Novel Approaches for Campylobacter Control in Poultry

Abstract

The Gram-negative bacterium Campylobacter is the most common bacterial cause of human gastroenteritis in the United States and many industrialized countries. Poultry, particularly chickens, is considered a major source of human campylobacteriosis. Thus, on-farm control of Campylobacter in poultry would reduce the risk of human exposure to this pathogen and have a significant impact on food safety and public health. To date, three general strategies have been proposed to control Campylobacter in poultry at the farm level: (1) reduction of environmental exposure (biosecurity measures), (2) an increase in poultry's host resistance to reduce Campylobacter carriage in the gut (e.g., competitive exclusion, vaccination, and host genetics selection), and (3) the use of antimicrobial alternatives to reduce and even eliminate Campylobacter from colonized chickens (e.g., bacteriophage therapy and bacteriocin treatment). Except for biosecurity measures, the other intervention approaches are currently not commercially available and are still under development. This review is focused on two promising strategies—vaccination and bacteriocin treatment. In particular, we extensively review recent research aimed at discovering and characterizing potent anti-Campylobacter bacteriocins to reduce Campylobacter load at the primary production level in poultry.

Introduction

Microaerophilic Campylobacter spp., including C. jejuni and C. coli, are the most common bacterial causes of human gastroenteritis in the United States and many industrialized countries (Friedman et al., 2000; Tauxe, 2002). Human Campylobacter illnesses are caused primarily by C. jejuni (∼90%) and secondarily by C. coli (∼10%). The estimated cases of campylobacteriosis in the United States are more than 2 millions per year (Mead et al., 1999). The medical and productivity costs resulting from C. jejuni infection are estimated at 1.5–8.0 billion dollars each year in the United States (Buzby et al., 1997; Buzby and Roberts, 1997). Poultry comprises the greatest concentration of Campylobacter and thus the main source of human campylobacteriosis (Friedman et al., 2000). A recent study using a novel population genetics approach further indicated that chicken is the major source of C. jejuni that is pathogenic to humans, whereas wild animal and environmental sources are responsible for only 3% of campylobacteriosis (Wilson et al., 2008). Quantitative risk assessment models have indicated that a reduction of C. jejuni numbers on a broiler carcass by 100-fold (or 2 log units) could result in a significant reduction (30 times less) in the incidence of campylobacteriosis (Rosenquist et al., 2003). Therefore, reduction or elimination of Campylobacter in the poultry reservoir is an essential step to control this food safety problem. Although there are multiple levels at which Campylobacter contamination can be targeted and implemented, on-farm control of Campylobacter would have the greatest impact because the intestine of living poultry is the only amplification point for Campylobacter throughout the food chain (Wagenaar et al., 2006, 2008).

Campylobacter is highly prevalent in poultry production systems, such as broilers, layers, turkeys, and ducks (Sahin et al., 2002). This review is focused on broilers, the largest poultry market sector being the primary research addressed. Campylobacter is a commensal organism that establishes persistent and benign infections in broilers with colonization level up to 10¹⁰ colony-forming units (CFU) per gram of feces (Sahin et al., 2002; Newell and Fearnley, 2003; Dhillon et al., 2006). Although Campylobacter can be isolated from most intestinal sites of broiler chickens, it is mainly found in the cecal and cloacal crypts, where it does not adhere to epithelial cells but is found in the mucous layer (Beery et al., 1988; Achen et al., 1998). In commercial conditions, Campylobacter is most often absent in broilers less than 2–3 weeks of age although experimental inoculation of newly hatched chicks with Campylobacter can establish colonization successfully (Stern et al., 1988; Newell and Wagenaar, 2000; Sahin et al., 2002). The reasons for this lag phase are unknown but might be attributed to multiple factors, such as presence of maternal antibodies, antibiotic feed additives, intestinal development, and intestinal microbial flora (Newell and Wagenaar, 2000; Sahin et al., 2002). Once the first bird in a flock becomes colonized, infection spreads to the entire flock in just a few days. This rapid spread of Campylobacter throughout the flock is a result of high levels of shedding and efficient fecal–oral transmission compounded by communal water and feed (Lee and Newell, 2006). Horizontal transmission from environmental sources is the primary route of flock infections by Campylobacter (Sahin et al., 2002). In broiler chickens, C. jejuni colonization can persist for the lifetime of the animal (6–7 weeks), consequently leading to carcass contamination at the slaughter facility. Together, Campylobacter can rapidly disseminate throughout the flock, and establish persistent and high-level colonization in broilers, which greatly challenges the development of effective farm-based intervention measures to reduce Campylobacter in poultry.

