REVIEWS

Role of the gut microbiota in host

appetite control: bacterial growth
to animal feeding behaviour
Sergueï O. Fetissov
Abstract | The life of all animals is dominated by alternating feelings of hunger and satiety — the
main involuntary motivations for feeding-related behaviour. Gut bacteria depend fully on their
host for providing the nutrients necessary for their growth. The intrinsic ability of bacteria to
regulate their growth and to maintain their population within the gut suggests that gut bacteria
can interfere with molecular pathways controlling energy balance in the host. The current model
of appetite control is based mainly on gut–brain signalling and the animal’s own needs to maintain
energy homeostasis; an alternative model might also involve bacteria–host communications.
Several bacterial components and metabolites have been shown to stimulate intestinal satiety
pathways; at the same time, their production depends on bacterial growth cycles. This short-term
bacterial growth-linked modulation of intestinal satiety can be coupled with long-term regulation
of appetite, controlled by the neuropeptidergic circuitry in the hypothalamus. Indeed, several
bacterial products are detected in the systemic circulation, which might act directly on
hypothalamic neurons. This Review analyses the data relevant to possible involvement of the gut
bacteria in the regulation of host appetite and proposes an integrative homeostatic model of
appetite control that includes energy needs of both the host and its gut bacteria.

Nutrition, Gut & Brain
Laboratory, Inserm UMR
1073, University of Rouen
Normandy, 22 Boulevard
Gambetta, 76183 Rouen,
France.
serguei.fetissov@univ-rouen.
fr
doi:10.1038/nrendo.2016.150
Published online 12 Sep 2016
Corrected online 18 Nov 2016
Feelings of hunger and satiety dominate the life of all
animals and are the main involuntary motivations for
feeding-related behaviour. Both feelings are of visceral
origin; that is, they involve the gastrointestinal tract
secreting molecules that reach the nervous system. The
brain integrates peripheral hunger-related and satiety-
related signals to generate motivated behaviour neces-
sary for obtaining, ingesting and digesting food. The
gut of all animals (except experimental germ-free ani-
mals) contains numerous bacteria, which are dependent
on host feeding behaviour for receiving the nutrients
necessary for maintenance of their population. The
gut, therefore, represents a stable ecological niche for
its inhabiting bacteria, which rely on host physiology to
maintain their basic biological processes such as feeding
and reproduction1–3. The complex interactions that
occur between gut bacteria and their host have been
conceptualized as symbiotic4–6. Indeed, bacterial contri-
butions to host physiology vary from digestion of die-
tary fibre and vitamin production to proper functioning
of the immune system, and even supplying ribosomal
machinery for protein synthesis7–9. Accumulating data
on the effects of the gut microbiota on brain function
and behaviour represents another intrinsic part of such
interactions10. However, the role of gut bacteria in the
regulation of host appetite has, until recently, received
little attention.
Understanding how the body generates and reg-
ulates appetite is of fundamental importance for the
treatment of chronic pathological conditions such as
obesity, anorexia nervosa and cachexia, in which the
normal regulation is lost. The current obesity epidemic
and the growing numbers of individuals with eating
disorders reflect the lack of efficient treatments and
highlight our insufficient understanding of the under-
lying causes of altered appetite11,12. Conversely, several
studies have revealed dysbioses of the gut microbiota
in obesity and anorexia nervosa13,14. Such associations
might reflect not only microbiota modifications sec-
ondary to malnutrition, but suggest a causal role of
gut bacteria in altered regulation of energy metabo-
lism and possibly feeding behaviour 15,16. Indeed, a 2016
study revealed that gut bacterial proteins can influence
host control of appetite dependent on the bacterial
growth phase17.
This Review discusses the evolving conceptual model
of appetite control supported by recent data, presents
the symbiotic arrangement viewed from both sides (the

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 11

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Key points nucleus (PBN), as well as by the central nucleus of amyg-

• Animals feed their gut bacteria, which are fully dependent on their host for providing
the nutrients necessary for growth and maintenance of the bacterial population
• Food intake is followed by cephalic reflex-mediated secretions of nutrients into the
gut, which activate the gut–brain satiety pathways via release of intestinal hormones
• Regular nutrient provision to cultured bacteria or nutrient infusion into the colon
stimulates immediate bacterial growth that lasts 20 min
• The dynamics of the regular nutrient-induced growth of bacteria are similar to the
dynamics of meal-induced intestinal satiety hormone (for example, PYY) release
• Bacterial molecules and metabolites, whose production depends on bacterial growth
phases, regulate intestinal release of satiety hormones
• Systemic bacterial molecules directly activate central appetite pathways that might
integrate the energy status of both the host and its gut microbiota
gut bacteria and the host) and describes their interac-
tions with molecular pathways that regulate appetite. A
new integrative model of appetite control based on the
homeo­static needs of both bacteria and the host is pro-
posed, which connects bacterial growth cycles with the
cycles of hunger and satiety in the host. The role of water
intake, another motivated ingestive behaviour, mechanis-
tically linked to food intake, is discussed elsewhere18. As
this Review concerns several fields of biological science,
it is not exhaustive in each individual topic but, rather,
aims to provide a transdisciplinary global view to con-
ceptually advance our understanding of the mechanisms
underlying appetite control.
Homeostatic model of appetite control
Host energy homeostasis and the brain
To place the emerging role of gut bacteria in host appetite
control in context, the current view of appetite control is
briefly reviewed, which provides a basis for understand-
ing bacteria–host molecular interactions. The homeo-
static model of appetite control was constructed on the
basis of pharmacological, histological and molecular
genetic studies19. This model is experimentally repro-
ducible, ascribing a specific role of relevant genes to the
regulation of appetite. These genes, encoding mainly pep-
tide messengers and their receptors, as well as proteins
necessary for neurotransmitter synthesis and signalling,
are expressed by cells connected via neuroendocrine and
neuronal pathways20. For simplicity, these pathways are
usually described separately as ‘hunger’ and ‘satiety’ or
‘appetite-suppressing’ pathways21,22, although each path-
way can be involved in the regulation of both hunger and
satiety. In the brain, the appetite pathways are organized
around the circumventricular organs such as the hypo-
thalamic arcuate nucleus (ARC), which is directly acces-
sible to circulating factors23,24, and the afferent centres of
the cranial nerves, such as the nucleus of the solitary tract
(NTS), which receives vagal input 25. Both ARC and NTS
neurons project to other hypothalamic, brainstem and
forebrain regions, forming a complex neuronal network
for autonomic control of appetite that is influenced by
cognitive factors26,27. Among these regions, important
roles have been ascribed to the hypothalamic paraven-
tricular (PVN) and ventromedial (VMN) nuclei, the
lateral hypothalamic area (LHA) and the parabrachial
dala (CeA)28. As discussed later, activation of these nuclei
by bacterial components suggests their involvement in a
central mechanisms of appetite control.
Although some building blocks and connections in
this ‘appetite network’ remain to be clarified, the key role
of two neuronal populations of the ARC has been well
established29–31. One population consists of the GABA-
ergic neurons, which express orexigenic neuropeptide
tyrosine (NPY) and agouti-related protein (AgRP)32–35.
The other population consists of a neighbouring group of
glutamatergic neurons expressing pro-opiomelanocortin
(POMC), a precursor of α-melanocyte-stimulating
hormone (αMSH), an anorexigenic neuropeptide36–38.
Activation of POMC neurons, followed by αMSH release
and binding to the melanocortin 4 receptor (MC4R),
leads to activation of the melanocortin satiety pathway 39.
These NPY, AgRP and POMC populations are often
viewed as the ‘first order’ neurons in the hunger and
satiety brain pathways19. These populations are anato­
mically and functionally interconnected locally in the
ARC as well as in their downstream projections sites40,
where they integrate appetite-related signals mainly
from humoral pathways.
A key neuronal signalling pathway in appetite con-
trol consists of cholinergic sensory vagal neurons that
release glutamate to excite NTS neurons, which, in
turn, relay gut-derived satietogenic signals to other
brain regions such as the PVN and CeA 41. Indeed,
NTS neurons express several transmitters and peptides
with anorexigenic properties, including catechola-
mines, glucagon-like peptide 1 (GLP1) and αMSH42–44.
Notably, ARC and NTS POMC neurons have reciprocal
connections, which demonstrates the redundancy of
humoral and neural pathways in satiety signalling 45,46.
Furthermore, both NTS and ARC neurons converge on
the PBN, which, in turn, sends anorexigenic calcitonin
gene-­related peptide (CGRP) projections to the CeA21.
Long-term control of appetite involves the hypo-
thalamic peptidergic network that originates in the
same ARC neurons involved in the hunger and satiety
pathways47. Activity of this network depends on the
long-term pattern of peptide hormone release from
the gastrointestinal tract and energy stores. Leptin,
which is produced in adipose tissue proportional to
body weight, is a major anorexigenic long-term signal
that promotes negative energy balance and activates
ARC POMC neurons both directly and indirectly 29,48.
The long-term activity of this peptidergic network is reg-
ulated by the gene expression levels of anorexigenic and
orexigenic neuropeptides and their receptors49. These
genomic and/or transcriptomic effects coupled with the
neuronal plasticity that forms synaptic and non-synaptic
communications within this network50 constitute the
molecular basis of ponderostatic and ponderodynamic
mechanisms; that is, for maintaining the body weight
set-point as well as its adjustment to the prevailing
nutritional environment, respectively.
According to the homeostatic model, the molec-
ular pathways driving appetite should be tuned to the
balance of energy intake and expenditure in the host.

