Protective mechanisms of a microbial oil against hypercholesterolemia: evidence from a zebrafish model

Abstract

A Western diet elevates the circulating lipoprotein and triglyceride levels which are the major risk factors in cardiovascular disease (CVD) development. Consumption of long-chain omega-3 fatty acids can stall the disease progression. Although these fatty acids can significantly impact the intestine under a hypercholesterolemic condition, the associated changes have not been studied in detail. Therefore, we investigated the alterations in the intestinal transcriptome along with the deviations in the plasma lipids and liver histomorphology of zebrafish offered DHA- and EPA-rich oil. Fish were allocated to 4 dietary treatments: a control group, a high cholesterol group and microbial oil groups with low (3.3%) and high (6.6%) inclusion levels. We quantified the total cholesterol, lipoprotein and triglyceride levels in the plasma. In addition, we assessed the liver histology, intestinal transcriptome and plasma lipidomic profiles of the study groups. The results suggested that higher levels of dietary microbial oil could control the CVD risk factor indices in zebrafish plasma. Furthermore, microbial oil-fed fish had fewer liver vacuoles and higher mRNA levels of genes involved in β-oxidation and HDL maturation. Analyses of the intestine transcriptome revealed that microbial oil supplementation could influence the expression of genes altered by a hypercholesterolemic diet. The plasma lipidomic profiles revealed that the higher level of microbial oil tested could elevate the long-chain poly-unsaturated fatty acid content of triglyceride species and lower the concentration of several lysophosphatidylcholine and diacylglycerol molecules. Our study provides insights into the effectiveness of microbial oil against dyslipidemia in zebrafish.

Keywords: DHA; EPA; RNA seq; bioactive compounds; cardiovascular disease; plasma lipidomics.

Figure 1

Alteration of the plasma lipids in zebrafish fed different experimental diets for a period of 12 weeks. Boxplots show total cholesterol (A), LDL cholesterol (B), HDL cholesterol (C) and total triglyceride (D) contents in the plasma of the fish. Blue dots inside boxplots indicate the mean values of the corresponding groups. CT-fish fed the control diet; HC-fish fed the high cholesterol diet (5.1% inclusion); HCA1-fish fed microbial oil (3.1% inclusion); HCA2-fish fed microbial oil (6.6% inclusion). ^***p < 0.001, ^**p < 0.01, ^*p < 0.05, and ^•p < 0.1. Each treatment group consisted of six biological replicates.

Figure 2

Alteration of the cardiovascular disease (CVD) risk indices in zebrafish fed different experimental diets for a period of 12 weeks. Boxplots show Castelli I index (A), Castelli II index (B), Atherogenic index (C) and Atherogenic coefficient (D) of plasma from the fish. Blue dots inside boxplots indicate the mean values of the corresponding groups. Scatter plot (E) shows the correlation between Castelli I index and total LDL cholesterol levels. Biplot (F) indicates the separation of the CT and HCA2 groups from the other 2 study groups and variables associated with each group. CT-fish fed the control diet; HC-fish fed the high cholesterol diet (5.1% inclusion); HCA1-fish fed microbial oil (3.1% inclusion); HCA2-fish fed microbial oil (6.6% inclusion). ^*p < 0.05 and ^•p < 0.1. Each treatment group consisted of six biological replicates.

Figure 3

Relative expression of selected genes in the liver of zebrafish fed different experimental diets. lecithin-cholesterol acyltransferase (lcat) (A); scavenger receptor class B, member 1 (scarb1) (B); carnitine palmitoyltransferase 1Aa (cpt1aa) (C); perilipin 2 (plin2) (D); acetyl-CoA acyltransferase 1 (acaa1) (E); ATP-binding cassette, sub-family A (ABC1), member 1A (abca1a) (F). Black dots indicate the relative expression of the respective genes in each sample. CT-fish fed the control diet; HC-fish fed the high cholesterol diet (5.1% inclusion); HCA1-fish fed microbial oil (3.1% inclusion); HCA2-fish fed microbial oil (6.6% inclusion). ^**p < 0.01 and ^*p < 0.05. Each treatment group consisted of six biological replicates.

