WO1998033923A1

WO1998033923A1 - MUTANT msbB or htrB GENES

Info

Publication number: WO1998033923A1
Application number: PCT/GB1998/000291
Authority: WO
Inventors: Duncan John Maskell; Gordon Dougan
Original assignee: Imperial College of London
Current assignee: Imperial College of London
Priority date: 1997-01-30
Filing date: 1998-01-30
Publication date: 1998-08-06
Anticipated expiration: 1999-07-30
Also published as: EP0973911A1; AU5873498A

Abstract

Nucleic acid for a mutant msbB or htrB gene derivable from Salmonella which results in loss of an msbB or htrB encoded protein, respectively, or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity.

Description

Mutant msbB or htrB genes The present invention relates to nucleic acid for a mutant msbB gene or a mutant htrB gene, a recombinant DNA construct comprising the nucleic acid, a micro-organism comprising a mutant msbB or htrB gene, an inactivated msbB or htrB gene or lacking a msbB or htrB gene, and uses thereof, particularly, but not exclusively, its use in a vaccine.

Lipopolysaccharide (LPS) forms the outer leaflet of the outer membrane of Gram negative bacteria. In most cases it is highly biologically active, being both immunogenic and the active principle in endotoxin. The structure of LPS may be considered to be divided into several domains. The domain of LPS which is required for the activities associated with endotoxin is lipid A. Lipid A molecules are able to induce the release of a number of cytokines as well as nitric oxide and it is through these mediators that the effects of endotoxin are seen.

Most lipid A molecules may be divided into hydrophilic and hydrophobic domains. The hydrophilic region consists of a 1-6 linked D-glucosamine (GlcN) disaccharide backbone substituted by phosphate groups at positions 1 and 4' , which may in turn be linked to, or replaced by, pyrophosphorylethanolamine or 4-amino-4- deoxy-L-arabinose. The hydrophobic region consists of fatty acids and these may vary between species. In S. typhimurium the lipid A has a fatty acylation pattern in which the 2 and 2' amino groups and the 3 and 3' hydroxyl groups on the diglucosamine are each linked to 3-hydroxytetradecanoic acid (3-OH- 14:0). The 2'- linked fatty acid is further substituted at the 3-hydroxyl group by dodecanoic acid (12:0) and the 3' fatty acid is again further substituted at the 3-hydroxyl group by tetradecanoic acid (14:0) (1,2). These have been called secondary acylations and will hereinafter be referred to as such.

It is believed that the pattern of fatty acylation in natural lipid A may have consequences for the biological activity (and thus toxicity) of the molecule. For example Rhodobacter sphaeroides lipid A is non-toxic and differs from toxic lipid A only in the pattern of fatty acyl substitutions (1,2). It is also known that treatment of lipid A with hydroxide ion cleaves the secondary acyl chains from the molecule with consequent detoxification. Indeed an acyloxyacyl hydrolase is present in neutrophils that catalyses precisely this cleavage and is probably one of the mechanisms responsible for detoxifying lipid A in vivo. However, both these systems are naturally occurring and there is no indication of how these observations could be applied to other systems.

Recently Raetz et al. (1,2) have made progress in understanding lipid A biosynthesis in E. coli, and have cloned, sequenced and mutagenized most of the genes necessary for it. The early stages of the pathway have been known for some time but the enzymes and genes required for the secondary acylations have only recently been discovered. The addition of secondary fatty acids to the hydroxyl groups of the 2'- and 3'-linked hydroxytetradecanoic acids completes lipid A biosynthesis and these reactions are catalysed by the products of the htrB and msbB genes. Very recently Somerville et al. (3) disclosed an msbB mutant of E. coli which was shown to have greatly reduced ability to induce cytokines (especially TNF alpha) and E-selectin expression in an in vitro system. Clearly, phenomena observed using in vitro models may have little significance in vivo. Further this work is limited to E. coli.

EP-A-0 650 733 describes an attenuated vaccine for avian species comprising a micro-organism which may be Salmonella or E. coli amongst others. The approach taken is to use a micro-organism which exhibits auxotrophy to one or more growth factors, such that it is incapable of growing on a minimal medium in the absence of said one or more growth factors.

We have taken a different approach to the problem of providing vaccines against virulent organisms. We employ an msbB mutant gene and/or htrB mutant gene, gene deletion and/or inactivation to give rise to a bacterium which makes a lipid A molecule which has reduced toxicity or which is non-toxic compared to the lipid A molecule produced by a wild-type bacterium. This approach has a substantial advantage over the approach described in, for example, EP-A-0 650 733 not least because it results in a vaccine which has a reduced endotoxicity and therefore a reduced reactogenicity. According to one aspect of the present invention there is provided nucleic acid for a mutant msbB gene derivable from Salmonella which results in loss of MsbB protein or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity compared to the wild-type lipid A molecule.

