During an investigation of amoeba rDNA sequences, a group I intron was identified in the 18S rDNA of Thecamoeba aesculea (Amoebozoa; Discosea) at position S956. Secondary structure of Thecamoeba intron and phylogenetic analysis suggest that it is affiliated with other S956 introns occurring in distantly related amoebae such as acrasid (Percolozoa; Heterolobosea) and myxogastrid (Amoebozoa; Mycetozoa) slime moulds. This is the first report of a nuclear intron occurring within the genus Thecamoeba.
Group I introns are mobile genetic elements that spread horizontally and/or vertically within the genomes of various organisms. They are self-splicing RNAs (ribozymes) that can also encode homing endonucleases allowing them to move to the DNA level [1,2]. In eukaryotes, group I introns occur mainly in the plastid and mitochondrial genomes, while their presence in nuclear genes is restricted to those of the ribosomal RNA operon in a wide and scattered set of microbial eukaryotes [1-3]. Group I introns have little sequence similarity but fold into conserved structures usually formed of nine paired elements (P1 to P9) and various loops. The overall structure and sequence variation in the conserved core (P3 to P8) allows their subdivision into five classes IA to IE [4,5]. Most nuclear introns are of IC and IE types, and they are particularly studied in fungi, various algae or plasmodial slime moulds (myxogastrid mycetozoans) because of their high frequency and variety [6-9]. On the other hand, they seem to occur more rarely in other protistan groups.
By studying the free-living amoeba Acanthamoeba (Amoebozoa; Discosea, Centramoebia), nuclear IC introns were found in the 18S rDNA of strains belonging to four distinct genotypes: T3 (A. griffini), T4A (A. castellanii complex), T5 (A. lenticulata), and T15 (A. jacobsi) [10-14]. During our studies on Acanthamoeba introns, an investigation of the available rDNA sequences of amoebae that may contain introns was conducted, based on data from the literature and BLAST analyses. Herein is reported the identification and characterization of a previously unnoticed intron in the 18S rDNA of Thecamoeba aesculea (Amoebozoa; Discosea, Flabellinia), a terrestrial amoeba isolated from epiphytic mosses and bark of the European horse-chestnut, Aesculus hippocastanum (Angiospermae, Sapindales) [15].
The intron characterized here is in the 18S rDNA of Thecamoeba aesculea (CCAP 1583/12) (GenBank ID JN247436, mislabelled as a mitochondrial sequence) [15,16]. Its identification was made during a wide screening of the available sequences to specifically detect nuclear introns. Once the intron was delimited, its secondary structure was deduced manually by identifying the conserved elements, therefore using Mfold for the remaining portions of the molecule, as described previously [13]. The phylogenetic relationships of the intron were deduced by analysing the extended conserved core (P3 to P8, P1’ and proximal sites of P2) [17] of selected nuclear introns of IC and IE classes. The sequences were manually aligned according to the secondary structure of the introns and phylogenetic trees were built with Maximum Likelihood (ML) (GTR+Γ model; 1000 replicates), Neighbour Joining (NJ) (Jukes-Cantor model; 2000 replicates) and maximum parsimony (MP; 1000 replicates), as described previously [13,14,17].
To verify whether the intron could have influenced the 18S rDNA-based phylogeny of the line, ML analysis (GTR+Γ+I:4; 1000 replicates) was performed on complete and partial (>1300 bp) sequences of Thecamoebidae (Sappinia, Stenamoeba, Thecamoeba, and Thecochaos), and the intron-free sequence of T. aesculea, using Dermamoebidae (Dermamoeba algensis and Paradermamoeba levis) as outgroup, previously aligned with MAFFT and refined for ambiguous sites using BioEdit, as described previously [18]. The same was done in parallel but with the intron-containing sequence of Thecamoeba. Both datasets were also analysed after realignment with MUSCLE.
The 18S rDNA sequence of T. aesculea deposited in GenBank is 2224 nucleotides, a fairly usual length for Discosea whose members often have expansion regions in their 18S rDNA. The present analysis revealed that an intron of 432 nucleotides is actually inserted at position S956, in the loop between helices 34 and 35. As a result, the available rRNA of T. aesculea is only 1792 nt, ending with an incomplete H43-1 region. The presence of the Thecamoeba intron is not indicated in the GenBank file, nor is it reported in any publication on these amoebae. No introns were found in the more than 50 additional 18S rDNA sequences from Thecamoebidae analysed here.
The secondary structure (Figure 1) and signature sequences indicate that the Thecamoeba intron belongs to the IC class. Phylogenetic analysis (Figure 2) suggests that it is close to other introns with the same insertion site, S956, occurring in distant organisms such as arcellinid testate amoebae (Hyaloshaenia, Nebela), Acrasis (Percolozoa; Heterolobosea, Acrasida), various myxogastrids (Amoebozoa; Mycetozoa) and Morellospora (Rozellomycota; Morellosporales). All have a similar secondary structure and size (~ 400 nt), except for the Didymium intron which encodes an endonuclease, and share 60 - 70% similarity with the Thecamoeba intron. It might be interesting to note that the intron-bearing Thecamoeba strain was isolated from tree bark [15], where acrasids and myxogastrids also live, and that Morellospora is a microsporidium-like intracellular parasite of amoebae, including slime moulds [19]. Close host-host interactions, such as those occurring during predation or parasitism, can facilitate lateral gene transfer. One could therefore suppose that these introns, belonging to the same lineage, can move horizontally between hosts sharing similar microhabitats and capable of interacting in various ways.
In 18S rDNA phylogeny based on intron-less sequences, T. aesculea emerges basal within the T. quadrilineata group (Figure 3), which also includes T. cosmophorea and T. astrologa, consistent with previous studies [20,21].
It should be noted that the presence of an intron in this sequence being missing, it could lead to misalignments. Introns insert at specific sites in conserved regions, normally causing large gaps in multiple alignments, making them easily recognizable. However, the same sequence may be treated differently by separate programs, and/or depending on the number and variety of other sequences analysed. In the special case of the Thecamoeba intron analysed here, it is noteworthy that the first eight nucleotides of the intron at its 5' end (P1 loop) are identical to the portion of the 18S rDNA located just after the S956 insertion site, on the 5' side of H35. In this way, the intron would not be recognized as a foreign element, instead appearing from its initial part as homologous to H35, thus inducing a misalignment, and its most divergent parts can be visualized as hypervariable regions or expansions of a few tens of not over the entire H34 to H44 portion of other Thecamoeba species. Such misalignment has indeed been produced in some cases, especially when few sequences from many distinct taxa are included. These regions will be removed during a phylogenetic analysis, thus having little effect on the result. Indeed, a virtually identical 18S tree topology can be obtained, differing for T. aesculea only in its branch length.
This is the first characterization of a group I intron present in the nuclear rDNA of a member of the genus Thecamoeba. Beyond Acanthamoeba spp., it is the only other member of the Discosea within which group I nuclear introns have been identified and characterized.
SignUp to our
Content alerts.
Are you the author of a recent Preprint? We invite you to submit your manuscript for peer-reviewed publication in our open access journal.
Benefit from fast review, global visibility, and exclusive APC discounts.