The present invention relates to vaccine compositions comprising attenuated pathogens.

Background

We discovered in the middle 1980's (Barrow et al, 1987) that oral inoculation of newly hatched chickens with a Salmonella strain resulted in massive mtestinal colonization. This prevented establishment of a second strain given orally 24 h later. The effect was genus specific such that colonization with E. coli or other closely related genera such as Citrobacter did not prevent gut colonization by Salmonella and vice versa. An in vitro model was developed in which 24 h nutrient broth cultures of a Salmonella are inoculated with small numbers of a closely related strain or with the same strain with a different marker. If the mixed culture is reincubated the second strain does not grow. However, if the first strain is E. coli and the second strain Salmonella, the Salmonella does grow (and vice versa). We carried out further work to try to characterise in more detail the practical aspects of the inhibition in vivo (Berchieri & Barrow, 1990) and in vitro (Berchieri & Barrow, 1991).

We have now taken this work further and have developed vaccine compositions comprising attenuated pathogens.

Summary of the Invention

One aspect of the invention provides a vaccine composition comprising an avirulent mutant of a cellular pathogen which colonizes a vertebrate mucosal surface (preferably the gut), the mutant being characterised by

having a functional deletion of a gene encoding a protein involved in the electron transport chain or ATP synthase.

We have found that such mutants provide effective attenuated/avirulent strains for raising an immune response whilst not causing serious disease, and also (in at least some cases) provide an exclusion effect in the mucosal surface, thereby inhibiting the growth of other (non-attenuated) strains of the same pathogen or other pathogens.

Surprisingly, such mutants do not exhibit this exclusion/inhibition effect in the in vitro model discussed above. Hence, Zambrano and Kolter (1993) disclosed that E. coli mutants (nuo A or nuoB) lacking NADH dehydrogenase I had a competitive disadvantage in stationary phase, which would not have suggested their use in a vaccine. Mutants with deletions in other genes (for example aroA (Griffin & Barrow (1993) Vaccine 11, 457-462; Barrow et al (1990) Epidemiol. Infect. 104, 413-426) and his pur) are satisfactorily attenuated but do not exhibit the inhibition of colonization by other strains/pathogens.

The electron transport chain and associated F ₀ F, ATP synthase are common to all organisms which respire and it is reasonable to suppose that the invention, demonstrated below in relation to E. coli and Salmonella typhimurium, is applicable to all cellular pathogens, for example bacteria, fungi and protozoa. The pathogen may, for example, be any Eubacterial pathogen, such as any of the Vibrio spp. , Campylobacter spp. , Neisseria spp. or Mycobacterium spp. Preferably, however, it is E. coli or a Salmonella, such as Salmonella typhimurium, S. enteritidis or S. gallinarum. The pathogen may generally be one which is transmitted vertically (ie from mother to offspring).

A protein is "involved in" the electron transport chain if functional absence of the protein selectively damages the operation of the electron transport chain.

The genes involved in the electron transport chain include those encoding all or a subunit of or regulating the function of NADH dehydrogenase I, flavoproteins, coenzyme Q and cytochromes such as cytochromes b, c, , c, a and a ₃ . Preferably, the gene encodes a pyridine-linked dehydrogenase such as an NADH dehydrogenase I or an NADPH dehydrogenase. In the operon for ATP synthase, uncH is a suitable gene for mutation. Many genes have already been identified as encoding a protein involved in the electron transport chain, for example all of the E. coli nuo genes encoding the various subunits of NADH dehydrogenase I. In addition, we disclose below the sequences of the S. typhimurium nuoG and nuoH genes. The invention may, of course, be applied to genes which have yet to be identified.

The mucosal surface which the pathogen colonizes is preferably the gut. In newly-hatched chickens, colonization of the gut by bacteria is extensive. Later, the main site is the lower end of the alimentary tract, where the flow rate of contents is slower. The crop is also colonized, albeit to a lesser extent. The organisms generally exist in the lumen and may have an association with the mucus which allows inoculation of fresh chyme as it enters the caeca (chick) or colon/caecum (calf).

Commercially, gut pathogens are particularly important in the rearing of calves, pigs, lambs and chickens, but the invention is generally applicable to any vertebrate, particularly mammals (including man) and birds (for example turkeys and ducks). The vaccines of the invention may be especially valuable in the protection of agammaglobulinaemic calves

(which have not acquired enough maternal IgG from the colostrum) against bacterial septicaemias. In a human context, the vaccines may be especially useful if the intestine is colonized by antibiotic-resistant organisms, such as Pseudomonas or Staphylococcus aureus following antibiotic therapy prior to bowel surgery.

The mutants may be made by any convenient means, for example by transposon mutagenesis using Tn phoA or bacteriophage P22, followed by appropriate screening, by site-directed mutagenesis or by insertion of anti- sense DNA. The mutation may cause the gene to produce no protein at all, for example by introducing a stop codon early in the coding sequence or by interfering with the promoter or some other regulatory region (including a gene which produces a factor that causes or enhances expression of the electron transport gene). Alternatively, it may cause non-functional protein to be produced.

