Disulfide Sensitivity in the Env Protein Underlies Lytic Inactivation of HIV-1 by Peptide Triazole Thiols

Abstract

We investigated the mode of action underlying lytic inactivation of HIV-1 virions by peptide triazole thiol (PTT), in particular the relationship between gp120 disulfides and the C-terminal cysteine-SH required for virolysis. Obligate PTT dimer obtained by PTT SH cross-linking and PTTs with serially truncated linkers between pharmacophore isoleucine-ferrocenyltriazole-proline-tryptophan and cysteine-SH were synthesized. PTT variants showed loss of lytic activity but not binding and infection inhibition upon SH blockade. A disproportionate loss of lysis activity vs binding and infection inhibition was observed upon linker truncation. Molecular docking of PTT onto gp120 argued that, with sufficient linker length, the peptide SH could approach and disrupt several alternative gp120 disulfides. Inhibition of lysis by gp120 mAb 2G12, which binds at the base of the V3 loop, as well as disulfide mutational effects, argued that PTT-induced disruption of the gp120 disulfide cluster at the base of the V3 loop is an important step in lytic inactivation of HIV-1. Further, PTT-induced lysis was enhanced after treating virus with reducing agents dithiothreitol and tris (2-carboxyethyl)phosphine. Overall, the results are consistent with the view that the binding of PTT positions the peptide SH group to interfere with conserved disulfides clustered proximal to the CD4 binding site in gp120, leading to disulfide exchange in gp120 and possibly gp41, rearrangement of the Env spike, and ultimately disruption of the viral membrane. The dependence of lysis activity on thiol-disulfide interaction may be related to intrinsic disulfide exchange susceptibility in gp120 that has been reported previously to play a role in HIV-1 cell infection.

Figure 1

Structures and dose response of the effects of 1 and 1a on HIV-1_BaL pseudovirus antiviral functions. (A) 1 is the lytic parent peptide of the library of peptide triazole thiol truncates. All PTs contain the signature pharmacophore, isoleucine (magenta)–azidoproline (blue)–tryptophan (green). (B) 1a is composed of Bis-Mal dPeg conjugated to the C-terminal sulfhydryl groups of two monomers of 1. (C) Inhibition of cell infection analyzed using a single round pseudotyped assay. The IC₅₀ values show that 1 and 1a inhibit HIV-1_BaL infection to the same extent. (D) Relative p24 release measured using ELISA. The calculated EC₅₀ values for 1 and 1a were and 1065 ± 40 nM and >100 000 nM respectively. The data were normalized using untreated virus as a negative control (<5% p24 release), and p24 release observed with 1% Triton X treated virus was taken as 100% p24 content. Sigmoidal curve fits of data were obtained using Origin v.8.1 (OriginLab, Northampton, USA). Error bars represent standard deviation of the mean, n = 3.

Figure 2

Structures of PTTs with serially truncated linker between IXW pharmacophore and C-terminal cysteine. The pharmacophore sequence, composed of isoleucine (magenta)–triazole-proline (blue)–tryptophan (green).

Figure 3

(A) ELISA competition of CD4-gp120 antagonism potencies of truncated peptide triazole thiols. (B) Dose response of the effects of peptide triazole thiol truncates on HIV-1_BaL pseudovirus. Inhibition of cell infection was analyzed using a single round pseudotyped assay. The IC₅₀ values are reported in Table 1. The data show that serially truncated PTTs progressively lose antiviral potency as the distance between the IXW pharmacophore and sulfhydryl group is shortened. (C) p24 release from HIV-1_BaL pseudotyped virus caused by PTTs. Relative p24 release was measured using ELISA. The data were normalized using untreated virus as a negative control (<5% p24 release), and p24 release observed with 1% Triton X treated virus was taken as 100% p24 content. Sigmoidal curve fits of data were obtained using Origin v.8.1 (OriginLab, Northampton, USA). Error bars represent standard deviation of the mean, n = 3. (D) Variation of EC₅₀/IC₅₀ ratio values for p24 release vs gp120 binding as a function of linker length. The IC₅₀ values for gp120 binding were from inhibition of virus cell infection (closed squares) or CD4 competition ELISA (open squares).

Figure 4

Molecular docking simulation of PTT, 8, in complex with gp120. The structure of 8 (carbons shown in cyan) was docked onto gp120 resulting in the binding model shown. Peptide 8 binds in a site overlapping the CD4 pocket (surface shown in gray), with the C-terminal cysteine (SH shown as CPK) trajectory to the conserved disulfide cluster (blue, C296–C331; yellow, C385–C418; violet, C378–C445; orange, C119–C205), which encompasses possible sites of disulfide interaction. The model shows that the peptide SH group is well solvated and with the conformational flexibility can reach any of the disulfide bridges shown.

Figure 5

Effect of protein ligands on p24 release by PTTs. (A) Effect of conformational antibody 2G12 on p24 release. HIV-1_BaL pseudovirus was treated in the presence and absence of 10 μM 1, 14 μM 2 or 50 μM 8 and serial dilution of 2G12 starting at 10 nM. (B) Effect of carbohydrate binding protein, cyanovirin, on p24 release. HIV-1_BaL pseudovirus was treated in the presence and absence of the respective EC₈₀ concentrations of 1, 2, and 8 and serial dilution of cyanovirin starting at 1 μM. (C) Effect of conformational antibody, 697-30D, on p24 release. HIV-1_BaL pseudovirus was treated in the presence and absence of 10 μM 1, 14 μM 2 or 50 μM 8 and serial dilution of 697-30D starting at 1 μM. (D) Effect of 2G12 on 1 binding to gp120: 17b Sandwich ELISA assay performed with constant concentration of 2G12(2.5 μg/mL) and serial dilutions of 1.

Figure 6

Effect of sulfhydryl reagents on PTT induced p24 release. (A) Effect of sulfhydryl reagents on HIV-1_BaL infectivity: DTT, TCEP, and IAAm treated HIV-1_BaL virions. Sulfhydryl reagents were removed by centrifugation prior to adding to HOS cells to test infectivity. (B) Effect of HIV-1_BaL treatment and removal of DTT, followed by 1 treatment and p24 release analysis. (C) Effect of HIV-1_BaL treatment and removal of TCEP, followed by 1 treatment and p24 release analysis. (D) Effect of HIV-1_BaL treatment of DTT serial dilutions and removal with subsequent treatment of 0.02 mM IAAm, followed by 1 treatment and p24 release analysis.

Figure 7

Effect of HIV-1_BaL disulfide mutations on p24 release by 1. Effect of HIV-1_BaL mutations C378S–C445S (cyan), C385V–C415S (blue), and C598A–C604A (red) on p24 release. Relative p24 release was measured using ELISA. The data were normalized using untreated virus as a negative control (<5% p24 release), and p24 release observed with 1% Triton X treated virus was taken as 100% p24 content. Sigmoidal curve fits of data were obtained using OriginPro8 (OriginLab). Error bars represent standard deviation of the mean, n = 3.