The Economist November 20th 2021 Science & technology 83ceptors as it does to those in human be
ings. These would be candidates for the
role of reservoirs.
To this end, they gathered molecular in
formation about every version of ace2that
they could get their hands on. Mostly,
these were from mammals—142 species of
them. They then used computer modelling
of the interatomic forces involved to work
out the strengths of the bonds likely to de
velop between sarscov2’s spike protein
and each version of ace2. As they expected,
based on news that broke while they were
conducting their work, the bond with
mink ace2was particularly strong. They
found a similarly strong affinity with ace2
from whitetailed deer, long before reports
of infections in that species emerged. Cats
and dogs also showed up as being at risk—
which reports then confirmed that they
were. And gorillas and macaques, which
have suffered a few cases in zoos, looked
susceptible as well.
Useful as this information was, Dr Han
was keen to look beyond these 142 species.
To delve deeper, the team built a database
of evolutionary traits shared by the species
with the most vulnerable ace2receptors.
This is a technique which has been used
successfully in the past on rodents and
bats, to assess their likelihood of acting as
reservoirs for viruses including Ebola and
Zika. It is based on the idea that particular
proteins of species with similar physiolo
gies and ways of life might be expected to
evolve in similar ways.
The researchers studied everything
they could—from breadth of diet, metabol
ic rate and age of sexual maturity to litter
size, lifespan, geographical range and phy
logenetic relationships—about more than
5,000 mammals for which little or no
ace2receptor information was available.
This enormous database completed, they
fed the outcome into a machinelearning
system that had been trained on the char
acteristics of the 142 species they had alrea
dy examined. The result was the revelation
of 540 species which seemed likely to have
vulnerable ace2receptors and thus the po
tential to function as covid reservoirs.
Most primates were on this list—which,
considering that people are primates too,
was expected. Nor, given suspicions about
sarscov2’s origins, was the inclusion of
35 types of bat a surprise. Surprises, how
ever, there were. Though the common
house mouse does not look to be a risk,
which is good news, two of its fellow ro
dents, the ricefield rat and the Malayan
field rat, both do. Since these species are
often preyed on by domestic cats, them
selves now known to be covidsusceptible,
that provides a route by which people
might become infected.
Dozens of other species were also
flagged up as potential reservoirs. These
included red foxes and raccoon dogs—two
creatures which, like mink, are sometimes
farmed for fur—and whitelipped pecca
ries (piglike creatures found in South and
Central America) and nilgai (a large Asian
antelope), both of which are farmed occa
sionally, and also hunted and eaten.
Among more widespread livestock, the
species of most concern is the water buffa
lo. There are reckoned to be over 200m of
these around the world, acting as both
beasts of burden and sources of milk. And
other frequently hunted animals, such as
the duiker (another antelope), the warty
pig and the mule deer were also reckoned
vulnerable, together with some rarities, in
cluding two critically endangered ante
lopes, the addax and the scimitarhorned
oryx (pictured below), which was once ex
tinct in the wild and is only now being re
introduced. In these cases the threat is less
to human beings than to the survival of the
species concerned.Paying the buffalo bill
The sheer range of species involved stag
gers Dr Han. “I never imagined that we
would ever see a virus with such a high
crossspecies infection potential,” she
says. “It appears that there are at least an
order of magnitude more species that are
susceptible to sarscov2 infection than
any other zoonotic virus I can think of.”
Forewarned, however, is forearmed.
And here there is perhaps a lesson on keep
ing the weapons in the arsenal sharp. One
reason Dr Han’s study took so long from in
ception to publication is the disparate na
ture of the sources she needed to draw on.
Scattered as they were around the world’s
naturalhistory collections, assembling
them took time. Many museums are now
in the business of making their collections
available electronically. To some, that
might sound alowpriority. Work like this
suggests it is not. nCovid’s next victims?AutismWhen theories go
belly up
T
he gutmicrobiome’s effect on health
is so big that it is sometimes referred to
as a forgotten organ. The trillions of crit
ters in the lower alimentary canal have an
influence that seems to reach all over the
body—helping regulate weight, for exam
ple. In recent years, though, one idea in
particular has intrigued researchers. This
is that certain mixtures of gut microbes
may help cause autism.
It is well known that children with au
tism suffer gastroenteric problems. They
often also seem to have simpler gut floras
than their neurotypical peers, and some
observations report the presence of unusu
al species. This has led to the suggestion
that these abnormal microecosystems
might actually be triggering autism. So
tempting is this thought that preliminary
experiments intended to tweak the micro
biomes of autistic children have been car
ried out, with apparently positive results.
Unfortunately, the theory that gut mi
crobes can cause autism is built on shaky
foundations. The best evidence comes
from work on mice, which is problematic
because identifying murine behaviour pat
terns that correspond to human autism is
as much an art as a science. Work on peo
ple, meanwhile, has been criticised for re
lying on samples too small to yield statisti
cally definitive conclusions.
But no longer. A study published in Cell
by Jacob Gratten of the University of
Queensland, Australia, and his colleagues
is reckoned to have enough statistical pow
er to answer the question robustly. And
that answer is “no”, gut microbes do not
cause autism. Dr Gratten’s work indicates
that, while there is indeed a correlation to
be explained, it is actually autism which
(albeit indirectly) affects the gut flora, not
the other way around.
Dr Gratten and his colleagues embarked
on their project in 2016, using stool sam
ples and dietary information collected as
part of two large studies, the Australian Au
tism Biobank and the Queensland Twin
Adolescent Brain Project. Of 247 children
they selected for inclusion, 99 had been di
agnosed with autism, 51 were neurotypical
siblings of some of these, and the remain
der were unrelated neurotypical children.
There was, they found, a clear connec
tion between individual children’s dietary
diversity and the diversity of their gut mi
crobes—with more adventurous diets be
ing associated with richer microbiomes.An unbalanced gut microbiome seems
not to be a cause of autism, after all