Menu

Close

LINE elements are a reservoir of regulatory potential in mammalian genomes

Maša Roller, Ericca Stamper, Diego Villar, Osagie Izuogu, Fergal Martin, Aisling Redmond, Raghavendra Ramachanderan, Louise Harewood, Duncan T. Odom, Paul Flicek

Preprint posted on May 31, 2020 https://www.biorxiv.org/content/10.1101/2020.05.31.126169v1

Dynamic turnover of tissue-specific and tissue-shared regulatory elements during mammalian evolution

Selected by Sergio Menchero

In an organism, different gene expression programs are activated in specific tissues and cell types to ensure particular developmental features and cellular functions. The control of those programs is driven by regulatory elements that enhance, repress, isolate or maintain gene expression in a spatial and temporal manner. Although many genes and specific expression programs are widely conserved among different phyla, the regulatory elements controlling those genes seem to have evolved more quickly (Villar et al. 2015). In this work, the authors characterise regulatory elements in four tissues of ten mammals that cover 160 million years of divergence in order to understand the principles underlying the evolution of tissue-specific and tissue-shared regulatory domains.

Scientists have classified regulatory elements according to their activity based on the presence of specific histone modifications. However, the same regulatory domain can have different activities depending on the cell type, the developmental timing and the species (Dao et al. 2017, Carelli et al. 2018). By means of ChIP-seq, the authors mapped three histone modifications whose combination is associated with specific regulatory activity. They focused on the identification of active promoters (characterised by the presence of H3K4me3 and H3K27ac), active enhancers (H3K4me1 and H3K27ac) and primed enhancers (H3K4me1) in four adult tissues: liver, muscle, brain and testis. In parallel, they performed RNA-seq in the same tissues to match transcriptional activity to the presence of the regulatory elements.

In order to understand the evolutionary stability of the regulatory regions, the authors quantified their evolutionary rates by calculating the fraction of enhancers and promoters that were maintained between pairs of species. The authors found that tissue-shared regulatory regions were more conserved (lower evolutionary rate) than their tissue-specific counterparts. Comparing the different types of regulatory regions, the authors demonstrated that primed enhancers have the highest evolutionary rate while active promoters are the most conserved (Figure 1). Although that is the general trend, there are some differences among tissues. For instance, in testis, active promoters have a higher evolutionary rate than enhancers. Also, brain-specific regulatory regions evolved the most slowly. Remarkably, these particular features correlate with previously reported gene expression evolutionary rates in which small changes in the brain and accelerated changes in testis were described (Cardoso-Moreira et al. 2019).

Figure 1. Tissue-specific regulatory regions have higher evolutionary turnover than tissue-shared regions. Evolutionary rates of alignable tissue-shared and tissue-specific regulatory regions estimated by linear regression of activity maintenance between all pairs of species. [Adapted from Figure 3 in this preprint, made available under a CC-BY 4.0 license].

 

The authors then focused on the evolutionary turnover of the identified regulatory regions. They defined intra-species dynamic regulatory regions as the regions that have different histone modification signatures across tissues of the same species, and between-species dynamic regulatory regions when the changes in histone modification signatures occurred between different species (in pairwise comparisons). Intra-species dynamic regulatory regions were relatively rare (7-11% of regions were characterised), and the ones identified were not maintained as dynamic regions in other species. In contrast, promoter-enhancer switching were much more common between species. Particularly high was the turnover between active and primed enhancers (44% of active enhancers aligned to a prime enhancer in another species). Taking into account the evolutionary distance, the authors showed that enhancers are more likely to evolve into active promoters than the reverse (six times more!). Nicely, those enhancer-to-promoter switches were correlated with higher transcriptional activity of the associated gene.

Finally, the authors investigate whether specific classes of transposable elements contribute to the evolution of regulatory elements. They compared the enrichment of annotated transposable elements in the identified tissue-shared and tissue-specific regulatory elements. Interestingly, tissue-specific regulatory elements were enriched in LINE L1 family of transposons while tissue-shared elements were enriched in transposons of the LINE L2 family (Figure 2). When comparing stable active enhancers versus evolutionary dynamic elements, the authors showed that stable enhancers were enriched in the LINE L1 family while dynamic elements were enriched with in LINE L2 family of transposons. This suggests that LINE L2 transposons may provide a more versatile potential for transcriptional regulation.

