WO2008046584A2 - Lepa as a target for antibacterial agents - Google Patents

Abstract

The present invention relates to methods, substances and compositions for retarding or inhibiting cell growth by inhibiting the activity of ribosomal factor LepA. Further, methods and kits for identifying antibacterial substances are disclosed.

Description

LepA as a target for antibacterial agents

Description

The present invention relates to methods, substances and compositions for retarding or inhibiting cell growth by inhibiting the activity of ribosomal factor LepA. Further, methods and kits for identifying antibacterial substances are disclosed.

The increasing bacterial resistance against antibiotics is a well-recognized problem in medicine and veterinary medicine worldwide. In most cases, antibiotic resistance in bacterial pathogens was identified soon after the introduction of antibiotics into clinical practice and occurs due to mutation or exogenous acquisition of appropriate genes. Whereas mutations leading to clinically significant levels of resistance are rather rare, mutations resulting in low levels of resistance to antibiotics are more frequent and represent an initial step in the development of the high-level-resistance strains found today. However, in the majority of cases, acquisition from exogenous sources represents the primary mechanism by which bacteria obtain genes encoding resistance to antibiotics. Upwards of a dozen different biochemical mechanisms of resistance have been characterized to date which are encoded by hundreds of different genes. By these means, the bacterial population has essentially developed a considerable armamentarium of genetic defenses against antibiotics [for review see Mazel and Davies, 1999]. Accordingly, there is a continuous need in the art to identify new targets for antibiotics active against bacteria.

One of such targets could be elongation factors involved in ribosomal mechanisms. The functions of ribosomes can be separated into four phases, each of which is governed by specific protein factors. The four phases are initiation, elongation and termination of protein synthesis, and the recycling phase, during which the ribosomes are dissociated into their subunits so that the small subunit is ready to re-enter the subsequent initiation phase [for review see Nierhaus and Wilson, 2004]. The details of the translation phases differ significantly between ribosomes from the three domains of life, viz. bacteria, archaea and eukarya, with the exception of the elongation phase. The elongation phase is at the heart of protein synthesis and consists of a cycle of reactions (hence elongation cycle), during which the nascent polypeptide chain is prolonged by one amino acid. The elongation cycle is governed by two universal elongation factors, termed EF-Tu and EF- G in bacteria, and EF-1 and EF-2 in archaea and eukarya. EF-Tu transports an aminoacyl-tRNA (aa-tRNA) in the ternary complex aa-tRNA^»EF-Tu^»GTP to the decoding center of the ribosomal A-site (A for aminoacyl-tRNA) on the small ribosomal subunit. After decoding EF-Tu hydrolyses GTP and leaves the ribosome as EF-Tu-GDP, whereas cognate aa-tRNA accommodates fully into the A-site. The next step, peptide-bond formation, does not require a translation factor. During this process, the peptidyl-residue at the ribosomal P-site (P for peptidyl-tRNA) is cleaved off of the peptidyl-tRNA and transferred to the aminoacyl-tRNA, with the result that now the peptidyl- tRNA resides at the A-site and is prolonged by one amino acid. The third step in the elongation cycle is the translocation reaction that is promoted by EF-G-GTP: The (tRNA)2^»mRNA complex is shifted by a codon length on the ribosome, moving the peptidyl-tRNA from the A to the P-site and the deacylated tRNA from the P to the E-site (E for exit). In higher fungi such as yeast and Candida albicans, a third elongation factor, EF-3, has been identified as being essential for viability. EF-3 is an ATP dependent E-site factor because ATP hydrolysis by EF-3 is necessary to open up the E-site enabling the EtRNA to be released upon A-site occupation [Triana-Alonso et al., 1995a].

In most bacteria, the lepA gene is the first cistron of a bi-cistronic operon. The second cistron, the leader peptidase or lep gene, encodes the signal peptidase Lep [March and Inouye, 1985a] - an integral membrane protein that is inserted into the inner cell membrane and cleaves off the N-terminal signal (leader peptide) from secreted and periplasmic proteins [Zwizinski and Wickner, 1980]. Dibb and Wolfe reported that a LepA knock-out in E. coli has no phenotype under the various growth conditions tested [Dibb and Wolfe, 1986]. Curiously, these null-results contrast with the fact that LepA is one of the most conserved proteins known in biology (Genebank Swiss-Prot: LepA from Escherichia coli: Entry name LEPA_ECO57; Primary accession number 60787; Genebank UniProt/TrEMBL: LepA orthologue from human: Entry name Q5XKM8JHUMAN; primary accession number Q5XKM8; protein name: FLJ 13220). Actually, after EF-Tu (in archaea and eukarya EF-1A) LepA is the second most conserved protein known with an amino-acid identity of 48 to 85% [Caldon et al., 2001].

In PCT/EP2006/006503 the present inventors describe methods, systems, compositions and kits for the synthesis of proteins in vitro, wherein the protein synthesis is carried out in the presence of LepA in order to improve the accuracy of protein synthesis. In conducting further studies on ribosomal factor LepA, it has has now surprisingly been found that LepA represents a third essential bacterial elongation factor with a novel function in translation, namely to induce "back-translocation" of mis-translocated tRNAs on the ribosome, that rationalizes the high conservation of this factor. On the other hand, it is demonstrated in the present application that a knock-out of the lepA gene is lethal under conditions of high ionic strength and that overexpression of LepA protein is toxic to the cell. In view of these results, it has been concluded that LepA (EF-4) is an elongation factor and thus represents a new target for antibiotics active against bacteria.

Thus, a first aspect the present invention relates to a method for retarding or inhibiting the growth of cells expressing ribsosomal factor LepA by inhibiting the activity of LepA.

In another aspect the present invention relates to a pharmaceutical composition comprising at least one substance inhibiting the activity of ribosomal factor LepA. - A -

In still another aspect the present invention relates to the use of a substance inhibiting the activity of ribosomal factor LepA for the manufacture of an agent for retarding or inhibiting bacterial cell growth.

In still another aspect the present invention relates to a method for identifying antibacterial substances, comprising the steps: a) providing a substance to be tested for its antibacterial properties b) providing a sample of bacterial cells c) bringing into contact the substance with the sample of bacterial cells, and d) determining whether the substance inhibits the activity of ribosomal factor

LepA in the bacterial cells.

In still another aspect the present invention relates to a kit for identifying antibacterial substances, comprising at least one sample of bacterial cells and means for detecting whether activity of LepA is inhibited by a tested substance.

In still another aspect the present invention relates to a method for preventing or treating diseases associated with bacterial pathogens, comprising administering to a patient in need thereof an effective amount of a pharmaceutical composition comprising at least one substance inhibiting the activity of ribosomal factor LepA, the pharmaceutical composition optionally additionally comprising pharmaceutical acceptable carrier means, dilution means or/and adjuvants.

According to the first aspect of the invention a method of retarding or inhibiting the growth of cells expressing ribsosomal factor LepA is provided, wherein retardation or inhibition of cell growth is effected by inhibiting the activity of LepA. In a preferred embodiment, the cells expressing ribsosomal factor LepA are prokaryotic cells. More preferably, the cells expressing ribsosomal factor LepA are bacterial cells which may be Gram-positive or Gram-negative and which comprise, for example, cells from Bordetella, Borellia, Brucella, Campylobacter, Chlamydia, Clostridium, Corynebacteria, Enterococcus, Escherichia coli, Haemophilus, Klebsiella, Legionella, Listeria, Mycobacteria, Mycoplasma, Neisseria, Proteus, Pseudomonas, Rickettsia, Salmonella, Shigella, Staphylococcus, Streptococcus, Treponema, Vibrio or Yersinia.

According to the present invention, the activity of LepA may be inhibited by a number of mechanisms. In a preferred embodiment, inhibition of the activity of LepA is effected by altering the regulation of LepA in the cell, blocking the binding of LepA to a ribosome or/and blocking the enzymatic activity of ribosomal factor LepA. As used herein, the term "regulation" refers to the control of the amount and timing of appearance of a functional gene product in a cell, which functional gene product may be a protein or an RNA molecule. On the other hand, the term "enzymatic activity" is used herein to mean the catalytic effect exerted by an enzyme, which is given by the moles of a substrate converted per time unit.

Preferably, the regulation of LepA in the cell is altered by inhibiting the synthesis of LepA in the cell or/and inhibiting the release of LepA from the cell membrane into the cytoplasm. As is decribed hereinafter, the cell membrane contains about 90% of the total cellular LepA and is therefore assumed to act as a storage reservoir for fast delivery of LepA. The molar ratio of LepA to 70S ribosomes in the cytoplasm of bacterial wildtype cells is about 0.1 :1, i.e. one LepA molecule per 10 ribosomes. In a preferred embodiment, inhibition of LepA yields a molar ratio of ribosomal factor LepA to 70S ribsomes of less than 0.05:1 in the cytoplasm, more preferably less than 0.01 :1.

Alternatively, the regulation of LepA in the cell may be altered by overexpressing LepA. An overexpression of LepA in the cell may be achieved by means known to a person skilled in the art and comprises, for example, transforming the cell with a plasmid containing a suitable promotor and a nucleotide sequence encoding the molecule of interest. Preferably, the overexpression of LepA yields a molar ratio of ribosomal factor LepA to 70S ribsomes of more than 0.25:1 in the cytoplasm, more preferably more than 0.5:1.

