Passive immunization to reduce Campylobacter jejuni colonization and transmission in broiler chickens

Campylobacter jejuni is the most common cause of bacterium-mediated diarrheal disease in humans worldwide. Poultry products are considered the most important source of C. jejuni infections in humans but to date no effective strategy exists to eradicate this zoonotic pathogen from poultry production. Here, the potential use of passive immunization to reduce Campylobacter colonization in broiler chicks was examined. For this purpose, laying hens were immunized with either a whole-cell lysate or the hydrophobic protein fraction of C. jejuni and their eggs were collected. In vitro tests validated the induction of specific ImmunoglobulinY (IgY) against C. jejuni in the immunized hens’ egg yolks, in particular. In seeder experiments, preventive administration of hyperimmune egg yolk significantly (P < 0.01) reduced bacterial counts of seeder animals three days after oral inoculation with approximately 104 cfu C. jejuni, compared with control birds. Moreover, transmission to non-seeder birds was dramatically reduced (hydrophobic protein fraction) or even completely prevented (whole-cell lysate). Purified IgY promoted bacterial binding to chicken intestinal mucus, suggesting enhanced mucosal clearance in vivo. Western blot analysis in combination with mass spectrometry after two-dimensional gel-electrophoresis revealed immunodominant antigens of C. jejuni that are involved in a variety of cell functions, including chemotaxis and adhesion. Some of these (AtpA, EF-Tu, GroEL and CtpA) are highly conserved proteins and could be promising targets for the development of subunit vaccines.


Introduction
Today, campylobacteriosis is the most frequently reported zoonotic diarrheal disease worldwide, with Campylobacter jejuni as the major causative agent [1]. Despite many efforts aimed at minimizing Campylobacter contamination of poultry, no effective, reliable intervention measures exist to reduce bacterial numbers in the broiler gut [2]. In general, broiler chickens become colonized from the age of two weeks onward [3]. Transportation of immunoglobulin (Ig) Y, the major Ig class in chickens, from hen to embryo via egg yolk is believed to play a key role in the protection of young chicks with an immature immune system against Campylobacter colonization during two to three weeks post-hatch [4,5]. From two weeks onward, the concentration of maternally derived anti-Campylobacter IgY drops significantly, which coincides with an increased colonization susceptibility of the broiler chicks. Passive immunization of chicks may be prolonged by feeding broilers with high levels of anti-Campylobacter antibodies recovered from immunized hens [4]. Following immunization, specific IgY is induced and transferred from the serum to the egg yolk [4], where it is accumulated at high levels [6]. Indeed, preincubating C. jejuni with IgY from immunized hens has previously been shown to reduce fecal C. jejuni counts in broilers experimentally inoculated with this mixture [7]. In this study, also the effect of IgY on colonization in already-colonized animals was assessed. Results indicated that IgY induced only a limited therapeutic efficacy. After an initial drop, fecal C. jejuni numbers regained their original counts when treatment was stopped. Nevertheless, these observations indicate that IgY preparations from egg yolks could be a promising candidate to reduce C. jejuni colonization in broilers, but need to be optimized because no studies regarding this topic were reported since. Moreover, in the study by Tsubokura et al. [7] fecal bacterial counts were determined, which may not perfectly correlate to determining cecal counts, which is more sensitive [8]. In the past two decades, some promising observations have been reported concerning the use of chicken egg yolk Igs in animal models and humans for the prevention and treatment of other bacterial and viral diseases, including streptococci [9], Helicobacter pylori [10], human rotavirus, enterotoxigenic Escherichia coli, Yersinia ruckeri and Salmonella species [4,11]. In these studies, protection was primarily observed after infection with the homologous strain.
Passive immunization using yolk IgY thus seems a promising strategy to control C. jejuni colonization in broiler flocks. For this purpose, the first aim of this study was to examine whether hyperimmune egg yolk could reduce or prevent cecal C. jejuni colonization in broiler chicks. The second objective was to identify the C. jejuni antigens reacting with C. jejuni-specific IgY raised in chicken eggs.

