FIELD OF THE INVENTION

The present invention relates to Salmonella mutant strains and their use as a vaccine for preventing Salmonella infection, in particular in eggs.

BACKGROUND OF THE INVENTION Salmonellosis is a worldwide occurring disease caused by bacteria belonging to the genus Salmonella. Salmonella enterica, subspecies enterica, are Gram-negative bacterial pathogens that are comprised of more than 2500 different serovars, of which only a limited number are associated with poultry. Salmonella enterica serovar Enteritidis (S. Enteritidis; SE) and S. Typhimurium are generally accepted as the most important serovars in chickens, with respect to human public health significance. Chickens infected with the aforementioned serovars appear mainly asymptomatic and continue to shed the bacteria for long periods with rare cases of systemic disease, except in young chicks. However, these serovars are regularly associated with human infections, which mostly lead to a self-limiting gastrointestinal disease, and exposure to poultry or poultry products is one of the major risk factors for human infection. Majowicz et al. (2010) estimated in 2009 that 93.8 million cases of gastroenteritis due to Salmonella species occur globally each year, with 155,000 deaths. More than 80 million cases were supposed to be foodborne, and a considerable part of these infections were caused by the serotype Enteritidis and egg consumption. Infection with S. Enteritidis or S. Typhimurium can become severe, requiring antibiotic treatment or even hospitalization. Hence, a massive burden is still placed on both the poultry industry and the healthcare system. In addition, with the emergence of multidrug resistant Salmonella strains, antibiotic treatment for human patients is becoming increasingly difficult. Thus, there is definitely a need for effective measures to control the prevalence of non-host-adapted Salmonella species in poultry flocks. Eggs are a main vehicle for the pathogen that causes spread to humans. Salmonella can be present on the shell surface due to the presence of Salmonella in the hen's environment or passage of the egg through the cloaca. In addition, the bacterium can also be contaminating internal eggs after reproductive tract colonization as a consequence of either shell penetration or colonization of the reproductive tract of laying hens and thus incorporation in the forming egg . In the latter case eggs are a 'box with Salmonella inside' that can't be eliminated using hygienic measures such as egg washing. Several lines of evidence however support the view that egg contamination with SE is more likely to take place during the formation of the egg in the reproductive organs than by eggshell penetration. The egg-associated pandemic reached a maximum in the mid 1990's to the early 2000's. In the European Union, legislation has been responsible for a serious reduction in Salmonella prevalence at laying hen farms, eggs and egg products and as a consequence human infections due to egg consumption. These legislations forced the member states to take action to monitor and control the pathogen, and reduction targets for prevalence have been produced. Over the past two decades, Salmonella control programs were

implemented by the European Union, including that a) that antimicrobials cannot be used to control Salmonella b) that member states with a prevalence of Salmonella Enteritidis in commercial laying hens higher than 10% are mandatory to vaccinate and c) that live vaccines can only be used during rearing. Regulation No. 1237/2007 (Anonymous, 2007) laid down restrictions for the trade of table eggs from flocks infected with Salmonella Enteritidis or Typhimurium. The latter states that eggs from Salmonella Enteritidis or Typhimurium positive flocks must be banned from the market, unless they are treated in a manner that guarantees that all Salmonella bacteria are destroyed.

Despite the decline in human cases, salmonellosis still is the second most commonly reported zoonotic disease, following campylobacteriosis. Although eggs are no longer the primary food vehicle causing salmonellosis, it appears that when one considers the risk related to the different sources weighted according to the tonnage of food available for consumption, the risk of Salmonella infection still remains the highest when consuming table eggs (EFSA, 2013).

Vaccination of chickens, along with other control measures as part of a comprehensive Salmonella control program, is an important strategy in lowering the prevalence of Salmonella. Vaccination of chickens harnesses the immune system of the hosts to decrease the levels of Salmonella species that are associated with poultry flocks upon infection rather than control disease. The Salmonella vaccines that have been tested are divided into three categories: live attenuated, inactivated and subunit vaccines (Desin T et al., 2013). Although some commercially available vaccines are in the killed bacteria form, a few registered S. Enteritidis live vaccines are commercially available for poultry. These live vaccines are developed on the principle of either metabolic drift mutations or auxotrophic double-marker mutants obtained through chemical mutagenesis implicating a higher risk for reverting to virulence (Van Immerseel F et al ., 2013). In addition, commercially available vaccines are developed with the focus on reducing shedding and colonization of host tissues such as spleen, liver and caeca, while it is known that Salmonella colonization in the reproductive tract is generally high and persistent. In several studies, SE was isolated from the reproductive tissue of infected birds, in the absence of intestinal colonization (Lister, 1988). It has been demonstrated that repeated in vivo passages through the reproductive tissues of chickens increase the ability of an SE strain to induce internal egg contamination, whereas serial passage through the liver and the spleen did not affect the ability of the strain to cause egg contamination (Gast et al., 2003). This is an indication that interaction of SE with the reproductive tissues may either induce or select for the expression of microbial properties important for egg contamination. SE is capable of persistence in reproductive tissues of naturally and experimentally infected hens, even though the animals generate an innate and adaptive immune response to the infection, indicating that the bacteria can reside intracellularly and escape the host defense mechanisms (Gantois et al., 2009). The deposition of Salmonella inside eggs is thus most likely a consequence of reproductive tissue colonization in infected laying hens (Keller et al., 1995; Methner et al., 1995; Gast & Holt, 2000a). WO2006129090 shows that vaccination of one-day old chicks with a S. Typhimurium tolC mutant strain results in a reduced shedding of the S. Typhimurium challenge strain together with a reduced colonization of liver and spleen tissues. However, WO2006129090 is completely silent on colonization of the reproductive tract and egg contamination. Studies documenting protection against egg contamination by vaccination of laying hens are limited (Gantois I et al., 2006). The efficacy of live vaccines in poultry has been tested in experimental and field studies but only a few studies have demonstrated a partial protective effect of immunization against egg contamination (Miyamoto T et al., 1999; Woodward MJ et al., 2002; Nassar TJ et al., 1994; Hassan JO et al., 1997; Gantois I et al., 2006).

Hence, although some Salmonella vaccines have been shown to be partially effective in reducing the rate of egg contamination, eggs from vaccinated hens cannot be guaranteed to be Salmonella free. Moreover, vaccine producers only claim a reduction in shedding of the bacteria in the faeces, not a protection against challenge infection or prevention of egg contamination.

The present invention provides a Salmonella vaccine that specifically counters the egg contamination and is not merely focused on the reduction of shedding. A further advantage of the present vaccine strain is that it is easy to administer and there is no risk of reversal to virulence, contrary to some commercial vaccine strains with undefined mutations. SUMMARY OF THE INVENTION

It is an objective of the present invention to provide a Salmonella mutant strain, having at least one genetic modification within the tolC gene or within one or more of the resistance-nodulation-division (RND) genes of the efflux pump system. In particular, the Salmonella mutant strain comprises a genetic modification of the tolC gene or of one or all of the acrAB, acrEF and mdtABC genes. Preferably the genetic modification is an artificially introduced genetic modification, in particular resulting in an inactivation of the gene, and more in particular said modification is a deletion of at least a part of said gene(s), and more in particular of the complete gene(s).

With the objective to obtain Salmonella mutant strains, the tolC and RND gene modifications as defined herein, can be applied in wild type Salmonella serovars. The Salmonella mutant strain as defined and used herein, includes Salmonella enterica and any serotype of the enterica subspecies, and is typically selected from the group consisting of Salmonella Enteritidis (S. Enteritidis), S. Typhimurium, S. Hadar, S. Virchow, S. infantis, S. Kentucky, S. Bredeney, S. Agona, S. paratyphi B and S. Gallinarum. In a more particular embodiment said strain is Salmonella ser. Typhimurium, Salmonella ser. Enteriditis, Salmonella ser. Infantis or Salmonella ser. Gallinarum. It is a further objective of the present invention to provide the use of a Salmonella mutant strain as described herein, in the manufacture of a vaccine and/or for preventing or reducing Salmonella infection in eggs.

In a further embodiment the present invention provides a composition, in particular a vaccine, comprising the Salmonella strain of the invention, and a pharmaceutically acceptable excipient, carrier and/or diluent, and optionally an adjuvant.

A further embodiment provides the Salmonella mutant strain, or the composition of the present invention for use as a medicament. More particular the invention provides the Salmonella mutant strain e.g. as part of a vaccine for use in the prevention or inhibition of Salmonella infection/colonization or a disease caused by such an infection in a subject and/or salmonellosis in humans, and in particular for prevention or (significant) reduction of Salmonella infection in eggs. Another embodiment provides the use of the mutant strain or composition of the present invention in the treatment or prevention of Salmonella infection, in particular for immunization of poultry, especially layer hens, against (disease or symptoms caused by) Salmonella infection.

It is also an object of the present invention to provide a method for treating, preventing, inhibiting and/or reducing the risk of (internal) Salmonella infection in eggs, as well as a method for immunising a subject against Salmonella disease, comprising administering a Salmonella mutant strain or a composition of the present invention, to a subject.