The binding of LepA to the ribsome may be blocked, for example, by blocking the binding-site of LepA on the ribosome or/and blocking the amino acid sequence(s) of LepA binding to the ribosome. In a preferred embodiment, the binding-site of LepA on the ribosome is blocked.

In another preferred embodiment, the enzymatic activity of LepA is blocked by blocking back-translocation of tRNAs present on a ribosome from the

POST state to the PRE state or/and blocking GTPase activity. As is decribed hereinafter, LepA was shown to exhibit a ribosome-dependent GTPase activity at least as strong as that of elongation factor EF-G, hitherto the strongest ribosome-dependent GTPase known, and to be responsible for back-translocating tRNAs present on a ribosome from the POST state to the

PRE state, thus being capable of correcting decoding errors induced at the ribosomes under certain conditions such as increased Mg²⁺ concentrations.

In a second aspect, the present invention provides a pharmaceutical composition comprising a substance inhibiting the activity of ribosomal factor LepA. In a preferred embodiment, the substance inhibiting the activity of ribosomal factor LepA alters the regulation of LepA, blocks the binding of

LepA to a ribosome, or/and blocks the enzymatic activity of LepA. According to the invention, this may be effected in particular by a substance inhibiting the synthesis of LepA in the cell, a substance inhibiting the release of LepA from the cell membrane into the cytoplasm, a substance blocking the binding-site of LepA on the ribosome or/and an inhibitor of GTPase activity.

Substances suitable for inhibiting the activity of ribosomal factor LepA comprise in particular antibodies which may be monoclonal or polyclonal. For the manufacture of antibodies a number of different hosts such as goats, rabbits, rats or mice may be immunised by injecting a protein or a suitable fragment or an oligopeptide thereof exhibiting immunogenic properties. In order to enhance the immunological response, different adjuvants may be used depending on the host species. In this context, peptides, fragments or oligopeptides having an amino acid sequence of preferbaly at least five amino acids, more preferably at least ten amino acids, are used to induce the production of antibodies to the protein.

According to the present invention, monoclonal antibodies are particularly preferred. Monoclonal antibodies may be produced using techniques which provide for the production of antibody molecules by means of continuous cell lines. These techniques comprise the hybridoma technique, in particular the human B-cell hybridoma technique and EBV hybridoma technique. The production of monoclonal antibodies by fusion of spleen cells derived from immunised mice and myeloma cells was described by Kohler and Milstein in 1975 (..Continuous cultures of fused cells secreting antibody of predefined specifity", Nature (1975), 256, 475-497). Techniques for the chemical selection of the hybridomas resulting from such fusion, as well as for subsequent isolation of cell clones secreting the distinct antibodies are known in the art.

Moreover, methods for the preparation of genetically produced antibodies such as chimeric antibodies may be employed. For this purpose, preferably constant regions of a murine antibody are replaced by constant regions of a human antibody. Alternatively, it is possible to use methods which allow the preparation of single-chain antibodies. These are obtained by expressing a construct of the gene segments of both variable antibody regions which are connected by a segment for the peptide.

According to the present invention it is also possible to employ a fragment of an antibody instead of using the full length antibody. As used herein, the term ..fragment" is meant to refer to any fragment of an antibody which maintains the antigen-binding function of the antibody. Such fragments are, for example, Fab, F(ab')₂, Fv, ScFv, as well as other fragments such as CDR-fragments (..complemetary determining region", hypervariable region) and fragments produced by means of a Fab expression library. Those fragments provide the binding specifity observed for the antibody and may be produced recombinantly by known methods. The F(ab')₂ fragments may be obtained by digesting the antibody molecule with pepsin, whereas Fab fragments are available by reducing the disulfide bridges of the F(ab')₂ fragments or by digesting the antibody molecule with papain. Alternatively, a Fab expression library may be constructed in order to enable a rapid and simple identification of monoclonal antibodies showing the desired specifity.

Alternatively, substances suitable for inhibiting the activity or ribosomal factor LepA are low molecular compounds that block the binding of GTP specifically to LepA or, alternatively, prevent LepA conformational changes triggered by GTP or GDP binding or by GTP hydrolysis on LepA, as has been shown for e.g. fusidic acid in the case of EF-G.

In a preferred embodiment, the pharmaceutical composition additionally comprises pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers. Additives which are appropriate for this purpose comprise, for example, detergents, solvents, antioxidants and preservatives. Dilution means suitable for use in such a pharmaceutical composition preferably comprise aqueous NaCI solution, lactose solution, mannitol solution, as well as water and alcohols. Suitable buffers comprise, without being intended to be limited to, TRIS, HCI, glycine and phosphate. An overview with regard to substances useful for such pharmaceutical compositions is given, for example, in the latest edition of Remington's Pharmaceutical Sciences (Maack Publishing Co., Easton, Pa.).

In another aspect, the present invention provides the use of a substance inhibiting the activity of ribosomal factor LepA for the manufacture of an agent for retarding or inhibiting bacterial cell growth. Substances which are particularly suitable to inhibiting the activity of LepA are those described above. In a preferred embodiment, the agent is formulated as a pharmaceutical composition which may optionally contain pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers. Suitable additives, dilution means and buffers comprise substances as decribed above or mentioned in Remington's Pharmaceutical Sciences (Maack Publishing Co., Easton, Pa.).

Preferably, the agent formulated as a pharmaceutical composition is for preventing or treating diseases associated with bacterial pathogens. In particular, these diseases are selected from the group consisting of infectious diseases and comprise, for example, adnexal infections, angina, anthrax, bacterial meningitis, bronchitis, brucellosis, campylobacteriosis, cat scratch disease, cholera, diphtheria, epidemic typhus, gonorrhea, gynecological infections, impetigo, legionellosis, laryngitis, leprosy, leptospirosis, listeriosis, Lyme borreliosis, melioidosis, MRSA infection, nocardiosis, pertussis, pharyngitis, plague, pneumococcal pneumonia, psittacosis, Q fever, rhinitis, Rocky Mountain spotted fever, salmonellosis, Scarlet fever, shigellosis, sinusitis, syphilis, tetanus, trachoma, tuberculosis, tularemia, typhoid fever, typhus and urinary tract infections.

In still another aspect, the present invention provides a method for identifying antibacterial substances. In this method it is determined whether a test substance is capable of inhibiting the activity of ribosomal factor LepA in the bacterial cells. The test substance may be derived from a chemical library of substances. The method may be a High Throughput Screening

Method wherein a plurality of test substances is screened in parallel. A substance which exhibits a significant inhibition of the activity of ribosomal factor LepA is a suitable candidate antibacterial agent.

In a preferred embodiment, the method is a molecular screening method or cellular screening method which allows determining the effect of a test substance on the LepA-activity with a suitable detection technology. The range of assay technologies supported for formatting molecular screens may include AlphaScreen, time resolved fluorescence (DELPHIA, and LANCE), fluorescence polarisation, steady-state fluorescence, photometry, chemiluminescence, ELISA, scintillation proximity, and filtration-based separations. For cell-based screens, supported assays may include reporter genes (luciferase, fluorescent proteins, alkaline phosphatase, beta- galactosidase), BRET (protein-protein interactions), or assays measuring biochemical responses such as cell-surface antigen expression, cytokine expression, cell proliferation and cytotoxicity.

In a preferred embodiment, the substance to be tested is formulated as a pharmaceutical composition which may optionally contain pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers. Suitable additives, dilution means and buffers comprise substances as decribed above or mentioned in Remington's Pharmaceutical Sciences (Maack Publishing Co., Easton, Pa.).

In still another aspect the present invention provides a kit for identifying antibacterial substances, the kit comprising at least one sample of bacterial cells and means for detecting whether activity of LepA is inhibited by a tested substance.

Furthermore, the present invention is to be explained in greater detail by the examples and figures hereinbelow.

Description of Figures

Figure 1

Phylogenetic analysis of LepA (EF-4) and EF-G. An unrooted NJ (Neighbor Joining algorithm) tree of EF-G and LepA proteins enables reciprocal rooting of each subfamily. Branch lengths reflect the estimated amino acid substitutions per every site (see scale bar). Numbers on internal branches indicate statistical support of clades based on 1000 bootstrap samples. Note that the total branch lengths of both GTPase families are comparable. In both families mitochondrial eukaryotic proteins are monophyletic. Fly EF-G and LepA proteins branch with their animal orthologs (confirmation of the coelomata hypothesis). A prospective secondary LepA protein of Arabidopsis branches with chloroplasts. There are no signs of inter-domain lateral gene transfer in the LepA family.

Figure 2

In silico analysis of LepA (EF-4) compared with EF-G. A, domain structures of LepA and orthologs in comparison to EF-G. LepA (E. coli) and orthologs Guf1 (yeast mitochondria) and Q5XKMB (human mitochondria) in red, corresponding EF-G domains in different colors. LepA has five potential structural domains (I, II, III, IV and LepA_C) according to the amino acid sequence (2-595), which have high consensus with E. coli EF-G domains I (purple), Il (blue), III and V (yellow). B, alignment of E. coli LepA with EF-G from T. thermophilus. Black and gray boxes indicate amino-acid identity and similarity, respectively. G' subdomain (blue) and domain IV (orange) are lacking in LepA, whereas LepA contains a specific C-terminal domain (red). C, comparison of the crystal structure of the ternary complex aa-tRNA*EF- Tu-GTP (PDB1TTT) with that of EF-G (PDB1WDT) and a homology model for E. coli LepA. The domains of EF-G are indicated with roman numerals, except the G' subdomain of domain I. Note that the EF-G domain IV corresponds to the anticodon stem-loop (ASL) of the aa-tRNA within the ternary complex and that LepA lacks the G' domain and domain IV but has a LepA specific C-terminal domain (CTD).

