Return to the PHS Forum:
Impact of Dietary and Environmental Factors on Microbial Communities of the Avian GI Tract
Juha Apajalahti1 and Michael Bedford2
1Danisco Cultor Innovation, 02460 Kantvik, Finland; 2Finnfeeds, PO Box 777, Marlborough, Wiltshire, UK. SN8 1XN.
© Copyright Apajalahti and Bedford (2000)
The gastrointestinal bacterial community is as metabolically powerful as any organ in the animal body. The gastrointestinal walls are the interface between this massive bacterial organ and the healthy, sterile tissues such as blood. If the immunological or other animal defence mechanisms fail, bacteria readily penetrate tissues rich in utilisable nutrients and cause infections. Chemical compounds and biomolecules on either side of the interface interact continuously and maintain the delicate balance, allowing nutrient absorption, but keeping bacteria out.
Bacteria in the gastrointestinal tract derive most of their energy for reproduction and growth from dietary compounds, which are either resistant to attack by digestive fluids or absorbed so slowly that bacteria can successfully compete for them. Since bacterial species have different substrate preferences and growth requirements, the chemical composition of the digesta largely determines the composition of the microbial community in the gastrointestinal tract. The bacterial community at a given time point, therefore reflects the fitness of the bacterial species to the chemical and physical environment and capabilities to compete against other bacterial residents and the defence system of the host.
By dietary means it is possible to affect the competitiveness of harmful and non harmful bacteria by changing gut dynamics. Specific species can be selected for by certain feed ingredients, which escape digestion by the host, but are readily available for the metabolic machinery of the target microbes. Products belonging to this group include prebiotics, dietary fibre, oligosaccharides etc. Probiotics, live fed bacteria, are another group of products, which are targeted to improve the health of the gastrointestinal tract, but these can only be effective if the requirements for their growth are present. The bacterial nutrient package will not succeed without the presence of the targeted, beneficial bacteria and likewise the live microbe product will not succeed if the environment into which it is introduced is unfavourable.
Not only the diet, but also the environment affects the microbial status of the chicken GI tract. Dirty litter and other animal management parameters affect microbial composition of the chicken GI tract both directly by providing a continuous source of bacteria and indirectly by weakening the physical condition and defence of the birds.
The ability to accurately monitor changes in the microflora depend upon the specific methods used. It is important to note that methods available for monitoring the total microbial community of the GI tract are few. All bacteria big enough can be seen under fluorescence microscopy, but only a few of them can be cultured under laboratory conditions. Indeed, most of the bacteria growing in such a complex community are dependent upon growth factors provided by other community residents or the secretions from the host tissues. Due to these complex requirements it is typical that less than 10% of the bacteria living in the gut can be cultured under laboratory conditions. As a consequence, most of the work and conclusions compiled until now reflect only minor members of the gastrointestinal microflora. It is worth noting that all the presently known pathogens are among this less fastidious, easily cultured minority group. Many diseases, the cause of which is not known today, may have a causative agent among the yet unknown, unculturable majority of the gastrointestinal tract microbes.
Methods using enrichment under selective conditions and/or direct selective plating are very efficient in discovering minority populations, such as some known pathogen species. Selective plating techniques may detect bacterial populations at much lower relative abundance than any DNA based method known today. Therefore, these techniques should not be underestimated when used by professionals, who are capable of restricting their interpretations correctly to the very observations made. Unfortunately, todays literature contains numerous examples of studies reflecting a lack of understanding of microbial physiology and ecology.
Our laboratory is using mainly DNA based approaches for analysing microbial communities in the GI tract. This means that the DNA of the total digestal bacterial community is recovered from the sample by using a combination of physical, chemical and enzymatic methods (Apajalahti, Sarkilahti et al., 1998). We have designed this process so that it should not discriminate for any bacterial types; that is, the method depicts faithfully the total rather than a partial bacterial community. Specificity, when desired, comes from the discriminating methods used for DNA analysis. The specificity ranges from the total bacterial community to bacterial strain level, depending on the technique used. Presently we are using techniques based on guanine + cytosine content of the bacterial DNA, hybridisation (DNA arrays), polymerase chain reaction methods with specific primers and sequencing of the 16S ribosomal DNA (order of increasing specificity). It is worth noting that no single technique can show the picture of the total bacterial community and be species specific at the same time. If both overall view and specificity are required, more than one method should be used for each sample.
In one of our trials, broiler chickens seemed to suffer from mild necrotic enteritis caused by Eimeria maxima challenge. Bacteria in the ileum of these 21 day-old broiler chicken were cultured on DRCM (differential reinforced clostridium medium) agar using strictly anaerobic techniques. Black colonies growing on the medium were randomly picked and analysed further by sequencing the 16S rDNA of the bacteria. Sequences obtained were compared to those deposited in the public RDP (Ribosomal Database Project) ribosomal data base and good matches were found to all the sequences compared (Sab scores higher than 0.9).
The medium used in this study is still considered in some laboratories as a medium, which selects for clostridia. In this study we found that only 14% of the colonies growing on the medium were clostridia. The major populations found were strains of Escherichia coli and Enterococcus spp (Figure 1). However, new selective media are being continuously developed to give more accurate identification (Dromigny, Bourrion et al., 1997).
