Transposition of Mboumar-9: Identification of a New Naturally Active mariner-Family Transposon

Abstract

Although mariner transposons are widespread in animal genomes, the vast majority harbor multiple inactivating mutations and only two naturally occurring elements are known to be active. Previously, we discovered a mariner-family transposon, Mboumar, in the satellite DNA of the ant Messor bouvieri. Several copies of the transposon contain a full-length open reading frame, including Mboumar-9, which has 64% nucleotide identity to Mos1 of Drosophila mauritiana. To determine whether Mboumar is currently active, we expressed and purified the Mboumar-9 transposase and demonstrate that it is able to catalyze the movement of a transposon from one plasmid to another in a genetic in vitro hop assay. The efficiency is comparable to that of the well-characterized mariner transposon Mos1. Transposon insertions were precise and were flanked by TA duplications, a hallmark of mariner transposition. Mboumar has been proposed to have a role in the evolution and maintenance of satellite DNA in M. bouvieri and its activity provides a means to examine the involvement of the transposon in the genome dynamics of this organism.

Transposable elements catalyze the movement of DNA from one genomic location to another. They are powerful forces of genetic change and have played a significant role in the evolution of many genomes. RNA transposons (class I) function via reverse transcription of an RNA intermediate, while DNA transposons (class II) generally move by a cut-and-paste mechanism in which the transposon is excised from one location and reintegrated elsewhere.

The mariner family of DNA transposons is probably the most widely distributed family of transposable elements in nature, represented in such diverse taxa as fungi, ciliates, rotifers, insects, nematodes, plants, fish and mammals.^1–3 Members of the family share three characteristics: a transposase with a DDD catalytic motif, short terminal inverted repeats (TIRs), and a TA dinucleotide target site that is duplicated upon insertion. During transposition, the two ends of the transposon are brought together by oligomerization of the bound transposase to generate a synaptic complex. The transposase then cleaves the DNA at each transposon end and promotes integration of the excised transposon at a new target site.^4–8

Satellite DNA (stDNA) consists of long arrays of tandemly repeated sequences located in genetically silent heterochromatic regions.⁹ Analysis of the stDNA of the ant, Messor bouvieri, revealed the presence of several copies of a mariner transposon termed Mboumar.¹⁰ The transposon is about 1287 bp in length, is flanked by 32-bp inverted repeats, and contains an open reading frame (ORF) coding for a transposase of 345 amino acids with up to 68% amino acid identity to the Drosophila mauritiana Mos1 transposase. Like other mariner transposases, the transposase has two domains: an amino-terminal region containing a helix–turn–helix motif necessary for the recognition and binding of TIRs, and a carboxy-terminal catalytic domain harboring a DD34D catalytic motif.

Miniature inverted-repeat transposable elements (MITEs) are short (80–500 bp) transposon-like elements present in large numbers in many eukaryotes, particularly plant species,^11,12 and occasionally in bacteria.^13,14 Although they have TIRs and are flanked by target site duplications, they generally lack transposase coding potential and are therefore presumably dependent on full-length autonomous transposons for mobility. Multiple copies of a 130-bp MITE-like element, termed IRE-130, have been detected in the stDNA of M. bouvieri.¹⁰ Five copies of the element were found to harbor copies of the Mboumar transposon inserted at the same TA dinucleotide.¹⁰ This suggested that stDNA and IRE-130 could represent hot spots for Mboumar insertions.

We have expressed and purified the Mboumar transposase from a naturally occurring copy of the transposon in M. bouvieri. We demonstrate here that the transposase is fully active and can perform transposition in vitro. Despite the wide phylogenetic distribution of mariner elements in nature, almost all harbor inactivating point mutations or deletions. Consequently, Mboumar is only the third naturally occurring, active, classical (DD34D) mariner transposase discovered to date (after Mos1^15,16 and Famar1 from the European earwig, Forficula auricularia¹⁷). The widespread presence of the Mboumar transposon in M. bouvieri stDNA suggests a positive role for the transposon in the development and/or maintenance of the DNA. Transposition has been hypothesized to play a role in the evolution of stDNA¹⁸ and the activity of the Mboumar transposase could provide us with the means to examine this idea.

In vitro excision of the Mboumar transposon by Mboumar-9 transposase

Several Mboumar transposons in M. bouvieri contain a full-length (complete) transposase ORF. We chose to study the ORF from the Mboumar-9 copy of the transposon because it has preserved all of the critical sequence motifs previously identified in other mariner transposases (the complete nucleotide sequence of the element was published by Palomeque et al.¹⁰). These include the DNA binding helix–turn–helix motif, the bipartite nuclear localization signal, the WVPHEL linker motif and the DD34D catalytic triad. The Mboumar-9 transposase was fused to the maltose-binding protein (MBP) affinity-purification tag by inserting the ORF into the pMAL-c2X plasmid. The MBP–Mboumar-9 fusion was expressed in Escherichia coli and purified as described (Fig. 1a).

