Monovalent and bivalent boosters elicit similar Ab binding responses against variant S trimers in pre-vaccinated mice.

A) Immunization schedule in BALB/c mice. Blood samples were collected on days 14, 56, 154, and 244. This panel was created using BioRender.com. B) Immunization regimens for each cohort. All cohorts were initially vaccinated with 2 µg of Wu1 S mRNA-LNP on days 0 and 28 by IM injection. Pre-vaccinated mice received different boost immunogens on day 230 by IM injection. Cohorts 1-4 received 2 µg of monovalent Wu1 S mRNA-LNP (cyan), monovalent Wu1 S-EABR mRNA-LNP (orange), bivalent Wu1/BA.5 S mRNA-LNP (blue), or bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon), respectively. C) Schematic showing the different boost immunogens that were administered on day 230. Wu1 S mRNA-LNP immunization results in cell surface presentation of membrane-anchored Wu1 S trimers. Wu1 S-EABR mRNA-LNP immunization results in dual presentation of Wu1 S trimers on cell surfaces and on eVLPs that are secreted after budding from the plasma membrane. eVLP formation is induced by EABR-mediated recruitment of host cell proteins from the ESCRT pathway. Bivalent Wu1/BA.5 S mRNA-LNP immunization results in cell surface presentation of membrane-anchored Wu1 and BA.5 S trimers. Wu1/BA.5 S-EABR mRNA-LNP immunization results in dual presentation of Wu1 and BA.5 S trimers on cell surfaces and eVLPs that bud from the plasma membrane. S trimers are shown as homotrimers for both monovalent and bivalent boost immunogens. This panel was created using BioRender.com. D-G) ELISA data from indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of area under the curve (AUC) values. ELISAs evaluated binding of purified Ss from the D) Wu1, E) BA.5, F) BQ.1.1, and G) XBB.1 variants. The log10-fold change in binding titers from day 154 to day 244 is also shown for each S (plotted as mean (bars) and standard deviation (horizontal lines)).

Bivalent S-EABR mRNA-LNP booster elicits potent Ab responses targeting Omicron RBDs.

A-D) RBD binding titers for antisera from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP (cyan), 2 µg of monovalent Wu1 S-EABR mRNA-LNP (orange), 2 µg of bivalent Wu1/BA.5 S mRNA-LNP (blue), or 2 µg of bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon) (immunization schedule in Fig. 1A-B). ELISA data are shown for indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of area under the curve (AUC) values. ELISAs evaluated binding of RBDs from the A) Wu1, B) BA.5, C) BQ.1.1, and D) XBB.1 SARS-CoV-2 variants. The log10-fold change in binding titers from day 154 to day 244 is also shown for each RBD (plotted as mean (bars) and standard deviation (horizontal lines)). Significant differences between cohorts linked by horizontal lines are indicated by asterisks: p<0.05 = *, p<0.01 = **.

Bivalent Wu1/BA.5 S-EABR mRNA-LNP booster elicits higher neutralizing responses against Omicron subvariants.

A-D) Neutralization titers for antisera from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP (cyan), 2 µg of monovalent Wu1 S-EABR mRNA-LNP (orange), 2 µg of bivalent Wu1/BA.5 S mRNA-LNP (blue), or 2 µg of bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon) (immunization schedule in Fig. 1A-B). Neutralization data are shown for indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of ID50 values. Results are shown for A) Wu1, B) BA.5, C) BQ.1.1, and D) XBB.1 pseudoviruses. Dashed horizontal lines correspond to the background values representing the limit of detection for neutralization assays (ID50 = 1:50). The log10-fold change in neutralization titers from day 154 to day 244 is also shown for each pseudovirus (plotted as mean (bars) and standard deviation (horizontal lines)). Significant differences between cohorts linked by horizontal lines are indicated by asterisks: p<0.05 = *, p<0.01 = **. E) Number of mice in each cohort that reached an ID50 ≤ 1:100 against each variant on days 56 (left), 154 (middle), and 244 (right).

