WO2006129090A2

WO2006129090A2 - Vaccines

Info

Publication number: WO2006129090A2
Application number: PCT/GB2006/001989
Authority: WO
Inventors: Laura Jane Violet Piddock
Original assignee: The University Of Birmingham
Priority date: 2005-06-02
Filing date: 2006-06-02
Publication date: 2006-12-07
Also published as: WO2006129090A3; GB0511203D0

Abstract

Medicaments comprising artificial proteins of the bacterial efflux pump AcrAB-TolC as well as medicaments comprising Gram-negative bacteria having defective efflux pump function. In particular vaccines comprising Gram-negative bacteria in which the function of the efflux pump component ToIC is defective.. Medicaments and vaccines of this sort are useful in the prevention and/or treatment of diseases such as Salmonellosis and Campylobacteriosis, and in the prevention and/or treatment of infection or colonisation of humans or poultry by Gram- negative bacteria.

Description

VACCINES

The present invention relates to medicaments suitable for use in the prevention and/or treatment of bacterial infection and/or colonisation. More particularly, the invention relates to vaccines of use in the prevention or treatment of diseases associated with Gram- negative bacteria.

Bacteria may be classified as either Gram-positive or Gram-negative based on their reaction to the Gram staining protocol. This protocol comprises the use of a crystal violet stain and safranin counterstain, the two staining steps being separated by an acetone- based decolourising step. Gram-negative bacteria are so named since they do not retain the Gram stain. Gram-negative bacteria are surrounded by a cell wall having a high lipid and low peptidoglycan content, and this cell wall composition allows the crystal violet stain to be washed away during decolourisation.

Gram-negative bacteria are frequently of medical or veterinary interest, since many Gram-negative species are pathogenic. The pathogenic nature of such bacteria is frequently associated with certain components of their cell walls, particularly the lipopolysaccharide (endotoxin) layer.

The class of Gram-negative bacteria encompasses many well-known bacterial genuses including Salmonella spp., Campylobacter spp., Escherichia coli, Klebsiella spp., Pseudomonas spp., Enterobacter spp., Serratia spp., Citrobacter spp., Proteus spp., Morganella spp., Acinetobacter spp., Stenotrophomonas spp., Haemophilus spp., and Moraxella spp.

The various species of Salmonella are among the most medically relevant Gram-negative bacteria. Many species of Salmonella are capable of infecting humans, birds, reptiles and other animals. Two Salmonella serotypes, S. Typhimurium and S. Enteritidis, are responsible for the majority of food poisoning incidents in humans. A further serotype of interest is S. Java, which has been linked with recent outbreaks in Scotland and in 60% of the Dutch poultry flocks. Salmonellosis is a major problem in most countries infecting over 160,000 individuals in the EU annually. Symptoms of salmonellosis include fever, headache, nausea, vomiting, abdominal pain and diarrhoea. Examples of foods involved in outbreaks of salmonellosis are eggs, poultry and other meats, raw milk and chocolate. Salmonella contamination of eggs was one of the main microbiological food safety issues of the 1990s. A Department of Health-funded survey in 1995/96 found that approximately 1 in 600 eggs were contaminated with Salmonella, with the majority of contamination thought to be on rather than in the egg. In recent years a number of legislations have been passed to control Salmonella entry into the food chain.

The U.S. Centre for Disease Control and Prevention estimates that 75 million people suffer food-borne illnesses each year in the United States. These cases account for 325,000 hospitalizations and more than 5,000 deaths. The economic impact of such illnesses is high, costing an estimated $5 to $6 billion in direct medical expenses and lost productivity.

Infectious intestinal disease causes substantial morbidity and economic loss in the United Kingdom and is responsible for over 300 deaths and 35,000 hospital admissions annually in England and Wales.

In March 1989 new Salmonella legislation was announced, which covered compulsory slaughter and mandatory reporting of results of tests for Salmonella. Further legislation was introduced in 1993, with the release of the Poultry Breeding Flocks and Hatcheries Testing Order, which required that hatcheries can only supply flocks that are Salmonella Enteritidis and Typhimurium negative. Following the introduction of these legislations the contamination of poultry products was significantly reduced.

Salmonella legislation is supported by codes of practice and assurance schemes, which are industry led arrangements that have been developed to ensure that standards of welfare, traceability, husbandry, storage and other aspects of production are met in order to help improve customer confidence in the integrity of products. The Department for the Environment, Food and Rural Affairs (DEFRA) and the Food Standards Agency (FSA) support the development of assurance schemes concerning egg and poultry production as they provide consumers with assurance on food safety and can also provide producers with the opportunity for better marketing, subject to clearance from the competition authorities. Assurance schemes include assured chicken production and Lion Code of Practice.

The assured chicken production scheme is a voluntary industry funded programme to cover conditions of poultry production and slaughtering. Under the conditions of bird health all parent birds must be vaccinated with a licensed Salmonella Enteritidis and Salmonella Typhimurium vaccine.

The Lion Code of Practice is applied at each point in the egg production chain. All breeding flocks are tested weekly for Salmonella, with flocks testing positive for Salmonella Enteritidis and Salmonella Typhimurium being slaughtered. Within rearing farms all pullets are vaccinated against Salmonella Enteritidis. The lion code only requires compulsory vaccination against Salmonella Enteritidis, but does encourage vaccination against Salmonella Typhimurium.

From 2004 the European Zoonoses Directive will demand the slaughter of any flocks infected with a salmonella species of significance to human health, this directive exerts extreme pressure on breeders, rearers, layers and broilers to maintain salmonella free flocks. Consequently the demand for a vaccine that is effective against a number of Salmonella strains will be in significant demand.

In the light of the above it will be appreciated that there remains a well-recognised need for new vaccines effective against Gram-negative bacteria. It is an object of certain aspects and embodiments of the present invention to provide improved vaccines for use in the prevention and/or reduction of bacterial colonisation and/or infection. It is also an object of certain aspects and embodiments of the invention to provide alternative vaccines for use in the prevention and/or reduction of bacterial colonisation and/or infection. In a first aspect of the present invention there is provided a vaccine comprising an artificial protein of the AcrAB-TolC efflux pump, or a variant, fragment or derivative thereof. The inventors have found that vaccines in accordance with this aspect of the invention (which may be referred to hereinafter as "peptide vaccines") are effective in preventing and/or treating bacterial colonisation or infection of host animals. Vaccines in accordance with the invention may be used to eliminate harmful bacteria from an animal that would otherwise be subject to bacterial colonisation or infection. The invention also provides the use of an artificial protein of the AcrAB-TolC efflux pump, or a variant, fragment or derivative thereof, in the manufacture of a medicament for the prevention of a disease caused by bacterial activity.

The use of vaccines in accordance with the first aspect of the invention is particularly beneficial as the AcrAB-TolC efflux pump proteins incorporated in the vaccines are expressed by many bacterial serovars, and hence the vaccines of the invention are able to confer immunity across a range of different serovars.

Suitable diseases caused by Gram-negative bacteria that may be prevented and/or treated using the vaccines, methods and medicaments of the invention include salmonellosis (which occurs as a result of infection or colonisation by Salmonella), campylobacteriosis (which occurs as a result of infection or colonisation with Campylobacter), and a range of disorders including diarrheal illness, hemolytic-uremic syndrome, HUS, which is a potentially devastating consequence of enteric infection with specific E coli strains, urinary tract infections (UTIs) and neonatal sepsis and meningitis (all of which may be caused by infection or colonisation by E Coli).

Except for where the context requires otherwise, references to "prevention" or "treatment" of infection or colonisation by Gram-negative bacteria should be considered to refer to a degree of prevention or treatment sufficient to prevent or treat a disease associated with infection or colonisation by the Gram-negative bacteria in question (i.e. such references should be taken to require "effective" prevention or treatment wherein the infection or colonisation is maintained below a threshold level at which pathological effects may otherwise occur, rather than "absolute" prevention or treatment in which no infection or colonisation occurs).

Vaccines according to the invention are for use in the prevention and/or treatment of a number of diseases caused by Gram-negative bacteria. Representative diseases that may be prevented and/or treated using the vaccines and methods of the invention can be selected from the group consisting of salmonellosis; campylobacteriosis; diarrheal illness; urinary tract infections (UTIs); and neonatal sepsis and meningitis. The skilled person will readily appreciate that, in the case of bacterial vaccines or methods of treatment of the invention, these may preferably be used in the prevention and/or treatment of diseases caused by bacteria of the type utilised in the vaccine itself (though obviously the diseases will normally be caused by bacteria having normal efflux pump function).

Except for where the context requires otherwise references to the use of "peptides" "proteins" or "artificial proteins" in vaccines in accordance with the invention should be taken to also encompass the use of variants of such proteins, fragments of such artificial proteins, and derivatives of such proteins or fragments. Furthermore, as is explained further below, a number of the components of the AcrAB-TolC efflux pump comprise lipoproteins. For the sake of clarity references to "proteins" herein should also be taken to encompass suitable lipoproteins, and particularly the amino acid portions of such lipoproteins.

Artificial proteins, or variants, or fragments or derivatives thereof suitable for use in accordance with the present invention will be immunologically effective proteins, variants, fragments or derivatives. The term "immunologically effective" will, for the purposes of the present invention, be taken to refer to those immuno-reactive proteins, variants, fragments or derivatives that are able to elicit an immunising response in an animal to which they are administered. Such an immunising response may suitably be elicited as the result of either the single or multiple administration of an immunologically effective protein, variant, fragment or derivative. These single or multiple administrations should be sufficient to provide an effective immunising dose (defined elsewhere in the specification) of the immunologically effective protein, variant, fragment or derivative. One skilled in the art of vaccine development will readily be able to identify suitable means by which the ability of proteins, variants, fragments and derivatives of interest to generate an immunising response may be assessed.

Bacterial efflux pumps are involved in the excretion of substances harmful to bacteria from the bacterial cytoplasm to the extra-bacterial space. This excretion thus prevents or reduces the potentially damaging effects of the harmful substance. Many bacterial efflux pumps are known, amongst the best characterised of which is the AcrAB-TolC efflux pump (found in. all bacteria in the Enterobacteriaceae family including Escherichia coli; Salmonella spp., Klebsiella spp., Serratia spp., and Haemophilus influenzae).

The AcrAB-TolC efflux pump is a tripartite complex, its three component members being AcrA, AcrB and ToIC. These three components associate to form a multi-component pump which functions to actively excrete harmful substances from the bacterial cytoplasm to the extra-bacterial space beyond the outer membrane.

AcrA is an approximately 42 kD membrane fusion lipoprotein also known as B0463, SipB, MbI, Lir, NbsA, MtcA and the acridine efflux pump. AcrA has a periplasmic location in intact bacteria, and is anchored to the outer surface of the inner membrane by the lipid moiety of the lipoprotein. It is believed that AcrA and AcrB interact to form a complex that is stable even in the absence of ToIC. A small region at the C-terminus of AcrA has been shown to be necessary for the interaction of the protein with AcrB. The amino acid sequence of AcrA of Salmonella Typhimurium is shown as Sequence ID No.l, and a nucleotide sequence encoding acrA of this serovar as Sequence ID No.2. In a preferred embodiment a vaccine of the invention may comprise an artificial AcrA protein (or a fragment or derivative thereof) such as the AcrA of Salmonella Typhimurium. In an alternative preferred embodiment a suitable vaccine may comprise artificial AcrA of E. coli. The amino acid sequence of AcrA of E. coli (which shares 81% identity with that of Salmonella Typhimurium) is shown as Sequence ID No. 3, and a nucleotide sequence encoding this protein as Sequence ID No. 4. AcrB is an approximately 113 kD RND-type permease also known as AcrE and B0462. AcrB is associated with the inner membrane of intact bacteria, and has twelve membrane- spanning α-helices. The presence of AcrB in the AcrAB-TolC efflux pump complex allows electrochemical-gradient energy to drive the active excretion of substances from the bacterial cytoplasm. The amino acid sequence of AcrB of Salmonella Typhimurium is shown as Sequence ID No.5, and a nucleotide sequence encoding acrB as Sequence ID No.6. In a further preferred embodiment a vaccine of the invention may comprise an artificial AcrB protein (or a fragment or derivative thereof) such as the AcrB of. Salmonella enterica. Alternatively a preferred vaccine may comprise an artificial AcrB of E. coli. The amino acid sequence of AcrB of E. coli is shown in Sequence ID No. 7 (which shares 89% identity with AcrB of Salmonella Typhimurium), and a nucleic acid encoding this protein is shown in Sequence ID No. 8.

Both AcrA and AcrB are encoded by the single acrAB locus. In intact bacteria AcrAB (the complex of AcrA with AcrB) is able to extrude substances from the bacterial cytoplasm without their accumulation in the periplasmic space, indicating that AcrAB functions in conjunction with an outer membrane channel. Genetic and co-localisation studies have indicated that this channel is provided by ToIC.

ToIC is an approximately 54 kD bacterial porin also known as B3035, WeeA, Toe, Refl, MukA and MtcB. In vivo ToIC is believed to form a functional trimer, with each monomer comprising a beta barrel of 18 membrane-spanning beta strands. The amino acid sequence of ToIC of Salmonella Typhimurium is shown as Sequence ID No.9, and a nucleotide sequence encoding tolC as Sequence ID No.10. In a still further preferred embodiment a vaccine of the invention may comprise an artificial ToIC protein (or a fragment or derivative thereof). A suitable ToIC may, for instance, be the ToIC of Salmonella Typhimurium. As an alternative, a further preferred vaccine may comprise an artificial ToIC protein derived from E. coli. The amino acid sequence of ToIC of E. coli (which has 83% identity with AcrA of Salmonella Typhimurium) is shown as Sequence ID No. 11, and a nucleotide encoding this protein as Sequence ID No. 12. The inventors believe that the use of ToIC in peptide vaccines of the invention is particularly advantageous, and that this use may give rise to vaccines that are surprisingly more effective than those that may be generated using outer membrane components of other efflux pumps.

The cytoplasm of Gram-negative bacteria is bounded by the plasma membrane, which is in turn enveloped by the outer membrane. Effective bacterial efflux pumps must allow the transport of substances from the interior of a bacterium to the extra-bacterial space. Accordingly, such pumps must enable the transport of substances across the two membranes surrounding bacteria. Suitable bacterial efflux pump proteins (or their fragments or derivatives) for use in accordance with invention may be selected with reference to the location of such proteins in intact bacteria.

The inventors have surprisingly found that efflux pump proteins that are associated with the bacterial inner membrane may be used as effective vaccines. This finding is surprising since it may be expected that such proteins, which are normally "hidden" from a host by the outer membrane of intact bacteria, would not normally constitute successful vaccine candidates since antibodies raised against such proteins would not have access to (and therefore not be able to react to) the protein in live bacteria. In the light of the inventors' finding it may be preferred that artificial bacterial efflux pump proteins (or their fragments or derivatives) for use in accordance with the present invention may be efflux pump proteins associated with the inner membrane of intact bacteria. For example, in the case of the AcrAB-TolC efflux pump, AcrA or AcrB may advantageously be used.

As an alternative, artificial bacterial efflux pump proteins (or their fragments or derivatives) suitable for use in accordance with the invention may be efflux pump components associated with the outer membrane of intact bacteria. In the case of the AcrAB-TolC efflux pump a suitable component may be the outer membrane channel ToIC.

Vaccines in accordance with the first aspect of the invention may comprise two or more artificial proteins of the AcrAB-TolC efflux pump. Vaccines containing two or more such artificial proteins may, for example, comprise one protein associated with the inner membrane of intact bacteria and one protein associate with the outer membrane of intact bacteria. Suitable examples of vaccines in accordance with this embodiment of the invention include those comprising:

i) AcrA in combination with ToIC; and ii) AcrB in combination with ToIC.

The vaccines of the invention provide a valuable alternative to vaccines already known in the prior art. It is generally recognised that it is advantageous to have multiple vaccines against any particular deleterious bacterium, since bacterial populations may occasionally undergo spontaneous mutation that may render known vaccines ineffective. By the development of an expanded range of usable vaccines the likelihood of being unable to treat a given bacterium, even one that has undergone mutation, is reduced. It is also recognised that it may frequently be preferred to employ vaccines, rather than antibiotic compounds, to control potentially harmful bacterial populations, since the promiscuous use of antibiotics is associated with the development of resistant bacterial populations that are not susceptible to treatment.

For the purposes of the present invention, by an "artificial protein" is meant a non-natural bacterial efflux pump protein (i.e. a bacterial protein that has not been isolated from bacteria naturally expressing said protein). For example, artificial proteins suitable for use in the vaccines of the invention may include recombinant proteins, which is to say bacterial proteins expressed by cells other than their natural bacterial sources. Techniques for the recombinant expression of bacterial proteins are well know to those of skill in the art, and include over-expression of the gene after cloning into an expression vector and subsequent purification of the protein via an affinity tag. A range of suitable vectors that may be used for this purpose are commercially available from various companies known to those skilled in the art. Many suitable protocols use E. colt strains to express artificial proteins (e.g. DH5α, topo 10). Illustrative examples of academic papers describing suitable methods for the expression of artificial proteins in bacteria include Motoi et al. (2005); Pillai et al. (2005) and Belli et al. (2004). Preferred sources for the production of recombinant bacterial efflux pump proteins include various strains of E. coli as described above and in the cited references. It should be recognised that bacterial efflux pump proteins expressed by bacteria other than their natural bacterial sources are encompassed by the term artificial protein as used in the present invention.

As mentioned above, expression of artificial proteins suitable for use in accordance with the invention may preferably be effected after attachment of an affinity tag (such as a "his"-tag) to the protein. This method allows artificial proteins expressed in this way to be purified on a column using well-known separation techniques. Furthermore, the presence of an affinity tag (such as a his-tag) in this manner allow artificial proteins suitable for use in accordance with the invention to be readily distinguished from naturally occurring proteins.

Thus in a preferred embodiment a suitable peptide vaccine in accordance with the invention may comprise an artificial protein of the AcrAB-TolC efflux pump bearing an affinity tag (such as a "his" tag). Even more preferably a suitable peptide vaccine may comprise artificial ToIC bearing an affinity tag.

As an alternative to the techniques described above, artificial proteins for use in accordance with the invention may be synthesised by any suitable method known to those skilled in the art. Suitable methods for the synthesis of artificial proteins may include fmoc solid phase synthesis or boc synthesis. It will be appreciated that the synthesis of artificial proteins in this manner may be of greatest advantage in the production of fragments or derivatives of artificial proteins suitable for use in the vaccines of the invention, rather than in the production of full-length proteins for use in such vaccines.

Fragments of the artificial proteins to be used in accordance with the invention may comprise any immuno-reactive portion derivable from a suitable full-length artificial protein. Suitable fragments may include those generated on enzymatic, or other, cleavage of the artificial protein. The use of such fragments may be preferred since the means for producing such fragments (suitable eixzymes having the requisite cleavage activity) are relatively cheap, widely available and readily capable of generating the quantities of peptide fragments necessary for commercial use.

Proteolytic enzyme cleavage sites are well known, and may be readily identified in bacterial efflux proteins. The analysis of bacterial efflux pump proteins in order to identify suitable enzyme cleavage sites may readily be undertaken using commercially available software. Suitable enzymes may then be employed to allow the generation of desired fragments. The skilled person will also appreciate that suitable proteolytic enzyme cleavage sites may be introduced into artificial bacterial efflux pump proteins in order to allow the enzymatic generation of fragments that would not otherwise be able to be generated. For example, artificial bacterial efflux pump proteins may be produced in which enzyme cleavage sites are introduced flanking amino acid residues having particular preferred immuno-reactive properties.

