SARS-CoV-2 Omicron is an immune escape variant with an altered cell entry pathway

Abstract

Vaccines based on the spike protein of SARS-CoV-2 are a cornerstone of the public health response to COVID-19. The emergence of hypermutated, increasingly transmissible variants of concern (VOCs) threaten this strategy. Omicron (B.1.1.529), the fifth VOC to be described, harbours multiple amino acid mutations in spike, half of which lie within the receptor-binding domain. Here we demonstrate substantial evasion of neutralization by Omicron BA.1 and BA.2 variants in vitro using sera from individuals vaccinated with ChAdOx1, BNT162b2 and mRNA-1273. These data were mirrored by a substantial reduction in real-world vaccine effectiveness that was partially restored by booster vaccination. The Omicron variants BA.1 and BA.2 did not induce cell syncytia in vitro and favoured a TMPRSS2-independent endosomal entry pathway, these phenotypes mapping to distinct regions of the spike protein. Impaired cell fusion was determined by the receptor-binding domain, while endosomal entry mapped to the S2 domain. Such marked changes in antigenicity and replicative biology may underlie the rapid global spread and altered pathogenicity of the Omicron variant.

Fig. 1. Spike amino acid changes, phylogeny and emergence of the Omicron variant.

a, Spike homotrimer in open conformation with locations of Omicron substitutions, deletions (Δ) or insertions (ins) present in the lineages BA.1 and BA.2 (superscripts 1,2). Mutated residues highlighted as spheres with opaque surface representation. Residues impacting RBD-specific antibodies of classes 1, 2 or 3; belonging to the NTD antibody supersite (NTD SS); or comprising the furin-cleavage site, are coloured, the remainder in grey. Inset shows the wider extent of these sites, with remaining areas of the protein shaded to show the three monomers. Mutations are annotated on the monomer with an ‘up’ receptor-binding domain with D614G (italicized), which is shared by common descent by all lineage B.1 descendants. Visualization uses a complete spike model based on a partial cryo-EM structure (RCSB Protein Data Bank (PDB) ID: 6VSB). b, Heat maps showing properties of amino acid residues in the Omicron variants BA.1 and BA.2. Structure-based epitope scores for residues in the spike structure in closed and open conformations are shown. For RBD residues, DMS studies show the escape fraction (quantitative measure of the extent to which a mutation reduced polyclonal antibody binding) for each mutant-averaged (‘plasma average’) and most sensitive plasma (‘plasma max’). Each mutation is classified as having mutations affecting neutralization by either monoclonal antibodies (mAbs)^,,,, or antibodies in convalescent plasma (infected or vaccinated^,,,). Membership of the furin cleavage site is shown. Distance to ACE2-contacting residues forming the receptor-binding site (RBS) is shown (‘RBS’ is residues with an atom <4 Å of an ACE2 atom in the structure of RBD bound to ACE2) (RCSB PDB ID: 6M0J). ACE2 binding scores represent the binding constant (Δlog₁₀ KD) relative to the wild-type reference amino acid from DMS experiments. c, Inferred evolutionary relationships of SARS-CoV-2 from NextStrain (https://nextstrain.org/ncov/gisaid/global), with the VOCs labelled. Tree tip colours correspond to the number of mutations causing spike amino acid substitutions relative to the Wuhan-Hu-1 (lineage B) sequence. d, The proportion of genome sequences from Scotland sampled between 1 September 2021 and 29 January 2022 is shown: ‘Delta’ includes all B.1.617.2 and AY assigned sequences that are not in the Delta sub-lineage AY.4.2 (defined by spike mutations Y145H and A222V), and BA.1, BA.1.1 and BA.2 are Omicron sub-lineages.

Fig. 2. Antibody responses elicited by SARS-CoV-2 vaccination.

