Growth performance
The growth performance results are summarized in Table 3. From d 1 to 23, SE infection alone showed a significant decrease in ADG, a reduced trend in ADFI and an increased trend in FCR as compared with the non-infected control (A) (P < 0.05). Compared with the SE-infected alone group, higher BWG and ADG was observed in infected chickens received three doses of EOA groups (P < 0.05) but had no significant difference in ADG among these three levels of EOA groups (P > 0.05). From d 1 to 23, ADFI exhibited a quadratic change with increasing levels of EOA. Moreover, infected birds given the medium-dose of EOA had higher ADFI than the other infected treatments, while lower FCR was found in the infected birds given low-dose EOA group.
During the grower and finisher period, SE challenge notably increased FCR compared with the non-infected control (P < 0.05). Nevertheless, FCR showed a linear decrease with the increase of EOA addition level. The infected birds fed with the high dose of EOA showed the lower FCR compared with the infected control B. There was no significant differences in ABW, ADG and ADFI among all groups from days 24 to 39 (P > 0.05).
During the overall period, although no significant differences in ABW, ADG and ADFI was found among all groups from days 1 to 39 (P > 0.05), dietary EOA addition showed a linear decrease in FCR with increasing EOA level (P < 0.05). Moreover, cumulative FCR from 1 to 39 days was better when broiler chickens was fed diets containing high dose of EOA compared to the infected B group. Additionally, no significant difference in mortality rate was observed among groups during the whole trial phase.
Ileal Morphology
As shown in Table 4 and Fig. 1. At 3 DPI, VH and VH/CD values of ileum was lower in single SE-infected B group than the non-infected A group (P < 0.05). Compared with the SE-infected B group, VH/CD values of the ileum linearly increase with the increasing levels of EOA addition (P < 0.05), moreover, the BM group had higher VH/CD value of the ileum, which was close to that of the non-infected A group (P < 0.05). VH of ileum in the three levels of EOA-treated groups showed a linear increased trend (P = 0.07), but had no significant difference compared with the SE-infected B group (P > 0.05).
At 10 DPI, higher CD and lower VH/CD were observed in the ileum of SE-infected B group (P < 0.05) than that in non-infected A group. Compared with the infected alone group, EOA addition significantly decreased crypt depth and notably increased VH/CD values of the infected birds (P < 0.05), and showed a linear changes in CD and VH/CD with increasing dose of EOA supplementation (P < 0.05).
Caecal bacterial colonization and internal organs Salmonella invasion
As shown in Table 5, compared with non-infected group (A), Salmonella was only detected in liver and spleen of the single infected chickens at 3 and 10 DPI and in liver of the BM groups at 3 DPI (P > 0.05). Moreover, infected chickens given low dosage and high dosage of EOA had no Salmonella found in the liver and spleen at 3 and 10 DPI. In addition, no significant difference in Salmonella concentration in the liver and spleen was observed among all groups (P > 0.05). However, single SE infection significantly increased the number of Escherichia coli in the liver at 3 and 10 DPI (P < 0.05), showed an increased trend for splenic Escherichia coli carriage at 3 DPI (P = 0.083), but had no notable influence on splenic Escherichia coli count at 10 DPI (P > 0.05), when compared with the non-infected group (A). Also, dietary EOA addition only numerically reduced Escherichia coli load in the liver and spleen of infected chickens, but had no significant difference in Escherichia coli counts compared to the infected group (B) (P > 0.05).
As summarized in Table 6, the numbers of Salmonella, Escherichia coli, Lactobacillus and Campylobacter in the infected positive group were significantly higher than that in the non-infected negative group at 3 DPI and 10 DPI (P < 0.05). Dietary supplementation of EOA exhibited a significant linear decrease in the number of Salmonella, Escherichia coli and Lactobacillus in cecal digesta at 3 DPI (P < 0.05). Also, infected chickens fed diets supplemented with different concentrations of EOA significantly decreased the concentration of Salmonella and Lactobacillus in the cecum compared with the single infected group at 3 DPI (P < 0.05). Interestingly, at 10 DPI, Lactobacillus counts in the cecum showed a quadratic change with supplemental levels of EOA. Moreover, moderate dosage and high dosage of EOA addition to diets of infected chickens remarkably decreased the number of Salmonella and Lactobacillus in the cecum compared with the infected positive group (P < 0.05).
