EP1082441A1 - Mycobacterial n-acetyltransferases - Google Patents

Abstract

Described are arylamine N-acetyltransferase proteins from Mycobacterium tuberculosis and an arylamine N-acetyltransferase gene and protein from Mycobacterium smegmatis together with antibodies raised against related gene products and proteins, and methods of detecting mycobacteria using such antibodies and mycobacterial strains expressing or lacking such proteins. Also described are methods of screeening for compounds which are ligands of arylamine N-acetyltransferase proteins from mycobacteria, for example substrates and inhibitors. Such ligands, examples of which are given, may be used in medicaments for treatment of mycobacterial infections or in anti-mycobacterial formulations and may be designed by use of the three-dimensional structures of the proteins of the invention. Methods of identifying nucleic acid sequences which relate to putative enzymes useful for the synthesis of, or as targets for, anti-mycobacterial drugs are also described.

Description

MYCOBACTERIAL N-ACETYLTRANSFERASES

Introduction

Arylamine N-acetyltransferases (NATs) catalyse the N- acetylation of arylamines, the N- and O-acetylation of arylhydroxylamines, the N-0 transfer of an acetyl group from acetylhydroxylamines and the N- acetylation of aromatic hydrazines (1 ,2). Their substrates include drugs (e.g. isoniazid, a hydrazine) and carcinogens (e.g. aminofluorene). NATs are found in a wide range of organisms and the enzymes form a unique family of proteins on the basis of sequence similarity. The enzymes from bacteria constitute a subgroup of the NAT enzymes on the basis of sequence comparison. All of the enzymes use acetylCoA as a cofactor in the acetylation reactions and the reaction mechanism is thought to involve the transfer of an acetyl group to an active site sulphydryl group in the protein followed by the transfer of the acetyl group from the acetylated intermediate to an aromatic amine or hydrazine substrate.

It was initially thought, from studies of activity measurements, that N-acetyltransferase activity in bacteria, initially referred to as N- hydroxyarylamine O-acetyltransferase (NhoA transferase) activity, was confined to Salmonella typhimunum. There has been interest in the activity of this bacterium because NAT activity has been demonstrated to be associated with the sensitivity of S. typhimunum strains when they are used as indicators of mutagenicity in the Ames test.

The NAT from S. typhimunum, when aligned with NATs from eukaryotes for maximum similarity, has a 20 amino acid gap corresponding to a region in eukaryotic NATs predicted to be random coil. The structural prediction suggests that eukaryotic NATs and S. typhimunum NAT each consist of alpha-helix in the N-terminal region and predominantly beta- sheet in the C-terminal half, with a loop on the C-terminal side between the two regions in eukaryotes which is missing in S. typhimunum NAT. It has been demonstrated that the N-terminal portion of human NAT catalyses the formation of the first step of the reaction generating an acetylated enzyme intermediate but does not transfer the acetyl group to the arylamine substrate. From this information and sequence comparisons related to substrate specificities, it appears that the C-terminal region and in particular the C-terminal 20 amino acids have an important role in controlling substrate specificity.

Isoniazid is used for the treatment of TB. Although the molecular mechanism which renders M. tuberculosis exquisitely sensitive to isoniazid has resulted in potential targets, one of which has been crystallised, it is not yet fully established how isoniazid acts. It has been proposed that isoniazid interferes with the synthesis of mycolic acids which are a characteristic component of the cell walls of mycobacteria, including M. tuberculosis. It is thought to be an oxidation product of isoniazid which is the active agent and isoniazid resistance has been associated with mutation in the katG (catalase-peroxidase gene). Other loci have been associated with isoniazid resistance. Ahpc (alkyl hydoperoxide reductase) expression is thought to counteract the action of activated isoniazid. A further protein, InhA which catalyses the NADH-specific reduction of enoyl- ACP in long chain fatty acid synthesis has been proposed as another target for isoniazid. The drug does not appear to bind to the InhA protein. Isoniazid resistance can only be partially accounted for by the currently identified potential causes. The inventors propose that the enzyme arylamine

N-acetyltransferase (NAT) has a role in modulating the activity of isoniazid in M. tuberculosis. Mutations which affect either the NAT protein itself or the expression of the NAT protein in M. tuberculosis may modify NAT activity in isoniazid resistant strains. The Invention

The inventors have determined that arylamine N-acetyltransferase (NAT) proteins are present in certain species of Mycobacteria. They believe that the NAT proteins of certain species of Mycobacteria, and drug-resistant Mycobacterial strains, contribute to the inactivation by acetylation of several anti-mycobacterial drugs. The inventors have isolated genes encoding such enzymes, produced recombinant enzymes encoded by these genes and raised antibodies against these recombinant enzymes and their component peptides. The use of these antibodies and enzymes will enable the discovery of new anti- mycobacterial compounds. Furthermore, the inventors have engineered both mycobacterial strains expressing heterologous NAT enzymes and strains lacking an NAT function and have developed model substrates for NAT enzymes that also restrict mycobacterial growth. The invention provides, in one aspect, the Mycobacterium tuberculosis arylamine N-acetyltransferase protein having the sequence of figure 1 or peptide fragments greater than 5, preferably greater than 10, contiguous amino acid residues thereof. The Mycobacterium tuberculosis arylamine N-acetyltransferase protein or peptide fragments thereof, may have an altered sequence from that shown in figure 1 , such that the amino acid at position 207 is arginine instead of glycine.

Also provided is the Mycobacterium smegmatis arylamine N- acetyltransferase protein having the amino acid sequence of figure 2 or peptide fragments greater than 5, preferably greater than 10, contiguous amino acid residues thereof.

The present invention also provides a protein or a fragment thereof or a polypeptide containing a mimetope of an epitope of a Mycobacterium tuberculosis or Mycobacterium smegmatis arylamine N- acetyltransferase protein or fragment thereof. Proteins and fragments thereof and polypeptides of the invention may be recovered from cells of organisms expressing an arylamine N-acetyltransferase gene or generated by expression of an arylamine N-acetyltransferase gene or coding sequence contained in a nucleic acid of the present invention in an appropriate expression system and host, or obtained by de novo synthesis or a combination thereof, by techniques well known in the art of recombinant DNA technology. The proteins, fragments thereof and polypeptides of the invention will contain naturally occurring L-α-amino acids and may also contain one or more non- naturally occurring α-amino acids having the D- or L-configuration.

Also envisaged according to the invention are variants of the said protein and fragments, e.g. variants having conservative substitutions of one residue for another one at one or more locations, having at least 80% sequence identity, with the said protein or fragments, and longer proteins and peptides containing the characteristic protein or fragments here defined, said longer protein or peptides having been artificially obtained.

Further provided is the Mycobacterium smegmatis arylamine N-acetyltransferase gene having the sequence of figure 2 or oligonucleotide fragments greater than 8, preferably at least 20, contiguous nucleotide residues thereof. The gene or oligonucleotide fragments may have sequences differing from the sequences of the arylamine N-acetyltransferase gene or oligonucleotide fragments, due to alternative codon usage and/or encoding alternative amino acids sequences, conservative or otherwise, or mimetopes. Single stranded DNA may be the transcribed strand or the non- transcribed (complementary) strand. The nucleic acids corresponding to the arylamine N-acetyltransferase gene or oligonucleotide fragments may be present in a vector, for instance a cloning or expression vector, such as a plasmid or cosmid or a viral genomic nucleic acid. RNA's of the invention include unprocessed and processed transcripts of DNA, messenger RNA (mRNA) containing the arylamine N-acetyltransferase gene and anti-sense RNA containing a sequence complementary to the arylamine N- acetyltransferase gene.