On-farm intervention measures to reduce Campylobacter in poultry have been comprehensively reviewed recently (Wagenaar et al., 2006, 2008; de Zoete et al., 2007; Connerton et al., 2008). Three general strategies have been proposed to control Campylobacter on the poultry farm: (1) reduction of environmental exposure (biosecurity measures), (2) an increase in poultry's host resistance to reduce Campylobacter carriage in the gut (e.g., competitive exclusion, vaccination, and host genetics selection), and (3) the use of antimicrobial alternatives to reduce and even eliminate Campylobacter from colonized chickens (e.g., bacteriophage therapy and bacteriocin treatment). The rationale and effectiveness of different intervention measures are briefly summarized in Table 1. Theoretically, reduction of environmental exposure of chickens to Campylobacter (the first general strategy in Table 1) should protect poultry against Campylobacter. However, effective implementation of this intervention strategy, such as biosecurity measures, relies on a better understanding of risk factors and sources of Campylobacter for poultry (Wagenaar et al., 2008). In addition, several practical limitations have impeded wide application of biosecurity measures (Table 1). To date, the remaining two general strategies (Table 1) are not commercially available and are still under development. This review will not discuss several specific measures, including competitive exclusion, host genetics selection, and bacteriophage therapy (Table 1), because of limited progress of these measures; the detailed information on these measures can be found in corresponding reviews or journal articles (Table 1). This review is primarily focused on two promising measures—vaccination and bacteriocin treatment. In particular, we will extensively review recent breakthroughs in the discovery and characterization of potent anti-Campylobacter bacteriocins because these bacteriocins dramatically reduced C. jejuni colonization in chickens and are being developed for on-farm control of Campylobacter in poultry.

Table 1.

Major On-Farm Strategies to Reduce Campylobacter jejuni in Poultry

Strategies	Rationale and effectiveness	References
I. Reduce environmental exposure
Biosecurity measures	Campylobacter is widespread in the farm environment. General biosecurity measures (e.g., hygiene and physical barriers) and specific biosecurity measures (e.g., single-species farming and partial depopulation) should protect poultry against Campylobacter. High biosecurity levels were correlated with absence of Campylobacter in poultry. However, it is difficult to assess the effectiveness of improving biosecurity in reducing Campylobacter in poultry. In addition, biosecurity measures may not work well for free-range flocks. The cost and practicality of such measures need to be determined.	Reviewed by Newell and Wagenaar (2000); Wagenaar et al. (2006, 2008)
II. Increase poultry's host resistance to reduce Campylobacter carriage in the gut
Competitive exclusion	Competitive exclusion (CE) products consist of defined or undefined bacterial agents from the microbiota of adult chickens. It is hypothesized that chickens fed with CE products should establish a protective enteric microbiota that prevents C. jejuni colonization. However, CE has only limited and inconsistent success for Campylobacter. For acceptance by the poultry industry and regulatory agencies, complete identification of complex species in CE products is needed but is challenging.	Reviewed by Mead (2000); Wagner (2006); Wagenaar et al. (2008)
Vaccination	There is a correlation between increasing levels of Campylobacter antibodies and reducing levels of C. jejuni colonization in poultry. Vaccination of chickens against Campylobacter has had only partial success. The challenges for an effective vaccine against Campylobacter in poultry are significant.	Reviewed by de Zoete et al. (2007)
Host genetics selection	It has been observed that different chicken lines showed different susceptibilities for Campylobacter colonization. Given significant difference in the susceptibility between individual chickens with same line under identical conditions, this approach requires an improved knowledge about interactions between Campylobacter and chicken.	Stern et al. (1990a); Boyd et al. (2005); Li et al. (2008)
III. Use antimicrobial alternatives to reduce/eliminate Campylobacter from colonized chickens
Bacteriophage therapy	Bacteriophages are viruses that can infect and kill susceptible bacteria. Some Campylobacter-specific phages reduced Campylobacter shedding about 2–3 log units in chickens under experimental conditions. The self-replicating nature and high specificity of phage make bacteriophages a potential alternative to conventional antibiotics. However, C. jejuni may develop resistance quickly in response to phage therapy and gain virulence genes via phage-mediated horizontal gene transfer. In addition, production of phage using Campylobacter host, a potential human pathogen that requires microaerophilic growth condition, raises concerns regarding safety and bulk production.	Reviewed by Connerton et al. (2008)
Bacteriocin treatment	Bacteriocins are antimicrobial peptides produced by bacteria with narrow or broad host ranges. The natural bacteriocins have considerable potential to fulfill the need for more effective antibiotics. Recently, several potent anti-Campylobacter bacteriocins have been identified in bacteria isolated from chicken intestine. These bacteriocins dramatically reduced Campylobacter colonization in chickens (>5–8 log reductions) and are being directed toward on-farm control of this pathogen in poultry.	Stern et al. (2005, 2006, 2008); Cole et al. (2006); Line et al. (2008); Svetoch et al. (2008)

Vaccination of Chickens Against Campylobacter

Campylobacter infections and chicken host immunity

Through oral ingestion, C. jejuni enters the host intestine and colonizes the distal intestine, primarily the cecum in chicken. Although Campylobacter was considered a commensal of the avian host, C. jejuni infection triggers both a systemic and mucosal immune response in chickens (de Zoete et al., 2007). C. jejuni–specific serum IgG, IgA, and IgM, and mucosal IgA and IgG increased after oral infection with C. jejuni (Myszewski and Stern, 1990; Cawthraw et al., 1994; Widders et al., 1996). Specifically, Campylobacter-specific serum IgG, IgA, and IgM levels were elevated gradually 2–3 weeks after experimental inoculation, and mucosal IgA rose 3–4 weeks after oral infection. The antibodies are directed against multiple Campylobacter antigens, among which flagellin is usually the first antigen to be recognized by all antibody isotypes (Cawthraw et al., 1994; Rice et al., 1997). The elevated levels of Campylobacter-specific antibodies are correlated with reduced colonization level of Campylobacter, suggesting a protective role of the antibodies in anti-Campylobacter infection in chickens. The Campylobacter maternal antibodies could also be vertically transferred from infected layer hens to newly hatched chickens (Sahin et al., 2001). The high-level of Campylobacter maternal antibodies in young chickens may partly contribute to the lack of Campylobacter infection in young broiler chickens in natural environments during the first 2 weeks of life, which was also supported by laboratory challenge experiments (Sahin et al., 2001, 2003). Together, these findings demonstrated the protective nature of Campylobacter-specific antibodies and supported the feasibility of development of immunization-based approaches to control Campylobacter infections in poultry.