Figure 4

Histomorphology of the liver of zebrafish fed different experimental diets for a period of 12 weeks. Representative histological images (A) of the liver of zebrafish. Dot-plot shows average number (B) and average size (C) of vacuoles in the liver of fish fed control (CT) diet, high cholesterol (HC) diet, HC diet supplemented with lower (HCA1) or higher (HCA2) levels of microbial oil. The scatter plot (D) shows correlation between average vacuole number and average vacuole size in the liver of zebrafish. ^***p < 0.001, ^**p < 0.01 and ^*p < 0.05. Each treatment group consisted of 9–12 biological replicates. Scale bar = 50 μm.

Figure 5

Network plot showing the link between enriched GO terms and the associated genes that were downregulated in the intestine of zebrafish fed high cholesterol diet. The significantly downregulated genes from the HC vs. CT comparison are written on the red nodes. Only the non-redundant GO terms are labelled in the network plot. For each GO term, the gradient color varies with the adjusted p value (Benjamini-Hochberg method). Adjusted p value (Benjamini-Hochberg method) < 0.05 and minimum gene count of 2 were set as cut-off parameters for each GO term.

Figure 6

Gene ontology (GO) term enrichment based on the downregulated genes in the HCA1 group. Chord diagram showing the link between the enriched GO terms and the associated downregulated genes from the HCA1 vs. CT group comparison. Genes that were considered for the enrichment analyses are shown on the left half of the circle. The gradient color bar intensity varies with the Log₂ fold change.

Figure 7

Network plot showing the link between enriched GO terms and the associated genes that were downregulated in zebrafish fed microbial oil (HCA2) supplemented diet. The significantly downregulated genes (HCA2 vs. CT) are written on the red nodes. Only the non-redundant GO terms are labelled and for each GO term, the gradient color varies with the adjusted p value (Benjamini-Hochberg method). Adjusted p value (Benjamini-Hochberg method) < 0.05 and minimum gene count of 2 were set as cut-off parameters for each GO term.

Figure 8

Alteration of genes related to cholesterol biosynthesis in zebrafish fed high cholesterol diet with and without microbial oil. Volcano plots highlighting the fold-changes in the intestinal cholesterol biosynthesis-related genes in (A) HC vs. CT, (B) HCA1 vs. CT and (C) HCA2 vs. CT transcriptome comparisons. Heatmap (D) showing hierarchical clustering of CT, HC, HCA1 and HCA2 groups. Each treatment group consisted of at least five biological replicates.

Figure 9

Differences in the transcriptome of the HCA2 and HC groups. Principal component analysis (A) and Hierarchical clustering (B) of the differentially expressed genes (DEGs) in the HCA2 group compared to the HC group. Transcripts with an adjusted p value (Benjamini-Hochberg method) < 0.05 and |Log₂ fold change| ≥ 1 were considered significantly differentially expressed. Dot-plot (C) showing the enriched GO terms, by considering the upregulated genes in the HCA2 vs. HC group comparison. Each treatment group consisted of at least five biological replicates.

Figure 10

Differences in the lipid classes and species in the plasma of zebrafish fed the high cholesterol diet with and without microbial oil for a period of 12 weeks. Pie chart (A) of lipid classes in the plasma of zebrafish fed the high cholesterol (HC) diet and microbial oil supplemented (HCA2) diet. PCA (B) of the lipidome data from the 6 plasma samples that were used for the analysis (n = 3 per group). The first two components captured 59 and 25% of the variation of the data, respectively. (C) Volcano plot illustrating the differentially abundant (|Log₂ fold-change| ≥ 1, p value < 0.05) lipid species in the HCA2 vs. HC diet group comparison. Red dots indicate downregulated and green dots indicate upregulated lipid species in the HCA2 group compared to the HC group. Heatmap (D) showing diet-group based hierarchical clustering of 40 differentially abundant lipid species.

Figure 11

Over representation analysis and correlation network of the differentially abundant lipid species. Network plot (A) showing the link between the enriched lipid classes and differentially abundant lipid species in the plasma of zebrafish. Over representation analysis was performed using a cut-off of p value < 0.05 after adjusting for multiple testing. The color of the nodes in the outer circle indicate the lipid species that had higher and lower abundance in the HCA2 group compared to the HC group. The gradient color varies with the p value for each enriched lipid class. Network plot (B) showing the correlation between the 40 differentially abundant lipid species. The color of the edges indicate positive (blue) or negative correlation. A particular lipid class is color coded and only the links with correlation coefficient greater than 0.75 or less than 0.75 and p value of < 0.05 are shown in the network.