Background of the invention Recently, the prevalence of multidrug resistant human pathogenic bacteria has increased dramatically. This increase correlates with an escalation of bacterial disease and related mortality. Also, anti-infective chemotherapy is becoming more difficult as the percentage of elderly and immunocompromised patients grows. The European Commission has already reacted to this situation by banning the use of certain anti-infective drugs as a growth promotant in livestock feed, among them the streptogramin virginiamycin, so as to limit the environmental spread of anti-infectives (thought to be a major driving force for selection of multidrug resistant pathogenic bacteria), thereby preserving the use of anti- infective compounds for human therapy. Therefore, new technologies are required for the (i) discovery of novel anti-infective drugs with antibacterial activity against multidrug resistant bacteria (Aubel et al., 2001, J. Antibiotic. 54, 44-55) and (ii) detection of anti- infective drugs in food and environmental samples to impose the ban of anti-infectives in stock farming (Kurittu et al., 2000, J. Agric. Food. Chem. 48, 3372-7).

Recently a method has been described how anti-infective compounds of the streptogramin class can be identified by using specific repressor proteins which bind to specific DNA operator sequences in an anti-infective molecule-dependent way (Aubel et al., 2001, J. Antibiotic. 54, 44-55, US Patent No. 6, 287, 813). In this specific setting, the Streptomyces coelicolor-derived PIP protein is transferred into mammalian cells, where it binds to a PIR-DNA sequence and thereby blocks transcriptional activity of a functionally linked mammalian promoter. However, in the presence of compounds of the streptogramin class (anti-infective molecules), these molecules bind the PIP protein and release it from its cognate PIR-DNA sequence, thereby derepressing the functionally linked promoter, which in turn leads to the expression of a reporter gene, the activity of which can be monitored. Therefore, expression of the reporter gene indicates the presence of anti-infective molecules of the streptogramin class. This setting can be used,

either to detect anti-infective molecules of the streptogramin class (e. g. virginiamycin, a growth promotant used in stock farming) in environmental samples or it can be used to discover new streptogramin-based molecules in chemical or biological libraries.

Patent application WO 00/65082 describes use of MarR peptides that bind to the marO (multiple antibiotic resistance) operon in E coli for identifying compounds interfering with antibiotic resistance. In a recent review (Grkovic et al., 2002, Microbiol. and Molec. Biol.

Reviews 66 (4), 671) it was indicated that anti-infective compounds interfering with MarR/marO binding do so over a phosphorylation requiring functional cells, but that a large variety of other compounds such as salicylate interact with MarR/marO in a cell-free system. A screen for tetracycline antibiotics using a cell system based on the tetR/tetO interaction and ß-galactosidase detection was described by Kirsch et al. (1991, J.

Antibiotics 44 (2), 210). In the recently published patent application WO 03/025184 a method of gene regulation based on the E protein/ETR binding partner interaction is described. The E protein/ETR interaction may also be used in a solid phase screen for antibiotics.

Summary of the invention The present invention is directed to a method for the in vitro detection, quantification and discovery of anti-infective compounds by measuring the change in interaction of a represser protein with a poiynudeotide-based protein binding partner in response to the addition of known or potential anti-infective compounds in a cell-free environment. The invention is further directed to in vitro assays useful in such a method, e. g. in vitro assays wherein the protein binding partner is immobilized on a solid support or in vitro assays wherein the interaction of the protein with its protein binding partner is measured in solution, and to kits for such cell-free in vitro assays. The invention is also directed to novel anti-infective compounds detected by using the method of the invention, and to methods of identifying new proteins and protein binding partners useful in said method.

Brief description of the Figures Figure 1: Setup of a cell-free in vitro anti-infective compounds assay using a solid support.

The operator DNA sequence designated"Biosensor-Operator" (BSO) is immobilized on a streptavidin (St)-coated solid support via a biotin molecule covalently linked to the operator DNA sequence (BSO). The repressor protein designated"Biosensor" (BS) is

bound to the BSO sequence. In the absence of an anti-infective compound (-A), the repressor protein BS remains bound to the BSO sequence and is subsequently detected with a primary antibody directed against the repressor protein BS (or a tag fused to BS) and a secondary antibody directed against the primary antibody. The secondary antibody is coupled to a horseradish peroxidase (HRP), which is used to generate a color reaction (C) by addition of chromogenic compounds, e. g. 3,3', 5, 5'-tetramethylbenzidine (TMB) in the presence of hydrogen peroxide. However, in the presence of an anti-infective compound (+ A) the repressor protein BS is subsequently released of its BSO binding partner. In the next washing steps (W), the repressor protein BS is eliminated and its absence prevents binding of the primary antibody and thus the subsequent color reaction.