The vaccine composition may be formulated and administered in any conventional way; administration to the gastrointestinal tract, for example by nasal spray or oral drench, is preferred to parenteral administration. The most preferred method (at least for chicks) is to spray them with an aqueous preparation of the vaccine containing 10 ⁵ -10 ⁷ cfii/ml of the mutant organism so that each chick receives 10 ³ -10 ⁵ cfu (colony forming units) by taking the drops off its fluff. The vaccines of the invention may be particularly useful if administered early (ie immediately after hatching) to chicks, for example to prevent or ameliorate infections caused by vertical transmission in hatcheries. Breeders and layers may be revaccinated by administering i.m. 0.05 ml containing about 10 ⁵ -10 ⁷ cfu per dose of a killed vaccine at, say, twelve to sixteen weeks.

A further aspect of the invention provides the newly-isolated Salmonella

nuoG and nuoH genes or variants thereof. Such genes are useful in designing constructs for deleting the genes.

Preferred aspects of the invention will now be described by way of example and with reference to the accompanying drawings, in which:

Figure 1 shows partial sequences of the nuoG gene and entire coding sequence of the nuoH gene of S. typhimurium F98. The sequence starts at residue 1840 of the sequence previously reported by Archer et al (3). Beneath the nucleotide sequence is an alignment of the deduced amino acid sequences of the nuoG and nuoH gene products (subunits of NADH dehydrogenase I) from S. typhimurium and E. coli, showing only residues that vary between the two species (identical residues being indicated by a dot). The putative S. typhimurium nuoG gene product contains an additional nineteen amino acids at the C-terminus not present in the E. coli homologue. Putative Fe-S clusters in the NuoG sequence are underlined. A putative ribosome-binding site (Shine-Dalgarno sequence) is double underlined.

Figures 2 shows the strategy used to generate S. typhimurium defmed mutants of n-.o:Km.

Figure 3 shows the inhibitory activity of 24 h LB cultures of S. typhimurium mutant AB145 (A, closed circles) or F98 (B, open circles) for F98 Spc ^r (open and closed diamonds) after incubation for 1-3 days.

Figure 4 shows the oxidase activity of NADH dehydrogenase from the

F98 wild-type strain and the F98 nuoGv.TxφhoA mutant. Membrane vesicles were prepared from F98 wild-type (closed symbols) and the nuoG: .TnphoA mutant (open symbols) and were assayed for oxidation of

NADH (upper) and dNADH (lower). Results were normalised to 1 and are therefore presented as relative absorbance.

Figure 5 (on 6 sheets) shows the sequence of the Salmonella typhimurium cyd operon and, for comparison, the E. coli sequence.

EXAMPLE 1: MATERIALS AND METHODS

Bacterial strains, plasmids and culture conditions

S. typhimurium strain F98 is a prototrophic isolate from diseased chickens whose virulence and colonisation characteristics in chickens and mice have been well characterised(5, 6, 9). Spontaneous mutants of this and other strains resistant to nalidixic acid (Nal ^r ) or spectinomycin (Spc ¹ ) were produced as described previously (39). Insertion mutant AB145 of F98 (11) was produced previously by TxφhoA mutagenesis using, as the donor plasmid, pRT733 wE. coli SM10. S. typhimurium C5 is prototrophic and virulent for mice (20, 28). S. gallinarum strain 9 is highly virulent for chickens of all ages (5, 8, 38). E. coli K12 strains SY327 lambda pir, a lysogen of SY327 ((lac pro) arg £(Am) rifnalA recA56) containing the pir gene of plasmid R6K, was the host for transformation of suicide plasmid pGP704 containing the R6K replicon (29) and SM10 thi thr leu tonA lacY supE recvl::RP4-2-Tc::Mu Km lambda pir (37) was used for conjugal transfer of this plasmid. Plasmid vector pBluescript (KS(-) was used for cloning the target gene of the TnphoA mutagenesis. Bacteriophage p22 Htl05/l z/ (35) was used for transduction of markers as described previously (7). Unless indicated otherwise bacterial cultures were made in 10 ml volumes of LB broth (Difco) incubating for 24 h in an orbital incubator (150 revs/min).

DNA manipulations, sequencing and reagents

Chromosomal DNA was prepared as described by Pitcher et al (32). Plasmid DNA was prepared using alkaline lysis (34). Restriction endonucleases, T4 DNA ligase and Taq DNA polymerase were obtained from Boehringer Mannheim (Germany) and used according to he manufacturer's instructions. DNA fragments cloned in pBluescript KS(-) and DNA from PCR products were sequenced using an oligonucleotide derived from the sequence of the alkaline phosphatase gene in TτφhoA. Sequencing was carried out using a cycle sequencing programme with an ABI 373A sequencing system according to the manufacturer's protocols (Applied Biosystems, Foster City, California) and the data analysed using the GCG software package (17).

