Salmonella in Chicken: Current and Developing Strategies To Reduce Contamination at Farm Level

Transcription

1 774 Journal of Food Protection, Vol. 73, No. 4, 2010, Pages Copyright G, International Association for Food Protection Review Salmonella in Chicken: Current and Developing Strategies To Reduce Contamination at Farm Level S. VANDEPLAS, 1 * R. DUBOIS DAUPHIN, 2 Y. BECKERS, 1 P. THONART, 2 AND A. THÉWIS 1 1 Animal Science Unit and 2 Bio-industry Unit, Gembloux Agro-Bio Tech, University of Liège, Passage des Déportés, 2, B-5030 Gembloux, Belgium MS : Received 28 September 2009/Accepted 6 December 2009 ABSTRACT Salmonella is a human pathogen that frequently infects poultry flocks. Consumption of raw or undercooked contaminated poultry products can induce acute gastroenteritis in humans. Faced with the public health concerns associated with salmonellosis, the European Union has established a European regulation forcing member states to implement control programs aimed at reducing Salmonella prevalence in poultry production, especially at the primary production level. The purpose of the present review article is to summarize the current research and to suggest future developments in the area of Salmonella control in poultry, which may be of value to the industry in the coming years. The review will focus especially on preventive strategies that have been developed and that aim at reducing the incidence of Salmonella colonization in broiler chickens at the farm level. In addition to the usual preventive hygienic measures, other strategies have been investigated, such as feed and drinking water acidification with organic acids and immune strategies based on passive and active immunity. Modification of the diet by changing ingredients and nutrient composition with the intent of reducing a bird s susceptibility to Salmonella infection also has been examined. Because in ovo feeding accelerates small intestine development and enhances epithelial cell function, this approach could be an efficient tool for controlling enteric pathogens. Feed additives such as antibiotics, prebiotics, probiotics, and synbiotics that modify the intestinal microflora are part of another field of investigation, and their success depends on the additive used. Other control methods such as the use of chlorate products and bacteriophages also are under study. Salmonella infection is a major cause of gastroenteritis in humans (salmonellosis) worldwide and is often associated with consumption of raw or undercooked poultry (35). Because salmonellosis is a public health concern, the European Union (EU) has established a regulation mandating its member states to set up surveys and control programs focused on Salmonella and other pathogenic agents in poultry. Decontamination treatments of meat or eggs as a final products is the easiest way of controlling Salmonella. However, carcass disinfectants are prohibited in the EU, and decontamination of fresh table eggs is difficult. Therefore, the EU control programs have to be applied as an integrated approach along the poultry production chain (112) and are focused on prevention and monitoring or eradicating Salmonella in live birds. Poultry products also must be controlled during transport, at the slaughterhouse, and/or at the production stage. In Belgium, as in most of the EU, the incidence of salmonellosis has decreased since 2003 after the implementation of national plans. This reduction may be explained by the reduced prevalence of Salmonella serovar Enteritidis (which is mostly associated with eggs) due to recent vaccination programs of flocks of breeder and laying hens. Nevertheless, vaccination appears to be only partly effective for controlling Salmonella in poultry, and * Author for correspondence. Tel: z ; Fax: z ; vandeplas.s@fsagx.ac.be. complementary strategies are needed, especially during chicken production. Fries (42), Dinçer and Baysal (28), and Rasschaert (102) discussed management practices that can be implemented to prevent contamination during catching, transportation, and slaughter operations and techniques to decontaminate raw poultry meat and meat products in the food processing industry. However, poultry carcass contamination cannot be avoided when chickens already are contaminated when they arrive at the slaughterhouse, even under optimal slaughtering and processing conditions (26), and prophylactic measures must target poultry flocks at the primary production level. The purpose of the present review article is to summarize the current research and suggest future directions in the area of Salmonella control in poultry, which may be of value to the industry in the coming years. This review focuses especially on preventive and curative strategies aimed at reducing the incidence of Salmonella colonization in broiler chickens at the farm level. The most widely used control measures against Salmonella are preventive hygienic measures and vaccination with dead and live Salmonella strains, as briefly summarized below. Another immune strategy is passive immunity of birds that have been fed specific antibodies produced from eggs of hyperimmunized hens. Cationic antimicrobial peptides such as avian gallinacins, which play an important role in innate host defense in chickens and have antimicrobial activity, can be

2 J. Food Prot., Vol. 73, No. 4 STRATEGIES TO REDUCE SALMONELLA PREVALENCE IN CHICKEN 775 added to feed to reduce Salmonella incidence in poultry (10). Other strategies include the use of genetically resistant chicken lines (30) and the acidification of feed and drinking water with short- and medium-chain fatty acids (124). Modification of the ingredients and nutrient composition of the diet also could be useful for reducing susceptibility to infectious pathogens, as suggested by Klasing (71). The most promising strategies in this field seem to be incorporation of high-fiber dietary sources and nonstarch polysaccharide hydrolyzing enzymes into the diet, which may modify the microfloral and physicochemical balance in the gastrointestinal tract (GIT). Amino acids, dietary antioxidants, and essential minerals also are important to maintain gastrointestinal integrity (1). In ovo feeding can accelerate the development of the small intestine and enhance function of epithelial cells (111). These effects may improve a bird s resistance to colonization by Salmonella and other pathogens. Feed additives such as antibiotics, prebiotics, probiotics, and synbiotics that modify the gut microflora also are being investigated, and the success of this approach differs with the additive used. Other control methods, in particular the use of chlorate products and bacteriophages, also have been investigated (2, 17). SANITARY BARRIER Biosecurity refers to an action plan designed to minimize the risk of introducing diseases or zoonotic