The effect of Bacillus cereus var . toyoi and avilamycin on the faecal microflora of turkeys

The effect of Bacillus cereus var. toyoi spores and avilamycin as bacterial and antibiotic growth promoters, respectively, on faecal microflora of turkeys was determined. The experiment was carried out on 108 turkeys randomly allocated to 3 groups: control and avilamycinor Bacillus spore-treated. After 12 weeks of feed supplementation and a subsequent 4-week period following withdrawal, selected groups of faecal microflora were determined, and the results were compared with those in the control group. Neither avilamycin nor B. cereus var. toyoi spores affected Bifidobacterium or Lactobacillus counts, whereas both supplements significantly increased the counts of E. coli by 0.84 and 0.69 log cycle (P ≤ 0.05 for both). B. cereus var. toyoi caused a significant increase in counts of spores of both aerobic and anaerobic proteolytic bacteria by 0.98 (P ≤ 0.001) and 0.83 (P ≤ 0.01) log cycle, respectively. At the end of the period following withdrawal, in the avilamycin group, counts of Bifidobacterium and Lactobacillus significantly decreased by 1.14 (P ≤ 0.01) and 1.97 (P ≤ 0.001) log cycle, respectively, as did the counts of anaerobic proteolytic bacteria spores, which decreased by 1.18 (P ≤ 0.001). In the B. cereus var. toyoi group, however, increased counts of Lactobacillus, by 0.91 (P≤0.01), E. coli, by 0.71 (P ≤ 0.05), anaerobic proteolytic bacteria spores, by 1.37 (P ≤ 0.001), and of anaerobic saccharolytic bacteria spores, by 0.72 (P ≤ 0.05) log cycle were found. The tested growth promoters did not affect the groups of beneficial gut microflora, whereas they stimulated the growth of opportunistic bacteria. The performance parameters of turkeys, better in groups administered avilamycin and B. cereus var. toyoi, were comparable with the control at the end of period after withdrawal.


INTRODUCTION
The number of antibiotics approved for use in animal feeding is being successively reduced, and now, in most European Union countries, only two of them, flavophospholipol and avilamycin, are allowed for poultry as growth promoters (Garshorn, 1998;Hughes and Heritage, 2002).The results of our preliminary studies suggest it may be possible to replace antibiotics with carefully selected probiotics for protection against zoonotic pathogens and enhancement of intestinal physiological functions (Bielecka et al., 2002a,b).One of the probiotics offered for farm animals is the spores of Bacillus cereus var.toyoi (CNCM I 1012 / NCIB-40112), a strain originally isolated from soil.The preparation of B. cereus var.toyoi spores is intended for use with sows and growing pigs, laying hens and broiler chickens, rabbit does and growing rabbits, calves and beef cattle at application rates of ~1 g/kg (i.e.~1 × 10 9 cfu/kg) of complete feedstuff.According to the declaration of the distributor, the addition of B. cereus var.toyoi has no adverse effect on the gastrointestinal microflora of any target species.
The best documented trends resulting from the addition of B. cereus var.toyoi to the diets were improved growth performance of piglets fed B. cereus var.toyoi to two months of age and a significant improvement in the survival rate and the weaning weight of piglets when the reproductive sows received that preparation (Report of SCAN, 2001).Other effects observed in some trials with animals fed spores of B. cereus var.toyoi included reduced severity of diarrhoea, decreased morbidity in piglets challenged with E. coli O149K and K99, improvement in average daily gain and feed conversion ratio (FCR).Some results also indicate that B. cereus var.toyoi has an effect on the morphology of the pig intestine, increasing the villus length in the small intestine and decreasing the number of goblet cells with 2,6-sialylated mucins in the large intestine (Baum et al., 2002).
The reason for undertaking the current study was the contradictory opinions about the usefulness of B. cereus var.toyoi as a probiotic in animal feeding expressed in publications, overinterpretation of some scientific data in commercial leaflets, as well as insufficient scientific confirmation of the usefulness of B. cereus var.toyoi in turkey nutrition.The aim of the study was to determine the effect of B. cereus var.toyoi and avilamycin as, respectively, bacterial and antibiotic growth promoters on selected performance indices as well as on the faecal microflora of turkeys.