Experimental animals
Commercial brown Leghorn laying hens and day-ofhatch Ross broiler chickens of both sexes from a local farm were raised in group until treatment. Birds were provided with a commercial feed and water ad libitum. Husbandry, euthanasia methods, experimental procedures and bio-safety precautions were approved by the Ethical Committee (EC) of the Faculty of Veterinary Medicine, Ghent University, Ghent, Belgium (EC numbers: 2010/174 and 2011/029). Chicks were examined for the presence of Campylobacter in mixed fecal samples using standard methods as described in [12] and proved to be free of Campylobacter.

Extraction of C. jejuni hydrophilic and hydrophobic proteins
To prepare the immunizing agent, cells from a C. jejuni culture were collected by centrifugation (5000 × g for 30 min at 4°C), washed in HBSS and sonicated on ice for 30 s in extraction buffer 1 (EB1; 40 mM Tris, supplemented with tributylphosphine solution, protease inhibitor cocktail, phosphatase inhibitors PP2 and PP3, DNase and RNase (Sigma-Aldrich, Steinheim, Germany) using a tip sonicator (Qsonica, Newtown, VS). After each sonication step, 12 in total, the sample was left on ice for 30 s. After centrifugation at 16 000 × g for 10 min at 4°C the supernatant (containing the hydrophilic proteins) was collected and the pellet resuspended in extraction buffer 2 (EB2; 40 mM Tris, supplemented with 5 M urea, 2 M thiourea, 2% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), 0.2% carrier ampholytes, 100 mM dithiothreitol (DTT) and the respective inhibitors as mentioned above). After sonication and centrifugation of the latter suspension at 5000 × g for 3 min to remove cell debris and non-lysed cells, the supernatant (containing the hydrophobic proteins) was either pooled with the first supernatant (to obtain the whole-cell protein lysate) or stored (−20°C) separately.

Immunization of laying hens
Campylobacter-free commercial Brown Leghorn chickens were randomly assigned to vaccination groups at the age of 19 weeks. Chickens were immunized with either 50 μg of the C. jejuni KC40 whole-cell lysate (EB1 + EB2) or 25 μg protein of the hydrophobic protein fraction only (EB2), by intramuscular injection of 250 μL of a 1:1 mixture with Freunds' Complete Adjuvant (FCA) at four different sites of the pectoral muscle. Chicks of the control group were immunized with a mixture of HBSS and FCA. Starting from the age of 21 weeks, three booster immunizations using Freunds' Incomplete Adjuvant (FIA) were given in a two-weekly time interval. Starting from one week after the second boost, eggs of the animals were collected and stored at 4°C.

Determination of egg yolk and white IgY, IgA and IgM titers
At weekly time intervals, starting from the day of the first immunization, eggs were collected from both groups. The egg yolk and white were separated and pooled per group, diluted 1:5 (wt/vol) in distilled water and mixed thoroughly. After overnight incubation at 4°C, the supernatant, containing the water-soluble fraction of the egg yolk and white, respectively, was collected for immunoglobulin quantification using enzyme-linked immunosorbent assay (ELISA). Wells of a 96-well Nunc-MaxiSorp microtiter plate were coated with 50 μL of a suspension of C. jejuni KC40 cells (2 × 10 7 cfu/ mL) in coating buffer (2.16 g Na 2 CO 3 .10H 2 O + 1.935 g NaHCO 3 in 500 mL H 2 O). Bacterial cells were either heat-killed or processed to antigenic fractions by sonication as described above. Plates were incubated overnight at 4°C, after which the wells were washed three times with HBSS followed by a final wash step using wash buffer (WB; PBS + 0.1% Tween 20). Next, the wells were incubated with 100 μL blocking buffer (BB; WB + 1% bovine serum albumin) at room temperature (RT) for 1 h to reduce non-specific binding of antibodies. Subsequently, 100 μL of two-fold dilutions (in BB) of the supernatant were added to the wells in triplicate. The plates were incubated at RT for 90 min. After incubation, the wells were washed three times using HBSS and once using WB and 100 μL of 1/10000 (in WB) of horseradish peroxidase (HRP)-labelled anti-chicken IgY (Sigma), 1/20000 HRPlabelled goat anti-chicken IgA (Bio-Connect, The Netherlands) or 1/10000 HRP-labelled goat antichicken IgM (Gallus Immunotech Inc., Canada) were added to each well. After incubation and washing as described above, 50 μL of 3,3′,5,5′-tetramethylbenzidine (TMB; Sigma) substrate was added. After 10 min the reaction was blocked by adding 50 μL 0.5 M H 2 SO 4 to the wells. The absorbance was then measured at 450 nm using an automated spectrophotometer. Antibody titres from egg yolks/whites of immunized hens were reported as the highest dilution where the optical density (OD) was greater than the OD + three standard deviations of wells containing yolk/white originating from birds that only received adjuvant.

