Infection and excretion of Salmonella Enteritidis in two different chicken lines with concurrent Ascaridia galli infection

Abstract

Studies on the impact of interaction of Salmonella enterica serovar Enteritidis and the parasitic nematode Ascaridia galli with the avian host were undertaken with particular emphasis on infection and excretion of these pathogens in two different layer lines. A total of 148 salmonella-free 1-day-old chickens (73 Hellevad and 75 Lohmann Brown) were randomly divided into five groups for each line. Group 1 served as an uninoculated control group. Groups 2 and 3 were infected with A. galli and S. Enteritidis, respectively. Group 4 was first infected with S. Enteritidis and subsequently with A. galli, and vice versa for group 5. The number of chickens excreting S. Enteritidis was significantly higher (P < 0.001) in the groups infected with both S. Enteritidis and A. galli compared with those only infected with S. Enteritidis over time. Furthermore, excretion of S. Enteritidis over time was significantly higher (P < 0.001) in the group first infected with S. Enteritidis and subsequently with A. galli compared with the group infected in the reverse order. No significant differences were observed between the two lines concerning excretion of S. Enteritidis over time in any group (P = 0.61 (group 3), P = 0.73 (group 4), P = 0.31 (group 5)). A. galli established itself significantly better (P = 0.02) in the group first infected with A. galli and subsequently with S. Enteritidis compared with the group infected in the reverse order. Furthermore, the A. galli infection rate was significantly higher (P = 0.02) in Hellevad chickens compared with Lohmann Brown chickens at the end of the experiment.

Des études sur l'impact de l'interaction de Salmonella enterica serovar Enteritidis et du nématode parasite Ascaridia galli chez l'hôte aviaire ont été entreprises avec une emphase particulière concernant l'infection et l'excrétion de ces agents pathogènes chez deux lignées différentes de pondeuses. Cent quarante-huit poussins d'un jour indemnes de salmonelle (73 Hellevad et 75 Lohmann roux) ont été répartis au hasard en cinq groupes pour chaque lignée. Le groupe 1 non inoculé a constitué le groupe témoin. Les groupes 2 et 3 ont été respectivement infectés par A. galli et S. Enteritidis. Le groupe 4 a d'abord été infecté par S. Enteritidis puis par A. galli, et vice-versa pour le groupe 5. Le nombre de poussins excrétant S. Enteritidis a toujours été significativement supérieur (P < 0,001) dans le groupe infecté par S. Enteritidis et A. galli comparé à ceux infectés par S. Enteritidis. De plus, l'excrétion de S. Enteritidis a toujours été significativement supérieure (P < 0,001) dans le groupe infecté en premier par S. Enteritidis puis par A. galli comparé au groupe infecté dans l'ordre inverse. Aucune différence n'a été observée entre les deux lignées concernant l'excrétion de S. Enteritidis tout au long de l'essai quels que soient les groupes (P = 0,61; P = 0,73; P = 0.31). A. galli s'est implanté significativement mieux (P = 0,02) dans le groupe infecté en premier par A. galli puis par S. Enteritidis comparé au groupe infecté dans l'ordre inverse. De plus, le taux d'infection à A. galli a été significativement supérieur (P = 0,02) chez les poulets Hellevad comparés aux poulets roux Lohmann à la fin de l'expérimentation.

Es wurden Untersuchungen durchgeführt zur Bedeutung von Interaktionen von Salmonella enterica Serovar Enteritidis und der parasitären Nematode Ascaridia galli mit dem aviären Wirtstier, wobei der Schwerpunkt dieser Untersuchungen auf der Infektion und Ausscheidung dieser Pathogene in zwei verschiedenen Legehennenlinien lag. Insgesamt wurden 148 Salmonellen-freie Eintagsküken (73 Hellevad und 75 Lohmann Brown) zufällig auf 5 Gruppen pro Hühnerlinie verteilt. Gruppe 1 diente als nicht inokulierte Kontrollgruppe. Gruppe 2 und 3 wurden mit A. galli oder S. Enteritidis infiziert. Gruppe 4 wurde zuerst mit S. Enteritidis und dann mit A. galli infiziert und umgekehrt wurde bei Gruppe 5 verfahren. Insgesamt war die Zahl der S. Enteritidis ausscheidenden Küken in den Gruppen, die sowohl mit S. enteritidis als auch mit A. galli infiziert worden waren, im Vergleich zu den nur mit S. enteritidis infizierten signifikant höher (p < 0,001). Außerdem war in den Gruppen, die zuerst mit S. Enteritidis und anschließend mit A. galli infiziert worden waren, die S- Enteritidis-Ausscheidung über den Gesamtversuchszeitraum im Vergleich zu den Gruppen mit umgekehrter Reihenfolge signifikant höher (p < 0,001). Hinsichtlich der S. Enteritidis-Ausscheidung gab es insgesamt keine signifikanten Unterschiede zwischen den Gruppen (p = 0,62; p = 0,73; p = 0,31). A. galli siedelte sich in den Gruppen signifikant besser (p = 0,02) an, die zuerst mit A. galli und dann mit S. Enteritidis inokuliert worden waren, als in den in umgekehrter Reihenfolge infizierten Gruppen. Weiterhin war am Versuchsende die A. galli-Infektionsrate bei de Hellevad-Küken im Vergleich zu den Lohmann Brown-Tieren signifikant höher.