Bivalent S-EABR mRNA-LNP booster induces polyclass Ab responses.

Sera for DMS studies were derived from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP, monovalent Wu1 S-EABR mRNA-LNP, bivalent Wu1/BA.5 S mRNA-LNP, or bivalent Wu1/BA.5 S-EABR mRNA-LNP (immunization schedule in Fig. 1A-B). DMS was performed for 6 serum samples for each cohort, which were selected based on XBB.1 neutralization activity (1st, 2nd, 3rd, 5th, 6th, and 10th best neutralizer from each cohort). For the Wu1 RBD library, only five DMS profiles were obtained for each cohort. A) Line plots for DMS results using Wu1 (top) or XBB.1.5 (bottom) RBD libraries for sera from pre-vaccinated mice that received the indicated boost immunogens. The x axes display the RBD residue numbers, and the y axes show the sum of the Ab escape of all mutations at a site (larger numbers indicating increased Ab escape). Each faint line shows one antiserum with heavy lines representing the average across the n = 5 (top) or n = 6 (bottom) sera in each cohort. Lines are colored differently for RBD epitopes within different RBD epitope classes33,40 (color definitions shown in legend above line plots; epitopes are shown in Fig. S2; gray color for residues not assigned to an epitope). B) The average site-total Ab escape for the Wu1 (top) or XBB.1.5 (bottom) RBD libraries for pre-vaccinated mice that received the indicated boost immunogens mapped to the Wu1 RBD surface (PDB: 6M0J). Gray coloring indicates no escape, and red coloring indicates escape. Lines with numbers colored according to epitope class indicate RBD positions with the most escape.

Heterotrimer S formation occurs for soluble constructs.

A) Schematic for the generation and purification of soluble SARS-CoV-2 Wu1/BA.1 (HT2) and Wu1/BA.1/Delta (HT3) S heterotrimers. HT2 heterotrimers were generated by co-transfection of StrepII-tagged Wu1 S and His-tagged BA.1 S constructs at a 1:1 DNA ratio, while HT3 heterotrimers were produced by co-transfecting StrepII-tagged Wu1 S, His-tagged BA.1 S, and D7324-tagged Delta S constructs at a 1:1:1 ratio. Each heterotrimer was sequentially purified using affinity chromatography specific for each tag (StrepII, His, or JR-52) to ensure incorporation of distinct protomers. This panel was created using BioRender.com. B) Cryo-EM density maps for Wu1/BA.1 HT2 S class I with 3 ‘down’ RBDs (3.8 Å resolution) and class II with 1 ‘up’ RBD (3.8 Å resolution). C) 4.0 Å and 3.9 Å resolution cryo-EM density maps for Wu1/BA.1/Delta HT3 S with 3 ‘down’ RBDs and 1 ‘up’ RBD. D) Table summarizing S heterotrimer and control homotrimer conformational states.

Wu1 and BA.5 S-EABR constructs induce efficient eVLP budding.

A) Design and sequence information for SARS-CoV-2 S-EABR constructs. Top: The Wu1 and BA.5 S (including a furin cleavage site, 2P stabilizing substitutions, the D614G substitution, and ΔCT, a cytoplasmic tail deletion) were fused to an EPM sequence, a (Gly)3Ser (GS) spacer, and an EABR sequence. EPM = Endocytosis prevention motif. GS = (Gly)3Ser linker. EABR = ESCRT- and ALIX-binding region. Bottom: EPM and EABR sequence information. B) DNA plasmids encoding Wu1 S-EABR (red) or BA.5 S-EABR (orange) constructs were transfected into HEK293T cells. 48 hours post-transfection, transfected cell culture supernatants were harvested, and eVLPs were purified by ultracentrifugation on a 20% sucrose cushion. Purified eVLP samples were serially diluted, directly coated onto 96-well plates, and eVLP-associated S protein levels were quantified by ELISA.

