Unraveling the impact of SARS-CoV-2 mutations on immunity: insights from innate immune recognition to antibody and T cell responses

Abstract

Throughout the COVID-19 pandemic, the emergence of new viral variants has challenged public health efforts, often evading antibody responses generated by infections and vaccinations. This immune escape has led to waves of breakthrough infections, raising questions about the efficacy and durability of immune protection. Here we focus on the impact of SARS-CoV-2 Delta and Omicron spike mutations on ACE-2 receptor binding, protein stability, and immune response evasion. Delta and Omicron variants had 3-5 times higher binding affinities to ACE-2 than the ancestral strain (KD_wt = 23.4 nM, KD_Delta = 8.08 nM, KD_BA.1 = 4.77 nM, KD_BA.2 = 4.47 nM). The pattern recognition molecule mannose-binding lectin (MBL) has been shown to recognize the spike protein. Here we found that MBL binding remained largely unchanged across the variants, even after introducing mutations at single glycan sites. Although MBL binding decreased post-vaccination, it increased by 2.6-fold upon IgG depletion, suggesting a compensatory or redundant role in immune recognition. Notably, we identified two glycan sites (N717 and N801) as potentially essential for the structural integrity of the spike protein. We also evaluated the antibody and T cell responses. Neutralization by serum immunoglobulins was predominantly mediated by IgG rather than IgA and was markedly impaired against the Delta (5.8-fold decrease) and Omicron variants BA.1 (17.4-fold) and BA.2 (14.2-fold). T cell responses, initially conserved, waned rapidly within 3 months post-Omicron infection. Our data suggests that immune imprinting may have hindered antibody and T cell responses toward the variants. Overall, despite decreased antibody neutralization, MBL recognition and T cell responses were generally unaffected by the variants. These findings extend our understanding of the complex interplay between viral adaptation and immune response, underscoring the importance of considering MBL interactions, immune imprinting, and viral evolution dynamics in developing new vaccine and treatment strategies.

Keywords: MBL; SARS-CoV-2; delta; immune imprinting; mannose-binding lectin; omicron; vaccine; variants of concern.

Figure 1

(A) Domain structure of the SARS-CoV-2 spike protein and location of the Delta, Omicron BA.1, and Omicron BA.2-defining mutations. Numbering and domain boundaries according to the spike wt. NTD, N-terminal domain; RBD, receptor binding domain; RBM, receptor binding motif; FCS, furin cleavage site; FP, fusion peptide; HR1/2, heptad repeats 1/2; TM, transmembrane domain; IC, intracellular domain. (B) Frequency distribution of sequences from SARS-CoV-2 VOCs in Denmark from October 2020 to 2^nd of January 2023. Modified from https://covariants.org. Only VOCs with frequencies above 0.05 are plotted. Data for panels A and B from the Nextstrain GISAID database (https://nextstrain.org/ncov/gisaid/global) (157, 158).

Figure 2

Biochemical impact of RBD mutations in Delta, BA.1, and BA.2 variants. (A–D) BLI sensorgrams of RBD wt (A), Delta (B), BA.1 (C), and BA.2 (D). ACE-2-Fc was immobilized on anti-human Fc capture sensors. ACE-2-immobilized sensors were dipped into serial dilutions of RBD (association 500 s), followed by only buffer (dissociation 500 s). (E) Kinetic parameters of single mutations and VOCs determined by BLI here and somewhere else (20, 21, 38). Data represent fold-change compared to the wt. Horizontal dotted black, yellow, and red lines signal the top KD, ka, and kdis values. Horizontal dashed grey line signals the baseline (no change compared to the wt). (F) Thermal denaturation curves of the RBD wt, Delta, BA.1, and BA.2 variants. Data are represented as individual first derivative curves of the 350:330 nm ratio from three repeats. Vertical dotted lines represent the inflection temperatures (Ti).

Figure 3

MS-based glycan analyses of the glycan shield of the spike wt, Delta, and Omicron BA.1. (A) MW determination by MALDI-MS. The two mayor peaks represent the intact spike peptide chain and probably the processed S1 subunit. (B) Released N-glycan profile after PNGaseF treatment and fluorescent labelling. The magnitude of the peaks is represented as relative units. (C) N-glycan site-specific occupancy, as determined by peptide mapping, for the spike wt, Delta, and Omicron BA.1. Positions that were not resolved are noted as “–”. N-glycans are represented as complex (oligomannose content ≤ 29%), hybrid (30–79%), and oligomannose (≥ 80%), from Watanabe et al., (40).

Figure 4

MBL interaction with spike. (A) Detection of rMBL bound to coated, 2-fold dilutions of spike wt, Delta (both in-house), BA.1, BA.2 (both from AcroBiosystems). BSA was used as negative control, and spike wt from NIBSC as positive control (ctrl). (B). Binding of native MBL from naïve sera from seven healthy, MBL-sufficient individuals (black dots). Naïve, MBL-defect serum was used as negative control (red dots). Friedman test with Dunn’s multiple comparisons. ***, p < 0.001. (C) Binding of native MBL from vaccinee plasma before and after total IgG depletion (n = 5). Data is normalized to mannan binding to correct for MBL loss after running the plasma through protein G agarose columns. (D–F) MBL-dependent complement deposition on spike. Serial dilutions of rMBL were applied to coated spike, mannan, or BSA, followed by naïve MBL-defect serum as a source of complement. Complement activation was measured as C4 (D), C3 (E), and TCC (F) deposition. (G) Calcium-dependent interaction between MBL and full-length spike, spike NTD, and spike RBD. EDTA was used to chelate calcium. No MBL was used as negative control. Mannan and BSA were used as positive and negative ligand controls, respectively (A–G). (H) MBL binding to spike N-glycan mutants from ExpiCHO supernatants captured with an anti-spike mAb (in-house) under calcium sufficient (MBL + Ca⁺⁺) or deficient (MBL + EDTA) conditions. Kruskal-Wallis test with Dunn’s multiple comparisons. (I) Structural representation of the spike protein (PDB ID: 6VXX (159)) with N-glycans (molecular surface) shown in grey. Highlighted in yellow are those positions evaluated in (G) by site-directed mutagenesis, and in red, those that impaired protein production and stability. Data from the Protein Data Bank (RCSB PDB) (https://www.rcsb.org/) (160). Created with Mol* Viewer (161). ns, not significant.

