All experiments were approved by the local genetic manipulation safety committee of Imperial College London, St. Mary’s Campus (center number GM77) and the Health and Safety Executive of the United Kingdom, under reference CBA1.77.20.1. SARS-CoV-2 reverse genetics was performed at CVR University of Glasgow.

hACE2 ^knock-in mice were purchased from The Jackson Laboratory (Line 035000) and bred in-house at Imperial College London. All animal studies were reviewed and approved by the Animal Welfare and Ethical Review Board (AWERB) at Imperial College London and approved by the UK Home Office in accordance with the Animals Act 1986 (Scientific Procedures) and ARRIVE guidelines.

First-wave SARS-CoV-2 (D614G, isolate of hCoV-19/England/IC19/2020), Delta (B.1.617.2, isolate of hCoV-19/England/SHEF-10E8F3B/2021), and Omicron (Omicron/BA.1, isolate of hCoV-19/England/PHEP-YYNTNOA/2021) were grown in African green monkey kidney cells overexpressing human ACE2 and TMPRSS2 (Vero-ACE2-TMPRSS2; VAT cells) (18). Reverse-genetics-derived viruses were generated at the CVR as previously described (11, 19, 20). Recombinant SARS-CoV-2 Wuhan cDNA genomes bearing the Omicron BA.1 spike-encoding sequence were assembled using the transformation-associated recombination (TAR) method as described previously (11). These genomes were used as a template to in vitro-transcribed viral genomic RNA and subsequently transfected into BHK-hACE2-N cells, which stably express SARS-CoV-2 N and hACE2 for virus rescue. The rescued viruses were passaged once in Vero E6 cells and sequenced using Oxford Nanopore to confirm their identity.

For infection, mice were lightly anesthetized and instilled intranasally (i.n.) with 2 × 10⁵ plaque-forming units (PFU) of SARS-CoV-2 or phosphate-buffered saline (PBS) in 100 µl.

SARS-CoV-2 titer was assessed in lungs at 0.75, 2, and 7 d.p.i. using a plaque assay. In brief, serial dilutions of lung homogenate in serum-free Dulbecco’s modified Eagle medium [DMEM, containing 1% non-essential amnio acids (NEAA), 100 U/mL penicillin, and 100 µg/mL streptomycin) were performed and inoculated onto VAT cells for 1 h at 37°C. The inoculum was then removed and replaced with overlay medium (1× MEM, 0.2% w/v BSA, 0.16% w/v NaHCO₃, 10 mM HEPES, 2 mM L-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, and 0.84% agarose). Plates were incubated for 3 days at 37°C before the overlay was removed, and cells were stained for 1 h at room temperature in 0.5% crystal violet solution. The virus plaques were counted and multiplied by the dilution factor to calculate titer as plaque-forming unit/mL (PFU/mL).

Mice were sacrificed at 0.75, 2, and 7 d.p.i., and lungs were perfused with PBS. To obtain lung leukocytes, lung lobes were cut into smaller pieces and incubated in complete DMEM (cDMEM, supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin), 1 mg/mL Collagenase D (Roche), and 30 µg/mL DNase I (Invitrogen) for 1 h at 37°C and then mashed through a 100-µm filter (BD). Red blood cells were lysed using ammonium chloride potassium buffer.

BAL was collected by flushing the lungs three times with 1 mL PBS supplemented with 5 mM EDTA (Life Technologies). The BAL cells and supernatant were separated by centrifugation, and the BAL supernatants were exposed to UV light for 2 min to inactivate SARS-CoV-2. Red blood cells were lysed using ammonium chloride potassium buffer.

Lung tissue was homogenized in TRIzol, and RNA extraction was performed according to the manufacturer’s instructions. After the chloroform step, the aqueous phase containing RNA was further processed using RNeasy Mini Kit (QIAGEN) according to the manufacturer’s instructions. Furthermore, 2 µg RNA was reverse-transcribed using a high-capacity RNA-to-cDNA kit (Applied Biosystems) according to the manufacturer’s instructions. To quantify the mRNA levels in lung tissue, quantitative RT-PCR was performed using QuantiTect Probe PCR Master Mix (Qiagen) and 7500 Fast Real-Time PCR System (Applied Biosystems). For Ifna, Cxcl1, Cxcl10, Mx1, Ccl2, and Il6 (all primer–probe mixes from Applied Biosystems) and SARS-CoV-2, N and E gene (21) expression was calculated relative to the expression of Gapdh. First, the ΔCT (CT = cycle threshold) between the target gene and Gapdh was calculated for each sample, followed by the calculation of 2^-ΔCT. The gene expression for Oas1, Viperin, Ifnb, and Ifnl was performed using primers and probes as previously described (22). For the absolute quantification of these genes, the exact number of copies of the genes was calculated using a plasmid DNA standard curve, and the results were normalized to the levels of Gapdh (Applied Biosystems).

