Covalent fusion inhibitors targeting HIV-1 gp41 deep pocket
- First Online:
- Cite this article as:
- Bai, Y., Xue, H., Wang, K. et al. Amino Acids (2013) 44: 701. doi:10.1007/s00726-012-1394-8
- 410 Views
Covalent inhibitors form covalent adducts with their target, thus permanently inhibiting a physiological process. Peptide fusion inhibitors, such as T20 (Fuzeon, enfuvirtide) and C34, interact with the N-terminal heptad repeat of human immunodeficiency virus type 1 (HIV-1) gp41 glycoprotein to form an inactive hetero six-helix bundle (6-HB) to prevent HIV-1 infection of host cells. A covalent strategy was applied to peptide fusion inhibitor design by introducing a thioester group into C34-like peptide. The modified peptide maintains the specific interaction with its target N36. After the 6-HB formation, a covalent bond between C- and N-peptides was formed by an inter-helical acyl transfer reaction, as characterized by various biophysical and biochemical methods. The covalent reaction between the reactive C-peptide fusion inhibitor and its N-peptide target is highly selective, and the reaction greatly increases the thermostability of the 6-HB. The modified peptide maintains high potency against HIV-1-mediated cell–cell fusion and infection.
KeywordsHIV-1Gp41PeptideSix-helix bundleCovalent inhibitor
Human immunodeficiency virus type-1 (HIV-1) envelope glycoprotein (Env)-mediated virus–cell membrane fusion is a critical step for HIV-1 infection and in vivo propagation (Cai and Jiang 2010; Eckert and Kim 2001). HIV-1 Env is composed of surface unit gp120 and transmembrane unit gp41, which are noncovalently associated with each other, form trimer and decorate on the viral surface as spikes (Eckert and Kim 2001). HIV-1 infection is initiated by the binding of gp120 to the cellular surface receptor CD4 and a co-receptor, CCR5 or CXC4, triggering a cascade of large conformational changes of the gp120/gp41 complex from a native state to a prehairpin fusion intermediate (PHI) state and then to a fusion state. The fusion core formed at the fusion state is a six-helical bundle (6-HB), in which three gp41 N-terminal heptad repeats (NHR) form a trimeric inner core, and three C-terminal heptad repeats (CHR) pack in an antiparallel fashion against the inner NHR trimer (Chan et al. 1997). The energy released by 6-HB formation drives the apposition and subsequent fusion of viral and target cell membranes. Peptides derived from the NHR and CHR can bind to their counterparts in gp41 to form an unproductive hetero 6-HB and prevent fusogenic 6-HB core formation, thus inhibiting HIV-1 host–cell membrane fusion and blocking viral infection (Jiang et al. 1993; Wild et al. 1992, 1994). Representative fusion inhibitors include: the first FDA-approved HIV-1 fusion inhibitor T20 (generic name: enfuvirtide; brand name: Fuzeon®) (Kilby et al. 1998; Lalezari et al. 2003), C34, T1249 (Eron et al. 2004; Lalezari et al. 2005), T2635 (Dwyer et al. 2007) and sifuvirtide (He et al. 2008b). Though highly efficient in drug-naive patients, peptide fusion inhibitors are prone to drug resistance due to the reduced binding affinities with the target in the mutated HIV-1 strains.
Covalent inhibitors have found increasing interests in drug design. (Singh et al. 2011) They act by forming irreversibly covalent adducts with their targets, thus permanently inhibiting the physiological processes (Jenkins et al. 2010; Yi et al. 2011; Jacobs et al. 2007). Covalent inhibitors interact with their targets in two steps. The first step consists of a specific binding that enables the inhibitor recognizing its target, and the second step is a followed covalent bond formation between the reaction group of the inhibitor and the specific site of the target. There is a close relationship between covalent bond formation and the molecular assembly of covalent inhibitors and their targets. The molecular assembly brings the functional groups involved into close proximity, so that the reactants are presented at high molar concentrations and in reachable positions. The assembly and folding also favors the conformational entropy and benefits the reaction thermodynamics. In the process of polypeptide biosynthesis, covalent bonds can be formed between peptides in physiological condition as a result of an inter-molecular acyl transfer reaction. (Sieber and Marahiel 2005; Fischbach and Walsh 2006) This process was recently reproduced in vitro by Ghadiri et al. who demonstrated the biomimetic catalysis of an inter-molecular acyl transfer of an amino acid ester and the bio-mimic synthesis of diketopiperazine based on a four-helical bundle structure (Leman et al. 2007; Wilcoxen et al. 2007; Huang et al. 2008).