a,b, Antibody responses were studied in three groups of individuals (n = 24 per group) receiving primary vaccination with either BNT162b2, ChAdOx1 or mRNA-1273 by MSD-ECL assay (a) or pseudotype-based neutralization assay (b). a, Responses were measured against full-length spike glycoprotein (Spike), RBD, NTD and N, and are expressed as arbitrary units (a.u. ml⁻¹). Horizontal bar represents group mean; between group comparisons by ordinary one-way analysis of variance (ANOVA), Tukey’s multiple comparisons test; ****P < 0.0001. b, NAb responses were quantified against B.1 lineage (D614G) or Omicron (BA.1) spike glycoprotein bearing HIV (SARS-CoV-2) pseudotypes. Each point represents the mean of three replicates, horizontal bar represents the group mean; between group comparisons by ordinary one-way ANOVA, Tukey’s multiple comparisons test; **P = 0.0075, ****P < 0.0001. c–e, To assess the effect of booster vaccines, antibody responses were studied in two groups of individuals primed with two doses of either BNT162b2 or ChAdOx1 and boosted with either BNT162b2 or mRNA-1273. Reactivity against SARS-CoV-2 antigens was measured by MSD-ECL assay (c), while neutralizing activity (d,e) was measured using HIV (SARS-CoV-2) pseudotypes, as above. Yellow data points represent those boosted with mRNA-1273, all others received BNT162b2. In d and e, ‘BA.1 neutralizing (%)’ refers to the proportion of serum samples that displayed neutralizing activity against Omicron pseudotypes. Source data

Fig. 3. Neutralization of Omicron variants by third dose sera.

a–c, Sensitivity of B.1, Omicron BA.1, BA.1.1 and BA.2 to neutralization by sera elicited following third dose booster vaccination. a, All third dose sera irrespective of vaccine type. Each point represents the mean of three technical replicates, bar represents the mean. Group means compared by one-way ANOVA, Tukey’s multiple comparisons test; ****P < 0.0001. b, Neutralizing activity between matched second and third dose vaccine sera. Box and whisker plots (median and range). c, Comparison of matched sera after second or third dose vaccination. Each point represents the mean of three technical replicates ± s.e.m. d,e, Activity of Ronapreve (casirivimab and imdevimab) (d) and Xevudy (sotrovimab) (e) against pseudotypes bearing B.1, BA.1, BA.1.1, BA.2 and SARS-CoV-1 spike glycoproteins. Each point represents the mean of three technical replicates ± s.e.m. CPS, counts per second. f, Interferon-ɣ production by PBMC from vaccinated individuals in response to B.1 or BA.1 S-derived peptides. Dose 3 boost with BNT162b2 (circles) or mRNA-1273 (squares). Source data

Fig. 4. Vaccine deployment and vaccine effectiveness.

a, Boxplots of date of first, second and third administered vaccine dose by vaccine product for the population of NHS GG&C aged 18 years and older. The box limits are the quartiles and the centre line is the median, with whisker length of 1.5 times the interquartile range. Outliers are shown as dots outside the whisker range. Data points are overlaid as a dot plot with points shown as black dots, with a random jitter along the x axis applied for visual clarity. b, Denominator (violin) plot showing populations of test positive and test negative cohorts in NHS GG&C, with the widths of the grey bands representing the populations in each group at each time point. VOC classifications of sequenced cases are overlaid as a dot plot, with points coloured by their VOC and a random jitter applied along the x axis for visual clarity. c, Error bar plot of estimated vaccine effectiveness against testing positive for Delta and Omicron BA.1 SARS-CoV-2 infection in the population of over 18 years in NHS GG&C who were tested between 6 and 26 December 2021. The points and corresponding numbers represent the estimated vaccine effectiveness (%) for each group and for each variant, with the error bars representing 95% confidence intervals.

Fig. 5. Omicron exhibits reduced syncytia formation and has switched entry route.

