Analysis of Tc1-Mariner elements in Sclerotinia sclerotiorum suggests recent activity and flexible transposases

Transposable elements (TEs) encompass a wide range of DNA sequences that can move to new sites in the genome. For many years following their discovery in the mid 1940s, TEs were thought to be a genetic rarity and later, pejoratively, as genomic parasites. More recently, a significant role for TEs in genomic evolution has been demonstrated [1]. Transposons are important tools for the evolution of several species because they increase genomic plasticity and diversity [2], modify gene structures [3],[4] and are important sources for regulatory sequences [5],[6].

Transposable elements can be divided into two classes that differ by the presence or absence of an RNA intermediate. Class I elements replicate by a copy-and-paste mechanism involving RNA intermediates that are subsequently reverse transcribed into double-stranded DNA by enzymes that are coded for by the transposable element (TE) itself. Class II elements, or DNA transposons, are divided into two subclasses. Subclass 1 consists of elements that transpose themselves by excision and integration, which results in both DNA strands being cleaved during the excision process. Transposons from subclass 2, on the other hand, duplicate before insertion. Subclass 1 contains two orders, the most widely known being the TIR (Terminal Inverted Repeated) order. This order contains nine superfamilies: Tc1-Mariner, Mutator, hAT, Merlin, Transib, P, PIF/Harbinger, CACTA and PiggyBac. Subclass 2 contains two orders: Helitron and Maverick[7]. Two groups of non-autonomous TEs that lack one or more genes necessary for transposition also exist: MITEs (Miniature Inverted-repeat Terminal Elements), which are categorized as class 2, SINEs, which are members of the non-LTR (Long Terminal Repeat) retrotransposon group, and TRIMs (Terminal-repeat Retrotransposon In Miniature) and LARDs (Large Retrotransposon Derivates), which are in the LTR retrotransposon group [8].

Of the subclass 1 superfamilies, Tc1-Mariner is likely the most prevalent in organisms [7]. Elements in this superfamily are generally between 1,300 and 2,400bp in length and have simple structures containing a single ORF that codes for the transposase protein and is flanked by terminal inverted repeats (TIRs) [9]. The transposase has a conserved, three-amino acid sequence containing two aspartic acid (D) residues and one glutamic acid (E) (DDE). In some cases, a third aspartic acid can be observed (DDD). The catalytic DDE/D motif performs the excision and insertion activities, but it must interact with a divalent cation, usually Mg⁺², to perform the transposition reaction [10]. The transposase also contains helix-turn-helix (HTH) DNA binding motifs that are responsible for recognizing the TIRs [11]. Due to the increasingly rapid availability of genomic sequences, identification of Tc1-Mariner elements and their potential evolutionary impacts have been shown in pathogenic fungi [12],[13].

The most prominent effect of transposons on the genome is the induction of mutations. Because of their mobility and ability to recombine, TEs can interrupt genes or generate several types of rearrangements such as deletions, duplications and inversions. Thus, cells have evolved mechanisms to silence TEs, e.g., silencing by Repeat Induced Point Mutation (RIP). This mechanism was first described in Neurospora crassa, where the introduction of mutations into the DNA of this species was related with the sexual cycle during meiosis. The RIP complex recognizes duplicated sequences that are larger than 400bp and have identity that is greater than 80% and introduces transitions that convert C:G to T:A in both copies [14]-[16]. RIP appears to be widely distributed in ascomycete fungi [17].

The mutagenic activity of TEs can affect genomic sequences, therefore, and they could have potentially negative effects on the fitness of the host. However, mutations caused by transposons play important roles in genomic organization and are, thus, beneficial under some conditions [18],[19]. Substantial evidence has shown that TEs can act as a dynamic reservoir for novel cellular functions, and many endogenous genes have incorporated coding and regulatory sequences from TEs during evolution [20]. Co-opting TEs to perform cellular functions can be considered an exaptation at the molecular level and has been observed in several species [21]. In fact, TEs represent a natural and abundant source of regulatory sequences for host genes [6].

Sclerotinia sclerotiorum is a necrotrophic fungus that is pathogenic to a wide range of species (>400 species) and can persist in the environment for many years due to its ability to produce sclerotia. The S. sclerotiorum genome is estimated to contain 38.3Mb, 7% of which is composed of TEs [22]. Analysis of the genetic diversity of TEs in S. sclerotiorum has suggested that a recent genomic remodeling event occurred that involved dramatic TE expansion [22]. Specifically, Tc1-Mariner elements exist at high copy numbers and show low genetic variability, suggesting recent transposition events in the genome, unlike retroelements, which have a high number of degenerate copies and unpaired LTRs and indicates limited expansion [22]. Due to the importance that the Tc1-Mariner element expansion may have on the organization and evolution of the S. sclerotiorum genome, this study sought to identify and characterize elements belonging to the Tc1-Mariner superfamily and to investigate the possible evolutionary impacts of these elements on the genome of this pathogen.