agents into a flock. Because everything introduced into the poultry house is potentially contaminated, an effective control program must encompass several biosecurity measures. The starting point is to ensure that poultry breeding flocks, house, feed, and litter are kept Salmonella free. Water also should be properly treated (30). The European Food Safety Authority (34) and Helm (55) have reported specific hygiene measures to be followed for equipment, animals, and people. However, such hygiene strategies are only partly effective for controlling pathogen colonization of the poultry house, and thus complementary measures are necessary to control Salmonella efficiently (80). Dietary modification. As stated by Klasing (71), dietary characteristics can affect a bird s susceptibility to infectious pathogens, and the level or type of ingredient may be of critical importance. Diverse dietary strategies have been developed or investigated for preventing or reducing the establishment of Salmonella in the avian GIT. The feed structure also affects the susceptibility of poultry to Salmonella infection. Whole wheat in the feed lowered the ph in gizzard contents significantly, which in turn decreased the Salmonella population (11). Huang et al. (62) investigated the effects of feed particle size and feed form on Salmonella colonization and found higher cecal Salmonella Enteritidis concentrations in broilers fed a pelleted diet than in those fed a mashed diet, whereas cecal volatile fatty acid (VFA) concentrations were significantly higher in broilers receiving the pelleted diet. Various raw feed materials also can inhibit the growth of pathogens in the GIT. The protein content and the nature of the protein source may help to protect the intestine from colonization. Plant protein based feeds especially contain various nonstarch polysaccharides that are fermented in the cecum into organic acids and thus are detrimental to pathogens, as shown for Campylobacter (121). High-fiber dietary approaches also have been proposed based on the assumption that dietary fiber is preferentially utilized by Lactobacillus and Bifidobacterium species to produce lactic acid and VFA, which promotes maintenance of normal microbial populations and low ph (31). For example, alfalfa is one of the most extensively studied high-fiber dietary sources that increase lactic acid production and have a potential inhibitory effect on Salmonella (29). Immunomodulation by diet is another important part of nutrition strategies to maintain animal health. Highmolecular-weight compounds such as b-glucans (77) and antioxidants (23) have been identified as having immunomodulation properties. Blount et al. (12) found an immune response to lutein, a yellow carotenoid, in avian species. b- glucans from yeast cell walls are used as feed additives and upregulate the innate immune response to Salmonella Enteritidis in immature chickens (77). Damage to the GIT structure may lead to increased risk of invasion by pathogens. Therefore, nutritional strategies that maintain gastrointestinal integrity are important. Several amino acids such as glutamine, serine, threonine, and arginine play an important role in maintaining GIT integrity and may have growth-limiting activity against Salmonella. These amino acids are fermented into lactic acid and VFAs (64), which are antagonistic to various pathogens. A chicken diet formulated with an optimal amino acid balance thus is necessary to control the proliferation of enteric pathogens in the gut. Other micronutrients such as iron may impair natural resistance to colonization, and free dietary iron can increase bacterial virulence (14). Incorporation of chelators such as phytic acid (84) and organic acids into the diet may improve the natural resistance of birds to Salmonella infection. However, the phosphorus in phytic acid is not bioavailable to nonruminant animals because these animals lack the digestive enzyme phytase. The unabsorbed phytic acid consequently passes unabsorbed through the GIT, elevating the amount of phosphorus in the manure and leading to environmental problems such as eutrophication. Rapid digestion and absorption of nutrients is an important aspect of dietary interventions because the amount of substrate remaining in the GIT that can be used by pathogenic microorganisms is reduced. Enzymes that hydrolyze nonstarch polysaccharides are used commonly to improve the nutritional value of feed cereals such as wheat and barley, but the effect of these enzymes on pathogen colonization in chickens has not been carefully studied (41, 53). Exogenous enzymes can alter microbial populations indirectly within the GIT through action on substrates that bacteria use as carbon sources. Santos (104) found that when a corn-based diet was supplemented with an enzyme product containing endoxylanase and alpha-amylase, Salmonella prevalence and cecal populations were significantly reduced. This product also improved growth performance in turkeys at 15 weeks of age.

3 776 VANDEPLAS ET AL. J. Food Prot., Vol. 73, No. 4 Chicks that were given a hydrated nutritional supplement during the first 2 to 3 days after hatching had earlier immune system development than did chicks that had fasted and were given no water over the same period, and the fed birds had improved resistance to disease challenge with coccidial oocysts (27). The early feeding concept was further expanded to in ovo feeding, i.e., the intra-amniotic administration of nutrients. During late embryogenesis, the injected solutions are consumed by the embryo and are digested and absorbed by the embryonic intestine before pipping. Because the immediate posthatch period is characterized by a transition from absorbing lipid-rich yolk to consuming exogenous feed that is rich in carbohydrates and proteins, solutions containing carbohydrates have been administered in ovo (111, 115). These solutions accelerated small intestine development and enhanced the functioning of enterocytes. Increased digestive capacity may improve growth performance in poultry; however, other effects linked to in ovo feeding have been described. Mousavi et al. (85) found that a solution containing 25 g/liter maltose, 25 g/ liter sucrose, 200 g/liter dextrin, and 1% threonine in 0.5% saline that was injected into the amniotic fluid of eggs on day 17.5 of incubation led to a significant increase in the body weight gain and food conversion ratio of posthatch broilers. Threonine was added because this important amino acid helps maintain function in the intestine and is one of the main precursors for intestinal mucin production. However, Smirnov et al. (111) injected a similar carbohydrate solution without threonine and observed that goblet cells containing acidic mucin increased by 50% and that mucin mrna expression was enhanced in comparison with that in the