Statistical analysis was performed using Prism 6 (GraphPad Software). One-way ANOVA with Tukey’s post-hoc test was used to compare multiple groups. Data are expressed as mean ± SEM, and for all tests, a value of P <0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001).

To investigate virus replication and interferon-stimulated gene (ISG) induction in hACE2 ^knock-in mice upon infection with SARS-CoV-2 Delta or Omicron (BA.1) variants compared to first-wave SARS-CoV-2 (D614G), mice were infected by inoculation with 2 × 10⁵ plaque-forming units (PFU) of each SARS-CoV-2 variant or mock infected with PBS. Lungs were harvested at days 0.75, 2, and 7 post-infection (p.i.) and homogenized ( Figure 1A ). As reported for other mouse strains (16, 23), infection with SARS-CoV-2 D614G, Delta, or BA.1 did not induce any weight loss in hACE2 ^knock-in mice ( Supplementary Figure S1A ). Plaque assays on Vero cells overexpressing ACE2 and transmembrane protease serine 2 precursor TMPRSS2 (VAT cells) revealed high titers, 10⁶ PFU or more, of infectious virus in the lungs of SARS-CoV-2 D614G and Delta VOC-infected mice on days 0.75 and 2 p.i. that were not completely cleared by day 7 p.i. ( Figure 1B ). In contrast, early infectious viral loads were significantly lower in the lungs of mice infected with SARS-CoV-2 Omicron BA.1 and below the detection limit at day 7 p.i. ( Figure 1B ). This pattern was confirmed by the gene expression analysis of SARS-CoV-2 nucleocapsid (N) and envelope (E) genes. While N and E gene expression was detected in all SARS-CoV-2-infected groups ( Figure 1C ; Supplementary Figure S1B ), viral gene expression was significantly lower in SARS-CoV-2 Omicron BA.1-infected mice compared to the other groups, and day 2 viral RNA loads were significantly higher for Delta than for the other variants ( Figure 1C ). Since the type I and III interferon (IFN) responses are the earliest host responses to be triggered upon viral infection, we next measured the gene expression of Ifna5, Ifnb, and Ifnl in lung tissue. For Ifna5 gene expression, there were no significant differences, although levels of mRNA were highest at day 2 in Delta-infected mice ( Figure 1D ). Correlating with the higher viral RNA loads ( Figure 1C ), Ifnb and Ifnl gene expression was significantly higher in SARS-CoV-2 Delta-infected mice at day 2 p.i. compared to D614G-infected mice, while no Ifn gene expression was detected in the lungs of SARS-CoV-2 Omicron BA.1-infected mice ( Figure 1D ). To investigate downstream ISG expression, we analyzed the mRNA levels of antiviral effectors Mx1, Oas1, and Viperin in lung tissue, which were all highly increased in SARS-CoV-2 D614G and Delta-infected mice at day 2 p.i. but returned to baseline by day 7, while gene expression was not detected in SARS-CoV-2 Omicron BA.1-infected mice ( Figure 1E ). The same gene expression pattern was detected for the type I IFN-induced chemokine Cxcl10 ( Supplementary Figure S1C ). We also assessed the gene expression of pro-inflammatory mediators Cxcl1, Ccl2, and Il6 and found that while gene expression was highly increased in the lungs of SARS-CoV-2 D614G-infected mice, it was significantly lower in SARS-CoV-2 Delta-infected mice and not induced in SARS-CoV-2 Omicron BA.1-infected mice ( Figure 1F ).

