Fig 1.
Diversity of uterine microbiota from the Healthy (n = 11), MNoFever (n = 12), and MFever (n = 11) groups.
(A) Number of reads. There is no significant difference among groups. (B) Chao1 richness index. Chao1 richness estimates the total number of species in a sample based on the number of singleton and doubleton taxa. (C) Shannon diversity index. Shannon index takes into account both the species richness and evenness. Data represent the mean ± SEM. Data labeled “a” are statistically different (P ≤ 0.05) from data labeled “b”.
Fig 2.
Dissimilarity of bacterial community.
(A) Principal coordinates analysis of uterine microbiota based on normalized values and Manhattan distance. This shows the ordination of the Healthy (healthy cows; green circles), MNoFever (metritic cows without a fever; blue squares), and MFever (metritic cows with a fever; red triangles). Variance explained by each PCO axis is given in parentheses. (B) Bray-Curtis dissimilarity of uterine microbiota using phylum level abundance of bacteria. An index of 0 indicates the most similarity between bacterial communities, whereas an index of 1 indicates the most dissimilarity between bacterial communities. Bars represent the mean ± SEM. (C) Phylum level assignment. (D) Genus level assignment. Bacterial phyla and genera having ≥ 1% abundance were presented in a stacked bar graph. Data labeled “a” are statistically different (P ≤ 0.05) from data labeled “b”.
Fig 3.
Heat map analysis with the dendrogram based on Euclidean distance.
Columns represent individual cows and rows represent 20 bacterial species. The relative abundance of bacteria was presented by color in each cell. Under the dendrogram the first bar shows whether cows had fever and the second bar indicates whether cows had metritis.
Fig 4.
Associations of Bacteroides pyogenes with a fever.
(A) Relative abundance of B. pyogenes based on metagenomic sequencing. Bars represent the mean ± SEM. (B) Absolute abundance of total bacteria and B. pyogenes based on droplet digital PCR. The log10 number of total bacteria (black circles) and B. pyogenes (blue squares) were quantified in uterine swab samples using the 16S rRNA gene and recA gene, respectively. Each symbol represents an individual cow and error bars indicate the means ± SEM. Data labeled “a” are statistically different (P ≤ 0.05) from data labeled “b”.
Fig 5.
(A) In vitro proportion of PMN with phagocytic activity (Healthy = 58, MNoFever = 19, MFever = 33). (B) In vitro proportion of PMN mediated oxidative burst (Healthy = 58, MNoFever = 19, MFever = 33). (C) In vivo serum TNFα (14 cows per group). (D) In vivo serum PGE2 metabolite (14 cows per group). Data represent the mean ± SEM. Data labeled “a” are statistically different (P ≤ 0.05) from data labeled “b”. See S1 Table for sample information.