MATERIAL AND METHODS
The experiments were carried out on 108 males of BUT-9 turkeys randomly assigned to 3 groups (3 pens in each).The birds were housed under the environ-mental and lighting conditions recommended by Faruga and Jankowski (1996) and fed ad libitum with mash diets (Table 1).The turkeys in the control group were receiving the basal diet, devoid of growth promoters.Avilamycin was administered from the 1st to the 8th week in the amount of 8 mg/kg and from the 9th to the 12th week in the amount of 6 mg/kg of feed, whereas B. cereus var.toyoi spores were used for twelve weeks at a dose of 150 mg/kg.Both diets were lacking these additives from week 13 to 16 of the experiment.The considered indices of turkey performance were live body weight (LBW) determined on weeks 4, 8, 12, and 16, FCR, and mortality.On weeks 12 and 16, i.e. at the end of administration of avilamycin and B. cereus var.toyoi, and at the end of the experiment, excrement was collected from 12 and 10 turkeys, respectively, from each group and its microflora analysed.All bacterial determinations were done immediately after sampling.Excrement samples of 1-2 g were weighted and mixed with 1% peptone water containing 0.5% (w/v) meat peptone (Peptobak, BTL, Łódź, Poland) and 0.5% (w/v) pancreatic casein hydrolysate (bio-Trypcase, bioMérieux, pH 7.0), as a diluter.After dispersion, serial decimal dilutions were made, avoiding aeration.
Lactobacilli were counted on MRS agar medium (pH 6.4 ± 0.1) using a double-layer technique, incubation at 37 o C/72 h and microscopic confirmation of cell morphology.
The most probable number (MPN) of E. coli was determined using brilliant green bile broth, after incubation at 37 o C/48 h, subsequent transfer of positive samples into tryptone water, followed by overnight incubation at 44 o C and indole test (Collins et al., 1995).
Spores were determined after heating of adequate dilutions at 80 o C/20 min and cooling.Anaerobic saccharolytic bacteria spores were counted as MPN using tubes containing nutrient agar with 1% glucose, 1% yeast extract and 1% peptone (fortified SPC) (Harrigan and Mc Cance, 1976) with pyrogallol stoppers (incubation at 37 o C up to 7 days); the number of anaerobic proteolytic bacteria spores was determined in tubes using broth gelatine (incubation at 22-25 o C up to 5 days) and gelatine stoppers; spores of aerobic proteolytic bacteria, on plates with fortified SPC agar supplemented with milk (incubation at 37 o C 24-48 h).
The results of the performance studies were subjected to statistical analysis using analysis of variance and STAT-1 software.The microbiological results were expressed as log colony forming units cfu/g excrement or MPN/g.Arithmetic means in groups of turkeys as well as significance of differences between groups were calculated with Student's t-test for less numerous groups.