In this report, covalent strategy was applied to fusion inhibitor design, based on N36/C34 6-HB and an inter-helical acyl transfer reaction (Bai et al. 2012). The precise site of the C-peptide was modified with a thioester that served as an acyl donor using an orthogonal peptide synthesis strategy. The modified C-peptide retains the ability to interact specifically with N36. After 6-HB formation, a covalent bond between C- and N-peptides was then formed by an inter-helical acyl transfer reaction, as characterized by various biophysical and biochemical methods. The covalent reaction between C-peptide fusion inhibitor and its N-peptide target is highly selective, and the covalent bond greatly increases the thermostability of the 6-HB. The modified peptide retains high potency against HIV-1-mediated cell–cell fusion and infection.
Peptide synthesis, purification and identification
Peptides were synthesized using a CS-Bio 136 automated peptide synthesizer (CS Bio Co., Menlo Park, CA) using a standard solid-phase Fmoc chemistry protocol. All protected amino acids used were purchased from GL Biochem Ltd. (Shanghai, China). Rink amide resin (0.38–0.45 mmol/g, Nankai Hecheng S&T Co. Ltd., Tianjin, China) was used. Coupling of the amino acids (AA) was achieved using O-benzotriazol-1-yl-N,N,N′,N′-tetramethyl-uronium hexafluorophosphate (HBTU, GL Biochem, Shanghai, China) and diisopropylethylamine (DIEA, Acrose) as an activator and base, respectively, in N,N-dimethylformamide (DMF) solution. The Fmoc protective group was removed using 20 % piperidine/DMF. Between each coupling or Fmoc removal step, the resin was washed five times with DMF and three times with dichloromethane (DCM). The carboxy termini were, respectively, amidated upon cleavage from the resin, and the amino termini were capped with acetic acid anhydride. The peptides were cleaved from the resin and de-protected with Reagent K that contained 82.5 % trifluoroacetic acid, 5 % thioanisole, 5 % m-cresol, 5 % water and 2.5 % ethanedithiol. The crude products were precipitated with cold diethyl ether and lyophilized.
For peptides possessing a side chain thioester, Fmoc-l-glutamic acid O-allyl ester [Fmoc-Glu(OAll)-OH] was used at the thioester-modified site. After all amino acids had coupled on the resin in the peptide synthesizer, the O-allyl group was removed manually by 1 eq tetrakis(triphenylphosphine)palladium with 10 eq 5,5-dimethyl-1,3-cyclohexanedione as scavenger in DCM/THF (1:1) solution. Then the resin was washed five times with 0.5 % DIEA in DMF and five times with 1 M sodium diethyldithiocarbamate in DMF. 4 eq 1-Ethyl-3-(3-dimethyllaminopropyl)carbodiimide hydrochloride (EDC) and 4 eq benzyl mercaptan were added to the resin for thioester formation. Finally, the resin was cleaved using Reagent K.
