Consequences of concurrent Ascaridia galli and Escherichia coli infections in chickens

Three experiments were carried out to examine the consequences of concurrent infections with Ascaridia galli and Escherichia coli in chickens raised for table egg production. Characteristic pathological lesions including airsacculitis, peritonitis and/or polyserositis were seen in all groups infected with E. coli. Furthermore, a trend for increased mortality rates was observed in groups infected with both organisms which, however, could not be confirmed statistically. The mean worm burden was significantly lower in combined infection groups compared to groups infected only with A. galli. It was also shown that combined infections of E. coli and A. galli had an added significant negative impact on weight gain.


Introduction
Ascaridia galli and Escherichia coli are both common causes of infections in confined as well as in free-range poultry productions systems (Dho-Moulin and Fairbrother, 1999;Permin et al. 1999). Pathogenic E. coli may cause airsacculitis, salpingitis, peritonitis, polyserositis, septicemia and other extra-intestinal diseases in chickens, turkeys and other avian species. However, E. coli also constitutes part of the intestinal microflora of healthy birds and most of the diseases associated with E. coli are considered secondary to environmental and host predisposing factors (Dho-Moulin and Fairbrother 1999). Clinical isolates of avian E. coli commonly belong to certain serogroups, i.e. O1, O2 and O78, and to a restricted number of clones (White et al. 1993). Experimental infections have shown that the air-exchange regions of the lungs and the airsacs are important sites of entry of E. coli into the bloodstream of birds during the initial stages of infection and that resistance to phagocytosis may be an important mechanism in the development of the disease (Gross 1990). It has also been demonstrated that F1 fimbriae are expressed in the respiratory tract, whereas P fimbriae are expressed in the internal organs of infected chickens (Vidotto et al. 1990). Unambiguous virulence factors associated with E. coli infections in avian species, remain to be identified. Diagnosis of E. coli infections is based on the clinical picture, lesions and isolation of E. coli (Dho-Moulin and Fairbrother 1999). A. galli may cause anorexia, weight loss, haemorrhages in the intestinal mucosa, obstruction of the intestinal lumen, altered hormone level and eventually death (Ackert 1931, Ikeme 1971, Roepstorff et al. 1999) in a wide range of avian species. The life cycle of A. galli is direct with a prepatent time of minimum 28 days under temperate climatic conditions . After ingestion of the infective egg, the egg hatches in the small intestine where the larva embeds in the mucosal layer of the duodenum for a varying period of 3-56 days depending on age and immunity of the bird (Herd and McNaught 1975). After maturation of the worm, it migrates to the intestinal lumen where it lives from intestinal contents and occasionally from host blood. The mature worms copulate and might start producing eggs after 28 days. Diagnosis of A. galli is based on faecal isolation of parasite eggs or direct identification of adult worms in the intestine (Permin and Hansen 1998). Few pathogen interaction studies have been carried out in poultry. Okulewicz and Zlotorzycka (1985) showed that A. galli exerted an inhibiting effect on the natural bacterial micro flora of the intestine of hens. The opposite situation, where the bacterial flora of the intestine inhibited the establishment of A. galli was demonstrated by Stefanski and Przyjakowski (1967). Chadfield et al. (2001) showed an interaction between the intestinal flora and A. galli, where the bacterium Salmonella enteriditis was incorporated into the eggs of A. galli. This finding, however, is in contrast to the finding of Baron et al. (1960), where the eggs of A. galli were found to be sterile. To the knowledge of the authors nobody has examined the interaction between E. coli and A. galli infections in chickens. Consequently this study has been designed to investigate the possible effect of two common concurrent infections; A. galli super imposed with E. coli and vice versa, on establishment of the respective infections, pathological lesions, mortality and impact on selected production parameters.

Experimental animals
Female Lohman Brown chickens were used for all experiments. The chickens were purchased as one-day-old chicks and kept for one week in a confined parasite free environment until the experiment. The chickens were given a commercial feed containing 20% protein and water ad libitum. The chickens were kept in one flock until infection whereafter they were placed in separate houses.

