A new study by researchers from Inserm (French National Institute of Health and Medical Research) has validated that cholesterol and ceramide lipids facilitate SARS-CoV-2 spike protein-mediated membrane fusion.
The researchers also found that Chlorpromazine (CPZ), a commonly used antipsychotic drug, could be used to inhibit this fusion and also prevent SARS-CoV-2 cellular entry.
The study findings were published on a preprint server and are currently being peer reviewed.
Cellular infection by SARS-CoV-2 starts with the fusion of its lipid envelope with the host cell membrane that leads to the delivery of its genomic RNA into the cell cytoplasm, production of new viral proteins by the host cell, and assembly/release of newly formed viruses.
Such fusion event can occur directly with the cell plasma membrane or with the endosomal membrane following endocytosis of SARS-CoV-2. Because it occurs early in the viral replication cycle, this early fusion step is an attractive target for the development of drugs and vaccines able to block virus entry1.
SARS-CoV-2 binding and fusion with the cell membrane is mediated by the Spike (S) protein, a class I viral fusion protein composed of two subunits, S1 and S2. The S1 subunit is involved in binding to the host cell plasma membrane via its interaction with the human angiotensin-converting enzyme (ACE2) receptor of the cell surface2–4, whereas the S2 subunit mediates the fusion of SARS-CoV-2 lipid envelope with the host cell membrane5.
For the S2 subunit to be available for fusion, the S protein must be cleaved at the S1/S2 interface by cellular proteases after S1 binding to ACE2.
These proteases can be either from the cell surface, e.g. the membrane-anchored serine protease TMPRSS26, when fusion occurs directly at the plasma membrane, or from endosomes, e.g. the cysteine protease cathepsin L7, when fusion occurs at the endosomal membrane after endocytosis of the viral particle. Several lines of evidence obtained with SARS-CoV-2 or the closely related coronaviruses SARS-CoV and MERS-CoV suggest that direct fusion with the plasma membrane is the preferred route, notably during infection of respiratory cells1. S protein produced by infected cells can also be
directly transferred to the cell surface8, leading to potential fusion with neighboring uninfected cells, which is believed to play a key role in SARS-CoV-2 pathogenicity9.
The S2 subunit of the SARS-CoV-2 S protein possesses an N-terminal fusion peptide (FP), followed by two heptad repeat domains (HR1 and HR2), and a transmembrane domain (TMD) that anchors the S protein into the viral membrane. Fusion by S2 starts with the insertion of its FP into the target cell membrane to establish a molecular bridge consisting of a three-helix coiled-coil complex composed of its HR1 and HR2 domains. S2 then folds back onto itself, in the form of a six-helix coiled-coil hairpin complex of the HR1 and HR2 domains, which brings the FP and TMD in molecular proximity along with the viral and cellular membranes in which they reside. This close membrane apposition combined with lipid bilayer destabilization produced by membrane insertion of the FP leads to fusion.
Isolated FPs have been very useful in elucidating the mechanisms of viral fusion because (i) they can themselves induce fusion, and (ii) single mutations within FPs can lead to complete loss of viral fusion and infection10,11. FPs usually consist of about 20 amino acids, mostly hydrophobic, and are highly conserved within a given viral family. Contrary to other type I viral fusion proteins, such as those of HIV and influenza viruses which possess a single FP, several potential FPs with membrane interacting and/or membrane fusion activity were identified within the S2 subunit of coronaviruses S protein12.
The region FP1 (residues 816- 833 in SARS-CoV-2) is however believed to be the functional FP based on its high sequence conservation (particularly among MERS-CoV, SARS-CoV and SARS-CoV-2) and its location immediately downstream of the S2’ cleavage site that is known to be critical for coronaviruses fusion6,10,13,14 (Fig. 1A).
In the case of SARS-CoV, FP1 was shown to induce membrane fusion and some mutations that abolished FP1-mediated liposome fusion in vitro also affected cell-cell fusion and viral infection in situ10. The region just C-terminal of FP1 (residues 834-855 in SARS-CoV-2, called FP2) was proposed to complement FP1 to form an
extended fusion peptide (FP1-FP2) with higher membrane affinity15. Recent experimental and computational structural studies showed that the FP1 of SARS-CoV-2 inserts deeply into lipid bilayer structures, whereas its FP2 stays around the bilayer surface16–18, suggesting strong membrane perturbing effect and thus fusion activity by FP1. But direct evidence that the FP1 of SARS-CoV-2 can induce membrane fusion is still missing.
The activity of viral FPs also depends on the lipid composition of the fusing membranes19. The FP of SARS-CoV was found to bind stronger and to penetrate deeper into membranes containing cholesterol (CHOL)20. Interestingly, a recent study showed that the enzyme cholesterol 25[hydroxylase could inhibit infection by MERS-CoV, SARS-CoV and SARS- CoV-2 by inducing depletion of plasma membrane CHOL, thus blocking viral fusion at the cell surface21.
Along the same lines, generation of ceramide (CER) membrane domains at the cell surface upon cleavage of sphingomyelin by the acid sphingomyelinase was found to facilitate SARS-CoV-2 entry into cells22. This suggests that repurposing drugs which can modify the structure or lipid composition of the host cell membrane could be an effective treatment against SARS-CoV-2 infection.
In this regard, antipsychotic (AP) drugs are interesting candidates that were proposed to exert some antiviral activity against coronaviruses23–25. Because of their amphiphilic property and planar geometry, AP drugs are known to intercalate into lipid bilayers and thus modify their biophysical properties26, which can in turn impact the ability of these lipid bilayers to fuse.
In this paper, we have investigated in vitro with a fluorescence resonance energy transfer (FRET)-based liposome fusion assay the capacity of FP1 and FP2 from SARS-CoV-2 S protein to fuse membranes of various lipid compositions (including or not CHOL and CER). We have also tested the ability of the AP chlorpromazine (CPZ) to inhibit FP-mediated membrane fusion. The effect of lipid composition and CPZ addition was also examined in situ on full-length S protein-mediated membrane fusion measured by a nanoluciferase-based cell fusion assay.