Evaluation of gastrointestinal and fecal microbial communities as markers of liver abscess risk in beef feedlot cattle

Abstract

Liver abscesses (LAs) are a prevalent and costly issue in the beef industry. Microbial translocation from the gastrointestinal tract is believed to be the underlying cause of LAs. However, little is understood about the factors that permit the passage of bacteria into the portal bloodstream, or where in the gastrointestinal tract this translocation is likely to occur. This study used 16S rRNA gene sequencing to characterize the microbial community composition of the rumen, small intestine, large intestine, and feces of steers with edible and abscessed livers of varying severity. The small intestine of steers with severe LAs had increased (P = 0.02) richness and evenness compared to cattle with edible livers. However, there were no differences in the alpha diversity among samples collected at other locations (P ≥ 0.08). Small intestine samples also had a reduced (P < 0.01) Firmicutes to Bacteroidota ratio for cattle with LAs compared to those with edible livers. Prevotellaceae and Synergistaceae family abundance differed in the small intestine of cattle with LAs. Differences in community composition were not identified in fecal, colon, or rumen samples in association with LA occurrence, and there were no differences related to bacteria that have been traditionally considered LA pathogens (e.g. Fusobacteria or Trueperella spp.). These findings suggest that the microbial communities of the small intestine may have an important influence on LA occurrence. However, differences in microbial communities were not identified in rumen and fecal samples that might be used in predicting LA occurrence. Although sampling techniques during the feeding period are still limited, advances in this area would greatly benefit LA research.

Keywords: 16S rRNA; Firmicutes to Bacteroidota ratio; gastrointestinal health; microbiome.

Figure 1.

Boxplots showing microbial alpha diversity measures across gastrointestinal regions by liver score classification. (A, C, E, G) Observed amplicon sequencing variants (ASVs) represent richness; (B, D, F, H) Shannon’s Diversity Index represents richness and evenness. Comparisons are shown for the rumen (A–B), small intestine (C–D), large intestine (E–F), and fecal samples (G–H). Liver scores include Edible, A−, A, and A+. P-values indicate statistical significance of differences between groups.

Figure 2.

Ordination plots by sample location comparing microbial community structure by liver score. Nonmetric multidimensional scaling (NMDS) of generalized Unifrac distances illustrate the differences in community structure by liver score in (A) rumen samples, (B) small intestine samples, (C) large intestine samples, and (D) fecal samples. Ellipses represent the 95% confidence interval for the group means. In Panel C, one outlier has been removed to reduce scaling issues and improve the visibility of the plot; ordination of the full dataset can be seen in Figure S6.

Figure 3.

Ordination of microbial community structure by treatment group. Nonmetric multidimensional scaling (NMDS) of generalized Unifrac distances illustrate the differences in community structure by treatment in (A) rumen samples, (B) small intestine samples, (C) large intestine samples, and (D) fecal samples. Ellipses represent the 95% confidence interval for the group means. Treatments were an industry-standard control diet and normal management (CONREG), a control diet and erratic management (CONERR), a low roughage, highly fermentable diet and normal management (HOTREG), or a low roughage highly fermentable diet with erratic management (HOTERR). In Panel C, one outlier has been removed to reduce scaling issues and improve the visibility of the plot; ordination of the full dataset can be seen in Figure S7.

Figure 4.

Hierarchical clustering and taxonomic composition of rumen samples at the family level. The upper panel shows a dendrogram based on Ward’s clustering of microbial community composition, with individual samples color-coded by liver score and treatment. The lower panel shows stacked bar plots of relative abundances for the same samples, highlighting the eight most abundant bacterial families. Each bar corresponds to an individual sample. Treatments include CONREG (control diet, regular management), CONERR (control diet, erratic management), HOTREG (high fermentable diet, regular management), and HOTERR (high fermentable diet, erratic management). This figure was made in R and legends were added using Biorender.com.

Figure 5.

Hierarchical clustering and microbial composition of small intestine samples at the family level. The top panel shows a dendrogram based on Ward’s distance, with samples color-coded by liver score and treatment group. The bottom panel displays corresponding stacked bar plots of family-level relative abundances for each sample. Each bar corresponds to an individual sample. The legend shows the 8 most abundant families. Treatments include CONREG (control diet, regular management), CONERR (control diet, erratic management), HOTREG (high fermentable diet, regular management), and HOTERR (high fermentable diet, erratic management). This figure was made in R and legends were added using Biorender.com.

Figure 6.

Hierarchical clustering and microbial composition of large intestine samples at the family level. The top panel displays a Ward’s distance dendrogram, with samples labeled by liver score and treatment. The lower panel shows stacked bar plots of family-level relative abundances. Each bar corresponds to an individual sample.The eight most abundant families are shown in the legend.Treatments include CONREG (control diet, regular management), CONERR (control diet, erratic management), HOTREG (high fermentable diet, regular management), and HOTERR (high fermentable diet, erratic management).This figure was made in R and legends were added using Biorender.com.

Figure 7.

Hierarchical clustering and microbial composition of fecal samples at the family level. The top panel shows a dendrogram constructed using Ward’s distance, with sample labels color-coded by liver score and treatment. The bottom panel displays stacked bar plots of relative abundance for each fecal sample. Treatments include CONREG (control diet, regular management), CONERR (control diet, erratic management), HOTREG (high fermentable diet, regular management), and HOTERR (high fermentable diet, erratic management). Each bar corresponds to an individual sample. The  most abundant families include Lachnospiraceae, Peptostreptococcaceae, Oscillospiraceae, Prevotellaceae, Atopobiaceae.

Figure 8.

Firmicutes:Bacteroidota ratios across liver score categories in different gastrointestinal sample types. Boxplots show the ratio of Firmicutes to Bacteroidota families in (A) rumen, (B) small intestine, (C) large intestine, and (D) fecal samples, plotted on a log scale. The ratio was calculated by dividing the total relative abundance of samples for Firmicutes families by that of Bacteroidota families. Samples are grouped by liver score (Edible, A−, A, A+). P-values indicate the statistical significance of group differences.

Figure 9.

Differentially abundant bacterial families identified when comparing animals with edible and A+ liver scores, across gastrointestinal sites. Panels display log-fold change (LogFC) values for bacterial familes in (A) rumen, (B) small intestine, and (C) large intestine samples. Black dots indicate taxa that were differentially abundant (P < 0.05), and red dots denote those that also passed sensitivity analyses. Positive values indicate higher abundance in cattle with A+ abscessed liveres compared to cattle with edible livers; negative values indicate reduced abundance in A+ cattle.

Figure 10.

Differentially abundant bacterial families in fecal samples identified when comparing animals with edible livers to those with different liver abscess scores: A− (Panel A), A (Panel B), and A+ (Panel C). Each black dot represents a bacterial family that were differentially abundant (P<0.05); no features shown passed additional sensitivity analysis. Positive values reflect decreased abundance in cattle with edible livers; negative values indicate greater abundance in cattle with edible livers.