Longitudinal analysis of exposure to a low concentration of oxytetracycline on the zebrafish gut microbiome

Abstract

Oxytetracycline, a widely produced and administered antibiotic, is uncontrollably released in low concentrations in various types of environments. However, the impact of exposure to such low concentrations of antibiotics on the host remains poorly understood. In this study, we exposed zebrafish to a low concentration (5,000 ng/L) of oxytetracycline for 1 month, collected samples longitudinally (Baseline, and Days 3, 6, 9, 12, 24, and 30), and elucidated the impact of exposure on microbial composition, antibiotic resistance genes, mobile genetic elements, and phospholipid metabolism pathway through comparison of the sequenced data with respective sequence databases. We identified Pseudomonas aeruginosa, a well-known pathogen, to be significantly positively associated with the duration of oxytetracycline exposure (Adjusted P = 5.829e^-03). Several tetracycline resistance genes (e.g., tetE) not only showed significantly higher abundance in the exposed samples but were also positively associated with the duration of exposure (Adjusted P = 1.114e^-02). Furthermore, in the exposed group, the relative abundance of genes involved in phospholipid metabolism had also decreased. Lastly, we characterized the impact of exposure on zebrafish intestinal structure and found that the goblet cell counts were decreased (~82%) after exposure. Overall, our results show that a low concentration of oxytetracycline can increase the abundance of pathogenic bacteria and lower the abundance of key metabolic pathways in the zebrafish gut microbiome that can render them prone to bacterial infections and health-associated complications.

Keywords: antibiotic exposure; environmental pollution; goblet cells; gut microbiome; longitudinal analysis; low concentration; oxytetracycline.

Figure 1

Graphical abstract of the study design. Zebrafish were grouped into control and exposure categories, and longitudinal samples were collected at Baseline, D03, D06, D12, D18, D24 and D30. The exposure group was exposed to 5000 ng/L of OTC for 30 days whereas no exposue was performed in the control group. The longitudinally collected samples were sequenced and data analysis was performed.

Figure 2

Trends in the α-diversity of the zebrafish gut microbiome during OTC exposure. (A–G) α-diversity (Shannon Index) comparisons between the control and OTC exposure groups at different timepoints during the one-month exposure window are shown. These results indicated the differences in the α-diversity were not statistically significant (Wilcoxon test, P > 0.05) between the two groups at all of the sampled timepoints. However, the mean α-diversity remained relatively higher in the OTC exposure group.

Figure 3

Distribution of the ten highly abundant microbial species in the zebrafish gut microbiome. Taxonomic analysis of the sequencing reads identified more than 300 microbial species in the zebrafish gut microbiome. However, for visual convenience, here we shown only the top ten microbial species. The complete list of species and their relative abundances are provided in Supplementary Table 2.

Figure 4

Five most significant associations between microbial species and the duration of OTC exposure. MaAsLin2 was used to evaluate the association between different microbial taxa with the duration of OTC exposure. We identified numerous positive and negative associations between the microbes and OTC exposure (Supplementary Table 4). However, Pseudomonas aeruginosa (A), Alicycliphilus denitrificans (B), Undibacterium parvum (C), Fimbriimonas ginsengisoli (D), and Acidovorax carolinensis (E) showed the strongest positive associations. Benjamini and Hochberg false discovery rate (FDR) controlling procedure was performed for the correction of P-values (Padj) and only Padj < 0.1 were considered statistically significant.

Figure 5

Composition of antibiotic resistance genes and their association with mobile genetic elements in the zebrafish gut microbiome. (A) Dendrogram based on the composition of ARGs indicates that the samples from control and OTC exposure groups clustered separately. (B) Relative abundance of the ARGs identified to be highly abundant in the OTC exposure group. (C,D) tet(E) and acrA showed significant positive correlation with the duration of exposure to OTC. (E) The hallagram shows clusters of significantly correlated antibiotic resistance genes and mobile genetic elements. Correlations were computed using Spearman's coefficient, Benjamini and Hochberg false discovery rate (FDR) controlling procedure was performed for the correction of P values (Padj) and only Padj < 0.1 was considered statistically significant. Clusters are ranked according to their significance (Padj) i.e., 1 represents the most significant correlation and vice versa. Here, only top 25 correlation clusters are shown while non-significant correlations and clusters are masked.

Figure 6

Impact of OTC exposure on microbial phospholipid metabolism and zebrafish intestinal goblet cells. (A) Heatmap showing the relative abundance of major genes involved in phospholipid synthesis or degradation. Majority of the genes involved in phospholipid synthesis (e.g., accB, accC, accD, and fabI) were highly abundant in control whereas we observed decreased abundance of these genes after OTC exposure. Furthermore, no major differences were observed in the relative abundance of phospholipid degrading genes (e.g., fadA, fadB, fadD, fadE, and fadL) in control and OTC exposure groups. (B–D) Histological examinations in zebrafish gut following exposure to OTC for 30 d (200× magnification, scale bars 50 μm). (B) Control group, (C) OTC exposure group and (D) The number of goblet cells (as indicated by the black arrows) in zebrafish gut (n = 5). The data presented are the mean ± SEM of triplicates. Significant differences between OTC groups and the corresponding control group are indicated by *P < 0.05, **P < 0.01.