SARS‐CoV‐2 B.1.617 Indian variants: Are electrostatic potential changes responsible for a higher transmission rate?

Abstract Lineage B.1.617+, also known as G/452R.V3 and now denoted by WHO with the Greek letters δ and κ, is a recently described SARS‐CoV‐2 variant under investigation first identified in October 2020 in India. As of May 2021, three sublineages labeled as B.1.617.1 (κ), B.1.617.2 (δ), and B.1.617.3 have been already identified, and their potential impact on the current pandemic is being studied. This variant has 13 amino acid changes, three in its spike protein, which are currently of particular concern: E484Q, L452R, and P681R. Here, we report a major effect of the mutations characterizing this lineage, represented by a marked alteration of the surface electrostatic potential (EP) of the receptor‐binding domain (RBD) of the spike protein. Enhanced RBD‐EP is particularly noticeable in the B.1.617.2 (δ) sublineage, which shows multiple replacements of neutral or negatively charged amino acids with positively charged amino acids. We here hypothesize that this EP change can favor the interaction between the B.1.617+ RBD and the negatively charged ACE2, thus conferring a potential increase in the virus transmission.


| INTRODUCTION
The coronavirus disease , caused by the new coronavirus SARS-CoV-2, continues to spread worldwide, with more than 163 million infections and about 3.5 million deaths as of May 17, 2021 (www.who.int). To fight this dreadful disease, several safe and efficacious vaccines against SARS-CoV-2 are being used with remarkable effectiveness in some countries and limited availability in others. In particular, the capacity of some countries to halt SARS-CoV-2 spread is still limited due to inadequate resources and vaccination infrastructures. 1,2 In this scenario, several SARS-CoV-2 variants have been identified and have become a global threat. Some of them have been classified as variants of concern (VOCs) due to their mutations in the S1 subunit of the spike (S) protein, particularly in its receptor-binding domain (RBD). [3][4][5] One of them, identified as B.1.1.7 (α), also known as the UK variant, bears a substitution of asparagine with tyrosine on the position 501 and deletion of two amino acids in the position 69-70 of the S1 subunit. This variant has quickly spread in several European countries to become globally dominant. 5 Other VOCs have been isolated in South Africa and Brazil and have been studied for their enhanced contagiousness and resistance to neutralization by antibodies from convalescent and vaccine-recipient subjects [6][7][8] (Table 1). Quite recently, a new variant under investigation (VUI) has been isolated from Maharashtra, India, in a setting of the highly diffusive epidemic with devastating proportions. This variant, identified as B.1.617, carries several nonsynonymous mutations. Two of them, the E484Q (or the P478K) and the L452R, are located in the RBD region, and they are critical sites for the binding with ACE2. Initial data suggest these mutations could confer increased transmission and immune evasion. 9 We then compare these changes with other VOCs to establish whether and to what extent those amino acid changes can influence the interaction of the spike protein with ACE2, thus potentially affecting SARS-CoV-2 transmission and immune-escape properties.

| METHODS
To perform a robust analysis three different three-dimensional structures of SARS-CoV-2 spike glycoprotein have been downloaded from Protein Data Bank with the following characteristics:  of the wild-type and VOCs SARS-CoV-2 spike receptor-binding domain (RBD) and evaluate potential differences in terms of molecular interaction with ACE2 receptor. The results have been reported within a range between −5 and +5 kT/e.

| Protein stability
In silico prediction of the mutation impact on the RBD stability has been carried out with DynaMut. Three alternative RBD structures denoted by the PDB codes 6M17, 6M0J, and 6XC4 have been tested.
These structures display small differences in the conformation of loops, especially in the one inside the receptor-binding motif (RBM).
According to the parameters of our in silico experiments, the output

| Surface and interface analysis
The | 6553 484, and 501 are within the RBM, containing residues that bind to ACE2. In contrast, mutant position 417 is located outside the motif. 15 According to our analysis, the mutations in positions 417, 484, and 501 might increase the spike binding affinity with the ACE2 receptor. In particular, the Tyr replacing Asn501 may form an aromatic interaction with ACE2 Tyr41, a hydrogen bond with ACE2 D38, and a potential cation-π interaction with ACE2 Lys353. In addition, substitutions of Glu484 with Lys or Gln may form hydrophobic interactions to Ile472, Gly482, Phe486, Cys488, and Tyr489.
Our data also indicate that replacing Lys or Gln with Glu484 abolishes the interfacial salt bridge between Glu484 and ACE2 Lys31.
Due to the fact that Lys417 is solvent-exposed and forms salt-bridge

| Electrostatic potential
We note that a major, global effect of the mutations characterizing  17 To further justify our assumption, we note that the intensely investigated D614G substitution of the spike protein, early reported in Italian isolates, [18][19][20] and subsequently attributed with increased virus transmissibility 21 was found to enhance the protein torsional ability and potentially affecting its stability. 21  which, in turn, would then increase affinity for the ACE2 receptor. All of these changes have the potential to eventually modify infectivity, pathogenicity, and virus spread. Regarding differential binding to neutralizing antibodies, previous studies suggested that VOCs RBD changes in the electrostatic potential surface could induce SARS-CoV-2 antibody evasion, and even single amino acid changes that marginally affect ACE2 affinity could greatly influence nAbs affinity. 27 Several factors have been demonstrated to affect the impact of VOCs. For example, it has been observed an increased effect at pH associated with nasal secretions (from 5.5 to 6.5). 28

ACKNOWLEDGMENTS
The authors are grateful to Alice Vinciguerra for her precious assistance. We also would like to thank all the authors who have kindly deposited and shared genome data on GISAID. Stefano Pascarella has been partially supported by the Sapienza grant RP120172B49BE24.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in bioRxiv at https://submit.biorxiv.org/.