Figure 3 Growth curves of E. coli strains. A, overexpression of LepA (EF-4) blocks growth. The strain BL21(DE3) containing the plasmid (pET+LepA) stops growth soon after IPTG induction (red, BL21_LepA) in contrast to wild type or overexpression of EF-G (BL21_EF-G). The yellow curve shows the growth of BL_LepA without induction. The arrows indicate the addition of IPTG (1 mM). After 10-fold dilution in the presence of IPTG the strain overexpressing LepA does not resume growth in contrast to wild type and the strain overexpressing EF-G. B, growth of the strain with knock-out of the lepA gene. Wild type strain MG 1655 (circles) in LB medium without and with 200 mM K⁺ or 100 mM Mg²⁺ (blue, yellow and red, respectively) show normal growth after a tenfold dilution in LB medium, whereas the strain with knocked-out lepA gene (triangles) shows growth defects already in LB medium, whereas in the presence of 200 mM K⁺ or 100 mM Mg²⁺ cells are early turning into the stationary phase and growth does not resume after a ten-fold dilution in LB medium indicating that the strain is not viable at this ionic strength.

Figure 4 Localization and quantification of LepA in wild type cells MG 1655 after growth in LB medium. Controls, defined amounts of LepA and proteins from 70S ribosomes were loaded onto SDS-PAGE and developed with antibodies against LepA and the ribosomal protein L2, respectively. The intensities of the bands were used to assess the amounts of LepA and ribosomes in the S30 and membrane fractions, from which a series of equivalent amounts were loaded onto the same gel. The obtained values were used to determine the distribution of LepA and ribosomes between the two fractions as well as the molar ratio LepA/70S in either fraction.

Figure 5

In vitro assays with LepA (EF-4). A, uncoupled ribosome-dependent GTPase activity of EF-G (green) and LepA (red). The GTPase activity is given as GTP hydrolysed in 5 min per ribosome, and plotted as a function of the molar ratio of factor to ribosome. B, tRNA binding and puromycin reaction with ribosomal complexes in different functional states (shown schematically) before and after incubation with LepA and GTP. The data are normalized to reactions per one ribosome (v). LepA does not influence the amounts of bound tRNAs (upright bars), but it abolishes the puromycin reaction (hanging bars) of the POST state rather than that of the Pi state. Red tRNAs and columns, Ac[¹⁴C]Phe-tRNA; blue tRNAs and columns, [³²P] tRNA_f ^Me'. As a control, the experiment was also performed in the presence of non-labelled AcPhe-tRNA, added either to the (i) PRE state (no chasing of the Ac[¹⁴C]-Phe-tRNA at the A-site was observed: 0.65 and 0.69 Ac[¹⁴C]Phe- tRNA at the A-site per ribosome in the absence and presence of non- labelled AcPhe-tRNA), or (ii) during back-translocation, when the non- labelled AcPhe-tRNA was added to the purified POST state before addition of LepA to induce back-translocation (0.56 and 0.54 of Ac[¹⁴C]Phe-tRNA per

5 70S in the absence and presence of non-labelled AcPhe-tRNA after back- translocation). C, puromycin reaction of the PRE state after incubation with various amounts of EF-G^»GTP (green curve) and the POST state after incubation in the presence of various amounts of LepA with GTP and GDPNP (red and orange, respectively). In the presence of GDPNPo quantitative blockage of the puromycin reaction is only achieved at a molar ratio LepA:70S = 1 :1. whereas with GTP full blockage is already seen at a ratio of (0.3-0.4):1. D, comparison of kinetics of EF-G dependent translocation and LepA-dependent back-translocation. The reactions were performed at 3O⁰C under single turnover conditions in the presence ofs GDPNP (0.5 mM; molar ratios of EF-G and LepA to 70S were 5:1 and 10:1 , respectively).

Figure 6

LepA (EF-4) induces back-translocation. A, primer extension analysis of0 DMS modified 16S rRNA from various ribosomal complexes in the absence (-) or presence (+) of LepA^«GTP. The band of the diagnostic A-site tRNA footprint at A1408 of 16S rRNA is indicated by an arrow. A₁ C denote dideoxy-sequencing lanes. Quantification of the DMS reactivity at A1408 in different ribosomal complexes is shown below the gel. The DMS reactivity at5 A1408 in the empty ribosome (70S) was taken at 1.00. Values shown represent the mean and the standard deviation of two independent DMS probing experiments. B, primer extension analysis of Pb²⁺ cleaved 23S rRNA. The cleavage efficiency at C2347 of 23S rRNA (arrow) was monitored in various ribosomal complexes in the absence (-) or presence (+) ofo LepA-GTP. The characteristic cleavage enhancement at C2347 in the POST state disappears upon LepA addition. A, C denote dideoxy-sequencing lanes. The Pb²⁺ cleavage efficiency of vacant ribosomes (70S) was taken as 1.00. Values shown represent the mean and the standard deviation of two independent Pb²⁺ cleavage experiments. C, toeprint assay with PRE and POST states. The PRE state (lane 1) was translocated with EF-G and GTP and the resulting POST state was purified by pelleting through a sucrose cushion. The purified POST state was then either toeprinted directly (lane 2) or after an incubation in the presence of LepA and GTP (lane 3). The relative amounts of the PRE and POST states given in percent were obtained by scanning the respective bands.

Figure 7 LepA (EF-4) effects on GFP synthesis in a coupled transcription-translation system in vitro. A, addition of various amounts of LepA. One aliquot of the reaction mixture was applied to a native gel and the fluorescence measured (upper panel, amount of active GFP), a sister aliquot was developed in an SDS-gel (middle panel, total amount of GFP). The total amount of GFP synthesized in the absence of LepA was designated as 100%, the relative amounts of total and active GFP determined and the active fraction calculated (lower graph). B, GFP synthesis as in A, but in the presence of various concentrations of the aminoglycoside paromomycin. Left panel, no LepA, right panel, in the presence of LepA (0.3 mole per mole 70S). C, same as B, but with increasing Mg²⁺ concentrations. "0" indicates the intrinsic Mg²⁺ concentration of 12 mM.

Figure 8

Model for LepA (EF-4) function. A, Under optimal growth conditions the translocation has a very low rate of error and therefore LepA is not so important under such conditions. Translocation involves the movement of tRNAs at the A and P-sites (PRE state) to the P and E-sites (POST). This reaction is catalysed by elongation factor G (EF-G, blue) and GTP. After dissociation of EF-G, the A-site is now free for binding of the next ternary complex aminoacyl-tRNA^»EF-Tu^»GTP (blue tRNA to blue A-site codon), which leads to release of the E-tRNA (cyan). B, In the rare case that EF-G malfunctions, a defective translocation complex may result. This is likely to occur more frequently under conditions of high ionic strength. The consequences of the defective translocation complex are two-fold: (i) ribosomes may incorrectly display the A-site codon, allowing binding of near-cognate ternary complexes and therefore misincorporation, as illustrated by right-hand pathway (binding of green tRNA to blue codon). (ii) Under extreme conditions the ribosome may even become stuck, thus precluding continued translation. The defective translocation state is recognized by LepA^»GTP (red), which induces a back-translocation, allowing EF-G a second chance to catalyse a correct POST state. In this way LepA reduces translational errors and relieves stuck ribosomes.

Examples

The materials used in the present invention were those as described in Marquez et al., 2004, the disclosure of which is herewith incorporated by reference.

Example 1 : Protein sequence and homology analysis

Database searches for orthologs of E. coli and yeast LepA and EF-G were carried out using BLASTP with standard parameters and protein databases of organisms with completely sequenced genomes that were downloaded from the Integrδ web site. Orthology was assigned based on best reciprocal hits. Additionally, the INPARANOID database and the NCBI non-redundant protein database were screened for additional homologs that were tested for orthology by phylogenetic analysis. Subcellular localization of proteins was predicted using CHLOROP and MITOP. Multiple alignments were constructed using the MAFFT software for each common structural domain of EF-G and LepA separately because domain order is only partially conserved between these proteins. Only regions with sufficient sequence similarity for unambiguous alignment were considered. Alignments for individual domains were concatenated and used for phylogenetic analysis. Phylogenetic analysis was carried out with the MEGA software (version 3.1). Pairwise sequence distances were obtained by Maximum Likelihood estimation on the basis of the JTT substitution rate matrix with the assumption of a uniform distribution of rates across sites. Phylogenetic trees were reconstructed using the Neighbor Joining algorithm. Statistical support values for internal branches of the tree were obtained from 1000 bootstrap samples and their analyses. Trees were calculated on a reduced set of organisms (Figure 1 ) and on a large set of organisms (data not shown). The essentials of the phylogenetic tree already emerged during the analysis of the smaller data set. Due to the higher number of taxa the second data set offers an enhanced resolution of internal branching patterns in the bacterial subtree and confirms results from the first data set.

