insects to humans and were
originally believed to be
maladaptive consequences
of the host response to
infection (Exton, 1997; Hart,
1988 ). Emerging evidence
now suggests that sickness
behaviors are, in fact,
evolved strategies to favor
host survival during infection
(Ayres and Schneider, 2009;
Hart, 1988; Murray and
Murray, 1979). But how
does it all work? In this
issue of Cell, Wang et al.
(2016)describe how anorexia
shapes the host metabolic
requirements of stress re-
sponses that are critical for
disease tolerance of bacterial
inflammatory states. But
before you choose the trick
over the treat, consider that
this protection comes at
a cost, rendering the host
less able to meet the meta-
bolic demands required for
disease tolerance of viral inflammatory
states.
Theability of sickness-inducedanorexia
to promote survival of infections is
dependent on how the fasted state
influences host defense strategies. The
defense response in animals can be
broken down into two distinct compo-
nents called resistance and tolerance
(Medzhitov et al., 2012; Schneider and
Ayres, 2008). Resistance involves the
execution of microbial killing mecha-
nisms through the induction of the im-
mune response, whereas disease toler-
ance enables the host to endure the
physiological damage that occurs during
infection, independent of microbial killing
mechanisms. As tolerance promotes
health of the host while having a neutral
to positive impact in microbial fitness,
this mode of defense is predicted to
lead to cooperative co-evolution between
host and microbial populations (Ayres,
2016 ). An important function of disease
tolerance is to regulate metabolic per-
turbations that occur during infections
(Dionne et al., 2006; Schieber et al.,
2015 ), and it is likely that the effects
of anorexia on metabolism will shape
tolerance defenses. Indeed, in a fruit fly
model, anorexia and diet restriction pro-
moted tolerance of a systemicSalmonella
infection, but the mechanisms by which
the fasted state can influence tolerance
remain unknown (Ayres and Schneider,
2009 ).
In the 1970s it was demonstrated
that interfering with anorexia induced
by the intracellular bacterium Listeria
monocytogenesusing force feeding led
to increased mortality in mice compared
to those that were allowed to develop the
anorexic response during infection (Mur-
ray and Murray, 1979). Following up on
these early studies, Wang et al. demon-
strate that oral administration of glucose
increased mortality in mice systemically
infected with Listeria, whereas admin-
istration of 2-deoxy-D-glucose (2-DG),
which competes with glucose utilization,
promoted survival of the mice. Thus
glucose is the food component that is
necessary and sufficient to mediate
lethality in a mouse model ofListeriainfec-
tion, when anorexia is blocked by caloric
supplementation. Because the effects
on lethality caused by administration of
glucose or 2-DG were associated with
changes inListeriainfection levels, the
authors turned to a Lipopolysaccharide
(LPS) endotoxemia model in which mor-
tality results from the systemic inflam-matory response, taking the
variable of pathogen growth
out of the equation. Force
feeding of LPS-injected mice
increased mortality. This
effect was dependent on
glucose but was largely inde-
pendent of the magnitude of
LPS-induced inflammation.
Bycontrast, the authors found
that glucose availability and
utilization are critical for sur-
viving influenza infection or
poly(I:C)-induced inflamma-
tion. Similar to LPS, these
effects were largely inde-
pendent of viral titers or the
magnitude of inflammation,
revealing that opposing meta-
bolic requirements are critical
fortissuetoleranceofdifferent
inflammatory states.
It is well established that
animals suffering from endo-
toxic shock exhibit clinical
and pathological neurological
issues, including seizure and
neuronal apoptosis. Consistent with this,
Wang et al. found that the protection
against LPS-induced mortality afforded
by 2-DG to mice was associated with
decreased occurrence of shrunken neu-
rons in the brain, while in mice suffering
from viral inflammation, 2-DG treatment
was associated with decreased heart
rate, respiratory rate, and body tempera-
ture that are suggestive of dysregulation
of central autonomic control. Collectively,
these observations led the authors to
explore the possibility that the differential
effects on glucose metabolism are asso-
ciated with neuronal dysfunction that
leads to death. In the case of viral
inflammation, they found that lethality is
dependent on type I IFN signaling, pre-
sumably in the brain. They propose that
glucose utilization is protective against
ER stress downstream of type I IFN
signaling and CHOP-mediated cellular
dysfunction, which ultimately prevents
neuronal dysfunction, dysregulation of
central autonomic control, and lethality.
In the case of bacterial inflammation,
glucose utilization inhibited ketogenesis,
leading to impaired tolerance of ROS-
mediated brain damage and death. Thus,
glucose availability and utilization sup-
ports a tolerance mechanism during viralFigure 1. Opposing Metabolic Requirements of Neurons Promote
Disease Tolerance during Bacterial and Viral Inflammatory States
(Left) During bacterial-induced inflammation, inhibition of glucose utilization
promotes ketogenesis, leading to increased tolerance of ROS-mediated
damage in neurons, protecting from lethality. (Right) During viral-induced
inflammation, neurons utilize glucose to inhibit ER stress and CHOP-
mediated cellular dysfunction, thereby preventing the dysregulation of
central autonomic control and lethality.Cell 166 , September 8, 2016 1369