Figure 1.
The hairpin and base flipping stages of transposition in different families of transposons.
A Different families of DDE transposons have hairpin intermediates of opposite polarity. Scissile phosphates are shown in red; transposon end and RSS, grey triangles. Left panel: Binding of Tn5 transposase creates a distortion in the DNA that destabilizes stacking of the T+2 base (green). The first step of the reaction is a nick to expose the 3′-OH at the end of the transposon. This facilitates flipping of the T+2 base from the helix in preparation for cleavage of the second strand by a direct transesterification reaction, generating a hairpin intermediate on the transposon end [35], [36]. Subsequently, the hairpin is resolved to yield a blunt transposon end. The insert shows the co-crystal structure of the Tn5 transposon end, with the flipped base at position +2 [1]. All of the residues of the bound transposase have been omitted except for two tryptophan residues. One acts as a probe inserted into the DNA helix, while the other provides stacking interactions to stabilize the flipped base. Right panel: In the hAT transposons and V(D)J recombination the polarity of the reaction is reversed. The first nick occurs on the top strand generating a 3′-OH on the flanking DNA end [2], [37], [38]. Transesterification yields a hairpin on the flanking DNA that is processed by the host. Residue -1 on the bottom strand is distorted and becomes permanganate sensitive after the first nick (green) [17]. B The amino acid sequence of the Tn5 transposase in the vicinity of the probe and stacking residues is aligned with the equivalent region from Tn10 transposase. The Tn5 transposase stacking tryptophan is at position 298 and a tryptophan occupies the equivalent position in Tn10 transposase. However, the Tn5 transposase probe tryptophan aligns with a methionine residue in the Tn10 transposase. The E residue of the YREK motif is also a member of the DDE triad of residues that coordinate the catalytic metal ions and are essential for catalysis.
Figure 2.
Base flipping after the first nick.
Transpososomes were assembled and treated with KMnO4. KMnO4 oxidizes thymine bases in distorted DNA, particularly if they are in an extra-helical position [23]. Oxidation converts the thymine to cis-thymine glycol which, upon piperidine treatment, undergoes further degradation leading to cleavage of the DNA strand. After quenching the DNA was analyzed on a DNA sequencing gel. A The substrate was either uncleaved or pre-nicked. The nucleotide sequence of the transposon end is given on the left, with the arrowhead indicating the location of the transposon end. UC, uncleaved substrate; N, pre-nicked substrate, T'ase, transposase; IHF, integration host factor; *, This band is an artifact that appears to be caused by a heterogeneity at the 5′-end of the DNA strand. Since the label is located at the 3′-end of the DNA strand, this artifact does not contribute to the sequencing ladder or the permanganate footprints. B As in part A except that the transpososomes were assembled on the pre-nicked substrate with the wild type and mutant proteins.
Figure 3.
Crosslinking between transposase and the flipped base.
The thymidine residue T+2 on the top strand of the DNA substrate was substituted with the zero-length crosslinking reagent iodouracil which reacts primarily with aromatic amino acid side chains. Transpososomes were assembled on 5′-end labeled uncleaved (UC) and/or pre-nicked (N) substrates. An aliquot was removed to monitor transpososome assembly by EMSA. The remainder was exposed to UV light and analyzed by SDS-PAGE. Autoradiograms are shown. A Transpososome assembly with wild type transposase was monitored by EMSA. As expected, assembly on linear DNA fragments requires the presence of the host protein IHF [29]. B UV crosslinking of the reaction mixtures from part A and analysis by SDS-PAGE. Crosslinking was detected only in the presence of transposase and was more prominent after the first nick. C and D The same as parts A and B respectively, except that the transpososomes were assembled on the pre-nicked substrate using wild type and mutant transposase.
Figure 4.
Cleavage reactions with transposase mutants and an abasic substrate.
Transpososomes were first assembled in the absence of divalent metal ions. The cleavage reaction was initiated by the addition of MgCl2 at time zero. Aliquots were withdrawn at the indicated times and the reaction halted by the addition of EDTA and SDS. The products were analyzed on a DNA sequencing gel and recorded and quantified by autoradiography on a phosphoimager. The DNA substrates were labeled at both 5′-ends so that all three phosphoryl transfer reactions could be observed in a single experiment. The steps of the cleavage reaction are shown in panel A of the figure below the gel panel. The flanking DNA is to the left and the transposon arm to the right of the half bracket that indicates the location of the transposon end. The positions of the radioactive labels are indicated by the asterisks. Since the reactions are analyzed on denaturing gels, the unlabeled DNA strands, illustrated in grey, are not detected in the autoradiograms. The identity of each band is indicated to the right of the gel in panel A. Bands I and IV each represent a single product of the reaction as indicated. Bands II and III each represent mixtures of more than one co-migrating product and/or substrate as indicated. A & B Cleavage reactions of wild type and abasic DNA substrates. The diagonal slashes indicate regions of the gels that have been removed because they contain no relevant information. Unaltered images of the gels are provided in Figure S1. The identity of the products are indicated next to each band: Band I is the hairpin intermediate; Band II consists the unreacted substrate plus the top strand of the nicked product; Band III contains the bottom strand of the nicked product and the bottom strand of the cleaved transposon end (the resolved hairpin); Band IV contains the top strand of the cleaved flanking DNA that is released upon hairpin formation. In panel B the substrate has an abasic residue at position +2 of the top strand. This was prepared by incorporating a uracil residue at that position by PCR and subsequently treating the substrate with uracil glycosylase. This approach was preferred over one in which the abasic site could have been incorporated during oligonucleotide synthesis. Tn10 transposon arms are folded during assembly of the transpososome [32], [39], [40], and the DNA fragments required are too long for convenient oligonucleotide synthesis. C-F Quantification of cleavage intermediates. The respective products are plotted as a percentage of the total substrate in the reaction. The amount of each intermediate present at each time point is indicated by the shading within the column. None of the conditions tested severely inhibit the nicking step of the reaction. Sixty minutes is sufficient time for all of the transpososome complexes present at the start of the reaction to achieve the first nick. The height of the column at the 60 minute time point is therefore equivalent to the efficiency of transposome assembly, which varied over a 3-fold range in the reactions presented in this experiment. Bands I and IV (corresponding to the hairpin and cleaved top strand, respectively) are unique and unambiguous products of the reaction and can be quantified directly from the gel by phosphorimager analysis. Other intermediates and/or substrate comigrate and therefore can not be quantified directly. They were calculated as follows: first strand cleavage (first nick) = Band III - (Band IV - Band I). Hairpin resolution = Band IV - Band I. These calculations rely on equal labeling efficiency at either end of the substrate. To determine the efficiency of labeling an aliquot of the substrate was cleaved into two parts by NdeI, and the ratio of label incorporated at each end of the fragment was determined by phosphoimager analysis. This ratio was used to adjust all quantifications described above.