Fig. 1.

Purification and excision activity of the Mboumar-9 transposase. (a) SDS-PAGE electrophoresis of fractions from various steps of the Mboumar-9 transposase purification procedure. An MBP–Mboumar-9 transposase fusion protein was expressed in E. coli Rosetta 2 (Novagen) using the pMAL-c2X expression system from New England BioLabs. Expression and purification were performed essentially as described in the supplied instruction manual. Briefly, cells were lysed by French press and centrifuged, and the soluble fraction was passed over amylose resin. MBP-transposase was eluted with maltose and then purified further by cation-exchange chromatography on a MonoS HR5.5 column (Amersham Pharmacia). Elution was with a 20 column volume gradient from 0.05 to 1 M NaCl in Hepes buffer. Lane 1, uninduced cleared cell lysate of E. coli Rosetta 2 cells harboring the MBP-transposase expression plasmid (pRC675); lane 2, cleared cell lysate from the same culture an hour after induction with IPTG; lane 3, eluate from the amylose column; lane 4, purified protein after cation-exchange chromatography. The MBP transposase fusion protein is 83.5 kDa. (b) In vitro cleavage assay. DNA cleavage was performed at 28 °C for 5 h in a total volume of 30 μl. The reaction contained 9 nM of the transposon donor plasmid pMboumar-9, which carries a complete wild-type copy of Mboumar-9. The standard reaction buffer was 25 mM Hepes (pH 7.9) supplemented with 12.5 mg/ml bovine serum albumin, 2 mM DTT, 100 mM NaCl, 10% glycerol and 10 mM MgCl₂ or MnCl₂. Lane 1, no transposase; lanes 2, 3, 4 and 5, reactions with 27, 9, 3 and 1 nM transposase, respectively.

To test whether the Mboumar-9 transposase was capable of transposon excision, a plasmid harboring a copy of Mboumar-9 (pMboumar-9) was incubated with purified transposase and the reaction products were examined by agarose gel electrophoresis (Fig. 1b). Excision of the transposon is expected to release a 3155-bp fragment corresponding to the plasmid backbone. Such a fragment is observed in the presence, but not the absence, of transposase (Fig. 1b). This result demonstrates that the Mboumar-9 transposase is proficient for the excision step of the reaction. We did not detect a fragment corresponding in size to the excised transposon, suggesting that it reacted further to produce integration products. Among the cut-and-paste transposons, the excised linear transposon is a transient species not usually observed at late time points such as these because it rapidly undergoes inter- and/or intramolecular insertions (e.g., Refs. 19–21).

Integration of Mboumar-9

We tested the Mboumar-9 transposase for transposition activity in a genetic in vitro “hop” experiment. In this assay, purified transposase is incubated with a transposon donor and a target plasmid. Transposition events, from donor to target, are recovered by genetic transformation of bacteria and selection of the appropriate antibiotic resistance markers (Fig. 2a and b).

Fig. 2.

Genetic assays for transposition. (a and b) Schematic representations of the two in vitro hop assays for transposition. (a) Transposition of Mboumar-9 from pMboumar-9 to a target plasmid, pGBG1. The cI gene on pGBG1²² acts as a trap for unmarked mobile genetic elements, as described in the text. (b) Transposition of a mini-Mboumar-9 transposon in which the TIRs flank a gene for kanamycin resistance (plasmid pRC766), as described in the text. Drug resistance markers are as follows: Tet^R, tetracycline resistance; Kan^R, kanamycin resistance; Amp^R, ampicillin resistance. ori, plasmid origin of replication; cI, lambda phage cI repressor gene. (c) DNA sequences of transposon integration sites in plasmid pGBG1. In vitro transposition reactions were incubated for 10 h at 28 °C in a 30-μl reaction volume containing 9 nM of each of the donor and target plasmids and 10 nM purified transposase in the standard reaction buffer defined in Fig. 1. Five microliters was used to transform E. coli DH5α cells. Plasmid DNA from the colonies obtained on selective medium was examined by restriction analysis for potential transposition products. DNA sequencing to confirm the transposon junctions was initiated from primer sites flanking the ends of the cI repressor gene. The underlined bases in the figure are extra (non-target) nucleotides at the insertion junctions. The mechanism by which these nucleotides were added is unclear.

