CORONAVIRUS
Molecular basis of immune evasion by the Delta and
Kappa SARS-CoV-2 variants
Matthew McCallum^1 , Alexandra C. Walls^1 , Kaitlin R. Sprouse^1 , John E. Bowen^1 , Laura E. Rosen^2 ,
Ha V. Dang^1 , Anna De Marco^3 , Nicholas Franko^4 , Sasha W. Tilles^5 , Jennifer Logue^4 ,
Marcos C. Miranda1,6, Margaret Ahlrichs1,6, Lauren Carter1,6, Gyorgy Snell^2 , Matteo Samuele Pizzuto^3 ,
Helen Y. Chu^4 , Wesley C. Van Voorhis^5 , Davide Corti^3 , David Veesler1,7*
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission leads to the emergence of
variants, including the B.1.617.2 (Delta) variant of concern that is causing a new wave of infections
and has become globally dominant. We show that these variants dampen the in vitro potency of vaccine-
elicited serum neutralizing antibodies and provide a structural framework for describing their immune
evasion. Mutations in the B.1.617.1 (Kappa) and Delta spike glycoproteins abrogate recognition by several
monoclonal antibodies via alteration of key antigenic sites, including remodeling of the Delta amino-
terminal domain. The angiotensin-converting enzyme 2 binding affinities of the Kappa and Delta receptor
binding domains are comparable to the Wuhan-Hu-1 isolate, whereas B.1.617.2+ (Delta+) exhibits
markedly reduced affinity.
T
he ongoing spread of severe acute res-
piratory syndrome coronavirus 2 (SARS-
CoV-2), the causative agent of the COVID-19
pandemic, is resulting in the continued
emergence of variants. The B.1.351 (Beta,b)
variant of concern was originally described in
South Africa and remains the isolate associ-
ated with the greatest magnitude of immune
evasion, as measured by reduced neutralizing
antibody (Ab) titers in vitro ( 1 – 3 ). Conversely,
the B.1.1.7 (Alpha,a) variant of concern, which
was first detected in the United Kingdom, has
a modest impact on neutralizing Ab titers but a
marked enhancement in angiotensin-converting
enzyme 2 (ACE2) receptor binding affinity and
transmissibility, relative to the Wuhan-Hu-1 iso-
late, which led to worldwide dominance in the
early months of 2021 ( 2 , 4 ).
The SARS-CoV-2 spike (S) glycoprotein is ex-
posed at the surface of the virus and mediates
entry into host cells; it is the main target of
neutralizing antibodies and the focus of most
vaccines ( 5 , 6 ). The S glycoprotein is subdivided
into two functional subunits, designated S 1 and
S 2 , that interact noncovalently after proteolytic
cleavage by furin during synthesis ( 5 , 7 , 8 ). The
S 1 subunit contains the receptor binding do-
main (RBD), which engages the receptor ACE2
( 5 , 7 , 9 , 10 ), and the N-terminal domain (NTD),
which recognizes attachment factors ( 11 – 13 ).
The S 2 subunit contains the fusion machinery
and undergoes large-scale conformational changes
to drive fusion of the virus and host mem-
branes ( 14 ), enabling genome delivery and
initiation of infection. Abs that bind to spe-
cific sites on the RBD ( 15 – 22 ), the NTD ( 23 – 26 ),
or the fusion machinery stem helix ( 27 – 31 )
interfere with receptor attachment or mem-
brane fusion. Serum neutralizing Ab titers
are a correlate of protection against SARS-
CoV-2 in nonhuman primates ( 32 – 35 ).
In late 2020, B.1.617 variants including B.1.617.1
(Kappa,k) and B.1.617.2 (Delta,d) were first
detected in India and caused devastating epi-
demics before spreading globally ( 36 , 37 ). The
B.1.617.1 variant S harbors T95I (Thr^95 →Ile),
G142D, E154K, L452R, E484Q, D614G, P681R,
and Q1071H substitutions, whereas the B.1.617.2
variant S carries T19R, G142D, E156G, L452R,
T478K, D614G, P681R, and D950N substitu-
tions and a deletion of residues 157 and 158
(157-158del) (table S1). Most of these muta-
tions localize to the RBD (residues 328 to 531)
and NTD (residues 14 to 305), which are the
major targets of neutralizing Abs in conva-
lescent and vaccinated individuals ( 18 , 38 ),
raising concerns about the efficacy of available
vaccines and therapeutic monoclonal anti-
bodies (mAbs) against these variants. More-
over, the K417N mutation was detected in
the B.1.617.2 lineage, known as the B.1.617.2+
(Delta plus,d+) variant, which is shared with
the B.1.351 (Beta,b) variant of concern and
was previously shown to reduce neutraliza-
tion potency of some mAbs ( 2 , 39 ).