Nucleic acids encoding Mycobacterial arylamine N- acetyltransferases are envisaged as useful primers for polymerase chain reactions (PCRs) conducted to ascertain the susceptibility of a particular strain of Mycobacterium to anti-mycobacterial drugs. Where nucleotide differences in the NAT gene or controlling regions of the NAT gene exist between isoniazid-resistant and sensitive strains, PCR methods, including the use of fluorescent primers for rapid screening, and restriction fragment length polymorphism (RFLP) analysis, can be used on TB isolates to allow a correct course of treatment to be carried out.

Also envisaged according to the invention are variants of the said gene or fragments having at least 80% sequence identity with the said genes and fragments; and longer polynucleotides containing the characteristic genes or fragment here defined, the said longer polynucleotide having been obtained by recombinant nucleic acid technology.

Further provided are mycobacterial strains transformed with one or more synthetic DNA constructs encoding one or more genes or proteins of the invention for example hybrid proteins containing amino acid sequence found in two or more arylamine N-acetyltransferase proteins. Also provided are mycobacterial strains transformed with a synthetic DNA construct which abolishes endogenous arylamine N-acetyltransferase activity.

In a further aspect, the invention provides antibodies raised against the Mycobacterium tuberculosis arylamine N-acetyltransferase protein having the sequence of figure 1 or peptide fragments greater than 5 contiguous amino acid residues thereof. Antibodies raised against Mycobacterium tuberculosis arylamine N-acetyltransferase protein or peptide fragments thereof, where the amino acid at position 207 is arginine instead of glycine, also form part of the present invention.

Also provided are antibodies raised against the Mycobacterium smegmatis arylamine N-acetyltransferase protein having the sequence of figure 2 or peptide fragments greater than 5 contiguous amino acid residues thereof.

Further provided are antibodies raised against the Salmonella typhimurium arylamine N-acetyltransferase protein having the sequence in figure 3 or peptide fragments greater than 5 contiguous amino acid residues thereof.

The antibodies provided by the invention may have been raised against a protein or a fragment thereof or a polypeptide containing a mimetope of an epitope of a Mycobacterium tuberculosis, Mycobacterium smegmatis or Salmonella typhimurium arylamine N-acetyltransferase protein or fragment thereof.

Antibodies against such proteins and fragments as well as fragments of such antibodies (which antibody fragments include at least one antigen binding site) including chemically derived and recombinant fragments of such antibodies, and cells, such as eukaryotic cells, for instance, hybridomas and prokaryotic recombinant cells capable of expressing and, preferably secreting antibodies or fragments thereof against such proteins or fragments, also form part of the present invention. Antibodies may be obtained by immunization of a suitable host animal and recovery of the antibodies, by culture of antibody producing cells obtained from suitably immunized host animals or by in vitro stimulation of B-cells with a suitable arylamine N-acetyltransferase protein, fragment or polypeptide or arylamine N-acetyltransferase mimetope, protein, fragment or polypeptide and culturing of the cells. Such cells may be immortalised as necessary for instance by fusion with myeloma cells. Antibodies may also be produced by 'phage-display' technology known in the art. Antibody fragments may be obtained by well known chemical and biotechnological methods. All these techniques are well known to practitioners of the arts of biotechnology.

Further provided by the invention is the use of the antibodies described herein for detecting mycobacteria in a sample, for example a biological sample, for example from an infected individual. Also provided is a method of detecting mycobacteria in a sample, for example a biological sample, which comprises the steps of providing a biological sample which has been treated so as to allow the release of cytoplasmic components of mycobacteria present in the sample, bringing said treated sample into contact with one or more antibodies raised against a mycobacterial arylamine N-acetyltransferase, and determining the binding of said antibody to one or more cytoplasmic components of the sample. The sample may treated by sonication to release the cytoplasmic components. In another aspect the invention provides a method of screening for compounds that bind to a Mycobacterial N-acetyltransferase protein by bringing one or more molecules of an N-acetyltransferase protein into contact with one or more compounds to be screened for binding and detecting said binding. Such compounds may be produced by combinatorial or rational design methods as known in the art.

One such method provided comprises the steps of immobilising one or more test compounds on a solid support; bringing an aliquot of N-acetyltransferase protein into contact with the said immobilised test compounds; removing unbound protein from the solid support; detecting the bound protein. The N-acetyltransferase protein used in the method may be arylamine N-acetyltransferase from Mycobacterium tuberculosis, Mycobacterium smegmatis or Salmonella typhimurium, or fragments thereof, or N-acetyltransferase proteins or fragments thereof from eukaryotes including humans. Where a non-mycobacterial NAT protein is used in the compound screen, the compound may also optionally be screened using a mycobacterial NAT.

Where a screened compound is found to bind to a mycobacterial N-acetyltransferase, it may be subsequently tested for binding to Human NAT1 or NAT2 or other eukaryotic NATs, for example, bovine or badger. This may be by the steps of: immobilising one or more screened compounds which bind to mycobacterial N-acetyltransferase on a solid support; bringing an aliquot of a human N-acetyltransferase protein into contact with the said immobilised test compounds; removing unbound protein from the solid support; detecting the bound protein. In this particular method, the detection of the bound protein may be by means of antibodies raised against complete molecules, fragments, or a polypeptide containing a mimetope of an epitope, of a human NAT protein .

Where human NAT protein is not used in the screening method, the detection of bound protein in such a method may be by means of antibodies raised against complete molecules, fragments or a polypeptide containing a mimetope of an epitope, of arylamine N- acetyltransferase from Mycobacterium tuberculosis, Mycobacterium smegmatis or Salmonella typhimurium. Other methods of protein detection known in the art, for example, the use of antibody fragments containing at least one epitope binding site and other ELISA and fluorescence techniques, are also envisaged.

In a preferred method the compounds being screened for are those that are ligands of NAT, herein defined as including compounds which are substrates of NAT or which modulate the enzymic activity of NAT such as inhibitors, for example inhibitors of the acetylation of aromatic amines and hydrazines by Mycobacterial N-acetyltransferase proteins. More preferably such compounds inhibit the acetylation of arylamines by Mycobacterial N-acetyltransferase proteins. Most preferably such compounds are ligands for Mycobacterial but not mammalian N-acetyltransferase proteins.

In a further aspect the invention provides a method of controlling the growth of a Mycobacterium by bringing said Mycobacterium into contact with a compound which is a ligand of a arylamine N- acetyltransferase protein expressed by said Mycobacterium. The ligand maybe a substrate or an inhibitor of the arylamine N-acetyltransferase. The method of controlling the growth of a Mycobacterium may comprise bringing the said Mycobacterium into contact with an aromatic amine or hydrazine and a compound, for example a NAT ligand, capable of diminishing the activity of an arylamine N-acetyltransferase protein expressed by said Mycobacterium for said aromatic amine or hydrazine. A preferred method of according to the invention controls the growth of

Mycobacterium tuberculosis, Mycobacterium leprae, Mycobacterium bovis or Mycobacterium avum, though the control of other pathogenic mycobacterial species is envisaged. The aromatic amine or hydrazine and the inhibitory compound may be applied to the mycobacterium separately or simultaneously.

The invention provides in a further aspect, the use of a compound which is a NAT ligand, in the manufacture of a medicament for the treatment of mycobacterial infections. The ligand may be a substrate or an inhibitor of mycobacterial arylamine N-acetyltransferase. The inhibitor may diminish the activity of a mycobacterial arylamine N-acetyltransferase protein towards a second compound which is an aromatic amine or hydrazine, which may be incorporated in the medicament also.

Also provided is an anti-mycobacterial formulation which contains an effective concentration of a compound which is a ligand of a mycobacterial N-acetyltransferase protein. The ligand may be a substrate or an inhibitor of the mycobacterial arylamine N-acetyltransferase. Also envisaged as an anti-mycobacterial formulation is a mixture of an effective concentration of an anti-mycobacterial aromatic amine or hydrazine and an effective concentration of a compound capable of diminishing the activity of a mycobacterial arylamine N-acetyltransferase protein. A preferred anti- mycobacterial hydrazine is isoniazid, though other compounds such as ethionamide, dapsone or p-aminosalicylate are envisaged.

