WO2005067939A1

WO2005067939A1 - Therapeutically useful antibacterial compounds

Info

Publication number: WO2005067939A1
Application number: PCT/GB2005/000131
Authority: WO
Inventors: Richard Bovill; Patrick Druggan
Original assignee: Oxoid Limited
Priority date: 2004-01-16
Filing date: 2005-01-17
Publication date: 2005-07-28
Also published as: GB0400988D0

Abstract

Disclosed is a compound which has antibiotic activity in vivo in a mammalian, preferably a human, subject, the compound comprising an uptake moiety linked to a bactericidal or bacteriostatic toxic moiety, said uptake moiety facilitating uptake of the compound by bacteria on which the toxic moiety exerts a bactericidal or bacteriostatic effect, wherein: (a) the compound is non-toxic for mammalian subjects; (b) the linkage between the uptake moiety and the toxic moiety is one which is cleavable by an enzyme or enzymes produced by the bacteria to be killed/inhibited, but is not readily cleavable by enzymes produced by the subject and to which the compound is exposed prior to contacting the bacteria; and (c) the toxic moiety, when cleaved from the uptake moiety, is toxic for the bacteria, but not toxic for the subject.

Description

Title: Therapeutically Useful Antibacterial Compounds

Field of the Invention

The present invention relates to therapeutically useful antibiotic compositions, certain novel compounds which may be useful as antibiotics, methods of making certain compounds, and use of certain compounds in the manufacture or preparation of therapeutically useful antibiotic compositions.

Background of the Invention

Allen et al (1978 Nature 272, 56-58) disclosed that phosphonopeptides possessed antibacterial properties. In particular, the compound L-alanyl-L-1-aminoethylphosphonic acid (called "alaphosphin" or "alafosfalin") was shown to be a reasonably potent antibacterial agent in vitro. Alaphosphin consists of the L stereoisomer of alanine, coupled to L-1-aminoethylphosphonic acid (AEP), the -COOH group of the alanine and the amino group of AEP condensing to form a peptide bond.

The structure of AEP is shown in Figure 1.

L-AEP has a high affinity for, and is an inhibitor of, L-alanine racemase, an enzyme involved in the synthesis of D-alanine, used in the production of the bacterial cell wall. Accordingly, once inside the bacterial cell L-AEP is toxic, disrupting cell wall synthesis leading to cell death. However, L-AEP is generally not readily taken up by bacteria. In the prior art, in order to exploit the toxic effect of L-AEP it is linked, via its free amino group, to an amino acid to form a peptide, which peptide is then actively taken up by the bacterial cell via a peptide permease uptake system. Once inside the cell, the peptide bond is cleaved by an appropriate amido-hydrolase, releasing the toxic AEP moiety. Alaphosphin is a particularly well-characterised example of an AEP-containing peptide. These original findings were further developed by Atherton et al, (1979 Antimicrob. Agents and Chemother. 15, 677-683) and by Allen et al, (1979 Antimicrob. Agents and Chemother. 16, 306-313). _, However, alaphospin was never widely adopted as an antibiotic, since in in vivo situations, the compound is of little or no efficacy. There are several reasons for this, including hydrolysis of the compound by peptidases in the subject before the drug can be taken up by bacteria, and rapid development of resistant mutant bacteria (Neuman 1984 J. Antimicrob. Chemother. 14, 309-311): permease-negative mutants arise at a high rate - about 1 in 10⁴-10⁵ (Barak & Gilvarg 1974 J. Biol. Chem. 249, 143-148; Diddens et al, 1979 The Journal of Antibiotics 32, 87-90).

WO 02/22785 is concerned with selective agents for biological cultures. The document discloses, in particular, selective agents which comprise AEP or the corresponding amino sulphonic acid/salt (e.g. 1-aminethyl sulphonic acid, AES) which may be coupled, via their α amino group to an N-sugar (page 8, paragraph 3). There is no disclosure of AEP or AES toxic moieties being coupled to carrier moieties by anything other than their amino group and, moreover, no disclosure or suggestion that the compounds might find any application as therapeutically useful antibiotic agents in vivo. Nor, in view of the lack of efficacy of compounds such as alaphosphin, would the person skilled in the art consider using such compounds as antibiotics in vivo.-

Summary of the Invention

In a first aspect the invention provides a compound which has antibiotic activity in vivo in a mammalian (preferably a human) subject, the compound comprising an uptake moiety linked to a bactericidal or bacteriostatic toxic moiety, said uptake moiety facilitating uptake of the compound by bacteria on which the toxic moiety exerts a bactericidal or bacteriostatic effect, wherein:

(a) the compound is non-toxic for mammalian subjects;

(b) the linkage between the uptake moiety and the toxic moiety is one which is cleavable by an enzyme or enzymes produced by the bacteria to be killed/inhibited, but is not readily cleavable by enzymes produced by the subject and to which the compound is exposed prior to contacting the bacteria; and (c) the toxic moiety, when cleaved from the uptake moiety, is toxic for the bacteria, but not toxic for the subject.

In a second aspect, the invention provides a pharmaceutical ^"composition comprising an active antibiotic compound in accordance with the first aspect of the invention. The pharmaceutical composition will typically comprise the active antibiotic together with a physiologically acceptable diluent, excipient or carrier. The pharmaceutical composition may be provided, for example, as a solution, a suspension, a gel, a cream, an ointment, a dry powder, or a solid (such as a tablet, capsule, pill etc). Suitable diluents, excipients and carriers are well known to those skilled in the art and include water, aqueous buffers or solutions (such as saline, phosphate-buffered saline), calcium carbonate, dextrose starches, and the like. The composition may be administered to the subject in any suitable manner e.g. by injection (including intravenously, intramuscular, subcutaneous), inhalation, orally, nasally, rectally, or topically and the composition formulated accordingly. A preferred route of administration (and hence a preferred formulation) is for oral delivery. Desirably the composition is provided as a tablet, pill, capsule or the like.

