Lytic Inactivation of Human Immunodeficiency Virus by Dual Engagement of gp120 and gp41 Domains in the Virus Env Protein Trimer

Abstract

We recently reported the discovery of a recombinant chimera, denoted DAVEI (dual-acting virucidal entry inhibitor), which is able to selectively cause specific and potent lytic inactivation of both pseudotyped and fully infectious human immunodeficiency virus (HIV-1) virions. The chimera is composed of the lectin cyanovirin-N (CVN) fused to the 20-residue membrane-proximal external region (MPER) of HIV-1 gp41. Because the Env gp120-binding CVN domain on its own is not lytic, we sought here to determine how the MPER_(DAVEI) domain is able to endow the chimera with virolytic activity. We used a protein engineering strategy to identify molecular determinants of MPER_(DAVEI) that are important for function. Recombinant mutagenesis and truncation demonstrated that the MPER_(DAVEI) domain could be significantly minimized without loss of function. The dependence of lysis on specific MPER sequences of DAVEI, determination of minimal linker length, and competition by a simplified MPER surrogate peptide suggested that the MPER domain of DAVEI interacts with the Env spike trimer, likely with the gp41 region. This conclusion was further supported by observations from binding of the biotinylated MPER surrogate peptide to Env protein expressed on cells, monoclonal antibody competition, a direct binding enzyme-linked immunosorbent assay on viruses with varying numbers of trimeric spikes on their surfaces, and comparison of maximal interdomain spacing in DAVEI to that in high-resolution structures of Env. The finding that MPER_(DAVEI) in CVN–MPER linker sequences can be minimized without loss of virolytic function provides an improved experimental path for constructing size-minimized DAVEI chimeras and molecular tools for determining how simultaneous engagement of gp120 and gp41 by these chimeras can disrupt the metastable virus Env spike.

Graphical abstract

The infection of host cells by HIV-1 occurs by virus–cell fusion driven by a cascade of interactions and conformational transformations programmed in the virus envelope trimeric glycoprotein (Env). HIV-1 Env is composed of receptor-binding gp120 and membrane-anchored gp41 subunits held together by noncovalent interactions.¹ gp120 interacts with the receptor (CD4) and coreceptor (CCR5/CXCR4),¹ triggering conformational changes in both gp120 and gp41.^2,3 After receptor binding, the fusion peptide at the N-terminus of gp41 is exposed and interacts with the target cell membrane to form a prehairpin intermediate that bridges the virus and cell membranes.⁴ The gp41 in this transient prehairpin reassembles into a six-helix bundle^5,6 to bring the two membranes sufficiently close together for virus–cell fusion^5–7 and formation of pores that allow viral contents to be released into the cell. The ability of receptor engagement to trigger energy-requiring membrane fusion for entry can be ascribed to the intrinsic metastability of the Env spike protein complex, with six-helix bundle formation providing an energetic driving force for membrane fusion.

We previously hypothesized that an engineered protein chimera composed of a gp120 ligand fused to MPER could hijack the intrinsic metastability of HIV-1 Env by anchoring on the virus Env spike and imparting sufficient stress on the virus membrane adjacent to the spike protein to disrupt the virus membrane and cause virus poration and consequent irreversible inactivation. We devised the chimeric molecule, termed DAVEI (dual-acting viral entry inhibitor), by cloning the gp120-binding lectin protein cyanovirin-N (CVN) joined to gp41 MPER using a flexible [(Gly₄Ser)_X] linker.⁸ DAVEI fusions could block HIV-1 infection in a pseudoviral infection assay at low nanomolar concentrations typical of CVN potency. Importantly, treating viral stocks with DAVEI molecules in the absence of target cells led to dose-dependent virolysis and release of intraluminal p24.⁸ In contrast to CVN-containing DAVEI, CVN alone binds with high affinity to glycans on gp120 and prevents binding of gp120 to CD4 but does not cause membrane poration.⁹ Hence, it is inclusion of the MPER domain in DAVEI that causes virus membrane lysis.

While the lytic inactivation function of DAVEI is striking, the mechanism by which the MPER_(DAVEI) domain allows virolysis has remained undetermined. The MPER peptide on its own has been found to disrupt membranes in model systems,^10–12 and initially, we envisioned that the MPER domain in DAVEI might similarly function by binding to the virus membrane. However, competition of DAVEI virolysis with the MPER peptide alone occurred at surprisingly low concentrations⁸ inconsistent with general MPER capping of the membrane surface to prevent DAVEI function. We sought to resolve the role of MPER_(DAVEI) in this study. We employed a combined mutagenesis and sequence redesign approach to identify minimal sequence surrogates of the MPER domain of DAVEI required for virolysis. The effects of MPER and linker simplification, together with binding measurements and correlation with Env protein structure, argued that the MPER_(DAVEI) domain interacts with Env protein itself. The results obtained open up possibilities for devising smaller DAVEI constructions and also a route to further structure–function analysis of the way in which dual gp120–gp41 engagement causes lytic inactivation of HIV-1.

MATERIALS AND METHODS

Reagents

Modified human osteosarcoma cells (HOS.T4.R5) were a gift from N. Landau. HEK293T cells were purchased from ATCC. The HIV-1 BaL.01 plasmid was a gift from J. M. Garcia. DNA purification was conducted using the Promega miniprep kit (catalog no. A1222). BamHI (catalog no. R0136S) and NdeI (R0111S) restriction enzymes and DNA ligase (catalog no. M0202S) were purchased from New England Biolabs for plasmid digestion and ligation. Mutagenesis was conducted using a Quick Change II XL Site Directed Mutagenesis kit (catalog no. 200522) with reactions conducted using PFU Ultra Polymerase (catalog no. 600380-51) obtained from Agilent technologies. All other reagents were purchased from Sigma-Aldrich unless otherwise specified.

Peptide Synthesis

Peptides were synthesized with a microwave peptide synthesizer (CEM LibertyBlue), using standard Fmoc chemistry on Rink amide resin. All peptides were purified to >95% homogeneity as judged by an analytical reverse phase C18 high-performance liquid chromatography column. Biotinylated-Trp3 peptide was purchased from Scilight-Peptide Inc. The integrity of purified peptides was confirmed by mass spectrometry; observed masses for Trp1, Trp2, Trp3, Trp4, MPER, and Bt-Trp3 were 446.5, 905.2, 1204.6, 2011.3, 2714.9, and 1561.11 Da, respectively, versus expected masses of 447.5, 905.02, 1205.34, 2011.22, 2715.11, and 1559.82 Da, respectively (Supporting Information 11).

Plasmid Constructs

Preparation of DAVEI-L4(ΔMPER) and DAVEI-L4(ΔCRAC) Constructs

Plasmids containing DAVEI-L4⁸ were confirmed for the base sequence. To investigate the virolytic role of the CRAC (cholesterol recognition amino acid consensus) sequence (LWYIK) located at the C-terminus of MPER, two DAVEI-L4 truncated proteins were constructed: DAVEI-L4(ΔCRAC) and DAVEI-L4-(ΔMPER) (Figure 1A). Because the viral membrane contains cholesterol, we hypothesized CRAC_(DAVEI)–cholesterol_(virus) interaction to be an important aspect of virolysis. Forward primer 5′-GAAATAACAGAATGGTAGTAGTGGATAAAATAGTAATAAAAGC-3′ and 5′-CATCATCATCATCATCATTAGAAATGGGCAAGTTTGTGG-3′ and their reverse complement primers were used to create DAVEI-L4(ΔCRAC) and DAVEI-L4(ΔMPER), respectively, using a standard mutagenesis protocol from Agilent technologies (Figure 1B).

Figure 1.

Assessment of the role of the CRAC domain in virolysis shown by the DAVEI molecule. (A) Different subdomains of gp41 subunit are highlighted. The amino acid sequence in the MPER region is shown. CRAC residues are colored red. (B) Cartoon representation of two derivatives of DAVEI-L4: DAVEI-L4(ΔCRAC) and DAVEI-L4(ΔMPER). (C) Detection of direct binding of DAVEI-L4, CVN, and DAVEI-L4(ΔCRAC) on BaL.01 viruses with trimers. Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control. Data were normalized relative to the positive control. Viruses devoid of trimers were taken as a negative control that showed no detectable binding to DAVEI derivatives. (D) HOS.T4.R5 cells were exposed to the HIV-1 BaL.01 pseudovirus with serial dilutions of the DAVEI-L4 derivatives. The inhibitory potencies of DAVEI-L4, CVN, DAVEI-L4(ΔCRAC), and DAVEI-L4(ΔMPER) were 1.2 ± 0.3, 0.9 ± 0.2, 1.1 ± 0.2, and 1.4 ± 0.3 nM, respectively. IC₅₀s were calculated using Origin version 8.1. (E) Release of p24 from HIV-1 BaL.01 pseudovirus upon incubation with DAVEI-L4 derivatives. A sandwich ELISA was conducted in which experimental p24 release was background-subtracted using PBS treated virus and then compared to virus lysed with 1% Triton X-100 (means ± SD; n = 3). The EC₅₀s for DAVEI-L4 and DAVEI-L4(ΔCRAC) protein were 29.8 ± 1.0 and 28.3 ± 2.1 nM, respectively. CVN and DAVEI-L4(ΔMPER) did not show p24 release. EC₅₀s of virolysis of HIV-1 BaL.01 pseudovirus were determined with Origin version 8.1.