12 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Inclination of this balance towards low food intake and
high energy expenditure must trigger appetite. The
model assumes that the energy balance controlling
hypothalamic centres must be able to measure availa-
ble energy on both short-term and long-term scales, in
order to meet immediate physiological energy needs and
to maintain the body weight set-point, respectively. This
homeostatic model does not, however, take into account
the energy needs of gut bacteria that are distinct from
those of the host and, hence, might independently affect
host control of appetite.
Homeostatic versus hedonic control
Whereas the homeostatic model explains increased
appetite and food intake as triggered by energy short-
age, the feeding-associated feeling of pleasure, involving
activation of the brain reward system, might underlie
the hedonic reason for eating. Although reward percep-
tion is a normal part of all types of motivated behav-
iour, including food intake, an abnormal hedonic drive
for eating can override homeostatic signals and might
be present in bulimia nervosa51 or potentially drive
increased intake of palatable energy-rich foods leading
to obesity 52,53. The key molecular pathways leading to
feeding reward involve the mesolimbic dopamine system
that originates in the ventral tegmental area (VTA). The
homeostatic appetite signals interact with the dopamine
system both at its source and in target areas of the brain,
including the nucleus accumbens and the hypothalamus;
the latter is involved in both homeostatic and hedonic
eating responses54–58. As the VTA dopaminergic neurons
are protected by the blood–brain barrier, the possible
effects of gut bacteria-derived circulating molecules
on this dopamine system should by relayed via other
neurons, such as those in the PBN59.
Intestinal signalling to the brain
Both the homeostatic and the hedonic brain systems
controlling appetite are activated by hormones derived
from tissues and organs that signal various metabolic
processes, including energy storage and nutritional
status60. The latter mainly involves mucosal entero­
endocrine cells in the gastrointestinal tract secreting
several peptide hormones into the systemic circu-
lation61–63. Ghrelin, the only known orexigenic hor-
mone so far, is produced by the stomach64. Among
the anorexigenic hormones, produced mainly in the
small and large intestine, are cholecystokinin, which
contributes to satiation, and peptide tyrosine tyrosine
(PYY), which induces postprandial satiety 65–67. GLP1,
a product of the proglucagon gene expressed in
enteroendocrine cells, acts as a satietogenic hormone
and further acts as an incretin, lowering postpran-
dial glucose levels by stimulating insulin secretion68.
Insulin has a major role in tissue glucose utilization
and also inhibits ghrelin production69. The gut-derived
peptide hormones are involved in both short-term and
long-term appetite control. For instance, while ghre-
lin secretion peaks before a regular meal, long-term
negative energy balance stimulates a sustained increase
in its baseline levels70. Although both types of appetite
control are functionally interrelated, the short-term con-
trol is responsible for the cyclic alternation of hunger
and satiety on a daily basis. By contrast, the long-term
control concerns appetite changes over extended periods
to meet new energy demands.
The current understanding of the mechanisms
responsible for short-term control of appetite implies
that the nutritional-status-dependent release of hunger
and satiety hormones from the gastrointestinal tract
underlies the corresponding sensations and activates
the brain pathways regulating feeding behaviour via
both humoral and neuronal pathways. Accordingly,
food intake is proposed to be triggered by increased
plasma concentrations of ghrelin, which occurs before
a meal, leading to activation of NPY and AgRP neu-
rons that express growth hormone secretagogue recep-
tor 1 (GHSR1)71. Next, the meal-induced increase in
plasma levels of PYY leads to inhibition of NPY and
AgRP neurons via direct binding of PYY (residues
3–36) to the neuropeptide Y receptor type 2 (Y2R)72.
This inhibition of NPY and AgRP neurons removes
their inhibitory GABA-ergic input on the ARC POMC
neurons, thus activating the melanocortin satiety path-
way 73. Furthermore, the vagal afferents in the gut also
express GHSR1 and Y2R and can be directly inhibited
or stimulated by ghrelin and PYY (residues 3 − 36),
respectively 74–76. These vagal afferents also mediate
satietogenic effects of cholecystokinin77.
The cyclic changes in gut-hormone production and
release associated with food intake, thus, seem to be key
signalling events in the brain for short-term regulation
of appetite. However, an open question remains on the
nature of mechanisms underlying the specific dynamics
of gut hormone release in both short-term and long-
term control of feeding. Another question concerns the
inconsistency of the homeostatic model with the short-
term regulation of appetite in regularly fed animals and
humans who continue to perfectly alternate their hun-
ger and satiety feelings under conditions of stable and
often excessive energy supply. As gut-bacteria-derived
signals might access the intestinal hormone signalling
pathways directly in the gut, and possibly also in the
systemic circulation, gut bacteria could conceivably use
these pathways to influence host control of appetite in
order to sustain bacterial population size.
Gut bacterial population dynamics
Nutrient-induced bacterial growth
Similar to the molecular pathways regulating energy
metabolism in the host, the gut microbiota is also sub-
jected to both short-term and long-term modifications.
The short-term daily changes include fluctuations
in bacterial numbers whereas the long-term changes
mainly concern microbiota composition. Potentially,
a functional link might exist between changes in the
number and composition of gut bacteria and the short-
term and long-term regulation of host appetite, respec-
tively. In this section, some bacterial and host-related
factors involved in the regulation of bacterial popula-
tion dynamics are discussed, as they potentially relate
to appetite control.

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 13

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

a Decline c
phase
Bacterial biomass

Stat phase Chemical digestive Bacterial content Time to Transit

∆109 cells

Exp factors produced (cells per g) stationary time

phase by host phase
Small population (<103 cells per ml) Stomach Stomach 101 8.8 h 1–3 h
Large population (>109 cells per ml) HCl pH ~1.4
• Pepsin
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 • Gastric lipase
Time (h)
Liver Duodenum 103 6.6 h 30–60 min
b Host Bile acids
Pancreas Jejunum 104 5.5 h 1–5 h
Ingested Secreted Chemical Peristalsis, • Trypsin
nutrients nutrients digestive contractions • Amylase
factors defecation • Carboxypeptidase
Small intestine Ileum 107 2.2 h 1–5 h
Bacterial growth Bacterial Bacterial Brush border
lysis elimination enzymes

Colon Colon 1012 20 min 10–60 h

No host digestive
factors
Increased cell numbers Decreased cell numbers

Figure 1 | Host factors influencing gut bacterial growth. a | Typical growth dynamics of a large versus small bacterial
Nature Reviews | Endocrinology
population illustrate the different durations of the exponential growth phase (Exp). b | Key host-related factors influencing
the balance between stimulation and inhibition of the bacterial cell number in the gut. The role of the immune system is
not shown, but it contributes by stabilizing the autochthonic community and neutralizing pathogenic invaders.
c | Presence of chemical and digestive factors and the transit time along the gastrointestinal tract might underlie the
increasing rostro–caudal gradient of bacterial content225. In the upper gut, the transit time is, apparently, shorter than the
time necessary for the bacterial population to reach the stationary growth phase (Stat), as calculated using the formula: t
(min) = G (generation time, 20 min, assumed based on in vitro experiments and in vivo infusions) × 3.3 log (minimal bacterial
number in the Stat phase, that is, 109) / bacterial number before multiplication (for example, 103 in the duodenum). Figure
1a modified from Cell Metab. 23 (2), Breton, J. et al. Gut commensal E. coli proteins activate host satiety pathways following
nutrient-induced bacterial growth. 324–334 © (2016), with permission from Elsevier.