Interaction between Campylobacter and chicken immune system

It is not surprising that Campylobacter-specific antibody response is slow and moderate in chickens because Campylobacter infection in chicken does not cause a strong inflammatory response or tissue damage in intestine. It is still largely unknown how Campylobacter interacts with the chicken immune system to trigger the immune response. Understanding the delicate interactions between Campylobacter and the chicken immune systems would greatly facilitate development of immunization-based approaches to control Campylobacter infections in poultry. In some studies, Campylobacter was also isolated from the spleen, liver, and blood in young chickens, suggesting that Campylobacter may invade intestinal epithelial cells and become systemic (Sanyal et al., 1984; Knudsen et al., 2006). Recent studies (Byrne et al., 2007; Van Deun et al., 2008) further demonstrated that C. jejuni could adhere to and invade chicken intestinal epithelial cells in vitro and in vivo. Notably, the in vitro invasiveness of C. jejuni was correlated with the magnitude of spleen colonization in C. jejuni–inoculated chickens. The C. jejuni strains that invaded chicken epithelial cells were not able to proliferate intracellularly, but quickly evaded from the cells. Therefore, Van Deun et al. (2008) proposed a novel colonization mechanism of C. jejuni by escaping rapid clearance through short-term epithelial invasion and evasion, combined with fast replication in the mucus. Interestingly, a recent report showed that C. jejuni also colonized the bursa of Fabricius of day-old chicks with 10⁴–10⁷ CFU/g of content in bursa for up to 28 days (Bingham-Ramos et al., 2008). Given that the bursa of Fabricius is an important immune organ in chickens, further examination of the colonization of C. jejuni in the bursa may provide novel information on the interaction between Campylobacter and the host immune system.

Some in vitro studies using chicken cells (e.g., primary chicken embryo intestinal cells, primary chick kidney cells, or chicken macrophage cell HD11) also provided compelling evidence that Campylobacter could stimulate the expression of proinflammatory cytokines and chemokines in chickens (Smith et al., 2005; Borrmann et al., 2007; Li et al., 2008). Recently, Smith et al. (2008) also reported that a significant induction of proinflammatory chemokin transcript was observed in both day-old and 2-week-old chickens upon infection with C. jejuni. These in vitro and in vivo studies indicated that C. jejuni could intimately interact with the chicken immune system to trigger an immune response although no pathological signs are observed for Campylobacter infection in chickens.

Antigenicity of Campylobacter components

Elucidation of immunogenic and protective antigens in C. jejuni is a primary step toward the design of effective vaccines. However, very limited studies have been done to characterize the immunological properties of Campylobacter components, primarily due to a lack of understanding of pathogenesis mechanisms and the antigenic complexity of this organism. Outer membrane proteins (OMPs) have been exploited and demonstrated as attractive targets of immune intervention in Gram-negative bacteria (Stern et al., 1990b; Lin et al., 2002a). Immunogenic OMPs identified in Campylobacter include flagellum (Fla) (Guerry, 1997), major outer membrane protein MOMP (Zhang et al., 2000), cell-binding factor Peb1 (Pei and Blaser, 1993), multidrug efflux pump component CmeC (Lin et al., 2002b, 2003, 2005b), and ferric enterobactin receptor CfrA (Zeng et al., 2008). Motility-mediating Fla is the best-characterized immunogenic protein shown to be required for Campylobacter colonization in birds and mammals (Morooka et al., 1985; Pavlovskis et al., 1991; Nachamkin et al., 1993; Wassenaar et al., 1993; Guerry, 1997). However, as an immunodominant protein in C. jejuni, Fla is modified by glycosylation and undergoes both phase and antigenic variation, which complicates the use of Fla for vaccination (Caldwell et al., 1985; Logan et al., 1989; Doig et al., 1996; Szymanski et al., 1999). Regarding another immunodominant protein MOMP, the definitive role of MOMP in microbe–host interaction is still not clear, and both conserved and variable regions were observed in MOMP (Zhang et al., 2000). The antigenicity of MOMP is also unique as reflected by predominant conformational epitopes in nature (Zhang et al., 2000). Peb1 functions as an outer membrane adhesin to mammalian cells and as an aspartate/glutamate-binding protein of an ABC transporter. However, it seems that Peb1 is localized mainly in the periplasm (de Zoete et al., 2007). CmeC is an essential OMP component of CmeABC efflux system that plays a critical role in multidrug resistance and pathogenesis (Lin et al., 2002b, 2003). Recent studies have shown that (1) CmeC is broadly expressed and highly conserved in C. jejuni, (2) CmeC is immunogenic in vivo, (3) CmeC is essential for colonization of Campylobacter in the intestine by mediating bile resistance, (4) expression of CmeC is dramatically induced by bile salts and highly upregulated in the intestinal tract, and (5) inhibition of CmeABC efflux pump by pump inhibitor increased susceptibility of C. jejuni to multiple antimicrobials and reduced in vivo colonization of C. jejuni in chicken (Lin et al., 2002b, 2003, 2005a, 2005b; Stintzi et al., 2005; Lin and Martinez, 2006; Martinez and Lin, 2006; Fakhr and Logue, 2007; Zeng and Lin, 2008). These findings strongly suggest that CmeC is an attractive and novel vaccine candidate that may not only prevent in vivo colonization of C. jejuni but also combat antibiotic resistance in C. jejuni. CfrA, an OMP component of ferric enterobactin iron acquisition system, is dramatically induced by iron-restricted condition and plays an essential role in colonization of C. jejuni in chickens (Palyada et al., 2004). Our recent work showed that CfrA is broadly distributed, expressed, and antigenically conserved among C. jejuni strains from various sources (Zeng et al., 2008). In addition, sera from Campylobacter-infected chickens showed vivid reaction with CfrA, indicating CfrA is also immunogenic in vivo (Zeng et al., 2008). Therefore, CfrA is another potential vaccine candidate against C. jejuni.