Se llevaron a cabo estudios del impacto en el huésped aviar de la interacción entre Salmonella enterica serovar Enteritidis y el parásito nematodo Ascaridia galli, con especial énfasis en la infección y excreción de estos patógenos en dos líneas distintas de ponedoras. Un total de 148 pollos libres de salmonella de 1 día de vida (73 Hellevad y 75 Lohmann Brown) se distribuyeron aleatoriamente en cinco grupos para cada línea. El grupo 1 sirvió como grupo control no inoculado. Los grupos 2 y 3 se infectaron con A. galli y S. Enteritidis, respectivamente. El grupo 4 se infectó en primer lugar con S. Enteritidis y posteriormente con A. galli, y viceversa en el grupo 5. El número de pollos que excretaron S. Enteritidis en el tiempo fue significativamente mayor (p < 0.001) en los grupos infectados con los dos S. Enteritidis y A. galli en comparación con aquellos que sólo se infectaron con S. Enteritidis. Además, la excreción en el tiempo de S. Enteritidis fue significativamente mayor (p < 0.001) en el grupo infectado primero con S. Enteritidis y posteriormente con A. galli en comparación al grupo infectado en el orden inverso. No se observaron diferencias significativas entre las dos líneas en relación a la excreción de S. Enteritidis en el tiempo en ningún grupo (p = 0.61, p = 0.73, p = 0.31). A.galli se estableció significativamente mejor (p = 0.02) en el grupo infectado primero con A.galli y después con S. Enteritidis en comparación con el grupo infectado en orden inverso. Además, el nivel de infección por A. galli fue significativamente mayor (p = 0.02) en los pollos Hellevad en comparación a los pollos Lohmann Brown al final de la prueba.

Introduction

Danish table egg production has traditionally been based on battery cage and deep litter production systems. However, today table eggs from free-range and organic systems account for approximately 20% of the marketed eggs in Denmark as a result of the increased interest for organic products and improved animal welfare (Anonymous, 2005a). Free-range chickens are regularly infected with endoparasites (Permin et al., 1999) and may also more easily acquire a range of bacterial and viral diseases because of their free access to outdoor areas and a resulting lack of biosecurity compared with conventional indoor systems where a high level of biosecurity can be obtained (Christensen et al., 1999; Permin et al., 2002). In addition, more effective cleaning and disinfection between stocks can be achieved in conventional indoor systems.

In relation to public health, Salmonella enterica serovar Enteritidis represents one of the most important bacterial infections in layers, and is one of the most frequently isolated serotypes of Salmonella from poultry. Introduction of the European Directive on food-borne zoonoses (Anonymous, 1992) and national legislation has led to considerable reduction in the number of Salmonella-infected flocks in Denmark (Anonymous, 2005b), which has been translated into a reduction in the number of cases of food poisoning due to S. Enteritidis in Denmark (Anonymous, 2005b). S. Enteritidis rarely results in disease in adult chickens. However, the asymptomatic colonization of the alimentary tract of poultry may result in human food poisoning cases associated with consumption of contaminated eggs.

Ascaridia galli infection is commonly observed in layers in conventional indoor systems, as well as in free-range and organic systems (Permin et al., 1999). Its short direct life-cycle and the resistance of its eggs favour infections under floor and free-range production systems (Permin et al., 1998a) where the chickens are not separated from their faeces. A cross-sectional prevalence study of gastrointestinal helminths in Denmark has shown that the flock prevalence of A. galli is 100% in free-range and organic systems, compared with 25% in confined indoor deep litter production systems (Permin et al., 1999). Furthermore, it has been shown that the eggs of A. galli might transfer Salmonella (Chadfield et al., 2001), thus representing an increased risk of the persistence of Salmonella in the environment.