Subcloning of DAVEI-L2 and DAVEI-L0 Constructs

BamHI restriction sites were introduced at two different positions within the DAVEI-L4 plasmid for removing glycine-4-serine repeats. Primer 5′-GGTACCCTGAAATACGAAGGATCCGGTGGCGGAGGGTC-3′ and 5′-GGTAGTGGTGGAGGCGGATCCCATCATCATCATCATCAT-3′ and their reverse complements were used to introduce two BamHI restriction sites. Forward primer 5′-GAAGGAGATATACATATGAAATACCTGCTGC-3′ and its reverse complement were used to introduce the NdeI restriction site for both of the modified plasmid constructs. The modified plasmids were digested with BamHI and NdeI enzymes for 2 h at 37 °C before being treated with shrimp alkaline phosphatase for 30 min at room temperature. The excised vectors were run on a 1% agarose gel (Supporting Information 8), purified using a gel extraction kit (Qiagen), and religated using T4 ligase (New England Biolabs M0202L). 5′-GGAGGGTCGGGCGGAGGTGGATCCGGTGGCGGAGGTGG-3′ and its reverse complement primer and 5′-GGTAGTGGTGGAGGCGGATCCCATCATCATCATCATCATC-3′ and its reverse complement primer were used to introduce two BamHI restriction sites for making the DAVEI-L2 construct. Forward primer 5′-GAAGGAGATATACATATGAAATACCTGCTGC-3′ and its reverse complement primer were used to introduce the NdeI restriction site for both modified plasmid constructs. Both mutated plasmids were digested with BamHI and NdeI (New England Biolabs) for 2 h at 37 °C before being treated with shrimp alkaline phosphatase for 30 min at room temperature. The excised vectors were run on a 1% agarose gel (Supporting Information 8), purified using a gel extraction kit (Qiagen), and religated using T4 ligase (New England Biolabs M0202L) overnight at room temperature. All ligated products were transformed on XL1-Blue Escherichia coli cells, plated on LB-kanamycin agar plates, and transformed overnight at 37 °C. Positive colonies were selected for miniprep, and plasmids were isolated and sequenced (Figure 4A).

Figure 4.

Determination of the minimal linker sequence required for virolysis by DAVEI. (A) Cartoon representation of two linker derivatives of DAVEI-L4: DAVEI-L2 and DAVEI-L0. (B) Detection of direct binding of DAVEI-L4, DAVEI-L2, and DAVEI-L0 on BaL.01 viruses with trimers. Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of trimers were taken as a negative control that showed no detectable binding to DAVEI derivatives. (C) Inhibition of HIV-1 viral infection by DAVEI-L4, DAVEI-L2, and DAVEI-L0 [means ± standard deviation (SD); n = 3]. The inhibitory potencies are as follows: IC₅₀ = 4.3 ± 1.7 nM for DAVEI-L0 and IC₅₀ = 2.1 ± 0.3 nM for DAVEI-L2). (C) The EC₅₀s for DAVEI-L2- and DAVEI-L4-induced virolysis were 182.6 ± 28.5 and 23.6 ± 2.5 nM, respectively, determined using Origin Pro version 8.1 with sigmoidal fits (means ± SD; n = 3). DAVEI-L0 did not induce p24 release at concentrations of up to 2000 nM. PBS-treated virus was used as a negative control, and virus lysed with 1% Triton X-100 was taken as a positive control (means ± SD; n = 3).

DAVEI-L4(ΔCRAC) Constructs with the MPER Mutation on Tryptophan

The DAVEI-L4(ΔCRAC) construct has an MPER sequence (lacking the CRAC sequence) at its C-terminus (DKWASLWNWFEITEW) with tryptophan at positions 3, 7, 9, and 15. Alanine scanning mutagenesis was performed to create W15A, W9,15A, W7,9,15A, and W3,7,9,15A mutations to determine the effect of tryptophan residues on the lytic behavior of DAVEI-L4. W3A, W7A, W9A, and W15A mutations were created using 5′-CATCATGACAAAGCGGCAAGTTTG-3′, 5′-CAAGTTTGGCGAATTGGTTTG-3′, 5′-GGCAAGTTTGTGGAATGCGTTTGAAATAACAGAATGG-3′, and 5′-GTTTGAAATAACAGAAGCGTAGTAGTGGATAAAATAG-3′ and their reverse complement primers, respectively. Because W7 and W9 are too close, a combination mutation was formed using 5′-GCAAGTTTGGCGAATGCGTTTGAAATAAC-3′ that had tryptophan to alanine mutations at both positions 7 and 9. Schematic representations for plasmid constructs are shown in Figure 2A.

Figure 2.

Assessment of the role of hydrophobic tryptophan residues of MPER in virolysis. (A) Cartoon representation of four derivatives of DAVEI-L4(ΔCRAC). (B) Detection of direct binding of DAVEI-L4(ΔCRAC) and DAVEI-L4(ΔCRAC)W3,7,9,15A tryptophan mutants on BaL.01 viruses with trimers. Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control. Viruses devoid of trimers were taken as a negative control that showed no detectable binding to DAVEI derivatives. (C) Inhibition of HIV-1 viral infection by DAVEI-L4(ΔCRAC) derivatives. The inhibitory potencies of DAVEI-L4(ΔCRAC) (IC₅₀ = 1.1 ± 0.2 nM), DAVEI-L4(ΔCRAC)W15A (IC₅₀ = 0.4 ± 0.1 nM), DAVEI-L4(ΔCRAC)W9,15A (IC₅₀ = 0.45 ± 0.1 nM), DAVEI-L4(ΔCRAC)-W7,9,15A (IC₅₀ = 0.36 ± 0.1 nM), and DAVEI-L4(ΔCRAC)-W3,7,9,15A (IC₅₀ = 0.78 ± 0.3 nM) are shown [means ± standard deviations (SD); n = 3]. (D) Release of p24 from HIV-1 BaL.01 pseudovirus upon incubation with DAVEI-L4(ΔCRAC) derivatives. EC₅₀s of virolysis were 29.8 ± 1.0, 46.8 ± 5.9, 49.7 ± 5.5, 43.1 ± 3.6, and 122.2 ± 11.6 nM for DAVEI-L4(ΔCRAC) and its single, double, triple, and quadruple mutants, respectively (means ± SD; n = 3).

Generation of DAVEI Derivatives with Shortened MPER Sequences

DAVEI-L4(ΔCRAC) has four tryptophans at positions 3, 7, 9, and 15. To identify the minimal MPER sequence required for lysis by DAVEI-L4, the MPER sequence was truncated by introducing stop codons after each tryptophan. Three truncated proteins, namely, DAVEI-L4-1Trp, DAVEI-L4-2Trp, and DAVEI-L4-3Trp, were formed using 5′-CATCATCATGACAAATGGTGAAGTTTGTGGAATTGG-3′, 5′-GACAAATGGGCAAGTTTGTGGTAGTGGTTTGAAATAAC-3′, and 5′-GGCAAGTTTGTGGAATTGGTAGGAAATAACAG-3′ and their reverse complements, respectively, as shown in Figure 3A. The number X in DAVEI-L4-XTrp represents the total number of tryptophans present on the MPER sequence of the DAVEI-L4 derivative (Figure 3A).

Figure 3.

