GB2389364A - Mycobacterial vaccines - Google Patents

Abstract

This invention relates to attenuated strains of the Mycobacterium tuberculosis complex, methods for producing and screening these, methods for assessing their effectiveness as vaccines and the use of such strains and derivatives of them as vaccines against tuberculosis in human and animal medical practice.

Description

MYCOBACT1:RIAL VACCINF.S

TECHNICAL FIELD

['his invention relates to attenuated strains of the.ycobucterizm tliherCZflO.YiS complex, methods for producing and screening these, methods for assessing their effectiveness as vaccines and the use ol' such strains and derivatives of them as vaccines against tuberculosis ill human and animal medical practice. l BACKGROUND ART

Mycobacteria are rod-shaped. acid-i'ast. aerobic bacilli that do not form spores. A moderate number of mycobacterial species are pathogenic to humans and/or 10 animals and determining improved methods of prevethl or trcati, the diseases they cause is of prime importance. for example tuberculosis is a worldwide human health problem whicl1 causes 2-3 million deaths per year At no highly eficctivc vaccine is available for its prevention. Tuberculosis in animals is also very widespread", particularly in developing countries.

15 'I'uberculosis in mammals is caused by species ol'thc Mycohacterun' taheiculosi.s complex. 'these species arc so closely related genetically that it has been proposed that they be combined into a single species (van Soolingen et al., 1997) and they are now generally treated at Icast informally in this way (Brosch et al., 2002).

l hrcc important members of the complex are M tzherculo.sis, the major cause of 20 human tuberculosis; Mycohacterir.m uiricumim' a major cause of human tuberculosis in some populations; and Mycobucterun? bovis, the cause of bovine tuberculosis. Other members of the complex include Mycohac/eri'ni microti.

Mycohucterum bov'.y SU6Sp. caprae icmann et al., 2002), Mycobacterinn? canettii (Pfyffer et al.7 1998) and a discrete group ol'isolates from seals (Brosch et 25 al., 2002). Nonc of the species ot'thc M taherc''losis complex is restricted to

being pathogenic for a single host species. ['or example, M hovi.s causes tuberculosis in a wide range of animals including, humans in which it causes a disease that is clinically indistinguishable from that caused by M tuberculosis.

Human tuberculosis is a major cause of mortality throughout the world, 5 particularly in less developed countries. it accounts for approximately eight million new eases of clinical disease and two million deaths each year. Bovine tuberculosis as well as causing a small percentac ol' these iunan cases, is a major cause of animal suffering and large economic costs in the animal industries.

There is a long-established vaccine for tuberculosis that is an attenuated form of 10 M. bovi.s known as BCG. This is very widely used but it provides incomplete protection (Bloom and Zinc, 1994). Recently, gene deletions that may contribute to the avirulence of BCG have been identified (Mahairas e' al. 1996; Behr et 1., 1999; Gordon el al.. 1999! and in one case a gene in one of these deleted regions has been shown to be important for virulence (Wards en al., 2000). 'I'here are 15 undoubtedly other genes associated with virulence of the M tuberculosis complex that have also been mutated in BCG but whose identity is not yet known or not yet reported. For example' we have identified a nonsense mutation in 13CG in the phol gene and showed that specific inactivation of this gene in a virulent M hovi.s strain caused it to become attenuated in guinea pigs (results not published).

20 L)espite this developing knowledge of gene mutations and deletions in BCC, that cause loss of virulence, the detailed mechanisms which enable BCG to have some vaccine ei:'fectiveness have not been detennined.

In the last ten years, there has been intense interest in producing better vaccines against tuberculosis. A large range of possible types of both living and non-living 2O vaccines are being actively researched (Collins, 2000; Snewin et al., 2000) and there is already some information on the ability of different types to stimulate the immune system in different ways. There is general agreement that a successful

vaccine will need to induce cell mediated immunity. but since no convincing correlates of protective immunity have yet been identified, it is not clear which lcaturcs of a particular type of vaccine are most important (Sncwin et al., 7000.

Live vaccine candidates include genetically modified forms of BCG,, cnetically 5 attenuated strains of the M. tuberculosis complex, and genetically engineered vaccinia virus and Salmonella strains. Non-living vaccine candidates include killed mycobacterial spccics, protein subunits and DNA vaccines. It may be that different types of vaccine will be uscl'ul for different purposes.

Thirteen years ago, we embarked on a programme lo produce a new tuberculosis 10 vaccine based on attenuated strains of the M. [zherculosis complex. Until the recent development of molecular genetic techniques for producing mutant strains ol'the M tahercu1Osis complex, the only attenuated strains available were either those which had been available for many years such as M havis BCG. M hovis ATCC35721, Mycobuclerium micron and M tuherculosis IT37Ra or new strains la made in the laboratory by mutapcnic techniques whicl1 produced,cnctic lesions that were not easily identifiable even after the application of modern molecular genetic approaches. In early work in our laboratory, we identified an attenuating mutation in M: bovis ATC('35721 (Collins et up., 1995), isolated avirulent strains of M. havis by Isoniazid selection (Banerjee et al 1994; Wilson et 'l., 1995) and 20 produced avirulent strains of M bovi.s by chemical mutagenesis (Buddle en al, 2002). For a variety of reasons, none of these mutant strains were suitable candidates for a new vaccine, although two of them in vaccination experiments did provide better protection than BCG against tuberculosis in cattle (huddle en al., 2002). Nevertheless, the work was useful because it clearly demonstrated that burl a better live vaccine than BCG could be produced.

The advent of a range of mycobacterial molecular genetic techniques in the last ten years has made it possible to produce both random and specific mutants of the

tuberculosis complex in which the genetic lesions can be easily identified (Collins, 2000; Collins and Gicquel 2000). These techniques are being used in two broadly different ways as the basis for identifying genes that are important for virulence. In the first approach, random mutants are produced by transposon mutagenesis or o illegitimate recombination, screened in some way to identify those that have lost virulence, checked for their loss of virulence, and the causative mutation fUr this IOSS ol' virulence is idenlilied. In the second approach, a gene that is thougllt likely to be important for virulence is inactivated by allelic exchange techniques and the virulence ol'the maiated strain then determined. Any strains found to be avirulent 10 by either approach can then be tested for their vaccine potential in an appropriate animal vaccine model. Many avirulent strains have now been produced by us and most have been tested for their vaccine ability.

All rel'erences, including any patents or patent applications cited in this specification are hereby incorporated by reference. No admission is made that any

15 reference constitutes prior art. The discussion of the references states what their

authors assert, and the applicants reserve the right to challenge the accuracy and pertinency of the cited documents. It will be clearly understood that, although a number of prior art publications are referred to herein, this reference does not

constitute an admission that any of these documents form part ol' the common 20 general knowledge in the art, in New Zealand or in any other country.

It is acknowledged that the term 'comprise' may, under varying jurisdictions, be attributed with either an exclusive or an inclusive meaning. For the purpose of I this specification, and unless otherwise noted, the term 'comprise' shall have an

inclusive meaning - i.e. that it will be taken to mean an inclusion of not only the 25 listed components it directly references, but also other non-specified components or elements. 'I'his rationale will also be used when the term 'comprised' or comprising' is used in relation to one or more steps in a method or process.

a It is an object of the present invention to address the foregoing problems or at least to provide the public with a useful choice.

Further aspects and advantages of the present invention will become apparent from the ensuing description which is given by way of example only.

5 Su MMARY OF l NVENT ION - The present invention relates to the identification, production, screening and use of attenuated mycobacterial species as live vaccines against mycobacterial diseases in man and animals. More particularly, it relates to mutant strains of the M. ubercul<'s.s complex, that arc sufficiently attenuated so that they do not cause 10 progressive disease when they are inoculated into animals that are susceptible to prog,rcssivc disease when infected with wildtypc species ot' the.tl. /uberenlosis complex. The mutant strains are produced by several methods including illegitimate recombination, transposon mutagenesis and allelic exchange and are screened by a variety of methods including ability oi'S-10 week-old cultures to lore grow when inoculated into minimal medium, visual changes in rmorpllology, severe limitation of growth in bovine alveolar macrophages, and inability to be rccovcrcd from guinea pig spleens in a Signature Tag Mutagenesis approach.

Mutants from the screening approaches are then tested individually for their ability to cause disease in guinea pigs and if they achieve a standard of attenuation, for 20 their ability to protect guinea pigs against tuberculosis in a guinea pig vaccine model. Strains that show similar or better protection in guinea pigs than BCG are in some cases then tested for their ability to protect Australian brushtailcd possums (7richosurus vulpecula) against infection. The site of mutation and the identity of the gene or genes af'f'ectcd by the mutation are also identilicd. 'I'hc 25 invention also claims the use of these strains' derivatives of these strains, and other strains with mutations in one or more identical or similar genes to these

G strains as vaccines against mycobacterial disease and more particularly as vaccines against tuberculosis in human and animal medical practice.

DISCLOSURE OF INVENTION

The practice of the present invention will employ, unless otherwise indicated, 5 conventional techniques ol' molecular biology, microbiology, recombinant DNA, immunology and vaccinology which are within the ordinary knowledge of a person skillful in the art. Such techniques arc explained folly in the literature.

According to a first aspect of the present invention there is provided an isolated Mycobacterian? species of the M. taherculosis complex which is attenuated and 1() comprises at least one inactivated gene selected Prom the group consisting ol' - ppiA; - glnA2; - Rv0()97; - any of the genes between and including Pv()445c (.sixK) - Rv()453 (PPE); 15 - any ol'the genes betwocn and including Rv ()175 - Rv ()Ifs'rS (h,lS'); - any of the genes between and including Rv Ol l l - Rv 0115; - Pv 3545c - any of the genes between and including Rv3 757c (pro W) - Rv3 760.

According to a second aspect there is provided an isolated Mycohacterium species 2() of M. /uberculo.ss complex which is attenuated and comprises inactivated or deleted genes selected from group consisting of:

- all genes between and including Rv0445c (.sigK) - Rv0453 HIPPY); all genes between and including RvOI 75 - Rv()l Y6 (hylS); all genes between and including RvOI I 1 - Rv()115; Preferably, the Mycohacterium species of the M tuberculosis complex may be G selected from existing or new members of the M: tubcrculos'As complex currently comprising: M taherculosis, M ufrcunum' M micrrti, M hovi.Y subsp. caprae, cunettii an! M hovi.Y.

Most preferably the species selected from the M. lUherCUlO.Yi.Y complex may be M hovi.Y. 10 According to a third aspect at the present invention there is provided all isolated Mycobucterium species substantially as described above wherein the strain is selected from the group consisting of: WAg759: M. hovis mutant with an inactivated ppiA gene; WAgS30: M hovis mutant with an inactivated glnA2 gene; 15 WAgS33: M horrid mutant with an inactivated Rv0097 gene; WAgS39: M bovis mutant with inactivation of the genes Rv()445c (sigK) -

Rv()453 (PPE); WAgS26: M hov.s mutant with inactivation of the genes Rv() 175 - RvOlY6 (bglLS); 20 WAgS69: M hovis mutant with inactivation oi the genes Rv()l l l - Rv()l I S. WAg570: M hovis mutant with an inactivated Rv3545c gene;

WAg537: M hot its mutant with an inactivated Rv(()97 gene; WAg566: Al: hovi.s mutant with inactivation of the genes Rv3757c (pro W) Pv3 76() According to a fourth aspect of the present invention there is provided an isolated 5 51ycobacferium species substantially as described above which also conprises an inactivated esafti gene.

Aeeording to a fifth aspect ol'the present invention there is provided an isolated Mycobucferium species substantially as described above wherein the strain comprises an inactivated glnA2 gene and also comprises one or more inactivated JO genes selected from the group consisting of Rv3N'74, 1?v38'75 and lv3rY76.

According to a sixth aspect of the present invention there is provided an isolated MycoLccferium species substantially as described above wherein the strain is WAgS30. I. According to a seventh aspect of the present invention there is provided an 15 isolated Mycohccerium species wherein the strain comprises an inactivated Rv2136 gene and also comprises one or more inactivated genes selected from the group consisting of Rv3874, Rv3Y75 and Rv3876.

According to an eighth aspect of the present invention there is provided an isolated Mycohacferium species substantially as described above wherein the 20 strain is WAg520.4.

According to a ninth aspect of the present invention there is provided an isolated Mycohc':cterium species which is attenuated and comprises one or more inactivated genes that are identical or have at least 70%-99 /0 nucleotide sequence homology to that of'the inactivated genes as claimed above.

According to a tenth aspect of the present invention there is provided an isolated Mycobucterium species of the M tubereulo.sis complex which is attenuated and comprises one or more mactvated gene(s) which is the same gene(s) as that inactivated by polar el'feets of any of the inactivated genes as claimed above.

5 According to an eleventh aspect ol' the present invention there is provided an isolated Mycobucterium species of the M tuberclosi.Y complex whiel1 is attenuated and comprises at least one inactivated gene selected from any one of the inactivated genes rel'erred to above.

According to a twell'th aspect of the present invention there is provided an isolated 10 Mycohaeterium species off the M. laherculo,Yi.s complex which is attenuated and comprises at least one inactivated gene that is identical or has at least 70-99% nucleotide sequence homology to any one of the inactivated genes referred to above. According to a thirteenth aspect of the present invention there is provided an 15 isolated Mycobucterium species substantially as described above wherein the mutant is further attenuated by inactivation of an additional gene.