Peptide crude products were purified by preparative reverse-phase HPLC using a Waters preparative HPLC system (PrepLC 4000): gradient elution of 30–50 % solvent B (0.1 % trifluoroacetic acid in 70 % CH3CN/H2O) in solvent A (0.1 % trifluoroacetic acid in H2O) over 60 min at 16 mL/min on a Waters X-bridge C8, 10 μm, 19.5 × 250 mm column. Analytical RP-HPLC was performed on an RP-C8 column (Zorbax Eclipse XDB-C8, 5 μm, 4.6 × 150 mm) with a gradient elution of 5–100 % solvent B in solvent A over 25 min at a flow rate of 1 mL/min. Compounds were detected by UV absorption at 220 nm with a SHIMADZU SPD-10A. All peptides were purified to >95 % purity. The molecular weight of the peptides was confirmed by MALDI-TOF-MS (Autoflex III, Bruker Daltonics).
Circular dichroism (CD) spectroscopy
The N-peptides were incubated with the respective C-peptides at 37 °C for 30 min or 48 h in PBS (phosphate-buffered saline, pH 7.2). Peptides were used at a final concentration of 10 μM. The mixture was then cooled to room temperature. CD spectra were acquired at room temperature (Biologic MOS-450 4.0 nm bandwidth, 0.1 nm resolution, 0.1 cm path length, 4.0 s response time and 50 nm/min scanning speed). The spectra were corrected by subtraction of the solvent blank.
Thermal midpoint analysis was performed to determine the temperature (Tm) at which 50 % of the 6-HB would decompose. The temperature was controlled using a Bio-logic TCU250 system. The final concentration of N- and C-peptides was 1 μM in PBS. CD spectra were monitored at 222 nm between 20 and 90 °C (2 °C/min).
Sedimentation velocity analysis (SVA)
Sedimentation velocity measurements were performed using a Beckman XL-A analytical ultracentrifuge (Beckman Coulter, Fullerton, CA) equipped with an An60 Ti rotor and a photoelectric scanner. Two-sectored aluminum centerpieces and windows were assembled according to the manufacturer’s instructions. N-peptides were incubated with C-peptides at 37 °C for 30 min in PBS (final concentration of N-peptide and C-peptide is 50 μM, respectively). The density of the sedimentation PBS buffer (0.999 g/mL) and partial specific volume [0.751 mL/g for SC35E(SBn)5H9 + N36 and SC29E(SBn)5H9 + N36, 0.757 mL/g for SC22E(SBn)5H9 + N36, and 0.755 mL/g for SC15E(SBn)5H9 + N36] were calculated using the SEDNTERP program (Laue et al. 1992). The 380 μL of the sample and 400 μL of corresponding buffer were loaded into cells. Data at 280 nm were collected at rotor speeds of 3,000 rpm initially and at 50, 000 rpm in a continuous mode with 3 min interval at 20 °C. Sedimentation coefficient distribution [c(s)] and molecular mass distribution [c(M)] were calculated from the data using the SEDFIT program (Schuck 2000).
Analysis of covalent bond formation between C- and N-peptides by tricine-SDS-polyacrylamide gel electrophoresis (tricine-SDS-PAGE)
20 % polyacrylamide gels and a BayGene Mini Cell were used for tricine-SDS-PAGE analysis. The cathode buffer was 0.1 M tricine, 0.1 M Tris and 1 % SDS, and the anode was 0.2 M Tris. N-peptide solutions were incubated with PBS at the indicated concentrations at 37 °C for 30 min before addition of C-peptides (final concentration of N- and C-peptides was 50 μM, respectively). After incubation at 37 °C for 0–48 h, the samples were mixed with Tris-SDS-glycine sample buffer (Invitrogen, Carlsbad, CA) at a ratio of 1:1 and then loaded onto the gels (20 μL/well). Gel electrophoresis was first carried out at 30 V constant voltage at room temperature for 1 h, and then at 150 V constant voltage at room temperature for 2 h. The gel was then stained with Bio-Safe Coomassie stain (Bio-Rad).
