New findings with the IBV decoy for cell entry inhibition of SARS-CoV-2, and unique structural data for soluble dimeric ACE2 bound to the viral S trimer

Abstract

[Background] The SARS-CoV-2 spike protein (S) mediates the interaction of the virus with cellular membrane receptor (angiotensin-converting enzyme 2, ACE2). In previous PTI meetings, we reported heterologous production in vitro of the ACE2 extracellular domains modified by site-directed mutagenesis to increase its affinity for the S protein, to enable it to be used as viral entry inhibitor (decoy) by competing with the membrane-bound cellular receptor. We now test the value of these decoys for: 1) binding to S variants that emerged during the evolution of the pandemic in viral lineages of concern; and 2) inhibiting experimental cellular infection by pseudotyped virus expressing these S variants. Cellular syncytia formation has been described in several organs as a manifestation of severe COVID-19, and likely has pathogenic impact. To test further our decoys’ effectiveness, we studied their impact on cellular syncytia formation within an experimental in vitro cell culture model. Searching for effective decoys, we produced monomeric and dimeric ACE2 proteins, depending on the respective absence/presence of the extracellular collectrin domain. Interestingly, there are no reported structures of dimeric soluble ACE2 bound to the S protein. After extensive knowledge-guided trial-and-error, we succeeded in visualizing by cryo-electron microscopy (cryoEM) this interaction (~7-Å-resolution), and in understanding the challenges inherent in determining such a complex structural organization.

[Methods] 1) Recombinant production and purification of the monomeric or dimeric ACE2, their decoys the receptor binding domain (RBD) and the S protein variants of interest. We used baculovirus/insect cells to produce ACE2s and RBDs, and human Expi293F cells for the S proteins. 2) Biolayer interferometry for assessing protein-protein interactions; 3) Use of a model system for monitoring viral cellular infection and its inhibition by decoys. We used a pseudotyped engineered vesicular stomatitis virus expressing and exposing at its surface the desired S protein variant, to infect appropriate SARS-CoV-2-susceptible mammalian cells; 4) Single-particle cryoEM; 5) Syncytia formation testing using an engineered cultured cell system in which heterologous surface expression of the S protein in one cell type induces syncytium formation in other cells expressing membrane-bound ACE2.

[Results] Our decoys proved highly effective in preventing cellular infection by pseudotyped virus expressing the S proteins of different SARS-CoV-2 variants of concern. Biophysical results have validated the maintained interaction between the decoy and the various S protein variants. When introduced into the cellular model system for syncytia formation, the decoys proved capable of decreasing such formation. Puzzlingly, the monomeric decoy was more effective than the dimeric one. The cryoEM images unveiled an ACE2 dimer configuration, where the subunits, resembling the previously reported monomer, were oriented at an angle of >60º, in which the vortex was the interlinked collectrin domains. Both catalytic domains engage with a single RBD of one subunit from different S trimers. The formation of a network at high stoichiometries of both components poses a challenge for structure determination by cryoEM.

[Conclusions] Unlike therapeutic antibodies, which proved ineffective on variants not initially used for their production, our decoys should be effective in preventing infection by all widely widespread SARS-CoV-2 variants.