DNA ends by retroviral integrases in vitro

We have designed novel substrates to investigate the first step in retroviral integration: the site-specific processing of two nucleotides from the 3' ends of viral DNA. The substrates consist of short duplex oligodeoxynucleotides whose sequences match those of the U3 and U5 ends of viral DNA but are covalently synapsed across the termini by short, single-strand nucleotide linkers. We show here that the optimal separation between termini in a synapsed-end substrate for avian sarcoma/leukosis virus (ASV) IN is 2 nucleotides. This places the two conserved 5'-CA-3' processing sites 6 nucleotides apart, a separation equal to the staggered cut in target DNA produced by this enzyme during the subsequent joining reaction. Based on estimates of initial reaction rates, this synapsed-end substrate is processed by IN at > 10-fold higher efficiency than observed with an equivalent mixture of U3 and U5 single-end (uncoupled) substrates. Enhanced processing is maintained at low IN concentrations, suggesting that the synapsed-end substrate may facilitate enzyme multimerization. Enhanced processing by HIV-1 IN, which produces a 5-bp stagger during integration, was observed with a synapsed-end substrate in which the separation between processing sites was 5 nucleotides. These observations provide estimates of the distances between active sites in the multimeric IN-DNA complexes of ASV and HIV-1. Our results also show that processing of paired U3 and U5 ends need not be coupled temporally. Finally, we observed that substrates that paired a wild-type with a mutated terminus were cleaved poorly at both ends. Thus, in vitro processing of the synapsed-end substrates requires specific recognition of the sequences at both ends. These findings provide new insights into the mechanism of integrative recombination by retroviral integrases and, by extension, other prokaryotic and eukaryotic transposases that are related to the viral enzymes.