Also envisaged according to the invention are compounds which modulate the expression of NAT. Such compounds can be screened by methods known in the art using the antibodies and/or gene sequences of the invention to monitor levels of one or more NAT gene expression products upon exposure of a target mycobacterium to candidate molecules. Such compounds may be used in the manufacture of medicaments for the treatment of mycobacterial infections or be contained in anti-mycobacterial formulations in an effective concentration. Multiple isoforms of NAT have been found in many eukaryotes including humans, chickens and mice. In humans, two isoforms exist, NAT1 and NAT2, which are 87% homologous in amino acid sequence but have different tissue distributions and substrate specificities. NAT1 is widely distributed and uses p-aminobenzoic acid (pABA) and suiphamethoxazole as substrates. NAT2 has been shown to metabolise isoniazid (INH) and arylamine sulphonamides.

The cytosolic NAT enzymes all share a high sequence identity and appear to have been conserved throughout evolution. A conserved amino acid sequence denoted as PFENL by the single letter code within the N-terminal region may be involved in substrate binding due to its capacity to make hydrophobic interactions with aromatic groups. An RGGdC region, where d is either Y or W, of the N-terminal half on the molecule is also highly conserved. Site directed mutagenesis has been used to show that the conserved cysteine in the RGGdC amino acid sequence found in all characterised NATs (C69 in Salmonella typhimurium) is essential for activity. The more variable C-terminal region may confer substrate specificity.

The invention provides, in a further aspect, a method of identifying nucleic acid sequences which at least partially encode, or amino acid sequences which at least partially comprise, putative enzymes which are potential targets for anti-mycobacterial drugs or putative enzymes useful for the synthesis of anti-mycobacterial drugs, by means of identifying within said sequences nucleotides encoding or amino acids comprising, protein sequences with similarity greater than 80% to the following sequence:

PFENL(X)_n=₅-₅oRGGdC where d is W or Y.

Preferably, similarity is greater than 90% and most preferably greater than 95%. When calculating the similarity between the above query sequence and similar sequences, the identity of the intervening (X)_n=₅-₅o residues is not considered. Preferably n is between 20 and 30, and more preferably n is equal to 24 such that the first R residue is 29 residues on the C-terminal side of the P residue. Using this method, sequences, for example those found in DNA, expressed sequence tag or protein databases, are screened to identify putative enzymes which closely resemble NAT. Such enzymes potentially share a similar acetylating function, and are therefore potentially useful as targets for anti-tubercular drugs, or useful for the synthesis of anti-mycobacterial drugs. Such a screening method can be carried out using sequence analysis software known in the art.

The invention also provides a method of isolating DNA sequences, from any source of genomic or cDNA, which at least partially encode putative enzymes which are potential targets for anti-mycobacterial drugs or putative enzymes useful for the synthesis of anti-mycobacterial drugs, by means of a primer extension reaction, such as the polymerase chain reaction, which uses one or more sense primer which encodes the amino acid sequence PFENL and one or more anti-sense primer which encodes the amino acid sequence RGGdC, where d is either W or Y. The product of the primer extension reaction may be subsequently labelled and used as a probe, by methods known in the art, to screen a DNA library, genomic or cDNA, which may be from the same organism as the source of the template DNA used in the primer extension reaction. Ideally a mixture of degenerate sense primers and a mixture of degenerate anti-sense primers, comprising all possible encoding sequences for the said amino acid sequences, are used in the primer extension reaction. Possible primer mixtures, as described using the IUPAC-IUB code, are: Sense primer: 5' CCNTTYGARAAYYTN 3';

Antisense primer: 5' RCAVYANCCNCCNCK 3'; where 'N' = A, T, G or C; = T or C; 'R' = G or A; 'K' = G or T; and V = A, G or C. More preferred mixtures of primers are those where the corresponding codon usage for the said amino acid sequences is similar or the same as that used by the organism from which the template DNA is derived.

The invention also relates to the use of the three dimensional structure of a Mycobacterial arylamine N-acetyltransferase protein for the design of compounds capable of binding to said arylamine N- acetyltransferase protein by methods known in the art. In addition, the invention further relates to the use of the three dimensional structure of a Mycobacterial arylamine N-acetyltransferase protein for determining or predicting the structure of a further arylamine N-acetyltransferase protein and using the structural information so obtained for the design of compounds capable of binding to said further arylamine N- acetyltransferase protein.

The invention also provides compounds which are substrates of arylamine N-acetyltransferase proteins which have the following formula:

where R = C1-C20 alkyl.

Such compounds may be used as model substrates in screens for compounds which modify NAT activity, or used in the manufacture of medicaments for treatment of mycobacterial infections, or contained in an anti-mycobacterial formulation in an effective concentration. NATs catalyse a classic ping pong reaction. The first step involves the acetylation of the enzyme. The enzyme starting state is restored by the transfer of the acetyl group to a suitable arylamine substrate, such as 4-aminoveratrole (4-AV), anisidine (ANS), p-aminosalicylic acid (PAS), dapsone, or hydrazine substrates such as isoniazid (INH). Organisms which do not appear to possess NAT may duplicate its activity using other enzymes. It was reasoned that any such activity by M. tuberculosis may affect INH efficiency, by competing with bioactivation of INH. It is also important to account for the difference in sensitivity to isoniazid between M. tuberculosis and M. smegmatis. Progress on the sequencing of bacterial genomes has demonstrated that sequences homologous to NAT are present in a series of bacterial species. The inventors have found NAT-like sequences in E. coli, B. subtilis and M. tuberculosis. The sequence in E. coli is extremely similar to the S. typhimurium sequence showing 74% nucleotide sequence identity.

The identification of a NAT-like sequence in M. tuberculosis is particularly interesting because the anti-bacterial agent isoniazid is a substrate for one of the human NAT isoenzymes and isoniazid is rendered inactive by acetylation. Humans who are fast acetylators (now identified as carrying one or two fast alleles at the NAT2 locus) inactivate isoniazid more rapidly than slow acetylators.

To determine whether the sequence which appeared to encode for NAT in M. tuberculosis encoded for a similar protein, a library from M. tuberculosis was screened. A clone was obtained with the expected sequence. The NAT region has subsequently been cloned into the expression vector pET28b and encodes for a protein of the expected size upon SDS-PAGE analysis (approximately 32 kDa, which includes a hexahistidine fusion tag). A clone from a library of the closely related M. smegmatis has been found and sequenced. It is 65% identical to putative NAT from M. tuberculosis. The M. smegmatis NAT, expressed in E. coli has the expected molecular weight and N-acetyltransferase activity with substrates, including isoniazid.

The examples that follow are more clearly described with reference to the following figures: Figure 1 shows the nucleotide and derived protein sequence of the M. tuberculosis NAT gene. Letters below sequence represent the predicted amino acids in the single-letter code. Two number code indicates nucleotide and derived amino acid number, respectively. The stop codon is marked by an asterisk. Amino acids in bold face represent areas of conservation typical of NAT genes. Previously reported conserved arginines t (; Biochem J, (1997) Delomenie et al) and cysteine § (; J. Biol. Chem, (1992) Dupret & Grant and J. Biol. Chem (1992) Watanabe et al) are shown.

Figure 2 shows the nucleotide and derived protein sequence of the M. smegmatis NAT gene. Letters below sequence represent the predicted amino acids in the single-letter code. Two number code indicates nucleotide and derived amino acid number, respectively. The stop codon is marked by an asterisk. Amino acids in bold face represent areas of conservation typical of NAT genes. Previously reported conserved arginines t (; Biochem J, (1997) Delomenie et al) and cysteine § (; J. Biol. Chem,

(1992) Dupret & Grant and J. Biol. Chem (1992) Watanabe et al) are shown.