Figure 2. Tissue-specific activity using the average ChIP-seq read enrichment for recently-evolved active promoters associated with LINE L1s and L2s and their flanking regions. [Adapted from Figure 5 in this preprint, made available under a CC-BY 4.0 license].

 

Why I chose this preprint and why I think this work moves the field forward:

The control of transcription mediated by the activity of regulatory elements has always been close to my own research directly or indirectly (by the work of some colleagues). The identification of regulatory elements is always complicated and we have to rely on the presence of histone marks or reporter activity, which is never 100% accurate. Despite all the variability and exceptions, researchers have managed to improve our understanding of this complex but robust system of gene regulation.

I find particularly fascinating the evolutionary perspective of the regulatory regions. Beautiful works from past years have shown how enhancers evolved from ancestral DNA exaptation (Villar et al. 2015) or how regulatory landscapes change their complexity during evolution to deal with whole-genome duplication and specialization events (Marlétaz et al. 2018). In this work, the characterisation of regulatory elements across ten different mammals and four different tissues allowed the authors to provide a wide view of the evolution of dynamic regulatory landscapes. Due to the complexity and variability of regulatory elements when studied one by one, it is important to first have a global picture of what are the behaviours that mostly underlie enhancer or promoter evolution. All these recent works, including this preprint, are important landmarks for future studies where more particular cases of gene and regulatory landscape evolution can be addressed.

 

Questions to the authors

  1. The authors identify regulatory elements based on the presence of specific histone marks for all the species. Have the authors studied if the degree of sequence conservation correlates with regulatory elements being more stable or dynamic?

 

  1. The authors have shown how turnover of regulatory elements affected the transcriptional activity of their associated genes. Do the authors think the evolution of genes can also affect the activity of regulatory elements? If a gene is not conserved in a species, due to exon erosion or transposon insertion for instance, can the associated regulatory elements be maintained or would they also disappear?

 

References

Carelli et al. Repurposing of promoters and enhancers during mammalian evolution. Nat Commun 2018 Oct 4;9(1):4066.

Cardoso-Moreira et al. Gene expression across mammalian organ development. Nature 2019 Jul;571(7766):505-509.

Dao et al. Genome-wide characterization of mammalian promoters with distal enhancer functions. Nat Genet 2017 Jul;49(7):1073-1081

Marlétaz et al. Amphioxus functional genomics and the origins of vertebrate gene regulation. Nature 2018 Dec;564(7734):64-70.

Villar et al. Enhancer evolution across 20 mammalian species. Cell 2015 Jan 29;160(3):554-66

 

Tags: enhancer, evolution, mammals, promoter, regulatory element, transposable element

Posted on: 22nd June 2020

doi: https://doi.org/10.1242/prelights.22018

Read preprint (No Ratings Yet)




  • Author's response

    The author team shared

    1. The authors identify regulatory elements based on the presence of specific histone marks for all the species. Have the authors studied if the degree of sequence conservation correlates with regulatory elements being more stable or dynamic?

    From our previous work (Schmidt et al. Science 2010; Ballester et al. eLife 2014), we know that if a transcription factor binding site is shared across more species, it has a higher sequence constraint than those shared across fewer species. For this project, however, we did not directly compare the sequence conservation of stable vs dynamic regulatory regions. This is a great question, and would be something interesting to look into.

     

    1. The authors have shown how turnover of regulatory elements affected the transcriptional activity of their associated genes. Do the authors think the evolution of genes can also affect the activity of regulatory elements? If a gene is not conserved in a species, due to exon erosion or transposon insertion for instance, can the associated regulatory elements be maintained or would they also disappear?

    The evolution of regulatory regions and genes can be highly intertwined processes. Gene evolution can change the evolutionary pressures on the regulatory landscape. If a gene evolves to lose its function in a species, for example, then that would release selective pressure on regulatory regions controlling that gene, which may cause them to disappear more readily or to evolve a different regulatory function. We have observed examples of this process including around the GULO gene, which is required to make vitamin C, but became a pseudogene in the primate ancestor (which is why primates must eat vitamin C, but rodents don’t have to). After pseudogenisation, all the of regulatory regions disappeared around GULO in human and macaque, although they remain in mouse, rat and dog where the gene is viable (data from Ballester et al. eLife 2014, but the GULO result did not make it into the paper). Regardless, this is a great follow-up question to our study, and hopefully our data will help in better understanding these types of processes in genome evolution.

    Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here
    Close