Sarbecovirus RBD sequence conservation.

A) Sequence conservation of 16 sarbecovirus RBDs calculated with ConSurf73 and plotted on a surface representation of a SARS-CoV-2 RBD (PDB 7BZ5). Anti-RBD class 1, 2, 3, 4, 1/4, and 5 Ab epitopes are outlined with dotted lines in different colors using structural information from representative structures of mAb bound to SARS-CoV-2 S or RBD (C102: PDB 7K8M; C002: PDB 7K8T, S309: PDB 7JX3; CR3022: PDB 7LOP; C118: PDB 7RKV; WRAIR-2063: PDB 8EOO). SARS-CoV-2 RBD substitutions (colored in red) in Omicron B) BA.1, C) BA.5, D) BQ.1.1, and E) XBB.1 variants plotted on the RBD sequence conservation plot.

Primary vaccinations and booster immunizations elicit potent anti-S2 binding responses.

S2 binding titers for antisera from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP (cyan), 2 µg of monovalent Wu1 S-EABR mRNA-LNP (orange), 2 µg of bivalent Wu1/BA.5 S mRNA-LNP (blue), or 2 µg of bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon). ELISA data for binding of Wu1 S2 are shown for indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of area under the curve (AUC) values. The log10-fold change in binding titers from day 154 to day 244 is also shown (plotted as mean (bars) and standard deviation (horizontal lines)).

Monovalent and bivalent S-EABR mRNA-LNP boosters elicit higher increases in BA.5 neutralization compared to conventional S mRNA-LNP boosters.

Neutralization titers for antisera from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP (cyan), 2 µg of monovalent Wu1 S-EABR mRNA-LNP (orange), 2 µg of bivalent Wu1/BA.5 S mRNA-LNP (blue), or 2 µg of bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon) (immunization schedule in Fig. 1A-B). BA.5 neutralization data are shown for indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of ID50 values. Dashed horizontal lines correspond to the background values representing the limit of detection for neutralization assays (ID50 = 1:50). The log10-fold change in neutralization titers from day 56 to day 244 is also shown (plotted as mean (bars) and standard deviation (horizontal lines)). Significant differences between cohorts linked by horizontal lines are indicated by asterisks: p<0.05 = *, p<0.01 **.

Monovalent Wu1 S-EABR mRNA-LNP booster elicits improved neutralizing responses against BA.1 variant.

Neutralization titers for antisera from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP (cyan) or monovalent Wu1 S-EABR mRNA-LNP (orange). Neutralization data are shown for indicated time points for antisera from individual mice (colored circles) presented as the geometric mean (bars) and standard deviation (horizontal lines) of ID50 values against BA.1 pseudovirus. Dashed horizontal lines correspond to the limit of detection for neutralization assays (ID50 = 1:50). The log10-fold change in neutralization titers from day 154 to day 244 is also shown (plotted as mean (bars) and standard deviation (horizontal lines)). Significant differences between cohorts linked by horizontal lines are indicated by asterisks: p<0.05 = *.

Depleting bivalent booster-induced antisera with Wu1 RBD reduces BA.5 neutralizing activity.

Day 244 serum samples obtained from pre-vaccinated mice that received booster immunizations with 2 µg of bivalent Wu1/BA.5 S mRNA-LNP (blue) or 2 µg of bivalent Wu1/BA.5 S-EABR mRNA-LNP (maroon) were depleted using a mock protein (HIV-1 gp120; left) or Wu1 RBD (right). The neutralizing activity of the depleted antisera was measured against BA.5 pseudovirus. Neutralization data are shown as ID50 values for individual mice (colored circles). Data points for individual mice following mock or Wu1 RBD depletions are connected by colored lines. Dashed horizontal lines correspond to the limit of detection for neutralization assays (ID50 = 1:50).

Bivalent S-EABR mRNA-LNP booster induces polyclass Ab responses against Wu1 RBD.