Figure 3 shows the nucleotide and derived protein sequence of the S.typhimurium NAT gene. Letters below sequence represent the predicted amino acids in the single-letter code. Two number code indicates nucleotide and derived amino acid number, respectively. The stop codon is marked by an asterisk. Amino acids in bold face represent areas of conservation typical of NAT genes. Previously reported conserved arginines t (Biochem J, (1997) Delomenie et al) and cysteine § (; J. Biol. Chem, (1992) Dupret & Grant and J. Biol. Chem (1992) Watanabe et al) are shown. Figure 4 shows the specific activity of heterologously expressed M. smegmatis NAT with 4-AV, ANS , p-aminobenzoic acid (at 0.2 mM) and isoniazid (at 0.015 mM) as substrates. The y-axis shows the activity in E. coli lysate relative to anisidine (1 equivalent to 14 nmol/min/mg protein).

Figure 5 shows the effect of expression of M. tuberculosis nat on the growth of M. smegmatis in the presence of isoniazid (INH). Results for cultures of M. smegmatis either transformed with pACE-1 alone (open circles) or with pACE-1 containing M. tuberculosis nat (solid circles) grown in minimal medium containing acetamide and cultures of pACE-1 containing M. tuberculosis nat grown in minimal medium containing glucose (triangles) are shown.

Figure 6 shows a sequence comparison between S. typhimurium NAT, M. tuberculosis NAT, M. smegmatis NAT and Human NAT2" '*' represents the active cysteine residue. Areas underlined represent conserved regions found in all NATs used to screen for M. smegmatis NAT. Dark shading shows conserved amino acids.

Figure 7 shows a western blot analysis of the recombinant and endogenous NAT gene products of M. smegmatis and S. typhimurium and also endogenous NAT in an M. smegmatis soluble cell extract.

Recombinant proteins were expressed in E. coli. Proteins were detected using a polyclonal antibody raised against purified recombinant

S. typhimurium NAT. Lane A = 500 ng total protein containing

M. smegmatis recombinant NAT. Lane B = 200 ng of total protein containing recombinant S. typhimurium NAT. Both recombinant proteins contained histidine tags to aid purification, these add an extra 2.1 kDa to the native protein size. Lane C = 40 μg of M. smegmatis soluble cell extract. The polyclonal antibody was used at a 1:100 000 dilution.

Figure 8 shows SDS-PAGE and Western blot analysis of bacterial lysates. Total lysate concentrations were adjusted to load equivalent amounts of recombinant protein. A. SDSPAGE of; Marker (M; kDa), BL21/PET28b (E. coli strain transformed with PET28b only), S. typhimurium NAT (BL21 cells transformed with PET28b and S. typhimurium nat), M. tuberculosis NAT (BL21 cells transformed with PET28b and M. tuberculosis nat), M. smegmatis NAT (BL21 cells transformed with PET28b and M. smegmatis nat) and Mycobacterial chimera NAT (BL21 cells transformed with PET28b and a chimera of M. smegmatis and M. tuberculosis nat). Western blot analysis using; B. Antiserum raised against pure whole protein of S. typhimurium NAT used at a dilution of 1 : 100, 000, C. Antiserum raised to a 13 mer C-terminal peptide specific forM. smegmatis NAT used at a dilution of 1 :5, 000, and D. Antiserum raised to pure whole protein of M. tuberculosis NAT used at a dilution of 1 : 100, 000.

Figure 9 shows a schematic representation of construct (3.6 kbp) used to ablate the endogenous nat gene of M. smegmatis. Gene knockout was achieved using the suicide vector approach. This was achieved by the electroporation of the construct ligated into a vector, incapable of replicating in mycobacteria, into M. smegmatis. Selection was by kanamycin resistance.

Figure 10 shows PCR analysis for potential nat gene knockouts and targeted integration. A representative sample of 40 kanamycin resistant colonies are shown. A. PCR amplification using primers specific for M. smegmatis nat demonstrates a lack of the gene (therefore gene knockout) in sample KO 9, as indicated by the lower arrow (828 bp). The upper arrow represents product obtained with kanamycin gene inserted into nat (2.1 kbp). As expected this is not present in MC2 155 (wild type M. smegmatis), this possesses only the nat gene. Lanes represent, M; (Hind Ill/Bam H 1 digested lambda phage marker), C; No template DNA, C2; vector containing M. smegmatis nat, C; knockout construct used to create nat-gene ablation in M. smegmatis, MC2 155; wild type M. smegmatis, KO 3, 4, 8, 16, 21 and KO 27 are single crossover (meioploidy) events, and KO 9; a double crossover or gene knockout event. B. PCR amplification of the same samples used in Fig. 2. A using primers representing the 3' terminus of the kanamycin insert running in a 5' direction together with primers designed to amplify outside of knockout cassette. This determines integration to the 3' end and/or a targeted event.

C¹, C², C³ and MC² 155 do not possess the required sequence to generate product. All others are at least integrated correctly at the 3' end of nat to yield a product size of 2.1 kbp.

Figure 11 shows 'A' a gel electrophoresis and 'B/C Southern blot analysis of selected potential knock out strains of M. smegmatis. Lanes represent, A; vector containing gene knock out construct (M. smegmatis nat gene with kanamycin insert), B; vector containing M. smegmatis nat gene only, C; blank lane, Mc2 155; wild type M. smegmatis, KO 3-27; selected strains of kanamycin resistant M. smegmatis. KO 9 clearly depicts a nat gene knock out event. Panel A demonstrates agarose gel analysis of equal amounts of digested chromosomal DNA using Eco Rl and Hind HI visualised using ethidium bromide. Panel B is a southern blot of Panel A using the nat gene open reading frame as a probe. Panel C is Panel B stripped and reprobed with the kanamycin gene.

Figure 12 shows the acetylation capacity of isoniazid by M. smegmatis lysates containing recombinant M. smegmatis NAT, endogenous M. smegmatis NAT and M. smegmatis with the nat gene ablated (KO 9). The y-axis shows acetylated isoniazid per mg total protein per hour. Experiments performed in triplicate.

Figure 13 shows sensitivity to increasing concentrations of isoniazid during growth (ODβoonm) in 7H9 media of M. smegmatis wild type (Mc2 155), M. smegmatis over expressing M. tuberculosis nat (TBNAT) and M. smegmatis with the nat gene knocked out (KO 9). TBNAT demonstrates an increased resistance to isoniazid and the nat gene knock out demonstrates an increased sensitivity to isoniazid.

Figure 14 is a plot of the rate of acetylation by NAT on a range of long chain (C2-C12) aminophenolethers of the formula:

R=Et, But, Hex, Oct, Dec.Dod 4, 3 and 2-isomers

The synthesised compounds were assayed with S. typhimurium NAT to determine the rate of acetylation. The 200 μl assay contained 9% DMSO, 91 % 20mM Tris-HCI buffer pH 7.5, 90 μM compound, 400 μM Acetyl CoA and 1 μg NAT. The rates of acetylation were compared with an identical solution without the NAT enzyme and the difference attributed to the enzyme.

Figure 15 shows the effect of on a range of long chain (C2- C12) aminophenolether compounds upon the growth of wild-type M. smegmatis. Cultures of M. smegmatis were grown in the presence of approximately 25 mg/ml of each compound and a measurement of growth via an optical density reading at 580nm was taken after 24 hours. The percentage inhibition of growth for the C4, C6, C8, C10 and C12 para and ortho compounds, compared to cultures in which the compounds were absent is shown.