Determination of the minimal MPER sequence (MPER_min) required for virolysis by the DAVEI molecule. (A) Cartoon representation of four derivatives of DAVEI-L4(ΔCRAC). (B) Truncated MPER peptides of corresponding lengths as the protein derivatives that are used as a control for virolysis. (C) Detection of direct binding of DAVEI-L4(ΔCRAC) and DAVEI-L4-3Trp truncates on BaL.01 viruses with trimers. Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of trimers were taken as a negative control that showed no detectable binding to DAVEI derivatives. (D) Inhibition of HIV-1 viral infection by DAVEI-L4(ΔCRAC) truncated proteins [means ± standard deviations (SD); n = 3]. The inhibitory potencies are as follows: IC₅₀ = 1.1 ± 0.2 nM for DAVEI-L4(ΔCRAC), IC₅₀ = 0.8 ± 0.2 nM for DAVEI-L4-1Trp, IC₅₀ = 2.3 ± 0.1 nM for DAVEI-L4-2Trp, and IC₅₀ = 2.2 ± 0.4 nM for DAVEI-L4-3Trp. (E) Release of p24 from HIV-1 BaL.01 pseudovirus upon incubation with DAVEI-L4-(ΔCRAC) truncations. EC₅₀s for DAVEI-L4(ΔCRAC), DAVEI-L4-3Trp, and DAVEI-L4-2Trp proteins were 28.3 ± 2.1, 36.1 ± 5.1, and 113.8 ± 11.1 nM, respectively, determined using Origin Pro version 8.1 with sigmoidal fits (means ± SD; n = 3). DAVEI-L4-1Trp did not show p24 release.

Protein Expression, Purification, and Validation of DAVEI Fusions

Purified DAVEI-L4(ΔCRAC) and all other derived plasmids were transformed on BL21(DE3) pLysS competent cells (Promega) and plated on LB-kanamycin agar plates for 16–18 h at 37 °C. Positive colonies were isolated and inoculated on 1 mL of LB-kanamycin for 16 h at 30 °C. A 1 mL culture was subcultured to a 4 L culture, grown for 6–8 h at 30 °C, and induced with 1 mM isopropyl D-thiogalactopyranoside (IPTG) when the optical density was in the range of 0.6–0.8. Cells were induced for 16–18 h at 16 °C and a shaking speed of 225 rpm. Grown cells were pelleted by centrifugation, and 60 mL of buffer A (50 mM Na₂HPO₄, 300 mM NaCl, and 10 mM imidazole) was added to the pellet. The mixture was sonicated using a microtip probe (Misonix Sonicator 3000) for three 1 min intervals at 70 V. To separate the protein from the cell debris, the sonicated mixture was spun down at 10000g for 45 min at 4 °C. The protein was purified with nickel-nitrilotriacetic acid (Ni-NTA) beads (Qiagen), followed by gel filtration with a 26/60 Superdex 200 prep-grade column (GE Healthcare) using an AKTA fast protein liquid chromatograph (GE Healthcare). The homogeneity of these proteins was assessed with protein eluates run on 18% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) gels (Supporting Information 9) followed by ELISA analysis. Fractions containing target proteins were concentrated and buffer exchanged with phosphate-buffered saline at pH 7.4 using a 3000 molecular weight cutoff (MWCO) spin filter (Amicon). The final concentration was determined using absorbance at 280 nm and extinction coefficients for various DAVEI constructs. Cyanovirin-N was produced in E. coli as previously reported.^13–15 The ability of fusion proteins to bind YU2 gp120 was analyzed by a sandwich ELISA. First, varying concentrations (500–0.5 ng) of CVN or fusion proteins were adsorbed to wells of 96-well high-binding polystyrene plates (Fisher Scientific) for 12 h at 4 °C. The plates were blocked with 3% bovine serum albumin (BSA) (Research Products International Corp.) in phosphate-buffered saline (PBS) for 2 h at 25 °C; 100 nM purified wild-type YU2 gp120 protein was loaded onto the plates and incubated for 1 h. The plates were washed with 0.1% (v/v) Tween 20 (PBS-T) for 15 min. All subsequent incubation steps were performed in 0.5% BSA in PBS. After a 15 min incubation period, 50 μL of sheep anti-gp120 (dilution factor of 1:3000; Aalto Bioreagents, D7324) was added to the plates for 1 h at room temperature. The plates were washed with PBS-T followed by addition of rabbit anti-sheep secondary antibody (dilution factor of 1:3000; life technologies, PJ209733). The plates were incubated for an additional 1 h at 25 °C and washed three times with PBS-T and one more time with PBS before the addition of OPD. Plates were developed for 30 min in the dark, and the final absorbance was measured at 450 nm using an Infinite m50 (Tecan) plate reader.

Expression and Purification of gp120 Protein

YU2 gp120 in the pcDNA3.1 plasmid was purified using a Qiagen MaxiPrep kit (Qiagen). The wild-type (WT) plasmid was transiently transfected into HEK293F cells according to the manufacturer’s protocol (Invitrogen). Five days post-transfection, cells were harvested and spun down, and the supernatant was filtered through 0.2 μm filters. Purification was performed over a 17b antibody column prepared using an NHS-activated Sepharose, HiTrap HP column (GE Healthcare). gp120 was eluted from the column using 0.1 M glycine buffer (pH 2.4). The pH of the eluted protein was rapidly neutralized by addition of 1 M Tris (pH 9.0). Purified proteins were run on 10% SDS gels for detecting the expression of protein. Eluted proteins were immediately dialyzed into 1× PBS and loaded onto a HiLoad 26/60 Superdex 200 HR prepacked gel filtration column (GE) for size exclusion chromatography. Eluted fractions of protein were run on 10% SDS–PAGE gels. Purified monomeric fractions of gp120 were concentrated and flash-frozen at −80 °C.

Direct Binding of DAVEI Derivatives to gp120 Monomeric Proteins for Functional Validation

Recombinantly purified YU2 gp120 protein (50 ng) was immobilized on ELISA plates. Plates were blocked overnight with 3% BSA at 16 °C. Varying concentrations of DAVEI derivatives were added to the plate, and the plate was incubated for 2 h followed by two 30 min washes with 1× PBS. Rabbit anti-CVN (Biosyn Inc.; dilution of 1:3000) was added followed by two washes with 1× PBS. The donkey anti-rabbit HRP conjugate (dilution factor of 1:3000; GE Biosciences, GE NA934V) was used as the secondary antibody, which was detected using an OPD solution.

Production of HIV-1 BaL.01 Pseudotype Virus

Recombinant pseudovirus were produced by cotransfection of two plasmids: (1) envelope plasmid that encodes the BaL.01 gp160 region and (2) backbone sequence corresponding to envelope-deficient pNL4-3 Luc⁺ Env⁻.¹⁶ Three million HEK293T cells were plated on a T75 flask (Corning Inc.). Cells were cotransfected 24 h after being plated with 4 μg of envelope DNA and 8 μg of backbone DNA, using PEI as a transfection reagent. Medium was changed 24 h post-transfection, and the cells were allowed to grow for an additional 48 h. After 48 h, the cell supernatant containing virus was collected and filtered using a 0.45 μm filter (Corning Inc.). The filtered supernatant was loaded onto a Iodixanol gradient (gradient range from 6 to 20%; Optiprep, Sigma-Aldrich) and centrifuged in a Sw41 Ti rotor (Beckman Coulter) at 110000g for 2 h at 4 °C. The five lower fractions (each 1 mL) were pooled together,¹⁷ and 400 μL aliquots were frozen at −80 °C. Purified pseudoviruses were tested for infectivity and p24 content postproduction.

Direct Binding of DAVEI Derivatives to Trimeric Spikes on BaL.01 Viruses

BaL.01 viruses were diluted (1:20) in 1× PBS and coated onto 48-well ELISA plates, sealed with parafilm, and kept at 4 °C overnight. Plates were blocked the following day with 200 μL of 3% BSA in 1× PBS for 2 h at room temperature. Blocking buffer was discarded (by pipetting), and DAVEI derivatives were added to the viruses at varying concentrations and incubated for 30 min. Plates were washed twice with 1× PBS, 10 min each. Viruses were fixed onto the ELISA plates with 100 μL of 1% PFA (paraformaldehyde) for <15 min. Plates were washed twice with 1× PBS followed by addition of rabbit anti-CVN (Biosyn Inc.; 1:3000 dilution) as the primary antibody, which was detected using the donkey anti-rabbit HRP conjugate (dilution factor of 1:3000; GE Biosciences, GE NA934V) as the secondary antibody. Plates were washed and detected using an OPD solution. Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of BaL.01 trimers produced with envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg) were used as a negative control for the experiment.

