WO2005047538A2 - Ligands of the mycobacterial repressor ethr, selection methods, and applications - Google Patents

Abstract

The invention relates to methods for selecting compounds able to bind to the mycobacterial repressor EthR, said bond preventing the repressor from repressing the expression of the mycobacterial gene ethA coding EthA, the bioactivator of ethionamide. The invention also relates to the compounds inhibiting the repressor EthR, and to the therapeutic applications thereof.

Description

 MYCOBACTERIAL REPRESSOR LIGANDS EthR, SELECTION METHODS AND APPLICATIONS

The present invention relates to the field of research into new drugs for the treatment of mycobacterial infections, in particular tuberculosis and leprosy. More specifically, the invention relates to methods for selecting compounds capable of binding to the mycobacterial repressor EthR, this binding preventing it from repressing the expression of the mycobacterial gene ethA encoding an activator of ethionamide (ETH). The present invention also relates to the compounds thus selected, as well as to their applications for the prophylaxis and / or the treatment of mycobacterial infections. It is currently estimated that approximately 7 to 8 million people contract tuberculosis each year. About 50 million people are currently infected with strains of

Mycobactθ um tuberculosis multi-resistant to antibiotics. Such strains are in particular resistant to both isoniazid (INH) and rifampicin (RIF), the two first-line antibiotics (ie, including the therapeutic index, representing the toxic dose / therapeutic dose ratio-high ) -used-for-tuberculosis (Espinal e ^' a / :,

2001). These patients are therefore subjected to treatment using second-line compounds (ie, whose therapeutic index is lower), less active, more expensive and or more toxic (ahmoudi and Iseman, 1993). The second-line antibiotic ethionamide (2-ethylthioisonicotinamide, ETH, Trescatyl) has been used clinically for more than 35 years. It is associated with significant gastrointestinal disorders and hepatotoxicity. Its prescription is therefore mainly limited to the treatment of patients who relapse due to strains of M. tuberculosis resistant to first-line antibiotics, or to treatment of patients with lepromatous leprosy, replacing clofazimine (Jenner and Smith, 1987). ETH is a structural analogue of INH. ETH and INH are prodrugs which must be converted into active compounds by specific mycobacterial enzymes (Baulard et a, 2000). These two antibiotics inhibit the biosynthesis of mycolic acids, essential constituents of the mycobacterial wall. They both target the enzyme enoyl-AcpM reductase (InhA). However, the mechanism of action of these two antibiotics follow different metabolic pathways: INH is activated by catalase peroxidase KatG, while activation of ETH is dependent on the enzyme EthA. Consequently 95% of the strains resistant to INH remain sensitive to ETH. Another major difference between these two antibiotics lies in their respective minimum inhibitory concentration (MIC). In order to ensure ETH concentrations in body fluids in excess of the MIC in vitro (2.5 μg / ml), it is recommended to use daily dosages of 10 mg / kg (Johnston et al. , 1967), 5 times more than the recommended dosage for INH. Recently, the ethA and ethR genes involved in the activation of the ETH prodrug have been identified (Baulard et al. 2000, DeBarber et a, 2000). EthA is an FAD-containing enzyme that catalyzes the activation of ETH in two stages (Vannelli et al. 2002). The expression of ethA is placed under the control of the adjacent ethR gene, insofar as the overexpression of ethR leads to resistance to ETH and the chromosomal inactivation of ethR is associated with hypersensitivity to ETH, thereby revealing that not all ETH molecules are activated in vitro in wild-type mycobacteria. The present invention aims to provide means which allow more efficient activation of ETH by mycobacteria. So, thanks to means according to the invention, it is now possible to optimize the therapeutic efficacy of ETH by reducing the dosages to be administered to patients and, consequently, by reducing the side effects linked to the administration of this antibiotic according to current protocols. . According to a first aspect, the present invention provides methods of selecting compounds which act as ligands for the repressor EthR, and which, when linked to said repressor, prevent the latter from repressing the expression of the ethA gene. follows, these compounds will also be called "EthR repressor inhibitors". In the context of the invention, by “selection of a compound” is meant the fact that one chooses to retain and use this compound because it has the property of interest sought: an inhibitor of the EthR repressor, in other words in the present case, as indicated below, it has the capacity, by attaching to the EthR repressor, to prevent the latter from repressing the expression of the ethA gene. The formulas “compound which acts as ligand for the repressor EthR” or “ligand for the repressor EthR” are equivalent and denote a “compound capable of binding to the repressor EthR”. Insofar as the compound in question blocks the activity of the EthR repressor by preventing it from repressing the expression of the ethA gene, this compound will also, and in an equivalent manner to the preceding formulas, be called “inhibitor of the EthR repressor” or “ inhibitor of EthR ”. In the present context, a "compound" can be any type of molecule, biological or chemical, natural, recombinant or synthetic. For example, a compound can be a nucleic acid (eg, a sense or antisense oligonucleotide such as an antisense RNA), a protein, a fatty acid, an antibody, a polysaccharide, a steroid, a purine, a pyrimidine, a molecule organic, chemical radical, etc. The term "compound" also covers fragments, derivatives, structural analogs or combinations thereof, as long as they have the ability, by attaching to the repressor EthR, to prevent the latter from repressing the expression of the ethA gene. This capacity represents the property of interest which must be presented by the compounds selected by the implementation of the methods which are the subject of the present invention. Here we mean by "repress", the fact of reducing, reducing, decreasing, negatively regulating, inhibiting, blocking, preventing, preventing. Conversely, one must understand by “activating”, the fact of increasing, increasing, regulating positively, promoting, improving, stimulating. According to a first embodiment, a selection method as targeted by the present invention comprises at least (method A): a) cloning of the gene encoding said EthR repressor on a first recombinant expression plasmid, and cloning of at least the promoter and operator region of the ethA gene upstream of a reporter gene on a second recombinant expression plasmid; b) transformation of a recombinant host cell with at least the plasmids according to step a); c) culturing said recombinant host cell in the presence of a candidate compound; d) measuring the level Ni of expression of the reporter gene in the presence of said candidate compound; e) comparing the level Ni measured in step d) with the level of expression of the reporter gene in the absence of said candidate compound; and f) if Ni is significantly greater than No, the selection of the compound. A second embodiment of the selection method according to the invention corresponds to the method described above, in which, however, not two recombinant expression plasmids are used (steps a) and b)), but only one, on which are cloned both the gene encoding the repressor EthR and at least the promoter and operator region of the gene ethA upstream of a reporter gene. Thus, such a method comprises at least (method B): a) the cloning of the gene coding for said EthR repressor, and the cloning of at least the promoter and operator region of the ethA gene upstream of a reporter gene, on a plasmid recombinant expression; b) transformation of a recombinant host cell with at least the plasmid from step a); c) culturing said recombinant host cell in the presence of a candidate compound; d) measuring the level Ni of expression of the reporter gene in the presence of said candidate compound; e) comparing the level Ni measured in step d) with the level N ₀ of expression of the reporter gene in the absence of said candidate compound; and f) if Ni is significantly greater than No, the selection of the compound. For the implementation of the steps mentioned above (cloning, transformation, culture, measurement of the expression of a reporter gene), the skilled person will use conventional techniques (see for example Sambrook and Russel, 2001 ). The promoter and operator region of the ethA gene to be cloned for the purposes of implementing a method according to the invention comprises at least the sequence SEQ ID No. 1, which corresponds to the promoter and operator region of the ethA gene on which fix 4 repressor molecules (see experimental part below). Preferably, this region comprises at least the sequence SEQ ID No. 2, to which 8 molecules of EthR are attached. More preferably, this region comprises at least the sequence SEQ ID No. 3, which corresponds to the intergenic sequence ethA-R. According to a particular embodiment, a recombinant expression plasmid containing the ethR gene (coding for EthR) is pMV261-ef /? ^: ? (see experimental part below). According to another particular embodiment, a recombinant expression plasmid containing at least the promoter and operator region of the ethA gene upstream of a reporter gene is pBSh-P _ethA - lacZ (see experimental part below). In the plasmid pBSh-P _ethA - lacZ, the reporter gene is lacZ.

Obviously, a person skilled in the art can use any other reporter gene known in the field of the invention to be usable in mycobacteria, in particular the xylE, gfp, lux genes and the gene coding for the protein glucuronidase. . In the methods according to the invention, said recombinant host cell is a mycobacterial cell. It is preferably a Mycobacterium bovis, M. smegmatis, M. tuberculosis or M. leprae cell. In particular, a person skilled in the art can choose a cell of M. bovis BCG 1173P2 (vaccine strain accessible on simple request from the World Health Organization), or of M. smegmatis mc ² 155 (Snapper et a, 1988 ; see experimental section below). According to a third embodiment, a selection method in accordance with the present invention comprises at least (method C): a) determining the level of binding of the EthR repressor on at least the promoter and operator region of the ethA gene, in the presence a candidate compound; b) comparing the Li level with the Lo level of binding of the EthR repressor on at least the promoter and operator region of the ethA gene, in the absence of said candidate compound; and c) if Li is significantly less than L ₀ , the selection of the compound. In particular, the binding levels L ₀ and U can be determined using a method chosen from surface plasmon resonance, detection of natural fluorescence and energy transfer by fluorescence resonance (FRET for

"Fluorescence resonance energy transfer") by which the human profession will determine the energy transfer between EthR and the promoter and operator region of ethA, in the presence and absence of a ligand (Jia et al., 1996; Pyles et al., 1998; Hillisch et al., 2001; Engohang - Ndong ef a /., 2004). According to a fourth embodiment, a selection process targeted by the present invention comprises at least (method D): a) bringing together the repressor EthR and at least two candidate compounds; b) crystallization of the mixture thus obtained; c) analysis of the structure of the crystal resulting from step b); d) if this analysis reveals that one of said compounds is linked to said EthR repressor, the solubilization of said crystal; e) elution and selection of said compound. The steps d) of solubilization of the crystal and e) of elution and selection of the compound call for conventional methods (see for example the experimental part below), well known in the technical field of the invention. According to a fifth embodiment, a selection method which is the subject of the present invention comprises at least (method E): a) bringing together the repressor EthR and a candidate compound; b) crystallization of the mixture thus obtained; c) analysis of the structure of the crystal resulting from step b); d) if this analysis reveals that said compound is linked to said EthR repressor, the selection of said compound (from the starting candidate compound). In the two preceding embodiments (methods D and

E), the stages of crystallization (stage b)), of analysis of the structure of the crystal (stage c)) and, where appropriate, that of the selected compound (as indicated below, it can be determined, in the context an additional subsequent step, the structure of the selected compound) can be carried out, for example using the techniques described in the examples below, or by any other conventional method known to those skilled in the art. It will be noted that at the end of the selection methods according to the invention described above (methods A, B, C, D and E), a step of determining the structure of the selected compound can be advantageously implemented. In addition, the selection methods according to the invention described above (methods A, B, C, D and E) can be implemented alone or in combination. Thus, for example, a person skilled in the art can choose a selection method according to one or other of the five above-mentioned embodiments. Preferably, one of the first three embodiments above (method A, B or C) will then be implemented. Alternatively, a person skilled in the art may first implement the fourth or fifth embodiment (method D or E) in order to select a compound, then any one of the first three embodiments (method A, B or C) in order to confirm that the compound thus selected exhibits the capacity, by binding to the repressor EthR, to prevent the latter from repressing the expression of the ethA gene. Thus, preferably, the method according to the fourth or fifth embodiment (method D or E) will be implemented prior to a method according to any one of the first three embodiments (method A, B or C). A second aspect of the present invention relates to a compound, as defined above, inhibitor of the EthR repressor, capable of being obtained by the implementation of a selection process according to the invention. Alternatively, a compound targeted by the invention is capable of being obtained by the successive implementation of several, of preferably two, different selection methods from those described above. According to one embodiment, a compound targeted by the present invention interacts with residues F110 and F114 of the repressor EthR (see experimental part below, paragraph ll-E). Preferably, such a compound interacts not only with residues F110 and F114, but also with residues W138 and W145 of EthR. According to another embodiment, a compound object of the invention is fixed in the binding pocket of the ligand of the EthR repressor, which is defined by the following EthR residues: R128, V134, G124, L76, Q125, F187, F184 , T121, W138, F114, L183, E180, M142, N179, N176, W145, F110, W207, T149, 1107, L87, V152, Y148, G106, W103 and M102 (see experimental section below, paragraph ll-E ). Note that hexadecyl octanoate (Garcia-Rubio et al.;