NADH dehydrogenase assay

The method was essentially that of Archer et al (3) with a number of differences. Cells were grown to late log phase (OD = 0.7) in LB broth and were disrupted at -70 °C with an X-press (Biox Ltd, Sweden). After removal of cell debris at 10,000 g for 10 min the protein concentrations of the preparations were measured spectrophotometrically (Pierce) and equalised before use.

Growth inhibition assay

The procedure has been described previously (11). Basically, a 24 h LB culture of one strain, resistant to one antibiotic, was inoculated with small numbers of a second strain, resistant to another antibiotic, followed by further incubation with enumeration of the second strain. The initial count of the second strain was ca. 10 ³ cfu/ml. Mutants were tested both as the

first and the second strain. Berchieri and Barrow (11) showed that inhibition was not related to resistance to either spectinomycin or nalidixic acid, the antibiotic resistances used in the assay.

Virulence assays

These were essentially following the protocols described by Barrow et al (5). Newly hatched chickens were inoculated orally with 0.1 ml volumes and three-week-old birds widi 0.3 ml of undiluted cultures or intramuscularly with 0.1 ml of decimal dilutions of cultures. Mice were inoculated orally with 0.05 ml volumes of cultures diluted in LB, while under light anaesthesia, or intravenously with 0.1 ml volumes of cultures diluted similarly. Animals which died or were killed after exceeding humane end points over periods of three weeks were scored. LD ₅₀ values were estimated (33).

The intestinal invasiveness of S. gallinarum 9 Nal ^r and its nuoG mutant (see results section) was assessed in two groups of chickens by assessing the rate at which organisms accumulated in the liver and spleen in the first three days after oral inoculation. This has recently been found to be a reliable indicator of mis characteristic (5). The behaviour of these two strains in the reticuloendothelial system of chickens was assessed by counting inoculated bacteria in the liver, spleen and blood following intravenous inoculation with 10 ⁴ cfu (S. gallinarum 9) or 10" cfu (nuoG mutant). Bacteria were counted on Brilliant Green agar (CM263, Oxoid, Basingstoke, United Kingdom) containing sodium nalidixate (20 μg/ml) and novobiocin (1 μg/ml).

RESULTS

Characterisation of the TnphoA insertion site in S. typhimurium F98 AB145

The non-inhibitory (in vitro) mutant of S. typhimurium F98, namely AB145, was partially rough as indicated by lipopolysaccharide profiles (results not presented). However, sufficient LPS was produced to enable the transposon-associated antibiotic-resistance marker to be transduced to the parent strain using bacteriophage P22. All kanamycin resistant transductants tested showed a similar non-inhibitory phenotype to AB145 (see below). The TnphoA insertion in AB145 was therefore likely to be the mutation responsible for the inability of a stationary phase LB culture to inhibit the growth of S. typhimurium F98 Spc ^r inoculated into the AB145 culture.

Previously reported work (11) indicated that the TnphoA inactivated gene(s) in AB145 was contained within a 11 kbp EcoRV fragment. Initial attempts at cloning the whole fragment into pBluescript (KS(-) were unsuccessful. H dIII digestion of AB145 chromosome DNA revealed hybridisation with a ca. 3.3. kbp fragment which contained the target gene-TNphoA junction. A section of an identical gel, corresponding to the position of this fragment, was excised. The DNA was purified and cloned into the compatible site of plasmid vector pBKS(-) and this was transformed into host strain XL 1-blue. Colonies containing the expected cloned 3.3 kbp H dIII fragment were identified by digestion of plasmid DNA followed by hybridisation using a ECL-labelled (ECL, Amersham) 1.3 kbp EcoKL-Xhol DNA fragment derived from the alkaline phosphatase gene of TnphoA .

The nucleotide sequence of the chromosomal fragment adjacent to the upstream TnphoA IS _50L was determined by cycle sequencing using a primer as described in the Materials and Methods section to allow sequencing outwards from the negative strand of TnphoA. The result revealed that TnphoA had inserted at nucleotide 1468 of nuoG, one of the genes in the nuo operon encoding NADH dehydrogenase I. The open reading frame of nuoG in S. typhimurium has not previously been completely sequenced. It was apparent that the orientation of TnphoA was such that the 5' end of the alkaline phosphatase gene was upstream from the 3 ' end of nuoG in AB145. In this orientation transcription of the phoA sequence would not have occurred from the nuo promoter. The TnphoA mutation in AB145 was therefore transduced by P22 to a phoN mutant of F98, which produces white colonies on LB agar containing 40 μg/ml X-P. All kanamycin resistant transductants were also white, indicating no detectable expression of phoA from the nuo or any other promoter.

Comparison between the nuoG and nuoH sequences of S. typhimurium F98 and E. coli

Both strands of a fragment containing parts of nuoG gene and a fragment containing nuoH gene, that was detected immediately downstream of nuoG, were determined by direct PCR sequencing. For this oligonucleotide primers, based on the sequence of nuoG, nuoH and nuol of E. coli, were used to amplify the genes from a colony of S. typhimurium F98. The deduced amino acid sequence encoding part of nuoG, and the whole of nuoH, together with the sequence of the same genes from E. coli is shown in Figure 1. The sequence data will appear in the EMBL/GENEBANK Nucleotide Sequence Data Libraries under the accession number L42521.

