Functional validation of transposable element-derived cis-regulatory elements in Atlantic salmon

Abstract

Transposable elements (TEs) are hypothesized to play important roles in shaping genome evolution following whole-genome duplications (WGDs), including rewiring of gene regulation. In a recent analysis, duplicate gene copies that had evolved higher expression in liver following the salmonid WGD ∼100 million years ago were associated with higher numbers of predicted TE-derived cis-regulatory elements (TE-CREs). Yet, the ability of these TE-CREs to recruit transcription factors (TFs) in vivo and impact gene expression remains unknown. Here, we evaluated the gene-regulatory functions of 11 TEs using luciferase promoter reporter assays in Atlantic salmon (Salmo salar) primary liver cells. Canonical Tc1-Mariner elements from intronic regions showed no or small repressive effects on transcription. However, other TE-CREs upstream of transcriptional start sites increased expression significantly. Our results question the hypothesis that TEs in the Tc1-Mariner superfamily, which were extremely active following WGD in salmonids, had a major impact on regulatory rewiring of gene duplicates, but highlights the potential of other TEs in post-WGD rewiring of gene regulation in the Atlantic salmon genome.

Keywords: Atlantic salmon; evolution; gene expression; transposable elements; whole-genome duplication.

Fig. 1.

Overview of functionally validated Tc1-Mariner elements. a) Overview of transposable element insertion relative to gene features and predicted TFBS bound by liver-expressed TFs. Jaspar TF motif names are indicated above each TE insertion. b) Within TE-family similarity distribution reflecting age of transposition activity. TE insertions in (a) marked with a thin horizontal line. Thick line marks expected genomic similarity for duplicated genomic regions after WGD.

Fig. 2.

Dual luciferase reporter assays using Tc1-Mariner TE regions. Barplots showing log₂ transformed mean fold change relative to sv40 empty vector across 3 experiments (a–c) with 3 replicates per experiment. a) Reporter assays testing whole TEs. b) Reporter assay testing accessible chromatin regions overlapping the TEs. c) Reporter assay testing regions of the TEs in open chromatin and predicted to be bound by liver-specific TFs. Positive controls are the CMV enhancer and the ATAC-peak closest (upstream) to the transcription start site (TSS) of the liver-expressed gene elovl5b (NC_059469.1; ssa28: 27245480..27256837, Assembly ICSASG_v2, Gene ID: 100192340) from Atlantic salmon.

Fig. 3.

Overview of functionally validated non-Mariner elements. a) Overview of transposable element insertion relative to gene features and predicted TFBS bound by liver-expressed TFs. Jaspar TF motif names are indicated above each TE insertion. b) TE copy similarity distribution for each TE consensus sequence reflecting age of transposition activity. TE insertions in Fig. 3a marked with a thin line. Thick line marks expected genomic similarity for duplicated genomic regions after WGD.

Fig. 4.

Dual luciferase assays on non-Mariner TE-CREs. Barplots showing log₂ transformed mean fold change relative to sv40 empty vector across 3 replicate experiments. Mean values of 3 biological replicates within each replicate experiment are indicated with points.