2002), as well as the compounds described by Fraaije et al. (2004) are already known as such. The known compounds are excluded from the compounds which are, as such, the subject of the present invention. According to a third aspect, the present invention provides means useful in the field of prevention and / or treatment of diseases, in particular bacterial infections and, preferably, mycobacterial infections such as tuberculosis and leprosy. Thus, the invention relates to a compound which inhibits the repressor EthR as a medicament, intended in particular for preventing and / or treating bacterial infections, preferably mycobacterial infections. In particular, such a compound can be obtained by at least one selection process described above. Indeed, since the knockout mutant ethR described in

Baulard et al. (2000) has a phenotype of significant growth slowdown, EthR alone constitutes a therapeutic target. An EthR ligand is likely to give rise to a comparable phenotype to that of the aforementioned mutant. Anyway, an EthR ligand can have advantageously bacteriostatic, even bactericidal, properties against mycobacteria. For example, in addition to the compounds capable of being obtained by at least one process according to the invention, the compound used as a medicament can be an analog or a derivative of the co-crystallized ligand described in the experimental part below (octanoate of hexadecyl). It can also be an analogue or a derivative of benzylacetone (see experimental part below). For the purposes of the present invention, an “analog” is any modified, pharmaceutically acceptable (in particular non-toxic) version of an initial compound, said modified version possibly being natural or synthetic, in which one or more atoms, such that atoms of carbon, hydrogen, oxygen, or heteroatoms such as nitrogen, sulfur or a halogen, have been added or removed from the structure of the initial compound, so as to obtain a new molecular compound pharmaceutically acceptable. The term “derivative” here designates any pharmaceutically acceptable compound which has an affinity for the repressor EthR, a resemblance or a structural motif, in common with a reference compound. This definition also includes, on the one hand, the compounds which, alone or with other compounds, can be precursors or intermediate products in the synthesis of the reference compound, by means of one or more chemical reactions, and on the other hand compounds which can be formed from the reference compound, alone or with other compounds, via one or more chemical reactions. In particular, ketones and esters obtained from a reference compound are included in the definition of the term "derivative". The expression “pharmaceutically acceptable” designates here what is suitable and useful for preparing a pharmaceutical composition or a drug, which is generally devoid of toxicity, especially in humans. Thus, at least some of the pharmaceutically acceptable compounds described by Fraaije et al. Are covered by the definitions of "analogs" and "derivatives" above. as either substrates or products of the mycobacterial activator EthA (see Table I in Fraaije et al. 2004; and see paragraph III-B of the experimental part below), and / or their analogs and / or pharmaceutically acceptable derivatives. Thus, according to a particular embodiment, a compound useful as a medicament targeted by the invention is chosen from analogs and pharmaceutically acceptable inhibitors of EthR and derivatives of 2-hexanone, 2-heptanone, 4-heptanone, 2 -octanone, 3-octanone, 2-decanone, 2-dodecanone (Fraaije et al. 2004). The invention also relates to a pharmaceutical composition comprising at least, as active principle, a compound which inhibits the repressor EthR and which is useful as a medicament (that is to say in particular pharmaceutically acceptable), as described above, in combination with the at least one pharmaceutically acceptable excipient. Those skilled in the art will choose one or more pharmaceutically acceptable excipients depending on the route of administration of the pharmaceutical composition. In addition, the form of the drug or pharmaceutical composition (eg, solution, suspension, emulsion, tablets, capsules, etc.) will depend on the route of administration chosen. Thus, within the meaning of the invention, a medicament or a pharmaceutical composition can be administered by any suitable route, for example the oral, local, systemic, intravenous, intramuscular or mucosal route, or alternatively by using a patch, or again in a form encapsulated in, or immobilized on, liposomes, microparticles, microcapsules, and the like. Mention may in particular be made, by way of nonlimiting examples of excipients suitable for oral administration, talc, lactose, starch and its derivatives, cellulose and its derivatives, polyethylene glycols, polymers d acrylic acid, gelatin, magnesium stearate, animal, vegetable or synthetic fats, paraffin derivatives, glycols, stabilizers, preservatives, wetting agents, dispersants, emulsifiers, penetration agents, solubilization, etc. The techniques for formulating and administering drugs and pharmaceutical compositions are well known in the art considered here (a person skilled in the art may in particular refer to the work "Remington's Pharmaceutical Sciences", latest edition) . The present invention also relates to the use of at least one compound inhibiting the EthR repressor and useful as a medicament (that is to say in particular pharmaceutically acceptable), as described above, for the manufacture of a medicament intended prevention and / or treatment of bacterial infections, preferably mycobacterials. In addition, the invention relates to products containing a compound which inhibits the repressor EthR and is useful as a medicament (that is to say in particular pharmaceutically acceptable), in accordance with the invention, and either ETH or one of its derivatives, an antibiotic activatable by the mycobacterial activator EthA, as a combination product for simultaneous, separate or spread over time in anti-tuberculosis or anti-leprosy therapy. The following figures are provided for illustrative purposes only and do not limit the present invention in any way. Figure 1: β-galactosidase activity in transformed mycobacteria. A: detection of β-galactosidase activity in M. smegmatis transformed with (a) pBSh-P _ethA - / acZ, (b) pBSh-P _ethA - / acZ + pMV261-e7? R, (c) pMV261-ef ? R, on agar plates containing LB medium enriched in X-Gal. B: quantification of β-galactosidase activity in M. bovis BCG transformed with (a) pBSh-P _ethA - / acZ, (b) pBSh-P _ethA - / acZ + pMV261-e /? F ?, (c) pMV261-ef? Each value is an average made over 3 replicas of cultures. Figure 2: EMSA of Hise-EthR binding to the ethA-R intergenic region

The labeled probe of 191 bp (100,000 cpm) was mixed with 0 μg (lane 1), 0.45 μg (lane 2), 0.90 μg (lane 3), 1, 35 μg (lane 4), 1, 8 μg (lane 5) and 3.6 μg (lane 6) His ₆ -EthR in the absence (A) or in the presence of an excess of the unmarked ethA-R intergenic region used as specific competitor (B) or still in the presence of an excess of non-specific competitor DNA (C).

B and C: EMSAs were carried out without (lane 1) or with 0.9 μg (other lanes) purified Hise-EthR. Figure 3: A and B: Dnase I protection of the ethA-R intergenic region by His ₆ - EthR.

The 2 probes (A: Hind \\\ - Sph \ fragment of 323 bp; B: Xho \ - HindlW fragment of 312 bp) were incubated without (lane 1) or with different amounts of His ₆ -EthR (A: lane 2: 0.0625 μg; lane 3: 0.125 μg; lane 4: 0.25 μg; lane 5: 0.5 μg; lane 6: 1 μg; B: lane 2: 1 μg; lane 3: 2.5 μg ; lane 4: 5 μg; lane 5: 7.5 μg; lane 6: 10 μg), before treatment with DNase I.

C: localization of the region protected by Hise-EthR against digestion by DNase I. D: analysis of primer extension. The coding strand sequence is shown and the +1 transcription point is underlined.

Figure 4: overall affinity and stoichiometry of the interaction between His ₆ -EthR and Pet A measured by SPR. A: structure of the ethA-R intergenic region.

IG-37 includes 2 imperfect direct repetitions (solid arrows) which include 2 imperfect indirect repetitions (bases in italics).

IG-36 includes 2 imperfect indirect repetitions (dotted arrows).

B: injection sensorgrams of 0.35; 0.70; 1, 39; 3.47 and 8.6 μM Hisβ-EthR, using the IG-106 fragment.

C: binding stoichiometry as a function of the Hise-EthR concentration.

Sensorgrams of 0.612; 1.25; 2.50 and 5 μM His ₆ -EthR, using the IG-62 (D), IG-37 (E) and IG-36 (F) fragments.

Figure 5: Hill representation for the binding of His ₆ -EthR on the ethA-R intergenic region.

Figure 6: gel filtration chromatography.

Elution profiles of a PCR fragment (0.108 nmole) overlapping the ethA-R intergenic region in the absence (A) or in the presence of 0.385 nmole

(B), 1.920 nmole (C) and 3.850 nmole (D) of purified His ₆ -EthR. E: 3.850 nmole elution profile of His ₆ -EthR in the absence of DNA.

Y-axis: relative absorbances (mAU for milliabsorbance unit) to

260 nm (dotted line) and 280 nm (solid line).

Figure 7: in vivo interaction between EthR molecules.

Swell boxes containing minimal medium, enriched with maltose and X-Gal. Figures 8 and 9: structure of the ligand linked to the homodimer of EthR, according to a respectively perpendicular view [Figure 8: structure of the crystal of the EthR repressor of the space group P2- | 2-ι2 (form 1)] or along (Figure

9) of the double axis of the dimer.

EthR consists of 9 helices: α1 (23-38), α2 (47-54), α3 (56-64), α4 (68-92), α5 (99-115), α6 (118-129), α7 (132-158), αδ (168-188) and α9

(192-212). The HeOc molecules occupy space across the entire ligand binding domain. The 2 recognition helices α3 and α3 'are separated by 52 A.

Figure 10: superposition of the central ligand binding region (helices α4 to α9) of EthR and that of the repressor QacR, equivalent from a topological point of view, in the induced form.

Residues 89 to 93 of QacR undergo, as illustrated, a turn-helix transition (TTH) in the α5 helix following binding of the ligand, this transition not being observed in the structure of the unbound repressor. This transition is also present in EthR. The HTH motifs, although relatively well conserved from a structural point of view between the two repressors (RMSD of 1 A for these isolated regions) do not overlap due to the different relative orientations of their respective ligand binding domains. Figure 11: recognition of the ligand by EthR.

Only the residues whose atoms lie in a 5A sphere around the ligand are indicated.

A water molecule is buried via a hydrogen bridge formed with the residue N179. The recognition of the ligand involves residues of the 6 helices which form the binding domain of the ligand.

Figure 12: superimposition of the central ligand binding region (helices α4 to α9) of EthR in complex with hexadecyl octanoate and that of QacR in complex with dequalinium. F110 apparently plays an important role in the induction of EthR, by analogy with that played by the Y92 region of QacR. The structure of QacR highlights the binding pocket comprising several sites. Although the two repressors have the same ligand entry portal and have partially overlapping adjacent binding sites, they are particularly divergent with respect to the overall binding cavities. The EthR ligand is deeply buried, primarily in the central hydrophobic region of the ligand binding domain, in a cavity which is located parallel to the axes of the propellers α4 and α7. On the other hand, the pocket connecting to several QacR sites is practically parallel to α6. The displacement of α5 in EthR compared to that of QacR closes this cavity in EthR. Figure 13: electron density map, performed at a level of 1.5 σ, calculated by excluding hexadecyl octanoate (HeOc) from the model. Figure 14: Identification of the EthR ligand by GC-MS after organic extraction. Chromatogram (retention time 21.00 min to 27.00 min) of the organic fraction of the crystals of EthR (A). Electron ionization spectrum (El) of the GC peak at 22.16 min (B). Using as input the masses measured from the GC-MS spectrum to screen an internal GC-MS library, the ligand was identified as a hexadecyl octanoate. EI reference spectrum of hexadecyl octanoate (C). It has a molecular peak of M + ions at m / z = 368 and characteristic ion peaks at m / z = 43, 57, 111, 127, 145, 269. The GC peak at 24.17 min corresponds to the octanoate hexadecyl. The GC peak at 25.99 min corresponds to non-specific contaminants (phthalates).