Each medium used for enumeration of pathogens or other defined bacterial species should be tested for its true selectivity as described above. In the study described, true representatives of C. perfringens were found only in broiler chickens challenged with E. maxima, This supported our earlier findings and some reports in the literature, suggesting that coccidia are a predisposing factor of necrotic enteritis (Al-Sheikhly and Al-Saieg, 1980) (Broussard, Hofacre et al., 1986) (Williams, Carlyle et al., 1999).
FIGURE 1. Relative abundance of ileal bacteria growing on differential RCM agar (inoculum from E. maxima challenged broiler chickens)
There are many reports on the bacterial numbers in the GI tract. All the studies where culture methods for fastidious anaerobic bacteria have been used are consistent in that there are 1011 to 1012 of bacteria in the lumen of an adult broiler chicken (Fuller, 1992). Maintaining such a population under anaerobic conditions requires a significant amount of substrates; 10 to 20 % of feed carbohydrates and protein is probably not an overestimate. The bulk of bacteria are distal to ileum, which means that compounds supporting their growth have to escape host digestion.
The GI tract of the broiler chicken harbours numerous bacterial species. We have an ongoing, global survey of bacteria in the chicken GI tract. In the survey we are using both a % guanine + cytosine profiling and 16S rDNA sequencing. In combination these methods have made it possible to create a database describing all the major bacteria present in the chicken GI tract. Figure 2 shows the mean of all the G+C profiles of caecal bacterial communities from all around the world analysed in our laboratory (n = ~500). This image of the bacterial profile shows that bacteria with %G+C close to 47 are at the highest relative abundance globally, followed by those with %G+C around 65 (Figure 2).
DNA profiles alone do not identify bacteria to species or strain level. Therefore, we amplified a variable region of the 16S rDNA by PCR using degenerate, conserved primers working on all known organisms. After cloning into E. coli, the representative DNA fragments were sequenced and analysed for relatedness. This global phylogenetic analysis revealed the presence of ~200 bacterial groups (strains) with different sequences. Definition of species and genus are far from clear, but as judged by the phylogenetic distances of some well studied species and genera, it seems that we have about 35 genus level bacterial clusters. This work is still ongoing, but has already now shown that approved bacterial names should only be used as nametags for different bacteria, not for indicating true relatedness. As an example, Clostridium perfringens is genetically unrelated to all other clostridia found in the chicken GI tract. Furthermore, most of the bacteria in the GI tract of the broiler chicken seem to represent species different from any of those deposited in the public sequence databases.
FIGURE 2. Mean profile of the caecal microbial community
In the global survey we found significant farm to farm differences in caecal microbial communities. Figure 3 shows mean bacterial profiles of two Finnish broiler farms, which were using the same commercial feed and located less than 100 km apart.
Profiles in the figure are means of three broiler chicken from each farm. Even though this profiling technique is not specific enough to identify bacterial species involved, it is obvious that bacterial composition in the caeca of the birds from these two farms were significantly different (Figure 3; Table 1).
FIGURE 3. Mean profile of the caecal microbial community in two Finnish broiler farms (bars indicate SE of the replicates)
TABLE 1. Statistical differences in chicken GI tract microflora between two Finnish farms
% G+C range
|35 - 39||40 - 44||50 - 54||60 - 64||65 - 69||70 - 74|
The reason for the differences observed is unclear. Chicks were from the same hatchery and the commercial feed used was the same. Therefore, the differences are likely to be caused by the growth environment, general management and/or hygienic conditions on the farms.
In order to see the effect of grain base on the microbial community profile, we analysed 144 caecal samples of birds being fed either wheat, corn or rye based diet. The % G+C profiling technique followed by statistical analysis was used. Figure 4 shows the mean profiles of the bacterial communities on each grain.
Each grain had its characteristic effect on the microbial community structure. The major feature for corn was that it increased the relative proportion of bacteria with %G+C between 25 and 30 by more than 70% (p<0.0001). Rye in the diet had its major effect on bacteria with the %G+C between 35 and 40. Rye increased the abundance of these bacteria significantly as compared to wheat and corn fed birds (p<0.0001). Wheat increased the proportion of bacteria with %G+C between both 55 and 59 (p<0.01) and between 69 and 69 (p<0.0001). Thus it seems that each of the grains favours some bacterial groups in the caecum. This analysis does not reveal the identity of the bacteria, but one possible scenario is that corn favours low G+C clostridia and campylobacteria, rye stimulates the growth of lactobacilli and enterococci and wheat favours propioni- and bifidobacteria.
FIGURE 4. Mean profile of the caecal microbial community in broiler chickens fed wheat, corn and rye based diets
Coccidial stress has been shown to sensitise the broiler chicken to necrotic enteritis, which suggests that the parasite is changing the rules of bacterial competition in the GI tract. We have run a number of challenge trials, in which E. maxima was given orally to 14 day-old broiler chickens fed wheat based diet. At the age of 20, 22 and 28 days we analysed short chain fatty acids and biogenic amines in the ileum and caecum of three replicate birds.