control group. Early and in ovo feeding could therefore be useful complementary techniques for preventing colonization of the chicken intestine by bacterial pathogens. These interventions stimulate immune system development and mucin production, which is essential to prevent bacteria from adhering to the gut epithelium and subsequently invading the body. Nevertheless, the effects of treatments with various carbohydrates and amino acids should be further investigated and compared to determine the relative importance of each component for growth and mucin production. Acidification of the chicken environment: shortand medium-chain fatty acids. VFAs and short-chain fatty acids (SCFAs), which are mainly acetic, propionic, and butyric acids, are produced by the normal anaerobic intestinal flora as end products of metabolism (79). As demonstrated in vitro (118), these acids exert bacteriostatic or bactericidal effects against gram-negative bacteria by entering the bacterial cell in undissociated form. Once inside the bacterial cell, the acid dissociates, causing the intracellular ph to decrease and anions to accumulate (130). Within practical limits, the acid concentration in undissociated form increases as ph decreases, so that the antibacterial effect is more efficient at low ph. Durant et al. (32) found this effect in vitro when they observed reduced growth rate of Salmonella Typhimurium at decreasing ph levels in the presence of VFAs. Many researchers have investigated these SCFAs for their use as acidifiers of feed or water for reducing Salmonella colonization of broilers. However, these studies, reviewed by Doyle and Erickson (30) and Van Immerseel et al. (130), have yielded conflicting results. The administration mode, the type of acid, and its concentration are very important, and colonization is not always affected by the treatment when the infection pressure is high or when the chickens are highly stressed. Several products containing VFA mixtures decrease Salmonella shedding (18, 60, 118), and these products appeared to be most effective in the gizzard and the crop (118, 124). Researchers also have developed microencapsulation methods for organic acids that prevent the absorption of these acids in the upper GIT and ensure their release further down. Most commercial products that are applied in the field contain propionic and formic acids, either in powder form or encapsulated in silica beads. Byrd et al. (18) evaluated the addition of 0.5% acetic, lactic, or formic acid in drinking water during an 8-h pretransport feed withdrawal period and found reduced numbers of Salmonella Typhimurium in the crop (0.79 versus 1.45 log CFU/g) after treatment with lactic and formic acids when compared with control birds. Coated butyric acid also was tested in the feed for its effect on Salmonella colonization (125). Although the number of chickens shedding Salmonella was significantly lower in the group receiving butyric acid, cecal colonization at slaughter age was equal to that of the control group of chickens. Van Immerseel et al. (129) used chickens challenged at 5 and 6 days posthatch with Salmonella Enteritidis at CFU to test the effect of feed supplementation with microencapsulated SCFAs on Salmonella invasion, which was measured in internal organs at 8 days posthatch. Although administration of the butyric acid impregnated microbeads resulted in significantly decreased colonization by Salmonella Enteritidis in the cecum, feed supplementation with acetic acid and to a lesser extent formic acid led to an increased colonization of the cecum and other internal organs. The results of some in vitro studies suggest that VFAs might also have an undesirable effect by promoting the invasiveness of Salmonella (126) or enhancing the resistance of Salmonella to acid (73). Although Salmonella normally do not survive at ph 3.0 in culture, Kwon and Ricke (73) found that a high number of bacteria survived for some hours after exposure to any VFA, suggesting that bacteria could be protected against gastric acid when passing through the GIT. This protection is due to the expression of genes involved in an acid-tolerance response and the synthesis of a series of acid shock proteins that are protective against extreme acidic conditions (73). Nevertheless, this resistance might be effective only when the pathogen is already in contact with the VFA before ingestion, i.e., in the feed, which is unlikely to occur, and this effect is not observed for all VFAs (126). SCFAs also may modulate the expression of Salmonella genes involved in virulence (32, 130). An acetate-induced increased invasion was detected for some Salmonella serovars in vitro, whereas propionate or butyrate decreased epithelial

4 J. Food Prot., Vol. 73, No. 4 STRATEGIES TO REDUCE SALMONELLA PREVALENCE IN CHICKEN 777 cell invasion (128). Butyrate down-regulated the expression of the Salmonella pathogenicity island 1 gene, which is crucial for bacterial invasion of the epithelial cells and for virulence (45). Medium-chain fatty acids (MCFAs), which contain 6 to 12 carbons (caproic, caprylic, capric, and lauric acids), appear to be much more effective than SCFAs against Salmonella. As little as 25 mm concentrations of 6- to 10- carbon acids were bacteriostatic against a Salmonella Enteritidis strain, whereas the same strain tolerated 100 mm concentrations of SCFAs (127, 128). When Salmonella serovars Enteritidis and Typhimurium were incubated with 5 mm monocaprin combined with an emulsifier, the bacteria did not survive (119). However, large-scale studies are needed to confirm the antibacterial activity of MCFAs against Salmonella. In conclusion, SCFAs and MCFAs can be effective products for reducing Salmonella bacteria in broilers and the infection pressure on the farm because these acids act fast and can be used in young chickens, in contrast to many other types of feed additives. However, selection of the correct type of organic acid to use in commercial products has been empirical, and little information is available concerning the best combination of acids in these products. Complementary studies are needed to analyze the effects of individual VFAs on Salmonella virulence. Acidification of the chicken environment: fermented liquid feed. Another type of acidified poultry diet is fermented liquid feed (FLF). The fermentation may happen spontaneously by inoculating the moistened feed at a certain temperature before feeding the birds or it may be induced. The induced fermentation can be achieved by backslopping or by inoculating with lactic acid bacteria. Consequently, FLF is characterized by high numbers of lactic acid bacteria, high concentrations of lactic acid, and a low ph of 3.5 to 4.5 (98). Fermented feed given to pigs influences the bacterial ecology of the GIT and reduces Enterobacteriaceae all along the