Those skilled in the art appreciate that even a substance regarded as safe and non-toxic can have detrimental effects if present at unusually high concentration. Thus, for present purposes, an antibiotic compound which is active in vivo in accordance with the first aspect of the invention is to be considered non-toxic for a mammalian subject if the compound can be tolerated by the subject, without significant ill effect, at a concentration which has a detectable bactericidal or bacteriostatic activity on bacteria present in the subject which bacteria it is desired to kill/inhibit. For present purposes, "killing" of bacteria is also considered to encompass preventing their growth, and therefore, unless the context dictates otherwise, references to "bactericidal" should be construed as encompassing "bacteriostatic". Bactericidal or bacteriostatic activity can be detected by measuring numbers of viable bacteria obtained from suitable samples (e.g. blood, plasma, urine or stool samples) from treated subjects and comparing said viable counts with those obtained prior to treatment or from untreated subjects.

In general terms, the invention is concerned with compounds with antibacterial activity in vivo. (Such in vivo activity cannot necessarily readily be predicted from in vitro activity, as illustrated by experience with alaphosphin). The invention is also concerned with pharmaceutical compositions comprising such compounds, and the use of such compounds to produce pharmaceutical compositions.

The invention is especially concerned with providing antibacterial compounds which are resistant to degradation by enzymes present in the (typically human) subject, especially degradation by peptidases. This is especially a problem where the toxic moiety is an amino acid analogue (such as a 1-aminoalkyl compound).

One solution is to join the amino acid analogue to an uptake moiety which is not an amino acid residue or an analogue thereof. Preferred non-amino acid residue carrier moieties include saccharides and derivatives thereof, as explained in detail below.

An alternative approach is to join the uptake moiety to the non-toxic moiety by a peptide bond which is not a peptide bond. Preferred examples include the use of phosphonate or sulfonate bonds, as explained in detail below.

The uptake moiety is preferably something which renders the compound a substrate for a bacterial uptake system. A large number of bacterial uptake systems are known to those skilled in the art. In general, such systems are intended to ensure that the bacterium obtains sufficient quantities of essential or desirable nutrients from its immediate environment. Examples of uptake systems include the dipeptide permease ("Dpp"), tripeptide permease ("Tpp") and oligopeptide permease ("Opp") systems.

The names given to these bacterial uptake systems are slightly misleading, in that whilst the Dpp and Tpp systems have higher affinity for dipeptides and tripep tides respectively, both the Dpp and Tpp systems will facilitate entrance into the bacterial cell of both dipeptides and tripeptides.

Generally similar peptide uptake mechanisms have been reported in Salmonella typhimurium (Gibson et al, 1984 J. Bacteriol. 160, 122-130) and Pseudomonas aeruginosa (Hulen & LeGoffic 1987 FEMS Microbiol. Letts 40, 103). Peptide uptake systems also exist in Gram positive bacteria such as Enterococcus faecalis (Nisbet & Payne 1982 J. Gen. Microbiol. 128, 1357-1364) and Lactobacillus lactis (Foucaud et al, 1995 J. Bacteriol. 177, 4652-4657).

The Opp system displays the broadest substrate specificity since, as its name suggests, it can transport peptides of various lengths. In practice, it can facilitate the uptake of almost any peptide containing from two to five amino acid residues, and some longer peptides. It can even transport peptides containing D-amino acid residues.

As mentioned above, the linkage between the uptake moiety and the toxic moiety is one which can be cleaved by enzyme/s produced by the bacteria to be killed/inhibited but which is substantially resistant to any host subject enzyme which might be encountered by the compound prior to uptake by the bacteria. The linkage preferably comprises a covalent bond or a plurality of covalent bonds.

In fact, a compound which utilises a carbohydrate uptake system is, in many instances to be preferred. This is because, whilst compounds comprising peptide bonds will resemble peptides (i.e. be structural analogues thereof) and so readily transported into bacteria by peptide permease systems, they will also be substrates for peptidases, which are common in mammals, especially in the gut (but also, to a lesser extent, in serum). Accordingly such compounds (especially if administered orally, the preferred route of administration) will be subject to a cleavage by enzymes of the subject and the toxic moiety will not be able to enter the bacterial cell (or at least, not so readily as the uncleaved compound) and will therefore be essentially ineffective. Indeed, it is believed that this is one of the major reasons why alaphosphin, although exhibiting bacterial properties at low concentration in vitro, was largely inactive in vivo.

In view of the abundance of peptidases in mammals, especially in humans, the present invention aims to avoid the use of compounds which are cleavable by human peptidases. This makes it difficult to devise an antibiotic which can make use of the bacterial peptide permease systems to enter the bacterial cell. Instead, in general (but not necessarily exclusively), the present inventors prefer to use compounds which make use of bacterial carbohydrate uptake systems. A number of carbohydrate uptake mechanisms exist in bacteria, including uptake systems for mono-, di- and trisaccharides. Uptake systems for molecules larger than trisaccharides are uncommon. Glycoside transport/uptake are systems are common in bacteria, (especially galactoside and glucoside uptake systems) and any of these may be conveniently utilised by antibiotic compounds in accordance with the invention. For present purposes a glycoside can be considered to comprise a saccharide moiety (optionally substituted), preferably a monosaccharide, covalently coupled to a non- saccharide moiety (i.e. an aglycon).