Figure 15: Synergistic effect of benzylacetone and ETH on the growth of M. smegmatis. The synergistic effect was tested by depositing 4 μL of dilutions in series (A: 10 ^'2 ; B: 10 ^"3 ; C: 10 ^-4 ; D: 10 ^{' 5} ) of a culture in the growth phase of M. smegmatis me ² 155 on dishes of LB medium not containing antibiotics, or containing 5 μg / mL of ETH alone (ETH), 1 mM benzylacetone alone (BA), or a combination of the two compounds (ETH + BA).

The experimental part below, supported by examples and references to the figures, aims to illustrate, without limiting, the present invention. EXPERIMENTAL PART

I- EthR suppresses the expression of ethA by binding directly to the promoter and operator region of ethA:

A. Materials and methods:

A.1. Bacterial strains and growing conditions

All strains were grown aerobically at 37 ° C. All the cloning steps were carried out in E. coli XL1-Blue (Stratagene).

The E. coli strains were cultivated in Luria-Bertani (LB) medium (Life

Technologies) with or without 1.5% agar (Difco).

The mycobacterial strains were cultivated in Sauton medium (Sauton

1912) enriched with 0.001% ZnS0 ₄ and 0.25% Triton WR1339 (Sigma) or on M7H11 agar enriched with 10% of catalase-dextrose-albumin-oleic acid mixture (Becton Dickinson). Mycobacterium smegmatis and

M. bovis BCG 1173P2 were transformed as described above

(Wards and Collins, 1996).

Large-scale cultures of mycobacteria were brought to the middle of the exponential phase (36 h for M. smegmatis mc ² 155; 11 to 16 days for M. bovis BCG).

If necessary, antibiotics (Sigma) were added to the medium at the following concentrations: kanamycin at 20 μg / ml, ampicillin at 100 μg / ml, hygromycin at 50 μg / ml, streptomycin at 20 μg / ml and gentamicin at 15 μg / ml.

After growth, the bacteria were harvested, washed with PBS phosphate buffer and stored at -20 ° C.

A.2. Plasmids and DNA manipulations All DNA manipulations have. been performed using conventional protocols, as described by Sambrook and Russell (2001). The “Fast-link ™ DNA ligation” system (Epicenter Technologies) was used for the ligations. a) Construction of pBSh -lacZ The Bam λ \ -Bgl \\ fragment of 3,678 kb, containing the open reading frame (ORF) coding for β-galactosidase, was isolated from pQEGM (Antoine et al., 2000) and inserted in pBSh-D (Picardeau et al., 2000) before being digested with SamHI. b) Construction of oBSH-PethA-lacZ

The promoter region P _{and A} was amplified by PCR using the primers 0-161: 5'-CCAGCGGACGGTCCTCTAGAAGGTTC-3 '(SEQ ID No. 4), 0-162: 5'-GTGAGATCTCATGGATCCACGCTATCAAGGTAA-3' (SEQ ID No. 5). The PCR product was inserted into pCR2.1Topo (Invitrogen) to give pCR2.P _and hA (2 + 4). The plasmid was redigested with Xba \ and BglII and the 191 bp fragment containing P _ethA was inserted into pBSh- lacZ, previously digested with Xba \ and BamHI to give pBSH-P _e tha- lacZ.

c) Plasmids for the double hybrid bacterial test

The region encoding EthR was amplified by PCR using the primers O-243: 5'-CTGCAGTGACCACCTCCGCGGCCAG-3 '(SEQ ID N ° 6) and O- 244: 5'-GGTACCTTATCTGTTCTCGCCGTAA-3' (SEQ ID N ° 7) . The amplified sequence was cleaved by Pst \ and / 4sp718, and inserted into pT25 digested with Psfl-, 4sp718 (Karimova et al., 1998), to give pT25-EthR. Alternatively, the same chromosomal region was amplified by PCR using the primers O-240: 5'-

GGTACCGACTACTTCCGCAGCCAGTCA-3 '(SEQ ID N ° 8) and 0-242: 5'- AAGCTTTCGCCGTAAATGCTGGTCA-3' (SEQ ID N ° 9), cleaved by Hind \\\ and Asp718, and inserted in pT18 (Karimova et a, 1998 ) previously digested with the same enzymes to give pT18-EthR. A strain of E. coli cyclase-deficient (DHP1) (Karimova et al., 1998) was co-transformed with these two plasmids. The capacity of the resulting clones to ferment the maltose was determined at 30 ° C. on boxes of freshly prepared solid medium M9 medium, containing 1% maltose as the sole carbon source and 40 μg / ml of 5-bromo-4-chloro -3- indolyl-β-D-galactopyranoside (X-gal, Promega).

A3. Beta-galactosidase tests

The β-galactosidase activities of the recombinant strains of M. smegmatis and M. bovis BCG were detected on M7H11 agar dishes enriched with 100 μg / ml of X-gal. Β-galactosidase activities have been quantified in liquid cultures, using a method adapted from Miller (1972). Briefly, the mycobacterial cells were cultured in Sauton medium (50 ml) to an optical density (OD) at 600 nm of 0.6. The cells were then harvested by centrifugation (4000 g, 15 min, 4 ° C), and the pellets were washed, then resuspended in 1 ml of PBS. After sonication of the bacterial suspensions, aliquots of the lysates were incubated with orthonitrophenyl-β-D-galactoside (ONPG, Sigma) at 30 ° C. The enzymatic reaction was followed by spectrophotometry at an absorbance of 420 nm (A ₂ o _n m), and the initial speed was calculated using the linear proportion of the kinetic curve. The β-galactosidase Specific Activity (AS) was determined as follows: AS = (ΔA ₂ onm min ^"1 ) / (DOβoonm I of culture). Triplicate samples were measured for each bacterial clone. assess the potential influence of ETH on the activity of the EthR repressor, M. smegmatis co-transformed with pBSh-P _ethA - / acZ and pM261-ef / 7f? was cultured up to an ODβoonm of 0.3 .

Variable concentrations of ETH (0-10-20-40 μg / ml) were then added to the culture (in duplicate), and the β-galactosidase activities were measured every 10 min for one hour. A.4. Production and purification of His ₆ -EthR The DNA encoding EthR was amplified by PCR using the chromosomal DNA of M. tuberculosis H37Rv as template and the 5. oligonucleotides 0-183: 5'-CATATGACCACCTCCGCGGCCAGT-3 '(SEQ ID No. 10) and 0-184: 5'-GGATCCGAGCACCCCCGACCGAGT-3 '(SEQ ID No. 11) as primers. The PCR product was inserted into pCR2.1 Topo so as to form pCR2Λ-ethR. The sequence of the fragment encoding EthR was determined on the two strands and, then, isolated from pCR2.1-0 ethR by digestion with Nde \ and BamHI, then inserted into pET-15b (Novagen), to give pET-15b -e # 7R. This plasmid codes for an EthR protein carrying an amino terminal tag having the following sequence: MGSSH ₆ SSGLVPRGSHM (SEQ ID No. 12). This plasmid was introduced into E. coli BL21 (DE3). E. coli BL21 (DE3) (pET-15b-etf? R) was cultured in LB medium (120 ml) to an OD of 0.6-0.7. Isopropylthiogalactoside (IPTG) was then added to a final concentration of 1 mM, and the culture was held for an additional 3 hours. The cells were harvested by centrifugation at 12,000 g at 4 ° C, resuspended in 10 ml of lysis buffer (50 mM NaH ₂ P04.0 300 mM NaCl, pH 7.5, 10 mM imidazole) and lysed by two passages through a French press at 6.2 × 10 ⁶ Pa. After centrifugation (20,000 g, 25 min, 4 ° C.), the supernatant was recovered, and His ₆ -EthR was separated from the lysate of total cells by chromatography on a Ni-NTA agarose column (Qiagen). After extensive washing with lysis buffer, Hise-EthR was eluted from the resin by 100 mM imidazole in lysis buffer, then dialyzed overnight against lysis buffer. About 900 μg / ml of purified proteins were obtained as determined using the “Bio-Rad protein assay” system. The purity of the proteins was determined by staining with Coomassie Blue after SDS-PAGE electrophoresis on a 12% polyacrylamide gel. The purified protein was stored in 10% glycerol at -20 ° C, then dialyzed before use.

AT 5. Electrophoretic mobility interference measurement (EMSA for “Electrophoretic mobility shift assays”)

The 191 bp PCR fragment obtained using oligonucleotides 0-161 and 0-162 and containing the promoter region of ethA was purified by agarose gel electrophoresis and labeled with 10 μCi of [γ- ³² P] ATP (3000 Ci / mmol; Amersham Biosciences) using T4 polynucleotide kinase (10 units, Invitrogen). The radiolabelled fragment ([ ³² P] -P _e thA) was purified using a MicroSpin G-25 column (Amersham, Biosciences). [ ³² P] -P _eth A (100,000 counts per minute; cpm) was incubated for 10 min at room temperature with variable amounts of HiSβ-EthR in a reaction volume of 20 μl containing 1 μg of poly- (dl- dC) (Amersham Biosciences) and 2 mM Tris-HCI (pH 8.0), 0.4 mM MgCI ₂ , 5 mM KCI, 0.2 mM DTT, 10% glycerol and 0.01% non-ionic detergent P40 (Igepal CA-630, Sigma).

After incubation, the mixture was loaded onto a 10% non-denaturing polyacrylamide gel and subjected to electrophoresis. The gel was then dried in vacuo and exposed overnight to Biomax X-ray film (Kodak).

The competition tests were carried out as described (Tang et al., 2001). Unmarked P _{and A} and the unmarked pHP45Ω streptomycin resistance cassette (Frey and Krisch, 1985) (in molar excess of 0 to 1000 times) were pre-incubated as specific and non-specific competitor DNA with Hise, respectively. -EthR for 10 min at room temperature, followed by the addition of labeled probes and incubation for 10 min at room temperature. The resulting protein-DNA complexes were subjected to electrophoresis and auto-radiography as described above. A.6. Gel filtration

A 94 bp DNA fragment overlapping P _ethA was amplified by PCR using the chromosomal DNA of M. tuberculosis H37Rv as template and O-270: 5'-CGGTCATGGATCCACGCTATCAAC-3 '(SEQ ID NO: 13) and O - 272: 5'-GGAGGTGGTCACCCTGGCAGC-3 '(SEQ ID No. 14) as primers. Varying amounts of His ₆ -EthR were mixed with 6.7 μg (0.108 mmol) of PCR products in buffer A (50 mM NaH ₂ P0 ₄ , pH 7.5; 250 mM NaCl) and filtered through a Superdex column 200 HR 10/30 connected to an AKTA FPLC system (Amersham Biosciences) at a flow rate of 0.3 ml per minute. Relative protein and DNA concentrations in the eluent were monitored by measuring absorbance at 280 nm and 260 nm. The apparent molecular mass of Hise-EthR in the absence of DNA was estimated using the standard curve V _e / V ₀ against the log of the molecular masses of the standard proteins (ribonuclease A, 13.7 kDa chymotrypsinogen A, 25 kDa ; ovalbumin, 45 kDa; albumin, 66 kDa aldolase, 158 kDa; catalase, 232 kDa; apoferritin, 440 kDa thyroglobulin, 669 kDa; molecular size standards for protein gel filtration; Amersham Biosciences). Vo is the dead volume and V _e is the elution volume.

