Science - 2020 08 07.pdf

RESEARCH | REPORT

this step allowed us to isolate C3′- and C5′- Grignard reagents in CH2Cl2 proved to be most ture of magnesium alkoxides 42a and 42b
protected NAs directly (e.g., 34 and 35, Fig. compatible, and executing the reactions at with ammonium chloride followed by a sub-
3D). To demonstrate that these acetonide- −78°C proved necessary because higher re- sequent Lewis acid–promoted AFD using InCl3
action temperatures promoted 1,2-hydride shift gave the anomeric a-D-NA 44. In this case,
protected NAs can be further derivatized using the magnesium alkoxides 42a and 42b cy-
standard protocols, several C2′-modified NAs and fluoride displacement. These reactions clize selectively using base or Lewis acid to
were prepared, including C2′-oxo (36), C2′- afford access to a-L– and a-D–configured NAs,
deoxy (37), C2′-3° alcohol (38), and C2′-epi generally gave mixtures of tertiary alcohols which are both unnaturally configured NAs
(39) (Fig. 3F). of contemporary interest to medicinal chem-
with a preference for addition to the re face ists (14–16).
Having ready access to a range of aFAR (24), and when the crude reaction mixture
products, we anticipated that the addition of was allowed to warm to room temperature, These results led us to examine the reac-
the intermediate magnesium alkoxide 42a tion of several additional Grignard reagents
organometallic reagents (rather than hydride underwent AFD to provide the C4′-modified with aldol adducts containing triazole, dea-
reduction) would lead directly to C4′-modified NA 43 directly. This short sequence enables zaadenine, thymine, pyrazole, or trifluoro-
NAs. Toward this goal, we examined reactions access to enantiomerically enriched C4′- methyluracil functions (Fig. 4A). We found
of fluorohydrin 41 with a range of organo- modified NAs in only three steps from simple that diastereoselectivity depends on both the
metallic reagents under a variety of reac-
achiral heterocycles and bromoacetaldehyde
tion conditions (see supplementary materials).
diethyl acetal. Alternatively, quenching the mix-

Fig. 2. A short synthesis of
the pyrazolyl NA 17. (A) Ribose-
last approach in a proposed prebiotic
synthesis of nucleosides inspired
a synthetic strategy to NAs.
(B) Proline-catalyzed aFAR, followed
by reduction and AFD. (2.5x),
2.5 equivalence. (C) Mechanistic
studies of the AFD process. NFSI,
N-fluorobenzenesulfonimide;
DMF, dimethylformamide; MeCN,
acetonitrile; OTf, triflate; equiv,
equivalence; ND, not determined.

Meanwell et al., Science 369, 725–730 (2020) 7 August 2020 3 of 6

RESEARCH | REPORT

solvent and the heterocycle, and reactions in from those generated in CH2Cl2. The addition that cyclized to the naturally configured
tetrahydrofuran (THF) generally gave mixtures of MeMgI to ketofluorohydrins substituted NA 48. Thus, subtle differences in chelation
of tertiary alcohols of different composition with triazole gave predominantly 1,3-syn-diols structures involving the heterocycle and/or

Fig. 3. Scope of nucleoside and
NA synthesis. (A) A simple
four-step reaction sequence
converts readily available starting
materials into enantioenriched and
naturally configured b-D-NAs.
(B) AFD to produce uracil, thymine,
pyrazolyl, and 5-pyrimidinyl
nucleosides; and NAs can be
promoted by NaOH. (C) AFD to
produce trifluoromethyl uracil,
triazolyl, phthalimidyl, deazaade-
nine, and adenosine nucleosides;
and NAs can be promoted by
the Lewis acids Sc(OTf)3 or InCl3.
(D) NAs protected at both the
C3′ and C5′ alcohol functions are
now readily available. (E) Using
D-proline to catalyze the aFAR
reaction, unnatural nucleosides
(L-enantiomers) are now readily
available. (F) Exploiting this conve-
nient process, we prepared a small
collection of C2′-modified NAs.
aTEMPO, BAIB, CH2Cl2 (92% from
34). bi) thiocarbonyldiimidazole,
THF; ii) Bu3SnH, azobisisobutyroni-
trile (55% over two steps from 35).
ci) TEMPO, BAIB, CH2Cl2; ii)
MeMgBr, THF, −78°C (80% over
two steps from 34). dDAST, CH2Cl2
then HCl, methanol (MeOH)
(53% from 35). TEMPO, 2,2,6,6-
Tetramethylpiperidin-1-yl)oxyl;
BAIB, bis(acetoxy)iodobenzene;
THF, tetrahydrofuran; DAST,
diethylaminosulfur trifluoride;
aq, aqueous; rt, room temperature;
SI, supplementary materials;
d.r., diastereomeric ratio.

Meanwell et al., Science 369, 725–730 (2020) 7 August 2020 4 of 6

RESEARCH | REPORT

Fig. 4. Rapid synthesis of C4′-modified and other NAs. (A to E) Diversification and large-scale production of aFAR products provide access to C4′-modified NAs,
iminonucleosides, and LNAs to support medicinal chemistry efforts. aYield from keto-fluorohydrin aldol adduct. bCombined yield of diastereomers. cProduct after
heating of crude reaction mixture to 50°C with camphorsulfonic acid (CSA) and dimethoxyacetone. dProduct after treatment of crude reaction mixture with aqueous
HCl. eStarting from a single fluorohydrin 59.