Vaccine development against Campylobacter in chickens

Vaccine development against Campylobacter in chickens has been comprehensively reviewed by de Zoete et al. (2007) recently. There is no vaccine available to date to control Campylobacter infections in poultry. A successful chicken vaccine should prevent colonization or cause a strong reduction of Campylobacter numbers in chickens (>2 log units) (de Zoete et al., 2007).

The following three approaches have been explored for developing effective and safe vaccine against Campylobacter in poultry:

1. Live attenuated vaccines. Because infection with wild-type C. jejuni strain induced anti-Campylobacter antibodies (Myszewski and Stern, 1990; Cawthraw et al., 1994; Widders et al., 1996), it is likely that a live attenuated vaccine will have a protective effect. However, experimental colonization with a noncolonizing C. jejuni strain did not protect upon homologous challenge (Ziprin et al., 2002). In addition, the paucity of information on the pathogenesis of the organism complicates this strategy.

2. Killed whole-cell vaccines. This type of vaccine could induce high protective immunity without the concern regarding potential pathogenesis to human. Vaccination with killed C. jejuni whole cells enhanced the immune responses and partly reduced colonization of C. jejuni in chickens (<2 log) (de Zoete et al., 2007).

3. Subunit vaccine. Successful development of subunit vaccine needs improved knowledge on immunogenic and protective antigens in C. jejuni.

Several studies have been focused on immunodominant antigen Fla with variable success (reviewed by de Zoete et al., 2007). However, Fla is modified by glycosylation and undergoes both phase and antigenic variation, which limits the application of Fla-based vaccines. The most encouraging vaccination study was published by a Polish group, in which oral vaccination of chickens with CjaA via a Salmonella carrier strain reduced C. jejuni colonization by 6 logs (Wyszynska et al., 2004). However, this finding is intriguing and needs to be confirmed because of the following two reasons. First, only two treatment groups (untreated chicken vs. vaccine treatment) were used in this study, and there was no Salmonella carrier strain control group included (Wyszynska et al., 2004). Therefore, it is likely that the protective effect observed in this vaccination trial was mediated by general boost of host immunity due to Salmonella infection instead of specific anti-CjaA antibodies. Second, a recent study (Wyszynska et al., 2008) indicated that CjaA is an N-glycosylated lipoprotein localized in the inner membrane of C. jejuni. Thus, it is difficult for specific CjaA antibodies to pass through outer membrane and gain access to CjaA, consequently conferring protection. Regarding CmeC and CfrA, the two promising vaccine candidates, the protective efficacy of these subunit vaccines needs to be determined in chicken in the future.

Oral delivery systems would be appropriate for Campylobacter vaccine in poultry as far as cost and simplicity of administration are concerned (Wagenaar et al., 2008). Particularly, successful identification of protective antigens as well as epitope mapping will lead to the development of inexpensive and practical oral vaccines for chickens to prevent Campylobacter infections using appropriate delivery systems, such as attenuated Salmonella-based vaccines (Curtiss et al., 1989) and genetically modified Lactobacillus (Mota et al., 2006). In conclusion, the short average life span of broiler chickens (∼6 weeks) poses a significant challenge to induce a strong antibody response against Campylobacter in chickens. To develop an effective vaccine against Campylobacter in poultry, three main challenges have been identified: (1) the identification of cross-protective antigens, (2) the induction of rapid and strong immune response, and (3) the development of novel adjuvants to further stimulate immunity against Campylobacter (de Zoete et al., 2007).

