Evaluation of extended-spectrum β-lactamase producing bacteria in feces of shelter dogs as a biomarker for altered gut microbial taxa and functional profiles

Abstract

Background: The USA is home to 83-88 million dogs, with 3-7 million living in shelters. Shelter dogs move through the supply chain from their geographical origin to adoptive homes, with possible exposure to pathogens and shift in their gut microbiota. However, research in this area is limited. This study examined the effects of intestinal colonization by ESBL bacteria on gut taxa abundance, diversity, and functions in 52 shelter dogs of various ages, sexes, and fertility statuses.

Methodology: We isolated fecal DNA, sequenced their 16S, processed the sequences using DADA2, identified taxa profiles in each dog by Phyloseq, and analyzed Chao1, Shannon, and Simpson alpha diversity by ggplot2 and Wilcoxon test. We analyzed beta diversity using Bray-Curtis dissimilarity matrix from the vegan package. Differential abundance of taxa, gut microbiome functions, and differential abundance of microbiome functions were analyzed using DESeq2, PICRUSt2, and ALDEx2, respectively, with Wilcoxon rank and Kruskal-Wallis tests for comparisons between dog groups.

Results: Firmicutes (69.3%), Bacteroidota (13.5%), Actinobacteriota (6.77%), Proteobacteria (5.54%), and Fusobacteriota (4.75%) were the major phyla in the gut of shelter dogs. ESBL bacteria colonized dogs had reduced gut microbiota alpha diversity than non-colonized dogs. The abundance levels of the following phyla (Proteobacteria, Deferribacterota, Bacteroidota, Fusobacteriota, and Spirochaetota), class (Gammaproteobacteria, Bacteroidia, Deferribacteres, Brachyspirae, and Fusobacteria), and families (Enterobacteriaceae, Peptostreptococcaceae, Lactobacillaceae, Lachnospiraceae, Prevotellaceae, and Peptostreptococcaceae) were significantly (p < 0.05) varied between the two dog groups. Further stratified analysis by age, sex, and spaying/neutering status influenced the abundance of taxa in ESBL bacteria colonized dogs, indicating these covariates act as effect modifiers. Most gut metabolic and biosynthetic pathways were downregulated in ESBL bacteria colonized dogs compared to non-colonized dogs. However, alpha-linolenic acid metabolism and shigellosis, fluorobenzoate degradation, allantoin degradation, toluene degradation, glycol degradation, fatty acid and beta-oxidation, and glyoxylate metabolism bypass pathways were increased in dogs colonized by ESBL bacteria.

Conclusion: Colonization by ESBL bacteria marks altered gut microbiota. Dog's demography and fertility status modify the alterations, indicating host factors and ESBL bacteria interplay to shape gut microbiota. ESBL bacteria or other factors reprogram gut microbiome functions through down and upregulating multiple metabolic and biosynthesis pathways to promote ESBL bacteria colonization.

Keywords: 16S amplicon sequencing; ESBL bacteria; alpha diversity; gut microbiota; microbiome function; shelter dogs.

Figure 1

Venn diagram of ASVs shared among the four groups: (a) sex, (b) fertility status (spayed females = FS, neutered males = MN, intact females = FI, intact males = MI), (c) age (1–3 months = age 1, 4–6 months = age 2, 7–12 months = age3, 12–60 months = age 4), and (d) ESBL status in shelter dogs. For sample size (n) of dogs in each group, look at sampling under the methodology above.

Figure 2

Boxplots of alpha diversity metrics with and without ESBL-bacteria colonization in the gut microbiota in shelter dogs. Boxplots display the Chao1, Shannon, and Simpson indices for ESBL-positive (n = 12) and ESBL-negative (n = 40) dogs. Statistical significance was determined using the Kruskal-Wallis test. Chao1 index showed a significant reduction in ESBL-positive dogs (p =  0.025), while Shannon (p = 0.086) and Simpson (p = 0.1517) indices showed no significant differences.

Figure 3

Boxplots of stratified alpha diversity analysis of ESBL-positive and ESBL-negative dogs by sex, fertility, and age. Alpha diversity metrics (Chao1, Shannon, Simpson indices) compared across ESBL-positive and ESBL-negative dogs, further stratified by sex (male vs. female), fertility status (intact vs. spayed/neutered), and age groups. Statistical significance was tested for significance using Kruskal-Wallis test.

Figure 4

Relative abundance and composition of the gut microbiota altered at phylum level in ESBL-positive and ESBL-negative dogs as stratified by sex, fertility, and age. The relative abundance of major bacterial phyla in ESBL-positive (n = 12) and ESBL-negative (n = 40) dogs. Statistical significance for group differences was determined using DESeq2 with Benjamini-Hochberg FDR correction (adjusted p < 0.05).

Figure 5

Relative abundance and composition of the gut microbiota altered at class level in ESBL-positive and ESBL-negative dogs as stratified by sex, fertility, and age. The bar plot displays taxonomic composition at the class level, comparing ESBL-positive and ESBL-negative dogs across different sex, fertility, and age groups. DESeq2 was used for statistical testing (p < 0.05).

Figure 6

Relative abundance and composition of the gut microbiota altered at family level in ESBL-positive and ESBL-negative dogs as stratified by sex, fertility, and age. Bar plot showing the relative abundance of bacterial families in ESBL-positive and ESBL-negative dogs. Statistical significance was assessed using DESeq2 with p < 0.05.

Figure 7

Relative abundance and composition of the gut microbiota altered at genus level in ESBL-positive and ESBL-negative dogs as stratified by sex, fertility, and age. Bar plot shows the relative abundance of bacterial genus in ESBL-positive and ESBL-negative dogs. Statistical significance was assessed using DESeq2 with p < 0.05.

Figure 8

Beta-analysis by NMDS Bray-Curtis distance for distinguishing dissimilarities in microbial community composition between ESBL-bacteria colonized and non-colonized dogs. The NMDS plot visualizes differences in microbial community structure between ESBL-positive (n = 12) and ESBL-negative (n = 40) dogs, based on Bray–Curtis dissimilarity. Statistical significance was determined using PERMANOVA (p < 0.05).

Figure 9

Gut microbiome functional pathways in KEGG database. PICRUSt2 analysis identified significantly altered pathways in ESBL-positive vs. ESBL-negative shelter dogs. Statistical testing was performed using Wilcoxon rank-sum test.

Figure 10

Gut microbiome functional pathways in MetaCyc database. Differentially abundant metabolic pathways identified via PICRUSt2 in ESBL-positive vs. ESBL-negative dogs. Statistical significance determined by Wilcoxon rank-sum test.