Meanwell et al., Science 369, 725–730 (2020) 7 August 2020 5 of 6

RESEARCH | REPORT

b-alkoxide function and organomagnesium of locked nucleic acids (LNAs) (34). These 17. P. Merino, Ed., Chemical Synthesis of Nucleoside Analogues
reagents play a notable stereodetermining conformationally restricted NAs demonstrate (Wiley, 2013).
improved stability, and their incorporation
role in these 1,2-addition reactions. Control- in antisense oligonucleotides can lead to sub- 18. M. Brodszki et al., Org. Process Res. Dev. 15, 1027–1032
stantial increases in specificity and potency. (2011).
ling this aspect of the process to selectively We evaluated the addition of alkynylmagne-
access b-D–configured NAs will be the sub- sium chloride to the thymine-containing aldol 19. M. McLaughlin et al., Org. Lett. 19, 926–929 (2017).
ject of future studies. In this study, a col- adduct 59 and found that the reaction gave 20. K. R. Campos et al., Science 363, eaat0805 (2019).
lection of deazaadenine-substituted NAs 43 two diastereomeric addition products 63 and 21. M. Peifer, R. Berger, V. W. Shurtleff, J. C. Conrad,
to 47 were readily accessed as both a- and 64. The major product was transformed di-
b-anomers. In general, and as noted in Fig. 3, rectly into the unusual LNA 65 by simply D. W. MacMillan, J. Am. Chem. Soc. 136, 5900–5903
base-promoted cyclizations resulted in C3′-OH– reacting with NaOH, which promoted both (2014).
and C5′-OH–protected NAs (e.g., 49 to 54), the AFD reaction and a subsequent cyclization 22. J. S. Teichert, F. M. Kruse, O. Trapp, Angew. Chem. Int. Ed. 58,
whereas Lewis acid–promoted cyclizations re- between the free alcohol function and alkyne. 9944–9947 (2019).
sulted in deprotection or protecting group This four-step total synthesis is comparable 23. M. Bergeron-Brlek, T. Teoh, R. Britton, Org. Lett. 15,
migration (e.g., 44 and 47). Densely function- to the 23-step route reported for the analo- 3554–3557 (2013).
alized L-configured C4′-modified NAs could gous uracil LNA 67 (35). We were also able 24. C. Grondal, D. Enders, Adv. Synth. Catal. 349, 694–702
be rapidly accessed from the corresponding to generate the unusual alkyne-functionalized (2007).
LNA 68, an unreported scaffold in nucleo- 25. W. Ren et al., Nat. Commun. 9, 3243 (2018).
ketofluorohydrin aldol adducts, including NAs side chemistry, by simply effecting an AFD 26. A. Quintard, J. Rodriguez, ACS Catal. 7, 5513–5517
of the 1,2-addition product 64. From here, (2017).
substituted with methyl, cyclopropyl, aryl, and formation of the 2,2′-anhydrothymidine fol- 27. R. Britton, B. Kang, Nat. Prod. Rep. 30, 227–236
lowed by deprotection and treatment with (2013).
alkynyl groups (Fig. 4). From this study, it is base in warm DMF (36) gave the LNA 68. This 28. F. A. Davis, P. V. N. Kasu, G. Sundarababu, H. Qi, J. Org. Chem.
clear that larger collections of C4′-modified scaffold is primed for further diversification 62, 7546–7547 (1997).
NAs (e.g., focused screening libraries) are now through standard click or Sonagashira cou- 29. D. D. Steiner, N. Mase, C. F. Barbas 3rd, Angew. Chem. Int. Ed.
pling reactions. 44, 3706–3710 (2005).
readily available. It is worth highlighting 30. E. M. Sánchez-Fernández et al., Chem. Eur. J. 18, 8527–8539
that each of the C4′-methyl, cyclopropyl, p- The demonstration of several aFARs on (2012).
methoxyphenyl, p-chlorophenyl, and alkynyl scales >10 g and up to 400 g supports the 31. T. C. Britton, M. E. LeTourneau, Process for Anomerizing
NAs 43 to 54 were prepared in only three use of this strategy in the process scale pro- Nucleosides, U.S. Patent 5,420,266A (1995).
or four steps total, which compares favorably duction of NAs. Several generations of medi- 32. E. De Clercq, J. Med. Chem. 59, 2301–2311 (2016).
cinal chemistry and total synthesis have 33. A. M. Hyde, R. Calabria, R. Arvary, X. Wang, A. Klapars,
to contemporary syntheses. provided reliable templates for nucleoside Org. Process Res. Dev. 23, 1860–1871 (2019).
and NA synthesis; however, this study pro- 34. M. A. Campbell, J. Wengel, Chem. Soc. Rev. 40, 5680–5689
Considering the potential for this process vides opportunities that should influence the (2011).
construction of diversity libraries and support 35. P. P. Seth, E. E. Swayze, 6-Disubstituted or Unsaturated
to affect the large-scale production of NAs, we future efforts in both drug discovery and Bicyclic Nucleic Acid Analogs, U.S. Patent 8278283 B2
development. (2008).
examined the synthesis of the D-uridine de- 36. T. Yamaguchi, M. Horiba, S. Obika, Chem. Commun. 51,
rivative 56 starting with 900 g of uracil. With- REFERENCES AND NOTES 9737–9740 (2015).
out additional optimization, we were able to
generate ~380 g of the aldol adduct 55 (Fig. 1. G. M. Blackburn, M. J. Gait, D. Loakes, D. M. Williams, Eds., ACKNOWLEDGMENTS
2B), which was converted into the protected Nucleic Acids in Chemistry and Biology (Royal Society of
uridine 56 in excellent yield. Oxidation of Chemistry Publishing, 2006). The authors thank D. McKearney (Simon Fraser University) and
the C2′-OH function followed by deprotection J. A. Newman (Merck & Co., Inc.) for carrying out x-ray
and addition of MeMgBr in THF gave the 2. C. M. Galmarini, J. R. Mackey, C. Dumontet, Lancet Oncol. 3, crystallographic analyses. Funding: R.B. acknowledges support
tertiary alcohol 57. This later compound is a 415–424 (2002). from the Canadian Glycomics Network (Strategic Initiatives
previously reported intermediate in the large- Grant, CD-46), the Natural Sciences and Engineering Research
scale production of MK-3682 (uprifosbuvir: 58) 3. E. De Clercq, Med. Res. Rev. 33, 1215–1248 (2013). Council (NSERC) of Canada (Discovery Grant, RGPIN-2019-
(33), an HCV NS5B RNA polymerase inhib- 4. L. P. Jordheim, D. Durantel, F. Zoulim, C. Dumontet, 06468), and Merck & Co., Inc. M.M. was supported by an NSERC
itor developed for the treatment of hepatitis C CGSD2 scholarship, and J.L. was supported by the Deutsche
Nat. Rev. Drug Discov. 12, 447–464 (2013). Forschungsgemeinschaft (DFG). Author contributions: M.M.
virus (HCV). 5. J. Shelton et al., Chem. Rev. 116, 14379–14455 and S.M.S. developed the methodology; M.M., S.M.S., J.L.,
B.A., and Y.W. performed the experiments; R.C. developed a
We also briefly assessed the utility of this (2016). protocol for assigning stereochemistry by nuclear magnetic
6. B. Ewald, D. Sampath, W. Plunkett, Oncogene 27, 6522–6537 resonance (NMR) spectroscopic and computational methods;
process for accessing an unusual class of NAs L.-C.C. and R.B. were responsible for project administration;
(2008). R.B. conceptualized the research; L.-C.C. and R.B. supervised
known as iminonucleosides, wherein the fura- 7. K. L. Seley-Radtke, M. K. Yates, Antiviral Res. 154, 66–86 the execution of experiments; and R.B. wrote the original
draft. Competing interests: Simon Fraser University and
nose oxygen is replaced by a nitrogen atom. In (2018). Merck & Co., Inc. have filed a patent application describing
8. M. K. Yates, K. L. Seley-Radtke, Antiviral Res. 162, 5–21 the synthesis of nucleoside analogs via the process presented
a single example (Fig. 4C), we observed that in this manuscript—U.S. provisional patent application
(2019). no. 62/994,349. Data and materials availability: All data
reductive amination of the fluorohydrin aldol 9. H. Ma et al., J. Biol. Chem. 282, 29812–29820 (2007). are available in the manuscript or the supplementary
adduct 59 (isolated as a single diastereomer, 10. E. P. Gillis, K. J. Eastman, M. D. Hill, D. J. Donnelly, materials. X-ray structures are deposited at the Cambridge
as shown) using benzyl amine and followed Crystallographic Data Centre under reference numbers 1955427,
by a basic work-up led directly to the b-D– N. A. Meanwell, J. Med. Chem. 58, 8315–8359 (2015). 2008890, and 1955420.
configured iminonucleoside 60 in good yield. 11. H. Ohrui, Chem. Rec. 6, 133–143 (2006).
To further demonstrate the advantages of 12. J. T. Witkowski, R. K. Robins, R. W. Sidwell, L. N. Simon, J. Med. SUPPLEMENTARY MATERIALS
this strategy, we prepared a C4′-modified
deoxy NA (Fig. 4D). In this case, C4′-allyl ri- Chem. 15, 1150–1154 (1972). science.sciencemag.org/content/369/6504/725/suppl/DC1
bothymidine 61 was accessed through the 13. J. Zeidler, D. Baraniak, T. Ostrowski, Eur. J. Med. Chem. 97, Materials and Methods
addition of allylmagnesium bromide to the Supplementary Text
fluorohydrin 59 followed by base-promoted 409–418 (2015). Figs. S1 to S14
AFD. A Barton-McCombie deoxygenation then 14. G. Ni et al., RSC Advances 9, 14302–14320 (2019). Tables S1 to S12
gave the C4′-allyl NA 62 in only six steps from 15. G. Gumina, G. Y. Song, C. K. Chu, FEMS Microbiol. Lett. 202, Liquid Chromatography Chromatograms
thymine. NMR Data
9–15 (2001). References (37–52)
Finally, we aimed to exploit this facile C4′- 16. H. Cui et al., J. Med. Chem. 55, 10948–10957 MDAR Reproducibility Checklist
functionalization strategy in the preparation
(2012). 15 February 2020; accepted 5 June 2020
10.1126/science.abb3231