Comparison of the two gene sequences reveals a high degree of homology, many of the amino acid differences being conservative substitutions. The only major difference between the E. coli K-12 and S. typhimurium F98 sequences occurs at the 3 '-end of the nuoG gene. The predicted Salmonella protein is 20 amino acids longer than that of E. coli. The comparable sequence in E. coli K-12 contains non-coding triplets which would result in premature termination of translation of the gene. Comparison of the nuoH gene between E. coli K-12 and S. typhimurium F98 revealed very similar sequences.

NADH and dNADH assays

The results of assessing membrane vesicle preparations of the parent F98 and the nuoG: .TnphoA for NADH and dNADH oxidase activity are shown in Figure 4. The reduced activity of the nuoGr.TnphoA mutant against NADH was not great, the residual activity probably being due largely to the activity of NADH dhll. NADH dhll is unable to oxidise dNADH as shown in Figure 4, indicating the NADH dhl activity had been virtually eliminated from the nuoG mutant.

Construction of a defined mutation in nuoG

A defined mutant of nuoG harbouring an insertion of a DNA cassette encoding kanamycin-resistance was constructed (Figure 2). A 1.259 kbp EcoRl-Xbal fragment of the nuoG gene was cloned into the compatible site of suicide vector pGP704. A kanamycin gene cassette, carried by pBSK, was removed with EcoRV and Spel. After end-filling, the resulting blunt- ended fragment was inserted into the EcoRV site within nuoG. The constructed plasmid, pGP704, containing nuoG with the kanamycin- resistance cassette insertion, was electroporated into E. coli SY327 lambda

pir. Plasmid DNA was prepared and transformed into E. coli SM10 lambda pir (34) enabling the plasmid carrying the mutated nuoG to conjugate back into the wild-type S. typhimurium F98 Nal ^r . The defmed nuoG mutant was selected for by allele exchange resulting in a kanamycin resistant, ampicillin-sensitive transcripient. The kanamycin cassette insertion in the nuoG gene was confirmed using PCR (data not shown).

Inhibitory activity of AB145 and defmed mutants

Mutant AB145 was compared with the parent F98 for the ability of a 24 h LB broth culture to inhibit growth of F98 Spc ^r . We also studied the growth inhibition of AB145 by 24 h broth cultures of F98. The results of the former are summarised in Figure 3. The growth curves of the parent and mutant are similar. However, unlike the parent strain (Figure 3b), AB145 failed to inhibit the multiplication of F98 Spc ^r inoculated into the culture (Figure 3a). The parent strain was able to prevent multiplication of AB145 when this was added (results not shown). The defined nuoG mutant also showed an identical phenotype to the TnphoA mutant, failing to inhibit the growth of strain F98 Spc ^r when F98 Spc ^r was inoculated into a stationary phase culture of the nuoG mutant.

The precise nuoG mutation was transferred by P22 transduction into S. typhimurium C5 and S gallinarum 9. These mutants showed the same non-inhibitory phenotype in vitro against Spc ^r mutants of the parent strains (results not presented).

Virulence of AB145 and defined nuoG mutants for chickens and mice

AB145 was partially rough and not surprisingly was avirulent when inoculated orally into newly-hatched chickens, in contrast to the parent

strain S. typhimurium F98 which killed 14/25 birds. To assess the role of nuo in virulence, the defmed nuoG :Km mutation was transduced, using bacteriophage P22, into the parent S. typhimurium F98 strain. Analysis of the LPS of the parent F98 strain and the nuoG: :Km mutant confirmed that these strains were smooth (results not shown). The smooth nuoGr.Km mutant of S. typhimurium F98 was considerably less virulent in chickens than the smooth parent strain (Table 1).

Table 1: Virulence of Salmonella strains and nuoG mutants for chickens

Virulence in

Newly-hatched chicks 3-week-old chickens

Serotype S. typhimurium S. gallinarum

Strain C5 F98 9

Route oral oral oral i/m

Parent strain 26/26 ^a 24/26 ^a 18/24 ^a < 0.38 ^b nuoG mutant 13/27 8/26 0/24 > 7.08 ^d Number c f chicks diec [/number inoci llated with 1 D ^B cftι m 0.1 π

Log _!0 LD ₁₀ value by intramuscular (i/m) routes

The nuoG: :km mutation was transduced into S. gallinarum strain 9. In comparison to the parental S. gallinarum strain 9 the isogenic nuoG mutant was highly attenuated for chickens by both oral and parenteral routes of inoculation. There appeared to be little difference in invasiveness to the liver and spleen from the alimentary tract following oral inoculation of chickens with 5. gallinarum 9 or its nuoG derivative. Both strains were found in similar numbers in the caeca soon after infection and appeared in the liver and spleen at similar intervals after inoculation (Table 2).