Infection material
A clinical nalidixic_acid_resistant O78 E. coliisolate originating from broilers suffering from respiratory distress (difficulties in breathing) was grown overnight in an enrichment media (LB-broth) to reach the desired infection dose (Lee and Arp 1998). The number of bacteria pr. ml (cfu) was determined by spectophotometry and plate counts. A. galli eggs were isolated from mature female worms obtained from layers and embryonated in 0.1N sulphuric acid according to the method described by Permin et al. (1997b).

Experimental design
In total, three experiments were carried out. The first experiment was conducted to determine the route and dose of the E. coli having clinical impact without killing all chickens, a prerequisite for the following two trials. The infection dose and infection route of A. galli was set to be 500 embryonated eggs given as a single oral infection in all three experiments according to previous studies by Permin et al. (1997a). The last two experiments were designed as 2 x 2 cohort studies, i.e., groups with or without A. galli and E. coli infections including a control group (Thrusfield 1995).

Determination of dose and infection route of E. coli
Fifty-two one-week-old chickens were purchased and were kept for seven days to acclimatise in the new environment. The chickens were divided into 9 groups of which 7 groups each consisted of 6 animals and 2 groups of each 5 animals. The animals were infected as given in Table 1. The primary infection with A. galli or E. coli took place on day 0 of the experiment. The secondary infection was carried out seven days after the first infection. All chickens were killed on day 14 and subjected to post mortem examinations (Permin and Hansen 1998).

Final experiments
Based on the results of the first experiment, two further experiments were set up, each with six groups. In total 299 four-week-old Lohman Brown female chickens were used for the experiments. The infection dose of E. coli was set to be 10 8 cfu in the two experiments, given as either an oral or a tracheal infection. In all experiments A. galli and E. coli were given either as single or combined infections. In the case of combined infections the chickens were inoculated with an interval of 7 days between the two infections. In the second experiment the animals were followed for 14 days after the second infection, whereas the animals in the third experiment were followed 4 weeks after the second infection. Group distributions and inoculation schemes are given in Tables 2 and 3.

Parameters measured
All chickens were weighed just as clinical signs were recorded. Furthermore, re-isolation and counting of the nalidixic-acid-resistant E. coli (Lee and Arp 1998) and A. galli larvae/adults (Permin and Hansen 1998) was carried out on all animals dying during the experiment and at the end of the experiment. Likewise, pathological lesions, if any, were recorded on all chickens.

Statistical analyses
All data were stored in the statistical programme GraphPad Prism (GraphPad Software Incorporated 2000). One-way analysis of variance (ANOVA), chi-square analysis (⑂ ⑂ 2 -analysis) or Students t-test were used to analyse the data.

Determination of dose and infection route of E. coli
Pathological lesions associated with an E. coli infection were seen in the group given E. coli by tracheal route with 10 8 cfu and in the groups infected with A. galli combined with a secondary E. coli infection given by tracheal route with a dose of 10 4 or 10 8 cfu. Pathological lesions were observed in four, three and two chickens respectively in these groups (Table 1). However, E. coli could only be re-isolated from three of these animals. One animal died in the group infected with A. galli and E. coli given by tracheal route with 10 4 cfu while two animals died in the group given only a tracheal oral dose of 10 8 E. coli. These three animals all tested positive for E. coli. A chi-square analysis showed no significant difference in mortality rates between the groups (p>0.05). A. galli larvae were recovered from all four groups infected with the parasite. An analysis of variance showed that the worm burdens were not significantly different between groups (p>0.05). The animals were weighed four times during the first experiment. The mean weight gains are given in Figure 1. One week after the first infection with either A. galli or E. coli differences were seen between the infected groups when compared to the control group (p<0.05), but there was no significant difference between the infected groups. At slaughter (fourth weighing) an analysis of variance showed that the group infected with A. galli and subsequently with E. coli given as a tracheal infection with 10 8 bacteria had a significantly lower (p<0.05) weight gain compared to all the other groups. But also the groups infected with E. coli given as an oral or tracheal infection with a dose of 10 8 bacteria and the group given A. galli and subsequently E. coli with 10 4 bacteria had a significantly lower weight gain (p<0.05) at time of slaughter compared to the control group. Based upon the overall results obtained in oneweek-old chickens, which are more susceptible to E. coli infections, 10 8 E. coli was chosen as the dose of infection in both oral and tracheal infections and four-week-old chickens were used instead (Dho-Moulin and Fairbrother 1999).

