Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published April 7, 2023 | Supplemental Material + Accepted Version
Journal Article Open

Germline-encoded amino acid–binding motifs drive immunodominant public antibody responses

Abstract

INTRODUCTION. Antibodies are generated by a DNA recombination mechanism occurring in the immunoglobulin heavy and light chain genes in which modular VDJ (for heavy) and VJ (for light) gene segments are combinatorially assembled. The vast complexity of the antibody repertoire allows many species to generate antibodies against virtually any protein. However, when different individuals are exposed to a given pathogen they often mount antibody responses to the same precise protein regions—or epitopes—from the pathogen. The mechanisms underlying these recurrent antibody responses to immunodominant "public epitopes" are not well understood. RATIONALE. We set out to identify a collection of immunodominant public epitopes that would allow us to study mechanisms underlying recurrent antibody responses. We employed VirScan—a phage display platform programmed to display peptides covering the entire human virome—to identify the epitopes of antiviral antibodies from a large cohort of individuals in a high-throughput manner. Additionally, we isolated B cell receptors from different individuals that bound to model public epitopes in order to investigate their determinants of specificity. Finally, we performed a systematic analysis of antibody–antigen structures from the Protein Data Bank (PDB) to search for recurrent modes of antigen recognition. RESULTS. We mapped 376 immunodominant public epitopes from 51 viral species to single-amino-acid resolution. Antibodies from different individuals that recognized the same public epitope often (i) shared light chain isotype (kappa or lambda) and (ii) bound the same precise critical residues in the epitope. Public epitopes showed biased amino acid composition, including a striking enrichment of lysine at the borders of public epitopes recognized by antibodies with lambda light chains. We examined 50 B cell receptors recognizing three model public epitopes in detail and observed conserved V gene segment usage but almost no conservation of heavy chain CDR3 sequences, indicating that key specificity determinants lay within the V gene segments themselves. Structural analysis of antibody–antigen complexes in the PDB uncovered 18 human V gene segments that harbor germline-encoded amino acid–binding (GRAB) motifs that specifically bind to particular amino acids. Among these were a family of six closely related lambda V gene segments with similar GRAB motifs specific for border lysines. We confirmed that the GRAB motifs we identified were critical for antibody recognition of two model public epitopes. Analysis of murine antibody–antigen structures revealed 21 V gene segment–encoded GRAB motifs that only partially overlapped with the human GRAB motifs, which may explain why there is little overlap between the public epitopes recognized across species. Thus, there appears to be a structural basis underlying the notable convergence in humoral immune responses to immunodominant public epitopes across humans and the differing public epitope selection across species. CONCLUSION. Recurrent antibody responses to immunodominant public epitopes are a general feature of humoral immunity. We propose that they are driven by GRAB motifs, a germline-encoded component of the architecture of the antibody repertoire that predisposes antibodies to recognize particular structures and thus influences epitope selection and composition. Public epitopes likely arise in part because they are best aligned for recognition by GRAB motifs and can thus be bound by a relatively large precursor pool of B cells. GRAB motifs may have evolved to ensure efficient antibody responses to pathogens; the recurrent responses they engender across populations likely exert selective pressure on pathogens and influence host–pathogen coevolution.

Additional Information

© 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. This is an article distributed under the terms of the Science Journals Default License. We thank L. Vandenberghe, N. Zabaleta, D. Barouch, and A. Chandrashekar for providing sera samples; S. Quake, D. Grishin, J. Bloom, C. O'Leary, C. Glassman, and M. Steffen for helpful discussions; A. Kohlgruber for designing initial versions of the schematics in Fig. 1B, Fig. 2A, and Fig. 4A; C. Araneo and the Immunology Flow Cytometry Facility and A. Ciulla Hurst and the BioPolymers Genomics Core Facility at Harvard Medical School for supporting this work; J. Q. O'Brien for contributing to revision experiments; and R. E. Shrock for reviewing the manuscript. Access to COVID-19 patient samples was facilitated by the MassCPR. We thank the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from NIH P41-GM103311 for help with molecular graphics. Fig. 8 and the print page summary figure were created with BioRender.com. This research was supported by the SARS-CoV-2 Viral Variants Program and the Value of Vaccine Research Network to S.J.E.; the MassPCR and the NIH 1P01AI165072 to S.J.E., A.G., and D.R.W.; and the NIH 1R01AI129784 to P.J.B. E.L.S. and T.K. were supported by the NSF Graduate Research Fellows Program. R.T.T. is supported by a Pemberton-Trinity Fellowship and a Sir Henry Wellcome Fellowship (201387/Z/16/Z). E.L.M. is supported by a Jane Coffin Childs Postdoctoral Fellowship. R.G. is supported by an NIH Pathway to Independence Award (K99/R00) K99DE031016. D.R.W. is also supported by AI139538, AI169619, AI170715, and AI170580. B.E.G. is supported by a Burroughs Wellcome Career Award in Medical Sciences. S.J.E. is an Investigator with the Howard Hughes Medical Institute. Author contributions: Conceptualization: S.J.E., E.L.S., and T.K. Formal analysis: E.L.S., R.T.T., T.K., A.P.W., I.L., A.A.C., L.G.A.M., C.B., and K. Investigation: E.L.S., T.K., R.G., A.A.C., L.G.A.M., C.B., K., Y.L., and M.L. Methodology: E.L.S., R.T.T., T.K., E.L.M., and A.P.W. Resources: E.F. and F.H. Supervision: S.J.E., P.J.B., B.E.G., A.G., and D.R.W. Visualization: E.L.S., T.K., E.L.M., I.L., A.A.C., R.G., A.P.W., L.G.A.M., C.B., and K. Writing – original draft: E.L.S. and S.J.E. Writing – review and editing: E.L.S., S.J.E., R.T.T., T.K., E.L.M., and P.J.B. Data and materials availability: All data necessary to interpret our findings are available in supplemental tables or on the Harvard Dataverse, digital object identifiers: 10.7910/DVN/AIXWW2; 10.7910/DVN/WZCLMB; 10.7910/DVN/DXWJ2Y. All reasonable requests for materials will be fulfilled. Competing interests: S.J.E. and T.K. are founders of TSCAN Therapeutics and ImmuneID. SJ.E. is a founder of MAZE Therapeutics and Mirimus, and serves on the scientific advisory board of Homology Medicines, TSCAN Therapeutics, MAZE Therapeutics, none of which impact this work. E.L.S. was a consultant for ImmuneID. S.J.E., and T.K. are inventors on a patent application filed by the Brigham and Women's Hospital (US20160320406A) that covers the use of the VirScan library to identify pathogen antibodies in blood.

Attached Files

Accepted Version - nihms-1901999.pdf

Supplemental Material - science.adc9498_mdar_reproducibility_checklist.pdf

Supplemental Material - science.adc9498_sm.v2.pdf

Supplemental Material - science.adc9498_tables_s1_to_s17.zip

Files

nihms-1901999.pdf
Files (123.5 MB)
Name Size Download all
md5:b59ab0a3ce2206627eeefa8e5580c8be
129.7 kB Preview Download
md5:d44c8905f0fb7b67382f8c65199a1fd9
2.1 MB Preview Download
md5:aa5518b474b9193e69a4c24cdd72e1c2
6.7 MB Preview Download
md5:f05148ccac71063ea7eb383007d0124c
114.7 MB Preview Download

Additional details

Created:
August 22, 2023
Modified:
December 22, 2023