One hundred and fifty-seven Tc1-Mariner elements were identified, and these included 50 potentially active elements. To our knowledge, this is the largest number of potentially active Tc1-Mariner elements that has currently been found in a fungal genome. This value is highly significant when compared to the potentially active Tc1-Mariner elements in other fungi such as Paracoccideoides[12], Verticillium spp. [13], Mycosphaerella fijiensis[25] and Lacaria bicolor[26]. In addition, the six MITE elements and other copies that have truncated ORFS but contain preserved TIRs can be mobilized in trans by enzymes coded for by an intact copy [7],[27]. DDE, HTH_psq and HTH_Tnp_Tc5 motifs were identified in all of the potentially active copies.

Tc1-Mariner elements may also have three types of functional sequences that are involved in transposition: cleavage sites at the ends of the TIRs that contain 47 nucleotides, UTRs between the TIRs and the ORF that increase transposition efficiency, and DRs (direct repeats) within the TIRS that act as transposase linkage sites [28]. All of the identified elements, except for elements belonging to the TcMar-Pogo family, have cleavage sites at their ends that contain ACGT, as found in elements from the DAHLIAE 1 and 2 families that were identified in Verticillium dahliae[13]. Symmetric and conserved UTR regions were also found in elements from every family; however, DRs in the TIRs were not found. Nevertheless, each end/transposase combination appeared to create subtle versions for mobilization, which guaranteed a certain amount of specificity during transposition [28].

TEs with high sequence identity between S. sclerotiorum and B. cinerea were found. Elements similar to Flipper, which was first identified in B. cinerea[23], were also identified in S. sclerotiorum. This result indicates a possible horizontal transfer of the Flipper between S. sclerotiorum and B. cinerea. Both species are notorious plant necrotrophic fungi and share extensive syntenic blocks [22]. Additionally, the Flipper element is widely used in genetic variability studies [29],[30] and, thus, can be analyzed as a molecular marker in S. sclerotiorum.

Various transposase transcripts of the same family were identified due to intron retention/excision. Introns within class II DNA transposons have been reported in plant pathogens [31],[32], and phylogenetic analysis of the Mariner-2_SS and Mariner-3_SS families demonstrated that elements that required the removal of the intron were ancestral. Therefore, the intron appeared to not be an evolved trait that was important to the element because it allowed the genome some control over transposition due to the dependence of the transposon on the host-splicing mechanism [6]. Conversely, the presence of alternative introns in transposase allows the elements an efficient strategy to overcome possible mutations that generate early stop codons. In addition, the existence of transposase sequences that may or may not maintain the intron in the mature mRNA shows that the transposase for Tc1-Mariner elements has a flexible coding sequence and, consequently, a flexible conformation. This flexibility is likely related to the complex, synaptic organization of transposition (transpososome) [33]-[35]. Consistent with this finding, Nesmelova and Hackett [35] demonstrated that the catalytic domains of DDE-transposases had few similar sequences and significantly different sizes, and they suggested that transposases must be flexible enough to allow conformational rearrangements of their DNA binding domains and to provide a catalytic site for each transposition step.

Analysis of insertion sites showed that class II TEs inserted themselves near the coding sequences of important proteins such as serine threonine kinase, the MFS multidrug transporter and polyprenyl 4-hydroxybenzoate transferase. Serine/threonine kinase is an essential component of several regulatory pathways in fungi, including the mechanism for creation of turgor pressure in the appresorium and pathogenicity [36]. The multidrug transporter protects the organism from toxic products such as fungicides [37], and the enzyme polyprenyl 4-hydroxybenzoate transferase is involved in ubiquitin biosynthesis [38]. TEs near these genes or sequences involved in the same biological processes that these proteins are involved in were also identified in M. fijiensis[25]. Because they are inserted downstream of these genes, the transposons can influence their expression. In fact, a TE sequence downstream and physically near the coding sequence of a NADB-Rossmann motif-containing dehydrogenase gene that is involved in a metabolic pathway such as glycolysis [39] has been shown to be co-expressed with the gene and detected in its EST. Analysis of genes in humans and rats have shown that the 3′ region of genes can be dynamically altered by TEs during evolution, which suggests that TEs can provide alternative polyadenylation sites when inserted downstream of endogenous genes [40]. However, the only analysis of polyadenylation sites in fungi has been performed in Aspergillus oryzae[41]. Therefore, because of the current lack of knowledge about the 3′ gene regulatory regions of fungi, additional studies are necessary to measure the possible involvement of TEs in the evolution of the 3′ ends of genes. Even if these insertions do not have any advantage for the host, they may be fixed in the population by genetic drift because strong evidence supports the idea that transposition is a significant source of exaptation events [6].