Fragments suitable for use in accordance with the invention also include fragments generated de novo, as opposed to on digest of a full-length sequence. Such fragments may be artificially synthesised using the methods considered above. The use of such artificial fragments may be preferred since they allow a greater range of potential fragments to be developed, since the artificial fragments that may be generated are not constrained by the presence in a full-length protein of suitable enzyme cleavage sites. Thus a desired artificial fragment may be produced irrespective of whether such a fragment is naturally bound by enzyme cleavage sites, and without the fragment produced being bound by residues left on cleavage by an enzyme. Artificial fragments produced in this way may preferably be fragments having preferred immuno-reactive properties, for example artificial fragments reproducing (or otherwise based on) epitopes having increased immunogenicity. Methods by which the immunogenicity of peptides, fragments or derivatives may be investigated are described further below.

Although artificial proteins of the AcrAB-TolC efflux pump represent preferred agents for use in vaccines of the invention, it will be recognised that there are contexts in which the sensitivity of peptide agents to degradation may be disadvantageous. There are many known techniques by which peptide derivatives may be produced that have greater resistance to degradation than do the original peptides from which they are derived.

Peptides based on AcrAB-TolC afflux pump proteins, and suitable for use in the vaccines of the invention, may be cyclised and/or stabilised using well-known techniques. The terminal amino acid residues of the peptides may also be subject to modification, for example the amino terminal residue may be acylated, and/or the amino acid residue at the carboxy terminal may be amidated.

Peptoids derived from AcrAB-TolC efflux pump proteins may be expected to have greater resistance to degradation than do the unmodified peptides, and such derivatives may be readily designed from knowledge of these peptides' structure. Commercially available software may be used to develop suitable peptoid derivatives according to well- established protocols. It will be appreciated that the well-characterized sequence of the native peptides facilitates the design and testing of peptoid and other derivatives.

Retropeptoids based on AcrAB-TolC efflux pump proteins (but in which all amino acids are replaced by peptoid residues in reversed order) may also represent suitable agents for use in the vaccines of the invention. A retropeptoid may be expected to bind a reactive antibody in the opposite direction, as compared to a peptide or peptoid-peptide hybrid containing one peptoid residue. As a result, the side chains of the peptoid residues are able to point in the same direction as the side chains in the original peptide.

D-amino acid forms of the artificial proteins, fragments or variants described above may also be used in the vaccines of the invention. In the case of D-amino acid forms, the order of the amino acid residues comprising the derivative is reversed as compared to those in the original peptide. The preparation of derivatives using D-amino acids rather than L-amino acids greatly decreases any unwanted breakdown of such an agent by normal metabolic processes, decreasing the amounts of agent which need to be administered, along with the frequency of its administration. Derivatives of AcrAB-TolC efflux pump proteins that may be used in accordance with the invention include artificial proteins containing conserved amino acid substitutions that retain the immunogenic activity of the original artificial proteins (as characterised by their ability to induce an immune reaction in a host suitable to prevent or reduce bacterial colonisation or infection in the host). It is preferred that conserved substitutions may be substitutions designed to remove protease cleavage sites, or other peptide structures that may be involved in the degradation or clearance of artificial proteins of the AcrAB-TolC efflux pump. Further details of suitable peptide variants and derivatives of that may be employed in the vaccines of the invention are provided below.

Suitable variant forms of artificial proteins of the AcrAB-TolC efflux pump may be ones in which certain of the native amino acids are replaced with amino acids having a side chain of similar biophysical properties to the amino acid it substitutes, to produce a conservative change. For example small non-polar, hydrophobic amino acids include glycine, alanine, leucine, isoleucine, valine, proline, and methionine. Large non-polar, hydrophobic amino acids include phenylalanine, tryptophan and tyrosine. The polar neutral amino acids include serine, threonine, cysteine, asparagine and glutamine. The positively charged (basic) amino acids include lysine, arginine and histidine. The negatively charged (acidic) amino acids include aspartic acid and glutamic acid.

Other modifications in protein sequences (such as those which occur during or after translation, e.g. by acetylation, amidation, carboxylation, phosphorylation, proteolytic cleavage or linkage to a ligand) may provide further variant forms of artificial proteins of the AcrAB-TolC efflux pump suitable for use in the vaccines of the invention.

Derivatives of artificial proteins of the AcrAB-TolC efflux pump suitable for use in the vaccines in accordance with the invention may include derivatives that increase or decrease the proteins' half-life in vivo. Examples of derivatives capable of increasing the half-life of artificial proteins of the AcrAB-TolC efflux pump include peptoid derivatives, D-amino acid derivatives and peptide-peptoid hybrids. It will be appreciated that, since artificial proteins of the AcrAB-TolC efflux pump or their fragments (as well as many of the possible variants and derivatives thereof) are proteins or may contain peptidyl components, they may be subject to degradation by a number of means (such as protease activity of hosts to whom vaccines based on such proteins are administered). Such degradation may limit the bioavailability of polypeptides of such vaccines, and hence the ability of the polypeptides to achieve their requisite immunogenic function. There are many well-established techniques by which peptide derivatives that have enhanced stability in biological contexts can be designed and produced. Such peptide derivatives may have improved immunogenicity as a result of ^• increased resistance to protease-mediated degradation. Preferably, a peptide derivative or analogue suitable for use in the vaccines of the invention is more protease-resistant than the AcrAB-TolC efflux pump protein from which it is derived.

Protease-resistance of a derivative of artificial proteins of the AcrAB-TolC efflux pump may be compared with protease-resistance of the artificial proteins themselves by means of well-known protein degradation assays. Suitable assays for the investigation of protease-resistance are described in the prior art.

In addition to the peptide vaccines described above, the invention further provides a vaccine comprising an isolated artificial protein of the AcrAB-TolC efflux pump, or a fragment, variant or derivative thereof. Isolated artificial proteins to be used in such vaccines may be purified by any means known from the prior art. Suitable proteins, fragments, variants or derivatives for use in accordance with this aspect of the invention may be as described with reference to the first aspect of the invention.

In a further aspect of the invention there is provided a method of treating or reducing a disease associated with bacterial colonisation or infection, the method comprising administering to a patient in need of such treatment or reduction an effective amount of a vaccine comprising an artificial protein of the AcrAB-TolC efflux pump, or a fragment, variant, or derivative thereof. The artificial protein, fragment, variant or derivative may be as described in any of the embodiments considered above. A patient in accordance with this aspect of the invention may be any human or non-human animal in need of treatment or reduction of disease.

According to a fourth aspect of the present invention there is provided a vaccine comprising a Gram-negative bacterium having defective efflux pump function. There is also provided the use of a Gram-negative bacterium having defective efflux pump function in the preparation of a medicament for the prevention or treatment of a disease caused by bacterial activity. It will be appreciated that the disease to be prevented or treated is preferably caused by activity of Gram-negative bacteria of the same species or serovar as those used in the medicament.

In a fifth aspect of the present invention there is provided a method of preventing or treating a disease caused by activity of Gram-negative bacteria, the method comprising administering to an individual in need of such treatment a therapeutically effective amount of a Gram-negative bacterium having defective efflux pump function.

The present invention is based on the inventors' finding that vaccines comprising Gram- negative bacteria having defective efflux pump function (referred to herein as "bacterial vaccines") are able to promote an effective immunising response capable of preventing or reducing subsequent bacterial infection and/or colonisation. Furthermore, the inventors have surprisingly found that bacteria having defective efflux pump function are not able to infect or colonise subjects to whom they are administered, or at least show much reduced ability to do so, and hence the bacteria are cleared from the host having provided a suitable immunising stimulus. Thus the invention provides bacteria that may be used as novel virulence-attenuated bacterial strains for use as vaccines.

Vaccines and medicaments in accordance with these aspects of the invention may comprise any suitable genus or species of Gram-negative bacteria. For example, suitable Gram-negative bacteria having defective efflux pump function may be selected from the group comprising Salmonella spp., Campylobacter spp., Escherichia coli, Klebsiella spp., Pseudomonas spp., Enterobacter spp., Serratia spp., Citrobacter spp., Proteus spp., Morganella spp., Acinetobacter spp., Stenotrophomonas spp., Haemophilus spp., Shigella spp., Yersinia spp., Vibrio cholerae, Pseudomonas aeruginosa, and Moraxella spp. Preferably Gram-negative bacteria for use in accordance with these aspects of the invention are selected from the group comprising Salmonella spp. and Campylobacter spp.

It will be appreciated that diseases caused by the activity of these listed bacteria also constitute a preferred group of dieses that may be prevented or treated using medicaments or methods in accordance with the present invention (and particularly using bacterial vaccines of the invention).

The inventors have found that vaccines comprising Gram-negative bacteria having defective efflux pump function provide a number of notable and surprising advantages over those vaccines that are available in the prior art.

For example, the inventors have found that vaccines comprising L 108 (a novel form of Salmonella enter ica Typhimurium SLl 344 in which the Acr AB-ToIC efflux pump is defective due to an inability to express functional ToIC) exhibit decreased persistence of the immunising bacteria in the host as compared to the persistence exhibited by existing vaccines such as Zoosaloral, vacT and χ3985. The decrease in bacterial persistence is statistically significant, and is discussed in greater detail in the Experimental Results section.

Furthermore, bacterial vaccines of the invention have been shown to decrease the level of colonisation and invasion that may be achieved by pathogenic bacteria (administered to hosts that have previously been immunised using bacterial vaccines of the invention), as compared to levels of colonisation and invasion that may be achieved after immunisation with known vaccines. For example, bacterial vaccines comprising Ll 08 were able to reduce colonisation and invasion of host tissues by pathogenic bacteria as compared to the results that may be achieved using the known vaccine SalenVacT, as discussed in more detail in the Experimental Results section. In a further advantage the inventors believe that bacterial vaccines in accordance with the present invention may be able to achieve an effective immunising response after only a single administration of the vaccine. Results using Ll 08 vaccines that support this belief are further described in the Experimental Results section.

None of these advantages have been reported, or even suggested, in the prior art. It will be appreciated that these advantages over the effects that may be achieved using currently available vaccines are of great benefit to those seeking means by which therapeutically effective immunisation against Gram-negative bacteria may be achieved.

Different Gram-negative bacteria (such as the illustrative genuses and species set out above) naturally express different efflux pump systems. For example, the AcrAB-TolC efflux pump system is expressed by Gram-negative bacteria such as Salmonella and Esherichia coli. This efflux pump system is well characterised, and accordingly bacteria having defective AcrAB-TolC efflux pump function represent preferred agents suitable for use in bacterial vaccines of the invention.

In addition to the AcrAB-TolC system, the inventors have found that bacteria with defective CmeABC efflux pump function may also be used in the preparation of vaccines of the invention. CmeABC is a tripartite efflux pump comprising CmeA, CmeB and CmeC, and is normally expressed by Campylobacter spp,

In one preferred embodiment of the invention a suitable bacterium may lack CmeA activity. The amino acid sequence of CmeA from Campylobacter jejuni is shown as Sequence ID No. 13, and a nucleotide encoding this protein as Sequence ID No. 14.

In a further preferred embodiment of the invention a suitable bacterium may lack CmeB activity. The amino acid sequence of CmeB from Campylobacter jejuni is shown as Sequence ID No. 15, and a nucleotide encoding this protein as Sequence ID No. 16. In another preferred embodiment of the invention a suitable bacterium may lack CmeC activity. The amino acid sequence of CmeC from Campylobacter jejuni is shown as Sequence ID No. 17, and a nucleotide encoding this protein as Sequence ID No. 18.

The inventors have found that other bacteria having defective efflux transporter activity investigated to date are capable of providing effective vaccines. For example, P. aeruginosa in which the activity of MexAB-OprM or MexCD-OprJ is defective may be suitable for use as vaccines. Preferably, suitable vaccines in accordance with the invention may comprise bacteria in which MexAB-OprM is defective. In a preferred embodiment suitable bacteria may lack MexA activity. The amino acid sequence of MexSA is shown as Sequence ID No. 19, and a nucleic acid encoding this protein shown in Sequence ID No. 20. In a further preferred embodiment suitable bacteria may lack MexB activity. The amino acid sequence of MexB is shown as Sequence ID No. 2I₅ and a nucleic acid encoding this protein shown in Sequence ID No. 22. In another preferred embodiment suitable bacteria may lack OprM activity. The amino acid sequence of OprM is shown as Sequence ID No. 23 and a nucleic acid encoding this protein shown in Sequence ID No. 24.

The inventors have also found that Stenotrophomonas maltophilia in which SmeDEF activity is defective may be used in accordance with the invention, as may Serratia marscecens in which the function of either SdeAB or SdeXY is defective; or Acenitobacter baumannii with defective AdeABC activity.

By defective efflux pump function is meant any decrease in the inherent efflux pump function of the bacterium. Bacterial efflux pump function may be readily assayed by means known to those skilled in the art. For example, the level of bacterial efflux pump function can be investigated by determining the effect of an efflux pump inhibitor on the susceptibility of a bacterial strain of interest to substrates including antibiotics. Such susceptibility may be analysed by minimum inhibitory concentration (MIC) testing of an antibiotic for test strains in the presence or absence of efflux pump inhibitor. A suitable example of an MIC assay is described in the Experimental Results section. Similarly the amount or concentration of antibiotic that accumulates within cells can be measured and the effect of efflux pump inhibitors thereby determined. In both cases the increase in susceptibility or antibiotic accumulation upon addition of the efflux pump inhibitor reveals the level of active efflux pump function for the test antibiotic and test strain. Further details of suitable methods are provided in publications by Andrews (2001) and Piddock et al. (1999). These methods may be readily adapted to allow assessment of inherent efflux pump function in bacteria of interest, and to compare the inherent efflux pump function of such bacteria of interest with activity of wild-type bacteria.

Preferably a bacterium suitable for use in accordance with the invention may, for example, have at least 25% less efflux pump function than comparable wild type bacteria, preferably, at least 50% less efflux pump function, more preferably at least 75% less efflux pump function, even more preferably at least 90% less efflux pump function, and most preferably 100% less efflux pump function.

Bacteria having defective efflux pump function suitable for use in accordance with the invention may include naturally occurring bacteria in which efflux pump function is reduced compared to wild type. The screening of bacterial populations using the methods described above may readily identify such bacteria having naturally defective efflux pump function.

Preferably bacteria having defective efflux pump function to be used in accordance with the invention are bacteria that have been engineered to have defective efflux pump function. For the purposes of the present invention engineering may be taken to encompass any artificial means by which bacteria having defective efflux pump function may be produced. Examples of suitable means will be readily apparent to those of skill in the art, and will include (by way of non-limiting example) bacteria having gene deletions and/or bacteria engineered to express appropriate anti-sense RNA. Preferably suitable engineered bacteria may be knockout bacteria.

By a knockout bacterium is meant a bacterium in which efflux pump function is ablated, by any suitable means. In other words, for the purposes of the present invention it is the function of the efflux pump that should be knocked out in order to render a bacterium suitable for use in accordance with the invention, irrespective of whether or not the bacterial efflux pump components (or the genes encoding such components) are present. Accordingly bacteria in which efflux pump components are present but have been rendered ineffective (for example by the addition of function-blocking moieties) may be considered to be knockout bacteria and to be suitable for use in accordance with the present invention.

Suitable means for the production of knockout bacteria may include knockouts constructed by allele replacement (for example using the method of Datsenko and Wanner) and transduced back into a wild-type background using P22 transduction. This technique relies on the construction of an artificial allele, having homology to the target gene, carrying a resistance cassette within the areas of homology that is inserted within the target gene rendering it inactive.

Preferred bacteria for use in the vaccines of the invention may be bacteria in which AcrAB-TolC activity is knocked out. Suitable bacteria may lack functional AcrA and/or functional AcrB and/or functional ToIC.

The AcrB knockout bacteria L643 and LI lO (AacrB) described in the Experimental Results section below represent preferred bacterium having defective efflux pump function that are suitable for use in the vaccines or medicaments of the invention. Suitable vaccines or medicaments in accordance with this embodiment of the invention may comprise L643 and/or Ll 10 either singly or in combination.

The ToIC knockout bacterium L108, L109, L141 and L142 and (AtolQ described in the Experimental Results section below constitute further preferred bacteria having defective efflux pump function suitable for use in the vaccines or medicaments of the invention. Vaccines or medicaments in accordance with this embodiment of the invention may preferably comprise Ll 08 and/or Ll 09, which may be provided either singly or in combination. The use of L 108 in the methods and medicaments (including vaccines) of the present invention is particularly preferred. In addition to the benefits outlined above (and described more fully in the Experimental Results section) the inventors have found that L108 does not spontaneously develop resistance to antibiotics (data not shown).

The Experimental Results section provides details of methods by which knockout bacteria having defective efflux pump function suitable for use in the vaccines of the invention may be produced. The skilled person will appreciate that the methods described in the Experimental Results section may be utilised not only in the preparation of bacteria with defective efflux pump function in which the AcrAB-TolC efflux pump is defective or inactive, but also in the production of bacteria in which the function of other efflux pump systems is impaired. Bacteria in which efflux pump function is rendered defective by the insertion of the aph gene (an aminoglycoside phosphotransferase gene that confers aminoglycoside resistance described more fully in Wright and Thompson 1999) into a bacterial gene necessary for efflux pump function represent preferred agents suitable for use hi accordance with the present invention.

As noted previously, examples of other preferred bacteria suitable for use in the vaccines of the invention may be bacteria in which CmeABC function is defective. Such bacteria may lack functional CmeA and/or functional CmeB and/or functional CmeC. The inventors have used Campylobacter allelic exchange to knock out CmeABC by introducing a suicide vector containing the CmeB gene carrying a kanamycin resistance cassette (produced in pGEMTEasy in E. colϊ) within the coding part of the gene. Accordingly, selection of kanamycin resistant colonies allowed knock out bacteria to be identified. Alternatively or additionally, the methods described in the Experimental Results section may be readily adapted to produce suitable bacteria exhibiting defective CmeABC function.

Bacteria suitable for use in accordance with the invention are preferably those that are associated with a deleterious condition or disease when present, either by infection or colonisation, in a host. Preferably, bacteria suitable for use in accordance with the present invention may be those associated with deleterious conditions of humans. In a preferred embodiment of the invention, bacteria of Salmonella spp. may be used in the vaccines of the invention (being considered bacteria of interest in that the presence of Salmonella bacteria may be associated with salmonellosis). Accordingly, vaccines comprising Salmonella bacteria having defective efflux pump function may be preferred for use in the prevention or reduction of diseases (such as salmonellosis) associated with infection or colonisation by Salmonella. Salmonella enterica bacteria having defective efflux pump function may be particularly preferred for use in accordance with the invention. Preferred Salmonella Enterica serovars suitable for use in bacterial vaccines of the invention include Salmonella Typhimurium, Salmonella Enteritidis and Salmonella Java.

In an alternative embodiment, bacteria of Campylobacter spp. having defective efflux pump function may be used in accordance with the invention. Campylobacter may be considered bacteria of interest in that Campylobacter infection or colonisation may lead to campylobacteriosis. This disease is characterised by symptoms of diarrhoea, cramping, abdominal pain and fever. It will be appreciated that vaccines comprising Campylobacter having defective efflux pump function may be preferred for use in the prevention or reduction of diseases (such as campylobacteriosis) associated with infection and/or colonisation by Campylobacter. Vaccines of the invention may particularly preferably comprise Campylobacter jejuni having defective efflux pump function.

It will be appreciated that vaccines in accordance with the invention should be administered so as to provide an effective immunising dose. For the purposes of the present invention an effective immunising dose should be considered to be a dose of a vaccine in accordance with the present invention sufficient to bring about the production of neutralising antibody and/or protection from bacterial infection, colonisation, or disease.

In the case of bacterial vaccines of the invention the vaccine should contain sufficient of the bacterium having defective efflux pump function to allow a subject to whom the vaccine is administered to mount an effective immunising response against native antigens expressed by the bacteria.

