Laboratory Cell Culture Footprints in the SARS-CoV-2 Glycan Shield V2

Laboratory Cell Culture Footprints in the SARS-CoV-2 Glycan Shield The origin of SARS-CoV-2 is still disputed. Structural and evolutionary evidence often point in different directions. Early arguments for a natural origin, especially those in Andersen et al. (2020), claimed that features such as the furin cleavage site and its predicted O-linked glycans could only have evolved under pressure from a living animal’s immune system. Later findings have weakened this claim. The predicted O-linked glycans do not form the protective shield originally described, and their presence does not rule out adaptation in cell culture. Laboratory passaging in different cell lines puts strong selective pressure on the spike protein, even without an immune system. In Vero E6 cells, which lack the TMPRSS2 enzyme, the furin cleavage site is quickly lost. In human lung and intestinal cells such as Calu-3 and Caco-2, the same site is kept and refined. At the same time, SARS-CoV-2 is missing a conserved glycan shield at position N370 because of a mutation called T372A. When researchers put this glycan back, the spike locks shut and the virus replicates more than 60 times less efficiently in human lung cells. Together with the furin cleavage site insertion, this change produces a coordinated functional outcome: greater spike openness and efficient, regulated processing that optimises entry into human airway cells. Early SARS-CoV-2 also carried additional Spike changes relative to the closest bat sarbecoviruses, including N519H and NTD glycan remodeling (S32F and related differences). Together these modifications favour more open Spike conformations and reduced compatibility with the low-pH bat gastrointestinal environment. Pre-2020 laboratory work had already established methods for removing specific glycan shields, including the precise structural equivalent of the N370 site. Vaccine-oriented studies by collaborators of the Wuhan Institute of Virology deliberately ablated this glycan (mutating the 372 site to alanine), and the 2018 DEFUSE proposal explicitly planned similar N-glycan manipulations to study host-range effects. Recent evolutionary analyses (Havens et al., 2026) found no clear signs of strong natural selection on the virus’s lineage just before it appeared in humans. This silent pre-human stem stands in contrast to the clear adaptive signals (elevated evolutionary rates and host-specific mutations) that appear whenever SARS-CoV-2 later enters other mammals such as mink and white-tailed deer. The pattern suggests that the SARS-CoV-2 already carried its key human-adapted features at the moment it entered the human population., When we consider together the Spike changes, the absence of a selective footprint on the pre-human stem, and the contrasting adaptive signals seen in secondary mammalian hosts, the data fits better with adaptation in laboratory cell culture (Calu 3, HAE), where immune and stomach constraints are missing, than with a prolonged natural journey through an as yet undetected intermediate animal host. Acknowledgements I would like to thank Yuri Deigin, Daoyu Zhang, Steven Massey, Valentin Bruttel and Gilles Demaneuf for their insights and help while I prepared this paper. Also thanks to Guy Gadboit for his enlightening feedback. (Received revisions in Version 2 to correct an oversight regarding codon usage statistics at the S1/S2 junction. The central thesis regarding dual-alanine functional and phenotypic synergy remains unchanged.)