Purification of IgY fraction from egg yolks
IgY was purified from the egg yolks according to the method of Bird and Thorpe [14].

Collection of broiler chicken intestinal mucus
Commercial and Campylobacter-free 14-d-old broiler chicks were euthanized and the small intestine was collected and gently rinsed with phosphate-buffered saline (PBS) to remove fecal material. The mucus was scraped from the mucosa with a glass slide covered in parafilm, diluted 1:3 with N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES, 25 mM, pH 7.4) and vortexed. The solution was centrifuged three times at 2000 × g for 15 min at 4°C. The pooled supernatant containing the crude mucus was centrifuged two times more at 5000 × g for 15 min at 4°C and stored at −80°C. Protein content was determined using a Coomassie (Bradford) Protein Assay Kit according to the manufacturer's instructions (Biorad, Nazareth, Belgium).

Mucus adhesion test
The effect of IgY on the adherence of C. jejuni to intestinal mucus was examined for the homolog C. jejuni strain KC40 as well as for five other C. jejuni strains (10kf-1.16, 7P6.12, 10C-6.1, 10kf-4.12 and 10VTDD-8). Bacteria were first incubated for 30 min at 42°C in HBSS only or HBSS supplemented with purified IgY from C. jejuni-immunized or HBSS/sham-immunized hens. Mucus was diluted to a final concentration of 250 μg protein/mL in coating buffer. One hundred μL were immobilized per well of a microtiter plate (Maxisorp, Nunc) and incubated overnight at 4°C. The wells were washed three times with HBSS and saturated with 1% (wt/ vol) bovine serum albumin (BSA) in HBSS for 1 h at RT. After washing the wells two times with HBSS, 100 μL of the C. jejuni-IgY mixture were transferred to the wells. After a 1 h incubation at 42°C, the wells were washed 15 times to remove unbound bacteria. Wells were thereafter treated with 200 μL 0.5% (vol/vol) Triton X-100 to release adherent bacteria and incubated for 30 min at RT while shaking. Next, 300 μL HBSS were added to each well and 10-fold dilutions of the wells were titrated on mCCDA plates.

Motility assay
To assess the influence of C. jejuni-specific IgY on C. jejuni motility of the six strains used in the study, bacteria were pre-incubated with purified IgY from yolks of either C. jejunior sham/HBSS-immunized laying hens and 20 μL were pipetted onto semi-solid Mueller-Hinton (MH) agar. After 24 h incubation at 42°C under microaerobic conditions, the diameter of Campylobacter growth from both conditions was measured.