The increased binding of the Omicron spike to mouse ACE2 has been extensively reported by others and has even led to suggestions of a rodent origin for the Omicron VOC (24). We were interested to test whether such increased binding would counteract the otherwise poor replication of Omicron BA.1 in mice, as seen in the hACE2 ^knock-in mice ( Figure 1 ). Therefore, we next infected C57BL/6 WT mice with the same titer of the three SARS-CoV-2 variants; first-wave D614G, Delta, or Omicron, and assessed viral load and innate immune responses at day 2 p.i. This time point was chosen since this was the time point that showed the most significant differences in the previous experiment ( Figure 2A ). Infectious virus in the lungs was detected by plaque assay ( Figure 2B ) and SARS-CoV-2 N and E gene expression by RT-PCR as before ( Figure 2C ). In contrast to the virus replication patterns in the hACE2 ^knock-in mice, Omicron replicated to the highest titers in WT mice, attaining titers between 10⁴ and 10⁵ PFU/mL lung extract. Furthermore, only infection with Omicron resulted in a significant induction of Ifna5 and Ifnb ( Figure 2D ), ISG-driven antiviral effectors Mx1, Oas1, and Viperin ( Figure 2E ), and pro-inflammatory mediators Cxcl10 and Il6 ( Supplementary Figure S2 ).

Since SARS-CoV-2 Omicron BA.1 efficiently infected WT mice and induced innate responses ( Figure 2 ), we next investigated the impact of type I IFN receptor signaling on the outcome of SARS-CoV-2 Omicron BA.1 infection. Using hACE2-AAV transduced Ifnar1^-/- and WT mice infected with SARS-CoV-2 D614G, we have previously shown that type I IFN receptor signaling impairment results in higher viral loads and altered innate immune responses, with particularly increased neutrophil and Ly6C^- inflammatory myeloid cell recruitment to the lung (21). Here WT or Ifnar1^-/- mice were infected with 2 × 10⁵ PFU SARS-CoV-2 Omicron BA.1, and lungs were harvested at day 2 p.i. as before ( Figure 3A ). While the infectious viral titers measured by plaque assay were not significantly higher in Ifnar1^-/- compared to WT mice ( Figure 3B ), SARS-CoV-2 N and E gene expression was significantly increased in Ifnar1^-/- mice ( Figure 3C ). The gene expression of type I IFNs—Ifna5 and Ifnb—was variable and not significantly increased in WT-infected versus uninfected animals, although it was consistently low in Ifnar1^-/- mice ( Figure 3D ). As expected in Ifnar1^-/- mice, the gene expression of the antiviral ISGs—Mx1, Oas1, and Viperin ( Figure 3E )—as well as Cxcl10 ( Supplementary Figure S3 ) was decreased compared to WT mice infected with SARS-CoV-2 Omicron BA.1. The gene expression of pro-inflammatory mediators Cxcl1, Ccl2, and Il6 was not significantly altered in this model at day 2 p.i. with SARS-CoV-2 Omicron BA.1 in either WT or Ifnar1^-/- mice ( Supplementary Figure S3 ).

To validate that the increased infection by the Omicron BA.1 in WT mice was indeed due to the mutations in the Spike protein, we used a chimeric virus in which the Omicron BA.1 S gene was substituted in the SARS -CoV-2 first-wave genome in place of homologous D614G S. WT mice were intranasally administered either 5 × 10⁴ PFU SARS-CoV-2 D614G, Omicron BA.1 or D614G^{[Omicron-spike]}, or PBS, and lungs were harvested at day 2 p.i. ( Figure 4A ). Note that this infection dose was lower than that used in the experiments in the previous figures due to the lower titers of the chimeric virus stock. Viral load was detectable in the lungs of SARS-CoV-2 D614G^{[Omicron spike]}- but not SARS-CoV-2 D614G-infected mice by both plaque assay ( Figure 4B ) and N and E gene expression assay ( Figure 4C ), and titers of chimeric virus were significantly increased compared to Omicron BA.1. While the gene expression of Ifna5 was variable, Ifnb gene expression was consistently and significantly increased in SARS-CoV-2 D614G^{[Omicron spike]}- as opposed to Omicron BA.1-infected mice ( Figure 4D ). Furthermore, SARS-CoV-2 D614G^{[Omicron spike]} infection resulted in a significantly increased expression of ISGs Mx1, Oas1, and Viperin ( Figure 4E ) as well as Cxcl10 ( Supplementary Figure S4 ), while pro-inflammatory mediators Cxcl1, Ccl2, and Il6 were not significantly altered in any group compared to the uninfected mice ( Supplementary Figure S4 ). These results confirm that attenuation of Omicron BA.1 in mice is at least partly mediated by non-spike genes. Combining the more efficient entry of virus bearing the Omicron BA.1 spike into cells expressing murine ACE2, with the less mouse attenuated non-spike backbone of the first-wave SARS CoV-2 virus, resulted in a more robust infection in WT mice.