Transposable elements as a potent source of diverse cis-regulatory sequences in mammalian genomes

Abstract

Eukaryotic gene regulation is mediated by cis-regulatory elements, which are embedded within the vast non-coding genomic space and recognized by the transcription factors in a sequence- and context-dependent manner. A large proportion of eukaryotic genomes, including at least half of the human genome, are composed of transposable elements (TEs), which in their ancestral form carried their own cis-regulatory sequences able to exploit the host trans environment to promote TE transcription and facilitate transposition. Although not all present-day TE copies have retained this regulatory function, the preexisting regulatory potential of TEs can provide a rich source of cis-regulatory innovation for the host. Here, we review recent evidence documenting diverse contributions of TE sequences to gene regulation by functioning as enhancers, promoters, silencers and boundary elements. We discuss how TE-derived enhancer sequences can rapidly facilitate changes in existing gene regulatory networks and mediate species- and cell-type-specific regulatory innovations, and we postulate a unique contribution of TEs to species-specific gene expression divergence in pluripotency and early embryogenesis. With advances in genome-wide technologies and analyses, systematic investigation of TEs' cis-regulatory potential is now possible and our understanding of the biological impact of genomic TEs is increasing. This article is part of a discussion meeting issue 'Crossroads between transposons and gene regulation'.

Keywords: enhancers; hourglass model and gene regulation; transcription factor binding; transposons.

Figure 1.

TF binding to TEs. (a) There are various possible outcomes from transposition of the ancestral TE (teal rectangles) that leads to variation in TF-binding motifs (orange motifs) observed in present-day TEs. When ancestral TEs contain functional TF-binding motifs (upper panel), they can spread these motifs across the genome, which might be co-opted and maintained, modified or lost by neutral substitution. Alternatively, ancestral TEs might serve as a substrate for the evolution of new or enhanced TF-binding motifs (lower panel). (b) TF binding is not only dependent on sequence but also on chromatin context. (Clockwise from top-left) TF binding can occur after chromatin re-modelling, through cooperation with another TF, through the binding of a pioneer TF to the nucleosome or through direct binding to a strong motif. (c) Differences in the TFs that bind TEs correspond to developmental stages. Preimplantation embryos express pluripotency TFs that can bind to the ancestral TE and also permit en masse TE entry into the genome. Alternatively, in somatic tissues, TEs might not have somatic TF-binding motifs but could evolve them via neutral substitutions. Ancestral TEs can contain suboptimal TF-binding motifs that become a bona fide binding site for TFs in somatic tissues, through a few nucleotide modifications.

Figure 2.

TE-derived cis-regulatory elements. (a) Here, we review TEs' role in gene expression via enhancers, promoters, boundary elements and silencers. (b) TE-derived enhancers act distally as either intergenic (upper panel) or intronic (lower panel). (c) TEs contribute promoters either as alternative promoters driving chimeric transcripts (upper panel) or as a replacement for the canonical promoter (lower panel). (d) TE-derived boundary elements contribute to topologically associated domains (TADs) by providing CTCF-binding sites (upper panel) and also maintain TADs by TE transcription (lower panel). (e) TEs can also act as a silencer by spreading heterochromatin (upper panel) or stalling Pol II elongation (lower panel).