Engineered phagemids and methods for antibacterial peptide delivery against bacterial pathogens in particular clostridioides difficile

Abstract

The present invention concerns engineered phagemids and methods for antibacterial peptide delivery against bacterial pathogens, in particular Clostridioides difficile.

Description

[0001] ENGINEERED PHAGEMIDS AND METHODS FOR ANTIBACTERIAL PEPTIDE DELIVERY AGAINST BACTERIAL PATHOGENS IN PARTICULAR CLOSTRIDIOIDES DIFFICILE

[0002] DESCRIPTION

[0003] Technical field of the invention

[0004] The present invention concerns engineered phagemids and methods for antibacterial peptide delivery against targeted bacterial strain, in particular a strain of Clostridioides difficile.

[0005] The present invention finds its applications mainly in C. difficile infection treatments and more generally specific targeting of bacterial pathogens or enteropathogens.

[0006] In the description below, references in square brackets ([ ]) refer to the list of references at the end of the text.

[0007] State of the art

[0008] Clostridioides difficile, an emerging human enteropathogen, is a major cause of life-threatening hospital-acquired infections associated with antibiotic therapy and a major burden worldwide, with annual healthcare costs estimated at $4.8 billion (Paparella et al., 2021 ) [1], Current treatment of C. difficile infections depends solely on three available antibiotics (metronidazole, vancomycine, fidaxomicine) and fecal transplantation (FMT) in the most serious cases after recurrence. Unfortunately the treatment failure rate is high and highly resistant strains are emerging all over the world, leading to rising recurrence rates up to 25-50%. Despite promising results, FMT remains a major operation and the long-term effects are unknown. In addition, large-scale commercial production of fecal transplants is very difficult, with the risk of transmission of bacterial or viral pathogens making the verification process tedious. In this regard, cases of FMT-related deaths have been reported in immunocompromised patients (Marcella et al., 2021 ) [2], On the other hand, recent efforts by major pharmaceutical companies (Sanofi-Pasteur / Pfizer / Merck) to develop vaccines have not achieved their objectives.

[0009] In this context, it seems essential to reinforce the current therapeutic arsenal with new strategies specifically targeting C. difficile. Bacteriophages, viruses that specifically infect their host bacteria species, are gaining huge interest in the scientific and medical community to treat a growing number of pathogenic and resistant bacteria. For example, Patent Application US 2020 / 0254035 proposes a strategy for efficient production of phage particles comprising components of a CRISPR / Cas system for killing target E. coli Nissle strain bacteria, as well as for Staphyloccocus aureus mobile genetic elements (MGEs, genomic islands) SaPI and S. aureus phages, although no experimental data are provided for suggested strategy. Thus the use of bacteriophages to target C. difficile appeared to be an attractive strategy for precision medicine and preservation of a healthy gut microbiota, considered as a major protective element against C. difficile infections and relapses. However, C. difficile natural phages have important limitations for the treatment of intestinal pathogens since efficient and strictly lytic phages, which are usually considered suitable for phage therapy in other bacteria, have never been identified in C. difficile. Whereas a large number of phages have been found to infect C. difficile, they are all temperate and inefficient to kill their sensitive bacterial host. These phages can inject their DNA inside the cell, but most of the time, the phage genome of a temperate phage integrates the bacterial genome, and become latent, without killing the bacteria. In addition, when the genome of a temperate phage integrates the bacterial genome, the bacterium becomes resistant to infection and lysis by other similar phages, thus reducing even more the potential of natural phages for treatment. Consequently, this feature of temperate phages to integrate the bacterial genome and become dormant has always been considered as a major obstacle and disincentive for their use as therapeutic agent against C. difficile infections. Also, in the hypothetical case a C. difficile natural lytic phage may be identified, studies have shown that there is a number of phage defence systems encoded by C. difficile, notably a large diversity of CRISPR-Cas systems conferring a protection to the bacteria against phages, further reducing the potential of natural phages as a treatment option. Usually, most of the anti-phage systems will not impede phage adhesion and phage DNA injection inside the bacteria, but will prevent the phage from replicating inside the bacteria, before new phages can be produced, and before the event of bacterial lysis (cell death).

[0010] Description of the invention The Inventors propose a new strategy against the human pathogen C. difficile using phage-derived vectors, called phagemids. Phagemid strategy allows to target C. difficile without affecting the healthy protective flora, which would allow a faster recovery of patients, and certainly reducing the risk of relapses.

[0011] Phagemids are DNA-based cloning vectors, which have both bacteriophage and plasmid properties. These vectors carry, in addition to the origins of plasmid replication (allowing its replication in a producing strain and a host strain), a sequence for encapsidation derived from the bacteriophage. Phagemids differ from plasmids by having the ability to be packaged into the capsid of a bacteriophage by specific genetic sequences (phage packaging signals) introduced on the plasmids (Figure 1A). The use of phagemids allows to keep the most interesting feature of natural phages, i.e. using the high target specificity of the phage, without the problems related to lysogeny (dormancy) and resistance to phage as above described.

[0012] So far, the Inventors successfully engineered phagemids, which once injected in C. difficile, produce and deliver in the cytoplasm a small peptide having antibacterial properties (ABP) targeting the bacterial membrane leading to highly effective inhibition of C. difficile growth (Figure 1 B). This small peptide is the toxin component of a type I (T1 ) Toxin-Antitoxin (TA) system produced by C. difficile which has recently been characterized (Maikova et al., 2018) [3], In addition, type I (T1 ) toxin-antitoxin (TA) systems were predicted or identified in other different bacteria including major pathogens (493 bacterial strains) with 25 families of T1TA modules (Tourasse et Darfeuille, 2021 ) [4], In this system, translation of the toxin can be prevented by the production of an antitoxin non coding RNA interacting with toxin mRNA.