To evaluate the effect on neutralization of
the mutations in the B.1.617.1, B.1.617.2, and
B.1.617.2+ S glycoproteins, we compared vaccine-
elicited serum neutralizing activity against
ancestral G614 S and these three variant S
pseudoviruses. We used a vesicular stomatitis
virus (VSV) pseudotyping system ( 40 ) withACE2-expressing HEK293T as target cells
( 41 ). We obtained plasma samples from
40 individuals, specifically from 13, 15, and
12 individuals who had received two doses
of Pfizer/BioNtech BNT162b2, two doses of
Moderna mRNA-1273, or a single dose of
Janssen Ad26.COV2.S, respectively (table S2).
The median time after final vaccination
was11daysforPfizer/BioNtechBNT162b2,
20 days for Moderna mRNA-1273, and 102 days
for Janssen Ad26.COV2.S. All three vaccines
contain D614 in the S glycoprotein. Twelve
sera samples from individuals vaccinated
with Janssen Ad26.COV2.S were evaluated,
but only eight of these individuals are shown,
because of a lack of neutralization when as-
sessed against G614 S pseudovirus for the re-
maining four samples.
Geometric mean neutralizing Ab titers (GMTs)
for the Pfizer/BioNtech BNT162b2–elicited
plasma were reduced 3.0-, 2.4-, and 4.1-fold
for B.1.617.1, B.1.617.2, and B.1.617.2+ S (GMT:
86, 110, and 63), respectively, compared with
G614 S (GMT: 260) (Fig. 1A and figs. S1 and
S2). Moderna mRNA-1273–elicited plasma GMTs
were reduced 4.1-, 2.6-, and 9.5-fold for B.1.617.1,
B.1.617.2, and B.1.617.2+ S (GMTs 170, 270, and
74), respectively, compared with G614 S (GMT:
700) (Fig. 1B and figs. S1 and S2). The average
neutralization potency of the Janssen Ad26.
COV2.S–elicited plasma was reduced 3.0-, 2.4-,
and 3.5-fold for B.1.617.1, B.1.617.2, and B.1.617.2+
S (GMT: 10, 13, and 8.8), respectively, compared
with G614 S (GMT: 31) (Fig. 1C and figs. S1 and
S2). These data demonstrate that all three var-
iants lead to reductions in neutralization
potency from vaccine-elicited Abs, with B.1.617.2+
causing the greatest decrease, on par with what
was observed for the B.1.351 (Beta) variant of
concern ( 1 , 3 , 42 ). Furthermore, polyclonal Abs
from half of the Janssen-vaccinated individuals
evaluated completely lost their ability to neutral-
ize one or multiple variants in our assay, likely as
a result of the moderate GMTs against G614 S
pseudovirus, which concurs with recent in vitro
and in vivo data ( 43 , 44 ).
Although the B.1.617.1 and B.1.617.2+ var-
iants have greater ability to evade vaccine-
elicited Ab neutralization than B.1.617.2—a
feature that might increase opportunities
for breakthrough infections ( 37 , 45 , 46 )—the
B.1.617.2 variant became dominant worldwide
byJune2021(Fig.1,DtoF).B.1.617.1and
B.1.617.2 sequences were first observed in late
2020, whereas B.1.617.2+ sequences were first
detected in April 2021. As of 30 September 2021,
6806 B.1.617.1, 1,415,761 B.1.617.2, and 3966
B.1.617.2+ sequenced genomes were deposited
in GISAID (table S1). The high incidence of
B.1.617.2 is consistent with recent studies
showing that the B.1.617.2 variant has en-
hanced transmissibility, replication kinetics,
and viral loads in oropharyngeal and nose-
throat swabs of infected individuals relativeSCIENCEscience.org 24 DECEMBER 2021•VOL 374 ISSUE 6575 1621
(^1) Department of Biochemistry, University of Washington,
Seattle, WA 98195, USA.^2 Vir Biotechnology, San Francisco,
CA 94158, USA.^3 Humabs Biomed SA, a subsidiary of Vir
Biotechnology, 6500 Bellinzona, Switzerland.^4 Division of
Allergy and Infectious Diseases, University of Washington,
Seattle, WA 98195, USA.^5 Center for Emerging and Re-
emerging Infectious Diseases, Division of Allergy and
Infectious Diseases, Department of Medicine, University of
Washington School of Medicine, Seattle, WA 98195, USA.
(^6) Institute for Protein Design, University of Washington,
Seattle, WA 98195, USA.^7 Howard Hughes Medical Institute,
University of Washington, Seattle, WA 98195, USA.
*Corresponding author. Email: [email protected]
RESEARCH | REPORTS