Figure 2: Dose-response characteristics of the solid support-based cell-free in vitro anti- infective compounds assay described in Example 1 following addition of increasing doses of tetracycline. The readout is given in optical density at 450 nm (OD 450 nm).

Figure 3: Readout of the solid support-based cell-free in vitro anti-infective compounds assay described in Example 2 with (+) and without (-) addition of the streptogramin pristinamycin. The readout is given in optical density at 450 nm (OD 450 nm).

Figure 4: Readout of the solid support-based cell-free in vitro anti-infective compounds assay described in Example 3 with (+) and without (-) addition of erythromycin. The readout is given in optical density at 450 nm (OD 450 nm).

Description of the invention For the purpose of the invention, a"polynucleotide-based protein binding partner"or "protein binding partner"is a molecule which binds a particular repressor protein in an anti-infective compound-dependent manner. The protein binding partner may be a naturally occurring polynucleotide sequence or a polynucleotide sequence derived from a naturally occurring polynucleotide sequence.

For the purpose of the invention, a"repressor protein"is a protein which changes its affinity to a polynucleotide-based protein binding partner in response to the presence of an anti-infective compound. A repressor protein may be identical, derived from or related to proteins naturally encoded by prokaryotes or eukaryotes, either on the chromosome or an episome.

By a polynucleotide-based protein binding partner"derived from"a naturally occurring polynucleotide sequence is meant, in this context, that the polynucleotide sequence of the protein binding partner contains modified nucleotides, base changes, modified bases and/or modified linkages between bases (e. g. phosphorothionate linkages), but still can bind the repressor protein in an anti-infective compound-dependent way.

By a repressor protein"derived from"naturally occurring proteins is meant, in this context, that the amino acid sequence of the repressor protein contains amino acid substitutions, preferably conservative amino acid substitutions, but remains at least 45%, preferably 60%, and more preferably 80% or more identical to the naturally occurring protein at the amino acid level.

By a repressor protein"related to"naturally occurring proteins is meant, for purposes of the invention, that the polynucleotide sequence which encodes the amino acid sequence of the repressor protein hybridizes to a naturally occurring polynucleotide sequence encoding a naturally occurring protein under at least low stringency conditions, more preferably moderate stringency conditions, and most preferably high stringency conditions.

"Conservative substitution"is known in the art and is described e. g. by Dayhof, M. D., 197$, blat. Biomea'. Res. Found., Washington, 1D ol. 5, Su, o. 3. Genetically encoded amino acids are generally divided into four groups : (1) acidic = aspartat and glutamate ; (2) basic = lysine, arginine, and histidine; (3) non-polar = alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, and tryptophan ; and (4) uncharged polar = glycine, asparagine, glutamin, cysteine, serine, threonin, and tyrosine. Phenylalanine, tryptophan and tyrosine are also jointly classified as aromatic amino acids. A substitution in a protein of one amino acid classified in a particular group with another amino acid in the same group is generally regarded as a conservative substitution.

The present invention is directed to a method for the in vitro detection, quantification and discovery of anti-infective compounds by measuring the change in interaction of a repressor protein with a polynucleotide-based protein binding partner in response to the addition of known or potential anti-infective compound in a cell-free environment.

Useful in the method of the invention is any pair of repressor protein and polynucleotide- based protein binding partner which binds to each other in the absence of a particular anti-infective compound or class of anti-infective compounds, and wherein the binding interaction is substantially weakened or interrupted in the presence of a particular anti- infective compound or class of anti-infective compounds. However, the reverse situation, wherein a repressor protein binds its polynucleotide-based protein binding partner in the presence of an anti-infective compound, and wherein the binding interaction is substantially weakened or interrupted in its absence, is also within the scope of the invention.

Such a weakening or interruption of binding interaction may be the result of competitive binding of an anti-infective compound or compound class to the repressor protein or polynucleotide-based protein binding partner, but preferably is based on a interaction of the repressor protein with the anti-infective compound or compound class leading to a conformational change in the repressor protein which weakens or interrupts its binding interaction with the polynucleotide-based protein binding partner.

