Science - USA (2022-06-10)

SCIENCE science.org

By Chadi M. Saad-Roy^1 , C. Jessica E.

Metcalf2,3, Bryan T. Grenfell2,3

D

espite much recent progress in mod-

eling the epidemiology and evolution

of acute viruses, a full quantitative

synthesis of viral eco-evolutionary

dynamics remains elusive. The severe

acute respiratory syndrome coronavi-

rus 2 (SARS-CoV-2) pandemic has stimulated

vast research efforts into measuring viral dy-

namics and genetics, across scales from indi-

vidual hosts to global circulation. In parallel,

understanding determinants of individual

immune protection against infection and se-

vere disease has been a major research focus.

However, the interaction between population

immunity and viral dynamics has been much

less studied—a crucial gap because this in-

teraction will strongly influence SARS-CoV-2

evolution over time. Clarifying the biology of

transmission at different scales, and in par-

ticular the impact of immunity on transmis-

sion, will define the epidemiological context

(current spread and population risk), as well

as the epidemiological and evolutionary im-

plications, of immune escape.

To probe likely evolutionary trajectories,

it is necessary to understand how immu-

nity intersects with transmission and thus

population outcomes, especially in partially

immune individuals. Building such a “phy-

lodynamic” synthesis ( 1 ) requires a revolu-

tion in cross-scale understanding of how

individual immune kinetics translate to im-

muno-epidemiology. The resulting frame-

works will provide a more mechanistic ba-

sis for exploring the predictability of viral

evolutionary dynamics. Such mechanistic

approaches synthesize findings across dis-

ciplines, and they will provide a principled

way to integrate information across scales

and inform data analyses.

Phylodynamic models meld the epide-

miological and evolutionary dynamics of

pathogens, often with underlying host im-

mune kinetics. Such frameworks have been

applied across a range of acute and chronic

pathogens ( 1 , 2 ), and often focus on influ-

enza eco-evolutionary dynamics as a model

for acute partially immunizing infections.

In particular, the immune escape of sea-

sonal variants of human influenza A virus

has been used to study the impact of popu-

lation immunity on viral population dy-

namics. The simplest conceptual qualitative

phylodynamic models for immune escape

( 1 ) posited that spread of escape variants

would be maximized at intermediate im-

mune pressure, through a trade-off between

transmission and immune selection. That

is, if there is no immune pressure, then vi-

ral abundance may be high but selection

for immune escape is absent. Conversely,

if there is strong immune pressure, then

selection may be high but viral abundance

very low; this would also limit adaptive evo-

lution for immune escape.

More detailed dynamic models for in-

fluenza have addressed quantitative phy-

lodynamic interactions across scales (from

within hosts to global spread). These models

also explored what limits viral diversity to

the observed phylogeny that emerges from

antigenic drift (changes in surface proteins

that lead to immune escape), given the huge

variation generated by error-prone viral re-

production ( 3 , 4 ). An array of related fram-

ings of immunodynamics in various guises

[such as strain-transcending immunity ( 3 )

and repeated selective sweeps driven by

herd immunity ( 4 )] have been explored as

candidates to explain these phylodynamic

patterns for influenza A virus.

Phylodynamic models have been widely

applied beyond seasonal influenza ( 2 ). In

particular, the dynamics of seasonal hu-

man coronaviruses (HCoVs) are becoming

increasingly salient, especially owing to

the continued spread of SARS-CoV-2. For

example, the 229E HCoV exhibits antigenic

drift ( 5 ), which could have implications for

the emergence of SARS-CoV-2 variants.

Indeed, the ongoing COVID-19 pandemic

underlines the major impact not only of

immune escape evolution, but also of rapid

selection for (surprisingly large) increases

in viral transmission rate. Although the

acquisition of human-to-human transmis-

sion by influenza viruses that caused pan-

demics is arguably an example of a selec-

tion on transmission, this is not as evident

for seasonal influenza.

A key issue for next-generation phylo-

dynamics is to determine how selection at

different levels (within-host, transmission

chains, population-level, globally) trans-

lates into population  outcomes, and how

it is modulated by host immunity. There

are important processes present at each

biological scale, from the emergence of vari-

ants to their global spread (supplementary

fig. S1). Within individual hosts , variants

arise through mutation and/or recombina-

tion ( 1 ). The ability of variants to replicate

(which is necessary for successful transmis-

sion) will be affected by their cellular tro-

pism and by the efficiency with which they

can enter cells and transmit across tissues.

For example, increases in angiotensin-con-

verting enzyme 2 (ACE2) avidity may lead

to enhanced SARS-CoV-2 transmissibility

( 6 ) and may also alter tissue tropism, e.g.,

for the upper respiratory tract ( 7 ). Tropism

could also depend on host immune re-

sponses; adaptive immunity (acquired

through infection or vaccination) may also

shape viral load trajectories [e.g., for influ-

enza ( 8 )] and lead to selection for immune-

escape variants [e.g., the Omicron variant

( 9 )]. Tropism and adaptive immunity also

likely play a role in the clinical severity of

infections (and reinfections) with variants.

The combination of individual immune

phenotypes and their impact on viral shed-

ding affects the transmission of variants

and determines whether they can cause

breakthrough infections in vaccinated in-

dividuals. Simultaneously, antigenic drift or

waning immunity may influence suscepti-

bility to (re)infection. In turn, these lead to

fitness advantages for variants with either

increased immune escape or transmissibil-

ity. Individual characteristics of immunity

might be particularly important in shaping

the features of the virus. For example, selec-

tion of influenza viruses within immunolog-

ically competent hosts is less strong owing

to asynchrony between viral growth and im-

mune response ( 8 ), whereas prolonged car-

riage in immunocompromised hosts could

result in variants ( 10 ); this may also be the

case for SARS-CoV-2 ( 11 ).

At the population level, the presence of

many immune individuals could limit vi-

ral spread through indirect protection. To

prevent the establishment of variants with

increased transmissibility (but with little

immune escape) once they have emerged,

a higher proportion of immune individu-

als are required than that needed to limit

transmission of the original virus. Indirect

VIEWPOINT: COVID-19

Immuno-epidemiology and the

predictability of viral evolution

Understanding viral evolution depends on a synthesis of

evolutionary biology and immuno-epidemiology

(^1) Lewis-Sigler Institute for Integrative Genomics, Princeton
University, Princeton, NJ, USA.^2 Department of Ecology and
Evolutionary Biology, Princeton University, Princeton, NJ,
USA.^3 Princeton School of Public and International Affairs,
Princeton University, Princeton, NJ, USA.
Email: [email protected]; [email protected]
10 JUNE 2022 • VOL 376 ISSUE 6598 1161