Meanwell et al., Science 369, 725–730 (2020) 7 August 2020 6 of 6

RESEARCH

CORONAVIRUS threefold greater than background staining ob-
served with a SARS-CoV–naïve donor sample
Broad neutralization of SARS-related viruses by (Fig. 1A). Cognate antibody heavy- and light-chain
human monoclonal antibodies pairs were amplified from 315 individual SARS-
CoV-2–reactive B cells by single-cell reverse tran-
Anna Z. Wec1, Daniel Wrapp2, Andrew S. Herbert3, Daniel P. Maurer1, Denise Haslwanter4, scription polymerase chain reaction (RT-PCR)
Mrunal Sakharkar1, Rohit K. Jangra4, M. Eugenia Dieterle4, Asparouh Lilov1, Deli Huang5, and subsequently cloned and expressed as full-
Longping V. Tse6, Nicole V. Johnson2, Ching-Lin Hsieh2, Nianshuang Wang2, Juergen H. Nett1, length immunoglobulin Gs (IgGs) in an engi-
Elizabeth Champney1, Irina Burnina1, Michael Brown1, Shu Lin1, Melanie Sinclair1, Carl Johnson1, neered strain of Saccharomyces cerevisiae (14).
Sarat Pudi1, Robert Bortz III4, Ariel S. Wirchnianski4, Ethan Laudermilch4, Catalina Florez4, Of the 315 cloned antibodies, 200 bound to
J. Maximilian Fels4, Cecilia M. O’Brien3, Barney S. Graham7, David Nemazee5, Dennis R. Burton5,8,9,10, SARS-CoV-2 S in preliminary binding screens
Ralph S. Baric6,11, James E. Voss5, Kartik Chandran4, John M. Dye3, (Fig. 1B). Sequence analysis revealed that
Jason S. McLellan2, Laura M. Walker1* about half of the clones were members of
expanded clonal lineages, whereas the other
Broadly protective vaccines against known and preemergent human coronaviruses (HCoVs) are half were unique (Fig. 1C). Moreover, about
urgently needed. To gain a deeper understanding of cross-neutralizing antibody responses, we mined the 30% of isolated antibodies displayed convergent
memory B cell repertoire of a convalescent severe acute respiratory syndrome (SARS) donor and VH1-69/VK2-30 germline gene pairing (Fig. 1C).
identified 200 SARS coronavirus 2 (SARS-CoV-2) binding antibodies that target multiple conserved sites As expected, almost all the antibodies were
on the spike (S) protein. A large proportion of the non-neutralizing antibodies display high levels of somatically mutated, with members of clo-
somatic hypermutation and cross-react with circulating HCoVs, suggesting recall of preexisting memory nally expanded lineages showing significantly
B cells elicited by prior HCoV infections. Several antibodies potently cross-neutralize SARS-CoV, higher levels of somatic hypermutation (SHM)
SARS-CoV-2, and the bat SARS-like virus WIV1 by blocking receptor attachment and inducing compared with unique clones (Fig. 1D). Finally,
S1 shedding. These antibodies represent promising candidates for therapeutic intervention and reveal consistent with the respiratory nature of SARS-
a target for the rational design of pan-sarbecovirus vaccines. CoV infection, index sorting analysis revealed
that 33% of binding antibodies originated from
I n December 2019, a novel pathogen emerged both SARS-CoV and SARS-CoV-2 functions as IgA+ MBCs and the remaining 66% from IgG+
in the city of Wuhan in China’s Hubei the receptor-binding domain (RBD) for the MBCs (Fig. 1E). We conclude that SARS-CoV
province, causing an outbreak of atypical shared entry receptor, human angiotensin- infection elicited a high frequency of long-lived,
converting enzyme 2 (hACE2) (6–10). The S2 cross-reactive MBCs in this donor.
pneumonia [a disease known as corona- subunit contains the fusion peptide, heptad
repeats 1 and 2, and a transmembrane do- We next measured the apparent binding af-
virus disease 2019 (COVID-19)]. The in- main, all of which are required for fusion of finities (KDApps) of the antibodies to prefusion-
the viral and host cell membranes. stabilized SARS-CoV and SARS-CoV-2 S proteins
fectious agent was characterized as a lineage (5). Although most antibodies (153 out of 200)
The S glycoprotein of HCoVs is the primary showed binding to both S proteins, a subset
B betacoronavirus, named severe acute respira- target for neutralizing antibodies (nAbs) (11). appeared to be SARS-CoV-2 S-specific (Fig. 2A).
SARS-CoV and SARS-CoV-2 share 76% amino This result was unexpected given that the
tory syndrome coronavirus 2 (SARS-CoV-2) and acid identity in their S proteins, raising the antibodies were isolated from a SARS-CoV–
possibility of conserved immunogenic surfaces experienced donor and may relate to differ-
shown to be closely related to SARS-CoV and on these antigens. Studies of convalescent sera ences between the infecting SARS-CoV strain
and a limited number of monoclonal anti- and the recombinant SARS-CoV S protein (Tor2)
several SARS-like bat CoVs (1). There are cur- bodies (mAbs) have revealed limited to no used for the binding studies. Alternatively, this
rently no approved vaccines or therapeutics avail- cross-neutralizing activity, demonstrating that result may be due to inherent differences in
conserved antigenic sites are rarely targeted the stability or antigenicity of recombinant
able for the prevention or treatment of COVID-19. by nAbs (5, 9, 12, 13). However, the frequencies, prefusion-stabilized SARS-CoV and SARS-CoV-2
specificities, and functional activities of cross- S proteins. Indeed, about 30% of antibodies
CoV entry into host cells is mediated by the reactive antibodies induced by natural SARS- that failed to bind recombinant SARS-CoV S
CoV and SARS-CoV-2 infection remain poorly displayed reactivity with SARS-CoV S expressed
viral S glycoprotein, which forms trimeric defined. on the surface of transfected cells, providing
some evidence for differences in the anti-
spikes on the viral surface (2). Each monomer We aimed to comprehensively profile the genicity of recombinant and cell-expressed
in the trimeric S assembly is a heterodimer cross-reactive B cell response induced by SARS- forms of S (fig. S1).
CoV infection by cloning an extensive panel
of S1 and S2 subunits. The S1 subunit is of SARS-CoV-2 S-reactive mAbs from the pe- Paradoxically, most of the highly mutated
ripheral B cells of a convalescent donor (do- and clonally expanded antibodies bound weakly
composed of four domains: an N-terminal nor 84) who survived the 2003 SARS outbreak. (KDApps > 10 nM) to both SARS-CoV and SARS-
To isolate cross-reactive antibodies, we ob- CoV-2 S (Fig. 2B). We sought to determine if
domain (NTD), a C-terminal domain (CTD), tained a blood sample from this donor about these antibodies originated from preexisting
and subdomains I and II (3–5). The CTD of 3 years after infection and stained purified MBCs induced by prior exposures to naturally
B cells with a panel of memory B cell (MBC) circulating HCoVs, which share up to 32%
1Adimab LLC, Lebanon, NH 03766, USA. 2Department of markers and a fluorescently labeled SARS- S amino acid identity with SARS-CoV and
Molecular Biosciences, The University of Texas at Austin, CoV-2 S protein. Flow cytometric analysis SARS-CoV-2. Accordingly, we assessed bind-
Austin, TX 78712, USA. 3U.S. Army Medical Research revealed that 0.14% of class-switched MBCs ing of the antibodies to recombinant S proteins
Institute of Infectious Diseases, Frederick, MD 21702, USA. were SARS-CoV-2 S-reactive, which was about of naturally circulating human alphacoronavi-
4Department of Microbiology and Immunology, Albert ruses (HCoV-NL63 and HCoV-229E) and beta-
Einstein College of Medicine, New York, NY 10462, USA. coronaviruses (HCoV-OC43 and HCoV-HKU1).
5Department of Immunology and Microbiology, The Scripps
Research Institute, La Jolla, CA 92037, USA. 6Department of
Epidemiology, The University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599, USA. 7Vaccine Research Center,
National Institute of Allergy and Infectious Diseases, National
Institutes of Health, Bethesda, MD 20892, USA. 8IAVI
Neutralizing Antibody Center, The Scripps Research Institute,
La Jolla, CA 92037, USA. 9Consortium for HIV/AIDS Vaccine
Development (CHAVD), The Scripps Research Institute, La
Jolla, CA 92037, USA. 10Ragon Institute of Massachusetts
General Hospital, Massachusetts Institute of Technology, and
Harvard, Cambridge, MA 02139, USA. 11Departments of
Microbiology and Immunology, The University of North
Carolina at Chapel Hill, Chapel Hill, NC 27599, USA.
*Corresponding author. Email: [email protected]