The homology model for E. coli LepA (Figure 2C) was generated based on the sequence alignment (Figure 2B) and the crystal structure for Thermus thermophilus EF-G (PDB1WDT) using the Protein Homology / analogY Recognition Engine (PHYRE) (http://www.sbg.bio.ic.ac.uk/~phyre/). Figure 2C was created using the PyMOL Molecular Graphics System (2002) from DeLano Scientific, San Carlos, CA₁ USA. (http://www.pymol.org).

Example 2: Generation of LepA-strains

E. coli lep A gene was cloned from genomic DNA using PCR primers that introduce Ndel and BamHI restriction sites for cloning into the expression vector pET14b (Novagen). The cultures of E. coli BL21(DE3) strain, or this strain transformed with either pET14b or pET+LepA were grown overnight with 150 rpm shaking at 37°C. Cells were diluted 1 :200 and grown for 2-3 h at 37°C. When the optical density reached an >A₅₈o of -0.4, the cells were induced with 1 mM IPTG. The LepA knock-out E. coli K-12 strain "b2569

LepA-" derived from the wildtype of MG 1655 was obtained from E. coli

Genome project (University of Wisconsin-Madison https://asap.ahabs.wisc.edu/asap/query_strains.php). The cultures of wildtype or knock-out strain were grown overnight in LB medium with 150 rpm shaking at 37°C. Cells were diluted 1 :200 with LB medium or LB plus

0.1 M MgCI₂ and grown for 6-7 h at 37°C until they grew in stationary phase for at least 1 h. Then, cells were diluted 1 :10 with LB medium. Example 3: Preparation of purified components for in vitro assays

Re-associated 70S ribosomes were prepared according to Blaha et al., 2000, the disclosure of which is herewith incorporated by reference. MF- mRNA described in Triana-Alonso et al., 1995b, and encoding Met-Phe (sequence: GGG(A4G)3AAAAUGUUC(A4G)3AAAU] was prepared according to Schafer et al., 2002, the disclosure of which is herewith incorporated by reference. EF-Tu and EF-G with C-terminal His-tags were isolated from E. coli as described previously for EF-Tu [Boon et al., 1992, the disclosure of which is herewith incorporated by reference] with the following changes: the cells were induced by a cell density of 0.5 A₅eo and incubated further for 4 h. The cells were pelleted and resuspended (1 ml/g) in a buffer containing 20 mM Hepes^«KOH (pH 7.6 at 0 ⁰C), 60 mM NH₄CI, 7 mM MgCI₂, 7 mM β-mercaptoethanol and 1 mM phenylmethylsulfonyl fluoride (PMSF), and in addition in the case of EF-Tu 50 μM GDP. The cells were disrupted with the microfluidizer (model M-110L; Microfluidics). After a centrifugation step (30,000 x g for 45 min) the supernatant was treated according to method 2 in Boon et al., 1992 (EF-Tu elution from the Ni²⁺- column at 80 mM and EF-G at 250 mM imidazole). After final dialysis against a buffer containing 20 mM Hepes-KOH (pH 7.6 at 0 ⁰C), 6 mM MgCI₂, 150 mM KCI, 1 mM DTE, 10 μM GDP and 10% glycerol. Crude or specific tRNAs were purchased from Sigma and charged according to Marquez et al., 2004. tRNA binding and dipeptide assays and translation of model-mRNAs were performed as described [Dinos et al., 2004, the disclosure of which is herewith incorporated by reference], with the final conditions of the standard buffer used for these experiments: 20 mM Hepes- KOH (pH 7.6 at 0⁰C), 4.5 mM Mg(acetate)₂, 150 mM NH₄acetate, 4 mM β- mercaptoethanol, 2 mM spermidine and 0.05 mM spermine.

Immuno-analysis of the LepA protein location in E. CoIi K12 MG 1655 was performed in order to detect the distribution of LepA in vivo. The overnight culture (in LB medium) was diluted 1 :200 and grew until -0.6 A600, harvested by centrifugation for 5 min at 4⁰C and 2000xg. The cells were resuspended in Quiagen cell resuspension buffer with 1 mg/ml lysozyme and freeze-and-thaw 3 times. The S-30 (cell lysate) fraction and membrane fraction were separated by centrifugation for 30 min at 4°C and 18.000xg. The membrane fraction was dissolved in protein loading buffer to yield the same volume as the S-30 fraction. Samples of 10μl of S30 fraction and 10μl

5 of membrane fraction were mixed with protein loading buffer and heated at 95°C for min before subjecting to a gel-electrophoresis. Controls with known amounts of purified LepA were also applied to the gel. A Western-blot analysis followed using standard techniques with the help of LepA polyclonal antibodies. The amount of LepA in the membrane and lysate fraction waso assessed by comparison with the known amounts of LepA present on the control lanes.

Example 5: GTPase activity

The GTPase assays were as described previously [Connell et al., 2003, thes disclosure of which is herewith incorporated by reference] except that reactions were set up to maintain the condition of the standard buffer with a final ribosome concentration of 0.2 μM, a final protein concentration of 0.02- 0.2 μM and [γ-³³P] GTP concentration of 50 μM. 0 Example 6: Preparation of defined ribosomal complexes

Pi complexes, Pre-translocational (PRE) and Post-translocational (POST) complexes were made as described previously [Marquez et al., 2004]. The Pi complex consisted of re-associated 70S ribosomes programmed with MF- ITIRNA, an Ac[¹⁴C]Phe-tRNA^Phβ in the P-site. The PRE complex consisted of5 re-associated 70S ribosomes programmed with MF-mRNA, a pPJdeacyl- tRNA_f ^Mβ' in the P-site and Ac[¹⁴C]Phe-tRNA^Phβ in the A-site, and subsequently were translocated by EF-G to yield the POST complexes. The complexes (1 ml) were then sedimented through sucrose cushion (I ml 10% sucrose in standard buffer) at 65,000 x g for 18 hours, 4°C in a TL-100 ultracentrifugeo (Beckman) to remove non-bound mRNA and tRNA in the case of Pi and PRE complexes, or EF-G in the case of POST complexes.

Example 7: Puromycin assay Defined ribosomal complexes (0.2 μM) in the standard buffer were incubated with or without 0.06 μM LepA and 250 μM GTP at 37°C for 10min. A puromycin reaction followed as described previously [Marquez et al., 2004].

Example 8: Structural probing

Prior to tRNA footprinting, 5 pmol of ribosomal complexes were incubated in 16.3 μl standard buffer at 37°C for 10min. Chemical probing with dimethyl sulfate (DMS) at 0⁰C for 30 min was performed as described [Bayfield et al., 2001 , the disclosure of which is herewith incorporated by reference]. Modification with 1 -cyclohexyl^-morpholino-carbodiimidemetho-p- toluensulfonat (CMCT) was initiated by the addition of 8.15 μl CMCT solution (84 mg/ml standard buffer) and was performed at 37°C for 15 min. Pb(OAc)₂ cleavage of 5 pmol ribosomal complexes was performed in 18 μl standard buffer for 5 min at 25°C as described [Polacek et al., 2000, the disclosure of which is herewith incorporated by reference]. Primer extension products of modified rRNAs [Polacek and Barta, 1998, the disclosure of which is herewith incorporated by reference] were separated on 6% polyacrylamide gels and quantified using a Molecular Dynamics Storm Phosphorlmager.

Example 9: Toe-Print

The following mRNA was used:

GGCAAAGGAGGUAUUAOOAAUGUUCAAACGAUCAAUCUACGUAUAA

£Kζτg which contains a Shine-Dalgarno sequence (bold underlined) and codes for MFKSIRYV (bold italic). The mRNA was annealed to a ³²P-5'-end-labeled primer (shaded region) as described previously [Hartz et al., 1988, the disclosure of which is herewith incorporated by reference], and then used to program ribosomes for pre- (PRE) and post-translocation (POST) complexes. Briefly, 200 pmol re-associated 70S were incubated with 5 pmol mRNA:primer and 400 pmol of each tRNA_f ^Mef and Ac-PhetRNAPhe in standard buffer. Aliquots of the reaction mixture with 5 pmol 70S were withdrawn before and after EF-G dependent translocation reaction and used for toeprinting assays. The remaining post-translocational mixture of 275 μl was centrifuged through a 1 ml 10% sucrose cushion in standard buffer (65,000 x g for 18 h). The pellet was resuspended in 90 μl standard buffer and aliquotized in 15 μl portions.

7.5 pmol POST complexes in 15 μl were incubated for 30 min at 37⁰C with 5 times excess of LepA and 200 times excess GTP (0.1 mM) and used for the toeprinting assay. The end-labeled primer on the mRNA was extended by 100 units of MuMLV reverse transcriptase (Fermentas) in the presence of dNTPs each 135 μM in standard buffer at 37⁰C for 15 min. The reaction was stopped by 20 μl of loading buffer (9 M Urea, 90 mM TRIS₁ pH 8.3 at room temperature, 90 mM boric acid, 15 mM EDTA, 0.05% xylene cyanol, 0.05% bromophenol blue) and heated at 95⁰C for 5 min. Toeprint reactions were analyzed on 8% urea-PAGE (8 M urea). The gels were quantified using a Molecular Dynamics Phosphor Imager.

Example 10: Coupled transcription-translation assay and quantification of fidelity The assay and quantification were described previously [Dinos et al., 2004] except that each reaction volume of 10 μl contained 0.1 μl of the plasmid solution with the GFP gene after the T7 promoter (plVEX2.2-GFPcyc3; 1 μg/μl) and 1.4 μl with LepA or/and antibiotics.