[0013] In the gut, such strategy blocking C. difficile growth will lead to a progressive decolonisation of C. difficile, as it will be outcompeted by the natural competitors of the gut microbiota, leading to its progressive elimination and decreasing the rate of recurrent infections.

[0014] An objective of the present invention is therefore a recombinant phagemid for expressing at least one gene expression cassette comprising a nucleic acid sequence encoding an antibacterial peptide that inhibits the growth of a targeted strain, in particular a strain of C. difficile, and a nucleic acid sequence encoding a phage packaging signal for encapsidation (SSE) of the recombinant phagemid into bacteriophage capsids, wherein the antibacterial peptide is the toxin component of a type I (T1 ) Toxin-Antitoxin (TA) system, in particular which is produced by a strain of the same species than the targeted strain, more particularly which is the T1TA system produced by a strain of C. difficile.

[0015] By “targeted strain” it is meant the bacterial strain that is specifically recognized by the phagemid particle and sensitive to the action of antibacterial peptide, i.e. the toxin component of the T1 TA system.

[0016] By “sequence for encapsidation (SSE)” it is meant the sequence corresponding to the signal used by a helper phage which is specific for the targeted strain.

[0017] Another objective of the present invention, is a method for producing a recombinant phagemid particle, the method comprising : i) introducing into a prokaryotic host cell, a recombinant phagemid of the present invention configured to persist inside the bacterial host; ii) transfer (e.g. by bacterial conjugation or transformation) of the recombinant phagemid to a recombinant phagemid producing strain, in particular a strain of C. difficile, expressing the antitoxin component of the type I (T1 ) Toxin-Antitoxin (TA) system corresponding to the toxin component of the recombinant phagemid; iii) infection of the recombinant phagemid producing strain from step ii), in particular a strain of C. difficile, with a helper phage comprising a nucleic acid sequence encoding bacteriophage structural proteins; iv) culturing the recombinant phagemid producing strain from step iii), in particular a C. difficile strain, under conditions which result in the doublestranded DNA of the recombinant phagemid being packaged by the structural proteins to form the recombinant phagemid particle; and v) bacterial lysis to recover the recombinant phagemid particle from recombinant phagemid producing strain from step iv), in particular a strain of C. difficile.

[0018] By “recombinant phagemid producing strain” it is meant a strain of the same species than the targeted strain, and used for the recombinant phagemid production, that can receive the phagemid expressing the toxin for example by bacterial conjugation or transformation, then be infected by specific helper phage but is immune to toxin action thanks to its constitutive antitoxin expression. By “bacteriophage structural proteins” it is meant all the proteins required for the production of phage capsid and phage particles.

[0019] According to a particular embodiment of the present invention, the gene expression cassette comprises a constitutive promoter, and optionally a marker gene.

[0020] According to a particular embodiment of the present invention, the procaryote host cell is an Escherichia coli strain.

[0021] Another objective of the present invention is a recombinant phagemid particle obtained by a method of the present invention, wherein the phagemid particle comprises a recombinant phagemid of the present invention.

[0022] Another objective of the present invention is a pharmaceutical composition comprising at least one recombinant phagemid particle of the present invention, and a pharmaceutically acceptable vehicule.

[0023] Another objective of the present invention is a recombinant phagemid particle or a pharmaceutical composition of the present invention, for use as a medicament.

[0024] Another objective of the present invention is a recombinant phagemid particle or a pharmaceutical composition of the present invention, for use in the prevention or treatment of a bacterial infection due to Clostridioides difficile, Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterococcus faecalis, Enterococcus faecium, Salmonella Typhi, Salmonella enterica, Shigella flexneri, Haemophilus influenzae, Streptococcus pneumoniae, Brucella abortus, Campylobacter jejuni, Helicobacter pylori, Clostridium botulinum, Clostridium perfringens, Listeria monocytogenes, Neisseria gonorrhoeae, Neisseria meningitidis, Vibrio cholerae, preferably a C. difficile infection.

[0025] Another objective of the present invention is a recombinant phagemid producing strain as defined above constitutively expressing the antitoxin component of the type I Toxin-Antitoxin (TA) system corresponding to the toxin component of the recombinant phagemid of the present invention.

[0026] According to a particular embodiment of the present invention, the recombinant phagemid producing strain of the present invention further comprises a recombinant phagemid of the present invention. According to a particular embodiment of the present invention, the recombinant phagemid producing strain of the present invention is a C. difficile strain.

[0027] Another objective of the present invention is a prokaryotic host cell comprising a recombinant phagemid of the present invention.

[0028] Definitions

[0029] Brief description of the figures

[0030] Figure 1 shows a schematic representation of (A) the production of phagemid particles encapsidating a phagemid thanks to the presence of DNA sequence encoding a phage packaging signal for encapsidation (SSE), and (B) the innovative phagemid particle strategy to inhibit the growth of a C. difficile strain by the production of antibacterial protein (ABP) toxin from a type I toxin-antitoxin module.