It is recognised that administration of an effective immunising dose may be achieved by way of a single administration (i.e. administration of a single dose of a vaccine, said dose constituting an effective immunising dose), or by way of multiple administration (i.e. administration of two or more doses of a vaccine, said two or more doses combining to constitute an effective immunising dose). The use of multiple administrations of vaccines (for example a primary dose followed by one or more booster doses) is well known, particularly in the context of live vaccines.

The inventors have found that bacterial vaccines in accordance with the present invention are capable of generating an effective immunising response after only a single administration of the bacterial vaccine. This is in contrast to the majority of vaccines know in the prior art, which require multiple administrations in order to achieve an effective immunising response. It will be appreciated that the ability of bacterial vaccines in accordance with the present invention to generate an effective immunising response after only a single administration confers significant advantages. For example, when effective immunisation may be achieved on administration of a single dose of a vaccine medicament, as compared to multiple doses, the cost associated with generating the immunising response is greatly decreased due to both the reduction in cost of medicaments administered, and also the reduction in the time and labour expended in repeated administration. Furthermore, the ability to generate an effective immunising dose through the use of a single administration of a vaccine medicament confers advantages in terms of the improved simplicity of the administration regime.

Accordingly, the present invention provides the use of a Gram-negative bacterium having defective efflux pump function, in the preparation of a medicament for the prevention of a disease caused by the activity of Gram-negative bacteria, wherein the administration pattern for the medicament consists of administering a single immunologically effective amount of the medicament comprising the Gram-negative bacterium. Preferably a medicament in accordance with this aspect of the invention may comprise Salmonella bacteria having defective efflux pump function. More preferably the Salmonella have defective AcrAB-TolC function, even more preferably the bacteria lack functional ToIC, and most preferably the bacteria comprise Ll 08.

It will be appreciated that derivatives of the bacteria disclosed herein as suitable for use in accordance with the invention will also themselves be suitable for use in accordance with the invention, as long as the derivatives retain the defective efflux pump function. The experimentally produced novel bacteria having defective efflux pump function described in the Experimental Results section of the present application are generally produced by the insertion of antibiotic resistance genes which serve to both disrupt efflux pump protein genes (and thereby cause defective efflux pump function) and also allow experimental selection (through the use of media containing antibiotics) of bacteria in which pump function has been disrupted. The skilled person will, however, readily appreciate that it may be preferred to use a bacterium that retains defective efflux pump function, but does not include a gene (or genes) conferring antibiotic resistance.

For example, in the case of novel bacteria produced in the manner described in the Experimental Results section (such as Ll 08 mentioned above), a preferred derivative may be one in which efflux pump genes are disrupted so as to impair efflux pump function, but where the introduced antibiotic resistance, gene is then removed (while retaining disruption of the efflux pump). The skilled person would be in no doubt as to how examples of such bacteria may be

Derivatives of Ll 08 represent preferred derivatives for use in accordance with the invention, and a particular preferred derivative of L 108 may lack ToIC, and may also lack the aph gene. Such a derivative may be accorded the name Salmonella enterica serovar Typhimurium AtolC, and this derivative represents a preferred bacterium having defective efflux pump function suitable for use in accordance with the present invention.

Preferably the medicament is for administration to agricultural animals, and more preferably is for administration to poultry subjects. It will be appreciated that, as far as possible, the medicament should be administered prior to exposure of the recipient to the Gram-negative bacterium responsible for causing the disease to be prevented. Preferably the medicament will be administered within the first five days of the recipient's life, more preferably within the first three days of the recipient's life, even more preferably within the first two days of the recipient's life, and most preferably during the first day of the recipient's life.

The inventors have found that when bacterial vaccines in accordance with the present invention are administered in the manner that is normally used for known vaccines (i.e. where the administration pattern for the medicament comprising administration of an initial primer dose, followed by a subsequent booster dose) the results that are achieved compare favourably with those that may be achieved using existing vaccines. The inventors have found that the administration pattern may preferably comprise administering the second ("booster") dose between ten and 20 days after the first administration of the medicament, more preferably between twelve and sixteen days after the first administration of the medicament, and most preferably fourteen days after the first administration of the medicament.

In keeping with the preceding description, the first administration of the medicament may preferably be within the first five days of the recipient's life, more preferably within the first three days of the recipient's life, even more preferably within the first two days of the recipient's life, and most preferably during the first day of the recipient's life.

In the light of the above, it will be appreciated that the invention provides a method of preventing a disease caused by bacterial activity, the method comprising administering a therapeutically effective amount of a vaccine or medicament of the invention to a patient in need of such prevention. The vaccine may be a peptide vaccine of the invention, or it may be a bacterial vaccine of the invention. A therapeutically effective amount of a vaccine or medicament of the invention will be an amount sufficient to produce an immunising response capable of preventing the disease in question. The administration of vaccines of the invention in accordance with the method described above may be in accordance with a regime established with reference to any of the considerations set out elsewhere in the specification. For example, the method may utilise multiple administrations of the vaccine or medicament. In particular^' the method may utilise a first administration (a "primer") followed by a second administration (a "booster"). The timing of the first and second administrations may be as set out in the passages above.

Alternatively the method of prevention outlined above may consist of only a single administration of the chosen vaccine or medicament of the invention. The timing of this single administration may be as set out in the paragraphs above.

Vaccines in accordance with the present invention may preferably be provided in the form of dosage units. For example, in the case where a vaccine is to be administered such that a single administration is sufficient to provide an effective immunising dose, such a single administration may constitute a suitable dosage unit. Alternatively, when multiple administrations of a vaccine are to be required in order to provide an effective immunising dose (for example a first "primer" and second "booster" administration) a suitable dosage unit may comprise sufficient vaccine to provide the necessary administrations.

The amount of an artificial peptide, fragment or derivative, or bacterium to be contained in a dose of a vaccine of the invention in order to provide an effective immunising dose may be determined with reference to a number of factors. Suitable factors may include the immunogenicity of the peptide, fragment or derivative, or bacterium contained in the vaccine; the route of administration by which the vaccine is administered; the relative size of the subject receiving the vaccine; and the extent of bacterial infection or colonisation from which the subject may suffer, either before or after administration of the vaccine.

In the case of vaccines of the invention containing artificial proteins, or fragments or derivatives thereof, it may be that an effective immunising dose may contain between 0.1 micrograms to 10,000 micrograms of the protein, fragment or derivative. Preferably an effective immunising dose may contain in the region of 1 microgram to approximately 1,000 micrograms of the protein, fragment or derivative, and more preferably may contain about 10 micrograms to about 500 micrograms.

In the case of bacterial vaccines in accordance with the invention an effective immunising dose may comprise between 10² and 10¹² colony forming units (CFUs), preferably between 10⁴ and 10¹⁰ CFUs, more preferably between 10⁶ and lO¹⁰ CFUs, and most preferably between 10⁷ and 10⁹ CFUs.

One method by which the suitability of artificial proteins (or their fragments or derivatives) or bacteria for use in accordance with the invention may be usefully investigated is by assessment of the ability of such agents to bind to antigen presenting cells (APCs). For example, a protein (or a fragment or derivative thereof) or bacterium intended for use in vaccines according to the present invention may be fluorescently labelled and incubated with suitable APCs. Examples of suitable APCs will include APCs derived from the species to which the vaccines are to be administered, and particularly APCs from tissues in which it is desired to induce immunity- Binding of proteins, fragments, derivatives or bacteria to the APCs, or the uptake of these agents by the APCs, may then be assessed. Preferred proteins (or their fragments or derivatives) or bacteria to be used in the vaccines of the invention may be those exhibiting increased uptake by, or binding to, the APCs.

It will be appreciated that the protocols outlined above may also be used to determine whether or not a variant, fragment or derivative of a protein that may be used in the vaccines of the invention exhibits greater APC binding than does the natural full-length artificial protein on which it is based. Such fragments or derivatives exhibiting increased APC binding may be selected as preferred agents compared to the full-length artificial protein. Furthermore, investigation of the binding characteristics of such derivatives or fragments may allow the design and production of derivatives or fragments having preferred APC binding characteristics. For example, such a procedure may involve the selection of fragments or derivatives having preferred APC binding characteristics; production of further fragments or derivatives ("second generation" fragments or derivatives) based on the selected derivatives; and selection of those second generation fragments or derivatives having improved binding characteristics. Such a procedure may be further repeated through third and subsequent generations of fragments or derivatives.

An analogous approach may be used in the production and selection of bacteria having preferred APC binding characteristics suitable for use in the vaccines of the present invention.

The inventors have found that vaccines in accordance with the present invention are particularly suitable for the prevention and/or treatment of bacterial infection and/or colonisation of poultry. Accordingly, the invention provides a live vaccine for the prevention and/or treatment of poultry infection and/or colonisation, comprising a live Gram-negative bacterium having defective efflux pump function. In a preferred embodiment, the invention provides a live vaccine for the prevention and/or treatment of poultry Salmonella spp. infection or colonisation comprising a live Salmonella spp. bacterium having defective efflux pump function. Bacteria suitable for use in accordance with this preferred embodiment of the invention may preferably have defective AcrAB- ToIC activity. Accordingly such preferred vaccines may make use of AcrA knock out bacteria; AcrB knock out bacteria; or ToIC knock out bacteria. The ToIC knock out bacteria L 108 represent particularly preferred bacteria having defective efflux pump function for use in connection with this aspect of the invention.

In an alternative preferred embodiment vaccines in accordance with the invention may be used in the prevention and/or treatment of Campylobacter spp. infection or colonisation. In accordance with this preferred embodiment, the invention provides a live vaccine for the prevention and/or treatment of poultry Campylobacter spp. infection or colonisation comprising a live Campylobacter spp. bacterium having defective efflux pump function. Preferred bacteria suitable for use in accordance with this aspect of the invention include bacteria having defective CmeABC activity. Thus it will be appreciated that suitable bacteria for use in such a live vaccine may include CmeA knock out bacteria; CmeB knock out bacteria; or CmeC knock out bacteria. In a still further preferred embodiment vaccines in accordance with the invention may be used in the prevention and/or treatment of E. coli infection or colonisation. In accordance with this embodiment of the invention there is provided a live vaccine for preventing and/or treating poultry E. coli. infection or colonisation comprising a live E. coli bacterium having defective efflux pump function. Preferred bacteria to be used in this manner include those having defective AcrAB-TolC function, and particularly ToIC knockout bacteria. In particular it is preferred that such bacteria may comprise Escherichia coli O78:K80 1117. Vaccines comprising such bacteria may be used in the prevention and/or treatment of diseases such as collibaccilosis.

Suitable bacteria that may be used in such live vaccines may be produced and selected using the protocols described elsewhere in the specification.

It is generally recognised that vaccines using isolated proteins (and particularly artificial proteins such as those produced by recombinant technology) may have the disadvantage of reduced immunogenicity as compared to more traditional "complex" vaccines. Accordingly, it may be preferred that vaccines in accordance with the present invention (and particularly in accordance with the 1st or 2nd aspects of the invention) may further comprise suitable adjuvants capable of increasing and prolonging the specific immune response to antigens in the vaccine. It is particularly preferred that vaccines in accordance with the present invention may further comprise known adjuvants that are capable of intensifying T-cell proliferation and the cellular immune response.

Examples of suitable adjuvants that may be used include inorganic adjuvants in gel form (aluminium hydroxide/aluminium phosphate, calcium phosphate); bacteria-derived adjuvants such as monophosphoryl lipid A and muramyl peptides; particulate adjuvants including immunestimulatory complexes, liposomes and biodegradable microspheres; adjuvants based on oil emulsions and emulsifiers, such as Freund's adjuvant or Incomplete Freund's adjuvant (IFA); saponines such as QS-21; squalene; synthetic adjuvants such as non-ionic block copolymers, muramyl peptide analogues, synthetic lipid A, synthetic polynucleotides and polycationic adjuvants such as polyarginine or polylysine. Vaccines in accordance with the present invention may use suitable buffering agents in order to produce vaccines that are isotonic with respect to the subject to whom the vaccines are to be administered. Suitable agents that may be used in ensuring that vaccines in accordance with the present invention are isotonic may include sugars, sugar alcohols, oligosaccharides, polysaccharides, polyhydric alcohols, amino acids or lipids.

Vaccines in accordance with the present invention may be buffered to an appropriate pH level. Suitable buffers known to those skilled in the art include phosphate-buffered saline (PBS) and HEPES-buffered saline (HBS).

Vaccines in accordance with the present invention, whether peptide vaccines or bacterial vaccines, may be administered to a subject by any suitable route of administration known in the prior art.

It may generally be preferred that vaccines in accordance with the present invention be administered orally. Orally administered vaccines are generally safer for patients or subjects receiving the vaccine, and the administration of such vaccines requires little by way of specialised training. Both physicians and patients may typically favour the use of oral vaccines since such vaccines are often cheaper and easier to deliver than injectable vaccines. In particular, the administration of oral vaccines does not generally require the use of sterilising equipment, and further involves no pain on administration of such vaccines to a patient or subject.

The use of orally administered vaccines is particularly preferred in the case that the vaccines are for the prevention and/or treatment of bacterial colonisation and/or infection occurring the in the digestive tract. Orally administered vaccines are also more readily able to give rise to beneficial "intestinal" immunity. Intestinal immunity may be preferred in cases where deleterious bacteria may otherwise infect or colonise the intestinal tract of a host animal. When vaccines in accordance with the present invention are to be administered by the oral route they may make use of any suitable formulation known to those skilled in the art. Suitable formulation may for example include an appropriate adjuvant, such as alum, or an immunommodulator, such as LT toxoid. Furthermore, it will be appreciated that the vaccines of the invention may be prepared for long term storage and may include appropriate protectants/excipients such as glycerol or sugars. Suitable vaccine formulation may make use of an isotonic solution such as phosphate buffered saline or saline.

Vaccines for oral administration may preferably be formulated such that the immunity- generating components within the vaccine (whether proteins, their fragments or derivatives, or live or dead bacteria) are protected from the harsh conditions of the digestive tract, which may otherwise give rise to degradation of the antigens administered before immunity may be conferred. Suitable methods by which such protection may be conferred are well known to those skilled in the art and include the use of methods considered above, as well as encapsulation, microencapsulation, and/or enteric coatings.

Oral administration of vaccines of the invention may be achieved by inoculation (such as by oral gavage) or by application of the vaccines in drinking water. Application in drinking water may represent a preferred route of oral administration since the vaccines may be administered to a relatively large number of subjects without extensive intervention by a person administering the vaccines.

As an alternative to their oral administration, suitably formulated vaccines may be administered to a subject by means of injection. In particular, vaccines in accordance with the present invention may be administered by intramuscular injection, intradermal injection subcutaneous injection, or intravenous injection. Formulations for use in the preparation of injectable vaccines are well known to those of skill in the art.

Vaccines in accordance with the present invention may also be administered by inhalation, for example via intranasal spray. It is well known to provide vaccines by nasal inhalation and such administration may be preferred since it lacks many of the undesirable effects associated with vaccination by injection (such as injection pain and the requirement for sterilising equipment). Suitable nasal spray formulations which may be used in the preparation of vaccines in accordance with the present invention will be known to those skilled in the art.

It has recently been shown that effective immunising dosages of vaccines may be administered to poultry through the use of whole body sprays. Surprisingly aerosol immunisation in this manner has been found to be suitable for the generation of a systemic immune response, not just a response associated with the respiratory tract. Accordingly, the use of such whole body sprays represents a preferred route of administration, particularly in the case of administration of vaccines in accordance with the present invention to poultry. ,

The use of routes of administration other than oral administration may be preferred in the case where it is desired to administer the vaccines of the invention to young poultry. Newly hatched chicks are known not to eat or drink for up to several days after their hatching, and so the use of injection, inhalation, or whole body spray may be particularly preferred in the case where it is desired to provide the vaccine to such relatively young poultry. Other methods of immunisation suitable for administration of vaccines of the present invention to poultry include administration by means of eye drop or injection.

For the purposes of the present specification by "agent" or "agent of the invention" is meant an artificial protein of the AcrAB-TolC efflux pump (or a fragment or derivative thereof) or a bacterium with defective efflux pump function suitable for use in the vaccines, methods or medicaments of the invention.

In order to be suitable for use in the vaccines of the invention an artificial protein of the AcrAB-TolC efflux pump (or a fragment or derivative thereof) or a bacterium having defective efflux pump function must retain the capability to induce an effective immunogenic response to Gram-negative bacteria that may otherwise infect or colonize a subject. The ability of potentially useful proteins or bacteria to achieve such an effect may be readily determined with reference to suitable in vitro or in vivo assays. Such assays will be well known to those skilled in the art and examples of assays that may be used are described in the Experimental Results section below.

It will be appreciated that the amount of a vaccine of the invention that must be administered to achieve an effective immunising dose depends on a number of factors including the biological activity and bioavailability of the agent present in the vaccine, which in turn depends, among other factors, on the nature of the agent and the mode of administration of the vaccine. Other factors in determining an effective immunising amount of a vaccine of the invention may include:

A) The half-life of the agent in the subject being treated.

B) The specific condition to be treated (e.g. immunisation to prevent bacterial colonisation or infection).

C) The age of the subject.

The frequency of administration will also be influenced by the above-mentioned factors and particularly the half-life of the chosen agent within the subject being treated.

Generally when vaccines in accordance with the invention are used to prevent bacterial colonisation or infection the vaccine should be administered as soon as the risk of bacterial colonisation or infection has been identified.

Frequency of administration will depend upon on, among other considerations, the biological half-life of the agent used. Typically a vaccine of the invention should be administered such that the concentration of the agent at a tissue at risk of bacterial colonisation or infection is maintained at a level sufficient to achieve an effective immunising dose.

Vaccines of the invention, may be administered by any suitable route capable of achieving the desired effect of preventing or reducing bacterial colonisation or infection, but it is preferred that the vaccines be administered orally, or by injection or inhalation. The inventors have found that the prevention or reduction of bacterial colonisation or infection niay be effected by the administration of a vaccine of the invention by injection. For instance, vaccines of the invention may be administered by means of intradermal intravenous or subcutaneous injection. Thus a preferred vaccine in accordance with the invention comprises an injectable solution of an agent of the invention. Suitable formulations for use in this embodiment of the invention are considered below.

Vaccines of the invention may take a number of different forms depending, in particular on the manner in which they are to be administered. Thus, for example, they may be in the form of a liquid, ointment, cream, gel, hydrogel, powder or aerosol. All of such compositions are suitable for topical application to a subject, which may be a preferred means of administering vaccines of the invention, particularly in the case of non-human subjects.

In one embodiment a pharmaceutical vehicle used in a vaccine of the invention may be a liquid and a suitable pharmaceutical composition would be in the form of a solution, hi another embodiment, the pharmaceutically acceptable vehicle is a solid and a suitable composition is hi the form of a powder or tablet.

A solid vehicle can include one or more substances which may also act as flavouring agents, lubricants, solubilizers, suspending agents, fillers, glidants, compression aids, binders or tablet-disintegrating agents; it can also be an encapsulating material. In powders, the vehicle is a finely divided solid which is in admixture with the finely divided agent of the invention, hi tablets, the agent of the invention is mixed with a vehicle having the necessary compression properties in suitable proportions and compacted in the shape and size desired. The powders and tablets preferably contain up to 99% of the agent of the invention. Suitable solid vehicles include, for example, calcium phosphate, magnesium stearate, talc, sugars, lactose, dextrin, starch, gelatin, cellulose, polyvinylpyrrolidine, low melting waxes and ion exchange resins.