According to another aspect of the present invention there is provided nucleic acid for a mutant htrB gene derivable from Salmonella which results in loss of HtrB protein or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity compared to the wild-type lipid A molecule.

In other words during the biosynthesis of a lipid A molecule the mutant msbB or htrB gene results in loss of MsbB or HtrB protein respectively, which in turn results in the biosysnthesis of a lipid A molecule with a reduced ability to induce cytokines.

The lipid A molecule is one which forms part of LPS. Whilst not wishing to be bound by any theory it is believed that the loss of the msbB encoded protein or the loss of function of the msbB encoded protein will give rise to a lipid A molecule lacking at least secondary acylation of the hydroxyl group of the 2 '-linked hydroxytetradecanoic acid of the lipid A. Similarly, it is believed that the loss of the htrB encoded protein or loss of function of the HtrB protein will give rise to a lipid A molecule lacking at least secondary acylation of the hydroxyl group of the 3'-linked hydroxytetradecanoic acid of the lipid A molecule. Thus, preferably the lipid A is deficient in at least one of the secondary acyl chains which are usually associated with a lipid A domain of a lipopolysaccharide. In a particularly preferred embodiment the lipid A molecule lacks both secondary acyl chains. In one embodiment the mutant is derivable, or derived, from Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli. As previously mentioned the mutant may be derivable, or in a particularly preferred embodiment is derived, from Salmonella. Thus the mutant can be arrived at by mutating a wild type Salmonella micro-organism or more specifically its msbB or htrB gene. However, synthetic nucleic acid fall within the scope of the present invention. Thus although the original mutant may have been derived by mutating Salmonella, the mutant may be sequenced and the nucleic acid of interest reproduced, e.g. synthetically, using techniques well known to the skilled worker. This is also true when the mutant is derived from a micro-organism other than Salmonella.

Any convenient technique which is, or becomes available, may be used to modify the gene. These will be known to workers skilled in the art. One preferred method uses genetic manipulation of msbB or htrB by insertion of a kanamycin resistance cassette to inactivate the gene, conjugation of the inactivated gene into the recipient to be mutated on a suicide vector, followed by P22 transduction into other recipients. In a particularly preferred embodiment of the present invention the microorganism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella sofia, Salmonella havana, Salmonella bovis-morbificans , Salmonella hadar. Salmonella arizonae or Salmonella anatum. In an especially preferred embodiment of the present invention the microorganism is S. typhimurium, and preferably the strain is C5, SL1344 or HWSH.

Preferably the mutation or loss of protein is not lethal for growth of a microorganism. This has the advantage that the micro-organism can be easily cultured without having to add supplements to the medium. In fact it is surprising that viable bacterial are produced after alteration of a component of the lipid A molecule. As previously mentioned, preferably the lipid A molecule has a reduced ability to induce a cytokine response. Ability to induce a cytokine response is a conventional toxicity measure. In general we have found that the lipid A molecules produced by the present invention have the ability to reduce cytokine induction down to about ^lA-^lA of that induced by wild-type lipid A molecules. More preferably the lipid A molecule and/or micro-organism induces less TNF-α and/or less IL-ljS and/or less NO. More preferably the lipid A molecule induces at least 5-fold less TNF-α and at least half as much IL-lβ as the corresponding wild-type. In another preferred embodiment the lipid A molecule induces at least half as much NO as the corresponding wild-type. Thus it will be appreciated that the present invention provides for the toxicity to be substantially reduced. In an especially preferred embodiment there is substantially no toxicity.

In one preferred embodiment, the micro-organism of the present invention kills a BALB/c mouse when the population of the micro-organism in the liver and/or spleen reaches about 10⁹ per organ. In fact it actually only kills a proportion of the infected mice, around 5-10%, even at such a high level of 10⁹ per organ. This can be compared to the wild-type where a micro-organism population of about 10⁸ per organ is sufficient to kill all mice infected. It is preferable to compare the reduced toxicity of the lipid A molecule arrived at using the present invention and/or toxicity of the micro-organism of the present invention against the toxicity of a lipid A molecule produced by the parent wild-type. By parent wild-type we mean the micro-organism from which the mutant was derived, e.g. the wild-type micro-organism which was used to produce the mutant, or the wild-type micro-organism in which the mutant was engineered.

Thus according to one preferred embodiment of the present invention there is provided nucleic acid derived from Salmonella and encoding for a mutant msbB gene or a mutant htrB gene which results in a lipid A molecule having reduced toxicity compared to the lipid A molecule produced by the respective msbB encoded protein or htrB encoded protein encoded for by the corresponding Salmonella msbB/htrB gene from which the mutant is derived. Whilst not wishing to be bound by any theory the mutant msbB and htrB genes of the present invention may result in a polypeptide which is truncated with respect to the polypeptide encoded by the non-mutated gene, or indeed loss of the peptide.

That being said the present invention also encompasses any polypeptide molecule encoded for by the nucleic acid of the present invention and/or produced by the micro-organism of the present invention.

