to one of claims 1 or 2, in particular a fragment having from 9 to 27 consecutive nucleotides of said molecule.

Polypeptide characterized in that it is encoded by a nucleic acid molecule according to any one of claims 1 to 3.

5. Polypeptide characterized in that it has an amino acid sequence chosen from:

(i) the amino acid sequence shown in Figure 14 or SEQ ID No. 2;

(ii) an amino acid sequence comprising the sequence (i) or an amino acid sequence constituting a fragment of the sequence of the gp17.1 protein of (i), in particular a fragment of at least 5 amino acids whose amino acid sequence is from residue 6 to residue 430 of gp17.1 of CLB_P1 (residues 6 to 430 in the sequence SEQ ID No. 2);

(iii) an amino acid sequence variant of the sequence (i) or of the sequence (ii) having an identity of at least 75%, in particular at least 78%, for example at least 80%, especially at least 85%, for example at least 90% or at least 95% with the sequence (i) said identity being measured along the length of the shortest sequence of the so-called variant sequences on the one hand and (i) or (ii) d 'somewhere else.

6. Polypeptide according to claim 5, characterized in that it comprises an epitope recognized by antibodies induced against the bacterial receptor recognition region of the bacteriophage gp17.1 protein CLB_P1 deposited at the CNCM under No. I -4534 or the gp17.1 fiber protein having the amino acid sequence SEQ ID NO: 2.

7. Virulent bacteriophage of the family Podoviridae capable of specifically infecting a bacterial strain of E. coli serotype O104, characterized by the presence in its genome of a sequence according to any one of claims 1 to 5 or a open reading frame (ORF) designated gp17.1 comprising a gene having the nucleotide sequence SEQ ID NO: 1 encoding a gp17.1 fiber protein.

8. Bacteriophage according to claim 7, characterized in that it is deposited in the CNCM Collection under No. 1-4534 and designated CLB_P1.

9. Bacteriophage according to claim 7 or claim 8, characterized in that its genome is mutated in the gp17 gene, said gp17 gene is no longer expressed or in that it comprises a fusion polynucleotide between the sequence coding the domain N-terminal of gp17 and the sequence encoding the receptor recognition domain at the bacterial surface of the gp17.1 protein, to allow the binding of the expressed recombinant protein, on the tail of the bacteriophage.

10. Bacteriophage according to claim 9, characterized in that it does not comprise the entire sequence of the gp17 gene, or does not include IORF17 or the gene 17.

1 1. Bacteriophage genome according to any one of claims 7 to 10.

An antibody induced against a polypeptide according to any one of claims 4 to 6.

Antibody according to claim 12, polyclonal or monoclonal.

14. Kit for the in vitro detection of an infection with a bacterium E.coli serotype 0104 comprising a bacteriophage according to any one of claims 7 to 10, or a polypeptide according to any one of claims 4 to 6, or antibodies according to claim 12 or 13.

15. A method for the in vitro detection of an infection with a bacterium E.coli serotype O104 comprising contacting a biological sample, in particular a biological fluid sample, with a bacteriophage according to any one of claims 7 to 10, or a polypeptide according to any of claims 4 to 6, or antibodies according to claim 13 or 14.

16. In vitro detection method according to claim 15, characterized in that it comprises an ELISA immunoenzymatic reaction with antibodies according to claim 13 or 14.

17. A method for the in vitro detection of an infection with a serotype 0104 E. coli bacterium comprising culturing a biological sample under growth conditions and optionally counting E. coli bacteria with a bacteriophage according to any one of claims 7 to 10 and the determination of the lysis of E. coli bacteria.

18. Therapeutic composition for the treatment of an infection with E. coli O104: H4 bacteria, comprising as active ingredient at least one bacteriophage CLB_P1 deposited under the number CNCM I-4534, CLB_P2 filed under the CNCM number I- 4675 or CLB_P3 deposited under the number CNCM I-4676, and a pharmaceutically acceptable vehicle.