The prototype HIV-1 maturation inhibitor, bevirimat, binds to the CA-SP1 cleavage site in immature Gag particles

Abstract

Background: Bevirimat, the prototype Human Immunodeficiency Virus type 1 (HIV-1) maturation inhibitor, is highly potent in cell culture and efficacious in HIV-1 infected patients. In contrast to inhibitors that target the active site of the viral protease, bevirimat specifically inhibits a single cleavage event, the final processing step for the Gag precursor where p25 (CA-SP1) is cleaved to p24 (CA) and SP1.

Results: In this study, photoaffinity analogs of bevirimat and mass spectrometry were employed to map the binding site of bevirimat to Gag within immature virus-like particles. Bevirimat analogs were found to crosslink to sequences overlapping, or proximal to, the CA-SP1 cleavage site, consistent with previous biochemical data on the effect of bevirimat on Gag processing and with genetic data from resistance mutations, in a region predicted by NMR and mutational studies to have α-helical character. Unexpectedly, a second region of interaction was found within the Major Homology Region (MHR). Extensive prior genetic evidence suggests that the MHR is critical for virus assembly.

Conclusions: This is the first demonstration of a direct interaction between the maturation inhibitor, bevirimat, and its target, Gag. Information gained from this study sheds light on the mechanisms by which the virus develops resistance to this class of drug and may aid in the design of next-generation maturation inhibitors.

Figure 1

Detergent-stripped immature HIV-1 virus-like particle purification. (A) Coomassie blue staining of SDS-PAGE of immature VLP before detergent treatment. (B) Coomassie blue staining of SDS-PAGE of immature particles after detergent treatment. Pr55 Gag and bovine serum albumin (BSA) are indicated by the arrows. Volumes of particle sample and nanograms of BSA standards are given at the top.

Figure 2

Peptide map of HIV-1 Gag after Arg-C, Glu-C, and Lys-C protease digestions. Only fully digested peptide fragments are shown. SP1 sequence is indicated in red. Black arrowhead points to the CA-SP1 cleavage site.

Figure 3

Photoaffinity bevirimat analogs. (A) Structures of bevirimat and the photoaffinity analogs. (B) Detail of mass spectra showing ion mass corresponding to the C-28 analog crosslinked to a synthetic peptide spanning the CA-SP1 region. Red and blue traces are spectra derived from the crosslinking reaction performed in 50% TFE or pure water, respectively. The theoretical monoisotopic mass of the synthetic peptide/C-28 analog adduct is 3302.79: monoisotopic mass of peptide + monoisotopic mass of C-28 - mass of N₂ gas + mass of proton = 2460.29 + 869.5 - 28 + 1 = 3302.79 Da, which is in good agreement with the monoisotopic mass of the ion shown.

Figure 4

Details of mass spectra corresponding to predicted masses for analog-peptide adducts. (A) Lys-C peptide, Gag residues 336-359, linked to the C-28 analog. (B) Lys-C peptide, residues 291-302, linked to the C28 analog. (C) Glu-C peptide, residues 292-307, linked to the C-28 analog. (D) Arg-C peptide, residues 295-299, linked to the C-28 analog. (E) Glu-C peptide, residues 346-365, linked to the C-30 analog. (F) Arg-C peptide, residues 362-384, linked to the C-30 analog.

Figure 5

Summary and analysis of analog-peptide adducts. (A) Pictorial representation of all labeled Gag fragments. The SP1 sequence is indicated in red. The MHR sequence is indicated in brown. The black arrowhead indicates the cleavage site between CA and SP1. Blue and red underlines represent C-28 and C-30 analog-peptide adducts, respectively; (B) Orientation of bevirimat on its target. Bevirimat is positioned parallel to the SP1 helix. Residues H358, L363, A364, A366, Q369, A370, and T371 are indicated. The last residue of the C-28 crosslinked region is indicated in blue, the whole C-30 crosslinked region is indicated in red. The remainder of the sequence is indicated in green. The model is shown in stick mode and dots show the van der Waals surfaces.

Figure 6

Hypothetical models of bevirimat binding to the CA-SP1 junction. Fitting of bevirimat into model A and B according to crosslinking and mutation data. Since the C-30 analog crosslinked within only four residues (VLAE) spanning the CA-SP1 cleavage site and the C-28 analog crosslinked within an area four residues upstream from the CA-SP1 cleavage site, we could orient the drug such that the C28 carboxylic acid group is toward the N-terminus and the C3 dimethylsuccinyl group is toward the C-terminus of the helix. Top panel: Top view of models A&B. In model A, hydrophobic and hydrophilic faces of one helix are making respective contacts with neighboring helices. In model B, the hydrophobic faces of all six helices are pointing toward the center of the bundle. Bevirimat is shown as gray ball-and-stick atoms. Interacting amino acids are also shown in ball-and-stick mode; hydrophobic residues are highlighted in red and hydrophilic residues are highlighted in blue. Models were constructed using six identical NMR structures of a peptide spanning the CA-SP1 region (PDB 1U57). Only sequence spanning the putative helix was used (GHKARVLAEAMSQVTNSATIMMQR). For simplicity, only sequence HKARVLAEAMSQVTNSAT is shown after the energy minimization process. Residue H358 is labeled in magenta. The most common in vitro resistant mutation, A364, is labeled in yellow. The most common resistant polymorphism, V370, is labeled in red. Bottom panel: Side view of models A&B.