GIT (98). Heres and coworkers (56, 57) tested the effect of FLF on the susceptibility of broiler chickens to Salmonella. Heres et al. (56) infected chickens orally with Salmonella Enteritidis at 10 2 to 10 5 CFU at 8 days of age and observed that the proportion of Salmonella-shedding chickens was significantly decreased in birds given FLF, regardless of the infection dose. In another experiment, a fermented feed containing high concentrations of lactic and acetic acids reduced Salmonella in the anterior parts of the GIT; this reduction was more pronounced in chickens inoculated with Salmonella Enteritidis at 8 or 22 days of age than in chickens fed dry feed (57). Similar results were obtained with Salmonella Typhimurium by Savvidou et al. (105), who found that FLF controlled Salmonella more effectively than did liquid feed acidified with 30.3 ml/kg lactic acid. However, several variables in the fermentation process, such as the composition of the feed and the production stage, may affect the advantages of FLF and should be further studied. Acidification of the chicken environment: litter treatment. Feed additives (food components other than feedstuffs) such as the organic acids discussed above are subject to strict control in Europe. Because of the cost associated with authorization of these additives, strategies such as litter treatment, with lower costs, could be more practical for the producer for preventing or reducing Salmonella contamination in broiler flocks. Litter treatments with various acidifying compounds have been reported. Their goal of such treatments was to reduce the emission of ammonia by modifying the litter ph. The effect of organic acids, formalin, sodium bisulfate, aluminum sulfate, and sulfuric acid on Salmonella and other pathogens has been described (65, 74, 99, 133). Poultry Guard litter amendment is a product based on 40 to 50% sulfuric acid and is used to control ammonia volatilization by converting litter ammonium to ammonium sulfate, thereby decreasing litter ph and resulting in pathogen reduction in the litter (132). Such treatment might lead to decreased Salmonella concentration in the chick cecum, as reported by Vicente et al. (132). However, Line (74) did not find any effect of two commercially available acidifying litter treatments, aluminum sulfate and sodium bisulfate, on Salmonella colonization in chickens raised on pine shavings. IMMUNE STRATEGIES Passive immunity conferred by specific antibodies. Maternal antibodies transferred from the yolk to the chick prevent chick colonization by Salmonella (54). This fact led to the hypothesis that feeding these antibodies to hatchlings would provide passive immune protection. Several researchers have investigated how hyperimmunized hens can produce pathogen-specific antibodies in large quantities in eggs (reviewed by Schade et al. (106)). In recent years, the efficacy of such antibodies against Salmonella has been the focus of in vitro (21) and in vivo (52, 101) experiments and has yielded promising results. In the study by Rahimi et al. (101), 3-day-old chicks that were challenged with Salmonella Enteritidis at CFU/ml and received purified yolk immunoglobulin (Ig) Y in drinking water showed significantly lower fecal shedding (0 versus 14% of chicks) and had lower levels of Salmonella Enteritidis in the cecum (0.27 versus 3.98 log CFU/g) at 28 days of age. These researchers also investigated the effect of IgY combined with a probiotic mixture of Lactobacillus acidophilus, Lactobacillus casei, Enterococcus durans, and Bifidobacterium thermophilus (Primalac, Star Labs/Forage Research, Clarksdale, MO) and obtained results similar to those obtained with the antibody alone. Tellez et al. (117) also found an inhibitory effect for a similar combination: Lactobacillus acidophilus and Streptococcus faecium (Avian Pac Plus, Loveland Industries, Inc., Greeley, CO) plus egg-source antibodies against Salmonella Enteritidis, Salmonella Typhimurium, and Salmonella Heidelberg. Products with IgY that are simultaneously directed against several Salmonella serovars have been effective (21). Nonimmunized egg yolk powder also was effective for eliminating and preventing Salmonella colonization in poultry (70). Egg yolk contains antiinfectious and antiadhesive factors other than IgY, such as ovotransferrin, and

5 778 VANDEPLAS ET AL. J. Food Prot., Vol. 73, No. 4 the antimicrobial activity may involve agglutination of the pathogen to the egg yolk, generating a competition for adhesion sites, or the stimulation of the immune system by egg yolk components. Lymphokines also may be used for passive immunity, as tested by McGruder et al. (78). These authors administered Salmonella Enteritidis immune lymphokines in ovo and challenged the day-of-hatch chicks with Salmonella Enteritidis at CFU. Organ invasion was significantly decreased in 2-day-old chicks treated with lymphokines. More information about passive immunity is available in the review by Chalghoumi et al. (20). Antimicrobial peptides. Antimicrobial peptides are small molecules with a molecular mass of 1 to 5 kda (68). Their structure usually contains elements that facilitate interaction with the negatively charged membrane of target pathogen organisms. In chickens, the cationic antimicrobial peptides, including avian b-defensins called gallinacins, are natural endogenous antibiotics and constitute bioactive molecules of the innate immune system (10). The antimicrobial activity of different cationic antimicrobial peptides against Salmonella has been described in some in vitro studies (36, 82, 123). Using a time-kill assay, Milona et al. (82) found that gallinacins 4, 7, and 9 had antimicrobial activity against strains of Salmonella Enteritidis and Salmonella Typhimurium, and there was evidence for the synergistic interaction of gallinacins 7 and 9 against Salmonella Enteritidis. Chicken and turkey heterophil peptides at concentrations of 2 to 16 mg/ml in 0.01% sterile acetic acid reduced the survival of Salmonella Enteritidis and Typhimurium strains by more than 90% (36). These antagonistic properties and the recent advances in cationic antimicrobial peptide technology may lead to development of new feed additives to prevent or limit poultry contamination by pathogens such as Salmonella. All living organisms, from bacteria to mammals, produce small antimicrobial peptides. Currently, chemical synthesis is too costly for large-scale production of peptides, and biological production should be attempted with microorganisms, tissue cultures, or transgenic plants and animals. Peptide-containing plant material also could be added to animal feed (68). However, the potential use of peptides as feed additives faces some important economic and ecological issues that must be resolved. For example, chemical modifications or encapsulation methods should