Recent studies of interactions between A. galli and Pasteurella multocida or Escherichia coli infections in chickens have shown that dual infections favour the establishment of bacterial infections and result in more severe disease than is seen with the bacterial infection alone (Dahl et al., 2002; Permin et al., 2006). Although interactions between gastrointestinal helminths and other pathogens appear to be important, little research has so far been carried out within this field. Control of food-borne human salmonellosis inevitably includes control of the infectious agent in the animal host. Thus, understanding the mechanisms of Salmonella infection, including intestinal colonization, invasion, excretion and persistence in layers dually infected with A. gall, is essential in order to identify appropriate measures to reduce infection of flocks and public health risk.

The objective of the present study was to investigate the interaction of dual infections of A. galli and S. Enteritidis on the colonization and infection with these agents, including subsequent excretion and persistence of S. Enteritidis in two different layer lines. Layer lines used for free-range table egg production, including Lohmann Brown, have been selected for high performance in individual cages without considering their ability to adapt to free-range and organic production systems. In addition to Lohmann Brown chickens, this study included Hellevad chickens, which are assumed to be better adapted to organic production systems (Worm, 2003).

Materials and Methods

Experimental animals and housing facilities

Hellevad (a cross of New Hampshire and White Leghorn lines) and Lohmann Brown chickens were used in the study (Worm, 2003). Seventy-three-day-old Hellevad chickens were obtained from a small Danish hatchery, while 75 day-old Lohmann Brown chickens were obtained from a commercial hatchery. Both hatcheries and parent flocks tested negative under the Danish Salmonella Control Programme (Feld et al., 2000; Wegener et al., 2003). Before infection chickens were placed together in a cleaned and disinfected house. They were neither vaccinated nor beak trimmed. Before and throughout the experiment, the chickens had free access to water and a standard commercial antibiotic-free premixed starter feed containing a coccidiostat (monensin-natrium) at a concentration of 100 mg/kg. At 11 days of age the chickens were wing-banded and randomly allocated into five groups of 15 for each line, with the exception of two groups of Hellevad chickens that consisted of 14 chickens. Each group was placed in a clean, disinfected house with a floor area of 6 m² and access to an outside pen by day. Three days prior to the primary infection at the age of 19 days, the Salmonella status of all chickens was evaluated by examining pools of faecal swabs for each group by standard bacteriological culture procedures performed as described later under bacteriological examinations.

Experimental design

The experiment was designed as a 2×2 + 1 cohort study (Thrusfield, 1995) in which group 1 for each line was kept as uninfected controls. Groups 2 and 3 were infected with A. galli and S. Enteritidis, respectively. Group 4 was first infected with S. Enteritidis and 1 week later with A. galli, and vice versa for group 5. The experiment was terminated 70 days after the primary infection.

Infections

S. Enteritidis phage type 4, isolated from a naturally infected Danish commercial layer flock, was used in this study (Aabo et al., 2002). It was stored in Luria Bertani (LB) broth containing 15% glycerol at −80°C. The strain was cultured aerobically overnight at 37°C on blood agar base (Oxoid, Denmark) with 5% citrated bovine blood. Five colonies were subsequently inoculated into LB broth and incubated aerobically overnight at 37°C with shaking. The challenge inoculum was prepared from the overnight broth culture, which was serially diluted with physiological saline to a concentration of approximately 1×10⁶ colony-forming units of bacteria per 0.1 ml. Confirmation of inoculate doses was verified by viable counts of serial 10-fold dilutions on LB agar.

A. galli eggs were isolated from the uterus of mature worms obtained from Salmonella-free naturally infected commercial layers. Embryonated eggs were prepared by cultivation of the eggs in 0.1 N sulphuric acid as described previously (Permin et al., 1997b). Prior to infection, the infectivity of the eggs were assessed visually by motility of larvae.

The experimental design is outlined in Table 1. All chickens in group 1 (control group) were sham infected orally with 0.1 ml of 0.9% saline at the ages of 19 days and 26 days. Chickens in group 2 were infected orally with 1000±50 infective A. galli eggs at the age of 19 days. All chickens in group 3 were infected orally with 1×10⁶ colony-forming units of S. Enteritidis at the age of 19 days. Both groups 2 and 3 were sham infected orally with 0.1 ml of 0.9% saline at the age of 26 days. Group 4 was infected with 10⁶ colony-forming units of S. Enteritidis at 19 days old and 7 days later with 1000±50 infective A. galli eggs, and vice versa for group 5. The same dose of A. galli infective eggs was used in a comparable study with concurrent infections with Pasteurella multocida and A. galli (Dahl et al., 2002). The specific age of the chickens was chosen to ensure intestinal colonization of S. Enteritidis without significant morbidity or mortality (Gast & Beard, 1989; Gorham et al., 1991) and the two pathogens were given with a 1 week interval to ensure a possible host response to the primary infection before the secondary infection.