We implemented two variations of this assay. The first version employed a “transposon trap” strategy (Fig. 2a). In this assay, the target plasmid (pGBG1) harbors the tetA (tetracycline resistance) gene under the control of the bacteriophage lambda pR promoter. The promoter is repressed by the product of the adjacent lambda cI repressor gene, rendering the host cell sensitive to the antibiotic. A mutation, such as a transposon insertion, in the cI gene will inactivate the repressor, rendering the host cell resistant to tetracycline. In vitro transposition was performed with pMboumar-9 as the donor and the pGBG1 trap as target. The reactions were first evaluated by agarose gel electrophoresis. As expected, a 3155-bp fragment corresponding to the plasmid backbone was detected, indicating that excision of the transposon had occurred (data not shown). The reaction also revealed a set of high molecular weight bands that may correspond to transposon integration products.

To assay for transposon integration into the target plasmid, the reaction mixture was used to transform Escherichia coli. Colonies were obtained only from complete transposition reaction mixtures. No colonies were obtained when the donor, target or transposase were omitted from the mixture. DNA was recovered from 27 colonies and examined by restriction analysis. All 27 plasmids had an insertion within the cI-tetA cassette of the size expected for Mboumar-9. The transposon–target junctions in six of these plasmids were determined by DNA sequencing. In all six cases, the junctions were precise and insertions had occurred into TA dinucleotide target sites in the cI gene (Fig. 2c). Transposition was also performed with Mn²⁺ instead of Mg²⁺ as the catalytic metal ion. In this condition, some of the transposon ends were imprecise and a few of the insertions occurred at target sites other than TA (Fig. 2c, Mn²⁺ panel).

The cI region of the trap plasmid represents a very limited target region for transposition. To better estimate the frequency of transposition, we used a second version of the in vitro hop assay that was designed to detect integration events in non-essential regions of a target plasmid (Fig. 2b). This strategy could also be used to identify potential hot-spot sequences in the target. We constructed a mini-Mboumar-9 donor plasmid in which the gene for kanamycin resistance is flanked by the TIRs of the transposon (pRC766). The relative frequency of transposition was measured by counting the number of colonies obtained after selection with kanamycin and ampicillin (resistance to the latter antibiotic is coded for by a gene on the target plasmid). To exclude the confounding effect of double transformation by both donor and target, the donor plasmid was based on a conditional R6K origin of replication, which is unable to function in the pir⁻ recipient strain. Similar to what was observed in the previous experiment, colonies were obtained only from complete reaction mixtures, and not when the transposase, donor or target were omitted from the reaction. Using this assay, we calculated the relative transposition efficiency (the proportion of target molecules carrying transposon insertions) to be 10^− 3 in the presence of Mg²⁺, and 10^− 4 when Mn²⁺ was substituted for Mg²⁺.

Taken together, these assays demonstrate that Mboumar-9 has easily detectable transposition activity in vitro. Since the reactions were performed in physiological Mg²⁺, pH and salt conditions, it seems likely that Mboumar-9 is active in its natural host M. bouvieri.

Target specificity of Mboumar-9

A previous study indicated that Mboumar is highly represented in the stDNA of M. bouvieri.¹⁰ Any number of different factors, such as a preference for heterochromatin, could potentially give rise to an insertional bias in vivo.²³ However, other transposons have insertion hot spots in vivo or in vitro that are probably determined by the structure of naked DNA.^{19,21,24–26} DNA structure is determined directly by nucleotide sequence and is dictated by factors such as AT/GC content and the phasing of repetitive regions.²⁷

To discover whether Mboumar-9 has enhanced affinity for M. bouvieri stDNA, we performed an in vitro hop assay using stDNA as a target. The target plasmid (pMBSAT) carries a 575-bp fragment consisting of seven tandem repeats of the 79-bp stDNA monomer. Transposition products were analyzed by their NotI restriction digestion pattern (not shown). Nine of the 45 transposition clones tested contained insertions within the stDNA fragment. The proportion of insertions within stDNA (0.20) is not statistically different from the ratio expected if insertions were random. This is true whether the theoretical ratio is based on the relative lengths of non-essential regions on the plasmid (0.19) (χ² ≤ χ_0.95²) or on the relative numbers of available TA dinucleotide target sites (0.26) (χ² ≤ χ_0.95²). As a control, transposition was also performed with the parental plasmid pGEM-T. Transposition efficiency was evaluated by measuring the proportion of total insertions in the lacZ gene. The proportion of insertions in lacZ was no different from the ratio expected if insertions were random (data not shown). Thus, the target plasmid does not contain any significant hot or cold spots that would influence insertion frequencies.