Table 2. Intestinal invasiveness of 5. gallinarum 9 and its nuoG mutant

Days Log ₁₀ viable count gm of parent strain or mutant in the following organs ³ after infection nuoG mutant parent

Liver Spleen Caeca Liver Spleen Caeca contents mucosa tonsil contents mucosa tonsil

c 1 _N N 2.4 2.6 0.8 N N 3.0 2.2 0.8 m

2 N N N N 1.4 N N 0.7 N 2.9

H C H 3 0.8 0.7 N 1.9 1.3 1.7 1.6 1.4 3.3 1.8 rn

4 0.9 1.0 N 1.5 1.5 2.0 1.9 1.2 2.4 2.0

3 10 ^a Mean of values from three animals r c m N = log ₁₀ < 0.5

IO 0)

The higher numbers of the parent strain in these two organs at four days after infection indicated multiplication of this strain whereas none seemed to have occurred of the nuoG mutant. This was also observed after intravenous inoculation.

15

Table 3. Behaviour of S. gallinarum 9 and its nuoG mutant in the tissues after intravenous inoculation

Days after Log ₁₀ viable count/gm of parent strain or mutant in the following samples ³ infection nuoG mutant parent

Liver Spleen Blood Liver Spleen Blood

0 4.5 6.3 3.6 2.5 4.3 0.7

2 5.1 5.7 0.8 4.6 5.0 N w c w 4 4.2 ^C 5.6 _N 4.9 ^C 5.4 ^C 1.0

CΛ

H 7 3.4 ^C 5.3 ^C N 5.5 ^C 5.0 ^C 1.7 C H m 10 2.7 ^C 4.6 ^C N 5.7 ^C 5.6 ^C 2.4 w z m 14 3.3 ^C 4.5 ^C N Dead m (-

H

10 21 3.3 ^C 5.0 ^C N c m 35 2.7 4.5 N ro

42 1.7 2.6 N

Mean of values from three animals

15 N = log ₁₀ < 0.5 necrotic lesions present in organs

The parent strain multiplied in the liver and spleen until a bacteraemia occurred and the animals died. Despite inoculation of 100 times more organisms of the nuoG mutant the chickens remained healthy. This strain persisted in the liver and spleen in considerable numbers during the course of the experiment.

The nuoGr.Km mutation was also transduced into the mouse- virulent S. typhimurium strain C5 and groups of BALB/c mice were orally challenged with doses of 10 ^ό or 10 ⁸ parental or nuoG: :Km mutant bacteria. Four out of five mice challenged with IO ⁶ and twelve out of twelve mice challenged with IO ⁸ wild-type S. typhimurium C5 died. However, all ten mice challenged with IO ⁶ and seventeen of twenty mice challenged with 10 ⁸ S. typhimurium C5 nuoGr.Km survived the challenge. Mice surviving the S. typhimurium C5 challenge harboured bacteria in their livers and spleens and some had small abscesses in these organisms. The number of bacteria per organ showed considerable variation between individual mice and the persistence pattern resembled that seen previously following infection with purE mutants (30).

DISCUSSION

This study describes some of the biological characteristics of nuoG mutants of S. typhimurium and S. gallinarum which have defective NADH dehydrogenase I activity. We have demonstrated mat such a defect attenuates virulence of these serotypes for mice and chickens and that it abolishes the genus-specific inhibition of growth seen in early stationary phase broth cultures. The mutation was detected while screening TnphoA mutants for their inability to inhibit the multiplication of S. typhimurium F98 Nal ^r Spc ^r when incubated as 24 h broth cultures. The original mutant, AB145, was partially rough. This rough phenotype was likely to

have been selected during conjugation, when the plasmid pRT733 containing TnphoA was introduced. However, it was sufficiently susceptible to bacteriophage P22 to allow retransduction to the parent strain. The phenotype was transferred to all recipients tested, indicating that the transposon insertion, rather than the partially rough phenotype, was responsible for the characteristics of this mutant. Production of a defined mutation showed mat the lesion responsible for the inhibitory phenotype was in nuoG, situated in the middle of this operon containing fourteen genes (nuoA-N). The large number of termination codons between nuoG and nuoH suggest that translation downstream of nuoG may be reduced normally. Whether this contributes to some form of regulation of nuoH to nuoH is not known. It is unclear why there should be a difference in the length of nuoG in S. typhimurium F98 and E. coli. This may be explained simply by a comparison of a wild-type and laboratory strain. Stop codons could have accumulated in the E. coli gene over many years of in vitro culture.

Mutants of S. typhimurium strains F98 and C5 and of S. gallinarum 9 which were nuoG showed reduced virulence for chickens and mice. Moreover, introduction of the nuoG mutation into S. gallinarum produced a great reduction in virulence both by oral and parenteral virulence. In this case the major affected stage of pathogenesis appeared to be the ability to multiply in the reticuloendothelial system rather than intestinal colonization and invasion. The difference in the degree of attenuation between these two serotypes may reflect more fundamental differences in their virulence attributes. Elimination of the virulence plasmid from S. gallinarum also attenuates this serotype to a much greater extent than occurs following the same manipulation of S. typhimurium (4, 8).