Table 1.  Experimental design

			Infection(s)
Group number	Group name	Group size^a	19 days old	26 days old
1	Control	15 + 15	None	None
2	A. galli	15 + 15	A.galli	None
3	S. Enteritidis	15 + 15	S. Enteritidis	None
4	S. Enteritidis + A. galli	14 + 15	S. Enteritidis	A. galli
5	A. galli+S. Enteritidis	14 + 15	A. galli	S. Enteritidis
^aPresented as Hellevad chickens + Lohmann Brown chickens.

Body weight

The body weight of individual chickens was recorded at 0, 7, 14, 21, 35, 49, 63 and 70 days after the primary infection.

Bacteriological examinations

To assess faecal excretion of S. Enteritidis, cloacal swabs were collected weekly from each chicken for bacteriology. The swabs from groups 1 and 2 (S. Enteritidis non-infected groups) were pooled in groups of five for each chicken line and each group but bacteriological examination was performed individually on chickens in groups 3, 4 and 5 (S. Enteritidis infected groups). The first time swabs were collected from group 5, these were pooled as for groups 1 and 2. All samples were checked for Salmonella using a method originally described by De Smedt et al. (1986) with minor modifications. Briefly, after overnight incubation in buffered peptone water (Merck, Germany) at 37°C, three drops of the pre-enrichment culture was inoculated onto Modified Semisolid Rappaport-Vassiliadis (Oxoid, Denmark) agar plates, and incubated at 42°C for 18 to 24 h. Bacteria putatively identified as Salmonella on Modified Semisolid Rappaport-Vassiliadis were subcultured on modified Brilliant Green Agar (Oxoid, Denmark) and incubated overnight at 37°C, followed by plating of typical colonies on blood agar incubated overnight at 37°C. Suspect Salmonella colonies were finally confirmed by slide agglutination using polyvalent anti-Salmonella serum (SSI, Copenhagen, Denmark). At the end of the experiment, selected isolates were serotyped according to Kauffmann (1972) for verification of S. Enteritidis.

Parasitological examinations

The excretion of A. galli eggs was monitored weekly from week 4 post infection by evaluating the faecal egg excretion using a modified McMaster method adapted to detect minimum egg counts of 20 eggs per gram of faeces (Permin & Hansen, 1998). Faecal samples from groups 1 and 3 (A. galli non-infected groups) were pooled in one sample for each chicken line and each group at all sampling dates. Samples from the A. galli-infected groups (groups 2, 4 and 5) were pooled in one sample for each group and each chicken line at the first sampling date (28 days post-inoculation (p.i.)). Thereafter, faecal samples were collected and examined as individual samples from the A. galli-infected groups (groups 2, 4 and 5). For this procedure, each chicken was housed separately in a cage for a short time and fresh droppings were taken from the cage floor.

At the end of the experiment all chickens were sacrificed and subjected to postmortem examination, during which all intestinal tracts were collected. The intestinal tracts were dissected longitudinally and the contents were washed in two sieves, the smallest with a mesh aperture of 38 µm. The sieve retentate was examined for the presence of mature and immature stages of A. galli using a stereomicroscope at 40x magnification. The numbers of immature and adult worms were counted and all adult worms were sexed by morphological parameters.

Statistical analysis

A repeated-measurements model (Diggle, 1988) was chosen to investigate the effect of the different groups on the body weight gain for the two chicken lines and to investigate the possibility of an interaction between the treatment groups and the chicken lines and an interaction between chicken lines and time and with baselines as covariates. To fit the model, the raw weight data were log-transformed. Data on S. Enteritidis excretion in the groups were analysed by logistic repeated measurement with individual weekly records as statistical units. The model included treatment groups and the chicken line as the main effect and the chicken line and time as interactions.

A Kruskal–Wallis test was performed to compare the worm burden (larvae + adult A. galli) between the three A. galli-infected groups, separately for Hellevad and Lohmann Brown chickens, respectively. The Kruskal–Wallis test was chosen due to the non-normal distribution of the raw data with no apparent way to transform to Gaussian distributed data. Logistic regression was carried out to investigate the probability for A. galli to establish and survive in the intestines of the chickens being related to the main effects of treatment groups and chicken lines, and the possibility that there is an interaction between the main effects.

