Mutants of human ACE2 differentially promote SARS-CoV and SARS-CoV-2 spike mediated infection.

SARS-CoV and SARS-CoV-2 encode spike proteins that bind human ACE2 on the cell surface to enter target cells during infection. A small fraction of humans encode variants of ACE2, thus altering the biochemical properties at the protein interaction interface. These and other ACE2 coding mutants can reveal how the spike proteins of each virus may differentially engage the ACE2 protein surface during infection. We created an engineered HEK 293T cell line for facile stable transgenic modification, and expressed the major human ACE2 allele or 28 of its missense mutants, 24 of which are possible through single nucleotide changes from the human reference sequence. Infection with SARS-CoV or SARS-CoV-2 spike pseudotyped lentiviruses revealed that high ACE2 cell-surface expression could mask the effects of impaired binding during infection. Drastically reducing ACE2 cell surface expression revealed a range of infection efficiencies across the panel of mutants. Our infection results revealed a non-linear relationship between soluble SARS-CoV-2 RBD binding to ACE2 and pseudovirus infection, supporting a major role for binding avidity during entry. While ACE2 mutants D355N, R357A, and R357T abrogated entry by both SARS-CoV and SARS-CoV-2 spike proteins, the Y41A mutant inhibited SARS-CoV entry much more than SARS-CoV-2, suggesting differential utilization of the ACE2 side-chains within the largely overlapping interaction surfaces utilized by the two CoV spike proteins. These effects correlated well with cytopathic effects observed during SARS-CoV-2 replication in ACE2-mutant cells. The panel of ACE2 mutants also revealed altered ACE2 surface dependencies by the N501Y spike variant, including a near-complete utilization of the K353D ACE2 variant, despite decreased infection mediated by the parental SARS-CoV-2 spike. Our results clarify the relationship between ACE2 abundance, binding, and infection, for various SARS-like coronavirus spike proteins and their mutants, and inform our understanding for how changes to ACE2 sequence may correspond with different susceptibilities to infection. Author summary: SARS-like coronaviruses, such as SARS-CoV-2, use their spike proteins to bind a common surface on the human ACE2 protein to gain entry and subsequently infect cells. We used site-specific genomic integration and expression of WT ACE2 or its missense variants, many of them previously observed in human exomes, to determine how ACE2 sequence and abundance correspond to infectability by SARS-CoV or SARS-CoV-2. We found that reduced binding only partially corresponded to infection, and mainly only at lower ACE2 abundance levels. We observed some human ACE2 variants differentially affect SARS-CoV, SARS-CoV-2, or SARs-CoV-2 N501Y spike variant pseudovirus entry, showing that each viral spike binds ACE2 in a unique manner. Our results provide improved quantitative understanding for how ACE2 sequence and abundance correlate with infectivity, with implications for how natural human ACE2 variants, or variants observed in related species, may impact susceptibility to infection. These genetic tools can be repurposed to characterize future SARS-CoV-2 spike variants, or to better understand how receptor protein sequences correspond with entry by zoonotic viruses during cross-species transmission events. [ABSTRACT FROM AUTHOR]