[0031] Figure 2 shows a schematic representation of (A) the sequence map of the phagemid p168, (B) the sequence map of the phagemid p360, (C) the sequence map of the phagemid p170, as illustrative examples of the phagemid sequences of the invention, and (D) the sequence map of the phagemid p167, (E) the sequence map of the phagemid p361 , (F) the sequence map of the phagemid p169, as control examples without toxin with respect to p168, p360, p170, respectively. Phagemid p168 carries the SSE of phage phiCD630-2 and the toxin component CD0977.1 of a type I Toxin-Antitoxin (TA) under the control of an anhydrotetracycline (ATc)- inducible promoter. Phagemid p360 carries the SSE of episomal phage phiCD38-2 and the toxin component CD0977. 7 of a type I Toxin-Antitoxin (TA) under the control of an anhydrotetracycline (ATc)-inducible promoter. Phagemid p170 carries the SSE of phage phi027 and the toxin component CD0977. of a type I Toxin-Antitoxin (TA) under the control of an anhydrotetracycline (ATc)-inducible promoter.

[0032] Figure 3 shows the inhibitory activity on the growth of C. difficile of phagemids p168 and p170 constructed to be packaged into the phiCD630-2 and phi027 phage capsids, respectively, and introduced by bacterial conjugation in C. difficile (A) CD24 strain and (B) R20291 Aphi027 strain deleted of its natural prophage phi027, in the absence or presence (+ATc) of the inducer. Plasmids p167 and p169 comprise the nucleic acid sequence encoding a DNA packaging signal but not the antibacterial protein and are used as controls in CD24 and (B) R20291 Aphi027 strains, respectively.

[0033] Figure 4 shows the inhibitory activity of phagemid p168 introduced in the lysogenized CD24::phiCD630-02 strain, in the absence or presence of the inducer (+ATc), on the growth of C. difficile. Plasmid p167 comprising the nucleic acid sequence encoding a DNA packaging signal but not the antibacterial protein is used as control.

[0034] Figure 5 shows a schematic representation of combined Figures 1 A and 1 B.

[0035] Figure 6 shows the impact of toxin CD0977.1 expression on growth of C. difficile strains CD24 (A) and CD126 (B, C) harboring the first generation of phagemids in TY broth at 37 °C supplemented with thiamphenicol (7.5 pg / ml) and in the presence of 10 ng / mL ATc. Plotted values represent means and error bars represent standard error of the means (3 biologically independent samples).

[0036] Figure 7 shows the sequence map of 2ndgeneration phagemids p362, p363, p364, p365, p366 and p367.

[0037] Figure 8 shows the impact of toxin CD0977.1 expression on growth of C. difficile strains CD24 (A, C) CD126 (B, D) harboring the second generation of plasmids in TY broth at 37°C supplemented with thiamphenicol (7.5 pg / ml) and in presence of 10 ng / ml ATc (A, B) and 250 ng / ml (C, D). Plotted values represent means and error bars represent standard error of the means (3 biologically independent samples). Toxicity effect of ATc at 250 ng / ml on growth was also evaluated. Higher concentration of ATc is needed for a complete growth inhibition of the strains harboring the second generation of phagemids. Figure 9 shows in vivo efficacy of C. difficile strains containing 2ndgeneration phagemids whose toxin expression is inducible with ATc (strains CD791 (CD24+p362) and CD792 (CD24+p363)) on gut colonization using the 0MM12 gnotobiotic mouse model with a stable microbiota of 12 bacterial strains. The upper panel depicts the mouse body weight changes as expected during the experiment and the lower panel shows the clearance kinetic of C. difficile strains up to 10 days post-challenge.

[0038] Figure 10 shows the sequence map of 3rdgeneration phagemids p311 , p312, p336, p368, p369 and p370.

[0039] Figure 11 shows protection of the strains against toxin action. (A) PCR analysis to confirm insertion of RCd11 and its native promoter in the vanS locus of CD126 (R20291AphiCD027) and CD156 (630Aem? AphiCD630-1AphiCD630-2) genomes. (B) RT-qPCR analysis results for RCd11 expression from the protected strains CD126-AT and CD156-AT compared to the WT. Expression normalized against a housekeeping gene rpoA, bars represent the mean of 3 biologically independent samples ± SD. (C) Growth of CD156-AT harboring toxin expressing plasmid (pTox+T1TA) and empty plasmid (p+T1TA).

[0040] Figure 12 shows the sequence map of phagemids p375, p349 and p176.

[0041] EXAMPLES

[0042] EXAMPLE 1 : DESIGN AND PRODUCTION OF PHAGEMIDS AND PHAGEMID PARTICLES

[0043] Plasmids

[0044] To construct phagemids, the plasmid pDIA6335 (Peltier et al., 2020) [5] expressing the toxin gene CD0977. 1 from a Toxin-Antitoxin (TA) system under the control of an anhydrotetracycline (ATc)-inducible promoter, thereby controlling the level of expression of the target gene, was used. In parallel, the plasmid pDIA6103 (Soutourina et al., 2013) [6], which is identical to the previous plasmid pDIA6335 except that it lacks the toxin gene CD0977. 1, was used as a control plasmid.

[0045] Sequence-Signal of Encapsidation (SSE) The encapsidation sequence (SSE) is a key sequence present on the DNA of bacteriophage genomes which allows the specific encapsidation of this DNA during the production of new bacteriophage particles. So cloning a SSE into a plasmid enables the plasmid to be encapsidated, as a phagemid. A phage that has encapsidated such a phagemid is therefore called a phagemid particle.

[0046] The SSE of phage phiCD630-2 from the C. difficile historical strain 630; and of phage phi027 and episomal phage phiCD38-2 from the C. difficile epidemic strain R20291 were identified using PhageTerm (Garneau et al., 2017) [7], and cloned into both plasmids pDIA6103 and pDIA6335 as described below.