Final experiments
In the second experiment A. galli was given as the primary infection followed by E. coli. The results of the second experiment are outlined in Table 2. Pathological lesions consistent with E. coli infections were seen in the group infected with A. galli followed by E. coli given as a tracheal dose of 10 8 cfu. in addition to the two groups infected by oral or tracheal route with only E. coli. Pathological lesions were observed in four, one and four chickens in these groups, respectively. Pathological lesions were not seen in the group infected with only A. galli or in the groups infected with first A. galli and subsequently with E. coli given as an oral dose of 10 8 cfu.
Mortality was encountered in the groups infected with E. coli given as an oral or tracheal infection and in the group with combined A. galli and tracheal E. coli infection. Mortality due to cannibalism was seen in the control group. A chi-square analysis for differences in mortality showed no significant differences be-  tween the groups (p>0.05). Pure isolates of E. coli were obtained from liver and spleen from the group with combined A. galli and tracheal E. coli infection and in the group infected with only tracheal E. coli. At slaughter, A. galli larvae were isolated from the three groups initially infected with A. galli. A significantly lower worm burden was seen in the A. galli group compared to the combined groups (p<0.01). It was not possible to recover the nalidixic-acidresistant O78 stain used for inoculation of the birds at time of slaughter. The mean weight gains for all groups are given in Figure 2. After the first infection with A. galli until the second infection a slight weight depression was seen in all groups including the control group. An analysis of variance between all groups at time of the first and second infec-tion showed no significant difference between the groups (p>0.05). However, after the second infection (and for the remaining time of the experiment), with E. coli given as a tracheal or oral infection, a significantly lower weight gain was seen in these groups compared to all other groups (p<0.05).
In the third experiment E. coli was given as the primary infection followed by A .galli. The results of the third experiment are outlined in  Ag+Ec4T Ag+Ec8O Ag+Ec8T Ec4O Ec4T EcO8 Ec8T Control single infected E. coli group had a weight gain similar to the other groups (p>0.05).