Particular pairs of repressor protein and polynucleotide-based protein binding partner are, for example, PIP protein and PIR binding partner (Fussenegger et al., 2000, Nat.

Biotechnol. 18,1203-8), being responsive to anti-infective compounds of the class of streptogramins, in particular pristinamycin (pristinamycin I and II), or the clinically applicable pristinamycin derivative Synercid, a combination of dalfopristin and quinupristin, virginiamycins A and B, and the like ; TetR protein and teto binding partner (Gossen and Bujard, 1992, Proc. Natl. Acad. Sci. USA, 89, 5547-51), being responsive to anti-infective compounds of the class of tetracyclines, in particular tetracycline, chlortetracycline, oxytetracycline, demeclocycline, methacycline, doxycycline, minocycline, and the like ; E protein and ETR binding partner (Weber et al., 2002, Nat.

Biotechnol, 20,901-7), being responsive to anti-infective compounds of the class of macrolides, in particular erythromycin, clarithromycin, azithromycin, tylosin, roxithromycin, oleandomycin, and the like ; the repressor protein encoded by the tcmR gene of Streptomyces glaucescens and the polynucleotide-based protein binding partner represented by a specific DNA sequence of the tcmA/tcmR promoter (Guilfoile et al., 1992, J. Bacteriol. 174,3651-58), being responsive to polyketide anti-infectives like tetracenomycin; the Streptomyces lividans TipAL protein and the ptipA DNA binding partner (Chiu et al., 1999, J. Biol. Chem. 274,20578-86), being responsive to thiostrepton anti-infective compounds like promothiocin B, geninthiocin, berninamycin A, thiostrepton

A, promoinducin and the like ; the rtTA protein and the tetO binding partner (Gossen et al., 1995, Science, 268,1766-69), being responsive to anti-infective compounds of the tetracycline class like tetracycline, oxytetracycline, chlortetracycline, anhydrotetracycline and doxycycline ; the VarR protein and the varS promoter region binding partner (Namwat et al., 2001, J. Bacteriol. 183,2025-31), being responsive to anti-infective compounds of the streptogramin class like virginiamycin S and the like ; the ScbR protein and the OscbR operator site, being responsive to compounds of the butyrolactone class (Takano et al., 2000, J. Biol. Chem. , 275,11010-16), the QacR protein and the nucleotide binding partner located in the qacA transcription start site (Grkovic et al., 1998, J. Biol. Chem. 273, 18665- 73), being responsive to anti-infective compounds like malachite green, rhodamine and tetraphenylphosphonium ; and the BmrR protein and the bmrR promoter region (Ekaterina et al., 2001, Nature 409, 378-82), being responsive to lipophilic cationic anti-infective compounds like TPP or TPSb.

However, the invention is not limited to these mentioned repressor proteins and polynucleotide-based protein binding partners. For example, other repressor proteins like RifQ, Orgy, DnrO, Actil and LanK (US Patent No. 6, 287, 813) are also within the scope of this invention and can be used according to the methods described herein. The invention also extends to any repressor protein and the respective polynucleotide-based protein binding partners which may be found by one of the methods described hereinafter.

New repressor proteins can be identified, for example, by sequence database searching for proteins derived from or related to known repressor proteins using e. g. a BLAST (Altschul et al., 1997, Nucleic Acids Res. 25, 3389-402) computer program. New polynucleotide-based protein binding partners can be found, for example, in promoters that drive anti-infective drug resistance genes, which can be identified by generating bacterial gene libraries in a heterologous host, and growing the heterologous host in the presence of particular anti-infective compounds. Emerging resistant clones containing anti-infective resistance determinants can be analyzed for putative repressor protein binding sites, for example, either by sequence comparison or by using the putative repressor protein binding site to identify the corresponding repressor protein as described hereinafter.

New repressor proteins can be isolated by binding to known protein binding partner DNA sequences or to protein binding partner DNA sequences discovered, for example, by the method described hereinbefore. For example, the polynucleotide with the protein binding

partner DNA sequence is immobilized on a matrix and ideally packed in a column.

Bacterial extracts are applied to the column under conditions which allow repressor protein derived proteins and related proteins to bind to the immobilized target sequence.

Following appropriate washing steps, the repressor protein-related-or repressor protein- derived proteins are eluted under suitable conditions, for example by addition of particular anti-infective compounds, and the sequence of the purified protein determined and the corresponding gene cloned.