A7. Surface plasmon resonance (SPR) binding test The BIAcore 2000 'device (BIAcore'AB, Uppsala, Sweden) was used for real-time analyzes of the molecular interactions between Hise-EthR and the promoter region of ethA. 4 DNA fragments were obtained by PCR using the chromosomal DNA of M. tuberculosis H37Rv as a template. A 106 bp fragment (Figure 4A; IG-106) overlapping the ethA-R intergenic region was obtained using O-270: 5'-CGGTCATGGATCCACGCTATCAAC- 3 '(SEQ ID NO: 13) and 0-271: 5' -biotin-CTGACTGGCCGCGGAGGTGGT- 3 '(SEQ ID No. 15). A 62 bp DNA fragment (IG-62) spanning a 55 bp region identified by DNase I fingerprinting experiments was amplified using 0-248: 5'-biotin-GATCCACGCTATCAACGTA-3 '(SEQ ID No. 16) and O-250: 5'-CTACGTGTCGATAGTGTCG-3 '(SEQ ID No. 17). A 37 bp fragment overlapping 2 direct repeats was amplified using 0-255: 5'-ATCAACGTAATGTCGAG-3 '(SEQ ID No 18) and 0-256: 5'-GTCGACATCTCGTTGACG-3' (SEQ ID No 19 ), then clone into pCR2.1Topo. A 104 bp biotinylated DNA fragment (IG-37) was then obtained by PCR using 0-255 and 5'-biotin-AATTGGGCCCTCTAGATGCAT-3 '(SEQ ID No. 20).

Finally, a 36 bp fragment overlapping 2 reverse repeats was amplified using 0-253: 5'-GTAATGTCGAGGCCGTCAA-3 '(SEQ ID No. 21) and 0-254: 5'-ATAGTGTCGACATCTCGTT-3' (SED ID N ° 22), then clone into pCR2.1Topo. A 103 bp biotinylated DNA fragment (IG-36) was then obtained by PCR using 0-253 and 5'-biotin-AATTGGGCCCTCTAGATGCAT-3 '(SEQ ID No. 23).

The amplified biotinylated fragments were purified on a Quiaquick column (Qiagen), then their sequence was determined before immobilization on a cell coupled to streptavidin CM5 Sensor Chip using the standard protocol provided with the "Amine Coupling" system (BIAcore '). Briefly, streptavidin was injected at 500 ng / μl in 10 mM sodium acetate (pH 3.5) for 12 minutes at a rate of 10 μl / min. Biotinylated DNA (approximately 69 kDa for the 106 bp fragment) was injected through each cell at 200 ng / ml to obtain a stable fixation of streptavidin of 205 resonance units (RU). The integrity and the quantity of DNA attached were checked by injection for 120 seconds of 1.6 μM of calf H1 histone (Sigma) at 10 μl / min. The binding of His ₆ -EthR to DNA was carried out at 25 ° C in .0.05 M NaH ₂ P0 ₄ (pH 7.15), 0.25 M NaCl, 0.002 M MgCl ₂ , 0.05 M KCI and 0.001 M DTT at a flow rate of 20 μl / min. His6-EthR was injected at the desired concentrations in circulation buffer until equilibrium was reached.

The Sensor Chip cells were regenerated by a 60-second injection of 0.03% SDS. The final curves were obtained by subtraction of the signal corresponding to a cell functionalized with an irrelevant double-stranded DNA fragment of 100 bp (fragment -276 to - 176 of the promoter of human stromelyin-1). The molar ratio between Hise-EthR (25.9 kDa) and the immobilized DNA were determined according to the following equation 1: n = (α RU Hise-EthR / β RUADN) X (Mr _A DN / Mr _H is6- EthR) in which n is the stoichiometry of the complex, RU is the measured response (in units) obtained at binding saturation, and Mr is the molecular weight, α = 1 pg of proteins bound by RU / mm ² and β = 0 , 73 pg of DNA bound by RU / mm ² . These factors come from the difference in molar refractive index between DNA and proteins, which results in a different plasmonic behavior (PharmaciaBiosensor, 1994; Speck et al., 1999). In order to determine the Hill coefficient and the overall Ko of the interaction, a second Sensor Chip cell containing the same biotinylated PCR product was created so as to obtain a DNA fixation on streptavidin equivalent to 100 RU. His ₆ -EthR was injected as described previously according to increasing concentrations ranging from 100 nM to 2 μM so as to obtain the binding equilibrium. For each concentration, the equilibrium signal, noted R _eq , was measured. The fractional saturation, θ = R _eq / Rm _a x, was calculated. R _max represents the maximum binding capacity of Hise-EthR estimated at 225 RU in this experiment. The linear portion (from θ = 0.2 to θ = 0.8) of the Hill representation, representing the log (Θ / 1-Θ) as a function of the log [His ₆ -EthR], was adjusted by linear approximation from Hill's equation, log (Θ / 1-Θ) = KH + ΠH [Hise-EthR]. KH and ΠH represent respectively the Hill constant and the Hill coefficient. For the semi saturation (θ = 0.5), the Hill equation is equal to 0, the global dissociation constant at equilibrium (Ko) is given by the relation Ko = io ^{kH / nH} . The same procedure and the same calculations were adapted and applied to the IG-62, IG-37 and IG-36 fragments.

AT 8. Primer extension analysis

Total RNA was isolated from M. bovis BCG 1173P2 as previously described (Mangan et al., 1997). The contaminating DNA was removed from the RNA preparation using DNase I (DNA-free ™ kit, Ambiom). Five pmoles of primer 0-273: 5'-GCTCTTGGTCGGGCAACGGTCCTG- 3 '(SEQ ID No. 24) were labeled at the 5' end using [γ- ³² P] ATP as described above.

10 μg of total RNA were mixed with 2 pmol of labeled primer and a mixture of dNTP (25 nmole for each nucleotide) in a final volume of 12 μl. After hybridization for 5 min at 70 ° C., the mixtures were rapidly cooled in ice, and 4 μl of First-Strand-Buffer [50 mM Tris-HCl, pH 8.3; 75 mM KCI and 3 mM MgCI ₂ , 10 mM DTT, DnaseOut (40 U; Invitrogen)] were added. The reaction mixture was then incubated at 37 ° C for 2 min before the addition of 200 units of M-MLV RT (Moloney murine leukemia virus reverse transcriptase; Invitrogen). The extension reaction was carried out at 37 ° C for 50 min, then the enzyme was inactivated at 70 ° C for 15 min. A sequencing reaction using the dideoxy chain terminator with primer 0-273 was carried out in parallel and loaded onto a sequence gel containing 8 M urea and 6% polyacrylamide, concomitantly with the extension products d 'primer, so as to determine the position of the starting point of the transcription.

A.9. DNase I fingerprint: DNase I protection experiments were carried out essentially as previously described (Reisenauer et al., 1999; Bellefontaine et al., 2002). A Hind \\\ - Sph \ fragment of 323 bp and an Xho \ -Hind \\\ fragment of 312 bp overlapping P _e thA were isolated from pCR2.1-P _e thA (2 + 4) (see paragraph A. 2 above). The first fragment was labeled at its end at the HindWl site using only 50 μCi [α- ³² P] dATP (3000 Ci / mmol; Amersham Biosciences) and 5 μl of Klenow fragment (Invitrogen). The Xho \ -Hind \\\ fragment was tagged at its end at the Xho \ site using the same protocol. The labeled DNA fragments (10,000 cpm) were incubated with purified Hisβ-EthR. Diluted DNase I (0.025 U, Roche) was added to each reaction mixture for 3 min at room temperature, then the reaction was stopped with 100 μl of stop solution (200 mM NaCl, 20 mM EDTA and 10% SDS ). The DNAse I digests were extracted with phenol, precipitated with ethanol and separated on a gel containing 8 M of urea and 6% of polyacrylamide. A sequencing standard generated with the universal primer and M13mp18 as a matrix was used to identify the position of the protected region.

B. Results:

B.1. Repression of EthA by EthR

It has been previously shown that overexpression of ethR leads to increased resistance to ETH in mycobacteria, while inactivation by mutation of ethR increases sensitivity to ETH (Baulard et al., 2000), suggesting that EthR negatively regulates the production of EthA, the putative mono-oxygenase required for activation of ETH.

In order to confirm this hypothesis, the promoter region of ethA was inserted upstream of the reporter gene lacZ in pBSh-D, a shuttle plasmid E. co // - mycobacterium which replicates in the form of a simple copy in mycobacteria ( Picardeau et al., 2000). The plasmid resulting, called pBSh-P _ethA - / acZ was electrotransformed in M. smegmatis mc ² 155 and M. bovis BCG 1173P2.

The colonies of recombinant M. smegmatis displayed an intense blue color on dishes of M7H11 enriched in X-gal (FIG. 1A), revealing the expression of ethA under these conditions. A pale blue color was observed for M. bovis BCG colonies (pBSh-P _ethA - / acZ) (data not shown). The two recombinant strains were then co-transformed with pMV261-eιf / ιR, a compatible multicopy plasmid comprising ethR (Baulard et al., 2000). The resulting colonies of M. smegmatis and M. bovis BCG were white on dishes enriched in X-gal, revealing that the presence of ethR repressed the expression of ethA.

In order to confirm the phenotype of M. bovis BCG, the β-galactosidase activity was quantified in the recombinant strains: Mycobacterium bovis BCG (pBSh-P _ethA - / acZ) exhibited a significantly higher β-galactosidase activity (15.7 U) compared to BCG co-transformed with pBSh-PethA- / acZ and pMV261 -ethR (0.51 U) (FIG. 1B). Thus, the overexpression of ethR in M. smegmatis and in BCG led to a strong inhibition of the expression of ethA. These results confirm the hypothesis that EthR negatively regulates the expression of ethA. By analogy with the capacity of tetracycline to inhibit the binding of TetR on its operator, the influence of ETH on the activity of the repressor EthR was evaluated. Mycobacterium smegmatis co-transformed with pBSh-P _ethA - / acZ and pMV261-ef /? R was cultured in the presence of variable concentrations of ETH, and the β-galactosidase activities were measured. Β-galactosidase activities did not appear to be influenced by treatment with ETH (data not shown), indicating that ETH does not interfere with the ability of EthR to suppress ethA, even at high concentrations. B.2. Binding of EthR to P _{and A} demonstrated by tests of modification of electrophoretic mobility

In order to determine whether the repression of ethA by EthR occurs via a direct physical interaction between the promoter region of ethA (P _e thA) and EthR, tests of modification of electrophoretic mobility (EMSA) were carried out on a DNA fragment. of 191 bp marked with [ ³² P] at its end, containing the ethA-R intergenic region, amplified by PCR from the chromosomal DNA of M. tuberculosis H37Rv. Complete and recombinant EthR was labeled with a histidine tag (His ₆ -EthR), at its N-terminal end, in order to facilitate the purification of the enzyme using Ni ²⁺ -chelate affinity chromatography. As shown in Figure 2A, the addition of increasing amounts of purified His ₆ -EthR to the [ ³² P] labeled DNA fragment resulted in delayed migration of the fragment on native polyacrylamide gels, revealing that EthR is capable to bind to the ethA-R intergenic region. Specific and non-specific DNA competitors were used to determine the specificity of binding of HiS ₆ -EthR to the promoter region of ethA (Figures 2B and 2C). The reappearance of a free probe was observed when a specific competitor (unlabeled probe) was added in excess of 500 to 1000 times to the mixture containing the labeled probe and His ₆ -EthR, while the same amount of non-specific competitor (see Materials and Methods, part A above) did not modify the interaction between Hise-EthR and the labeled probe.

B.3. Determination of EthR binding sites

In order to precisely locate the operator of ethA, DNase I fingerprint experiments were performed on the two strands of DNA. Two fragments, a Hind \\\ - Sph \ fragment of 323 bp and a XΛol-HindWl fragment, containing the intergenic region ethA-R were labeled at their ends on the opposite strands. For both fragments, purified Hise-EthR was able to protect the same region by 55 bp after digestion moderate to DNase I (Figures 3A and 3B). This region is located from 5 nucleotides upstream from the ethA initiation codon to 16 nucleotides upstream from the ethR initiation codon, thus covering 55 bp beyond the 75 bp intergenic region (Figure 3C). Thus, EthR binds to an unusually long region of DNA, suggesting that it interacts with its operator in the form of a functional multimer. Interestingly, DNase I protection experiments performed with either low (Figure 3A) or high (Figure 3B) concentrations of His ₆ -EthR resulted in equivalent fingerprints, suggesting a cooperative association of the protein with the operator.