RESULTS
The LBW of 4-and 8-week-old turkeys was significantly higher (P<0.01) in the group administered avilamycin than in the others (Table 2).Twelve-week-old turkeys from that group were still significantly heavier (P<0.05) in comparison with the control.At the end of the experiment, the LBW of the 16-week-old turkeys was not significantly differentiated, showing only a tendency to be higher in the groups administered avilamycin and B. cereus var.toyoi (with means 11.08 kg and 10.94 kg, respectively) than in the control (10.66 kg).The FCR value was the lowest in the group of turkeys fed with the antibiotic growth promoter (2.54 kg/kg LBW), and was lower by 0.13 and 0.10 kg than in the control and in the group receiving B. cereus var.toyoi, respectively.Two and one mortalities per 36 individuals in avilamycin-and Bacillus groups, respectively, occurred.
After 12 weeks of the experiment, the Bifidobacterium and Lactobacillus counts were determined as markers of gut microflora beneficial to host health, the E. coli count as opportunistic or potentially pathogenic bacteria, and the spore count as an indicator of aerobic and anaerobic spore-forming bacteria of different biochemical properties, proteolytic or saccharolytic.The next determination was performed 4 weeks after withdrawal of avilamycin and B. cereus var.toyoi from the diet, i.e. on week 16 of the experiment.
At the end of supplementation of the diet, in the control group, the populations of Lactobacillus and Bifidobacterium had the highest counts with means of 8.76 ± 0.32 log cfu/g and 8.72 ± 0.43 log cfu/g, respectively (Figure 1).The E. coli count (mean 5.67 ± 0.79 log MPN/g) was lower by ~3 log cycles than of Lactobacillus and Bifidobacterium.The lowest were the counts of spores: 2.21 ± 0.37 log cfu/g for aerobic proteolytic bacteria spores, and 2.23 ± 0.91 log cfu/g for anaerobic spores.Among the spores, anaerobic saccharolytic bacteria spore counts (mean 1.09 ± 0.40 log MPN/g) were the lowest.
In the groups of turkeys fed the diets with avilamycin or B. cereus var.toyoi, Bifidobacterium counts equalled 8.59 ± 0.90 and 8.76 ± 0.53 log cfu/g, Lactobacillus counts, 9.18 ± 0.61 and 8.74 ± 0.44 log cfu/g, respectively.Neither were significantly different in comparison with the control.In contrast, in both avilamycin-and B. cereus var.toyoi-fed groups, the E. coli counts, 6.50 ± 0.87 and 6.36 ± 0.53 log MPN/g, were significantly higher, by 0.84 and 0.69 log cycle, than those in the control (both P ≤ 0.05), respectively.Proteolytic bacteria spore counts in the avilamycin group were 2.22 + 0.58 log cfu/g for aerobes and 2.81 ± 0.46 log cfu/g for anaerobes.They remained on a level comparable with their controls.In contrast, their counts in the B. cereus var.toyoi group, at a level of 3.19 ± 0.38  for aerobes and 3.05 ± 0.33 log cfu/g for anaerobes, were significantly higher, by 0.98 (P ≤ 0.001) and 0.83 log cycle (P ≤ 0.01), than in the control.The counts of anaerobic saccharolytic bacteria in groups administered avilamycin or B. cereus var.toyoi were 1.54 ± 0.83 and 0.91 ± 0.52 log MPN/g, respectively, and were not significantly different from the control.
On week 16, in the group from which avilamycin was withdrawn, the determined counts were as follows: Bifidobacterium 7.45±0.48log cfu/g, Lactobacillus 7.21±0.50log cfu/g, E. coli 6.13±0.81log MPN/g, aerobic proteolytic bacteria spores 2.41±0.57log cfu/g, anaerobic proteolytic bacteria spores, 1.63±0.39log cfu/g, and anaerobic saccharolytic bacteria spores 1.61±0.50log MPN/g (Figure 2).The majority of bacterial populations were not affected by withdrawal of avilamycin, except for a significant decrease of Bifidobacterium populations by 1.00 log cycle (P≤0.05)in comparison with the control, and a tendency for the Lactobacillus count to decrease (by 0.75 log cycle; P≤0.1).In comparison with the data from week 12, the populations of three groups of bacteria were significantly reduced: Bifidobacterium by 1.14 (P≤0.01),Lactobacillus by 1.97 (P≤0.001), and anaerobic proteolytic bacteria spores by 1.18 (P≤0.001)log cycle (Table 3).