Other methods to identify new repressor proteins comprise, for example, hybridization of labeled oligonucleotides encoding known repressor proteins with genomic libraries of a collection of unknown or known bacteria. Genomic sequences hybridizing to oligo- nucleotides encoding known repressor proteins are subcloned, and the putative new repressor protein analyzed.

The invention therefore also relates to a method of identifying a new repressor protein, characterized in that a protein related to or derived from a known repressor protein is reacted with the polynucleotide-based protein binding partner to the known repressor protein, and the change of interaction between said protein and the protein binding partner on addition of an anti-infective compound is measured, and to a method of identifying new repressor proteins and polynucleotide-based protein binding partners, characterized in that promoters that drive anti-infective drug resistance genes in bacteria are isolated, immobilized and reacted with bacterial extracts, the anti-infective drug is added, and the new repressor protein thereby released and the immobilized promoter being a polynucleotide-based protein binding partner are isolated.

The repressor protein is chemically synthesized or expressed in a suitable host and purified according to standard techniques, e. g. using a Ni-column, as described in the Examples. Likewise the polynucleotide-based protein binding partner is synthesized according to standard techniques, e. g. as outlined in the Examples.

In a particular pair of repressor protein and polynucleotide-based protein binding partner, the naturally occurring repressor protein may be replaced by a repressor protein derived from the naturally occurring repressor protein as defined hereinbefore, or by a repressor protein related to a naturally occurring repressor protein as defined hereinbefore. Such derived or related repressor proteins are e. g. the variants of the tetracycline-responsive repressors TetR (A), TetR (B), TetR (C), TetR (D), TetR (E), TetR (G) and TetR (H)

(Schnappinger et al., 1998, EMBO J. 17,535-543) which bind to their cognate protein binding partner polynucleotide sequence tetO in response to the addition and withdrawal of anti-infective compounds of the tetracycline class as listed hereinbefore. Those repressor proteins share between 45 and 80% sequence identity on the amino acid level.

Further to that, in a particular pair of repressor protein and polynucleotide-based protein binding partner, the naturally occurring polynucleotide-based protein binding partner may be replaced by a protein binding partner derived from a naturally occurring protein binding partner as defined hereinbefore. Such derived polynucleotide-based protein binding partners are e. g. the modified polynucleotide sequences, which allow binding of TetR- derived or related proteins in response to the addition and withdrawal of anti-infective compounds of the tetracycline class as listed hereinbefore (Baron et al., 1999, Proc. Natl.

Acad. Sci. USA 96, 1013-18).

The derived or related repressor proteins are likewise chemically synthesized, modified or obtained by expression of DNA coding for a derived or related protein in a suitable host and purified according to standard techniques. Likewise the derived polynucleotide-based protein binding partner is synthesized or modified according to standard techniques.

In one aspect of the invention, the repressor protein or the polynucleotide-based protein binding partner is immobilized on a solid support. Solid supports considered are, for example, glass surfaces such as glass slides ; microtiter plates, membranes or beads consisting of nitrocellulose, nylon, polyvinylidene fluoride (PVDF) or functionalized polymers; chemically modified oxidic surfaces, e. g. silicon dioxide, tantalum pentoxide or titanium dioxide; or also chemically modified metal surfaces, e. g. noble metal surfaces such as gold or silver surfaces.

In one preferred embodiment the polynucleotide-based protein binding partner is immobilized on a solid support e. g. using members of a specific binding pair, one member being fused to the polynucleotide-based protein binding partner and the other member being attached or attachable to the solid support, either covalently or by any other means.

A specific binding pair considered is e. g. biotin and avidin or streptavidin. Further examples of binding pairs are e. g. maltose and maltose binding protein, glutathione and glutathione-S-transferase (GST), any hapten and a specific antibody such as digoxigenin and anti-DIG-IgG, poly-histidines and immobilized nickel ions, complementary oligo- nucleotide sequences, which specifically bind through base pairing, and the like. The

polynucleotide-based protein binding partner may also be immobilized using reactive substituents allowing chemoselective reaction between such substituent with a complementary functional group on the surface of the solid support. Examples of such pairs of reactive substituents and complementary functional group are e. g. amine and activated carboxy group forming an amide, azide and a propiolic acid derivative undergoing a 1, 3-dipolar cycloaddition reaction, amine and another amine functional group reacting with an added bifunctional linker reagent of the type of activated bis- dicarboxylic acid derivative giving rise to two amide bonds, a sulfhydryl and a maleimide group that yields a thioether linkage upon coupling, any enzyme, which forms a covalent and irreversible linkage with its substrate like the 06-alkylguanine-DNA-alkyltransferase (AGT) together with its cognate substrates (06-benzylguanine derivatives) or the thymidilate synthase, which irreversibly binds fluorodeoxyuridylate in the presence of methylenetetrahydrofolate, or other combinations known in the art.