Bacteria that live and thrive in communities are
capable of intrinsically regulating their own growth
rate and population size. Bacterial duplication results in
exponential growth with the division rate determined
by the time necessary for DNA replication and cell
division. Under conditions of regular nutrient supply,
which keeps cell-division machinery in a ‘ready’ state78,
it takes about 20 min for bacteria such as Escherichia coli
to duplicate after provision of nutrients. After bacterial
multiplication to a level of ~109 cells per ml, the bacte-
rial population enters a stationary growth phase that is
characterized by maintenance of bacterial numbers over
several hours. Accordingly, the larger the initial bacterial
population, the shorter the time it takes to get to the
stationary phase and vice versa (FIG. 1a) The stationary
phase is followed by a decline phase that is character-
ized by a progressive decrease in the number of bacteria
owing to the natural process of cell lysis79. The molecu-
lar mechanisms responsible for onset of the stationary
phase involve the production of autoinducers, which are
bacterial signals used in quorum sensing 80 (BOX 1).
Depending on growth dynamics, the production
of 109 new bacterial cells in 20 min requires a starting
population of at least 109 cells, as only one division of
each bacteria is possible in the time frame of 20 min.
This cell-number generation serves as a critical mass for
induction of the stationary phase, as higher initial num-
bers of E. coli still result in the same ~109 cells per ml
increase in bacterial numbers after nutrient provision
in vitro17. This phenomenon has also been reproduced
in rats, where the stationary phase of gut bacteria was
induced 20 min after nutrient infusion into the large
intestine17. An earlier study (in 1983) showed that the
doubling time of E. coli implanted into the small or large
intestine of mice was 2 h with an initial bacterial con-
centration of 2.4 × 104 cells per ml81. Nutrient-induced
bacterial growth, thus, displays specific dynamics that
are defined by the intrinsic properties of a bacterial
population (including the gut microbiota).
Host control
In contrast to in  vitro conditions, where bacterial
growth is regulated only by supplied nutrients and
quorum sensing, in the gut, besides this bacteria-
intrinsic regulation, the growth of bacterial popula-
tions is constantly influenced by diet 82 and limited by
the host production of chemical and physical factors.
A balance between these positive and negative influences
determines the relative stability of a bacterial population
at each part of the gastrointestinal tract (FIG. 1b). Among
the chemical factors that cause bacterial lysis and prevent
mucosal adhesion83 are digestive juices containing gas-
tric and bile acids and various digestive enzymes (FIG. 1c).
The main physical factors are the intestinal peristalsis and
contractions of the colon and rectum, which lead to regu-
lar elimination of bacteria by defecation. As an example,

14 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Box 1 | Quorum sensing: a coordinated behaviour of bacterial populations
Autoinducers
Signalling molecules produced by bacteria that act on neighbouring bacteria.
Exponential growth phase
Rapid increase in the number of cells, where each new cell further divides. Duration of
the exponential growth phase of a bacterial population depends on the initial number
of bacteria, as cell division stops once the number of cells reaches a specific threshold.
Stationary growth phase
Maintenance of the size of a population by equilibrium between dividing and lysing
cells. Duration of the stationary phase under in vitro conditions depends mainly on the
availability of nutrients. In viscera, the stationary phase is shortened by host-regulated
bacteria-lysing factors (FIG. 1).
Growth dynamics
The growth dynamics of a bacterial population are regulated by both the availability of
nutrients and by quorum sensing. Growing bacteria release autoinducers that act on
the same bacteria by inducing arrest of cell division and initiating the stationary phase.
Three main types of autoinducers have been described: specific oligopeptides;
acylhomoserine lactones; and the S‑ribosylhomocysteine lyase (LuxS; also known as
autoinducer‑2 production protein LuxS and AI‑2 synthesis protein) inducer
4,5‑dihydroxy 2,3‑pentanedione. Production of these autoinducers is characterized by
high, medium and low energy cost, respectively, and their action is inversely related to
their specificity80. In addition to regulating the same bacterial species, autoinducers can
also act on other neighbouring bacterial species, thus coordinating changes in the
complex microbiota222. Furthermore, they can also act on host cells, for instance,
AI‑2 from E. coli interferes with cytokine production223 and some quorum-sensing
peptides are able to cross the blood–brain barrier224. Whether autoinducers influence
host pathways involved in appetite control remains to be determined.
a daily faecal output of a standard white man contains
~15 g of bacteria, which corresponds to 55% of the dry
weight of faeces84. The increasing rostro–caudal gas-
trointestinal tract gradient of bacterial numbers might,
therefore, reflect the decreasing antibacterial power of
both chemical and physical host-derived factors (FIG. 1c).
In fact, the colon, which contains the highest density of
bacteria, does not secrete digestive enzymes and has no
peristalsis. Inversely, the transit time of ingested nutri-
ents in the upper gut is shorter than the time necessary
for bacteria to enter the stationary phase after nutrient-
induced bacterial growth (FIG. 1c). Spontaneous bacterial
lysis is also present in the large intestine, and has been
estimated to occur at a rate of approximately three to
one cells in the rat 85. Notably, the colonic transit time is
highly variable whereas decreases in the transit time are
associated with increased microbial richness and diver-
sity 86. The constant dynamics of bacterial growth, lysis
and elimination, thus, keep the gut bacterial population
at a fairly stable level, which indicates that both bacteria
and the host contribute to the maintenance of energy
homeostasis in a gut-bacterial population. The immune
system also influences gut-bacterial homeostasis via both
innate and adaptive immunity, for example by producing
secretory antibacterial IgA87–89.
Links with microbiota composition
Host feeding and energy metabolism
Gut microbiota and long-term changes in energy
metabolism. Long-term regulation of appetite serves
the physiological needs of the organism, which vary
with age, sex, reproductive status and the environment.
Although influenced by diurnal rhythms and by short-
term effects of different foods or medicines, the com-
position of the gut microbiota is fairly stable over long
periods of time, thereby demonstrating the ability of the
gut microbiota to maintain population homeo­stasis90–92.
Nevertheless, the gut microbiota also undergoes long-
term changes characteristic of age, sex, nutritional habits
and geographical location93–96. Moreover, various chronic
pathological conditions are characterized by dysbioses of
the gut microbiota and are often accompanied by altered
appetite and food intake, which, in turn, can worsen the
underlying disease, such as is the case for obesity, dia-
betes mellitus and cancer, as well as liver, kidney and
cardiac diseases15,97–100. Although the specific bacterial
species that might contribute to altered appetite and
body weight in these different diseases remain to be dis-
covered, a decreased bacterial diversity has been found in
both obesity 101 and anorexia nervosa102. Moreover, stud-
ies of microbiome composition have found associations
between the gut microbiota and BMI and adiposity 103.
For instance, increased presence of E. coli correlates
inversely with BMI after bariatric surgery, as well as in
patients with anorexia or obesity 104,105. An increased pres-
ence of Enterobacteriaceae has been linked to low body
weight in severely or chronically malnourished chil-
dren88,106, and the presence of Akkermansia muciniphila
is associated with anti-obesogenic effects107. An initial
finding of increased ratios of Firmicutes to Bacteroidetes
in obesity 108 has, however, not been reproduced in other
studies109. The causality between changes in the micro-
biota composition and host feeding behaviour remains
to be established. Although changes in body weight
gain following gut microbiota transplantation in germ-
free mice support this link, these experiments were not
always accompanied by immediate modification of feed-
ing behaviour in recipients. For instance, while germ-
free mice receiving gut microbiota from hyperphagic
toll-like receptor 5 (TLR5)-deficient mice, indeed dis-
played increased food intake110, the hyperphagic trait
was not transmittable by microbiota transfer from
leptin-deficient ob/ob mice13. Similarly, transplanta-
tion of gut microbiota from malnourished children
impaired normal weight gain in recipient mice with-
out causing notable modification of food intake111.
Another example comes from bariatric surgery,
showing that its therapeutic effects of decreasing
appetite and body weight in obesity can be due to
modifications of the gut microbiota after the proce-
dure112. Furthermore, growth dynamics of gut bacteria
can be associated with altered metabolic control, even
without notable changes in their relative abundance;
for example, an increased cell division rate of E. coli
correlates with occurrence of type 2 diabetes mellitus113.
Diurnal rhythms of gut microbiota and host feeding.
Bacterial growth and host-related rhythmic factors,
including regular nutrient supply, underlie the diurnal
fluctuations in bacterial composition that have been
described in the past few years as circadian rhythms
of the gut microbiota 114–117. Indeed, metagenomic
analysis revealed that all gut bacteria in both mice