Gene Expression Of Tight Junction Protein Genes And Immune-related Genes In The Ileum
Table 7 presents the results of ileal barrier-related gene expression in broilers. On 3 DPI, SE infection alone significantly downregulated the mRNA levels of Claudin-1, Occludin, ZO-1 and MUC-2 (P < 0.05), but had no effect on FABP-2 mRNA level (P > 0.05) in the ileum as compared with the non-infected A group, indicating that SE infection damage intestinal barrier function. Nevertheless, no significant differences in Claudin-1, Occludin, ZO-1, MUC-2 and FABP-2 mRNA levels was observed in the ileum among the BL, BM, BH and B groups at 3 DPI (P > 0.05).
At 10 DPI, our data showed that the gene expression of Claudin-1, Occludin, ZO-1 and MUC-2 (P < 0.05) in the ileum was significantly downregulated by Salmonella infection (P < 0.05), but the gene expression of Occludin (P = 0.05) and ZO-1 (P = 0.09 < 0.10) showed a decreased trend after SE challenge compared with non-infected control. These data indicated that Salmonella infection caused intestinal barrier injury. However, the mRNA levels of Claudin-1, Occludin, ZO-1, MUC-2 and FABP-2 were significantly upregulated by EOA administration at 10 DPI (P < 0.05). Moreover, dietary supplementation of EOA linearly and quadratically increased Claudin-1, Occludin, ZO-1, MUC-2 and FABP-2 mRNA levels at 10 DPI with increasing EOA levels in the diet (P < 0.05).
The results of immune-related gene expression was listed in Table 8. SE infection alone significantly upregulated TNF-α mRNA levels in the ileum (P = 0.012), and showed an increased trend for the mRNA levels of IL-1β (P = 0.10) and NF-κB (P = 0.08) at 3 DPI compared with the non-infected chickens. Intriguingly, at 10 DPI, no significant difference for immune-related genes (TLR4, MyD88, NF-κB, IL-1β, TNF-a, IL-6 and IFN-γ) was observed in the ileum between the infected positive group and non-infected negative group at 10 DPI (P > 0.05). The data possibly indicated that SE infection could induce intestinal inflammatory response at the early infection stage, but during the middle and later phase of infection, chicken might show immune tolerance to SE bacteria. However, compared with the single infected control, at 3 DPI, dietary addition of EOA tend to linearly reduce the gene expression level of TLR4, NF-κB and IL-1β (0.05 < P < 0.10). Furthermore, the BM group significantly reduced ileal IL-1β and TNF-α mRNA levels (0.05 < P < 0.10), but the BH and BL group showed a decreased trend for these two genes mRNA levels (0.10 < P < 0.05).
At 10 DPI, dietary supplementation of EOA showed a significant linear decreasing effect on TLR4 mRNA level, displayed a quadratic effect on NF-κB mRNA levels and had a significant linear and quadratic influence on MyD88 and IL-6 mRNA levels (P < 0.05). Moreover, dietary different dosage of EOA administration all significantly reduced TLR4 mRNA levels in the ileum (P = 0.004). Besides, infected chickens given the BL diet exhibited a notable upregulation for ileal MyD88, NF-κB and IL-6 genes expression levels (P < 0.05) compared with the infected alone control and the BH group. Supplemental EOA at moderate dosage only remarkably increased NF-κB and IL-1β genes expression levels in the ileum of the infected chickens compared with the single infected control (P < 0.05), but had no significant difference in NF-κB and IL-1β mRNA levels when compared with the BL group. However, infected chickens given diets supplemented with higher level of EOA had no significant influence on ileal TLR4, MyD88, NF-κB and IL-1β genes expression levels (P > 0.05), but significantly reduced TNF-α mRNA levels (P < 0.05) when compared with the infected positive control.