The solid support carrying the polynucleotide-based protein binding partner is then loaded with an optionally suitably labelled corresponding repressor protein. The repressor protein is added to the immobilized polynucleotide-based protein binding partner and the corresponding interaction is monitored after addition of a known or a potential anti- infective compound. A change in the binding state between repressor protein and polynucleotide-based protein binding partner indicates the presence of an anti-infective compound, and may also be used to quantify the amount of a known anti-infective compound. After reaction with the known anti-infective compound or the suspected anti- infective compound the supernatant is decanted from the solid support and labelled repressor protein is detected directly, such as by surface plasmon resonance (e. g.

Biacore0 technology) or quartz microbalances (e. g. rupture event scanning), or based on the properties of the label either in the supernatant or still bound to the solid support through interaction with the polynucleotide-based protein binding partner.

Examples of suitable labels for the repressor protein include a spectroscopic probe such as a fluorophore, a chromophore, a magnetic probe or a contrast reagent; a radioactively labelled molecule ; a molecule which is one part of a specific binding pair which is capable of specifically binding to a partner; a biomolecule with desirable enzymatic, chemical or physical properties; or a molecule possessing a combination of any of the properties listed above.

When the label is a fluorophore, a chromophore, a magnetic label, a radioactive label or the like, detection is by standard means adapted to the label. If the label is a member of a specific binding pair, the other member is preferably attached or attachable to a solid support, an enzyme, fluorophore, chromophore, radioactive label or the like, either covalently or by any other means. A specific binding pair considered is again e. g. biotin and avidin or streptavidin, or maltose and maltose binding protein, the bindings pairs mentioned before or the like. When the label is a biomolecule with desirable enzymatic, chemical or physical properties, the labelled repressor protein is detected by standard methods to detect such enzymatic, chemical or physical properties using methods known in the art, e. g. color reaction observable on enzymatic reaction with a convenient enzyme substrate.

Alternatively, a repressor protein may be detected by an antibody directed to an epitope on the repressor protein. The antibody to this epitope may carry a detectable further label such as described hereinbefore, or may be detected by a second antibody binding to the first antibody and carrying a detectable further label as described hereinbefore.

As an example, the polynucleotide protein binding partner fused to biotin is immobilized on a solid support like a 96-well microtiter plate coated with streptavidin. Subsequently the repressor protein is added, allowing its binding to the protein binding partner poly- nucleotide sequence. To this setup, different possible anti-infectives-containing test compound mixtures can be added, and the presence of an anti-infective compound results in disruption of the repressor protein/protein binding partner polynucleotide sequence interaction. Following a washing step, unbound repressor protein is eliminated.

Subsequently, the bound repressor protein is detected using a primary antibody directed against the repressor protein or against a molecule covalently or non-covalently linked to the repressor protein. The primary antibody is detected by addition of a secondary antibody labeled with an enzyme like a peroxidase to enable a subsequent enzymatic reaction for quantification. In this setup, a high enzyme activity corresponds to the absence of anti-infective compounds in the test mixture. However, upon anti-infectives presence in the test compound, the amount of repressor proteins bound to the protein binding partner polynucleotide sequence is decreased, leading to an attenuated enzyme activity after completion of the assay. The corresponding setup is shown in Figure 1.

In another preferred embodiment, the repressor protein is immobilized on a solid support by technologies known in the art as described hereinbefore for the immobilization of

polynucleotide-based protein binding partner. The polynucleotide-based protein binding partner optionally carrying a label as described hereinbefore for labeling the repressor protein is added to the immobilized repressor protein and the corresponding interaction is monitored after addition of a known or a potential anti-infective compound. A change in the binding state between repressor protein and polynucleotide-based protein binding partner indicates the presence of an anti-infective compound, and may also be used to quantify the amount of a known anti-infective compound. After reaction with the known anti-infective compound or the suspected anti-infective compound the supernatant is decanted from the solid support and the optionally labelled polynucleotide-based protein binding partner is either detected directly such as by surface plasmon resonance (Biacore0) or quartz microbalances (e. g. through rupture event scanning), or based on the properties of the label or using an antibody against an epitope of the polynucleotide- based protein binding partner as described hereinbefore.