[0047] Cloning SSE in plasmids to produce phaqemids

[0048] To clone SSE of phage phiCD630-2 and phage phi027 in plasmids pDIA6103 and pDIA6335, both of them were first linearised by reverse PCR using the primers JP403 (CTGTATCGTAACTAGAGAACCAAAC, SEQ ID NO: 1 ) and JP404 (GGGATATGCTTATATTGAGTTATAGTAC, SEQ ID NO: 2). To clone SSE of phage phiCD38-2 in plasmids pDIA6103 and pDIA6335 background, plasmids p169 and p170 were first linearized by reverse PCR using the primers IQC003 (CTCAACACCTATATGCGGAGGTACTAATCAGTGCTTAATATAG, SEQ ID NO: 7) and IQC004 (TGGTATTTCTGTTCATGTGCCGATATAGATAAGTTTAGATAG, SEQ ID NO: 8).

[0049] The SSE of phage phiCD630-2 and phages phiO27 / phiCD38-2 were directly amplified from the genomes of C. difficile strain 630, R20291 or R20291 carrying phiCD38-2, respectively, as the DNA of these bacteriophages is integrated into the chromosomal DNA or maintained as an episome (prophage). The SSE of phage phiCD630-2 was amplified by PCR from chromosomal DNA of C. difficile strain 630 using the primers JP816

[0050] (ACTCAATATAAGCATATCCCCTTTCAATTTTCTTTTGAAATTGC, SEQ ID NO: 3) and JP817

[0051] (GTTCTCTAGTTACGATACAGCTAACTATCATTTCTATCACCATAATC, SEQ ID NO: 4). The SSE of phage phi027 was amplified by PCR from chromosomal DNA of C. difficile strain R20291 using the primers JP818 (ACTCAATATAAGCATATCCCAAGTTGGATACTGAGGGTATAAAG, SEQ ID NO: 5) and JP819 (GTTCTCTAGTTACGATACAGCTGCATACGTTGAATACTTATG, SEQ ID NO: 6). The SSE of episomal phage phiCD38-2 was amplified by PCR from the DNA of C. difficile strain R20291 carrying phiCD38-2 episomal prophage using the primers IOC001 (CTCCGCATATAGGTGTTGAGAGAGTAAC, SEQ ID NO: 9) and IOC002 (GCACATGAACAGAAATACCACTATC, SEQ ID NO: 10).

[0052] The amplified SSE sequences were then all cloned by Gibson Assembly in linearized plasmids pDIA6103 and pDIA6335. The phagemids resulting from these constructs are as follows: p167 (pDIA6103+SSE of phiCD630-2), p168 (pDIA6335+SSE of phiCD630-2), p169 (pDIA6103+SSE of phi027), p170 (pDIA6335+SSE of phi027), p360 (pDIA6335+SSE of phiCD38-2) and p361 (pDIA6103+SSE of phiCD38-2). The table 1 below summarizes the first generation phagemid constructs.

[0053] Table 1

[0054] As an example, Figure 2 shows (A) the sequence map of phagemid p168 constructed to be packaged into the phiCD630-2 phage capsid, and thereafter transferred (injected) in the CD24 target strain, (B) the sequence map of phagemid p360 constructed to be packaged into the phiCD38-2 phage capsid, and thereafter transferred (injected) in the R20291 target strain, and (C) the sequence map of phagemid p170 constructed to be packaged into the phi027 phage capsid, (D) the sequence map of phagemid p167 (control without toxin) as for p168, (E) the sequence map of phagemid p361 (control without toxin) as for p360, and thereafter transferred (injected) in the CD126 target strain, and (F) the sequence map of phagemid p169 (control without toxin) as for p170.

[0055] First generation of phagemid particles The surface protein A (SIpA) of C. difficile acts as a general receptor for bacteriophage infection. The sequence of this protein varies from strain to strain and several types have been defined (Royer et al., 2023) [8], Phages of C. difficile often recognize only one specific variant of SIpA protein. Therefore the phage phiCD630- 2 and the phage phi027 / episomal phage phiCD38-2 can only infect C. difficile strains producing type 7 (e.g. CD360, riboypes 012) and type 4 (e.g. R20291 , ribotypes 027) SIpA protein, respectively.

[0056] Introduction of phaqemids into specific strains of C. difficile

[0057] To produce the phagemid particles of interest, the phagemids produced as above-described were first introduced into the Escherichia coli conjugating strain HB101 (RP4) allowing their subsequent transformation into specific strains of C. difficile by bacterial conjugation.

[0058] Both p168 and p167 phagemids, encoding or not the toxin CD0977.1 from a Toxin-Antitoxin (TA) system, respectively, were introduced into the C. difficile strain CD24 because this strain produces a type 7 SIpA protein and is sensitive to infection by the natural phage phiCD630-2 in which the phagemid sequence will be packaged to become the final phagemid particle. The resulting strains containing the p168 or p167 phagemid have been named CD348 (CD24+p167) and CD350 (CD24+p168).

[0059] On the other hand, p170 and p169 phagemids, encoding or not the toxin CD0977.1 , respectively, were introduced into the hypervirulent C. difficile strain R20291 deleted from prophage phi027 (R20291Aphi027, CD126). The R20291Aphi027 strain (CD126) produces a type 4 SIpA protein and is thus sensitive to infection by the phage phi027. The resulting strains containg the p169 or 170 phagemid have been named CD352 (R20291Aphi027+p169) and CD354 (R20291Aphi027+p170).

[0060] On another hand, p360 and p361 phagemids, encoding or not the toxin CD0977.1 , respectively, were introduced into the C. difficile strain R20291 deleted from prophage phi027 (R20291Aphi027 (CD126)). The R20291Aphi027 strain produces a type 4 SIpA protein and is thus sensitive to infection by the phage phiCD38-2. The resulting strains containing the p360 or p361 phagemid have been named CD789 (R20291Aphi027+p360) and CD790 (R20291Aphi027+p361 ).

[0061] The inhibition of the growth of strains CD24 and CD126 by production of the toxin CD0977.1 from the phagemid p168, p170 or p360 was tested. Figure 3A shows that the strain CD348 (CD24+p167) that does not produce the toxin CD0977.1 , is able to grow in the absence and presence of the ATc inducer whereas the strain CD350 (CD24+p168) cannot grow in the presence of the ATc inducing the toxin CD0977.1 expression (also see Figure 6A). Figure 3B shows that the same applies to the strain CD352 (R20291Aphi027+p169) that does not produce the toxin CD0977.1 and the strain CD354 (R20291Aphi027+p170) that produces the toxin CD0977.1. The CD352 strain grows in the absence and presence of ATc whereas the strain CD354 only grows in the absence of ATc (when toxin CD0977.1 is not expressed) (also see Figure 6B). The same results are obtained with the strains CD789 and CD790 (Figure 6C).

[0062] Encapsidation of phagemids by phages

[0063] This step was only carried out with the phage phiCD630-2.

[0064] Particles of the phage phiCD630-2 were obtained by inducing prophages of the strain 630 of C. difficile with mitomycin C. Phages were then isolated, purified and amplified, and then used to infect strains CD348 and CD350. During the encapsidation step, the phage DNA and the phagemids possessing the SSE were packaged indiscriminately in the new capsids produced. Following bacterial lysis, a mixed population of wild phages and phagemid particles having encapsidated either the p167 phagemid (pDIA6103+SSE of phiCD630-2) or the p168 phagemid (pDIA6335+SSE of phi630-2), was obtained and purified.

[0065] Second generation of phagemid particles

[0066] To directly evaluate the efficiency of phagemids in reducing intestinal colonization of C. difficile an additional modification of initial phagemids has been introduced. The vector for use in vivo has been modified by increasing its stability by adding a complete type I toxin-antitoxin (T1TA) module to the vectors constructed at the first stages to eliminate the selection with antibiotic resistance markers. A new seguence composed of the complete T1TA module CD0956.2 / RCd10 has been added to the first generation of phagemids. Indeed, this T1TA module has been shown to confer plasmid stabilization in the absence of selective pressure (Peltier et al., 2020) [5], The resulted phagemids were suitable for first in vivo experiment in which the use of thiamphenicol antibiotic is avoided. Cloning CD0956.2 / RCd10 T1TA module in phagemids for stabilization

[0067] To clone CD0956.2 / RCd10 T1TA module in phagemids p167, p168, p169, p170, p360 and p361 , all of them were first linearized by reverse PCR using the primers IOC014 (TTGAGGCGGCGCATTGGAAATGCTAGATTC, SEQ ID NO: 11 ) and IOC013 (AAAAAAAATGAAGCACCTACTAATTAGGATTC, SEQ ID NO: 12).

[0068] The amplified CD0956.2 / RCd10 T1TA module sequence was then cloned by Gibson Assembly linearized phagemids p167, p168, p169, p170, p360 and p361. The phagemids resulting from these constructs are as follows: p362 (p167+T1TA), p363 (p168+T1TA), p364 (p169+T1TA), p365 (p170+T1TA), p366 (p361 +T1TA) and p367 (p360+T1TA).

[0069] The table 2 below summarizes the second generation phagemid constructs and Figure 7 shows the sequence map of resulting phagemids.

[0070] Table 2

[0071] Introduction of phagemids into specific strains of C. difficile

[0072] New p362, p363, p364, p365, p366 and p367 phagemids, encoding or not the toxin CD0977.1 , respectively, were introduced into the C. difficile strain CD24 (for phagemids with SSE of phiCD630 phage) or R20291 deleted from prophage phi027 (strain CD126) (for phagemids with SSE of phi027 or phiCD38-2 phages). The resulting strains containing the p362, p363, p364, p365, p366 and p367 phagemids have been named CD791 (CD24+p362), CD792 (CD24+p363), CD793 (R20291 Aphi027+p364), CD794 (R20291 Aphi027+p365), CD795

[0073] (R20291 Aphi027+p366) and CD796 (R20291Aphi027+p367).

[0074] The inhibition of the growth of strains CD24 and R20291 Aphi027 by production of the toxin CD0977.1 from the phagemid p363, p365, or p367 was tested. The Figure 8 shows the efficient growth inhibition in the presence of ATc inducer for strains carrying phagemids producing CD0977.1 toxin.

[0075] In vivo testing of phagemids to lower intestinal colonization of C. difficile

[0076] Initial in vivo experiments have been performed in a mouse model, using C. difficile strains containing 2ndgeneration phagemids whose toxin expression is inducible with ATc (strains CD791 (CD24+p362) and CD792 (CD24+p363)). These experiments mimic the final stage of the phagemid therapeutic approach.

[0077] To test in vivo phagemid efficacy on C. difficile gut colonization, the 0MM12 model was used, a gnotobiotic mouse model with a defined stable microbiota of 12 bacterial strains (Brugiroux et al., 2016) [9], known as a tool for investigating C. difficile infection within a functional microbial ecosystem. Mice orally infected with C. difficile spores were monitored daily for their body weight and symptoms, and fecal samples collected to assess the intestinal level of C. difficile. The study indicates a slight drop in body weight for both the control (no toxin, CD791 (CD24+p362)) and inducer (toxin expressed, CD792 (CD24+p363)) groups on day 2, which stabilizes over time (Figure 9, up). No symptom was observed for both groups. Concerning C. difficile colonization monitored from fecal samples along infection, a gradual decline was observed in the C. difficile population for the control group as usually observed in mice model (competitive impact of the commensal strains), while in the case of inducer group, C. difficile was not seen even at an early point (Figure 9, down). This indicated that when harboring toxin-producing phagemid, C. difficile cells could not grow in the 0MM12 model and confirm by mimicking the last stage of the therapeutic strategy ( / .e., transfer of the phagemid into the target C. difficile strains) that the system works in vivo.