The invention further encompasses a method of producing Salmonella free eggs by immunising laying hens with the Salmonella mutant provided herein. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 : The percentage of Salmonella positive samples in spleen (A), caeca (B), oviduct (C) and ovary (D) in non-vaccinated animals and animals vaccinated at day 1 , week 6 and week 16 with Salmonella Enteritidis 147 AtolC or Salmonella Enteritidis 147 AacrABacrEFmdtABC strains, challenged at 3 weeks post-infection with Salmonella Enteritidis S1400/94, after enrichment. Statistical significant differences (p<0.05) in percentage of positive organ samples between vaccinated groups and the non-vaccinated control group are marked with an asterix. Figure 2: OD values of the ELISA detecting LPS antibodies in the sera of 18 week old laying hens, vaccinated at day 1 , week 4 and week 16 with Salmonella Enteritidis 147 AtolC or Salmonella Enteritidis 147 AacrABacrEFmdtABC. The cut-off OD value is 0.55 and is calculated as the mean obtained from the sera from the Salmonella free chicks (control group) plus five times the standard deviation. Figure 3: Percentage of cloacal swabs positive for the Salmonella Enteritidis and Salmonella Typhimurium tolC or acrABacrEFmdtABC (Δ7) deletion mutant strains after administration of these strains to one day old broilers. Broilers were inoculated with both Salmonella Enteritidis and Salmonella Typhimurium tolC deletion mutant strains or with Salmonella Enteritidis and Salmonella Typhimurium acrABacrEFmdtABC deletion mutant strains on the first day of life. Cloacal swabs were then weekly taken to monitor shedding of these strains.

Figure 4: Spleen colonization by Salmonella Enteritidis and Salmonella Typhimurium tolC or acrABacrEFmdtABC (Δ7) deletion mutant strains after administration to one day old broilers. Broilers were inoculated with both Salmonella Enteritidis and Salmonella Typhimurium tolC deletion mutant strains or with Salmonella Enteritidis and Salmonella Typhimurium acrABacrEFmdtABC deletion mutant strains on the first day of life. Represented values are logio CFU/g sample. Samples were taken on day 7, 21 and 36. The error bars represent the standard error of the means (SEM).

Figure 5: Percentage of spleen and caecum samples positive for Salmonella Enteritidis and Salmonella Typhimurium tolC or acrABacrEFmdtABC (Δ7) deletion mutant strains after enrichment. Broilers were inoculated with both Salmonella Enteritidis and Salmonella Typhimurium tolC deletion mutant strains or with Salmonella Enteritidis and Salmonella Typhimurium acrABacrEFmdtABC deletion mutant strains on the first day of life.

Figure 6: TolC, acrA, acrB, acrD, acrF, acrE, mdsB, mdsA, mdtA, mdtB, and mdtC coding sequences.

Figure 7: Lohman Brown laying hen body weight after oral inoculation with 10 ⁶ CFU of a Salmonella Gallinarum 9R (SG9R) strain or 10 ⁶ CFU of a Salmonella Gallinarum tolC (SG tolC) deletion mutant strain on day 35 of life. Groups treated with either of the strains consisted of 20 animals, and the error bars shown in the figure represent the standard deviation of the mean.

Figure 8: Necrotic foci scores after post-mortem examination of spleen and liver of Lohmann Brown laying hens that were orally inoculated with a Salmonella Gallinarum 9R (SG9R) strain or a Salmonella Gallinarum AtolC (SGtolC) strain. Animals were treated on day 35 of life, liver and spleens were collected and examined on day 63 of life. Necrotic foci scores were determined as described by Matsuda et al. (201 1 ). Necrotic foci scores for the spleen were determined according to the following macroscopic findings: score 0: no foci, score 1 : fewer than 5 foci, score 2: fewer than 20 foci, score 3: greater than 20 foci. Scores for necrotic foci in the liver were determined according to macroscopic findings: score 0: no foci, score 1 : fewer than three foci, score 2: fewer than ten foci, score 3: greater than ten foci. Horizontal bars represent the mean and the standard error of the mean. No statistically significant differences could be observed between both groups.

Figure 9: Lohman Brown laying hen spleen and liver weight after oral inoculation with 10 ⁶ CFU of a Salmonella Gallinarum 9R (SG9R) strain or 10 ⁶ CFU of a Salmonella Gallinarum tolC (SGtolC) deletion mutant strain. Animals were treated on day 35 of life, liver and spleens were collected and weighed on day 63 of life. Horizontal bars represent the mean and the standard error of the mean. No statistically significant differences could be observed between both groups.

Figure 10: Caecal (A & C) and spleen (B & D) colonization by Salmonella Enteritidis (A & B) or Salmonella Typhimurium (C & D) wild-type strains on day 7 of age after experimental infection of broiler chickens treated with a CI culture. The CI culture was administered on day one of life, and consisted of 10 ⁸ CFU of a Salmonella Enteritidis AacrAbacrEFmdtABC strain and 10 ⁸ CFU of a Salmonella Typhimurium

AacrAbacrEFmdtABC strain administered simultaneously by oral gavage. The chickens were experimentally infected on day 2 of life by administering them 10 ⁵ CFU of the respective challenge strain by oral gavage. The values shown represent Iog10 of the CFU/g sample. The horizontal lines represent the mean, the error bars represent the standard error of mean (SEM). The number of samples equals 10 in all groups.

DESCRIPTION OF THE INVENTION The present invention relates to a method of preventing Salmonella infection, in particular Salmonella infection of the reproductive organs (e.g. oviduct, ovary), and even more particular Salmonella infection in eggs. Previous studies demonstrate that the correlation between intestinal colonization and colonization of the reproductive tissue is unclear since it has been shown that Salmonella was isolated from the reproductive tissue of infected birds while being absent in the intestinal organs. Hence, existing Salmonella vaccines focusing on a reduction in shedding of the bacteria in the faeces will not inevitably result in a protection against infection of the reproductive organs, and in particular of (internal) egg contamination.

The invention is based on the finding that vaccines comprising Salmonella bacteria having a genetic modification in tolC gene or in one or more of the resistance- nodulation-division (RND) genes of the efflux pump system are able to promote an effective immune response capable of preventing or reducing subsequent bacterial infection and/or colonisation of the reproductive organs, thereby preventing or reducing vertical transmission to and Salmonella contamination ofthe forming eggs. Furthermore, it was demonstrated that said Salmonella bacteria are not able to infect or colonise the reproductive tract and eggs of the 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 and local immunising stimulus. Hence, the disclosed methods and compositions not only reduce pathogen infection in the bird but remarkably also reduce incidence of pathogen contamination in eggs produced by laying birds/hens.

Efflux pumps are found in almost all bacterial species and genes encoding this class of proteins can be located on chromosomes or plasmids. According to their composition, number of transmembrane spanning regions, energy sources and substrates, bacterial efflux pumps are classified into five families: the resistance- nodulation-division (RND) family, the major facilitator superfamily (MFS), the ATP (adenosine triphosphate)-binding cassette (ABC) superfamily, the small multidrug resistance (SMR) family (a member of the much larger drug/metabolite transporter (DMT) superfamily), and the multidrug and toxic compound extrusion (MATE) family. Except for the RND superfamily which is only found in Gram-negative bacteria, efflux systems of the other four families: MFS, ABC, SMR and MATE are widely distributed in both Gram-positive and negative bacteria. A study by Nishino K. et al. (2006) has shown that S. enterica serovar Typhimurium has nine functional drug efflux pumps (AcrAB, AcrD, AcrEF, MdtABC, MdsAB, EmrAB, MdfA, MdtK and MacAB) (see Figure 1 of Horiyama et al., 2010). These efflux pumps in S. enterica are classified into four families on the basis of sequence similarity: the major facilitator (MF) family (EmrAB and MdfA); the RND family (AcrAB, AcrD, AcrEF, MdtABC and MdsAB); the multidrug and toxic compound extrusion (MATE) family (MdtK); and the ATP-binding cassette (ABC) family (MacAB). TolC is a major outer membrane channel involved in siderophore export and is part of the multidrug resistance pumps (MDR).

The present invention provides mutant strains of Salmonella that are useful as a live or attenuated vaccine for inducing immunological protection against Salmonella, and that are characterized in that they prevent or reduce Salmonella infection and/or colonization of the host tissues in a subject, especially of the reproductive organs, and more in particular in eggs and/or meat. As such, the risk for salmonellosis in humans is reduced or absent. The mutant strains of the present invention are characterized in that they contain at least one genetic modification within the tolC gene or within one or more of the resistance-nodulation-division (RND) genes, i.e. acrAB, acrD, acrEF, mdtABC and mdsAB, and especially the acrAB, acrEF and mdtABC genes. The present invention thus provides a Salmonella strain in which at least one genetic modification within the tolC gene or within one or more of the acrAB, acrEF and mdtABC genes was introduced. In particular, the tolC mutant does not comprise any further artificial genetic modifications within (e.g. deletions of) one or more of the RND genes. In a further embodiment, the acrAB, acrEF and mdtABC mutant comprises an unmodified/complete tolC gene and/or other RND genes.