In another aspect of the invention, the interaction of the repressor protein with its polynucleotide-based protein binding partner is directly monitored in solution, for example by analyzing the conformation of the repressor protein or its polynucleotide-based protein binding partner by methods known in the art like spectroscopic or resonance techniques.

Additionally or alternatively, at least one of the binding partners can be covalently or non- covalently linked to at least one other molecule for monitoring the interaction between repressor protein and polynucleotide-based protein binding partner.

As an example the repressor protein is incubated together with its polynucleotide-based protein binding partner in an appropriate buffer system. The conformation of the repressor protein is analyzed by spectroscopic methods known in the art, e. g. CD/OD spectra. Upon addition of a compound having anti-infective properties, the conformation of the repressor protein is changed, subsequently resulting in a different spectrometric readout.

Alternatively, the repressor protein or its polynucleotide-based protein binding partner can be labeled with fluorophores suitable for Fluorescence Resonance Energy Transfer (FRET). Therefore, a change in the binding state, as induced by addition of a compound with anti-infective properties, can easily be monitored by a changed fluorescence behavior. Possible fluorophores for FRET analysis comprise FITC, FAM, Cy5, Cy3 and the like or fluorescent and luminescent proteins like green fluorescent protein (GFP) and luciferase (luc). Other known fluorescence techniques for measuring molecule-molecule interactions, e. g. fluorescence quenching, are also within the scope of this invention.

Alternatively, the interaction of the repressor protein with its polynucleotide-based protein

binding partner can be monitored by measuring the size of the corresponding protein- poylnucleotide complex, e. g. with standard size exclusion chromatography, capillary electrophoresis or by electromobility shift assays (EMSA). Other methods known in the art for assaying DNA-protein interactions like those described by Pavski and Le (Pavski, V. and Le, X. C. , 2003, Curr. Opin. Biotechnol. 14,65-73) and the references cited therein are also applicable for determination of the corresponding protein-polynucleotide interactions in response to anti-infective compounds.

The invention is further directed to cell-free in vitro assays useful in the method of the invention as described hereinbefore, e. g. in vitro assays wherein the protein binding partner is immobilized on a solid support or in vitro assays wherein the interaction of the protein with its protein binding partner is measured in solution, and to kits useful in such an assay. In vitro assays in solution are particularly considered for the method using E protein and ETR binding partner.

A kit according to the invention comprises e. g. a solid support coated with a polynucleotide-based protein binding partner, repressor protein (labelled or non-labelled) in solid form or in solution, and optionally reagents for detecting the repressor protein, buffer solutions, materials for handling the solid support, repressor proteins and the reagents, and a description how to use the components of the kit. In another embodiment, the kit comprises a solid support coated with a repressor protein, polynucleotide-based protein binding partner (labelled or non-labelled) in solid form or in solution, and optionally reagents, solutions and materials as described hereinbefore. In a further embodiment for use in an assay in solution, the kit contains repressor protein and polynucleotide-based protein binding partner, and further optional components as described hereinbefore.

The invention is also directed to novel compounds detected by using the method of the invention. Such novel compounds may belong to a known class of therapeutic substances like anti-infectives, e. g. macrolides, streptogramins, polyketides, thiostreptones or tetracyclines, or may represent a member of a new class with the same or other possible therapeutic applications, e. g. with antibiotic or cytostatic activity against mammalian cells.

In particular, a test solution containing a compound suspected of being such a therapeutic substance of a known or new class is brought into contact with a pair of repressor protein and its polynucleotide-based protein binding partner as described hereinbefore, and the change of interaction is measured. Depending on the pair of repressor protein and polynucleotide-based protein binding partner, the substantial weakening or interruption of

the binding interaction (or a newly arising binding interaction) between repressor protein and polynucleotide-based protein binding partner is indicative of a novel compound of a class of therapeutic substances, e. g. in the anti-infectives area.

The invention having been described, the following examples are offered by way of illustration and not limitation.

Example 1: Detection of tetracycline in vitro by using TetR-His and tetO_ DNA immobilization on a solid support.

