Surviving in a toxic world: transcriptomics and gene expression profiling in response to environmental pollution in the critically endangered European eel

Abstract

Background: Genomic and transcriptomic approaches have the potential for unveiling the genome-wide response to environmental perturbations. The abundance of the catadromous European eel (Anguilla anguilla) stock has been declining since the 1980s probably due to a combination of anthropogenic and climatic factors. In this paper, we explore the transcriptomic dynamics between individuals from high (river Tiber, Italy) and low pollution (lake Bolsena, Italy) environments, which were measured for 36 PCBs, several organochlorine pesticides and brominated flame retardants and nine metals.

Results: To this end, we first (i) updated the European eel transcriptome using deep sequencing data with a total of 640,040 reads assembled into 44,896 contigs (Eeelbase release 2.0), and (ii) developed a transcriptomic platform for global gene expression profiling in the critically endangered European eel of about 15,000 annotated contigs, which was applied to detect differentially expressed genes between polluted sites. Several detoxification genes related to metabolism of pollutants were upregulated in the highly polluted site, including genes that take part in phase I of the xenobiotic metabolism (CYP3A), phase II (glutathione-S-transferase) and oxidative stress (glutathione peroxidase). In addition, key genes in the mitochondrial respiratory chain and oxidative phosphorylation were down-regulated at the Tiber site relative to the Bolsena site.

Conclusions: Together with the induced high expression of detoxification genes, the suggested lowered expression of genes supposedly involved in metabolism suggests that pollution may also be associated with decreased respiratory and energy production.

Figure 1

Relative abundance of European eel BLAST hits in main taxonomic groups using a simplified Tree of Life diagram.

Figure 2

Functional annotation of European eel contigs associated to GO terms. GO classes were grouped into 112 GO-slim terms for biological process (BP), molecular function (MF) and cellular component (CC) ontologies.

Figure 3

Pearson-correlation based heatmap using the 2,000 genes with higher expression and variability across samples. Individuals are colour-coded according to sex (pink, females; blue, males) and pollution (green, low; orange, high).

Figure 4

Mapping of zebrafish genes homologous to European eel transcripts to oxidative phosphorylation. Green boxes represent KEGG nodes specific to the considered organism. Red stars indicate enriched nodes, which may represent one or more genes.

Figure 5

Mapping of zebrafish genes homologous to European eel transcripts to gap junction. Green boxes represent KEGG nodes specific to the considered organism. Red stars indicate enriched nodes, which may represent one or more genes.

Figure 6

Mapping of zebrafish genes homologous to European eel transcripts to RNA polymerase. Green boxes represent KEGG nodes specific to the considered organism. Red stars indicate enriched nodes, which may represent one or more genes.

Figure 7

Alignment of the amino acid sequence of European eel CYP3A and other vertebrate CYP3A members with high amino acid sequence similarity using the CLUSTAL_W method. Amino acids shared by all species are indicated by a stark mark and amino acids shared by all fish species by a dot mark. Putative recognition sites SRS are indicated by light shade. A gene tree was constructed using the Neighbour-Joining tree method in MEGA5. Genbank accession numbers for the sequences used are as follows: zebrafish CYP3A65, AY452279; killifish CYP3A330, AF105068; medaka CYP3A40, AF251271; rainbow trout CYP3A27, U96077; human CYP3A4, NP059488.