The HIV gp41 pocket binding domain enables C-terminal heptad repeat transition from mediating membrane fusion to immune modulation

The initiation of HIV infection begins with viral entry into the host cells [1]. Entry is promoted by the trimeric envelope (ENV) glycoprotein, consisting of the two subunits gp120 and gp41, which respectively bind membrane receptors and physically facilitate membrane fusion [2–5]. Gp41 comprises the fusion peptide (FP), the N- and C-terminal heptad repeats (NHR and CHR respectively) and the transmembrane domain (TMD) [6,7]. The six-helix bundle (SHB) is constructed of three NHRs packed into three CHRs, the assembly of which is crucial for successful fusion [8,9]. Therefore, the formation of the SHB is a primary target for HIV fusion inhibitors [10,11].

An additional aspect of HIV infection is immune evasion by the virus, such as immune activation resulting in the generation of additional host cells and by viral reservoirs created as the virus enters a latent state in CD4⁺ T-cells, macrophages and dendritic cells. At a molecular level, the HIV protein Vif negates the blocking of viral replication by the cellular protein APOBEC3G [12,13]. A more recently discovered evasion mechanism is obstructing T-cell receptor (TCR) complex assembly and thus its signalling [14]. The TCR is crucial for the function, proliferation and survival of T-cells [15] promoted by the assembly of the TCRα and TCRβ chains with the three CD3 signalling dimers that leads to an array of signalling cascades that drive gene expression [16]. Gp41 has been shown to utilize membranotrophic regions such as its FP, TMD and loop region to interfere with the TCR signalling cascade via various mechanisms [17–20].

One such segment encompasses the interface between the loop and CHR region and binds the TCR α-chain TMD by intra-membrane interactions between acidic and basic residues [20,21]. Intriguingly, situated in this region at the CHR N-terminus is the conserved pocket binding domain (PBD) comprised of the residues Trp⁶²⁸, Trp⁶³¹ and Ile⁶³⁵, which bind a conserved groove in the NHR, which stabilizes SHB formation [8]. Trp⁶²⁸ and Trp⁶³¹ have been suggested to play a role in immunosuppression, although no direct evidence has been given to support this claim and their exact role is yet unknown [20]. These reports implicate the PBD as a possible player in immune suppression, albeit via an alternative mechanism than direct TCR binding due it its hydrophobic nature. Although immunosuppression by the region housing the PBD is suggested to occur in the membrane via TCR binding [21], SHB formation is proposed to occur in the virological synapse [4,9,22]. For the PBD to function in both cases, it would need to alternate between being buried in the SHB to binding the TCR in the membrane.

We aimed to elucidate the role of the CHR, specifically its PBD, in immunosuppression. Furthermore, we investigated how the PBD alternates between its roles in fusion and as an immunosuppressant. By employing cellular studies together with a biochemical and a biophysical approach, we show that CHR immunosuppression of T-cells, via binding to the TCRα TMD, is mediated by its PBD. Additionally, we reveal that the CHR alternates between the SHB conformation and TCRα TMD binding by membrane association leading to SHB disassembly. Importantly this step is maintained by the PBD that stabilizes CHR α-helical structural integrity in the membrane. These results shed light on the way HIV utilizes its fusion protein in a multifunctional manner and in the process overcomes the changing environments that it encounters during its lifecycle.

Peptides were synthesized on Rink Amide MBHA resin by using the Fmoc strategy as previously described [23]. Several peptides contain a lysine residue at their C-terminus with an MTT side-chain protecting group (Nova biochem AG) that requires a special deprotection step under mild acidic conditions (2×1 min of 5% trifluoroacetic acid (TFA) in dichloromethane (DCM) and 30 min of 1% TFA in DCM). This enables the conjugation of a lipid moiety to the C-terminus. Conjugation of hexadecanoic (palmitic) acid (PA) (C₁₆) (Sigma Chemical) to the C-terminus of selected peptides was performed using standard F moc chemistry. Addition of a rhodamine [5(6)-carboxytetramethylrhodamine] fluorescent probe (Chem-Impex International) to the N-terminus of selected peptides was performed by standard Fmoc chemistry. All peptides were cleaved from the resin by a TFA/DDW/TES (93.1:4.9:2, by vol.) mixture, and purified by reverse-phase HPLC (RP-HPLC) to >95% homogeneity. The molecular mass of the peptides was confirmed by platform LCA electrospray mass spectrometry.