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 15

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

and humans undergo rhythmic changes over the day
with ~20% of species exhibiting diurnal fluctuation
in their relative abundance118. A bi‑directional link
between hypothalamic clock mechanisms and micro-
biota composition suggests an effect of sunlight on gut
bacterial physiology 116,117. Bacterial rhythms can also
be driven by the prevailing daily distribution of feed-
ing; changing the feeding pattern between the dark and
the light phases shifts the diurnal distribution of bac-
terial abundance114,115. It is of a practical interest that
imposed alterations in circadian rhythms are accom-
panied by obesogenic changes in gut microbiota114,119.
Inversely, obesity is characterized by increased eating
in the resting phase. For instance, whereas increased
light-phase resting eating underlies hyperphagia in
obese Zucker rats120, nocturnal hyperphagia is often
observed in humans with obesity and is called the
‘night-eating syndrome’ (REF. 121). An involvement of
dysbiotic microbiota in such mistimed overeating is
suspected, on the basis of improved metabolic param-
eters in obese rodents after targeting their gut bacteria
using antibiotics122,123 or in humans with obesity after
bariatric surgery 124,125. Associations between diurnal
rhythms of eating and microbiota composition might,
thus, underlie different metabolic requirements dur-
ing the active and resting states, which is a longer-term
parameter than alternations in hunger and satiety. The
peak amplitude during 24 h changes in relative bacte-
rial abundance in human faecal samples is ~50%114,
which points to a high turnover rate of the gut bacterial
population.
Energy intake as food
Used by gut Used by host
bacteria
During starvation,
gut bacteria receive
energy only from
host energy stores
Bacterial growth Metabolic needs
Energy dissipation
Host energy use:
Direct Indirect
Figure 2 | Distribution of Nature
food-derived
Reviewsenergy
| Endocrinology
between the host and gut bacteria. Energy derived
from ingested nutrients is directly and indirectly available
to both gut bacteria and the host. Gut bacteria use this
energy for bacterial multiplication, which results in
population maintenance. In turn, bacteria generate energy
as a part of bacterial catabolism and nutrient processing,
which releases energy that is available to the host. The
host, therefore, receives energy derived both directly from
nutrient digestion and indirectly from gut bacteria to meet
its metabolic needs. Part of the energy generated by both
gut bacteria and the host is dissipated as heat and waste.
Alternatively, during starvation, gut bacteria receive
energy only from host energy stores.
Bacterial and host energy homeostasis
Energy needs of gut bacteria
The gut microbiota, as well as bacteria inhabiting other
niches in the body, depend fully on nutrients and water
provided by their host. Some ingested nutrients can be
used directly by gut bacteria, such as in the upper gut,
but the majority of intestinal bacteria (located in the
large intestine) receive food-remnant nutrients that have
already been digested and absorbed (with the exception
of non-digestible fibre). When feeding starts and food
reaches the oral cavity, nutrients in the circulation and
tissues, and water, are released into the intestinal lumen
via the cephalic phase response126, which indicates that
animals feed their gut microbiota by secreting nutrients
into the gut as often as they eat and, even under peri-
ods of fast, the body continues to secrete these factors.
Considering the remarkable long-term stability of the
gut microbial community, gut bacteria should be able to
regulate their own energy balance, that is, to use the host
mechanisms for their energy needs.
Gut bacteria compose a metabolically active organ
weighing between 1 kg and 2 kg in adult humans. By
calculating how much energy this organ might con-
sume, one can determine its contribution to the daily
needs of the energy intake of the host. Nutrient-induced
bacterial growth in a large population, such as in the
colon, results in simple duplication of bacteria. Bacterial
duplication is energetically costly, requiring 1 mol of
ATP per 10.3 g of cells127. As the complete hydrolysis
of ATP yields ΔG° = −10.9 kcal/mol, it can be roughly
estimated that 1 g of gut bacteria needs 1 kcal (4.18 kJ) of
energy for duplication. Accordingly, to duplicate 1–2 kg
of gut bacteria, the energy needs would be 1,000–2,000
kcal. The latter numbers represent about half of the
recommended daily energy intake for a healthy indi-
vidual, depending on age, sex and level of physical
activity. Luckily, because of quorum sensing, bacterial
division is limited, and the total weight of 109 cells per
ml of newly formed bacteria in a 1,000 ml volume of
the large intestine is only ~1.0 g. Bacterial growth in the
colon after each meal-induced nutrient provision will,
therefore, consume ~1 kcal. However, for the daily
energy consumption by gut bacteria, this value should
be increased by including the energy needs of dividing
bacteria in other parts of the gut and multiplied by the
number of meals, snacks and also presence of non-food
related factors.
To exactly calculate the energy needs of gut bacteria,
additional studies are required, focusing on their daily
turnover as ~15 g of bacteria excreted daily in human
faeces should represent only a fraction of lost bacteria
beside those lysed in the gastrointestinal tract. A 27%
decrease in metabolic rate (oxygen consumption) shown
in germ-free mice128 might be indicative of an approx-
imate contribution of the gut microbiota to the total
energy balance in mice. The example of the constitu-
tive role of gut bacteria in energy consumption, thus,
illustrates that it should be taken into account in the
energy balance of the host (FIG. 2). Moreover, it implies
that alterations in the dynamics of bacterial growth, for
example, those resulting in bacterial overgrowth, will

16 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

be accompanied by increased energy demands of gut
bacteria, which suggests that it might lead to increased
appetite in the host.
Gut bacteria as a source of energy
Energy exchange through the food chain represents a
universal link between all organisms129. The energy (in
the form of ATP) produced by gut bacteria is used not
only for their own growth but also becomes available
to the host. Bacteria release enzymes and metabolites
that help digestion of nutrients, including fibre, that are
not digestible by the host. For instance, E. coli proteins
released after nutrient-induced bacterial growth can
synthesize ATP that is used by the host 17. Moreover, the
continuously renewing bacterial biomass, upon bacterial
lysis, liberates in the gut all their structural molecules,
including proteins constituting ~24% of the bacterial
cell. This indirect bacteria-mediated supply of energy
and nutrients to the host seems a more efficient way of
energy provision than direct extraction from nutrients
by the host digestive system. Indeed, germ-free mice
fed a Western diet display increased food intake but
low adiposity with otherwise similar body weight, and
can be resistant to such high-fat diets (HFDs)117,128,130.
a Satiety perception b E. coli and colonic bacteria growth
Bacterial biomass
However, an absence of the gut microbiota in mice
does not always protect them from gaining weight and
adiposity on different types of HFD, which suggests a
role for diet composition in energy assimilation131,132.
Hyperphagia, which can be a compensatory response to
inefficient energy assimilation, is present in patients after
small bowel resection133,134. The bacterial role in host
energy extraction is further highlighted by an increased
gut-mucosal surface in obesity that does not, however,
prevent hyperphagia135. Moreover, the bacterial capacity
of energy extraction depends on the individual compo-
sition of the gut microbiota, as has been shown to be
increased in obesity 13.
During host starvation, gut bacteria receive nutrients
only from the host, slowly depleting energy reserves
in fat, liver and muscle. Although host starvation (for
example in anorexia nervosa) leads to an altered compo-
sition of gut microbiota102,136, gut bacteria continue their
metabolic activities and maintain their community at
the expense of the host. Changes in the composition of
the gut microbiota during starvation favours species that
are better adapted to low energy supply, which, in turn,
might help the host to survive energy deprivation137.
Gut bacteria might, thus, supply the host with energy
more efficiently than the host extracting it directly from
nutrients using its own metabolic processes. The fact
that germ-free mice can regulate food intake to achieve
the body weight set-point indicates that gut bacteria are

not necessary for animal feeding behaviour. However,

Fullness rating

Stat the peculiar physiology of germ-free mice cannot be

fully extended to natural conditions of bacteria–host
Exp
coexistence. Although it might seem as facultative, the
contribution of gut bacteria to host energy metabolism
cannot be disregarded as, throughout evolution, bacteria
have always been present in the animal gut.
20 40 120 20 40 120
Meal Nutrients
Host appetite control
c GLP1 d PYY Bacterial growth dynamics-based model
In order to secure the energy necessary for stable main-
tenance of a bacterial population, the gut microbiota
should be linked to the host mechanisms of energy
Plasma levels

