Relationship Between Poultry Gut Microbiota and Diet
The animal gut microbiome describes the microorganisms of the gut as well as genomes and genes belonging to those microorganisms. Since the gut microbiome is involved in the regulation of multiple host metabolic pathways, a deep understanding of the relationships between microbiome and host provides new strategies to fight diseases and improve animal health.
By using shotgun metagenomic sequencing, it is possible to characterize the impact of the administration of different feed additives on the gut microbiota of commercial chickens in order to find nutritional strategies to be able to fight the foodborne pathogens colonization in the animals, thereby improving the safety of poultry meat. The impact of the dietary supplementation of Lactobacillus acidophilus and proteases on the chicken gut microbiota was reported.
Knowledge of the microorganisms having genes coding for enzymes associated to specific metabolic pathways allows improving (or decreasing) those pathways driving specific microbial populations by fit for purpose nutrition strategies and investigating how the metabolic pathways are correlated to one another in the different microorganisms.
The symbiotic equilibrium between a host and its microbiota is established immediately after the animal birth, then during the early stages of life its development is influenced by exposure to microorganisms present in the environment. The intestinal microbiota is an extremely dense and complex ecosystem, which plays a relevant role in the maintenance of the animal’s well-being through the production of biologically relevant metabolites. The microbiota can influence the usage of nutrients by the host, the optimal development of his intestinal mucosa (Scott et al., 2013, Doré and Blottière, 2015) and the prevalence of pathogens (Cogen et al., 2008, Yoon et al., 2015). For these reasons it is already common to introduce feed additives in animal diets in order to positively modulate the gut microbiota (Angelakis and Raoult, 2010, Huyghebaert et al., 2011, Kiarie et al., 2013, Pan and Yu, 2014, Doré and Blottière, 2015), especially after the ban of the antibiotic growth promoters within the European Union (Butaye et al., 2003, Huyghebaert et al., 2011). By using traditional culture methods the gut microbiota remains largely unexplored. Therefore, in order to predict its nutritional and ecological roles, researchers have turned to new culture-independent approaches and shot gun metagenomic sequencing is one of those. In this study the metagenomic sequencing has been applied to investigate the effects of the dietary supplementation of Lactobacillus acidophilus as well as proteases in diets with different percentages of protein on broiler caecum microbiota. Moreover, preliminary results of the administration of phytase alone or combined with inositol on the gut microbiota of broilers are presented.
Materials and methods
The experimental workflow followed in the different trails included (1) the collection of broiler caeca and crop contents; (2) the DNA extraction from the samples using the procedure described by Danzeisen et al. (2011), slightly modified; (3) the library preparation using Nextera XT Library Preparation Kit (Illumina); (4) shotgun metagenomic sequencing using HiScanSQ sequencer (Illumina) at 100 bp in paired-end mode. For the data analysis all the sequences, based on the results of the FastQC report, were trimmed (TrmGalore programm) in order to optimize the quality of the data. The MG-RAST pipeline (Meyer et al., 2008) was used to identify bacterial taxa performing a BLAST similarity search for the longest cluster representative against the M5rna database, integrating SILVA (Pruesse et al., 2007), Greengenes (De Santis et al., 2006) and RDP (Cole et al., 2003). Moreover, reads were assigned to functional groups using the Kyoto Encyclopedia of Genes and Genome (KEGG) database (www.genome.jp/kegg/) (Kanehisa, 2002). The samples heterogeneity was investigated using the Pielou’s definition (Pielou, 1966) and the Bray-Curtis distance, respectively. Principal Coordinate Analysis (PCoA) was computed with python module skbio 0.4.2 in order to assess which species mostly contributes to the samples Bray-Curtis distances, SIMilarity PERcentages analysis (SIMPER) (Clarke et al., 1993) was computed with R package vegan 2.3-5, setting 100000 permutations.