Antigen-specific T-cell lines were selected in vitro [24] from primed lymph node cells derived from C57Bl/6J mice that had been immunized 9 days before with antigen [100 μg of myelin peptide, MOG (myelin oligodendrocyte glycoprotein)- (35–55)] emulsified in complete Freund's adjuvant (CFA) containing 150 μg of Mycobacterium tuberculosis H37Ra (Difco Laboratories). All T-cell lines were maintained in vitro in medium containing IL-2 (interleukin 2) with alternate stimulation with the antigen, every 10–14 days. T-cells specific to MOG-(35–55) [mMOG (mouse myelin oligodendrocyte glycoprotein)-35–55 T-cells] were plated on to round 96-well plates in medium containing RPMI 1640 supplemented with 2.5% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin, 50 μM 2-mercaptoethanol and 2 mM L-glutamine. Each of the 96 wells had a final volume of 200 μl and contained 20×10³ T-cells, 5×10⁵ irradiated (25 Gy) syngeneic spleen cells as antigen-presenting cells (APCs), without or with 5 μg/ml MOG-(35–55). In addition, the relevant HIV-derived peptide was added. In order to exclude interaction between the examined peptides and the MOG-(35–55) antigen, initially the MOG-(35–55) antigen was added to the APCs in a test tube, and in a second test tube the examined peptides were added to the T-cells. After 2 h, the APCs were mixed with the T-cells and were co-incubated for 48 h in a 96-well flat-bottomed plate. The T-cells were pulsed with 1 μCi of [³H] thymidine, with a specific activity of 5.0 Ci/mmol, and after overnight incubation, the [³H] thymidine incorporation was measured using a Matrix 96 Direct Beta Counter (Packard Instruments).

Effector cells were the ENV-expressing cells HL2–3, a HeLa- derived cell line which constitutively expresses the HXB2 strain of the HIV-1 ENV glycoprotein along the Tat protein, and as target cells, TZM-bl cells were used. The fusion of HL2–3 cells with TZM-bl cells was assessed through luciferase expression. The TZM-bl cells were seeded at 2×10⁴ cells/well overnight in a 96-well plates. The medium was then aspirated from each well and replaced with serum-free Dulbecco's modified Eagle's medium (DMEM) containing 40 μg/ml DEAE-dextran. Stock dilutions of each peptide were prepared in DMSO so that each final concentration was achieved with 1% dilution. Upon addition of the peptides, the HL2–3 cells were added to the TZM-bl cells in serum-free DMEM containing 40 μg/ml DEAE–dextran at a 1 : 1 cell ratio. The cells were co-cultured at 37°C for 6 h to allow the fusion to occur. Luciferase activity was analysed using the Steady-Glo Luciferase assay kit (Promega).

Aliquots of 2.5×10⁴ cells of either TZM-bl or HL2–3 cells were distributed on to a 96-well plate in the presence of peptides for 4 h. Wells in the last two columns served as blank (medium only), and 100% survival controls (cells and medium only). After incubation, the 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT) reaction solution (benzene sulfonic acid hydrate and N-methyl dibenzopyrazine methyl sulfate, mixed in a proportion of 50:1), was added for a further 2 h. Optical density was read at a 450-nm wavelength in an ELISA plate reader. For MOG-(35–55) specific T-cells that divide once every 24 h, incubation with the peptides was for 16 h, then XTT reaction solution was added and optical density was read every 1 h for a period of 12 h. The time point with the biggest difference between blank and 100% survival control was chosen for analysis. The percentage of toxicity was calculated relative to the control, 2.5×10⁴ cells in medium with no peptide added.

Circular dichroism (CD) measurements were performed by using an Applied Photo physics spectropolarimeter. The spectra were scanned using a thermostatic quartz cuvette with a path length of 1 mm. Wavelength scans were performed at 25°C; the average recording time was 7 s, in 1-nm steps, in the wavelength range of 190–260 nm and recordings were done in triplicate. Each peptide concentration of 25 μM was tested in Hepes (5 mM, pH 7.4) and a membrane mimetic environment composed of Hepes with 1% lysophophatidylcholine (LPC). Secondary structure analysis was performed with the CDNN software [25].