Results

In sHico analyses of LepA: Conservation and domain structure

LepA is one of the most highly conserved proteins known; the amino-acid identity of LepA among bacterial orthologs ranges from 55 to 68%, which compares well with the corresponding values for EF-Tu₁ EF-G and IF-2 that are 70-82%, 58-70% and 35-49%, respectively (see Table 1 ).

Table 1

Database searches revealed that LepA orthologs could be found in all bacteria and nearly all eukaryotes. It is only missing in eukaryotes that have lost mitochondria and have only retained mitochondrial remnants without ribosomes like Encephalitozoon cuniculi or Giardia lamblia [Knight, 2004]. In all plants with completely sequenced genomes (rice, mouse-ear cress, and red algae) two forms of LepA were found. Whereas one form branches with other mitochondria LepA sequences in the phylogenetic analysis according to this invention, the second form branched with cyanobacterial orthologs, indicating its subcellular targeting to chloroplasts in plants (see Figure 1). This suggests that LepA is essential for bacteria, mitochondria and plastids.

As well as having high conservation with EF-G, LepA also exhibits a conspicuous similarity in terms of the domain structure with EF-G in that it contains equivalents to EF-G domains I to V, with the exception that domain IV is absent (see Figure 2A and B). In addition, LepA has a unique C- terminal domain (CTD). This domain arrangement of LepA is found in bacteria and mitochondria from yeast to human. Due to the high conservation between EF-G and LepA, it is possible to generate a homology model for LepA based on the known EF-G structure. From the representation seen in Figure 2C, it is obvious that LepA lacks the G' subdomain of EF-G domain I as well as the complete domain IV.

In vivo analyses of LepA The E. coli lepA gene was cloned into pET14b, transformed into a E. coli BL21(DE3) strain and thus expression of LepA could be induced with IPTG (see Experimental Procedures). Even without induction, a prolonged lag phase was observed, and the viable cells (BL21_l_epA in Figure 3A) turn earlier into the stationary phase than the control expression of EF-G. Induction of LepA expression by addition of 1 mM IPTG severely affected growth, indicated by an early entry into stationary phase, whereas overexpression of EF-G as a control shows a growth behavior similar to wildtype. SDS-PAGE analysis reveals "leaky" expression of uninduced BL21_LepA and strong expression of both induced LepA and EF-G (data not shown). After a 10-fold dilution in the presence of IPTG the strain overexpressing LepA hardly grows at all, whereas the control EF-G strain again shows almost wildtype growth (Figure 3A). These results demonstrate that overexpression of LepA is toxic to the cell.

Although a strain where the lepA gene has been chromosomally disrupted is viable [Dibb and Wolfe, 1986], the generation time of this knock-out strain is slightly retarded (from 30 min to 40 min), even in rich LB medium (Figure 3B). Furthermore, this effect can be reproduced after 10-fold dilution of cells that have reached the stationary phase (Figure 3B). When the growth was repeated in LB medium, but now in the presence of 200 mM K⁺, the wildtype strain (circles in 3B) grew normally after a short adaptation phase and maintained wildtype-like growth patterns after a ten-fold dilution (20 mM K⁺ final concentration). In striking contrast, cells bearing an inactivated lepA gene (triangles in 3B) exhibited a serious growth defect in the presence of the high potassium or magnesium (200 and 100 mM, respectively); growth completely ceased after 5 hours and moreover it did not resume following tenfold dilution (Figure 3B). A detailed analysis revealed that magnesium concentrations of ≥20 mM or potassium concentrations of ≥200 mM in the LB medium were lethal for E. coli strains that lack LepA (data not shown). It was therefore concluded that LepA is essential for cell viability under conditions of high ionic strength. In vitro analyses of LepA: Distribution of LepA

In order to determine the distribution of LepA and ribsomes between the membrane and cytoplasm fractions in a wild-type E. coli strain, the presence of LepA and and the presence of ribosomes was monitored using a polyclonal antibody raised against E. coli LepA and antibodies to the large subunit ribosomal protein L2, respectively. In this context, 90% of the total cellular LepA was found to be present in the membrane fraction, whereas only 10% was detected in the cytoplasm (see Figure 4). In contrast, ribosomes were equally distributed between both fractions, leading to a (0.3- 0.4):1 ratio of LepA:70S ribosomes in the cell. However, in the cytoplasmic fraction the lower amount of LepA results in a ratio of 0.1 :1 , i.e. one LepA molecule per 10 ribosomes. For comparison, the molar ratio of EF-G and EF-Tu is roughly 1 :1 and 10:1 , respectively [Gordon, 1970; Furano, 1975].

It is assumed that the membrane-bound LepA functions as a storage reservoir for fast delivery of LepA under conditions when the intracellular ionic strength is suddenly increased. The membrane-bound LepA thus seems to belong a unique regulation system designed for fast tuning the free LepA concentration in the cell. On the other hand, commercially available transcription-translation systems almost quantitatively lack LepA protein since the membranes are removed during the preparation of the extracts.

In vitro analyses of LepA: GTPase activity and assays related to translocation The high similarity between the domain structure of LepA and EF-G prompted the present inventors to test the ribosome-dependent GTPase activity of LepA. EF-G is known to have the strongest ribosome-dependent GTPase activity among all characterized G-proteins involved in translation. When the ribosome stimulation of GTP cleavage is not coupled to protein synthesis, it is referred to as being uncoupled GTPase activity. Figure 5A demonstrates that LepA not only exhibits an uncoupled GTPase activity, but that this activity is stimulated by ribosomes to same extent as that of EF-G. Next it was tested whether or not LepA can affect peptide-bond formation. To this end three functional states of the ribosome were constructed: (i) The Pi state with only one tRNA on the ribosome, namely the peptidyl-tRNA analogue AcPhe-tRNA^Phe at the P-site (i for initiation, since this state compares with the 70S initiation complex), (ii) the pre-translocational complex (PRE state) with a [³²PJtRNA/*" at the P and an Ac[¹⁴C]Phe-tRNA at the A-site, and (iii) the post-translocational state (POST state) with the same tRNAs, but now located at E and P-sites, respectively. Puromycin, an analogue of the 3'-end of an aminoacyl-tRNA that binds to the A-site region of the peptidyl-transferase centre, reacts quantitatively with the Pi and POST states (0.72 and 0.73, respectively, in Figure 5B), but not with PRE state ribosomes (0.00) as expected. Surprisingly, in the presence of LepA the POST state does not react with puromycin any more (0.01), whereas the Pi state still does (0.71 ; PM reaction in Figure 5B). At the same time LepA does not affect the amount of tRNAs bound to the programmed ribosomes (binding values for both tRNAs are the same in the presence and absence of LepA). Dipeptide analysis also support the puromycin results suggesting that the addition of LepA to a POST state ribosome prevents dipeptide formation through precluding binding of aa-tRNA to the A-site (data not shown).

Figure 5C shows an additional detail: In the presence of GTP, LepA works catalytically, in a similar fashion as EF-G, saturating at 0.4 molecules per 70S ribosome (the corresponding number for EF-G is 0.3). However, in the presence of the non-hydrolysable GTP analog GDPNP, the LepA action becomes stoichiometric, saturating at ~1 molecule per 70S ribosomes. Therefore, GTP cleavage seems to be required for dissociation of LepA from the ribosome and thus the factor behaves like a typical G-protein [reviewed by Bourne et al., 1991]. In order to test whether the rates of EF-G dependent translocation and LepA dependent back-translocation are comparable, kinetics at 30⁰C were performed in order to slow down the reactions. Single turnover conditions were applied (LepA and EF-G in a 5- and 10-molar excess of ribosomes, respectively, in the presence of GDPNP). Figure 5D shows that both reactions occur with similar rates, indicating that the LepA- dependent reaction can be incorporated into an elongation cycle without a significant delay of protein synthesis.

A possible explanation for the puromycin and dipeptide observations is that LepA induces a "back-translocation" of the POST state to the PRE state, since the occupation of the A-site by the AcPhe-tRNA after back- translocation would prevent both the puromycin reaction as well as binding of a cognate ternary complex to the A-site (dipeptide formation). In contrast, the Pi state cannot be back-translocated because the A-site of the ribosome cannot be occupied in a stable fashion without a tRNA in the adjacent P-site [Rheinberger et al., 1981]. This explains why the Pi state remains puromycin reactive even in the presence of LepA. To verify this back-translocation hypothesis three strategies were employed: (i) chemical probing of the tRNA positions through analysis of diagnostic base protections in the 16S rRNA, (ii) monitoring a POST state-specific 23S rRNA conformation marker by Pb²⁺ cleavage, and (iii) a direct test of the ribosome movement on the mRNA with the toeprint assay.

A-site bound tRNAs protect a set of characteristic bases in the 16S and 23S rRNA from chemical modifications [Moazed and Noller, 1989; Moazed and

Noller, 1990]. In order to unravel the ribosomal location of AcPhe-tRNA in the various complexes, two known A-site tRNA footprints in 16S rRNA were screened. At position A1408 in the decoding center, AcPhe-tRNA in the PRE state ribosome produced the expected A-site protection, which however was lost upon translocation to the P-site in the POST state (Figure 6A).

Significantly, the addition of LepA^»GTP to the POST state re-established the

A1408 footprint. An essentially identical tRNA footprinting pattern was observed at position U531 of 16S rRNA (data not shown). These data are compatible with the notion that AcPhe-tRNA re-occupies the A-site upon LepA addition to POST state ribosomes.

