Mutational analysis of the integrase protein of human immunodeficiency virus type 2.

Purified integrase protein (IN) can nick linear viral DNA at a specific site near the ends and integrate nicked viral DNA into target DNA. We have made a series of 43 site-directed point mutants of human immunodeficiency virus type 2 IN and assayed purified mutant proteins for the following activities: site-specific cleavage of viral DNA (donor cut), integration (strand transfer), and disintegration. In general, the different activities were similarly affected by the mutations. We found three mutations that (almost) totally abolished IN function: Asp-64-->Val, Asp-116-->Ile, and Glu-152-->Leu, whereas 25 mutations did not affect IN function. A few mutations affected the different activities differentially. Near the amino terminus a zinc finger-like sequence motif His-Xaa3-His-Xaa20-30-Cys-Xaa2-Cys is present in all retroviral IN proteins. Two mutations in this region (His-12-->Leu and Cys-40-->Ser) strongly inhibited donor cut but had less effect on strand transfer. The central region of IN is most highly conserved between retroviral INs. Three mutants in this region (Asn-117-->Ile, Asn-120-->Leu, and Lys-159-->Val) were inhibited in strand transfer but were inhibited less strongly in donor cut. Mutation of Asn-120 (to glycine, leucine, or glutamate) resulted in changes in integration-site preference, suggesting that Asn-120 is involved in interactions with target DNA. We did not find a mutant in which one activity was lost and the others were unaffected, supporting the notion that IN has only one active site for the catalysis of donor cut and strand transfer.

Identification of amino acids in HIV-2 integrase involved in site-specific hydrolysis and alcoholysis of viral DNA termini.

The human immunodeficiency virus integrase (HIV IN) protein cleaves two nucleotides off the 3' end of viral DNA and subsequently integrates the viral DNA into target DNA. IN exposes a specific phosphodiester bond near the viral DNA end to nucleophilic attack by water or other nucleophiles, such as glycerol or the 3' hydroxyl group of the viral DNA molecule itself. Wild-type IN has a preference for water as the nucleophile; we here describe a class of IN mutants that preferentially use the 3' hydroxyl group of viral DNA as nucleophile. The amino acids that are altered in this class of mutants map near the putative active-site residues Asp-116 and Glu-152. These results support a model in which multiple amino acid side-chains are involved in presentation of the (soluble) nucleophile. IN is probably active as an oligomeric complex, in which the subunits have non-equivalent roles; we here report that nucleophile selection is determined by the subunit that supplies the active site.

Complementation between HIV integrase proteins mutated in different domains.

HIV integrase (IN) cleaves two nucleotides off the 3' end of viral DNA and integrates viral DNA into target DNA. Previously, three functional domains in the HIV IN protein have been identified: (i) the central catalytic domain, (ii) the C-terminal DNA binding domain, and (iii) the N-terminal region, which is also necessary for activity. We have now investigated whether IN proteins mutated in different domains can complement each other. Mutant D116I does not contain an intact active site, but does bind DNA, whereas the C-terminal deletion mutant C delta 73 does not bind DNA, but does have an intact active site. Neither mutant protein mediates site-specific cleavage or integration. However, a mixture of both proteins is active, suggesting that IN functions as an oligomer, and that two subunits can have different functions; one subunit binds the (viral) DNA and another subunit provides the active site. We found three classes of mutants, corresponding to the three domains mentioned above. Mutants from different classes, but not from the same class, can complement each other. However, complementation is most efficient when the N- and C-termini are present on the same molecule.