Large unilamellar vesicles (LUVs) were prepared as described previously [26] from egg phosphatidylcholine (PC), cholesterol (Chol) and egg-yolk sphingomyelin (SM) (Sigma Chemical). A dried film of lipids containing a total of 3 mg of PC/SM/Chol (1:1:1, by vol.) was suspended in PBS and vortex-mixed for 1.5 min. The lipid suspension underwent five cycles of freezing–thawing and then extrusion through polycarbonate membranes with 1 μm and 0.1 μm diameter pores 21 times and added to a PBS−/− solution (no Ca²⁺ or Mg²⁺) reaching an LUV final concentration of 100 μM. Fluorescence spectra were obtained via the Cytation 5 plate reader Lumitron using a black 96-well plate, at room temperature, with excitation set at 467 nm (10-nm slit) and emission scan at 500–600 nm (10-nm slits) and a gain of 130. An 7-nitrobenz-2-oxa-1,3-diazole (NBD)-labelled peptide was added first from a stock solution in DMSO at a final concentration of 0.1 μM and a maximum of 0.25% (v/v) DMSO to a dispersion of PC/SM/Chol (1:1:1) LUVs (100 μM) in PBS−/−. This was followed by the addition of rhodamine (Rho)-labelled peptide (stock in DMSO) in several sequential doses. Fluorescence spectra were obtained before and after addition of the Rho-labelled peptide. The fluorescence values were corrected by subtracting the corresponding blank (buffer with the same vesicles concentration).

A one-way ANOVA test was used where appropriate. P<0.05 was considered significant (*P≤0.05, **P≤0.01, ***P≤0.001). Results are displayed as means±S.E.M.

Multiple studies suggest a role for the gp41 CHR, specifically its PBD, in fusion facilitation and T-cell inhibition [3,8,9,14,20–22]. We attempted to examine the role of the CHR and its PBD in suppression of T-cell activation and the transition from an SHB stabilizer to TCR signalling inhibitor. To this aim we first assessed whether the PBD affects the CHR's suppressive activity by utilizing consecutive and overlapping wild-type and mutant CHR-based peptides (Table 1), and assessing their ability to inhibit T-cell proliferation (Figure 1) upon activation via the TCR. Proliferation was evaluated on a MOG-(35–55)-specific murine T-cell line [mMOG-(35–55) T-cells] as previously described [20]. Significant inhibition is regarded as over 20% T-cell proliferation [19–21]. The results are not due to toxicity as active peptides were not toxic at the levels used in the study and initial toxicity of the peptides was 2-fold higher than concentrations used in proliferation assays (Supplementary Figure S1A). The wild-type peptides consisted of the CHR (termed C34), PBD and T-20, an antiviral drug used in the clinic that contains the CHR but lacks the PBD (Figure 1A). We found that at 5 μM C34 and the PBD conjugated to a palmitic acid side chain (PBD-PA) inhibited T-cell proliferation, whereas T-20 showed no effect (Figure 1B). Initially, at 10 μM the PBD exhibited a low inhibitory effect (Figure 1C). We surmised that the PBD could not sufficiently reach the membrane as it lacks a hydrophobic region that is present in C34 (Figure 1D). It has been shown before that hydrophobic regions can be replaced via lipid conjugation [27]. Therefore, we conjugated the PBD to a C₁₆ PA via an added lysine residue, which highly increased the inhibitory ability of the PBD (Figure 1C). Furthermore, PBD-PA exhibited activity similar to that of C34 (Figure 1B). The suppressive effect of the added PA to PBD and lysine addition was evaluated by assessing PA alone and adding lysine to C34, termed C34K. Results show that both did not significantly affect the suppressive function of the peptides as no significant difference was found between C34 and C34K, and PA alone shows no effect (Figure 1B). However, lipidation may result in a higher peptide concentration at the membrane [28,29]. To assess whether T-cell suppression is due to elevated peptide levels at the membrane due to PA addition, we evaluated a PBD-PA mutant termed PBDmut-PA. PBDmut-PA showed significantly less suppressive ability than PBD-PA (Figure 1C), indicating that high-membrane-bound peptide levels do not affect T-cell activation. The results obtained using C34, PBD and T-20 pinpoint the N-terminus of the CHR as comprising its suppressive abilities.