a, Virus-induced fusion of co-cultured cells expressing split GFP (GFP-10 and GFP-11, respectively) reconstitutes fluorescence (image generated in BioRender, agreement SX23HJ6GY8SX23HJ6GY8). b, Co-cultured GFP-10 and GFP-11 A549-ACE2-TMPRSS2 cells were infected with B.1, Delta and Omicron/BA.1, photomicrographs taken at 22 h post infection: GFP, green; nucleocapsid (N), red; DAPI nuclei, blue; scale bars, 100 µm. c, Fluorescence over 20 h (n = 4 technical repeats, 3 independent experiments, 2 virus stocks). d, Calu-3 cell infection with B.1, Delta and Omicron/BA.1. Supernatants were assessed by RT-qPCR (n = 3 technical repeats, 2 independent experiments). e, SARS-CoV-2 entry routes: route 1 – fusion at the cell surface following processing by TMPRSS2; route 2 – fusion occurs post endocytosis after processing by cathepsins. Routes 1 and 2 are inhibited by Camostat and E64d, respectively. f, SARS-CoV-2 pseudotype infection of cell lines, mean luciferase values (1 experiment, n = 8 technical repeats, representative of 3 independent experiments). Route 1 predominates in Calu-3, route 2 in HEK, and both routes are supported by A549-ACE2-TMPRSS2. Control is Pangolin CoV (route 2 only), negative control is no glycoprotein pseudotypes (No). g, Relative pseudotype infection (compared to untreated) of 10 μM protease inhibitor-treated cells; mean of 4 biological repeats, ****P < 0.0001, one-way ANOVA, Dunnett’s multiple comparisons test. h, Camostat (left) and E64d (right) titration against Delta, Omicron/BA.1 and Pangolin CoV in A549-ACE2-TMPRSS2; mean relative infection compared to untreated control (n = 2 biological repeats). i, Immunoblot of spike-expressing HEK lysates with anti-spike (S2) and anti-actin. j, HEK infection by Omicron/BA.2 pseudotype; mean luciferase values (1 experiment, n = 8 technical repeats, representative of 3 independent experiments). k, hNEC infection with Delta and Omicron/BA.1 clinical isolates, supernatants assessed by RT-qPCR; mean of 2 independent biological repeats. l, Immunoblots of uninfected (left) and infected (right) hNEC and Calu-3 lysates. Immunoblots probed for ACE2, TMPRSS2, Cathepsin-L, GAPDH, Spike S1 and N. m, Cell–cell fusion (as in a) in Omicron/BA.1-transfected cells treated with trypsin. Plot displays mean fluorescence (1 experiment, 3 technical repeats, representative of 2 independent experiments). n, Cell–cell fusion in Delta or Omicron/BA.1-transfected cells, ±0.3 µg ml⁻¹ trypsin. Data are relative to fusion by Omicron/BA.1 spike (−trypsin, 8 h), mean of 2 biological repeats. All error bars represent s.e.m. Source data

Fig. 6. Determinants of Omicron spike biology.

a, Top: schematic representation of domain swap constructs between WT (lineage B) and Omicron spike. Bottom: linear representation illustrates the junction points between the chimeric proteins. b, Domain swap pseudotype infection of HEK cells. Data represent mean luciferase values from a representative experiment (n = 8 technical repeats, data are representative of 3 independent experiments). c, Sensitivity of domain swap pseudotypes to camostat and E64d in A549-ACE2-TMPRSS2 cells. Data are expressed relative to the untreated control for each virus. Values represent the mean of 3 biological repeats. d, Immunoblot analyses of lysates from HEK cells producing domain swap pseudotypes. Blots were probed for spike (S1 and processed S2), actin loading control and HIV Gag p55 transfection control. The actin control relates to the S1 blot, while the p55 control relates to the processed S2 blot. e, Quantification of spike proteolytic processing in WT (lineage B), Omicron BA.1 and the reciprocal RBD swaps. Data are expressed relative to the total for a given spike and represent the mean of 5 independent blots. f, Cell fusion assay using cells transfected with WT (lineage B) (left), Omicron BA.1 (right) and the reciprocal RBD swap spikes. Plot displays mean GFP fluorescence signal from 1 representative experiment. g, Cell–cell fusion (as in f). Data are expressed relative to fusion by WT (lineage B) spike (at 16 h) and are the mean of 4 independent experiments. One-way ANOVA, *P < 0.01, ****P < 0.0001. In all plots, error bars represent s.e.m. Source data

Extended Data Fig. 1. HCoV reactivity following two doses of SARS-CoV-2 vaccine.

Antibody responses were studied in three groups of individuals (n = 24 per group) vaccinated with either BNT162b2, ChAdOx1 or mRNA-1273 by MSD-ECL assay. Responses were measured against full-length spike glycoprotein (Spike) from HCoVs 229E, OC43, NL63 and HKU1 and are expressed as MSD arbitrary units (AU/ml). The response to OC43 was significantly higher in BNT162b2 vaccinates than in ChAdOx1 vaccinates. Bar represents group mean. Group means compared by one-way ANOVA, Tukey’s multiple comparisons test, **** significantly different p < 0.0001. Source data

Extended Data Fig. 2. Influenza reactivity following two doses of SARS-CoV-2 vaccine.

Antibody responses were studied in three groups of individuals (n = 24 per group) vaccinated with either BNT162b2, ChAdOx1 or mRNA-1273 by MSD-ECL assay. Responses were measured against haemagglutinins from influenza viruses; influenza A Michigan H1, Hong Kong H3 and Shanghai H7, and influenza B Phuket HA and Brisbane and are expressed as MSD arbitrary units (AU/ml). No significant differences were detected between the vaccine groups for each of the antigens. Bar represents group mean. Source data

Extended Data Fig. 3. HCoV reactivity following third dose of SARS-CoV-2 vaccine.