Viral Infection Inhibition Assays

Infection inhibition assays were conducted for all BaL.01 viruses produced as described previously.^18,19 Approximately 7000–8000 HOS.T4.R5 cells were seeded in 96-well plates on day 1. Twenty-four hours later, BaL.01 viruses were added to the plated cells in the presence or absence of DAVEI compounds at various dilutions. The amount of viruses to be added to each well was determined on the basis of the titer experiment. For the optimal signal, virus stocks were diluted in growth medium such that the final dilution of virus showed a signal of 10⁶ luminescence counts. After the addition of the virus and compound, the plates were incubated for 24 h at 37 °C before the medium was changed. Cells were allowed to grow for an additional 24 h, after which the medium was removed and the cells were lysed using Passive Lysis Buffer (Promega). Cell lysates were transferred to a white well plate (Greiner) and mixed with 1 mM luciferin salt (Anaspec) diluted in 0.1 M potassium phosphate buffer containing 0.1 M magnesium sulfate. The luminescence was measured using a Wallac 1450 Microbeta luminescence reader at a wavelength of 490 nm.

Sandwich ELISA for the Detection of p24 Release

The release of p24 from viruses was measured using the protocol described previously.¹⁷ In brief, a p24 titer experiment was conducted to estimate the dilution factor of the virus required to observe the optimal p24 signal, equivalent to the signal shown by 50 ng of recombinantly purified p24. The final dilution factor for the viral stock used was found to be 1:40. After the titer experiment, high-binding polystyrene ELISA plates were coated overnight at 4 °C with 50 ng of mouse anti-p24 and blocked with 3% BSA for 2 h at room temperature on a rocker. The blocked plate was rinsed three times with PBS-T (PBS and 0.05% Tween 20). Viral stocks were diluted 10-fold using PBS (pH 7.2), and 100 μL of diluted stocks was treated with 100 μL of serially diluted compound; 100 μL of virus treated with 100 μL of 1% Triton X-100 was used as a positive control to ensure complete lysis and release of p24. Recombinant p24 (50 ng) was also used as a positive control. The virus/compound stocks were incubated at 37 °C for 0.5 h and spun down in 1.5 mL tubes for 2 h at 4 °C and 21130g. The top 100 μL of the supernatant was removed carefully without disturbing the pelleted virus fractions. This supernatant was then diluted [1:1 (v/v)] in PBS containing 1% BSA and 1% Triton X-100; 50 μL of sample per well was then loaded onto the prepared ELISA plates. The plate was incubated at 4 °C overnight (16 h) on a rocker before being washed three times with PBS-T. The primary rabbit anti-p24 antibody (dilution factor of 1:3000; Abcam, ab63913) was added to the plate, followed by the secondary antibody, the donkey anti-rabbit HRP conjugate (dilution factor of 1:3000; GE Biosciences, GE NA934V). Both antibodies were incubated for 1 h at room temperature. The plate was washed three times with PBS-T and one more time with PBS before the addition of OPD. The plate was developed for 30 min in the dark, and the final absorbance was measured at 450 nm using an Infinite m50 (Tecan) plate reader.

Competition Virolysis Assay

The p24 lysis titer experiment was first performed with DAVEI-L4-3Trp protein to determine the minimal concentration of protein required to observe maximal lysis.⁸ This concentration was determined to be 50 nM. Purified BaL.01 virus was incubated with 50 nM DAVEI-L4-3Trp protein in the presence of various concentrations of MPER peptide (DKWASLWNWFEITEWLWYIK) or Trp3 peptide (DKWASLWNW) for 30 min at 37 °C. UM15 peptide was taken as a negative control because its binding does not compete with CVN binding.¹³ The incubated mixture was centrifuged for 2 h at 4 °C and 21130g. The top 100 μL of supernatant was carefully removed without disturbing the virus pellet and then further diluted [1:1 (v/v)] in PBS containing 1% BSA and 1% Triton X-100. This sample was used to conduct the same sandwich p24 ELISA as described above. Viruses treated with PBS and with 1% Triton X-100 were used as negative and positive controls, respectively. A p24 standard (Abcam, Ab9071, 50 ng) was also used as a positive control in each experiment. Experiments were conducted in triplicate, and the percentage of p24 released as compared to the positive control was plotted versus the concentration of MPER and Trp3 peptide used. Data were plotted using Origin version 8.1 to determine the IC₅₀.

Cell-Based Enzyme-Linked Immunosorbent Assay (ELISA)

A cell-based ELISA was optimized to measure the direct binding of Bt-Trp3 peptide to the HIV-1 envelope glycoprotein trimers expressed on cells. The protocol was based on early experiments published by Haim et al.²⁰ and was optimized for our assay with minor modifications. HEK293T cells (30000) were seeded on 24-well plates. Cells were transfected the next day with 0.4 μg of a plasmid expressing the envelope glycoprotein trimers [Env(−)ΔCT] using the Fugene transfection reagent (Promega, E2311). The medium was changed the following day, and 2 days later, cells were used for the direct binding experiment. In brief, cells were blocked once with blocking buffer [35 mg/mL BSA, 140 mM NaCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 25 mM Tris (pH 7.5)] for 30 min. After being blocked, cells were washed with wash buffer [140 mM NaCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 25 mM Tris (pH 7.5)], and the following set of experiments were conducted.

Direct Binding of mAbs

F105 (AIDS Repository, catalog no. 857), 4E10 (AIDS Repository, catalog no. 10091), 2G12 (AIDS Repository, catalog no. 1476), 10E8 (AIDS Repository, catalog no. 12294), and 35O22 (AIDS Repository, catalog no. 12586) antibodies were each diluted in blocking buffer at various concentrations (0.0001–1000 nM), added to the HEK293T cells that expressed envelope trimers, and incubated for 45 min at 37 °C. Cells that were not transfected with Env trimer-expressing plasmids were taken as a negative control. Cells were blocked once with blocking buffer for 10 min followed by two 15 min washes with wash buffer. Anti-human HRP (dilution factor of 1:3000; Millipore, AP101P) was added to the cells, and the cells were incubated for 45 min. The plate was washed twice with wash buffer, and an OPD solution was added. After incubation for 15 min in the dark, the final absorbance was measured at 450 nm using an Infinite m50 (Tecan) plate reader. A schematic depiction of the experimental setup is shown in Figure 7A.

Figure 7.

Assessment of the interaction of the trimer with various antibodies and biotinylated Trp3 peptide by a cell-based ELISA. (A) Cartoon representation showing trimer expression on the HEK293T cell surface and its interaction with antibodies. (B) Interaction of gp120 and MPER specific antibodies with trimers expressed on the HEK293T cell surface [F105 (EC₅₀ = 4.5 ± 0.3 nM), 35O22 (EC₅₀ = 6.8 ± 1 nM), 4E10 (EC₅₀ = 9.6 ± 0.5 nM), 10E8 (EC₅₀ = 6.7 ± 0.7 nM), and 2G12 (EC₅₀ = 2.7 ± 0.5 nM)]. F105 does not bind to HEK293T cells not transfected with the JRFL Env(−)ΔCT plasmid. (C) Cartoon representation showing trimer expression on the HEK293T cell surface and its interaction with biotinylated Trp3 peptide. (D) Interaction of biotinylated Trp3 peptide with HEK293T cells transfected with the JRFL Env(−)ΔCT plasmid [biotinylated Trp3 (EC₅₀ = 129 ± 14 nM)]. Biotinylated Trp3 peptide does not bind to cells not transfected with the JRFL Env(−)ΔCT plasmid.

Direct Binding of Biotinylated Trp3 Peptide

Biotinylated Trp3 peptide was diluted in blocking buffer at various concentrations (10 μM to 0.0001 nM), added to the adherent HEK293T cells expressing envelope trimers, and incubated for 45 min at 37 °C. The Bt-Trp3 peptide solution was removed, and the plate was washed twice, 15 min each, with wash buffer. Streptavidin HRP (dilution factor of 1:3000; Anaspec, 0668) was added to the cells, and the cells were incubated for 45 min. The plate was again washed twice with wash buffer; the OPD solution was added, and an end point measurement was taken at 450 nm. Cells that were not transfected with Env trimers expressing plasmids were taken as negative controls. The schematic depiction of the experimental setup is shown in Figure 7C.

Competition of Bt-Trp3 with gp120-Binding Proteins

Transfected cells were pretreated with each of F105, gp120 (purified in the laboratory), 2G12, CVN (purified in the laboratory), and DAVEI-L4-3Trp proteins at 1000, 100, 10, and 0 nM for 45 min at 37 °C followed by two buffer washes. Biotinylated Trp3 (200 nM) was added to the cells (Figure 8A). The concentration of biotinylated Trp3 was chosen on the basis of prior analysis to achieve ∼80% binding to the transfected cells. Plates were washed twice with buffer, 15 min each. Streptavidin HRP (dilution factor of 1:3000; Anaspec, 0668) was added to the cells, and the cells were incubated for 45 min. Plates were washed twice with buffer; the OPD solution was added to the plate and the end point absorbance measured at 450 nm.