According to a further aspect of the present invention there is provided a recombinant DNA construct comprising the DNA of the present invention cloned into a cloning or expression vector. According to a further aspect of the present invention there is provided a recombinant micro-organism comprising the recombinant DNA construct of the present invention.

It will be appreciated that as well as using a mutant msbB or htrB gene the same effect may be achieved by actually deleting the msbB or htrB gene from the genome or by inactivation. Any convenient technique which is, or becomes available may be used to delete or inactivate the gene. These will be known to workers skilled in the art. Thus, according to one aspect of the present invention there is provided a Salmonella micro-organism comprising a mutant msbB or htrB. an inactivated msbB or htrB gene or lacking a msbB or htrB gene and having reduced toxicity compared to the parent wild-type, i.e. the Salmonella micro-organism from which it is derived. According to a preferred embodiment there is provided a recombinant Salmonella micro-organism transformed with a mutant msbB or htrB gene, comprising an inactivated gene or from which the gene has been deleted and having reduced toxicity compared to the micro-organism prior to transformation.

Preferably the Salmonella micro-organism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella ofia, Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar, Salmonella arizonae or Salmonella anatum. More generally, the present invention provides a micro-organism comprising a mutated Salmonella msbB or htrB gene, an inactivated msbB or htrB gene or a micro-organism from which the msbB or htrB gene has been deleted.

Also the present invention provides a micro-organism comprising an inactivated msbB or htrB gene: a mutated Salmonella msbB or htrB gene or from which the gene has been deleted, and which results in loss of an msbB encoded protein or htrB encoded protein, respectively; or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity.

According to a preferred embodiment of the present invention the microorganism is Salmonella, Shigella, Klebsiella, Enterobacter. Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella. Pseudomonas, Acinetobacter Moraxella,

Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli.

In a particularly preferred embodiment of the present invention the microorganism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C. Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella sofia, Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar. Salmonella arizonae or Salmonella anatum.

In an especially preferred embodiment of the present invention the micro- organism is S. typhimurium, and preferably the strain is C5, SL1344 or HWSH.

According to yet another aspect of the present invention there is provided a live vaccine comprising an attenuated or avirulent micro-organism having a mutated msbB or htrB gene, inactivated msbB or htrB gene or lacking the gene and having reduced toxicity in accordance with the present invention. The mutation may be introduced into live attenuated vaccine strains of, e.g.

Salmonella, thus reducing their endotoxicity and thereby reducing their reactogenicity. This would generate safer vaccine strains that would be more acceptable to the licensing authorities and to the general public. The same strategy might be used for all live attenuated Gram negative bacterial vaccines. A prime example here would be the new live attenuated Shigella vaccines. The same effect may arise with an inactivated or deleted gene.

According to a further aspect of the present invention there is provided a method of immunising a subject comprising administering a vaccine of the present invention. Preferably the vaccine is against infection caused by a micro-organism which is Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas. Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli. In a more preferred embodiment the micro-organism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis , Salmonella montevideo, Salmonella newport, Salmonella enteritidis. Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella sofia, Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar, Salmonella arizonae or Salmonella anatum. The subject may, for example, be a mammal or avian. Examples of such mammals include humans, cattle, swine and ovine species. Examples of such avians include chickens, ducks, turkeys, geese, bantams, quail and pigeons.

In order to prepare the vaccine of the present invention the micro-organism must be attenuated or rendered avirulent. The vaccine composition of the present invention may be administered by injection or orally, and the composition must be suitable for the desired administration route. Suitable vaccine compositions are well known to those skilled in the art.

Bacteria with mutations in the msbB or htrB gene, an inactivated gene or lacking the gene would provide excellent background strains for the production of proteins and nucleic acid for vaccines and therapeutics, substantially removing the requirement for downstream processing to remove the erstwhile toxic LPS molecules.

Thus according to yet another aspect of the present invention there is provided use of a micro-organism having a mutant msbB or htrB gene, an inactivated gene or lacking said gene and having reduced toxicity in the recombinant production of a protein or gene of interest.

The isolated LPS made by these mutants may be useful as an endotoxin antagonist.

A sample of S. typhimurium C5 having a msbB mutation in accordance with the present invention has been deposited under the Budapest Treaty at NCIMB on 17

January 1997 and accorded the deposit number NCIMB 40856. Further characteristics of this deposited micro-organism are given below in the Examples.

Mutants lacking both htrB and msbB may synthesise Lipid IV _A - KDO₂ which is a non-toxic antagonist of lipid A. These mutants will thus be a source of this molecule which may be used to treat septic shock resulting from endotoxaemia.