Sera for DMS studies were derived from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP, monovalent Wu1 S-EABR mRNA-LNP, bivalent Wu1/BA.5 S mRNA-LNP, or bivalent Wu1/BA.5 S-EABR mRNA-LNP (immunization schedule in Fig. 1A-B). Line (left) and logo plots (right) are shown for DMS results for individual sera from pre-vaccinated mice that received the indicated boost immunogens. DMS was performed using a Wu1 RBD library. The x axes of the line and logo plots display the RBD residue numbers. The y axis of the line plots shows the sum of the Ab escape of all mutations at a site (larger numbers indicating increased Ab escape). Sites with the strongest Ab escape are shown in logo plots, with tall letters representing the most frequent mutations at a site. Logo plots are colored differently for RBD epitopes within different classes33,40,41 (epitopes are shown in Fig. S2; gray color for residues not assigned to an epitope).

Bivalent S-EABR mRNA-LNP booster induces less pronounced class 3 Ab responses against XBB.1.5 RBD.

Sera for DMS studies were derived from pre-vaccinated mice that received booster immunizations with 2 µg of monovalent Wu1 S mRNA-LNP, monovalent Wu1 S-EABR mRNA-LNP, bivalent Wu1/BA.5 S mRNA-LNP, or bivalent Wu1/BA.5 S-EABR mRNA-LNP (immunization schedule in Fig. 1A-B). Line (left) and logo plots (right) are shown for DMS results for individual sera from pre-vaccinated mice that received the indicated boost immunogens. DMS was performed using an XBB.1.5 RBD library. The x axes of the line and logo plots display the RBD residue numbers. The y axis of the line plots shows the sum of the Ab escape of all mutations at a site (larger numbers indicating increased Ab escape). Sites with the strongest Ab escape are shown in logo plots, with tall letters representing the most frequent mutations at a site. Logo plots are colored differently for RBD epitopes within different classes33,40 (epitopes are shown in Fig. S2; gray color for residues not assigned to an epitope).

Cartoon diagram illustrating inter- and intra-S crosslinking.

Left: model showing potential inter-S crosslinking through the interactions between both Fabs of a single IgG and two RBDs of adjacent viral S proteins. Right: model showing intra-S crosslinking through the interaction between both Fabs of a single IgG and two RBDs on the same viral S trimer. The hinge regions between the Fabs and Fc are modeled with flexible linkers, shown as dotted lines, as these regions were not resolved in the crystal structure of an intact IgG74. Disulfide bonds in the hinge region are shown as black lines. These diagrams are modeled based on the cryo-EM structure of SARS-CoV-2 S complexed with human C118 Fab (PDB ID 7RKV) and the crystal structure of human Fc (PDB ID 1FC1). IgGs shown in this figure represent either circulating Abs or membrane-bound BCRs.

Cryo-EM data collection and processing for heterotrimeric and control SARS-CoV-2 S structures.

Representative micrographs, workflow for single-particle data processing, and final reconstructions are shown for SARS-CoV-2 (A) Wu1/BA.1 S HT2 heterotrimer, (B) Wu1/BA.1/Delta S HT3 heterotrimer, and homotrimer S controls including (C) Wu1 His/StrepII double-tag control, (D) StrepII-tagged Wu1, (E) His-tagged BA.1, and (F) D7324-tagged Delta constructs.

Bivalent S-EABR mRNA-LNP booster induces consistent polyclass anti-RBD Ab responses.

The DMS profiles for antisera from individual mice against the Wu1 and XBB.1.5 RBD libraries shown in Fig. S6 and S7 were classified as polyclass (total escape peaks <1 at all sites; green), moderate escape (total escape peaks between 1 and 2 at one or more sites; yellow), or strong escape (total escape peaks >2 at one or more sites; red) profiles. Total escape peaks represent the sum of escape fractions for all mutations at a specific site45. For moderate and strong escape profiles, the RBD epitope classes with the highest total escape peaks are specified.