B.4. Identification of the point of initiation of transcription of the ethA gene

In order to understand the mechanisms of repression of ethA by EthR, the +1 point of transcription of ethA was identified using a primer extension analysis. When the total RNA prepared from M. bovis BCG 1173P2 was subjected to a primer extension analysis, a unique product was identified, starting at nucleotide A located 13 bases upstream of the initiation codon of the translation of ethA (Figure 3D). This result indicates that the initiation point of the transcription of efM and the RNA polymerase recognition sites overlap with the operator of ethA (FIG. 3C).

When the sequence upstream from the initiation point of the transcription of efM was compared with other mycobacterial promoter sequences, no putative conserved box -10 or -35 could be identified (FIG. 3C). The efM promoter does not resemble the consensus of σ70 from E. coli and therefore should be classified in group C of the regulated promoters proposed by Gomez and Smith (2000).

B.5. Stoichiometry of the interaction between His ₆ -EthR and the ethA-R intergenic region A 106 bp biotinylated double stranded DNA fragment (Figure 4A; IG-106) spanning the ethA-R intergenic region was used as a ligand to measure the binding stoichiometry of Hisβ-EthR by surface plasmon resonance (SPR). Increasing concentrations of His ₆ -EthR were injected into a Sensor Chip cell functionalized with 205 RU of biotinylated DNA fragment until the equilibrium ligand is saturated (FIG. 4B). Since Hise-EthR has a calculated M _r of 25,920 and 205 RU of DNA fragment (69 k Da) have been fixed on the surface of the Sensor Chip cell, the Hise-EthR: DNA ratio has been estimated using Equation 1 for each concentration of Hise-EthR. A stoichiometry of 8 molecules of .His ₆ - EthR per DNA molecule was deduced. Interestingly, this stoichiometry was very quickly reached at a Hise-EthR concentration threshold (Figure 4C). This experiment was repeated using a 62 bp biotinylated double stranded DNA fragment (Figure 4A; IG-62) overlapping the 55 bp region identified by the DNase I fingerprint experiments. An identical stoichiometry of 8 molecules of His ₆ -EthR per DNA molecule was measured (Figure 4D), confirming that all of the repressor molecules bind in the region protected by DNAse I.

Assuming that a homo-octamer is likely to have some link symmetry, the center of the DNase I protected region (the G in position 32 of the ethA-R intergenic region) seems to split two copies of a long direct repetition (7-CA-4-CGt / aa / g- ATGTCGA; SEQ ID N ° 25 and 26) which itself includes imperfect inverted repetitions (bases in italics, Figure 4A). When a fragment of 37 bp (Figure 4A, IG-37) overlapping these two _repeats, was directly attached to the cell surface Sensor chip, an average of 4.5 molecules of cooperative binding Hise-EthR per molecule DNA was observed (Figure 4E). When a 36 bp PCR fragment (Figure 4A; IG-36) overlapping the two nearly perfect inverted repeats (TAa / gTGTCGA) centered on base 34 was used, the binding of His ₆ -EthR was greatly reduced (Figure 4F) .

β. 6. Determination of the Hill coefficient (n ») and of the Ko of the interaction between Hise-EthR and the intergenic region ethA-R In order to assess the affinity of Hise-EthR for the intergenic region ethA-R, analyzes of the equilibrium interaction were performed. 100 RU of the 106 bp biotinylated DNA fragment mentioned above (IG-106) were immobilized on a Sensor Chip cell. This small amount of immobilized DNA allowed the protein to reach a binding equilibrium for a wider range of protein concentrations (from 0.1 to 2 μM) in the maximum possible injection volume for the BIAcore device. 2000. Under these conditions, the capacity of the Sensor Chip cell was estimated at R _ma χ = 225 RU, which corresponds to the same stoichiometry of 8 His ₆ -EthR molecules as that deduced from the previous SPR analyzes (see paragraph B.5). . The fractional saturation of the DNA at equilibrium R _eq / Rmax (where R _eq is the RU value measured at equilibrium) was determined for each concentration of proteins. The function of Hill (Fersht, 1985), has been reported as a function of the log of [His ₆ -EthR] (Figure 5). Analysis of the linear part of the Hill function (from 0.2 to 0.8) by linear regression (R ² = 0.9904) provided an overall equilibrium binding constant (KD) of 146 nM and a Hill coefficient (ΠH) of 3.46. The KD value represents an average value for the 8 binding sites. The maximum theoretical value of the Hill coefficient is 8 and corresponds to the number of binding sites of His ₆ -EthR. The value of the Hill coefficient of 3.46, greater than 1, indicates a cooperative binding of HiSβ-EthR on the DNA probe (Fersht, 1985). B.7. Need for operator presence for octomerization of His6-EthR

The SPR experiments described above (see section B.5) indicated that EthR binds cooperatively in the form of a homo-octamer to the operator of ethA.

In order to determine whether the multimerization of EthR requires the presence of this DNA target, Hise-EthR was subjected to size exclusion analytical chromatography on a Superdex 200 HR 10/30 column in the presence of a PCR fragment of 94 bp containing the ethA-R intergenic region (see materials and methods, part A above).

The 94 bp DNA fragment eluted as a single peak with maximum absorbance at 260 nm at an elution volume (V _e ) of 11.11 ml (Figure 6A). Adding small amounts of purified His ₆ -EthR to the 94 bp DNA fragment resulted in the elution of a single peak with maximum absorbance at 260 nm at a V _e of 10.27 ml (Figure 6 B ). As these results were obtained under the same salt conditions as those used for the SPR analyzes, the decreasing elution volume of the DNA in the presence of Hise-EthR most certainly corresponds to an increase in the molecular mass of the DNA complex. 94 bp due to the binding of the protein octamer to the operator site. The addition of increasing concentrations of His ₆ -EthR to the DNA fragment led to the gradual appearance of an additional peak eluting 15.30 ml after loading, with maximum absorbance at 280 nm (Figures 6C and 6D) . This peak corresponds to a molecule which has a Stokes radius practically equivalent to that of the standard of molecular mass chymotrypsinogen A (25,000 Da) (Amersham Biosciences). These data were confirmed by determining the elution volume of Hise-EthR in the absence of DNA fragment (Figure 6E), and strongly suggest that under in vitro conditions which allow Hise-EthR to bind to its operator, the unbound repressor does not adopt a large multimeric form, but rather a monomeric form. However, being since size exclusion chromatography is based on the Stokes radius rather than the molecular weight per se, the fact that Ηis ₆ -EthR dimerizes in solution in the absence of its operator site cannot be excluded.

B.8. Oligomerization of EthR in vivo

In order to determine whether EthR can dimerize in the absence of its target DNA, the bacterial double hybrid system (BACTH) was used (Karimova et al., 1998). In this system, the interaction of two proteins leads to functional complementation between the two domains of adenylate cyclase (CyaA) from Bordetella pertussis, leading to the synthesis of cAMP. As a soluble regulatory molecule, cAMP is then able to activate the transcription of dependent cAMP genes. ethR was thus cloned in phase with the region coding for the 25 kDa domain of CyaA, giving pT25-EthR. ethR was then fused with the region coding for the 18 kDa domain of CyaA so as to produce pT18-EthR. A strain of E. coli cyclase-deficient (DHP1) (Karimova et al., 1998), was transformed with different combinations of plasmids (Figure 7). When they were co-transformed with pT25-EthR and pT18-EthR, the colonies regained their capacity to grow on minimum medium enriched in maltose. None of the clones obtained by the co-transformation of the various control plasmids were able to grow, demonstrating that the functional complementation of T25-EthR and T18-EthR was specifically linked to the interaction of two molecules of EthR (Figure 7). Since no DNA homologous to the operator of EthR has been detected in the genome of E. coli, this result suggests that EthR is a dimer when it is not linked to its operator. II- Purification and crystallization of the Hise-EthR repressor:

A. Preparation of the Hisg-EthR protein

The entire ORF encoding the EthR repressor of M. tuberculosis was subcloned into the expression vector pET15b (Novagen) in order to allow the production of the protein fused to an amino-terminal histidineβ tag. E. coli BL21 (DE3) was transformed and the recombinant clones were selected on LB containing 100 μg / ml of ampicillin. Five ml of LB containing 100 mg / ml of ampicillin were inoculated with E.coli BL21 (DE3) (pET15b-His ₆ EthR) and cultured with stirring at 37 ° C overnight. The preculture was used to inoculate 2 liters of LB medium in a D0 ₆ oo of 0.1. At DOβoo = 0.6-0.7, the expression of ethR was induced by adding 1 mM IPTG. After 2 hours of induction at 37 ° C., the bacteria were recovered by centrifugation at 7,500 rpm, with a Sorval RC 5B PLUS centrifuge, rotor SS-34. The bacterial pellet was then frozen at -20 ° C. The pellet was resuspended in 10 ml of buffer A (50 mM Tris, pH 7.5, 10 mM imidazole, 250 mM NaCl) and then lysed by passage through the French press (Bioritech) under a pressure of 6.2 × 10 ⁶ Pa. The lysate was centrifuged for 30 minutes at 4 ° C at 18,000 rpm in a Sorvall RC-5B PLUS centrifuge, SLA-3000 rotor and the supernatant recovered.

25 ml of affinity resin (chelating Sepharose Fast Flow - Amersham Biosciences) were washed with 5 ml of double-distilled water, then loaded with Nickel by adding 5 volumes of NiS0 ₄ 100mM. The resin was washed with 5 volumes of double-distilled water and equilibrated with 5 volumes of buffer A. The resin was loaded on a disposable column. The bacterial lysis supernatant was loaded onto the affinity column at a flow rate of 2 ml / min using a P-3 peristaltic pump (Pharmacia Biosciences). The column was washed with 10 volumes of 50 mM Tris pH 7.5 buffer; 50 mM imidazole; 250 mM NaCI. The His ₆ -EthR protein was eluted with 1.5 volumes of elution buffer (50 mM Tris pH 7.5; 250 mM imidazole; 250 mM NaCI). The recovered fractions were analyzed on a 12.5% SDS-PAGE gel. The fractions with sufficient purity were pooled and dialyzed twice in 10 mM Tris pH 7.5; 200 mM NaCl, at 4 ° C. The solution was then concentrated on a Centriplus YM-3 column (Amicon), then with Macrosep 3K (Pall), in order to reach a final concentration of 50 mg / ml of protein.

Crystallization: Crystallization was carried out at 20 ° C, by the method of the drop suspended in vapor diffusion (2 μl of protein solution, 2 μl of precipitant solution and 0.75 ml of reservoir). Hisβ-EthR crystals were obtained in 7-10 days with solution 15 of Cryokit for protein crystallization (Sigma) (0.17 M ammonium sulfate; 0.085 M Na-cacodylate pH 6.5; 15% glycerol; 25.5% PEG8000). These crystals are called form 2 in the remainder of the protocol.