Thus it will be appreciated that the present invention also extends to constructs and micro-organisms comprising (i) a mutant msbB gene which results in loss of MsbB protein or the loss of function of the protein; an inactivated msbB gene; or which lacks the msbB gene, in combination with (ii) a mutant htrB gene which results in loss of HtrB protein or loss of the function of the protein; an inactivated htrB gene; or which lacks the htrB gene. The present invention also includes the use of such a so-called msbB/htrB double mutant as a vaccine and pharmaceutical compositions comprising it, together with its use in producing genes and proteins of interest. Preferably the mutations in accordance with the present invention are mutations which are substantially incapable of reversion. A substantially non- reversible mutant has a reversion frequency preferably of < 10^"8, more preferably < 10 ^"9, even more preferably < 10^"10, and most preferably a mutant with zero reversion.

Various preferred features and embodiments of the present invention will now be described by way of non-limiting example and with reference to the accompanying drawings in which:

Figure 1 is a graph showing growth curves of wild-type and msbB mutant S. typhimurium in BALB/c mice. The two growth curves are indistinguishable in the first week of infection. All the mice infected with wild-type organisms died by 1 week post-infection, whereas most of the mice infected with the msbB mutant survived. Subsequently the msbB mutant was cleared from the livers and spleens of infected animals;

Figures 2a and 2b are graphs representing the in vitro analysis of TNF-α and IL-ljS. 2xl0⁶ cultured J774 macrophage-like cells were incubated with 10⁵ msbB mutant or wild-type Salmonella both of which had been heat-killed. A time course of release of TNF-α and IL-1/3 from these cells in response to the bacteria was determined. Mutant Salmonella induce 5-fold less TNF-α (Figure 2a) and half as much IL-1/3 (Figure 2b) as the wild-type organism;

Figure 3 is a graph representing NO generation in vitro. 2xl0⁶ cultured J774 macrophage-like cells were incubated with 10⁷ msbB mutant and wild-type Salmonella that had been heat-killed. Following 24 hours incubation, the culture medium was assayed for NO by the Griess reaction, which detects NO by determining nitrate/nitrite in the medium. Mutant Salmonella induced half as much NO as wild-type bacteria;

Figures 4a and 4b are graphs representing an in vivo study of cytokines. Serum samples were taken at 24 hours from mice infected with wild-type or msbB mutant organisms. These samples were assayed for TNF-α (Figure 4a) and IL-10

(Figure 4b) by ELISA. Analysis of the results showed that the mutant was inducing approximately 4-fold less TNF-α and 2-fold less IL-1_/3 than the wild-type bacteria. These results correlate precisely with the observed reduced lethality of the mutant;

Figure 5 is a graph representing the results of Example 3, an oral vaccination study using an aroA mutant of S. typhimurium in BALB/c mice: and

Figure 6 is another graph representing the results of Example 3, an oral vaccination study using an msbB/aroA mutant of S. typhimurium in BALB/c mice. Whilst not wishing to be bound by any theory we believe that the fatty acyl substitutions in a lipid A molecule of the LPS domain of a bacterium, to a large extent, determine the toxicity of the molecule and, furthermore, if alterations in fatty acid substitution could be engineered, then previously toxic LPS molecules may be detoxified. Thus, genetic manipulation of bacteria such that at least one of the secondary acyl chains is not added should reduce the molecule's toxicity and furthermore engineering a bacterium such that both (or all) secondary acyl chains are not added should make the lipid A molecule substantially non-toxic. Work using natural Salmonella infections in mice to assess the precise role of lipid A and endotoxin in disease is described below.

Salmonella typhimurium causes a severe invasive disease in mice, which shares many features in common with typhoid fever, caused by S. typhi in humans. Mouse typhoid has been extensively investigated, generating a vast amount of data regarding virulence and immunogenicity (4). Using parenteral inoculation into inbred mice, several patterns of growth of the bacteria in vivo have been observed, and this growth is controlled by a number of host genetic systems. The best studied of these is that regulated by the Ity gene, which has recently been cloned and named nramp. After intravenous inoculation of S. typhimurium into mice over 90 % of the inoculum is killed within the first few hours of infection, but the survivors then live and grow within macrophages of the mononuclear phagocyte system (MPS). The rate of growth of the bacteria over the first few days of infection is controlled by nramp such that inbred mice may be divided into susceptible and resistant types. Susceptible mice (e.g. BALB/c) allow the growth of typical virulent Salmonellae at a rate of increase of about ten-fold per day per liver or spleen. Resistant mice (e.g.

A/J) allow only half this rate of increase.

In lethal infections the Salmonellae grow too fast to be controlled by primary host responses and rapidly reach levels of approximately 10⁸ per organ. The mice then die with extensive organ damage. The precise mechanism of tissue damage leading to death in mouse typhoid is unknown. It has been speculated that endotoxin levels in the livers, spleens and other infected organs become so high that tissue is damaged, but there is currently no direct evidence to support this. Tissue damage may be induced, in the presence of high levels of endotoxin, by cytokine-mediated responses.

