Figure 1 shows the cell–cell fusion assay in hACE2-expressing cells. Very little GFP expression was observed in cells lacking S proteins. When SARS-CoV-1 and SARS-CoV-2 S proteins were expressed in effector cells, GFP expression was significantly increased. Compared to SARS-CoV-1 S, all SARS-CoV-2 S-expressing cells showed increased GFP expression, indicating that SARS-CoV-2 S induces cell–cell fusion more efficiently. Cells expressing SARS-CoV-2 variant S proteins showed a non-significant increase in GFP expression compared to those expressing SARS-CoV-2 reference strain S proteins. Among variants, Alpha and Delta showed higher fusion activity than Beta and Gamma.
Next, we investigated cell–cell fusion in cells expressing endogenous levels of hACE2 (Fig. 2). In cells expressing endogenous levels of hACE2s, the Delta variant showed the highest fusion activity. Other variants showed levels of fusion kinetics similar to those of the reference viruses. SARS-CoV-1 showed lower fusion activity than all SARS-CoV-2 strains. The results indicated that the level of hACE2 expression can affect the fusogenicity of S proteins and that the Delta variant can utilize very low amounts of hACE2 for S-mediated membrane fusion.
S-mediated membrane fusion is a crucial step for virus entry and cell–cell spread. Receptor interaction is needed to induce conformational changes in S proteins to expose the S2` cleavage site and trigger fusion [2, 18]. We present two possible reasons that SARS-CoV-2, especially the Delta S protein, exhibits high cell–cell fusion. First, we speculate that the high binding affinity between the S protein and hACE2 may promote cell–cell fusion. The binding affinity of S proteins and hACE2 has been well documented in previous reports. The N501Y (Alpha, Beta, Gamma), E484K (Beta, Gamma), L452R (Delta), and T478K (Delta) mutations increased binding affinity to hACE2, whereas K417N resulted in weaker binding [14, 16, 27]. RBDs with triple mutation of N501Y, E484K, and K417N showed only slightly enhanced binding affinity [16]. These results are consistent with our observations that the Alpha and Delta variants generally induced greater cell–cell fusion than the Beta and Gamma variants.
Additionally, S mutations outside the RBD may influence the dynamics of the conformational changes that occur after receptor binding. Previously, it was shown that communication between the amino acid at position 330 (in S1) and the amino acid at position 1114 (in S2) is important for the efficient fusion activity of murine hepatitis virus in BHK-R2 cells [19]. Although both the Alpha and Delta RBDs have mutations that increase hACE2 binding affinity, only Delta S proteins showed significantly enhanced cell–cell fusion. Unlike Alpha variant S proteins, Delta variant S proteins contain additional mutations in the NTD domains of S1 and S2. Thus, it is possible that these mutations affect the kinetics of conformational changes in S proteins. To demonstrate the molecular mechanisms of increased cell–cell fusion of Delta S proteins, further studies with single S mutations are needed.
Recently, it was demonstrated that SARS-CoV-2 can spread through cell–cell fusion transmission [31]. Cell-to-cell transmission is critical for the robust replication of SARS-CoV-2 in human lung and other tissues, as well as the rapid spread of SARS-CoV-2, including some variants of concern. In a previous study, the Alpha and Beta variants, as well as the D614G mutant, exhibited similar efficiencies of cell-to-cell transmission compared with the wild type [29, 31]. Here, we found that the cell–cell fusion promoted by the Alpha, Beta, and Gamma variants was similar to that of reference strains. However, Delta variants exhibited stronger cell–cell fusion than all the tested strains. More importantly, the Delta variant seems to induce cell–cell fusion even in the presence of very low amounts of hACE2, as evidenced in our results.
In conclusion, we evaluated the cell–cell fusion ability of SARS-CoV-2 variants in comparison with SARS-CoV-1 and SARS-CoV-2 reference strains. We found that S has evolved to favor receptor-induced conformational changes and that the facilitated conformational changes induce accelerated cell–cell fusion. Our results improve the molecular understanding of earlier S variants underlying virus transmissibility and infection and further provide a molecular basis of our knowledge and prediction of SARS-CoV-2 evolution.