Furthermore, the 5OS subunit also shows structural evidence for a LepA- promoted back-translocation. Previously, it was demonstrated that the 5OS conformation of the post-translocational ribosome is different to that of the pre-translocational ribosome - a difference that could be monitored by site- specific Pb²⁺ cleavage of 23S rRNA [Polacek et al., 2000]. A diagnostic cleavage was detected at position C2347, which was significantly enhanced in the POST compared to the PRE state. Figure 6B demonstrates that LepA brings the strong signal observed in the POST state down to the level of the PRE signal, suggesting that upon binding of LepA^«GTP, the ribosome adopts a PRE configuration.

Additionally, the back-translocation ability of LepA was confirmed using the toeprinting assay. In this assay, the programming mRNA carries a complementary pPJ-labeled DNA primer annealed to the 3' end, located downstream of the ribosome. The primer is prolonged by reverse transcription until the polymerase clashes with the ribosome. In this way, the length of the transcript provides a measure of the distance between the primer and the ribosome. During translocation the ribosomes move by a codon length towards the primer position and thus the reverse transcript becomes shorter by three nucleotides [Hartz et al., 1990]. Conversely the transcript will be longer by three nucleotides after the putative back- translocation. A translocation of a PRE state shows a decrease in the length of the reverse transcript by three nucleotides, while the addition of LepA-GTP to a POST state increases the length of the transcript to that of the PRE state again (Figure 6C), proving that LepA is a back-translocator.

An alternative possibility is that in fact no back-translocation, but rather a complete release of AcPhe-tRNA from the POST state coupled with quantitative rebinding of the tRNA to the A-site to form a PRE state is observed. To exclude this the back-translocation experiment was repeated in the absence and presence of a two-molar excess of non-labelled AcPhe- tRNA over ribosomes. If LepA triggers a release of Ac[¹⁴C]Phe-tRNA from the P-site and rebinding at the A-site, then the presence of non-labelled AcPhe-tRNA would reduce the ribosome-bound Ac[¹⁴C]Phe-tRNA dramatically. However, no reduction in the [¹⁴C] label during the back- translocation was observed (numbers given in the legend to Figure 5B).

In vitro analyses of LepA: Effects on synthesis of active protein

As has been demonstrated previously by the present inventores bacterial coupled transcription-translation systems can produce large amounts of protein (e.g. 4 mg/ml GFP, green fluorescent protein), but under standard conditions (3O⁰C incubation) the active fraction (50 ± 20%) is unsatisfactorily low [Dinos et al., 2004]. The experimental setup is that the total protein amount is assessed via SDS-PAGE₁ since the reporter protein GFP does not overlap in a Coomassie stained gel with any other protein present in the cell lysate. This enables the GFP band to be scanned and an accurate determination of the total amount. In parallel, the same samples are loaded onto native gels and the active amount is revealed via the fluorescence of the GFP band (Figure 7A) thus allowing a precise assessment of the active fraction.

In the presence of increasing amounts LepA, the total GFP amount increases and peaks at a ratio of 0.1 molecules LepA added per 70S. Further addition of LepA leads to a rapid reduction in the GFP production, eventually blocking completely the total synthesis at a molar ratio of LepA:70S = (>0.5):1 , in agreement with the toxic effects of overproduced LepA in vivo (Figure 3A). In contrast, the native gel reveals that the active GFP amount increases to attain the same levels as the total GFP amount at LepA stoichiometrics of ≥0.2 LepA per 70S. In other words, addition of LepA promotes the synthesis of fully active proteins (also demonstrated with luciferase, data not shown). It is concluded that LepA cannot only increase the total protein yield, but more importantly improves the activity of the produced protein.

The best-known drugs that induce mis-incorporations are aminoglycosides, such as paromomycin, which bind directly in the decoding center to impair the tRNA selection process [Ogle et al., 2001]. Since LepA was shown to increase the active fraction of proteins, it was questionable as to whether LepA would also correct paromomycin-induced translational errors. Figure 7B₁ left panel, demonstrates that paromomycin severely decreases GFP synthesis, blocking it completely at 2 μ M. On the native gel, the fluorescence of GFP was even more strongly reduced by paromomycin addition,

5 indicating that paromomycin causes a drop in the active GFP fraction. In the presence of LepA, a similar paromomycin-dependent reduction in the active GFP fraction is observed (Figure 7B, right panel) indicating that LepA cannot counteract the paromomycin-induced errors. Equivalent results were obtained with other aminoglycosides, such as streptomycin and neomycin,o that also bind to the decoding center of the A-site, as well as with the E-site antibiotic edeine (data not shown).

The fidelity of protein synthesis is very sensitive to changes in magnesium concentration, such that an increase of only 5 mM (from 12 to 17 mM)s reduces the total synthesis of GFP to 40% and the active fraction from 50% to 25% (Figure 7C, left panel). Addition of LepA dramatically alters the picture: with a Mg²⁺ increase up to 3 mM the total synthesis is not reduced, in fact a small but significant increase is observed (up to 120%) and the active fraction is maintained at -100% (Figure 7C, right panel). It follows thato the dominant effect of LepA is seen at a Mg²⁺ increase of 2-3 mM, where the total protein synthesis is doubled with an active fraction of virtually 100%.

Discussion 5 The conserved domain structure of LepA, where the first four domains correspond to the EF-G domains I to V but lacking EF-G domain IV, coupled with the presence of a fifth unique LepA domain (Figure 2) makes it easy to trace LepA through the three domains of life. In this context, LepA domain III and IV (corresponding to III and V from EF-G) as well as the unique CTDo were used as probes in order to avoid false positives caused by EF-G or the corresponding factor EF2 in archaea and the cytoplasm of eukarya. LepA orthologs were found in all bacteria and eukaryotes with mitochondria, but not in archaea. This observation suggests that LepA does not contribute to eukaryotic cytoplasmic translation, but is probably essential for correct mitochondrial translation. LepA is probably also ubiquitous in chloroplasts, since LepA with apparent chloroplast import sequences was found to be nuclear encoded in the three plant genomes that have been completely sequenced, viz. the dicotyledon Arabidopsis thaliana, the monocotyledon Oryza sativa and the red alga Cyanidioschyzon merolae (data not shown). It is noted that LepA phylogeny largely reflects the canonical species phylogeny and shows no signs of inter-domain horizontal gene transfer (HGT). In this respect LepA behaves like ribosomal proteins rather than tRNA synthetases that frequently undergo HGT [Wolf et al., 1999].

The lack of EF-G domain IV in the LepA structure is intriguing. The finding according to the present invention that LepA is a back-translocator fits with the early suspicion that domain IV of EF-G has a "door-stop" function, i.e. by occupying the decoding region of the A-site after the tRNAs have been translocated from A and P-sites to the P and E-sites, respectively, domain IV of EF-G prevents a back movement of the tRNAs [Nierhaus, 1996a, see also Figure 2C]. In this respect, LepA would reduce the activation barrier between PRE and POST states similar to EF-G, but due to the absence of domain IV it catalyzes a back-translocation rather than a canonical translocation.

In addition to EF-G domain IV LepA also lacks the G' subdomain (Figures 2B and 2C). It has been speculated that the function of G' might be to promote the GDP-GTP exchange, as EF-Ts does for EF-Tu [Czworkowski et al., 1994]. However, the GDP-GTP exchange on EF-G can also be explained without the help of an additional factor or G" subdomain [Nierhaus, 1996b]. Despite the absence of the G' subdomain, LepA shows an uncoupled GTPase activity in the presence of 70S paralleling that of EF-G (Figure 5A). This argues against the assumption that this subdomain is involved in GDP-GTP exchange

The first experimental hint for the back-translocation activity of LepA came from two separate functional tests, the puromycin reaction and dipeptide formation. Both Pi and POST states with an AcPhe-tRNA donor at the P-site usually act as equally good substrates for peptide-bond formation using puromycin, or an aminoacyl-tRNA, as an A-site acceptor. The essential point is that LepA prevents peptide-bond formation exclusively of the POST state, while leaving the Pi state unaffected (Figures 4B and Supplementary Figure 2). The most likely interpretation for this is that LepA induces a back- translocation by shifting the tRNAs from E and P-sites back to the P and A- sites, respectively. The A-site is now filled with AcPhe-tRNA, this prevents binding of both puromycin and aa-tRNA, and thus prevents peptide-bond formation with both substrates.

This interpretation could be substantiated by three structural assays monitoring (i) the tRNA occupancy of the A-site via protection of diagnostic rRNA bases of the A-site, (ii) the functional state - PRE or POST - of the ribosome via conformation-specific Pb²⁺ cleavage, and (iii) the movement of the ribosome on the mRNA via toeprinting. Protection of residues A1408 and U531 of the 16S rRNA is diagnostic for the presence of a tRNA at the A- site [Moazed and Noller, 1990]. POST state ribosomes have an empty A-site and therefore show no A-site tRNA footprints, However, upon administering LepA'GTP to such a POST state, protection of these A-site specific positions was observed, thus arguing for the re-occupation of the A-site by the peptidyl-tRNA (Figure 6A). Pb²⁺ cleavages occur within distinct binding pockets of RNAs and are therefore very sensitive to conformational changes. Cleavage at position C2347 of 23S rRNA is strong in the POST and weak in the PRE state [Polacek et al., 2000], and LepA reduces the cleavage level of the POST state to that of the PRE state (Figure 6B). Finally, the toeprinting assay [Hartz et al., 1991] was used to demonstrate directly the back-movement of the ribosome on the mRNA by one codon upon adding LepA to the POST state (Figure 6C). Such a back-translocation cannot take place with a Pi state: a single tRNA on the ribosome cannot move from the P to the A-site, since the resulting complex (A-site occupied, P-site free) is unstable [Rheinberger et al., 1981]. The fact that LepA functions only with the POST state, rather than with a Pi state, means that the function of this factor depends on the ribosome having an occupied E- site. This requirement is a strong indication that the E-site also exists in mitochondrial ribosomes, for which the number of tRNA-binding sites has not yet been assessed.