In sublethal infections bacterial growth in the livers and spleens is suppressed around day 4 to 5 of infection by a T-cell-independent mechanism and bacterial counts reach a "plateau" . This requires recruitment of bone marrow-derived cells and coincides with the formation of focal lesions then granulomas. The induction of plateau requires the involvement of a number of cytokines, including TNFα and IFN7(5-7). TNFα acts to recruit monocytes to the site of infection and IFN7 is required for macrophage activation. An IL-12 requirement in this induction has also recently been established, with it probably acting as a positive modulator of IFNγ production (8). LPS, and more specifically its lipid A domain, has been described as a potent inducer of all three of these cytokines in many systems. It is possible that the signal inducing the host to begin synthesising these cytokines, and subsequently to control the infection in mouse typhoid, is dependent, at least in part, on the lipid A domain of LPS. This hypothesis has not been established previously for this model.

An intriguing feature of immunisation of mice with live attenuated S. typhimurium vaccine strains is that 7 days after immunisation of C3HeB/FeJ mice there is profound suppression of responses to B- and T-cell mitogens and suppression of the capacity of spleen cells to mount primary in vitro plaque-forming- cell responses to sheep erythrocytes. This inhibition is mediated by nitric oxide (9, 10). The bacterial inducer of the nitric oxide response leading to immunosuppression in this model is unknown, but it is likely that endotoxin is involved.

In summary, in mouse typhoid it seems probable that low-level induction of cytokine responses by low levels of LPS, likely to be present in sublethal infections, induces a protective mechanism that enables the animal to survive (i.e. plateau) while the tissue responses induced by the high levels of LPS likely to be present in the last stages of a lethal infection are damaging and eventually lead to organ failure and death. Example 1 We have cloned and partially sequenced the msbB gene from S. typhimurium.

Briefly, a probe based on E. coli msbB DNA sequence was generated using the polymerase chain reaction (PCR), cloned and radiolabelled. This was used to probe a Southern blot of Salmonella typhimurium DNA, identifying a 3.2kb Dral fragment. In addition the oligonucleotides were also used in a PCR using S. typhimurium DNA as template. This generated an approximately lkb piece of DNA which was cloned into pGEM-T. On sequencing from either end of this construct it was clear from amino acid and DNA sequence homology that this was msbB. To generate an msbB mutant in the S. typhimurium chromosome it was first necessary to insert an antibiotic resistance marker into the msbB coding sequence. To do this new oligonucleotides, based on the Salmonella DNA sequence, were generated and used to PCR the gene from the pGEM-T clone. This was then treated with Klenow enzyme to blunt the ends of the DNA and digested with Sail to cut the DNA into 450bp and 550bp fragments. The Sail site is in the coding sequence of the Salmonella msbB gene. A gene cassette encoding kanamycin resistance (Pharmacia) was also cut with Sail. The two fragments of the PCR product, the kanamycin resistance cassette and pBluescript that had been digested with EcoRV were then mixed and ligated. This was then transformed into E. coli with selection on ampicillin and kanamycin. Resultant clones were screened for the correct plasmid product. One of these was chosen for further studies. The entire insert from this plasmid was removed using Pvull and cloned into the suicide vector pCVD442 which had been digested with Smal. This was transformed into E. coli carrying the pir gene to allow pCVD442 to replicate. Resultant plasmids were again checked for the correct insert size. One of these was chosen to be used in making the mutant. E. coli donor bacteria were conjugated with S. typhimurium LB 5010 recipients using standard methods. After incubating the conjugation mixture, the bacteria were harvested and plated onto selective media containing kanamycin and sucrose. pCVD442 contains the sacB gene, the product of which confers sensitivity to the presence of sucrose in the medium. Plating on media containing sucrose thus selects against the presence of vector sequences. Of the colonies that grew on the selection plates, one was picked for further study. To check if the msbB gene has been mutated, chromosomal DNA was prepared and used as template in a PCR using the msbB-specific oligonucleotides. This showed a 2.3kb band in the mutant, with a 1 kb band appearing in the wild-type control, confirming that the mutant had a rearranged msbB gene. The LB 5010 strain is not virulent for mice. It serves as an intermediate in making mutants, since it is mutated in its DNA restriction system, but not its DNA modification system. DNA that passes through LB 5010 is thus modified, which allows a better frequency of introduction of the DNA into its final recipient. To move the chromosomal mutation from LB 5010 into the wild type virulent S. typhimurium strains, P22 transduction was used. Briefly, LB 5010 msbBv.Km was infected with P22 HT101 int and plated. At the appropriate dilution confluent lysis was observed after overnight incubation at 37° C. These plates were harvested and the bacteriophage recovered as a plate stock. This stock was used to infect S. typhimurium strains C5, SL1344 and HWSH with subsequent selection on kanamycin plates. Again, resulting colonies were checked for chromosomal rearrangement by PCR. The mutant bacteria gave a band at 2.3kb and the wild-type controls a band at 1 kb as expected, confirming that msbB mutants had been generated. Before biological assays were performed with these mutants, the LPS molecule was checked to see if the mutation affecting the lipid A domain had had any effects on the more distal parts of the molecule. An SDS-PAGE gel of the LPS from the wild-type and mutant bacteria was performed and stained with silver. No difference was observed between the LPS molecules in terms of ladder pattern of the O-antigen, or intensity of staining. Furthermore, the msbB mutant was also susceptible to infection with bacteriophage P22. which requires O-antigen for binding to the bacterial surface, confirming that the msbB mutants were able to synthesise full length LPS molecules. Example 2 - Infection study in BALB/c mice