Multilocus sequence typing
To determine the relatedness between the C. jejuni strains used in this study, all strains were characterized by multilocus sequence typing (MLST) according to [15]. For determination of the sequence type (ST) and clonal complex (CC) all allelic sequences were queried against the online C. jejuni MLST database [16].
Effect of in-feed hyperimmune egg yolk on transmission of and cecal colonization with C. jejuni in two-week-old broilers In trial 1, day-of-hatch Campylobacter free broiler chicks (n = 22) were raised in group. At 6 days of age the chicks were randomly assigned to 2 groups (n = 11/group) and housed in separate isolating chambers. Animals of group 1 were provided with feed containing 5% (wt/wt) egg yolk (mixed manually through the feed) from hens sham-immunized with HBSS/adjuvant, while birds of group 2 were fed 5% (wt/wt) egg yolk from hens immunized with a whole-cell C. jejuni lysate/adjuvant mixture.
Egg yolks were added to the feed for the remainder of the experiment. Equal amounts of feed and drinking water were provided for each group during treatment and care was taken that all animals had unlimited access to the feed and water. At the age of 10 days, three chicks of both groups were orally inoculated with approximately 8 × 10 3 cfu of C. jejuni strain KC40. At 13 days of age all animals were euthanized (as described above) and the ceca as well as their contents were collected for C. jejuni enumeration (see below).
In trial 2, day-of-hatch broiler chicks (n = 54) were raised in group. At 6 days of age the chicks were randomly assigned to 6 groups (n = 9/group) and housed in separate isolating chambers. Animals of groups 1, 2 and 3 were provided with feed containing 5% (wt/wt) egg yolk from hens immunized with HBSS/adjuvant, while birds of groups 4, 5 and 6 were fed 5% (wt/wt) egg yolk from hens immunized with the hydrophobic protein fraction of C. jejuni mixed with adjuvant. At the age of 10 days, three chicks of each group were orally inoculated with approximately 3 × 10 4 cfu of C. jejuni strain KC40. At 13 days of age all animals were euthanized (as described above) and the ceca as well as their contents were collected for C. jejuni enumeration (see below).

Cecal Campylobacter jejuni enumeration
Ceca and contents were cut into small fragments, weighed, and diluted 1:9 (wt/vol) in NB2 with supplements. After homogenization, a 10-fold dilution series was made in HBSS. Of each dilution, 100 μL were spread on mCCDA plates. After 22 h incubation at 42°C under microaerobic conditions, characteristic colonies were counted. For enrichment, diluted cecal samples in NB2 were incubated at 37°C under microaerobic conditions. After 24 h, samples were plated on mCCDA and incubated at 42°C in a microaerobic environment. After 24 and 48 h plates were examined for the presence or absence of C. jejuni.

Two-dimensional gel-electrophoresis and Western blot analysis
To identify the immunodominant antigens of C. jejuni recognized by yolk IgY of immunized hens, the whole C. jejuni lysate was separated by iso-electric focussing (pH 3-10), followed by sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE) with a 4% stacking gel and 10% separating gel at 150 V for 30 min and 200 V for 60 min as previously described [17]. The separated proteins were transferred electrophoretically (50 V) onto nitrocellulose membranes (Bio-Rad) for 3 h. After washing with milliQ water, the membrane blots were blocked in 5% (wt/vol) milk in PBS for 1 h at RT. Subsequently, the blots were incubated overnight with purified egg yolk IgY, diluted to 1:5000 (vol/vol) in 5% (wt/vol) milk. After washing with 0.05% (vol/vol) Tween-20, the blots were incubated with horseradish peroxidaseconjugated goat-anti-chicken immunoglobulin G at a dilution of 1:50 000 in 5% (wt/vol) milk for 1 h. The blots were washed and proteins were immunodetected by enhanced chemiluminescence, using Supersignal West Dura Extended Duration Substrate (Pierce), and scanned and digitized using the Versa Doc imaging system. Blotting experiments were performed in duplicate to identify antigens recognized by IgY in yolk from immunized/ sham-immunized laying hens.

In-gel digestion and mass spectrometric analysis
Spots of interest (C. jejuni antigens reacting with hyperimmune egg yolk IgY) were cut out of the gel and processed as previously described [18]. Prior to mass spectrometry the isolated peptides were separated on a U3000 nano-HPLC device (Dionex) as previously described [19].
Identification of the peptides was performed on an electrospray ionization quadrupole time of flight (ESI-QUAD-TOF) spectrometer (Ultima, Waters). Peptides were acquired in the range of 450 to 1650 m/z in MS. The seven most abundant multiply charged ions with a minimum intensity of 60 counts per second were subjected to MS/MS (between 50 and 2300 m/z). m/z ratios selected for MS/MS were excluded for 150 s.
Data analysis was performed against the Campylobacter protein databank (124864 sequences) from NCBI using the in-house search engine Mascot Daemon (2.3, Matrix Science, London, UK). Carbamidomethyl (N-term), deamidation (NQ) and oxidation (M) were set as variable modifications. Peptide mass tolerance and fragment mass tolerance were set at 0.35 Da and 0.6 Da, respectively. Maximum two miss cleavages were allowed. Proteins were only considered to be positively matched if the significance was below 0.01 (p ≤ 0.01) and at least one peptide passing the required bold red criteria from Mascot Daemon, indicating that at least one peptide had rank 1 and a significance below 0.01.