Although mutant AB145 did not produce inhibition of growth in stationary

phase LB broth cultures it was nevertheless inhibitory in vivo (11) demonstrating that in vivo and in vitro inhibition mechanisms are different or are stimulated by different environmental conditions.

Growth suppression in the absence of nutrient starvation could be mediated by inter-bacterial signalling at high bacterial density. Mutations in nuo could conceivably affect this in a number of ways. For example, the reduction in aerobic metabolism which would be characteristic of AB145 would result in a higher than normal oxygen concentration present in early stationary phase. Regulatory proteins sensitive to such changes and indicator molecules which reflect such metabolic changes, such as acetyl phosphate (27) or internal cellular pH, could all separately or together be involved in such a mechanism.

The central role of the electron-transport chain in changes that occur in early stationary phase is supported by the fact that a second non-inhibitory mutant of S. typhimurium F98 has been found to have an insertion in the cyd operon, encoding cytochrome d oxidase. Table 4 shows that this mutant is non-inhibitory as a 24 h broth culture for small numbers of Stm F98 Nal ^r , even though it appears to be inhibitory in vivo.

Table 4. Multiplication over 4 days of S. typhimurium F98 Spc ^r (Spc ^r mutant of parent strain) in 24 h broth culture of cyd mutant of 5. typhimurium F98 Nal ^r .

Time (days) after Log ₁₀ viable numbers Log ₁₀ viable numbers inoc. of challenge of cyd mutant of challenge strain strain

0 9.60 3.59

1 9.69 7.53

4 9.78 9.30

EXAMPLE 2: CONSTRUCTION OF A DEFINED MUTATION IN cydA

This may be done in two ways:

1. Method 1 is to clone into a suicide plasmid such as pGP704 vector the cyd operon (cydA and E), amplified by the PCR with oligonucleotides 1 and 4. This fragment is then digested with EcoRV Αt base 1242 (in em _. ba:eccyd). A kanamycin gene cassette, carried by pBSK, was amplified witii oligonucleotides 5 and 6 which have Kpn\ sites included at their 5' ends. After end-filling, the resulting blunt-ended fragment was inserted into the EcoRV site within cydA . This constructed plasmid containing cloned cydA and B with the kanamycin cassette insertion was electroporated into E. coli SY327 λ pir (Ref 29). Plasmid DNA was prepared and transformed into E. coli SM10 λ pir (Ref 37) enabling the plasmid carrying the mutated cyd operon to conjugate back into the wild- type S. typhimurium strain. The defined mutation was selected for by allele exchange resulting in a kanamycin-resistant, ampicillin-

sensitive transcipient. The insertion is confirmed by PCR using oligonucleotides 1 and 4.

Oligonucleotides from any parts of the sequence may be used to check by PCR whether the gene has been disrupted, for example by insertion of an antibiotic resistance cassette. Oligonucleotides prepared from the extreme ends of the sequence will give a fragment approximately 2750-2800 in size depending on the size of the oligonucleotide. Insertion of a cassette will either disrupt this or will create an enlarged fragment.

2. Method 2 is to amplify single fragments by PCR from the N- terminal end of cydA (oligonucleotides 1 and 2) and from the C- terminal end of cydB (oligonucleotides 3 and 4). The two fragments have Kpnl sites with which they may ligate to each other and EcoRl and Xbal sites for ligation into pGP704. This plasmid is transferred sequentially into E. coli SY327 λ pir and E. coli SM10 λ pir. Allele exchange is used selecting for ampicillin sensitive transcipients. The deletion incorporating part of cydA and cydB is confirmed by PCR using oligonucleotides 1 and 4.

Oligonucleotide primers for cydA and B taken from em_ba:eccyd

Primer 1 base 10-29 with £cøRI site added to 5' end Primer 2 base 1146-1155 with Kpnl site added to 5' end

Primer 3 base 1877-1896 with Kpnl site added to 5' end

Primer 4 base 3582-3601 with Xbal site added to 5' end

Kanamycin cassette oligonucleotides

Primer 5 GAATTCGGTACCCGCTGAGGTCTGCCTCGTGAAGG Primer 6 GAATTCGGTACCAAAGCCACGTTGTGTCTAAAATC

EXAMPLE 3; CONSTRUCTION OF A DEFINED MUTATION IN uncH (ATP SYNTHASE)

Method 2 was followed in an identical way. The oligonucleotides were taken from the E. coli sequence deposited with the EMBL nucleotide data library (no em_ba:ecuncol).

N-terminal end {primer 1 base no 1990-2014 Xbal site added to 5' end

{primer 2 base no 2760-2785 Kpnl site added to 5' end C-terminal end {primer 3 base no 3411-3436 Kpnl site added to 5' end

{primer 4 base no 4045-4069 £cøRI site added to 5 ' end

ATP synd ase is concerned with ATP generation rather than the release of protons back with the cells. The mutant created here is non-inhibitory in vitro. In other words, when inoculated in small numbers into a 24 hour broth culture of the parent strain, it does not inhibit growth of the parent strain.