Figure 8.

Binding of biotinylated Trp3 peptide with the JRFL Env(−)ΔCT trimer expressed on the HEK293T cell surface in the presence of various mAbs and other proteins. (A) Biotinylated Trp3 peptide competed against gp120 and gp120 specific antibodies, including CVN and DAVEI-L4-3Trp protein, in a cell-based ELISA. Bt-Trp3 peptide binding was competed by DAVEI-L4-3Trp protein but not by F105, gp120, 2G12, or CVN. (B) Biotinylated Trp3 peptide competed against the gp120–gp41 interface specific antibody (35O22) and gp41 specific antibodies (4E10 and 10E8) in a cell-based ELISA. 35O22, 4E10 and 10E8 competed with Bt-Trp3 peptide binding (n = 3).

Competition of Bt-Trp3 with gp41 and Trimer Specific Antibodies

Transfected cells were pretreated with each of the 4E10, 10E8, and 35O22 antibodies at 1000, 100, 10, and 0 nM for 45 min at 37 °C followed by two buffer washes. F105 was used as a positive control; 200 nM biotinylated Trp3 was added to the cells (Figure 8B). The plate was washed twice with buffer, 15 min each. Streptavidin HRP (dilution factor of 1:3000; Anaspec, 0668) was added to the cells, and the cells were incubated for 45 min. The plate was washed twice with buffer; the OPD solution was added to the plate and the end point absorbance measured at 450 nm.

Competition of DAVEI-L4-3Trp-Induced Virolysis with gp41 Specific mAbs

Purified BaL.01 virus was incubated with 50 nM DAVEI-L4-3Trp protein in the presence of various concentrations of F105, 4E10, 10E8, and 35O22 for 30 min at 37 °C. The incubated mixture was centrifuged for 2 h at 4 °C and 21130g. The top 100 μL of supernatant was carefully removed without disturbing the virus pellet and then further diluted [1:1 (v/v)] in PBS containing 1% BSA and 1% Triton X-100. This sample was used to conduct the sandwich p24 ELISA as described above. Viruses treated with PBS and 1% Triton X-100 were used as negative and positive controls, respectively. A p24 standard (50 ng) was also used as a positive control in each experiment. Experiments were performed in triplicate, and the percentage of p24 released as compared to the positive control was plotted against the antibody concentration used. Data were plotted using Origin version 8.1 for the determination of IC₅₀.

Production of Viruses with Variable Numbers of Trimeric Spikes

The number of BaL.01 trimeric spikes was varied on the viruses produced by transfecting varying amounts of the envelope plasmid that encodes BaL.01 gp160 (4, 0.4, and 0 μg), without changing the amount of backbone plasmid corresponding to envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg).¹⁶ As a negative control, we produced viruses without the BaL.01 gp160 trimer as well as VSV viruses. Viruses were purified as described above. Because the amount of plasmid corresponding to envelope-deficient pNL4-3 Luc⁺ Env⁻ was not varied, we expected that each virus would incorporate the same amount of p24 and other proteins inside the virus particles with varying densities of trimers on their surface. This was validated by comparing the level of p24 in all of the viruses prepared. Initially, all viruses were separated by gradient centrifugation and pooled to yield equal volumes. Assuming each virus produced contains 1500 p24 molecules as reported previously,²¹ the p24 level was quantified for all BaL.01 and VSV viruses produced. Recombinantly purified p24 (50 ng) was used for quantitation of p24 levels in viruses. The amount of p24 on each virus type was taken as a standard for comparing the number of viruses for performing direct-binding virus-based ELISAs.

Virus-Based ELISA

Each virus type was coated onto an ELISA plate. The dilution factor for BaL.01 gp160 virus (4, 0.4, and 0 μg) and VSV was 1:20 (calculated by comparing the p24 signal) in all cases. ELISA plates were sealed with parafilm and kept at 4 °C overnight. Plates were blocked with 3% BSA (pH 7.5) the following day for 2 h at room temperature, and then assays were performed.

Binding of VRC01 to Trimers on BaL.01 Virus

Varying concentrations of the VRC01 antibody (AIDS Repository, catalog no. 12033) diluted in 1× PBS (0.5% BSA) were added to the ELISA plates and incubated for 30 min. Plates were washed with 1× PBS for 5 min. Viruses were fixed with 100 μL of 1% paraformaldehyde for 15 min onto the ELISA plates. The plate was washed twice, 15 min each with 1× PBS. VRC01 was conjugated to anti-human HRP (dilution factor of 1:3000; Millipore, AP101P), followed by OPD, and end point measurements were taken at 450 nm.

Direct-Binding BaL.01 Virus with Bt-Trp3 Peptide

Blocking buffer was discarded, and varying concentrations of biotinylated Trp3 peptide were added to the ELISA plates and the plates kept at room temperature for 30 min. Plates were washed with 1× PBS for 5 min. Viruses were fixed with 1% paraformaldehyde for 15 min onto the ELISA plates. The plate was washed twice, 15 min each with 1× PBS. Streptavidin HRP (dilution factor of 1:3000; Anaspec, 0668) was added to the fixed viruses, and the viruses were incubated for 30 min. The plate was washed twice with buffer, followed by addition of OPD, and the end point absorbance measured at 450 nm.

Cyanovirin Fusion Models, Energy Minimization, and Distance Assessment

The fusion models for DAVEI-L2 were made using the Maestro 10.3 interface, provided by Schrodinger Inc. The 1IIY structure was downloaded from the Protein Data Bank. The additional histidine tag, linker, and MPER sequence were added to the C-termini of the 1IIY structure using the BUILD toolbar provided by the Maestro 10.3 interface. The growth direction was set on default mode from the N-to-C direction, with trans-joining geometry. The secondary structure was set in extended mode for the linker and the histidine tag addition and an α-helical conformation for the addition of MPER sequence. The α-helical conformation of MPER was chosen on the basis of a Macromodel Conformational Search using the Maestro 10.3 interface (Schrodinger Inc.). Fusion models were subjected to energy minimization using the Macromodel OPLS-2005 and the Polak–Ribiere conjugate gradient algorithm. The minimization was terminated when the RMS energy gradient reached a value of 0.05 kJ mol⁻¹ Å⁻¹.

RESULTS

Role of CRAC Sequence for DAVEI-L4 Lytic Function

The CRAC sequence located at the C-terminus of MPER is known to associate with cholesterol,^22–25 which is a crucial component of the viral membrane. We constructed two derivatives of DAVEI-L4, namely, DAVEI-L4(ΔCRAC) and DAVEI-L4(ΔMPER), to investigate the role of CRAC sequence in DAVEI-induced lysis (Figure 1B). DAVEI derivatives bound at nanomolar concentrations to monomeric gp120 (Supporting Information 1A) as well as to BaL.01 gp120 trimers expressed on viruses (Figure 1C). Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of BaL.01 trimers produced using envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg) were used as a negative control for the experiment. DAVEI derivatives did not bind to viruses devoid of trimers produced using envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg) (data not shown). Infection inhibition assays showed that both DAVEI-(ΔMPER) (IC₅₀ = 1.4 ± 0.3 nM) and DAVEI(ΔCRAC) (IC₅₀ = 1.1 ± 0.2 nM) inhibited pseudoviral infection to a degree comparable to those of CVN protein (IC₅₀ = 0.9 ± 0.2 nM) and DAVEI-L4 (IC₅₀ = 1.2 ± 0.3 nM) (Figure 1D). In contrast, while viral stocks treated with DAVEI-L4(ΔCRAC) exhibited a virolytic property (EC₅₀ = 28.3 ± 2.1 nM) comparable to that of DAVEI-L4 (EC₅₀ = 29.8 ± 0.9 nM) (Figure 1E), CVN and DAVEI-L4(ΔMPER) did not show any virolysis (Figure 1E). These data indicated that the CRAC sequence and its potential interaction with pseudovirus membrane cholesterol are not required for DAVEI-induced virolysis.

