Extended Data Fig. 3: hACE2 knock-in mouse generation, and infection with SARS-CoV-2^D614 and SARS-CoV-2^G614.

a, Design and generation of hACE2 knock-in mice. The human ACE2 cDNA was inserted in-frame with the endogenous initiation codon of mouse Ace2 in exon 2, which was deleted. The human ACE2 cDNA was flanked 5′ with a loxP site (black triangle) and 3′ with a FRT–neomycin–FR–-loxP cassette. The targeting construct included a negative selection cassette (PGK-DTA) to improve selection of clones with homologous recombination. Chimeric male mice transmitting the targeted locus were crossed with Flp deleter female mice to generate the floxed human ACE2 knock-in allele. This allele can be used: (1) without further Cre-mediated recombination (as shown here) to study hACE2 knock-in mice, in which the human ACE2 cDNA is expressed in place of mouse Ace2; (2) after crossing with a Cre-deleter mouse line to generate constitutive Ace2-knockout mice; and (3) after crossing with a tissue-specific Cre line. Ubiquitous and tissue-specific knockout mice can be crossed with conventional hACE2 transgenic mice to remove the endogenous mouse ACE2, which could confound pathogenesis studies (owing to heterodimerization of ACE2). b, Loss of body weight at indicated time points after infection of hACE2 knock-in mice (n = 8) and wild-type mice (n = 9), and for mock-infected wild-type mice sampled identically (n = 10). c, d, RT–qPCR analysis (c) and viral titres (d) of tissue homogenates of inoculated hACE2 knock-in and wild-type mice at indicated time points. Olf., olfactory. e, Percentage of sequencing reads encoding S or S(D614G). Each square represents data for one individual mouse in a competition experiment. For each time point, a linear regression model was generated on the basis of the sequence read counts for S and S(D614G). P values were calculated for the group (variant) coefficient. P values (left to right): ***P = 0.0009; **P = 0.0020; NS, P = 0.7875; and *P = 0.0180. f, Cat’s eye plot illustrating the relative replicative fitness values of SARS-CoV-2^D614 over SARS-CoV-2^G614 from oropharyngeal swabs of hACE2 knock-in mice in the competition experiment shown in Fig. 2. Ratios of SARS-CoV-2^G614 to SARS-CoV-2^D614 were measured after competition using the MinION sequencing platform at the time point indicated on the plot. Each dot represents one infected hACE2 knock-in mouse, the centre line represents the mean and the shaded area represents the s.d.

Source data