Regulation of mariner transposition: the peculiar case of Mos1

Abstract

Background: Mariner elements represent the most successful family of autonomous DNA transposons, being present in various plant and animal genomes, including humans. The introduction and co-evolution of mariners within host genomes imply a strict regulation of the transposon activity. Biochemical data accumulated during the past decade have led to a convergent picture of the transposition cycle of mariner elements, suggesting that mariner transposition does not rely on host-specific factors. This model does not account for differences of transposition efficiency in human cells between mariners. We thus wondered whether apparent similarities in transposition cycle could hide differences in the intrinsic parameters that control mariner transposition.

Principal findings: We find that Mos1 transposase concentrations in excess to the Mos1 ends prevent the paired-end complex assembly. However, we observe that Mos1 transposition is not impaired by transposase high concentration, dismissing the idea that transposase over production plays an obligatory role in the down-regulation of mariner transposition. Our main finding is that the paired-end complex is formed in a cooperative way, regardless of the transposase concentration. We also show that an element framed by two identical ITRs (Inverted Terminal Repeats) is more efficient in driving transposition than an element framed by two different ITRs (i.e. the natural Mos1 copy), the latter being more sensitive to transposase concentration variations. Finally, we show that the current Mos1 ITRs correspond to the ancestral ones.

Conclusions: We provide new insights on intrinsic properties supporting the self-regulation of the Mos1 element. These properties (transposase specific activity, aggregation, ITR sequences, transposase concentration/transposon copy number ratio...) could have played a role in the dynamics of host-genomes invasion by Mos1, accounting (at least in part) for the current low copy number of Mos1 within host genomes.

Figure 1. Mariner transposition cycle.

A representative mariner element is depicted at the top of the figure, with its main components, i.e. the transposase coding sequence (in grey), the inverted terminal repeats (5′ITR and 3′ITR, in orange), and the TA dinucleotide flanking the element (landmark for transposition). According to published data , Mos1 transposition consists of five main steps: (1) dimerization of MOS1 proteins, the Mos1 transposase (green circle) for subsequent ITR binding, thus forming SEC2 (Single-end complex 2). (2) Synaptic complex assembly is obtained by the addition of the second ITR to SEC2, thus forming the PEC (Paired-end complex). (3) DNA strands are then cleaved by the transposase, promoting the excision. Once the PIC (Pre-integration complex) has been produced, the capture of the target DNA occurs (4), followed by the integration of the element into a TA target dinucleotide (5). The results presented in this study argue for generalized this model to all mariner elements.

Figure 2. The PEC contains two ITRs and two transposases. A.

Sequence of the double-strand oligonucleotide (short-PC: short pre-cleaved ITR) used in EMSAs. TS: transferred strand; NTS: non-transferred strand. The 3′ITR is shown in bold, with a 3-bases overhang in 3′ of the TS . Inner sequence: narrow letters. The long-PC oligonucleotide (long pre-cleaved ITR) has the same sequence, with a longer inner sequence. An asterisk marks the position of the ³²P labeling. B. Short/long transposase analyses. EMSAs were performed with 250 nM of short-PC (short pre-cleaved) labeled ITR (as a probe), and 250 nM of purified MBP-MOS1. Lane 1: probe alone, Lane 2: complexes assembly without factor-Xa treatment. Lanes 3 to 5, complexes were subjected to factor-Xa cleavage (1H, 2H and 5H respectively) before electrophoresis. The MOS1 dimer in the PEC was assayed here. We have taken advantage of the fact that the MBP-MOS1 fusion protein contains a cleavage site for factor-Xa between MBP and MOS1. If the PEC contains MBP-MOS1 dimer, then a three-band pattern is expected after cleavage of factor-Xa: one band containing uncleaved MBP-MOS1 (in native PEC, as seen at T = 0), one band containing one cleaved and one uncleaved MOS1 in the complex, and one band containing two cleaved MOS1s in the complex. SEC2 disappears as a result of the factor-Xa cleavage, since it contains two MBP-MOS1 molecules that are converted into MOS1 molecules by the release of the MBP moiety, giving bands with faster mobility in electrophoresis. The proteins present in the various PECs are drawn on the right. C. Short/long ITR analyses. EMSAs were performed with 250 nM of purified MBP-MOS1 and 250 nM of short/long ITR combinations (as indicated). The number of ITRs in the PEC is assayed here. The ITRs present in the complexes are drawn on the right. Short-PC: short pre-cleaved ITR. Long-PC: long pre-cleaved ITR. S*: labeled short-PC. L*: labeled long-PC.