Results

Weight gain

All chickens gained weight during the experimental period and no interactions between chicken line and treatment groups were observed (P=0.62). For both lines the control group (group 1) performed significantly (P<0.05) better than the groups infected with either A. galli (group 2) or S. Enteritidis (group 3) and the group infected first with S. Enteritidis and subsequently with A. galli (group 4). For both lines, also, no significant difference (P=0.06) was demonstrated between the control group (group 1) and the group infected first with A. galli and subsequently with S. Enteritidis (group 5) or between the groups dually infected with S. Enteritidis and A. galli in reverse order (P=0.1078) (data not shown).

Excretion of S. Enteritidis

Before inoculation, no Salmonella bacteria were detected in the pooled cloacal swabs from any of the groups. In addition, Salmonella was not isolated from any cloacal swabs from group 1 (control group) or group 2 (infected with A. galli only) throughout the experiment.

Faecal excretion rates of Salmonella for the groups inoculated with S. Enteritidis are presented in Table 2. The number of chickens excreting S. Enteritidis in each group varied considerably between sampling dates. The number of birds excreting was higher in groups 4 and 5 (infected with S. Enteritidis and A. galli), and the birds eliminated the infection more slowly than those in group 3 (infected with S. Enteritidis only) for both lines. The highest Salmonella shedding rate for Hellevad chickens in group 3 was 40%, and excretion stopped between 49 and 56 days after infection. In the same group, the highest shedding rate for Lohmann Brown chickens was less than 30% and excretion stopped between days 28 and 35 post infection, except for a single chicken. The highest percentage of chickens shedding Salmonella (100%) and the longest excretion time (i.e. to the end of the experiment) was observed for those infected with S. Enteritidis at the age of 19 days and subsequently infected with A. galli at the age of 26 days (group 4). The Lohmann Brown chickens given the dual infection in the reverse order (group 5) appeared to stop excreting Salmonella by 56 p.i. and only one Hellevad chicken was excreting Salmonella by 63 days p.i. In total, excretion of S. Enteritidis was demonstrated in 46.7% (7/15), 100% (14/14) and 100% (14/14) of Hellevad chickens from groups 3, 4 and 5, respectively, compared with 40.0% (6/15), 100% (15/15) and 100% (15/15), respectively, for Lohmann Brown chickens. No statistically significant difference was observed between Hellevad and Lohmann Brown chickens in any of the groups (P = 0.61 (group 3), P = 0.73 (group 4), P = 0.31 (group 5)).

Table 2.  Faecal excretion of S. Enteritidis following oral inoculation of Hellevad and Lohmann Brown chickens with S. Enteritidis

	Group 3^a, S. Enteritidis				Group 4^a, S. Enteritidis+A. galli				Group 5^a, A. galli + S. Enteritidis
	Hellevad		Lohmann Brown		Hellevad		Lohmann Brown		Hellevad		Lohmann Brown
Days p.i. with S. Enteritidis	Positive/15^b	%	Positive/15	%	Positive/14	%	Positive/15	%	Positive/14	%	Positive/15	%
7	1	6.7	0	0.0	5	35.7	3	20.0	7	50.0	10	66.7
14	1	6.7	1	6.7	5	35.7	9	60.0	13	92.9	14	93.3
21	0	0.0	1	6.7	14	100.0	15	100.0	13	92.9	10	66.7
28	6	40.0	4	26.7	14	100.0	15	100.0	10	71.4	8	53.3
35	1	6.7	0	0.0	13	92.9	15	100.0	4	28.6	5	33.3
42	2	13.3	0	0.0	11	78.6	10	66.7	6	42.9	4	26.7
49	1	6.7	0	0.0	9	64.3	10	66.7	5	35.8	1	6.7
56	0/13^c	0.0	1	6.7	12	85.7	10	66.7	2	14.3	0	0.0
63	0/13^c	0.0	0	0.0	5	35.7	4	26.7	1	7.1	0	0.0
70	0/13^c	0.0	0	0.0	6	42.9	13	86.7	N.D.^d	N.D.	N.D.	N.D.
^aChickens in groups 3 and 4 were infected with Salmonella at 19 days of age; group 5 were infected at 26 days of age. ^bNumber of chickens positive/number tested. ^cThirteen chickens tested because two escaped on day 50 after S. Enteritidis inoculation. ^dNo data. No significant difference was observed in the number of positive cloacal swabs between the two lines in any of the groups (P=0.61 (group 3), P=0.73 (group 4), P=0.31 (group 5)). A significant difference was observed in the number of positive cloacal swabs between group 3 and groups 4 and 5 for both lines over time (P<0.001). Significantly more positive cloacal swabs were demonstrated in group 4 compared with group 5 for both lines over time (P<0.001).

A statistically significant difference (P < 0.001) in the incidence of faecal excretion was observed between group 3 and groups 4 and 5 for both lines. Furthermore, the incidence of faecal excretion was statistically higher (P < 0.001) for the group infected first with S. Enteritidis and subsequently with A. galli (group 4) than the group infected in the reverse order (group 5) for both lines.

A. galli egg excretion

As there was no obvious way of transforming the raw eggs per gram data to produce acceptable normally distributed data, no statistical test was performed for comparison of groups or lines. Control chickens (group 1) and chickens infected only with S. Enteritidis (group 3) remained negative for A. galli eggs in the pooled faecal samples throughout the experiment. All pooled samples from the A. galli-infected groups (groups 2, 4 and 5) taken 28 days p.i. were negative. The mean eggs per gram and the percentage of positive samples in the different groups on the sampling dates are presented in Table 3. Positive faecal egg counts were detected the first time individual samples were taken (35 days p.i.) in group 2 for both lines and for Hellevad chickens in group 5. The Lohmann Brown chickens in group 5 had the first positive faecal egg counts 42 days p.i. In group 4, the Hellevad chickens were detected positive for the first time 49 days p.i., while no Lohmann Brown chickens were detected positive at any sampling dates. In total, 13.3% (12/90), 12.9% (9/70) and 29.3% (25/84) of the samples from the Hellevad chickens were found positive for A. galli eggs in groups 2, 4 and 5, respectively. A. galli eggs were shed on at least one sampling date in 40.0%, 35.7% and 71.4% of the Hellevad chickens in groups 2, 4 and 5. In total, 17.8% (16/90), 0% (0/90) and 18.9% (17/90) of the droppings from the Lohmann Brown chickens were found positive for A. galli eggs in groups 2, 4 and 5, respectively. A. galli eggs were shed on at least one sampling date in 46.7% and 60.0% of the chickens in groups 2 and 5 for Lohmann Brown chickens.

Table 3.  Parasitological parameters of Hellevad and Lohmann Brown chickens infected with 1000 ± 50 embryonated A. galli eggs

	Group 2^a, A. galli		Group 4^a, S. Enteritidis+A. galli		Group 5^a, A. galli + S. Enteritidis
Parameter	Hellevad	Lohmann Brown	Hellevad	Lohmann Brown	Hellevad	Lohmann Brown
Number of chickens	15	15	14	15	14	15
Mean (standard deviation) eggs/g faeces
35 days p.i.	1.3 (5.2)	8.0 (31.0)	0.0 (0.0)	0.0 (0.0)	17.1 (64.1)	0.0 (0.0)
42 days p.i.	21.3 (77.3)	4.0 (15.5)	0.0 (0.0)	0.0 (0.0)	38.6 (128.1)	1.3 (5.2)
49 days p.i.	17.3 (56.5)	9.3 (31.0)	4.3 (11.6)	0.0 (0.0)	15.7 (36.1)	1.3 (5.2)
56 days p.i.	13.3 (51.6)	18.7 (61.6)	17.1 (25.8)	0.0 (0.0)	48.6 (133.1)	18.7(32.5)
63 days p.i.	34.7 (118.2)	49.3 (123.7)	22.9 (66.0)	0.0 (0.0)	87.1 (123.9)	37.3 (61.8)
70 days p.i.	38.7 (144.3)	20.0 (52.9)	N.D.^b	N.D.	107.1 (114.0)	30.7 (61.3)
Percentage of positive faecal egg counts
35 days p.i.	6.7	6.7	0.0	0.0	7.1	0.0
42 days p.i.	13.3	6.7	0.0	0.0	14.3	6.7
49 days p.i.	20.0	13.3	14.3	0.0	21.4	6.7
56 days p.i.	6.7	20.0	35.7	0.0	28.6	33.3
63 days p.i.	20.0	40.0	14.3	0.0	50.0	33.3
70 days p.i.	13.3	20.0	N.D.	N.D.	57.1	33.3
^aThe chickens in groups 2 and 5 were A. galli-infected at 19 days of age; group 4 were infected at 26 days of age. ^b No data.

Worm counts

Occasionally, expelled worms were found in faecal droppings of the chickens. At slaughtering, adult female worms were harboured in 20.0% (3/15), 21.4% (3/14) and 57.1% (8/14) of the Hellevad chickens of groups 2, 4 and 5, respectively, and in 5/15 (33.3%), 2/15 (13.3%) and 5/15 (33.3%) of the Lohmann Brown chickens of groups 2, 4 and 5, respectively. The total numbers of A. galli (larvae + adult) recorded at the end of the experiment are presented in Table 4. There was great variability in the number of worms per chicken. The Kruskal–Wallis test showed no significant differences (P = 0.07) in the total number of A. galli (larvae + adults) between the three Hellevad groups infected with A. galli. Likewise, no significant difference (P = 0.40) was observed between the three Lohmann Browns groups infected with A. galli. However, logistic regression showed that the probability of A. galli to establish was significantly higher (P = 0.02) in the group that was first infected with A. galli and subsequently with S. Enteritidis (group 5) compared with the group infected in the reverse order (group 4), and that A. galli established and survived differently in Hellevad and Lohmann Brown chickens, independently of the groups, as the infection rate of A. galli to establish was significantly higher (P = 0.02) in Hellevad chickens compared with Lohmann Brown chickens. No interaction between chicken line and treatment group was demonstrated (P=0.54). In the A. galli-infected groups no other gastrointestinal parasites except A. galli were demonstrated. All chickens in the non-infected groups remained free of A. galli and other gastrointestinal parasites throughout the experimental period.

Table 4.  Mean worm burden of A. galli (larvae + adult) in infected chickens at slaughter

	Group 2, A. galli		Group 4, S. Enteritidis+A. galli		Group 5, A. galli + S. Enteritidis
	Hellevad	Lohmann Brown	Hellevad	Lohmann Brown	Hellevad	Lohmann Brown
Number of chickens	15	15	14	15	14	15
Number infected	8	5	6	4	12	7
Mean worm burden (standard deviation)	1.4 (2.5)	1.4 (2.1)	1.6 (2.5)	0.5 (0.9)	2.7 (2.7)	1.5 (2.1)
Median (interquartile range)	1 (0 to 2)	0 (0 to 4)	0 (0 to 3)	0 (0 to 1)	2 (0 to 2)	0 (0 to 1)
Minimum to maximum	0 to 10	0 to 5	0 to 7	0 to 3	0 to 10	0 to 7

Clinical observations, mortality and pathological changes

No signs of clinical disease were observed during the study. During the observation period no mortality was observed in any group, and gross postmortem lesions were not observed at necropsy. Two Hellevad chickens from group 3 escaped from the outside pen before the experiment was finished.

Discussion

Neither clinical signs nor mortality were observed in Hellevad and Lohmann Brown chickens infected with S. Enteritidis in the present study. This confirms previous studies (Gast & Beard, 1989; Gorham et al., 1991) showing that susceptibility to disease of chickens infected with paratyphoid Salmonella spp. decreases rapidly during the first week after hatching. In addition, no clinical signs and gross lesions were observed in the A. galli and control groups, or in the groups with dual infection of S. Enteritidis and A. galli. Low level infections with A. galli do not normally cause clinical signs (Permin et al., 1998a). A reduction in growth rate has previously been associated with A. galli (Ackert & Herrick, 1928; Reid & Carmon, 1958; Ikeme, 1971a,b; Permin et al., 1998b), but in this study no clear conclusions could be drawn on the effect of A. galli on weight gain.

The low level of Salmonella infection observed in the Hellevad and Lohmann Brown chickens infected only with S. Enteritidis was also seen in a preliminary pilot study with a similar groups of animals (data not shown). In the present study both the percentage of chickens shedding Salmonella and the duration of excretion were significantly influenced by concurrent A. galli infection. For both lines of chicken, the number of birds excreting S. Enteritidis was significantly higher in the group infected first with S. Enteritidis and subsequently with A. galli than in the group infected in the reverse order.

To the best of our knowledge this is the first experimental study to demonstrate that A. galli may play an important role in determining the outcome of a concurrent Salmonella infection. The consequences on the persistence of Salmonella in the environment through infected A.galli eggs (Chadfield et al., 2001) are obvious but the mechanisms leading to an asymptomatic Salmonella carrier state and the higher rate of Salmonella shedding in the dual-infected groups remains to be elucidated. Concurrent infections with A. galli and other bacteria have demonstrated that A. galli predisposed to infection and a subsequent carrier state of P. multocida and E. coli in chickens (Dahl et al., 2002; Permin et al., 2006). It is possible that this is related to a polarization of the immune response. The immune response of mammals is known to polarize into so-called type 1 (Th1) or type 2 (Th2) immune pathways depending on the type of pathogen encountered (Cox, 2001), and recently this has also been demonstrated for chickens (Degen et al., 2005). Thus, a helminth infection might suppress the Th1 response and indirectly favour the establishment of bacterial infection, and vice versa. However, it could also be speculated that lesions associated with the histotrophic phase of A. galli infection might facilitate intestinal colonization and persistence of subsequent bacterial infections. Infection with Eimeria spp. has been identified as a factor that enhances the establishment and persistence of concurrent infection with S. Typhimurium in the intestinal tract of chickens (Stephens et al., 1964; Stephens & Vestal, 1966; Arakawa et al., 1981; Baba et al., 1982), suggesting that S. Typhimurium persists in and penetrates the damaged mucosa of the intestine of the chickens infected with coccidia. For the same reason, the histotrophic phase of A.galli may play an important role in developing a salmonella carrier state in chickens.