Statistical analysis
Data were analyzed by SPSS 17.0 software for Windows. The significance level α was set at 0.05. Campylobacter counts were first transformed to log 10 counts before statistical analysis. For the in vivo trials, a nonparametric Mann-Whitney U test was carried out to compare the means of log 10 transformed counts in chicken cecal contents (of seeders, contact animals or both) of all groups (treated and control groups). For the mucus adhesion test, a one-way analysis of variance (ANOVA) was carried out to compare the means of log 10 transformed adherent bacteria of all groups (treated groups and control groups). Significant differences were assessed by Bonferroni Post Hoc tests. P-values below 0.05 were considered significantly different.

Results
Determination of antibody titers from egg C. jejuni-specific IgY were detectable in egg yolk of hens immunized with a C. jejuni lysate, resulting in titers of up to 1:16000 as determined by ELISA (results not shown). In contrast, C. jejuni-specific IgY in egg white and C. jejuni-specific IgA and IgM in egg white and yolk were not detectable after immunization.

In vitro anti-Campylobacter properties of egg yolk IgY
Pre-treatment with purified IgY from yolks of C. jejuni KC40-immunized hens significantly (p < 0.05) promoted bacterial binding of the homologous C. jejuni strain to chicken intestinal mucus by approximately log 10 2 compared to IgY from HBSS/sham-immunized (control) animals ( Figure 1). Increased bacterial binding to mucus after pre-treatment with KC40-derived IgY compared to control IgY was also observed for C. jejuni strain 10kf-1.16 (p < 0.05), whereas the binding capacity of the other four C. jejuni strains (7P6.12, 10C-6.1, 10kf-4.12 and 10VTDD-8) remained unaltered (p > 0.05). All six C. jejuni strains tested belong to a different clonal complex, as proved by MLST analysis (Table 1), indicating that C. jejuni strain KC40 and C. jejuni strain 10kf-1.16 are "true" heterologs. C. jejuni KC40-specific IgY did not affect C. jejuni motility of any of the tested strains (data not shown).