Plasma levels

intake. To optimize this goal, the dynamics of bacterial

20 40 120 20 40
Meal Time (min) Meal
Figure 3 | Satiety, bacterial growth and satiety hormone release. Meal-induced
Nature Reviews | Endocrinology
changes in satiety perception in humans (part a) temporally overlap with both
nutrient-induced bacterial growth dynamics in vitro and in viscera in rats (part b), and
with meal-induced plasmatic changes in intestinal satiety hormone release in humans
(parts c and d). Release of glucagon-like peptide 1 (GLP1; part c) is associated with the
exponential growth phase (Exp) and release of peptide tyrosine tyrosine (PYY; part d)
with the stationary growth phase (Stat). Figure 3a modified with permission from The
American Physiological Society © Labouré, H. et al. Am. J. Physiol. Regul. Integr. Comp.
Physiol. 282, R1501–R1511 (2002). Figure 3b modified from Cell Metab. 23 (2), Breton, J.
et al. Gut commensal E. coli proteins activate host satiety pathways following
nutrient-induced bacterial growth. 324–334 © (2016), with permission from Elsevier.
Figure 3c and d modified with permission from The American Physiological Society ©
Gerspach, A. C. et al. Am. J. Physiol. Endocrinol. Metab. 301, E317–E325 (2011).
growth should be synchronized with host feeding behav-
iour and feelings of hunger and satiety. Accordingly, a
decline in the size of the bacterial population should be
associated with hunger and, inversely, its stability with
120
satiety.
Time (min) One of the key questions concerning the mechanisms
controlling appetite is why we start to feel satiated after
~20 min of eating138, when the ingested nutrients have
not yet been digested or absorbed (FIG. 3a). The most
probable answer might implicate stomach distention
and increased release of intestinal satiety hormones that
activate anorexigenic pathways in the brain. Indeed,
the typical dynamics of plasma levels of GLP1 and PYY
in humans show an increase ~15–30 min after a meal,
depending on the assay intervals used (FIG. 3c,d). Similar
dynamics for the release of satiety hormones are also
present in the rat 139. Such an increase cannot be the
consequence of direct stimulation of nutrient recep-
tors on the surface of enteroendocrine cells by ingested

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 17

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Prandial phase Postprandial phase Preprandial phase be considered with caution, as although the circadian
fluctuations of bacterial faecal content correlate with
↑ Satiety signalling feeding behaviour, the direct effect of a single spontane-
Meal to the host New meal
ous meal on bacterial growth still remains to be shown.
The molecular mechanisms supporting this model are
Stat discussed in the next section.
dynamics in the colon

Decline From the point of view of maintenance of bacte-

Bacterial growth

rial energy homeostasis, the stationary growth phase

↑ Satiation
↑ Hunger signalling
to the host
Exp
20 min
~5 h
Time
Figure 4 | Bacterial growth dynamic-based model of Nature appetite control.
Reviews Hypothetical
| Endocrinology
model linking bacterial growth phases with host-feeding cycles. During regular feeding
schedules, the meal-induced exponential growth phase (Exp) of bacterial populations in
the large intestine should be terminated after 20 min, a time usually associated with
feeling full and with activation of satiety pathways. When the bacterial population size
declines postprandially due to the natural lysis and elimination of bacteria, the feeling of
satiety also declines resulting in a renewed feeling of hunger and the onset of the next
meal. The inter-meal interval in regularly and spontaneously-fed healthy humans lasts
~5–6 h226. This time corresponds with the duration of the bacterial stationary phase (Stat)
and the beginning of the decline phase accelerated by in viscera conditions. A new meal
will reset the cycle of bacterial growth resulting in long-term maintenance of the gut
bacterial population.
food as, in both humans and rodents, several hours of
intestinal transit are required, during which time most
nutrients are absorbed by the small intestine. In fact, the
only way for nutrients, shortly after a meal, to activate
enteroendocrine cells in the distal gut, is if they are deliv-
ered directly into the gut lumen. Such a secretory reflex
was, indeed, discovered more than 100 years ago by Ivan
Pavlov using sham feeding of dogs140. Pavlov showed that
abundant intestinal secretion starts 1–3 min after a meal
or its anticipation140. If nutrients present in the intestinal
secretions, or ‘appetite juice’ in Pavlov’s terms, can
immediately activate enteroendocrine cells, they will
also trigger bacterial growth that will last for exactly
20 min in the large intestine, similar to in vitro conditions
of a rich bacterial population (FIG. 3b).
The overlap of the dynamics of nutrient-induced
bacterial growth with meal-related satiety perception
and plasma concentrations of gut satiety hormones, in
particular PYY, is striking. This overlap suggests that
bacterial growth dynamics might be causally linked to
signalling of meal-induced satiety. The inverse relation-
ship is improbable as bacterial growth dynamics are an
intrinsic feature of bacterial populations regulated by
quorum sensing (BOX 1). The short-lasting exponential
and long-lasting stationary growth phases of gut bac-
teria, apparently, fit temporally to the host prandial
and postprandial phases of feeding behaviour, respec-
tively. Accordingly, a theoretical model of host appe-
tite control based on gut bacterial homeostasis can be
proposed, in which the specific dynamics of bacterial
growth are responsible for the alternation of feelings of
hunger and satiety in the host (FIG. 4). This model should
should reflect the maximal entropy state of a microbial
community characterized by no immediate needs in
nutrient supply. Bacteria-derived molecules produced
during the stationary phase might signal energy equi-
librium to the host, which is converted to a feeling of
satiety. Conversely, a decline phase of the gut bacterial
population might be accompanied by a decrease of such
stationary phase-related signals, which is perceived by
the host as an energy deficit and a feeling of hunger.
Furthermore, growing bacteria might send the host
signals of anticipation of the stationary phase, that is,
they might prepare the host for nutrient assimilation, for
instance by regulating insulin secretion. The energy sta-
tus of a bacterial population might, thus, drive alternate
activation of hunger and satiety pathways.
Depending on the occurrence of the stationary phase,
bacterial communities occupying different compart-
ments of the gastrointestinal tract might make different
contributions to appetite control. In fact, given the fairly
low number of bacteria found in the upper gut (FIG. 1c),
stationary growth phase might not be reached, owing
to lysis and elimination. In the large intestine, however,
numerous bacteria thrive and quickly reach the stationary
phase upon feeding. Accordingly, nutrient-induced divi-
sion of bacteria in the upper gut is more likely to transmit
signals only from their exponential growth phase, which
can be associated with an acute release of upper-gut hor-
mones such as cholecystokinin141. By contrast, in the large
intestine, gut bacteria might send signals from both the
exponential and stationary phases, underlying both acute
and long-lasting release of gut hormones, respectively.
Mechanistic effects of gut bacteria
Gut bacteria in short-term control of appetite. What are
the molecular signals that could link bacterial growth
phases with host pathways signalling hunger and satiety?
A stable increase in bacterial biomass after each nutrient-
induced bacterial multiplication, limited by the sta-
tionary phase, suggests that a simple and reproducible
concentration gradient of several bacteria-derived mol-
ecules might be perceived by the gut sensory system
upstream of the satiety pathways. Such bacterial mol-
ecules can be more or less specific, that is, produced
by certain bacterial species or by large representative
families of gut bacteria, respectively. Typically, proteins
and their derived peptides transmit more specific signals
than lipids and sugars.
For the activation of short-term pathways of appetite
control, bacterial molecules should be detected by the
chemical sensory elements of the gut epithelium located
directly on enteroendocrine cells or in the vagal afferents
of the gut 63,142. Alternatively, bacterial molecules might
activate enteroendocrine cells via modulation of paracrine

18 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Systemic number of host-like neuroactive compounds, including