Bacteriocins to Reduce Campylobacter in Poultry

Bacteriocins

Bacteriocins are designated as the antimicrobial peptides (AMPs) produced by bacteria with narrow or broad host ranges (Hechard and Sahl, 2002; Riley and Wertz, 2002; Cotter et al., 2005). Bacteriocins are ribosomally synthesized, produced, and exported by almost every bacterial species examined to date for the apparent purpose of destroying their competitors (Riley and Wertz, 2002). Many bacteriocin-producing bacteria (e.g., lactic acid bacteria) are commensals in intestine (Riley and Wertz, 2002; Cotter et al., 2005; Sit and Vederas, 2008). Therefore, the intestinal bacteriocin–producing bacteria may achieve competitive advantage and function as an innate barrier against pathogens in the gut. Bacteriocins are classified into modified bacteriocins (Class I bacteriocins, such as nisin) and unmodified bacteriocins (Class II bacteriocins, such as the anti–C. jejuni bacteriocins described below) (Hechard and Sahl, 2002; Riley and Wertz, 2002; Sit and Vederas, 2008). Despite the existence of a broad diversity in bacteriocin sequences and structures, it has been widely accepted that bacteriocins and other host defense peptides share a common theme in the mechanism of killing action by disruption of membrane integrity (Hechard and Sahl, 2002; Riley and Wertz, 2002; Yeaman and Yount, 2003). Generally, AMPs directly interact with target cells via initial electrostatic and hydrogen bond attraction, and then disrupt the structure or function of the bacterial membrane by permeating lipid bilayers, forming a transmembrane pore, and ultimately leading to cell death. However, transmembrane pore formation is not the only mechanism of bacterial killing by bacteriocins (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). For example, nisin, a bacteriocin widely used for food biopreservation, also has other modes of antimicrobial action, such as inhibition of cell-wall biosynthesis, inhibition of lipid bilayer function, inhibition of spore outgrowth, and activation of autolytic enzyme (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). Detailed information on bacteriocin evolution, structure–function relationships, and mode of action are available in several excellent reviews (Hechard and Sahl, 2002; Riley and Wertz, 2002; Cotter et al., 2005; Peschel and Sahl, 2006).

Potential of bacteriocins as new antimicrobials

Bacterial pathogens are increasingly resistant to currently available antibiotics, and new antimicrobials are needed to combat multidrug resistance (Walsh, 2003). Bacteriocins have considerable potential for the design and production of new antimicrobials (Cleveland et al., 2001; Cotter et al., 2005; Kirkup, 2006; Galvez et al., 2007; Rossi et al., 2008; Sahl and Bierbaum, 2008; Sit and Vederas, 2008). In contrast to traditional antibiotics, bacteriocins are considered natural and nontoxic on eukaryotic cells because they are found commonly in food animal products and thus have been consumed for centuries (Cleveland et al., 2001; Galvez et al., 2007). In fact, two bacteriocins, nisin and pediocin PA1/AcH, have been widely used in the food industry for food biopreservation, and no toxicity due to these bacteriocins has been demonstrated (Cleveland et al., 2001; Cotter et al., 2005; Galvez et al., 2007).

From standpoint of antimicrobial development, the emergence of bacteriocin resistance is a concern, either for food preservation or for therapeutic treatment. Because nisin is the only bacteriocin licensed as a food preservative and many potential bacteriocins are still under development, limited information is available directly addressing the development and mechanisms of bacteriocin resistance. Both Gram-positive and Gram-negative bacteria can develop resistance to bacteriocins (e.g., nisin), and the mechanism of bacteriocin resistance appears to be complex and involves various structural and physiological changes in the bacterial cell envelope (Ennahar et al., 2000; Cleveland et al., 2001; Breukink and de Kruijff, 2006; Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). Intriguingly, it seems that bacteria have not developed highly effective mechanisms to resist natural AMPs, including bacteriocins (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). This feature is very different from many therapeutic antibiotics for which bacteria can develop high-level of resistance. Recently, it has been proposed that bacteriocins may have multiple low-affinity targets and cause pleotropic effects on various bacterial targets. Therefore, it is possible that such low-affinity interactions of bacteriocins with multiple targets are not favorable for the development of bacterial resistance. In contrast, many therapeutic antibiotics act on a single, high-affinity target, which makes it comparatively easy for bacteria to develop resistance, particularly high-level resistance (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). Together, bacteriocins have considerable potential to fulfill the need for more effective antimicrobial agents. Unlike the antibiotics that act on a single target, there is less in the way of resistance development for bacteriocin-based antimicrobials.

Anti-Campylobacter bacteriocins

In the past 3 years, significant progress has been made toward isolation of chicken commensal bacteria inhibitory to Campylobacter and characterization of associated bacteriocins from these bacteria (Stern et al., 2005, 2006; Svetoch et al., 2005, 2008; Cole et al., 2006; Line et al., 2008; Nazef et al., 2008). Several potent anti-Campylobacter bacteriocins have been purified and characterized in bacteria isolated from the chicken intestinal tract, which includes SRCAM 602 from Paenibacillus polymyxa (Stern et al., 2005; Svetoch et al., 2005), OR-7 from Lactobacillus salivarius (Stern et al., 2006), and E-760 and E 50–52 from Enterococcus spp. (Line et al., 2008; Svetoch et al., 2008). These bacteriocins also dramatically reduced C. jejuni colonization in poultry and are being developed for on-farm control of Campylobacter to protect public health. The bacteriocins that reduced Campylobacter colonization in poultry are summarized in Table 2.