Wec et al., Science 369, 731–736 (2020) 7 August 2020 1 of 6

RESEARCH | REPORT

A Gated on CD19+ swIg+ B cells Donor 84 BSARS-CoV-2 S Binding (nm) analysis revealed that between 0.06 and
0.12% of total B cells in the three naïve donors
105 NaïveSARS-CoV-2 S | APC 0.14% 1.5 n=315 displayed SARS-CoV-2 reactivity (fig. S5B).
More than 350 SARS-CoV-2–reactive MBCs
4 0.051% 1.0 were sorted and amplified by single-cell RT-
PCR, and 141 variable region of Ig heavy chain
10 (VH)–variable region of Ig light chain (VL) pairs
were cloned and expressed as full-length IgGs.
3 63.5% Although a limited number of SARS-CoV-2 S
10 binding antibodies (3 to 22) were isolated from
IgA+ all three naïve donors, they displayed signif-
0 0.5 IgG+ icantly lower levels of SHM, clonal expansion,
0.0 and KDApps for both SARS-CoV and SARS-
3 CoV-2 S compared with the cross-reactive
-10 antibodies identified from donor 84 (Fig. 2, F
and G, and fig. S5C). Altogether, these results
33 45 33 45 suggest that SARS-CoV infection likely led to
-10 0 10 10 10 -10 0 10 10 10 the activation and expansion of preexisting
cross-reactive MBCs induced by circulating
SARS-CoV-2 S | PE SARS-CoV-2 S | PE HCoV exposure in this donor.

C D **** E To map the antigenic sites recognized by
the SARS-CoV and SARS-CoV-2 cross-reactive
40 100 antibodies isolated from donor 84, we per-
formed binding experiments using a panel of
200 #VH nucleotide substitutions 30 % SARS-CoV-2+ MBCs 75 recombinant S protein subunits and individ-
20 50 ual domains. Because of the inherent technical
challenges associated with measuring binding
10 25 of low-affinity antibodies to monomeric pro-
teins, we analyzed only the 64 high-affinity
VH1-69/VK2-30 0 0 binders (KDApps < 10 nM) to SARS-CoV-2 S
Other germlines Expanded (Fig. 2, A and B). We first evaluated binding
Unique to recombinant SARS-CoV-2 S1 and S2 sub-
units and observed that 75% of the antibodies
Fig. 1. Isolation of SARS-CoV-2 S-specific antibodies. (A) Frequency of SARS-CoV-2 S-reactive B cells recognized epitopes within S1, whereas the
remaining 25% bound to epitopes within S2
in donor 84 and a SARS-CoV–naïve donor. Fluorescence-activated cell sorting plots are gated on (Fig. 3A). Two of the S2-directed antibodies
CD19+CD20+IgD−IgM− B cells. swIg, switched immunoglobulin. (B) Binding of 315 isolated antibodies to also showed strong reactivity with OC43 S,
suggesting recognition of a conserved anti-
SARS-CoV-2 S, as determined by BLI. The dashed line indicates the threshold for designating binders genic site (fig. S6). We next evaluated the
49 S1-directed antibodies for reactivity with
(0.1 nm). (C) Clonal lineage analysis. Each lineage is represented as a segment proportional to the lineage individual SARS-CoV-2 RBD and NTD pro-
teins and found that 21 (43%) and 28 (57%) of
size. The total number of antibodies is shown in the center of the pie. Clonal lineages were defined based on the S1-specific antibodies recognized the RBD
and NTD, respectively (Fig. 3A).
the following criteria: identical VH and VL germline genes, identical CDR H3 (third complementarity-
determining region of the heavy chain) length, and CDR H3 amino acid identity ≥80%. (D) Somatic To further define the epitopes recognized
mutation load, expressed as the number of nucleotide substitutions in VH, in unique antibodies and members by the 21 RBD-directed antibodies, we per-
of expanded clonal lineages. Red bars indicate medians. Statistical comparisons were made using the formed competitive binding studies with re-
Mann-Whitney test (****P < 0.0001). (E) Proportion of SARS-CoV-2 S binders derived from IgG+ and IgA+ combinant hACE2 and a previously described
B cells, as determined by index sorting. antibody, CR3022, that targets a conserved
epitope that is distinct from the receptor
More than 80% of the low-affinity (KDApps > MBCs, indicating a mucosal origin, whereas binding site (Fig. 3B and fig. S7) (15). Six of
10 nM) SARS-CoV and SARS-CoV-2 cross- most of the SARS-CoV and SARS-CoV-2 cross- the antibodies competed only with hACE2,
reactive antibodies reacted with one or more reactive antibodies originated from IgG+ MBCs three competed only with CR3022, four com-
of the HCoV S proteins, suggesting that SARS- (Fig. 2E). Finally, all of the broad binders peted with both hACE2 and CR3022, and seven
CoV infection may have boosted a preexisting lacked polyreactivity, demonstrating that did not compete with hACE2 or CR3022 (Fig.
MBC response induced by circulating HCoVs their cross-binding is not due to nonspecific 3B). Thus, these antibodies delineate at least
(Fig. 2B). Consistent with this hypothesis, the cross-reactivity (fig. S4). four adjacent and potentially overlapping sites
broadly cross-reactive antibodies showed sig- within the RBD. Most of the antibodies that
nificantly higher levels of SHM and clonal To investigate whether the above results competed with recombinant hACE2 binding to
expansion compared with those that only were due to an original antigenic sin phenom- SARS-CoV-2 RBD in the biolayer interferome-
recognized SARS-CoV and SARS-CoV-2 (Fig. enon, or rather simply due to avid binding try (BLI) assay also interfered with binding of
2, B to D). Furthermore, 72% of the broadly of circulating HCoV-specific B cell receptors full-length SARS-CoV-2 S to endogenous ACE2
binding antibodies used VH1-69/VK2-30 germ- to the SARS-CoV-2 S tetramers used for cell expressed on the surface of Vero E6 cells
line gene pairing, suggesting germline-mediated sorting, we assessed whether similarly broadly (Fig. 3C). The four antibodies (ADI-55951,
recognition of a common antigenic site (Fig. 2B binding antibodies were also present in SARS-
and fig. S2). Although we were unable to finely CoV– and SARS-CoV-2–naïve donors that had
map the epitopes recognized by these anti- been exposed to endemic HCoVs. We obtained
bodies, none of them bound to recombinant peripheral blood mononuclear cell (PBMC)
SARS-CoV-2 S1, suggesting that they likely samples from three healthy adult donors with
target epitopes within the more conserved S2 serological evidence of circulating HCoV expo-
subunit (fig. S3). Index sorting analysis re- sure and no history of SARS-CoV or SARS-
vealed that the majority of the broadly cross- CoV-2 infection and stained the corresponding
reactive antibodies were derived from IgA+ B cells with a fluorescently labeled SARS-
CoV-2 S probe (fig. S5A). Flow cytometric