EXAMPLES

Example 1 Mycobacterium tuberculosis arylamine N-acetyltransferase recombinant protein production

Expression of Recombinant M tuberculosis N-acetyltransferase

A search of the genomic sequence of Mycobacterium tuberculosis by the inventors, led to the surprising observation of a N-acetyltransferase like sequence. Many of the known NAT proteins, including the S. typhimurium enzyme, acetylate isoniazid, an anti- mycobacterial drug to which some strains of M. tuberculosis are resistant. This putative gene for arylamine N-acetyltransferase (NAT), once isolated from an isoniazid sensitive strain of M. tuberculosis, was cloned into expression vector pET28b (Novagen Inc.). This vector contains a T7 polymerase promoter, regulated by IPTG via a lac operon, which transcribes any insert with a T7 polymerase promoter. No published data on recombinant M. tuberculosis NAT exists. In the case of this insert the nucleotide sequence was 56% homologous to a gene from S. typhimurium with proven NAT activity and was calculated to have an open reading frame (ORF) corresponding to a 32 kDa protein. In order to determine that the M. tuberculosis gene encoded an enzyme possessing NAT activity, it was expressed and tested for activity using a colourimetric assay (3). Combined supernatants and pellets from cultures of BL21 E. coli containing the cloned M. tuberculosis NAT in pET28b appeared to be enzymatically inactive when assayed colourimetrically with anisidine, an arylamine substrate.

To determine if soluble M. tuberculosis NAT could be produced further expression studies were performed under a variety of conditions. In case the concentrations of soluble protein were too small to be resolved by SDS page, the colourimetric assay was used to determine NAT activity. Initially, expression of insoluble protein was shown to increase with time after induction. Similarly, expression of insoluble protein over time decreased with temperature down to 12°C whilst no corresponding increase in soluble protein was observed.

Expression of the M. tuberculosis NAT clone produced inactive inclusion bodies of the 32 kDa protein, as visualised by denaturing SDS-PAGE. This is the size expected for NAT, although it showed no NAT activity. The protein was not soluble and cytosolic enzymes of this type only show activity when they are expressed in a soluble form. The formation of inclusion bodies allowed the protein to be purified using simple physical techniques. N-terminal amino acid sequencing of the purified recombinant protein further confirmed its identity.

To determine whether recombinant M. tuberculosis NAT could metabolise arylamine substrates, such as isoniazid, it was necessary to refold the insoluble protein into an active state. The M. tuberculosis NAT was denatured in 6 M urea; a standard denaturant for refolding inclusion bodies. M. tuberculosis NAT at 50 mg/ml is almost entirely soluble in 6M urea. Urea at 6 M did not affect NAT activity but did alter the absorbance of the solution at 450 nm in a linear fashion. The detection limits of the apparatus allowed measurements to be made in up to 60 mM urea, with greatest confidence at urea concentrations of less than 10 mM. A solution of M. tuberculosis NAT dissolved in 6 M urea was dialysed into refolding buffer (50 mM Tris-HCI, 2 mM EDTA) at pH 8 to a final urea concentration of 9mM. A significant specific activity was seen with anisidine with confidence limits of greater than 95%.

Thus when the expressed 32 kDa protein was denatured in 6 M urea and dialysed it showed NAT activity with anisidine. This demonstrates that the cloned gene encodes a M. tuberculosis NAT. When M. tuberculosis NAT is expressed in M. smegmatis, however, it does appear to have activity with isoniazid (Fig. 5). When M. smegmatis (strain MC² 155) was transformed by electroporation with an inducible mycobacterial expression construct containing M. tuberculosis NAT, it was found that the transformed M. smegmatis containing NAT from M. tuberculosis had 50 fold increased resistance to isoniazid (Fig. 5) compared with wild type M. smegmatis. This strongly indicates that NAT from M. tuberculosis is active with isoniazid under these circumstances and supports the concept that NAT is involved in determining isoniazid sensitivity. Other experiments have demonstrated that NAT from M. smegmatis is even more active with isoniazid than the NAT from

M. tuberculosis when over-expressed in M. smegmatis using the inducible mycobacterial expression construct. NAT from M. smegmatis or from M. tuberculosis was expressed in either E. coli or in M. smegmatis. The lysates were centrifuged at 11 ,600 g for 30 min and the amount of protein corresponding to the additional band at 30 000-33 000 mol. weight on SDS- PAGE as determined by Coomassie blue staining was estimated in samples of supernatant and resuspended pellet. The percentage solubility in each host is shown below:

Although both the M. smegmatis and M. tuberculosis NAT enzymes are expressed in a soluble form in M. smegmatis, the total amount of protein induced from the two constructs is different, with the M. smegmatis construct resulting in expression of approximately five times more protein than the M. tuberculosis construct. Using this expression system (as opposed to expression in E. coli) then both M. smegmatis NAT and M. tuberculosis NAT are equally soluble. The optimum conditions for the over-expression of M. tuberculosis and M. smegmatis NAT in M. smegmatis using a vector containing either of the genes cloned downstream of an acetamidase promoter were found to be as follows: the use of minimal media containing 68 mM acetamide and 50 μg/ml hygromycin; inoculation of the minimal media from a starter culture in a 1 :200 dilution of unwashed cells; harvesting of cells at OD₆oonm = 1.1 ; sonication 4x5 min with 2 min rest intervals. Example 2 Cloning of a gene from Mycobacterium smegmatis encoding arylamine N-acetyltransferase

M. smegmatis is a very closely related non-pathogenic species which is used as a model mycobacterial system; its genetics are understood and plasmid expression systems are being developed (4,5).

To determine if M. smegmatis also possesses NAT, a g ridded library representing the M. smegmatis genome was screened using M. tuberculosis NAT cDNA. A 400 bp probe template from the M. tuberculosis NAT open reading frame was chosen because it encodes the PFENL and RGGYC regions conserved in NATs.

The 400 bp probe template was radiolabelled with a conventional kit (Amersham-Pharmacia Biotech) and used to probe a gridded library representing the M. smegmatis whole genome. A low stringency screen was performed to detect clones containing the PFENL and RGGYC regions hybridising to the probe. Six clones which hybridised strongly to the 400 bp probe were obtained from the screen. Restriction digestion with BamHI was performed to test for insert size. Results were confirmed using a Southern blot to check for hybridisation to the original probe. Most clones appeared sufficiently large to contain an entire NAT gene.

The gridded genomic library was screened. A Southern blot confirmed that all six clones of the clones obtained appeared to contain DNA encoding for PFENL and RGGdC, where d is W or Y, sequences which are also conserved in NATs. The gridded library manufacturing process, which utilises blunt end religation, creates a SatnHI site 1 in every 4 occasions. All of the clones had at least one BamHI site and so were linearised.

Clone '20F22' contained two BamHI sites and it was thus completely excised from its vector to give a smaller fragment. Clone 20F22 was selected for sequencing along with a second representative clone from the remainder (5115). Plasmids previously obtained from M. smegmatis clones 5115 and 20F22 were digested with Xho\ to excise the inserts from the pBluescript vectors (Stratagene). The fragments were gel purified and analysed. The clones were also digested with Λ/ofl to give a set of overlapping fragments. Clone 20F22 contained an internal Λ/o l site which gave two fragments. The excised inserts were ligated into pGEM11Z, (Promega). Constructs were sequenced with M13 forward and reverse primers.

The separate sequence data obtained from each subcloned fragment were pieced together into a contiguous sequence for each clone. Automated sequencing improved significantly with the use of a proprietary polymerase, optimised for the high GC content found in mycobacteria (Fidelase). The final sequences were confirmed by primer walking along both strands of DNA. Translation of the DNA from clone 20F22 revealed that it was a false positive although it showed a high degree of homology to NAT at the nucleotide level, it contained neither the conserved PFENL or RGGdC regions. Subsequent specific PCR confirmed that clone 20F22 had no NAT gene. In contrast, the translation of the DNA sequence of clone 51 15 contained 275 amino acid residue open reading frame which shows both the conserved PFENL and an RRGGYC sequence. This open reading frame is shown in Figure 2. Expression of this putative NAT from clone 51 15 allowed its identity to be confirmed (see Figs. 4 and 7).

