Nonhelical leash and alpha-helical structures determine the potency of a peptide antagonist of human T-cell leukemia virus entry

Abstract

Viral fusion proteins mediate the entry of enveloped viral particles into cells by inducing fusion of the viral and target cell membranes. Activated fusion proteins undergo a cascade of conformational transitions and ultimately resolve into a compact trimer of hairpins or six-helix bundle structure, which pulls the interacting membranes together to promote lipid mixing. Significantly, synthetic peptides based on a C-terminal region of the trimer of hairpins are potent inhibitors of membrane fusion and viral entry, and such peptides are typically extensively alpha-helical. In contrast, an atypical peptide inhibitor of human T-cell leukemia virus (HTLV) includes alpha-helical and nonhelical leash segments. We demonstrate that both the C helix and C-terminal leash are critical to the inhibitory activities of these peptides. Amino acid side chains in the leash and C helix extend into deep hydrophobic pockets at the membrane-proximal end of the HTLV type 1 (HTLV-1) coiled coil, and these contacts are necessary for potent antagonism of membrane fusion. In addition, a single amino acid substitution within the inhibitory peptide improves peptide interaction with the core coiled coil and yields a peptide with enhanced potency. We suggest that the deep pockets on the coiled coil are ideal targets for small-molecule inhibitors of HTLV-1 entry into cells. Moreover, the extended nature of the HTLV-1-inhibitory peptide suggests that such peptides may be intrinsically amenable to modifications designed to improve inhibitory activity. Finally, we propose that leash-like mimetic peptides may be of value as entry inhibitors for other clinically important viral infections.

FIG. 1.

HTLV-1 TM and the recombinant TM fusion protein. (A) Structure of the trimer-of-hairpins motif of HTLV-1 TM. The central triple-stranded coiled coil is shown in space-filling form with the extended antiparallel peptide and C-helical region shown in color. (B) Representation of the recombinant trimeric MBP-TM fusion protein used in this study as a model of the exposed coiled coil. The structure is displayed with each monomer of TM in ribbon format, with the three MBP monomers shown as a white space-filling model. (C) Detail of the HTLV-1 TM leash and α-helical region (shown in ribbon form with side-chains represented as sticks). The structure shows the residues that are included in the N terminus of the peptide inhibitor (P^cr-400) of HTLV envelope-catalyzed membrane fusion. The amino acid coordinates denote residue positions in the envelope protein precursor; this numbering system is retained in describing the mimetic peptides for ease of comparison. Residues replaced in this study are shown in red. All structures were modeled from Protein Data Bank ID 1MG1 using MacPymol software (; http://www.pymol.org).

FIG. 2.

Deletions disrupt the activities of peptide inhibitors of membrane fusion. (A) HeLa target cells were cocultured with mock-transfected (top left) or envelope-expressing HeLa cells in the presence of control peptide (HTLV-derived P-80 or HIV-derived C34) or in the presence or absence of P^cr-400 or the truncated derivatives P^cr-ΔN and P^cr-ΔC. The cells were stained with Giemsa and imaged by low-power light microscopy. (B) The numbers of syncytia per low-power field were scored (means ± standard deviations of triplicate assays).

FIG. 3.

Disulfide bonding and cysteine residues are not required for the inhibitory activity of the peptide. (A) Cys401 of Env, equivalent to the second cysteine of the mimetic peptide (blue sticks), lies in a groove on the surface of the coiled coil (green; space-filling model). The arrowhead indicates the position of Leu403. (B) Comparison of the ability of an N-terminally biotinylated P^cr-400 peptide to bind to immobilized recombinant HTLV-1 coiled coil in the presence (+DTT) or absence (−DTT) of 5 mM DTT. (C) The inhibitory properties of the parental peptide (P^cr-400), the control peptide (P-80), and the alanine-substituted peptide (P^cr-CC/AA) were compared in syncytium interference assays. The error bars represent standard deviations.

FIG. 4.

Specific leucine residues in the leash and α-helical segments of the mimetic peptide are critical to inhibitory activity. (A) The parental peptide (P^cr-400) and the alanine-substituted peptides (P^cr-L403A, P^cr-L424A, and P^cr-L429A) were compared for the ability to inhibit membrane fusion in syncytium interference assays. (B) Peptide P^cr-400 and the alanine-substituted peptides were examined for the ability to bind to immobilized recombinant core coiled coil of HTLV-1 in competition with a biotinylated peptide (Bio-P^cr-400). The datum points represent the means ± standard deviations (SD) of triplicate assays. The cavities on the coiled coil of TM occupied by Leu403 of the C-terminal leash (C) and by α-helix residue Leu413 (D) and leash residue Leu419 (E) are shown (the coiled coil is represented in blue, gray, and red space-filling form with key residues labeled and shown as sticks; the TM residues mimicked by the inhibitory peptide are illustrated as green, blue, and red sticks). The arrowhead in panel E indicates V350 from the adjacent TM monomer. (F) The parental peptide (P^cr-400) and peptides carrying alanine substitutions, P^cr-L413A and P^cr-L419A, were examined for the ability to inhibit membrane fusion in syncytium interference assays. (G) Peptides P^cr-400, P^cr-L413A, and P^cr-L419A were compared for the ability to compete with Bio-P^cr-400 for core coiled-coil binding (means ± SD of triplicate assays).

FIG. 5.

Two isoleucine residues modulate inhibitory peptide activity. (A) Ile405 of the peptide (yellow, red, and blue sticks) makes contact with a deep hydrophobic pocket on the coiled coil (blue and red space-filling model; key residues are labeled and shown as sticks; the arrowhead indicates the position of N367 of the coiled coil). (B to D) P^cr-400 and the peptides P^cr-I405A (B) and P^cr-I412A (C) were examined for the ability to inhibit membrane fusion in syncytium assays and for the ability to compete with Bio-P^cr-400 (D) for binding to immobilized core coiled coil (means ± standard deviations of triplicate assays).

FIG. 6.

Replacement of specific amino acids in C helix-based peptide inhibitors has profound effects on the inhibition of viral entry. HIV particles pseudotyped with envelope of HTLV-1 were incubated with target cells in the presence or absence of the peptide inhibitors P^cr-400, P^cr-L413A, P^cr-L419A, and P^cr-I412A at the concentrations indicated. After infection (40 to 48 h), the cells were examined for transduced luciferase activity. The datum points represent the means (± standard deviations) of triplicate assays from three independent cultures. AU, arbitrary units.

FIG. 7.

Replacement of Ile412 with alanine improves peptide binding to the coiled coil by removal of a steric clash. (A) Image of the region mimicked by the inhibitory peptide shown in green-stick format, with the α-helical region shown as a ribbon, I412 (blue stick format, with dots to show the space filled) clashes with Gln356 (red) on the sidewall of the groove on the coiled coil (gray; space-filling form). (B) This clash is removed by the alanine substitution in peptide P^cr-I412A; the model presented is consistent with the derived data, but alternative peptide-docking solutions are possible. The arrowhead indicates the position of Leu413.