Peptide Inhibitors of HIV-1 Integrase Dissociate the Enzyme Oligomers ² Richard G. Maroun, ‡,§ Ste ´phanie Gayet, ‡ Mohamed S. Benleulmi, ‡ Horea Porumb, ‡ Loussine ´e Zargarian, ‡ Hayate Merad, ‡ Herve ´ Leh, | Jean-Franc ¸ ois Mouscadet, | Fre ´de ´ric Troalen, ⊥ and Serge Fermandjian* ,‡ De ´ partement de Biologie et Pharmacologie Structurales, Laboratoire de Physicochimie et de Pharmacologie des Macromole ´ cules Biologiques, and Laboratoire de Microchimie et d’Immunologie Mole ´ culaire, De ´ partement de Biologie Clinique, UMR 8532 CNRS, Institut GustaVe Roussy, 94805 Villejuif, France, and De ´ partement des Sciences de la Vie et de la Terre, Faculte ´ des Sciences, UniVersite ´ Saint Joseph, CST-Mar Roukos, B.P. 1514, Beyrouth, Liban ReceiVed June 25, 2001; ReVised Manuscript ReceiVed August 30, 2001 ABSTRACT: Integration of HIV-1 genome into host cell chromosome is mediated by viral integrase (IN). The IN catalytic core (CC, IN 50-212 ) dimerizes through mutual interactions of its R1 and R5 helices. Peptides INH1 and INH5 reproducing these helix sequences strongly inhibited IN. For instance, an IC 50 of 80 nM was determined for INH5 in integration assays using wild-type IN (wtIN). In size exclusion chromatography, INH1 and INH5 perturbed the association-dissociation equilibrium of both dmIN (IN 1-288 / F185K/C280S) and CC, leading to monomers as surviving species, while in circular dichroism, binding of peptides to dmIN altered the protein conformation. Thus, enzyme deactivation, subunit dissociation, and protein unfolding are events which parallel one another. The target of INH5 in the enzyme was then identified. In fluorescence spectroscopy, C 0.5 values of 168 and 44 nM were determined for the binding affinity of INH5 to IN and CC, respectively, at 115 nM subunit concentration, while interaction of INH5 with INH1 was found stronger than interaction of INH5 with itself (23 times larger in term of dissociation constants). These results strongly suggested that the R1 helix is the privileged target of INH5. The latter could serve as a lead for the development of new chemotherapeutic agents against HIV-1. The virus-encoded human immunodeficiency virus type 1 (HIV-1) 1 integrase (IN) is essential for the viral replication cycle (for a review, see refs 1-8). It is thus an obvious candidate for antiviral chemotherapy besides the commonly used targets reverse transcriptase and protease. Yet, progress with the design of integrase inhibitors is still slow, a major difficulty being the absence of compounds displaying good specificity for IN (for a review, see refs 9-12). HIV-1 IN is a 32 kDa polynucleotidyltransferase that catalyzes the integration of the DNA copy of the viral genome in a two-step reaction, each proceeding by direct transesterification (3). The 3′ processing occurs in the cytoplasm and the DNA strand transfer in the nucleus of the infected cell. A wealth of experimental results suggests a three-domain structure for HIV-1 IN (7): the N-terminal domain (residues 1-51); the central domain CC (residues 52-210) containing the highly conserved catalytic acidic residues Asp-64, Asp-116, and Glu-152; and the C-terminal domain (residues 220-288) rich in basic residues. The whole protein is required for processing and DNA strand transfer reactions, while CC is able to carry out the disintegration reaction on its own (13). The three-dimensional structure of each domain is now well-known (5, 14-20). Each forms a dimer in solution, although the full enzyme is likely to function as at least a tetramer (4). Several crystal structures of the CC domain (residues 50-212) have been determined (5, 14-17). More recently the first IN multidomain crystal structure has been reported, including the CC and the C-terminal domains (residues 52-288) (21). The protein assumes a Y-shaped dimer structure, where the CC domains are at the dimer interface and the C-terminal domains point away from each other. In all the structures so far reported, the dimer interface involves the strong helix-to-helix contacts R1:R5′ and R5: R1′, where both hydrophobic and electrostatic interactions contribute to dimer stabilization (Figure 1) (15). Here, we report on the inhibitory properties and inhibition mechanism of 2 synthetic peptides: INH1 that reproduces the amino acid sequence of the R1 helix (17 residues from amino acids 93-107); and INH5 that incorporates the R5 helix and a part of the loop separating the R4 and R5 helices (21 residues from amino acids 167-187). INH1 and INH5 present mostly unordered structures in aqueous solution, but they gain in helicity upon addition of trifluoroethanol (TFE), ² This work was supported by SIDACTION and the Agence Nationale de la Recherche sur le SIDA, France. R.G.M. and L.Z. are holders of grants from SIDACTION. * To whom correspondence should be addressed at the De ´partement de Biologie et Pharmacologie Structurales, UMR 8532 CNRS, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France. Tel: +33 1 42 11 49 85. Fax: +33 1 42 11 52 76. E-mail: sfermand@igr.fr. ‡ De ´partement de Biologie et Pharmacologie Structurales, Institut Gustave Roussy. § De ´partement des Sciences de la Vie et de la Terre, Universite ´ Saint Joseph. | Laboratoire de Physicochimie et de Pharmacologie des Macromol- e ´cules Biologiques, Institut Gustave Roussy. ⊥ Laboratoire de Microchimie et d’Immunologie Mole ´culaire, Institut Gustave Roussy. 1 Abbreviations: HIV, human immunodeficiency virus; IN, integrase; dmIN, double mutant integrase; wtIN, wild-type integrase; CC, catalytic core; SEC, size exclusion chromatography; CD, circular dichroism; TFE, 2,2,2-trifluoroethanol; CHAPS, 3-[(3-cholamidopropyl)dimethyl- ammonio]-1-propanesulfonate; EDTA, ethylenediaminetetraacetic acid. 13840 Biochemistry 2001, 40, 13840-13848 10.1021/bi011328n CCC: $20.00 © 2001 American Chemical Society Published on Web 10/26/2001