According to a t'ourteenth aspect of the present invention there is provided a method of producing a live attenuated vaccine strain of any Mycohacterium species including M. avium and its subsp. avizm, parataherculos.s and Si/vatiC?im 20 and M. ulcercm.s' comprising steels of inactivating in these species one or more genes that are identical or have at least 50%-99% nucleotide sequence homology to that of the inactivated genes relcrred to above.

According to a fifteenth aspect of the present invention there is provided a vaccine to prevent or treat tuberculosis infection in mammals wherein the vaccine 25 comprises an isolated Mycabucterium species substantially as described above.

Preferably, the vaccine also includes a pharmaceutically or veterinarily suitable carrier or diluent.

Most preferably, the vaccine may also include an adjuvant or other immuno stinulant. 5 According to a sixteenth aspect of the present invention there is provided a vaccine substantial Iy as described above wherein the; LIycobucterium species comprises one or more foreign genes that are capable of enhancing the ability of the vaccine to stimulate the immune system of the diseased marginal to increase the effectiveness of said vaccine.

10 Preferably, the foreign gene may encode a polypeptide antigen, a cytokine or other immuno-stimulant. -I According to a seventeenth aspect of the present invention there is provided a composition to prevent or treat tuberculosis infection in mammals wherein the composition comprises an isolated Mycohaclerium species substantially as 15 described above together with a pharmaceutically or vetcrinarily suitable carrier or 3 dilucnt. According to an eighteenth aspect of the present invention there is provided a method of screening mutants of a Mycobacierium species of the M tUbe/ACll{J.YiS complex for attenuation comprising the steps of: 20 a) isolating a Mycoh6cterium species to be screened; b) growing the isolated mutants for 5-10 weeks in a complete medium, c) transferring the 5-] 0 week old mutants to a minimal media; and d) selecting mutants based on their inability to grow in a minimal media.

According to a nineteenth aspect of the present invention there is provided a method of producing and screening mutants of a Mycobuceri''m species of the M tuherculo.s.Y complex for attenuation comprising the steps of: a) constructing a number of unique tagged suicide vectors containing a 5 marker gene (e.g. antibiotic resistance gene); b) introducing the vectors to a culture of a 'lfycohaeferzan species and optionally leaving a predetermined amount of time; e) plating onto a suitable medium to select tor marker gene; d) selecting mutants which contain marker gene, combining mutants together, 10 and probing with DNA to identify the presence of all mutants in an Input pool; e) injecting thc'input pool' into a test mammal; I) collecting mutant colonies *om test mammals to create an 'output pool'; g) probing 'output pool' to identii'y mutants which are no longer present (ie.

15 attenuated mutants).

(he methods for introducing vectors into a myeobacterium may be performed by using methods well known in the art. Such methods may include but should not be limited to eleetroporation and phage transduction.

lithe test mammal may generally be selected from the groups consisting of', mice, 20 guinea pigs, rats, possums. cattle, deer, t'errets, badgers and rabbits.

Preferably the test mammal may be a guinea pig.

According to a twentieth aspect of the present invention there is provided a use of mycobacterium substantially as described above in the manufacture ot a vaccine or composition to treat or prevent tuberculosis infections in mammals.

According to a twenty-first aspect of the present invention there is provided a 5 method of determining whether a candidate drug is capable of inhibiting a polYpeptide involved in mycobacterial iniection comprising the steps of: a) Providing an isolated polypeptide encoded by one of the genes claimed above; b) Providing a candidate drug; 10 c) Use ol an assay which measures biological activity of the polypeptide in a); d) Measuring inhibition of biological activity of the polypeptide in a).

According to a twenty-second aspect of the present invention there is provided a method of determining whether a compound is an effective drug candidate against tuberculosis comprising the steps of 15 a) producing a stationary phase culture of a strain of the M. therculo.si.s complex b) subculturing it in a minimal medium with the addition of a compound that is a potential drug c) comparing its growth in b) to that of a similar subculture in the same 20 medium without the added compound the strains may he subcultured by means well known to those skilled in the art.

As used herein "stationary phase" refers to the plateau of the growth curve after log growth in a culture, during which cell number remains substantially constant such that new cells are produced at a generally commensurate rate to that at which cell death occurs or in which cell numbers slovvly decline as cell death occurs 5 fstcr than new cells arc produced.

According to a twenty-third aspect of the prescot invention there is provided a method of detecting whether a drug is effective against mycobacterial int'cction conprising the steps of: a) administering to an animal having a mycobactcrial infection a candidate 10 drug as selected by the methods ol'the present invention; and b) assessing effectiveness ot'the drug at reducing or eliminating, the clinical symptc\ms of said infection.

The present invention provides methods to screen for attenuated mutants of the cobacterium tuberculosis complex some of which will have utility as vaccines 15 against tuberculosis. The present invention also provides the attenuated mutants detailed in Table I that have been shown to have vaccine utility as well as any strains of the M tuberculosis complex with similar or related mutations that also have vaccine utility.

In addition, the present invention also provides the attenuated mutants in 1ablc 2 2() that have not yet been shown to have vaccine utility.

As used herein the term i attenuated mutants" denotes strains that have been tested l'or their ability to cause disease in animals and have been shown not to cause any disease or to cause a much diminished disease than wildtype strains of the M tuberculosis complex.

More particularly, these attenuated mutants have been shown not to cause any grossly visible tuberculosis lesions in the spleens of guinea pigs when 1Os _ 10 colony forming units (CLU) of the mutants were Inoculated sub-cutaneously on one occasion into three guinea pigs each and the animals were autopsied 8-10 5 weeks later (except in the case of WAg570, where one ol'thc three guinea pigs had one very small lesion in the spleen). This does not necessarily imply that such strains would never cause any lesions in guinea pigs but was a convenient criterion by which to identify those mutants that were highly attenuated relative to wildtype strains. In comparison, moderately virulent strains such as M hovis 10 ATCC35723 and M tuberculosis H37Rv will cause a mean oi'3-15 grossly visible tuberculosis lesions per spleen when inoculated into 3 guinea pigs and recent wiLltype isolates ol' M h'vis such as WAg20(), WAg20], WAg203 and AF2122/97 will cause a mean of greater than 30 such lesions per spleen when inoculated into 3 guinea pigs. If a mutant strain is attenuated, this indicates that 15 the expression ol' one or more genes that arc associated with virulence of' the wildtype organism has become altered and most likely that the insertion of foreign DNA that occurred into the chromosome has disrupted and prevented the expression of a gene or genes at or near where the inserted DNA integrated. These genes whose expression is disrupted or prevented are associated with one or more 2() virulence properties of the microorganism. 'l'hese properties arc listed in 'I'ablc 3.

111 the strains listed in Tables 1 and 2 contain kanamycin and/or hygromycin resistance genes as a result of the selection methods used in their production. In order to licence a vaccine for use it may be desirable to remove these antibiotic resistance genes or to produce vaccines strains of the M: tuherculosis complex 25 without using such genes. 'his patent covers the production of vaccines without such antibiotic genes. however that its accomplished. Approaches could include use of other selection genes such as heavy metal resistance (Silver and Phung,

1996; Bau]ard et al., 1995), phage incompatibility factors (pratfall et al., 1994), phage excision or resolution factors (Lewis and llatfull, 2() 00, Steyn e' al., 2002) or counter-selection techniques (Parish and Stoker' 2t'00; 2eyrat ct al., i998).

This patent covers the further modification of vaccines claimed in this patent by 5 over expressing one or more genes ol' the M. tuberculosis complex. Such an approach has been shown to improve the protection provided by M bovis BCG (Horwitz et at, 2000) and a similar approach could be applied to the strains claimed here.

Apart from antibiotic resistance genes, none of the strains listed in Tables I and 2 10 contain foreign genes. This patent covers the further modification of vaccines claimed in this patent by incorporating into them l'oreign genes that enhance their ability to stimulate protection against tuberculosis or enable the vaccine to have additional desirable properties such as the ability to induce protection against other disease organisms or enable speci tic products of the M. tuberculosis 15 complex to be under-expressed, over-expressed or expressed in a modified form An extensive range of foreign genes have already been expressed in M. hovis BCG with this intention and one or more of the vaccines claimed here may offer advantages over E3(:G for these purposes. Such foreign genes could be antigens, cytol;ines, chemoL:ines or other immuno-stimulants or other proteins whose 20 expression modifies the Functional properties of strains of the AzI. tuberculosis complex (Murray and Young, 1998; Pless and Kaufmann' 1')99; Ohara and Yamada, 2001). Such foreign genes are at present commonly incorporated using antibiotic resistance genes but the presence of these antibiotic resistance genes in the final vaccine could be avoided by use of the methods in the previous paragraph 25 or specific gene integration (Knipfer Ed al., 1997).

This patent also covers methods for screening compounds for their use as drugs against mycobacterial inflection. In one approach, a stationary phase culture of a strain of the JI. t'herc'lo.si.s complex is subculttrcd in a minimal medium with the addition of a compound that is a potential drug and the inhibition ol' growth of 5 the strain under these conditions is compared to that which occurs in minimal medium without the added compound. The concept of screening compounds for their potential use as drugs is well established ((global Alliance for 'l'B Drug; Development, 2001). The discovery in the work described here, that stationary phase cultures ol' random mutants that fail to grow on minimal medium are often 10 avirulent, indicates that genes important for virulence that are at a range of different loci must be active in order for this growth to occur. If a compound that specifically inhibits one of these important genes or its product is added at the right concentration to a subculture in minimal medium of a stationary phase strain of the M tuberculosis complex, it will inhibit the said gene or product and the 1, organism will not grow. Importantly, this approach targets genes and their products in the tuberculosis complex that are involved in enabling organisms to recommence cell division after being in stationary phase. This group of genes and their products are regarded as a particularly important group of targets against which to produce new drugs (Young, 2001; McKinney et al., 2000). In another 20 approach, this patent also provides a method for identifying polypcptides and using these to produce an assay nor determining the activity of candidate drugs (Dessert et al., 1995; Global Alliance for TB Drug Development, 2001). Isolating the polypeptide enables its structure to be determined and analysis of this structure assists in the design of new drugs and the development of'said assays (Sharma et 2o al., 2000; Huang et al., 2002).

The assays suitable for measuring the biological activity of a polypeptide t'rom M tuberculosis will be readily apparent to a person skilled in the art and will generally he selected dependent on: - the actually or putative biological "'unction of the polypeE'tide 5 - the actual or putative 3-D shape of the polypeptide; target molecules known to be associated with the polypeptide. The term "inactivated'' means that the ncleotide sequence of the pcne(s)

has been disrupted for example mutated or removed front the chromosome such that a functional gene product (c.g. protein) normally produced by the gene or a group of' 10 genes is no longer produced.

The term 'gene" as used herein refers to a nucleic acid molecule comprising an ordered series of nucleotidcs that encodes a gene product (i.e. specific protein).

The term "protein (or polypcptide or peptide)" ret'ers to a protci encoded by the nucleic acid molecule of the invention, including l'rag, ments, mutations and 15 homologs or analogs having the same biological activity i.e. ovulation manipulation activity. The polypeptide of the invention can be isolated from a natural source, produced by the expression of a recombinant nucleic acid molecule, or can be chemically synthesized.

he designation Rvxxxx(c) used to denote genes of the M taherculo.sis complex 20 where xxxx is a number varying from OOOl to 3924 and c refers to genes encoded on the complementary strand refers to the annotation of the complete genorne sequence of M: tuberculosis H37Rv (at http://www. sanger.ac.uk). Essentially the same genes are designated in different ways in the sequcoces of other strains of

the it tlherclosi.s complex e.g. M tllberclilosiAY CDC1551 (at I]l'll1:'', ''\\ \\ \\ [it'll' cart''') The term "polar effect" refers to a mutation which in addition to affecting the gene in which it occurs, reduces partially or completely the expression of any gene(s) in 5 the same operon on the "promoter" distal (downstream) side of the mutation.

The term 'operon" refers to a controllable unit of transcription consisting ol a number of structural genes that are transcribed together.

al he term "transcription" and "transcribed" refer to the first step in gene expression i.e. the synthesis of an RNA copy from a sequence of DNA (a gene).

10 the term "promoter" refers to a site on L)NA to which RNA polymerase (an enzyme) will bind and initiate transcription of [)NA (a gene) to produce mRNA.

The tend ' Immuno-stimulant' refers to any class of compounds capable of eliciting or otherwise activating or helping to activate/elicit an immune response in a mammal. An immuno-stimulant may include cytokincs, chenokines, peptide 1.r containing and non-peptide-containing products produced in strains of the M. tuberculosis complex, and peptidecontaining and non-peptide-containing bacterial, protozoan or metazoan products but this list should not be seen as liniting. An inmunostimulant may include: polypeptides: polypeptidc-

including chemicals, or other non- protein based chemicals: but an immuno 2() stimulant should not be limited to this list.

The term "isolated" means substantially separated or purified away from contaminating sequences in the cell or organism in which the nucleic acid naturally occurs and includes nucleic acids purified by standard purification techniques as well as nucleic acids prepared by recombinant technology, including cart PCR technology, and those chemically synthesized. Preferably. the nucleic acid

molecule is derived prom genomic DNA or the mRNA of the lovicola ovis chewing douse.