To test the virulence of the S. typhimurium msbB mutant it was injected into nram -susceptible BALB/c mice and its growth in vivo was followed. Mice infected with wild-type organisms died as expected after 7 days of infection with counts in livers and spleens reaching approximately 10⁸ per organ (Figure 1). Intriguingly, the msbB mutants grew at exactly the same rate as the wild-type (w.t.) bacteria, but only caused approximately 5 % of the infected animals to die. Death only occurred when the bacterial counts had reached very high levels (approximately 10⁹ per organ).

After this, the mutant Salmonellae were gradually cleared until there were no counts in livers and spleens (Figure 1).

To investigate likely reasons why the mutant bacteria were not as lethal as their wild-type parents, cytokine release and iNOS induction in vitro and in vivo in response to the bacteria were measured. First the release of TNF-α and IL-lβ from

J774 macrophage-like cells was measured. The cultured cells were incubated with 10⁵ heat-killed wild-type or msbB mutant bacteria and the two cytokines were measured in the culture medium by ELISA. It can be seen from Figure 2 that mutant bacteria induced 5-fold less TNF-α and half as much IL-lβ as their wild-type parents. This is as expected given that the msbB mutant has a reduced toxicity lipid A molecule. The release of NO from J774 cells was next determined. Again cultured cells were incubated with 10⁷ heat-killed wild-type or msbB mutant bacteria and culture medium was assayed for NO by the Griess' reaction. The mutant bacteria induced half as much NO as the wild-type bacteria (Figure 3).

Finally, cytokine levels in vivo were measured. At 24 hours post- infection, serum samples were taken and assayed for TNF-α and IL-lβ using ELISA. These results show that the mutant induced 4-fold less TNF-α and half as much IL-lβ than the wild-type bacteria. These results correlate precisely with the in vitro results and strongly suggest that the reduced lethality of the msbB mutants is due to their reduced ability to induce potentially harmful cytokine responses: in short, because the toxicity of their lipid A molecule has been reduced. This is the first direct evidence that endotoxin is responsible for lethality in this infection.

Studies using the msbB, aroA/msbB and aro mutants of S. typhimurium - Vaccination and Challenge Studies Example 3 - Oral Vaccination and Challenge

30 mice/group were inoculated with 10⁸ or 10⁹ oral by gavage tube S. typhimurium aroA or S. typhimurium aroA/msbB. Animals were left for 30 days and then challenged with wild type 5. typhimurium 10⁸ oral by gavage tube. Two separate groups of mice were inoculated as described above but animals in the two groups were killed on days 1, 3, 5, 9, 14, 21 and 28. The livers and spleens of infected mice were homogenised and viable counts performed on the surface of agar plates (see Figures 5 and 6).

Challenge with the wild type strain in the group receiving S. typhimurium aroA resulted in all the animals surviving (100% protection, 30/30 mice alive). In the group given S. typhimurium aroA/msbB and challenged with S. typhimurium wild type 16/30 mice survived the wild type challenge (40% protection). This result indicates that the aroA/msbB mutation in S. typhimurium is more attenuated than the aroA mutation alone.

Example 4 - Intravenous Challenge of BALB/c mice with S. typhimurium aroA and 5. typhimurium msbB (i.v. LD50).

Groups of mice were challenged intravenously with 5. typhimurium aroA or S. typhimurium msbB at doses of 10², 10³, 10\ 10⁵, 10⁶, 10⁷, 10⁸, and 10⁹. Mice were left and observed for deaths in all groups. S. typhimurium aroA and aroA/msbB all lived at doses up to 10⁶ i.v. All of both sets of mutants died at 10⁷ i.v. but the msbB group lived significantly longer than the aroA infected group. S. typhimurium aroA infected mice died at day 7 but the msbB group lived up to 3 weeks after the aroA group died.

References

1. Raetz CRH. Biochemistry of endotoxins. Annu. Rev. Biochem. 1990; 59: 129-170.

2. Raetz CRH. The enzymatic synthesis of lipid A. In Levin J, Alving CR, Munford RS and Stutz PL (eds) Bacterial Endotoxin: Recognition and Effector

Mechanisms. Endotoxin Research Series Vol. 2. Elsevier, Amsterdam, 1993; pp. 39-48.