The "genetic modification" may be an insertion, a deletion, and/or a substitution of one or more nucleotides in said genes. Such a genetic modification results in a (total) decrease in the inherent efflux pump or gene 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. Preferably a bacterium suitable for use in accordance with the invention may, for example, have at least 50% less efflux pump function than comparable wild type bacteria, preferably, at least 75% less efflux pump function, more preferably at least 90% less efflux pump function, and even more preferably 100% less efflux pump function. Mutants with inactivated genes or deletion mutants (of the complete gene or (substantial) part thereof) are preferred. The genetic modifications or mutations may be introduced into the microorganism using any known technique. Preferably, the mutation is a deletion mutation, where disruption of the gene is caused by the excision of nucleic acids. Alternatively, mutations may be introduced by the insertion of nucleic acids or by point mutations. Methods for introducing the mutations into the specific regions will be apparent to the skilled person and are preferably created using the one step inactivation method described by Wanner and Datsenko (2000). Other methods can be applied to achieve a site directed mutagenesis (eg. using suicide plasmids), however the one-step inactivation method is generally accepted as the best and fastest way to achieve a knock-out deletion mutant.

Preferably, the mutants of the present invention contain a deletion of (at least part of) the tolC gene or one or more of the RND genes of the efflux pump system, including the acrA, acrB, acrD, acrF, acrE, mdsB, mdsA, mdtA, mdtB, or mdtC gene. As used herein, the tolC, acrA, acrB, acrD, acrF, acrE, mdsB, mdsA, mdtA, mdtB, and mdtC gene is meant to include any homolog or artificial sequence that is substantially identical, i.e. at least 85%, 90%, 91 %, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, and preferably 100% identical to the corresponding tolC, acrA, acrB, acrD, acrF, acrE, mdsB, mdsA, mdtA, mdtB, and mdtC gene as found in Salmonella enterica subsp. enterica serovar Typhimurium str. LT2 (NCBI: NC_003197.1 Gl:16763390). In said reference sequence the tolC gene is characterized by Gene ID: 1254709 and encodes the TolC outer membrane channel protein. The acrA gene is characterized by Gene ID: 1251996 and encodes the AcrA acridine efflux pump. The acrB gene is characterized by Gene ID: 1251995 and encodes the AcrB RND family acridine efflux pump. The acrD gene is characterized by Gene ID: 1254003 and encodes the AcrD RND family aminoglycoside/multidrug efflux pump. The acrF gene is characterized by Gene ID: 1254914 and encodes the AcrF multidrug efflux protein. The acrE gene is characterized by Gene ID: 1254913 and encodes the AcrE multidrug efflux protein. The mdtA gene is characterized by Gene ID: 1253647 and encodes the MdtA multidrug resistance protein. The mdtB gene is characterized by Gene ID: 1253648 and encodes the MdtB multidrug resistance protein. The mdtC gene is characterized by Gene ID: 1253649 and encodes the MdtC multidrug resistance protein. The mdsA gene is characterized by Gene ID: 1251871 and encodes the MdsA cation efflux system protein. The mdsB gene is characterized by Gene ID: 1251870 and encodes the MdsB cation efflux system protein. The nucleic acid sequences of the tolC, acrA, acrB, acrD, acrF, acrE, mdsB, mdsA, mdtA, mdtB, and mdtC genes are given in Figure 6 (SEQ ID NO: 1 -1 1 ).

The percentage identity of nucleic acid and polypeptide sequences can be calculated using commercially available algorithms which compare a reference sequence with a query sequence. The following programs (provided by the National Center for Biotechnology Information) may be used to determine homologies/identities: BLAST, gapped BLAST, BLASTN and PSI BLAST, which may be used with default parameters.

In one embodiment, the present invention encompasses a Salmonella mutant strain comprising a deletion of the tolC gene, as compared to the corresponding wild type sequence as found in Salmonella enterica subsp. enterica serovar Typhimurium str. LT2 (NCBI: NC_003197.1 Gl:16763390). In a further embodiment, the present invention encompasses a Salmonella mutant strain comprising a deletion of all of the acrAB, acrEF and rndtABC genes, as compared to the corresponding wild type sequence as found in Salmonella enterica subsp. enterica serovar Typhimurium str. LT2 (NCBI: NC_003197.1 Gl:16763390).

Although any serotype of S. enterica may be used to produce the mutant strain, in preferred embodiments, the modifications are inserted into Salmonella serovars most common in poultry, including serovars belonging to serogroup B such as S. Agona, S. Bredeney, S. Paratyphi B, S. Typhimurium, and monophasic strains of S.

Typhimurium; serogroup D such as S. Enteritidis and S. Gallinarum; and serogroup C such as S. Hadar, S. Virchow, S. Infantis, and S. Kentucky. The combination of one or more of the mutant strains in one composition or vaccine is also envisaged by the present invention (e.g. mono-, bi-, tri or tetravaccine).

In a particular embodiment said modification(s) are inserted in a Salmonella spp. selected from the group comprising Salmonella Salmonella enterica subsp. enterica serovar Enteritidis, Salmonella enterica subsp. enterica serovar Typhimurium or Salmonella enterica subsp. enterica serovar Infantis. Salmonella enterica subsp. enterica serovar Enteritidis is a serovar of the D1 serogroup. S. Enteritidis is the most common serovar in the United States and Europe. Salmonella enterica subsp. enterica serovar Typhimuriunn is a serovar of the B serogroup. S. Typhimuriunn is a widely distributed serovar, which represent the second most common serovar isolated from humans in the United States and Europe. Salmonella enterica subsp. enterica serovar Infantis is a serovar of the C1 serogroup. S. Infantis is commonly found in chickens and broiler flocks.

A "subject" as used herein includes a human or an animal, in particular birds, more in particular poultry, and even more in particular chickens, especially laying hens (layers), breeders and/or broilers.

"Laying hen" or "egg-laying hen" is a common term for a female chicken that is kept primarily for laying eggs and includes young chickens that are reared for egg production. Some chickens are raised for meat (called "broiler" chickens), while others are primarily for eggs (used for human consumption). Raising laying hens is a different process than raising chickens for meat. Broiler chickens typically take less than six weeks to reach slaughter size while most laying hens are kept for one to three laying cycles (up to 200 weeks) before they are replaced with a new flock. Layers typically start laying eggs around 20 weeks of age. Layer feeds are formulated for chickens laying table eggs (those used for human consumption). Broiler feeds are formulated for those chickens producing hatching eggs ("breeders"). The diets are basically the same, but the breeder diets typically have slightly more protein and are fortified with extra vitamins for proper embryo development.

In a specific embodiment, the Salmonella mutant strains of the present invention are used as attenuated live vaccines. It is well established that live attenuated micro- organisms are highly effective vaccines; immune responses elicited by such vaccines are often of greater magnitude and of longer duration than those produced by non- replicating immunogens. One explanation for this may be that live attenuated strains establish limited infections in the host and mimic the early stages of natural infection. In addition, unlike killed preparations, live vaccines are often more potent in inducing mucosal immune responses and cell-mediated responses, which may be connected with their ability to replicate in epithelial cells and antigen-presenting cells, such as macrophages, respectively. However, concerns remain over the safety of using live- attenuated vaccines. There may also be a risk of the attenuated strain reverting to virulence, thus having the potential to cause disease and abortion in the vaccinated animal. However, it was demonstrated that the mutant strains of the present invention are safe (no clinical symptoms and not persistently colonizing the host) and do not revert to virulence. It is an object of the present invention to provide the use of the Salmonella mutant strains of the present invention for preparing a medicament which is employed for the prophylactic and/or therapeutic treatment of Salmonella infection in animals, in particular poultry, more particular chickens, and even more particular in layers. In a preferred embodiment the present invention provides the mutant strains of Salmonella as defined herein for use as a medicament. In particular, the present invention encompasses the (use of the) mutant strains of Salmonella as described herein for use in protecting against egg contamination. Hens' eggs produced by the immunized hens are substantially free from Salmonella. Remarkably, the present mutant strains have been shown to significantly reduce colonization of the reproductive organs. The oviduct can be subdivided into five functional regions. Starting from the ovary, there are the infundibulum, magnum, isthmus, uterus and vagina. The infundibulum captures the ovulatory follicles, the magnum produces the albumen, the isthmus deposits the eggshell membranes, the uterus forms the eggshell and the vagina is involved in oviposition. Salmonella colonizing the oviduct could be incorporated into the albumen, the eggshell membranes or the eggshell itself, depending on the site of colonization (magnum, isthmus and uterus, respectively). Although SE has been isolated from both the yolk and the albumen, according to several studies, the albumen is most frequently contaminated, pointing to the oviduct tissue as the colonization site. However, some studies found the yolk to be primarily contaminated, suggesting the ovary to be the primary colonization site (Gantois et al., 2009). It is thus an aim of the invention to provide Salmonella mutants strains for use in preventing or reducing colonization/infection of the oviduct tissues and/or the ovary.

In a further embodiment, the Salmonella mutant strains are used to manufacture a (pharmaceutical) composition, in particular a vaccine, which may be administered to the subject via the parenteral, mucosal or oral route. Live vaccines can be produced using art known procedures and typically include a (pharmaceutically) acceptable excipient, carrier or diluent, and optionally an adjuvant. The present invention provides a pharmaceutical composition or a vaccine against Salmonella egg infection comprising:

- one or more of the mutant strains according to the invention; and

- a pharmaceutically acceptable carrier or diluent.

