Biophysical studies of HIV-1 glycoprotein-41 interactions with peptides and small molecules – Effect of lipids and detergents

https://doi.org/10.1016/j.bbagen.2020.129724Get rights and content

Highlights

  • 19F NMR and fluorescence are sensitive measures of ligand binding to HIV-1 gp41.

  • DPC is a poor substitute for membrane when studying interactions of gp41.

  • The gp41 N-heptad repeat does not unfold in liposomes composed of POPC and cholesterol.

  • Lipid binding competes with protein binding for small molecule hydrophobic pocket inhibitors.

  • Observed antiviral activity is most closely correlated with the aqueous fraction of the inhibitors.

Abstract

Background

The hydrophobic pocket (HP) of HIV-1 glycoprotein-41 ectodomain is defined by two chains of the N-heptad repeat trimer, within the protein-protein interface that mediates 6HB formation. It is a potential target for inhibitors of viral fusion, but its hydrophobic nature and proximity to membrane in situ has precluded ready analysis of inhibitor interactions.

Methods

We evaluated the sensitivity of 19F NMR and fluorescence for detecting peptide and small molecule binding to the HP and explored the effect of non-denaturing detergent or phospholipid as cosolvents and potential mimics of the membrane environment surrounding gp41.

Results

Chemical shifts of aromatic fluorines were found to be sensitive to changes in the hydrogen bonding network that occurred when inhibitors transitioned from solvent into the HP or into ordered detergent micelles. Fluorescence intensities and emission maxima of autofluorescent compounds responded to changes in the local environment.

Conclusions

Gp41ligand binding occurred under all conditions, but was diminished in the presence of detergents. NMR and fluorescence studies revealed that dodecylphosphocholine (DPC) was a poor substitute for membrane in this system, while liposomes could mimic the membrane surroundings.

General significance

Our findings suggest that development of high potency small molecule binders to the HP may be frustrated by competition between binding to the HP and binding to the bilayer membrane.

Introduction

HIV-1 glycoprotein-41 (gp41) is a transmembrane protein that forms part of the spike on the viral surface and that is involved in fusion between HIV-1 and host cells. It consists of an ectodomain, transmembrane domain and cytoplasmic domain [1,2], and exists in a metastable state prior to fusion. Fusion is triggered by viral attachment and release of glycoprotein-120 (gp120), followed by collapse of trimeric N-heptad repeat (NHR) and C-heptad repeat (CHR) helices of the ectodomain into a six helix bundle (6HB) [1,[3], [4], [5], [6]]. Inhibitors targeting gp41 have been studied for their ability to prevent entry and infection. A hydrophobic pocket (HP) defined by two chains of the NHR trimer is contained within the protein-protein interface that mediates 6HB formation [7]. The pocket is considered to be a potential target for low molecular weight fusion inhibitors [[8], [9], [10], [11], [12], [13]]. Nevertheless, small molecule fusion inhibitors with drug-like potency have not been found. The most effective (nM or better) fusion inhibitors discovered to date are peptides based on the CHR domain of gp41 [14], including T20 (enfuvirtide), an FDA approved fusion inhibitor. They act in a dominant-negative fashion by preventing the conformational changes in gp41 required for fusion.

It has proved difficult to study gp41 – small molecule interactions by biophysical methods as an aid to structure based drug design. Isolated gp41 ectodomain adopts the collapsed 6HB form in which the hydrophobic pocket is obscured (Fig. 1A) [15,16]. In order to study binding in the pocket, we designed reverse hairpin proteins in which a truncated CHR preceded the NHR in sequence with a short connecting loop (Fig. 1B) [17,18]. C28(L4)N50, named according to the length of the domains, has an exposed HP in solution. We also prepared CHR peptide containing the HP binding domain (HPbd) attached to the small globular domain GB1, allowing for straightforward bacterial synthesis and labeling for NMR. The resulting 15N-labeled constructs CPi635 (shown in Fig. 1B) was shown by HSQC to bind to C28(L4)N50 [17].

While C28(L4)N50 proved to be trimeric and highly helical [18], it gave broad low quality HSQC spectra (Supplementary Material Fig. S1), evidence of significant dynamics on the NMR chemical shift time scale. Furthermore ligands are hydrophobic and complexes with protein are poorly soluble at concentrations required for structural studies in aqueous solution. Therefore protein-detected ligand binding and structural analysis by NMR was not feasible for this system.