In order to further examine the PBD's role in immune suppression, we created two C34 peptides with the PBD mutated into either three alanine residues, termed C34-3A or three glycine residues, termed C34-3G, and evaluated their inhibitory effect. As alanine is hydrophobic and greatly stabilizes α-helices, C34-3A is useful to assess the amino acid function without affecting secondary structure. Glycine, on the other hand, adds flexibility to the helix, leading to destabilization of the helical structure, aiding us to evaluate the importance of the secondary structure. Both mutants exhibited a baseline inhibitory effect and significantly reduced C34’s ability to inhibit T-cell proliferation by 50% (Figure 1E), suggesting an important role for the specific amino acids comprising the PBD. Taken together, these results show that the CHR can inhibit T-cell activation and that the PBD plays a significant role in mediating this effect.

The results detailing the PBD's function in immunosuppression are interesting as the accepted primary function of the CHR and the PBD is to create and stabilize the SHB respectively [8,9,22]. We asked whether the PBD is vital for SHB formation as it is for immune suppression. All peptides, wild-type and mutants, showed similar secondary structures in an aqueous solution (Table 1), thus not hinting on their SHB-forming abilities. We attempted to assess the PBD's importance to membrane fusion via a cell–cell fusion assay (Figure 2). C34 is derived from the SHB, therefore its ability to bind the NHR can be observed by its inhibitory effect on gp41-mediated cell–cell fusion, as previously described [8,30]. As expected, the peptides C34 and C34K containing wild-type PBDs exhibited strong dose-dependent inhibition with an IC₅₀ of 6 and 12 nM respectively and the alanine mutant C34-3A showed limited inhibitory activity with an IC₅₀ over 1500 nM. Surprisingly, in contrast with the C34-3A, the glycine mutant C34-3G had an IC₅₀ of 12 nM, similar to the wild-type peptides (Figure 2B). Active peptides were not toxic at the levels used in the study (Supplementary Figure S1B). Although the PBD is a conserved motif, in some cases C34 with mutant PBDs can still act as powerful membrane fusion inhibitors.

Gp41-mediated membrane fusion entails SHB formation stabilized by the PBD [9] that occurs in the virological synapse [4,9,22]. However, when in contact with the membrane the SHB disassembles, possibly into individual monomers, losing its hexameric form [31,32]. This suggests that the CHR encounters the TCR complex not in its SHB form. We validated this attribute for our peptides via CD spectroscopy. It has been shown that the ratio between θ222 and θ208 can distinguish between oligomeric states of α-helices [33–35]. By this method, we show that in a membrane mimetic environment, the θ222/θ208 ratio of wild-type and mutant SHBs are lower than 0.8, suggesting that the SHB cannot assemble (Table 2). These results suggest that when encountering the TCR complex, the CHR and its PBD are not in a full SHB form.

Inhibition of TCR complex assembly has been reported to occur via interactions in the membrane between basic residues in the TMD of TCRα and acidic residues in the CHR loop overlap of gp41 [20,21]. Since the PBD incorporates highly hydrophobic residues, one might assume that it mediates immunosuppression via an alternative mechanism rather than directly binding TCRα. We hypothesized that the PBD influences the secondary structure of the CHR, as during fusion, the PBD is important for SHB stability, albeit in a different environment. Based on the results in Table 2, we assessed the structure of all the CHR-based peptides alone in an aqueous environment and a membrane mimetic one. Using CD, we observed that when alone and in an aqueous solution all peptides tended to adopt a more antiparallel β-sheet structure. In a membrane mimetic solution the peptides with wild-type PBDs shifted their structure to a more α-helical one while losing their antiparallel β-sheet structure. This was in direct contrast with the PBD mutants that did not show any structure shift (Table 1). These results suggest that the PBD provides an environment-dependent flexibility, promoting an active α-helical form of the CHR in the membrane.