Liquid vehicles may be used in preparing solutions, suspensions, emulsions, syrups, elixirs and pressurized compositions. The agent of the invention can be dissolved or suspended in a pharmaceutically acceptable liquid vehicle such as water, an organic solvent, a mixture of both or pharmaceutically acceptable oils or fats. The liquid vehicle can contain other suitable pharmaceutical additives such as solubilizers, emulsifϊers, buffers, preservatives, sweeteners, flavoring agents, suspending agents, thickening agents, colors, viscosity regulators, stabilizers or osmo-regulators. Suitable examples of liquid vehicles for oral and parenteral administration include water (partially containing additives as above, e.g. cellulose derivatives, preferably sodium carboxymethyl cellulose solution), alcohols (including monohydric alcohols and polyhydric alcohols, e.g. glycols) and their derivatives, and oils (e.g. fractionated coconut oil and arachis oil). For parenteral administration, the vehicle can also be an oily ester such as ethyl oleate and isopropyl myristate. Sterile liquid vehicles are useful in sterile liquid form compositions for parenteral administration. The liquid vehicle for pressurized compositions can be halogenated hydrocarbon or other pharmaceutically acceptable propellant.

Liquid pharmaceutical compositions which are sterile solutions or suspensions can be utilized by for example, intramuscular, intrathecal, epidural, intraperitoneal, intradermal or subcutaneous injection. Sterile solutions can also be administered intravenously. The agent of the invention may be prepared as a sterile solid composition which may be dissolved or suspended at the time of administration using sterile water, saline, or other appropriate sterile injectable medium. Vehicles are intended to include necessary and inert binders, suspending agents, lubricants and preservatives.

In the situation in which it is desired to administer an agent of the invention by means of oral ingestion, it will be appreciated that the chosen agent will preferably be an agent having an elevated degree of resistance to degradation. For example, the agent of the invention may be protected (for instance using the techniques described above) so that its rate of degradation in the digestive tract is reduced.

Compositions of agents of the invention are suitable to be administered to the eye, in which case a vaccine of the invention may be formulated as an eye drop. Known procedures, such as those conventionally employed by the pharmaceutical industry (e.g. in vivo experimentation, clinical trials etc), may be used to establish specific formulations of compositions comprising agents of the invention and precise therapeutic regimes for administration of such compositions (such as effective immunising doses of the active agent and suitable frequency of administration).

The optimal concentration of peptide or bacterium to be used in a particular medicament will be determined by a range of factors, including the nature of the medicament, the route of administration, and the tissue in which immunisation is to be achieved. The ways in which preferred concentrations may be calculated based on such factors are conventional, and will be well known to those skilled in the art.

Agents of the invention may be used to prevent or reduce bacterial infection or colonisation as a monotherapy (e.g. through use of vaccines of the invention alone). Alternatively the methods or medicaments of the invention may be used in combination with other compounds or treatments to prevent or reduce bacterial infection or colonisation. Suitable treatments that may be used as parts of such combination therapies will be well known to those skilled in the art, and may include the use of applicable antibiotic and antimicrobicidal agents.

Preferably a vaccine in accordance with the invention may be a water-soluble pharmaceutical composition. A suitable vaccine may be a physiological solution which, for the purposes of the present invention, may be taken to be a solution having physiological concentration or osmolarity. For example, a soluble vaccine in accordance with the invention suitable for injection in mammals may preferably formulated to achieve an osmolarity of approximately 290 milliosmolar

One of the factors determining whether or not a composition is considered suitable for safe injection into mammals such as humans is the concentration of sodium ions present in the composition at the time of administration. Accordingly, injectable vaccines of the invention may preferably be formulated such that they contain a sodium concentration of between 130 and 160 mEq/L. Preferably, the quantities of the source of sodium ions and of the sugar are such that on dissolution of the composition to produce a sodium concentration of between 130 and 160 mEq/L the concentration of the sugar is greater than 50mg/ml. Sources of sodium ions suitable for use in injectable pharmaceutical compositions will be well known to those of skill hi the art.

As described above, vaccines hi accordance with the invention may be formulated for intradermal injection, or may be formulated for subcutaneous injection. Typical formulations suitable for intradermal injection and/or subcutaneous injection will be well known to those skilled in the art.

Vaccines in accordance with the present invention may preferably comprise aqueous solutions. Suitable compositions may be liquid compositions, or alternatively may further comprise thickening agents such that a gel or semi-solid composition is formed. Suitable thickening agents are well known to those skilled in the art, and include methylcellulose. The use of thickening agents in vaccines of the invention may be advantageous in a number of contexts. Particularly, the use of thickening agents in compositions of the invention may be advantageous in ensuring that the composition is retained in the site at which it is administered and at which it is desired for the composition to have its biological and/or therapeutic effect. It will be appreciated that thixotropic formulations of vaccines of the invention may be beneficial in a number of applications, for example in the use of nasal sprays or other means (such as whole body sprays) by which vaccines may be administered through inhalation.

The vaccines of the invention may preferably be provided in pre-fϊlled vessels containing the composition. Such pre-filled vessels provide advantages hi terms of their ability to readily deliver vaccines, and particularly pre-sterilised vaccines, to the location at which the vaccine is to be employed. Suitable vessels may be selected with reference to the chosen formulation and the method or route by which the vaccine is to be administered, and will typically include vessels such as vials or syringes.

Vaccines in accordance with the invention may be provided in readily stored forms, such as in sealed vessels. Such readily stored forms, and/or sealed vessels, may advantageously be sterilised or steiilisable. Suitable sealed vessels may be selected with reference to the manner in which the vaccine contained therein is to be used. By way of example, in the case of vaccines to be administered by injection, suitable sealed vessels may include vials, ampoules, prefilled syringes, or the like.

Vaccines of the invention may be administered in the form of a liquid or suspension. In the event that vaccines in accordance with the invention are stored in a dried form vaccines may then be dissolved or suspended in a suitable diluent, such as distilled water prior to their use.

It will be appreciated that vaccines in accordance with the present invention may be used in the production of animal-based foodstuffs, since the vaccines are able to prevent and/or treat bacterial colonisation and/or infection that may otherwise give rise to the harmful presence of bacteria in foodstuffs.

Accordingly, in a further aspect, the invention provides a method of preparing meat for human consumption, the method comprising: i) administering to an animal intended for meat production an amount of a vaccine in accordance with the invention sufficient to render the animal substantially free of a Gram- negative bacterium harmful to humans; ii) growing the animal to a size suitable for harvesting for meat production; and iii) harvesting the animal for meat production.

In a still further aspect, the invention provides a method of preparing food for human consumption, the method comprising: i) administering to an animal intended for food production an amount of a vaccine in accordance with the invention sufficient to render the animal substantially free of a Gram- negative bacterium harmful to humans; ii) growing the animal to a size suitable for harvesting food from the animal; and iii) harvesting food from the animal for human consumption. Animal-based foodstuffs in the context of the present invention may be taken to encompass both the meat of animals and other foodstuffs produced by animals, such as eggs, milk or the like. The term animal-based foodstuffs should also be taken to encompass products derived from foods produce by animals, such as dairy produce derived from milk.

These methods may be used for any suitable food- and/or meat-producing animal, including, but not limited to, cattle, sheep, pigs, poultry and farmed fish.

It is known that vaccination of food animals by injection may be associated with the development of lesions in the carcasses of animals so injected. Therefore it may be preferred that vaccines administered in accordance with the methods of the preceding aspects of the invention be administered by a needle-free route of administration.

In a still further aspect of the invention there is provided a method of rendering a subject substantially free of a deleterious bacterium, the method comprising administering to the subject an effective amount of a vaccine in accordance with the present invention. It should be recognised that in the context of this aspect of the invention a deleterious bacterium is a bacterium capable of having deleterious effects on humans, particularly through infection or colonisation of human hosts. The deleterious bacterium need not have a deleterious effect on the subject to whom the vaccine is administered.

In the context of the present invention "substantially free" should be construed as meaning that the subject or animal (if food producing, at slaughter) will be sufficiently free of the deleterious bacterium that the amount of the deleterious bacterium present in the subject or animal will not give rise to ill effects in a human (for instance on consumption of a meat or food derived from a host animal treated in accordance with these aspects).

In a further aspect the invention provides a nucleic acid construct suitable for use in the production of a bacterium having defective efflux pump activity, wherein the construct comprises one or more of the primers set out in accompanying Figure 4. The present invention will now be further described by way of illustration with reference to the accompanying Experimental Results section, Figures and Tables in which:

Figure 1 shows the results (reported in both Studies A and B) investigating adhesion to and invasion of human gut epithelial cells (INT-407) by naturally occurring control Salmonella and mutant strains. The results of the adhesion assay are shown in Figure IA, and the results of the invasion assay in Figure IB.

Figure 2 shows the results of studies investigating numbers of Salmonella enterica shed by experimental poultry populations. Panel A shows total S. enterica numbers shed by birds of Group A; Panel B shows numbers of the ToIC knockout L 108 shed by birds of Group A; Panel C shows total S. enterica numbers shed by birds of Group B; and Panel D show numbers of the AcrB knockout L643 shed by birds of Group B.

Figure 3 illustrates the amino acid and nucleic acid sequences of a number of bacterial efflux pump proteins relevant in the context of the vaccines of the invention (either through their provision in protein vaccines, or their defective function in bacterial vaccines).

Figure 4 sets out sequences of PCR primers used in the generation of mutant bacteria, as described in the Experimental Results section.

Figure 5 illustrates S. enterica serovar Typhimurium L 108 vaccine shedding from vaccinated birds of Study C over three weeks prior to challenge with L696. Black arrow indicates when the booster vaccine was given. Broken black line indicates the limit of detection (3.I x IO² cfu g^"1 faeces).

Figure 6 illustrates gastrointestinal shedding of S. enterica serovar Typhimurium L696 from unvaccinated and vaccinated birds of Study C. Squares and triangles indicate the total S. enterica serovar Typhimurium count for individual birds from unvaccinated and vaccinated birds, respectively. Solid and dotted bars represent the median values for unvaccinated and vaccinated birds, respectively. Black arrow indicates when the challenge strain was administered. Broken line indicates the limit of detection (3.1 x 10 cfu g^"1 faeces).

Figure 7 illustrates serum IgM whole cell antigen (black), IgG whole cell antigen (grey) and IgM LPS antigen (dashed) antibody responses from un-vaccinated (open square) and vaccinated (filled squares) birds over 45 days, as described in Study C.

Figure 8 illustrates results reported in Study D. S. Typhimurium SL1344 (solid colour) and L 108 (tolCv.aph) (diagonal hatched colour) shedding from faecal pellets collected over five days from mice inoculated with 2.3 x 10⁶ CFU/ml L354 and 1.8 x 10⁶ CFU/ml L108.

Figure 9 illustrates results reported in Study E. Growth kinetics of E. coli O78:K80 (1115) and tolC disrupted mutant (1117). Growth was measured by absorbance at όOOnm at 37°C. Data are expressed as mean absorbance (n = 6).

Figure 10 illustrates results reported in Study E, showing association (top panel - A) and invasion (lower panel - B) of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants respectively, to a human gut epithelial cell line (INT 407). Data are presented on a logarithmic scale and are expressed as mean CFU/ml ± S. E. (Standard error of the mean).

Figure 1 T illustrates results reported in Study E, showing association (top panel - A) and invasion (lower panel - B) of S. Typhimurium SLl 344 and E. coli O78:K80 and tolC disrupted mutants respectively, to a mouse macrophage cell line (RAW 264.7). Data are presented on a logarithmic scale and are expressed as mean CFU/ml ± S. E. (Standard error of the mean).

Figure 12 illustrates results reported in Study E, showing Association (top panel - A) and invasion (lower panel - B) of S. Typhimurium SLl 344 and E. coli O78:K80 and tolC disrupted mutants respectively, to a avian macrophage cell line (HDI l). Data are presented on a logarithmic scale and are expressed as mean CFU/ml ± S. E. (Standard error of the mean).

Table 1 sets out oligonucleotide primers used in the studies.

Table 2 shows minimum inhibitory concentrations of naturally occurring control Salmonella and mutants in response to antibiotics, dyes and detergents.

Table 3 compares gastrointestinal shedding of naturally occurring control Salmonella and mutant strains.

Table 4 shows the results of investigations (reported in Study B) to establish the minimum inhibitory concentrations of antibiotics, dyes and detergents for wild type bacteria and bacteria engineered to have defective efflux pump function.

Table 5 shows the results of investigations reported in Study C, and illustrates the recovery of S. Typhimurium Ll 08 vaccine strain from tissues taken from birds dosed with vaccine at one day of age and again at 15 days of age.

Table 6 shows the results of investigations reported in Study C, and illustrates the recovery of S. enterica serovar Typhimurium L696 challenge strain from tissues taken from un- vaccinated and vaccinated birds.

Table 7 shows the results of investigations reported in Study C, and illustrates the number of S. Typhimurium in faecal pellets of experimental mice from this Study.

Table 8 shows the results of investigations reported in Study D, and illustrates the recovery of S. Typhimurium SL1344 and L108 vaccine strain from livers taken from mice dosed at five weeks of age.

Table 9 illustrates results reported in Study E, and compares susceptibility of Escherichia coli O78-.K80 in which tolC disrupted (1117) with the susceptibility of its parent strain, (1115). Table 10 compares respectively the association of S. Typhimurium SLl 344 and E. coli O78:K80 and tolC disrupted mutants of these bacteria, to a human gut epithelial cell line (INT 407).

Table 11 compares invasion of S. Typhimurium SLl 344 and E. coli O78:K80 and tolC disrupted mutants, respectively, in a human gut epithelial cell line (INT 407).

Table 12 compares association of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a mouse macrophage cell line (RAW264.7).

Table 13 compares invasion of S. Typhimurium SLl 344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a mouse macrophage cell line (RAW264.7).

Table 14 compares association of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to an avian macrophage cell line (HDl 1).

Table 15 compares invasion of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to an avian macrophage cell line (HDl 1).

EXPERIMENTALRESULTS.

STUDYA

1. Materials and Methods

1.1. Bacterial strains.

Salmonella enterica serovar Typhimurium SLl 344 was used as a control throughout the following experimental procedures. This strain bacterial strain is well studied and is considered representative of this serovar (Wray and Sojka, 1978).

1.2. Construction of mutants.

The acrD and tolC genes were inactivated in S. Typhimurium SL 1344 using the one-step technique previously described for E. coli (Datsenko and Wanner, 2000) and S. Typhimurium (Eaves et al., 2004). The disclosures of both of these publications are incorporated herein by reference, in particular the descriptions provided by these documents of methods for the production of mutant bacteria.

The method utilised provides a simple and highly efficient method to disrupt chromosomal genes in bacteria such as Salmonella and Escherichia coli. In this method PCR primers provide the homology to the targeted gene(s). Recombination requires the phage A Red recombinase, which is synthesized under the control of an inducible promoter on an easily curable, low copy number plasmid.

Suitable PCR products may be generated by using primers with 36- to 50-nt extensions that are homologous to regions adjacent to the gene to be inactivated and template plasmids carrying antibiotic resistance genes that are flanked by FRT (FLP recognition target) sites. By using the respective PCR products, it is readily possible to produce disruptions of chromosomal genes of interest. Mutants in which the gene of interest has been disrupted are isolated as antibiotic-resistant colonies after the introduction into bacteria carrying a Red expression plasmid of synthetic (PCR-generated) DNA. The resistance genes may then be eliminated by using a helper plasmid encoding the FLP recombinase which is also easily curable. Briefly, in the present case Ampicillin- and kanamycin-resistant transformants were selected on Lennox agar containing 80 μg of ampicillin per ml or 25 μg of kanamycin per ml, respectively. A 1.7-kb PCR product was amplified from template pKD4 by using primers acrBplpKD4 and acrBp2pKD4. The linear PCR product was recombined into the chromosome of SL1344 by the use of helper plasmid pKD46. A 1.8-kb PCR product was amplified from the resulting kanamycin-resistant intermediate strain by using primers usacrB and dsacrB. Primer pairs usacrB-kl and dsacrB-k2 amplified products of 880 and 1.2 kb, respectively. A PCR product with a size (3.7 kb) identical to that amplified from SL 1344 was amplified from the resulting acrB mutant, mutant L644, with primer pair usacrF-dsacrF. The kanamycin resistance cassette was excised from the chromosome with helper plasmid CP20, which encodes FLP recombinase under the control of a temperature-sensitive promoter. Afterwards, the PCR amplimers amplified from L644 with primers usacrB and dsacrB were reduced from 1.8 kb to 600 bp, as predicted for the loss of the 1.2-kb kanamycin resistance cassette from the deleted acrB site. No products were obtained with primer pairs containing either kl or k2, also consistent with the loss of the kanamycin resistance gene from the chromosome.

Sequences of all primers used are shown in accompanying Figure 4 of the specification.

Although the above technique is described with reference to the inactivation of components of the AcrAB-TolC bacterial efflux pump it will be appreciated that it may be modified, in manners readily apparent to those of skill in the art to, achieve the inactivation of components of any E. coli or Salmonella efflux pump of interest, and indeed is applicable across a range of Gram-negative bacterial types.

L356 (AacrF) and L643 (AacrB) were also produced according to above method, as previously described (Eaves et al., 2004).

Primers used for the inactivation of the genes encoding the AcrD and ToI-C proteins are shown in Table 1. Insertion of the aph gene in the correct location was confirmed by PCR and sequencing using further primers also identified in Table 1. For further verification, AacrB (L643) and AtolC (L108) were transduced back into SL1344 via P22 transduction to give LIlO and Ll 09 respectively, and the phenotype confirmed.

1.3. Determination of MIC of antibiotics, dyes, detergents and disinfectants.

The minimum inhibitory concentration (MIC) values of various antimicrobial agents in respect of the control and mutant strains were determined using a previously described doubling agar dilution method (Andrews, 2001). All MIC values were determined on at least three independent MIC occasions.

1.4. Adhesion and invasion/intracellular survival assays.

The adhesion and invasion/intracellular survival properties of control and mutant bacterial strains were investigated using previously described tissue culture assays (Dibb-Fuller et al., 1999). Briefly, tissue culture cells were maintained in Dulbecco's modified essential media (DMEM; Sigma) supplemented with 10% heat-inactivated foetal calf serum (Sigma), 2mM L-glutamine (Sigma) and 50μg/ml gentamicin (Sigma).

Confluent monolayers of human embryonic intestine cells (INT-407) and mouse monocyte macrophage (RAW 264.7) were prepared in 24- well tissue culture plates (Nunc) and pre- washed in Hank's balanced salt solution (HBSS; Sigma) prior to adhesion and invasion/intracellular survival assays. Cultures of bacteria were grown in LB broth for 18h at 37⁰C and were washed and diluted in DMEM prior to the adhesion and invasion assays.

For the adhesion assay, each cell line was infected with approximately 5 x 107 CFU/ml of pre-washed bacteria and incubated for 2h at 37⁰C (5% CO2). Monolayers were washed six times with HBSS before disrupting with 1% (v/v) Triton (Sigma). Colony forming units (CFU) was determined by serial dilution plated onto LB agar. Adhesion values are calculated as the number of adhered bacteria minus the number of invaded bacteria.

Invasion/intracellular survival assays were performed in duplicate 24-well plates. After allowing the bacteria to adhere to the cells of the monolayer, wells were washed three times before adding LB broth containing lOOμg/ml gentamicin to kill all external bacteria. Plates were incubated for 2h at 37°C (5% CO2) and washed twice. The monolayers were then disrupted with 1% (v/v) Triton and CFU were determined as described above.

Each adhesion and invasion/intracellular survival assay was performed at least three times (four wells per test), and the data analysed with the Student's two-tailed t-test. P values < 0.05 were taken as significant.

1.5. Colonisation and persistence of Salmonella enterica in vivo.

All animal procedures were performed under the United Kingdom Animals Scientific Procedures Act (1986).