Active TEs in the Tc1-Mariner superfamily have been reported in fungi [42],[43]. Here, analyses to detect ORFs and nucleotide diversity have suggested that these elements were recently introduced and are potentially active. However, the presence of transposase sequences in S. sclerotiorum ESTs database only provides the information that TEs are transcribed. So, Western blot analysis for transposases of S. sclerotiorum should be performed in future work to suggest the mobility of Tc1-Mariner elements in this genome. Interestingly, the Mariner-4_SS family includes the largest number of potentially active elements (14). The negative and non-significant Tajimas D test for sequences that code for this familys transposase indicates that selection against genotypes carrying deleterious mutant alleles occurred. However, deviations from the infinite allele model are not only due to natural selection because a population that is growing will also contain an excess of rare alleles.

Despite strong evidence for Tc1-Mariner transposon activity, no evidence for gene-silencing mechanisms similar to RIP was found. Clutterbuck et al. [17] analyzed seven Tc1-Mariner elements in the S. sclerotiorum genome and did not find strong evidence that these copies were affected by RIP; however, they have suggested that gene silencing might be present because CpA and CpG dinucleotides are more commonly mutated than CpY dinucleotides. Here, two indices for detecting RIP-like mutations, TpA/ApT and (CpA + TpG)/(ApC + GpT), were used and indicated that RIP-like mutations were absent from the analyzed sequences. In addition, sequences from transposase coding regions have low nucleotide diversity, meaning few mutations occur between them, and they have high haplotype diversity, which indicates that these mutations are unique and, according to the Tajimas D test, neutral. Therefore, these results do not provide any evidence for RIP silencing in Tc1-Mariner elements. However, the absence of RIP-like mutations in Tc1-Mariner elements does not indicate the absence of the RIP mechanism in the genome of S. sclerotiorum because differences in the intensity with which RIP acts between the different transposable elements, within the same genome, has been reported for various genomes of fungi as Stagnospora nodorum[44], Aspergillus niger[45] and Cochliobolus heterostrophus[46]. Therefore, another type of activity control in Tc1-Mariner elements likely exists. Four other types of regulation for Tc1-Mariner elements in the S. sclerotiorum genome can be suggested. First, some elements depend on host regulatory factors for transposition, such as transcription factors, the existence of poly(A) sequences, epigenetic regulation and splice sites [6]. In this case, one type of control could be observed because transposition of the TE copies with intron excision could be regulated by its dependence on the host splicing machinery [47],[48]. Second, TEs can be repressed by DNA methylation [49]. Third, transcription of complete elements or MITEs that can form double-stranded RNA (dsRNA) can be controlled due to the presence of TIRs. These dsRNAs would then be processed by the short interfering RNA (siRNA) machinery and could silence copies of transcribed elements [50]. Fourth, because many TE sequences in the S. sclerotiorum genome remain potentially active, monomers of transposase could form inactive or less-active oligomers that decrease transposition activity [51].

Consistent with the in silico evidence of recent activity, the analysis of seven S. sclerotiorum cDNA libraries showed that Tc1-Mariner element sequences were expressed under various conditions and, thus, were likely active in the genome. This fact suggests important ideas about the evolution of the S. sclerotiorum genome. First, it provides evidence that several elements could be transposing in the genome. Thus, an element could insert itself in a new location and inactivate a gene [20]. In addition, when a Tc1-Mariner element transposes, it generates a double-stranded break in the DNA; thus, homologous recombination events that are catalyzed by the DNA repair system could occur [12]. Second, complete transcription of the Tc1-Mariner elements or MITEs can form hairpins due to the complementarity of the TIRs and form a region of dsRNA that can be processed by the enzymatic machinery to form short siRNAs that can, in turn, silence these elements [50]. Third, miRNA originating from transposons are evolutionarily new regulators that are involved in the regulation of endogenous genes [6],[50],[52]. In conclusion, the activity of these TEs may be allowed over evolutionary time in S. sclerotiorum because it provides the fungus with a large range of genetic variability that allows or has allowed the pathogen to parasitize a wide range of hosts.

Fourteen novel Tc1-Mariner families were characterized. Some families had evidence of introns, which might or might not be excised depending on the family or element in question, and this finding demonstrates a possible strategy for overcoming possible mutations that generate premature stop codons in a RNA sequence. This observance also indicates variation in the sequence and conformation of the transposase, which is likely due to the synaptic complex transposition (transpososome). Apparently, Tc1-Mariner TE activity occurred recently or has been tolerated throughout S. sclerotiorum evolution. The presence of these elements near gene regulatory regions may lead to exaptation of these elements by natural selection or genetic drift, and the activity of these transposons may result in recombination, inactivation or changes in gene expression that could provide an important source of genetic variability that allows the fungus to adapt to various stress conditions or exploit a wider range of hosts.

This study was conceptualized planned by MFS and MVQ. MFS performed the in silico and preparation of the manuscript. MVQ coordinated and guided the research, assisted with data analysis and interpretation and helped to prepare the manuscript. EFA and ESGM assisted with the manuscript preparation and were co-mentors for MFS. JCFS assisted with data analysis. All authors have read and approved the final manuscript.