In vivo anti-Campylobacter properties of egg yolk IgY
In in vivo trial 1, the overall cecal C. jejuni numbers ( Figure 2) of inoculated (seeder) broilers receiving hyper-immune egg yolks from C. jejuni whole-cell lysate-immunized layers were reduced by > 5 log 10 cfu (p < 0.01) compared to control seeders, receiving yolks from HBSS/sham-immunized hens (3.3 ± 1.2 vs. 8.4 ± 0.6 log 10 cfu/g cecal contents). Moreover, transmission to non-seeder (contact) chicks was completely prevented, in contrast to the control group where seeder and nonseeder birds (8.4 ± 0.6 vs. 7.3 ± 1.1 log 10 cfu/g cecal  contents) were colonized to a similar (p > 0.05) degree. Together, this lead to a dramatic reduction (p < 0.001) in the overall cecal C. jejuni count by log 10 6.7 for birds receiving hyperimmune yolk (0.9 ± 1.6 log 10 cfu/g cecal contents) compared to birds receiving control yolks (7.6 ± 1.1 log 10 cfu/g cecal contents). In in vivo trial 2, the protective effect of IgY induced against the outer-membrane proteins of C. jejuni was investigated. In two of the three replicate experiments (Figure 3A and B) a significant (p ≤ 0.05) reduction in cecal C. jejuni counts was observed in seeder animals (by respectively 3.3 and 1.6 log 10 cfu/g cecal contents) fed hyperimmune egg yolk compared to control birds. In the third replicate ( Figure 3C), counts were reduced (p > 0.05) by 1.1 log 10 cfu. In replicate 1, transmission to contact birds was completely prevented, while in replicate 2 and 3 cecal counts were significantly reduced by respectively 3.6 and 6.1 (p < 0.01) log 10 cfu/g cecal contents) when birds were fed hyperimmune egg yolk. Together, this resulted in an overall reduced cecal C. jejuni count in the replicate experiments of respectively 5.6 (p = 0.001), 2.9 (p = 0.001) and 4.5 (p < 0.05) log 10 cfu/g. Averaged over three replicates, cecal C. jejuni numbers of inoculated broilers receiving hyper-immune egg yolks from layers immunized with the hydrophobic protein fraction of C. jejuni were reduced by two log 10 cfu (p = 0.002) compared to control animals (6.0 ± 1.8 vs. 8.0 ± 0.4 log 10 cfu/g cecal contents), while transmission to contact birds was greatly (p < 0.001) reduced (1.5 ± 2.5 vs. 7.0 ± 1.1 log 10 cfu/g cecal contents). Together, this lead to a dramatic reduction (p < 0.001) in the overall cecal C. jejuni count by log 10 4.4 for broilers receiving hyperimmune yolk (3.0 ± 3.2 log 10 cfu/g cecal contents) compared to birds receiving control yolks (7.4 ± 1.0 log 10 cfu/g cecal contents).

Identification of immunodominant C. jejuni KC40 antigens
Western blots from 2-dimensional separated C. jejuni proteins immunostained with IgY from egg yolk of C. jejuni whole-cell lysate-immunized laying hens revealed several immunodominant antigens ( Figure 4A), while blots stained with IgY from HBSS/sham-immunized hens showed only marginal reactivity (data not shown). Spots that were preferentially present on the Western blot immunostained with purified IgY from yolks of laying hens immunized with C. jejuni were linked to the 2-dimensionally separated C. jejuni proteins on the gel ( Figure 4B), digested in peptides with trypsin and subjected to mass spectrometric (MS) analysis. Proteins with a Mascot score of < 56 (p < 0.01) or present in only one of duplicate experiments were excluded from the study. MS analysis identified several of the immunodominant antigens of C. jejuni KC40, represented in Table 2.