Figure 5

IgG and IgA responses after infection and vaccination. (A) Overview of the vaccinee cohort. Donors were grouped in infection-naïve (n = 20), wt infection (n = 20), Delta infection (n = 18), and Omicron infection (n = 20). Blood samples were collected before vaccination, after the first, second, and third doses of the BNT162b2 vaccine. (B, C) Evolution of IgG (B) and IgA (C) RBD-specific titers, reported as arbitrary units (AU)/ml, after infection and vaccination. Horizontal solid lines represent median. Horizontal dotted lines represent the threshold for positivity. (D, E) Avidity maturation of IgG (D) and IgA (E). Ordinary Two-way ANOVA. Data is represented as mean ± SD. (F) IgG and IgA nAbs contribution to the neutralization potency of hybrid immune sera (n = 5), plotted as the ratio (in percentage) of IgG, IgA, and IgG+IgA depleted to non-depleted sera. Two-way ANOVA with the Geisser-Greenhouse correction. Data is represented as mean ± SD.

Figure 6

Neutralization of RBD wt, Delta, BA.1, and BA.2 after vaccination and infection. (A) Neutralization potency of sera, reported as IU/ml, from infection-naïve individuals (n = 20), wt infection (n = 20), Delta (n = 18), and Omicron (n = 20), against RBD wt, Delta, BA.1, and BA.2. Two-way ANOVA with the Geisser-Greenhouse correction. Horizontal grey lines represent the median. Horizontal dotted lines represent the threshold for positivity. (B) Comparison between the neutralization potency of sera against RBD wt from infection-naïve individuals (n = 20) and sera from those with a previous infection (n = 20) from panel (A) Two antigen exposures refer to complete vaccination or infection + first dose. Three exposures refer to complete vaccination + booster or complete vaccination + infection. Mann-Whitney tests, with a two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli. Data are presented as median with interquartile range. (C) Antibody evasion gains of the Delta and Omicron variants plotted as the ratio of the neutralization of RBD wt to RBD Delta, BA.1, and BA.2 from panel (A) Kruskal-Wallis with Dunn’s multiple corrections. Horizontal grey lines represent the mean. Dashed lines indicate the highest and lowest mean values. Outliers identified by ROUT with Q = 1% (G [n = 7], H [n = 3]). (D, E) Affinity (D) and neutralization potency (E) of a panel of murine mAbs (n = 14) towards RBD wt, Delta, BA.1, and BA.2. Horizontal dotted lines indicate the maximum concentration of mAbs used in the assays. Friedman tests with Dunn’s multiple comparisons. (F, G) Antibody titers (F) and neutralization potency of sera (G) in a murine model of heterologous prime-boost vaccination. Mice were divided into wt (three doses of spike wt, n = 4), Delta (two doses of wt followed by a boost with spike Delta, n = 4), and Omicron (two doses of wt followed by a boost with spike Omicron, n = 4). Kruskal-Wallis with Dunn’s multiple corrections. ns, not significant.

Figure 7

Cellular immunity to SARS-CoV-2 wt, Delta, and Omicron. (A) IFN-γ release after whole blood stimulation from infection-naïve (n = 11), individuals with a wt infection (n = 6) or a recent Omicron infection (n = 14), with peptide MPs covering the wt, Delta, and Omicron spike proteins, and CD4 T cell peptides restricted to the remainder of the proteome (CD4RE). Multiple Kruskal-Wallis tests with Dunn’s multiple comparisons corrections. Dotted line represents the threshold for positivity. (B) Ratio (%) of T cell responses after whole blood stimulation with spike peptide pools from panel A in infection-naïve, individuals with a wt infection, or an Omicron infection. Friedman tests with Dunn’s multiple comparisons. (C) IFN-γ release from splenocyte cultures from mice immunized thrice with spike wt (Wt) (n = 4), twice with spike wt followed by spike Delta (Delta) (n = 4), or twice with spike wt followed by spike Omicron BA.1 (Omicron) (n = 4). Two-way ANOVA with the Geisser-Greenhouse correction. (D) Variant-specific T cell responses from panel C plotted as the ratio of IFN-γ release after spike Delta MP stimulation to spike wt (left) or as the spike Omicron MP stimulation to spike wt (right). Kruskal-Wallis with Dunn’s multiple comparisons. Only statistically significant differences are plotted (A–D). (E) Median response frequency (RF) scores for CD4 and CD8 epitopes from the SARS-CoV-2 spike protein. Data from the Immunome Browser (www.iedb.org) (45). Delta and Omicron mutated residues are marked in purple and orange, respectively. (F, G) Waning of T cell responses against spike wt (F) or CD4RE (G) peptide pools monitored for eight months after an Omicron infection (n = 9). Solid lines represent 3-knot smoothing splines. Horizontal dotted lines indicate the 50% response.