Example 3 Mycobacterium smegmatis arylamine N-acetyltransferase recombinant protein production The 825 bp clone (5115) corresponding to the putative 275 residue NAT open reading frame was inserted into a pET28b recombinant protein expression vector (Novagen) and used to transform BL21 E. coli (Promega). The open reading frame is relatively short for a NAT protein and, given that the similar M. tuberculosis NAT ORF contains an extra 24 base pairs a second transformation was performed using a clone of the open reading frame and a further 105 base pairs beyond the termination codon. Recombinant soluble protein with a relative molecular weight of 32,000 was expressed from both transformed bacterial colonies consisting of 30,000 of M. smegmatis NAT protein and 2,100 from the fusion hexahistidine affinity tag. The specific activity of recombinant M. smegmatis with a variety of substrates was measured using a 60 minute assay for soluble NAT expressed from the 875 bp M. smegmatis insert. These assays, the results of which are shown in Figure 4, demonstrated that the recombinant enzyme had strong activity for isoniazid, in addition to activity for 4-AV and anisidine but poor activity with p-aminobenzoic acid.

Example 4 Raising antibodies to Mycobacterium tuberculosis arylamine N-acetyltransferase

Polyclonal antibodies raised against M. tuberculosis NAT provide a useful tool for specific protein detection. For instance the detection of native protein expression from lysates isolated from drug resistant or sensitive M. tuberculosis would be essential for understanding NAT'S ability to metabolise isoniazid in vivo.

M. tuberculosis NAT inclusion bodies were isolated using sonication. Nonionic detergents were used to remove membrane proteins by the following method:

The resuspended pellets were spun down at 10,000 x g for 10 minutes at 4°C. The resuspension buffer (50 mM Tris-HCI, 2 mM EDTA, 4 mM DTT and 1 mM Pefabloc (protease inhibitor) at pH 8.0) was replaced with an equal amount of detergent buffer (200 mM NaCl, 1% (w/v) deoxycolic acid, 1% (v/v) Nonidet P40. The pellet was resuspended and spun at 10,000 x g for 10 minutes at 4°C. The supernatant was discarded without disturbing the pellet or the gelatinous strands of material. The pellet was carefully resuspended in Triton-EDTA solution (1% (v/v) Triton X-100, 1 mM EDTA) and spun at 10,000 x g for 10 minutes at 4 °C.

Careful removal of the supernatant and resuspension in Triton-EDTA was repeated up to three times until the gelatinous layer was no longer evident. The pellet was resuspended in resuspension buffer.

This preparation was further purified on a 8% acrylamide gel. Gel fragments containing the protein, were mixed with an equal amount of Freund's incomplete adjuvant and injected into rabbits. Rabbits were injected using the following protocol:

Event Analysis Result

Pre-bleed & Dot Non-specific protein binding

Immunization i

Boost i

First Bleed Dot/Western No apparent antibody activity seen x3 i

^L 6 week Boosts

I

Second Bleed ELISA Detection up to 1 : 100,000 dilution

2 ml of blood were removed during bleeds to test antibodies using dot or Western blots or ELISA. 1 ml of M. tuberculosis NAT in acrylamide, mixed with Freund's incomplete adjuvant, was injected at each boost and for the priming.

Freund's complete adjuvant relies on components derived from mycobacterial cell walls to stimulate a sufficient humoral immune response to produce antibodies. To prevent non-specific recognition of other mycobacterial proteins or cell wall components, Freund's incomplete adjuvant was used. An adequate immune response was achieved by leaving the protein in acrylamide. The resulting antibodies have been shown to strongly bind M. tuberculosis NAT and cross react weakly with S. typhimurium NAT.

Antibodies to a peptide of M. tuberculosis NAT were also raised. The hydrophobicity plots generated from the sequences of M. tuberculosis NAT and M. smegmatis NAT reveal differences between the C-terminal regions and a region between residues 160-180. The peptides CDELLARQPGADAP from the C-terminus of M. tuberculosis NAT and CDVQARVAEVLDT from the C-terminus of M. smegmatis NAT were synthesised and coupled to soybean trypsin inhibitor as a carrier protein for immunisation (750 μg) with Freund's Incomplete Adjuvant in rabbits, with two subsequent boosts (600 μg). Final bleed antiserum for the C-terminal peptide of M. tuberculosis NAT was found to titrate out (was no longer detectable when used as the primary antibody) against M. tuberculosis NAT whole protein in ELISA studies at a dilution of 1 :500-1 :2000, but at >1 :100 against M. smegmatis NAT and S. typhimurium NAT whole proteins. In turn the antisera raised against the C-terminus of M. smegmatis NAT titrated out at 1 :10000-1 :50000 against M. smegmatis NAT whole protein but at >1 :100 against M. tuberculosis NAT whole protein. Thus the antisera raised against these peptides are capable of distinguishing between the NAT proteins of M. tuberculosis and M. smegmatis. The C- terminal region of both proteins is an important antigenic epitope of NAT. Antibodies raised against this region are therefore useful in the detection and identification of mycobacteria in samples, for example which have been treated to release proteins or protein fragments from within the mycobacterial cell wall.

Example 5 Raising antibodies to M. smegmatis NAT

M. smegmatis NAT inclusion bodies were isolated using sonication. Triton X100 and lgepal-630 detergents were used to remove membrane proteins by the same method outlined in the previous example: This preparation was further purified on a 8% acrylamide gel. Gel fragments containing the protein, were mixed with an equal amount of Freund's incomplete adjuvant and injected into rabbits. Rabbits were injected using the same protocol tabulated in the previous example. 2 ml of blood were removed during bleeds to test antibodies using dot or Western blots or ELISA. 1 ml of M. smegmatis NAT in acrylamide, mixed with Freund's incomplete adjuvant, was injected at each boost and for the priming.

Freund's complete adjuvant relies on components derived from mycobacterial cell walls to stimulate a sufficient humoral immune response to produce antibodies. To prevent non-specific recognition of other mycobacterial proteins or cell wall components, Freund's incomplete adjuvant is used. An adequate immune response is achieved by leaving the protein in acrylamide. The antibodies raised against whole M. smegmatis NAT protein were found to cross reactive with the M. smegmatis NAT protein.

Example 6 Raising antibodies to S. typhimurium NAT

Pure NAT from S.typhimurium (0.6 mg) was used to immunise rabbits as a primary injection in Freund's complete adjuvant followed by two booster injections (0.5 mg) firstly in adjuvant and then in saline at two week intervals. A weak response was obtained. The animals were boosted with 1 mg of pure protein in Freund's incomplete adjuvant on up to 4 occasions at up to monthly intervals before the final bleed. The antibody reacts extremely strongly at 1 :100,000 dilution and monospecifically with NAT from S. typhimurium and with NAT from M. smegmatis, see figure 7

Example 7 Engineering of a Mycobacterial strain which lacks NAT activity

A M. smegmatis strain lacking a functional nat gene was created by use of a DNA construct (shown in figure 9) capable of ablating the endogenous nar gene. Gene knockout was achieved using a suicide vector containing the construct which was placed into M. smegmatis via electroporation. The vector was incapable of replicating in mycobacteria and transformed strains were selected by kanamycin resistance. A representative sample of 40 kanamycin resistant colonies were tested by PCR analysis for potential nat gene knockouts and targeted integration, this is shown in figures 10 and 11. PCR amplification using primers specific for M. smegmatis nat were used to demonstrate a lack of the gene (therefore gene knockout) in sample one sample designated KO 9. PCR amplification of the same samples was to determine integration to the 3' end and/or a targeted event.

Southern blot analysis of selected potential knock out strains of M. smegmatis was also carried out - strain KO 9 clearly depicted a nat gene knock out event - see figure 11.

