An intriguing relationship between teleost Rex3 retroelement and environmental temperature

Abstract

The movement and accumulation of transposable elements (TEs) exert a great influence on the host genome, e.g. determining architecture and genome size, providing a substrate for homologous recombination and DNA rearrangements. TEs are also known to be responsive and susceptible to environmental changes. However, the correlation between environmental conditions and the sequence evolution of TEs is still an unexplored field of research. Among vertebrates, teleosts represent a successful group of animals adapted to a wide range of different environments and their genome is constituted by a rich repertoire of TEs. The Rex3 retroelement is a lineage-specific non-LTR retrotransposon and thus represents a valid candidate for performing comparative sequence analyses between species adapted to diverse temperature conditions. Partial reverse transcriptase sequences of the Rex3 retroelement belonging to 39 species of teleosts were investigated through phylogenetic analysis to evaluate whether the species' adaptation to different environments led to the evolution of different Rex3 temperature-related variants. Our findings highlight an intriguing behaviour of the analysed sequences, showing clustering of Rex3 sequences isolated from species living in cold waters (Arctic and Antarctic regions and cold waters of temperate regions) compared with those isolated from species living in warm waters. This is the first evidence to our knowledge of a correlation between environmental temperature and Rex3 retroelement evolution.

1. Introduction

Repetitive DNA is represented by sequences repeated up to 1000 times in eukaryotic genomes and in many cases involves more than half of the whole DNA content in the cell nucleus [1,2]. Repetitive DNA includes satellite DNAs made up of tandemly arranged sequences and transposable elements (TEs) constituted by interspersed sequences. TEs are genetic elements that are able to proliferate and insert themselves in novel locations of the host genome. They are classified as retrotransposons (Class I) or DNA transposons (Class II) on the basis of their transposition mechanism. The genomes of vertebrates have been shown to be significantly rich in TEs that affect genome size and architecture. TEs are also known to be highly responsive and susceptible to environmental changes [3–5]. However, the correlation between environmental conditions and the sequence evolution of TEs is still an unexplored and fascinating field of research.

Among vertebrates, teleosts represent a successful group of animals adapted to a wide range of different environments and their genome is characterized by a highly diverse TE content between lineages and even between species [6,7]. In particular, Rex3 is a teleost lineage-specific non-long terminal repeat (non-LTR) retrotransposon with a full length of about 2000 bp identified for the first time by Volff and co-workers [8] in the fish model Xiphophorus maculatus. Its essential features are the presence of a reverse transcriptase (RT) domain, no LTR flanking regions, a 3′-end consisting of GAA repeats followed by GATC tandem repeats, the latter considered to be the signature feature for the Rex3 element [8]. Most of the studies performed on these retroelements have investigated the location of Rex elements on fish chromosomes with the primary aim to clarify the role of this TE family in the diversification of teleost genomes [9]. Undoubtedly, the great number of results achieved in the last two decades clearly suggests their key role in karyotype evolution, especially in chromatin organization, in sex chromosome differentiation and in the formation of supernumerary chromosomes [9,10].

In this context, Rex3, for its aforementioned features and its wide distribution in teleosts, represents a valid candidate on which to perform comparative sequence analyses between species adapted to diverse temperature conditions.

Therefore, the aim of this work is to evaluate whether the species' adaptation to different environments led to the evolution of different Rex3 temperature-related variants. To that end, partial RT sequences of Rex3 retroelements belonging to 39 species of teleosts were investigated through phylogenetic analysis. The findings obtained highlighted an intriguing behaviour of the analysed sequences, showing clustering of Rex3 retroelements isolated from species living in cold waters (Arctic and Antarctic regions and cold waters of temperate regions) compared with those isolated from species living in warm waters. This strongly supports the hypothesis of a putative correlation between environmental temperature and Rex3 retroelement evolution.

2. Material and methods

Eleven Arctic species were collected during an expedition in 2010 (Careproctus reinhardti, Sebastes mentella, Liparis fabricii, Leptagonus decagonus, Hippoglossoides platessoides, Leptoclinus maculatus, Lycodes frigidus, Lycodes reticulatus, Lycodes eudiplerostictus, Lycodes polaris and Lycodes pallidus) and an Antarctic species (Trematomus bernacchii) was collected from Tethys Bay in 2013–2014 during the 29th expedition. Chelidonichthys lucerna was collected from the Adriatic Sea for comparison with polar Perciformes. A specimen of Gasterosteus aculeatus was sampled from Posta Fibreno lake, located in Central Italy and characterized by cold waters, in order to make a sequence comparison with those deposited in GenBank (figure 1).

Figure 1.

Geographical distribution of specimens collected for this study. (Online version in colour.)

Total RNA was isolated from liver tissue with TRIzol reagent (Invitrogen, Carlsbad, CA). First-strand cDNA was obtained with reverse transcription using SuperScript™ III First-Strand Synthesis SuperMix (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. cDNA was amplified using Platinum Taq DNA Polymerase (Invitrogen, Carlsbad, CA) under the following reaction conditions: 30 cycles with 1 min at 94°C, 1 min at 52°C and 1 min at 72°C. A partial RT-encoding region of about 460 nt in length for Rex3 was amplified using primers RTX3-F3 (CGGTGAYAAAGGGCAGCCCTG) and RTX3-R3 (TGGCAGACNGGGGTGGTGGT) [8]. The PCR products were sequenced, and features related to Rex3 were searched in REPBASE using CENSOR (https://www.girinst.org/censor/).