be developed to make the peptides more resistant to proteolysis in animals. Some of the peptides such as defensins or bactencin have toxic effects (69, 100). Properties and applications of antimicrobial peptides have been reviewed by van t Hof et al. (131). Vaccination. Development of effective vaccines and disease control strategies requires an understanding of host immunological mechanisms against the pathogen. Little is known about the poultry immune response to virulent and attenuated Salmonella strains. The importance of cellmediated immunity for tissue clearance (9, 87) and systemic clearance (38) of virulent Salmonella strains is generally recognized. Vaccines have been designed with killed and live attenuated Salmonella strains and live attenuated vaccine strains developed by nonreverting mutations of genes involved in metabolic functions or in virulence factors. The protective immunity provided by killed pathogen vaccines is inferior to that provided by live pathogen vaccines because the killed pathogen stimulates mainly antibody production and is rapidly destroyed and eliminated from the host system, and relevant antigens are destroyed during vaccine preparation (8). Killed vaccines generally fail to induce cytotoxic T cells (87) and secretory IgA responses, which are potentially important for protecting mucosal surfaces (9). Although a number of different live Salmonella strains have been tested for their efficacy, only a few are registered and commercially available for use in poultry in Europe. One example is the rough strain Salmonella Gallinarum 9R (40), which is registered for prophylactic use against Salmonella Enteritidis. More information about the inhibitory effect of the different types of vaccines against Salmonella is available in the detailed review by Barrow (8). Subunit vaccines, such as those based on the outer-membrane protein (81) or toxoids (83), also have been used in poultry. Microflora-modulating feed additives: antibiotics. In modern poultry production, therapeutic antibiotics are used to control economically important infections such as coccidiosis and bacterial enteritis caused by Salmonella. Several types of antibiotic agents are available with different modes of action based on inhibition or alteration of bacteria metabolic functions (46). In addition to their curative use, antibiotics also have been given to poultry in feed to promote growth and prevent pathogen colonization. Fluoroquinolones, salinomycin sodium, trimethoprim, and polymyxin B were tested as feed additives before Salmonella challenge and either eliminated Salmonella Enteritidis from the poultry flocks or at least reduced Salmonella incidence (13, 50, 109). Most growth promoters act by reducing mainly gram-positive bacteria, which are known to depress animal growth either directly or indirectly through their metabolic activities (46). In recent years, concerns about using antibiotics in livestock have arisen. Antibiotic feed additives have been linked to the emergence of antibiotic resistant bacteria (134), the presence of undesired antibiotic residues in meat, and environmental contamination. Consequently, all prophylactic and growth-promoting antibiotics used in animal husbandry have been banned in the EU since January 2006 according to the Precautionary Principle. However, research in this area had produced conflicting conclusions and is still debated (19, 97). Casewell et al. (19) stated that the suppression of growth-promoting antibiotics has led to deteriorating animal welfare and increasing infections, which in turn substantially increase the use of therapeutic antibiotics. Alternatives methods for coping with pathogens have thus become popular. Unlike antibiotics, alternatives such as probiotics, prebiotics, and synbiotics have been designed partly for promoting the growth of some microflora communities.

6 J. Food Prot., Vol. 73, No. 4 STRATEGIES TO REDUCE SALMONELLA PREVALENCE IN CHICKEN 779 Microflora-modulating feed additives: probiotics. A probiotic is defined as a live microbial feed supplement which beneficially affects the host animal by improving its intestinal microbial balance (44). The administration of probiotic strains actually induces changes in the bacteria community structure (89) and stimulates the immune system and the production of host enzymes, which may lead to a beneficial nutritional and growth-promoting effect (96), as Jin et al. (66) observed. Probiotic bacteria may be especially useful in poultry under stress (e.g., from feed withdrawal, heat stress, or colonization by an enteric pathogen such as Salmonella), when sensitive antimicrobial components of the microflora such as lactobacilli and bifidobacteria tend to decrease and are no longer able to provide sufficient resistance to enteric pathogens (96). Results of many studies suggest that inoculation of cultures of one or several probiotic strains into broiler chickens may inhibit Salmonella contamination (3, 59, 122). Several potential mechanisms allowing probiotic cultures to exclude enteric pathogens have been proposed, including competition for receptor sites, competition for limiting nutrients, and production of antimicrobial metabolites (e.g., bacteriocins, VFAs, and hydrogen peroxide) (37, 92, 110). Microorganisms used as probiotics in animal feed are mainly bacterial strains belonging to the genera Lactobacillus, Enterococcus, Pediococcus, and Bacillus, although microscopic fungi such as Saccharomyces also have been used. However, much in vitro selection research has focused on the Lactobacillus species isolated from various sources (67, 120, 122) and is based on their various modes of action against pathogens. Jin et al. (67) found that Lactobacillus fermentum and L. acidophilus significantly reduced the attachment of Salmonella Typhimurium and Salmonella Pullorum, respectively, to ileal epithelial cells through exclusion (lactobacilli prior to addition of salmonellae) and competition (lactobacilli versus salmonellae). Two other strains of L. acidophilus and L. fermentum also were selected in vitro for their ability to adhere to intestinal epithelium cells from poultry and their inhibition effect against Salmonella Typhimurium due to production of lactic acid (120). In the study by Van Coillie et al. (122), Lactobacillus reuteri R and Lactobacillus johnsonii R inoculated into the proventriculus at CFU significantly decreased Salmonella Enteritidis colonization in cecum, liver, and spleen of 6-day-old chicks infected at 2 days of age with 10 4 CFU. Inoculation of Enterococcus spp. also protected chickens against Salmonella challenge because of the combined effects of lactic acid and bacteriocins (3). In all the in vivo trials described here, the probiotic was inoculated orally in one or a few steps, although the probiotic also could be incorporated into