Figure 3. The PEC is sensitive to transposase concentration in EMSAs. A.

Relationships between PEC assembly and MOS1 concentration. EMSAs were performed with 250 nM of short-PC labeled ITR, excepted in lane 8, and various amount of purified MBP-MOS1, ranging from 0 to 2.5 µM. Lane 1: no MOS1, lane 2: 50 nM, lane 3: 100 nM, lane 4: 250 nM, lane 5: 500 nM, lane 6: 1 µM, lane 7: 2.5 µM. Lane 8: 2.5 µM of short-PC and 2.5 µM MOS1. Short-PC: short pre-cleaved ITR. B. Relationships between PEC assembly and ITR concentration. Left panel: EMSAs were performed with various amount of short-PC labeled ITR, and 100 nM of MOS1. Lane 1: 5 nM ITR, lane 2: 10 nM, lane 3: 50 nM, lane 4: 100 nM, lane 5: 150 nM, lane 6: 250 nM, lane 7: 500 nM, lane 8: 750 nM, lane 9: 1 µM. Right panel: the amount of ITR in the PEC (nM) was plotted against the concentration of free ITR (nM). The maximum amount of bound ITR obtainable for MOS1 concentration of 100 nM, Bmax, is indicated (dotted line). Short-PC: short pre-cleaved ITR.

Figure 4. pBC-3T3 transposition rates.

Top panels: Transposition rates were assayed with various amounts of pBC-3T3 (1.6 to 16 nM) and three MOS1 concentrations: 10 nM (grey bars), 100 nM (checkerboard bars), 1 µM (hatched bars). Each bar is the mean (+/− SD) of at least five independent assays. Kruskal-Wallis and post hoc tests were used to monitor the significance of the differences. ns: no statistic differences. (*) p<0.05. (**) p<0.005. Bottom panel: Transposition rates (from the top panel) are plotted as a function of the amount of pBC-3T3 used in the assay, for each transposase concentration. 10 nM MOS1: black line, 100 nM: gray line, 1 µM: dotted line.

Figure 5. pBC-5T3 transposition rates.

Top panels: Transposition rates were assayed with various amounts of pBC-5T3 (3.2 to 16 nM) and three MOS1 concentrations: 10 nM (grey bars), 100 nM (checkerboard bars), 1 µM (hatched bars). Each bar is the mean (+/− SD) of at least five independent assays. Kruskal-Wallis and post hoc tests were used to monitor the significance of the differences. ns: no statistic differences. (*) p<0.05. (**) p<0.005. Bottom panel: Transposition rates (from the top panel) are plotted as a function of the amount of pBC-5T3 used in the assay, for each transposase concentration. 10 nM MOS1: black line, 100 nM: gray line, 1 µM: dotted line.

Figure 6. ITR sequence analyses.

Sequences of the Mos1-tribe ITRs (Tables 1 and 2) were aligned to obtain 5′ and 3′ITRs logos and consensus. The majority rule was used to define variable positions, which are 1, 2, 3, 14, 18, and 26 in the 5′ITR and 1, 2, 3, and 14 in the 3′ITR. Doing so, the 24 conserved positions in Mos1 ITRs are conserved in the consensus ITRs (black letters). When compared to each other, 5′ and 3′ consensus ITRs contain three clear differences at positions 1, 18, and 26, which are the same as that found in Mos1 ITRs (blue letters). Position 16 (purple letter) in the 5′ consensus remains ambiguous, being either a G (as in the 5′ Mos1 ITR) or a T (as in the 3′ Mos1 ITR).