Pro- and eukaryotes evolved ingenious ways to shuffle DNA fragments within or between their genomes. One such way is called DNA transposition. During transposition, a DNA fragment called transposon is inserted into a different location in the genome or transferred from a plasmid to a genome of a cell. Among various mechanisms of transposition, the “paste-and-copy” mechanism does not only move the transposon but simultaneously replicates it as if the initial DNA fragment harboring a transposon remained intact while a copy of the transposon becomes inserted into the target DNA. The replication makes the transposition by the “paste-and-copy” mechanism very efficient which is probably why this mechanism of transposition is responsible for the fast spread of antibiotic resistance.
In collaboration with the laboratory of Prof. Bernard Hallet at Université Catholique de Louvain, we studied “paste-and-copy” transposition in the so-called Tn3 family transposons. To accomplish transposition, the pieces of donor DNA containing transposon and target DNA are brought together by a protein called transposase or in the case of the Tn3 family TnpA. Once all pieces of DNA are assembled, TnpA transfers one strand of the transposon into the target DNA thus accomplishing the first and critical step of transposition.
We determined the structures of TnpA transposase in apo form and in a complex with the transposon ends that contain inverted DNA sequence repeats, the signature of transposons. We observed that TnpA is made of 10 structural domains, most of which interact with DNA specifically or unspecifically. Interestingly, the binding of transposon ends leads to large rearrangements in TnpA. These rearrangements are needed for TnpA to be able to recognize transposon ends which it binds specifically and also activates the catalytic domain of TnpA which mediates the strand transfer.
Unlike the transpososome, the TnpA-transposon-target DNA complex, the complex for which we solved the structure did not contain target DNA. However, in the structure, we can observe a large hollow space in between the flanking DNA fragments (yellow), the RNase-H-like catalytic domains (dark red), and the dimerization domains (dark green). The shape and properties of this large cavity are suggestive of the target DNA binding site. We modeled the putative target DNA and showed that the cavity can indeed accommodate a strongly bent double-helical DNA molecule.
Probably the most unusual finding was the behavior of the catalytic RNase-H-like domain. In the apo state, it is partially unfolded and catalytically inactive, whereas the binding of transposon ends induced conformational transition which led to a change in the fold of the RNase-H-like domain, transforming it into a catalytically active form as can be seen in the movie below. Such a change in the fold of proteins or protein domains is called metamorphism, a condition enabling proteins to have more than one well-defined fold.
Another interesting TnpA-specific feature we observed was an alpha-helix that switched its conformation upon binding transposon ends. We called it the switch helix. It is located just next to the catalytic domain, and it carries two very conserved in most transposases arginine residues: R899 and R901. In the DNA-bound form, these arginine residues interact with the DNA backbone and appear to have a critical role in sensing the presence of bound DNA and switching the protein conformation between apo and DNA-bound forms.