Gut epithelium Gut lumen Submucosal layer circulation
Partly digested
nutrients and fibres Vagal
Digestion afferents
EEC
Bacterium Energy substrates:
ATP, lactate,
butyrate, etc.
Enteric
neuron
Bioactive
molecules Satiety
hormones
LPS, 5HT,
indole, etc. PYY, GLP1,
etc.
Mimetics of peptide
hormones (ClpB)
Figure 5 | Gut bacteria-derived chemical signals that mightNature activate
Reviewsintestinal
| Endocrinology
satiety pathways. During their life in the gut, bacteria metabolize non-digestible fibre
and digestible nutrients and produce several energy substrates such as ATP, lactate and
butyrate. Upon bacterial lysis, bioactive molecules such as lipopolysaccharide (LPS)
and proteins are released that continue their enzymatic activities synthesizing bioactive
metabolites, such as 5‑hydroxytryptamine (5HT), or acting directly as mimetics of peptide
hormones, such as caseinolytic peptidase B protein homologue (ClpB). All the
bacteria-derived chemical signals, along with nutrients, have direct contact with the gut
epithelium that carries chemical sensors. Direct or indirect (via the enterocytes)
activation of enteroendocrine cells (EECs) by bacterial signals triggers local and systemic
release of peptide tyrosine tyrosine (PYY) and glucagon-like peptide 1 (GLP1), thereby
transmitting satiety. Paracrine actions of bacteria-derived molecules on EEC and
entrochromaffin cells, producing 5HT, might activate the enteric nervous system
regulating intestinal motility and gut barrier permeability, including the access of
bacterial signals to the vagal afferents.
signals released from enterocytes143. The enteroendocrine
cells will then secrete hormones into the systemic circula-
tion, as well as locally, to activate vagal afferents. Indeed,
intraluminal application of Lactobacillus increases electri-
cal activity of vagal afferents144. However, the vagal nerve
endings, although numerous in the gut mucosal layer,
do not penetrate the lamina propria of the epithelial cell
layer 145. Therefore, for a direct action on vagal afferents,
bacterial molecules should pass across the intestinal epi-
thelial barrier. Interestingly, germ-free rats have highly
elevated plasma levels of PYY and enteroglucagon146,
which suggests that an absence of gut bacterial stimulation
deregulates enteroendocrine-cell physiology.
The functional permeability of the gut barrier involves
the coordinated action of the enteric nervous system,
which is activated by locally secreted transmitters such
as serotonin (also known as 5‑hydroxytryptamine and
5HT)147,148. Whereas increased intestinal permeability
is observed in dysbiotic HFD-induced obese mice149,150,
some probiotics and prebiotics might enhance the func-
tion of the intestinal barrier 151–153. Bacteria produce a
biological amines154, that contribute to the biosynthesis
of 5HT in the gut epithelium155,156. Interestingly, indole, a
precursor of tryptophan and 5HT, is released during the
growth of several bacteria, including E. coli, as it is used
by bacteria as a chemical signal to protect them from
stress157. Indole was shown to modulate GLP1 secretion
depending on its concentration158 and is also a ligand of
the aryl hydrocarbon receptor that can transmit bacte-
rial signals to host intestinal immunity 159. Moreover,
the presence of the aryl hydrocarbon receptor pathway
in the hypothalamic NPY/AgRP and POMC neurons160
suggests a role of bacterial-derived indole in central
appetite regulation.
Although the identity of bacterial molecules that
might activate enteroendocrine cells following nutrient-
induced bacterial growth is unknown, we can speculate
about molecules that are potentially relevant for short
term-appetite control (FIG. 5). Polysaccharides, pepti-
doglycan and lipoproteins are among the less specific,
but most abundant, bacterial molecules that might
activate toll-like receptors (TLRs) expressed by the
gut epithelium, including enteroendocrine cells161,162.
Lipopolysaccharide is a product of lysis of Gram-negative
bacteria (as reflected by its other historical name ‘endo-
toxin’), in contrast to releasable toxins from live bacte-
ria163. Lipopolysaccharide was shown to synchronize gene
expression in the intestinal epithelium164 and its ingestion
decreased the taste response to sucrose165. Presence of the
lipopolysaccharide receptor TLR4 in the vagal afferents166
also supports a role of gut lipopolysaccharide in signalling
satiety, which is in line with its anorexigenic effects167.
Besides their cellular constituents, bacteria expose
the host to their metabolites as a result of enzymatic pro-
cessing of nutrients and prebiotics such as non-digestible
fibre. Butyrate, propionate and acetate are the main
short-chain fatty acids (SCFAs) produced by gut bac-
teria from various substrates, including non-digestible
fibre168, which exert multi-organ effects on host energy
metabolism169,170. Receptors for SCFAs are present on
enteroendocrine cells and propionate can directly acti-
vate PYY and GLP1 release from L‑cells both in vitro
and in vivo in humans171–173. A putative role of SCFAs
generated following nutrient-induced bacterial growth
in signalling satiety is consistent with the known, albeit
modest, effects of dietary fibre in reducing appetite174
and stimulating secretion of satiety hormones175.
In addition to their enzymatic role in nutrient catab-
olism, some bacterial proteins might interact specifi-
cally with the host in their role as signalling molecules
owing to their molecular mimicry with peptide hor-
mones. The first such molecule was identified in E. coli
as a conformational antigen mimetic of αMSH176. The
protein is the caseinolytic peptidase B protein homo-
logue (ClpB), containing a discontinuous αMSH-like
epitope. Indeed, animals exposed to ClpB-producing
bacteria or those immunized against ClpB generate
antibodies against ClpB that are crossreactive with
αMSH176,177. ClpB expression is increased in the bac-
terial stationary growth phase17, in agreement with its
role as a disaggregation chaperone protein in bacteria178.

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 19

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

ClpB and other bacterial proteins have been detected in
the gut mucosa17,179 and, hence, might act directly on the
intestinal epithelium, including enteroendocrine cells.
Remarkably, colonic infusion in rats of E. coli proteins
from the stationary, but not exponential, growth phase
increased portal levels of PYY17. By contrast, proteins
from the exponential phase increased portal levels
of GLP1 (REF. 17).
Such differential effects of bacterial proteins sug-
gest their involvement in regulation of the sequential
release of satiety hormones; acute release of GLP1 with
an incretin function during the first 20 min of the bacte-
rial exponential growth phase, and a prolonged release
of PYY initiated after the onset of the stationary phase17.
Although these effects of bacterial proteins might not
only depend on ClpB, this protein seems to have a key
role in the host appetite control effects of E. coli. Indeed,
intragastric gavage of ClpB-deficient E. coli failed to
reduce food intake and body weight in mice, effects
observed with the wild-type ClpB-expressing strain of
E. coli 176. The types of melanocortin receptor (MCR)
that might possibly transmit ClpB action on the gut
satiety pathway remain to be clarified. In fact, although
L‑cells express MC4R180, a 14‑amino acid fragment of
ClpB containing the αMSH-like epitope was shown to
activate MC1R but not MC4R181. The gut epithelium
abundantly expresses functional MC1R, which has a
protective role against inflammation182. However, while
MC1R is probably present on enterocytes, it might not be
expressed by the enteroendocrine cells180. In addition to
the demonstrated effects of gut bacterial proteins on the
intestinal satiety pathway, future studies should explore
their effects on the hunger pathway, that is, the release of
ghrelin from the stomach. In fact, ghrelin production has
been associated with presence of Helicobacter pylori in the
gastrointestinal tract183.
Gut bacteria in long-term control of appetite. The caus-
ative associations between host energy metabolic pheno-
types and composition of the gut microbiota, including
its long-term stability, point to the existence of molecular
links between gut bacteria and the hypothalamic pon-
derostatic network. To directly influence this neuronal
circuitry, bacterial components should be present in the
circulation where they might reflect the relative abun-
dance of a gut bacterial population. Although bacterial
DNA has been detected in many organs and tissues184, data
on potential functional effects of bacterial components
circulating in the host are limited.
Lipopolysaccharide is probably the only bacterial
fragment that has been relatively well studied con-
cerning its systemic effects; plasma lipopolysaccharide
levels increase postprandially after a high-fat meal
and are associated with increased energy intake185,186.
Although high doses (≥0.1 mg/kg) of lipopolysaccha-
ride are known to decrease food intake in animals, inde-
pendent from inducing fever 187, its physiological role in
the regulation of long-term energy balance is unclear.
Lipopolysaccharide, via its downstream messenger
IL‑1β induces a chronic inflammatory state that includes
hypothalamic inflammation, with a similar pattern of
neuropeptide expression as observed in both obesity,
anorexia nervosa and cachexia188. Lipopolysaccharide-
induced low-grade inflammation might also interfere
with the long-term regulation of energy metabolism
via increased production of endocannabinoids, such as
anandamide189. These lipid mediators are elevated in the
hypothalamus of obese rodents and are known for their
orexigenic effects by activating cannabinoid CB1 recep-
tors expressed in inhibitory presynaptic GABA terminals
in orexigenic neurons190,191.
Other products of bacterial metabolism relevant to
appetite control and sensed by the host include ATP and
lactate, which are ubiquitous energy substrates. Beside
its key role as an energy substrate, circulating ATP might
also signal the state of energy balance192. Catabolism of all
macronutrients results in increased production of ATP,
whereas anabolism consumes ATP and results in a rela-
tive increase of AMP. Changes in the ATP to AMP ratio
regulate 5ʹ‑AMP-activated protein kinase, which is linked
to central appetite pathways193. ATP receptors are present
on nerve cells, including the NPY/AgRP neurons194. The
efficient production of ATP by E. coli proteins in vitro
suggests that nutrient-induced increases in bacterial bio-
mass in the gut, including the enzymes necessary for ATP
production, might increase both local and systemic
ATP levels and lead to modulation of the activity of cells
constituting appetite-regulating pathways.
Lactate, the preferred substrate for neurons195,196,
is abundantly produced in the gut by Lactobacilli, and
also by Enterobacteriaceae and Bifidobacteria, from fer-
mentable carbohydrates197. Lactate levels are increased
postprandially in the portal blood, whereas portal infu-
sion of lactate reduces meal size, thus supporting a role
for gut-derived lactate in satiation198,199. SCFAs can also
participate in long-term regulation of energy metabo-
lism; systemic administration of acetate in mice failed
to stimulate secretion of PYY and GLP1 but decreased
appetite through central effects200, whereas butyrate and
propionate stimulated intestinal gluconeogenesis201.
An immunoassay for Enterobacteriaceae ClpB has
been developed and used to show that this bacterial pro-
tein is naturally present at variable levels in plasma of
both rodents and humans17,202. Although plasma levels
of ClpB in rats remained stable after colonic infusion of
nutrients or E. coli proteins, plasma ClpB levels correlated
directly with ClpB DNA levels in the gut micro­biota17.
These results, coupled with ClpB activation of ARC
POMC neurons17, support a ClpB-mediated humoral
link between the gut microbiota and hypothalamic neuro-
peptidergic circuitry. Furthermore, E. coli proteins in the
ClpB-rich stationary phase increased c‑Fos in the ARC
POMC, VMN and CeA neurons, demonstrating activa-
tion of both the sensory and integrative centres of the ano-
rexigenic circuitry in the brain. Such anorexigenic effects
of E. coli proteins are in agreement with data associating
the relative abundance of Enterobacteriaceae with human
metabolic phenotypes, including obesity and anorexia
nervosa. Nevertheless, a 2016 study in patients with eating
disorders found that plasma levels of ClpB did not corre-
late significantly with BMI, but did correlate with other
feeding-related parameters such as bulimia nervosa202.

