TMEM106B-mediated SARS-CoV-2 infection allows for robust ACE2-independent infection in vitro but not in vivo

Angiotensin-converting enzyme 2 (ACE2) is the primary entry receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), but ACE2-independent entry has been observed in vitro for strains with the spike-E484D substitution. Here, we conduct a whole-genome CRISPR-Cas9 knockout screen using SARS-CoV-2 mouse adapted 1 (SARS-CoV-2_MA1), which carries spike-E484D, to identify the ACE2-independent entry mechanisms. SARS-CoV-2_MA1 infection in HEK293T cells relies on heparan sulfate and endocytic pathways, with TMEM106B, a transmembrane lysosomal protein, the most significant contributor. While SARS-CoV-2_MA1 productively infects human brain organoids and K18-hACE2 mouse brains, it does not infect C57BL/6J or Ifnar^-/- mouse brains. This suggests that ACE2-independent entry via TMEM106B, which is predominantly expressed in the brain, does not overtly increase the risk of SARS-CoV-2 neuroinvasiveness in mice with endogenous Ace2 expression. Importantly, SARS-CoV-2_MA1 does not replicate in the Ace2^-/- mouse respiratory tract. Overall, this suggests that robust ACE2-independent infection by SARS-CoV-2_MA1 is likely an in vitro phenomenon with no apparent implications for infection in vivo.