Table 2.

Overview of Bacteriocins That Reduced Campylobacter Colonization in Poultry

Bacteriocin-producing bacteria
Name	Species	Bacteriocin name	Effecta	References
NRRL B-30509	Paenibacillus polymyxa	SRCAM 602	ND (4.6–6.3 log reduction in 10-day-old chickens)	Stern et al. (2005); Svetoch et al. (2005)
			ND (>4 log reduction in 13-day-old turkeys)	Cole et al. (2006)
NRRL B-30514	Lactobacillus salivarius	OR-7	>6 log reduction in 10-day-old chickens	Stern et al. (2006)
			ND (>4 log reduction in 13-day-old turkeys)	Cole et al. (2006)
NRRL B-30745	Enterococcus durans/faecium/hirae	E-760	ND (>6.6 log reduction in 10-day-old chickens with dose as low as 31.2 mg/kg feed)	Line et al. (2008)
			ND (2.2–5.0 log reduction in 42-day-old chickens with dose of 125 mg/kg feed)
NRRL B-30746	Enterococcus faecium	E 50–52	ND (>6.4 log reduction in 15-day-old chickens with dose as low as 31.2 mg/kg feed)	Svetoch et al. (2008)
			>5.3 log reduction with one day treatment of 35–41-day-old broilers (12.5 mg of E 50–52/liter of drinking water)

^aND, no Campylobacter was detected in all birds after bacteriocin treatment with minimum detection of 100 CFU/g cecal contents. Unless specifically clarified, treated birds were provided specific bacteriocin at a dose of 250 mg/kg feed for 3 consecutive days. The bacteriocins were mixed with polyvinylpyrrolidone power to produce microencapsulated bacteriocins, which were used to make a medicated feed.

Svetoch et al. (2005) first reported the identification and characterization of novel anti-Campylobacter bacteriocins from bacteria isolated from chicken intestine. This study involved screening 365 representative Bacillus and Paenibacillus isolates for activity against C. jejuni. One B. circulans and three P. polymyxa strains displayed strong antagonism to C. jejuni. The anti-Campylobacter activity was later observed to result from the secreted protein component of these antagonists. Stepwise purification using ammonium sulfate precipitation, Superose-12 gel filtration, and Mono Q anion-exchange chromatography resulted in the identification of a short peptide (∼3.5 kDa) with purity up to 98.8%. The purified peptides displayed potent anti-Campylobacter activity and were stable at high temperature (100°C, 15 min) and a wide pH range (3.0–9.0). These peptides lost their activity after being treated with β-chymotrypsin, proteinase K, and papain but retained activity when treated with lysozyme or lipase. Amino acid sequences of these peptides indicated that these peptides are consistent with class IIa bacteriocins because of the conserved N-terminal sequence of Tyr-Gly-Asn-Gly-Val and two cysteine amino acids forming a disulfide bridge at the N-terminal portion of the peptide (Svetoch et al., 2005). The biochemical and anti-Campylobacter activities of these bacteriocins indicate that these peptides represent a new and unreported group of bacteriocins.

To determine if above novel bacteriocins could reduce Campylobacter colonization in poultry, one bacteriocin, SRCAM 602 produced by P. polymyxa NRRL B-30509, was evaluated in chickens (Stern et al., 2005) and Turkeys (Cole et al., 2006). For both trials, the purified SRCAM 602 bacteriocin was microencapsulated in polyvinylpyrrolidone and then mixed with commercial feed to produce medicated feed with final bacteriocin concentration of 250 mg/kg feed. For the chicken studies, 1-day-old chickens were orally inoculated with one of four C. jejuni strains (Stern et al., 2005). When colonization of C. jejuni in chickens was well established by 7 days of age, the chickens were provided with either nonmedicated feed or bacteriocin-embedded feed (10 birds per group) for 3 consecutive days. Chickens receiving nonmedicated feed displayed high levels of colonization with C. jejuni (6.6–8.3 log CFU/g of feces). However, none of the bacteriocin-treated chickens were colonized with C. jejuni (detection limit of 2 log CFU/g of feces). This finding was highly consistent in duplicated experiments for four different C. jejuni strains. The efficacy of this bacteriocin was also observed in turkey (Cole et al., 2006). Turkey poults were orally challenged with a mixture of three C. coli isolates at 3 days of age. From day 10–12 posthatch, the turkeys received either nonmedicated feed or the bacteriocin-emended feed (10 birds per group). Oral administration of the purified bacteriocin SRCAM 602 eliminated detectable cecal Campylobacter colonization in all turkeys in three separate trials. Together, these findings strongly suggest that bacteriocins are effective in vivo and that bacteriocin treatment of colonized poultry may represent an effective intervention strategy against Campylobacter colonization in poultry.