The obtained sequences were aligned with another 64 Rex3 sequences retrieved from NCBI and Ensembl (electronic supplementary material, file S1). For some species, more sequences (electronic supplementary material, table S1) were analysed to evaluate an intraspecific divergence. The alignment was performed with MAFFT (https://www.ebi.ac.uk/Tools/msa/mafft/) using default parameters. The phylogenetic analysis was carried out with MrBayes-3.2 [11]. ModelTest v. 3.7 (Akaike information criteria, AIC) was employed to determine the best-fit model of DNA substitution: TVM + G [12]. The analysis was performed using all parameter values provided by ModelTest (gamma distribution shape parameter = 0.90; substitution model: rate matrix A–C 1.21, A–G 5.21, A–T 1.37, C–G 0.51, C–T 5.22, G–T 1.00, base frequencies A 0.22, G 0.24, C 0.27, T 0.27). The Markov chain Monte Carlo was run for 1 000 000 generations, sampling every 100 generations (burn-in = 25%). Stationarity was defined as when the standard deviation of split frequencies reached 0.0084. The phylogenetic tree includes a total of 78 sequences belonging to 39 teleost species. GenBank accession numbers for all sequences used in the analysis are listed in electronic supplementary material, table S1. Moreover, the search of multiple sequence variants was performed in the genomes available in the Ensembl database for G. aculeatus and Oryzias latipes by BLAST analysis. The obtained sequences were downloaded and aligned with those used in the phylogenetic analysis. The sequence divergence was obtained through the distance matrix calculated with PAUP [13]. Finally, variance partitioning analysis was performed using vegan package 2.5.5 [14] to evaluate possible horizontal transfer events (for details, see electronic supplementary material, file S2).

3. Results

An extended phylogenetic analysis was performed on 78 nucleotide sequences related to the partial RT-encoding region of Rex3 belonging to 39 teleost species. The analysed region of about 460 bp (RT full-length about 1100 bp) was chosen following the paper published by Ozouf-Costaz and colleagues [15]. The overall average distance between the analysed sequences is 19.4%. The tree topology shows two main clades supported by a posterior probability value of 1.00: clade A, which includes groups 1 and 2, corresponding to 31 sequences belonging to species adapted to cold waters (Arctic and Antarctic regions and cold waters of temperate regions) and clade B, which includes groups 3 and 4, corresponding to 47 sequences belonging to species adapted to warm waters typical of subtropical, tropical and temperate regions (figure 2). Group 1 is made up of sequences of species belonging to the orders Perciformes, Gadiformes and Pleuronectiformes while group 2 includes sequences of species belonging only to Perciformes. Regarding clade B, the sequences of species belonging to Cichliformes form a monophyletic clade corresponding to group 3, while group 4 comprises sequences belonging to species of various orders: Tetraodontiformes, Cyprinodontiformes, Characiformes, Centrarchiformes, Esociformes, Cypriniformes, Perciformes and Beloniformes. Overall, the phylogenetic analysis highlighted a non-monophyletic distribution of sequences belonging to Perciformes as well as those belonging to Cypriniformes and Tetraodontiformes.

Figure 2.

On the upper right, phylogenetic analysis of partial reverse transcriptase-encoding region for Rex3 retroelement. Species adapted to cold regions are grouped by light blue dashed box. Coloured vertical bars indicate teleost orders. On the lower left, current cladogram reporting the evolutionary relationships between orders of teleost species investigated in this study. (Online version in colour.)

As regards the intraspecific divergence, the multiple sequences analysed for some species were found to be either monophyletic or paraphyletic, with the exception of Tetraodon nigroviridis, whose sequences are polyphyletic but all confined within group 4. Moreover, in order to corroborate these findings, an extensive search for the presence of multiple sequence variants of Rex3 was performed in the whole sequenced genomes of G. aculeatus and O. latipes, chosen as representatives for the cold and warm waters, respectively. To that end, a Rex3 sequence of stickleback (G. aculeatus) was used to perform a BLAST analysis on the medaka (O. latipes) genome. The Rex3 sequences retrieved from the O. latipes genome were closely related to those of the same species reported in our phylogenetic analysis. The same procedure was performed on the stickleback genome using a Rex3 sequence of medaka, obtaining similar results.

To evaluate a possible role of horizontal transfer events in the lack of congruence between teleosts and Rex phylogeny, a variance partitioning analysis was performed. For each pair of specimens, both sequence divergence and sequence topological position in the phylogenetic tree have been compared with the data related to water temperature (cold or warm waters) and to physical distance. Results of this analysis do not seem to support the hypothesis that phylogenetic tree topology may be due to horizontal transfer of Rex3 retroelements (electronic supplementary material, figure S1).

