SARS-CoV-2 spike-protein D614G mutation increases virion spike density and infectivity

Coronaviruses utilize the trimeric spike (S) protein to bind host receptors and mediate membrane fusion. For SARS-CoV-2, S engages angiotensin-converting enzyme 2 (ACE2) via the receptor-binding domain (RBD) in S1, while S2 mediates fusion. SARS-CoV-2 S is cleaved at the S1–S2 junction by furin in producer cells and further processed at S2′ in target cells. Early genomic surveillance identified a prevalent S-protein mutation, D614G, within the S1 domain near the S1–S2 interface. Global sequence analyses (GISAID) showed that the G614 genotype was absent in January–February 2020, increased to 26% in March, and rapidly predominated by April–May (65–70%), suggesting a potential transmission advantage. Some reports associated G614 with higher viral loads in patients, although linkage to other mutations complicates interpretation. The study’s research question is whether D614G alters S-protein function to increase viral entry/infectivity, and by what mechanism (e.g., ACE2 affinity, spike stability, incorporation, S1 shedding), with implications for transmissibility.

Prior work established ACE2 as the receptor for SARS-CoV and SARS-CoV-2 and characterized the S protein’s structure, processing, and entry mechanisms, including roles of furin, TMPRSS2, and endosomal cathepsins. Structural studies suggested that D614G might shift S toward an open conformation favorable for ACE2 binding. Epidemiological sequencing documented rapid emergence of G614 globally and correlations with higher patient viral loads. However, definitive mechanistic links between D614G and infectivity or transmission required controlled virological assays to distinguish effects on ACE2 affinity, spike stability, or virion incorporation.

• Sequence surveillance: Downloaded SARS-CoV-2 S sequences (GenBank), filtered for full length (>1272 aa) and by release date; computed residue 614 genotype frequencies (R, Biostrings) and generated logo plots (WebLogo). Counts by month included January (17), February (33), March (293), April (1511), May (254).
• Pseudovirus (PV) production: Generated MLV-based pseudoviruses encoding eGFP or luciferase and pseudotyped with SARS-CoV-2 S variants (D614, G614, and a furin cleavage knockout, FKO) using HEK293T producer cells and calcium phosphate transfection. Some S constructs carried N- and/or C-terminal FLAG or N-Myc tags for detection.
• Entry assays: Infected HEK293T cells transduced to express human ACE2 (hACE2-293T), with or without TMPRSS2 co-expression, and lung epithelial cells engineered to express ACE2. Measured entry by GFP fluorescence or luciferase activity 24 h post-infection; quantified PV genome copy numbers by RT-qPCR.
• PV purification and analysis: Purified PVs by pelleting through 20% sucrose. Assessed PV yields by qPCR. Analyzed S incorporation and processing by western blot using anti-FLAG M2; used anti-MLV p30 as a PV loading control. Performed densitometry (ImageJ) to quantify total S (S1+S2) and S1:S2 ratios, normalizing for p30.
• Virus-like particles (VLPs): Produced SARS-CoV-2 VLPs in HEK293T cells by co-expressing native M, N, E, and S (D614 or G614). Purified as for PVs. Detected S with anti-FLAG M2 and N with pooled convalescent plasma.
• Stability assay: Stored PVs at room temperature up to 48 h; compared decay of infectivity for PV^D614, PV^G614, and PV^FKO.
• ACE2 binding assays: Expressed S variants on HEK293T cells; measured binding of engineered ACE2-Fc (hACE2-NN-Ig) by flow cytometry, normalizing to total S (FLAG) or S1 (Myc) levels. Performed surface plasmon resonance with immobilized S1-Fc and monomeric ACE2 at multiple concentrations to derive kinetic parameters (ka, kd) and Kd.
• Neutralization: Assessed neutralization of PVs pseudotyped with D614 or G614 S by serial dilutions of human convalescent plasma; measured entry via luciferase.
• Statistics: Analyses in GraphPad Prism; two-tailed tests; one-way ANOVA with multiple comparisons where appropriate; unpaired two-sided Student’s t-tests for pairwise comparisons; significance threshold p<0.05.