FIGURE 5. Residual concentration of SCFA in the GI tract of Eimeria challenged and non-challenged broiler chicken (average of three replicate birds)
As shown in the Figure 5 the level of SCFA in the ileum of E. maxima challenged chicken was two times higher than in the corresponding non-challenged broiler chicken 6 days after the challenge. Two days later (day 8), the acid concentration in the ileum of challenged birds reached 100 mM, whereas in healthy birds the level remained constantly at 30 mM level. Similar increases in SCFA concentration was observed in the caecum. Again, challenged birds had SCFA levels almost twice that of the non-challenged ones at day 8 after the administration of the oocysts. Six days later SCFA levels in the GI tract of the challenged birds had decreased to the level of the non-challenged birds (Figure 5).
These results indicate that at maturation E. maxima is significantly changing microbial fermentation in the GI tract of the broiler chicken. Presumably the damage to the intestine following Eimeria challenge is incremental with time until the immune system deals with the pathogen. Initial changes in gut environment chemistry would therefore be confined to the site of infection, ie the ileum. As the Eimeria challenge reaches peak, sloughed cells and undigested feed would reach the caecum which may explain why the SCFA levels in the caeca of the challenged and unchallenged birds appear to diverge at least 2 days after those of the ileum.
FIGURE 6. Residual concentration of biogenic amines in the GI tract of Eimeria challenged and non-challenged broiler chicken (average of three replicate birds)
Like SCFA, biogenic amine levels were also affected by the coccidial challenge. In the ileum the concentration of the amines remained at or below 2 mM independent of the challenge. However, in the caecum, E. maxima challenge caused a 400% increase in the level of total biogenic amines (Figure 6). This suggests that unlike in the ileum, Eimeria caused a peak of putrefaction in the caecum 8 days after introduction, approximately the same point in time as maximum SCFA levels. This could be as a result of impaired amino acid absorption in the ileum and/or protein released from the intestinal tissue damaged by the parasite. As in the case of SCFA, biogenic amine levels also became normal after the acute phase of the cocci invasion had passed (Figure 6).
As shown above, the metabolic profiles in the intestine changed after a single Eimeria challenge. In the same trial we followed changes in the microbial community structure at different time points. Figure 7A shows %G+C profiles of the caecal bacteria from unchallenged birds, and the Figure 7B shows the corresponding time points in the challenged chicken. In the control chicken no significant changes were observed in the microbial community profile between days 20 and 28. However, in the Eimeria challenged broiler chicken a significant temporary shift in the caecal microbial community structure was observed at day 22 from hatching (8 days after Eimeria challenge). It was characteristic of this shift that the relative abundance of bacteria with %G+C 30 - 40 and 60 - 70 increased (Figure 7). This is consistent with the increase of numbers of lactobacilli and bifidobacteria and the timing coincides with the metabolite concentration maxima.
It is worth noting that under practical farm conditions challenge is often continuous due to recycling of the infectious oocysts and other pathogenic agents. Therefore, it is likely that both the chemical and microbial changes observed in this study 8 days after the challenge are present, but less dramatic under most commercial conditions where the challenge is expected to be less acute.
Clearly there are dramatic shifts in the whole microfloral community following either dietary or disease challenge. Such changes can now be followed using the techniques described. Since much of the community has yet to be identified it is clear that we are at the beginning of a lengthy learning process. The incentive to understand such interactions is particularly great in light of the decision in the European community to remove many growth promoters which apparently have controlled much of the shifts in populations described here. The removal of such products has resulted in a significant increase in enteric disorders which will be discussed in future work.
FIGURE 7. Mean profile of the caecal microbial community in unchallenged (A) and E. maxima challenged (B) broiler chickens at 20, 22, and 28 days of hatching (6, 8 and 14 days after Eimeria challenge, respectively).
Al-Sheikhly F, Al-Saieg A (1980) Role of Coccidia in the occurrence of necrotic enteritis of chickens. Avian Dis. 24:324-333
Apajalahti J, Sarkilahti LK, Maki BRE, Heikkinen P, Nurminen P, Holben WE (1998) Effective recovery of bacterial DNA and percent-guanine-plus-cytosine-based analysis of community structure in the gastrointestinal tract of broiler chickens. Appl.Env.Microbiol. 64:4084-4088
Broussard CT, Hofacre CL, Page RK, Fletcher OJ (1986) Necrotic enteritis in cage-reared commercial layer pullets. Avian Dis. 30:617-619
Dromigny E, Bourrion F, Rugraf Y, Bolton FJ, Leden N (1997) New media for detection and counting of Clostridia in foods. Lett.Appl.Microbiol. 24:19-22
Fuller R (1992) Probiotics, the scientific basis. The University Press, Cambridge
Williams RB, Carlyle WWH, Bond DR, Brown IAG (1999) The efficacy and economic benefits of Paracox (R), a live attenuated anticoccidial vaccine, in commercial trials with standard broiler chickens in the United Kingdom. International Journal for Parasitology 29:341-355