Glycosides may be taken up either via the phosphotransferase system (PTS) or the Glycoside-Pentoside-Hexuronide (GPH) cation symporter system (Saier et al, 1999 J. Mol. Microbiol. Biotechnol , 1, 257-279; and Tchieu et al, 2002, "The complete phosphotransferase system in Escherichia coli". In: The Bacterial Phosphotransferase system, M H Saier, M.H. , Jr. (Ed), Horizon Scientific Press, Wymondham, UK). In both cases a wide range of aglycons may be attached to the sugar. For transport by the GPH system in E. coli the preferred size of the aglycon has been found to be hexose > benzene ring > methyl group > no aglycon > disaccharide > trisaccharide. However, neither the specific structure of the aglycon nor its relative hydrophobicity are important (Olsen and Brooker, 1989 J. Biol. Chem. , 264, 15982-15987). The structure of the sugar itself is a much more rigid requirement (although there is a degree of overlap) and some bacteria such as the streptococci possess over twenty PTS permeases, each specific for a different subset of sugars (Parr and Saier, 1992 Res. Microbiol, 143, 443-447.) The PTS transport system is thought to be found exclusively in prokaryotes (Parr and Saier, 1992) and the use of phosphorylated sugars linked to a toxic moiety e.g. AEP or AES, may provide a preferred means of delivering the toxic agent to bacteria without prior hydrolysis in the gut by mammalian enzymes. In particular therefore, derivatised sugar- based (e.g. phosphorylated sugar-based) antibiotics may be especially useful in treating infections of the gastro-intestinal tract.

Hydrolysis of glycosides by exoglycosidases is again almost exclusively dependent on the structure of the sugar and conformation of the glycosidic bond (Faure, 2002 Appl. Environ. Microbiol. , 68, 1485-1490). This will ultimately determine the toxicity of the antibiotic and a broad spectrum antibiotic may comprise, a glycoside that many bacteria can hydrolyse or a cocktail of glycosides that cover a range of bacteria.

Glycosidases are present in the blood and tissues of mammals, but are not as common as peptidases, especially in the gut. Accordingly the inventors consider that glycoside compounds should be far less sensitive to enzymatic breakdown in vivo than the peptide- based compounds of the prior art. Thus, glycoside antibiotics in accordance with the invention should prove especially useful in the treatment of enteric infections (e.g. caused by organisms such as Salmonella spp; Shigella spp - especially Shigella dysenteriae; and verocytotoxigenic E. coli - such as E. coli 0157). In addition, glycosidases are widespread among bacterial species, so a range of suitable compounds should be available, at least one of which can be cleaved (to release a toxic moiety) by a particular bacterial organism which it is desired to kill/inhibit.

Accordingly, it is highly desirable that the uptake moiety of the present invention is one which facilitates uptake of the compound by a bacterial carbohydrate (especially glycoside) uptake system.

A preferred uptake moiety comprises a saccharide or a saccharide derivative. For present purposes, the term "saccharide" encompasses monosaccharides, disaccharides and oligosaccharides. The term "oligosaccharide" means a moiety comprising from 3 to 10 sugar units. Preferably an oligosaccharide will comprise from 3-8 sugar units, more preferably from 3-6 sugar units. Desirably the uptake moiety is a monosaccharide or monosaccharide derivative.

The term "saccharide derivative" encompasses any pharmaceutically acceptable derivative of the aforementioned saccharides. In particular, for example, substituted saccharides are considered as saccharide derivatives. The mammalian gut does comprise some glucosidases and galactosidases which might be expected to cause some hydrolysis of glucoside and galactoside-based antibiotics. Accordingly it is preferred either to avoid using galactose or glucose uptake moieties (for example by using a hexose other than glucose or galactose, or using a pentose, e.g. a ribose, uptake moiety) or else to derivatise a glucose/galactose moiety in such a way as to render it more resistant to gut enzymes.

Preferred toxic moieties include inhibitors of L-alanine racemase such as 1-aminoalkyl compounds, especially acids and pharmaceutically acceptable salts and esters thereof. Particularly preferred are amino lower alkyl moieties (i.e. those comprising a C_M alkyl group, most desirably an ethyl group). Specific examples of preferred toxic moieties include 1 -amino alkyl phosphonic acid, 1-aminoalkylsulfonic acid (especially 1- aminoethylphosphonic acid and 1-aminoethylsulfonic acid) and salts and esters of the aforementioned compounds. Preferred salts are the sodium and ammonium salts.

Thus, in one preferred embodiment, the present invention provides a compound having antibiotic activity in vivo in a mammalian (preferably human) subject, the compound comprising a saccharide or saccharide derivative uptake moiety, covalently linked to an arninoalkyl toxic moiety, preferred examples of which are as stated in the preceding paragraph.

WO 02/22785 discloses, inter alia, 1-aminoalkyl compounds such as AEP or AES, or salts thereof, covalently coupled via their α-amino group to N-sugars for use as selective agents in vitro. Since these compounds per se are disclosed in the prior art, they are specifically excluded from the scope of the invention insofar as the present invention relates to compounds per se or to their use as selective agents in vitro. However, the use of such compounds as antibiotics in vivo is neither taught nor suggested in the prior art. Accordingly, the other aspects of the invention do encompass the use etc. of such compounds in pharmaceutical compositions and in methods of treatment, and the like.

In one preferred embodiment, the antibiotic compound comprises a sugar or sugar derivative which is covalently linked to an aminoalkyl moiety (preferably a 1-aminoalkyl moiety) such as an amino alkyl phosphonic acid (or salt thereof) or amino alkyl sulfonic acid (or salt thereof) via the amino group thereof (e.g. as exemplified schematically in Figure 2). Preferred amino alkyl moieties are amino ethyl phosphonic acid (or pharmaceutically acceptable salts thereof) and amino ethyl sulfonic acid (or pharmaceutically acceptable salts thereof).