Previous studies show that the loop–CHR interface binds the TCR [20], suggesting that the CHR exerts its immunosuppressive effect by interacting with the TCRα TMD. This raises a question about the nature of the SHB during the interaction, i.e. either fully formed, partially disassociated or fully disassociated. Our previous results together with previous studies [31,32] suggest that the SHB disassembles losing its hexameric form when in contact with the membrane. To gain further insight into SHB structure during immunosuppression we conducted a FRET assay between either SHB or its individual components (N36 or C34) labelled with Rho and the TMD labelled with NBD (Figure 3). The FRET pair consists of Rho (acceptor) and NBD (donor). FRET appears as a reduction in the maxima of NBD fluorescence at 530 nm and in some cases an elevation of the Rho signal at 570 nm. NBD–TMD was added to LUVs consisting of PC/Chol/SM (1:1:1, by vol.), then a Rho-labelled peptide was added in increasing concentrations. As expected, FRET was achieved with Rho–C34 and not Rho–N36 (Figures 3A, 3B and 3E). At the highest donor acceptor ratio of 1:1 Rho/-N36 showed an elevation in FRET. However, this was significantly lower than Rho–C34. In addition, when tested for inhibitory activity N36 showed only a baseline inhibitory effect, similar to C34 mutants (Supplementary Figure S2). Next we attempted to assess whether an assembled SHB is present when C34 interacts with the TMD. The SHB was assembled as previously described [26] with either the N36 or the C34 peptide labelled with Rho (SHB–RhoN36 and SHB–RhoC34 respectively), then added to NBD–TMD as described above. The SHB exhibits a diameter of 35 Å (1 Å=0.1 nm) [22], whereas the Förster distance (R₀) of the NBD–Rho pair, at which the FRET efficiency is 50%, is 56 Å [36]. This indicates that if an assembled SHB labelled on a single component were to interact with the TMD, FRET would be observed, regardless of which component is physically interacting. We found that whereas SHB–RhoN36 did not exhibit FRET, SHB–RhoC34 exhibited a high FRET similar to C34 (Figures 3D, 3C and 3E). This proposes that the SHB disassembles completely while interacting with the TMD, indicating the importance of the PBD in maintaining CHR structural integrity.

Gp41 contains several immunosuppressive domains [14], of which the most recently discovered one spans the CHR and loop regions of the gp41 core [20]. Interestingly, this region of the CHR houses the conserved PBD, whose proposed primary function is to stabilize SHB formation via packing of three hydrophobic residues, two tryptophan and an isoleucine [8], suggesting a dual function for the CHR. We explored the CHR's double functionality through a CHR-derived fusion inhibitor, termed C34. We found that the conserved PBD motif specifically mediates immune suppression of antigen-specific MOG-(35–55) T-cells and maintains an α-helical structure of C34 in a membrane mimetic environment. In contrast, we show that cell–cell fusion inhibition by C34 is not exclusively dependent on the PBD. Furthermore, C34 interacts with the TCRα TMD independently of the SHB, shedding new light on the post-fusion immune suppression process.

Previously, a gp41 loop domain derived-peptide was shown to inhibit TCR activation by obstructing receptor assembly [20]. This is achieved by intra-membrane binding of acidic residues in this region to basic residues of the TCRα TMD [20,21]. Although the activity is not attributed to the CHR, the functional residues are all encompassed at the CHR N-terminus [20]. Therefore we tested the CHR's suppressive ability and found that it highly suppresses antigen-specific T-cell activation. By utilizing C34 wild-type and N-terminal mutated peptides, we show that the CHR N-terminus is the region responsible for the CHR-mediated suppression. T-20, an HIV fusion inhibitor used in the clinic, which lacks the CHR N-terminus, showed no suppressive activity. In contrast, C34, a potent fusion inhibitor in its own right, highly suppressed T-cells. This makes C34 inadequate as a fusion inhibitor as it would inhibit the patient's T-cells thus possibly hastening HIV progression.