[0078] Third generation of phagemid particles

[0079] All phagemid used for the initial and optimization steps (1stand 2ndgeneration) correspond to those expressing the toxin of the type I TA system under the control of an inducible promoter. For the development of the final phagemid product we constructed the phagemid vector expressing constitutively the toxin by cloning the toxin-encoding gene under the control of a promoter that is constitutive to avoid the need of induction, in particular under the control of its own promoter / native promoter. The toxic effect of this new phagemid on C. difficile growth was verified in vitro by the absence of transconjugants due to the lethal expression of the toxin, unlike the control vector. This type of phagemid particle (and its derivatives) is the one that must be used in therapeutic trials.

[0080] For the phagemids of first and second generation, ATc induction is needed in order to express the toxin. To avoid using the inducer ATc, a third generation of plasmids has been constructed, in which the tetracycline promoter system is replaced with CD0977. 1 native promoter for a constitutive expression of the toxin.

[0081] Cloning CD0977.1 gene with its own promoter into phagemids

[0082] To clone CD0977.1 gene with its own promoter in phagemids p168 and p170, both of them were first linearized by reverse PCR using the primers JP722 (GTGAAAGTGGGTCTTAAGGTAC, SEQ ID NO: 15) and JP420 (GGATCCTATAAGTTTTAATAAAACTTTAAATAG, SEQ ID NO: 16). The CD0977. 1 gene with its own promoter has been amplified by PCR from the plasmid DNA pDIA6787 (pMTL84121 + T1TA CD0977.1 module, Peltier et al, 2020) [4] using primers AH001

[0083] (TTTATTAAAACTTATAGGATCCAAAGATGTAAATTCGTATCAAAAAC, SEQ ID NO: 17) and AH002

[0084] (CTTAAGACCCACTTTCACTAACATATTTAGTATATACCTATGTA, SEQ ID NO: 18). The amplified CD0977.1 seguence was then cloned by Gibson Assembly in linearized phagemids p168 and p170. The phagemids resulting from these constructs are p311 (p168+promoter-CD0977. 1) and p312 (p170+ promoter- CD0977. 7).

[0085] The resulting p312 phagemid was linearized by reverse PCR using the primers IQC003 (CTCAACACCTATATGCGGAGGTACTAATCAGTGCTTAATATAG, SEQ ID NO: 7) and IQC004

[0086] (TGGTATTTCTGTTCATGTGCCGATATAGATAAGTTTAGATAG, SEQ ID NO: 8) to introduce SSE for phiCD38-2 phage giving p336 phagemid as previously described. The p336 phagemid was in turn linearized by reverse PCR with primers IQC014 (TTGAGGCGGCGCATTGGAAATGCTAGATTC, SEQ ID NO: 11 ) and IOC013 (AAAAAAAATGAAGCACCTACTAATTAGGATTC, SEQ ID NO: 12) to introduce complete CD0956.2 / RCd10 T1TA module giving p370 phagemid.

[0087] The p311 and p312 phagemids were linearized by reverse PCR with primers IOC014 (TTGAGGCGGCGCATTGGAAATGCTAGATTC, SEQ ID NO: 11 ) and IOC013 (AAAAAAAATGAAGCACCTACTAATTAGGATTC, SEQ ID NO: 12) to introduce complete CD0956.2 / RCd10 T1TA module giving p368 and p369 phagemids, respectively.

[0088] The table 3 below summarizes the third generation phagemid constructs and Figure 10 shows the sequence map of resulting phagemids. Table 3

[0089] EXAMPLE 2: VALIDATION OF PHAGEMID PARTICLES PRODUCTION AND

[0090] ACTIVITY Validation of phagemid particles production

[0091] In order to confirm that phagemid particles have indeed been generated, transduction experiments were carried out to determine the phagemid particles titre. In this experiment, the lysogenized CD24 strain that integrated the phage phiCD630-2 into the bacterial chromosome, giving the strain CD24::phiCD630-2, was used. The advantage of this strain is that it can be infected by phagemids but will be insensitive to infection by the wild-type phage phiCD630-2. The titre of the phagemid particles was then determined by spreading the CD24::phiCD630-2 strain after infection with phagemid particles on a rich medium containing thiamphenicol. Indeed phagemids p167 and p168 have a thiamphenicol resistance cassette, which allows them to be selected.

[0092] This test showed that both phagemids p167 and p168 were able to be transferred into C. difficile cells with an average titre of 5.75 x104transductants / ml. Two strains of C. difficile CD24::phiCD630-2 with phagemids p167 and p168 introduced by transduction with the phagemid particles, were thus isolated. The resulting strains containing the p167 or p168 phagemid have been named CD24167 (CD24::phiCD630-2+p167) and CD24168 (CD24::phiCD630-2+p168).

[0093] Validation of the inhibitory activity of phagemid particles on C. dfff / c / / e growth

[0094] The growth inhibition of the CD24168 strain (CD24::phiCD630-2+p168) by production of the toxin CD0977.1 was then tested.