Effects of the dietary supplementation of Lactobacillus acidophilus on broiler caecum microbiota
In the day-old chickens more than 95% of crop and caecal bacterial population was represented by Firmicutes (80.12 in the crop and 77.03% in the caecum) and Proteobacteria (15.15 in the crop and 21.63% in the caecum). Firmicutes were largely represented both in the crop and caeca of birds at 14 and 35 days. The relative frequency of abundance of Firmicutes in the caecum of day-old chicks was significantly lower than that observed in all groups at 14 and 35 days (P=0.002309), whereas Proteobacteria were significantly higher (P=0.000167). On the contrary no statistically significant differences were reported in the microbiota of crops in the different experimental groups. Within the phylum of Firmicutes, in the caecum of all groups, except for 1 day old chicks, Clostridia was the most abundant class, followed by Bacilli with values ranging between 36.75% (day-old chicks) to 86.41% (High dose of LA 35 d) and 4.94% (High dose of LA 35 d) to 40.28% (day-old chicks). On the contrary within the phylum of Firmicutes, in the crop of all groups Bacilli was the most abundant class, followed by Clostridia, with values ranging between 52.86% (day-old chicks) to 96.20% (Control at 35 d) and 0.43% (Control at 35 d) to 26.45% (day-old chicks). In day-old chicks, as well as all treated groups and control at 14 and 35 days, Gammaproteobacteria was the most representative class of the Proteobacteria phylum in the crop and caecum. This class in the caecum was the only one significantly higher in day-old chicks (21.33%) in comparison to control groups and treated groups at 14 and 35 days of age (P=0.000942).
Moreover, in day-old chicks Lactobacillaceae was the most statistically representative family in the crop (46.40) as well as in the caecum (36.22%). A similar trend was observed in the crop of all experimental groups with percentages of abundance of 75.48, 93.25, 79.52, 87.72, 84.93 and 85.30% for the Control 14 d, Control 35 d, High dose of LA at 14 d, High dose of LA at 35 d, Low dose of LA at 14 d, Low dose of LA at 35 d respectively. On the contrary, in the caeca of all experimental groups at 14 and 35 days, the relative abundance of the same family was as low as 2.94, 9.51, 3.28, 1.80, 3.70 and 3.17% (P<0.001).
Lachanospiraceae was the most represented family identified within the Clostridia class in one day old chicks (12.98%), control group at 14 d (29.08%), control group at 35 d (22.39%), High dose of LA at 14 d (29.81%) and Low dose of LA at 14 d (29.92%). On the contrary, in High dose of LA at 35 d and Low dose of LA at 35 d Ruminococcaceae was the most represented family (40-58 and 40.13%, respectively). Table 1 and 2 report the first 5 bacterial species identified in different groups.
Faecalibacterium, Subdoligranulum and butyrate-producing bacterium SM4/1 identified in the caeca positively impact gut health through the production of butyrate and other short chain fatty acids. In terms of microbiota, the relative abundance of Lactobacillus acidophylus was high in the crop, but low in the caeca and was comparable with the one of the control group. This result might be explained taking into account the colonization preference of the administered strain for the crop and the small intestine, as shown in table 2. In fact, statistically significant differences (P<0.001) were reported in the relative abundance of Lactobacillus acidophilus in the crop, with values of 1.58, 1.49, 5.70, 17.28, 5.48, 2.84 and 4.74% for 1 day-old chicks, Control 14 d, Control 35 d, High dose of LA at 14 d, High dose of LA at 35 d, Low dose of LA at 14 d, Low dose of LA at 35 d, respectively. Beside the lack of colonization of LA in the broiler caeca, the results of this study seem to suggest that the metabolic activity of supplemented Lactobacillus acidophilus, and in particular the lactic acid production, positively affect the microbial species producing butyric acid by a cross feeding mechanism.