the feed, as described by Jin et al. (66) and Line et al. (75). Line et al. (75) observed that continuous supplementation of a standard feed with live yeast cells of Saccharomyces boulardii reduced cecal colonization from 1.64 to 0.15 log CFU/g in broiler chickens that had been cochallenged with Salmonella Typhimurium and Campylobacter jejuni at 4 days of age. Microflora-modulating feed additives: prebiotics. Prebiotics are feed ingredients that are not digested or metabolized or are metabolized very little as they pass through the upper portion of the GIT so they are available to interact with the flora of the large intestine. In the lower intestine, prebiotics serve selectively as a substrate to stimulate the growth and/or activity of bacteria in the GIT that are beneficial to host health (51). These characteristics are summarized in the definition of prebiotics by Gibson and Roberfroid (47). In principle, nondigestible fermentable feed carbohydrates and sugar alcohols may be considered prebiotics. Diverse types of prebiotics were listed by Šušković et al. (114). The nondigestible carbohydrates are divided according to their molecular weight, and the most important prebiotics are the di- and oligosaccharides (86). Prebiotics confer beneficial health effects on animals and humans, especially by inhibiting intestinal colonization by enteric pathogens such as Salmonella (51). This effect can be direct by binding to pathogens in the intestinal lumen and therefore blocking the adhesion of these bacteria to the epithelial cells (113). The effect also may be indirect; fermentable prebiotics may provide a substrate for the metabolism and growth of the normal intestinal flora, thus inhibiting pathogen colonization by competitive exclusion (CE), and may stimulate production of antibacterial metabolites such as lactic acid, VFAs, and bacteriocins (114). However, studies of the ability of prebiotics to control colonization by enteropathogens have produced inconsistent results, which were detailed in the review by Rehman et al. (103). Lactose, lactulose, and lactosucrose are natural disaccharides, or isomerization products, that have been reported to have prebiotic effects in chickens (22, 116). Tellez et al. (116) found a significant reduction in the total number of organ invasions by Salmonella Enteritidis at 14 and 19 days after infection in chicks fed with lactose. In contrast, Barnhart et al. (7) found that 2.5% lactose added to the drinking water of broilers during the last 5 to 11 days of growth before slaughter failed to reduce Salmonella populations in the crop and cecum. Fructooligosaccharides and mannanoligosaccharides are the most extensively studied oligosaccharides in chickens with respect to their prebiotic effect and their activity against Salmonella. Fructooligosaccharides (short-chain polymers of b 1-2- linked fructose units) produced commercially by hydrolysis of inulin or by enzymatic synthesis from sucrose or lactose (86) reduced chicken intestinal colonization by Salmonella in a way that differed with the mode of preparation and at what point they were added to the experimental diet (22). Mannanoligosaccharides (mannose-based carbohydrates) occur naturally in many products such as yeast cell walls and gums (113). Spring et al. (113) found that when chicks were infected at 3 days of age with Salmonella Dublin at CFU, the number of chicks in which this pathogen was present in the cecum at 10 days of age was significantly reduced (by 34%) when mannanoligosaccharides was incorporated at 4,000 ppm in the diet. More recently, other oligomers such as bacterial isomaltooligosaccharides produced from Leuconostoc mesenteroides fermentation (24) or

7 780 VANDEPLAS ET AL. J. Food Prot., Vol. 73, No. 4 xylooligosaccharides produced by hydrolysis of arabinoxylans from cereal (33) have been studied for their ability to inhibit Salmonella. Although mannose has not been considered a prebiotic because it can be metabolized by the host, it blocked Salmonella Typhimurium adherence to chicken intestinal epithelial cells in vitro (94) and decreased cecum colonization of 3-day-old chickens in vivo (93). Microflora-modulating feed additives: synbiotics. Because prebiotics are able to stimulate some beneficial microflora populations such as bifidobacteria or lactobacilli (114), another approach was to develop synbiotics, which are combinations of probiotic strains and prebiotic substrates (25, 108). According to Schrezenmeir and de Vrese (108), synbiotic is reserved for products in which the prebiotic compound selectively favors the probiotic strain, as observed in vitro by Kontula et al. (72), who found that b-glucooligosaccharides and xylooligosaccharides may be fermented by three specific lactic acid bacteria. This fermentation results in advantages to the host offered by the antimicrobial activity of the probiotic supplemented by the prebiotic effect. For example, the combination of fructose with a Bifidobacterium strain is a potentially effective synbiotic as is the use of lactilol or lactulose in conjunction with lactobacilli (25). The results of much research (48, 76, 114) have indicated that the combination of probiotics and prebiotics has a synergistic effect on the fecal microflora of experimental animals. This effect was demonstrated by the increase in total counts of anaerobes, aerobes, lactobacilli, and bifidobacteria and by the reduction of bacteria considered harmful, such as clostridia, Escherichia coli, Enterobacteriaceae, and Salmonella. Nevertheless, these results are based on experiments carried out on humans and rats. Most data available for poultry concern the effect of different synbiotics on growth performance and intestinal metabolism (4, 15). When Nisbet et al. (90) used a combination of dietary lactose and CE flora to reduce Salmonella colonization in chickens, these authors found that this combination was more effective for reducing Salmonella colonization than was lactose or the competitive flora alone. In contrast, Fasina et al. (39) failed to find any beneficial effect from the combination of pectin and a commercially available probiotic on cecal contamination in chickens infected at 4 days of age with Salmonella Typhimurium at 10 6 CFU. Table 1 summarizes the studies that have been conducted on the effects of feed additives based on pre-, pro-, and synbiotics on Salmonella colonization of the chicken intestine. Concerning synbiotic use in poultry, no studies are currently available with probiotic strains, but some were carried out with CE cultures of bacteria (Table 1). Nevertheless, some research is being carried out in our laboratories on another kind of synbiotic that combines a probiotic with an enzyme in a feed additive. By specifically metabolizing the prebiotics supplied by the enzyme from the hydrolysis of the broiler