Anti- Salmonella specific IgA and IgG concentrations
As presented in Table 9, single SE infection significantly enhanced the production of specific IgA against Salmonella in the ileum digesta at 10 DPI, but had no notable effects on serum specific anti-SE IgG relative to the non-infected control, showing that SE infection induced intestinal mucosal immune inflammatory responses (P = 0.033). The medium and higher dosage of EOA addition notably increase serum anti-SE IgG level (P = 0.020) and ileal anti-SE IgA (P = 0.014) concentration at the early infection stage when compared with the single infected positive group and the lower dose of group (P < 0.05).
Concentration Of Short-chain Fatty Acids In Cecal Content
As illustrated in Table 10, SE infection alone significantly reduced the concentration of isobutyric acid in the cecum digesta compared with the non-infected control and the EOA-treated groups (P < 0.05). However, adding EOA in the diet linearly changed iso-butyric acid concentration in cecal digesta of infected broilers (P < 0.05). Also, the concentration of iso-butyric acid in the BL and BH groups were significantly increased compared with the positive group (P < 0.05), and an increased trend for iso-butyric acid level in the BM group.
Cecal Microbiome Analysis By 16s Rrna Sequencing And Bioinformatics
In this study, 551 OTUs were obtained from ceca contents samples of the four groups based on 97% sequence similarity level. Venn diagram (Fig. 2, a) indicated 421 common core OTUs were shared by the four groups, while 20, 14, 6 and 6 OTUs were unique to groups A, B, BM and BH, respectively. There were no significant differences (P > 0.05) in ACE index, Chao1 index, Simpson index and Shannon index among all dietary treatments (Fig. 2, b-e), indicating that cecum microbial α diversity was not influenced by EOA treatment or Salmonella challenge. In order to study the similarity or difference of cecum microbial community structure in different samples, the β-diversity of cecal microorganisms was assessed by PCA analysis and PcoA analysis. PCA analysis showed that there was significant separation in cecal microbial community structure among the four groups (P = 0.006) (Fig. 2, f-g), especially between the infected control and non-infected control, and between the infected control and the BM group.
As presented in Fig. 3, a. At the phyla level, ceca microbiota was dominated by Firmicutes, Bacteroidota, Actinobacteriota, followed by Verrucomicrobiota and Proteobacteria for all treatments, with no significant differences in the relative abundance among four treatment groups (P > 0.05). At the genus taxa, the top 10 genera in abundance were Lactobacillus (21.44%), Faecalibacterium (10.17%), Alistipes (9.05%), Bacteroides, (8.44%), unclassified_f_Lachnospiraceae (7.10%), norank_f_norank_o_Clostridia_UCG-014 (4.72%), Ruminococcus torques group (3.59%), UCG-005 (3.42%), followed by norank_f_norank_o_Clostridia_vadinBB60_group (3.32%) and Butyricicoccus (2.70%) (Fig. 3, b). The comparison of cecal bacterial compositions among four groups showed that the relative abundance of unclassified_f__Lachnospiraceae was significantly (P < 0.05) increased in the single SE-infected group, while the relative abundance of Butyricicoccus was significantly (P < 0.05) increased in BM group. In addition, the relative abundances of Norank_f__Oscillospiraceae, Eisenbergiella and Flavonifractor were significantly (P < 0.05) increased in the non-infected group, BH group and BM group respectively (Fig. 3, c). Salmonella infection also significantly (P < 0.05) decreased the relative abundance of norank_f_norank_o_Oscillospiraceae, norank_f_norank_o_Rhodospirillales and Eggerthella. However, dietary EOA treatment significantly (P < 0.05) increased relative abundance of Butyricicoccus, unclassified_f__Oscillospiraceae, Anaerotruncus, unclassified_f__Bacillaceae and Enterococcus, whereas decreased relative abundance of unclassified_f_Lachnospiraceae, norank_f__norank_o__Clostridia_vadinBB60_group, Eisenbergiella, UCG-009 and Merdibacter (P < 0.05).