The preparation of pharmaceutical compositions including pharmaceutical carriers are well known in the art, and are set out in textbooks stitch as Reming,ton's 5 Phannaceutica] Sciences, l 9th Edition, Mack Publishing Company, Easton, Pennsylvania. USA.

The compounds and compositions of the invention nary be administered by any suitable route. and the person skilled in the art will readily be able to determine the most suitable route and dose for the condition to be treated. Dosage will be at 10 the discretion of the attendant physician or veterinarian. and will depend on the nature and state ol' the condition to be treated, the age and general state ol' health of the subject to be treated, the route of administration. and any previous treatment which may have been administered.

The carrier or diluent, and other excipients, will depend on the route ol' 15 administration, and again the person skilled in the art will readily be able to determine the most suitable formulation for each particular case.

The homology between two nucleotide sequences can be assessed by a variety of methods. Prel'erably' the nucleotide sequences may be assessed by programs such as BLAS'I'N as available front NCFlI lttp:.!iN\\\.nchi. l':.nih.,ov or GAP and 20 BESTFIT programmes of CiCG (Wisconsin Package Version 10.2, Genetics Computer Group, Madison, Wisconsin) Thus, the homology of the nucleotile sequences may be assessed by comparing on BLASTN or other programmer the nucleotide sequence of interest against those of the relevant gene of the invention as can he determined from the complete genome

of M. taherc'losis (Cole et al, 1998) available at 1-ttl::!:N\:v..rncr. acail</ or at http.//www.pasteur.fr/mycdb/. The unique tagged suicide electors may be made Irom a variety of different vectors provided they arc compatible with species of the Al tzherculo.siAy complex. In irk particular, suitable vectors may include any vector that replicates in Pcherichia; cold but either cannot replicate in mycobacterial species or does not replicate in mycobactcrial species under specified conditions such as grovth temperature or the presence of a compound in the growth medium and which contains a variable region of nucleotides called a "lag" such as pUH1604.

10 BRIF.F DESCRIPTION OF DRAWINGS -

L'urtller aspects of the present invention will become apparent from the following description which is given by way of example only and with rchercnce to the

accompanying drawings in which: Figure 1 Shows a plasmid pYUB553.1 used for transposon mutagenesis.

15 'f'ransposon TnS367 is composed of TnpR, Ah (Lanamycin resistance), and'l'npA; Figure 2 Shows the sites of transposon mutagenesis in M bovis WAg9()0 to produce WAg533 and in Al hovi.s WAg201 to produce WAg537; Figure 3A Shoves a DNA fragment of' the al?pC.' locus of M bovis containing a 20 kanamycin resistance gene (neo); Fiurc 3B Shows a ONA fragment of Rv3.44 locus containing a kanamycin resistance gene (neo); figured Shows a Southern blot hybridization after EcoRI digestion and probing with neo ol' recombinants that were unable to grow in

minimal medium (lanes 1-5, 7-9) and one recombinant that failed the screen and grew in minimal medium (lane 6); Figure 5A Shows the site of illegitimate recombination and size of deletion in M ho,i.s ATCC35723 to produce WAgS26; 5 Figure 5B Shows the site of illegitimate recombination and size of deletion in 1. bovi.Y A l-CC'35723 to produce WAg>30; Figure 5(: Shows the site of illegitimate recombination and size of Relation in M hovis ATCC35723 to produce WAg569: }figure GA Shows y'UHA600 used as intermediate construct in signature tag 1 0 experiment; lain- ure 6B Shows!inearised nl JHA600 alter digestion with BsrBI; positions of selected restriction sites in the plasmid are given in brackets; figure 7 Shows plasmid pL1HA604 which is comprised ot pUHA601 used as intermediate construct in signature tag experiment together with TO inserted tag; Figure 8A Shows the 89 Up DNA fragment made for producing signature tags; Figure 8B Shows tile 81 Up ONA product produced by BssHII and lgi11 digestion of the 89 bp fragment in ig 8A and inserted into pUlIA601 to produce pUIIA6()4 plasmids containing signature 2() tags, }figure 9A Shows a site of illegitimate recombination and size ol deletion in M. hotic WAg2()() to produce WAúg539;

Figure 9B Shows a site of illegitimate recombination and size of deletion in M Avis WAg20 1 to produce WAg566; Figure 9C Shows a site of illegitimate recombination in M. hovis WAg201 to produce WAg570; 5 Figure 10 Shows a linearized diagram ol plasnid pUllA772 used for allelic exchange of pp1 In M Loves ATCC35723. pUHA9 = pBluescripl KS II+ with a Pcrcl-Bcll-Pacl cassette inserted at the BomHI site; hype is a hygromycin resistance gene; Figure I IA Shows a Southern blot hybridization ol PvuIIdigested M: bovine 10 recombinants with a probe that spans across the ppiA gene sequence into which the hype gene was inserted in the suicide vector pUHA772 (fig.!()) with the size of the wildtype fragment (2.6 kb) and the pUI 1A772 fragment (4.1 kb) in the reeombinants indicated by arrows. Lane I, ATC(235723 parent strain; lanes 2-6, 8, 10-16, illegitimate mutants and single homologous recombinants; lanes 7, 9, allelic exchange mutants; Figure I lB Shows a picture of an agarose gel electrophoresis separation of DNA products lrom M bouts A1CC35723 and selected M hovis reeombinants subjected to PCtR with L)MC3 12 and DMC3 13 20 (Table 6) which amplify a ONA fragment of 0.4 kb in wildtype strains and].9 kb in strains where prig has been interrupted with the 1.5 kb try,; gene. Lanel, DNA standards, lane 2,,/iT(:C35793 parent strain; lanes 3-6, four ppi/1 allelie exchange mutants (WA,759); lane 7, single homologous recombinant; lane 8, 25 illegitimate recombinant; and

Figure 12 Shows alignment of the taherculosis genomc at the elate locus with the DN/Y construct used for csc't5 knockout by homologous recombination. The deleted Iragmcut Included all of the esu'6 gene and small parts of the ill' and Rv3876 genes.

5 BEST MOI)ES FOR CARRYlN(J OUT THE INVENTION Experimental Non-limiling cxanples illustrating the invention will now be provided. It will be appreciated that the above description is provided by way of example only and

variations in the materials and technique used which arc known to those skilled in 10 the art arc contemplated.

Bacterial culture All bacteria were cultured at 37 C-38"C:. The,ll. havis strains used for this wore: and the carbon sources routinely used for their liquid and solid culture arc listed in Table 4. Antibiotics were added where appropriate at the concentrations given in 15 Table 5. Strains were cultured on a variety of di l'fercnt liquid and solid mycobacterial media. Standard liquid culture was in 3 - 100 ml media in static sealed glass containers. Two media were used: Tween-albumin broth (Kent and Kubica, 1985) which does not require an additional carbon source, and Middlebrook 7H9 (Difco) supplemented with albumin/dextrose complex (ADC; 20 Dil'co), 0.05% 'I'ween 80 and the appropriate carbon source (able 4). 'l'he standard solid medium used was Middlebrook 71-11 I (Difco) supplemented with ADC and the appropriate carbon source ('['able 4), sometimes with the addition of 2.5% bovine scrub. In some cases, M. bovis mutants were isolated on Middlebrook 7tI10 (Dif'ro) medium supE, lemcnted with 0.085% Na(:l, 0.2% 26 glucose, and 0.2% casamino acids. Solid culture was performed in standard plastic

Petri dishes. M hovis being prepared lor electroporation was cultured in roller bottles at 1 rpm in 100 m] Middlebrook 71-19 (Difco) supplemented with albumin/dextrose complex (AL)(; Disco) 0.05% i wean 8(), 0. 0 sodium pyruvatc and 0.5% glycerol. For selection of mutants with impaired growth in 5 minimal medium, 5 p'1 of a 5 - 10 week old culture in supplemented Middlebrook 7H9 medium was inoculated into supplemented Middlebrook 7H9 medium and also into minimal Proskaer and Beck liquid medium (L)arz.ms' 1958j to which 0.05% Tween 80 and 0.2% sodium pyruvate had been added.

Escherichi cold strains used were XLI-F'luc-MR, JM109 and t)H10B. They Nvere 10 grown in Luria-Bertani liquid medium or solid agar with the appropriate antibiotic concentrations (Table 5).

Production and screening of M. hovis mutants The fallowing four methods were successfully used to produce mutants of M. hovis' some of which were subsequently found after screening, virulence testing 15 and vaccine testing to have vaccine efficacy against tuberculosis in experimental animals. I hose mutant strains of M: havis that showed vaccine efficacy in guinea pigs that was similar or better than BCG under the standard vaccine challenge conditions used are given in Table 1. In the first three methods, random mutagenesis is used to make a large number of mutants that are subsequently 20 screened to select a tew candidates while in the fourth metllod, allelic exchange, a specific gene to be inactivated is chosen and then deleted. The advantages of randon, mutation approaches are that they name iw assumptions about what genes arc important for virulence and only reveal those genes that are amenable to inactivation. Its major disadvantacs are that the mutation methods may not be 25 s:uffciently random to inactivate all virulence genes and because many thousands and even millions of mutants can be made, there is great difficulty in selecting out

all of the mutants that have become attenuated. Some of the mutants claimed herein were produced by illegitimate recombination. This can produce mutants t comparable to those produced by transposon mutagenesis in which a single gene is deleted but in many cases, illegitimate recombination causes deletion of several genes. This may have advantages for producing a dive vaccine that cannot revert.

In the experiments reported here, highly-attenuated ilk. hove rccombinants with good vaccine efficacy were obtained by screening transposon mulants lor morphological changes. and by screening illegitimate recombinants for survival in macrophages and for impaired ability of' 5-lO week-old cultures to grow in 10 minimal medium. While all screening methods were not used on either transposon mutants or illegitimate recombinants this should not be taken to imply that they cannot be so used. One skilled in the art would expect that any of the screening methods used here as well as other screening netl1ods not used such as standard auxolrophic selection (Hondalus et al., 2000) could potentially be successfully 15 applied to mutants of the M hberculo.si.y complex produced by different methods in order to acquire highly attenuated mutants with some degree ol' vaccine el'l'icacy. One slcillcd in the art would also expect that mutants of the M. ubercJlJsis complex with identical or related genes inactivated to those strains] claimed in this patent would in many cases also have vaccine el'lcacy and that :20 they could be developed by random mutagenesis or by the use of any of a large number of dit't'ercut methods to identity one or more genes associated with I virulence, immune modification or essential metabolic pathways (Collins and Gicquel, 2000; Collins. 200O, 2001; DesJardin and Schlesinger, 2000) followed by specil'ic gene inactivation, for example by allelic exchange.

25 1. Transposon mutagenesis end morphological screening Jransposon mutagenesis Strain M bovi.s WAg200 was mutated by transposon mutagenesis using the

transposon Tn5367 in plasmid pYUB553.1 (Fig. 1) obtained from Professor W. R. Jacobs, Albert Einstein College of Medicine, New York. The technique followed was similar to that described by McAdam et ui. (1995) except that the suicide plasmid used in this work was electroporated into M Morris WAg2OO using a high 5 efficiency electroporation technique as described by Wards and Collins (1996).

Morphological.screening Men thousand individual kanamycin resistant colonies were plated at approximately 1 o2 CI7U / plate on both 7T I I O and 7Hl I media and after 4-6 weeks culture were assessed by eye l'or colony morphology. Colonies that appeared 10 different from the norm were recultured on two separate occasions and compared again to colonies with normal morphology. Three Al: hovis mutants appeared to show stable, sn1all dil'ferences *om normal morphology and these were tested for virulence in guinea pigs. One of these mutants, WAg533, was found to be avirulent and subsequently was shown to have vaccine efficacy at least 15 comparable to that of BCG (Table 7A).

Determinulion oJ.site of mutation To determine the site of insertion of the tran'sposon in the M. havis chromosome, chromosomal DNA was digested with L'coRI, a restriction enzyme that does not have a site within Tn5367, ligated into the k:coRI site of pHluescript KSll+.

20 electroporated into E. cold Xl,l-Blue MR and plated on Luria-Bertani medium containing 25 g/ml kanamycin and 50 g/ml ampicillin. The junction regions of the fragment and chromosome were sequenced, using primers DMC153 and DMC l S4 (Table 6) directed outwards from rln5367 into the mycobacterial sequence. DNA sequences were analysed using the programmes of the Genetics 25 Computer Group and compared to GenBank (www.nchi.nlm. nih.gov) and Sanger (www.sanger.ac.uk) latabases.

?'milar mutant -

Subsequently, a transposon mutant (WAg537) from a different parent, M hovis WAg201. was selected by screening for its inability to grow in microaerophilic conditions and to be scasitive to levels oi'cycloscrine that do not inhibit the parent strain. WAg537 was found to be avirulent in guinea pigs and on analysis was found to have a transposon insertion in Rv()097 (Fig. 2), the same gene that is inactivated in WAg533 (Fig. 2). The chromosomal site of' inactivation in Rv()()97 is different in the two strains; Mtcy95 position 146] 0 for WAg537 and 147'30 for WAg533 (Tables I and 2). WAg537 was not originally screened lor different 10 colonial morphology and WAg533 was not originally screened for its ability to grow in microuerophilic conditions or its cycloserinc sensitivity. Ilowever, when \NAg533 and WAg537 were compared using all three screening methods they were found to have. identical phenotypes. This clearly indicates that in at least some cases, different screening methods will identity dit'fcrent mutants of the 11.