The IRE-130 MITE is 130 bp in length and contains only four TA dinucleotides that are potential target sites for Mboumar integration. Five of the Mboumar copies identified in M. bouvieri stDNA were inserted at the same TA dinucleotide in different copies of IRE-130.¹⁰ To determine whether this TA dinucleotide represents a hot spot for Mboumar insertion, an in vitro hop experiment was performed using a target plasmid (pIRE130) harboring a single copy of IRE-130. We examined 53 independent insertions, but only 3 were present within IRE-130. The remaining insertions were elsewhere in the plasmid. Only one of the IRE-130 insertions was at the TA dinucleotide favored in vivo. As in the previous experiment, the proportion of insertions in IRE-130 (0.06) was no different from that expected if insertions were random (0.05 and 0.04, theoretical ratios based on the lengths of non-essential regions on the plasmid, and on the numbers of TA dinucleotide target sites, respectively [χ² ≤ χ_0.95²]). These results suggest that Mboumar is not strongly biased towards the intrinsic, sequence-dictated structure of M. bouvieri stDNA or the IRE-130 MITE. Any factors responsible for targeting these sequences in vivo must therefore reside at a higher level of chromosome organization.

Implications for satellite DNA evolution

Mariner transposons are arguably one of the most successful transposon families in existence, as evidenced by their widespread distribution in diverse eukaryotic taxa. Yet prior to this report, experimental evidence for transposition activity had only been reported for two naturally occurring classical (DD34D) mariner transposons: Mos1 from D. mauritiana,^15,16 and Famar1 from the European earwig, F. auricularia.¹⁷ In contrast, the well-characterized active mariner transposon, Himar1, represents the consensus sequence derived from numerous inactive elements in the horn fly, Hematobia irritans.⁴ The human Hsmar1 mariner transposon, notably incorporated into the SETMAR gene by a domestication event, has also been similarly resurrected using a reconstructed ancestral sequence deduced from inactive copies.^28–30

We have shown that Mboumar-9 is capable of transposition in vitro. Transposon insertions are precise and occur at TA dinucleotide sites, which are duplicated upon insertion. Consistent with data from other mariners, transposition of Mboumar requires no proteins or cofactors other than Mg²⁺ and the transposase itself. The efficiency of transposition for Mboumar in our genetic in vitro hop assay (in the presence of Mg²⁺) is comparable to that obtained for Mos1 and Himar1 under similar experimental conditions,^25,31 suggesting that their activities may be comparable. Transposition was also examined with the non-physiological ion, Mn²⁺, instead of Mg²⁺. Although transposition is often enhanced with Mn²⁺,^28,32 transposition of Mboumar was reduced 10-fold in the presence of an equal concentration of this cofactor. This has also been observed with Mos1.³¹ Mn²⁺ also caused a relaxation in target site specificity. This has been seen for many transposition systems and is explained by evidence indicating that Mn²⁺ permits more flexible DNA strand positioning in the active site than does Mg²⁺.^28,32

Transposition has been postulated to play an important role in the evolution of stDNA.¹⁸ The recombinogenic nature of transposons, their ability to proliferate and to transpose to locally restricted target sites are properties well suited to a role in the expansion, diversification and homogenization of stDNA. Indeed, some stDNA families are believed to have originated from transposable elements.^33,34 Mboumar has been identified in three ant species to date: M. bouvieri, Messor structor and Messor barbarus.¹⁰ In all three species, multiple copies are localized to stDNA. Unusually for mariners, the vast majority of which are inactivated by multiple mutations, several copies of Mboumar contain intact ORFs.¹⁰ This suggested that Mboumar activity may have recently contributed to, and may continue to act on, the stDNA structure in these ant species.

The association of Mboumar with stDNA and the IRE-130 MITE suggested that they might represent hot spots for insertion. However, our present results suggest that Mboumar has no intrinsic affinity for these DNA sequences, although its affinity for DNA may be quite different in vivo. One alternative explanation is that Mboumar insertions favor stDNA because insertions are excluded from euchromatin either by counterselection or by the physical properties of the different genomic regions, as suggested for other transposons.^33,35 There is also the possibility that Mboumar activity in stDNA provides a benefit of some kind. Current experiments directed towards understanding Mboumar function in vivo may provide an answer to these questions.

Previous analysis of Mboumar insertion sites in stDNA has provided at least one example where co-transposition of adjacent Mboumar copies could have led to mobilization of the intervening stDNA.¹⁰ The excision and integration activities demonstrated here for Mboumar provide the tools to address this prediction and unearth the role of transposition, if any, in the evolution of stDNA. It might also provide us with evidence of ongoing transposition activity that continues to shape the M. bouvieri genome.

Edited by J. Karn