It has been noted by the inventors that colonies of M. smegmatis lacking a NAT function display altered morphology. In addition, sub-cultures display improved growth - a similar phenomenon to isoniazid sensitisation seen in clinical isolates of M. tuberculosis, where isoniazid treatment which is effective initially becomes less effective upon continued exposure to isoniazid. It is thought that different metabolic pathways might be invoked in order to bypass the reduced NAT function in both NAT knockout strains of M. smegmatis and strains of M. tuberculosis which have been exposed to isoniazid for prolonged periods.

Example 8 Testing of isoniazid with engineered strains of M. smegmatis with altered NAT expression.

Lysates were obtained from cultures of wild M. smegmatis (Mc2 155), a recombinant strain that was over-expressing NAT and strain KO9 in which the nat gene has been ablated. The ability of these lysates to carry out the acetylation of isoniazid was investigated. The results are shown in figure 12. It can be seen from in figure 12 that acetylation is increased four-fold in the NAT over-expressing strain compared to the wild type strain. The K09 strain shows a depletion in acetylation relative to the other two strains.

The growth of M. smegmatis strains in the presence of isoniazid was also investigated. A wild-type strain (Mc2 155), a strain over- expressing M. tuberculosis NAT (TBNAT) and strain KO9, lacking a NAT function were cultured in 7H9 media supplemented with acetamide, in different concentrations of isoniazid (1 , 2 and 10 μg/ml). The results are shown in figure 13. It can be seen that TBNAT demonstrates an increased resistance to isoniazid and the nat gene knock out demonstrates an increased sensitivity to isoniazid.

Example 9 Compounds shown to be NAT substrates and anti- mycobacterial agents

To determine whether it would be possible to use an aliphatic chain to attach compounds to a gel and use this method to determine association constants between compounds and NAT. A range of long chain (C2-C12) aminophenolethers were synthesised according to Nodzu et al. (J.Pharm.Soc.Jap. 1954, 74, 872-5). The compounds synthesised were the ethyl (2 carbon atoms, abbr. Et), butyl (4, But), hexyl (6, Hex), octyl (8, Oct), decyl (10, Dec) and dodecyl (12, Dod) with the chain in the 2, 3 and 4 positions.

(4, 3, 2-isomer) R=Et, But, Hex, Oct, Dec.Dod 4, 3 and 2-isomers The synthesised compounds were assayed with S. typhimurium NAT to determine the rate of acetylation. The 200 μl assay contained 9% DMSO, 91 % 20mM Tris-HCI buffer pH 7.5, 90 μM compound, 400 μM Acetyl CoA and 1 μg NAT. The rates of acetylation were compared with an identical solution without the NAT enzyme and the difference attributed to the enzyme. The results were as follows:

These are illustrated on the graph in Figure 14. The 4-isomer compounds also show activity with NAT from E. coli, M. smegmatis and M. tuberculosis

The para (4-isomer) and ortho (2-isomer) compounds were additionally tested for an inhibitory effect upon the growth of M. smegmatis. Cultures of M. smegmatis were grown in the presence of approximately 25 mg/ml of each compound and a measurement of growth via an optical density reading at 580nm was taken after 24 hours. The percentage inhibition of growth for the C4, C6, C8, C10 and C12 para and ortho compounds, compared to cultures in which the compounds were absent is shown is shown in figure 15. It can be seen that the C8 para-isomer compound that has highest NAT activity of the group also has the highest inhibitory effect upon the growth of M. smegmatis. It is therefore shown that compounds which are good substrates of NAT may be used to inhibit mycobacterial growth. The prevention of NAT from acting upon endogenous substrates as a result the exogenous substrate being present in the culture may be the mechanism behind the impaired growth observed.

Compounds as described above can be derivatised by methods known in the art, in order to allow coupling to a solid support for example in a screen. Not only is it desirable to find compounds which prevent mycobacterial NAT from acetylating anti-mycobacterial drugs which have targets elsewhere in the bacterium, but these experiments show that compounds which are acetylated by NAT can act as anti-mycobacterial drugs.

Example 10 Detection of NAT mutants of M. tuberculosis

NAT genes from 16 clinical isolates of M. tuberculosis isoniazid resistant strains were amplified by PCR, cloned and both strands sequenced. Two strains displayed a genuine allelic variant of the wild-type

TB NAT sequence as follows:

These base changes with respect to the wild-type M. tuberculosis NAT sequence were observed in sequencing data of both DNA strands, and confirmed using RFLP analysis with restriction enzyme Bs AI.

To carry out the RFLP analysis the M. tuberculosis NAT sequence open reading frame was amplified from DNA obtained from M. tuberculosis clinical isolates displaying isoniazid resistance using the following primers:

5'GACGAGGTCAGAATGGCAAC3' and δ'GGGGTTCGTTTGTTCGGATAS'

The resulting amplification products were digested with

BsmAI (New England Biolabs). The 619 G to A (207Gly to Arg) mutation introduces an extra BsmAI site in mutant sequences, thus the expected fragments from the restriction digest are as follows:

Electrophoresis on a 1.5% agarose gel of a 1 kb molecular weight marker ladder, the undigested product and the restriction digest products showed that 4 isolates out of a total of 14 possessed the 619 G to A mutation.

Thus the analysis of mutation in the NAT gene can give an indication of how sensitive a particular mycobacterial strain is to isoniazid and other NAT ligands.

Conclusion

The sequence comparison (Figure 6) of the mycobacterial NATs show that the NAT from mycobacteria are slightly different in size from other known NATs. The mammalian NATs have 290 amino acids, that from S. typhimurium and E. coli have 281 amino acids. The M. tuberculosis NAT has 283 amino acids and the M. smegmatis has 275 amino acids. In the mycobacterial NATs the active site cysteine and the arginine which has been implicated in catalysis are conserved Comparative hydrophobicity plots show a 12 residue region, 162-174 where M. tuberculosis NAT is significantly more hydrophobic than M. smegmatis NAT. In the absence of a 3-D structure it is difficult to tell whether this region could induce misfolding during E.coli expression. However it is immediately adjacent to the proposed interdomain loop in eukaryotic NATs. It is therefore possible that these residues form an exposed hydrophobic region which could cause aggregation or otherwise prevent correct folding of M. tuberculosis NAT.

Whilst the NAT enzyme from M. tuberculosis is not active when expressed in lysates of E coli, it could be refolded to a form which was active with anisidine as substrate. The NAT enzyme from M. tuberculosis is active with isoniazid as substrate when expressed in another mycobacterium (M. smegmatis) (Figure 5).

The stability of the two enzymes may be different. These differences may contribute to the inherent differences of M. tuberculosis and M. smegmatis in sensitivity to isoniazid. Mutations which affect either the NAT protein itself or the expression of the NAT protein in M. tuberculosis may increase NAT activity in isoniazid resistant strains of TB. Administration of ligands of mycobacterial NAT, together with isoniazid may increase its effectiveness against other pathogenic mycobacteria, such as M. leprae which causes leprosy or M. bovis which infects cattle. The current work gives new insights into the mechanism of isoniazid metabolism in mycobacteria and may help tackle the current resurgence in tuberculosis, the largest cause of preventable deaths in the world. References

1. Weber WW and Hein DW (1985) Pharmacological Reviews 37:25- 79

2. Hein DW et al. (1993) Carcinogenesis 14:1633-1638 3. Andres et al. (1985) Anal. Biochem. 145:367-375

4. Delomenie C et al. (1997) Biochemical Journal 323:207-215

5. Dupret JM & Grant DM (1992) J. Biol. Chem. 267:7381-7385

6. Watanabe M et al. (1992) J.Biol. Chem. 267:8429-8436

Claims

1. The Mycobacterium tuberculosis arylamine

N-acetyltransferase protein having the sequence of figure 1 or peptide fragments greater than 5 contiguous amino acid residues thereof.