15 tabcrculosi complex in which the same gene is inactivated.

Vaccine testing I'he standard vaccine challenge method used for assessing the protective effect induced by WAgS33 and the other mutants detailed below was essentially as described in de Lisle el al., (1999) except that instead of the animals being 20 challenged by intratracheal inoculation they were challenged by aerosol inoculation. Briefly, 4-6 guinea pigs were inoculated subcutaneously in the flank with approximately 1Os (:FU of the M hovels mutant to be tested. At the same time, a group of 4-6 guinea pigs was inoculated hi the same manner with M. /'ovis BCG and a control group of 4-6 guinea pigs was not inoculated. The number of )r, CFU inoculated was approximated by appropriate dilution following estimation of turbid cultures. The exact number of CFIJ used was determined retrospectively by

plating 10-fold dilutions on supplemented Middlebrook 7H11 agar. Animals were fed food and water ad lihtum and housed in a controlled environment biocontainment unit. Approximately X weeks alter inoculation. the guinea pips were challenged with an aerosol containing a single cell suspension of M. hov.s 5 WAg201 with a known quantity of organisms. Single cell suspensions ol' the isolate were prepared using a mollification of a method described by Grover et al. (1967) and stored at -7()".'. For preparing these suspensions the bacterial cells were dispersed by sonication for 30 sec and filtered through an 8 lam membrane filter. Guinea pigs were infected via tile respiratory route by using an aerosol 10 chamber which produces droplet nuclei of the size appropriate for entry into alveolar spaces (McMurray et al., 1 9X5 Wiegeshaus et al. 1970) . The concentration of viable M bovis in the nebuliser fluid was empirically adjusted to result in the inhalation and retention of 2-10 viable organisms per guinea pig (huddle and de 1 isle unpublished). 'I'he final solution that was aerosolised 15 contained approximately 0.048 x 10t' CFU/ml. This challenge dose had previously been estimated from the number of primary tubereles observed grossly in the lungs of non- vaccinated guinea pigs at 4 vveeks T,ost-infection. A similar procedure had been reported previously to result in reproducible uniform infection of the lungs of guinea pigs (Wiegeshaus et aI. 1970; Smith et al.. 1970).

2() The aerosol infection and subsequent maintenance and manipulation of infected guinea pigs were performed under strict isolation conditions in a biohazard facility. Approximately 5 weeks alter challenge the animals were euthanased and autopsied and body weight and gross pathology was recorded. Samples of spleen and lungs were subjected to nycobacterial culture and enumeration. Delayed type 25 hypersensitivity to tuberculin was measured immediately before vaccination and prior to the animals being sacrificed. Bovine purified protein derivative (4 units; MAE Central Animal Health l aboratory Nevv Zealand) was injected intradermally and the diameter of the erythema was measured 24 h later. For

WAg533 and all the other Ill hovis recombinants described in this patent' no delayed type hypersensitivity reactions were observed when the animals were tested prior to being inoculated with mycobacteria. Hi animals hatl pusiiive responses immediately prior to sacrifiec. The vaccine efficacy of the mutants and of BCG was determined by comparison of the gross lesions and CFU in the test animals to those in the control animals.

When WAg533 was tested for its ability to protect guinea pigs against tuberculosis using the above method it provided protection that was comparable to that provided by BCG (Table 7A).

10 2. Illegitimate recombination and screening in minimal medium and macro p h ages Previo1',<..'Qr,k illegitimate recombination in h I. hovr. s B(:C and M tuberculosis u as first described by Kalpana et al. ( 1991) and occurs when there is recombination 1r' between an introduced DNA fragment and a region of the chromosome with which it has little homology. v'c adapted this finding to produce mutants of M hovis (Wilson et 1., 1997). In that study. we used a linear DNA fragment of the ahpC locus of M. hovis with a kanamycin resistance gene (neo) inserted into it (Fig. 3A) . In contrast to transposon mutagenesis which Ilroduces mutants in which 20 insertion at a single point without any deletion occurs, a variable degree ol deletion usually occurs with illegitimate recombination at the site of insertion in the chromosomal DNA and sometimes also at either end of the inserting DNA. In our earlier work, the deletion in the chromosomal DNA varied from 2 bp to 64.G kb. Four illegitimate mutants that were selected from their inability to grow in 2'ry minimal medium were shown to be attenuated in guinea pigs. When challenged with virulent M. h'vi.s, two of these mutants induced a IeveJ ot protection that was

similar to that induced by BCG (dc Lisle en al., 1999) in one of these mutants, a putative undecaprenol kinase gene was interrupted by illegitimate insertion of a ONA fragment that accompanied a 2 hp chromosomal deletion. In the other mutant, DNA fragment insertion was accompanied by a large DNA deletion of] 5 kb containing 12 genes. In none of the lout mutants were any of the interrupted genes identified as being involved or having close homology to genes encoding enzymes for common metabolic pathways that are often associated with auxotrophy such as amino acid, purine, pyrimidine or co-factor synthesis.

Nevertheless, at the time we reported these experiments, the mutants selected on 10 minimal medium were described as auxotrophs.

Stal'rnclry growth phclse screening Subsequently, we discovered that these mutants were not ausotrophs in the strict sense as 2-week-old actively growing cultures of the mutants in complete medium were ab]c to grow when inoculated into minimal medium. Rc-examination of our 15 procedures revealed that the mutants in the study had been subcultured into minimal medium 1rom 5-10 week old cultures in complete medium, at a stage when they would have been in stationary growth. Since the ability of strains of the M laherculosi.s' complex to survive in the host in a stationary or lowered metabolic state is regarded as a crucial facet of tuberculosis pathogenesis, we 20 reasoned that it might be desirable to have a vaccine strain without this ability, with the added consideration that if that was the only ability it had lost, it might provide a better stimulation of host immune systems than a true auxotrophic mutant which required a specific metabolite to achieve adequate growth.

Regardless of this explanation, since the use of 5-10 week old cultures had 25 produced two mutants with vaccine eitectiveness at least equal to that ol BCG, we continued using similar approaches to produce avirulcnt mutants. In one new study (Collins et a., 2002), we used either an 'hpC' fragment interrupted with neo

as in the earlier work or a second fragment in which Rv3844, a gene of unknown function, was interrupted with neo (Fig. 3B). A Southern blot hybridization including DNA from each of the 8 mutants obtained from this experiment is given in Fig. 4. Five of tl1esc mutants (WAg526, WAgS7, WAg599, WAg530 and 5 WAgS31) were found to be highly attenuated when tested for virulence in guinea pegs. Screeningf'r severe limitation of growth in bovine alveolar inacrophcges Alveolar macrophages were obtained by ravage from freshly excised lungs of tuberculosis-free cattle and dispensed into flat-bottomed 96-vvell microtitre plates 10 (Nalge Nunc International) as described previously (Aldwell ei at., 1996). After washing to remove non-adherent cells, they were cultured at 37 C under a 5% CO2-95% air atmosphere in RPMI 1640 supplemented with 4 mM 1 glutamine, 1 mM non-essential amino acids 1 mM sodium pyruvate. 2.9 mM sodium bicarbonate, 50 {J/nl penicillin G and 10% bovine serum. Sixty illegitimate 15 recombinants of M hovis ATCC35723 were cultured in TAB medium. 'the bacteria were serially diluted and 5 pi and 10 pr1 aliquots of' each culture containing 1-5 x 107 Ct;U/ml were placed in duplicate into microtitre wells containing macrophages. Since each well contained approximately I x l (I> macrophages per well' the multiplicity of infection ranged from 0.5:1 - 5:1. An 20 aliquot of each bacterial culture was also subcultured in a mierotitre well containing'l'AB. Alter 5 days, the cells in both the TAB and maerophage cultures were labelled overnight with [3H]uracil, harvested as described previously (Aldwell el al., 1996) and the level of radioactivity was counted in a 'I'rilux MicroBeta 1450 scintillation counter. The level of radioactivity was used as an 2; approximate measure of bacterial cell numbers based on previous comparisons (Aldwell et al., 1996). The ratio ol'3H counts in macrophages compared to the same strain in TAB were calculated and three strains with the lowest ratios were

tested f or virulence in guinea pigs. One of these strains. WAg569, was found to he highly attenuated in guinea pigs.

Vlccination results The highly attenuated strains developed in these experiments were tested for their 5 ability to protect guinea pigs against a standard vaccine challenge as described earlier and three of these (WAg526, W530 and W,\5('')) were found to provide comparable protection to that provided by BCG (Table 7A, 7B).

Identification oJ.sites of mutation The sites ol mutation and the size of the deletions that had occurred during, 10 illegitimate recombination were determined for WAg526, WAg530 and WAg569 using previously described methods (Wilson et al., 1997) and are shown in Fig. 5A' Fig. 5B and Fig. 5C respectively. Brielly, chromosomal DNA was digested with a restriction enzyme that did not cut within the inserted fragment (Xiio! for the Rv38J? fragment and F,coR1 or BamT]I for the ukp(' (ragmenL), ligated into the 15 appropriate site of pBlucscript KSII+, elcctroporated into E cold Xl,lBlue MR and plated on Luria-Bertani medium containing 25,ug/ml kanamycin and 50 g/ml ampicillin. The junction regions of the fragment and chromosome were sequenced, using appropriate primers directed outwards from the vector sequence into the mycobacterial sequence. NINA sequences were analysed using Bl,AS'J' 20 (httl:/iw\w.chi.lnl.nil-.o\) and the protrrammcs of the Genetics Computer Croup and compared to GenBank (\\ww. ncbi.nlm.nih.gov) and Sander (www.sanger.ac.uk) databases.

3. Signature Tag Mutagenesis based on illegitimate recombination Apprcclch used in Lois study'

Signature Tag Mutagenesis (STM) combines the production of mutant libraries with the direct identification of individual mutants that have become attenuated. Unlike the other random mutant approaches that arc used in this patent to

produce attenuated strains of the.U laherGz/lo.vi.s complex, in which mutants are produced? 5 screened in various ways and then tested individually for loss of virulence; S'I'M combines these procedures in a co-ordinated way. IN Srl'M, pools of mutants are produced In wlicl each member of the pool is tagged with a unique DNA sequence. Susceptible hosts are intccted with as pool of mutants called the 'input" pool. Subsequently, an '-output.' pool of disease-causing organisms is recovered 1() from the infected host and the absence of' a member(s) of the input pool in the output pool indicates which particular mutant(s) was attenuated. 'l'he method was first applied to Salmonellu typhmurium infection in mice (Hensel et al.. 1995) and since then has been applied to a considerable number of different pathogens (Meccas, 200')). 'I'wo groups used the technique to identify transposon mutants of 15 M. {uberculo.Yi.s that are unable to replicate in the lungs of mice (Cox emus., 199); Camacho e' al., 1999). Onc group (Camacho et al., 1999) identified 14 dii'fercnt genetic loci whose interruption caused attenuation and both groups found that many of the affected loci arc inN:olvcd in lipid metabolism. In particular, a 50 kb region of the chromosome that contains 13 genes involved in biosynthesis or 20 transport of phthiocerol and phenolphthiocerol and in mycocerosic acid synthesis was identified in several mutants by both groups.

A substantially difl'erent approach was used in this study. First, we made a tagged suicide vector (pUH1604, Fig. 7) as described below and produced mutants of M bovis by illegitimate recombination instead of by transposon mutagenesis. Second, 25 we used a guinea pig model in which pools of mutants were inoculated sub cutaneously anal mutants that could not be isolated shsequently from pooled colonies Irom spleens of three animals were identified and reinoculated singly into

three guinea pigs each to idcntil'y those that were most attenuated. This contrasts with both the reported studies with M luheiAcuZo.si.y mutants where the bacterial pools were inoculated intravenously into two mice and the mutants were tested for their ability to survive in mouse lungs for three weeks.

5 Preparation of tagged suicide plasmid pIJUA604 1. A 203) bp plasmid (pljllA600, Fig. 6A) was derived from the 5()11 bp integrating mycobacterial shuttle plasmid pYUB178 (Pascopella et al., 1994) by digestion with BumHI and Bell which each CUt once in this vector followed by selt:ligation of the smaller fragment which contains a 10 kanamycin resistance gene (ape) and an origin of replication in cold (orir). 2. Plasmid pUHA600 was digested with R.srRT which cuts the plasmid once between the end ol'oriE and the start of the ciph gene to produce a blunt end linear full-length plasmid pUHA600 as shown in Fig. 6El.

15 3. A short blunt-ended double-stranded piece of DNA was made by allowing DMC125 and DMC126 ('I'able 6) to anneal by heating both primers to 90 C and allowing them to cool slowly over 30 min to 40 C to form a linker that contained BssHII (itarlicisel) and BAIT (lower case) sites as shown below: 20 TTG GC.'G C.'GC'CAA GAa get ctT C AAC (.tG('GCC; GT'I' CTt cta gaA G 4. Linearised and phosphatased pUI IA600 (500 ng) was ligated with I pig ol' this double-strand DNA and the ligation mixture was eleetroporatcd i!'t".4, colt and plated onto solid medium containing lanamycin. i'1asmid DNA

extraction was performed on selected colonies and a plasmid of the expected size that gave a linear fragment of approximately 2.1 kb on digestion with either 13/11 fir Bs.sl III was designated pUHA6()1.