HPLC analysis of covalent bond formation between C- and N-peptides
The reactions were carried out in 1.5 mL Eppendorf tubes. A standard solution of 50 μM tryptophan (Trp) was prepared and stock solutions of N- and C-peptides were prepared by dissolving the appropriate peptides in PBS. In a typical experiment, N- and C-peptides were incubated at a 1:1 ratio at room temperature and then PBS and Trp solutions were added. The common final concentrations of the internal Trp standard, and C- and N-peptides were 25 μM. The reaction mixture was incubated at 37 °C for 0–48 h. At the indicated time, a portion of the 50 μL reaction mixture was removed and then quenched immediately with 5 μL 10 % TFA solution. Samples were frozen at −20 °C prior to HPLC analysis. Reverse-phase analytical HPLC was performed using an RP-C8 column (Agilent Zorbax Eclipse XDB-C8, 5 μm, 4.6 × 150 mm) connected to an Agilent 1200 series HPLC system with an automatic sampler. Binary solvent A and B gradients were employed at a flow rate of 1 mL/min with monitoring at 280 nm. The peptide concentrations were determined by comparison to the internal Trp standard.
Cell–cell fusion assay
The inhibitory activity of the peptides against HIV-1 env-mediated cell–cell fusion was measured as described (Wexler-Cohen and Shai 2007; Chen et al. 2011). HL2/3 cells which stably express HIV Gag, Env, Tat, Rev and Nef proteins were obtained from the AIDS Reference and Reagent Program (NIH, Dr. Barbara Felber and Dr. George Pavlakis) and used as target cells. TZM-bl cells that stably express large amounts of CD4 and CCR5 were also obtained from the AIDS Reference and Reagent Program (NIH, Dr. John C. Kappes, Dr. Xiaoyun Wu and Tranzyme Inc.). TZM-bl cells were prepared at a concentration of 5 × 105/mL (50 μL/well) in 96-well plates (Corning Costar) and incubated at 37 °C in 5 % CO2. After a 24 h incubation, 50 μL of HL2/3 cells (2 × 106 cells/mL per well) were added to the TZM-bl cells in the presence or absence of 20 μL test inhibitors. Instead of HL2/3 cells and inhibitors, equal volumes of DMEM medium was used to establish the background signal. HL2/3 and TZM-bl cells were co-incubated for 6–8 h at 37 °C in a 5 % CO2 atmosphere. Afterward, the medium was aspirated and the monolayer was washed carefully with PBS twice, followed by incubations with 50 μL/well of cell culture lysis reagent (Promega Corporation) for 5 min. Suspensions were then transferred to a 96-well white polystyrene plate (Corning Costar) before adding the Promega Luciferase Assay System (Promega Corporation). Fusion was measured by detecting the luciferase activity using a plate reader (Molecular Devices SpectraMax M5). Experiments were performed in duplicate and normalized to the appropriate fusion signal in the absence of inhibitors. IC50 values were computed by fitting a four-parameter nonlinear regression model with Origin software.
HIV pseudoviruses were generated as described previously (He et al. 2008a). Briefly, 293T cells (60–70 % confluent) were co-transfected with 4 μg pSF162 and 4 μg pNL4-3.luc.R-E- plasmid into six-well plate using Lipofectamin2000. The plasmid pSF162 contains the Env of HIV-1SF162 strain, while pNL4-3.luc.R-E- contains an Env and Vpr defective, luciferase-expressing HIV-1 genome. Seventy-two hours after transfection, the culture supernatants were harvested and centrifuged at 2,000 rpm for 5 min. Aliquots were stored at −70 °C until use. The amount of pseudotyped particles was quantitated using the HIV-1 p24 ELISA kit (Retro-Tek, Buffalo, NY). For measuring the inhibitory activity of the test peptide against the infection of HIV pseudovirus, 100 μL U87.CCR5 cells (1 × 104/well) were seeded in 96-well plates and grown overnight. Peptide at indicated concentration was incubated with the pseudovirus (1 ng p24/well) for 30 min at 37 °C. Subsequently, the virus–peptide mixture was transferred to the cells and incubated for an additional 48 h. Cells were washed with PBS and lysed with the lysing reagent included in the luciferase kit (Promega, Madison, WI). Aliquots of cell lysates were transferred to 96-well flat bottom luminometer plates (Costar), followed by addition of luciferase substrate. The luciferase activity was measured in a microplate luminometer (Genios Pro, Tecan, US).