Alanine Scanning in Hydrophobic Tryptophans for Lytic Investigation

Apart from the CRAC domain, the hydrophobic tryptophan residues of MPER have also been reported to have an affinity for cholesterol-rich membranes.^26–28 We asked whether these tryptophans may be involved in a similar interaction with the viral membrane as a component of DAVEI-induced lysis. We introduced alanine scanning mutations in place of the tryptophan residues of the DAVEI-L4(ΔCRAC) construct. Tryptophans were mutated to alanines starting from the C-terminus (Figure 2A, colored red). DAVEI derivatives bound to both monomeric YU2 gp120 (Supporting Information 1B) and BaL.01 gp120 trimers expressed on viruses (Figure 2B). Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of BaL.01 trimers produced using envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg) were used as a negative control for the experiment. DAVEI derivatives did not bind to Env-deficient viruses (data not shown). All DAVEI-L4(ΔCRAC) derivatives inhibited pseudovirus cell infection: IC₅₀ = 0.4 ± 0.1 nM for the single mutant, IC₅₀ = 0.45 ± 0.1 nM for the double mutant, IC₅₀ = 0.36 ± 0.1 nM for the triple mutant, and IC₅₀ = 0.78 ± 0.3 nM for the quadruple mutant. These potencies were comparable to that of DAVEI-L4-(ΔCRAC) (IC₅₀ = 1.1 ± 0.2 nM) (Figure 2C). Substitution of tryptophan with alanine did not show any significant effect on lytic function for the single mutant (EC₅₀ = 46.8 ± 5.9 nM), double mutant (EC₅₀ = 49.7 ± 5.5 nM), or triple mutant (EC₅₀ = 43.1 ± 3.6 nM) (Figure 2D). When all four tryptophans were mutated to alanine, we observed a partial loss of lytic function, with the EC₅₀ values decreasing fractionally by 3-fold for the quadruple mutant (EC₅₀ = 122.2 ± 11.6 nM) (Figure 2D). These data argued that possible interactions of tryptophan(MPER) with the membrane do not play an essential role in DAVEI-induced HIV-1 lysis.

Minimal MPER_(DAVEI) Sequence Required for DAVEI Lytic Function

The lytic functionality of DAVEI-L4-(ΔCRAC) and tryptophan to alanine mutants suggested that other residues must be required. We used serial truncations to minimize the MPER region to identify residues important for virolysis. Three DAVEI-L4(ΔCRAC) derivatives were constructed in which a stop codon was introduced after each tryptophan (Figure 3A). MPER-truncated peptides with the corresponding lengths were used as controls for the virolysis and infection inhibition analyses of the truncated proteins (Figure 3B). DAVEI derivatives bound to YU2 gp120 monomers (Supporting Information 1C) as well as to trimers (BaL.01) expressed on viruses (Figure 3C). Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control, and viruses devoid of BaL.01 trimers produced using envelope-deficient pNL4-3 Luc⁺ Env⁻ (8 μg) were used as a negative control for the experiment. No binding was observed on Env-deficient viruses (data not shown). Infection inhibition assays showed that all of the truncated DAVEI proteins inhibited pseudoviral infection at low nanomolar concentrations: IC₅₀ = 0.8 ± 0.2 nM for DAVEI-L4-1Trp, IC₅₀ = 2.3 ± 0.1 nM for DAVEI-L4-2Trp, and IC₅₀ = 2.2 ± 0.4 nM for DAVEI-L4-3Trp (Figure 3D). Importantly, DAVEI-L4(ΔCRAC) (EC₅₀ = 28.3 ± 2.1 nM) and DAVEI-L4-3Trp (EC₅₀ = 36.1 ± 5.1 nM) proteins were fully virolytic (Figure 3E). In addition, DAVEI-L4-2Trp was substantially lytic (Figure 3E, dark cyan line). In contrast, a complete loss of virolytic function was observed with DAVEI-L4-1Trp (Figure 3E). While MPER and Trp4 peptide showed similar low-level though experimentally detectable lytic activity at the highest concentration (2 μM), the Trp3, Trp2, and Trp1 peptides were all fully nonlytic (Figure 3E, dotted lines). Overall, these experiments showed that DAVEI-L4-3Trp protein is equipotent to DAVEI-L4 in the virolytic assay, while the corresponding Trp3 peptide (DKWASLWNW) is not. Hence, we were able to decrease the length of MPER significantly in the DAVEI molecule without loss of virolytic function. The results suggest that non-Trp residues in the MPER-derived 3Trp sequence of DAVEI-L4-3Trp are important for DAVEI-induced virolysis.

Minimal (G4S)_x Linker Required for Virolytic Function

Prior investigation showed that DAVEI molecules having four and eight repeats of glycine₄-serine linker sequence were both active in virolytic assays. We investigated if reducing the size of the (Gly₄Ser)₄ linker would have any effect on virolysis. Two shorter DAVEI constructs were cloned with smaller linker lengths, namely, DAVEI-L2 with two (glycine₄-serine) repeats and DAVEI-L0 with no linker (Figure 4A). Cloned constructs were recombinantly produced, purified, and tested for infection inhibition and lytic activities. Both DAVEI derivatives with shorter linkers bound to monomeric YU2 gp120 (Supporting Information 1D) as well as to BaL.01 trimers on viruses (Figure 4B) but not to Env-deficient viruses (data not shown). Monomeric gp120 (50 ng) directly immobilized onto an ELISA plate was taken as a positive control for this experiment. Both purified proteins inhibited infection at nanomolar concentrations (Figure 4C) (IC₅₀ = 4.3 ± 1.7 nM for DAVEI-L0 and IC₅₀ = 2.1 ± 0.3 nM for DAVEI-L2). Virolysis analysis showed that DAVEI-L2 protein (EC₅₀ = 182.6 ± 28.5 nM) was active in triggering p24 release, while DAVEI-L0 was not (Figure 4D). These results argue that there is a minimal CVN-MPER spacing required for DAVEI function that is satisfied by the L2 linker.

Competition of DAVEI-Induced Virolysis by MPER and Trp3 Peptides

In our earlier study, we found that free MPER peptide was able to compete with DAVEI-L4-induced virolysis at low nanomolar concentrations.⁸ In the investigation presented here, we were able to significantly reduce the size of MPER from 20 amino acids (DAVEI-L4) to nine amino acids (DAVEI-L4-3Trp) with retention of infection inhibition and virolytic function. Because the size of MPER was significantly reduced, we examined if shortening the length of MPER on the DAVEI-L4 molecule caused a change in competition activity. We compared the effects of MPER peptide and Trp3 peptide on the virolytic activity of DAVEI-L4-3Trp protein. Both free MPER peptide and Trp3 peptide competed with DAVEI-L4-3Trp protein (Figure 5). While free MPER peptide (16.8 ± 1.6 nM) was slightly more potent than Trp3 peptide (52.6 ± 5.4 nM), strong competition of virolysis was nevertheless seen for both peptides, implying the possibility of a common virolytic mechanism of DAVEI-L4-3Trp and DAVEI-L4 proteins. Inhibition of DAVEI-induced virolysis by MPER peptide and Trp3 peptide (Figure 5) suggested that these peptides bind either to the DAVEI molecule or to the virus to prevent virolytic action of DAVEI.⁸ However, we observed no direct binding between DAVEI and either of the lysis competitors, MPER peptide or Trp3 peptide (Supporting Information 6), suggesting a direct interaction of these peptides with virus.

Figure 5.

Free MPER peptide competition of virolysis induced by DAVEI derivatives. Free MPER and Trp3 peptide were able to compete out the virolysis exhibited by DAVEI-L4 and DAVEI-L4-3Trp proteins; 50 nM DAVEI-L4 and DAVEI-L4-3Trp proteins were incubated with BaL.01 virus in the absence or presence of serial dilutions of both MPER and Trp3 peptide. IC₅₀s for inhibition of DAVEI-L4-3Trp-induced virolysis for MPER peptide and Trp3 peptide were 16.8 ± 1.6 and 52.6 ± 5.4 nM, respectively, measured using a p24 ELISA (means ± SD; n = 3).

Spatial Comparison between DAVEI and Env Protein-Binding Regions

CVN is known to have one high-affinity binding site and another low-affinity site.²⁹ At nanomolar concentrations, CVN binds to oligosaccharides only through its high-affinity site, while at high micromolar to millimolar concentrations the oligosaccharides are bound through both sites, leading to divalent protein−carbohydrate interaction.²⁹ Because our assays were conducted at low nanomolar concentrations, we expected the CVN component of the DAVEI fusion to interact via its high-affinity binding site localized around residues Asn42, Asn53, Thr57, and Gln78 of CVN. A distance estimate of 70 Å was made from the high-affinity binding site²⁹ of cyanovirin-N to the center of mass of MPER present in the lytically active DAVEI-L2 molecule (Figure 6A). This distance between the functional domains in DAVEI-L2 can be compared (Figure 6B) with the approximate distance of 99 Å between the center of mass of CVN-binding glycans of gp120 (approximately around N332) and the membrane-proximal C-terminus of gp41 (residue D664) in the gp120−gp41 protomer of the recently reported cryo-EM structure of native, fully glycosylated, cleaved HIV-1 JR-FL EnvdCT trimer [Protein Data Bank (PDB) entry 5FUU].³⁰ This correlation is consistent with the view from the DAVEI mutant results that the DAVEI MPER region is more likely to interact with components in the Env trimer spike protein on the virus than with the virus membrane itself.