The particular pharmaceutically acceptable carriers or diluents employed are not critical to the present invention, and are conventional in the art. Examples of diluents include: buffer for buffering against gastric acid in the stomach, such as citrate buffer (pH 7.0) containing sucrose, bicarbonate buffer (pH 7.0) alone, or bicarbonate buffer (pH 7.0) containing ascorbic acid, lactose, and optionally aspartame. Examples of carriers include: proteins, e.g., as found in skimmed milk; sugars, e.g. sucrose; or polyvinylpyrrolidone.

The particular adjuvants employed are not critical to the present invention, and are conventional in the art. Examples of adjuvants include, but are not limited to, tensoactive compounds (such as Quil A), mineral salts (such as aluminium hydroxide), micro-organism derived adjuvants (such as muramyl dipeptide), oil-in- water and water-in-oil emulsions (such as Freund's incomplete adjuvant), particulate antigen delivery systems (such as liposomes, polymeric atmospheres, nanobeads, ISCOMs and ISCOMATRIX), polysaccharides (such as micro-particulate inulin), nucleic acid based adjuvants (such as CpG motivs), cytokines (such as interleukins and interferons), activators of Toll-like receptors and eurocine L3 en N3 adjuvantia. As is known to the skilled person, the dose or amount varies according to the route of administration. Those skilled in the art may find that the effective (immunizing) dose for a vaccine administered parenterally may be smaller than a similar vaccine which is administered via drinking water, and the like. The number of microorganisms that are required to be present in the formulations can be determined and optimised by the skilled person. However, in general, a subject may be administered approximately 10 ⁴ -10 ¹⁰ colony-forming units (CFUs), preferably between 10 ⁵ -10 ⁹ CFUs in a single dosage unit, and more preferably between 10 ⁷ -10 ⁹ CFUs in a single dosage unit.

As already mentioned hereinbefore, the mutant strains and vaccine compositions of the present invention may be prepared by known techniques.

The choice of particular Salmonella enterica microorganism, can be made by the skilled person without undue experimentation. A preferred microorganism is selected from the group consisting of Salmonella Enteritidis (S. Enteritidis), S. Typhimurium, S. Hadar, S. Virchow, S. Infantis, S. Kentucky, S. Bredeney, S. Agona, S. paratyphi B and S. Gallinarum. In one embodiment the microorganism is Salmonella Typhimurium; more in particular the Salmonella Typhimurium strain 1 12910a (Van Parys et al., 2012; De Cort et al., 2014). In a further embodiment the microorganism is Salmonella Enteritidis; more particular the Salmonella Enteritidis strain 147 (Methner et al., 1995; Bohez et al., 2008; De Cort et al., 2013). In an even further embodiment, the microorganism is either Salmonella Infantis or Salmonella Gallinarum. In a particular embodiment of the present invention, the mutant strains are either tolC deletion mutants, or acrABacrEFmdtABC deletion mutants of Salmonella Typhimurium strain 1 12910a, or of Salmonella Enteritidis strain 147.

The Salmonella mutant strains as described herein are especially useful as vaccines, in particular (for use in a method in order) to prevent or (significantly) reduce Salmonella infection and/or colonization of the host tissue and/or whereby said mutant strain is capable of preventing or reducing (internal) egg contamination. A further embodiment provides the composition or vaccine of the present invention for use in the immunization of chickens, especially layers and broilers, against Salmonella infection. It is also an object of the present invention to provide a method for treating, reducing or preventing a Salmonella infection, comprising administering a Salmonella mutant strain as provided herein or a composition or vaccine of the present invention, to a subject in need thereof.

Furthermore, the invention is directed to reduce or prevent salmonellosis (e.g. gastroenteritis, vomiting, fever) in humans by the use of the Salmonella mutant strain and the methods as provided herein. In other words, the invention relates to the use of the Salmonella mutants strain for preventing or reducing egg contamination, e.g. by immunising or vaccinating hens in order to reduce colonization of the reproductive tissue. By such method, the contamination of the eggs is limited or absent and hence also the risk of salmonellosis and/or the number of food borne Salmonella infections in humans. Hence the method of the invention is especially useful to produce

Salmonella free eggs.

In a particular embodiment, the Salmonella Gallinarum mutant strain is characterized in that it contains at least one genetic modification within the tolC gene or within one or more of the resistance-nodulation-division (RND) genes, i.e. acrAB, acrD, acrEF, mdtABC and mdsAB, and especially the acrAB, acrEF and mdtABC genes. Said S. Gallinarum mutant strain is especially useful in protecting layers or broilers against fowl typhoid, a severe septicaemic disease, in particular against clinical disease and internal organ colonization by S. Gallinarum. Clinical symptoms include anorexia, diarrhea, anemia, a decreased laying percentage but the major issue is the high mortality it can induce in both chicks and adult hens

It is recognized that administration of an effective (immunizing) 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 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 dose). The use of multiple administrations of vaccines (for example a primary dose followed by one, two or more booster doses) is well known, particularly in the context of live vaccines, and is hence an embodiment of the present invention.

Oral administration of the strains or compositions of the invention may be achieved by inoculation (such as by oral gavage) or by application in drinking water. In one embodiment, the invention relates to (poultry) food comprising the Salmonella mutant(s) as described herein. As an alternative to their oral administration, suitably formulated strains or compositions may be administered to a subject by means of injection. In particular, strains or compositions 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.

Strains or compositions 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 sterilizing 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 also been shown that effective immunizing dosages of vaccines may be administered to poultry through the use of whole body sprays. Aerosol immunization 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.

The mutant strains as provided herein can be part of a vaccination kit comprising a dispensing device and an (immunologically) effective amount of the vaccine strain. The dispensing device is preferably adapted for spray, aerosol delivery or ocular eye drops. The invention will be described in further details in the following examples and embodiments by reference to the enclosed drawings. Particular embodiments and examples are not in any way intended to limit the scope of the invention as claimed. The rationale of the examples given here for the serotype S. Enteritids are equally well applicable to other Salmonella enterica serotypes infecting poultry, such as for example S. Typhimurium, S. Hadar, S. Virchow, S. Infantis, S. Kentucky, S. Bredeney, S. Agona, S. Paratyphi B and S. Gallinarum.

EXAMPLES

Example 1 : Prevention of egg contamination by Salmonella Enteritidis after oral vaccination of laying hens with Salmonella Enteritidis AtolC and AacrABacrEFmdtABC mutants

MATERIALS AND METHODS Vaccine and challenge strains

The vaccine strains AtolC and AacrABacrEFmdtABC are defined mutants of Salmonella Enteritidis 147 phage type 4. The wild type strain 147 was originally isolated from egg white and is resistant to streptomycin. The strain is known to colonize the gut and internal organs to a high level (Methner, al-Shabibi et al. 1995, Bohez, Dewulf et al. 2008). All mutations were constructed according to the one step inactivation method previously described by Datsenko and Wanner (Datsenko and Wanner, 2000).

The challenge and vaccine strains were incubated overnight with gentle agitation at 37°C in Luria Bertani (LB) medium (Sigma, ST. Louis, MO, USA). To determine bacterial titers, ten-fold dilutions were plated on brilliant green agar (BGA, Oxford, Basingstoke, Hampshire, UK) for the challenge strain. The vaccine strains were plated on LB supplemented with 1 % lactose, 1 % phenol red and 100 g/ml streptomycin to determine the titer. The vaccine and challenge strains were diluted in HBSS (Hanks Balanced Salt Solution, Invitrogen, Paisley, England) to 10 ⁸ cfu/ml.

Experimental birds

Ninety (90) day-old Lohmann Brown laying hens (De Biest, Kruishoutem, Belgium) were randomly divided into 3 groups and housed in separate units. The lighting program provided by the commercial supplier was implemented. Commercial feed and drinking water was provided ad libitum. The animal experiment in this study followed the institutional guidelines for the care and use of laboratory animals and was approved by the Ethical Committee of the Faculty of Veterinary Medicine, Ghent University, Belgium (EC2013/135). Euthanasia was performed with an overdose of sodium pentobarbital in the wing vein.

Experimental setup

Two different groups (n=30) were orally immunized at day of birth, at 6 weeks of age and at 16 weeks of age through crop instillation of 0.5 ml containing 10 ⁸ cfu of Salmonella Enteritidis 147 AtolC (group 1 ) or Salmonella Enteritidis 147 AacrABacrEFmdtABC (group 2). A third group of birds (n=30) was kept as non- immunized but Salmonella challenged positive controls (group 4). At the age of 18 weeks, serum samples were taken for quantification of Enteritidis and Typhimurium antibodies in an LPS-ELISA (Desmidt, Ducatelle et al . 1996). At the same time, cloacal swabs were taken in each group and bacteriologically analyzed for the presence of the vaccine strains. At 21 weeks of age, all the hens were in lay and eggs were collected daily during 3 weeks for bacteriological detection of the vaccine strain in the egg content. At 24 weeks of age, all the animals were intravenously inoculated in the wing vein with 0.5 ml containing 5 x 10 ⁷ cfu of the Salmonella Enteritidis strain S1400/94. This protocol was already used to produce high levels of internal egg contamination (De Buck, Van Immerseel et al . 2004, Gantois, Ducatelle et al. 2006). The eggs were collected daily during 3 weeks and analyzed for the presence of the challenge strain. Three weeks after challenge inoculation, all the animals were euthanized by an overdose of pentobarbital in the wing vein. Samples of the spleen, oviduct, ovary, uterus and caecum were aseptically removed for bacteriological quantification of challenge and vaccine strain bacteria. ELISA to quantify anti-LPS antibodies

Serum samples taken at week 18 were analyzed for the levels of

LPS antibodies using a previously described indirect ELISA protocol (Desmidt et al., 1996). Four 96 well-plates (Sigma, St. Louis, MO, USA) were coated with 100μΙ of an LPS solution (10pg/nnl) in 0.05M carbonate-bicarbonate (pH 9.6; coating buffer) and incubated for 24 hours at 4°C. The LPS was purified from Salmonella Enteritidis PT4, strain 76Sa88 and Salmonella Typhimurium, strain 742Sa91 . The plates were rinsed four times with phosphate buffered saline (PBS, Sigma, St. Louis, MO, USA) supplemented with 0.1 % Tween-20 (Sigma, St. Louis, MO, USA; washing buffer) between each step. In the first step, 100 μΙ PBS (Sigma, St. Louis, MO, USA) supplemented with 1 % bovine serum albumin (BSA, Sigma, St. Louis, MO, USA; blocking buffer) was added to the wells for one hour at 37°C. The blocking buffer was then removed. Secondly, serum samples of animals from the different groups were diluted in blocking buffer (1 :200) and added to the plates (100μΙ). As a negative control, serum from a Salmonella free chicken was used. Serum from a chicken that had been infected experimentally with Salmonella Enteritidis PT4, strain 76Sa88, was used as a positive control. The plates were allowed to shake for 2 hours at 37°C. Thirdly, peroxidase-labelled rabbit anti-chicken IgG (100μΙ, Sigma, St. Louis, MO, USA) was diluted (1 :2000) in blocking buffer and added to the wells for 1 hour and 30 min while shaking at 37°C. Finally 50 μΙ of TMB substrate (Fisher Scientific, Erembodegem, Belgium) was added to the wells. When a blue color started to appear the reaction was blocked with 50μΙ of sulfuric acid (0.5M). The absorbance was measured by the ELISA reader at 450nm. Every sample was analyzed in duplicate The cut-off OD value was calculated as the mean obtained from the sera from the Salmonella free chicks (the non-vaccinated birds) plus five times the standard deviation (OD = 0.55).

Bacteriological examination of the challenged birds

The cloacal swabs taken at week 18 were incubated overnight at 37°C in buffered peptone water (BPW, Oxoid, Basingstoke, Hampshire, UK). Afterwards a loopful was plated on LB plates supplemented with 1 % lactose, 1 % phenol red and either 100pg/nnl streptomycin (Sigma, St. Lous, MO, USA) for the detection of the Salmonella Enteritidis 147 Atol C and AacrABacrEFmdtABC vaccine strains.

Samples of caecum, spleen, ovary, oviduct and uterus were homogenized in BPW (10% weight/volume suspensions) and 10-fold dilutions were made in HBSS (Invitrogen, Paisley, England). Six droplets of 20 μΙ of each dilution were plated on BGA (for quantification of the challenge strain) or on LB supplemented with 1 % lactose, 1 % phenol red and the appropriate antibiotics (for quantification of the vaccines). After overnight incubation at 37°C, the number of cfu/g tissue was determined by counting the number of bacterial colonies for the appropriate dilution. Samples that tested negative after direct plating for the challenge strain were pre- enriched in tetrathionate brilliant green broth (Oxoid, Basingstoke, UK) by overnight incubation at 37°C. After incubation, a loopful of the tetrathionate brilliant green broth was plated on BGA.

Egg production and bacteriological examination of eggs

Eggs were collected daily for 6 weeks from week 18 onwards and the egg production was determined. Each day, eggs of six chicks per group were pooled in one batch, yielding an egg per batch number that varied between one and six. Upon collection, lugol solution and 95% ethanol were subsequently used to decontaminate the surface of the eggshell . After decontamination of the eggshell, the eggs were broken aseptically and the total content of the eggs was pooled and homogenized per batch. A volume of 40 ml of BPW was added for each egg to the pooled egg content and incubated for 48h at 37°C. To detect the vaccine strains, a loopful of the BPW broth was plated on LB plates supplemented with 1 % lactose, 1 % phenol red and 100pg/nnl streptomycin. To detect the challenge strain, a loopful of the BPW broth was plated on BGA. Additionally, further enrichment was done overnight at 37°C in tetrathionate brilliant green broth and after incubation, a loopful of broth culture was streaked onto BGA.

Statistical analysis

GraphPad Prism 5 software was used for statistical analysis. Data of cfu Salmonella/gram tissue of the caecum, spleen, ovary, oviduct and uterus were log- transformed and analyzed by an anova test followed by a Dunnet post hoc test to determine differences between the groups. After enrichment samples were classified as either positive or negative. A Fisher's exact test was used to determine significant differences. Cloacal swabs and batches of eggs were categorized as either positive or negative. As such a Fisher's exact test was also done to determine significant differences. For all tests, differences with p-values below 0.05 were considered to be statistically significant. RESULTS

Detection of anti-Salmonella LPS antibodies in serum

Data derived from the LPS-ELISA show that 26/30 and 19/30 chicks contained anti- Salmonella LPS antibodies in the group of animals vaccinated with the Salmonella Enteritidis 147 AtolC and Salmonella Enteritidis 147 AacrABacrEFmdtABC strain, respectively (Figure 2).

Analysis of cloacal swabs and eggs for the presence of vaccine strains

No cloacal swabs were found positive in the groups vaccinated with the Salmonella Enteritidis 147 AtolC and Salmonella Enteritidis 147 AacrABacrEFmdtABC strains. No swabs were positive in the non-vaccinated control group. None of the vaccine strain was isolated from the egg content samples.

Clinical signs and egg production after challenge

Over the whole experiment, there was no reduction in feed and water intake in either of the groups. The egg production rate after infection in the unvaccinated control group dropped to 59% in the first week post-infection (pi) and raised to 75% and 86% in the second and third week pi. The egg production rate also decreased in the vaccinated groups. No significant differences were detected. The egg production percentages in the group vaccinated with the AtolC strain was 60%, 100% and 90%, and 56%, 70%, 68% for the AacrABacrEFmdtABC strain in the first, second and third week pi respectively. Some eggs were thin-shelled and malformed during the first week of infection. At the end of the experiment 1 1 chicks died in the group of animals vaccinated with the Salmonella Enteritidis 147 AacrABacrEFmdtABC strain because of cannibalism.

Isolation of the challenge strain from egg contents

The non-vaccinated hens laid significantly more Salmonella positive eggs compared to the vaccinated animals during the whole 3-week follow-up period. Three egg batches were Salmonella positive in the control group while the batches from the vaccine strains were negative after direct plating. Not a single positive egg batch was detected for animals vaccinated with the Salmonella Enteritidis 147 AtolC and Salmonella Enteritidis 147 AacrABacrEFmdtABC strains. No positive egg batches were found in the third week pi.

Table 1 : The percentage of egg content batches positive for the challenge strain Salmonella Enteritidis S1400/94 in non-vaccinated animals and animals vaccinated at day 1 , week 6 and week 16 with Salmonella Enteritidis 147 AtolC or Salmonella Enteritidis 147 AacrABacrEFmdtABC strains, during the two weeks following infection. Results are shown after incubation of the egg content in BPW (48h, 37°C). Results between brackets show the percentage of batches positive after enrichment in tetrathionate brilliant green broth (37°C, overnight). Different superscripts indicate significant differences between the groups (p<0.05).

Group Week 1 Week 2

Non-vaccinated 70 ^a (74 ^a ) 0(1 If

AtolC 0 ^C (0) ^C 0(0) ^c

AacrABacrEFmdtABC 0 ^C (0) ^C 0(0) ^c Isolation of the challenge strain from the organs at 3 weeks post-infection

No samples were positive at direct plating. No significant differences in Salmonella colonization were seen for the uterus (data not shown). Figure 1 presents the percentage of Salmonella positive samples in the spleen, caeca, oviduct and ovary in non-vaccinated animals and animals vaccinated at day 1 , week 6 and week 16 with either the Salmonella Enteritidis 147 AtolC or the Salmonella Enteritidis 147 AacrABacrEFmdtABC strains, at 3 weeks pi with Salmonella Enteritidis S1400/94 after enrichment. Vaccination with the Salmonella Enteritidis 147 AtolC and AacrABacrEFmdtABC strain both significantly decreased the number of Salmonella positive samples in the spleen, caeca, oviduct and ovary against the control group. Additionally in the AacrABacrEFmdtABC vaccinated group, the number of Salmonella positive samples in the oviduct was significantly lower than the group vaccinated with AtolC. Example 2: A Salmonella Enteritidis and Salmonella Typhimurium tolC and acrABacrEFmdtABC deletion mutant are safe for use as live vaccine strains in broilers

MATERIAL & METHODS Chickens

One-day-old Ross broiler chickens were obtained from a local hatchery and housed in isolation. Experimental groups were housed in separate rooms in containers on wood shavings. Commercial feed and drinking water were provided ad libitum. Experiments were performed with the permission of the Ethical Committee of the Faculty of Veterinary Medicine, Ghent University, Belgium.

Vaccine strains

Salmonella Enteritidis 147 Strep ^R (SE147) is a well-characterized strain originally isolated from egg white and was used for the production of the deletion mutants (Methner et al. 1995; Methner et al. 1995; Bohez et al . 2008). A spontaneous nalidixic acid-resistant mutant of Salmonella Typhimurium strain 1 12910a, originally isolated from a pig stool sample (Van Parys et al. 2012), was used for the production of the other deletion mutants. This antibiotic resistance has previously been shown to have no impact on the in vivo results (Barrow et al. 1987). Deletion of the tolC gene or the acrAB, acREF and mdtABC genes was done using the one-step inactivation method described by Datsenko and Wanner (Datsenko and Wanner 2000; Bohez et al. 2006). This yielded a Salmonella Enteritidis Strep ^R tolC deletion mutant, a Salmonella Enteritidis 147 Strep ^R acrAbacrEFmdtABC deletion mutant, a Salmonella Typhimurium Nal ^R tolC deletion mutant and a Salmonella Typhimurium Nal ^R acrAbacrEFmdtABC deletion mutant.

Experimental design

Analysis of the colonisation pattern of Salmonella Enteritidis and Salmonella Typhimurium AtolC or AacrABacrEFmdtABC mutant strains in broilers: evaluation of safety. One hundred and twenty one-day-old chicks were divided into 2 groups of 60 and each housed in a container of 2,4 m ² . One group was given 0.5 ml of a mixture containing 2 x 10 ⁸ CFU/ml of the Salmonella Enteritidis AtolC strain and 2 x 10 ⁸ CFU/ml of the Salmonella Typhimurium AtolC strain by oral gavage. The other group was given 0.5 ml of a mixture containing 2 x 10 ⁸ CFU/ml of the Salmonella Enteritidis AacrAbacrEFmdtABC strain and 2 x 10 ⁸ CFU/ml of the Salmonella Typhimurium AacrAbacrEFmdtABC strain by oral gavage. To evaluate colonisation by the deletion mutant strains, their numbers in caecum and spleen were determined for 20 animals at days 7, 21 and 36. Shedding of the strains was evaluated during the experiment by bacteriological analysis of cloacal swabs taken on days 2, 9, 16, 23 and 30.

Bacteriological analysis

Cloacal swabs were directly inoculated on Lysogeny Broth (LB) plates with 20 g/ml nalidixic acid (Sigma-Aldrich, St. Louis, MO, USA) or 100 g/ml streptomycin (Sigma- Aldrich, St. Louis, MO, USA). Samples negative after direct inoculation were pre- enriched in buffered peptone water (BPW, Oxoid, Basingstoke, England) and incubated overnight at 37 °C. One ml of this suspension was further enriched by adding 9 ml tetrathionate-brilliant green broth (Merck, Darmstadt, Germany). After overnight incubation at 37 °C, this suspension was plated on LB plates supplemented with the appropriate antibiotic. Samples of caecum and spleen were homogenized in BPW and 10-fold dilutions were made in HBSS. Six droplets of 20 μΙ of each dilution were plated on LB plates supplemented with 20 g/ml nalidixic acid or 100 g/ml streptomycin. After overnight incubation at 37 °C, the number of CFU/g tissue was determined by counting the number of bacterial colonies on the plates. Negative samples were enriched as described above. RESULTS

Administration of the Salmonella Enteritidis and the Salmonella Typhimurium tolC deletion mutants and the Salmonella Enteritidis and the Salmonella Typhimurium acrABacrEFmdtABC deletion mutants to one day old broilers did not induce clinical symptoms in the animals. In the group treated with the Salmonella Enteritidis and the Salmonella Typhimurium tolC deletion mutants 2 animals died, while in the group treated with the Salmonella Enteritidis and the Salmonella Typhimurium acrABacrEFmdtABC deletion mutants 5 animals died. This does not differ significantly from average mortality (5%) when rearing broilers. (GraphPad Prism 5 software was used for statistical analysis. A Fisher's exact test (one-sided) was used to analyse mortality rates within differently treated groups.)

As shown in Figure 3, nearly all cloacal swabs taken one day after inoculation were positive. However, shedding declined quickly with only a limited number of animals shedding the tolC deletion strains on day 16, and no animals were shedding any of the deletion mutant strains from day 23 onwards.

None of the strains could be detected in the caecum after direct plating on day 7, 21 or 35. In the spleen however, the tolC and the acrABacrEFmdtABC deletion mutant strains colonized the spleen on day 7, and the acrABacrEFmdtABC deletion mutant strains still colonized the spleen on day 21 (Figure 4). However, by slaughter age (earliest at day 36), the Salmonella Enteritidis and Salmonella Typhimurium tolC and the Salmonella Enteritidis and Salmonella Typhimurium acrABacrEFmdtABC deletion mutant strains could no longer be found in the spleen or caecum. Enrichment of caecum and spleen samples confirmed these findings (Figure 5), as both the tolC and acrABacrEFmdtABC deletion mutant strains could be found in the spleens of a high percentage of the animals on day 7, and the acrABacrEFmdtABC deletion mutant strains still colonized the spleen on day 21 . However, by day 36, none of the strains could still be found in the spleens of any of the animals. In addition, the tolC and acrABacrEFmdtABC deletion mutant strains could only be found in a small number of the caeca after enrichment, and there were no caeca positive for any of the deletion mutant strains at slaughter age.

These results indicate that both the Salmonella Enteritidis and the Salmonella Typhimurium tolC deletion mutants and the Salmonella Enteritidis and the Salmonella Typhimurium acrABacrEFmdtABC deletion mutants are safe for use in broilers, and that they are cleared by slaughter age. As a consequence, these strains can thus be used as live vaccine strains in broilers.

Example 3: Evaluation of the safety of a Salmonella Gallinarum tolC deletion mutant strain for use as a vaccine strain offering protection against Salmonella Gallinarum infections in poultry

MATERIAL & METHODS

Chickens

One-day-old Lohmann Brown laying hens were obtained from a local hatchery and housed in isolation. Experimental groups were housed in separate rooms in containers of 2.4 m ² on wood shavings. Commercial feed and drinking water were provided ad libitum. Experiments were performed with the permission of the Ethical Committee of the Faculty of Veterinary Medicine, Ghent University, Belgium.

Vaccine strains

Salmonella Gallinarum strain 9 (SG9) was used for the production of the deletion mutants. This strain was originally isolated in the United Kingdom (Van Immerseel et ai, 2013). Deletion of the tolC gene was done using the one-step inactivation method described by Datsenko and Wanner (Datsenko and Wanner, 2000; Bohez et ai, 2006). This yielded a Salmonella Gallinarum tolC deletion mutant. In addition, Salmonella Gallinarum 9R (SG9R) was used in this study as well. This strain is frequently used in practice to control Salmonella Gallinarum infections in poultry (Van Immerseel et ai, 2013), and was used as a control strain to compare the Salmonella Gallinarum tolC deletion mutant strain to. Experimental design

Forty one-day old laying hens were randomly divided into two groups of twenty chickens and housed in separate rooms. They were reared for 5 weeks, until the animals were 35 days old. On day 35 of life, all animals in the first group were orally inoculated with 1 ml of a mixture containing 10 ⁶ CFU/ml of a Salmonella Gallinarum tolC deletion mutant. The other group was orally inoculated with 1 ml of a mixture containing 10 ⁶ CFU/ml of the SG9R strain. The weight of the animals was monitored for four weeks, and when the animals were 63 days old (9 weeks) samples were taken from liver and spleen to evaluate colonisation by the Salmonella Gallinarum tolC deletion mutant strain and the SG9R strain through bacteriological analysis. Post-mortem examination of liver and spleen was performed as well, scoring enlargement and necrotic foci in liver and spleen as described by Matsuda et al. (Matsuda et al., 201 1 ). In addition, the weight of livers and spleens was determined as well.

Bacteriological analysis

Samples of liver and spleen were homogenized in buffered peptone water (BPW, Oxoid, Basingstoke, England) and 10-fold dilutions were made in Hank's Balanced Salt Solution (HBSS, Invitrogen, Paisley, England). Six droplets of 20 μΙ of each dilution were plated on Lysogeny Broth (LB) plates supplemented with 20 g/ml nalidixic acid or 100 g/ml streptomycin. After overnight incubation at 37 °C, the number of CFU/g tissue was determined by counting the number of bacterial colonies on the plates. Negative samples were enriched by adding 9 ml tetrathionate-brilliant green broth (Merck, Darmstadt, Germany) to one ml of the samples homogenized in BPW. After overnight incubation at 37 °C, this suspension was plated on LB plates supplemented with the appropriate antibiotic.

Statistical analysis

GraphPad Prism software (Version 5.0, GraphPad Software Inc., La Jolla, CA) was used for statistical analysis of the obtained data. A Mann-Whitney test was used to analyse the difference in weight and the difference in enlargement scores and necrotic foci scores between the two groups. RESULTS

Administration of the Salmonella Gallinarum tolC deletion mutant or the Salmonella Gallinarum 9R strain to 5 week old laying hens did not induce clinical symptoms in the animals. No animals died during the experiment, indicating that both strains are severely attenuated when compared to wild-type Salmonella Gallinarum strains.

The average body weight of the laying hens before vaccination (on day 35) did not differ significantly between the groups (Figure 7). After oral inoculation with either a SG9R strain or a Salmonella Gallinarum tolC deletion mutant, no statistical significant differences could be observed between both groups during the experiment, except on day 51 when there was a statistically significant difference between the two groups. However, this difference was most probably due to the animals being reared in separate rooms, as this was the only day a difference could be observed and the average weight of the animals in the group treated with the Salmonella Gallinarum AtolC strain tented to be lower throughout the entire experiment, even prior to treatment.

