Jump ahead with a twist: DNA acrobatics drive transposition forward

Abstract

Transposases move discrete pieces of DNA between genomic locations and had a profound impact on evolution. They drove the emergence of important biological functions and are the most frequent proteins encoded in modern genomes. Yet, the molecular principles of their actions have remained largely unclear. Here we review recent structural studies of transposase-DNA complexes and related cellular machineries, which provided unmatched mechanistic insights. We highlight how transposases introduce major DNA twists and kinks at various stages of their reaction and discuss the functional impact of these astounding DNA acrobatics on several aspects of transposition. By comparison with distantly related DNA recombination systems, we propose that forcing DNA into unnatural shapes may be a general strategy to drive rearrangements forward.

Figure 1

Transposition pathways catalyzed by DNA transposases. (a–c) Schematics of the main steps of transposon excision and integration in distinct transposase families. Examples for which high-resolution transposase-DNA complex structures are available are listed at the bottom. The color scheme (beige: transposon DNA; orange: transposon ends; grey: flanking donor DNA; violet: target DNA) is retained throughout. (a) Main pathways used by DDE transposases. In the cut-and-paste process, the transposon is excised from its original location through DNA double strand breaks. Integration occurs by attack of the liberated 3′-OH groups on a target DNA. In replicative transposition, the element is only nicked on both ends and integration creates a so-called Shapiro intermediate. This is then resolved by replication, generating a new transposon copy at the target site. Some transposases combine features of these main routes, for example, utilizing replication to proceed via excised circular intermediates. (b) Transposition by Y-transposases and S-transposases. Excision creates a double-stranded circular intermediate with the transposon ends abutted. Y-transposases enclose a short stretch (5–7 base pairs) of flanking DNA between the ends. The donor DNA is simultaneously resealed. Recombination of the transposon circle with target DNA, usually in a new bacterial cell, leads to integration. (c) Pathway of HUH-like (Y1-/Y2-) transposases. A single-stranded transposon DNA circle is excised and integrated. Replication re-generates the second DNA strand. (d) Schemes of double strand DNA cleavage in DDE enzymes. The DNA strand that contains the 3′-OH on the transposon end used for subsequent integration is denoted as transferred strand (TS); the complementary strand is labeled non-transferred strand (NTS). The TS is cleaved always precisely at the transposon end, while the site of NTS cleavage varies. Adapted from [13].

Figure 2

DNA distortion promotes cleavage in RAG and the Hermes transposase. (a–f) Transposase-DNA complex structures are shown with the RNase H-like catalytic domain (CAT) in blue, the insertion domain in green and other domains in grey. DNA is colored analogous to Figure 1, with the RSS/transposon ends in orange and the flank in grey. (a) Cryo-EM structure of the RAG-DNA complex before first strand cleavage (PDB: 6DBL). Two RAG1-RAG2 protein pairs synapse two intact DNA substrates (colored in different shades). RAG1 is colored as above, RAG2 is beige. (b) The shape of intact unwound (left) and nicked (right) DNA in the RAG structures (PDB: 6DBL and 6DBI, respectively). DNA nucleotides involved in the distortions are highlighted in red. (c) Close-up of the RAG active site with intact unwound DNA. Two calcium ions (green balls) and the catalytic DDE triad (sticks) assemble around the scissile phosphate (orange sphere). Distorted DNA segments (as in B) are shown as sticks. (d) Crystal structure of Hermes with a nicked DNA substrate arranged for hairpin formation (PDB: 6DWW). Two DNA molecules are bound in a transposase dimer (colored in different shades). DNA is sharply kinked at the boundary of transposon and flank. (e) Cartoon of the Hermes DNA, with flipped-out or unpaired nucleotides shown in red. (f) Close-up of one Hermes active site. The catalytic residues (DDE) are shown as sticks in yellow. Two manganese ions (purple balls) interact with the scissile phosphate (orange sphere). Deformed DNA parts are shown as sticks and transposase residues from the insertion domain that help shape the DNA for hairpin formation are in green.

Figure 3

Target DNA bending is required for integration in the Mu and Mos1 transpososomes. (a) Crystal structure of the Mu strand transfer complex (PDB: 4FCY). Four MuA molecules bind the X-shaped DNA product of integration. Catalytic domains of the two active MuA subunits are colored in shades of blue, other domains and the accessory protein subunits are grey. Transposon end DNA is orange and target DNA is violet. (b) The DNA shape with nucleotides in distorted parts shown in red. (c) Close-up view of one active site. Catalytic residues are shown as yellow sticks. The coiled-coil domain (IIIα) on the concave side of the bent target is labeled. (d) Crystal structure of Mos1 strand transfer complex (PDB: 5HOO). Two transposase molecules bind the integration product DNA (colored as in a). (e) Cartoon of the DNA alone. (f) Close-up of the transposase active site. Unpaired and flipped nucleotides, catalytic residues and amino acids that stabilize base flipping are shown as sticks. The magnesium ion is green and the scissile phosphate is marked with an orange sphere.

Figure 4

DNA melting enables conjugative transposons to insert at diverse genomic sites. (a) Proposed pathway for transposon circle (orange, grey) integration into target DNA (black) by Y-transposases (blue circles). Four DNA strands are cut and exchanged in pairs, proceeding through a four-way Holliday junction (HJ) intermediate. The DNA region that connects the transposon ends in the circle (grey) originates from the donor site and its sequence is different from the target. Thus, strand exchange creates a HJ with no base pairing in the center. (b) Structure of the Tn1549 transposase (Int) with a circular transposon DNA intermediate mimic (PDB: 6EMZ). Two protein molecules (different shades) bind at the transposon ends (orange) with the melted crossover region in the middle (grey). Core DNA binding domain is grey and the catalytic domain is blue. (c) Cartoon of the DNA alone, with nucleotides in the distorted part in red. (d) Close-up of the Int active site. The unwound and melted DNA segments, and the residues involved in catalysis (including the nucleophile tyrosine) or base flipping are show as sticks.