After incubation of the mixture the counts of the parent strain at various times of sampling were as follows:

0d 3 x 10 ² cfu/ml

Id 2.1 x IO ⁴ cfu/ml

2d 6 x 10 ⁵ cfu/ml

3d 2.7 x IO ⁶ cfu/ml

d 2.5 x IO ⁷ cfu/ml d 1.3 x lO ⁸ cfu/ml

REFERENCES

1. Anraku, Y. (1988) "Bacterial electron transport chains" Ann. Rev. Biochem. 57, 101-132. 2. Anraku, Y. and R.B. Gennis (1987) "The aerobic respiratory chain of Escherichia coir Trends Biochem. Sci. 12, 262-266.

3. Archer, CD. , X. Wang and T. Elliott (1993) "Mutants defective in the energy-conserving NADH dehydrogenase of Salmonella typhimurium identified by a decrease in energy dependent proteolysis after carbon starvation" Proc. Natl. Acad. Sci. USA 90,

9877-9881.

4. Baird, G.D., E.J. Manning and P.W. Jones (1985) "Evidence for related virulence sequences in plasmids of Salmonella dublin and Salmonella typhimurium" J. Gen. Microiol. 131, 1815-1823. 5. Barrow, P. A. , M.B. Huggins and M.A. Lovell (1994) "Host specificity of Salmonella infection in chickens and mice is expressed in vivo primarily at the level of the reticuloendothelial system" Infect. Immun. 62, 4602-4610.

6. Barrow, P. A. , M.B. Huggins, M.A. Lovell and J.M. Simpson (1987) "Observations on the pathogenesis of experimental

Salmonella typhimurium infection in chickens" Res. Vet. Sci. 42, 194-199.

7. Barrow, P. A. , J.O. Hassan and A. Berchieri Jnr. (1990) "Reduction in faecal excretion of Salmonella typhimurium strain F98 in chickens vaccinated widi live and killed S. typhimurium organisms" Epidemiol. Infect. 104, 413-426.

8. Barrow, P. A. , J.M. Simpson, M.A. Lovell and M.M. Binns (1987) "Contribution of the Salmonella gallinarum large plasmid towards virulence in fowl typhoid" Infect. Immun. 55, 388-392. 9. Barrow, P. A. , J.F. Tucker and J.M. Simpson (1987) "Inhibition

of colonisation of the chicken alimentary tract with Salmonella typhimurium by Gram-negative facultatively anaerobic bacteria" Epidemiol. Infect. 98, 311-322.

10. Berchieri, A. Jnr. and P. A. Barrow (1990) "Further studies on the inhibition of colonisation of the chicken alimentary tract with

Salmonella typhimurium by pre-colonisation with an avirulent mutant" Epidemiol. Infect. 104, 427-441.

11. Berchieri, A. Jnr. and P. A. Barrow (1991) "In vitro characterisation of intra-generic inhibition of growth in Salmonella typhimurium" J. Gen. Microbiol. 137, 2147-2153.

12. Calhoun, M. and R.B. Gennis (1993) "Demonstration of separate genetic loci encoding distinct membrane-bound respiratory NADH dehydrogenases in Escherichia coli" J. Bacteriol. 175, 3013-3019.

13. Chen, C.Y. , N.A. Buchmeier, S. Libby, F. Fang, M. Krause and D.G. Guiney (1995) "A central regulatory role for the RpoS sigma factor in expression of Salmonella dublin plasmid virulence genes" J. Bacterial. 177, 5303-5309.

14. Curtiss, R. and S.M. Kelly (1987) "Salmonella typhimurium deletion mutants lacking adenylate cyclase and cyclic AMP receptor protein are avirulent and immunogenic" Infect. Immun. 55, 3035-

3043.

15. Dorman, CJ. and N. Ni Bhriain (1992) "Global regulation of gene expression during environmental adaptation: implications for bacterial padiogens" pages 193-230. In: C.E. Hormaeche, C. . Penn and CJ. Smyth (eds.) Molecular Biology of Bacterial

Infection (Society for General Microbiology Symposium 49), Cambridge University Press, Cambridge, UK.

16. Fang, F.C, S.J. Libby, N.A. Buchmeier, P.C Loewen, J. Switala, J. Harwood and D.G. Guiney (1992) "The alternative sigma factor KatF (RpoS) regulates Salmonella virulence" Proc.

Natl. Acad. Sci. USA 89, 1 1978-11982.

17. Genetics Computer Group (1991) Program Manual for the GCG Package, Version 7 (Genetics Computer Group, Madison, WI).

18. Graham-Smith, G.S. (1920) "The behaviour of bacteria in fluid cultures as indicated by day estimates of the numbers of living organisms" J. Hyg. 19, 133-204.

