summary - Reports 3 sarbecoviruses (BANAL-52, -103, -236) sampled from Rhinolophus bats in caves in northern Laos (645 bats sampled across 6 families/46 species), the closest known relatives of SARS-CoV-2 identified to date: BANAL-52 is 96.8% identical to SARS-CoV-2 across the genome, edging out RaTG13’s 96.2%. The spike gene is a mosaic - different BANAL genomes are each closest to SARS-CoV-2 across different spike sub-regions (S1-NTD, S1-RBD, S2). BANAL-236’s receptor-binding domain binds human ACE2 with affinity similar to SARS-CoV-2’s, and the virus replicates in human cells in vitro - direct lab evidence that SARS-CoV-2-like spike/RBD combinations capable of infecting human cells circulate naturally in wild bats. None of the three BANAL viruses carries a furin cleavage site.
relevance_note - Strengthens the natural-reservoir case (a human-ACE2-binding, human-cell-infectious SARS-CoV-2-like virus exists in nature, independent of any lab), but the absence of a furin site in even these closest known relatives leaves the furin site’s own origin unresolved by this data alone - cited affirmatively by zoonosis-favoring arguments, and (more cautiously) noted by lab-leak-favoring commentary on the latter point. Per the recurring-dataset note: a distinct “closest relative” data-basis from RaTG13 (different expedition, institution, and country) - don’t conflate the two.