Figure 6.

Spatial comparison of the DAVEI fusion with the ectodomain of the HIV-1 JRFL gp120–gp41 protomer. (A) Simulated structure of DAVEI-L2 constructed computationally by fusion of the CVN molecule (PDB entry 1IIY) with an extended (glycine₄-serine) linker and MPER using Maestro version 10.3. (B) Cryo-electron microscopy structure of the gp120–gp41 protomer in the cleaved wild-type HIV-1 JR-FL Env trimer complex (PDB entry 5FUU), with distance measurements made from the center of CVN-binding glycans in gp120 (N332) to the C-terminus of the gp41 ectodomain close to the membrane as shown in the structure (D664).

Binding of the Biotinylated Trp3 Peptide to the Envelope Trimer

Mutagenesis (Figures 1 and 2), truncation (Figure 3), competition of virolysis (Figure 5), distance assessment (Figures 4 and 6), and potential membrane surface binding site analysis (Supporting Information 3) suggest a lack of interaction between the MPER component of the DAVEI molecule and the viral membrane. We examined the alternative that the Trp3 component of DAVEI binds to the envelope protein trimer. A cell-based ELISA was used to assess the direct binding of the biotinylated Trp3 peptide and the envelope trimer expressed on the HEK293T cell surface. The biotinylated Trp3 peptide was used as a substitute for DAVEI-L4-3Trp protein for this assay to rule out interaction of CVN_(DAVEI) with glycans present on Env or on other proteins on the HEK293T cell surface. Bt-Trp3 did not inhibit BaL.01 infection at concentrations of up to 10 μM (Supporting Information 5). Unlike MPER, biotinylated Trp3 peptide was not virolytic (Supporting Information 2). HEK293T cells were transfected with a trimer-coding plasmid, and the expression of Env protein on the cell surface was confirmed by F105 binding (mAb that binds to the CD4-binding site of gp120). F105 bound to cells transfected with the envelope trimer encoding plasmid [JRFL Env(−)ΔCT] in a dose-dependent manner (EC₅₀ = 4.5 ± 0.3 nM), but not to nontransfected cells, thus validating Env expression on the cell surface (Figure 7B). Similarly, mAbs such as 2G12, 35O22, 4E10, and 10E8 bound to cells that expressed Env in a dose-dependent manner (Figure 7B), with EC₅₀ values of 2.7 ± 0.5, 6.8 ± 1, 9.6 ± 0.5, and 6.7 ± 0.7 nM, respectively, confirming proper exposure of gp120 and gp41 subunits on the cell. Biotinylated Trp3 peptide bound to JRFL Env(−)ΔCT-expressing HEK293T cells (EC₅₀ = 129 ± 14 nM) but not to nontransfected cells (Figure 7D). The results argue that the MPER_(DAVEI) region in DAVEI interacts with the trimeric spike protein, but not with the membrane alone, to elicit virolysis.

Competition of Biotinylated Trp3 Peptide Binding to Env Spike by gp120 Specific Antibodies

We examined the competition of Bt-Trp3 binding by MPER specific mAbs (4E10 and 10E8), a gp120 outer domain mAb (2G12), a mAb specific for the gp120−gp41 interface (35O22), and a CD4 site specific mAb (F105) in a cell-based ELISA. Direct binding of Bt-Trp3 peptide to these mAbs (2G12, F105, 35O22, 4E10, and 10E8) was examined before a competition cell-based ELISA was performed to ensure that the competition is due to binding of the antibody to the trimeric spike but not to the Bt-Trp3 peptide. This was important because Bt-Trp3 has a sequence similar to that of the MPER region of gp41. As expected, antibodies F105, 35O22, and 2G12 did not bind to Bt-Trp3 peptide (data not shown). In addition, Bt-Trp3 did not bind to the gp41 MPER-directed antibodies, 4E10 and 10E8 (Supporting Information 4A). In contrast, Bt-Trp3 did bind to 2F5 (Supporting Information 4A), consistent with the presence of the epitope for 2F5 in Bt-Trp3 (Supporting Information 4B). Therefore, 2F5 was not used for competition cell-based ELISAs.

Competition cell-based ELISA experiments showed that gp120 protein as well as gp120-binding mAbs (F105 and 2G12) do not compete with binding of Bt-Trp3 to trimeric spike (Figure 8A). DAVEI-L4-Trp3 protein strongly competed with Bt-Trp3 binding, but CVN protein did not (Figure 8A). DAVEI-L4-3Trp protein competing with Bt-Trp3 binding confirms that Bt-Trp3 peptide and the 3Trp moiety of DAVEI-L4-3Trp protein can both interact with gp41 in a similar fashion. Importantly, MPER specific mAbs strongly competed with binding of Bt-Trp3 to the trimeric spike [4E10, >65% inhibition at 10 nM; 10E8, >60% inhibition at 10 nM (n = 3)] (Figure 8B). Similarly, the gp120−gp41 interface specific antibody (35O22) competed with binding of Bt-Trp3 to the trimeric spike, as well (35O22, >60% inhibition at 100 nM) (Figure 8B).

Inhibition of DAVEI-L4-3Trp-Induced Virolysis by gp41 Specific Antibodies

The cell-based competition ELISA performed on JRFL Env(−)ΔCT suggested gp41 as an interactor for biotinylated Trp3 peptide. Because 35O22, 4E10, and 10E8 competed for binding of biotinylated Trp3 peptide with trimeric spike on JRFL Env(−)ΔCT, we examined the competition of DAVEI-L4-3Trp-induced virolysis in BaL.01 viruses. JRFL gp120 shows an 89.26% sequence identity with the BaL.01 gp120 region (Supporting Information 7), with all eight CVN specific glycosylation sites being conserved in both BaL.01 and JRFL Env(−)ΔCT (Supporting Information 7), thus justifying the comparison. In addition, both BaL.01 and JRFL Env(−)ΔCT show 100% sequence identity in the MPER region (Supporting Information 7) as well as 35O22 binding epitopes (N88, N230, N241, and N625). F105 was taken as a negative control for the virolysis assay. The results obtained showed that 35O22, 4E10, and 10E8 compete with DAVEI-L4-3Trp-induced virolysis in a dose-dependent fashion (IC₅₀ = 196.5 ± 22.4 nM for 35O22, IC₅₀ = 85.5 ± 6.6 nM for 4E10, and IC₅₀ = 25.8 ± 15.2 nM for 10E8) (Figure 9A).

Figure 9.

Inhibition of DAVEI-L4-3Trp-induced virolysis by gp41 specific antibodies. (A) DAVEI-L4-3Trp protein (50 nM) was incubated with BaL.01 virus in the absence or presence of serial dilutions of 35O22, 4E10, and 10E8 antibodies. The IC₅₀ values for inhibition of DAVEI-L4-3Trp-induced virolysis for 35O22, 4E10, and 10E8 were 196.5 ± 22.4, 85.5 ± 6.6, and 25.8 ± 15.2 nM, respectively (means ± SD; n = 3). (B) gp120–gp41 protomer bound to the 35O22 antibody, produced by a structural overlay of the crystal structure of the ectodomain of JRFL EnvΔCT (PDB entry 5FUU) with the 35O22-bound crystal structure of SOSIP 664 (clade G x1193.c1, PDB entry 5FYJ). The gp120 subunit is colored gray, and gp41 is colored wheat. Cyanovirin specific glycosylation sites are colored green (N160, N339, N386, N392, N448, N136, and N332), and 35O22 specific glycosylation sites (N88, N241, and N625) are colored red. N230, a common glycosylation site shared by CVN and 35O22, is colored orange, and the tentative binding site of 4E10 and 10E8 (D664) is colored blue. Figures were produced using Pymol graphics version 1.4.1.