Lactic acid bacteria such as Lactobacillus spp. are widely used probiotic organisms. Many lactic acid bacteria produce bacteriocins with different spectra ranges of inhibition (Cotter et al., 2005; Galvez et al., 2007). Therefore, Stern et al. (2006) also evaluated anti-Campylobacter activity among >1200 isolates of different lactic acid bacteria. One isolate, Lactobacillus salivarius NRRL B-30514, displayed highest anti-Campylobacter activity. Bacteriocin OR-7 from this strain was purified using ammonium sulfate precipitation, followed by SP Sepharose cation exchange and Octyl-Sepharose hydrophobic interaction chromatography. The purified bacteriocin OR-7 was also resistant to high temperature (90°C, 15 min) and a wide pH range (3.0–9.1). The amino acid sequence of OR-7 bacteriocin (55 aa residues) was also consistent with class IIa bacteriocins. The inhibitory effect of OR-7 on C. jejuni colonization in chicken was then evaluated with the same experimental design as that for SRCAM 602 described above. Bacteriocin OR-7 treatment consistently reduced Campylobacter colonization more than 1 million fold in chicken. Therefore, the bacteriocin OR-7 from L. salivarius NRRL B-30514 also has significant potential to reduce C. jejuni load in poultry.

Recently, the same research group identified another two novel bacteriocins, E-760 and E 50–52, which are produced by two different Enterococcus spp. isolated from broiler ceca (Line et al., 2008; Svetoch et al., 2008). The highly purified bacteriocins were obtained from culture supernatants by initial ammonium sulfate precipitation followed by ion-exchange and hydrophobic-interaction chromatography. Amino acid sequence analysis indicated that these bacteriocins also belong to Class IIa bacteriocins. These two bacteriocins also displayed tolerance to high temperature and a wide range of pH. The biochemical characteristics of all anti-Campylobacter bacteriocins are summarized in Table 3. Both E-760 and E 50–52 bacteriocins not only displayed potent anti-Campylobacter activity but also showed strong antibacterial activity against a broad spectrum of foodborne pathogens, such as Salmonella spp., E. coli O157:H7, Listeria spp., and Shigella spp. Chicken trials further showed that treatment of E-760–embedded feed eliminated detectable cecal Campylobacter colonization in both young chickens (10-day-old) and market-aged broiler chickens (42-day-old) (Line et al., 2008; Svetoch et al., 2008). With respect to bacteriocin E 50–52, young chickens received E 50–52–emended feeds at different dose (31.2, 62.5, or 125 mg/kg feed) from day 4 to day 7. All birds were sacrificed on day 15, and cecal samples were collected for enumeration of viable C. jejuni. Despite high levels of C. jejuni colonization in control chickens (8.40 log₁₀ CFU/g feces), no Campylobacter was detected in all chickens treated with E 50–52 even at 8 days after termination of the bacteriocin treatment, suggesting that the bacteriocin treatment completely eliminated C. jejuni from chicken intestine (Svetoch et al., 2008). E 50–52 was also very effective to reduce C. jejuni colonization in adult bird (35–41-day-old broilers). All the market-aged birds were environmentally colonized by C. jejuni at a level of about 8.00 log₁₀ CFU/g feces, and 1 day treatment via drinking water (12.5 mg of E 50–52/liter of drinking water) dramatically reduced C. jejuni colonization in chicken intestine (>5.3 log reduction).

Table 3.

Biochemical Characteristics of Anti-Campylobacter Bacteriocins

				Retained activity when treated with
Bacteriocina	Lengthb	Molecular mass (Da)	Isoelectric point	Enzymec	Heat	pH	Reference
SRCAM 602	39	3864	7.2	Lysozyme lipase	100°C, 15 min	3.0–9.0	Stern et al. (2005); Svetoch et al. (2005)
OR-7	54	5123	9.5	Lysozyme lipase	90°C, 15 min	3.0–9.1	Stern et al. (2006)
E-760	62	5362	9.5	Lysozyme lipase	100°C, 5 min	5.0–8.7	Line et al. (2008)
E 50–52	39	3340	8.0	Lysozyme lipase	100°C, 15 min	3.0–8.4	Svetoch et al. (2008)

^aAll anti-Campylobacter bacteriocins belong to class IIa bacteriocins.

^bTotal number of amino acid residues of purified active bacteriocin.

^cSpecific bacteriocin was incubated with enzyme for 3 h of incubation at 37°C.

In summary, four anti-Campylobacter bacteriocins, produced by bacteria isolated from chicken intestine, have been successfully purified and characterized (Table 3). Oral administration of these bacteriocins dramatically reduced C. jejuni colonization in chicken intestine (Table 2). Therefore, these natural bacteriocins have been proposed as effective alternatives to therapeutic antibiotics and were being directed for on-farm control of Campylobacter in poultry (Casewell et al., 2003; Stern et al., 2005, 2006; Svetoch et al., 2005, 2008; Line et al., 2008).

Key issues on bacteriocins-based intervention strategy

Although the above anti-Campylobacter bacteriocins are very effective in reducing C. jejuni colonization in poultry, several important issues (e.g., production, safety, and development of resistance) need to be addressed for future regulatory approval and public acceptability of this intervention measure.

