Fig 1.
Analyses of the gut microbiome of sailfin mollies acclimated to either freshwater or seawater for 14 days.
Taxonomic bar plots displaying the relative abundance (%) of the top five most abundant bacterial phyla (a) and the top ten most abundant bacterial families (b) across the four conditions of: FW14-Anterior, FW14-Posterior, SW14-Anterior, and SW14-Posterior are shown. Alpha-diversity (Shannon Diversity Index; c) was compared across conditions using a Kruskal-Wallis test and conditions that share letters are not significantly different. Beta-diversity was assessed using Bray-Curtis dissimilarity and visualized with principal coordinate analysis (Axis 1 vs. Axis 2; d and Axis 1 vs. Axis 3; e) were also observed across the four conditions.
Fig 2.
LDA of the gut microbiomes of freshwater and seawater sailfin mollies at the family level.
The anterior (a) and posterior intestines (c) were compared across the freshwater and seawater environments and the respective cladograms (anterior intestines: b; posterior intestines: d) are also shown. Patterns of intestinal zonation in the gut microbiome of freshwater sailfin mollies (e) and seawater sailfin mollies (g) are plotted alongside the respective cladogram for each environment (freshwater: f; seawater: h).
Fig 3.
Predicted differential KEGG pathways across environmental salinities and intestinal sections.
The top 15 (based on highest mean proportions) predicted differential KEGG pathways (Level 2 based on the full set of abundance data) between the freshwater and seawater environments for the anterior (a) and posterior (b) intestines; as well as intestinal zonation between the anterior and posterior intestines in freshwater (c) and seawater (d). The bar plot on the left depicts each KEGG pathway’s mean proportion (%) and on the right, the difference in mean proportion (%) between the two groups, the 95% confidence interval, and p-value are shown.
Fig 4.
Predicted functional differences in the gut microbiome of freshwater and seawater sailfin mollies are shown, with significant variations observed in the proportion of sequences (%) for several key metabolic functions.
Significant differences were observed in the proportion of sequences (%) for the following functions: metabolism (a); digestive system (b); transport and catabolism (c); protein digestion and absorption (d); amino acid metabolism (e); and glyoxylate and dicarboxylate metabolism (f). The star symbols (∂) represent the mean of each group and the addition symbols (+) represent outliers. Each of the four groups (FW14-Anterior, FW14-Posterior, SW14-Anterior, SW-14-Posterior) have an n = 5. A two-way ANOVA with Tukey post-hoc test was used for each panel. Within each panel, groups that share letters are not significantly different.
Fig 5.
Oxalate concentrations in the plasma (µM;a), and in the following tissues (µmol mg wet tissue weight-1): anterior intestines (b), posterior intestines (c), and kidneys (d), liver (e), and urine (f) across the FW14, SW14, and SW28 conditions.
The total, free, and precipitated oxalate concentrations are represented by black, dark grey, and light grey bars respectively. The bars represent column means ± SEM (n-values). Within each panel, bars that share letters are not statistically different. For panels a, d, e, and f, mixed-effects linear model regressions were employed. For panels b and c, the datasets did not pass normality and generalized additive models were used. In all analyses, individual fish were treated as the unit of observation, and tank identity and sex were included in the models to account for potential clustering; no significant tank or sex effects were detected, so fish were treated as independent observations for inference.
Fig 6.
SLC26A3 (striped bars) and SLC26A6 (cross-hatched bars) gene expression levels (%) in the anterior intestines (a), posterior intestines (b), and kidneys (c) across the FW14, SW14, and SW28 conditions.
The relative gene expression levels are expressed as percentages with the control (FW14) condition set to 100%. The bars represent column means ± SEM (n-values). Within each panel, bars that share symbols are not statistically different. For SLC26A3, datasets in both the anterior and posterior intestines (a & b) did not pass normality and generalized additive models were used. For SLC26A6 in the anterior intestine (a), all model assumptions were met, and a mixed-effects linear model regression was used. For SLC26A6, datasets in the posterior intestine (b) and kidney (c) were log(10)-transformed to meet normality and then mixed-effects linear model regressions were employed. In all analyses, individual fish were treated as the unit of observation, and tank identity and sex were included in the models to account for potential clustering; no significant tank or sex effects were detected, so fish were treated as independent observations for inference.
Fig 7.
Oxalate concentrations in the plasma (µM; a), and in the following tissues (µmol mg wet tissue weight-1): anterior intestines (b), posterior intestines (c), and kidneys (d), liver (e), and urine (f) in both the control and antibiotic treatment conditions.
Both the control and antibiotic exposures were run for 14 days in seawater. The total, free, and precipitated oxalate concentrations are represented by black, dark grey, and light grey bars respectively. The bars represent column means ± SEM (n-values). Within each panel, bars that share letters are not statistically different. For panels a, d, e, and f, mixed-effects linear model regressions were utilized. For panels b and c, the datasets did not pass normality and thus, generalized additive models were used. In all analyses, individual fish were treated as the unit of observation, and tank identity and sex were included in the models to account for potential clustering; no significant tank or sex effects were detected, so fish were treated as independent observations for inference.
Fig 8.
SLC26A3 (striped bars) and SLC26A6 (cross-hatched bars) gene expression levels (%) in the anterior intestines(a), posterior intestines (b), and kidneys (c) in both the control and antibiotic treatment.
Both the control and antibiotic exposures were run for 14 days in seawater. The relative gene expression levels are expressed as percentages with the control (SW14) condition set to 100%. The bars are means ± SEM (n-values). Within each panel, bars that share symbols are not statistically different. For SLC26A3, the dataset for the anterior intestine (a) was log(10)-transformed to meet normality and subsequently a mixed-effects linear model regression was used. For the posterior intestine (b), the SLC26A3 dataset did not require any transformation prior to utilizing a mixed-effects linear model regression. For the anterior intestine (a), the SLC26A6 dataset did not require any transformation prior to using a mixed-effects linear model regression. For SLC26A6, the datasets for the posterior intestine (b) and kidney (c) were log(10)-transformed to meet normality and mixed-effects linear model regressions were used. In all analyses, individual fish were treated as the unit of observation, and tank identity and sex were included in the models to account for potential clustering; no significant tank or sex effects were detected, so fish were treated as independent observations for inference.