Cloning of a prokaryotic vector for expression of a TetR-His6 fusion protein (hexahistidine- tagged TetR, Gossen and Bujard, 1992, Proc. Natl. Acad. Sci. USA, 89,5547-5551) : The TetR fusion protein encoded on plasmid pSAM200 (US Patent No. 6,287, 813) is amplified with primers OWW302 (5'-CGGAATTCCCACCATGCATATGTCTAGATTAGATAAAAG-3') and OWW303 (5'-GCTCTAGAGCAAGCTTTTAATGGTGATGGTGATGATGGGATCCAC GCGGAACCAGACCGGACCCACTTTCACATTT-3') and pSAM200 as a template, and cloned (Ndel/Hindlil, partial digestion) into a pRSET T7-polymerase-driven expression vector (Zisch et al., 2001, J. Control. Release 72, 101-113). This vector is used to transform E coli BL21*. Following transformation, the bacteria are grown in LB medium in the presence of appropriate selection antibiotics (100 µg/ml ampicillin, 34 ug/ml chlor- amphenicol) up to an optical density (600 nm) of Ode9= 1. 2, followed by an IPTG (1 mM)-induced expression for another 3-4 h. The cells are harvested by centrifugation and stored over night at-80°C. The following day cell lysis is performed by lysozyme treatment (0.2 mg/ml) for 45 min at 37°C, and the DNA is fragmented by sonication (5 bursts, 75W, 2 seconds).

ELISA plates (Corning, NY, Cat. No. 3590) are coated over night at 37°C, 250 rpm with 0. 2 ug streptavidin in 100 pi water per well before blocking for 2 h in block buffer (blocking reagent No. 1'096'176, Roche Molecular Biochemicals, Rotkreuz, Switzerland, 1% in TBS, pH 7.6, 10 mM EDTA). The blocking solution is withdrawn and the plates are washed 4 times with TBST (20 mM Tris/HCI, pH 7.6 ; 150 mM NaCl, 0.05% Tween20). All subsequent incubations are performed at room temperature and 350 rpm.

The polynucleotide-based protein binding partner in this example is the tetO7 encoding polynucleotide sequence, which is bound by the TetR-His6 protein in a tetracycline- dependent way. For immobilization of the tet07encoding polynucleotide sequence a

biotinylated tet07-containing DNA fragment is constructed by PCR using the biotinylated primer (Microsynth, Balgach, Switzerland) OWW64 (Biotin-5'- GGGGTTCCGCGCACATTTCCCC-3') and the unbiotinylated primer OWW22 (5'- GCTAGAATTCCGCGGAGGCTGGATCGG-3') and plasmid pMF111 (US Patent No.

6,287, 813) as template (pMF111 encodes the tet07 sequence between the two primer annealing sites). The PCR product is purified using a PCR purification kit (Roche Molecular Biochemicals, Rotkreuz, Switzerland, No. 1'732'668). This biotinylated tetO7- containing DNA fragment (0. 2 ug/well) is applied in 100 ul block buffer for 1 h. After washing with TBST 30 ul crude cell extract containing TetR-His6 fusion protein (approx.

0.5 pg total protein) in a total of 100 pi block buffer are added per well and incubated for 1 h at 23°C. The plates are washed and filled with 100 ui/well TBS (20 mM Tris/HCI, pH 7.6 ; 150 mM NaCI), 1 % bovine serum albumin (BSA), 70 mM MgCl2 before addition of different amounts of tetracycline. After washing, the wells are incubated with a 1: 200 dilution of monoclonal mouse anti-His6 antibody (Novagen, Madison, WI) in block buffer.

After 1 h the plates are washed and the secondary antibody (anti mouse IgG x HRP, Amersham Life Science, NJ) is applied at a 1 : 200 dilution in block buffer without EDTA.

The color reaction is performed by addition of 100 pl/well TMB substrate solution (Sigma, St. Louis, MO, No. T-8665) and stopped by addition of 25 u11 M H2SO4. Absorbance is read at 450 nm.

The optical readout for different tetracycline concentrations is shown in Figure 2.