Semi-quantitative competitive exclusion experiments were performed using specific pathogen-free white Leghorn chickens in accordance with previously described protocols (La Ragione and Woodward, 2003). Seven groups (A - G), each containing at least 11 birds, were housed in separate isolators and given feed and water ad libitum.

Groups A to D of the birds were inoculated when 24hrs old by oral gavage with 0.1ml of a mixture of control and mutant bacteria (1:1 ratio) suspended in PBS. The total concentration of bacteria administered was approximately 105 CFU/ml. Group A were inoculated with a mixture of control and Ll 08 (AtolC) bacteria; Group B with a mixture of control and L643 (AacrB) bacteria; Group C with a mixture of control and Ll 06 (AacrD) bacteria; and Group D with a mixture of control and L356 (AacrF) bacteria.

In contrast, Groups E to G of the birds were inoculated at 14 days old by oral gavage with 0.1ml of a mixture of control and mutant bacteria (1:1 ratio) suspended in PBS. In these cases the total concentration of bacteria administered was approximately 107 CFU/ml. Group E were inoculated with a mixture of control and L643 (AacrB) bacteria; Group F with a mixture of control and Ll 06 (AacrD) bacteria; and Group G with a mixture of control and L356 (AacrF) bacteria.

The intestinal bacterial populations of each group of birds were investigated by cloacal swabbing. Swabbing was conducted twice a week for at least five weeks. The swabs were aseptically transferred to ImI sterile PBS and the CFU determined by serial dilution onto (i) Brilliant Green Agar (BGA; Oxoid) plates, (ii) BGA supplemented with 8mg/ml kanamycin and (iii) LB (Oxoid) plates supplemented with 8μg/ml kanamycin. Plating onto the Salmonella selective agar BGA gave the total Salmonella count, whereas plating onto BGA supplemented with kanamycin grew only the mutant Salmonella containing the aph gene. Two mutants, AacrB and AtolC, were susceptible to the dye in the BGA₃ and so experiments with these strains used LB (Oxoid) plates supplemented with 8μg/ml kanamycin. Colonies growing on LB agar were confirmed as S. Typhimurium by agglutination test using 04 antisera. The remaining PBS, including the swab, was transferred to 10ml selenite broth and incubated at 37⁰C for seven days. If no colonies were observed on any of the plates (i) - (iii) (above) then the selenite broth was streaked onto BGA and/or LB agar supplemented with kanamycin. Data are expressed as CFU per swab.

2. Results.

2.1. Antimicrobial susceptibility.

The results of MIC experiments to investigate the antimicrobial susceptibility of the parent and mutant strains are shown in Table 2. These results show that the control parental strain S. Typhimurium SL 1344 was sensitive to all the agents tested. Mutant strains L356 (AacrF) and L 106 (AacrD) exhibited comparable susceptibility to all agents to that shown by the control parent strain.

In contrast L643 (AacrB) and L108 (AtolC) and SL1344 transductants (LI lO and L109, respectively) exhibited up to 4 fold more susceptibility to all antibiotics, dyes, disinfectants and detergents (including triclosan, novobiocin and fusidic acid) than did the parental control strain SL 1344.

All mutants created exhibited the same susceptibility to amikacin, tobramycin, gentamicin and SDS as the parental strain SL1344 (data not shown). 2.2. In vitro adhesion, invasion and intracellular survival of mutants.

S. Typhimurium SLl 344 adhered to both human embryonic intestine cells (INT-407) and mouse monocyte macrophages (RAW 264.7) in similar numbers (3.5x105 - 5.1x105 CFU/ml). The results obtained in this study are similar are similar to those reported by other observers in studies using wild-type S. Enteritidis and INT-407.

Results comparing the adhesion of S. Typhimurium SLl 344 and the mutants Ll 06 (AacrD), L356 (AacrF) and L643 (AacrB) to the human embryonic intestine cells are shown in Figure IA. As illustrated in this Figure, the level of adhesion of these three mutants and the parental strain were broadly comparable.

In contrast the mutant L108 (AtolC) adhered very poorly, compared with SL1344 (exhibiting only about 5.5% of the adhesion exhibited by the parental strain). The counts obtained for L108 (AtolC) were significantly different to those for SL1344 (P = 0.002).

In the human embryonic intestine cells mutants Ll 06 {AacrD) and L356 (AacrF) invaded and survived in similar numbers to the control parental strain SL 1344, as shown in Figure IB. In contrast, mutants L643 (AacrB) and Ll 08 (AtolC) exhibited only 12% and 7% respectively of the invasion level exhibited by SLl 344. These values were significantly different to those obtained using SL1344 (P = 0.028 and P = 0.021, respectively).

Mutants L 106 (AacrD) and L356 (AacrF) both adhered to mouse monocyte macrophages in similar numbers to the control parental strain SL1344 (data not shown).

Mutants L643 (AacrB) and L108 (AtolC) respectively exhibited 36% and 11% of the adhesion measured in respect of the control parental strain SLl 344. These adhesion values for L643 and L108 were significantly different to those for SL1344 (P = 0.02 and P = 0.0009, respectively).

In the invasion and intracellular survival experiments, SL 1344, L 106 (AacrD) and L356 (AacrF) invaded and survived intracellularly in mouse monocyte macrophages in similar numbers (data not shown). Mutants L643 (AacrB) and Ll 08 (AtolC) exhibited invasion levels 19% and 15%, respectively, those of SLl 344. Statistical analysis revealed that the invasion counts obtained in respect of mutants L643 and L 108 were significantly different to those obtained for SL1344 (P = 0.00005 and P = 0.00004, respectively). .

2.3. Infection of one-day-old chicks.

The results of competitive index experiments performed are shown in Table 3 and Figure 2 (Panels A to D)

Group A birds were inoculated with Ll 08 (AtolC) and SLl 344. Between days 1 and 7 the total count of S. enterica increased until day 10 after which the count decreased (Figure 2A and Table 3). The counts obtained on BGA supplemented with kanamycin allowed distinction between SL1344 and L108. Low numbers of L108 (AtolC) were observed throughout the experiment (Figure 2B and Table 32). The highest count was obtained on day 10, after which it decreased.

Group B birds were inoculated with L643 (AacrB) and SLl 344. Between days 1 and 10 the total count of S. enterica increased until day 14 after which there was a decline (Figure 4C and Table 2). Low numbers of L643 (AacrB) mutant were observed throughout the experiment (Figure 2D and Table 3). The highest count was obtained on day 3, after which it decreased.

Group C birds were inoculated with the Ll 06 (AacrD) and SLl 344. Between days 1 and 10 the total count of S. enterica increased until day 14 after which there was a decline (Table 3). Between days 1 and 7 the numbers of Ll 06 shed increased, after which they decreased. Between days 17 and 24 an increase in numbers of Ll 06 was observed, which subsequently decreased from day 28.

Group p birds were inoculated with L356 (AacrF) and SL1344. Between days 1 and 7 the total count of S. enterica increased until day 10, after which it decreased (Table 2). Between days 1 and 7 the shedding of L356 increased until day 14, where it decreased (Table 3). 2.4. Infection of two-week-old chicks.

To establish the ability of each mutant to colonise, persist and compete with the normal flora in the avian GI tract and compare directly to the parent strain, competition index experiments were performed (Table 3). Due to the poor colonisation observed in Group A birds, Ll 08 (AtolC) was not included in this study as it was deemed unable to compete with SL 1344 plus the normal gut flora. This property of Ll 08 (and seemingly of all bacteria having defective ToIC function) indicates that these mutant bacteria represent particularly suitable candidates for use in vaccines, since the bacteria are able to give rise to an effective immunising response, but are not able to maintain a population in the gut, thereby providing a vaccine having a rapid clearance rate after immunisation.

Group E birds were inoculated with L643 (AacrB) and SLl 344. Between days 1 and 7 the total & enterica count increased, after which it decreased (data not shown). Low numbers of L643 were observed throughout the experiment. The highest count was obtained on day 1, after which it decreased.

Group F birds were inoculated with L 106 (AacrD) and SL 1344. The total S. enterica count increased from days 1 to 4, after which it decreased. Between days 1 and 4 the numbers of L 106 shed increased reaching a maximum on day 7, after which the numbers decreased.

Group G birds were inoculated with L356 (ΔacrF) and SL1344. Between days 1 and 4 the S. enterica count increased reaching a maximum on day 7. After day 7, the numbers of S. enterica shed decreased. The numbers of L356 shed increased between days 1 and 7, after which the numbers decreased.

3. Discussion

Disruption of the efflux pump genes acrD and acrF had no significant effect upon the ability of S. Typhimurium to adhere, invade and survive in human embryonic intestine cells or mouse monocyte macrophages. These data suggest that AcrD and AcrF play little or no role in the ability of S. Typhimurium to cross the host cell epithelium or in intracellular survival. The mutants that lacked acrB were significantly less able to adhere, invade and survive in mouse monocyte macrophages. Slightly different data were obtained with the human embryonic intestine cells; disruption of acrB had no effect on the ability of this mutant to adhere, but was required for invasion and intracellular survival.

Disruption of tolC abolished the ability of S. Typhimurium Ll 08 to adhere, invade and survive in both cell types. These data suggest that functional AcrB and ToIC proteins are important for the adhesion, invasion and survival of S. Typhimurium in eukaryotic cells.

A competitive index assay was used to determine the ability of each mutant to colonise and persist in the avian GI tract compared directly to the parent strain. The total count of S. enterica declined in numbers after day 10, which could be due to the acquisition of the normal avian gut flora, which would compete for space and nutrients. The tolC mutant colonised and persisted in the avian digestive tract poorly compared to SL1344. This mutant was also hyper-susceptible to antibiotics, dyes, detergents and bile and was deficient in adherence and invasion in tissue culture. The poor ability to colonise and persist may be due to the mutant's bile hypersensitivity.

The AacrB mutant was able to colonise the day old and two week old chicks, but was not able to persist gastro-intestinally, which is in agreement with the in vitro tissue culture human embryonic intestine cell data. This suggests that efflux via AcrB is not required for colonisation of Salmonella in the avian gut, but is important in gastro-intestinal survival.

In conclusion, data from the present study indicate an important role for the AcrAB-TolC efflux system in the colonisation, persistence and invasion of Salmonella Typhimurium in the host. These data illustrate that tri-partite efflux pumps allow survival of bacteria within hostile environments, and hence contribute to the pathogenicity of the organism.

The Experimental Results reported above clearly indicate that Gram-negative bacteria having defective efflux pump function (such as AcrB or TolC knockouts) represent suitable candidates for use as vaccines. These bacteria are able to induce an effective immunising response in subjects to whom they are administered, but as a result of their reduced virulence are rapidly cleared from the subject without producing resident bacterial colonies.

STUDY B

Study B is an expansion of the study reported as Study A above. Accordingly some of the results reported in Study B are shared in common with Study A.

Bacterial strains.

Salmonella enterica serovar Typhimurium L354 (SL1344) was used as a control throughout. This strain is well studied and considered representative of this serovar (Wray & Sojka, 1978). L356 (acrFraph) and L643 {acrBraph) were described previously (Eaves et al, 2004).

Construction of mutants.

The acrD and tolC genes were inactivated in S. Typhimurium L354 using the one-step technique previously described for E. coli (Datsenko & Wanner, 2001) and S. Typhimurium (Eaves et ah, 2004). Primers used for the inactivation of these genes are shown in Table 3. Insertion of the aph gene in the correct location was confirmed by PCR and sequencing using the primers in Table 1. In addition, acrB::aph and tolC::aph were transduced back into L354 via P22 transduction and the phenotype confirmed, aph was excised from the chromosome using the method described by Datsenko and Wanner, (2000). Essentially, the transduced gene disrupted mutants/strains were transformed with helper plasmid CP20, which encodes the FLP recombinase under the control of a temperature sensitive promoter. After transformation with the pCP20 helper plasmid, kanamycin sensitivity could not be restored to either the transduced tolCr.aph mutant or the non-transduced tolC::aph mutant. Therefore, two extra tolCr.aph disrupted mutants were constructed following the method of Eaves et ah, (2004). One tolCr.aph mutant was constructed with aph expressed 5' to 3' (L141) and to ensure that no polar effects were seen another mutant expressing aph 3' to 5' (L142) was constructed in the same position. The insertion of the aph gene in the correct orientation and location was confirmed by PCR and sequencing using primers in Table 1. RT-PCR showed that the acrD: : aph mutant expressed 15% more (P = 0.09) acrB than L354. Finally, acrB and tolC were complemented in all mutants. The complements were prepared by cloning wild-type acrB and tolC PCR amplimers from L354 into pWSK30 (Wang & Kushner, 1991) after digestion with HindIII to create pacrB and ptolC, respectively. Inserts were verified by DNA sequencing and MICs determined.

Determination of MIC of antibiotics, dyes, detergents and disinfectants.

The minimum inhibitory concentration (MIC) of each agent was determined using the doubling agar dilution method as previously described (Andrews, 2001). AU MIC values were determined at least three independent MIC occasions.

Adhesion and invasion/intracellular survival assays.

The tissue culture assays were performed as previously described (Dibb-Fuller et al., 1999). Briefly, tissue culture cells were maintained in Dubecco's modified essential media (DMEM; Sigma) supplemented with 10% heat-inactivated foetal calf serum (Sigma), 2mM L-glutamine (Sigma) and 50μg/ml gentamicin (Sigma). Confluent monolayers of human embryonic intestine cells (INT-407) and mouse monocyte macrophage (RAW 264.7) were prepared in 24- well tissue culture plates (Nunc) and pre- washed in Hank's balanced salt solution (HBSS; Sigma) prior to adhesion and invasion/intracellular survival assays. Cultures of bacteria were grown in LB broth for 18h at 37°C and were washed and diluted in DMEM prior to the adhesion and invasion assays. For the adhesion assay, each cell line was infected with approximately 5 x 10⁷ CFU/ml pre-washed bacteria and incubated for 2h at 37°C (5% CO₂). Monolayers were washed six times with HBSS before disrupting with 1% (v/v) Triton (Sigma). Colony forming units (CFU) was determined by serial dilution plated onto LB agar. Adhesion is reported as the number of adhered bacteria minus the number of invaded bacteria.

Invasion/intracellular survival assays were preformed in duplicate 24-well plates. After allowing the bacteria to adhere with the monolayer, wells were washed three times before adding LB broth containing lOOμg/ml gentamicin to kill all external bacteria. Plates were incubated, for 2h at 37°C (5% CO₂) and washed twice. The monolayers were then disrupted with 1% (v/v) Triton and CFU were determined as described above. Each adhesion and invasion/intracellular survival assay was performed at least three times (four wells per test), and the data analysed with the Student's two-tailed t-test P values < 0.05 were taken as significant.

Electron microscopy.

Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) were performed with mouse monocyte macrophages and human gut epithelial cells. TEM and SEM were exactly performed as the adherence assay except that after 2h incubation with S. Typhimurium, the monolayers were washed three times with phosphate buffered saline (PBS) and fixed with 3% glutaraldehyde. Representative electron micrographs were taken for each strain.

Colonisation and persistence of S. enterica in vivo.

All animal procedures were performed under the United Kingdom Animals Scientific Procedures Act (1986). Semi-quantitative competitive exclusion experiments were performed using specific pathogen-free white Leghorn chickens as described previously (La Ragione & Woodward, 2003). Seven groups (A - G) containing at least 11 birds were housed in separate isolators and given feed and water ad libitum. The first set of birds (A to D) were inoculated when they were 24hrs old by oral gavage with 0.1ml of a 1:1 ratio of parent and each mutant (suspended in PBS) giving approximately 10⁵ CFU/ml total bacteria. The second set of birds (E to G) was inoculated at 14 days old by oral gavage with 0.1ml of a 1:1 ratio of parent and each mutant (suspended in PBS) giving approximately 10⁷ CFU/ml total bacteria.

Each group of birds were sampled by cloacal swabbing, twice a week for at least five weeks. The swabs were aseptically transferred to ImI sterile PBS and the CFU determined by serial dilution onto (i) Brilliant Green Agar (BGA; Oxoid) plates, (ii) BGA supplemented with 8μg/ml kanamycin and (iii) LB (Oxoid) plates supplemented with 8μg/ml kanamycin. Plating onto the Salmonella selective agar, BGA, gave the total Salmonella count, whereas plating onto BGA supplemented with kanamycin grew only the mutant Salmonella as they contained the aph gene. Those mutants with acrB or tolC disrupted were susceptible to the dye in the BGA, and so experiments with these strains used LB (Oxoid) plates supplemented with 8μg/ml kanamycin. Colonies growing on LB agar were confirmed as S. Typhimurium by agglutination test using 04 antisera. The remaining PBS, including the swab, was transferred to 10ml selenite broth and incubated at 37°C for seven days. If no colonies were observed on any of the plates (i) - (iii) (above) then the selenite broth was streaked onto BGA and/or LB agar supplemented with kanamycin. Data are expressed as CFU per swab.

Results

Antimicrobial susceptibility.

S. Typhimurium L354 was sensitive to all the agents tested (Table 4). L356 (acrFr.aph) and L106 (acrDr.aph) were also susceptible to all agents (Table 1). L643 (acrB::aph) and L108 (tolC::aph) and L354 transductants were up to 4 fold more susceptible to all antibiotics, dyes, disinfectants and detergents including SDS, triclosan, novobiocin and fusidic acid than L354 (Table 4). AU mutants had the same susceptibility to amikacin, tobramycin and gentamicin as L354 (data not shown). These data extend those previously observed for mutants lacking components of tri-partite efflux pumps (Lacriox et al., 1996; Eaves et al., 2004). To ensure that no polar effects are seen in mutants contracted with the method of Datsenko & Wanner (2001), using the templates between the FRT sites an identical scar that has stop codons in all six reading frames is created such that deletions remove only an N-terminal coding region and express a C-terminal protein domain. To futher confirm no polar effects were observed the kan^R (aph) cassette was removed from L643 (acrBr.aph); this gave rise to a kanamycin susceptible strain (L644), the remainder of the phenotype was unchanged. We were not successful at removing the kan^R (aph) cassette from L108. Therefore, two further mutants (L141 and L142) were constructed by inserting the kan^R cassette (aph) both forward and backwards and inserting further into tolC. The phenotype was the same as for Ll 08 (Table 4). Complementation of acrB in L643 and tolC in L 108 by addition of plasmid encoded wildtype genes cloned from L354 into pWSK30 reversed the phenotype to that of the parent strain L354 (Table 4).

In vitro adhesion, invasion and intracellular survival of mutants. S. Typhimurium L354 adhered to both human embryonic intestine cells (INT-407) and mouse monocyte macrophages (RAW 264.7) in similar numbers (3.5x10⁵ - 5.IxIO⁵ CFU/ml). These are similar numbers to those observed by Dibb-Fuller et ah, (1999) with wild type S. Enteritidis and INT-407.

The tissue culture experiments involve lysing the eucaryotic cells with 1% v/v triton X- 100. As Ll 08 and L643 are hyper-susceptible to this detergent by MIC testing, experiments were performed to ensure that any data obtained were genuine and not experimental artefacts. The effect of Triton X-100 was investigated by simulating use in the cell culture assays where the detergent is added once the bacteria are growing logarithmically and they are only exposed to the detergent for 15 minutes, after which the culture is diluted thereby removing the detergent. Under these conditions growth was not inhibited for one hour (four times longer than in the tissue culture experiment) after addition of this detergent. In addition, viable cell counts were determined after exposure to water or 1% triton; the values obtained in water were similar to those in triton (data not shown). These data indicate that Triton X-100 had no detrimental effect on cell viability during the time-span of the experiments.

