HIV integrase (IN) is a viral enzyme required for integration of virus DNA into the host chromosome. This type of integration is highly specific for retroviruses, and HIV-1 IN is therefore a target for antiviral therapy. Disruption of integrase function is also desirable in design of live attenuated HIV vaccines. We have expressed the IN gene in E. coli in both full-length and truncated forms. Hybridomas have been prepared using the full-length IN clone and the MAbs are reactive with HIV-1 but not HIV-2 by Western blot. IN possesses a specific DNA cleavage activity and a non-specific DNA-binding activity. We have previosly shown that only the C-terminus of IN was capable of binding DNA in a Southwestern blot. Construction of a series of C-terminal deletion clones and analysis of the proteins demonstrated that the DNA binding domain could be localized to within amino acids 180-248. This region contains an array of basic amino acids (lysines and arginines). Site-directed mutagenesis of some of these residues has now been done and the resultant mutant proteins are being purified and will be analysed for their DNA binding ability as well as in a specific oligonucleotide cleavage assay. This newest set of clones was constructed in a gluathione transferase fusion vector, which provides for a simplified purification by affinity chromatography, and additionally, the possibility of producing labelled proteins for use in protein-protein association studies. IN has been postulated to act as a dimer in vivo, but the region of the protein required for this interaction is unknown. We have also purified large quantities of the IN MAb for David Davies to use to generate Fab fragments for X-ray crystallography of IN; attempts to crystallize full- length IN alone have thus far not been successful due to its tendency to aggregate. This may eventually facilitate design of enzyme inhibitors by providing information about IN~s 3-dimensional structure.
