Last,weassessedtheRBD-andNTD-
directed antibodies isolated from both donor
cohorts for reactivity with multiple SARS-CoV-2
VOCs and variants of interest (VOIs). Against
pre-Omicron VOCs and VOIs (Beta, Gamma,
Delta, Kappa, and Lambda), 41% (12 of 29) of
RBD-directed antibodies derived from ChAdOx1-
boosted individuals displayed reduced bind-
ing activity (greater than twofold) to two or
more of these variants as compared with
only 20% (8 of 40) of antibodies isolated
from mRNA-1273–boosted individuals, sug-
gesting that mRNA-1273 booster vaccination
activates a larger proportion of broadly reac-
tive B cells (Fig. 4A and fig. S15A). Further-
more, the mRNA-elicited antibodies bound
to both WT and pre-Omicron VOC and VOI
RBDs with significantly higher affinities (me-
dianKd= 1.4 to 3.7 nM) relative to the
ChAdOx1-induced antibodies (medianKd=
4.9 to 20.3 nM) (Fig. 4B), potentially explain-
ing their increased breadth of binding. How-
ever, in contrast to earlier variants, the Omicron
variant broadly escaped recognition by RBD-
directed antibodies induced by both booster
regimens, with 24 of 29 (83%) and 29 of 40
(73%) antibodies derived from ChAdOx1- and
mRNA-1273–boosted individuals, respectively,
showing significantly reduced binding reac-
tivity (<10% of wild type) to the Omicron RBD
(Fig.4A).Amongthe26neutralizinganti-
bodies identified from both donor cohorts, only
three maintained binding to the Omicron RBD
within 10-fold of the Wuhan-1 RBD (Fig. 4C).
This result is consistent with the large number
of mutations within the frequently targeted
class 1/2/3 RBD epitopes and provides a mo-
lecular explanation for the extensive resistance
of this variant to neutralization by convalescent
and vaccinee sera ( 21 , 22 ).
Consistent with the overall increased breadth
of recognition by RBD-directed antibodies in-
duced by heterologous booster vaccination,
52 to 58% of NTD-directed antibodies isolated
from mRNA-1273–boosted donors retained
bindingtoeachofthefourVOCstested(Beta,
Gamma, Delta, and Omicron) compared with
only 31 to 37% of antibodies isolated from
ChAdOx1-boosted donors (Fig. 4D and fig. S15B).
Furthermore, the NTD-specific antibodies iso-
lated from mRNA-1273–boosted donors trended
toward higher binding affinities across all
VOCs tested as compared with antibodies de-
rived from ChAdOx1-boosted donors (Fig. 4E).
Overall, a significantly larger fraction of anti-
bodies to RBD and NTD (49%) derived from
mRNA-boosted donors retained reactivity (Kd<
50 nM) with all VOC NTDs or RBDs relative
to ChAdOx1-boosted donors (26%) (Fig. 4F).
Thus, heterologous mRNA-1273 immunization
appears to skew the early secondary B cell
response toward higher-affinity clones with
improved breadth of variant recognition com-
pared with homologous ChAdOx1 immuniza-
tion, although the Omicron variant broadly
escapes neutralizing antibodies induced by
both booster regimens.
Heterologous ChAdOx1:mRNA-1273 prime-
boost immunization induces significantly
broader and more potent serum neutralizing
antibody and MBC responses against WT SARS-
CoV-2 and VOCs relative to homologous
ChAdOx1 vaccination, and this difference
appears to be driven by both the magnitude
and quality of the early secondary B cell re-
sponse. Expression of WT S by ChAdOx1 ap-
pears to distract the B cell response away from
neutralizing sites of vulnerability present on
prefusion S, and homologous booster vacci-
nation further expands these non-neutralizing
specificities. By contrast, heterologous booster
immunization with mRNA-1273, which en-
codes S-2P, redirects the B cell response toward
epitopes expressed on prefusion-stabilized
S. Furthermore, mRNA-1273 activates B cells
with higher affinity for prefusion S and greater
breadth of reactivity relative to ChAdOx1. The
molecular basis for this difference remains to
be determined but could potentially be asso-
ciated with differences in cell-surface expres-
sion of S-2P relative to WT S or the distinct
innate immunostimulatory properties of mRNA
versus adenoviral particles ( 23 , 24 ). Last, al-
though heterologous ChAdOx1:mRNA-1273
prime-boost immunization shows superior
immunogenicity relative to two-dose ChAdOx1,
the B cell response induced by both vaccination
regimens is dominated by non-neutralizing
antibodies, and the vast majority of neutral-
izing antibodies fail to recognize the Omicron
variant. Although studies in animal models
have demonstrated that non-neutralizing anti-
bodies can contribute to protection, serum
neutralizing antibody titer strongly correlates
with vaccine-induced efficacy against symp-
tomatic COVID-19 in humans ( 25 , 26 ). Thus,
rationally designed immunogens that focus
the B cell response on conserved, neutralizing
epitopes within the RBD and S2 subunit may
enhance the potency, breath, and durability
of protection against SARS-CoV-2, future
emerging VOCs, and potentially preemergent
b-coronaviruses ( 27 , 28 ).REFERENCES AND NOTES- L. J. Abu-Raddad, H. Chemaitelly, A. A. Butt, National Study
Group for COVID-19 Vaccination,N. Engl. J. Med. 385 , 187– 189
(2021). - S. A. Madhiet al.,N. Engl. J. Med. 384 , 1885–1898 (2021).
- C. H. Hansenet al., medRxiv 21267966 [Preprint] (2021).
- C. R. Simpsonet al.,Nat. Med. 27 , 1290–1297 (2021).
- J. B. Schulzet al.,Ann. Neurol. 90 , 627–639 (2021).