S. Typhimurium L354 and the mutants L106 (acrDraph), L356 (acrFr.aph) and L643 (acrB::aph) adhered to the human embryonic intestine cells in similar numbers (Figure IA). Ll 08 {tolCr.aph) adhered very poorly, compared with L354 (5.5% compared with 100%). The counts obtained for L108 (tolCr.aph) were significantly different to those for L354 (P = 0.002). In the human embryonic intestine cells, L354, Ll 06 (acrD::aph) and L356 (acrFr.aph) invaded and survived in similar numbers (Figure IB). L643 (acrBraph) and L108 (tolCr.aph) invaded 12% and 7%, respectively, to that of L354 (100%), and the counts were significantly different to those for L354 (P = 0.028 and P = 0.021, respectively).

L354, Ll 06 (acrD::aph) and L356 (acrFr.aph) all adhered to the mouse monocyte macrophages in similar numbers (data not shown). L643 (acrB::aph) and Ll 08 (tolC::aph) adhered 36% and 11%, respectively, compared with L354. The values for L643 and L108 were significantly different to those for L354 (P = 0.02 and P = 0.0009, respectively). In the invasion and intracellular survival experiments, L354, Ll 06 (acrD::aph) and L356 (acrFr.aph) invaded and survived intracellularly in mouse monocyte macrophages in similar numbers (data not shown). L643 (ocrB::aph) and L108 (tolC::aph) invaded 19% and 15%, respectively, to that of L354 (100%), and the counts were significantly different to those for L354 (P = 0.00005 and P = 0.00004, respectively).

Electron microscopy.

When observed by SEM S. Typhimurium L354 and all four mutants caused actin cytoskeleton mediated membrane protrusions (ruffles) in the human embryonic intestine cells. Semi-quantification of the infected human embryonic intestine cells indicated that L354, L106 (acrD::aph) and L356 (acrF::aph) caused > 5 ruffles per field compared to 1-2 ruffles per field caused by L643 (acrBr.apK) and Ll 08 (tolCr.apH). SEM analysis of the mouse monocyte macrophages revealed similar data.

TEM analysis of the infected mouse monocyte macrophages showed that S. Typhimurium L354 and all four mutants invaded these cells. Two hours post infection there were > 30 bacteria per eukaryotic cell of L354, L 106 and L356. These bacteria were able to replicate within the SCV, with the number of daughter cells often exceeding 10 cells per SCV. There were ~20 bacteria per eukaryotic cell of L108 (tolC::aph), but there were only 1 or 2 cells per vacuole and bacterial cell degradation was visible, suggesting that these cells had not invaded, but been phagocytosed. There were ~l-2 bacteria per eukaryotic cell of L643 (acrBr.apK), again suggesting phagocytosis rather than invasion.

Infection of one-day-old chicks. To establish the ability of each mutant to colonise and persist in the avian GI tract compared directly to the parent strain, competitive index experiments were performed (Table 3 and Figure 2).

Group A birds were inoculated with Ll 08 (tolC::aph) and L354. Between days 1 and 7 the total count of S. enterica increased until day 10 after which the count decreased (Figure 2A and Table 3). The counts obtained on BGA supplemented with kanamycin allowed distinction between L354 and L108. Low numbers of L108 (tolC::aph) were observed throughout the experiment (Figure 2B and Table 3). The highest count was obtained on day 10, after which it decreased. Group B birds were inoculated with L643 (acrB::aph) and L354. Between days 1 and 10 the total count of S. enter ica increased until day 14 after which there was a decline (Figure 2C and Table 3). Low numbers of L643 (acrB::aph) mutant were observed throughout the experiment (Figure 2D and Table 3). The highest count was obtained on day 3, after which it decreased.

Group C birds were inoculated with the L106 (acrD::aph) and L354. Between days 1 and 10 the total count of S. enterica increased until day 14 after which there was a decline (Table 3). Between days 1 and 7 the numbers of Ll 06 shed increased, after which they decreased. Interestingly, between days 17 and 24 an increase in numbers of Ll 06 was observed, which subsequently decreased from day 28.

Group D birds were inoculated with L356 (acrFr.aph) and L354. Between days 1 and 7 the total count of S. enterica increased until day 10, after which it decreased (Table 3). Between days 1 and 7 the shedding of L356 increased until day 14, where it decreased (Table 3).

Infection of two-week-old chicks.

To establish the ability of each mutant to colonise, persist and compete with the normal flora in the avian GI tract and compare directly to the parent strain, competition index experiments were performed (Table 3). Due to the poor colonisation observed in Group A birds, Ll 08 (toIC::aph) was not included in this study as it was deemed unable to compete with L354 plus the normal gut flora.

Group E birds were inoculated with L643 (acrB::aph) and L354. Between days 1 and 7 the total S. enterica count increased, after which it decreased (Table 3). Low numbers of L643 were observed throughout the experiment. The highest count was obtained on day 1, after which it decreased.

Group F birds were inoculated with Ll 06 (acrD::aph) and L354. The total S. enterica count increased from days 1 to 4, after which it decreased. Between days 1 and 4 the numbers of Ll 06 shed increased reaching a maximum on day 7, after which the numbers decreased (Table 3).

Group G birds were inoculated with L356 (acrFy.aph) and L354. Between days 1 and 4 the S. enterica count increased reaching a maximum on day 7. After day 7, the numbers of S. enterica shed decreased. The numbers of Ll 06 shed increased between days 1 and 7, after which the numbers decreased.

Discussion

Disruption of the efflux pump genes acrD and acrF had no significant effect upon the ability of S. Typhimurium to adhere, invade and survive in human embryonic intestine cells or mouse monocyte macrophages. These data suggest that AcrD and AcrF play little or no role in the ability of S. Typhimurium to cross the host cell epithelium or in intracellular survival.

The mutants that lacked acrB were significantly less able to adhere, invade and survive in mouse monocyte macrophages._. Slightly different data were obtained with the human embryonic intestine cells; disruption of acrB had no effect on the ability of this mutant to adhere, but was required for invasion and intracellular survival.

Disruption of tolC abolished the ability of S. Typhimurium to adhere, invade and survive in both cell types. These data suggest that functional AcrB and ToIC proteins are important for the adhesion, invasion and survival of S. Typhimurium in eukaryotic cells.

SEM and TEM data suggested that L106 (acrD::aph) and L356 (acrF::aph) were able to adhere, invade and survive intracellularly in both cell lines as well as the parent strain (L354). TEM analysis confirmed these data. SEM analysis of L643 (μcrBr.apli) and L108 (tolC::aph) indicated fewer ruffles compared to L354 suggesting that these mutants were unable to invade as well as L354. The low number of L643 cells in the TEM micrograph also suggests that this mutant invades poorly. These data are in agreement with those from the tissue cell culture experiments, where little adhesion and invasion was observed. However, the number of L108 (toIC::aph) cells observed by TEM micrograph was to some extent contradictory with the tissue cell culture data, where little invasion was detected.

One hypothesis is that the bacteria were phagocytosed, rather than invaded, by the macrophage. Other than defects in SPI-I genes, delayed and reduced ruffle formation and invasion is associated with inactivated flagella. However, all mutants used in this study were fully motile. Normally, lysosomes fuse with the phagosome after bacterial ingestion, however, via expression of the SPI 2 genes (namely sifA and spiC), S. enterica can prevent the phagosome from entering the normal endocytic pathway, and thus bacterial replication within the vacuole can occur. It may be that Ll 08 may still replicate inside the macrophage without causing membrane ruffle-type invasion. Salmonella survival within the vacuole requires several secreted factors to overcome the harsh environments (nutrient and oxygen deprivation), and it is thought that ToIC plays a role in this secretion. Deletion of the outer membrane protein ToIC could abolish the secretion of one or more of these factors, resulting in impaired cell survival even though these Salmonella can still replicate. Poor invasion could also be due to secretion of host antimicrobials, such as basic peptides, to which the AcrAB-TolC pump usually provides defence; the MtrCDE system of gonococci confers resistance to protegrin-1.

A competitive index assay was used to determine the ability of each mutant to colonise and persist in the avian GI tract compared directly to the parent strain. The total count of S. enterica declined in numbers after day 10, which could be due to the acquisition of the normal avian gut flora, which would compete for space and nutrients. Mutants with tolC disrupted colonised and persisted in the avian digestive tract poorly compared to L354. These mutants were also hyper-susceptible to antibiotics, dyes, detergents and bile, and deficient in adherence and invasion in tissue culture. The poor ability to colonise and persist could be due to the bile hypersensitivity. The bile concentrations at different sites in the chicken digestive tract; duodenal, jejunal and cecal extracts have previously been determined. The concentrations of bile in these tissue samples were 1.75, 7 and 0.085 mg/ml, respectively. Alternatively, it could be due to its poor growth compared to L354; this mutant also does not survive acidic pH. As the crop and the gizzard are the most acidic compartments within a chicken, this may also contribute to the observed in vivo data.

Mutants with acrB disrupted were able to colonise the day old and two week old chicks, but unable to persist gastro-intestinally, which is in agreement with the in vitro tissue culture human embryonic intestine cell data. This suggests that efflux via AcrB is not required for colonisation of Salmonella in the avian gut, but is important in gastrointestinal survival. Eaves et al., (2004) showed that deletion of either acrB or acrF resulted in increased expression of the other RND pumps. The differences in substrate specificity of each RND pump could explain why a functional Acr AB system is important for intracellular survival whereas the other efflux pumps have a limited role. Previously a S. Typhimurium mutant with Tn PhoA kan inserted into acrB; has been shown to poorly colonised C57B 1/6 mice.

The Ll 06 (acrD::aph) count mirrored that of the total count up to 21 days post infection when another peak in numbers isolated was observed. This second peak could be due to the mutant not only competing with L354 but also with the natural gut flora. The increased invasion of L106 (acrD::aph) compared with L354 could be due to the 15% increase in expression of acrB (already implicated in gastro-intestinal survival) in this mutant (unpublished data).

Over the last decade there have been sporadic reports implying that components of tripartite efflux pumps play a role in the pathogenicity of other microorganisms, and data our study supports and extends these reports. For example, in previous studies decreased cellular invasion was observed with mutant P. aeruginosa PAOl with inactivated mexB (76% similarity and 70% identity to AcrB in S. Typhimurium). The same studies also showed that a hmexAB-oprM double knockout of P. aeruginosa PAOl, which includes the deletion of the outer-membrane protein component of the tri-partite efflux pump system, OprM, significantly reduced invasion in MDCK cells compared to the parent strain. A functional MtrCDE efflux system of Neisseria gonorrhoea enhanced survival in the female mouse model of genital tract infection, and it has been demonstrated that the CmeABC system plays a role in the colonisation of Campylobacter jejuni in poultry. Although there is evidence that a major role for the AcrAB-TolC efflux pump in S. enterica is the ability to survive bile, further roles of efflux pumps that contribute to overall pathogenicity should also be considered. To survive intracellularly, S. enterica must be able to overcome the toxic compounds found in lysosomes; it may be that a natural function of the AcrB-TolC efflux system is to export such compounds from the bacterial cell. Furthermore, it is postulated that the observed decrease in adhesion and/or invasion by the mutants with acrB or tolC disrupted could be accounted for by loss of secreted adhesins, which may be substrates of this system. However, it is also possible that the observed effects may not be due to lack of a functional AcrAB-TolC efflux system, but that lack of the genes leads to altered expression of virulence factors. The inventors have previously shown that lack of one efflux pump gene leads to increased expression of other efflux pump genes or mar A (Eaves et ah, 2004).

In conclusion, data from the present study indicate an important role for the AcrAB-TolC efflux system in the colonisation, persistence and invasion of Salmonella Typhimurium in the (avian) host. Bacteria in which efflux pump function is defective constitute surprisingly useful candidates for use in the production of vaccines.

STUDY C

Materials and Methods

Bacterial strains.

A mutant Salmonella enterica serovar Typhimurium SL1344 with tolC disrupted, L108, was used as the vaccine strain. Construction of Ll 08 is described in Study A above. The challenge strain was a nalidixic acid resistant mutant (MIC of 256μg/ml nalidixic acid, 0.25μg/ml ciprofloxacin) S. enterica serovar Typhimurium SLl 344, L696 (GyrA Gly87). L696 was chosen as it is resistant to nalidixic acid and thus can be differentiated from Ll 08 and native gut flora. In a competitive index model L696 has previously been shown to colonise and persist in chickens as well as S. enterica serovar Typhimurium SLl 344. Both the vaccine and challenge strains were grown in LB media.

Gastrointestinal shedding of vaccine strain

To illustrate the efficiency of Ll 08 to protect chickens against further colonisation by Salmonella (thus illustrating the ability of bacterial vaccines comprising Gram-negative bacteria having defective efflux pump function - such as Ll 08 - to prevent infection or colonisation by pathogenic bacteria), chickens were vaccinated at 24hrs old and given a booster vaccination at 15 days old (i.e. day 14 of the present study). At four weeks of age the chickens were exposed to the challenge strain, L696. All animal procedures were performed under the United Kingdom Animals Scientific Procedures Act (1986).

On day zero, 77 specific pathogen-free White Leghorn one day old chicks were randomly divided into seven isolators containing 11 birds each and given feed and water ad libitum. Two isolators contained birds that were not vaccinated; the remaining 5 isolators (55 birds) were vaccinated with L108. The chicks were vaccinated by oral gavage with 0.1ml washed and diluted (1/10 in phosphate buffered saline (PBS)) overnight culture giving 2 x 10⁷ CFU/ml per bird. A secondary (booster) vaccination was given when the birds were 15 days old. Each bird was inoculated with 0.1ml of a washed and 10-fold concentrated overnight culture of Ll 08 by oral gavage giving 1 x 10⁹ CFU/ml per bird. The inoculum was confirmed by viable counting. Ten birds, randomly chosen, were sampled by cloacal swabbing thrice weekly for least four weeks. Swabs were weighed before and after cloacal swabbing to ascertain the log g^'1 faeces. The swabs were aseptically transferred to ImI sterile PBS and the CFU determined by serial dilution onto LB plates supplemented with 16μg/ml kanamycin. , Colonies growing on LB agar were confirmed as S. Typhimurium by agglutination test using 04 antisera. The remaining PBS, including the swab, was transferred to 10ml selenite broth and incubated at 37⁰C for seven days. If no colonies were observed on any of the LB plates (above) then the selenite broth was streaked onto LB agar supplemented with kanamycin. Data are expressed as log CFU per gram faeces (CFU g^"l). The limit of detection for this method was <2.5 log cfu g^"1 faeces and determined by the lowest weight of faecal material (0.0032g) observed on a swab, in which it was assumed one viable bacterium of L 108 was recoverable.

Gastrointestinal shedding of challenge organism

AU vaccinated and un-vaccinated birds were exposed at four weeks old to S. Typhimurium L696. All birds were challenged with L696 at four weeks old by first treating the birds with 0.5ml 10% sodium bicarbonate, to neutralise the gizzard pH, then giving 0.5ml of an overnight diluted culture diluted 1/10000 in PBS giving 1 x 10⁵ CFU/ml per bird. The inocula were confirmed by viable counting. Ten unvaccinated birds and 10 vaccinated birds randomly chosen, were sampled by cloacal swabbing thrice weekly for least three weeks. Swabs were weighed before and after cloacal swabbing to ascertain the log g^'1 faeces. The swabs were processed exactly as described above except that dilutions were sub-cultured on Brilliant Green Agar plates supplemented with 8 μg/ml nalidixic acid to detect L696 and LB plates supplemented with 16μg/ml kanamycin to detect L 108. Tissue enumeration for vaccine and challenge strains

Prior to challenge with L696, three randomly chosen vaccinated birds were killed by cervical dislocation on days 2, 4, 7, 17, 18 and 21 days of age for post mortem examination. Four tissues samples were extracted; liver, spleen, caeca and caecal tonsils. Up to one gram of each tissue was aseptically weighed, diluted 1/10 with sterile PBS and homogenised. Serial dilutions were made from each tissue sample and lOOμl was plated out on LB plates supplemented with 16μg/ml kanamycin. The plates were incubated at 37° C for 18-24hrs before the colonies were counted and the log CFU g^'1 tissue calculated. Post-challenge at least five vaccinated and un-vaccinated birds (randomly chosen) were culled on days 29, 31, 35 and 45 days of age. Tissue samples were analysed as described above.

Immune response in vaccinated birds.

At least seven birds from each of the vaccinated and un-vaccinated groups were 'wing bled' at 11, 25, 37 and 45 days old. Blood samples were left for 4 hours at room temperature to clot, then the sera extracted. Both the IgM and IgG antibody responses to Salmonella whole cell antigen and LPS antigen were measured by enzyme-linked immunosorbant assay (ELISA). Control wells were seeded with Enter obacter aerogenes ATCC 13048, Escherichia coli K12 LlOOl (Promega) and Klebsiella pneumoniae. For the ELISA the bird sera was diluted 1/50, and lOOμl. Sera from a previous study investigating colonisation of chickens with S. enterica serovar Typhimurium SLl 344 was used as a positive control (Randall et ah, 2005). Chicken anti-IgG/IgM antibody conjugated to horseradish peroxidase (Serotec Ltd, UK) was diluted 1/50,000 (as per manufacturers instructions). The absorbance of each well was read using a spectrophotometer (Anthos HT2) at 450nm. A cut-off point was calculated as being the mean value plus two standard deviations of the absorbance for the negative control wells. Each ELISA was repeated three times and data expressed as mean absorbance at 450nm. . Results

Gastrointestinal shedding of vaccine and challenge strains L108 and L696.

Shedding of L108 (SL1344 tolC::aph) peaked at 5 x 10³ cfu g^"1 faeces one day after oral dosing by gavage and declined rapidly thereafter (Figure 5). The booster vaccine was given when the chickens were 15 days old (day 14 post vaccination) and resulted in a rise in shedding at 18 days of age to 1.8 x 10⁴ cfu g^"1 faeces. L108 was not detected at 21 days of age or any day thereafter in any of the cloacal swabs from the birds sampled. No L 108 were detected from the unvaccinated birds prior to challenge.

Both the unvaccinated and vaccinated birds were challenged at four weeks of age by oral gavage with 1 x 10⁵ cfu/ml of L696 (nalidixic acid resistant isogenic mutant of SLl 344, GyrA Gly87). Shedding of L696 by the un-vaccinated birds peaked at 1.1 x 10⁶ cfu g^"1 faeces at 35 days of age. Thereafter the amount of shedding remained relatively constant throughout the rest of the trial with median shedding values of 5.7 x 10⁴ to 7.6 x 10⁵ cfu g^" ¹ faeces (Figure 6). Shedding of L696 from the vaccinated birds peaked at 3.5 x 10⁵ cfu g^" ¹ faeces at 35 days of age; the amount of shedding remained relatively constant throughout the rest of the trial with median shedding values of 2.5 x 10⁴ to 1.1 x 10⁵ cfu g^" ¹ faeces. At 30, 32 and 35 days of age the number of vaccinated birds colonised with L696 was significantly lower than the un-vaccinated birds (Fisher's exact test P = 0.0054, 0.0027 and 0.0433, respectively).

Recovery of vaccine and challenge strains L108 and L696 from internal organs.

Seventy-six tissue samples [(four tissue types from three birds (four birds on at 20 days of age) taken at six separate time points)] were examined and Ll 08 was detected in 29 (Table 5). At 2 and 7 days of age, Ll 08 was detected in only two and one liver samples, respectively. Similarly, Ll 08 , was detected in only one and two spleen samples, 'respectively (Table 5). No caecal tonsil samples were taken when the chickens were two days old as the tissues had not developed sufficiently to be accurately identified. At four days of age, L 108 was detected in three caecal tonsil samples, but not in any at seven days ofage (Table 5).