19. Harrison, J.A. , D. Pickard, A. Khan, S.N. Chatfield, CJ. Dorman, G. Dougan and CE. Hormaeche (1993) "Reduced virulence of Salmonella typhimurium with mutations in global regulatory genes" In: Cabello, F., C. Hormaeche, P. Mastroeni and L. Bonina. Biology of Salmonella Plenum Press, New York.

20. Hormaeche, C.E. (1979) "Natural resistance to Salmonella typhimurium in different inbred mouse strains" Immunology 37, 311-318. 21. Jenkins, D.E. , J.E. Schulz and A. Matin (1988) "Starvation- induced cross-protection against heat or H ₂ 0 ₂ challenge in Escherichia coir J. Bacteriol. 170, 3910-3914.

22. Lange, R. and R. Hengge-Aronis (1991) "Identification of a central regulator of stationary phase gene expression in Escherichia coli" Mol. Microbiol. 5, 49-59.

23. Matin, A. (1990) "Molecular analysis of the starvation stress proteins in Escherichia coli" FEMS Microb. Letters. 74, 185-196.

24. Matin , A ( 1991 ) " The molecular basis of carbon-starvation-induced general resistance in Escherichia coli" Mol. Microbiol. 5, 3-10. 25. Matsushita, K. , T. Ohnishi and H.R. Kaback (1987) "NADH- ubiquinone oxidoreductases of the Escherichia coli aerobic respiratory chain" Biochemistry 26, 7732-7737.

26. McCann, M.P. , J.P. Kidwell and A. Matin (1991) "The putative

(f factor KatF has a central role in development of starvation- mediated general resistance in Escherichia coli" J. Bacteriol. 173,

4188-4194. 27. McCleary, W.R. , J.B. Stock and A.J. Ninfa (1993) "Is acetyl phosphate a global signal in Escherichia coli" J. Bacterial. 175, 2793-2798. 28. Miller, I. , D. Maskell, C Hormaeche, K. Johnson, D. Pickard and G. Dougan (1989) "Isolation of orally attenuated Salmonella typhimurium following TnphoA mutagenesis" Infect. Immun. 57, 2758-2763.

29. Miller, V.L. and J.J. Mekalanos (1984) "Synthesis of cholera toxin is positively regulated at the transcriptional level by toxR" Proc.

Natl. Acad. Sci. USA 81, 3471-3475.

30. O'Callaghan, D. , D. Maskell, F.Y. Liew, CS.F. Easmon and G. Dougan (1988) "Characterisation of aromatic-dependent and purine- dependent Salmonella typhimurium: Studies on attenuation, persistence and ability to induce protective immunity in BAL:B/c mice" Infect. Immunol. 56, 419-423.

31. Penfold, W.J. (1914) "On the nature of bacterial log" J. Hyg. 14, 215-241.

32. Pitcher, D.G. , N.A. Saunders, and R.J. Owen (1988) "Rapid extraction of bacterial genomic DNA with guanidium thiocyanate"

Letters in Appl. Microbiol. 8, 151-156.

33. Reed, L.J. and H. Muench (1938) "A simple method of estimating fifty percent endpoints" Amer. J. Hyg. 27, 493-497.

34. Sambrook, J. , E.F. Fritsch and T. Maniatis (1989) "Molecular cloning" 2nd ed. Cold Spring Harbor Laboratory Press, Cold

Spring Harbor, New York, USA.

35. Schmieger, H. (1972) "Phage P22 mutants with increased or decreased transduction abilities" Mol. Gen. Genet. 119, 75-88.

36. Schultz, J.E. , G.I. Latter and A. Matin (1988) "Differential regulation by cyclic AMP of starvation protein synthesis in

Escherichia coli" J. Bacteriol. 170, 3903-3909. 37. Simon, R. , U. Priefer and A. Puhler (1983) "A broad host range mobilisation system for in vivo genetic engineering: transposon mutagenesis in Gram negative bacteria" Biotechnology 1 , 784-789. 38. Smith, H.W. (1955) "Observations on experimental fowl typhoid"

J. Comp. Pathol. 63, 37-54. 39. Smith, H.W. and J.F. Tucker (1980) "The virulence of Salmonella strains for chickens: their excretion by infected chickens" J. Hyg.

84, 479-488. 40. Weidner, U. , S. Geier, A. Plock, T. Friedrich, H. Lief and H.

Weiss (1993) "The gene locus of the proton-translocating NADH: ubiquinone oxidoreductase in Escherichia coli: organization of the

14 genes and relationship between the derived proteins and subunits of mitochondrial complex I" J. Mol. Biol. 233, 109-122. 41. Zambrano, M.M. and R. Kolter (1993) "Escherichia coli mutants lacking NADH dehydrogenase I have a competitive disadvantage in stationary phase" J. Bacteriol. 175, 5642-5647. 42. Zambrano, M.M. , D.A. Siegele, M. Almiron, A. Tormo and R.

Kolter (1993) "Microbial competition: Escherichia coli mutants that take over stationary phase cultures" Science 259, 1757-1759.