We observed that non-denaturing detergents helped to solubilize some of the more hydrophobic inhibitors and their mixtures with protein. We reasoned that lipids or membrane mimetic components could contribute to the mechanism of action in the milieu of the in situ fusion reaction. Here we examined this possibility using 19F NMR and fluorescence experiments. 19F NMR is a useful probe of molecular interactions due to the sensitivity of 19F-resonances to changes in local environment [19], as is deuterium NMR [20], and the low concentrations required for measurement. Additionally, 19F NMR is readily applicable to the study of proteins in the presence of detergents, lipids or stabilizing agents, because of the absence of background signals from these compounds. We examined both peptide and low molecular weight inhibitor binding to the HP, as well as the effect of phospholipid detergent dodecylphosphocholine (DPC) and Tween-20. Tween-20 is a mild non-ionic non-denaturing detergent that is used to solubilize membrane proteins without disrupting structure [21,22], DPC is a zwitterionic detergent that has been used in NMR studies of membrane-associated peptides and proteins as a membrane mimetic. The small size of DPC micelles together with the availability of DPC in perdeuterated form have favored its use for solution state proton NMR studies [23], with the caveat that functional activity should be checked on a case by case basis [24]. Previous studies of peptides and constructs of the 6HB and associated domains have resulted in disparate interpretations of their interactions with detergent micelles and lipid bilayers [2,25,26]. Some studies have indicated that DPC can alter the monomer-trimer equilibrium in gp41, which would disrupt the HP binding site [[27], [28], [29]]. Other studies have suggested that DPC is a suitable membrane mimetic in which to evaluate structure and fusion-inhibitory agents targeting gp41 domains [[30], [31], [32]]. Interpretation is likely compounded by the difficulties inherent in biophysical studies of the hydrophobic peptides, making the outcomes sensitive to exact composition, concentration and conditions. Overall, it appears that the 6HB, stable in aqueous solution, is destabilized or more dynamic in DPC micelles, while the structure is retained in zwitterionic lipid bilayers representative of the outer leaflet of the plasma membrane [33,34]. Isolated NHR domain but not 6HB has also been shown to disrupt lipid ordering [25,26].

Since our ligands contain indole or indazole groups with intrinsic fluorescence, we additionally studied binding by examining changes in fluorescence spectra. Fluorescence studies were also conducted in the presence of liposomes that could more accurately represent the bilayer interface, giving us the opportunity to examine whether DPC micelles were adequate mimetics for this protein – lipid – small molecule system.

Section snippets

Material and methods

Experimental Procedures are described in Sections S1 – S3 of the Supplementary Material. These include synthetic routes of synthesis of ligands, binding and antiviral assays, NMR and fluorescence experiments and preparation of liposomes.

C-heptad repeat domain binding in the hydrophobic pocket is readily detected by 19F NMR

We first established the sensitivity of 19F NMR to the binding interaction in the HP by studying the validated CPi635 – C28(L4)N50 interaction (Fig. 1) [17]. CPi635 and C28(L4)N50 were labeled with 5-F-tryptophan by addition of 5-F-indole to the growth media [19]. Tryptophan residues predominate in the CHR-NHR interaction in the HP. CPi635 contains three tryptophan residues, two in the HPbd and one within GB1. The NHR receptor C28(L4)N50 contains a single tryptophan residue in the HP. Fluorine

Conclusions

This study demonstrated that HPbd peptide binding and small molecule binding in the hydrophobic pocket were readily detected by 19F NMR and fluorescence. 19F resonances of fluorine-labeled peptides and small molecules shifted downfield upon addition of protein and/or detergent, indicating an increase in hydrophobic environment. In peptides, 19F spectra of 5-F-Trp were very sensitive to binding. However, in the small molecules, sensitivity of 19F chemical shifts was highest for 19F atoms that

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by NIH grants GM087998 and AI122847 to MG.

References (46)

  • D.C. Chan et al.

    Core structure of gp41 from the HIV envelope glycoprotein

    Cell

    (1997)
  • J.R. Lakowicz et al.

    Binding of an indole derivative to micelles as quantified by phase-sensitive detection of fluorescence

    J. Biol. Chem.

    (1983)
  • S.L. Veatch et al.

    Separation of liquid phases in giant vesicles of ternary mixtures of phospholipids and cholesterol

    Biophys. J.

    (2003)
  • L.R. Whitby et al.

    Discovery of HIV fusion inhibitors targeting gp41 using a comprehensive alpha-helix mimetic library

    Bioorg. Med. Chem. Lett.

    (2012)
  • L. Cai et al.

    Biochemistry and biophysics of HIV-1 gp41 - membrane interactions and implications for HIV-1 envelope protein mediated viral-cell fusion and fusion inhibitor design

    Curr. Top. Med. Chem.

    (2011)
  • R.M. Markosyan et al.

    HIV-1 envelope proteins complete their folding into six-helix bundles immediately after fusion pore formation

    Mol. Biol. Cell

    (2003)
  • G.B. Melikyan et al.

    Evidence that the transition of HIV-1 gp41 into a six-helix bundle, not the bundle configuration, induces membrane fusion. Comment In: J Cell Biol. 2000 Oct 16;151(2):F9–14

    J. Cell Biol.

    (2000)
  • J. Chen et al.

    N- and C-terminal residues combine in the fusion-pH influenza hemagglutinin HA(2) subunit to form an N cap that terminates the triple-stranded coiled coil

    Proc. Natl. Acad. Sci. U. S. A.

    (1999)
  • D.C. Chan et al.

    Evidence that a prominent cavity in the coiled coil of HIV type 1 gp41 is an attractive drug target

    Proc. Natl. Acad. Sci. U. S. A.

    (1998)
  • B.D. Welch et al.

    Design of a potent D-peptide HIV-1 entry inhibitor with a strong barrier to resistance

    J. Virol.

    (2010)
  • F. Yu et al.

    Approaches for identification of HIV-1 entry inhibitors targeting gp41 pocket

    Viruses

    (2013)
  • M.A. Luftig et al.

    Structural basis for HIV-1 neutralization by a gp41 fusion intermediate-directed antibody

    Nat. Struct. Mol. Biol.

    (2006)
  • G. Zhou et al.

    Structure-activity relationship studies of Indole-based compounds as small molecule HIV-1 fusion inhibitors targeting glycoprotein 41

    J. Med. Chem.

    (2014)
  • Cited by (1)

    View full text