diet, the probiotic could efficiently compete in the GIT and its antagonistic activity could be more effective. Microflora-modulating feed additives: CE. In intensive production systems, newly hatched chicks are very sensitive to enteric infections because they have had little opportunity to develop a normal intestinal microflora in the clean housing conditions in which they are reared. Nurmi and Rantala (91) suggested that a suspension of alimentary tract contents from adult birds administrated orally to newly hatched chicks may establish an adult-type microflora in the chicks, which would protect them from Salmonella colonization. The process through which normal intestinal microflora protects the host against invading pathogens is called CE. This treatment is generally used as a prophylactic measure aiming at increasing the resistance of young chicks to Salmonella (79), but it also can be used after antibiotic therapy restore the normal microbiota (109). CE treatment is normally given to newly hatched chicks or turkey poults as soon as possible after hatching. It is either added to the first drinking water or sprayed in the hatchery or somewhere on the farm (6, 95, 107). The environmental spray method has been designed to overcome the problems linked to the reduced viability of anaerobic strains in water and to the variable drinking water uptake in the first 24 h after hatching (79). At present, various CE products are commercially available, e.g., AviFree (Alltech Ltd., Stamford, UK), Aviguard (Bayer AG, Leverkusen, Germany), Broilact (Orion Corporation, Espoo, Finland), MSC (Continental Grain Co., Arlon, Belgium), and Preempt or CF-3 (Milk Specialties Co., Dundee, IL). All are mixed cultures derived from the cecal contents and/or gut wall of domestic fowl (107). CE treatment was focused originally on controlling Salmonella infections; however, these treatments can protect chicks against other pathogens such as E. coli, Yersinia enterocolitica, and C. jejuni (107). Nakamura et al. (88), Bailey et al. (6), and Palmu and Camelin (95) found protective effects of Aviguard, MSC, and Broilact, respectively, against overwhelming colonization by salmonellae. Very little is known about the action mechanism of CE flora. The effect on Salmonella in the cecum is mainly bacteriostatic rather than bactericidal (79) and was partly explained by the VFA production by the bacteria (107). Hume et al. (63) found a significant increase in cecal propionic acid concentrations in chicks challenged with Salmonella Typhimurium at 10 4 CFU 4 h after CF-3 treatment by oral gavage. Other proposed mechanisms are competition for receptor sites in the intestinal tract and competition between pathogens and native microflora for nutrients (37, 79). Recently developed strategies. Byrd and coworkers (16, 17) recently investigated the use of an experimental chlorate product against Salmonella Typhimurium contamination in broiler chickens. Salmonella respires under anaerobic conditions by converting nitrate into nitrite with nitrate reductase. Because nitrate reductase does not distinguish between nitrate and chlorate, chlorate is reduced by this enzyme to chlorite, which builds up to toxic levels and kills susceptible bacteria. Byrd et al. (16, 17) found that a chlorate product added to control feed (5 to 18.5%) orto drinking water (7.5 to 30 mm chlorate ion) led to significantly lower incidence and levels of Salmonella Typhimurium in the crop and cecum. Sodium chlorate has

8 J. Food Prot., Vol. 73, No. 4 STRATEGIES TO REDUCE SALMONELLA PREVALENCE IN CHICKEN 781 TABLE 1. In vivo studies of the inhibitory effect of prebiotics, probiotics, and synbiotics against Salmonella in chickens Feed additive Salmonella challenge Observed effects Reference Probiotics Saccharomyces boulardii a at 100 g/kg of feed Lactobacillus reuteri R and L. johnsonii R at CFU (oral gavage) Enterococcus faecium J96 at 10 9 CFU (oral gavage) Eleven lactic acid bacteria isolates b at, CFU (oral gavage) CFU Salmonella Typhimurium at 4 days 10 4 CFU Salmonella Enteritidis at 2 days CFU Salmonella Pullorum at 4 days 10 4 CFU Salmonella Typhimurium or Salmonella Enteritidis at 1 day Prebiotics 0.4% arabinoxylooligosaccharides CFU Salmonella Enteritidis at 14 days 5% fructooligosaccharides, 5% lactulose, or 5% lactosucrose 10 7 CFU Salmonella Typhimurium at 1 day 0.4% mannanoligosaccharides 10 4 CFU Salmonella Typhimurium at 3 days 0.4% mannanoligosaccharides 10 4 CFU Salmonella Dublin at 3 days 0.75% fructooligosaccharides 10 9 CFU Salmonella Typhimurium at 7 days 10% lactose 10 8 CFU Salmonella Enteritidis at 13 or 18 days Synbiotics Anaerobic culture of encapsulated cecal bacteria z 2% lactose 0.02 ml of competitive exclusion culture (oral gavage) z 0.1% fructooligosaccharides Continuous-flow derived bacterial culture z 2% lactose 10 4 CFU Salmonella Typhimurium at 3 days 10 8 CFU Salmonella Enteritidis at 7 or 21 days 10 4 CFU Salmonella Typhimurium at 3 days 65% reduction in number of colonized birds; reduction of Salmonella counts in cecum from 1.64 to 0.15 log CFU/g Significant reduction of Salmonella counts in cecum, liver, and spleen Increased resistance to Salmonella infection (25% versus 50% mortality for the control) Reduction of the incidence of Salmonella Enteritidis by 60 70% and of Salmonella Typhimurium by 89 95% in the cecal tonsils at 25 h postchallenge; reduction of total cecal Salmonella Enteritidis by 2.9 log CFU at 25 h postchallenge Significant reduction of frequency of positive cloacal swabs from 1 to 11 days postinfection; significant reduction of cecal Salmonella Reduction of Salmonella scores at 6 wk, from 1.13 to 0.61, 1.08, and 1.11 for fructooligosaccharides, lactosucrose, and lactulose c Reduction of cecal Salmonella levels from 5.40 to 4.01 log CFU/g No. of birds with positive cecal samples reduced by 34% at 10 days No. of colonized chickens reduced by 42% Significant reduction in the total number of Salmonella-positive organ samples 1 day after infection Reduction of Salmonella counts in cecum by 3.4 to 5.3 log CFU at 10 days Significant reduction of mean Salmonella counts in chicks at 1 and 7 days postinoculation Significant reduction of Salmonella counts in cecum at 10 days; protection factor of 9.26 versus 2.49 for control d Line et al. (75) Van Coillie et al. (122) Audisio et al. (3) Higgins et al. (59) Eeckhaut et al. (33) Chambers et al. (22) Spring et al. (113) Spring et al. (113) Bailey et al. (5) Tellez et al. (116) Hollister et al. (61) Fukata et al. (43) Nisbet et al. (90) a Levucell SB20, Lallemand, Inc., Rexdale, Ontario, Canada. b FM-B11, IVS-Wynco LLC, Springdale, AR. c Score for Salmonella per cecal swab: 0, 0 CFU; 1, 1 to 100 CFU; 2,.100 CFU. d Protection factor is the Salmonella level (log CFU) in chicks fed the control diet divided by the level of the treatment group.