When comparing the necrotic foci scores for the spleen between the differently treated groups, no statistically significant difference could be observed between both groups (Figure 8). In both groups, only one liver had more than ten foci. For no other livers necrotic foci could be observed. As such, there was no statistically significant difference between the two groups for liver necrotic foci score (Figure 8).

No statistically significant differences could be observed when comparing the average weight of livers and spleens of laying hens treated with a SG9R or a Salmonella Gallinarum AtolC strain (Figure 9). In addition, no statistically significant differences could be observed when comparing the enlargement scores of liver and spleen. All spleens and livers in the group treated with the SG9R strain received a score equal to zero, while one spleen received a score equal to one, and one spleen a score of two in the Salmonella Gallinarum AtolC strain treated group. One liver in the group treated with the Salmonella Gallinarum AtolC strain received a score equal to two. However, when comparing the two groups, these differences were not statistically significant. The SG9R and the Salmonella Gallinarunn tolC deletion mutant strain could not be detected in liver or spleen after bacteriological analysis of the samples, even after enrichment, indicating that they are cleared from vaccinated laying hens within 4 weeks after administration if the strains are administered on day 35 of life.

These results indicate that the Salmonella Gallinarunn tolC deletion mutant is an at least as safe vaccine strain as the commonly used SG9R strain, as there were no statistically significant differences in remaining virulence between both strains. As a consequence, the Salmonella Gallinarunn AtolC strain can be used as a live vaccine strain in laying hens.

Example 4: Protection offered by a culture consisting of Salmonella Enteritidis and Salmonella Typhimurium AacrABacrEFmdtABC mutant strains against experimental Salmonella Enteritidis and Typhimurium infection in broilers: evaluation of efficacy

MATERIAL & METHODS

Chickens

One-day-old Ross 308 broiler chickens were obtained from a local hatchery and housed in isolation. Experimental groups were housed in separate rooms in containers on wood shavings, while commercial feed and drinking water were provided ad libitum. The chickens were examined daily for clinical symptoms following inoculation with Salmonella strains. Experiments were performed with the permission of the Ethical Committee of the Faculty of Veterinary Medicine, Ghent University, Belgium.

Salmonella strains

Salmonella Enteritidis 147 strepR (SE147) is a well-characterized strain originally isolated from egg white and was used for the production of the deletion mutants (Methner et al., 1995a; b; Bohez et al., 2008). A spontaneous nalidixic acid-resistant mutant of Salmonella Typhimurium strain 1 12910a, originally isolated from a pig stool sample (Van Parys et al., 2012), was used for the production of the Salmonella Typhimurium deletion mutants. This antibiotic resistance has previously been shown to have no impact on the in vivo results (Barrow et al., 1987). Deletion of the tolC gene or the acrAB, acREF and mdtABC genes was done using the one-step inactivation method described by Datsenko and Wanner (Datsenko and Wanner, 2000; Bohez et al., 2006). This yielded a Salmonella Enteritidis strepR tolC deletion mutant, a Salmonella Enteritidis 147 StrepR acrABacrEFmdtABC deletion mutant, a Salmonella Typhimurium naIR tolC deletion mutant and a Salmonella Typhimurium naIR acrABacrEFmdtABC deletion mutant. Salmonella Enteritidis strain 76Sa88 naIR is a well-characterized nalidixic acid resistant strain which was originally isolated from a poultry farm (Van Immerseel et al., 2002) and was used as a challenge strain in this study. Salmonella Typhimurium MB2136, a streptomycin resistant wild-type strain originally isolated from swine (De Cort et al., 2015), was also used as a challenge strain in this study.

Experimental design

Forty one-day-old chicks were divided into 4 groups of 10 and each housed in a container of 1 .2 m2. Two groups were given 0.5 ml of a mixture containing 2 x 108 CFU/ml of the Salmonella Enteritidis AacrABacrEFmdtABC strain and 2 x 108 CFU/ml of the Salmonella Typhimurium AacrABacrEFmdtABC strain by oral gavage on day 1 of the experiment. The two other groups were given 0.5 ml of Hank's Balanced Salt Solution (HBSS, 14175053, Invitrogen, Paisley, England) by oral gavage as a control on day one of the experiment. On day two of the experiment, one control group and one group treated with the CI mixture were given 0.5 ml of a solution containing 2 x 105 CFU/ml of the Salmonella Enteritidis 76Sa88 naIR challenge strain by oral gavage, while the other two groups were challenged by administering 0.5 ml of a solution containing 2 x 105 CFU/ml of the Salmonella Typhimurium MB2136 streptR challenge strain by oral gavage. To evaluate colonization by the challenge strains, their numbers in caecum and spleen were determined at day 7 of the experiment. Shedding of the challenge strains was evaluated by bacteriological analysis of cloacal swabs taken on days 3 and 7. Bacteriological analysis

Cloacal swabs taken were directly inoculated on xylose lysine deoxycholate agar (XLD; Oxoid, Basingstoke, England) plates supplemented with 20 g/ml nalidixic acid or 100 g/ml streptomycin. Because the Salmonella Enteritidis AacrABacrEFmdtABC strain and the Salmonella Typhimurium AacrABacrEFmdtABC strain are unable to grow on XLD agar, XLD agar was used for the detection of the challenge strains. Samples negative after direct inoculation were pre-enriched in buffered peptone water (BPW; Oxoid, Basingstoke, England) and incubated overnight at 37 °C. One ml of this suspension was further enriched by adding 9 ml tetrathionate-brilliant green broth (Merck, Darmstadt, Germany). After overnight incubation at 37 °C, this suspension was plated XLD plates supplemented with the appropriate antibiotic. Samples of caecum and spleen were homogenized in BPW and 10-fold dilutions were made in HBSS. Six droplets of 20 μΙ of each dilution were plated on XLD plates, supplemented with 20 g/ml nalidixic acid or 100 g/ml streptomycin. After overnight incubation at 37 °C, the number of CFU/g tissue was determined by counting the number of bacterial colonies on the plates. Negative samples were enriched as described above.

Statistical Analysis

GraphPad Prism software (Version 5.0, GraphPad Software Inc., La Jolla, CA) was used for statistical analysis of the obtained data. A chi-square test was used to analyze differences in mortality between groups. A Fisher's test was used to analyze statistical differences between groups in the number of Sa/mone//a-positive cloaca swabs and in the number of spleen and cecum samples positive for Salmonella. Bacterial counts in cecum and spleen were converted into logarithmic form for statistical analysis. Samples of cecum and spleen that were negative after direct plating were rated as Iog10 = 0. Differences between groups were analyzed using a Mann-Whitney test. Differences with P-values lower than 0.05 were considered to be significant.

Results

No animals died after during the experiment and as such, there are no statistical differences in mortality between groups treated with the CI culture and the control groups.

Faecal shedding of the Salmonella Enteritidis challenge strain after experimental infection was the same in the control group and the CI culture treated group, with 5 out of 10 chickens shedding the strain in both groups on day 3 of the experiment. On day 7 of the experiment, only 6 out of 10 chickens in the CI treated group were shedding the challenge strain, while 10 out of 10 chickens in the control group were shedding the Salmonella Enteritidis challenge strain. Faecal shedding of the Salmonella Typhimurium challenge strain was initially higher in the CI treated group where 5 out of 10 animals were shedding the strain, while in the control group, only one chicken out of 10 was shedding the strain. However, on day 7 of the experiment, 10 out of 10 animals were shedding the Salmonella Typhimurium challenge strain in control group, and 9 out 10 chickens were shedding the challenge strain in the CI treated group. After direct plating of the caecal samples, the Salmonella Enteritidis challenge strain could not be detected in the group treated with the CI culture (Figure 10). However, in the control group, the Salmonella Enteritidis challenge strain could be detected in high numbers in several samples. The Salmonella Enteritidis challenge strain could not be detected in any of the spleen samples, in neither one of the groups. The Salmonella Typhimurium challenge strain could be found in significantly lower amounts in the group treated with the CI culture when compared to the control group (Figure 10). In the spleen however, there was no significant difference between the treated and the untreated group in colonization by the Salmonella Typhimurium challenge strain.

After enrichment of the caecal samples, the Salmonella Enteritidis challenge strain could be detected in all samples in both Salmonella Enteritidis-challenged groups. Similarly, the Salmonella Typhimurium challenge strain could be detected in all caecal samples from both the control and the CI treated group (Table 2). After enrichment of the spleen samples, a significantly higher amount of spleens were positive for the Salmonella Enteritidis challenge strain in the control group when compared to the group treated with the CI culture. There was no significant difference in number of spleen samples positive for the challenge strain between the groups that were experimentally infected with the Salmonella Typhimurium challenge strain (Table 2).