3. Somerville JE, Cassiano L, Bainbridge, B, Cunningham MD and Darveau RP. A novel Escherichia coli Lipid A mutant that produces an a i inflammatory lipopolysaccharide. J. Clin. Invest. 1996; 97: 359-365.

4. Hormaeche CE, Villareal B Mastroeni, P, Dougan G. Chatfield SN. Immunity mechanisms in experimental salmonellosis. In Cabello F, Hormaeche CE, Mastroeni P, Bonina L (eds) Biology of Salmonella. New York: Plenum Press, 1993: pp 223-335. 5. Hormaeche CE, Mastroeni P, Arena A, Uddin J, Joysey HS. T-cells do not mediate the initial suppression of a Salmonella infection in the RES. Immunology 1990; 70: 247-250.

6. Mastroeni P, Arena A, Costa GB, Liberto MC, Boninaq L, Hormaeche CE. Serum TNF8α in mouse typhoid and enhancement of the infection by anti-TNFα antibodies. Microb. Pathogen. 1991; 11 : 33-38.

7. Muotiala A, Makela PH. The role of IFN7 in murine Salmonella typhimurium infection. Microb. Pathogen. 1990; 8: 135-141.

8. Mastroeni P, Harrison JA, Chabalogoity JA, Hormaeche CE. Effect of interleukin-12 neutralization on host-resistance and gamma-interferon production in a mouse typhoid. Infect. Immun. 1996; 64: 189-196.

9. Eisentein TK, Huang D, Meissler JJ, Alramadi B. Macrophage nitric-oxide mediates immunosuppression in infectious inflammation. Immunobiol. 1994; 191: 493-502.

10. Huang D, Schwacha MG, Eisenstein TK. Attenuated Salmonella vaccine- induced suppression of murine spleen cell responses to mitogen is mediated by macrophage nitric oxide-quantitative aspects. Infect. Immun. 1996: 64: 3786-3792.

Claims

1. Nucleic acid for a mutant msbB gene derivable from Salmonella which results in loss of an msbB encoded protein or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity.

2. Nucleic acid for a mutant htrB gene derivable from Salmonella which results in loss of an htrB encoded protein or loss of function of the protein, which in turn results in a lipid A molecule having reduced toxicity.

3. Nucleic acid according to claim 1 or claim 2 wherein the lipid A molecule has reduced toxicity compared to a lipid A molecule resulting from a protein coded for by a wild-type gene.

4. Nucleic acid according to any preceding claim wherein the lipid A molecule is substantially non-toxic.

5. Nucleic acid according to any preceding claim wherein the lipid A molecule has one secondary acyl chain.

6. Nucleic acid according to any one of claims 1 to 4 wherein the lipid A molecule has no secondary acyl chains.

7. Nucleic acid according to any preceding claim wherein the mutant gene is not lethal for growth.

8. Nucleic acid according to any preceding claim wherein the lipid A molecule has a reduced ability to induce cytokines compared to a lipid A molecule resulting from an MsbB protein coded for by a wild-type msbB gene.

9. Nucleic acid according to any preceding claim wherein the mutant gene is derivable or derived from Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas, Acinetobacter , Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella,

Haemophilus or Escherichia coli.

10. Nucleic acid according to claim 9 wherein the mutant gene is derived from Salmonella.

11. Nucleic acid according to claim 10 wherein the mutant gene is derivable or derived from Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or

C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi , Salmonella dublin. Salmonella sofia. Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar, Salmonella arizonae or Salmonella anatum.

12. Nucleic acid according to any preceding claim derived by genetic manipulation of a Salmonella msbB or htrB gene comprising insertion of kanamycin resistance cassette to inactivate the gene, conjugation into a recipient to be mutated on a suicide vector, followed by P22 transduction into another recipient.

13. Nucleic acid for a mutant msbB gene having the characteristics of the mutant msbB gene contained in NCIMB Deposit Number 40856.

14. Isolated nucleic acid according to any preceding claim.

15. Nucleic acid according to any preceding claim which is cDNA or synthesised DNA.

16. A polypeptide which is coded for by the nucleic acid of any preceding claim.

17. A polypeptide according to claim 16 which is isolated from any other proteins with which it is naturally associated.

18. A recombinant DNA construct comprising the nucleic acid of any one of claims 1 to 15.

19. A recombinant DNA construct comprising the nucleic acid of any one of claims 1 to 15 cloned into a cloning or expression vector.

20. A recombinant micro-organism comprising the recombinant DNA construct of claim 18 or claim 19.

21. A recombinant micro-organism of claim 20 wherein the micro-organism is Salmonella, Shigella. Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium,

Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli.

22. A recombinant micro-organism of claim 21 wherein the micro-organism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella sofia, Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar, Salmonella arizonae or Salmonella anatum.

23. A micro-organism having the characteristics of NCIMB Deposit Number 40856.

24. A micro-organism comprising nucleic acid for a mutant msbB gene derivable from Salmonella which results in loss of an msbB encoded protein or loss of function of the protein; an inactivated msbB gene; or which lacks an msbB gene, and having reduced toxicity.