Wec et al., Science 369, 731–736 (2020) 7 August 2020 2 of 6

RESEARCH | REPORT

ADI-55993, ADI-56000, and ADI-56035) that formed neutralization assays using both murine results were observed in the VSV-based pseu-
showed stronger competition in the BLI assay leukemia virus (MLV)– and vesicular stomatitis dovirus assay (fig. S8). Of the eight RBD-
displayed weak binding affinities for SARS- virus (VSV)–based pseudotype systems as well directed nAbs, four targeted epitopes overlapping
CoV-2 S (fig. S12), which likely explains their as authentic SARS-CoV-2. Because of the large with both the hACE2 and CR3022 epitopes
lower level of competition in the cell-surface number of antibodies, we first measured infec- and the other four recognized epitopes over-
assay. Thus, SARS-CoV infection elicited high- tion inhibition of authentic SARS-CoV-2 at a lapping only the hACE2 epitope, suggesting
affinity cross-reactive antibodies to a range single concentration of purified IgG. Only 9 the existence of two partially overlapping neu-
of antigenic sites within both the S1 and S2 out of 200 antibodies displayed neutralizing tralizing epitopes within the RBD (Fig. 3B).
subunits. activity at the 100 nM concentration tested, Neutralization titration studies revealed that
eight targeted the RBD, and the remaining the median inhibitory concentrations (IC50s)
To evaluate the neutralization activities of one recognized the NTD (Fig. 3D). Similar of the RBD-directed nAbs ranged from 0.05
the SARS-CoV-2 binding antibodies, we per-

Fig. 2. Binding properties of A B K App [nM] L2SSSOHNGi2LAeAnHKCr96e4RRUME3am31lSSig-ne2e
SARS-CoV-2 S-specific D
antibodies. (A) Apparent n.b.
binding affinities (KDApps) of p.f. n.b.
SARS-CoV-2 S-specific IgGs
for prefusion-stabilized >100-1000 1
SARS-CoV and SARS-CoV-2
S proteins, as determined by >10-100
BLI. Low-affinity clones for
which binding curves could not [M] 1-10
be fit are designated as “poor
fit” (p.f.) on the plot. n.b., <1 10
nonbinder. (B) IgG KDApps for
SARS-CoV-2, SARS-CoV, 229E, AppK 10-7 Germline 20
HKU1, NL63, and OC43 S D othe r 30
proteins. Germline gene usage, 40
clonality, and SHM are S VH1-6 9 / 50
presented in the three left- VK2-30
most columns. SHM load SARS-CoV-2 10-8
is represented as the number Lineage
of nucleotide substitutions 10-9 unique
expanded

SHM

40

30

10-10 10-9 10-8 10-7 p.f. n.b. 20
10-10
E 100 10 60

SARS-CoV S K App [M] 0
D
70

C **** D Broad IgA+ 80
IgG +
40
# VH nucleotide substitutions
in VH. (C) Load of somatic 30 75 90
mutations in broadly cross-
reactive and SARS-CoV– 77
% Antibodies
and SARS-CoV-2–specific 20 SARS-1/2 only 50 100
antibodies. Red bars indicate 10 25 110
medians. (D) Degree of clonal
expansion in broadly cross-

reactive and SARS-CoV– 0 SARS-1/2 78 Broad 120
and SARS-CoV-2–specific Reactivity Broad only SoAnlRyS-1/2 130
antibodies. Each lineage 0
is represented as a segment Reactivity

proportional to the lineage F **** G SARS-CoV-2 S SARS-CoV S 140
size. The total number of anti- # VH nucleotide substitutions 150
bodies is shown in the center 40 (coDomnbNioarnïev8de4) 0.8 ********* ********** 160
of the pie. (E) Proportion of 170
broadly cross-reactive and 30 0.6Binding (nm)
SARS-CoV– and SARS-CoV-2–
specific antibodies derived 20 0.4
from IgG+ and IgA+ B cells, as
determined by index sorting.

(F) Load of somatic mutations 10 0.2 180
in SARS-CoV-2 S-reactive

antibodies isolated from three 0 0.0 190
naïve donors and donor 84.

Antibodies from healthy DDDDoooonnnnorooorrr82314
DDDDoooonnnnorooorrr82134
donors were combined for this 200
analysis. (G) Binding activity

of antibodies isolated from

SARS-CoV-2 S-reactive B cells in donor 84 and three naïve donors to SARS-CoV and SARS-CoV-2 S proteins, as determined by BLI. Statistical comparisons

were made using the Mann-Whitney test (**P < 0.01; ***P < 0.001; ****P < 0.0001).