In general, the uptake moiety preferably comprises a saccharide or saccharide derivative, as hereinbefore defined. Most preferably the uptake moiety comprises a monosaccharide or derivative thereof. The monosaccharide (or derivative thereof) may conveniently be a triose, tetrose, pentose, hexose or heptose. Pentoses, hexoses and heptoses are generally preferred. Examples of suitable monosaccharides include ribose, xylose, arabinose, ribulose, xylulose, allose, altrose, glucose, mannose, gulose, idose, talose, psicose, fructose, sorbose and tagatose. Generally speaking, D-sugars are preferred.

Saccharide derivatives include those compounds comprising substitution of one or more H or OH groups (e.g. with OH or H, respectively, or with halides, C,.₃ alkyl, C,.₃ alkoxy, acetyl, carboxyl N-acetyl, phosphate etc. or any other pharmaceutically acceptable moiety). The substitution may be at any one or more of the carbon atoms. A preferred substitution is at the C₆ position of a hexose, such as glucose. In particular the inventors consider substitution of H or OH at C₆ with a phosphate group to be desirable. Specific saccharide derivatives that may be useful include acidic sugars such as glucuronides. It is also possible, although not necessarily preferred, to substitute carbon atoms of the saccharide e.g. with N, Si. Compounds according to these preferred embodiments will tend to enter the bacterial cell via a carbohydrate (especially a glycoside) uptake pathway. The inventors believe that once inside the bacterium, an appropriate bacterial enzyme, such as a glycosidase, will then cleave the linkage between the uptake moiety and the toxic moiety, allowing the toxic moiety to exert its bactericidal or bacteriostatic effect.

The inventors believe that glycoside-based antibiotics should not suffer from the disadvantages associated with prior art peptide-based compounds, because :-

1. Bacteria possess a multitude of substrate specific permease systems for the uptake of carbohydrates. Mutants multiply defective in these systems would occur only at a very low frequency and such mutants would be severely metabolically compromised. The use of a cocktail of different toxic glycosides would, therefore, not give rise to the so-called "superbugs" that are multiply-resistant to conventional antibiotics.

2. Apart from the activity of the gut microflora, digestion of carbohydrates in the small intestine of humans is restricted to the hydrolases of the brush-border membrane (lactase, sucrase, maltase and trehalase). These enzymes tend to be limited to glucoside or galactoside substrates. This would mean that AEP-glycosides containing sugars other than glucose or galactose would be expected to be resistant to digestion in the gut.

In another embodiment the invention provides a compound which has antibiotic activity in vivo in a mammalian (preferably a human) subject, the compound comprising a toxic moiety linked to an uptake moiety; wherein the toxic moiety, when cleaved from the uptake moiety, is an inhibitor of L-alanine racemase; and wherein the uptake moiety is such as to facilitate uptake of the compound by bacteria on which the toxic moiety exerts a bactericidal or bacteriostatic effect; the toxic moiety being covalently linked to the uptake moiety according to the formula R-Y-X wherein R is the uptake moiety and X is the toxic moiety and Y is a polyvalent atom or group, the covalent linkage being cleavable by an enzyme or enzymes produced by the bacteria to be killed/inhibited but not readily cleavable by enzymes produced by the subject and to which the compound is exposed prior to contacting the bacteria. Typically Y is a bivalent atom, conveniently O or S. Preferably Y = O. An example of such a compound is illustrated in Figure 3. Typically the enzymatic cleavage comprises breaking at least one of the bonds involving an oxygen atom.

In one particular embodiment, the uptake moiety comprises a pyroglutamyl moiety linked, according to the formula R-Y-X, to a toxic moiety. An example of such a compound is illustrated in Figure 4. In this and other embodiments, amino ethyl phosphonic acid or amino ethyl sulfonic acid are preferred toxic moieties. Pyroglutamate has been linked to AEP (via the amino group of AEP) in the prior art but, to the best knowledge of the present inventors, pyroglutamate has not previously been linked to AEP via the phosphate group.

The inventors have surprisingly found that phosphonate-linked pyroglutamyl-AEP and pyroglutamyl- AES are reasonably stable in aqueous environments and toxic for bacteria. Without wishing to be bound by any particular theory, the inventors believe that the explanation for their finding lies in the Dpp bacterial uptake system.

The Dpp uptake system has an absolute requirement for a free primary amine group. Since pyroglutamate has no free amine group, and the amine group of the amino-linked compound is used to form the link between AEP and the pyroglutamate, amino-linked pyroglutamyl-AEP cannot act as a substrate for the Dpp system. In contrast, in phosphate- linked pyroglutamyl-AEP, there is a free primary amine (on the AEP moiety) and so the compound can be successfully taken up by the dipeptide permease. Once inside the cell, there are many bacteria which possess a pyroglutamyl amidase suitable to cleave the compound, thus releasing the toxic moiety, such as AEP or AES. In contrast, peptides with an N-terminal pyroglutamate residue are substantially resistant to hydrolysis by enzymes of the gastrointestinal tract of mammals (especially humans) and the compound should therefore be active in vivo and clinically useful. Thus, the present invention provides, in a third aspect, the use of an active antibiotic compound in accordance with the invention in the preparation of a pharmaceutical antibiotic composition to treat or prevent a bacterial infection in a mammalian (preferably human) subject. Preferably the compound comprises a saccharide or saccharide derivative uptake moiety, covalently coupled to an aminoalkyl toxic moiety. Preferably the saccharide or saccharide derivative is covalently attached, conveniently via the amino group, to a 1 aminoalkyl group such as amino ethyl sulfonic or phosphonic acid (or pharmaceutically acceptable salts thereof).

In another aspect the invention provides a method of treating or preventing a bacterial infection in a mammalian (preferably human) subject, the method comprising the step of administering an effective dose of an antibiotic compound in accordance with the invention. Typically the antibiotic will be administered as a dose of a pharmaceutical composition in accordance with a further aspect of the invention. In particular, the method preferably comprises administering the antibiotic orally to the subject. For present purposes, asymptomatic colonisation of a subject by bacteria which are potentially pathogenic can also be considered as an "infection". Thus, for instance, the method of the invention encompasses administering an antibiotic in accordance with the invention to a "carrier" subject (e.g. a carrier of MRS A or the like).