- J. Barros-Martinset al.,Nat. Med. 27 , 1525–1529 (2021).
- J. Normarket al.,N. Engl. J. Med. 385 , 1049–1051 (2021).
- P. Nordström, M. Ballin, A. Nordström,Lancet Reg Health Eur.
11 , 100249 (2021). - A. Rössler, L. Riepler, D. Bante, D. von Laer, J. Kimpel,
N. Engl. J. Med.NEJMc2119236 (2022). - M. A. Tortorici, D. Veesler,Adv. Virus Res. 105 , 93–116 (2019).
- D. Wrappet al.,Science 367 , 1260–1263 (2020).
- J. Yuet al.,Science 369 , 806–811 (2020).
- F. Amanatet al.,Cell 184 , 3936–3948.e10 (2021).
- R. C. Breweret al.,Nat. Immunol. 23 , 33–39 (2022).
15. M. Mukhamedovaet al.,Immunity 54 , 769–780.e6 (2021).
16. J. Wrammertet al.,Nature 453 , 667–671 (2008).
17. D. F. Robbianiet al.,Nature 584 , 437–442 (2020).
18. Z. Wanget al.,Nature 592 , 616–622 (2021).
19. C. O. Barneset al.,Nature 588 , 682–687 (2020).
20. A. J. Greaneyet al.,Nat. Commun. 12 , 4196 (2021).
21. M. McCallumet al.,Science 375 , XXX (2021).
22. W. Dejnirattisaiet al.,Cell 185 , 467–484.e15 (2022).
23. R. Boset al.,NPJ Vaccines 5 , 91 (2020).
24. N. Chaudhary, D. Weissman, K. A. Whitehead,Nat. Rev.
Drug Discov. 20 , 817–838 (2021).
25. D. Liet al.,Cell 184 , 4203–4219.e32 (2021).
26. D. S. Khouryet al.,Nat. Med. 27 , 1205–1211 (2021).
27. D. Pintoet al.,Science 373 , 1109–1116 (2021).
28. A. A. Cohenet al.,Science 371 , 735–741 (2021).
29. C. G. Rappazzoet al.,Science 371 , 823–829 (2021).
ACKNOWLEDGMENTS
We thank study nurse I.-L. Persson and personnel at the Clinical
Research Center at Umeå University Hospital for support,
enrollment of study subjects, and sampling. We also thank
L. Vikström, M. Järner, and M. Lagerqvist at Umeå University for
processing of samples and A. Edin at Umeå University Hospital for
the summary of clinical data. We also acknowledge Parexel for
assistance with figure preparation and E. Krauland, J. Nett, and
M. Vasquez for helpful comments on the manuscript. All IgGs were
sequenced by Adimab’s Molecular Core and produced by the
High Throughput Expression group.Funding:M.N.E.F is funded by
grants from the Swedish Research Council (2020-06235) and from
the SciLifeLab National COVID-19 Research Program (VC-2020-
0015), financed by the Knut and Alice Wallenberg Foundation. C.A.
is funded by the Swedish Research Council (2021-04665). J.N. is a
Wallenberg Center for Molecular Medicine Associated Researcher.
M.G., W.C., and J.K. are funded by Swedish Research Council
(#2020-05782) and SciLifeLab/KAW (#VC-2021-0026).Author
contributions:L.M.W. and M.N.E.F. conceived and designed the
study. C.A. and J.N. are principal investigators for the clinical trial
and supervised sample collection. M.G., W.C., and J.K. performed
Omicron serum neutralization assays. C.I.K. performed serum
and B cell analyses, single B cell sorting, and antibody characterization.
C.E.J. produced recombinant IgG antibodies for characterization.
E.R.C. designed and performed developability and biolayer
interferometry assays. C.I.K. and M.S. developed, designed, and
performed pseudovirus neutralization assays. C.I.K. and M.E.A.
developed, designed, and performed multiplexed flow cytometry
assays. C.I.K., E.R.C. and L.M.W. analyzed the data. C.I.K. and L.M.W.
wrote the manuscript, and all authors reviewed and edited the
paper.Competing interests:C.I.K., E.R.C., C.E.J., M.S., and L.M.W.
are employees of Adimab and may hold shares in Adimab. L.M.W.
is an employee of Adagio Therapeutics and holds shares in Adagio
Therapeutics. J.N., M.G., C.A., W.C., J.K., and M.N.E.F. declare no
competing interests. L.M.W., C.I.K., and M.N.E.F. are inventors
on a provisional patent application (US patent application no.
63/307,230) describing the SARS-CoV-2 antibodies.Data and
material availability:Antibody sequences have been deposited in
GenBank (accession codes OL697893 to OL698712). All other
data are available in the manuscript or supplementary materials.
IgGs are available from L.M.W. under a materials transfer agreement
(MTA) from Adagio Therapeutics. Requests for clinical samples
should be addressed to [email protected]. Sharing is regulated
by the Swedish law of bio banking and requires a collaboration with
Umeå University (represented by MNE Forsell) and must comply with
EU regulations for GDPR. This work is licensed under a Creative
Commons Attribution 4.0 International (CC BY 4.0) license, which
permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited. To view a copy of this
license, visit https://creativecommons.org/licenses/by/4.0/. This
license does not apply to figures/photos/artwork or other content
included in the article that is credited to a third party; obtain authorization
from the rights holder before using such material.SUPPLEMENTARY MATERIALS
science.org/doi/10.1126/science.abn2866
Materials and Methods
Figs. S1 to S15
Tables S1 and S2
References ( 30 – 35 )
MDAR Reproducibility Checklist19 November 2021; accepted 2 February 2022
Published online 10 February 2022
10.1126/science.abn2688SCIENCEscience.org 4 MARCH 2022•VOL 375 ISSUE 6584 1047
RESEARCH | REPORTS