Wec et al., Science 369, 731–736 (2020) 7 August 2020 3 of 6

RESEARCH | REPORT

A S2 hACE2 B AAAAAAAAAAAAAAAAAAAADDDDDDDDDDDDDDDDDDDDIIIIIIIIIIIIIIIIIIII--------------------5555555555555555555565656655555556566665060009776606960600091805354994789954925509146059098143820018 % Neutralization
NTD 64 Competitor? at 100 nM
hACE2
C Yes CR3022 100
No
100
Competitor? 75
Yes 50
No 25

SARS-CoV-2 0
authentic

RBD

% SARS-CoV-2 S Binding 10 RBD D % SARS-CoV-2 Neutralization
hACE 2 at 100 nM
1 hACE2/CR3022 100
CR3022 80 non-hhAACCEE22
Other 60 NTD
40 S2
NTD 20 Other
S2 0

*

AAAAAAAAAAAAAAAAAAAAAANoDDDDDDDDDDDDDDDDDDDDDDIIIIIIIIIIIIIIIIIIIIIIC----------------------AR5555555555555555555555nKt66666555565356656555566i000006609Z60770060969600g052359945925437580985914e262812048904295180391006n

E Authentic MLV pseudotype F RBD NTD
S2
SARS-CoV SARS-CoV hACE2
hACE2/CR3022 Other
SARS-CoV-2 SARS-CoV-2 CR3022 G

WIV1 100 180

10 % SARS-CoV-2 Neut. at 100 nM 160 ADI-55688
ADI-55689
Virus neutralization 1 75 140 % Binding relative to baseline ADI-55993
IC50 / PRNT50 (µg/mL) 120 5 ADI-56000
30 ADI-56046
0.1 50 100 60 ADI-55690
80 110200 ADI-56010
180 ADI-55951
0.01 25 60 240 ADI-56020
CR3022
0.001 0 40 ADI-56040
20 ADI-56032
1 10 100/w.b. ADI-56006

AAAAAAAADDDDDDDDIIIIIIII--------5555555556556655660060998941590868031009 On-cell SARS-CoV-2 S Time (min)
binding EC (nM)

50

Fig. 3. Epitope mapping and neutralization screening. (A) Proportion of SARS- MLV pseudovirus (strain n-CoV/USA_WA1/2020) using HeLa-ACE2 target cells, and
CoV-2 S-specific antibodies targeting each of the indicated antigenic sites. neutralization of authentic SARS-CoV, SARS-CoV-2, and WIV1-CoV using Vero E6
(B) Heat map showing the competitive binding profiles of the RBD-directed target cells. Data represent two technical replicates. (F) Binding EC50s for
antibodies, as determined by BLI (top) and percent neutralization of authentic cell-surface SARS-CoV-2 S are plotted against the percent neutralization
SARS-CoV-2 at a 100 nM concentration (bottom). (C) Antibody inhibition of of authentic SARS-CoV-2 at 100 nM. Background binding was assessed using
SARS-CoV-2 S binding to endogenous ACE2 expressed on Vero E6 cells, as mock-transfected HEK-293 cells. Data points are colored according to epitope
determined by flow cytometry. Antibodies were mixed with recombinant SARS-CoV-2 specificity. RBD-directed antibodies are further categorized based on their
S bearing a Twin-Strep tag at a molar ratio of 10:1 before adding to Vero E6 cells. competition group: hACE2 indicates hACE2-only competitors; CR3022 indicates
An anti-ebolavirus antibody (KZ52) was used as an isotype control. The “no CR3022-only competitors; hACE2/CR3022 indicates antibodies that compete
antigen” control indicates secondary-only staining. The asterisk indicates that no with hACE2 and CR3022; other indicates hACE2 and CR3022 noncompetitors.
detectable binding was observed. Bars are colored according to epitope specificity, Antibodies with cell binding EC50s >100 nM are designated as weak binders (w.b.) on
as determined in the BLI competition assay. Data represent three technical the plot. (G) Antibody binding activity to cell-surface SARS-CoV-2 S over time,
replicates. (D) Percent authentic SARS-CoV-2 neutralization in the presence of as determined by flow cytometry. IgGs were incubated with cells expressing WT
100 nM IgG. Antibodies are grouped according to epitope specificity. RBD-directed SARS-CoV-2 over the indicated time intervals. Binding MFI was assessed at
antibodies that compete or do not compete with ACE2 are designated as ACE2 and 240 min for all samples. CR3022 is included for comparison. Curves are colored by
non-ACE2, respectively. (E) Antibody neutralization of SARS-CoV and SARS-CoV-2 epitope specificity, as in (F). Data represent two technical replicates.

Wec et al., Science 369, 731–736 (2020) 7 August 2020 4 of 6

RESEARCH | REPORT

A tein, but neutralizing activity is primarily
restricted to RBD-directed antibodies that
ADI-55688 ADI-55689 ADI-55690 ADI-55993 ADI-56000 ADI-56010 ADI-56046 interfere with receptor binding and promote
S1 dissociation.
B hACE2 C
To structurally characterize the epitopes
binding site ADI-56046 recognized by the RBD-directed nAbs, we
performed negative-stain electron microscopy
ADI-55689 (EM) to observe each of these Fabs bound to
the SARS-CoV-2 S protein. Many of the two-
CR3022 dimensional (2D) class averages that we ob-
binding site tained displayed obvious heterogeneity in the
number of Fabs that were bound to a single
Fig. 4. Structures of cross-neutralizing antibodies bound to SARS-CoV-2 S. (A) Negative-stain EM 2D S trimer, which is likely due to dynamic in-
class averages of SARS-CoV-2 S bound by Fabs of indicated antibodies. The Fabs have been pseudocolored for accessibility of RBD epitopes and substoi-
ease of visualization. (B and C) 3D reconstructions of Fab:SARS-CoV-2 S complexes are shown in transparent chiometric binding of S at the low protein
surface representation (light gray) with the structure of the SARS-CoV-2 S trimer (white surface) and Fabs concentrations used to prepare grids (Fig. 4A)
(ribbon) docked into the density. S-bound Fabs of ADI-55689 (B) and ADI-56046 (C) are colored in orange and (5, 18). The 3D reconstructions of these com-
purple, respectively. The hACE2 and CR3022 binding sites on S are shaded in red and light blue, respectively. plexes support the results of our biophysical
competition assays and show that the RBD-
to 1.4 mg/ml against SARS-CoV-2 and 0.004 to combinant and cell surface S, but none were directed nAbs recognize a single region on
0.06 mg/ml against SARS-CoV in the MLV neutralizing (Fig. 3F). Surprisingly, even within the solvent-exposed surface of the RBD with
assay (Fig. 3E and fig. S9). Comparable neu- the group of hACE2-blocking nAbs, we did not overlapping footprints. ADI-55689, which po-
tralization IC50s were observed in authentic observe a strong correlation between binding tently neutralizes and competes with hACE2,
SARS-CoV and SARS-CoV-2 neutralization to cell surface–S or recombinant-S and neu- appears to bind at the edge of the hACE2
assays (Fig. 3E and fig. S9). By contrast, the tralization, suggesting that antibody potency binding site, close to the more structurally
VSV–SARS-CoV-2 neutralization IC50s were is governed at least in part by factors beyond conserved core domain of the RBD, without
substantially lower (8- to 35-fold) than those binding affinity (Fig. 3F and figs. S12 and S13). overlapping with the CR3022 epitope (Fig.
observed for live SARS-CoV-2 (figs. S9 and To determine whether the hACE2 competitor 4B). ADI-56046, which exemplifies the group
S10). To assess the breadth of neutralization antibodies neutralized by inducing S1 shedding of antibodies that compete with both hACE2
against representative preemergent SARS-like and premature S triggering (17), we incubated and CR3022, binds slightly farther away from
bat CoVs, we measured infection inhibition human embryonic kidney (HEK)–293 cells ex- the flexible tip of the RBD, and thus its epitope
of authentic WIV1-CoV using a plaque reduc- pressing WT SARS-CoV-2 S with saturating spans both the hACE2 binding site and the
tion assay (16). All eight antibodies neutralized concentrations of antibody and measured the CR3022 epitope (Fig. 4C). Our structural anal-
WIV1-CoV, with median plaque reduction neu- median fluorescence intensity (MFI) of anti- ysis suggests that all of the nAbs recognize a
tralization titers (PRNT50s) ranging from 0.076 body binding over time by flow cytometry. single patch on the surface of the RBD with
to 1.7 mg/ml, demonstrating their breadth of Indeed, all of the hACE2-blocking antibodies overlapping footprints. These antibodies po-
activity (Fig. 3E and fig. S11). Crucially, none of showed substantially decreased binding over tently cross-neutralize SARS-CoV, SARS-CoV-2,
the antibodies left an unneutralized viral frac- time, consistent with induced S1 dissociation, and WIV1, suggesting that this antigenic sur-
tion in any of the assays (figs. S9 and S11). whereas antibodies recognizing the NTD, S2 face exhibits extensive conservation among
stem, and RBD epitopes outside of the hACE2 the SARS-like coronaviruses.
We observed little to no correlation between binding site displayed either no change or an
apparent binding affinity for wild-type (WT) increase in binding over time (Fig. 3G). We The potent cross-neutralizing antibodies de-
SARS-CoV-2 cell surface S and neutralizing conclude that SARS-CoV infection induces scribed here bind to conserved epitopes over-
activity. For example, all of the S2-directed high-affinity cross-reactive antibodies targeting lapping the hACE2 binding site, thus illuminating
antibodies and a subset of NTD-directed anti- multiple distinct antigenic sites on the S pro- this antigenic surface as a promising target for
bodies bound with high avidity to both re- the rational design of pan-sarbecovirus vac-
cines. For example, the RBD epitope(s) defined
by this class of antibodies could be presented
on conformationally stable protein scaffolds to
focus the antibody response on this site, as
previously demonstrated for the motavizumab
epitope on respiratory syncytial virus F (19).
Furthermore, the nAbs themselves, alone or in
combination, represent promising candidates
for prophylaxis or therapy of SARS, COVID-19,
and potentially future diseases caused by new
emerging SARS-like viruses.

