SARS-CoV-2 Fusion Peptide has a Greater Membrane Perturbating Effect than SARS-CoV with Highly Specific Dependence on Ca2

Abstract

Coronaviruses are a major infectious disease threat, and include the zoonotic-origin human pathogens SARS-CoV-2, SARS-CoV, and MERS-CoV (SARS-2, SARS-1, and MERS). Entry of coronaviruses into host cells is mediated by the spike (S) protein. In our previous ESR studies, the local membrane ordering effect of the fusion peptide (FP) of various viral glycoproteins including the S of SARS-1 and MERS has been consistently observed. We previously determined that the sequence immediately downstream from the S2' cleavage site is the bona fide SARS-1 FP. In this study, we used sequence alignment to identify the SARS-2 FP, and studied its membrane ordering effect. Although there are only three residue differences, SARS-2 FP induces even greater membrane ordering than SARS-1 FP, possibly due to its greater hydrophobicity. This may be a reason that SARS-2 is better able to infect host cells. In addition, the membrane binding enthalpy for SARS-2 is greater. Both the membrane ordering of SARS-2 and SARS-1 FPs are dependent on Ca²⁺, but that of SARS-2 shows a greater response to the presence of Ca²⁺. Both FPs bind two Ca²⁺ ions as does SARS-1 FP, but the two Ca²⁺ binding sites of SARS-2 exhibit greater cooperativity. This Ca²⁺ dependence by the SARS-2 FP is very ion-specific. These results show that Ca²⁺ is an important regulator that interacts with the SARS-2 FP and thus plays a significant role in SARS-2 viral entry. This could lead to therapeutic solutions that either target the FP-calcium interaction or block the Ca²⁺ channel.

Keywords: COVID-19; ESR; ITC; hydrophobicity; ion specificity.

Graphical abstract

Figure 1

(A). A model of the coronavirus FP or “platform” interacting with a lipid bilayer (adopted from our previous model using the structural information from CD spectroscopy in this study, and the molecular dynamics model. A Heliquest Plot for the first 11 amino acids is shown. (B) The structure of spin labeled lipids used: DPPTC, 5PC, 10PC and 14PC. (C) The sequence alignment of SARS-2, SARS-1, and MERS FPs, and their grand average of hydropathy (GRAVY) values (calculated using GRAVY Calculator, http://www.gravy-calculator.de/index.php). The residues in red highlight the difference between SARS-2 and SARS FP. The residues in bold are the negatively charged residues which are potential Ca²⁺ binding sites.

Figure 2

(A) The SDS-PAGE gel of the SARS-2 FP after purification using a His-tag affinity column; the arrow shows the bands for the FP. The fractions were the combined and further purified using size-exclusive chromatography. (B) CD Spectra of SARS-2 FP in solution (thin lines) and POPC/POPG/Chol = 3/1/1 SUVs (thick lines) in pH 5 buffer (black) or pH 7 buffer (red) at 25 °C, and SARS-1 FP in SUVs at pH5 (blue). (C) CD spectra for SARS-2 FP in SUVs at pH5 show that the alpha helix content increases with the increase of Ca²⁺ concentration. (D) The alpha helix content of the FP in membranes at different Ca²⁺ concentrations, which was calculated from (C) using K3D2 server. The data were fit with the built-in logistic function of Origin, which shows the X₅₀ = 0.83 mM. (E, F) The CD spectra of SARS-2 FP1 (E) and FP2 (F) in SUV in pH5 buffer with 0 mM (black) and 2 mM (red) CaCl₂. The percentage of alpha helix and beta strand was calculated using K3D2.

Figure 3

Plots of order parameters of DPPTC (A), 5PC (B), 10PC (C) and 14PC (D) versus peptide:lipid ratio (P/L ratio) of SARS-2 FP, SARS FP and MERS FP in POPC/POPG/Chol = 3/1/1 MLVs in buffer at pH5 with 150 mM NaCl at 25 °C. Black, SARS-2 FP with 1 mM Ca²⁺ and at pH 5; red, SARS-2 FP with 1 mM EGTA; blue, SARS FP with 1 mM Ca²⁺; green, MERS FP with 1 mM Ca²⁺. The curves were fitted with the logistic function (Table 1A and 1B) (E). Plots of local order parameters of DPPTC versus P/L ratio as in A-D with pH indicated. Black, SARS-2 FP at pH 5; red, SARS-2 FP triple mutant (E4A_D5A_D15A) at pH 5; blue, SARS-2 FP at pH 7; and green, a peptide with shuffled sequence of SARS-2 FP at pH 5. (F) Plot of local order parameters of DPPTC with and without 1% peptide binding (ΔS₀) versus Ca²⁺ as in A-D. Black, SARS-2 FP; blue, SARS FP and green, MERS FP. The experiments were typically repeated three times. The typical uncertainties from simulation found for S₀ range from 1–5 × 10⁻³, while the uncertainties from repeated experiments were 5–8 × 10⁻³ or less than ± 0.01. We show in Panels A and B the bars for the standard deviation. Statistical significance analyses were performed using two-tailed Student’s t-test on the S₀′s of 0% SARS-2 FP and 5% SARS-2 FP at the “1mM Ca²⁺” condition, *** ≤ 0.001, highly significant.

Figure 4

ITC analysis of Ca²⁺ binding to FP of the coronaviruses. (A) SARS-2 FP, (B) SARS-2 FP triple mutant (E4A_D5A_D15A). (C) SARS-1 FP. (D) MERS FP. The peptides were titrated with CaCl₂ in the pH5 buffer. The integrated data represent the enthalpy change per mole of injectant, ΔH, in units of kcal/mol as a function of the molar ratio. The data were fitted using a one-site model (parameters shown in Table 3B). Data points and fitted data are overlaid.

Figure 5

The specificity of Ca²⁺. (A) Plots of local order parameter changes of DPPTC (ΔS₀) versus the P/L ratio with 1 mM ion concentration for a variety of ions and in buffer with 150 mM NaCl at pH5 and at 25 °C. (B) The same with 2 mM ion concentration. (C) Plots of local order parameter changes of DPPTC (ΔS₀) versus the concentration of the ions with 1% P/L ratio and at pH5. (D) ITC analysis of ions binding to SARS-2 FP in solution. The peptides were titrated with CaCl₂ in pH5 buffer.