Discussion
In total, three experiments were carried out to examine the effect of various combinations of A. galli and E. coli infections in growing chickens of layer type. Characteristic pathological lesions due to E. coli were seen in all the groups tracheally infected with E. coli as previously described by Dho- Moulin and Fairbrother (1999) and Nakamura et al. (1985), while lesions were absent in those inoculated orally. Pathological lesions were not observed in relation to the A. galli infection. This is presumably due to the rather low worm burdens observed in the chickens (Ikeme 1971, Permin et al. 1997a. The combined infections of E. coli and A. galli did not produce more pathological lesions, which is unexpected as simultaneous parasitic infections often create more severe pathological lesions (Christensen et al. 1987). However, a trend for increased mortality rates was seen in the groups infected with the two pathogens, but it was not confirmed statistically.
Significantly different worm burdens were isolated from the intestinal tract of the A. galli and E. coli infected groups compared to the A. galli infected groups. With A. galli given first followed by an oral or a tracheal E. coli infection, significantly higher worm burdens were observed in both groups. Johnson and Reid (1973) (Stefanski and Przyja kowski 1967) whereas Okulewicz and Zlotorzycka (1985) showed that A. galli inhibited the natural bacterial micro flora of the intestine of hens. The mechanisms for these phenomena are not known, but possibly related to the development of immunity. A recent paper by Pritchard and Brown (2001) has indicated that although cellular response mechanisms of bacteria and parasites are related to each their pathway (Th2 for parasites and Th1 for bacteria) there is a balance between the two pathways. Thus a parasite infection might favour the Th2 cell development and indirectly suppress the establishment of bacteria, or vice-versa. Furthermore, lower worm burdens were detected in the third experiment which ran for additional weeks. Similar observations were made by Tongson and McCraw (1967) where a nonspecific age related immunity developed in growing chickens around the age of 3 months. The mechanism could be a self-cure mechanism as recently described in chickens in rela- tion to A. galli (Permin and Ranvig 2001). Balic et al. (2000) discussed the expulsion of trichostrongyle nematodes after primary infections in rodents speculating that the mechanism behind is genetically related as also described by Behnke et al. (1992). The expulsion is mainly seen in rodents and not in larger animals (Balic et al. 2000).
The nalidixic-acid-resistant E. coli strain was only isolated from the chickens which died during the experiment and not from any of the animals at slaughter. Similar findings have been reported by Leitner and Heller (1992), who could not isolate an orally inoculated nalidixic-acidresistant O78 from the trachea or from the blood in stressed turkeys. However, in orally infected broiler chickens, stress resulted in bacteremia and mortality.
In the experiments significantly lower weight gains were seen in the groups given E. coli as a tracheal infection. Weight depression as a result of tracheal E. coli infections is in accordance with the findings of other researchers (Dho-Moulin and Fairbrother 1999). It was further shown that the combined infection with A. galli had a significant added negative impact on weight gain. Interestingly a primary infection with A. galli followed by an oral infection with E. coli also had a significantly negative impact on the weight gain. Young birds (4-8 weeks) may have a brief period of anorexia and depression after infection with E. coli followed by acute septicaemia with mortality. However, weight depression was also seen after an oral E. coli infection when A. galli eggs were given as the primary infection. This may be related to damage of the intestinal mucosa leading to loss of blood and, probably, establishment of a secondary infection such as E. coli (Herd and McNaught 1975). Likewise, infections with A. galli have been reported to cause reductions in the growth rate, weight loss and mortality in broilers (Ackert and Herrick 1928, Reid and Carmon 1958, Ikeme 1971, He et al. 1990). The severity of the intestinal lesions may depend on the number of worms established in the intestine (Ikeme 1971). However, in this study only moderate weight losses were seen due to the parasite and only in the very young birds (1-3 weeks) whereas the older birds apparently were able to compensate for the infection. This is in contrast to earlier findings (Ackert andHerrick 1928, Reid andCarmon 1958). Permin and coworkers (unpublished) have observed similar findings in growing chickens where the animals apparently compensated for the loss due to the parasites by an increased feed intake. The findings of this study indicate a negative relationship between concurrent infections of E. coli and A. galli. The mechanisms behind the observed relationship are not known, but might be related to immune mechanisms (Pritchard and Brown 2001). Leitner and Heller (1992) investigated the potential of pathogenic E. coli to penetrate the bloodstream via the intestinal mucosa in normal and stressed turkeys and chickens, but did not examine this in relation to stress caused by parasites. Their studies showed that, in orally infected turkeys, the pathogenic bacteria (a nalidixic-acid-resistant O78) remained in the intestine where it replaced 10% to 50% of the native coliform flora. But in orally infected broiler chickens, stress resulted in bacteremia and mortality. In our study significant weight depressions were seen in the orally infected chickens, which indicates that A. galli, when given as a primary infection, has an damaging effect on the intestinal mucosa (Herd and McNaught 1975) enabling E. coli to establish when it is given as an oral infection. However, an increased mortality was not seen. An additional effect of A. galli was seen in the group secondarily infected with A. galli. This could be related to an immunosuppressive effect of A. galli (Malviya et al. 1988, Sharma 1997, Roepstorff et al. 1999.

Conclusion
In conclusion combined infections of A. galli and E. coli can have a significant impact on confined and free-range chickens keeping in mind that both infections are common in such production systems. Further studies are needed in order to determine the underlying mechanisms of combined infections of A. galli and E. coli in growing and adult chickens.