2. The Mycobacterium tuberculosis arylamine N-acetyltransferase protein of claim 1 wherein the amino acid at position 207 as shown in figure 1 is arginine.

3. The Mycobacterium smegmatis arylamine N-acetyltransferase protein having the sequence of figure 2 or peptide fragments greater than 5 contiguous amino acid residues thereof.

4. The Mycobacterium smegmatis arylamine N-acetyltransferase gene having the sequence of figure 2 or oligonucleotide fragments greater than 8 contiguous nucleotide residues thereof.

5. Antibodies raised against the Mycobacterium tuberculosis arylamine N-acetyltransferase protein having the sequence of figure 1 , wherein the amino acid at position 207 may be one of glycine or arginine, or peptide fragments greater than 5 contiguous amino acid residues thereof.

6. Antibodies raised against the Mycobacterium smegmatis arylamine N-acetyltransferase protein having the sequence of figure 2 or peptide fragments greater than 5 contiguous amino acid residues thereof.

7. Antibodies raised against the Salmonella typhimurium arylamine N-acetyltransferase protein having the sequence of figure 3 or peptide fragments greater than 5 contiguous amino acid residues thereof

8. A method of screening for compounds that bind to a

Mycobacterial N-acetyltransferase protein by bringing one or more molecules of an N-acetyltransferase protein into contact with one or more compounds to be screened for binding and detecting said binding.

9. A method as claimed in claim 8 comprising the steps of :

(i) immobilising one or more test compounds on a solid support;

(ii) bringing an aliquot of N-acetyltransferase protein into contact with the said immobilised test compounds;

(iii) removing unbound protein from the solid support; (iv) detecting the bound protein.

10. The method of claim 8 or claim 9 wherein the

N-acetyltransferase protein is arylamine N-acetyltransferase from a Mycobacterium or a fragment thereof.

11. The method of claim 8 or claim 9 wherein the N-acetyltransferase protein is arylamine N-acetyltransferase from Mycobacterium tuberculosis or a fragment thereof, or Mycobacterium smegmatis or a fragment thereof.

12. The method of claim 8 or claim 9 wherein the N-acetyltransferase protein is arylamine N-acetyltransferase from Salmonella typhimurium or a fragment thereof.

13. The method of claim 8 or claim 9 wherein a screened compound that is found to bind to a Mycobacterial N-acetyltransferase is subsequently tested for binding to Human NAT1 or NAT2 or other eukaryotic NATs.

14. The method of claim 13 wherein detection of bound human NAT1 or NAT2 protein in the subsequent test is by means of antibodies raised against a Human arylamine N-acetyltransferase or a fragment thereof.

15. The method of any one of claims 8 to 13 wherein detection of bound protein is by means of antibodies raised against arylamine

N-acetyltransferase from Mycobacterium tuberculosis or a fragment thereof.

16. The method of any one of claims 8 to 13 wherein detection of bound protein is by means of antibodies raised against arylamine

N-acetyltransferase from Mycobacterium smegmatis or a fragment thereof.

17. The method of any one of claims 8 to 13 wherein detection of bound protein is by means of antibodies raised against arylamine N-acetyltransferase from Salmonella typhimurium or a fragment thereof.

18. A method as claimed in any one of claims 8 to 17 wherein the compounds being screened for are ligands of a Mycobacterial arylamine N-acetyltransferase protein.

19. A method as claimed in any one of claims 8 to 18 wherein the compounds being screened for are those that modulate the acetylation of aromatic amines and hydrazines by said Mycobacterial arylamine N-acetyltransferase protein.

20. A method of controlling the growth of a Mycobacterium by bringing said Mycobacterium into contact with a compound which is a ligand of a arylamine N-acetyltransferase protein expressed by said Mycobacterium.

21. The method of claim 20 wherein the ligand is a substrate of the arylamine N-acetyltransferase.

22. The method of claim 20 wherein the ligand is an inhibitor of the arylamine N-acetyltransferase.

23. The method of claim 20, wherein the Mycobacterium is Mycobacterium tuberculosis, Mycobacterium leprae or Mycobacterium bovis.

24. The use of a compound which is a ligand of a mycobacterial arylamine N-acetyltransferase protein, in the manufacture of a medicament for the treatment of mycobacterial infections.

25. An anti-mycobacterial formulation which contains an effective concentration of a compound which is a ligand of a mycobacterial arylamine N-acetyltransferase protein.

26. The anti-mycobacterial formulation of claim 25 wherein the ligand is a substrate of the arylamine N-acetyltransferase protein.

27. The anti-mycobacterial formulation of claim 25 wherein the ligand is an inhibitor of the arylamine N-acetyltransferase protein.

28. The use of a compound which diminishes the activity of a

Mycobacterial arylamine N-acetyltransferase protein for a second compound which is an aromatic amine or hydrazine, in the manufacture of a medicament for the treatment of mycobacterial infections.

29. An anti-mycobacterial formulation which comprises a mixture of an effective concentration of an anti-mycobacterial aromatic amine or hydrazine and an effective concentration of a compound capable of diminishing the activity of a mycobacterial arylamine N-acetyltransferase protein.

30. The anti-mycobacterial formulation of claim 29 wherein the anti-mycobacterial aromatic amine or hydrazine is at least one of the group consisting of isoniazid, ethionamide, dapsone and p-aminosalicylate.

31. Use of the antibodies of any one of claims 5 to 7 for detecting mycobacteria in a biological sample.

32. A method of detecting mycobacteria in a biological sample which comprises the steps of: a) providing a biological sample which has been treated so as to allow the release of cytoplasmic components of mycobacteria present in the sample, b) bringing said treated sample into contact with one or more antibodies raised against a mycobacterial arylamine N-acetyltransferase, c) determining the binding of said antibody to one or more cytoplasmic components of the sample.

33. The method of claim 32 wherein the sample is treated by sonication.

34. A method of identifying nucleic acid sequences which at least partially encode, or amino acid sequences which at least partially comprise, putative enzymes which are potential targets for anti-mycobacterial drugs or putative enzymes useful for the synthesis of anti-mycobacterial drugs, by means of identifying within said sequences nucleotides encoding or amino acids comprising, protein sequences with similarity greater than 80% to the following sequence:

PFENL(X)_n=_5-5oRGGdC.

35. The method of claim 23 wherein "n" is between 20 and 30.

36. A method of isolating DNA sequences, from any source of genomic or cDNA, which at least partially encode putative enzymes which are potential targets for anti-mycobacterial drugs or putative enzymes useful for the synthesis of anti-mycobacterial drugs, by means of a primer extension reaction which uses one or more sense primer which encodes the amino acid sequence PFENL and one or more anti-sense primer which encodes the amino acid sequence RGGdC, where d is either W or Y.

37. The method of claim 36 wherein the product of the primer extension reaction is subsequently labelled and used as a probe to screen a DNA library.

38. The method of claim 36 wherein the sense primers used are a mixture corresponding to: 5" CCNTTYGARAAYYTN 3'; and the antisense primers used are a mixture corresponding to: 5' RCAVYANCCNCCNCK 3'.

39. The use of the three dimensional structure of a Mycobacterial arylamine N-acetyltransferase protein for the design of compounds capable of binding to said arylamine N-acetyltransferase protein.

40. The use of the three dimensional structure of a Mycobacterial arylamine N-acetyltransferase protein for determining or predicting the structure of a further arylamine N-acetyltransferase protein and using the structural information so obtained for the design of compounds capable of binding to said further arylamine N-acetyltransferase protein.

41. Mycobacterial strains transformed with a synthetic DNA construct encoding a protein of any one of claims 1 to 3.

42. The mycobacterial strains of claim 41 wherein the DNA construct encodes a hybrid protein of the proteins of any one of claims 1 to 3.

43. Mycobacterial strains transformed with a synthetic DNA construct which abolishes endogenous arylamine N-acetyltransferase activity.