Example 2 : Detection of the streptogramin pristinamycin in vitro by using PIP and PIR DNA immobilization on a solid support The in vitro assay for the detection of the streptogramin pristinamycin is performed as illustrated in Example 1 except for the following modifications : The streptogramin-sensing protein PIP (US Patent No. 6, 287, 813) fused to a hexahistidine tag for detection by anti- His6-antibodies is expressed with an expression vector constructed by PCR using primers OWW300 (5'-CGGAATTCCCACCATGCATATGAGTCGAGGAGAGGTGCGCAT-3') and OWW301 (5'-GCTCTAGAGCAAGCTTTTAATGGTGATGGTGATGATGGGATCCACGC GGAACCAGACCGGCCTGTTCGACCATCGCGT-3') and plasmid pMF150 (US Patent No. 6,287, 813) as a template. The PCR fragment is ligated (NdellHindlll) into a T7 polymerase driven pRSET-based expression vector (see Example 1). The PIP-His6 fusion protein is expressed according to the same protocol as the TetR-His6 protein described in Example 1. The biotinylated polynucleotide-based protein binding sequence PIR is

constructed by PCR using the same oligos as for the Biotin-tetO7 sequence (Example 1), except that plasmid pBP33 is used as a template, which encodes the PIR binding sequence between the two primer annealing sites.

Different concentrations of the streptogramin pristinamycin are added. For the detection of the PIP-His6 fusion protein a monoclonal mouse anti-His6 antibody (Novagen, Madison, WI) is used. Figure 3 shows the optical density with and without the addition of the streptogramin pristinamycin.

Example 3: Detection of the macrolide erythromycin using E-Hiss and ETR DNA immobilization on a solid support.

The cell-free in vitro assay for the detection of macrolide anti-infective compounds is performed as illustrated in Example 1 except for the following modifications: The macrolide-sensing protein E (Weber et al., 2002, Nat. Biotechnol. 20,901-7) fused to a hexahistidine tag (E-His6) for detection by anti-His6-antibodies is expressed with an expression vector constructed by PCR using primers OWW61 (5'-CGGAATTCCCACCATGCATATGCCCCGCCCCAAGCTCAAG-3') and OWW307 (5'-GCTCTAGAGCAAGCTTTTAATGGTGATGGTGATGATGGGATCCACGCGGAAC CAGACCCGCATGTGCCTGGAGGAGTTGGAA-3') and plasmid pWW35 (Weber et al., 2002, Nat. Biotechnol. 20, 901-7) as a template. The PCR fragment is listed (NdeI/HindIII) into a T7 polymerase driven pRSET-based expression vector (see Example 1). The E-His6 fusion protein is expressed according to the same protocol as the TetR- His6 protein described in Example 1. The biotinylated repressor protein binding sequence ETR is constructed by PCR using the same oligos as for the Biotin-tet07sequence (Example 1), except that plasmid pWW37 (Weber et al., 2002, Nat. Biotechnol. 20,901-7) is used as a template, which encodes the ETR binding sequence between the two primer annealing sites.

Different concentrations of erythromycin are added in block buffer. For the detection of the E-His6 fusion protein a monoclonal mouse anti-His6 antibody (Novagen, Madison, WI) is used. Figure 4 shows the optical density with and without the addition of the macrolide erythromycin.

Example 4: Anti-infective compound detection in solution by using spectroscopic methods A bacterial expression vector (pWW312) for hexahistidine-tagged PIP protein (the repressor protein in the streptogramin-dependent PIP/PIR interaction) is constructed as described in Example 2. The hexahistidine-tagged PIP protein is expressed according to Example 1 and the bacterial lysate is subjected to affinity chromatography on a Ni2+- loaded metal chelation resin (Novagen, Cat. No. 69670) according to the manufacturers protocol. The eluted hexahistidine-tagged PIP protein (PIP-His6) is dialyzed against TBS (20 mM Tris/HCI, pH 7.6, 150 mM NaCI) for 24 h with buffer exchange after 12 h at 4°C.

The purified PIP-His6 protein is concentrated on a 5 kDa cut-off spin column and subjected to fluorescence labeling using FITC (fluorescein-isothiocyanate) with subsequent size exclusion chromatography to remove unbound fluorescent dye. The PIP- specific PIR DNA sequence is synthesized as fluorescein-labeled oligonucleotide and hybridized with its complementary strand. The fluorescently (FITC) labeled PIP-His6 protein is now incubated with the fluorescent double stranded PIP sequence in the presence of increasing doses of the streptogramin pristinamycin 1. The binding between the PIP-His6 protein and PIR sequence is monitored through fluorescence resonance energy transfer (FRET) between the fluorescein fluorophores bound to the DNA and the protein.