20 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Long-term regulation of appetite by gut bacteria might
also involve the immune system. Diet-induced changes in
microbiota and microbial antigens has profound effects
on gut mucosal immunity stimulating maturation of
the immune system and antibody production9,203–206.
Immunoglobulins are among the most stable biomol-
ecules; the 26 day half-life of IgG207 is a characteristic
suitable for long-term effects. Circulating immunoglob-
ulins that react with several appetite-regulating peptide
hormones, including αMSH and ghrelin, are present
in healthy individuals and their plasma levels correlate
with behavioural traits208–212. In particular, αMSH-reactive
IgG and IgM are associated with eating disorder-related
traits, which suggests that they have a role in modulation
of αMSH signalling via MC4R209. Such a modulatory
role of IgG is not necessarily inhibitory but might, con-
versely, reduce the threshold of MC4R activation213,214.
Furthermore, circulating IgG might protect highly unsta-
ble peptide hormones from degradation by functioning
as a peptide carrier. Such a role has been shown for
ghrelin-reactive IgG, which has slightly increased micro­
molar affinity for ghrelin in individuals with obesity,
thereby enhancing the orexigenic effects of ghrelin215.
Altered affinity kinetics of IgG for ghrelin might
be a response of the immune system to changes in
the composition of the gut microbiota induced by
an obesogenic diet; however, bacterial ghrelin-like
antigens have yet to be identified216. Nevertheless,
the identification of ClpB as an antigenic E.  coli-
derived protein responsible for production of immu-
noglobulins crossreactive with αMSH provides an
example of molecular mimicry underlying a role of
the gut microbiota at the origin of peptide-hormone-
reactive IgG176. Furthermore, an altered immune response
to some bacterial antigens might also be involved in the
pathophysiology of eating disorders. For instance,
the crossreactivity of antibodies against ClpB and αMSH
could alter αMSH-mediated signalling of satiety and
anxiety, and possibly lead to eating disorders176,202,209,217.
Specific bacterial antigens in the gut microbiota are,
thus, plausible long-term factors regulating appetite
control under both normal and pathological conditions.
Integration of gut bacterial growth
From the point of view of co‑evolution of animals
and their microbiota218, animals have probably phylo­
genetically developed their molecular pathways of
hunger and satiety as an adaptation to bacterial chem-
ical signals. In this symbiotic relationship, interactions
between the microbiota and their host should not be
in conflict, but instead, underlie a supraorganismal
physiology of energy metabolism regulation. The cur-
rent homeostatic model of alternation of hunger and
satiety is based on activation of orexigenic and ano-
rexigenic pathways in the brain that depend on the
energy status of the body. Considering data linking
gut bacteria to host regulation of feeding behaviour, a
new integrative model of appetite control can be pro-
posed that incorporates gut bacterial growth dynamics
with the host molecular pathways for homeostatic
control of appetite (FIG. 6).
According to this model, the central sites responsive
to short-term and long-term signals controlling appetite
might integrate both host-derived and bacteria-derived
signals, that is, they will respond to the homeostatic needs
of both host and gut bacteria. Integration of short-term
control might occur in the gut, where gut sensory systems
are activated by nutrients released by the host in response
to nutrient ingestion, as well as by signals derived from
gut bacteria during their different growth phases. The
hypothalamus and other central sites might integrate
the circulating long-term signals of host energy storage
and bacterial community energy status, in part assisted
by antibodies against gut bacterial proteins that are
crossreactive with hunger and satiety hormones.
This new model could help to resolve some con-
ceptual problems inherited from the model of a closed,
feedback-driven self-regulatory system. For instance, the
apparent inability of the central sites to homeostatically
respond to host-derived signals during chronic altera-
tions of host energy balance is observed in obesity. In fact,
increased appetite and food intake in obesity cannot be
explained by energy shortage as, on the contrary, excess
energy is available in the body. A constitutive role of gut
bacteria in the regulation of host energy homeostasis
might explain such inconsistencies. In conditions such
as obesity, bacteria-derived signals might dominate over
host-derived signals to enable gut bacteria to maintain
their own energy homeostasis despite discrepancy with
host energy needs. For instance, bacterial signals from the
gut might compete with increased plasma levels of ano-
rexigenic homeostatic signals in obesity such as leptin219,
whereas modification of the gut microbiota by prebiotics
in obese mice improves leptin sensitivity 220. This possibil-
ity is in line with a theory of genetic competition between
the host and its microbiota for food sources221.
Conclusions
This Review has focused on the molecular pathways of
hunger and satiety in the context of gut bacterial biology,
and proposes a new approach to an old basic physiolog-
ical question concerning the mechanisms of appetite
control. A new model is justified by the accumulating
data implicating the gut microbiota in host metabolic
phenotypes, including eating behaviour. The key con-
ceptual change is the suggestion that nutrient-induced
growth dynamics of the gut microbiota might partici-
pate, on a regular daily basis, in the host regulation of
appetite so as to maintain bacterial homeostasis. In fact,
the energy necessary for gut bacterial multiplication
could constitute an important part of the daily energy
requirements of a healthy person and might shed light
on the mechanisms underlying increased food intake
during obesity. These molecular mechanisms include
bacteria-derived chemical signals that might transmit to
the host the energy status of the gut microbiota that has
a constitutive role in short-term regulation of appetite.
Moreover, the microbiota composition might be a deter-
mining factor in fine tuning the regulation of host energy
metabolism (specifically in the long term), including sspe-
cific immune changes. This new model implies that pres-
ence of a ‘healthy’ gut microbiota is necessary to ensure