The core crystal structure of the HIV-1 gp41 shows that a positively charged residue Lys574 (the 29th residue of N36, 29Lys) in the pocket-forming region of the NHR (serves as target in current study) interacts with a negatively charged Asp632 (the 5th residue of C34, 5Asp) residue in the pocket-binding domain of the CHR (serves as inhibitor in current study) to form an inter-helical salt bridge (Chan et al. 1997; Jiang and Debnath 2000; He et al. 2007, 2008a). Mutation experiments showed that the salt bridge played critical roles in the 6-HB stability, virus infectivity and fusion inhibitory activity (He et al. 2008a). When 5Asp in the C-peptide inhibitor was mutated to a Glu, the 6-HB stability was further enhanced (He et al. 2008a). The existence of the Asp(Glu)–Lys salt bridge suggests that the side chains of these two residues are in close proximity in the 6-HB and interact specifically with each other. So, ε-amino of 29Lys in N36 was selected as an acyl acceptor, and 5Asp in C34 was modified to function as an acyl donor. We expected that an inter-helical covalent bond would be formed between the modified C34 inhibitor and the target after the 6-HB assembly. A His residue located at the i+4 site of the acyl donor, which has been shown to accelerate the acyl transfer reaction, was also introduced (Wilcoxen et al. 2007; Erben et al. 2011; Bai et al. 2012). To further increase the binding affinity, double salt bridges were introduced into the C34 to increase the helical interaction and the peptide’s solubility (Otaka et al. 2002; Naito et al. 2009; Nishikawa et al. 2009). The resulting covalent inhibitor SC35E(SBn)5H9 (peptide 1) contained the SC35 scaffold, its 9th site (i+4 site) was mutated to His, and the 5th residue of SC35 was mutated to Glu for modification with a side chain benzyl thioester. Shorter covalent fusion inhibitors were also designed for structure–activity relationship study, resulting in 29-mer, 22-mer and 15-mer peptide covalent inhibitors. Unmodified peptides with the same sequences were also synthesized as controls. The designed peptide sequences are presented in Fig. 1.
Thioester-modified C-peptides interact with N36 to form 6-HB complex
CD spectroscopy data of modified C-peptides, N-peptide and their complexes
Tm after 48 h (°C)a
C34+N36 Bai et al. (2011)
SC22E5H9 + N36
Thioester-modified C-peptides and N36 complex SVA results
Sedimentation coefficient (S)
Observed MW (Da)
Calculated 6-HB MW (Da)
SC35E(SBn)5H9 forms covalent adducts with N-peptides after 6-HB formation through an inter-helical acyl transfer reaction
The rate constant and t1/2 of the inter-helical acyl transfer reaction
k2 (M−1 s−1)
The inter-helical acyl transfer reaction could only occur after the 6HB formation
Covalent bond formation greatly increases the thermostability of 6-HB
Covalent peptide fusion inhibitors exhibit high potency against HIV-mediated cell–cell fusion and HIV infection
Anti-fusogenic activity assessment using cell–cell fusion and pseudovirus assays
IC50 (nM, cell–cell fusion)
IC50 (nM, pseudovirus)
1.59 ± 0.10
4.24 ± 1.31
1.02 ± 0.33
3.50 ± 0.50
2.24 ± 0.78
4.33 ± 0.65
5.79 ± 1.27
14.46 ± 8.10
4.68 ± 1.18
4.27 ± 1.39
928 ± 144
1,290 ± 130
27,800 ± 9,790
62,200 ± 14,300
We then tested the anti-HIV-1 activities of the covalent fusion inhibitors using a pseudovirus assay. Consistent with cell–cell fusion assay, the covalent fusion inhibitors showed similar activities against HIV-1 infection in the pseudovirus assay (Table 4). Combined with cell–cell fusion assay results, we concluded that the covalent fusion inhibitors retained the ability to interact with gp41 target and inhibit HIV-1 infection.