REFERENCES AND NOTES

1. F. Wu et al., Nature 579, 265–269 (2020).
2. F. Li, Annu. Rev. Virol. 3, 237–261 (2016).
3. A. C. Walls et al., Nature 531, 114–117 (2016).
4. M. A. Tortorici, D. Veesler, Adv. Virus Res. 105, 93–116

(2019).
5. D. Wrapp et al., Science 367, 1260–1263 (2020).
6. W. Song, M. Gui, X. Wang, Y. Xiang, PLOS Pathog. 14,

e1007236 (2018).

Wec et al., Science 369, 731–736 (2020) 7 August 2020 5 of 6

RESEARCH | REPORT

7. J. Lan et al., Nature 581, 215–220 (2020). and Infectious Diseases (NIAID) Vaccine Research Center. accession codes MT565750 to MT566149. Standard material
8. M. Hoffmann et al., Cell 181, 271–280.e8 (2020). Funding: This work was funded in part by National Institutes of transfer agreements (MTAs) for scientific materials will be given to
9. Q. Wang et al., Cell 181, 894–904.e9 (2020). Health (NIH)–NIAID grants awarded to J.S.M. (R01-AI127521) and academic researchers through The Scripps Research Institute.
10. F. Li, J. Virol. 89, 1954–1964 (2015). K.C. (U19 AI142777). D.Hu. and J.E.V. were supported by Yeast-produced IgGs are available from the corresponding author
11. S. Jiang, C. Hillyer, L. Du, Trends Immunol. 41, 355–359 (2020). R01AI132317 and R01AI073148 (to D.N.). J.E.V. was also supported under a MTA from Adimab. This work is licensed under a Creative
12. X. Ou et al., Nat. Commun. 11, 1620 (2020). by the Bill and Melinda Gates Foundation (OPP 1183956 to Commons Attribution 4.0 International (CC BY 4.0) license, which
13. H. Lv et al., Cell Rep. 31, 107725 (2020). J.E.V.). Author contributions: L.M.W., J.S.M., and A.Z.W. permits unrestricted use, distribution, and reproduction in any
14. M. S. Gilman et al., Sci. Immunol. 1, eaaj1879 (2016). conceived and designed the study. A.Z.W., M.Sa., and D.P.M. medium, provided the original work is properly cited. To view a copy
15. M. Yuan et al., Science 368, 630–633 (2020). performed ELISA and cell binding assays. J.S.M., N.W., and D.W. of this license, visit https://creativecommons.org/licenses/by/4.0/.
16. V. D. Menachery et al., Proc. Natl. Acad. Sci. U.S.A. 113, performed the structural studies. A.S.H., D.Ha., R.K.J., M.E.D., This license does not apply to figures/photos/artwork or other
D.Hu., L.V.T., N.V.J., C.-L.H., R.B., A.S.W., E.L., C.F., J.M.F., and content included in the article that is credited to a third party; obtain
3048–3053 (2016). C.M.O. performed neutralization experiments. A.L., E.C., I.B., authorization from the rights holder before using such material.
17. A. C. Walls et al., Cell 176, 1026–1039.e15 (2019). M.B., S.L., and M.Si. performed biolayer interferometry assays.
18. J. Pallesen et al., Proc. Natl. Acad. Sci. U.S.A. 114, J.H.N., C.J., and S.P. expressed and purified IgGs. B.S.G. provided SUPPLEMENTARY MATERIALS
convalescent PBMC samples. D.P.M., D.Hu., L.V.T., D.R.B., M.B.,
E7348–E7357 (2017). R.S.B., J.E.V., K.C., J.M.D., J.S.M., A.Z.W., D.W., A.S.H., and science.sciencemag.org/content/369/6504/731/suppl/DC1
19. B. E. Correia et al., Nature 507, 201–206 (2014). L.M.W. analyzed the data. L.M.W. and A.Z.W. wrote the paper, and Materials and Methods
all authors reviewed and edited the paper. Competing interests: Figs. S1 to S13
ACKNOWLEDGMENTS A.Z.W., A.L., E.C., I.B., M.B., S.L., M.Sa., J.H.N., C.J., L.M.W., M.Si., References (20–26)
D.P.M., and S.P. are employees of Adimab, LLC, and may hold MDAR Reproducibility Checklist
We thank E. Krauland and M. Vasquez for helpful comments on the shares in Adimab, LLC. L.M.W. and A.Z.W. are listed as inventors
manuscript. We also thank C. Kivler, C. O’Brien, and E. Platt for on pending patent applications describing the SARS-CoV-2 View/request a protocol for this paper from Bio-protocol.
antibody expression and purification; M. Hagstroem, E. Worts, antibodies. D.R.B. is on the science advisory board of Adimab and
and A. Gearhart for antibody sequencing; and R. Niles and holds shares in Adimab, LLC. Data and materials availability: 14 May 2020; accepted 11 June 2020
K. Canfield for providing mammalian culture support. We Antibody sequences have been deposited in GenBank under Published online 15 June 2020
acknowledge the generous provision of PBMCs from a SARS 10.1126/science.abc7424
survivor provided by I. Gordon, J. Ledgerwood, W. Kong, L. Wang,
K. Corbett, and other members of the National Institute of Allergy

Wec et al., Science 369, 731–736 (2020) 7 August 2020 6 of 6

JEFFERSON SCIENCE FELLOWS PROGRAM ICYS Research Fellow at ICYS, NIMS, Japan online @sciencecareers.org