25. A micro-organism according to claim 24 which is Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas. Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli, with the proviso that when the micro-organism is Escherichia coli, the mutant msbB gene is not from Escherichia coli.

26. A micro-organism comprising nucleic acid for a mutant htrB gene derivable from Salmonella which results in loss of an htrB encoded protein or loss of function of the protein; an inactivated htrB gene; or which lacks an htrB gene, and having reduced toxicity.

27. A micro-organism according to claim 26 which is Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli

28. A micro-organism comprising (i) nucleic acid for a mutant msbB gene derivable from Salmonella which results in loss of an msbB encoded protein or loss of function of the protein; an inactivated msbB gene; or which lacks an msbB gene, and (ii) nucleic acid for a mutant htrB gene derivable from Salmonella which results in loss of an htrB encoded protein or loss of function of the protein; an inactivated htrB gene: or which lacks an htrB gene, which in turn results in a lipid A molecule having reduced toxicity.

29. A micro-organism according to claim 28 which is Salmonella, Shigella, Klebsiella. Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas. Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli.

30. A micro-organism according to any one of claims 24 to 29 wherein the micro-organism is Salmonella typhimurium, Salmonella typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum. Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin,

Salmonella sofia, Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar. Salmonella arizonae or Salmonella anatum.

31. A micro-organism according to any one of claims 24 to 30 which is substantially non-toxic.

32. A micro-organism according to any one of claims 24 to 31 in which the said nucleic acid is not conditionally lethal for growth.

33. A micro-organism according to any one of claims 24 to 32 which has a reduced ability to induce cytokines compared to a corresponding wild-type microorganism.

34. A micro-organism according to any one of claims 24 to 33 which produces a lipid A molecule which has one secondary acyl chain.

35. A micro-organism according to any one of claims 24 to 33 which produces a lipid A molecule which does not have a secondary acyl chain.

36. A micro-organism according to any one of claims 24 to 35 comprising the nucleic acid of any one of claims 1 to 15 or in the form of a recombinant microorganism of any one of claims 20 to 22.

37. A live vaccine comprising an attenuated or avirulent micro-organism comprising a mutant msbB gene, an inactivated msbB gene or lacking an msbB gene and which exhibits a reduced toxicity.

38. A live vaccine comprising an attenuated or avirulent micro-organism comprising a mutant htrB gene, an inactivated htrB gene or lacking an htrB gene and which exhibits a reduced toxicity.

39. A live vaccine comprising an attenuated or a virulent micro-organism comprising (i) a mutant msbB gene; an inactivated msbB gene; or lacking an msbB gene; and (ii) a mutant htrB gene; an inactivated htrB gene; or lacking an htrB gene, and which exhibits a reduced toxicity.

40. A vaccine according to any one of claims 37 to 39 wherein the microorganism exhibits a reduced toxicity compared to a corresponding wild-type microorganism.

41. A vaccine according to any one of claims 37 to 40 wherein the microorganism is substantially non-toxic.

42. A vaccine according to any one of claims 37 to 41 wherein the mutated micro-organism is not lethal for growth.

43. A vaccine according to any one of claims 37 to 42 wherein the microorganism produces a lipid A molecule which has one secondary acyl chain.

44. A vaccine according to any one of claims 37 to 42 wherein the micro- organism produces a lipid A molecule which does not have a secondary acyl chain.

45. A vaccine according to any one of claims 37 to 44 wherein the microorganism is Salmonella, Shigella, Klebsiella, Enterobacter, Serratia, Proteus, Yersinia, Vibrio, Aeromonas, Pasteurella, Pseudomonas, Acinetobacter, Moraxella, Flavobacterium, Bordetella, Actinobacillus, Neisseria, Brucella, Haemophilus or Escherichia coli.

46. A vaccine according to claim 45 wherein the micro-organism is Salmonella typhimurium, Samonells typhi, Salmonella paratyphi A or C, Salmonella schottmulleri, Salmonella choleraesuis, Salmonella montevideo, Salmonella newport, Salmonella enteritidis, Salmonella gallinarum, Salmonella pullorum, Salmonella abortusovi, Salmonella abortus-equi, Salmonella dublin, Salmonella sofia,

Salmonella havana, Salmonella bovis-morbificans, Salmonella hadar, Salmonella arizonae or Salmonella anatum.

47. A vaccine according to any one of claims 37 to 46 derived from a microorganism according to any one of claims 20 to 36, except that with reference to claim 25 the proviso does not apply.

48. A method for immunising a subject comprising administering an effective amount of a vaccine of any one of claims 37 to 47.

49. A pharmaceutical composition comprising a micro-organism according to any one of claims 20 to 36, except with reference to claim 25 the proviso does not apply.

50. Use of a micro-organism according to any one of claims 20 to 36, except with reference to claim 25 the proviso does not apply.