Interpreting viral evolution

 

Jesse Bloom

Fred Hutch Cancer Center / HHMI

 

These slides at https://slides.com/jbloom/viral-evol-2025

 

The Faroe Islands

"Measles had not prevailed on the Faroes since 1781, then it broke out early in April 1846."

"Of the 7782 inhabitants, about 6000 were taken with measles."

"Of the many aged people still living in the Faroes who had measles in 1781, not one was attacked the second time."

Panum was describing immune memory, which provides lifelong protection from measles

What was reason for thinking coronavirus immunity would not be affected by viral mutations?

Coronaviruses have lower mutation rate than other RNA viruses

Coronaviruses are only RNA viruses with proofreading activity in their polymerase, and so have ~5- to 10-fold lower mutation rate than influenza virus

The average single-nucleotide mutation to SARS-CoV-2 had occurred >10,000 independent times in human-transmitted SARS-CoV-2 by third year of the pandemic.

But evolution of SARS-CoV-2 (and other RNA viruses) is not mutation-limited

Evolution is not just mutation: also involves selection on phenotypic effects of mutations

 

CoV-229E causes common colds and has been circulating in humans for a long time.

 

 

How do human coronaviruses evolve?

Evolution of CoV-229E spike

We experimentally generated CoV-229E spikes at ~8 year intervals to study in the lab:

- 1984

- 1992

- 2001

- 2008

- 2016

Evolution of CoV-229E spike erodes neutralization by human serum antibodies

Ideally vaccines would elicit evolution-resistant antibodies (like those made by person at right) rather than evolution-sensitive antibodies (like those made by person at left)

Evolution erodes neutralization by sera from different people at different rates

We now know SARS-CoV-2 evolves to erode neutralizing antibodies, just like CoV-229E

neutralization from original COVID-19 vaccine

Original vaccine induced high neutralizing antibody titers against early viral strains

We now know SARS-CoV-2 evolves to erode neutralizing antibodies, just like CoV-229E

newer viral variants

neutralization from original COVID-19 vaccine

New variants (with more antibody escape) replace old ones in SARS-CoV-2 evolution

Main regions where neutralizing antibodies bind

Molecular basis of this antigenic evolution?

Molecular basis of this antigenic evolution?Mutations where antibodies bind

Main regions where neutralizing antibodies bind

Sites of mutations in recent (BA.2.86) SARS-CoV-2 strain relative to early 2020 strain

Can we predict evolutionary success of new viral variants?

SARS-CoV-2's spike binds receptor and mediates viral entry

viral membrane

cell membrane

spike

spike conformational change

Image adapted from here

ACE2

antibody

Image adapted from here

SARS-CoV-2's spike binds receptor and mediates viral entry

Deep mutational scanning to measure effects of mutations 

RBD

fluorescent ACE2

yeast

fluorescent tag on RBD

Yeast display of spike receptor-binding domain (RBD) to measure ACE2 binding

cell sorting

Deep mutational scanning: selections on libraries of mutants, read out by sequencing

Experiments identified N501Y as a mutation that increases ACE2 binding

RBD

fluorescently labeled antibody

yeast

fluorescent tag on RBD

Yeast display to measure how all RBD mutations escape antibody binding

site in RBD

antibody escape

Experiments identified mutations at E484 as causing most antibody escape

484

These anecdotes suggest experiments can identify mutations subsequently selected by evolution

Yeast display only measures binding, which is just part of spike's biological function

cell sorting

Viral infection is more biologically relevant measure of mutation effects

However, we are cognizant of biosafety concerns of mutating actual virus

actual SARS-CoV-2 virion: pathogen capable of spread in humans

pseudotyped lentiviral particle: not a pathogen, cannot spread in humans

actual SARS-CoV-2 virion: pathogen capable of spread in humans

pseudotyped lentiviral particle: not a pathogen, cannot spread in humans

Therefore we use pseudoviruses rather than actual replicative SARS-CoV-2

Pseudoviruses are modified viruses that can only undergo single round of infection

We need to link phenotype (spike mutant) and genotype (sequence encoded by virus)

To link genotype to phenotype, we encode spike in viral genome with barcode

We then create genotype-phenotype linked libraries by two-step process

Result is library of spike mutant pseudoviruses with identifying barcodes

  • Captures full cell-entry function of spike
  • Each mutant identified by short barcode
  • Pseudoviruses safe at biosafety-level 2
  • Broadly applicable to viral entry proteins

Measured how spike mutations affect three molecular phenotypes

Measured how spike mutations affect three molecular phenotypes

Measured how spike mutations affect three molecular phenotypes

Neutralization by soluble ACE2 is proportional to ACE2 binding affinity

Measured how spike mutations affect three molecular phenotypes

Full workflow to measure effects of mutations on antibody neutralization

Deep mutational scanning correlates well with traditional neutralization assays

How do these experimental measurements relate to actual evolution?

Estimating variant fitness (multinomial logistic regression)

Estimating variant fitness (multinomial logistic regression)

We analyze changes in clade growth rather than raw clade growth

change in clade growth

clade growth

Deep mutational scanning measurements correlate with SARS-CoV-2 variant fitness

Combining phenotypes offers best predictions (because mutations can involve tradeoffs)

We can explain ~55% of the variance in growth of different clades, with largest fraction of variance uniquely explained by sera escape.

Can partially predict outcome of real-world evolutionary competition from experiments!

But human population in which SARS-CoV-2 evolves is changing...

Initially, most people were exposed to early strain, imprinting antibody response (     )

First exposed to early strain in original vaccine

First infected by an early strain (pre-Omicron)

Imprinting to early strain continues to shape antibody response after later exposures

But antibody response of infants is imprinted by more recent strain

Imprinted by recent strain

We measured which spike mutations erode neutralization by antibodies from people imprinted with different strains

Adults vaccinated with original strain followed by various exposures

Dadonaite et al (2025); adult sera from Helen Chu's HAARVI study; infant sera from Mary Staat's IMPRINT cohort

6-12 month infants first infected by XBB*

Adult sera have similar escape (mostly in RBD)

site in spike

escape caused by mutations at site

Sites of escape from sera of adults imprinted with original vaccine, then exposed to various infections and vaccinations.

adult 1

adult 2

adult 3

adult 4

adult 5

adult 6

2021

site in spike

escape caused by mutations at site

Sites of escape from infants with only single infection with XBB*

infant 1

infant 2

infant 3

infant 4

infant 5

infant 6

2023

Infant sera have similar escape (mostly in NTD)

There are dramatic differences in sites of escape from adult versus infant sera

escape caused by mutations at site

site in spike

Average of sera from six infants with only a single infection by XBB* in 2023

Average of sera from ten adults imprinted by original vaccine in 2021

How does it affect evolution if viral mutations only affect immunity of some individuals in the population?

To examine role of population immune heterogeneity, I will switch focus to human seasonal influenza

Human influenza evolves similarly to SARS-CoV-2, but has circulated in humans for decades, so imprinting and population immunity) are highly heterogeneous.

 

As described on the next few slides, we developed a new method to characterize immune heterogeneity for influenza by efficiently measuring neutralization of >50 viral strains by ~100 human sera.

Traditional neutralization assays are low-throughput

Assays measure just one virus against one serum in each row (or column) of a 96-well plate

We developed new assay to make 1000s of neutralization measurements in single plate 

Our high-throughput multiplexed neutralization assay uses barcoded virions

Our high-throughput multiplexed neutralization assay uses barcoded virions

Full viral library assayed in each row of plate

Full viral library assayed in each row of plate

Full viral library assayed in each row of plate

Neutralization of all the different viruses is read out simultaneously be sequencing

Neutralization of all the different viruses is read out simultaneously be sequencing

Neutralization of all the different viruses is read out simultaneously be sequencing

Neutralization of all the different viruses is read out simultaneously be sequencing

Neutralization of all the different viruses is read out simultaneously be sequencing

Neutralization of all the different viruses is read out simultaneously be sequencing

Overall we generate triplicate neutralization curves like one below for 50-100 different viral strains in each row of the 96-well plate

different recent H3N2 viral strains

Each row of plate measures titers to many viruses. Below are titers for one child serum.

this child has low titers to these viral strains

Person-to-person variation in per-strain titers

Heterogeneity within and between age groups

How do these heterogeneous neutralization titers for different humans shape viral evolution?

Unpublished data available here

We measured neutralization for a large set of sera in two ways

sera from 95 individuals of different ages

Measured neutralization by each individual serum

Unpublished data available here

sera from 95 individuals of different ages

Measured neutralization by each individual serum

Measured neutralization by pool of all sera

We measured neutralization for a large set of sera in two ways

Growth rates of influenza strains correlate with fraction of sera with titer below cutoff

But viral growth rates do not correlate with titers of a pool of all serum

Population immune heterogeneity is important for viral evolution.

 

What matters is fraction of population susceptible to each viral variant.

Conclusions

Some viruses evolve rapidly to erode antibody immunity.

 

We can use deep mutational scanning to measure how mutations affect the function and antibody neutralization of viral proteins.

 

These measurements help explain success of viral variants in real world.

 

However, imprinting means there is lots of person-to-person variation in how antibodies target viruses, and hence how viral mutations erode immunity.

 

Sequencing-based neutralization assays can measure that heterogeneity at scale, and improve our understanding of how immunity drives viral evolution.

Experimental approaches I have described are applicable to many viral entry proteins

Bloom lab

Thanks

Fred Hutch Cancer Center

Trevor Bedford

 

University of Washington

Helen Chu and HAARVI cohort

Neil King

David Veesler

 

Cincinnati Children's Hospital

Mary Staat

David Haslam

Allie Burrell

 

University of Pennsylvania

Scott Hensley

 

Seattle Children's Hospital

Janet Englund

Tyler Starr (now at Utah)

Allie Greaney (now at UCSF)

Also: Kate Crawford, William Hannon, Caelan Radford

Bernadeta Dadonaite

Rachel Eguia

Caroline Kikawa

Andrea Loes

viral-evol-2025

By Jesse Bloom

viral-evol-2025

Interpreting viral evolution

  • 391