[0095] As shown in Figure 4, the CD24167 strain (CD24::phiCD630-2+p167) that does not produce the toxin CD0977.1 was able to grow in the absence and presence of the inducer ATc whereas the CD24168 strain (CD24::phiCD630-2+p168) could not grow when the toxin CD0977.1 is produced in presence of ATc.

[0096] EXAMPLE 3: PHAGEMID PRODUCING STRAINS

[0097] To produce phagemid particles constitutively expressing the "toxin", a phagesensitive strain of C. difficile immune to the toxin is needed, i.e., whose expression is neutralized by the antitoxin of the TA system. Such strain was produced by integrating the antitoxin gene into the chromosome under the control of its own promoter for a strong constitutive expression.

[0098] To produce the phagemids, phages have to infect the strains harboring the plasmids expressing the toxin. During the infection, the encapsidation sequence will be recognized leading to the integration of the plasmid in the capsid when new phage particles are produced and assembled. However, plasmids of the third generation are toxic to the strains due to the strong constitutive expression of the toxin. To achieve the survival of strains harboring these plasmids, they must be protected from the toxic effect. A strategy involving the insertion of the antitoxin gene RCd11 into the vanS locus of the C. difficile strain's genome was employed. Antisense RNA RCd11 expressed from its own promoter neutralizes the toxin CD0977.1 (Peltier et al., 2020) [5], Thus, by overexpressing the antitoxin, the toxic effect of the plasmid can be mitigated, enabling the strain to survive and to produce phagemids.

[0099] The strain expressing RCd11 antitoxin from its own promoter was constructed using improved allelic exchange procedure (Peltier et al., 2020) [5], Plasmid p375 was constructed by Gibson Assembly to introduce vanS upstream region as left homology arm (800 bp) for recombination, RCd11 region with its own promoter and vanS downstream region as right homology arm (800 bp) for recombination. For cloning of vanS homology arms and RCd11 gene with its native promoter the plasmid pMSR185 was first linearized by reverse PCR using the primers JP1 (AAACTCCTTTTTGATAATCTCATGACC, SEQ ID NO: 19) and JP2 (AAACTTAGGGTAACAAAAAACACCG, SEQ ID NO: 20). The RCd11 gene surrounded by vanS homology arms was amplified by PCR from the plasmid DNA of p349 using the primers JP806

[0100] (CGGTGTTTTTTGTTACCCTAAGTTTCATATGGATTTTATATTCTTTTATATTGTA TG, SEQ ID NO: 21 ) and JP813

[0101] (TCATGAGATTATCAAAAAGGAGTTTGATTCTGAGAAACAGCATC, SEQ ID NO: 22) and cloned by Gibson Assembly. The plasmid p349 that served as a template for this amplification was generated as follows: the plasmid p176 was linearized by reverse PCR using the primers JP812 (TTACTTTTCTGGGAGGATGATTATATAAAGCTTAAATTTATAAATC, SEQ ID NO: 23) and IQC005 (GCTCCCTGTACAGCTGTTATTTCAATAG, SEQ ID NO: 24) to clone by Gibson assembly the RCd11 gene with its own promoter amplified by PCR from CD630Aer / 7? genome using the primers JP811 (TATAATCATCCTCCCAGAAAAGTAAAAAGCC, SEQ ID NO: 25) and IQC006 (ATAACAGCTGTACAGGGAGCTTTTTAGGACAAAAAC, SEQ ID NO: 26). The initial p176 plasmid was constructed by Gibson assembly from five fragments that include the pMSRO plasmid (Peltier et al 2020) [5] linearized by Pmel digestion; the vanS homology upstream arm amplified by PCR from C. difficile R20291 strain genome using the primers JP806

[0102] (CGGTGTTTTTTGTTACCCTAAGTTTCATATGGATTTTATATTCTTTTATATTGTA TG, SEQ ID NO: 21 ) and JP807

[0103] (TTTAATAATCCAGCTGTACAGCTGTTATTTCAATAG, SEQ ID NO: 27); the vanS homology downstream arm amplified by PCR from C. difficile R20291 strain genome using the primers JP812 (TTACTTTTCTGGGAGGATGATTATATAAAGCTTAAATTTATAAATC, SEQ ID NO: 23) and JP813

[0104] (TCATGAGATTATCAAAAAGGAGTTTGATTCTGAGAAACAGCATC, SEQ ID NO: 22); the Pcwp2 promoter sequence amplified by PCR from CD630Aem? genome using the primers JP808

[0105] (ACAGCTGTACAGCTGGATTATTAAAAAAAATTACAATTTTG, SEQ ID NO: 28) and JP809

[0106] (ATACGAATTTACCTTATTTACCAATTATAATATATATTTGATATTATTTC, SEQ ID NO: 29), and RCd11 gene sequence amplified by PCR from CD630Aer / 7? genome using the primers JP810

[0107] (ATTGGTAAATAAGGTAAATTCGTATCAAAAACAAAAAAAG, SEQ ID NO: 30) and JP811 (TATAATCATCCTCCCAGAAAAGTAAAAAGCC, SEQ ID NO: 31 ).

