PHOTO: UNIVERSITY OF MEDICINE AND DENTISTRY OF NEW JERSEY/SCIENCE SOURCE
SCIENCE science.orgBy J oel O. WertheimT
he evolution of virulence—the degree
to which a pathogen sickens, kills, or
otherwise reduces its host’s fitness—
depends on the biology of infection
and transmission ( 1 ). A more virulent
virus may be less transmissible be-
cause in killing its host, it reduces the op-
portunity for transmission. But virulence
and transmissibility can be intrinsically
linked, so that to maintain or increase infec-
tiousness, a virus must be virulent. On page
540 of this issue, Wymant et al. ( 2 ) describe
the emergence of a more virulent and trans-
missible variant of HIV that has spread
to 102 known cases, mostly in
the Netherlands, over the past
decade. This finding raises
questions about the selective
pressures and molecular mech-
anisms that drive increased vir-
ulence and transmission.
To appreciate the nuances in
the evolution of virulence, it is
worth revisiting the apocryphal
tale of how a bioweapon—the
extraordinarily lethal myxoma
virus, which was released in
Australia to exterminate the
European rabbit infestation—
quickly attenuated to become a
benign infection. Myxoma virus
is a vector-borne pathogen that,
when introduced to Australia,
had a >99% fatality rate. Less
virulent myxoma virus evolved,
but it continues to circulate and
kill rabbits in Australia with a 50% mortal-
ity rate. Ample highly virulent myxoma virus
strains persist, indicating that evolution to-
ward attenuation did not drive the most vir-
ulent variants extinct ( 3 ). The key to main-
taining virulence may lie in the inability of
myxoma virus to replicate in its insect vec-
tors; high titers of the virus must be main-
tained in rabbits to ensure transmissibility.
Thus, virulence (which achieves high viral
load) is critical to its evolutionary success.
In HIV, the biology is clear: There can be
no decoupling of virulence and transmis-
sibility ( 4 ). Upon initial HIV infection, the
viral load peaks and then stabilizes to pro-duce a prolonged, symptom-free chronic in-
fection that can last a decade or more. This
stabilized viral load is known as the set-
point viral load and can be quite variable
across infected people, differing by several
orders of magnitude. This variability can
partly be traced back to genetic variation in
the virus itself. People with higher set-point
viral load, left untreated, progress to AIDS
more rapidly ( 5 ). Those same people will
also be more infectious during this time pe-
riod because higher viral load leads to more
viral transmission.
By contrast, people with lower set-point
viral load will live symptom-free longer and
be less infectious over this period. Thereinlies the trade-off from the perspective of
the virus. Burn bright and be brief, or smol-
der and persist? For HIV, natural selection
favors a middle ground, but the trajec-
tory depends on the host population and
the initial starting conditions. In Uganda,
where heterosexual HIV transmission pre-
dominates, there are two major subtypes
of HIV: A and D. Subtype D was the most
prevalent variant in the 1990s and is associ-
ated with higher viral load and more rapid
progression to AIDS than that of subtype A.
Over the past two decades, subtype D has
been gradually outcompeted in Uganda
by the comparatively less virulent subtype
A, which itself may be becoming even less
virulent ( 6 ). The selective driver behind this
change remains unknown.HIV in the United States is evolving in
the opposite direction. Viral load at diagno-
sis has been increasing every decade since
the pandemic was first identified in 1981 ( 7 ),
and higher viral loads are found in people
belonging to larger HIV transmission clus-
ters ( 8 ). Notably, this trend toward more
virulent HIV is not emanating from a single
cluster of transmission. Like the myxoma
virus, this adaptation appears to be found
across the spectrum of HIV genetic diver-
sity in the United States. By contrast, the
virulent cluster described by Wymant et
al. emanated from a single cluster on the
phylogenetic tree, suggesting a single adap-
tive event in HIV evolution. However, the
specific mutations that are re-
sponsible for increased viru-
lence were not identified. In
both Europe and the United
States, the genetic mechanisms
underpinning viral adaptation
remain unclear.
Antiretroviral therapy upon
diagnosis reduces viral load to
a point at which it is undetect-
able. Treatment not only im-
proves survival among those
infected with HIV but also
largely prevents onward sexual
transmission. Evolutionary
modeling suggested that this
intervention—which has been
standard of care in the United
States for a decade—would se-
lect for increasingly virulent
HIV. By decoupling the oppor-
tunity for transmission from
disease progression, HIV with higher viral
load would be favored. However, HIV in the
United States apparently began increasing
in virulence before the test-and-treat era.
Why, after more than a century of human-
to-human transmission, is HIV virulence
still evolving, and how will this virus evolve
in response to efforts to “end the epidemic”
through substantial reduction of transmis-
sion over the next decade?
Observing the emergence of more viru-
lent and transmissible HIV is not a public
health crisis. Standard public health ac-
tion—including molecular HIV surveillance
( 9 ), facilitating linkage to care, and partner
notification—are still the best options when
faced with a rapidly growing cluster of more
virulent HIV. Let us not forget the overreac-Natural selection favors virulence when it is coupled with increased viral transmission
Department of Medicine, University of California, San
Diego, La Jolla, CA, USA. Email: [email protected]Colored transmission electron micrograph showing HIV-1 virions
emerging from T cells (red), continuing the infection cycle. Virulence and
transmission of HIV-1 are influenced by viral load.INFECTIOUS DISEASEWhen viruses become more virulent
4 FEBRUARY 2022 • VOL 375 ISSUE 6580 493