Discussion and conclusions
In antiretroviral therapy, drug resistance is the main issue and drug-resistant HIV-1 isolates can emerge rapidly, usually as early as in clinical trials. For HIV-1 fusion inhibitors, the target gp41 NHR, though highly conserved among HIV-1 isolates in drug-naive patients, undergoes fast mutation when treated with fusion inhibitors. These mutations weaken the interaction between fusion inhibitors and the gp41 NHR target. Covalent inhibitors may provide an alternative for peptide fusion inhibitor design to overcome the weak binding between the fusion inhibitor and target associated with drug-resistant mutations.
The designed covalent HIV-1 fusion inhibitor SC35E(SBn)5H9 fulfills two criteria for a covalent inhibitor: it retains the ability to specifically interact with its target, and it can form a covalent bond to link the target thereafter. For a covalent inhibitor, the activity of the reaction groups plays a dual and controversial role. Basically, it should be active enough to facilitate the covalent formation after recognizing the target; on the other hand, reaction groups that are too active hold the potential of nonspecific reaction with unrelated sites. The covalent bond is readily formed between designed covalent inhibitors of different lengths as long as the peptide can specifically interact with the gp41 NHR target. The covalent formation is highly selective since a short peptide contains the same thioester group, but it is too short to recognize its target. Therefore, the short thioester-modified peptide will not form a covalent bond with the NHR target. SC35E(SBn)5H9 also could not react with unassembled peptides in spite of Lys residues inside.
Covalent fusion inhibitors retain their activity in HIV-1 gp41-mediated cell–cell fusion assay and infection assays using HIV-1 psudoviruses. Other reported covalent HIV-1 fusion inhibitors, such as maleimide modified C34 (Jacobs et al. 2007; Yi et al. 2011), showed similar antiviral activity with C34. Those maleimide modified C34 can form a covalent bond with gp41 and permanently attach the viral membrane, as evidenced by a temperature-arrested state (TAS) prime-wash assay. However, the authors mutated the 28Lys on the maleimide modified C34 to prevent the reaction between the maleimide and the 28Lys on C-peptide implying that there may be a problem of selectivity for maleimide to amino groups. SC35E(SBn)5H9, with similar interacting mechanism, should act by permanently attaching to gp41 NHR target and irreversibly inhibit HIV-1 infection. The high activity of SC35E(SBn)5H9 suggests its potential application in antiretroviral therapy; the detailed mechanism studies of its interaction with the gp41 NHR target may justify this class of covalent fusion inhibitors used as tools to study the mechanism of HIV-1 gp41-mediated virus–cell membrane fusion for new fusion inhibitor design.
In conclusion, covalent fusion inhibitors were designed based on a well-studied C-peptide inhibitor, C34, by introducing thioester as the reactive group. These covalent fusion inhibitors retain the ability to recognize their target HIV-1 gp41 NHR. After proper assembly, an inter-helical covalent bond is formed between the inhibitors and their target via an acyl transfer reaction. These covalent fusion inhibitors also retain high potency against HIV-1 gp41-mediated cell–cell fusion and replication. The covalent bond formed between the inhibitor and target may make the inhibitor permanently attach to the target and irreversibly inhibit the HIV-1 gp41 function, and thus have potential to maintain high activity against drug-resistant HIV-1 isolates. However, there were no obvious improvements of the modified peptides in their IC50 in cell–cell fusion and pseudovirus assays with their corresponding unmodified peptides. The velocity of inter-helical acyl transfer reaction should be enhanced in future studies for designing covalent inhibitors with higher potency.
This work was support by the Natural Science Foundation of China Grants (No. 81072581 and No. U0832001) and Key Tech. of National S&T Major Project of Original New Drug Research grant (2012ZX09301-003-001).
Conflict of interest
All the authors declare that they have no conflict of interest.