Table 2: The number of caecal and spleen samples positive for Salmonella Enteritidis or Salmonella Typhimurium wild-type strains on day 7 of age after experimental infection of two days old broiler chickens treated with a CI culture. Challenge serotype: Salmonella Enteritidis Salmonella Typhimurium

Group: Control CI treated Control CI treated

Caecum 10 ^a /10 ^b 10/10 10/10 10/10

Milt 10710 2710 7/10 8/10

a Number of positive samples after enrichment

b Total number of samples

^* Significant difference between control and CI treated groups (P-value < 0.05) The CI culture was administered on day one of life, and consisted of 108 CFU of a Salmonella Enteritidis AacrAbacrEFmdtABC strain and 108 CFU of a Salmonella Typhimurium AacrAbacrEFmdtABC strain administered simultaneously by oral gavage. The chickens were experimentally infected on day 2 of life by administering them 105 CFU of the respective challenge strain by oral gavage.

Conclusion

A CI culture consisting of the AacrABacrEFmdtABC strains is able to offer protection against Salmonella Enteritidis and Typhimurium after experimental infection. As such, these strains can be used to help reduce Salmonella prevalence in broilers and eventually reduce the number of food borne Salmonella infections in humans.

REFERENCES

Allen-Vercoe, E. and M.J. Woodward, The role of flagella, but not fimbriae, in the adherence of Salmonella enterica serotype Enteritidis to chick gut explant. J Med Microbiol, 1999. 48(8): p.771 -80.

ANONYMOUS. 2007. Regulation (EC) No. 1237/2007 of 23 October 2007 amending regulation (EC) No 2160/2003 of the Europian Parliament and of the Council and Decision 2006/696/EC as regards the placement of the market of eggs from Salmonella infected flocks of laying hens. off. J. Eur. Union, L280: 5-9. Barrow, P. A., J. F. Tucker, et al. (1987). "Inhibition of colonization of the chicken alimentary tract with Salmonella typhimurium gram-negative facultatively anaerobic bacteria." Epidemiol Infect 98(3): 31 1 -22.

Bohez, L, J. Dewulf, et al. (2008). "The effect of oral administration of a homologous hilA mutant strain on the long-term colonization and transmission of Salmonella Enteritidis in broiler chickens." Vaccine 26(3): 372-8.

Bohez, L, R. Ducatelle, et al. (2006). "Salmonella enterica serovar Enteritidis colonization of the chicken caecum requires the HilA regulatory protein." Vet Microbiol 116(1 -3): 202-10. Bohez L, Dewulf J, Ducatelle R, Pasmans F, Haesebrouck F, Van Immerseel F. The effect of oral administration of a homologous hilA mutant strain on the long-term colonization and transmission of Salmonella Enteritidis in broiler chickens. Vaccine 2008;26:372-8. Datsenko, K. A. and B. L. Wanner (2000). "One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products." Proc Natl Acad Sci U S A 97(12): 6640-5.

Davison, C.E. Benson, D.J. Henzler, R.J. Eckroade, Field observations with Salmonella Enteritidis bacterins, Avian Dis., 43 (1999), pp. 664-669. De Buck J, Van Immerseel F, Haesebrouck F, Ducatelle R. Protection of laying hens against Salmonella Enteritidis by immunization with type 1 fimbriae. Vet

Microbiol. 2005 Jan 31 ;105(2):93-101 .

De Cort W, Geeraerts S, Balan V, Elroy M, Haesebrouck F, Ducatelle R, et al. A Salmonella Enteritidis hilAssrAfliG deletion mutant is a safe live vaccine strain that confers protection against colonization by Salmonella Enteritidis in broilers. Vaccine 2013;31 :5104-10.

De Cort, W., S. Geeraerts, V. Balan, M. Elroy, F. Haesebrouck, R. Ducatelle, and F. Van Immerseel. A colonisation-inhibition culture consisting of Salmonella Enteritidis and Typhimurium AhilAssrAfliG strains protects against infection by strains of both serotypes in broilers. Vaccine 2014;32;4633^1638.

De Cort, W., Haesebrouck, F., Ducatelle, R., Van Immerseel, F. Administration of a Salmonella Enteritidis AhilAssrAfliG strain by coarse spray to newly hatched broilers reduces colonization and shedding of a Salmonella Enteritidis challenge strain. Poultry Science, 2015, 94(1 ):131 -135.

Desin TS, Koster W, Potter AA. Salmonella vaccines in poultry: past, present and future. Expert Rev Vaccines. 2013 Jan;12(1 ):87-96.

Desmidt, M., R. Ducatelle, J. Mast, B. M. Goddeeris, B. Kaspers, and F. Haesebrouck. 1998. Role of the humoral immune system in Salmonella enteritidis phage type four infection in chickens. Vet. Immunol. Immunopathol . 63:355-367.

EFSA 2013. EU summary report on zoonoses, zoonotic agents and food-borne outbreaks 201 1 . The EFSA Journal, 1 1 , 19-73.

Gantois I, Ducatelle R, Pasmans F, Haesebrouck F, Gast R, Humphrey TJ and Van Immerseel F. FEMS Microbiology Reviews; Volume 33, Issue 4, pages 718-738, July 2009.

Gantois I, Ducatelle R, Timbermont L, Boyen F, Bohez L, Haesebrouck F, Pasmans F, van Immerseel F. Oral immunisation of laying hens with the live vaccine strains of TAD Salmonella vac E and TAD Salmonella vac T reduces internal egg contamination with Salmonella Enteritidis. Vaccine. 2006 Sep 1 1 ;24(37-39):6250-5.

Gast RK, Guard-Petter J & Holt PS (2003) Effect of prior serial in vivo passage on the frequency of Salmonella Enteritidis contamination in eggs from experimentally infected laying hens. Avian Dis 47; 633-639.

Hassan JO, Curtiss R 3rd. Efficacy of a live avirulent Salmonella typhimurium vaccine in preventing colonization and invasion of laying hens by Salmonella typhimurium and Salmonella enteritidis. Avian Dis. 1997 Oct-Dec;41 (4):783-91 .

Lister SA (Ί 988,) Salmonella Enteritidis infection in broilers and broiler breeders. Vet Rec 123: 350.

Majowicz SE, Musto J, Scallan E, Angulo FJ, Kirk M, O'Brien SJ, Jones TF, Fazil A, Hoekstra RM; International Collaboration on Enteric Disease 'Burden of Illness' Studies. The global burden of nontyphoidal Salmonella gastroenteritis. Clin Infect Dis. 2010 Mar 15;50(6):882-9. Matsuda, K., Chaudhari, A. A. and Lee, J. H., 201 1 . Comparison of the safety and efficacy of a new live Salmonella Gallinarum vaccine candidate, JOL916, with the SG9R vaccine in chickens. Avian Dis 55, 407-12.

Methner, U., S. al-Shabibi, et al. (1995). "Experimental oral infection of specific pathogen-free laying hens and cocks with Salmonella enteritidis strains." Zentralbl Veterinarmed B 42(8): 459-69.

Methner, U., S. al-Shabibi, et al . (1995). "Infection model for hatching chicks infected with Salmonella enteritidis." Zentralbl Veterinarmed B 42(8): 471 -80.

Miyamoto T, Kitaoka D, Withanage GS, Fukata T, Sasai K, Baba E. Evaluation of the efficacy of Salmonella enteritidis oil-emulsion bacterin in an intravaginal challenge model in hens. Avian Dis. 1999 Jul-Sep;43(3):497-505.

Nandre R, Matsuda K, Lee JH. Efficacy for a new live attenuated Salmonella Enteritidis vaccine candidate to reduce internal egg contamination. Zoonoses Public Health. 2014 Feb;61 (1 ):55-63. Nishino K, Latifi T, Groisman EA. Virulence and drug resistance roles of multidrug efflux systems of Salmonella enterica serovar Typhimurium. Mol Microbiol 2006; 59: 126-41 .

Tsukasa Horiyama, Akihito Yamaguchi and Kunihiko Nishino. TolC dependency of multidrug efflux systems in Salmonella enterica serovar TyphimuriumJ Antimicrob Chemother 2010; 65: 1372-1376.

Van Immerseel F, Studholme DJ, Eeckhaut V, Heyndrickx M, Dewulf J, Dewaele I, Van Hoorebeke S, Haesebrouck F, Van Meirhaeghe H, Ducatelle R, Paszkiewicz K, Titball RW. Salmonella Gallinarunn field isolates from laying hens are related to the vaccine strain SG9R. Vaccine. 2013 Oct 9;31 (43):4940-5.

Van Immerseel, F., De Buck, J., De Smet, I., Haesebrouck, F. and Ducatelle, R. (2002). The effect of vaccination with a Salmonella Enteritidis aroA mutant on early cellular responses in caecal lamina propria of newly-hatched chickens, Vaccine, 20(23-24) : 3034-3041 . Van Parys, A., F. Boyen, et al. (2012). "Salmonella Typhimurium induces SPI-1 and SPI-2 regulated and strain dependent downregulation of MHC II expression on porcine alveolar macrophages." Vet Res 43: 52.

Woodward MJ, Gettinby G, Breslin MF, Corkish JD, Houghton S. The efficacy of Salenvac, a Salmonella enterica subsp. Enterica serotype Enteritidis iron-restricted bacterin vaccine, in laying chickens. Avian Pathol. 2002 Aug;31 (4):383-92.

Woodward, M.J., E. Allen-Vercoe, and J.S. Redstone, Distribution, gene sequence and expression in vivo of the plasmid encoded fimbrial antigen of Salmonella serotype Enteritidis. Epidemiol Infect, 1996. 117(1 ): p. 17-28.