Antibody responses were studied in four groups of individuals primed with two doses of either ChAdOx1 or BNT162b2, followed by a booster of BNT162b2 or mRNA-1273. Responses were measured by MSD-ECL assay against full-length spike glycoprotein (Spike) from HCoVs 229E, OC43, NL63 and HKU1 and are expressed as MSD arbitrary units (AU/ml). Bar represents group mean. Source data

Extended Data Fig. 4. Influenza reactivity following third dose of SARS-CoV-2 vaccine.

Antibody responses were studied in four groups of individuals primed with two doses of either ChAdOx1 or BNT162b2, followed by a booster of BNT162b2 or mRNA-1273. Responses were measured by MSD-ECL against haemagglutinins from influenza viruses; influenza A Michigan H1, Hong Kong H3 and Shanghai H7, and influenza B Phuket HA and Brisbane and are expressed as MSD arbitrary units (AU/ml). Bar represents group mean. Group means compared by one-way ANOVA, Tukey’s multiple comparisons test, * significantly different p = 0.0413. Source data

Extended Data Fig. 5. Effect of third dose of SARS-CoV-2 vaccine on neutralising antibody titres.

Two groups of healthy volunteers vaccinated with two doses of either ChAdOx1 or BNT162b2, were sampled two weeks following a third dose of either BNT162b2 or mRNA-1273. Each point represents the mean of three replicates. Where dose 2 and dose 3 samples were available from the same individual, points are joined by a solid line.

Extended Data Fig. 6. Vaccine effectiveness estimates considering time since second dose.

Error bar plot of estimated vaccine effectiveness against testing positive for Delta and Omicron SARS-CoV-2 infection in the population of over 18 s in NHS GG&C who were tested between 6^th and 26^th December 2021. The points and corresponding text represent the estimated vaccine effectiveness (%) for each group, for each variant, with the error bar endpoints representing the endpoints of the corresponding 95% CIs. The modelling process was identical to the main vaccine effectiveness estimation reported in the main document, but the vaccine status variable had additional levels for 2^nd dose within 12 weeks of start of study period or before 12 weeks of start of study period. Note that the estimates for infection with Delta after previous confirmed SARS-CoV-2 infection and 2 x BNT162b2 < 12 weeks and 2 x mRNA-1272 < 12 weeks are calculated for a group with no infections and are therefore unreliable.

Extended Data Fig. 7. Comparison of syncytia formation by clinical isolates and reverse-genetics live virus.

GFP-10 and GFP-11 A549 ACE2 TMPRSS2 cells were co-cultured and infected with Delta and Omicron BA.1 clinical isolates or with live reverse-genetics virus in which the Delta or Omicron BA.1 spike is presented in the context of the ancestral wildtype B lineage genome. Fusion was quantified by measuring GFP signal, as in Fig. 5. Source data

Extended Data Fig. 8. Sensitivity of live SARS-CoV-2 to protease inhibitors.

SARS-CoV-2 infection of A549 ACE2 TMPRSS2 cells in the presence of 10 µM Camostat or E64d, data is expressed relative to untreated control, values represent mean across two independent experiments, asterisks indicate statistical significance (Two tailed T-test) between E64d treated Delta and Omicron infections. Error bars indicate standard error of the mean. Source data

Extended Data Fig. 9. Characterisation of Omicron BA.2 spike.

a, Relative SARS-CoV-2 pseudotype infection (compared to respective untreated controls) treated with 10 μM protease inhibitors. Data represent mean of three biological repeats, error bars indicate standard error of the mean. b, cell fusion assay using cells transfected to express WT (B lineage), Delta, Omicron BA.1 and BA.2 spike, values represent mean GFP fluorescence signal from one representative experiment error bars indicate standard error of the mean (3 technical repeats, representative of 4 biological repeats). Source data

Extended Data Fig. 10. Isolation of Omicron in cell culture.

a, Vero ACE2 TMPRSS2 (VAT) and BHK-hACE2 cells were inoculated with diluted clinical samples. Viral progeny was quantified in the medium 5 dpi by RT-qPCR. b, Aliquots of the medium from samples named 204 and 205 were used to generate a P1 in BHK-hACE2 and Calu-3 cells and, limited to sample 205, a P2 in Calu-3 and Caco2 cells. Viral stocks were quantified by RT-qPCR. Source data