DISCUSSION
A meaningful positive influence of the antibiotic growth promoter on the body weight gain was observed in the first 8 weeks of feeding.In the subsequent periods this effect was gradually reduced.The effect of B. cereus var.toyoi spores did not reach statistical significance, although a positive tendency towards enhancement was observed, and was rather stable during administration.The greatest influence of the antibiotic growth promoter on BWG of turkeys was also found by Plavnik and Wax (1998) during the first eight weeks of life, whereas Garshorn (1998) did not observe any significant differences between turkeys fed diets without a growth promoter as compared with an antibiotic (flavophospholipol) or B. cereus var.toyoi.
Neither avilamycin nor B. cereus var.toyoi affected the counts of Bifidobacterium and Lactobacillus, intestinal bacteria beneficial to host health, nor the counts of spore-forming saccharolytic bacteria during supplementation.On the contrary, they both significantly increased the counts of E. coli.Avilamycin had no influence on the count of proteolytic spore-forming bacteria, either aerobic or anaerobic, whereas B. cereus var.toyoi significantly increased their counts.All these changes ranged from 0.7 to 1.0 log cycle.
A significant increase in the E. coli count by 0.69 log cycle in the B. cereus var.toyoi group contradicted the leaflet information that E. coli cells are systematically eliminated from the intestines.In several studies on the efficacy of B. cereus var.toyoi, no changes of microflora were observed, although on occasion a decrease in the numbers of E. coli was reported, accompanied by a small increase in lactobacilli counts (Report of SCAN, 2001).Kyriakis et al. (1999) used spores of B. cereus var.toyoi for effective control of post-weaning diarrhoea syndrome of piglets caused by an enterotoxigenic strain of Escherichia coli.
The significantly increased count of aerobic proteolytic bacteria spores might include the introduced spores of Bacillus cereus var.toyoi.Perhaps the germinated spores through their vital activity modified the intestinal environment in a direction favourable for anaerobic proteolytic bacteria, the number of which was also increased in the group of turkeys fed the diet with B. cereus var.toyoi.If the spores of Bacillus cereus var.toyoi are able to germinate and temporarily colonise the anaerobic intestinal environment, another explanation is possible entailing the anaerobic metabolism, as suggested by Casula and Cutting (2002) for the probiotic strain of Bacillus cereus in murine intestines.
Our results partially resemble those of Gedek et al. (1993) who showed no effect of B. cereus FH1457 administered to piglets in the amount of 10 7 , 10 8 , or 10 9 cfu/kg of feed during two periods of a 42-day feeding trial, 21 days each, on the counts of Lactobacillus/Bifidobacterium, Eubacteria, and Bacteroidaceae in the duodenum, jejunum, ileum, caecum, and colon.In different periods of the trial and in different parts of the gastrointestinal tract, the counts of E. coli and Enterococcus increased or decreased.The B. cereus count increased only at the highest dosage used in that experiment.
After withdrawal of avilamycin from the diet, some unfavourable changes occurred.A significant decrease in Bifidobacterium and Lactobacillus populations was observed, especially with reference to their counts at the end of the supplementation period, but also in comparison with the control at week 16.Such changes did not happen after withdrawal of B. cereus var.toyoi-the Bifidobacterium and Lactobacillus counts were unchanged from weeks 12 to 16 of the experiment, moreover, on week 16, the Lactobacillus count was higher than in the control group.
During the experiment the spore counts generally did not exceed 10 4 /g with means from 10 1 to 10 3 /g.It is obviously difficult to imagine that a dormant life form, especially in low amounts, may exhibit any probiotic effect, which points to vegetative cells as the active agents, although scientific opinions are divergent.On the basis of studies on non-pathogenic strains of B. subtilis, B. cereus and B. clausii used as probiotics, Spinosa et al. (2000) suggested that the spores of Bacillus are the forms responsible for any claimed probiotic effect, as they did not find detectable amounts of vegetative cells in intestinal samples.The opposite point of view was expressed by Jadamus et al. (2001).The authors observed repeated germination and sporulation of B. cereus var.toyoi during intestinal passage and concluded that the spores germinated rapidly in broiler chickens and piglets, which was a necessary prerequisite for their possible probiotic effects.Moreover, germination and in vivo sporulation of vegetative cells indicated that the probiotic strain was metabolically active in the intestine of both animal species.That opinion was also confirmed by Hoa et al. (2001) and Casula and Cutting (2002) for probiotic strains of B. subtilis in mice.

CONCLUSIONS
The levels of microflora in the control and avilamycin supplemented group suggest that administration of antibiotic growth promoter did not improve the profile of faecal microflora, moreover, it decreased the counts of beneficial bacteria, of Bifidobacterium (significantly) and Lactobacillus (a tendency) in the period after their withdrawal.Performance was better during feeding supplemented diets, but it was not different at the end of the study.The spores of B. cereus var.toyoi did not alter the level of beneficial intestinal bacteria in comparison with the control, except for a significantly higher count of Lactobacillus after being withdrawn, whereas they increased the counts of E. coli and proteolytic spores, both during administration and after withdrawal.Generally, the microbiological results of B. cereus var.toyoi supplementation seemed to be better than avilamycin, especially at the end of the period following their withdrawal, whereas performance parameters, better in groups during administration of avilamycin and B. cereus var.toyoi, were comparable with the control after these supplements were withdrawn.

Figure 1 .
Figure 1.Faecal microflora of turkeys at the 12th week of the experiment, i.e. at the end of supplementation of diet with avilamycin and B. cereus var.toyoi ○ -counts in single samples ■ -means + SD groups of turkeys: C -control, A -administered with avilamycin, B -administered with B. cereus var.toyoi counts different from the control at significance level: *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001

TABLE 2
, A -fed diets with avilamycin, B -fed diets with B. cereus var.toyoi a,b,A,B means in rows with no common superscripts were significantly differed at: a,b P≤0.05; A,B P≤0.01

TABLE 3
Changes in microflora from the 12th to 16th week of the experiment, i.e. within 4 weeks after deprivation of avilamycin and B. cereus var.toyoi from diets