B. Labeling of the EthR protein with seleniated methionines

E.coli BL21 (DE3) (pET15b-His ₆ EthR) was cultured at 37 ° C in 2 liters of M9 medium containing 6 g / l of Na ₂ HP0 ₄ , 3 g / l of KH ₂ P0 ₄ , 0, 5 g / l NaCl; and added, at the start of the culture, of 4 g / l of glucose, 1 g / l of NH ₄ CI, 2 ml of MgSO ₄ 1 M, 2 ml of CaCl ₂ 100 mM, 20 ml of the mixture of vitamins MEM ( Sigma) and ampicillin at 100 mg / l., The methionine biosynthetic pathway of E.coli BL21 (DE3) (pET15b-HiS ₆ EthR) was blocked at Dθ6oo = 0.5 (Doublie, 1997), by adding amino acids that inhibit aspartokinases: lysine 100 mg / l, threonine 100 mg / l, phenylalanine 100 mg / l, leucine 50 mg / l, isoleucine 50 mg / l, valine 50 mg / l and selenomethionine at 75 mg / l. Incubation at 37 ° C was continued to reach a Dθ ₆ oo = 0.6, then 1 mM of IPTG was added in order to induce the expression of ethR. The culture was stirred at 37 ° C for an additional 3 hours. The bacterial pellet, recovered by centrifugation at 7500 rpm in a Sόrvall RC-5B PLUS centrifuge, SLA-3000 rotor, was resuspended in 20 ml of lysis buffer (50 mM Tris pH 7.5, 10 mM imidazole, 250 mM NaCl). 1 ml of Complete anti-protease mixture (Roche) was added before breaking the bacteria with the French press. The lysate was centrifuged at 18,000 rpm, 30 minutes, at 4 ° C, with a Sorvall RC-SB PLUS centrifuge, rotor SS-34. The supernatant was loaded using a P-3 peristaltic pump (Amersham Biosciences) onto 25 ml of affinity resin (Chelating Sepharose Fast Flow- Amersham Biosciences) pre-balanced with 5 volumes of lysis buffer. The Hise-EthR protein was eluted with 1.5 volumes of elution buffer (50 mM Tris pH 7.5, 250 mM imidazole, 250 mM NaCl). The recovered fractions were supplemented with 10 mM of Dithiothreitol (DTT), 0.2 mM of ethylenediaminetetraacetic acid (EDTA), then analyzed by 12.5% SDS-PAGE gel stained with Coomassie blue. The pure fractions were combined and dialyzed twice against 10 mM Tris, pH 7.5, 200 mM NaCl, 10 mM DTT and 0.2 mM EDTA. The solution was then concentrated as described above to reach a spectral concentration of 10 mg / ml, then stored at -80 ° C.

Crystallization of EthR-selenomethionine: Before being used for crystallization, Hisβ-EthR / met-sel was thawed on ice and concentrated to 50 mg / ml using 3K microcons (Amicon), at 4 ° C. The seleniated protein was crystallized by the method of drops suspended in vapor diffusion (2 μl of protein solution, 2 μl of precipitant solution, 0.75 ml of reservoir), at a temperature of 20 ° C. Crystals appeared between 7 and 10 days in the drops containing 0.17 M ammonium sulfate (Fluka), 0.085 M Na-cacodylate pH 6.5 (Sigma), 15% glycerol (ICN) and 25% polyethylene glycol 8000 (Fluka). These crystals were used to determine the three-dimensional structure of His ₆ -EthR / met-sel. They are called form 1 crystals in the remainder of the protocol. C- Obtaining a second crystalline form

Co-crystallization tests were carried out using the protein HiS ₆ - EthR / met-sel in the presence of 5 different DNA fragments 15 bp long and covering the zone of binding of the protein to the operator of ethA (see section I above). The protein was thawed on ice and concentrated to 50 mg / ml. The protein-DNA ratio was fixed at 2: 1. The mixture was subjected to co-crystallization by the method of drops suspended in vapor diffusion (1 μl of protein solution, 1 μl of DNA solution, 2 μl of crystallizing agents, 0.75 ml of reservoir), at a temperature of 20 ° C. Crystals were obtained after 2 to 3 days of incubation, for the His ₆ -EthR / met-sel / DNA fragment 1 mixture (SEQ ID No. 27) with a crystallization solution containing 0.17 M sulfate d ammonium (Fluka), 0.085 M Na-cacodylate pH 6.5 (Sigma), 15% glycerol (ICN) and 35% PEG8000 (Fluka). These crystals are called form 3 in the remainder of the protocol.

D- Acquisition of diffraction data and determination of the structure of EthR

The crystals of form 1 (crystals produced from Hisβ-EthR / met-sel protein containing selenomethionines) were used to determine the crystallographic structure.

The diffraction data were acquired on the BM14 beamline at the European synchrotron in Grenoble (ESRF). The procedure for acquiring data and determining the structure was as follows.

A crystal of form 1 was frozen using an Oxford brand cryogenics system. The diffraction data were measured by a CCD detector, of the Marresearch brand, 133 mm in diameter, positioned 150 mm from the crystal. Diffraction patterns have been recorded for a crystal oscillation of 1 °. The exposure time per exposure was set to 20 seconds. 360 images were measured at the wavelength of 0.97865A. The orientation matrix of the crystal in the X-ray beam was determined with the XDS software. The diffraction data were integrated and reduced with the same software. The crystals obtained belong to the space group P2- | 2-ι2 (Figure 8) and have the following lattice parameters: a = 57.57 A; b = 85.94 A; c = 100.41 A. They contain two monomers per asymmetric unit. As shown in Figure 8, the structure reveals two EthR repressor monomers associated in a symmetrical dimer. The monomers are associated by their ligand binding domain.

Each monomer (amino acids 22 to 215) is entirely helical and comprises 9 helices (respectively α1 to α9 and 1 'to α9', see Figure 8). The crystal revealed the presence of one ligand molecule per monomer, embedded in a tunnel formed by the helices α4, α5, α6, al, α8 and α8 '(from the second monomer). The centers of the HTH (“helix-turn-helix”) domains for attachment to the DNA of this structure are 52 A. The determination of the phases associated with each structural factor was determined by the “SAD” method (“ Single Wavelength Anomalous Diffraction ”). This method is based on the use of the abnormal signal of selenium incorporated into the protein via selenomethionines (see preparation of form 1 crystals). This signal is maximum at the wavelength used (0.97865 A), wavelength selected on the basis of an analysis of fluorescence emission by the crystals. The position of the selenium atoms in the asymmetric unit of the crystal was determined with the Shelxd software. These positions have been refined and the initial phases calculated with Sharp software. The phases were then improved by electronic density modification procedures implemented in Solomon and Dm software, from the CCP4 software suite. The molecular model was initially built with the software Arp / Warp and the TURBO-FRODO software. The final structure was obtained by energy minimization (refinement procedure) with the CNS software and manually inspected with the TURBO-FRODO software. The parameters describing the topology of the EthR ligand were obtained with the HIC-UP software.

The crystals of form 2 are similar to the crystals of form 1. These, however, were obtained with non-seleniated recombinant protein. The crystallization conditions are identical, but the source of the crystallization solutions is different. The diffraction data were acquired at the Lille Institute of Biology, using equipment from the macromolecular crystallography laboratory. The procedure for acquiring data and determining the structure was as follows. A crystal of form 2 was frozen using an Oxford brand cryogenics system. The diffraction data were measured by an Imaging Plate detector, of the Marresearch brand, of 345 mm in diameter, positioned at 250 mm from the crystal. Diffraction patterns were recorded for a crystal oscillation of 1 °. The exposure time per exposure was set to 120 seconds. 180 images were measured at the wavelength of 1.5418 A. The orientation matrix of the crystal in the X-ray beam was determined with the XDS software. The diffraction data were integrated and reduced with the same software. The phases associated with each structure factor were determined by isomorphic positioning of the molecular model of HiS _δ -EthR / met-sel (obtained from type 1 crystals). The final structure was obtained by energy minimization (refinement procedure) with the CNS software and manually inspected with the TURBO-FRODO software. The parameters describing the topology of the EthR ligand were obtained with the HIC-UP software. The crystals of form 3 belong to the space group P4ι2- | 2, with the following lattice parameters: a = 122.2 A; b = 122.2 A; c = 33.6 A. These were obtained with recombinant seleniated protein (His ₆ - EthR / met-sel). The diffraction data were acquired at the Lille Institute of Biology, using equipment from the macromolecular crystallography laboratory. The procedure for acquiring data and determining the structure was as follows.

A crystal of form 3 was frozen using an Oxford brand cryogenics system. The diffraction data were measured by an Imaging Plate detector, of the Marresearch brand, of 345 mm in diameter, positioned at 250 mm from the crystal. Diffraction patterns were recorded for a crystal oscillation of 1 °. The exposure time per exposure was set to 600 seconds. 100 images were measured at the wavelength of 1.5418 A. The orientation matrix of the crystal in the X-ray beam was determined with the XDS software. The diffraction data were integrated and reduced with the same software. The determination of the phases associated with each structure factor was carried out by the molecular replacement method (implemented in the AMoRe software) using as a model the structure of EthR obtained from type 1 crystals. The final structure was obtained by energy minimization (refinement procedure) with CNS software and manually inspected with TURBO-FRODO software. The parameters describing the topology of the EthR ligand were obtained with the HIC-UP software. The crystals of form 3 contain only one monomer per asymmetric unit, confirming the strict symmetry of order two in the homodimer of EthR in the presence of a lipophilic inducer.

E. Analysis of the structure

Analysis of the structure of EthR confirms that this transcriptional repressor belongs to the family called "TetR / CamR". At the present time, the structures of two members of this family, TetR and QacR, have been determined, in the induced conformations and not induced. TetR is the transcriptional repressor of TetA, the protein responsible for resistance to tetracycline in gram negative bacteria. The repression function of TetR is inhibited by tetracycline, which attaches symmetrically to each monomer on the TetR homodimer. QacR is the transcriptional repressor of QacA, the protein responsible in Staphylococcus aureus for resistance to many antiseptics and disinfectants. QacR and TetR are homodimers, of different topology although they are both composed only of α helices. EthR, like TetR and QacR, is a homodimer. Each monomer has a topology similar to that of QacR and consists of 9 α helices. The three N terminal helices of each monomer make up the HTH type DNA binding motif (α2, α3), stabilized by α1 (Figure 8). The other 6 α helices make up the ligand binding domain as well as the dimerization zone (Figure 9). The dimerization pattern consists of the helices α8 and α9 (with αδ 'and α9') in EthR, just like in QacR. The dimerization covers an area of 1562 A ² per monomer of EthR, a contact area comparable to that observed for QacR (1530 A ² ). Although the ligand binding domains of QacR and EthR are similar, they exhibit weak sequence homology, which is reflected by the different relative orientations of the helices forming these domains (overall RMSD of the main 2 A chains for α4 helices to α9). The α4, α7 and α9 helices are oriented similarly in QacR and EthR, while the α5, αδ and αδ helices have more divergent axial orientations, causing displacements of up to 5 A between the topologically equivalent residues. This observation reflects the specificity of repressors vis-à-vis their respective respective inducers. The relative orientation of HTH patterns and Ligand binding domains are also slightly different (Figure 9), which results in a more flattened EthR dimer. It is specified that the terms “ligand” and “inducer” are here synonymous and refer to the definition given above. In addition, the “liganded” or “induced” form of the EthR repressor corresponds to the repressor “linked to a ligand”.

Similarly, in the context of the invention, the terms "attachment" and "connection" are used interchangeably. The conformational changes resulting from the binding of ligands to QacR have been described. The binding of ligands causes changes in conformation which result in a relative reorientation of the two HTH domains within the dimer. This reorientation produces a separation of the HTH motifs so that they can no longer bind in the major groove of the target DNA molecule. The binding of ligands to QacR produces a turn-helix transition which extends the α5 helix. This movement has repercussions on the surrounding propellers. Note that only one ligand molecule binds per QacR dimer, resulting in an asymmetric dimer. The monomer on which the ligand binds undergoes the most drastic conformational change (elongation of the α5 helix, and repercussion on the adjacent helices). However, the binding of a ligand in a monomer affects to a lesser extent the conformation of the second monomer of the repressor, the signal propagating via the dimerization zone. Overall, these conformation modifications cause an increase in the “center to center” distance of the recognition helices of the two HTH α3-α3 'motifs in QacR dimers from 37 A (free form or linked to DNA) to 4δ A (liganded or induced form in the presence of ligands). This increase in distance explains the inability to bind the induced form of QacR to DNA. The three crystalline forms of EthR obtained all reveal the presence of a lipophilic ligand associated with each subunit of the dimer (octanoate hexadecyl (HeOc); for the specific characterization of this ligand, see part III below).