9 782 VANDEPLAS ET AL. J. Food Prot., Vol. 73, No. 4 low toxicity for animals, but such a nutritional strategy probably would face serious regulatory obstacles. Bacteriophages also were tested for their inhibitory effect against Salmonella. These viruses specifically infect and replicate in bacteria. Bacteriophage treatments were first applied to decontaminate poultry carcasses and products (49, 58). Higgins et al. (58) found that application of 5.5 ml of a single bacteriophage at 10 8 or PFU/ml significantly reduced the frequency of Salmonella recovery in carcasses inoculated with Salmonella Enteritidis. Subsequently, Andreatti Filho et al. (2) studied the effect of such treatment on Salmonella contamination in vivo. After treating day-of-hatch chicks via oral gavage with cocktails of 4 or 45 bacteriophages at 10 8 PFU per chick 1 h after oral challenge with Salmonella Enteritidis at CFU, recovery of Salmonella Enteritidis from cecal tonsils was significantly reduced 24 h posttreatment. However, no significant difference was reported at 48 h. These results suggest that further investigation is necessary to determine the efficiency of bacteriophages for reducing Salmonella in poultry over long periods. CONCLUSION Human gastroenteritis caused by Salmonella is a major food safety concern. Because foods of poultry origin are an important source of this pathogen, European regulations were established to force member states to implement programs to efficiently control Salmonella in poultry. These regulations are based on an integrated approach and include control measures applied throughout the poultry production chain. At the primary production level, biosecurity measures are only partly effective, and subtherapeutic antibiotics, which were used as growth promoters but also helped to prevent pathogen contamination, have been banned in the EU since January Therefore, other strategies have been investigated and are mostly based on diet modulation and feed supplementation with different types of additives other than feedstuffs. Vaccination was efficient only in laying hens and does not seem applicable in meat-producing birds, but passive immunity techniques, which are still under development, should offer broader applications. Acidification and microbial-based approaches have been studied for about 10 years but have produced conflicting results. In addition to CE microflora, defined bacterial strains such as probiotics have in vitro and in vivo antagonistic effects against Salmonella, mostly by producing organic acids and antimicrobial metabolites. Prebiotics, which are nondigestible carbohydrates fermented by specific bacteria in the microflora, also may be efficient additives for Salmonella control, as are synbiotics, where prebiotics can be used specifically as substrate by probiotics. None of these treatments are effective for completely eliminating Salmonella contamination from poultry flocks. Therefore, combinations of several measures seem to be the most efficient strategy for reducing the prevalence of this pathogen at the farm level. The most important criterion for selecting appropriate control measures is efficacy, which depends on many breeding parameters, but economical costs and aspects of Salmonella resistance, as observed in response to some organic acids, also must be taken into account. ACKNOWLEDGMENT The authors thank the Ministry of the Walloon Region DGTRE, Division for Research and Scientific Cooperation, Jambes, Belgium, for financial support. REFERENCES 1. Adams, C. A Nutrition-based health in animal production. Nutr. Res. Rev. 19: Andreatti Filho, R. L., J. P. Higgins, S. E. Higgins, G. Gaona, A. D. Wolfenden, G. Tellez, and B. M. Hargis Ability of bacteriophages isolated from different sources to reduce Salmonella enterica serovar Enteritidis in vitro and in vivo. Poult. Sci. 86: Audisio, M. C., G. Oliver, and M. C. Apella Protective effect of Enterococcus faecium J96, a potential probiotic strain, on chicks infected with Salmonella Pullorum. J. Food Prot. 63: Awad, W., K. Ghareeb, and J. Böhm Intestinal structure and function of broiler chickens on diets supplemented with a symbiotic containing Enterococcus faecium and oligosaccharides. Int. J. Mol. Sci. 9: Bailey, J. S., L. C. Blankenship, and N. A. Cox Effect of fructooligosaccharide on Salmonella colonization of the chicken intestine. Poult. Sci. 70: Bailey, J. S., N. J. Stern, and N. A. Cox Commercial field trial evaluation of mucosal starter culture to reduce Salmonella incidence in processed broiler carcasses. J. Food Prot. 63: Barnhart, E. T., D. J. Caldwell, M. C. Crouch, J. A. Byrd, D. E. Corrier, and B. M. Hargis Effect of lactose administration in drinking water prior to and during feed withdrawal on Salmonella recovery from broiler crops and ceca. Poult. Sci. 78: Barrow, P. A Salmonella infections: immune and nonimmune protection with vaccines. Avian Pathol. 36: Barrow, P. A., and T. S. Wallis Vaccination against Salmonella infections in food animals: rationale, theoretical basis and practical application, p In A. Wray and C. Wray (ed.), Salmonella in domestic animals. Cab International, Oxford. 10. Bennoune, O., M. Melizi, K. Khazal, and R. Bourouba Avian cationic antimicrobial peptides in health and disease: a mini review. J. Anim. Vet. Adv. 8: Bjerrum, J., K. Pedersen, and R. M. Engberg The influence of whole wheat feeding on Salmonella infection and gut flora composition in broilers. Avian Dis. 49: Blount, J. D., N. B. Metcalf, T. R. Birkhead, and P. F. Surai Carotenoid modulation of immune function and sexual attractiveness in Zebra finches. Science 300: Bolder, N. M., J. A. Wagenaar, F. F. Putirulan, K. T. Veldman, and M. Sommer The effect of flavophospholipol (FlavomycinH) and salinomycin sodium (SacoxH) on the excretion of Clostridium perfringens, Salmonella enteritidis, and Campylobacter jejuni in broilers after experimental infection. Poult. Sci. 78: Bullen, J. J., H. J. Rogers, P. B. Spalding, and C. G. Ward Iron and infection: the heart of the matter. FEMS Immunol. Med. Microbiol. 43: Buteikis, G., P. Matusevičius, A. Januškevičius, J. Jankowski, D. Mikulski, J. Blok, and K. Kozłowski Use of synbiotic preparations in turkey diets and their effect on growth performance. Vet. Zootech. 43: Byrd, J. A., R. C. Anderson, T. R. Callaway, R. W. Moore, K. D. Knape, L. F. Kubena, R. L. Ziprin, and D. J. Nisbet Effect of experimental chlorate product administration in the drinking water on Salmonella Typhimurium contamination of broilers. Poult. Sci. 82: Byrd, J. A., M. R. Burnham, J. L. McReynolds, R. C. Anderson, K. J. Genovese, T. R. Callaway, L. F. Kubena, and D. J. Nisbet Evaluation of an experimental chlorate product as a preslaughter feed supplement to reduce Salmonella in meat-producing birds. Poult. Sci. 87: Byrd, J. A., B. M. Hargis, D. J. Caldwell, R. H. Bailey, K. L. Herron, J. L. McReynolds, R. L. Brewer, R. C. Anderson, K. M.