CALL FOR APPLICATIONS International Center for Young Scientists (ICYS) of National Institute for
Materials Science (NIMS) invites applications for ICYS Research Fellow
Established by the Secretary of State in 2003, Jefferson positions. ICYS will offer you the freedom to conduct independent and
Science Fellows program engages the American academic self-directed research in various area of materials science with full access
science, technology, engineering, and medical communities in to NIMS advanced research facilities.
U.S. foreign policy and international development.
Common language in ICYS is English. Clerical and technical support in
Administered by the National Academies of Sciences, English will be given by the ICYS staffs. An annual salary of approx-
Engineering, and Medicine, this fellowship is open to tenured, imately 5.35 million yen is guaranteed, which may be increased to the
or similarly ranked, faculty from U.S. institutions of higher maximum of ~5.88 million yen depending on the performance of the
learning who are U.S. citizens. After successfully obtaining Research Fellow*. In addition, research grant of 2 million yen per year
a security clearance, selected Fellows spend one year on will be provided to each Research Fellow. The initial contract term is
assignment at the U.S. Department of State or the U.S. Agency two years, which may be extended for another year depending on one’s
for International Development serving as advisers on issues of performance.
foreign policy and international development.
All applicants must have obtained a PhD degree within the last ten years.
The deadline for applications is October 16, 2020 at 5 PM Applicants should submit an application form including research pro-
ET. To learn more about the program and to apply, visit www. posal during the ICYS term, CV Header, CV with a list of publications
nas.edu/jsf. and patents, a list of DOI of journal publications, PDF fles of three
signifcant publications to ICYS Recruitment Desk by September 30,
2020 JST. The format for the application form and CV header can be
downloaded from our website. The selection will be made on the basis
of originality and quality of the research proposal as well as the research
achievements. Please visit our website for more details.

*Approximately 23% of annual salary will be deducted as social insurance
premium, residence tax and income tax.

ICYS Recruitment Desk
National Institute for Materials Science
https://www.nims.go.jp/icys/recruitment/index.html

Confused about DEPARTMENT OF PHARMACOLOGY
your next career move? AND PHYSIOLOGY CHAIR

Download Free Career DEPARTMENT CHAIR
Advice Booklets! Pharmacological and Physiological Science
ScienceCareers.org/booklets
Saint Louis University, a Catholic Jesuit institution dedicated to
education, research, service and health care, has started a national
search for the next William Beaumont Professor and Chair of the
Department of Pharmacological and Physiological Science. The
department encompasses multidisciplinary research particularly
strong in the cardiovascular, endocrine and neuroscience areas and
has an outstanding record in graduate education supported in part by
an NIH-T32 training grant currently in its 30th year. The department
seeks an individual with the vision and scientifc stature to build
and maintain robust basic and translational research programs, and
continue the strong commitment to graduate and medical education.
The successful applicant will have a Ph.D. and/or M.D. degree, a
strong record of academic achievement, a solid level of extramural
research funding and experience in both graduate student and medical
student education. Interested candidates should submit a cover letter,
curriculum vitae and a description of their leadership vision to
Enrico Di Cera, M.D., Chair of the Search Committee (enrico.dicera@
health.slu.edu).

Saint Louis University is an equal opportunity/affrmative action
employer. All qualifed candidates will receive consideration for
the position applied for without regard to race, color, religion, sex,
age, national origin, disability, marital status, sexual orientation,
military/veteran status, gender identity, or other non-merit factors. We
welcome and encourage applications from minorities, women,
protected veterans, and individuals with disabilities (including
disabled veterans). If accommodations are needed for completing
the application and/or with the interviewing process, please contact
Human Resources at 314-977-5847.

WORKING LIFE

By Ruofan Yu

Building bridges

I never thought I would miss the days when visa delays were my primary worry. But recently, my
experience as a Chinese postdoc working in the United States has taken a darker turn. In Febru-
ary, I was forced to cut my family vacation in China short after hearing the U.S. government was
about to issue a ban on travelers from China. Since then, I’ve been stunned how my people have
suffered verbal and physical attacks, how my home country is being used as the name of a virus,
and how my local Chinese consulate—which I’ve visited frequently—was ordered closed by the U.S.
government. Last month, a graduate student I know was grilled on suspicion of espionage for 2 hours
before boarding a plane to China. Imagining the fear he must have experienced, I was left in tears.

I moved from China to the United So I wrote a letter sharing my

States 6 years ago to start my Ph.D. thoughts and anxieties and sent it

program. Having spent time in to my college’s president. I men-

Canada before, I didn’t face much tioned it to other Chinese trainees,

of a language barrier or culture and I was impressed when some

shock. My main source of stress expressed interest in sending a

was the visa approval process. On second, co-signed letter; normally,

my first attempt, I received my visa Chinese trainees are reticent to

in 5 days. But when I went to China voice concerns openly. I echoed

to renew it last year—amid the the idea, and a few days later we

turmoil of the trade war—the same sent a joint letter to school offi-

process took an entire month. All cials, including the president and

things considered, I was quite the head of the graduate program.

lucky, as I know of others who have We were pleasantly surprised

had to wait months—or even years. when they responded promptly

Some of my friends simply avoid and organized virtual meetings

going back home for the entire “I felt better knowing that my to discuss our concerns. We re-
duration of their Ph.D. or postdoc ceived a lot of kind words and

years because they don’t want to understanding, and a few days
later, the institution issued pub-
academic community supports me.”risk leaving the United States and

not being able to get back on time. lic statements in support of all

Both countries have taken a progressively harder stance foreign trainees. The world did not change: The pandemic

toward each other in recent years—and as U.S.-China tensions continued, and U.S.-China tensions did not go away.

worsen, I am left wondering what will come next. In May, the But I felt better knowing that my academic community

U.S. government announced that Chinese STEM trainees who supports me.

work on topics related to “military-civil fusion strategy” will So I want to say this to my fellow foreign trainees: I

no longer be allowed to enter the country. The announcement understand that for a lot of us, it is hard to speak up and

doesn’t affect me directly—I study baking yeast—but it does give voice to our feelings. It is scary to stand out as strangers

make me worried for the future. Will all Chinese STEM train- in a foreign land, with limited rights and no family around.

ees be looked at with suspicion from now on? Nevertheless, it is important to share your concerns and

A few days after the May announcement, I received an ask for help. Send emails and letters. Speak up at town hall

email from my department chair, advising employees who meetings. Open up to colleagues.

are Chinese citizens not to leave the country until more By sharing challenges and engaging in honest dialogue,

was known about the policy. I appreciated the advice, but I we can educate the scientific community about the unique

wanted more empathy and emotion—some acknowledgment issues we face. Such conversations might not move the ILLUSTRATION: ROBERT NEUBECKER

that my university administrators understood what I was needle on foreign policy—but on a personal level, they can

going through. I was not alone: In chat groups and on social help build bridges and tear down walls. j

media, other Chinese trainees were sharing anger, frustra-

tion, and a yearning for support. But people outside our Ruofan Yu is a postdoctoral researcher at the Baylor College of Medicine

community seemed unaware. in Houston, Texas. Send your career story to [email protected].

738 7 AUGUST 2020 • VOL 369 ISSUE 6504 sciencemag.org SCIENCE

Published by AAAS