• PVs bearing S^G614 exhibited markedly higher entry into ACE2-expressing HEK293T cells than S^D614, irrespective of TMPRSS2 expression; similar enhancement observed in ACE2-expressing lung epithelial cells.
• Increased PV^G614 entry correlated with decreased S1 shedding and higher virion S incorporation. Western blots and densitometry showed approximately 4.17-fold higher total S (S1+S2) in PV^G614 versus PV^D614 when normalized by p30; S1:S2 ratios were ~3.5-fold higher for G614 preparations under matched S2 band conditions.
• Similar trends were observed in SARS-CoV-2 VLPs: VLP^G614 displayed higher S1:S2 ratios and greater total virion S than VLP^D614; findings held for untagged S constructs and for PVs produced in lung epithelial cells.
• Stability: No detectable differences in loss of infectivity at room temperature up to 48 h among PV^D614, PV^G614, or PV^FKO, suggesting differences in spike levels do not alter PV decay kinetics under these conditions.
• ACE2 binding: Cell-based binding assays and SPR indicated that D614G did not increase ACE2 affinity; binding levels, when normalized to S or S1 expression, were comparable between D614 and G614.
• Neutralization: PV^G614 did not exhibit increased resistance to neutralization by human convalescent plasma compared to PV^D614.
• Overall: D614G enhances infectivity by increasing the number of functional S trimers on virions and reducing S1 shedding, rather than by increasing ACE2 affinity or altering neutralization sensitivity.

The study demonstrates that the D614G mutation enhances entry of SARS-CoV-2 S-pseudotyped virions by promoting greater incorporation of intact S and reducing S1 shedding, thereby increasing functional spike density. This mechanism explains higher infectivity without requiring changes in ACE2 binding affinity or neutralization resistance. The lack of differences in PV stability over time indicates that pre-release events (e.g., S1 shedding during assembly/trafficking) likely underlie the altered spike composition. While these virological findings are consistent with epidemiological observations of G614 predominance and higher patient viral loads, they do not alone prove increased transmissibility in humans. The physical basis of reduced S1 shedding and increased incorporation remains unresolved; possibilities include altered S1–S2 interdomain interactions, effects on post-translational processing, or changes mediated via the cytoplasmic tail affecting incorporation. The requirement for a functional furin cleavage site to observe pronounced differences suggests interplay between D614G and S1/S2 processing. Evolutionarily, the rise of D614G occurred in the context of SARS-CoV-2’s acquisition of a furin cleavage site, distinguishing it from SARS-CoV and related bat sarbecoviruses that rely on TMPRSS2 or endosomal proteases for S1/S2 cleavage.

D614G in the SARS-CoV-2 spike increases pseudovirus and VLP infectivity by reducing S1 shedding and increasing the density of functional S trimers on virions. The mutation does not increase ACE2 affinity or confer resistance to neutralization by convalescent plasma. These results provide a mechanistic basis for the observed epidemiological dominance of the G614 variant and suggest that enhanced virion spike incorporation contributes to transmission fitness. Future work should elucidate the structural basis for altered S1–S2 interactions, assess effects in authentic SARS-CoV-2 across diverse cell types and in vivo models, and determine the mutation’s impact on pathogenesis and transmission dynamics.

• Use of MLV-based pseudoviruses and VLPs may not fully recapitulate authentic SARS-CoV-2 assembly, maturation, and entry pathways in vivo.
• Physical/structural basis for reduced S1 shedding and increased S incorporation with G614 remains undefined.
• Epidemiological inference of transmission advantage is indirect; the study does not measure transmission in humans or animal models.
• Some constructs utilize epitope tags; although untagged controls were tested, tagging could still introduce artifacts.
• D614G often co-occurs with other genomic changes; effects were isolated here to S but linkage in nature may confound clinical correlations.
• Stability assays at room temperature may not reflect physiological conditions; timing suggests some S1 shedding occurs before assay windows.