5. Plasmid pUIlA601 (5 fig) was digested with an excess ol'BssIIII and loll irk and the large linear fragment was separated from uncut vector and the tiny (7-11 bp) linking piece between the two restriction sites by electrophoresis and gel extraction ((,eneclean, HiolOI).

6. Microgram quantities of an 89 bp T)NA fragment (I ig. SA) were produced by polymerase chain amplification (l'CR) of DMC120 with DMC121 and 10 DMC 123 ('['able 6). The size of'the product was checked on a 4 /0 Nusieve + 1% normal agarose gel. This DNA fragment contained a variable 40 bp sequence that was used for uniquely tagging individual strains of Ad' hovix.

7. The 89 bp DNA product (I;ig. 8A) was digested with Bs.sHII and ByIII to produce a sticky-ended 81 bp DNA product (Fig. 8B) that was then 1 5 deE,hosphorylated.

8. 'I'he sticky-ended 81 bp product was ligated lo the large pUHA6() 1 fragment prepared in step 5 by digestion with BssHI1 + Bglll and talc mixture was electroporatcd into E: ok and plated on solid medium containing kanamycin to produce a cloned family of pUMA604 plasmids 20 each with a different 40 be tag (Fig. 7).

First experimental approach In the first experimental approach, plasmid DNA was extracted from a liquid culture made from a combined pool of IO, OOO CFU of recombinant E cold containing plJHA604 plasmids. The plasmid DNA was digested with Quell or 25.SpeI and clectroporated into M. hovi.s WAg200 which was then plated onto solid

medium containing kanamycin. Two thousand individual colonies were recovered, cultured in liquid medium and stored at room temperature until needed.

Restriction digestion ot' LENA from 20 individual clones, followed by electrophorcsis, Sout}1crn blotting and hybridization with a probe of pUHA6()0, Irk indicated that insertion had occurred into diff'ercnt parts of the chromosome. DNA from 20-50 clones was digested individually with Hindlll, subjected to electrophoresis ( 1.5% agarose, 4.5 V/cm, t)0 min) and transi'errcd to nylon membrane (Zeta Probe, 13iorad) by Southern Blotting. A probe of LENA pooled from all the M. hovi,.s recombinanLs on the membrane was prepared by PCR using 10 the primers DMC122 and DMC124 in a Geneamp 9600 PCR system (Perkin Elmer Cetus) under the following temperature conditions: 94"C for 3 min initial denaturation followed by 30 cycles of 62"C for 30 see, 72"C For 10 scc, and 94"C for 30 sec The standard PCR reaction mixture consisted of 50 mM KCI, 10 mM Tris-liCI pH 8.3, 1.5 mM MgCl2, 0.2 mM each of dA'l'E', dCTP, dGTI' and dTTP, 16 0.5 'm each of oligonucleotide primers and 2.5 units of AmpliTaq L)NA polymerasc (Perkin-Elmer Cetus). 'I'he 1)NA product was separated by electroplloresis on a 1.6% agarosc gel (Seaplaquc (TUG, I;MC 13io- l'roducts) and extracted using a QlAquick gel extraction kit (Qiagen). If a strong 81 bp product was not obtained, the PCR was repeated using a different concentration of MgCI2.

20 The probe was labellcd with 32p by nick translation. input pools of sizes ranging from 20-50 recombinants were constructed by separately culturing recormbinanLs, estimating their concentration based on their turbidity and adding equal numbers of each recombinant to make the pool. Each pool was inoculated info three guinea pigs. After 7 - 9 weeks, the animals were euthanased and in all cases had multiple 25 grossly-visible tubcrculous lesions in their spleens l'rom which M pelvis was isolated. Between 40, 000 and 200, OOO colonies from the spleens of animals infected with a pool of recombinants were combined together and extracted for DNA using cetyl-trimetllyl ammonium bromide (van Soolingen el al., 1991).

labelled probe was made from this output-pool DNA in the same way as the probe for the input-pool DNA and hybridiscd to a Southern blot of DNA from each member of the input peril in the same way as described earlier. Recombinants whose DNA hybridised much less strongly to the output-pool probe than to the 5 input-pool probe were tested individually for their attenuation in three guinea pigs.

Considerable technical difficulty was encountered with this approach because DNA from many rccombinants hytridised poorly to a pooled DNA probe and the presence of mycobacterial DNA in both the probe and the Southern blot gave significant background hybridisation. Nevertheless, 145 recombinants in various

10 size pools were inoculated into guinea pigs and two attenuated recombinants were detected (WAg539 which gave no lesions in guinea pig spleens and a second strain, WAg54O, which gave a small number of lesions in the spleen of one guinea pig). <Seconc! experinZentul approach 15 In the second experimental approach. a total of 132 plasmids in three batches of 39, 4() and 53 plasmids each containing a different tag were prepared, subjected individually without restriction enzyme digestion to agarose gel electrophoresis (1% agarose7 60 min. 9 V/cm) and transferred to nylon membranes (zeta Probe, Biorad) by Southern blotting. Forty-five of the plasmids gave an acceptable 20 positive signal when hybridiscd to a probe made from their respective combined pools and did not cross-react when half the pool was labclled and hybridised against the other half. The probe was prepared in the same way as in the first experimental approach except that the template DNA used was plasmid DNA instead of DNA from M hovis recombinants. To make M hovis rccombinants 26 using these plasmids, each plasmid was digested with Hoell, and electroporated separately into M. bovis WAg20 I which was then plated on solid medium containing kanamycin. Twenty-scven individual colonies ol recombinant M. '5ovi.s

from each electroporation were cultured in liquid Realism and stored until needed at -70"C. Pools of 42 - 45 M hovis recombinants containing approximately 106 CL'[J were inoculated Into three guinea pigs each. A pool was constructed by separately culturing 42-45 rccombinants each with a dit'ferent tag, estimating their concentration based on their turbidity and adding equal numbers of each recombinant to make the pool. After 7 weeks, the animals were euthanased and in all cases had multiple grosslyvisible tuberculous lesions in their spleens from which M. hari.s was isolated. Between 40, 0()0 and 200, 000 colonies from the spleens of animals int'ected with a pool of recombinants were combined 10 together and extracted for DNA using cetyl-trimethyl ammonium bromide (van Soolingen et al.,]991). Rebelled probes were made from both the inputpool DNA and the output-pool 1)NA in the same way as in the first experimental approach and used to probe a Southern blot of ONA samples from the original 45 plasmids. This gave less background than in the first experimental approach where

15 Southern blots of DNA from M bovis' recombinants were used. Recombinants whose DNA hybridised much less strongly to the output-pool probe than lo the input-pool probe were tested individually for their attenuation in three guinea pigs.

A total ot'l l lO M horrid recombinants were tested. Those that were found to have reduced virulence arc given in flyable 8. Two of these strains arc named in Claim 3:

20 viz. WAg566 and WAgS70.

Identification of Sites of muialion The sites of mutation and the size of the deletions that had occurred during illegitimate recombination were determined for WAy,539, WAg566, and WAg570 using an approach similar to that described above for other illegitimate 25 recombinants (Wilson et al. ,]997) and are shown in Fig. 9A, Fig. 9B, and Fig. 9C respectively and summarised in l'ablcs 1 and 2. In some cases, there were tandem copies of the inserting suicide plasmid at the mutation site which necessitated

subcloning of the recovered constructs containing the kanamycin resistance gene and chromoson1al junction fragments in order to get unique sites for the sequencing primers.

Vaccination r esult.s 5 One (WAg539) Ott the three highly attenuated strains (WAg539, WAg566, WAg57()) developed hi these STM experiments was tested for its ability to protect guinea pigs against a standard vaccine challenge and was l'ound to provide comparable protection to that provided by BC(T (Table 7A).

4. Allclic exchange 10 Allelic exchange in they M tuberculo.si.s complex Allelic exchange has two important roles in the development of attenuated strains of the M. t''herculosi.s complex with vaccine potential. First, there are a large range of dil'ferent methods that can be applied to strains al' the M luberculo.si.s complex in order to identify genes that may be involved in the pathogenesis of To tuberculosis or whose inactivation might be expected to affect the virulence of these nycobacteria (Collins and Gicquel, 2000; Collins, 2000, 2001; DcsJardin and Schlesinger, 2000). These genes can then be inactivated by allelic exchange techniques and the virulence and vaccine properties ol'the mutants formed can be determined. This approach was used to produce WAg759 in this present study.

20 Second, allelic exchange can be used in a wide variety of ways to modify strains of the Off. tuberculo.s.s complex that have already been mutated. For example, it can be used to produce inactivation of a second gene in a mutant and this was done for WAg520.4 and WAgS30.1 in this present study, or it can be used to eliminate a previously inserted selection gene or to modify a mycobacterial gene 25 in a way that alters the gene's properties but does not inactivate the germ.

Initial attempts to perform allelic exchange in the M tuberculosis complex were unsuccessful and resulted in the production ol'illegitimate recornbinants (Kalpana car al., 199 l; Aldovini e' us., 1993). More recently, different groups have developed a variety of methods to perform allelic exchange (I'elicic et al., 1997; 5 Hinds et al., 1999; Wards et al., 2000; Raman et al., 2001) and a relatively large number of genes have now been inactivated.

Method oJallelic exchange used in this.suly The method of allelic exchange used in this study has beets used to inactivate a number of genes in strains of M hovis including esat6 (Wards e' al., 2000), ahp(' 10 (Collins, 2()(11,) and Rv2136 (Collins and (]icquel, 2000). ldriefly, a 2-3 kb DNA sequence incorporating the gene of interest is cloned into an E. cali vector and an antibiotic resistance gene is inserted into the coding sequence of the gene so that nonnal expression will be prevented. The vector is denatured and electroporated as a suicide vector into a strain of the M. taherculosi complex and antibiotic 15 resistant clones are recovered and analysed to detect an allelic exchange mutant.

Production of WAg759 In the case of the ppil gene, a 2219 bp DNA fragment incorporating the coding sequence of ppi.4 at position 482-1 l9l of the fragment was amplified by PCR from chromosomal DNA of M bovi.s ATCC35723 by using the primers DMC330 20 and DMC33 1 ('I'able 6) which contain a restriction site near their 5' ends for I'acI, an enzyme that has no sites in the chromosome of A4; taherczlosis. The fragment WAS digested with PacI and cloned into pUHA9 (a derivative of the cloning vector pBluescript II KS (Stratagene) in which a DNA cassette containing the restriction enzyme sites Pac l-Bcll-l'ccI replaces the BamHl site) to form pUI-IA774.

25 pUlIA774 was digested with Esprit which cuts once in the plasmid, at 658 by in the 2219 bp M: hovis DNA fragment. The S,fiI digest was treated with '1'4 DNA

polymerase to blunt end the plasn1;d by removing the three-nucleotide 3' overhanging ends due to the StfiI dicst, and was dephosphorylated. A 1.5 kb Noel - istl DIVA fragment containing a hygromycin resistance gene was blunt-ended using Klcnow fragment, ligated to blunt-ended pULIA774 and transformed into E. ire cold to produce pUIIA772 (shown in linear form in Fig. 10). This plasmid was denatured in 0.2 M NaOT] for 5 min. neutralized with half its volume of I M gallium acetate pli 5.0, and preciptatcd and washed with ethanol. The denatured plasmid was elcctroporated into M bovis ATCC35723 using a high efficiency elcctroporation technique (Wards and Collins, 1L)9G) and cells were plated onto 10 medium containing hygromycin. DNA was isolated from liquid subcultures ol 72 hygromycin-resistant colonies (van Soolingen el al., 1991), 1ig,ested with /'v'TI and characterized by Southern blot hybridization with a probe that spans across the mycobacterial DNA site containing the inserted antibiotic resistance acne.

Four recombinants gave Southern blot hybridisation patterns indicating that allelic lard exchange had occurred. An autoradiogram of one Southern blot hybridisation revealing two of the allclic exchange mutants is shown in L'ig. 11A. All four allelic exchange mutants were confirmed by PCR using DMC312 and DMC313 (Table 6) which amplify a DNA fragment of 0.4 kb in wildtypc strains and 1.9 kb in strains where ppiA has been interrupted with the I.5 kb hype gene (fig. 1 1 B). In 20 both the Southern blot hybridizations and the PCR product gel the presence ol two fragments indicates single homologous rccon1bination or illegitimate recombination has occurred, the presence of a single fragment with a size 1.5 kb more than the wildtypc iragn1ent h1dicatcs that allclic exchange has occurred. One of these allelic mechanic mutants WAg759 wax shown to be avirulent in guinea 25 pigs.

Production of WAgS20.4 and WAgS30 1 The production of M: bovi.s WAg520 has been described previously (Wilson et al., 1997) and the production of M hovis WAg530 is described earlier in this patent.

To delete the e.sut6 gene locus in WAg520 and WAg530 a similar approach was 5 used to that described in Wards et al. (2000) except that the suicide plasmid contained a hygromycin resistance gene inserted into the deleted esat6 locus (consisting of all the esut6 gene and parts of the two adjacent genes, Rv3Y74 and lv3876) instead of'a kanamycin resistance gcnc. An outline of the construct used tor allelic exchange of the esut6 locus is given in Fig;. 12. Both WAg520.4 and 10 WAg530.1 were shown to be highly attenuated in guinea pigs and, using the vaccine testing method described earlier, were t'ound to provide comparable protection to that provided by BC(, and by their respective parent strains ('['able 7B). Advantages of deleting e.sat6 locus in already attenuated strains of the M 15 taherculosis complex The advantages of eliminating the esat6 locus as has been done for WAg520.4 and WAg530.1 are threefold. First, the region deleted prevents the production by the strains of two proteins, ESAT6 and CI;P10 (also called MTSAIO, the product of Ihp also known as Rv3874), which together stimulate strong cell-mediated 20 immune responses in infected hosts (Renshaw et al., 2002). The removal of'these proteins from a live vaccine enables the use of testing regimes for tuberculosis infection that combine tests of cell immune responses to ESAT6 and/or CFP10 as well as other tests for tuberculosis infection (Colangeli et al., 2000; van Pinxteren el al., 2000; Vordermeier et al, 2001). The results of these tests can distinguish 25 hosts infected with wildtype strains of the M. tuberculosi.s. complex from hosts vaccinated with attenuated strains of the M tuberculosis complex lacking

production of ESAT6 and CT4P10. Second, removal of the esat6 locus greatly reduces virulence (shards et al., 2000) and therefore provides an added safety factor against reversion to virulence of any live vaccine strain that like W>\g520.4 and WAg530.1 also contains another attenuating mutation. Third, removal ol the o i'SNt6 locus may directly improve the vaccine properties of an already attenuated strain ol the M. uberculosis complex if the absence of ESAT6 and/or CI7PIO from the vaccine strain allows a more protective or long-lived immune response to be established in the host.

Future use of live vaccines 10 It is anticipated that the vaccine strains described here or vaccine strains with improved properties based on inactivation of genes described here will have wide utility tor protecting against tuberculosis and other mycobacterial diseases in mammalian hosts. Already some of these vaccine strains have been tested for their ability to protect Australian brush-tailed possums (Trichosurus vulpecula) and in lfi some cases have demonstrated protection that is at least as good as B(( (Table 9). This demonstrates that the approach being used with guinea pigs is likely to produce attenuated strains o l the M. tuberculosis complex with vaccine effectiveness against tuberculosis in many dillerent mammalian hosts including humans. Table 1. Attenuated mutants of M. bovis shown to have vaccine efficacy that is not significantly different from BCG Mutant Parent strain Method of Primary Genes Position of insertion Size of production screen inactivated and where deletion (*) applicable deletion (bp) relative to Sanger sequence of M. tuberculosis H37Rv _ W Ao75', A TCC35723 Allciic Southern ppzA 9390 - 9392 of 3 exchange e blotting Mtcy I OH4 W Ag530 A lCC35723 111egitirnate Growth ginA2 5330 5359 of 30 recombination in Mtcy427 _

| mlnlmal medium WAg533 WAg200 I ransposon Morphol Rv0097 14790 o f Mtcy251 mutagenesis | ogy _ __ WAgS39 WAg200 I ilegtimate Signature Rv044' 7900 - 18040 of 10,141 recombination Tag (S'gK) - Mtv037 Rv0453 (PPE) WAg526 ATCC35723 111egitimate (irowth Rv0175- 16862 - 28088 of 11,227 recombitiatton in Rv0186 Mtc128 minimal (hgls) _ medium _ W I;6569 AT CC35723 illegtunate Poor RvOI l l - 53 'S - 8bb2 of 3,328 recombulation growth in RvO115 Mtv031 amaecSroph .. WAg520 A l CC35723 Allelic Southern Rv2136 3877 3878 of 2 4 exchange blotttng and Mtcy270 Rv38'74 - 5 75 Rv3876 9754 - 10329 of Mtv027 WAgS3() ATCC35723 Allelic Southern gin/12 and 5330 - 5359 of 30 I exchange blotting Rv3874 - Mtcy427 575 Rv36'76 t)754 - 10329 of Mtv027 _. * Cicnc designaLioni,are those ttsed in annotatin the complete 'enome sequence of M taherculo.ss H37Rv (Cole e cl., 1998) and can bc found at lltt:/w\\\ sar.ic ul and in MycDE! at http://www.pasteur. fr/mycdb/ Tablc 2. Attenuated mutants of M. bovis whose vaccine efficacy has not yet been determined , Mutant I Parent I Method of I Primary I Genes | Position ol | Dcletoi1 | stram | prodtction | screen | inactvated | Insertion and, | size (bp) | | where applcable, | | deletton relative | | | to Sanger | sequence of M taherculosz,s l l 1 1 1 1 1 H37Rv 1 1 1 1 1 1 1 1 -

WAg537 | WAg201 | Transposon | Cycloserine. | Kv()097 | 14610 of | 0 mulagenesis I Reduced I | Mtcy251 I | | oxygen WAg566 1 WAg201 1 Illegitimate 1 Signature 1 Rv3757c 1 110334 - 112676 1 9,343 recombnation I Tag | (prr>W) - | of Mtv()25 1 1 1 1 Rv3 760 1 1 1 I I I,,

WAgS70 | WAg201 1 111egttimate | Signature I Rv3545c | 25861 of | 0 10 I I recombination I Tag I | Mtcy3C7 1..

Table 3. Properties associated with virulence of the M. tuberculosis complex | | Infectious, c'-tpable of being spread llom one individual mammal to another 2 I Capable of entering mammalian host cells | 3 | Capable of surviving or escaping phagocyte ccilular defences

4 Capable of multiplying in host cells 5 Capable of spreadtl1g fron1 an ufected ceil to an unu1fected cell 6 Capable of causing or enabilng cell InJury that results in pathology 7 _ Capable in some circumstances of killing thc Int'ected host Table 4. Origin and properties of M. bovis strains used in these experimcuts M /,ov/s Virrlencc Orgin Carbon source stran in guinea for culture DigS _ _.. _

ATCC35723 Moderate Originally Is<'lated fiom ulfected bovne. 0 5 /O Glycerol Obtained from American Type Culture Coliecton. WAg200 Full AgResearch Wallaceville strain 89/5276. 0 4% PyruYate Isolated in this laboratory from infected bovine. WAg201 Full A,Rcscarch Wallaceville strain 86/5701. 0.4 /O Pyruvate Isolated in ths laboratory from infected bovine _ __

BCG Avn ulent Pasteur strain 1 173 P2; Statens Serumins;titut, 0.5% Glycerol _ Copenhagen, L)enmark _ _ Table 5. Concentration of antibiotics used for selection Antibiotic L' coil M bovs Kanamycln Hygromycin | 200 /ml | 50 tigim | Ampicillin | 100lg/ml I Not used Table 6. DNA oligonucleotides used in this work Oli onlcleotide Sc uence 5' - 3' g 4 _ DMC 120 T T G G C G C G C T A C A A C C T C A A G C 'r T [ N W 12(, A A G C T TGGTTAGAA'l'GACiATCTTCA DMC121 TTGGCGCGC1'ACAACCT

DMC 122 CC,CTACAACCTCAAGCT

DMC123 TGAAGATCTCATTCTAACC

DMC 124 ATCTCATTCTAACCAAGC'I'

DMC125 T T G G C (i C G C C A A G A A G A 'I C T T C DMC126 GAAGATCTTC1''1'G(iCGCGCCAA DMC 153 CACACCGCGAAAGCAGC

DMC 154 CTATCCCGGCACCGACG

DMC3 12 AACAGCCCCCT'I'GCGAC

DMC3 13 GGTTCAGGT<iCGGAGTC DMC330 GC(G TTAATTAACCCCGCGAAGATCCAG AG G'l'Ci DMC331 GCGTTAAT'I'AAC2C'AGCAGCAGCCCCGCCAG

Table 7A. Vaccination results for some illegitimate recombinants | AY boYs Spleen weght Spleen lesons Spleen M bovs recombinant mean (g) SF. mean No SE log',, CFU SE Non-vaccinated 2500.33" 46-lOa 4.974 0.17', BCG 1.29 0.09t' 0.25 0.18b 1.53 + O 53, WAg526 i.l70.1()i' 00-gob 1 21 - ()21t' WAg530 l.lS=O.106 00 T ool, ND WAg533 1 23 0.24 0 0 0 0 1.00 1 O 00 WAgS39 1 08 0 12b 0 0 0 ob 1.40 + 0.30 WAg569 1.07 t () 06b 0.33 1 0 33'' ND WAC,759 1.18 0.07 0.33 0 33i' 1 62 1 0 52 ilS, standard error b' values within the same column with the same suffix are not signil'icantly di l'ferent ND, not determined; Signil'icance for spleen lesions calculated on log> transformed clata Tablc 7B Vaccination results I'or some illegitimate recombinants bov.s strain Splecn weight Spleen lesions Spleen M: bcvs Lung M bov.y mean (g) + SlS mean No. + logO (:FIJ SE log' CIIJ SE SE Non-vaccillated 2.15 0.30t' 59 + 28a 4.96 0.30'1 5.18 + 0.14 13(G i.35 + 0.1 Uh (J. 1 7 0.00t' 2.33 = 0.3O" 3:: u 54! WAg520.28 0.1 3b 0,00 + 0.00].46 0.46 3 49 0 12b WAg520 4 1.53 _ 0.11b 0.00 + 0.00 1.54 + 0.43 4.02 0.26 Non-vaccinated 3.57 0.43a 86 + 21 a 5.63 0.09" 5.41 0. 13a BCG 1.45 + 0.11 b 0.67 + 0.49 2.55 0.49 3 71 i 0.30: WAg530 1.12 0. 07 1.0 0.5 2.69 _ 0.62 4.73 0.28a WAg530.1 1.58 _ 0 09b 2.3 0.7 3.81 0. |2b 4.47 0 3X 10 SE, standard error ah Values within the same column with the same suf'fix are not signil'icantly different Table 8. M; bovis recombinants with greatly reduced virulence from Signature 15 Tag Mutagenesis experiments Numbers and sizes of tuberculous Icsions * Parcnt Strain {;uinea Cuinca Guinea Livers Lungs Dcgrec oi Pi2 1 Di2 2 Pi2 3 virulence ... WAg201 I WAgS66 NVL NVL NVL NVL NVI.

WAg201 WAc567 NVL NVL NVL NVI. NVL _ _ WAg201 WAgS68 NVL NVL NVL NVL NVL -

WAg200 WAgS39 NVL NVL NVL NVL NVL__ WAg201 WAg570 NVL NVL lxlmm NVI. NVL + WAo201 WAgS71 NVL NVL 3 x 3mm NVL NVL + _ WAg201 WAgS72 NVL NVL 2 x <Imm NVL NVL _. WAg20i WAgS73 2 x imm NVL 3xi-2mm NVL NVL + W/\g201 WAgS74 3 x 1-2mm NVL Sxlmm NVL NVL + WAg200 WAg540 100 x I NVL NVL manyin I NVI. ++ animal _

* NVL: No visible lesions Tahle 9 Protection of possums against bovine tuberculosis bYt vaccination | Vaccine Change in body weight after Lung weight Spleenbacterial | _ challenge (g)a(g)b _ countc i WAgS20-92 14644 7 6 1 73 - O 33 _ WAg526995 13444 a- 7 0 1.64 0 25 _ Em_ WAc530 85*32 4.4* 0 87 0.17* WA,53342 i 49*_ 36 8.2 () 83 - O 13* WA, 'sl933 + 6Q*31 5 6* 1.13 0!4 _ _Q-. _

1 BCC; -108 77 31 5.3* 1 14 0.90

1 Non- -267 106 51 7 X 1.95 0 43 I vaccinated _ _ _.. a Mean change in body weight OF.

h Mcan lung weight SE Mean bacterial count, log 1 () colony forming units/g spleen:t SE * Significantly different to non-vaccinated group 10 Aspects of the present invention have been described by way of example only and it should be appreciated that modifications and additions may be made thereto without departing from the scope thercoi as defined in the appended claims.

References Aldovini, A., Husson, R.N., Young, R.A. (1993) The uraA locus and homologous recombination in Mycohaoterium bovis BCG..11Jacteriol 175: 7282-7089.

Aldwcll, F.F.., Wedlock, D.N., E3uddle, 13.M (1996) Bactcrial metabolis?n, r cytokine mRNA transcription and vibility of boN"!e alveolar macropha:, cs infected with Mycobuclerium lori.s BC:G or virulent M. hovi.. Immunol C.'ell Biol 74: 45-51.

Banerjce, A., lDubnau, E., Quemard, A., Balasubra?nanian, V., lm, K.., Wilson, 1., Collins, D., dc Lisle, (.,.lacobs, W.R. (1994) inhA, a gene encoding a target 10 tor isoniazid and ethionan1ide in Mytohacterium uberculosis. Science 263: 227 230. 13aulard, A., Kremcr, L., Supply, P., Vidaud, D., Bidart, J.M., Bellet, D.. Locht, C. (1996) A new series of mycobacterial expression vectors for the development of live recombinant vaccines. Cene 176: 149-154.

15 Behr' M.A., Wilson, M.A., Gill, W.l3., Salamon, H., Schooluik, G.K.,Rane, S., Sn?all, P.M. (1999). Comparative genomics of BCG vaccines by whole-genomc DNA microa?-ray. ÀScience 284: 1520-1523.

Bloom, B.R., Fine, P..M. (1994) The BCG experience: implications for fi?ture vaccines against tuberculosis. In: Bloom B.R. (ed), Tubcrculosis. I''thogene.si.s, 20 lrotection and Confrol, Washington DC: American Society for Microbiology, pp. 53 1 -557.

Brosch, R., Gordon, S.V., Marmicsse, M., Brodin, P., E3uchrieser, C., Liglmcier, K., Garnier,T., Gutierrez, C., Hewinson, G., Krcmer, K., Parso?ls, 1,.M. Pym, A.S. Samper, S., van Soolingen, D., Colc, S.T. (2002) A new evolutionary

scenario for the MycohacterinnZ tuberculosis complex. Proc Atntl Acacl. 7ci U S A 99: 3684-3689.

Buddle, B.M., Wards, B.J., Aldwell, 17.. Collins, 13.M., de Lislc, (,.W. (2002) influence of sensitisation to environmental mycobacteria on subsequent 5 vaccination against bovine tuberculosis. Vaccine 20: 1126ll33 (:amacho, L.R., lLnserg,ueix7 D., Perez, E., Giccluel, B., Guilhot, C. (1999) Identification of a virulence gene cluster of Mycohacterium [uherculoYIs by signature-tagged transposon mutagenesis. Iklol Microhiol 34: 257-67.

Colangeli, R.7 Spcncer' J.S. Bifani, P., Williams. A., Lyashchenko, K., Kcen, 1() M.A.. frill, P.J. Belisle7 J.7 Gennaro, M.L. (2()00) MTSA-IO, the product of the Rv3874 gene ol' MycabucteriZlm laherculosi.s, elicits tuberculosis-specific.

delayed-type hN:perseusitivit, ir,uinea pigs Ir',f. c 1imu,, 58: 990-993.

Cole, S.T.' Brosch, R., Parkhill, J., Gamier, '1.7 Churcher7 C.' Harris. D.7 Gordon.

S.V.7 Eiglmeier, K., Gas, S., Barry, C'.E. 3rd, Tekaia, L'., Badcock, K.. 13asham, D.7 15 Brown7 D., Chillingworth7 T.7 Connor7 R.7 Oavies7 R.7 Devlin, K., Feltweli7 '1'.7 Gentles, S., tiamlin, N., Ilolroyd, S., Hornsby, T., Jagels, K.7 Barrell, B.G., et al (1998) Deciphering the biology of Mycobacteriun1 tuberculosis from the conplete genome sequence Naure 393: 537-544.

Collins, L).M., Kawakami, R.P., de Lisle7 G.W.? Pascopella7 L.7 [3100m7 B. R., 20 Jacobs, W.R. (1995) Mutation of the principal sigma factor causes loss of virulence in a strain of the Mycobacterium uberculo.si.Y complex. Proc Nutl Acud Sci USA 92: 8036-8040.

Collins, D.M. (2000) New tuberculosis vaccines based on attenuated strains of' the Mycohaclerium tzherculosis complex. Immunol C'ell Biol 78: 342348.

so Collins, D.M. (2001) Virulence factors ol Mycobuclerium bovi.s. Tuberculo.is 81: 97-102.

Collins, D.M., Gicquel, B. (2000) Genetics of Mycobacterial Virulence. In: Hatful, G.F., Jacobs Jr., W.R. (eds), Molecular Geneiic.s of Mycohaceri.

5 Washington DC, American Society for Microbiology l'ress, pp. 265-78.

(Collins, 1)'M,, Wilson, T., Campbell, S., Buddle, B.M.. Wards,13.J., l lotter, G., dc Lisle, ().W. (200?) Production of avirrlent mutants of Mycolucteriur'' hovis with vaccine properties by the use ol illegitimate recombination and screening of stationary phase cultures. Microbiology 148: 3019-3027.

10 Con, J.S., Chen, B., McNcil, M., Jacobs.Ir.. W.R. (1999) Complex lipid determines tissue-specilc replication ot Mycohaclerium laherculosis in mice.

Are 402. 79-83.

Darzins 13. (1958) The Bacteriology of Tuberculosis. Minneapolis, University of Minnesota Press.

15.de Lisle, G.W. Wilson, T., Collins, D.M., Buddle, FI.M. (1999) Vaccination of gr,uinca pigs with nutritionally impaired avirulent mutants of Mycobucterum hovis protects against tuberculosis. Infecl Immun 67: 2624-2626.

DesJardin, L.F:., Schlesinger, L.S. (2000) Idcntifing,lIycobaclerium luberculosi.s virulence determinants - new technologies for a difficult problem. 7,en1.s 2() Microhiol 8: 97-99.

Dessen, A., Quenard, A., Blanchard, J.S.,.lacobs, W.R. Jr., Sacchettini, J.C.

(1995) Crystal structure and function of the isoniazid target of Mycohaclerium luberc2!osis Science. 267: 1638-1641.

Global Alliance for TB Drug Development (2001) Tuberculosis. Scientilc blueprint for tuberculosis drug development. T''herculosis (l:dfnh) 81 Suppl. 1: 1-

52. Gordon, S.V., FIrosch, R., Billault, A., Gamier' T., higlmeier, K., Cole, S.T.

b (1999) Identification of variable regions in the genomes of tubercle bacilli using,, bacterial artificial chromosome arrays. Idol Acrohil 37: 643-655.

Grover, A.A., Kim, H.K., Wiegcshaus, L.H. and Smith' D.W. ( 1')67) Host-

parasitc relationships in experimental airborne tuberculosis 11. Reproducible infection by means of an inoculum preserved at 70 C2.,I BaccJriol 94: 832-835.

1() Hatfull. Gall.. Barsom. L., Chant,, L., Donnelly-Wu, M., Lee, M.H., I, evin, M., Ncsbit, C.. Sarkis. C].J. (19')4) Bacteriopha',es as tools Air vaccine development.

Nero Dial Stan 80: 43-47.

* flensel, M.. Shea, J.E., Cileeson, C., Jones,, M.D.. Dalton, IS., Flolden, I>.W. (1995) Simultaneous identification of bacterial virulence genes by negative selection.

15 Science 269: 400-403.

Hess,.1., Kaufmann, S.H. (1999) Live antigen carriers as tools for improved anti tuberculosis vaccines. IrEMSImmunol Med,crobiol 23: 165-173.

Hinds, J., Mahentlliraling,am, E., Kempsell, K.., Duncan, K., Stokes, E(. W., I'arish, T., Stoker, N.G. (]999) Enhanced gene replacement in mycobacteria.

20 Microhiol 145: 519-527.

Hondalus, M.K., Bardarov, S., Russell, R., Chan.1., Jacobs fir., W.R., Bloom, B.R. (2000) Attenuation ol and Protection Induced by a Leucine Auxotroph ol A6'cokocteriz.m Iz.bcrc ulosis. Ir.fc?ct À'nznzufz 6 ,: 208'0,-28,9 .

Horwitz, M.A., Harth, G., Dillon. B.J., Maslesa-Galic, S. (2000) Recombinant bacillus calnL:tte-guerin (BCG) vaccines expressing the Mycohacterium uberculo.si.s 3(-kDa major secretory protein induce preater protective immunity against tuberculosis than conventional BCG vaccines in a highly susceptibJe [; animal model. ProcNutlAcadSci US'A t)7: 13X5313858.

Huang, C.C., Smith, C.V.. (Jlickman. M..7.lacobs' W.R. Jr., Sacchettini.. l.C.

(2002) Crystal structures of mycolic acid cyelopropane syntheses trom Mycobacierium taherculosis..JBJiol Chem 277: 11559-11569.

Kalpana, G.V., Bloom, B.R., Jacobs Jr, W.R. (1991)1nsertiona1 mutagenesis and 10 illegitimate recombination in mycobacteria. Broc Nal Acud tSci (J, 5 1 88: 5433 5437. Kent, P.T. and K',bica, G.P. (1985) Public!4ea!4h Mycob..cieriologj, A 5uide for the Level Ill 1,aboratory. l]S Departmenl of llealth and Iluman Services, Public I lealth Service, Centers for Disease Control, Atlanta, Georgia.

15 Knipter, N., Seth, A., Shrader, T.E. (1997) IJnnarked gene integration into the chromosome of Mycohlcerium smegmutis via precise replacement of the pyrF gene. Plasmic! 37: 29-1 40.

Lewis, J.A., Hatfull, G.F. (2000) Identification and characterization of mycobaeteriophage L5 excisionase. Mol Microhzol 35: 350-360.

20 Mahairas, G.G., Sabo, P.J., I-lickey, M..l., Singh, I).C, Stover, C.K. ( 1996) Molecular analysis of genetic differences between Mycobac'rizm hovi.s BCG and virulent M hovis. JBucteriol 17X: 1274-12X2.

McAdam, R A. Weisbrod,!.R.,..4ain, J., Scuderi, J.I)., Brown, A.M., Cirillo, J.l:)., Bloom, B.R., Jacobs, W.R..Ir (1995) ln vivo growth characteristics of

leucine and methionine auxotrophic mutants of Mycohacterium bovis BCG generated by transposon mutagenesis. Infect Immun 63: 1004- 1012.

McMurray, D.N., Carlomagno? M.A., Minizer, C.L. and 'I'et;:la0', C.L. (1985) Mycohacterium /,ovis BE'( vaccine ['ails to protect proteildeficient guinea pigs 5 against respiratory challenge with virulent Mycohacterium tuberculosis. Inf'ct Immun 5(): 555-559.

McKimey? J.D., [loner zu Bentrup, K., Muno:-Elias, E.J. et al. (2000) Persistence of Mycohacterium taherczllosi.s in macrophag,es and mice requires the glyoxylate shunt enzyme isocitrate Iyase..7\1ture 406: 735738.

10 Mecsas, J. (2002) Use of signature-tag nutagenesis in pathogenesis studies. (.'rr Opinion Microhiol 5: 33-37.

Murray, I'.J., Young, R.A. (1998) in In: Parish T. and Stoker N.G. (eds).

Mycohacteri Protocol.s, 'Iotowa, New.Iersey: llumana Press, pp. 275-284.

Niemann, S., Richler, F... Rusch-Gerdes, S. (2002) Biochemical and genetic 1o evidence l'or the transfer of Mycohacterium tuberculoss subsp. caprce Aranaz et al. 1999 to the species lllycobacterium hovis Karlson and Lessel 1970 (approved lists 1980) as Mycob'cterium bovis subsp. caprae comb. nov. Int JSist Eol Mzerohiol 52: 433-436.

Ohara, N., Yamada, T. (2001) Recombinant F3CG vaccines. Vflccine 19: 40X9 i20 4098. Parish, '1'., Stoker, N.G. (2000) 1Je of a flexible cassette method to generate a double unmarked MycoAccterium tuberculo.sis tlyA plcABC mutant by gene replacement. A4icr'hiolo,' 146: 1969.

Pascope]la, L., Collins, F.M., Martin,.l.M., Lee, M.H., lIatfull, G.., Stover, C.K., Bloom, B.R., and Jacobs Jr., W.R. (1994) IJse of in vivo complementation in = Scohaclerium uberculus.y to identify a genomic fragment associated with virulence. Ink ect Imrnun 62: 1 3 1 3 - I 3 1'3.

6 Pelicic, V., Jackson, M., Reyrat, J.M., Jacobs, W.R. Jr., Gicqucl, B., (uilhot, C. (]997) Efficient allelic exchange:n1 transposon mutagenesis in.;coLc1cteri1m tuberculosis. I'rocNatlAcacl'i USA 94: 10955-10960.

Pfyffer, G.., Auckenthaler, R., van F,mblen, J.D., van Soolingen, D. (1998) Mycohacterum cuneii, the smooth variant of M tuberculosis, isolated Irom a 10 Swiss patient exposed in Africa. Emory Infect Lois 4: 631-634.

Raman, S., gong, I., Puyang, X., Bardarov, S., Jacobs, W.R..Ir., Husson, R.N.

(?00!) The alternative sienna tractor Sigh regulates major co;npoicnts of oxidati; e and heat stress responses in Myco/'cterium /uherculo.si. J Bceleriol 183: 6119-

6125. A' Renshaw, 1'.., Panagiotidou, P., Whelan, A., Gordon, S.V., Hewinson, G.R., Williamson, R.A., Carr, M.D. Conclusive evidence that the major T-cell antigens of the M uberc'lo.sis complex ESA1-6 and CFP-1O form a tight, 1:1 complex and characterization of the structural properties of ESAT-6, Cl;P-lO and the ESA1-6-CFP-10 complex: implications for pathogencsis and virulence..1 Biol 20 Chem, in press.

Reyrat' J.M., I'clicic, V., Gicquel, B., Rappnoli, R. ( 1998) Countcrselectable markers: untapped tools tor bacterial genetics and pathogencsis. Inflict Imm'n 66: 401 1-401.

Sharma, V., Sharma, S., fIoener zu Bcntrup, K., McKinney. J.D., Russell. D.G., Jacobs, W.R. JrSacchettini,.l.C. (2000) Structure of isocitratc Iyase, a persistence factor of M,cobucleri'in t'.hercul'si.. A:ulure ruct Biol 7: 663-668.

Silver, S., Phung, L.l. (1996) Bactcrial heavy metal resistance: new surprises.

Annu lVev lVicrohiol 50: 753-7X9 Snewin, V., Stewart, G., Young, D. (20() 0) Cjenetic strategies lor vaccine development. In: TTatful, G..,.Tacobs Jr., W.K. (eds), Molecular Gcnctics ol Mycobacteria. Washington DC, American Society lor Microbiolog,y Press, pp. 279-296.