This induced form of EthR with its lipophilic ligand reveals structural characteristics similar to those observed in the structure of QacR in the induced form. The α5 helix of EthR has a length similar to the α5 helix of QacR in the ligand-binding monomer.

A detailed analysis of the structure of EthR indicated that the DNA recognition helices α3 and α3 'are separated by 52 A. Such a distance between the HTH domains is, by analogy with QacR, incompatible with the binding of the repressor on its target DNA (Schumacher et al., 2001 and Schumacher et al., 2002). EthR fixes a ligand per monomer, symmetrically, unlike QacR which fixes only one ligand per dimer. The binding of the ligand causes a turn-helix transition of the residues Tδ9 to Y93, only at the level of the subunit linked to the ligand, so that the C-terminal end of α5, but not of α5 ′, is extended by a tower. Y92 and Y93 are therefore capable of interacting with the ligand. The turn-helix transition is essential not only for ligand binding, but also for induction, since the formation of the extra turn in the α5 helix results in repositioning of αδ and the DNA binding domain which is attached to him. As indicated above, the movement of α6 causes the translation and rotation of the DNA binding domain which releases QacR from its operator site. The length of the α5 helix, an imprint in the induction of QacR, is comparable in EthR and QacR linked to their ligand, which suggests similar induction mechanisms.

However, due to the presence of one ligand molecule per monomer in EthR, both the α5 and α5 'helices undergo the putative HT-helix transition and displacement of the HTH motif, which results in a more open conformation of the recognition propellers separated by 52 A.

This induction mechanism differs from that of TetR, which binds to 2 ligands per dimer and uses an induction mechanism, involving several movements "through helix", in which the binding of tetracycline increases the distance center to center DNA recognition helices of only 3 A (Orth et al., 2000). Furthermore, the ligand binding domain of EthR is unique and does not present a molecular analogy with that of QacR. The ligand binding site in EthR crosses the domain defined by the α4 to α9 helices. It is delimited by the propellers α4, α5, α6, al, α8 and αδ '(coming from the second monomer) and is in the form of a tunnel or pocket extending parallel to the propellers α4, α5 et al. Thus, this pocket consists of residues belonging to all the helices of the ligand binding domain of EthR. The only apparent portal of entry into the ligand binding site is formed by the divergence of the α6, α7 and α8 helices.

As shown in Figure 11, one end of the ligand is almost parallel to α6 (C1 to C6 in the HeOc ligand molecule described below), near the surface of the protein. Then, it winds around the Trp13δ residue (C7 to C15) and sinks into the ligand binding domain, parallel to the axes of the α4, α5 and α7 helices. The structure indicates that the C1 to C15 region of HeOc is divided into 4 segments with different curvatures (C1 to C6; C6 to C9; C9 to C12 and C12 to C15), and which bypass the aromatic cycle of the W133 residue (Figure 11 ). HeOc adopts a buried stretched conformation from position C15 to its end C24, over approximately 12 A. Thus, the HeOc ligand essentially adopts a stretched conformation, over a distance of 21 A between carbons C1 and C24, with a coiled region (C1 to C15) then a linear carbon chain (C15 to C24). The ligand binding pocket is defined by the following EthR residues (see Figure 11, in the direction C1 to C24): R128, V134, G124, L76, Q125, F187, F184, T121, W138, F114, L183, E180 , M142, N179, N176, W145, F110, W207, T149, 1107, Lδ7, V152, Y148, G106, W103 and M102. The binding cavity of the region C15 to C24 is rich in hydrophobic (M142, 1107, Lδ7, V152, M102) and aromatic amino acids (W145, F110, W207, Y14δ, W103). It also contains 2 asparagines, both located opposite the He17 C17 and C13 carbons (Figure 11). A water molecule is also buried via a hydrogen bridge with the N179 residue near the C17 carbon of the ligand. 4 aromatic residues play a particularly important role in the curvature of segments C6 to C15: F114, F110, W13δ and W145. W133 is crucial for delineating the ligand binding domain. It is stabilized by the interactions of its side chain NE1 with OE1 of the residue E160. E1δ0 is very stabilized, since it also interacts with NE of the residue R1δ1 belonging to the other monomer present in the dimer of EthR, near the axis of double symmetry.

Although the structure of unbound EthR has not yet been elucidated, it has been observed that F110 and F114, which belong to the last round of α5, point towards the ligand binding site and interact with the ligand. By analogy with the change at Y92 when QacR is linked to its ligand, it appears that an increase in the center-to-center distance of the HTH domains when EthR binds to its ligand may be due to conformational changes arising in the portion distal to α5, in the form of a turn-helix transition which results in the aromatic change near the ligand binding site. QacR is induced by binding to mono and bivalent cationic lipophilic ligands exhibiting great structural diversity. The structures of the QacR repressor linked to various cytotoxic ligands revealed the presence of 2 separate but potentially overlapping binding sites within from a single pocket. These 2 binding sites are practically parallel to αδ of QacR and end in α5 which closes the cavity. The comparison of the binding sites of QacR and EthR revealed that the 2 ligand binding domains have the same ligand entry portal. HeOc partially overlaps site 1 of QacR, adjacent to the ligand gateway. However, site 2 of QacR is not found in the HeOc-EthR complex due to the different orientations of the α5 and αδ helices which narrow the binding cavity of EthR. Instead, EthR has a second cavity, not observed in QacR, which extends over approximately 15 A, parallel to α4 and α7 (Figure 12). Site 2 is delimited from site 1 of EthR by the positions W13δ and E1δ0. This second binding site in EthR is not completely occupied by HeOc since it also contains a molecule of water buried near the C17 position of HeOc. In general, in QacR, the linking cavities not occupied by the various ligands also contain water molecules.

III- Characterization of the ligands of the EthR repressor:

A- Materials and methods: In order to characterize the ligand, 750 ng of co-crystal of EthR-ligand were washed using the crystallization buffer, transferred individually into chloroform in order to denature the proteins and to extract therefrom the ligand. Analysis of the extraction chloroform by GC / MS made it possible to identify hexadecyl octanoate (HeOc) as a major product present in the solution. The structure of HeOc was then superimposed on the X-ray analysis spectrum of the crystal and was found to be compatible with the electron density corresponding to the ligand. B- Results: The lipophilic ligand co-purified and co-crystallized with the repressor EthR was identified, both by mapping the electron density (Figure 13) and by GC-MS / MS analysis (Figure 14). The equivalent of 750 ng of co-crystallized EthR was extracted by immersion of the crystals in chloroform. The elution profile of gas chromatography has a specific peak at 22.16 min. The electron ionization spectrum (EI) of this peak has been compared to a spectra bank and corresponds to hexadecyl octanoate (HeOc). This spectrum has a peak corresponding to the molecular ion M + at m / z = 36δ. The ions having the characteristics m / z = 41, 55, 57, 69, 63, 97, 111 are characteristic of the molecules CnH _2n -ι. The ions m / z = 43, 57, 71 correspond to CnH2n + ι. The ion m / z = 144 represents a Me Lafferty-type rearrangement of the ester and m / z = 269 corresponds to the C ₁₆ H ₃₃ 0-CO ⁺ ion (Me Lafferty and Turecek, 1993).

The co-crystallization of a ligand has been described for the first time for the crystal structure of the RXR α (F31δA) / RAR α complex, revealing the presence of a fatty acid responsible for the property of the constituent agonist of this mutant (Bourguet et al., 2000). More recently, the crystal structure of the binding domain of the ROR β ligand has been resolved in connection with a stearate ligand (Potier et al., 2003). The stearate molecule acts as a "filler" rather than a true ligand, since a small percentage of bag occupancy and a partially disordered conformation have been observed, and, moreover, that no activating effect or antagonist on the activity of ROR β has been detected.

On the other hand, the homodimer of EthR systematically co-crystallizes with HeOc according to a stoichiometry of 2: 2, with total occupation and a well defined electron density, which indicates an essential role of HeOc in the stabilization of the induced conformation. of EthR. Furthermore, interestingly, it has recently been shown that EthA is a Baeyer-Villiger type monooxygenase, capable of converting a wide range of ketones, including long chain ketones, into their corresponding esters (Fraaije et al., 2004). Given the regulation of the ethA gene by the EthR repressor, it appears that substrates of EthA could serve as ligands for the EthR repressor.

IV - Increase the sensitivity of mycobacteria to ETH: therapeutic perspectives:

A. Synergy tests:

M. smegmatis me ² 155 was grown in 250 ml Erlenmeyer flasks containing 100 ml Sauton medium. The cells were incubated at 37 ° C with shaking for 2 days. During this experiment, the synergy between ETH and a given ketone was determined to be the ability of this ketone to promote the bacteriotoxicity of ETH at a concentration below the minimum inhibitory concentration (15μg / ml) . The phenotypic effect of ketones was evaluated by depositing serial dilutions of the culture on LB medium containing a low concentration of ETH (5 μg / ml), 1 mM of ketones (benzylacetone, 2-octanone, 3-octanone or 2- decanone), or a combination of ETH (5μg / ml) and ketones (1 mM). The appearance and the number of colonies for each deposit were analyzed after three days of growth at 37 ° C.

B. Results

Repression of the ethA gene by the EthR repressor reduces the sensitivity of mycobacteria to ETH. Conversely, an EthR-deficient mutant showed an increased sensitivity to ETH compared to the mother strain (Baulard et al., 2000). Interference with the activity of EthR therefore leads to an increase in sensitivity to ETH.

Interference with EthR in vivo was tested using synthetic hexadecyl octanoate. However, hexadecyl octanoate is a compound too hydrophobic to be used in combination with a biologically compatible solvent.

Other related, more soluble compounds have therefore been sought. As noted above, EthA has recently been described as being capable of converting a wide range of ketones, including long chain alkanones and aromatic ketones, to their corresponding esters, suggesting a possible role for this activator in acid metabolism mycolic (Fraaije et al., 2004). The capacity of EthR to bind to a long hydrophobic ester such as hexadecyl octanoate is compatible with this hypothesis and proves the link between the regulation of the ethA gene and the activity of the monoA oxygenasease EthA.

It has therefore been envisaged that some of the molecules described by Fraaije et al. could constitute interesting EthR ligands within the meaning of the present invention. In order to determine whether some of these molecules could actually interfere with the activity of the EthR repressor, and thereby modify the sensitivity of mycobacteria to ETH, 4 different ketones were tested for their ability to increase the sensitivity of M. smegmatis at ETH. This ability was tested on the growth of mycobacteria, on agar dishes supplemented with 5 μg / ml of ETH, which is a concentration significantly lower than its minimum inhibitory concentration, which is 15 μg / ml.

One of them, benzylacetone (BA), showed a significant bacteriostatic effect when it was co-administered with a low dose of ETH (5μg / ml). No antimycobacterial effect of benzylacetone was observed in the absence of ETH, which therefore demonstrates the strict synergy of this compound with ETH (see Figure 15). Strong inhibitors of EthR could therefore lead to a drastic reduction in the therapeutic dose of ETH required in therapy and, consequently, reduce the toxicity associated with the treatment of tuberculosis with ETH and related drugs. In addition to ETH, EthA is also involved in the activation of other thioamides, such as thioacetazone and isoxyl (De Barber et al., 2000). Despite its effectiveness, the World Health Organization does not recommend the use of thioacetazone, in particular in individuals infected with HIV due to the risk of a severe dermal reaction (Blanc et al., 2003). Depression of the ethA gene by EthR ligands could therefore reduce the therapeutic dose of thioacetazone and isoxyl. The latter compound has recently been shown to target a lipid desaturase (Phetsuksiri et al., 2003), which represents an anti-mycobacterial target that is not yet exploited for therapeutic purposes.