[0108] The p375 editing plasmid was conjugated to the strain CD126 (R20291 AphiCD027) or CD156 (630Aem? AphiCD630_2 AphiCD630_1 ) and selection of double homologous recombination events resulted in the strain CD801 (CD126 (R20291 AphiCD027) var?S::native promoter+RCd11 ) and CD802 (CD156 (630Ae / 77? AphiCD630_2 AphiCD630_1 var?S::native promoter+RCd11 ). The insertion of RCd11 antitoxin sequence with its own promoter into vanS locus on the chromosome of resulting strains was confirmed by PCR amplification with primers JP814 (CAATCCTGACTTTACTCAAGTGC, SEQ ID NO: 32) and JP815rev (GCTCAAGTTTTTCCTCTGGC, SEQ ID NO: 33) (Figure 11 A) and sequencing and the strong constitutive expression of RCd11 antitoxin from this new chromosome location was confirmed by qRT-PCR analysis using primers IQC017 (GCCTAGAGTGAAGTTTCATTAAC, SEQ ID NO: 34) and OS658 (AGTGACTTTGAGCTTGATGTG, SEQ ID NO: 35) (Figure 11 B). The antitoxin expressing CD802 strain was conjugated with p364 and p368 phagemids expressing or not the toxin and the viability of the transconjugants verified by the monitoring of bacterial growth over 24h time period (Figure 11 C). Figure 11 C shows that the antitoxin-protected strain expressing the toxin from the phagemid is able to grow, however the growth rate and the growth yield of this strain harboring phagemid that constitutively expresses the toxin are affected as compared to the control strain carrying the phagemid without the toxin gene. Despite the slower growth, this result clearly demonstrates that the protected strain is suitable for the phagemid particles production. The table 4 below summarizes the plasmid constructs used for the generation of the strain constitutively expressing the antitoxin from the chromosome and Figure 12 shows the sequence map of resulting plasmids.

[0109] Table 4

[0110]

[0111] Table 5

[0112] List of references

[0113] 1. Paparella et al., Nat Commun., 12(1 ): 6285, 2021

[0114] 2. Marcella et al., Aliment Pharmacol. Then, 53(1 ): 33-42, 2021 3. Maikova et al., Nucleic Acids Research, 46(9): 4733-4751 , 2018

[0115] 4. Tourasse et Darfeuille, RNA, 7: 1471-1481 , 2021

[0116] 5. Peltier et al., Communications Biol, 3(1 ): 718, 2020

[0117] 6. Soutourina et al., PLoS Genet., 9(5): e1003493, 2013

[0118] 7. Garneau et al., Sci. Rep., 7, 8292, 2017 8. Royer et al., Microbiol Spectr., 11 (2):e0389422, 2023.

[0119] 9. Brugiroux S. et al, Nature Microbiol, 2 : 16215, 2016.

Claims

CLAIMS1 . Recombinant phagemid for expressing at least one gene expression cassette comprising a nucleic acid sequence encoding an antibacterial peptide that inhibits the growth of a targeted strain and a nucleic acid sequence encoding a phage packaging signal for encapsidation of the recombinant phagemid into bacteriophage capsids, wherein the antibacterial peptide is the toxin component of a type I Toxin- Antitoxin (TA) system.

2. Recombinant phagemid according to claim 1 , wherein the toxin component of the type I Toxin-Antitoxin (TA) system is produced by a strain from the same species than the targeted strain.

3. Recombinant phagemid according to any of claims 1 or 2, wherein the targeted strain is a strain of C. difficile.

4. Recombinant phagemid according to any of claims 1 to 3, wherein the gene expression cassette comprises a constitutive promoter, and optionally a marker gene.

5. Method for producing recombinant phagemid particles, the method comprising : i) introducing, into a prokaryotic host cell, a recombinant phagemid as defined in any of claims 1 to 4; ii) transfer of the recombinant phagemid to a recombinant phagemid producing strain expressing the antitoxin component of the type I (T1 ) Toxin- Antitoxin (TA) system corresponding to the toxin component of the recombinant phagemid; iii) infection of the recombinant phagemid producing strain from step ii) with a helper phage comprising nucleic acid encoding bacteriophage structural proteins; iv) culturing the recombinant phagemid producing strain from step iii) under conditions which result in the double-stranded DNA of recombinant phagemidbeing packaged by the structural proteins to form the recombinant phagemid particle; and v) bacterial lysis to recover the recombinant phagemid particles from the recombinant phagemid producing strain from step iv).

6. Method according to claim 5, wherein the recombinant phagemid producing strain is a C. difficile strain.

7. Method according to any of claims 5 or 6, wherein the prokaryote host cell is an Escherichia coli strain.

8. Recombinant phagemid particle obtained by a method according to any of claims 5 to 7, wherein the recombinant phagemid particle comprises a recombinant phagemid as defined in any of claims 1 to 4.

9. Pharmaceutical composition comprising at least one recombinant phagemid particle according to claim 8, and a pharmaceutically acceptable vehicule.

10. Recombinant phagemid particle according to claim 8 or pharmaceutical composition according to claim 9, for use as a medicament.

11. Recombinant phagemid particle according to claim 8 or pharmaceutical composition according to claim 9, for use in the prevention or treatment of a bacterial infection due to C. difficile, E. coli, S. aureus, K. pneumoniae, A. baumannii, P. aeruginosa, E. faecalis, E. faecium, S. Typhi, S. enterica, S. flexneri, H. influenzae, S. pneumoniae, B. abortus, C. jejuni, H. pylori, C. botulinum, C. perfringens, L monocytogenes, N. gonorrhoea , N. meningitidis, V. choleras, preferably a C. difficile infection.

12. Recombinant phagemid producing strain constitutively expressing the antitoxin component of the type I Toxin-Antitoxin (TA) system corresponding to the toxin component of the recombinant phagemid as defined in any of claims 1 to 4.

13. Recombinant phagemid producing strain according to claim 11 , further comprising the recombinant phagemid according to any of claims 1 to 4.

14. Recombinant phagemid producing strain according to any of claims 12 or 13, wherein the recombinant phagemid producing strain is a strain of C. difficile.

15. Prokaryotic host cell comprising a recombinant phagemid according to any of claims 1 to 4.

Images