Discussion
Immunizing laying hens and subsequently collecting their eggs is a cheap and straightforward method to obtain high amounts of specific antibodies [6,11]. In our study, C. jejuni-specific IgY was dramatically induced in egg yolk of hens immunized with a C. jejuni whole-cell lysate, resulting in titers of up to 1:16 000 as determined by ELISA. In contrast, specific IgY in egg white and specific IgA and IgM in egg white and yolk were not significantly induced after immunization. This indicates that only transfer of IgY to egg yolk is biologically relevant in the overall transfer of immunoglobulins into eggs after Figure 2 Reduced cecal Campylobacter jejuni colonization with yolks from hens immunized with a whole-cell lysate of C. jejuni. Individual and mean (▬) cecal C. jejuni counts of two-week-old co-housed broiler seeders (•) and sentinels (○) receiving standard feed supplemented with 5% (wt/wt) egg yolk from either HBSS/ sham-immunized (left) or C. jejuni immunized (right) layers, from day six until the end of the experiment, are shown. Values are represented as log 10 cfu/g cecal content. At 10 days of age, seeder animals were inoculated with approximately 8 × 10 3 cfu of C. jejuni KC40.
immunization of hens, which is in line with Dohms et al. [20].
In the first in vivo trial, using the whole-cell lysate of C. jejuni, broiler chicks were provided with feed supplemented with (hyperimmune) egg yolks from day seven onward. At 10 days of age three out of 11 birds were orally inoculated with approximately 8 × 10 3 cfu of C. jejuni. Colonization of these seeder animals is supposed to result in transmission of C. jejuni to the remainder of the birds. Three days after inoculation, cecal C. jejuni numbers of seeder birds receiving hyperimmune egg yolk were significantly reduced compared to control seeders without IgY (by over 5 logs). Moreover, none of the contact birds were colonized with C. jejuni, while chicks receiving control eggs carried high (ca. log 10 7 cfu/g) bacterial numbers in their ceca. It is not clear at what site (the cecum or more proximal in the GI tract) the IgY fraction was active, but clearly it captured ingested C. jejuni bacteria, disabling them to colonize and transmit to other birds. In addition, specific IgY increased bacterial binding of C. jejuni to chicken intestinal mucus, suggesting that it promotes bacterial uptake in the mucus layer for enhanced mucosal clearance. Because several of the identified immunodominant antigens are believed to be associated with the bacterial outer-membrane, a second (analogous) in vivo trial was performed using egg yolks containing IgY against only the hydrophobic protein fraction of C. jejuni. Although the effects observed in this second in vivo trial were less pronounced compared to the results obtained in the first in vivo trial using IgY against the whole-cell lysate of C. jejuni (probably due to the higher inoculation dose of the seeders (3 × 10 4 cfu vs. 8 × 10 3 cfu)), hyperimmune yolks reduced cecal bacterial counts of seeders by 2 log 10 cfu and dramatically reduced C. jejuni transmission to contact birds, indicating that the hydrophobic protein fraction is important for immunization but that also non-hydrophobic proteins may be involved in this immune response. Previous passive immunization studies with Salmonella enterica by other researchers [21] indicated no significant effect on cecal colonization of broilers receiving feed supplemented with freeze-dried egg yolk powder containing anti-Salmonella enterica IgY. Cecal Salmonella counts in birds receiving feed supplemented with 5% (wt/wt) yolk powder on day 3 post-inoculation were, however, (non-significantly) reduced by three logs compared to control birds. These authors suggested that the antibodies were denatured and degraded along the GI tract, thereby reaching the ceca at insufficient concentrations, but it cannot be ruled out that a significant reduction could be achieved using more animals per group. In our study, egg yolks were administered as such to exclude degradation during freeze-drying. This might indicate that egg yolks form a protective matrix allowing IgY to survive the digestive enzymes and the low pH along the GI tract [11]. Other factors that may be responsible for the differential outcome between our study and that of Chalghoumi et al. [21] include (1) age of the animals at challenge, (2) experimental set-up, since we used a seeder experiment while all birds were challenged with Salmonella, (3) immunization protocol and (4) bacterium-specific differences between Salmonella enterica and C. jejuni. We identified the most likely C. jejuni proteins responsible for the induction of specific IgY. Because a C. jejuni whole-cell lysate was used, specific IgY could be induced against both membrane-bound as well as cytoplasmic proteins. Although our obtained results are fairly consistent with those reported in other studies examining antigenic proteins of C. jejuni, there was a striking absence in our study of Omp18, Cme, Cja and PEB proteins, identified in another study [22]. These differences might be in part explained by C. jejuni strain differences. In addition, it is not clear whether repeated immunization of chickens in their breast muscle could lead to a different immune response compared to that observed during natural gut colonization with C. jejuni.