The mitochondrial membrane potential depends on the respiratory activity of the mitochondria [Petit et al., 1990], which in turn might influence the intraorganelle ionic strength, creating a requirement for LepA. It is, however, noted that this must be true only under specific and as yet unknown conditions, because a knock-out of the LepA ortholog GUF1 in yeast mitochondria exhibits no clear phenotype [Kiser and Weinert, 1995]. Be it as it is, the extreme conservation of both the domain structure and the amino- acid sequence in all currently available sequences of mitochondrial LepA orthologs signals that an important function for this protein must also exist in this organelle.

Although LepA seems to work like a typical G-protein (Figure 5C) one note of caution must be added: The binding of LepA to the ribosome was monitored in the absence of nucleotides and in the presence of GTP or GDPNP via pelleting the ribosomes through a sucrose cushion and determining the presence of LepA in an SDS gel. In this context, 0.20, 0.19 and 0.51 LepA bound per 70S ribosome, respectively, was observed. In a second experiment, back-translocation of a purified POST state was analyzed using a toeprinting assay in the presence of LepA with and without GTP. Surprisingly, it was observed that LepA promoted back-translocation, even in the absence of GTP, however, the level was about 50% of that observed in the presence of GTP (data not shown). One explanation might be that, even in the absence of nucleotide, a fraction of the LepA molecules have retained the GTP conformation. Whether or not this "apo" LepA can work catalytically has yet to be determined. Such a scenario would go some way to explaining why a 40% reduction of the puromycin reactivity was observed in the presence of GDPNP, when the LepA concentration was only 10% of that of the ribosomes (Figure 5C): While half of this discrepancy can be accounted for by the fraction of ribosomes in the POST state (60%), the rest may result from the possible catalytic action of "apo" LepA.

In order to reconcile the in vivo and in vitro effects of LepA to provide a complete molecular description of its function, the following explanations are available. One possible explanation of the LepA mechanism may be slowing down the translational rate thus improving both cotranslational folding of proteins and the active fraction of the synthesized proteins. This possibility cannot be excluded, but at the moment an alternative scenario is favored:

In the present invention is has been demonstrated that LepA improves the fidelity of translation and induces back-translocation of POST state ribosomes, which suggests there is a link between translocation and activity of the synthesized protein. EF-G dependent translocation is probably not successful in 100% of cases, particularly at higher Mg²⁺ concentrations, where the ribosome may not reach the canonical POST state. Translocation of tRNAs occurs at the interface between the small and large subunits and involves a ratchet-like movement of one subunit relative to the other [Frank and Agrawal, 2001]. It has long been known that Mg²⁺ and ionic strength influence subunit interaction [Hapke and Noll, 1976] and therefore it is easy to envisage that high ionic strength could hinder EF-G action or/and induce a defective POST state in the ribosome. This defective post-translocation state might have two consequences: (i) A suboptimal display of the A-site codon in such a way that as to promote mis-incorporation. This is evident from the coupled transcription-translation system, where the addition of low amounts of LepA leads to dramatic increases in the accuracy of GFP synthesis, particularly in conditions of high ionic strength (Figure 7A and C). (ii) In some cases a ribosome might even become stuck during the course of a translocation reaction. This would explain the lethal phenotype of the LepA knock-out mutant under conditions of high ionic strength (Figure 3B). In this respect, it should be noted that the increased misincorporation seen at higher Mg²⁺ concentrations (Figure 7C) cannot alone explain the lethal effect, since E. coli strains harboring a ram mutation in S4 or S5 exhibit tenfold higher misincorporation rates, i.e. equivalent to those induced by streptomycin, and are still viable [Zimmermann et al., 1971].

Based on the results of the present application it is proposed that low concentrations of LepA (≤0.3 molecules per 70S ribosome) specifically recognize ill-translocated ribosomes, back-translocate them, and thus provide EF-G a second chance to catalyse a proper translocation reaction (Figure 8). At higher concentration (~1 molecule per 70S) LepA loses its specificity and back-translocates every POST ribosome, therewith turning the translational machinery into a non-productive mode. This is seen by the inhibition of the coupled transcription-translation system at high concentrations of LepA (>1.5 μM, equivalent to 1.5 μM ribosome concentration in the RTS; Figure 7A)₁ as well as explaining the toxicity of overexpressing LepA in vivo (Figure 3A). From these results it is clear that the intracellular level of LepA must be precisely tuned and regulated to restrict it to the narrow beneficial concentration window. It is finally noted a potential application for LepA derived from results of the coupled transcription-translation system, namely that the addition of a small, defined amount of LepA to bacterial lysates significantly improves the protein output, combining both high yield and full activity of the synthesized protein. This illustrates not only the importance of LepA for the protein synthesis in the bacterial cell, but paves the way to the development of more efficient in vitro transcription-translation systems.

List of References

Bayfield, M. A., Dahlberg, A. E., Schulmeister, U., Dorner, S., and Barta, A. (2001). A conformational change in the ribosomal peptidyl transferase center upon active/inactive transition. Proc. Natl. Acad. Sci. USA 98, 10096-10101.

Blaha, G., Stelzl, U., Spahn, C. M. T., Agrawal, R. K., Frank, J., and Nierhaus, K. H. (2000). Preparation of functional ribosomal complexes and the effect of buffer conditions on tRNA positions observed by cryoelectron microscopy. Methods Enzymol. 317, 292-309.

Boon, K., Vijgenboom, E., Madsen, L. V., Talens, A., Kraal, B., and Bosch, L. (1992). Isolation and functional analysis of histidine-tagged elongation factor Tu. Eur. J. Biochem. 210, 177-183.

Bourne, H. R., Sanders, D. A., and McCormick, F. (1991). The GTPase superfamily: conserverd structure and molecular mechanism. Nature 349, 117-127.

Caldon, C. E., Yoong, P., and March, P. E. (2001). Evolution of a molecular switch: universal bacterial GTPases regulate ribosome function. MoI. Microbiol. 41 , 289-297.

Colca, J. R., McDonald, W. G., Waldon, D. J., Thomasco, L. M., Gadwood, R. C, Lund, E. T., Cavey, G. S., Mathews, W. R., Adams, L. D., Cecil, E. T., et al. (2003). Crosslinking in the living cell locates thE-site of action of oxazolidinone antibiotics. J. Biol. Chem. 278, 21972-21979.

Connell, S. R., Trieber, C. A., Dinos, G. P., Einfeldt, E., Taylor, D. E., and Nierhaus, K. H. (2003). Mechanism of Tet(O)-mediated tetracycline resistance. EMBO J. 22, 945-953. Czworkowski, J., Wang, J., Seitz, T. A., and Moore, P. B. (1994). The crystal structure of elongation factor G complexed with GDP, at 2.7 A resolution. EMBO J. 13, 3661-3668.

Dibb, N. J., and Wolfe, P. B. (1986). lep operon proximal gene is not required for growth or secretion by Escherichia coli. J. Bacteriol. 166, 83-87.

Dinos, G., Wilson, D. N., Teraoka, Y., Szaflarski, W., Fucini, P., Kalpaxis, D., and Nierhaus, K. H. (2004). Dissecting the ribosomal inhibition mechanisms of edeine and pactamycin: the universally conserved residues G693 and C795 regulate P-site tRNA binding. MoI. Cell 13, 113-124.

Frank, J., and Agrawal, R. (2001 ). Ratchet-like movements between the two ribosomal subunits: their implications in elongation factor recognition and tRNA translocation. Cold Spring Harb. Symp. Quant. Biol. 66, 67-75.

Furano, A. V. (1975). Content of elongation factor Tu in Escherichia coli. Proc. Natl. Acad. Sci. USA 72, 4780-4784.

Gordon, J. (1970). Regulation of the in vivo synthesis of the polypeptide chain elongation factors in Escherichia coli. Biochemistry 9, 912-917.

Hapke, B., and Noll, H. (1976). Structural dynamics of bacterial ribosomes: IV.CIassification of ribosomes by subunit interaction. J. MoI. Biol. 105, 97- 109.

Hartz, D., Mc Pheeters, D. S., and Gold, L. (1990). From polynucleotide to natural mRNA translation initiation: Function of Escherichia coli initiation factors. In The Ribosome: Structure, Function and Evolution, W. E. Hill, A. Dahlberg, R. A. Garrett, P. B. Moore, D. Schlessinger, and J. R. Warner, eds. (Washington, USA, American Society of Microbiology), pp. 275-280.

Hartz, D., McPheeters, D. S., Green, L., and Gold, L. (1991). Detection of Escherichia coli ribosome binding at translation initiation sites in the absence of transfer-RNA. J. MoI. Biol. 218, 99-105.

