There are 9 sites that differ between SARS-CoV-2 spike glycoproteins from the original Wuhan strain (Wuhan- Hu-1) and UK variant B.1.1.7 [1] (Figure 1, Table 1). There are 2 sites with deletions (positions 69-70, 144-145) and 7 sites with amino acid substitutions (4 positions in the S1 subunit and 3 positions in S2). Most substitutions in the spike protein that distinguish UK variant B.1.1.7 are located at sites of intermolecular interaction (4 out of 7 substitutions). One mutation (N501Y) enhanced the affinity of the spike protein with ACE2 [2]. N501Y was modeled based on the cryoEM structure of the Wuhan- Hu-1 spike protein/ACE2 complex [3] (Figure 2A, PDB 6M17), indicating that the gain in affinity likely results from aromatic interactions (π stacking) between Tyr501 and Tyr41 of ACE2 (Figure 2B).

Table 1: Locations, surface accessibility and interactions of mutations that distinguish the B.1.1.7 and Wuhan-Hu-1 spike proteins.

Strikingly, 3 mutations in the spike protein that distinguish UK variant B.1.1.7 are located at interfaces between subunits of the trimeric protomer. Unlike the mutation at position 501 that increased affinity for ACE2, the 3 substitutions at spike trimer interfaces likely reduce intermolecular binding affinity. The mutations likely increase spike protein lability in a manner that enhances dynamic virus processes that include spike protein cleavage, structural rearrangement and host cell fusion mechanisms.

Intermolecular interactions between individual chains of the SARS-CoV-2 spike glycoprotein were observed at positions A570 (Figure 3A), D614 (Figure 3B), and S982 (Figure 3C) in the original Wuhan strain (Wuhan- Hu-1). The A570D substitution in the UK variant B.1.1.7 variant introduces steric clash with the backbone amide of K964 (Figure 3D). The D614G substitution results in the formation of a distinctive cavity at the interface of spike protein subunits in the UK variant B.1.1.7 trimer (Figure 3E). The S982A in UK variant B.1.1.7 lacks intermolecular hydrogen (H) bonding potential between spike protein subunits at this site (Figure 3F). Collectively, these data suggest that: 1) the UK variant B.1.1.7 exhibits a change that enhances affinity for the coronavirus receptor ACE2 (N501Y), and 2) mutations may enhance dynamic virus fusion mechanisms by reducing intermolecular stability of spike protein subunits (A570D, D614G, S982A).

P681H represents a potentially important difference between Wuhan-Hu-1 sequence and UK variant B.1.1.7. Position 681 is located adjacent to the RRAR proprotein convertase motif considered a hallmark of high pathogenesis (PRRAR in Wuhan-Hu-1 [4], HRRAR in B.1.1.7). This site is cleaved by furin and other proteases to separate the S1 and S2 subunits of the spike protein, which undergo structural rearrangement and fusion with host cell membranes mediated by heptad repeat domains in S2. Since endosomal S1/S2 cleavage occurs in an acidified environment, a protonated histidine at position 681 of the UK variant B.1.1.7 has the potential to influence the rate of spike protein cleavage and subsequent membrane fusion mechanisms to gain cell entry.

Since SARS-CoV-2 variants with D614G, such as UK variant B.1.1.7, currently predominate globally, we asked if the interface between individual chains of the spike protein trimer at position 614 may be druggable. A model of the spike protein trimer of the UK variant B.1.1.7 was used as the basis for molecular docking. 1,207 FDA approved drugs were docked to the interface site formed between position D614G and T859 of neighboring chains in the spike protein trimer. Drugs were estimated to bind the UK variant B.1.1.7 at the D614G site, such as the antileprosy drug sulfoxone, ΔG -24.4 kcal/mol (Figure 4). These data suggest that drugs may be developed to target highly transmissible SARS-CoV-2 strains including the UK variant B.1.1.7.