The mean worm burden demonstrated in all groups was relatively low, with a mean establishment rate of less than 0.3% for all groups. This is somewhat lower than observed in comparable studies (Permin et al., 1997a,b, 1998b, 2006; Gauly et al., 2001, 2005; Permin & Ranvig, 2001; Schou et al., 2003). It is possible that expulsion of worms had some impact on the mean worm burden as the proportion of chickens that had at least one positive faecal egg count during the study period was significantly higher than the proportion harbouring female worms at slaughtering in all groups, with the exception of Lohmann Brown chickens in group 4 where 13.3% (2/15) were found to harbour mature female worms at slaughtering, although all faecal samples were negative for parasite eggs. However, it is possible that the McMaster method used on the faecal samples was not sufficiently sensitive to detect some positive samples. The experiment should be repeated using more birds to support the results regarding the low number of A. galli-positive birds and the low worm burden demonstrated.

The worm counts of A. galli (larvae + adults) were not significantly different between infected groups of Hellevad and Lohmann Brown chickens; however, the probability of establishing an A. galli infection was significantly higher for chickens first infected with A. galli and subsequently with S. Enteritidis compared with those infected in the reverse order. The observed difference might be due to lesions associated with early infection with S. Enteritidis. However, it is also possible that it is related to a polarization of the immune response directed by the sequence of infection. It is therefore possible that a primary bacterial infection might suppress the Th2 response and indirectly favour the establishment of a secondary helminth infection. In the studies involving P. multocida, Dahl et al. (2002) showed that this bacterium had a significant impact on establishment of A. galli. Thus chickens first infected with A. galli and subsequently with P. multocida had a lower percentage of A. galli-infected birds than those infected in the reverse order or those infected only with A. galli.

In addition to the impact of S. Enteritidis on the success of A. galli establishment it was shown that the probability for A. galli to establish infection was significantly higher in Hellevad chickens than Lohmann Brown chickens at the end of the experiment. This might indicate a difference in genetic resistance to A. galli, supporting earlier studies that found commercial chickens to be less susceptible to A. galli than more outbred chicken lines (Permin & Ranvig, 2001; Schou et al., 2003) such as the Hellevad line. Based upon faecal shedding of Salmonella, no differences between the two lines were observed. A considerable amount of work has shown variations in resistance/susceptibility to colonization of intestines and caeca in outbred (Guillot et al., 1995; Protais et al., 1996; Duchet-Suchaux et al., 1997) and inbred (Barrow et al., 2004; Sadeyen et al., 2004) lines of chickens infected with S. Enteritidis. Recently, heritability estimates for resistance to caecal colonization of S. Enteritidis have indicated genetic influences on carriage in the caeca (Beaumont et al., 1999).

In conclusion, this appears to be the first experimental study demonstrating an interaction between A. galli and S. Enteriditis in chickens, and suggests that, under field conditions, A. galli may increase the colonization rate and prolong the duration of faecal excretion of S. Enteriditis and increase the risk of persistence in the environment through infected A. galli eggs as previously reported (Chadfield et al., 2001). Furthermore, this study indicated that some degree of genetic resistance against A. galli may exist in Lohmann Brown chickens compared with Hellevad chickens, while no difference with regard to faecal Salmonella excretion was observed. Control of A. galli in chickens may represent an important key to reduce the potential for spread of Salmonella infection. Further research is needed to fully understand the mechanisms behind the interactions observed, and genetic resistance to A. galli may improve the possibilities of preventing S. Enteriditis infections in humans.

Acknowledgments

The authors gratefully acknowledge Michael Fink, Mette Pedersen, Signe Andersen, Johnny Jensen, Pernille Ginsbo, Tony Bönnelykke, Pia Mortensen, Margrethe Anne-Gurri Pearman, Frederik Andersen and Jørgen Olesen for their technical assistance. Financial support was given by the Institute Agronomic for Overseas, the Department of Animal Production, Italy and the Ministry of Agriculture (the FØJO II programme), Denmark.