Natarajan H, Crowley AR, Butler SE, Xu S, Weiner JA, Bloch EM, Littlefield K, Wieland-Alter W, Connor RI, Wright PF, Benner SE, Bonny TS, Laeyendecker O, Sullivan D, Shoham S, Quinn TC, Larman HB, Casadevall A, Pekosz A, Redd AD, Tobian AAR, Ackerman ME.
mBio · April 20, 2021 · DOI: 10.1128/mBio.00765-21
How to cite
APA
Natarajan, H., Crowley, A. R., Butler, S. E., Xu, S., Weiner, J. A., Bloch, E. M., Littlefield, K., Wieland-Alter, W., Connor, R. I., Wright, P. F., Benner, S. E., Bonny, T. S., Laeyendecker, O., Sullivan, D., Shoham, S., Quinn, T. C., Larman, H. B., Casadevall, A., Pekosz, A., Redd, A. D., Tobian, A. A. R., & Ackerman, M. E. (2021). Markers of polyfunctional SARS-CoV-2 antibodies in convalescent plasma. mBio, 12(2), e00765-21. https://doi.org/10.1128/mBio.00765-21
Natarajan and colleagues profiled 126 SARS-CoV-2 convalescent plasma donations from a Johns Hopkins (Baltimore/Washington, DC) cohort using a multiplex Fc Array assay that captures both antigen-binding specificity and Fc-receptor engagement, then ran four parallel functional assays — live-virus neutralization, antibody-dependent cellular phagocytosis (ADCP), FcγRIIIa activation as a surrogate for ADCC, and antibody-dependent complement deposition (ADCD). A small panel of biophysical features — chiefly S-2P– and S2-specific IgG and FcγR-binding antibodies — identified the polyfunctional UMAP cluster of donors with discriminatory accuracy superior to standard S1-based clinical ELISAs, and the discrimination held in a 20-donor Dartmouth-Hitchcock validation cohort — reframing convalescent-plasma donor selection from titer to function-predictive.
In this publication:
- 126 convalescent plasma donations profiled head-to-head across four antiviral functions (neutralization, ADCP, FcγRIIIa activation, ADCD) with a Johns Hopkins discovery cohort and a 20-donor Dartmouth-Hitchcock validation cohort.
- Top nine Fc Array features outperformed S1-based clinical ELISAs at picking polyfunctional donors by ROC-AUC, with the discrimination holding in the independent validation cohort.
- UMAP partitioned 126 donors into 4 clusters; cluster 2 (n=27) showed elevated activity across all four functions, defining the operationally interesting “polyfunctional plasma” donor pool.
- ADCP and FcγRIIIa activation correlated tightly (R = 0.82) and moderately with neutralization, while complement deposition was the most independent axis — meaning a single titer measurement under-represents the breadth of antiviral function present in a unit of plasma.
Early in the COVID-19 pandemic, plasma donated by people who had recovered from SARS-CoV-2 infection — convalescent plasma — was deployed at scale as one of the first available antibody therapies. The U.S. FDA issued an emergency authorization after more than 100,000 patients were transfused, and on average outcomes were better when high-titer units were used early in disease. But efficacy data were mixed, and units with lower neutralizing titers were sometimes effective too — which suggested that activities beyond direct virus neutralization were doing meaningful work.
This study set out to measure those other activities head-to-head. The team profiled 126 convalescent plasma donations from a Johns Hopkins (Baltimore/Washington, DC) donor cohort using a multiplex Fc Array assay that captures both the variable region of antibodies (which antigen they recognize) and the constant Fc region (which immune effector functions they can engage). They then ran functional assays measuring four activities: live-virus neutralization, antibody-dependent cellular phagocytosis (ADCP), FcγRIIIa activation as a surrogate for antibody-dependent cellular cytotoxicity (ADCC), and antibody-dependent complement deposition (ADCD). A separate 20-donor cohort from Dartmouth-Hitchcock served as an independent validation.
Three findings stand out. First, the four antiviral activities were correlated but distinct — ADCP and FcγRIIIa activation were tightly correlated with each other and moderately correlated with neutralization, while complement deposition was the most independent — meaning a single titer measurement under-represents the breadth of antiviral function actually present in a unit of plasma. Second, an unsupervised UMAP analysis of the biophysical data revealed four distinct donor clusters; one cluster (cluster 2, n=27 of 126) showed elevated activity across all four functions and represents the operationally interesting “polyfunctional plasma” donors. Third, receiver-operator-characteristic analysis identified a small set of multiplex binding features — chiefly S-2P– and S2-specific IgG and FcγR-binding antibodies — that discriminate polyfunctional from non-polyfunctional plasma with accuracy superior to standard S1-based clinical ELISAs, and the discrimination held in the independent Dartmouth-Hitchcock validation cohort. The translational implication is that a high-throughput multiplex assay can identify donors whose plasma carries broad antiviral function — not just high binding titer — opening a path to function-predictive donor selection for convalescent plasma, hyperimmune globulin, and polyclonal-antibody therapy programs, with downstream relevance for vaccine and therapeutic-antibody programs targeting durable cross-variant protection.