An effective dose of antibiotic may be considered for present purposes as a dose which exerts a detectable antibacterial effect on the infection. The effect may be detected, for example, by a discernible amelioration in symptoms exhibited by the subject (e.g. reduction in fever), or by a reduction in amount of bacterial antigen or other bacteria- specific component, or by a decline in viable numbers of bacteria in samples obtained from the site of infection.

What constitutes an effective dose will vary to some extent depending on circumstances e.g. the identity of the antibiotic compound, the identity and phenotype of the bacteria causing the infection, the sensitivity of the bacteria to the^' compound, the site of the infection etc. However, for those skilled in the art it is a matter of routine trial-and-error to establish an effective dose for any particular situation - by starting at a low dose and gradually administering larger doses until a discernible antibacterial effect is obtained without causing unacceptable side effects.

By way of guidance, a suitable dose of active antibiotic compound would typically be in the range 10mg-2000mg, more preferably 50-1000mg, to be administered from once a day up to about 6 times a day, more typically 2 to 4 times a day. If given by intramuscular injection, for example, a suitable effective dose might be in the range 10-500mg of active agent, preferably 25-250mg.

In a further aspect the invention provides for use of the antibiotic compound of the invention in a biological medium (e.g. to prevent growth of bacteria in tissue culture media). The medium may comprise a single antibiotic compound in accordance with the invention, or a plurality of such compounds. The medium may optionally additionally comprise one or more conventional antibiotics. More especially, an antibiotic in accordance with the invention may be used as a selective agent in a bacterial growth medium.

Thus, in a further aspect the invention provides a biological medium comprising one or more antibiotic compounds in accordance with the invention. Desirably the medium is a selective bacterial growth medium (i.e. one which favours the growth of some bacteria relative to other bacteria which might be expected to be present in a single sample). A specific example of such a medium might be one which facilitates the growth of Salmonella spp. whilst tending to inhibit the growth of other gut bacteria e.g. coliforms.

Such selective media might be, for instance, a diagnostic medium for clinical use or else used to detect faecal contamination (as indicated by the presence of coliforms) in environmental samples (e.g. water, food, or swabs from kitchen work surfaces etc). This aspect of the invention explicitly excludes the amino-linked AEP/ AES glycosides or amino-linked AEP/AES pyroglutamyl compounds, the use of which as selective agents is explicitly taught in WO 02/22785. For the avoidance of doubt it is hereby expressly stated that features of the invention described as "preferred", "desirable", "convenient" or "advantageous" and the like may be used in the invention in isolation, or in combination with any other feature of the invention so described, unless the context dictates otherwise.

The invention will now be further described by way of illustrative example and with reference to the accompanying drawings in which:

Figure 1 is a schematic representation of the structure of 1 -amino ethyl phosphonic acid (AEP), a preferred toxic moiety for use in antibiotic compounds in accordance with the invention;

Figure 2 is a schematic representation of the structure of an antibiotic compound in accordance with the invention, being a glycoside comprising AEP linked, via its amino group, to a glucose residue;

Figure 3 is a schematic representation of the structure of an antibiotic compound in accordance with the invention, being a glycoside comprising AEP linked, via its phosphate group, to a glucose residue; and

Figure 4 is a schematic representation of the structure of an antibiotic compound in accordance with the invention being AEP linked, via its phosphate group, to a pyroglutamyl uptake moiety.

Examples

Example 1 - Synthesis of phosphate linked and N-linked AEP-β-D-galactopyranosides During the synthesis, reactions and products were checked by thin layer chromatography (tic) on silica gel sheets (manufactured by Machery-Nagel, Germany, available from Sigma) using one of two mobile phases :-

(a) Toluene/ ethyl acetate/ acetone/ dichloroacetic acid (5:6:1 :2); or (b) Dichloromethahe/ methanol dichloroacetic acid (8:5:2).

Compounds were detected using ninhydrin or sulphuric acid charring (the sheets were dried, sprayed with concentrated sulphuric acid and heated to 110°C).

All chemicals were obtained from Sigma- Aldrich unless otherwise stated.

1.1 Synthesis of phosphate linked AEP-β-D-galactopyranoside

N-Carbobenzoxy-L-1-aminoethylphosphonic acid (1).

5g of R-(l)-aminoethylphosphonic acid (40mM) were dissolved in 40ml of deionised water and the pH adjusted to 9.5 with 4M sodium hydroxide. To the solution was added 20ml of diethyl ether and 10. Ig (60mM) of benzyl chloroformate. The mixture was stirred vigorously in an ice bath keeping the temperature below 2°C.

Amounts of 4M sodium hydroxide were added to keep the pH in between 9.0 and 9.5. When the pH stopped dropping (about 7 hours) the solution was extracted once with diethyl ether and the extracted discarded. The pH of the aqueous layer was then decreased to 1.2 by stirring with strong cation exchange resin (TR-1200 H⁺) and it was then extracted 6 times with diethyl ether. The ether extracts were pooled, dried with magnesium sulphate and solvent evaporated off at room temperature to leave a syrup. Toluene was evaporated from the syrup twice and a white solid resulted (yield 6.2gms, 59.8%; Mpt. 129°C; t.l.c. system b, rf = 0.42).