C34 with PBD mutants, replaced by either alanine (C34-3A) or glycine (C34-3G) residues, showed a drastic and significant fall in inhibitory activity, despite no changes to the acidic residues crucial for binding to the TCRα TMD. This pinpoints the PBD as a key modulator of C34 immune suppression. The PBD is a highly conserved motif in the CHR's N-terminus that binds a hydrophobic pocket in the NHR, playing a pivotal role in SHB stabilization, therefore fusion promotion [8,9,22]. Due to the nature of hydrophobic interactions one might suppose that other residues would be suited for such an interaction, raising the possibility that the constraints leading to conservation are due to additional functions. Our T-cell inhibition results demonstrate that immune suppression could be the key factor of PBD conservation. This hypothesis is reinforced by potent inhibition of cell–cell fusion by both C34 and C34-3G, suggesting that the PBD is dispensable for successful core formation and membrane fusion.

TCR assembly is mediated by charged residues [37] and its disruption by the gp41 loop is attributed to competition via charged residues [20]. Therefore it is unlikely that the tryptophan- and isoleucine-containing, hence highly hydrophobic, PBD shares a similar mechanism. This is further supported by the crystal structure of the PBD region showing that the PBD creates a hydrophobic face opposite the charged residues in the CHR [8]. Direct and indirect studies have shown that the SHB does not assemble in a hydrophobic environment [31,32,38]. In addition, the present study and others [14,21,39] show that the loop region of the SHB can insert into the membrane. It is therefore plausible that at post-SHB-formation stages (Figure 4A), by loop insertion, the SHB comes into close proximity to the membrane (Figure 4B) leading to a disassociation of the structure (Figure 4C). A recent direct study focusing on events leading to SHB formation by Roche et al. [32] showed that the single SHB components can interact with the membrane, suggesting that after SHB disassociation, the NHR and CHR components remain membrane bound (Figure 4C). It is at this stage that the CHR, mediated by its PBD, may be free to insert into the membrane and bind the TCRα TMD, thus leading to deficient TCR complex assembly (Figure 4D).

Secondary structure analysis showed that the PBD promotes a more α-helical conformation of C34 when in contact with the membrane. Tryptophan shows a statistical preference for the aqueous–lipid interface in membranes, suggested to be due to its side chain that is suited for both environments [40]. This could explain the PBD's role as a structure stabilizer when in contact with the membrane. This is supported by our findings that the C34 interacts with the TCRα TMD independently of the SHB, as the PBD is buried in a hydrophobic pocket of the NHR [8] until exposed by SHB disassociation, freeing it to interact with the membrane.

The present study reveals a dual role of the CHR, mediated by the PBD, providing insights into residue conservation and functionality. Although successful membrane fusion does not seem to entirely rely on the PBD, taking into account its role in T-cell suppression described herein, this motif may serve as a promising future drug target as it mediates two key processes in the viral infection cycle.

Yoel Klug wrote the paper and together with Gal Kapach, prepared materials, performed and designed most of the experiments and analysed the data. Etai Rotem performed and designed experiments. Benjamin Dubreuil prepared materials. Yechiel Shai directed the project, designed the experiments and advised on writing the paper.

We thank Ron Rotkopf for his help with the statistical analysis and Roland Schwarzer for his valuable input.

This work was supported by the Israel Science Foundation [project number 1409/12].

 
• APC

  antigen-presenting cell

 
• CD

  circular dichroism

 
• Chol

  cholesterol

 
• CHR

  C-terminal heptad repeat

 
• DCM

  dichloromethane

 
• DMEM

  Dulbecco's modified Eagle's medium

 
• ENV

  envelope

 
• FP

  fusion peptide

 
• FRET

  Förster resonance energy transfer

 
• gp

  glycoprotein

 
• LPC

  lysophophatidylcholine

 
• LUV

  large unilamellar vesicle

 
• mMOG

  mouse myelin oligodendrocyte glycoprotein

 
• MOG

  myelin oligodendrocyte glycoprotein

 
• NBD

  7-nitrobenz-2-oxa-1,3-diazole

 
• NHR

  N-terminal heptad repeat

 
• PA

  palmitic acid

 
• PBD

  pocket binding domain

 
• PC

  phosphatidylcholine

 
• Rho

  rhodamine

 
• RP-HPLC

  reverse-phase HPLC

 
• SHB

  six-helix bundle

 
• SM

  sphingomyelin

 
• TCR

  T-cell receptor

 
• TFA

  trifluoroacetic acid

 
• TMD

  transmembrane domain

 
• XTT

  2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide