Can mapping viral antibody escape mutations inform vaccines?
Jesse Bloom
Fred Hutch Cancer Center / HHMI
These slides at https://slides.com/jbloom/escape-calc-for-vax
Recap of basic principles of human coronavirus evolution
SARS-CoV-2 imprinting and antibody responses
Possible future evolutionary steps in XBB
Predictive cocktail vaccines?
Recap of basic principles of human coronavirus evolution
SARS-CoV-2 imprinting and antibody responses
Possible future evolutionary steps in XBB
Predictive cocktail vaccines?
Other human CoVs cause annual waves
- Seasonal waves starting late Oct / early Nov, peaking Jan to mid Feb
- Often only one each of the alpha- and beta-coronaviruses cause large wave
Human CoV spikes evolve to escape neutralizing antibodies
Human sera from 1985-1990 neutralizes contemporaneous CoV-229E spikes quite well, but as virus evolves into the "future", neutralization is eroded. There is substantial heterogeneity in the rate of erosion across sera from different human individuals.
Antibody selection concentrated in RBD
Sites of evolutionary change in spike of CoV-229E over last four decades
Sites of mutations in SARS-CoV-2 Omicron BQ.1.1 spike relative to Wuhan-Hu-1
Recap of basic principles of human coronavirus evolution
SARS-CoV-2 imprinting and antibody responses
Possible future evolutionary steps in XBB
Predictive cocktail vaccines?
Imprinting: even after Omicron infection, neutralization highest against early strain
Imprinting: Omicron infection mostly boosts pre-existing B-cells to old strain
Omicron specific antibodies
Wu cross-reactive antibodies
But in Chinese cohort, two Omicron infections mostly breaks imprinting
Omicron specific antibodies
Wu cross-reactive antibodies
Immunological imprinting and CoV-229E?
People imprinted with first CoV-229E that infects them, but still generally neutralize later strains to which they are exposed. Blue shading shows timeframe each individual alive prior to serum collection.
Recap of basic principles of human coronavirus evolution
SARS-CoV-2 imprinting and antibody responses
Possible future evolutionary steps in XBB
Predictive cocktail vaccines?
Dominant variant currently is XBB
Current vaccine recommendation is XBB.1.5. But by time vaccine is in use, there will almost certainly be more mutations. Can we forecast what they might be?
Deep mutational scanning to map RBD mutations that escape antibody binding
RBD
fluorescently labeled antibody
yeast
fluorescent tag on RBD
Experiments combine flow cytometry and deep sequencing of a library of yeast expressing all RBD mutants
These measurements can be made on ~10,000 to ~100,000 mutants all at once
Escape map from one antibody (LY-CoV555, ie bamlanivimab)
484
452
490
This turned out be an unfortunate choice of an antibody for Eli Lilly to develop as drug, as mutations at sites 484 and 452 were prevalent by early to mid 2021
Infection / vaccination elicit polyclonal antibodies
From Yunlong Cao, we have deep mutational scanning data on ~1000s antibodies that neutralize XBB
Text
See here for summary of how successful escape calculator has been so far in pinpointing sites of evolution.
Sites of escape in XBB
Interactive escape calculator: https://jbloomlab.github.io/SARS2-RBD-escape-calc/
ACE2 affinity from: https://jbloomlab.github.io/SARS-CoV-2-RBD_DMS_Omicron/RBD-heatmaps/
Fitness effects from: https://jbloomlab.github.io/SARS2-mut-fitness/S.html
Interactive escape calculator: https://jbloomlab.github.io/SARS2-RBD-escape-calc/
Clear differences between one and two Omicron infections: eg, site 498 only major site of escape after two Omicron infections
We will probably see viral evolution at many of these sites of escape
See here for summary of how deep mutational scanning has predicted many of sites of antigenic evolution in the SARS-CoV-2 RBD over last few years
Recap of basic principles of human coronavirus evolution
SARS-CoV-2 imprinting and antibody responses
Possible future evolutionary steps in XBB
Predictive cocktail vaccines?
What if we make vaccine that is just mix of a bunch of these predicted future mutants?
Is this a brilliant idea, or a dumb idea? It's hard to tell...
Design of a "future mutant" cocktail vaccine
Design cocktail using the following principle:
- Pick RBD mutations that maximize the product of:
- antibody escape as quantified by escape calculator
- ACE2 affinity and RBD expression in yeast display deep mutational scanning
- fitness effects estimated from all human SARS-CoV-2 and just recent XBB clade
- Design cocktail to favor inclusion of new mutations in each component
Current cocktail vaccine
Current designs are available at https://github.com/jbloomlab/SARS2-cocktail-vax (if you plan to use designs, let me know and I will update)
Current components, not validated for folding are below. For a N-component cocktail, take the first N entries off the list:
- XBB.1.16
- XBB.1.16_F456L_K440R_K356T_K444T
- XBB.1.16_F456L_P445A_S490V_T346K
- XBB.1.16_K440R_Y453F_F456L_T346S
- XBB.1.16_K444R_K440Y_Q493L_H505Y
- XBB.1.16_L452M_S494P_P445A_L455S
- XBB.1.16_L452R_K440R_A484K_N405G
- XBB.1.16_S490L_K444M_K460Y_E340D
- XBB.1.16_A484S_K444T_Y453F_K356R
- XBB.1.16_A484Q_K444R_K460T_E340Q
Bloom lab
Bernadeta Dadonaite
Kate Crawford
Caelan Radford
Tyler Starr
Allie Greaney
Rachel Eguia
rest of lab
Peking University
Yunlong Cao
Fanchong Jian
Penn
Scott Hensley and Grace Li
Drew Weissman
Thanks