N-Carbobenzoxy-L,l-aminoethylphosphonate-β-D-galactopyranoside tetraacetate (2) 2g (7.72 rnmol) of (1) was dissolved in anhydrous dichloromefhane containing 5ml of anhydrous 2, 4, 6 collidine with 2g of silver trifluoromethanesulphonate and 4 grams of 3 A pore size molecular sieves (Fluka). To this solution was added 3.17g (7.71mmol) of tetraacetyl-α-D-galactopyranosyl bromide. The solution was stirred at room temperature in the dark for 2 days. It was then filtered though silica gel contained in a Buchner funnel, the funnel washed 3 times with dichloromethane and the dichloromethane solution washed twice with 0.1m hydrochloric acid and once with water. It was then dried with magnesium sulphate and solvent evaporated off at 30°C to leave a syrup (3.1g; 5.3 rnmol; 68.7%). (t.l.c. system a, rf = 0.64).

L,l-AminoethyIphosphonate-β-D-galactopyranoside (3)

3.1g (7.6 rnmol) of (2) was dissolved in 50ml of dry methanol and 10ml of sodium methoxide solution (0.5g sodium in 100ml dry methanol) added. The suspension was left for 30 minutes at room temperature then LR-1200H⁺ resin was added decreasing the pH to 1.2. The solution was decanted from the resin and 2g of palladium on charcoal (hydrogenation catalyst) and 0.5ml of glacial acetic acid added. Hydrogen was slowly bubbled through the suspension for 4 hours. The palladium on charcoal was then filtered off and 1.38g (7.6 rnmol) of dicyclohexylamine added to the filtrate.

Solvent was evaporated off and toluene added and evaporated from the syrup twice. A solid formed which was recrystalized from ethanol diethyl ether to give pale yellow crystals of the dicyclohexylammonium salt of (3) (yield l.lg, 3.83 mM, 49.7%, Mpt. 132°C, t.l.c. system b, rf= 0.57).

Anal. Calc. for C₈H₁₈NO₈P: C, 33.46; H, 6.32; N, 4.88; P, 10.78. Found: C, 33.38;H, 6.39 ;

N, 4.81; P, 10.71.

Diethyl N-carbobenzoxy-L-1-aminoethylphosphonate (4)

A stirred solution of 2.8g (10.8 rnmol) of (1) and 20g (136mmol) of triethyl orthoformate" was heated at 135°C for 1 hour removing the by-products of the reaction by distillation. The mixture was then cooled and filtered and the excess diethyl orthoformate removed from the filtrate under vacuum leaving 3.1g (9.81 rnmol; 90.8%) of the product as a viscous gum. Diethyl-L-1-aminoethylphosphonate (5)

3.1g (9.8mmol) of (4) was dissolved in 150ml of methanol and 2g of 10% palladium on charcoal catalyst was added. Hydrogen was bubbled through the mixture at R.T. for 4 hours then the catalyst was filtered off and the solvent evaporated at 30°C. Toluene was twice evaporated from the syup that was formed and the product was finally obtained as crytals. RecrystaUisation from ethanol afforded the pure product as off-white crytals, yield 1.3g, (7.2 mmol; 73.5%).

N-linked, (diethyΙ-L-l-aminoethyΙphosphonate)-β-D-galactopyranoside (6)

2.0g (11.0 mmol) of (5) was dissolved in 100ml of dry dichloromethane with 3.3g (12.0 mmol) of silver carbonate and 4.52g of acetobromogalactose (11.0 mmol) and 4g of 3 A molecular sieves. The mixture was stirred in an ice bath in the dark overnight. It was then filtered though silica gel contained in a Buchner funnel and the silica gel washed 6x with 100ml quantities of dichloromethane. The dichloromethane was extracted 3x with 0.1M hydrochloric acid and once with deionised water and dried with magnesium sulphate. Solvent was then evaporated off at R.T. to give the product as a syrup. The protected product was not isolated but instead deprotected immediately.

Thus, to the syrup was added 100ml of a 45% solution of hydrogen bromide in glacial acetic acid. The solution was stirred at R.T. for 2 hours then 600ml of ether were added. A gummy mass precipitated after storing for 2 hours at -18°C. Solvents were decanted from the gum and toluene evaporated twice form it. A syrup resulted which was dissolved in 100 ml of dry methanol and 75ml of a sodium methoxide solution (0.5g in 100ml of dry sodium methoxide) added. The suspension was left for 30 minutes at room temperature then LR- 1200H⁺ resin was added decreasing the pH to 1.2. The solution was decanted from the resin and 4.0g (22.0 mmol) of dicyclohexylamine added to the filtrate. Solvent was evaporated and the product rerytallised from ethanol. Yield 4.4Sg, 6.2mmol, 56.4%; Mpt. = 163°C.

1.2 Alternative method for the synthesis of N-linked AEP-β-D-galactopyranoside

Diethyl-L-1-aminoethylphosphonate-β-D-tetra-O-benzyl-galactopyranoside (7) 1.5 g of (5) (0.0083 mol) was dissolved in 60ml of dry dichloromethane with 5.67g (0.00083 mol) 2,3,4,6-tetra-O-benzyl-α-D-galactopyranosyl trichloroacetimidate. Boron trifluoride diethyl etherate (1.4ml; 1.53g; 0.011 mol) was then added and the mixture stirred at room temperature for 16 hours. The organic solution was extracted with 0.1M sodium bicarbonate and then dried with magnesium sulphate. Solvent was evaporated off and ethanol added to the syrupy mass. The product was filtered off as white crystals (5.8g; 0.0076; 92.1 %). Anal. Calc. for C₄₀H₄₂NO₁₂P: C, 63.24; H, 5.57; N, 1.84; P, 4.08. Found: C, 63.21; H, 5.59; N, 1.78; P, 4.02.