LEfSe analysis (Fig. 4) showed that g__norank_f__Oscillospiraceae, g__Lachnospiraceae_NK4A136_group, g__Eggerthella, f__norank_o__Rhodospirillales, g__norank_f__norank_o__Rhodospirillales, o__Rhodospirillales and c__Alphaproteobacteria were significantly (P < 0.05) enriched in the non-infected group, while g__unclassified_f__Lachnospiraceae and g__UCG-009 were significantly (P < 0.05) enriched in the positive B group. Moreover, g__Butyricicoccus、f__Butyricicoccaceae、g__Anaerotruncus、g__norank_f__norank_o__Oscillospirales、g__unclassified_f__Bacillaceae、o__Bacillales、f__Bacillaceae、g__Flavonifractor、f__Enterococcaceae and g__Enterococcus were significantly (P < 0.05) enriched in the BM group, and g__Eisenbergiella and g__Anaerofilum were significantly (P < 0.05) enriched in the BH group (Fig. 6, b). Interestingly, Lefse analysis also found that g__Butyricicoccus, g__unclassified_f__Oscillospiraceae, g__Anaerotruncus, g__unclassified_f__Bacillaceae, g__Enterococcus, g__Eggerthella and g__Eubacterium were important bacteria that distinguished the BM group from the single SE-infected group (P < 0.05).
PICRUSt analysis exhibited that functions related to microbial infection and anti-infection such as Salmonella infection, Shigellosis, nucleotide oligomerization domain-like (NOD-like) receptor signaling pathway, streptomycin biosynthesis, prodigiosin biosynthesis, acarbose and validamycin biosynthesis, biotin metabolism, ascorbate and aldarate metabolism, biosynthesis of vancomycin group antibiotics and insulin signaling pathway, were significantly enhanced in Salmonella-infected group compared with the non-infected A group (P < 0.05) (Fig. 5, a). Comparing with the single SE-infected group, D-arginine and D-ornithine metabolism, ethylbenzene degradation, furfural degradation, alpha-linolenic acid metabolism, microbial metabolism in diverse environments, fatty acid metabolism, bacterial secretion system and biosynthesis of unsaturated fatty acids were significantly enhanced in EOA-treated group (P < 0.05), while Salmonella infection, thiamine metabolism, Shigellosis, NOD-like receptor signaling pathway, flagellar assembly and biosynthesis of vancomycin group antibiotics were significantly enriched in infected group (P < 0.05) (Fig. 5, b and Fig. 5, c).
It is vital to construct a network between the differential microbiota and the expressions of intestinal tight junction protein genes and immune-related genes together with SCFA concentration of cecal content to understand how the intestinal host–microbial relationship regulates host defense and inflammation (Fig. 6). Results of the spearman’s correlation coefficients showed that the relative abundances of unclassified_f_Lachnospiraceae (significantly enriched in Salmonella-infected chickens) was markedly negatively correlated with the relative mRNA expression levels of MUC-2, FABP-2 and MyD88, and concentration of isobutyric acid and isovaleric acid in cecal content (P < 0.05 or P < 0.01). The Butyricicoccus showed a positive regulatory effect on the mRNA expression of Claudin-1, Occludin, FABP-2, NF-κB, MyD88, IL-6 and IFN-γ (P < 0.05 or P < 0.01), while the relative abundance of g__norank_f__Oscillospiraceae had a negative correlation with the relative mRNA expression of FABP-2, but displayed a positive correlation with the concentration of valeric acid. In addition, the significant positive correlation between the relative abundances of g__Flavonifractor and the relative mRNA expression of FABP-2 was observed (P < 0.05).