After administration of the booster vaccination, Ll 08 was not detected in any liver samples and only in 2/10 spleen samples. Of the caecal samples Ll 08 was detected in 7/9 samples at 4 and 17 days of age (Table 5). After administration of the booster vaccination, Ll 08 was detected in 8/10 caecal samples at 17 and 21 days of age, and in the caecal tonsils only at 17 days of age.

Post-exposure to L696, 88 tissue samples (four tissue types from >5 birds at four separate time points) were taken from the un-vaccinated birds. L696 was detected in 44 (Table 6). At 31 and 35 days of age, L696 was detected in two and five liver samples, respectively, and in one and five spleen samples, respectively. L696 was detected in 18/22 caecal samples and 13/22 caecal tonsil samples between 29 and 45 days of age (Table 6).

Post-exposure to L696, 100 tissue samples (four tissue types from >5 birds at four separate time points) were taken from the vaccinated birds. L696 was detected in 36 (Table 6). At 35 days of age L696 was detected in only two liver and one spleen sample. L696 was detected in 17/22 caecal samples and 16/25 caecal tonsil samples between 29 and 45 days of age (Table 6). L696 was detected from significantly fewer liver and caeca samples from vaccinated birds at 31 days of age than from un-vaccinated birds (Fishers exact test P = 0.083); similarly L696 was detected in significantly fewer spleen samples positive from vaccinated birds at 35 days of age than from un-vaccinated birds (Fisher's exact test, P = 0.024).

Serum antibody detection

Serum samples were taken from un-vaccinated and vaccinated birds at 11, 25, 38 and 45 days of age. At all sampling times IgM specific for whole cell antigen was the predominant immunoglobulin present (Figure 7). After primary and secondary vaccination the sera from vaccinated birds had comparable titres of whole cell IgM specific antibody, to the sera from un-vaccinated birds. After challenge, un-vaccinated birds had higher titres of IgM compared to sera taken from vaccinated birds. Prior to challenge no LPS specific IgM immunoglobulin was detected from sera from un- vaccinated or vaccinated birds. After challenge at 28 days of age, the IgM titres increased and higher titres were observed for the un-vaccinated birds compared to the vaccinated birds. IgG antibody titres against Salmonella whole cell antigen was high for all birds at all sampling times.

Discussion and Conclusions

Ll 08 was administered to 55 chickens and was well tolerated causing no morbidity and no diarrhoea. Two days after primary vaccination 7/10 birds shed detectable numbers of Ll 08, whereas 16 days after vaccination (17 days of age) only 3/10 birds shed detectable numbers. L 108 is acid and bile hyper-susceptible compared to SLl 344 in vitro (Buckley et at, 2006), so the low number of birds shedding Ll 08 could be due to bile sensitivity and the increasing acidity of the crop and/or gizzard, which develops as the bird matures. Certainly these results indicate that bacterial vaccines of the invention comprising bacteria such as L 108 having defective efflux pump function exhibit reduced persistence in the bodies of hosts to whom they are administered, indicative of reduced colonisation or infection compared to that achieved by wild-type bacteria.

L 108 was administered at ~ 10⁷ CFU/ml to each bird, similar to the recommended doses for Zoosaloral, vacT and χ3985 poultry vaccines (3xlO⁷, 8xlO⁶ and 2xlO⁷ CFU/ml to each bird, respectively) (Barbezange et ah, 2000). Previously reported studies in chickens with an S. enterica serovar Enteritidis aroA- mutant (CVL30) used an initial dose of 10⁹ CFU/ml (Cooper et ah, 199 A). Another potential vaccine, S. enterica serovar Typhimurium SL 1344 dam^'phoP^' mutant, used three inoculating doses of 10⁷ - 10⁹ CFU/ml given to three-day-old chicks and boosted with the same inocula two weeks later (Du & Wang, 2005). However, data with Ll 08 suggest that a single vaccine dose to chicks at one-day-of-age may be sufficient to provide protection. It will immediately be appreciated that the ability to confer protective immunisation via only a single administration of vaccine is a notable advantage of the vaccines of the invention. Despite not detecting L 108 on cloacal swabs at 21 days of age (after primary and secondary vaccination) Ll 08 was recovered from caecal tissue and from 8/36 liver and spleen tissue samples, albeit at low numbers obtained post mortem, indicating that a small number of L 108 invaded and disseminated to these tissues. However, the persistence of the Ll 08 was markedly reduced compared to other marketed vaccines. Barbezange et al., (2000) showed by cloacal swabbing that Zoosaloral was shed for up to 5 weeks post primary vaccination, and vacT and χ3985 were still shed 8 weeks post primary vaccination. χ3985 was also still isolated from liver and spleen samples at week 6, indicating similar systemic survival to wild-type S. enterica serovar Typhimurium (Barbezange et ah, 2000). CVL30 (S. enterica serovar Enteritidis aroA^') was isolated and persisted in liver and spleen samples up to 21 days post vaccination (Cooper et ah, 1994; Van Immerseel et al., 2002). At doses ranging from 10⁷ to 10⁹, the & enterica serovar Typhimurium damphoP^' mutant persisted gastrointestinally to ~ 6 weeks and persisted in spleen and liver samples up to week 5 (Du & Wang, 2005).

The challenge strain was a nalidixic acid resistant (with decreased susceptibility to fluoroquinolones) mutant of SLl 344 (GyrA Gly87). This was chosen as GyrA mutants of Campylobacter jejuni are fitter than wildtype strains in poultry (Luo et al, 2005), and previous work by ourselves has also indicated that S. enterica serovar Typhimurium possessing substitutions in GyrA were as fit, if not more so, than wildtype salmonella (Randall et al, 2005). Accordingly this was a 'worst case' challenge for vaccines of the invention comprising Ll 08. Ability of such vaccines to produce a successful immunisation response capable of preventing or reducing colonisation or infection by L696 would indicate that these vaccines would be more than capable of achieving such results when exposed to challenge by wild-type pathogenic bacteria.

After challenge, the un-vaccinated birds were quickly colonised with L696, which persisted to the end of the trial. L696 was detected in the liver and spleen samples from 80% of the birds, up to seven days after challenge. Up to six days after challenge the vaccinated birds showed a significant reduction in the number of (i) L696 recovered, (ii) birds colonised with L696, (iii) shedding of L696, and (iv) liver and spleen tissues in which L696 was detected. This delay in colonisation by L696 is unlikely to be due to 'competition exclusion' with L108, as L108 was not detected on cloacal swabs a week before challenge. Instead these data show that vaccination induced sufficient immunity (which the inventors believe to be both mucosal and systemic) to confer effective protection against infection or colonisation by the challenge bacterial strain. This reduction may prove a valuable trait in this vaccine as it could protect chickens against the main route of Salmonella dissemination through the faecal-oral route. These data are more favourable than those obtained after vaccination with SalenVacT and challenged with S. enteήca serovar Typhimurium strain 2391 (nalidixic acid resistant), where the challenge strain of 10⁶ cfu/ml/bird colonised and invaded the liver, spleen and caeca after three days (Clifton-Hadley et al, 2002).

Immunity to S. enterica colonisation and infection in chickens depends upon several factors such as dose, route of inoculation and virulence of strain. Immunity to an intracellular pathogen such as S. enterica serovar Typhimurium is reliant on a good ThI response and the recruitment of T cells to the site of infection to activate macrophages (delayed-type hypersensitivity) (Babu et al., 2003). A ThI response involves CD4⁺ T cells recognising Salmonella antigen bound MHC class II molecules and secreting IL-2, IFN-γ and TNF cytokines, therefore priming and activating CD8⁺ T cells and macrophages for microbe killing (Hess & Kaufmann, 1993; Withanage et al., 2005). However, several studies have also indicated a role for humoral response in S. enterica serovar Typhimurium infection and gastrointestinal clearance (Lee et al., 1983; Hassan & Curtiss, 1990).

Pre-challenge, both the un-vaccinated and vaccinated birds had raised whole cell specific IgM responses, which increased after challenge. This initial high antibody titre was not due to an immune response specific to S. Typhimurium LPS as no LPS specific IgM antibodies were detected pre-challenge. However, it may reflect responses to common core LPS antigen of other commensal Enterobacteriaceae that form part of the chicken gastrointestinal flora. The increase in IgM titres post challenge observed for whole cell antigen could be accounted for by the increase in LPS specific IgM antibodies. Suppressed IgG and IgM responses of vaccinated birds after challenge have been noted before (Cooper et al., 1994). This suggests protective responses were not related entirely to humoral antibodies and could be due to other arms of the immune system.

Data presented show the potential of Ll 08 as a vaccine candidate for the use in poultry: it is administered easily via the oral route and yet is cleared easily from chickens and it does not become antibiotic resistant unlike some of the current licensed vaccines.

STUDYD

This Study illustrates that the attenuation of Salmonella enterica serovar Typhimurium SLl 344 tolC::aph (observed in poultry in the Studies outlined above) is also applicable in mammalian hosts, as illustrated using the following mouse model.

Introduction

Salmonella enterica serovar Typhimurium SLl 344 tolC::aph (Ll 08) poorly colonises poultry and has subsequently been shown to act as a vaccine against later Salmonella infection or colonisation. To establish whether the activity noted in poultry is also applicable to other animals, including humans, the standard Balb/c mouse model was used as a surrogate for human infection.

Materials and Methods

On day zero, 30 female BALB/c five week old mice were divided into three equal groups of 10 and housed in wire roofed cages with food and water ad libitum. Group A mice were inoculated with sterile phosphate buffered saline (control group), group B mice were inoculated with S. Typhimuirum SL1344 (L354) and group C mice were inoculated with S. Typhimuirum SL1344 tolCv.aph mutant (L108).

Overnight cultures of L354 and L108 were prepared by inoculating 100ml pre-warmed

(37°C) LB broth with a single colony from a purity plate and incubated at 37⁰C for 18hrs

with shaking (150rpm). Group B and C five-week-old mice were inoculated by oral gavage with 0.1ml washed and diluted (1/100) overnight culture giving 2.3 x 10⁶ CFU/ml/mouse and 1.8 x 10⁶ CFU/ml/mouse, respectively. Group A received 0.1ml sterile PBS by oral gavage. The actual inoculum given to the mice was determined by viable counts from serial dilutions.

24hrs after dosing and daily thereafter, fresh faecal pellets (-20 from each group) were collected from each group and analysed for S. Typhimurium content. Faecal pellets were weighed (upto Ig of faecal matter was used) and diluted 1/10 in BPW, then vortexed vigorously and incubated at room temperature for lhr. The samples were vortexed again to ensure complete homogenisation and the appropriate serial dilutions made in PBS and

lOOμl plated out onto BGA and LB agar supplemented with 16μg/ml kanamycin. The

plates were incubated overnight at 37⁰C and the CFU/g^"1 faecal matter determined. ImI

of the macerated faecal matter was also inoculated into 9mls of selenite broth and

incubated up to seven days at 37°C.

Four and five days post inoculation, up to five mice from each group were humanely culled by cervical dislocation and post mortemed. At post mortem, the liver, spleen and caeca tissues were aseptically removed and Ig tissue, where possible, was homogenised using rotating blades, which were cleaned and then sterilised between each tissue sample.

Serial dilutions were made from each tissue sample and lOOμl plated out onto BGA and

LB agar supplemented with 16μg/ml kanamycin and incubated as before. ImI of the

homogenised organ tissue was also inoculated into 9ml of selenite broth and incubated up

to seven days at 37°C. Results and Discussion

SLl 344 was inoculated at a similar concentration as to previous studies. Typically salmonella enter the host via M cells and then systemically colonise the organs and tissues leaving few bacteria to be shed in to the faeces. Our data matches previous observations (Table 7 and Figure 8). It can be seen clearly that while low numbers of SLl 344 were shed from the faeces, these remained constant over the period of study. However, Ll 08 was only detected in the faeces after one day post-inoculation and not thereafter.

Post-mortem analysis of liver samples revealed that SLl 344 was present in higher numbers than Ll 08, consistent with Ll 08 being attenuated (Table 8).

These data indicate that SLl 344 tolC::aph mutants (Ll 08) are attenuated in mice compared with the parent strain. As such L 108 should be a viable option as a vaccine candidate for mammals and warrants further experimentation.

STUDY E

This Study illustrates that the biological characteristics that make Salmonella having defective efflux pump function suitable for therapeutic use in accordance with the invention are also observed in other bacterial species engineered to have defective efflux pump function. These biological characteristics were found to be applicable across in vitro models using human, murine and avian cell lines. This clearly indicates the broad therapeutic utility of E. coli having defective efflux pump function, and the suitability of such bacteria for use in the methods and medicaments of the invention.

Experimental design

Using the same technology (Datsenko & Wanner, 2001; Eaves et ah, 2004; Buckley et ah, 2006) as used previously to construct Salmonella enterica serovar Typhimurium SL1344 with tolC disrupted (Ll 08), a strain of Escherichia coli O78:K80 (1115) was constructed with tolC disrupted (1117). Escherichia coli O78:K80 is a cause of collibacillosis in poultry, an infection that can have huge economic impact to the poultry farmer, hence a vaccine effective against this infection would also be marketable. E. coli O78:K80 tolCr.aph (1117) was hyper-susceptible to various antimicrobial agents as found previously for L108 compared with its parent strain SL1344 (L354) (Table 9). In . addition, 1117 grew more slowly than its parental strain, 1115 (Figure 9).

No animal experiments were performed as the Home Office Licence required for the work is still pending, therefore to determine whether E. coli O78:K80 tolCr.aph (1117) was attenuated in vivo, the ability of 1117 compared with 1115 to associate with, and invade, two types of cell lines were investigated, human epithelial cells (INT407) and mouse monocyte macrophages (RAW 264.7).

Compared with S. Typhimurium SLl 344 the number of parental E. coli O78:K80 (1115) that associated with the human epithelial cells (INT407) was low (Table 10; Figure 10A). E. coli O78:K80 tolCr.aph (1117) associated in significantly fewer numbers than its parental strain, Il 15. When the ability to invade the epithelial cells was determined it was again observed that compared with S. Typhimurium SLl 344 the number of parental E. coli O78:K80 (1115) that invaded the cells was much lower (Table 11; Figure 10B). No invasion of the epithelial cells by E. coli O78:K80 tolCr.aph (1117) was seen.

Compared with S. Typhimurium SL1344 the number of parental E. coli O78:K80 (1115) that associated with the mouse monocyte macrophages (RAW 264.7) was similar (Table 12; Figure HA). E. coli O78:K80 tolCr.aph (1117) associated in significantly fewer numbers than its parental strain, 1115. When the ability to invade the epithelial cells was determined it was again observed that compared with S. Typhimurium SLl 344 the number of parental E. coli O78:K80 (1115) that invaded the cells was lower (Table 13; Figure HB). Invasion of the macrophages by the E. coli O78:K80 tolC::aph (1117) was significantly less than that observed for the parent strain, Il 15.

These data are similar to those observed for Salmonella enterica serovar Typhimurium Ll 08, disruption of tolC in E. coli O78:K80 impaired both association with, and invasion of, host cells.

In addition, a third cell line was investigated, avian macrophages, HDI l. New data for both salmonella and E. coli were obtained:

S. Typhimurium SL1344 tolCv.aph associated with, and invaded, the avian macrophages poorly (Tables 14 and 15; Figures 12A and 12B). However, S. Typhimurium SL1344 acrBv.aph associated and invaded the avian macrophages in similar numbers to the parent strain.

Compared with S. Typhimurium SL1344 the number of parental E. coli O78:K80 (1115) that associated with the avian macrophages (HDI l) was similar (Table 14; Figure 12A). E. coli O78:K80 tolCr.aph (1117) associated in significantly fewer numbers than its parental strain, 1115. When the ability to invade the epithelial cells was determined it was observed that compared with S. Typhimurium SL 1344 the number of parental E. coli O78:K80 (1115) that invaded the cells was similar (Table 15; Figure 12B). Invasion of the macrophages by the E. coli O78:K80 tolCr.aph (1117) was significantly less than that observed for the parent strain, Il 15.

These results indicate that E. coli having defective efflux pump function share characteristic biological activity with Salmonella lacking efflux pump function. These Salmonella have been shown to constitute useful vaccines, and the common activity indicates that E. coli having defective efflux pump function will also be useful as vaccines. These vaccines (such as vaccines comprising 1115) may be particularly preferred for the prevention and/or treatment of collibaccilosis.

Table 1. Primers used in this study

Amplinier size

Gene Primer Sequence (5'→3')^a Reference (bp)

Creation of S. Typhimurium SL1344 mutants

CGCGATAATGGCATTGGTATCGGCGAACCGCATGAGC acrD knockout forward

CTGGTGTAGGCTGGAGCTGCTTC acrD N/A This study

TTCAACCATGAGAGCAGCGGCGGAGTCGGCATCGGTA acrD knockout reverse

ATCGGGAATTAGCCATGGTCCAT

TCATATGCGGCAATACGAATCTGTTCGACCAGCACCA tolC knockout forward OO

CCTGTGTAGGCTGGAGCTGCTTC tolC N/A This study

GAAATTGAAGCGAGAAAAGGCGAGAATGCGGCGGAAT tolC knockout reverse

AGCGGGAATTAGCCATGGTCCAT

PCR verification acrD upstream forward AATTGTGCGTGAAGCGGTCC 2066 acrD This study acrD downstream reverse GCGACCGAACAACATTCCGT (4801)^b tolC upstream forward CTTCTATCATGCCGGCGACC 1745 tolC This study tolC downstream reverse CGCTTGCTGGCACTGACCTT (2249)^b

Sections underlined show the PKD4 homologous regions of the primers used to create the S. Typhimurium SL1344 mutants. ' Numbers in brackets show ampliiner size before homologous recombination with the kanamycin resistance determinant.

Table 2. Minimum inhibitory concentrations of antibiotics, dyes and detergents.

Strain Genotype MIC (μg/mi) number CIP NAL CHL TET AMP ACR EtBr BILE CTAB TRIC NOV FUS

SL1344 Wt 0.03 8 4 2 1 128 512 >2048 >256 0.12 64 128

L643 AacrB 0.015 1 1 1 0.5 32 64 1024 64 0.03 16 16

L356 AacrF 0.03 8 4 2 0.5 128 256 >2048 256 0.12 >256 >256

Ll 06 AacrD 0.03 8 4 2 1 128 512 >2048 >256 0.06 >256 >256

L108 ΔtolC 0.015 1 1 0.5 0.5 32 32 512 64 0.015 1 4

CIP, ciprofloxacin; NAL, nalidixic acid; CHL, chloramphenicol; TET, tetracycline; AMP, ampicillin; ACR, acridine orange; EtBr, ethidium bromide; CTAB, cetyiximethylammoniiimbromide, TRIC, triclosan; NOV, novobiocin; FUS, fusidic.

Table 3. Gastrointestinal shedding of mutants compared with wildtype parent strain SL1344.

Mutant/Parent ratio (CFU/swab)^a

Day old Chicks Colonisation Gastro-intestinal Persistence (Days 1-3) (Days 7-17)

SL1344 and L108 (Δto/Q 0.005 0.01

SL1344 and L643 (Δαcr.5) 1.6 0.009 SLl 344 and Ll 06 (AacrD) 2.3 1.3

SL1344 and L356 (ΔαcrF) 1.3 0.8

Colonisation Gastro-intestinal Persistence

Two week old chicks (Days 1-4) (Days 7-20)

SL1344 and L643 (AacrB) 0.15 0.019

SLl 344 and Ll 06 (AacrD) 0.6 0.4

SL1344 and L356 (AacrF) 0.7 0.4

A, Ratio of numbers of mutant growing on BGA + kanamycin agar: total numbers viable Salmonella Typhimurium growing on BGA above. Bold text indicates values that are significantly different (P<0.05) to those of the total count.

Table 4. Minimum inhibitory concentrations of antibiotics, dyes and detergents.