As mutations that distinguish the UK variant B.1.1.7 may influence immune responses, we analyzed solvent accessibility of positions that differ with Wuhan-Hu-1 (Table 1). Although N501Y has the potential to influence neutralizing antibody binding, this semiconservative difference is located at the edge of the ACE2/spike protein interface, therefore not expected to dramatically alter neutralizing antibody responses. T cells recognize SARSCoV- 2 peptides in the context of multiple HLA molecules [5], suggesting that differences between peptides derived from the Wuhan-Hu-1 spike protein or UK variant B.1.1.7 are not expected to radically influence the overall function of polyclonal T cell responsiveness to infection. However, surveillance of neutralizing antibody responses and T responses to peptides derived from the UK variant B.1.1.7 in COVID-19 patients will be required to understand functional effects on immune recognition.

Over the past year, the virus has adapted variant forms with improved fitness. Mutations that distinguish UK variant B.1.1.7 increase virus fitness through mechanisms that likely include: 1) increased affinity for binding to the coronavirus receptor ACE2, and 2) alteration of intermolecular contacts between subunits of the spike protein trimer. Changes in intermolecular contacts are expected to improve dynamic mechanisms involved in spike protein cleavage, structural rearrangement and host cell membrane fusion. Emerging mutations such as D614G can serve as the basis for drug discovery efforts to target specific highly transmissible variants such as UK variant B.1.1.7.

Sequences of UK variant B.1.1.7 and reference Wuhan-Hu-1

The Wuhan-Hu-1 sequence was used as a reference (GENBANK accession number MN908947). A Threat Assessment Brief, December 20, 2020, was used to define mutations that distinguish the UK variant B.1.1.7 (https://www.ecdc.europa.eu/en/publications-data/threatassessment-brief-rapid-increase-sars-cov-2-variantunited- kingdom).

Modeling mutations in the UK variant B.1.1.7 spike glycoprotein

The cryoEM structure of the SARS-CoV-2 (Wuhan-Hu-1) RBD/ACE2-B0AT1 complex [3] (PDB 6M17) was used as the basis for modeling N501Y. The prefusion SARS-CoV-2 (Wuhan-Hu-1) spike glycoprotein with a single receptorbinding domain up [6] (PDB 6VSB) was used as the basis for modeling D570A, D614G, T716I, S982A and D1118H. Side chains were mutated in COOT [7] using rotamers that represent a local energy minimum of torsional angles.

Molecular docking

We used the atomic model of the UK variant B.1.1.7 spike protein as the basis for molecular docking. To prepare the site for docking, all water molecules were removed. We explored the molecular surface of the structure using sets of spheres to describe potential binding pockets at the UK variant B.1.1.7 interface between the equivalent of chains A and B of PDB 6VSB. The sites selected for molecular docking were defined using the SPHGEN program in DOCK [8,9], which generates a grid of points that reflect the shape of the selected site, then filtered through CLUSTER. The CLUSTER program groups the selected spheres to define the points that were used by DOCK6.7 (UCSF) to match potential ligand atoms with spheres. Intermolecular AMBER energy scoring (van der Waals plus columbic), contact scoring, and bump filtering were implemented in the DOCK program algorithm. Atomic coordinates for 1,207 FDA approved small were positioned in a structural pocket located at position G614 of chain A, and position T859 in chain B. Each drug was docked in 1,000 different orientations and scored on the basis of predicted polar (hydrogen bond) and nonpolar (van der Waals) interactions. The most favorable orientation and scores (contact and electrostatic) were calculated. PyMOL (https://pymol.org/2/) was used to generate molecular graphic images and animation (Supplemental material).

The Global Virus Network and Christian Brechot are thanked for providing the UK variant B.1.1.7 sequence. Supported by a COVID-19 Rapid Response Pilot Project, University of Florida Clinical and Translational Science Institute, which is supported in part by the NIH National Center for Advancing Translational Sciences under award number UL1TR001427.