4. Discussion

Teleosts are a successful group of animals adapted to a wide range of different environments. Their genomes are characterized by the presence of a rich repertoire of mobile elements representing a high fraction of their DNA content. Several papers have underlined the key role of TEs in teleost evolutionary success [1,7,16,17]. TEs are known to be a powerful evolutionary tool, able to generate raw genetic material for adaptation and innovation [7,16,17], as well as being responsive and susceptible to environmental changes [3,4,5]. Although most studies focus on the activity of TEs in relation to environmental conditions [4,5,18], the correlation between environmental temperature and the sequence evolution of TEs had yet to be explored. In this context, we chose to investigate the Rex3 non-LTR retrotransposon, which is a mobile element widely distributed in teleost genomes.

In this work, an extensive phylogenetic analysis based on the partial RT-encoding region of Rex3 was performed. In particular, 78 sequences of 39 teleost species were included (figure 2). The tree topology showed that Rex3 phylogeny does not reflect species phylogeny (figure 2; see also electronic supplementary material, figure S2). The current teleost phylogeny based on the molecular and genomic data of 2000 fish species has shown that Perciformes are closely related to Centrarchiformes and Tetraodontiformes and that there is also a close relationship between Cichliformes, Beloniformes and Cyprinodontiformes as well as between Cypriniformes and Characiformes [19] (figure 2). On the contrary, our results highlighted an intriguing behaviour of the analysed sequences showing clustering of Rex3 sequences isolated from cold water species compared with those isolated from warm water species (figure 2). This allowed us to strongly support a correlation between Rex3 and environmental temperature. Clade A clearly shows clustering of sequences belonging to species living in polar regions, together with those of Battrachocottus baikalensis, a species living in the Russian Lake Baikal, and those of G. aculeatus, a species having a circumarctic distribution. The position of the Rex3 sequence obtained from a specimen of G. aculeatus sampled in Posta Fibreno lake located in Central Italy is interesting. The waters of Posta Fibreno lake come from the mountains in the Abruzzo National Park and, after travelling for a long distance in underground layers, flow into the lake via numerous springs characterized by cold waters. Our phylogenetic analysis placed this sequence in clade A close to those of the same species obtained from nucleotide databases and to those of other Perciformes living in polar regions. This finding could be explained as a result of phyletic constraints or environmental adaptation. To test this hypothesis, a Rex3 sequence was isolated from C. lucerna, another Perciformes collected from the Adriatic Sea and thus adapted to temperate waters. The tree topology shows that this sequence lies in clade B, closely related to sequences of species belonging to Beloniformes, Cypriniformes, Esociformes and Centrarchiformes and one Tetraodontiformes sequence. This finding clearly allowed us to exclude the involvement of phyletic constraints responsible for the position of the G. aculeatus Rex3 sequence from Posta Fibreno in the tree and supports the assumption that specific variants of Rex3 have been selected in the analysed species in relation to environmental temperatures.

The choice of a particular variant of the retroelement in organisms living at a given temperature might be due to natural selection, and thus the presence of a common sequence variant in species adapted to live in environments characterized by similar features might represent an event of evolutionary convergence of Rex3 retroelements. However, it is possible that a specific environmental temperature might have activated transposition and consequently determined the amplification of a particular variant of the Rex3 retroelement. This implicates the possibility that this variant was already present in the common ancestor. The relationship between temperature and the presence of a specific sequence variant in species adapted to cold environments has also been reported for satellite DNAs [20,21]. To our knowledge, this is the first time that a correlation has been shown between the selection of a specific TE variant and environmental temperature. Although the expression levels and functionality of the Rex3 retroelement have not been investigated in this work, it has already been demonstrated that transposition can be activated through temperature-dependent mechanisms [4,18,22] and that TEs can regulate nearby genes in response to abiotic conditions [5]. As the sequence analysed here is related to RT, it is possible that the sequence variant selected leads to an optimal functionality of this enzyme at low temperatures.

The lack of congruence between teleost and Rex phylogenies could be explained as a result of horizontal transfer events that occurred during evolutionary history of teleost species. Indeed, pathogens, parasites and other vectors could have favoured the regional homogenization of Rex retroelements in species sharing the same habitat but phylogenetically distant [15,23–26]. However, the results here obtained suggest that the evolutionary pattern of Rex3 retroelements has probably been influenced by an adaptive selection related to abiotic factors.

In conclusion, our results represent a starting point for further experiment and functional analyses that will allow a better understanding of the correlation between repetitive elements and environmental conditions.

Data accessibility

The sequences have been deposited in GenBank under the accession numbers (MK183734–MK183747).

Authors' contributions

M.B. and A.C. conceived the study. M.F. and F.C. performed the analysis, with substantial assistance from M.B., M.A.B. and A.C. M.A.B. performed the phylogenetic and bioinformatic analyses. The paper was written with the input from all authors. All authors contributed to revising the manuscript and approved the final manuscript. All authors agree to be accountable for the content of this work.

Competing interests

The authors declare no competing interests.

Funding

This work was supported by a grant from Ministero della Ricerca e dell'Istruzione, project nos: 468-2014 and 491-2015.