10 Steyn, A.J.C., Collins, D.M., Hondalus. M.K., Jacobs, W.R., Kawakami, R.P., Bloom, t3.R. (2002),livcohacerium taherculosis WhiB3 interacts with RpoV to affee' host survival but is dispensable for in vi.:o gr.\ih. r)r. N!a/! Ac.',';'., T,TC'., 99: 3147-3152.

van Pinxteren, L.A., Rav. 1.. Agger E.M., Pollock, J., Anlersen, P. (2000) 1S Diagnosis of tuberculosis based on the two specific antigens ESA 1-6 and CFF 10.

Clin Dilgn Lab Immunol 7: 155-160.

van Soolingen, D.. Hcrmans, P.W.M., de Haas, P.F..W., Soll, D.R., van EmUcien, J.D A. (1991) Occurrence and stability ol insertion sequences in Adycohaeterium uberculosis complex strains: F:valuation of an insertion sequence-dependent DNA 2) polymorphism as a tool in the epidemiology of tuberculosis..J Clin Microbiol 29: 2578-2586.

van Soolingen, D., Hoogenboozcn1, T., dc Haas, P.E., llermans' P.W., Koedam, M.A., Teppema, K.S., Brennan, 13.J., Besra, G.S., Portaels. I;., lop, J., Schouls, L.M., van Embden, J.D. (19')7) A novel pathogenic taxon of the Mycohaclerium

tfFerczlo.sis complex. Canctti: characterization of an exceptional isolate from Africa Int.J.Syst Bacteriol 47:] 236-1245.

Vordermeier, H.M. Whelan, A., (cockle, P.J., I;arrant, L., Palmer, N., Hewinson, R.G. (2001) Use of synthetic peptides derived from the antigens ESAT-6 and 5 CFP-10 for differential diagnosis ol bovine tuberculosis in cattle. C'lin Diugn Lab i lmmunol 8: 571-578.

Wards, B.J., Collins, D.M. ( l 996) F,lectroporation at elevated temperatures substantially improves transformation efficiency of slowgrowing mycobacteria.

FF:S Microbial Lelt 145: 101 -1 05.

10 Wards, By., de l,isle, G.W.' Collins, D.M. (2000) An esal6 knockout mutant of Mycohaclerium bovis produced by homologous recombination will contribute to the rlevelopment o! a live tuberculosis vaccine. Tuber' /e Lang Hi. 80 I X5-l 89.

Wedlock, D.N., Aldwcll, F.E.' Collins, D.M., de Lisle, G.W.? Wilson, T. and Buddle, B.M. ( 1999) Immunc responses induced in cattle by virulent and 15 attenuated Mycobucterium hovis strains: correlation of the delayedtype hypersensitivity with ability of strains to grow in macrophages. Infect Immzm 67: 2172-2177.

Wiegeshaus, E.H., McMurray, D.N., (drover, A.A., Harding, G.E. and Smith, FEW. (1970) Host-parasite relationships in experimental airborne tuberculosis Ill.

20 Relevance of microbial enumeration to acquired resistance in guinea pills. timer Rev Resp Dis 102: 422-429.

Wilson, l.M., de l.isle, G.W., Collins, D.M. (1995) Effect of inhA and kalG on isonia:.id resistance and virulence of Mycohaclerium hovi.v. Mol Microbial 15: 1009-1015.

Wilson, T., Wards B J., White, S.J., Skou, B., do lisle, G.W., Collins I). M.

(1997) Production of avirulent Myo'hacterium /'ovi.s strains by illegitimate recombination with DNA fragments containing an interrupted ahp( gene.

T''6ercle Lung Lois 78: 229-23 5.

O Young, D (2001) Letting the genome out ol the bottle: prospects for new drug development Ann N YAct7:],Yc i 93 146-1 50

Claims (1)

1. WHAT WE CLAIM IS:

   1 An isolated MycohacterzzJm species of the M. z/herculo.si.s complex which is attenuated and comprises at least one inactivated gene selected lrom the group consisting of: I 5 'pi.4; glnAt2; Rv()OY7; any of the genes between and including Rv()445c (sig,K) - Rv(J453 (PPE);

   any of the genes between a!ld in:!!di!lsg Rig 0175 - Rv 076 (h,rl.; any ofthe genes between and including Rv ()111 - Rv 0115; - Rv 3545c any of the genes between and including Rv3757c (pro W) - Rv376().

   2. An isolated Mycoh-clerzzJm species of M: tulercul:.si.s complex which is 15 attenuated and comprises inactivated or deleted genes selected from group COllSiStill of: all genes between and including Rv0445c (si,gK) R\0453 (PPE); all genes between and including Rv()l 75 - Rv()1fJ6 (hylS); all genes between and including Rv()l l l - Rv()l lS;

   3. An isolated Mycobacterium species as claimed in claims 1-2 wherein the 'cohacterum species of the M taherculoss complex is selected from existing or new members of the Of taherculo.si: complex currently comprising: M tuberculosis M cfrcanum, M microti, M hovis sbsp.

   5 C'aprae; M ccrnetii and M hovi5 A,A,n isolated Ill l3;cohcc/,ri?/m species as claimed in claim 3 wherein tle species selected lrom the M taherculosis complex is 1. bouts 5. An isolated Mycohacerium species as claimed in claim 3 wherein tile strain is selected from the group consisting of: 10 - WAg759: M hovis mutant with an inactivated ppiA gene; WAg530: M bovis mutant with an inactivated glib 4 genie; - VvTAg,533:.1! bovis mutant with an inactivated Rvnt)97 gene; WAg539: 111; hovis mutant with inactivation of the genes Rv()45c (sink) - Rv()3 (PPES); 1 - WAg526: M boviY mutant with inactivation of the genes RvO175 R0186 (hilly); WAg569: M. bovis mutant with inactivation of the genes RvOI I 1 Rv01 1 5, WAgS70: M hovis mutant with an inactivated Rv3545c cnc; 20 WAg537: by. hovis mutant with an inactivated Rv0(197 gene; WAgS66: M. bovi.s mutant with inactivation of the genes Kv3757c (pro W) - Rv376().

   6. An isolated Mycohacterium species as claimed in claim 5 which also comprises an inactivated esat6 acne.

   7. An isolated Mycobuclerium species as claimed in claim 3 wherein the strain comprises an inactivated glut gene and also comprises one or more 5 inactivated genes selected from the group consisting of Rv3974, Rv3375 and Rv3876. 8. An isolated Mycohac serum species as claimed in claim 7 wherein the strain is WAg530. 1.

   9. An isolated Mycohacterium species as claimed in chains 3 wherein the strain 10 comprises an inactivated Pv2i35 gene and also comprises one or more inactivated genes selected from the group consisting ol Rv3874, Rv3. Y75 and Rv3876. 10. An isolated Mycohacterium species as claimed in claim 9 wherein the strain is WAg520.4. 15 11. An isolated Mycobucteri''m species which is attenuated and comprises one or more inactivated genes that are identical to or have at least 70 /0-99 /O nucleotidc sequence homology to that of the inactivated genes as claimed in claims I - I 0.

   12. An isolated Mycohaclerum species oJ the M tuberculosis complex which is 20 attenuated and comprises one or more inactivated Bends) which is the same gene(s) as that inactivated by polar effects ol any of the inactivated genes as claimed in any one of the preceding claims.

   13 An isolated Alycobucierium species as claimed in claim 3 wherein the species is attenuated and comprises at least one inactive Irene selected tram any one of the preceding claims.

   14. An isolated Mycobucterium species ok the M tuberculosis complex which is 5 attenuated and comprises at least one inactivated gene that is identical or has at least 70-99% nucleoiide sequence homology to any one of the inactivated genes as claimed in claim].

   15. An isolated Mycobaclerzm species as claimed in claims 1-14 wherein the mutant is further attenuated by inactivation of an additional gene.

   10 l 6 A method of producing a live attenuated vaccine strain of any Mycoh'cterium species including A/ album and its subsp. opium, paralaherculos'.s anti isil'ctlicz!m and,i,I 1'!Ce.'-'!.1S COmpLiSillQ steps of inactivating in these species one or more genes that are identical to or have at least 50%-99% nucleotidc sequence homology to that oi the inactivated genes as claimed in any one of 15 the preceding claims.

   17. A vaccine to prevent or treat tuberculosis infection in mammals wherein the vaccine comprises an isolated Mycobucterium species as claimed in any one of claims 1-15.

   18. A vaccine as claimed in claim 17 wherein the vaccine also includes a 20 pharmaceutically or veterinarily suitable carrier or diluent.

   19. A vaccine as claimed in claim 18 vvllerein the vaccine also includes an adjuvant or other immuno-stimulant.

   20. A vaccine as cIaime_ in claims 17-19 wherein the Mycfacterum species comprises one or more foreign genes that are capable of enhancing the ability

   of the vaccine to stimulate the immune system of the diseased mammal to increase the effectiveness of said vaccine.

   21. vaccine as claimed in claim 20 wherein the foreign gene may encode a polypeptide antigen, a eytokine or other immuno-sLimulant.

   0 22. A composition to prevent or treat tuberculosis infection in mammals wherein the composition comprises an isolated Mycobacterium species as claimed in cluing 1-]5, together with a pharmaceutically or veterinarily suitable carrier or dilucnt.

   23. A method of screening mutants of a Myccbactcrim species of the Al 10 tz'erculo.sis complex for attenuation comprising the steps ol a) isolating a Mycohacerium species to be screened; b) growing the isolated mutants for 5-10 weeks in a complete medium; c) transferring the 5-10 week old mutants to a minimal media; and 15 d) selecting mutants based on their inability to grow in a minimal media. 24. A method of producing and screening mutants ol a Mycobccterium species of the M. taherculosis complex tor attenuation comprising the steps of a) constructing a number of unique tagged suicide vectors containing 2() a marker gene (e.g. antibiotic resistance gene); b) introducing the vectors to a culture of a MycoAclcterizfm species and optionally leaving a predetermined amount of time;

   c) plating onto a suitable medium to select for marker gene; selecting mutants which contain marker gene, combining mutants together, and probing with DNA to identify the presence ol' all mutants in an 'input pool; 5 e) injecting the 'input pool' into a test mammal; f) collecting mutant colonies from test mammals to create an 'output pool'; and g) probing 'output pool' to identify mutants which arc no longer present (ie. attenuated mutants).

   10 25. A method of producing and screening mutants of,lfycohuctcriu,'' species of AS t'!hercl!csis connplex fair attenuation as claimed in either claim 23 or 24 wherein the test mammal is be selected from the group consisting of', mice guinea pigs, rats, possums. cattle, deer, ferrets, badgers and rabbits 26. A method of producing and screening mutants ol' Mycohaclerum species of 16 M tuberculosis complex for attenuation as claimed in either claim 23 or 24 wherein the test mammal is a guinea pig.

   27. 'l'he use ol' an isolated Mycohac/erum species as claimed in claims l]S in the manufacture of' a vaccine or composition to treat or prevent taherculosis inflections in mammals.

   20 28. A method of determining vvhether a candidate drug is capable of inhibiting a :olypeptide involved in mycobacterial infection comprising the steps of: a) providing an isolated polyneptiie encoded by one of the cencs as claimed in any one of claims l-l S;

   b) providing a candidate drug; c) use of an assay which measures biological activity of the polypeptide in a); and d) measuring inhibition of biological activity of the polypeptide in a).

   S 29. A method of determining whether a compound is an effective drug candidate against tuberculosis comprising the steps of: a) producing a stationary phase culture of a strain of the M uberc'lo.si.s complex; b) subculturing it in a minimal medium with the addition of a 10 compound that is a potential drug; and c) comparing its growth in b) to that of a similar subculture in tile same medium without the added compound.

   30. A method of detecting whether a drug is eficctive against mycobacterial infection comprising the steps of 15 a) administering to an animal having a mycobacterial infection, a candidate drug as selected by any one of the methods as claimed in claims 28 or 29.

   b) assessing effectiveness of the drug at reducing or eliminating the clinical symptoms of said infection.

   20 31. An isolated Mycob1cterium species as described herein with reference to any example thereof: 32. A method of producing a live attenuated vaccine strain of any Mycobacterium

   species as described herein with reference to any example thereof.

   33. A vaccine to prevent or treat tuberculosis infection in mammals as described herein with reference to any example thereof 34. composition to prevent or treat tuberculosis inf'cction in mammals as 5 described herein with reference to any example thereof.

   35. A method of screening mutants of a Mycobuclerium species t'or attenuation as described herein with ret'erencc to any example thereof'.

   36. A method of producing and screening mutants of a A1ycoEccteritin species for attenuation as described herein with reference to any example thereof'.

   10 37. IJsc of an isolated MycrJlucte/Ainn species in the nanufcturc ol'a vaccine or composition to treat or prevent tlhe'r!./io.si.s infections in mammals as described herein with ret'erence to any example thereof.

   38. A method of determining whether a candidate drug, is capable ol'inhihiting a polypeptide involved in mycobacterial infection as described herein with 1.r rel'erence to any example thereof: 39. A method of determining whether a compound is an effective drug candidate against uberczlo.sis as described herein with reference to any example thereof. 40. A method of determining whether a drug is el'l'ective against mycobactcrial 20 infection as descriLcd herein with reference to any example thereof.