Hartz, D., McPheeters, D. S., Traut, R., and Gold, L. (1988). Extension inhibition analysis of translation initiation complexes. Methods Enzymol. 164, 419-425.

Kiser, G. L., and Weinert, T. A. (1995). GUF1 , a gene encoding a novel evolutionarily conserved GTPase in budding yeast. Yeast 11 , 1311-1316.

Knight, J. (2004). Giardia: not so special, after all? Nature 429, 236-237.

March, P. E., and Inouye, M. (1985a). Characterization of the lep operon of Escherichia coli. Identification of the promoter and the gene upstream of the signal peptidase I gene. J. Biol. Chem. 260, 7206-7213.

March, P. E., and Inouye, M. (1985b). GTP-binding membrane protein of Escherichia coli with sequence homology to initiation factor 2 and elongation factors Tu and G. Proc. Natl. Acad. Sci. USA 82, 7500-7504.

Marquez, V., Wilson, D. N., Tate, W. P., Triana-Alonso, F., and Nierhaus, K. H. (2004). Maintaining the ribosomal reading frame: The influence of the E- site during translational regulation of release factor 2. Ce// 118, 45-55.

Mazel, D., and Davies, J. (1999). Antibiotic resistance in microbes. Ce//. MoI. Life Sci. 56, 742-754.

Moazed, D., and Noller, H. F. (1989). Interaction of tRNA with 23S rRNA in the ribosomal A, P, and E-sites. Ce// 57, 585-597.

Moazed, D., and Noller, H. F. (1990). Binding of tRNA to the ribosomal A and P-sites protects two distinct sets of nucleotides in the 16S rRNA. J. MoI. Biol. 211 , 135-145. Nierhaus, K. H. (1996a). Protein synthesis - An elongation factor turn-on. Nature 379, 491-492.

Nierhaus, K. H. (1996b). The tricks of ribosomal elongation factors. Angew. Chem. Int. Ed. Engl. 35, 2198-2201.

Nierhaus, K. H., and Wilson, D. N., eds. (2004). Protein Synthesis and Ribosome Structure: Translating the Genome (Weinheim, WILEY-VCH Verlag GmbH & Co.).

Ogle, J. M., Brodersen, D. E., demons Jr, W. M., Tarry, M. J., Carter, A. P., and Ramakrishnan, V. (2001). Recognition of cognate transfer RNA by the 3OS ribosomal subunit. Science 292, 897-902.

Petit, P. X., O'Connor, J. E., Grunwald, D., and Brown, S. C. (1990). Analysis of the membrane potential of rat- and mouse-liver mitochondria by flow cytometry and possible applications. Eur. J. Biochem. 194, 389-397.

Polacek, N., and Barta, A. (1998). Metal ion probing of rRNAs: evidence for evolutionarily conserved divalent cation binding pockets. RNA 4, 1282-1294.

Polacek, N., Patzke, S., Nierhaus, K. H., and Barta, A. (2000). Periodic conformational changes in rRNA: Monitoring the dynamics of translating ribosomes. MoI. Cell 6, 159-171.

Rheinberger, H.-J., Stembach, H., and Nierhaus, K. H. (1981). Three tRNA binding sites on Escherichia coli ribosomes. Proc. Natl. Acad. Sci. USA 78, 5310-5314.

Schafer, M. A., Tastan, A. O., Patzke, S., Blaha, G., Spahn, C. M., Wilson, D. N., and Nierhaus, K. H. (2002). Codon-anticodon interaction at the P-site is a prerequisite for tRNA interaction with the small ribosomal subunit. J. Biol. Chem. 277, 19095-19105.

Triana-Alonso, F. J., Chakraburtty, K., and Nierhaus, K. H. (1995a). The elongation factor 3 unique in higher fungi and essential for protein biosynthesis is an E-site factor. J. Biol. Chem. 270, 20473-20478.

Triana-Alonso, F. J., Dabrowski, M., Wadzack, J., and Nierhaus, K. H. (1995b). Self-coded 3'-extension of run-off transcripts produces aberrant products during in vitro transcription with T7 RNA polymerase. J. Biol. Chem. 270, 6298-6307.

Wolf, Y. I., Aravind, L., Grishin, N. V., and Koonin, E. V. (1999). Evolution of aminoacyl-tRNA synthetases-analysis of unique domain architectures and phylogenetic trees reveals a complex history of horizontal gene transfer events. Genome Res. 9, 689-710.

Zimmermann, R. A., Garvin, R. T., and Gorini, L. (1971). Alteration of a 3OS ribosomal protein accompanying the ram mutation in Escherichia coli. Proc. Natl. Acad. Sci. USA 68, 2263-2267.

Zwizinski, C, and Wickner, W. (1980). Purification and characterization of leader (signal) peptidase from Escherichia coli. J. Biol. Chem. 255, 7973- 7977.

Claims

Claims

1. A method for retarding or inhibiting the growth of cells expressing ribsosomal factor LepA by inhibiting the activity of LepA.

2. A method as claimed in claim 1 , wherein the cells expressing ribosomal factor LepA are prokaryotic cells.

3. A method as claimed in claim 1 or 2, wherein the cells expressing ribosomal factor LepA are bacterial cells.

4. A method as claimed in any one of claims 1-3, wherein the activity of LepA is inhibited by a) altering the regulation of LepA in the cell, b) blocking the binding of LepA to a ribosome, or/and c) blocking the enzymatic activity of LepA.

5. A method as claimed in claim 4, wherein the regulation of LepA in the cell is altered by inhibiting the synthesis of LepA or/and inhibiting the release of LepA from the cell membrane into the cytoplasm.

6. A method as claimed in claim 5, wherein inhibition of LepA yields a molar ratio of LepA:70S ribosomes of less than 0.05:1 in the cytoplasm, in particular less than 0.01 :1.

7. A method as claimed in claim 4, wherein the regulation of LepA in the cell is altered by overexpressing LepA.

8. A method as claimed in claim 7, wherein overexpression of LepA yields a molar ratio of LepA:70S ribosomes of more than 0.25:1 in the cytoplasm, in particular more than 0.5:1.

9. A method as claimed in claim 4, wherein the binding of LepA to the ribosome is blocked by blocking the binding-site of LepA on the ribosome.

10. A method as claimed in claim 4, wherein the enzymatic activity of LepA is blocked by blocking back-translocation of tRNAs present on a ribosome from the POST state to the PRE state or/and blocking GTPase activity.

11. Pharmaceutical composition comprising a substance inhibiting the activity of ribosomal factor LepA.

12. Pharmaceutical composition as claimed in claim 11 , wherein the substance inhibiting the activity of ribosomal factor LepA a) alters the regulation of LepA, b) blocks the binding of LepA to a ribosome, or/and c) blocks the enzymatic activity of LepA.

13. Pharmaceutical composition as claimed in claim 11 or 12, wherein the substance inhibiting the activity of ribosomal factor LepA is an inhibitor of LepA synthesis, a substance inhibiting the release of LepA from the cell membrane into the cytoplasm, a substance blocking the binding-site of LepA on the ribosome, or/and an inhibitor of GTPase activity.

14. Pharmaceutical composition as claimed in any one of claims 11-13, additionally comprising pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers.

15. Use of a substance inhibiting the activity of ribosomal factor LepA for the manufacture of an agent for retarding or inhibiting bacterial cell growth.

16. Use as claimed in claim 15, wherein the substance inhibiting the activity of ribosomal factor LepA a) alters the regulation of LepA, b) blocks the binding of LepA to a ribosome, or/and c) blocks the enzymatic activity of LepA.

17. Use as claimed in claim 15 or 16, wherein the substance inhibiting the activity of ribosomal factor LepA is an inhibitor of LepA synthesis, a substance inhibiting the release of LepA from the cell membrane into the cytoplasm, a substance blocking the binding-site of LepA on the ribosome, or/and an inhibitor of GTPase activity.

18. Use as claimed in any one of claims 15-17, wherein the agent is formulated as a pharmaceutical composition, the pharmaceutical composition optionally comprising pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers.

19. Use as claimed in claim 18, wherein the pharmaceutical composition is for preventing or treating diseases associated with bacterial pathogens.

20. Use as claimed in claim 19, wherein the diseases associated with bacterial pathogens are infectious diseases.

21. A method for identifying antibacterial substances, comprising the steps: a) providing a substance to be tested for its antibacterial properties b) providing a sample of bacterial cells a) bringing into contact the substance with the sample of bacterial cells, and b) determining whether the substance inhibits the activity of ribosomal factor LepA in the bacterial cells.

22. A method as claimed in claim 21 , wherein the method is a cellular screening method or a molecular screening method.

23. A method as claimed in claims 21 or 22, wherein the substance to be tested is formulated as a pharmaceutical composition, the pharmaceutical composition optionally comprising pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers.

24. A kit for identifying antibacterial substances, comprising at least one sample of bacterial cells and means for detecting whether activity of LepA is inhibited by a tested substance.

25. A method for preventing or treating diseases associated with bacterial pathogens, comprising administering to a patient in need thereof an effective amount of a pharmaceutical composition comprising a substance inhibiting the activity of ribosomal factor LepA, the pharmaceutical composition optionally additionally comprising pharmaceutical acceptable carrier means, adjuvants, additives, dilution means or/and buffers.