Purification of the four bacteriocins has been well established and standardized (Svetoch et al., 2005, 2008; Stern et al., 2006; Line et al., 2008). The purification generally involved two steps: (1) crude bacteriocin (∼9% purity) preparation from the supernatant using ammonium sulfate precipitation, and (2) bacteriocin purification from the crude preparation using two different chromatography columns, which finally results in bacteriocin purity up to 98.8%. To make bacteriocin usage in poultry economically feasible, it is important to improve the yield during production. Recently, Stern et al. (2008) observed that bacteriocin yield was dramatically increased when bacteriocin-producing strains were cultured together with a novel inducer strain and inducer peptide. For example, approximate 214–225 mg of high-purity bacteriocin OR-7 could be purified from 1 L of culture liquid under modified culture condition using the novel inducer strain and peptide. Therefore, this modified procedure provides an economic method for producing large quantities of bacteriocins for commercial use.

Bacteriocins are widely considered natural and safe to animals and humans (Cleveland et al., 2001; Galvez et al., 2007). Some bacteriocins have been used as food preservatives for a long time (Cleveland et al., 2001; Cotter et al., 2005; Galvez et al., 2007). All identified anti-Campylobacter bacteriocins are produced by commensal bacteria from chickens that have been consumed by humans for centuries. Therefore, these bacteriocins are likely nontoxic to both poultry and humans. However, to address this issue definitively, more research is needed to determine toxicity of the bacteriocins using cell culture and/or animal model systems. In addition, to date it is still not clear if oral administration of the bacteriocins could cause significant absorption of bacteriocin in intestine, consequently leading to the presence of the bacteriocin in chicken blood and tissues. Given the molecular mass of the bacteriocins (3.3–5.4 kDa; Table 3), absorption and diffusion of intact bacteriocins through the epithelial barrier may not be efficient. Svetoch et al. (2008) observed that oral administration of bacteriocin E 50–52 in chicken resulted in the significant reductions of Salmonella enteritidis in the liver and spleen, which suggests that the bacteriocin can enter the systemic system by intestinal absorption. However, it is also possible that bacteriocins, like many other AMPs, could function as a potent immune modulator and directly enhance the innate immune response to fight Salmonella infections (Finlay and Hancock, 2004). This hypothesis remains to be studied in the future.

The development of antibiotic resistance is inevitable in bacteria, and every antibiotic that is introduced into market to date has a limited shelf life (Walsh, 2003). However, unlike high-affinity small-molecule antibiotics, bacteria have not developed highly effective mechanisms to resist natural AMPs, including bacteriocins, which is likely because of the multiple modes of action of AMPs (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). Therefore, compared to conventional antibiotics that act on a single target, there is less in the way of resistance development for peptide antibiotics, such as bacteriocin-based antimicrobials (Peschel and Sahl, 2006; Sahl and Bierbaum, 2008). Our recent study (Hoang et al., 2008) also supported this hypothesis. Examination of 146 Campylobacter isolates of various origins revealed only one isolate displaying resistance to bacteriocin OR-7 (MIC = 64 μg/mL), while all other isolates showed low MICs ranging from 0.25 to 1.0 μg/mL. We also observed that Campylobacter could develop resistance to bacteriocin OR-7 at low frequency in vitro. All in vitro–selected mutants only displayed low-level resistance (4–16-fold increase in MIC) to OR-7. The multidrug efflux pump CmeABC contributes to the intrinsic and acquired resistance of C. jejuni to bacteriocin OR-7 (Hoang et al., 2008). Despite this recent progress in understanding the development and mechanism of bacteriocin resistance in Campylobacter, it is unknown if therapeutic usage of bacteriocins in poultry would promote the emergence of bacteriocin-resistant Campylobacter mutants in vivo. If so, can Campylobacter develop high-level bacteriocin resistance in response to therapeutic treatment with bacteriocins? It is also unclear if the bacteriocin-resistant Campylobacter could persist in the absence of antimicrobial selection pressure in vitro and in vivo. The specific targets of anti-Campylobacter bacteriocins and the molecular basis of bacteriocin resistance in Campylobacter are still largely unknown. Answering these questions should greatly improve our understanding on the modes of action of bacteriocins, provide helpful information for risk assessment, and facilitate the development of more sustainable bacteriocins for on-farm control of Campylobacter in poultry.

Conclusion

It is widely accepted that contamination of poultry by Campylobacter is a significant risk factor of human campylobacteriosis. Thus, the prevention and control of C. jejuni in poultry would reduce the risk of human exposure to Campylobacter and is an important food safety issue. However, there is no effective, reliable, and practical intervention measure available to reduce C. jejuni in poultry to date. Biosecurity measures are practical, but hygiene barriers could still be broken through. All other potential measures are still under development. Vaccination of chickens against C. jejuni is a feasible strategy but requires a good deal of basic research to reveal protective antigens, to examine delicate interaction between C. jejuni and chicken immune system, and to optimize vaccination regimen (e.g., mucosal adjuvant and delivery systems). Anti-Campylobacter bacteriocin treatment is clearly an effective and feasible strategy to reduce C. jejuni load in market-aged chickens. More research is needed to address several key issues (e.g., production, safety, and development of resistance) of bacteriocin application for future regulatory approval and public acceptability of this promising intervention measure.

Acknowledgments

We thank Dr. Norman J. Stern (USDA) for providing us bacteriocins and C. jejuni isolates for examination of bacteriocin resistance in Campylobacter. The research in our laboratory described in this review was supported by NIH/NIAID R21 AI072551-01A1 and Tennessee Agricultural Experiment Station.

Disclosure Statement

No competing financial interests exist.