Influence of DNA sequence and structure on TN4430 transposition complex assembly

Details

Supervisors

Hallet, Bernard

Faculty

Faculté des sciences

Degree label

Master [120] en biochimie et biologie moléculaire et cellulaire, à finalité approfondie

Abstract

enTn4430 is a replicative bacterial transposon of the Tn3 family, for which a new model termed ‘replication hijacking’ has recently been proposed. Multiple lines of evidence indeed suggest that insertion of this element occurs at DNA replication intermediates, thereby enabling direct recruitment of the host replication machinery at the time of transposition. More specifically, a first key determinant of target selection by Tn4430 appears to be replication-fork stalling sites. In addition to this replication-linked bias, Tn4430 integration sites display a weak core consensus (5′-CTNNNAG-3′) embedded within flanking GC/AT oscillations with an ∼10.5-bp periodicity, pointing to the involvement of both direct and indirect readout. In this context, the aim of this thesis was to further dissect how DNA sequence and DNA shape contribute to the assembly of the Tn4430 transposition complex. To assess whether the determinants identified in vivo are recapitulated in vitro, an “extended consensus” target was first derived from an alignment of the 5,000 most frequently targeted sites in the E. coli chromosome and compared with a scrambled control sequence. Deep-learning–based DNA-shape prediction and indicated that the extended consensus is intrinsically predisposed to adopt a mildly bent conformation. In addition, the extended consensus displayed markedly elevated cyclizability relative to the scrambled sequence. This anisotropic bendability may favor transposase binding, consistent with structural evidence that the target DNA adopts a bent conformation in the transposition complex and protein-induced bendability analysis. To test this, electrophoretic mobility shift assays (EMSAs) were performed using intrinsically curved substrates generated by mismatches or bulges. The transposase did not show a clear preference for these duplexes, although it appeared to interact transiently and/or with low apparent affinity with substrates carrying large bulges. By contrast, the extended-consensus duplex increased apparent binding relative to the scrambled control, whereas a replication fork–mimicking substrate remained the highest-affinity target. Competition assays further indicated that transposon ends (IRs) and target DNA compete for transposase binding in vitro, with no stable ternary complex detected. Nevertheless, a functional transposition complex is still transiently formed, as in vitro strand transfer from the IRs was detected with the fork-like target. Collectively, these results tend to indicate that multiple factors contribute to target choice by Tn4430, including replication as well as the sequence and structure of the target DNA.