Letter to the Editor: ASTER-REP, a Database of Asteraceae Sequences for Structural and Functional Studies of Transposable Elements

Le résumé fourni par la source

Transposable elements (TEs) are interspersed repetitive DNA sequences that can move independently within the genome through specific transposition mechanisms. In eukaryotes, TEs are divided into two classes: class I TEs (retrotransposons, REs), which use an RNA intermediate for the transposition, and class II TEs (DNA transposons), which move through DNA excision (Wicker et al. 2007). These classes are grouped into orders and lineages, according to sequence homology and to the ability to encode their own transposition machinery or not. Traditionally, TEs have been poorly studied also because their identification has been challenging; nevertheless, these sequences have tremendous implications for genome stability and function: first, TEs drive structural variation, as their transposition can cause deletions, inversions (i.e., homologous recombination of TEs with an opposite orientation) and duplications (Cordaux and Batzer 2009), making TEs activity strongly associated with dynamics of genome reduction/expansion. Second, TEs may drastically affect the expression of other genes (Lisch 2013, Fambrini et al. 2018), either by disrupting the coding sequence or by altering the gene regulation when transposing into the upstream region with different effects, such as the disruption of regulatory sites, epigenetic silencing of neighboring regions or providing new regulatory elements (Morgante et al. 2007, Slotkin and Martienssen 2007). Finally, TEs can originate novel genes through exaptation mediating neofunctionalization with a selective advantage to the host (Ventimiglia et al. 2022). For these reasons, the detection of such elements is now considered an essential step to achieve high-quality genomes, more so for eukaryotes since TEs represent a considerable fraction of their genomes. However, the accurate identification of TEs remains a challenging and largely manual endeavor due to the diversity in structures and sequences, which greatly vary across species. There are three strategies for repetitive sequences annotation: homology-based, ab initio, and structural-based prediction (Bergman and Quesneville 2007, Saha et al. 2008). Of these, the homology-based one is the most commonly adopted approach, although it requires libraries of well-curated, representative repeats from related organisms, which might be lacking. Here, we present ASTER-REP, a comprehensive TEs database of full-length sequences belonging to species of the Asteraceae family that represents a novel resource to facilitate the homology-based annotation of TEs. We focused on the family Asteraceae for different reasons: it is an enormous clade, comprising 8% of all plant species distributed across the globe, including economically important species, such as sunflower artichoke and lettuce. On average, the genomes of Asteraceae display a high proportion of TEs (70%), representing an excellent system to study plant genome evolution (Staton and Burke 2015). However, only the sunflower, Helianthus annuus, has been deeply investigated focusing on the TE content (Natali et al. 2013, Mascagni et al. 2015, 2020), revealing to have a large proportion of repeated sequences accounting for more than 80% of the genome itself. The principal aim of this database is to collect and carefully annotate full-length TEs producing a useful tool for starting studies on transposon diversity and dynamics in this important plant family. Furthermore, given its size, detailed annotation, and user-friendly implementation, ASTER-REP will be a useful platform to study transposon variability, helping to unravel the genome structure and improving transposon annotation for other plant genomes. The ASTER-REP database (accessible at https://aster-rep.agr.unipi.it) contains original data of interspersed repeat sequences, organized with hierarchical criteria (Supplementary Fig. S1). De novo interspersed repeat libraries were developed to identify TEs using structural-based methods on sequence assemblies of six Asteraceae species. In particular, H. annuus, Lactuca sativa, and Cynara cardunculus var. scolymus, assembled at chromosome level, and Artemisia annua, Carthamus tinctorius, and Chrysanthemum seticuspe, assembled at contig level, were chosen (see also Supplementary Table S1). All TEs were retrieved using structure-based software, by searching for their characteristic features: i.e., high copy number, specific domains, and typical sequence motifs (see Supplementary Data). The approach used to identify and collect Asteraceae TEs has been consistently applied in all six species, harmonizing the resulting datasets. One feature that differentiates ASTER-REP from other existing TE databases is that it is composed of full-length elements. Additional efforts were put into cross-validating all libraries belonging to different TE orders to reduce the misclassification of individual entries or the presence of any other sequence differing from the one reported in the entry (e.g., captured gene fragments or nested TEs). A total of 328,696 unique full-length TEs are presently included in the database (Supplementary Table S2). The interface of ASTER-REP website allows users to search, browse, display, and retrieve sequences (Fig. 1). Users can obtain the desired data with automatic search options by selecting checkboxes for species, whereas TE class, order, superfamily and lineage are selectable from drop-down menus. ASTER-REP database search function. Checkboxes and drop-down menus permit the customization of search parameters (A), and the resulting elements are listed and visualized (B). Search results can be downloaded locally using the temporary links (C). The hierarchical structure of sequence classification was exploited to allow a straightforward selection of the desired repeat pool with a single query (i.e., selecting class I, only short interspersed nuclear elements, and/or long terminal repeat elements REs will be selectable). Selected entries can then be visualized and downloaded locally. Finally, we consider the sequence similarity–based access to TE data among the most useful aspects of ASTER-REP, for which we integrated a Basic Local Alignment Search Tool service linked to the database sequences. Users can exploit this tool for analyzing their own sequences and/or for the study of transposons in species not belonging to the Asteraceae family. Actually, beacuse of the general conservation of sequence characteristics among plant TEs, the data collected in ASTER-REP can be used as a template to produce new and more customized datasets. Considering its size and annotation level, ASTER-REP collection represents an important resource for repeat annotation of plant genomes and a step forward compared to unspecific sequence databases. Furthermore, the user-friendly database we implemented will facilitate the exploitation of these sequences, also providing a stable platform to be periodically updated through the integration of newly identified repeat elements from other Asteraceae genomes. Supplementary data are available at PCP online. The data underlying this article are available in ASTER-REP database at https://aster-rep.agr.unipi.it. This work was supported by University of Pisa, Project Plantomics (grant number 569999_2017). The authors have no conflicts of interest to declare.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.