Diethyl-L-1 -aminoethylphosphonate-β-D-galactopyranoside (8)

Palladium charcoal catalyst (l.Og) was added to a suspension of 5.2g (0.0068 mol) of (7) in 100ml of ethanol containing 5ml of deionised water. Hydrogen was gently bubbled through the mixture, at a rate so as to maintain the catalyst in suspension, for 24 hours. The catalyst was then filtered off and the filtrate concentrated in vacuo to give 2.24g (0.0065 mol; 95.3 %). Anal. Calc. for C₁₂H₂₆NO₈P: C, 41.98; H, 7.63; N, 4.08; P, 9.02. Found: C, 42.01 ; H, 7.59; N, 4.02; P, 8.97.

N-linked, aminoethylphosphonate-β-D-galactopyranoside

To a stirred solution of (8) (1.8g; 0.0052 mol) in 30ml of glacial acetic acid was added 30ml of a 45 % solution of hydrogen bromide in acetic acid. Stirring was continued for 3 hours. Anhydrous ether (500ml) was then added and the solution stored at 4°C for 4 hours. Solvent was then decanted form the syrupy mass and it was washed once with a small amount of anhydrous ether. Anhydrous ethanol (100ml) was added and as the syrup dissolved a crystalline precipitate formed. The mixture was left overnight at -20°C and the precipitated was then filtered off and dried. Yield l. lg; 0.0038 mol; 73.7% . Anal. Calc. for C_sH₁₈NO₈P: C, 33.46; H, 6.32; N, 4.88; P, 10.78. Found: C, 33.40; H, 6.34; N, 4.85; P, 10.76. nmr (D2O) δ 1.0 (d, 3, methyl), 3.40 (m, 1, proton on C4), 3.46 (m, 1, proton on C3), 3.54 (m, 1, methine proton on C6), 3.6 (m, 1, methine), 3.61 (m, 1 , proton on C2), 3.76 (m, 1, proton on C5), 3.74 (m, 1, methine proton on C6), 4.81 (s, 5, glycosidic proton and OH protons). Example 2 - in vitro assay of antibacterial activity

Experiments were performed to demonstrate inhibition of the growth of a range of different organisms by phosphate linked L-pyroglutamyl (Pyr-AEP) and phosphate linked AEP-β-D-galactopyranoside. Thus, the respective compounds were dissolved in deionised water (llmg/ml), filter sterilised, and an amount of the solution added to an autoclaved defined medium ("DM") for the former compound or modified DM (mDM) for the latter to give final concentrations from 0.5 to lOOμg/ml for the former and 10-500μg/ml for the latter.

(DM contained, in grams/litre, 1-arginine 0.1, 1-aspartic acid 0.1, 1-cysteine 0.1, glycine 0.1, 1-histidine 0.1, 1-isoleucine 0.1, 1-leucine 0.1, 1-lysine.HCl 0.1, 1-methionine 0.1, 1- phenylalanine 0.1, 1-proline 0.1, 1-serine 0.1, 1-threonine 0.1, 1-tryptophan 0.1, 1-tyrosine 0.1, 1-valine 0.1, adenine 0.01, cytosine 0.01, guanine 0.01, uracil 0.01, ammonium sulphate 1.0, calcium chloride 0.005, calcium pantothenate 0.05, choline chloride 0.005, dipotassium hydrogen phosphate 4.72, magnesium sulphate heptahydrate 0.1, myo-inositol 0.009, nicotinamide 0.005, potassium dihydrogen phosphate 3.77, pyridoxal hydrochloride 0.005, riboflavin 0.0001, sodium citrate 0.5, thiamine hydrochloride 0.005, yeast extract 0.1, glucose 5.0, sodium pyruvate 0.5. mDM was identical except that glucose was present at 0.8gm/l instead of 5.0gm/l).

Volumes (300μl) of these antibiotic-containing solutions were then pipetted into the wells of Bioscreen microtitre plates and 30μl quantities of 1 in 10,000 dilutions of overnight cultures added. (The overnight cultures were prepared by inoculating Tryptic Soy Broth with an isolated colony of the relevant organism and incubating at 37°C without agitation.) Plates were then covered with a lid, incubated at 37°C for 18 hours, and the opacity of the organism suspensions measured using a Bioscreen C instrument (available from Thermo Electron Corporation, Waltham MA, USA).

Negative and positive controls were included. (Negative control - uninoculated medium; positive control - inoculated medium without antibiotic.) Using this assay technique growth curves (in terms of OD against time) for various organisms could be obtained, indicating whether the organisms where sensitive or resistant to a particular antibiotic at a particular concentration.

Table 1 shows the results for the P-linked pyroglutamyl-AEP compound. Those organisms which exhibited significantly reduced O.D. values (relative to the positive control) at antibiotic concentrations of 2 μg/ml or less were considered sensitive, and organisms which exhibited an O.D. value substantially unchanged at this concentration of antibiotic were considered resistant.

Table 1. Antibacterial activity of phosphate linked Pyroglutamyl- 1-aminoethylphosphonic acid

(S = Sensitive; R = Resistant) Strains of Salmonella were inhibited by phosphate linked Pyr-AEP only by levels of 8 μg/ml or above (50 μg/ml for 3 of the strains) and were essentially resistant whereas the non-Salmonella strains tested were affected by levels of 2.0 μg/ml or below (Table 1).

Table 2 shows the results obtained with the P-linked AEP-β-D galactopyranoside compound at a concentration of 500 μg/ml. Even at this high a concentration of antibiotic, the one Salmonella strain tested was resistant, whilst strains of Enterobacter and Klebsiella were both sensitive.

Table 2. Sensitivity of various strains to 500 μg/ml of phosphate linked AEP-β-D- galactopyranoside

A phosphate-linked AEP-β-D-glucopyranoside was prepared in a manner analogous to that described in Example 1 for AEP-β-D-glucopyranoside. The sensitivity of various organisms to the glucosides was then tested, using the Bioscreen C instrument and the assay format described above. The antibiotic was tested at concentrations of 0, 10, 20, 100 and 300 μg/ml. The results are shown in Table 3, which compares the sensitivity of the organisms to a concentration of 100 μg/ml of the antibiotic.