Strain Genotype MIC (μg/ml) number CIP NAL CHL TET NOV FUS ACR EtBr BILE CTAB SDS TX lOO TRIC

Wt

L354 0.03 8 4 2 64 128 128 512 >2048 >256 >256 >2048 0.12 (SL1344)

L643 acrBr.aph 0.015 1 1 ^' 1 16 16 32 64 1024 64 256 1024 0.03

L643 +

Wt (pαcr£) 0.03 4 2 2 64 64 128 512 >2048 128 >256 2048 0.12 00 pacrB

L644 AacrB* 0.015 1 2 1 8 16 32 64 >2048 64 256 1024 0.06

L141 tolC::aph^a 0.015 0.5 1 0.5 8 16 16 32 512 64 16 256 0.015

L142 tolC::aph^b 0.015 0.5 1 0.5 8 16 16 32 512 64 16 256 0.015

L108 tolC::aph 0.015 1 1 0.5 1 4 32 32 512 64 64 512 0.015

L108 +

Wt (ptolQ 0.015 4 2 1 64 64 128 512 >2048 128 >256 1024 0.12 ptolC

L106 acrD::aph 0.03 8 4 2 >256 >256 128 512 >2048 >256 >2048 >2048 0.06

L356 acrF::aph 0.03 8 4 2 >256 >256 128 256 >2048 256 1024 >2048 0.12

CIP, ciprofloxacin; NAL, nalidixic acid; CHL, chloramphenicol; TET, tetracycline; NOV, novobiocin; FUS, fusidic acid; ACR, acridine orange; EtBr, ethidium bromide; CTAB, cetytrimethylammoniumbromide; SDS, sodium dodecylsulphate; TX-IOO, Triton X-IOO; TRIC, triclosan. * in L644 the aph kanamycin resistance cassette has been removed. ^a tolC gene disrupted with kanamycin resistance cassette expressed 5' to 3'. ^b tolC gene disrupted with kanamycin resistance cassette expressed 3' to 5'. Two fold differences between MIC values were considered relevant as those reported above are the mode value obtained from >10 experiments.

Table 5. The recovery of S. Typhimurium L108 vaccine strain from tissues taken from birds dosed with vaccine at one day of age and again at 15 days of age.

^' Liver tissue Spleen tissue Caeca tissue Caeca tonsil tissue

Age of chickens

CFU (logio) Number CFU (log₁₀) Number CFU (IOg₁₀) Number CFU (log₁₀) Number

g ^la positive¹* g ¹ positive g ¹ positive g^'1 positive

2 3.00 ± 2.60 2/3 4.60 1/3 5.06 ± 5.06 2/3 ND ND

00

4 0 0/3 0 0/3 4.99 ± 4.70 3/3 4.58 ± 4.43 3/3

7 2.78 1/3 3.75 ± 3.65 2/3 3.48 ± 2.98 2/3 0 0/3

17 0 0/3 0 0/3 3.60 ± 3.19 3/3 5.18 ± 5.03 3/3

18 0 0/3 2.30 1/3 2.95 1/3 0 0/3

21° 0 0/4 3.26 1/4 3.46 ± 3.23 4/4 0 0/4

Double line indicates when the booster vaccine was administered. ND; Not done (the caeca tonsils had not developed sufficiently). ^a The mean of the counts cfu g^"1 of L108 recovered from positive organs is given to log base 10. The standard error of the counts of L108 recovered was calculated for values from positive organs only, and is given to log base 10. ^b The number of organs positive relative to the number of birds tested.

⁰ Day 21 one extra bird was culled due to 'sour crop' affliction.

birds.

Liver < issue Spleen tissue Caeca tissue Caeca tonsil tissue

Age of

Group chickens CFU Number CFU Number CFU Number CFU Number

(logio) g ^la positive⁵ (logio) g^"1 positive (logio) g^"1 positive (logio) g^"1 positive

29 0 0/5 0 0/5 4.30 1/5 0 0/5

31 3.70 ± 3.36 2/5 • 3.42 1/5 7.53 ± 7.34 5/5 6.41 ± 6.24 3/5

35 3.10 ±2.80 5/5 3.51 ± 3.30 5/5 7.64 ± 7.47 5/5 6.49 ± 6.27 5/5

45^C 0 0/7 0 0/7 8.20 ± 8.11 111 6.89 ± 6.85 5/7

29 0 0/5 0 0/5 0 0/5 0 0/5

1 31 0 0/5 0 0/5 5.64 ± 5.62 2/5 4.60 ± 4.48 3/5

« 1 35 3.76 ± 3.62 2/5 3.00 1/5 7.34 ± 7.29 5/5 5.68 ± 5.36 4/5

45^C 0 0/10 0 0/10 7.48 ± 6.76 10/10 6.25 ± 5.66 9/10

^a The mean of the counts cfu g^"1 of L696 recovered from positive organs is given to log base 10. The standard error of the counts of L696 recovered was calculated for values from positive organs only, and is given to log base 10.

The number of organs positive relative to the number of birds tested. ^c Day 45 all birds remaining were culled.

Table 7. Number of S. Typhimurium in mouse faecal pellets.

CFU/g^"1 faecal LOG CFU/g-¹

Days post inoculation Group matter faecal matter

PBS <100^a <2

SLl 344 40000

4.60206

L108 100 2

PBS <100 <2

SLl 344

400 2.60206

L108 <100 <2

PBS <100 <2

SLl 344

200 2.30103

L108 <100 <2

PBS <100 <2

SLl 344

600 2.778151

L108 <100 <2

PBS <100 <2

SLl 344

700 2.845098

L108 <100 <2

¹ The limit of detection is 100 CFU/g faecal matter. Table 8.

Recovery of S. Typhimurium SL1344 and L108 vaccine strain from livers taken from mice dosed at five weeks of age.

Liver tissue

Days post infection Group Strain CFU Number

-la

(logio) g positive

A PBS 0.00 0/4

4 days B SLl 344 4.88 2/5

C SL1344 tolC.-.aph 4.78 1/5

A PBS 0.00 0/5

5 days B SLl 344 4.34 1/5

C SL1344 tolCwaph 0.00 0/5 ^a The mean of the counts CFU g^" of S. Typhimurium recovered from positive organs is given to log base 10. ^b The number of organs positive relative to the number of mice tested.

Table 9

Susceptibility of Escherichia coli O78:K80 with tolC disrupted (1117) compared with its parent strain (1115).

MIC (μg/ml)

Strain NaI Cip ChI Tet Kan EtBr Fus AF Bile SDS

1115 4 0.03 4 2 2 1024 >256 256 >2048 >256 1117 1 <0.03 2 1 >128 16 64 8 256 16

Bold text indicate hypersusceptibility. The aph gene confers kananiycin resistance.

Table 10.

Association of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a human gut epithelial cell line (INT 407).

Percent

LOG

Strain Mean SD SE compared T-test (p) mean to parent

L354 506666.7 5.704722 217938 65710.78 100 L108 1666.667 3.221849 616.9328 186.0122 0.328947^a 0.000006^a L643 164166.7 5.215285 71789.89 21645.47 32.40132^a 0.0002^a 1115 311666.7 5.49369 87368.95 26342.73 61.51316^b 0.01^b 1115 311666.7 5.49369 87368.95 26342.73 100 1117 2083.333 3.318759 297.9729 89.84222 0.668449° 0.00000009° K12 1366.667 3.135663 494.2089 149.0096 0.269737 0.000006 ^a Salmonella compared to SL1344. L108 is tolCwaph, and L643 is acrBv.aph. ^b 1115 compared to L354.

° Il 17 is O78:K80 tolCv.aph and so has been compared with 1115. Table 11.

Invasion of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a human gut epithelial cell line (INT 407).

Percent

Strain Mean LOG mean SD SE compared T-test (p) to parent

L354 336666.7 5.527200119 103426.2 31184.16 100

L108 82.75 1.917768002 26.97516 8.133316 0.024579³ 0.0000002^a

L643 127500 5.105510185 65383.48 19713.86 37.87129³ 0.0002^a

1115 1.5 0.176091259 1.445998 0.435985 0.000446^b 0.0000002^b

1115 1.5 0.176091259 1.445998 0.435985 100

1117 0 0 0 0 0.004°

K12 1.142857 0.057991947 0.651339 0.196386 0.000339 0.0000002

^a Salmonella compared to SL1344. L108 is tolCv.aph, and L643 is acrBv.aph. ^b Il 15 compared to L354.

⁰ Il 17 is O78:K80 tolCv.aph and so has been compared with Il 15.

Table 12.

Association of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a mouse macrophage cell line (RAW264.7).

Percent

LOG

Strain Mean SD SE compared T-test (p) mean to parent

L354 735000 5.866287 218819.1 65976.45 100

L108 4450 3.64836 1576.821 475.4293 0.605442³ 0.0000002^a

L643 690000 5.838849 121430.9 36612.78 93.87755^a 0.6^a

1115 509166.7 5.70686 131940.6 39781.6 69.27438 ^b 0.02^b

1115 509166.7 5.70686 131940.6 39781.6 100

1117 5116.667 3.708987 671.272 202.3961 1.00491° 0.00000004°

K12 2166.667 3.335792 578.9227 174.5518 0.294785 0.0000002

^a Salmonella compared to SL 1344. L 108 is tolCv.aph, and L643 is acrBv.aph. ^b Il 15 compared to L354.

⁰ Il 17 is O78:K80 tolCv.aph and so has been compared with Il 15.

Table 13.

Invasion of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to a mouse macrophage cell line (RAW264.7).

Percent

LOG

Strain Mean SD . SE compared T-test (p) mean to parent

L354 498333.3 5.69752 277352.2 83624.83 100

L108 2066.667 3.31527 967.0323 291.5712 0.414716^a 0.00007^a

L643 187500 5.273001 130462.5 39335.91 37.62542^a 0.002^a

1115 23833.33 4.377185 19093.52 5756.912 4.782609^b 0.0001^b

1115 23833.33 4.377185 19093.52 5756.912 100

1117 2891.667 3.461148 850.0891 256.3115 12.13287° 0.003°

K12 775 2.889302 245.4125 73.99464 0.155518 0.00007

^a Salmonella compared to SL1344. L108 is tolCv.aph, and L643 is acrBv.aph. ^b Il 15 compared to L354. ^c Il 17 is O78:K80 tolCv.aph and so has been compared with. Il 15.

Table 14.

Association of S. Typhimurium SL1344 and E. coli O78:K80 and tolC disrupted mutants, respectively, to an avian macrophage cell line (HDIl).

Percent

LOG

Strain Mean SD SE compared T-test (p) mean to parent

L354 924166.7 5.96575 154828.8 46682.65 100

L108 5275 3.722222 . 946.8848 285.4965 0.570784^a 0.0000000004^a

L643 790833.3 5.898085 147983.9 44618.83 85.57259^a 0.01^a

1115 1695000 6.22917 699772.7 210989.4 183.4085^b 0.002^b

1115 1695000 6.22917 699772.7 210989.4 100

1117 5616.667 3.749479 979.6412 295.3729 0.331367° 0.000004°

K12 1916.667 3.282547 671.272 202.3961 0.207394 0.0000000004

^a Salmonella compared to SL1344. L108 is tolCv.aph, and L643 is acrBwaph. ^b Il 15 compared to L354. ^c Il 17 is O78:K80 tolCv.aph and so has been compared with Il 15.

Table 15.

Invasion of S. Typhimurium SL1344 and E. coli O78.-K80 and tolC disrupted mutants, respectively, to an avian macrophage cell line (HDIl).

Percent

LOG

Strain Mean SD SE compared T-test (p) mean to parent

L354 150833.3 5.178497 80165.55 24170.82 100

L108 1341.667 3.127645 510.7184 153.9874 0.889503^a 0.00005^a

L643 125833.3 5.099796 70124.35 21143.29 83.42541 ^a 0.3^a

1115 87500 4.942008 58949.13 17773.83 58.01105^b 0.1^b

1115 87500 ^; 4.942008 58949.13 17773.83 100

1117 1683.333 3.22617 989.7964 298.4349 1.92381° 0.0004°

K12 2316.667 3.364864 2892.257 872.0483 1.535912 0.00006

Salmonella compared to SLl 344. Ll 08 is tolC::aph, and L643 is acrBv.aph.

Il 15 compared to L354.

Il 17 is O78:K80 tolCv.aph and so has been compared with Il 15.

SELECTED REFERENCES.

Motoi Y, Sato K, Hatta H₅ Morimoto K₅ Inoue S, Yamada A. Production of rabies neutralizing antibody in hen's eggs using a part of the G protein expressed in Escherichia coli. Vaccine. 2005 Apr 27;23(23):3026-3032.

Pillai S₅ Howell A₅ Alexander K, Bentley BE, Jiang HQ₅ Ambrose K₅ Zhu D₅ Zlotnick G. Outer membrane protein (OMP) based vaccine for Neisseria meningitides serogroup B. Vaccine. 2005 Mar 18;23(17-18):2206-9.

Belli SI₅ Mai K, Skene CD, Gleeson MT₅ Witcombe DM₅ Katrib M₅ Finger A₅ Wallach MG₅ Smith NC. Characterisation of the antigenic and immunogenic properties of bacterially expressed, sexual stage antigens of the coccidian parasite, Eimeria maxima. Vaccine. 2004 Oct 22,22(31-32):4316-25.

Andrews JM. Determination of minimum inhibitory concentrations. J Antimicrob Chemother. 2001 Jul;48 Suppl 1 :5-16.

Piddock, LJV₅ Jin YF, Ricci V, Asuquo AE. Quinolone accumulation by Pseudomonas aeruginosa, Staphylococcus aureus and Escherichia coli. J Antimicrob Chemother. 1999 Jan;43(l):61-70.

Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A. 2000 Jun 6;97(12):6640-5.

Eaves DJ₅ Ricci V and Piddock LJ Expression of acrB, acrF, acrD, marA, and soxS in Salmonella enterica serovar Typhimurium: role in multiple antibiotic resistance. Antimicrob Agents Chemother. 2004 Apr;48(4): 1145-50.

Wright GD₅ Thompson PR. Aminoglycoside phosphotransferases: proteins, structure, and mechanism. Front Biosci. 1999 Jan 1;4:D9-21

Claims

CLAIMS.

1. A vaccine comprising a Gram-negative bacterium having defective efflux pump function.

2. A vaccine according to claim 1, wherein the bacterium has been engineered to have defective efflux pump function.

3. A vaccine according to any preceding claim, wherein the bacterium is selected from the group consisting of Salmonella spp. and Escherichia coli.

4. A vaccine according to claim 3, wherein the bacterium is selected from the group comprising Salmonella Typhimurium; Salmonella Enteritidis; Salmonella Java and Escherichia coli O78:K80.

5. A vaccine according to claim 4, wherein the bacterium is selected from the group consisting of Salmonella enterica serovar Typhimurium SLl 344 and Escherichia coli O78:K80

6. A vaccine according to any preceding claim, wherein the bacterium has defective AcrAB-TolC activity.

7. A vaccine according to claim 6, wherein the bacterium is a ToIC knockout bacterium.

8. A vaccine according to claim 7, comprising Salmonella enterica serovar Typhimurium SL1344 L108.

9. A vaccine according to claim 7, comprising Salmonella enterica serovar Typhimurium SLl 344 L 109.

10. A vaccine according to claim 7, comprising Escherichia coli O78:K80 Il 17.

11. A vaccine according to claim 6, wherein the bacterium is an AcrA knockout bacterium.

12. A vaccine according to claim 6, wherein the bacterium is an AcrB knockout bacterium.

13. A vaccine according to claim 12, is selected from the group consisting of L643 and Ll lO (ΔacrB).

14. A vaccine according to either claim 1 or 2, wherein the bacterium comprises a Campylobacter spp.

15. A vaccine according to claim 14, wherein the bacterium has defective CmeABC activity.

16. A vaccine according to claim 15, wherein the bacterium is a CmeA knockout bacterium.

17. A vaccine according to claim 15, wherein the bacterium is a CmeB knockout bacterium.

18. A vaccine according to claim 15, wherein the bacterium is a CmeC knockout bacterium.

19. A vaccine according to any preceding claim for use in the prevention and/or treatment of a disease selected from the group consisting of salmonellosis; campylobacteriosis; collibaccilosis; diarrheal illness; urinary tract infections (UTIs); and neonatal sepsis and meningitis.

20. A vaccine comprising an artificial protein of the AcrAB-TolC efflux pump, or a fragment or derivative thereof.

21. A vaccine according to claim 20, wherein the artificial protein of the AcrAB-TolC efflux pump is a component of the AcrAB-TolC efflux pump associated with the inner membrane.

22. A vaccine according to claim 21, comprising an artificial AcrA protein, or a fragment or derivative thereof.

23. A vaccine according to claim 22, wherein the artificial AcrA protein, fragment or derivative is derived from Salmonella.

24. A vaccine according to claim 21, comprising an artificial AcrB protein, or a fragment or derivative thereof.

25. A vaccine according to claim 24, wherein the artificial AcrB protein, fragment or derivative is derived from Salmonella.

26. A vaccine according to claim 20, wherein the artificial protein of the AcrAB-TolC efflux pump is a component of the AcrAB-TolC efflux pump associated with the outer membrane.

27. A vaccine according to claim 26, comprising an artificial ToIC protein, or a fragment or derivative thereof.

28. A vaccine according to claim 27, wherein the artificial ToIC protein, fragment or derivative is derived from Salmonella.

29. A vaccine according to any one of claims 20 to 28, wherein the artificial AcrAB- TolC protein, fragment or derivative is a recombinant protein, fragment or derivative.

30. A vaccine according to any one of claims 20 to 29, wherein the artificial protein fragment or derivative has greater resistance to protease degradation than does the artificial protein from which it is derived.

31. A vaccine comprising an isolated artificial protein of the AcrAB-TolC efflux pump, or a fragment, variant or derivative thereof

32. A vaccine according to any preceding claim for use in poultry.

33. A vaccine according to any one of claims 1 to 31 for use in mammals.

34. A vaccine according to claim 33, for use in humans.

35. A live vaccine for preventing or treating poultry Salmonella spp. infection or colonisation comprising a live Salmonella spp. bacterium having defective efflux pump function.

36. A live vaccine for preventing or treating poultry Campylobacter spp. infection or colonisation comprising a live Campylobacter spp. bacterium having defective efflux pump function.

37. A live vaccine for preventing or treating poultry E. coll infection or colonisation comprising a live E. coli bacterium having defective efflux pump function.

38. A vaccine according to any preceding claim for administration by injection.

39. A vaccine according to any preceding claim for oral admim^'stration.

40. A vaccine according to any preceding claim for administration by whole body spray.

41. A vaccine according to any preceding claim further comprising an adjuvant.

42. A method of preparing meat for human consumption, the method comprising: i) administering to an animal intended for meat production an amount of a vaccine according to any preceding claim sufficient to render the animal substantially free of a Gram-negative bacterium harmful to humans; ii) growing the animal to a size suitable for harvesting for meat production; and iii) harvesting the animal for meat production.

43. A method of preparing food for human consumption, the method comprising: i) administering to an animal intended for food production an amount of a vaccine according to any preceding claim sufficient to render the animal substantially free of a Gram-negative bacterium harmful to humans; ii) growing the animal to a size suitable for harvesting food from the animal; and iii) harvesting food from the animal for human consumption.

44. A method according to claim 42 or claim 43, wherein the animal is selected from the group comprising fish, cows, pigs, sheep or poultry.

45. A method according to claim 45, wherein the food is selected from the group comprising eggs and milk.

46. A method of treating or reducing a disease associated with bacterial colonisation or infection, the method comprising administering to a patient in need of such treatment or reduction an effective amount of a vaccine comprising an artificial protein of the AcrAB-TolC efflux pump, or a fragment, variant, or derivative thereof.