Effect of the administration of proteases in diets with different percentages of protein on the chicken gut microbiota
robiota colonizing the caecum contents of broilers fed with the different diets and quantified by the Pielou alpha-diversity was similar in the control group (i.e., diet A) in comparison to the groups fed with diets B, C and D at 14 days, whereas the diet C at the end of the rearing period resulted in a more diverse species composition in the caeca. The bacteria accounting for up to 70% of the differences observed between all pairwise comparisons (i.e., diet vs time) at 14 days belonged to the phyla Firmicutes, Proteobacteria and Actinobacteria, whereas at 42 days to the same phyla as well as Bacteroidetes. Within the phyla Firmicutes, the most abundant species, accounting for up to 70% of the differences observed at both 14 and 42 days, was Faecalibacterium prausnitzii followed by Subdoligranum variabile at 14 days and Pseudoflavonifactor capillosus at 42 days. The distribution of this species in the caeca of broilers fed with diets B, C and D was significantly lower in comparison to the control at 14 days and increased over time in the caeca of broilers fed with diets A and D. Within the phyla Bacteroidetes, the species Alistipes putredinis accounted for the majority of the differences observed between 14 and 42 days. In the second sampling time the other most abundant species causing those differences were Alistipes sp. HGB5 and Alistipes shahii. Within the phyla Proteobacteria, Escherichia coli was the most abundant species causing differences between pairwise comparisons at both sampling time, whereas within the phyla Actinobacteria the most abundant species at both sampling time, independently from the diet, were Bifidobacterium longum and Bifidobacterium adolescentis, followed by Bifidobacterium bifidum at 14 days and Slackia heliotrinireducens at 42 days. Escherichia coli was significantly lower in the caeca of broilers fed with diets A and D at 14 days and decreased over time, except for group D. The families Clostridiaceae, Ruminococcaceae, Lachnospiraceae, Eubacteriaceae and Erysipelotrichaceae were similarly distributed among diets and sampling time, whereas Enterobacteriaceae decreased over time. The ANOVA two way analysis, assessing the impact of diet, time and their interaction on the abundance of the phyla colonizing the caeca, showed that abundances of Proteobacteria and Actinobacteria were mainly affected by the diet as well as interaction between diet and time. On the contrary, the abundances of Firmicutes and Bacteroidetes were mainly affected by the age of the birds.
At 14 days in the caeca of broilers fed with diet B, the species showing in the t test a p value < 0.05 in comparison to the control and with an average abundance ≥ 0.025% belonged the phyla Firmicutes and were mainly represented by Eubacterium cylindroides and Lactobacillus fermentum. In the birds belonging to group C the most abundant species were Butyrate producing bacterium SM4/1, belonging to the phylum Firmicutes, and Provotella ruminicola, belonging to the phylum Bacteroidetes. Finally, in group D the most abundant species were Thermoanaerobater brockii, Bacillus pumilus and Pelotomaculum thermopropionicum within the phylum Firmicuets, Ralstonia solanacearum and Desulfuromonas acetoxidans belonging to the phylum Proteobacteria and Provotella ruminicola belonging to the Bacteroidetes. At the end of the rearing period (i.e., 42 days) Lactobacillus lactis, Bacillus licheniformis and Lactobacillus ruminis belonging to the phylum Firmicutes, showed a significantly higher abundance in the caeca of birds fed with diet B in comparison to the control. In the birds belonging to group C the most abundant species were Epulopiscium sp. among Firmicutes, Pseudomonas fluorescens, Janthinobacterium sp. and Herbaspirillum seropedicae among Proteobacteria. Finally, in group D the most abundant species was Lactococcus lactis and Solobacterium moorei belonging to Firmicutes and Acholeplasma laidlawii belonging to Tenericutes.
Preliminary results of the administration of phytase alone or combined with inositol on the gut microbiota of broilers
No statistically differences were reported in relation to the phyla identified in the different experimental groups with around 85% belonging to Firmicutes and Bacterodetes. The only statistical difference was reported for Bacilli class and Lactobacillaceae family between control group vs treated groups (P=0.010). Within the species identified statistically significant differences were observed for Lactobacillus johnsonii in control group vs B and C (P=0.048), Ruminococcus torques in control group vs D (P=0.021), Lactobacillus crispatus, helveticus and acidophilus in control group vs E (P=0.038, 0.02 and 0.02 respectively), Clostridium saccharolyticum in groups control, C and D vs E (P=0.016), Escherichia coli in groups B vs C (P=0.048), butyrate producing bacterium SL7/1 in groups C vs E (P=0.040). As reported in Table 3 the first five species identified in the different groups are quite similar with high abundance of Faecalibacterium prausnitzii ranging between 10.67 (group E) to 13.27 (group D).
The results obtained in the two trials show that using shotgun metagenomic sequencing is possible to detail the microbiota composition of chicken fed with different molecules. This knowledge can drive the nutritional strategies to achieve a specific microbial population in the chicken gut or even support specific microbial targets able to fight the colonization by foodborne pathogens improving safety of poultry meat.