We inspected the crystal structure of one of the protomers of the JRFL ectodomain trimer (PDB entry 5FUU) to ascertain how inhibition of virolysis might be exhibited by these antibodies (Figure 9B). On the basis of binding site analysis of mAbs, we envision that 4E10 and 10E8 inhibit virolysis because of steric hindrance conferred by these antibodies to MPER_(DAVEI) when they interact with the gp41 subdomain. Inhibition by the 35O22 mAb could derive from steric hindrance of MPER_(DAVEI) binding, or binding of the CVN component of DAVEI, or both simultaneously, because 35O22 and CVN share a common glycan-binding site on gp120, namely N230 (Figure 9B, colored orange). If DAVEI engages N230 for virolytic function, 35O22 might inhibit virolysis by inhibiting CVN engagement. Similarly, the bulky light and heavy chains of 35O22 could sufficiently block engagement of MPER_(DAVEI) to prevent virolysis, as well (Figure 9B). The uncertainty with 35O22 notwithstanding, these overall results reinforce the conclusion that gp41 interacts with the 3Trp moiety of DAVEI-L4-3Trp protein.

Validation of Virus with Varying Numbers of Trimeric Spikes on Surfaces

Viruses with varying amounts of BaL.01 trimeric spikes on their surfaces were purified (Figure 10A) and tested for p24 content. All virus types (BaL.01 and VSV) contained equal amounts of p24 (Supporting Information 10A) at corresponding dilutions. The infectivity of BaL.01 viruses decreased as the number of spikes on the virus surfaces was decreased by lowering the plasmid concentration during transfection (Supporting Information 10B). VSV viruses had infectivity that was 33% higher than that of BaL.01 at 4 μg of plasmid transfection. No infectivity was observed for BaL.01 viruses produced in the absence of the trimer-coding plasmid.

Figure 10.

Direct binding of VRC01 and biotinylated Trp3 peptide to viruses with varying amounts of trimeric spikes. (A) Cartoon representation showing three different types of BaL.01 viruses produced each with varying numbers of trimeric spikes on their surfaces and VSV virus, as well. (B) Interaction of each virus type with the VRC01 antibody. VRC01 binds to BaL.01 viruses having trimers on their surfaces (>50% binding at a concentration of 10 nM) but does not bind to viruses without trimers. No binding of VRC01 was seen with VSV viruses. (C) Interaction of each virus type with Bt-Trp3 peptide. Bt-Trp3 bound to BaL.01 viruses with trimers on their surfaces (∼100% binding at a concentration of 1000 nM with BaL.01 with a maximal number of spikes). A reduction in the number of spikes weakened binding of Bt-Trp3 (∼55% binding at a concentration of 1000 nM with BaL.01 with a reduced number of spikes). No significant binding of VRC01 was seen with BaL.01 viruses with no trimeric spikes or VSV viruses (n = 3).

Virus ELISA with VRC01 and Bt-Trp3 Peptide

In addition to infectivity, recombinantly purified BaL.01 and VSV viruses (Figure 10A) were assayed with the VRC01 antibody to confirm the differential in the expression of functional HIV-1 spikes. VRC01 bound to BaL.01 viruses containing trimeric spikes in a dose-dependent fashion (∼50% binding at a concentration of 10 nM) but not to BaL.01 viruses without trimers or VSV viruses (Figure 10B). Similar experiments were performed with Bt-Trp3 peptide. Bt-Trp3 also bound to BaL.01 viruses with trimers on their surfaces (∼100% binding at a concentration of 1000 nM to BaL.01 viruses with the maximal number of spikes) (Figure 10C). The reduction in the number of spikes reduced the level of binding of Bt-Trp3 (∼55% binding at a concentration of 1000 nM to BaL.01 viruses with a reduced number of spikes) (Figure 10C). No significant binding of Bt-Trp3 was observed for BaL.01 viruses with no trimeric spikes or for VSV viruses (Figure 10C). The lack of competition for binding of Bt-Trp3 to virus Env by gp120 (Figure 8) argues that Bt-Trp3 does not bind to the gp120 components of Env trimers. Hence, the binding of Bt-Trp3 to virus in proportion to the amount of Env spike content, combined with the results from a competition cell-based ELISA, further confirms that binding of the 3Trp moiety occurs with the gp41 domain of trimeric spike.

DISCUSSION

DAVEIs initially designed as fusions of CVN and MPER are potent inhibitors of HIV-1 cell infection and are capable of inducing lytic inactivation of both fully infectious and BaL.01 pseudotype virus.⁸ Because a mechanistic understanding of DAVEI-induced virolysis has been lacking, we used a protein engineering approach to investigate the role of the MPER domain that is required for the virolytic function of the DAVEIs. We showed that both the CRAC sequence and tryptophan residues in MPER can be compromised without loss of DAVEI function, thus making it unlikely that membrane interaction is required for virolysis by DAVEI. Additionally, spatial assessment of CVN to MPER distance in DAVEI-L2 and glycan-to-membrane distance in the gp120−gp41 protomer of trimeric Env(+)ΔCT suggested that if MPER_(DAVEI) interaction with the membrane were required for virolysis, the trimeric spike would require bending for DAVEI-L2MPER to interact with the membrane while CVN binds to gp120 glycans. Because no evidence for such bending of the trimer toward the membrane has been reported to date, we conclude that MPER_(DAVEI) interacts with spike protein. These results, combined with results from a direct-binding virus ELISA, a competition cell-based ELISA with antibodies, and competition virolysis experiments, argue in favor of gp41 as the binding site of MPER_(DAVEI). While the CVN domain of DAVEI is the source of infection inhibition activity by gp120 interaction, it is association of MPER_(DAVEI) with the gp41 subunit that causes virolysis. Furthermore, we determined that the N-terminal nine residues in MPER, DKWASLWNW, provide a minimal molecular signature for virolytic function.

The proposed mechanism of MPER in DAVEI as a gp41-interacting domain contrasts with recent observations with MPER-derived peptides or gp41 fusions containing MPER sequences. Membrane-bound gp41 proteins containing fusion peptide and MPER, CHR, and NHR groups have the capacity to induce content leakage from cholesterol-rich liposomes.³¹ In fact, MPER peptide alone has the capacity to induce leakage.^10–12 Apart from HIV-1 virus, fusion peptide analogues in other viruses such as influenza have also shown vesicular lytic properties,³² preferentially in the trimeric form.³³ The membrane disruptive property for both fusion peptide and MPER is attributed to their hydrophobic amino acid sequences. MPER-induced membrane disruption and fusion depend on the presence of an unusually high percentage of tryptophan residues.^10–12 These tryptophan residues are segregated along the axial length of α-helical MPER to form interactions with the membrane.^26–28 The MPER Trp residues are known to be involved in infectivity³⁴ as well as fusion pore formation and expansion.³⁵ In addition, the cholesterol recognition motif, also known as the CRAC sequence (LWYIK) located at the MPER C-terminus,^22–25 has strong affinity for cholesterol, an essential component of the HIV-1 envelope membrane.^22,24,25 Mutations or truncations in the CRAC region of virus interrupt interaction with cholesterol and lead to loss of fusion activity.^36,37

The results of this work leave open the question of what specific epitopes in gp41 are the lysis-enabling targets for MPER_(DAVEI) binding. MPER has been reported to have a tendency to self-associate.^31,38 In addition, a recent mutagenesis study by Yi et al.³⁹ reported the possibility of interaction of MPER with the fusion peptide and NHR domain. The observations described above allow the possibility that MPER_(DAVEI) could bind to gp41 MPER, FP, or NHR to induce virolysis. Hypothetically, interactions of MPER_(DAVEI) with regions of gp41 concurrent with interaction of CVN with gp120 could cause conformational distortion, inducing membrane stress and hence lysis. The “spring-loaded” conformation of gp41 in the unliganded state has long been recognized to hold strong potential energy, which is released during the fusion process.^6,40 It may be hypothesized that the strong potential energy of the metastable trimer is exploited by DAVEI molecules to elicit virolysis when gp120 and gp41 are simultaneously engaged. In this hypothesized view, interactions of DAVEI with gp41 can be sterically hindered by antibodies such as 4E10, 10E8, and possibly 35O22, all of which inhibit virolysis (Figure 9). Further work is needed to define the interaction pathway of the lysis-inducing DAVEI engagement with gp120 and gp41 sites on the Env spike.

Overall, the results of this work demonstrate that simplifications in both linker and MPER domains can be introduced into DAVEI molecules with retention of lytic function. This leads to the follow-up question of whether the CVN domain of DAVEI also could be simplified and indeed whether other gp120-binding ligands, including smaller molecules, can be substituted. The potential for smaller molecule DAVEI lytic inactivators of HIV-1 remains an intriguing possibility for future investigation.

ABBREVIATIONS

DAVEI

dual-acting virucidal entry inhibitor

ELISA

enzyme-linked immunosorbent assay

HIV-1

human immunodeficiency virus

CVN

cyanovirin-N

MPER

membrane-proximal external region