Mycobacterium leprae is also sensitive to ETH and derivatives such as prothionamide, but is not sensitive to clinically significant levels of INH, due to multiple lesions in the katG gene of this organism. This is why INH cannot be used to treat leprosy. However, since both ethR and ethA are conserved in the reduced genome of M. leprae, the development of EthR ligands that inhibit its repressor activity, as drug additives to ETH and its derivatives, could therefore allow improve chemotherapy for leprosy. REFERENCES

Antoine et al. (2000) J. Bacteriol 182: 5902-5905 Aramaki et al. (1995) Protein Eng 8: 1259-1266

Baulard et al. (2000) J. Biol. Chem. 275: 26326-28331

Bellefontaine et al. (2002) Mol. Microbiol. 43: 945-960

DeBarber et al. (2000) Proc. Nati. Acad. Sci. 97: 9677-9662

Espinal et al. (2001) N. Engl.J. Med. 344: 1294-1303 Fersht et al. (1965) In Enzyme Structure and Mechanism. Fersht, A (ed.).

Freeman, 263-292

Frey et al. (1965) Gene 36: 143-150

Gomez et al. (2000) In Molecular Genetics of Mycobacteria. Hatfull, G. F and

Jacobs, W.R. (eds). Washington, D.C .: American Society for Microbiology Press.

Jenner et al. (1967) Lepr. Rev. 58: 31-37

Jonhston et al. (1967) J. Pharm. Pharmacol. 19: 1-9

Karimova et al. (1996) Proc. Natl. Acad. Sci. 95: 5752-5756

Mahmoudi et al. (1993) JAMA 270: 65-6δ Mangan et al. (1997) Nucleic Acids Res. 25: 675-676

Miller et al. (1972) Experiments in Molecular Genetics. Cold spring

Harbor, NY: Cold Spring Harbor Laboratory Press, 352-355

PharmaciaBiosensor (1994) Biotechnology Handbook. Uppsala:

Amersham Biosciences Picardeau et al. (2000) Microbiology 146: 305-313

Reisenauer et al. (1999) J. Bacteriol 181: 2430-2439

"Remington's Pharmaceutical Sciences", Mack Publication Co., Easton,

PA, latest edition

Sambrook and Russel (2001) Molecular cloning: a Laboratory Manual: Cold Spring Harbor, NY: Cold Spring Laboratory Press

Sauton, MB (1912) CR Acad. Sci. Paris 155: 660-861 Speck et al. (1999) EMBO J 18: 6169-6176 Snapper, et al. (1988) PNAS 85: 6967-6991 Tang et al. (2001) J. Biol. Chem. 276: 19631-19639 Vannelli et al. (2002) J. Biol. Chem. 277: 12624-12629 Wards et al. (1996) FEMS Microbiol Lett 145: 101-105 Pyles et al. (1996) Biochemistry 37: 5194-5200 Jia et al. (1996) J. Biol. Chem. 271: 30451-30456 Hillisch et al. (2001) Curr. Opin. Struc. Biol. 11: 201-207 Doublie, S. (1997) Methods Enzymol. 276: 523-530 Fraaije et al. (2004) J. Biol. Chem. 279: 3354-3360

Schumacher et al. (2002) Embo J 21 (5): 1210-δ.

Schumacher et al. (2001) Science 294 (5549): 215δ-63.

Orth et al. (2000) Nat Struct Biol 7 (3): 215-9. Bourguet, et al. (2000) Mol Cell 5 (2): 2δ9-9δ.

Potier et al. (2003) Protein Sci 12 (4): 725-33.

Garcia-Rubio et al. (2002) J. Chem. Ecol. 28 (4): 769 -762

McLafferty et al. (1993) Interpretation of Mass Spectra; University Science

Books: Mill Valley, CA, pp 146-149. Blanc et al., (2003) Treatment of tuberculosis: Guidelines for National

Programs. In report WHO / TB / 2003.313, WHO, ed., (Geneva, Switzerland).

Phetsuksiri et al., (2003) J. Biol. Chem. 278: 53123-30

Claims

1. Method for selecting a compound capable of binding to the mycobacterial repressor EthR, this bond preventing said EthR repressor from repressing the expression of the mycobacterial gene ethA encoding an EthA activator of ethionamide, said method comprising at least: a) cloning the gene encoding said EthR repressor on a first recombinant expression plasmid, and cloning at least the promoter and operator region of the ethA gene upstream of a reporter gene on a second recombinant expression plasmid; b) transformation of a recombinant host cell with at least the plasmids according to step a); c) culturing said recombinant host cell in the presence of a candidate compound; d) measuring the N- _t level of expression of the reporter gene in the presence of said candidate compound; e) comparing the level Ni measured in step d) with the level of expression of the reporter gene in the absence of said candidate compound; and f) if Ni is significantly greater than N ₀ , the selection of the compound. 2. Method for selecting a compound capable of binding to the mycobacterial repressor EthR, this bond preventing said repressor EthR from repressing the expression of the mycobacterial gene ethA coding for an EthA activator of ethionamide, said method comprising at least: a) cloning of the gene encoding said EthR repressor, and cloning of at least the promoter and operator region of the ethA gene upstream of a reporter gene, on a recombinant expression plasmid; b) transformation of a recombinant host cell with at least the plasmid from step a); c) - the culture of said recombinant host cell in the presence of a candidate compound; d) measuring the level Ni of expression of the reporter gene in the presence of said candidate compound; e) comparing the level Ni measured in step d) with the level N ₀ of expression of the reporter gene in the absence of said candidate compound; and f) if Ni is significantly greater than No, the selection of the compound. 3. Method according to claim 1 or 2, characterized in that said promoter and operator region of the ethA gene comprises at least the sequence SEQ ID No. 1, preferably the sequence SEQ ID No. 2, more preferably the sequence SEQ ID # 3. 4. Method according to claim 1 or 2, characterized in that the recombinant expression plasmid containing said gene encoding the repressor EthR is pMV261-e # 7R. 5. Method according to claim 1 or 2, characterized in that said reporter gene is chosen from lacZ, xylE, gfp, lux and the gene encoding glucuronidase. 6. Method according to claim 1 or 2, characterized in that the recombinant expression plasmid containing at least said promoter and operator region of the ethA gene upstream of a reporter gene is pBSh-Pet _h A - lacZ. 1. Method according to claim 1 or 2, characterized in that said recombinant host cell is a mycobacterial cell, preferably a cell of Mycobacterium bovis, M. smegmatis, M. tuberculosis or M. leprae. δ. Method according to claim 7, characterized in that said cell is a cell of M. bovis BCG 1173P2 or of M. smegmatis mc ² 155. 9. A method of selecting a compound capable of binding to the mycobacterial repressor EthR, this bond preventing said repressor EthR from repressing the expression of the mycobacterial gene ethA encoding an EthA activator of ethionamide, said method comprising at least: a) determining the Li level of binding of the EthR repressor on at least the promoter and operator region of the ethA gene, in the presence of a candidate compound; b) comparing the Li level with the Lo level of binding of the EthR repressor on at least the promoter and operator region of the ethA gene, in the absence of said candidate compound; and c) if Li is significantly less than Lo, selection of the compound. 10. Method according to claim 9, characterized in that said detection is carried out using a method chosen from surface plasmon resonance, detection of natural fluorescence and energy transfer by fluorescence resonance. 11. A method of selecting a compound capable of binding to the mycobacterial repressor EthR, this binding preventing said repressor EthR from repressing the expression of the mycobacterial gene ethA encoding an EthA activator of ethionamide, said method comprising at least: a) bringing together said EthR repressor and at least two candidate compounds; b) crystallization of the mixture thus obtained; c) analysis of the structure of the crystal resulting from step b); d) if this analyzed reveals that one of said compounds is linked to said EthR repressor, the solubilization of said crystal; e) elution and selection of said compound. 12. A method of selecting a compound capable of binding to the mycobacterial repressor EthR, this bond preventing said repressor EthR from repressing the expression of the mycobacterial gene ethA encoding an EthA activator of ethionamide, said method comprising at least: a) bringing together said EthR repressor and a candidate compound; b) crystallization of the mixture thus obtained; c) analysis of the structure of the crystal resulting from step b); d) if this analysis reveals that said compound is linked to said EthR repressor, the selection of said compound. 13. Method according to claim 11 or 12, characterized in that it is implemented prior to a method according to any one of claims 1 to 10. 14. Method according to any one of claims 1 to 13, characterized in that it further comprises a step of determining the structure of the selected compound. 15. Compound inhibitor of the EthR repressor capable of being obtained by the implementation of a selection process according to any one of claims 1 to 14. 16. Compound according to claim 15, characterized in that it interacts with residues F110 and F114 of the repressor EthR. • - 17. Compound according to claim 16, characterized in that it also interacts with the residues W13δ and W145. 1δ. Compound according to any one of claims 15 to 17, characterized in that it is fixed in the binding pocket of the ligand of the repressor EthR, said pocket being defined by the following residues: R128, V134, G124, L76, Q125, F187, F184, T121, W138, F114, L183, E180, M142, N179, N176, W145, F110, W207, T149, 1107, L87, V152, Y148, G106, W103 and M102. 19. Pharmaceutical composition comprising at least one compound inhibiting the EthR repressor as an active ingredient, ^" in combination with at least one pharmaceutically acceptable excipient. 20. Pharmaceutical composition according to claim 19, characterized in that said compound is chosen from analogs and pharmaceutically acceptable derivatives of hexadecyl octanoate, or of benzylacetone. 21. Pharmaceutical composition according to claim 20, characterized in that said compound is chosen from analogs and derivatives which are inhibitors of EthR and which are pharmaceutically acceptable for 2-hexanone, 2-heptanone, 4-heptanone, 2-octanone, 3-octanone, 2-decanone, 2-dodecanone. 22. Products containing a compound that inhibits the repressor EthR and ethionamide or one of its derivatives, as a combination product for simultaneous, separate or spread over time in anti-tuberculosis therapy. 23. Products containing an EthR repressor inhibitor compound and an antibiotic activatable by the mycobacterial activator EthA, as a combination product for simultaneous, separate or spread over time in anti-tuberculosis therapy. 24. Products containing a compound that inhibits the repressor EthR and ethionamide or one of its derivatives, as a combination product for simultaneous, separate or spread over time in leprosy therapy. 25. Products containing an EthR repressor inhibitor compound and an antibiotic activatable by the mycobacterial activator EthA, as a combination product for simultaneous, separate or spread over time in leprosy therapy. 26. Products according to any one of claims 22 to 25, characterized in that said compound is chosen from analogs and pharmaceutically acceptable derivatives of dexadecyl octanoate, or of benzylacetone. 27. Products according to claim 26, characterized in that said compound is chosen from analogs and derivatives which are inhibitors of EthR and which are pharmaceutically acceptable for 2-hexanone, 2-heptanone, 4-heptanone, 2-octanone, 3 -octanone, 2-decanone, 2-dodecanone. 2δ. EthR repressor inhibitor compound as a drug. 29. Compound according to claim 2δ, characterized in that it is chosen from analogs and pharmaceutically acceptable derivatives of dexadecyl octanoate, or of benzylacetone. 30. Compound according to claim 29, characterized in that it is chosen from analogs and derivatives which are inhibitors of EthR and which are pharmaceutically acceptable for 2-hexanone, 2-heptanone, 4-heptanone, 2-octanone, 3 -octanone, 2-decanone, 2-dodecanone. 31. Use of at least one inhibitor compound for the EthR repressor for the manufacture of a medicament intended for the prevention and / or treatment of bacterial infections, preferably mycobacterials. 32. Use according to claim 31, characterized in that said compound is chosen from analogs and pharmaceutically acceptable derivatives of dexadecyl octanoate, or of benzylacetone. 33. Use according to claim 32, characterized in that said compound is chosen from analogs and derivatives which are inhibitors of EthR and which are pharmaceutically acceptable for 2-hexanone, 2-heptanone, 4-heptanone, 2-octanone, 3 -octanone, 2-decanone, 2-dodecanone.