A first group of proteins that reacted with the purified egg yolk IgY are flagellar proteins: flagellar hook protein FlgE, major flagellin FlaA and minor flagellin FlaB. FlgE and FlaA are known to be immunogenic and are needed for full motility of C. jejuni [22,23]. In addition, FlgE mediates flagellar assembly and protein secretion in C. jejuni [22], while FlaA mutation may result in reduced colonization in chicks [24]. These results are consistent with Shoaf-Sweeney et al. [22] who reported the presence of maternal antibodies reacting with these flagellar proteins in the serum of laying hens. Despite these flagellar proteins, which were shown to be highly immunogenic in this study, specific IgY did not affect C. jejuni motility. The major outer-membrane protein (MOMP) of C. jejuni, encoded by the porA gene, was identified in A total protein extract of C. jejuni KC40 was separated by 2-dimensional gelelectrophoresis and separated proteins were detected by SYPRO® RUBY Protein staining. Immunoreactive antigens are indicated with a number and were excised from the gel and sequenced using mass spectrometry. Identified proteins are indicated by the spot numbers given in Table 2. several gel pieces, pointing to its abundance in the C. jejuni proteome [25]. In addition to being immunogenic [26], MOMP is involved in adhesion and transmembrane ion transport in C. jejuni [25]. However, MOMP is extremely genetically diverse and several conformational epitopes have been implicated in the induction of protective immunity [26], thereby possibly hindering its use in vaccine applications. Methyl-accepting chemotaxis (MCP) proteins are transmembrane receptors for chemotactic stimuli [23]. Attached to these transmembrane receptors is the C. jejuni chemotaxis protein V (CheV), a coupling protein involved in transducing chemotactic signals in C. jejuni [27], which was also identified in this study. Another immunodominant protein was the branched-chain amino acid ATP-binding cassette transport protein LivJ, a periplasmic binding protein probably involved in transporting amino acids into the bacterial cell [28] and crucial for chick colonization [22]. An additional role for LivJ in the interaction of C. jejuni with the chick cecum was suggested by Ribardo and Hendrixson [28]. ATP synthase is abundantly present and highly conserved among bacteria [29]. The protein is associated with the membrane where it performs its role in energy metabolism. In this study, the alpha subunit of ATP synthase F1 (AtpA) was found to be immunogenic in C. jejuni, as already reported for C. concisus [29]. Several of the identified antigens in this study were proteins mainly located in the bacterial cytoplasm where they perform their respective functions. Some of these antigens may nevertheless be interesting candidates for vaccine development. GroEL is an extremely conserved [30] 60-kDa heat shock protein that plays a crucial role in the C. jejuni stress response and has previously been shown to be an immunodominant antigen of C. jejuni [31]. Although mainly located in the bacterial cytoplasm, in Salmonella Typhymurium (and other bacteria) GroEL is suggested to be expressed on the cell surface as well since it has been shown to mediate Salmonella adhesion [32]. The highly conserved translation elongation factor thermo unstable (EF-Tu) was found to be abundantly present in the C. jejuni whole-cell lysate, which is in line with observations in other bacteria [33]. Despite its cytoplasmic role during protein synthesis, EF-Tu was shown to be translocated to the surface in several bacteria, where it mediates adhesion and invasion of host cells [33]. Immunization of mice with Burkholderia pseudomallei EF-Tu resulted in a potent immune response that was partially protective against melioidosis. Finally, a putative secreted carboxyl-terminal protease of C. jejuni, CtpA, was identified. This protein has been shown to be highly prevalent and conserved in Burkholderia mallei, another Gram-negative bacterium [34]. The exact function, as well as prevalence and conservation of CtpA in C. jejuni is less understood, but the immunodominance of this protein indicates CtpA might be an interesting candidate for subunit vaccin development.
This study reports some promising observations, but further research is needed before a commercial product can be promoted. First of all, it needs to be established whether this protective effect is maintained over periods exceeding three days post-inoculation. Second, the immunization protocol and the yolk dose in the feed need to be optimized. Further research will also have to elucidate whether this strategy is capable of providing crossprotection against heterologous C. jejuni strains. The increased mucosal adherence observed for at least one of the heterologous C. jejuni strains tested in this study, suggests that (partial) cross-protection may be attained, but in vivo trials are necessary to proof this hypothesis. Eventually, this may lead to the development of a passive immunization strategy that can successfully control C. jejuni in chickens. If the relevant protective antigens of C. jejuni could be identified and studies regarding conservation and prevalence among C. jejuni strains are conducted, subunit vaccines able to target a wide range of C. jejuni strains could be developed and used for immunization of hens in order to prepare yolks for passive immunization applications.
To conclude, we here demonstrate for the first time that feeding broilers IgY-rich yolks from hens immunized with C. jejuni dramatically reduces both C. jejuni numbers in the ceca after challenge with the homologous strain and transmission to non-inoculated contact birds, providing a solid base for further research regarding passive immunization to control C. jejuni colonization in broiler flocks.