Table 3. Sensitivity of various strains to 100 μg/ml of phosphate linked AEP-β-D- glucopyranoside

(Symbols: R = resistant; S = sensitive; OCC = Oxoid culture collection)

These results are explainable by considering the distribution of hydrolase activity that is displayed among the bacteria tested. Salmonella do not possess the pyroglutamyl amidase necessary to hydrolyse Pyr-AEP, thus the toxic AEP moiety cannot be released inside the cell. Citrobacter, Enterobacter, Escherichia and Klebsiella all do possess the relevant hydrolytic enzyme and are therefore sensitive to Pyr-AEP. Similarly, the resistant organisms in Tables 2 and 3 would be expected not to cleave the respective galactoside and glucoside substrates whereas sensitive organisms would.

An example of a cocktail of AES or AEP-glycosides that could be used to inhibit a broad selection of bacteria is a mix of n-Acetyl-β-glucosaminidase, β-galactosidase and β- glucosidase. This would inhibit Citrobacter freundii, Edwardsiella hoshinae, Escherichia coli, Morganella morgani and Salmonella enteritidis (Table 4).

Table 4. Distribution of enzymes in members of the Enter obacteriaceae

Data from: Kampfer et al, (1991). Journal of Clinical Microbiology, 29, 2877-2879.

Experiments were performed to test the antibacterial activity in vitro of N-linked AEP-β- galactopyranoside. The protocol was as described previously as above. The results are summarised in Table 5 below. It was found that the compound was active at a concentration as low as lOμg/ml against sensitive organisms.

Table 5. Sensitivity of various strains to lOμg/ml of N-linked AEP-β- galactopyranoside

From Table 5 it can be seen that the N-linked galactoside is a much more potent inhibitor under the test conditions (at lOμg/ml) than the phosphate linked analogue. This is presumably due to the galactosidase enzyme having a much greater affinity for the N- linkage than for the P - version.

Claims

1. A compound which has antibiotic activity in vivo in a mammalian, preferably a human, subject, the compound comprising an uptake moiety linked to a bactericidal or bacteriostatic toxic moiety, said uptake moiety facilitating uptake of the compound by bacteria on which the toxic moiety exerts a bactericidal or bacteriostatic effect, wherein:

(a) the compound is non-toxic for mammalian subjects; (b) the linkage between the uptake moiety and the toxic moiety is one which is cleavable by an enzyme or enzymes produced by the bacteria to be killed/ inhibited, but is not readily cleavable by enzymes produced by the subject and to which the compound is exposed prior to contacting the bacteria; and (c) the toxic moiety, when cleaved from the uptake moiety, is toxic for the bacteria, but not toxic for the subject.

2. A compound according to claim 1, other than: a 1-aminoalkyl compound covalently coupled via its α-amino group to an N-sugar or to a pyroglutamine residue.

3. A compound according to claim 1 or 2, wherein the toxic moiety, when cleaved from the uptake moiety, is an inhibitor of L-alanine racemase.

4. A compound according to claim 3, according to the formula R-Y-X, wherein R is the uptake moiety, X is the toxic moiety and Y is a polyvalent atom or group, and wherein the covalent linkage between R and Y is cleavable by an enzyme produced by the bacteria to be killed.

5. A compound according to claim 4, wherein Y is S or O.

6. A compound according to any one of the preceding claims comprising amino ethyl phosphonic acid, or a pharmaceutically acceptable salt thereof as a toxic moiety, linked via a phosphonate bond to an uptake moiety.

7. A compound according to any one of claims 1-5, comprising amino ethyl sulphonic acid or a pharmaceutically acceptable salt thereof as a toxic moiety, linked via a sulphonate bond to an uptake moiety.

8. A compound according to any one of the preceding claims, wherein the compound is a glycoside.

9. A compound according to claim 8, wherein the glycoside comprises a sugar residue uptake moiety which is other than galactose or glucose.

10. A compound according to claim 8, wherein the glycoside comprises a galactose or glucose residue uptake moiety, which is derivatised so as to be resistant to galactosidases or glucosidases present in the human gut.

11. A compound according to any one of claims 1-7, wherein the uptake moiety comprises a pyroglutamyl residue.

12. A pharmaceutical antibiotic composition comprising an active agent in accordance with any one of the preceding claims and a pharmaceutically acceptable excipient, diluent or carrier.

13. A composition according to claim 12, in the form of a cream, ointment or gel formulated for topical application to human skin.

14. A composition according to claim 12, in the form of a tablet, capsule, pill or other unitary dose form.

15. Use of a compound according to any one of claims 1-11 as an active agent in the preparation of a medicament to treat or prevent a bacterial infection in a mammalian, preferably human, subject.

16. A use according to claim 15, wherein the medicament is formulated for the treatment or prevention of an enteric infection.

17. A use according to claim 15 or 16, wherein the active agent comprises a glycoside.

18. A use according to claim 17, wherein the glycoside comprises a 1-aminoalkyl compound covalently linked, via its amino group, to a saccharide.

19. A method of treating or preventing a bacterial infection in a mammalian, preferably human, subject, the method comprising the step of administering an effective dose of an active antibiotic compound in accordance with any one of claims 1-11.

20. A method according to claim 19, wherein the bacterial infection is an enteric infection.

21. A method of making a biological medium, the method comprising the step of adding to the medium an effective amount of an active compound in accordance with any one of claims 2-11.

22. A biological medium comprising an effective amount of an active compound in accordance with any one of claims 2-11.

23. A medium according to claim 22, wherein the medium is a selective bacterial growth medium, such that the active antibiotic prevents or inhibits the growth of undesirable bacteria whilst permitting the growth of selected desired bacteria.