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 21

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Brain Anorexigenic pathways Satiety

PBN → CeA: CGRP

Fullness
PVN: OT, CRH Meal Meal
Food VMN: BDNF
ingestion
Hunger
Hypothalamus 20 min 5h

NPY
ARC and/or 2 POMC NTS
AgRP

TH
Nodose ganglion
Fat

Stomach Ghrelin Leptin

Antibody n. Vagus Long-term

Glucose Insulin control
utilization
PYY afferents
(Stat) 3
αMSH ClpB
Pancreas GLP1
(Exp)

Gut epithelium (enterocyte) MC4R

Short-term
control
EEC
1
Bacteria-derived
Cephalic phase signals
Food digestion nutrient,
and absorption electrolyte and
water secretion Stat
Exp
Gut microbiota
Gut bacteria composition
20 min 5h
Bacterial growth dynamics

Small intestine Large intestine

Figure 6 | Bacteria–host integrative homeostatic model of appetite control. Hypothetical model of appetite control
that integrates gut bacteria-derived signals into host molecular pathways that control energyNature Reviews | Endocrinology
homeostasis. According to
this model, activation of host intestinal satiety pathways by nutrients is integrated with nutrient-induced dynamics of gut
bacterial growth (1). For example, glucagon-like peptide 1 (GLP1) is stimulated during the exponential growth phase (Exp)
and functions as an incretin via stimulation of insulin; conversely, peptide tyrosine tyrosine (PYY) is stimulated during the
stationary growth phase (Stat), which occurs 20 min after nutrient supply and leads to activation of the anorexigenic
circuitry. The contribution of the gut microbiota to long-term control of appetite might involve systemic effects of
bacterial components whose plasma levels depend on the composition of the microbiota. For example, caseinolytic
peptidase B (protein homologue ClpB), an E. coli-derived antigen-mimetic protein of α-melanocyte stimulating hormone
(αMSH), can activate POMC ARC neurons in a similar way to leptin, which suggests that the hypothalamic peptidergic
network might integrate both host and gut microbiota energy states (2). Furthermore, while immunoglobulins modulate
long-term stability and the functional activity of hunger and satiety hormones, such as ghrelin and αMSH, their plasma
levels and affinity are influenced by molecular mimicry of gut bacterial antigens, such as between ClpB and αMSH (3).
These natural antibodies might serve as molecular ‘bridges’ in communication between the gut microbiota and host appe-
tite-controlling pathways as well as for other homeostatic functions. AgRP, agouti-related protein; ARC, arcuate nucleus;
BDNF, brain-derived neurotrophic factor; CeA, central nucleus of amygdala; CGRP, calcitonin gene-related peptide; CRH,
corticotropin-releasing factor; EEC, enteroendocrine cell; MC4R, melanocortin receptor type 4; NPY, neuropeptide
tyrosine; NTS, nucleus of the solitary tract; OT, oxytocin; PBN, parabrachial nucleus; POMC, pro-opiomelanocortin; PVN,
paraventricular nucleus; TH, tyrosine hydroxylase; VMN, ventromedial nucleus.

optimal regulation of host appetite both for short-term
and long-term goals. Determining the key bacterial spe-
cies characteristic of ‘healthy’ versus dysbiotic microbiota,
and the underlying mechanisms of their action, is a hot
topic of research. The integrative model of appetite control
presented here might provide a conceptual framework to
help guide future studies on the involvement of bacterial
signals in the regulation of host energy homeostasis. The
identification of additional bacterial molecules involved in
appetite control should help to create the knowledge nec-
essary for improved prevention and treatment of appetite
problems and their metabolic consequences.

22 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

1. Turnbaugh, P. J. et al. The human microbiome project.
Nature 449, 804–810 (2007).
2. Sekirov, I., Russell, S. L., Antunes, L. C. & Finlay, B. B.
Gut microbiota in health and disease. Physiol. Rev.
90, 859–904 (2010).
3. Dethlefsen, L., McFall-Ngai, M. & Relman, D. A. An
ecological and evolutionary perspective on human–
microbe mutualism and disease. Nature 449,
811–818 (2007).
4. Gilbert, S. F., Sapp, J. & Tauber, A. I. A symbiotic view
of life: we have never been individuals. Q. Rev. Biol.
87, 325–341 (2012).
5. Bukharin, O. V. & Perunova, N. B. Microsymbiocenosis
(UrDepart RAS, 2014).
6. Bäckhed, F., Ley, R. E., Sonnenburg, J. L.,
Peterson, D. A. & Gordon, J. I. Host–bacterial
mutualism in the human intestine. Science 307,
1915–1920 (2005).
7. Fergus, C., Barnes, D., Alqasem, M. & Kelly, V. The
queuine micronutrient: charting a course from
microbe to man. Nutrients 7, 2897–2929 (2015).
8. Sonnenburg, E. D. et al. Diet-induced extinctions in
the gut microbiota compound over generations.
Nature 529, 212–215 (2016).
9. Hooper, L. V., Littman, D. R. & Macpherson, A. J.
Interactions between the microbiota and the immune
system. Science 336, 1268–1273 (2012).
10. Cryan, J. F. & Dinan, T. G. Mind-altering
microorganisms: the impact of the gut microbiota on
brain and behaviour. Nat. Rev. Neurosci. 13,
701–712 (2012).
11. Gautron, L., Elmquist, J. K. & Williams, K. W. Neural
control of energy balance: translating circuits to
therapies. Cell 161, 133–145 (2015).
12. Aigner, M. Treasure, J., Kaye, W., Kasper, S. & The
WFSBP Task Force on Eating Disorders. World
Federation of Societies of Biological Psychiatry
(WFSBP) guidelines for the pharmacological treatment
of eating disorders. World J. Biol. Psychiatry 12,
400–443 (2011).
13. Turnbaugh, P. J. et al. An obesity-associated gut
microbiome with increased capacity for energy
harvest. Nature 444, 1027–1131 (2006).
14. Armougom, F., Henry, M., Vialettes, B., Raccah, D. &
Raoult, D. Monitoring bacterial community of human
gut microbiota reveals an increase in lactobacillus in
obese patients and methanogens in anorexic patients.
PLoS ONE 4, e7125 (2009).
15. Rosenbaum, M., Knight, R. & Leibel, R. L. The gut
microbiota in human energy homeostasis and obesity.
Trends Endocrinol. Metab. 26, 493–501 (2015).
16. Fetissov, S. O. & Déchelotte, P. The new link between
gut–brain axis and neuropsychiatric disorders. Curr.
Opin. Clin. Nutr. Metab. Care 14, 477–482 (2011).
17. Breton, J. et al. Gut commensal E. coli proteins activate
host satiety pathways following nutrient-induced
bacterial growth. Cell Metab. 23, 1–11 (2016).
18. Thornton, S. N. Thirst and hydration: physiology and
consequences of dysfunction. Physiol. Behav. 100,
15–21 (2010).
19. Schwartz, M. W. Central nervous system control of
food intake. Nature 404, 661–671 (2000).
20. Hökfelt, T., Bartfai, T. & Bloom, F. Neuropeptides:
opportunities for drug discovery. Lancet Neurol. 2,
463–472 (2003).
21. Carter, M. E., Soden, M. E., Zweifel, L. S. &
Palmiter, R. D. Genetic identification of a neural circuit
that suppresses appetite. Nature 503, 111–114 (2013).
22. Atasoy, D., Betley, J. N., Su, H. H. & Sternson, S. M.
Deconstruction of a neural circuit for hunger. Nature
488, 172–177 (2012).
23. Norsted, E., Gömüç, B. & Meister, B. Protein
components of the blood–brain barrier (BBB) in the
mediobasal hypothalamus. J. Chem. Neuroanat. 36,
107–121 (2008).
24. Langlet, F. et al. Tanycytic VEGF‑A boosts blood–
hypothalamus barrier plasticity and access of
metabolic signals to the arcuate nucleus in response
to fasting. Cell Metab. 17, 607–617 (2013).
25. Schwartz, G. J. Brainstem integrative function in the
central nervous system control of food intake. Forum
Nutr. 63, 141–151 (2010).
26. Waterson, M. J. & Horvath, T. L. Neuronal regulation
of energy homeostasis: beyond the hypothalamus and
feeding. Cell Metab. 22, 962–970 (2015).
27. Richard, D. Cognitive and autonomic determinants of
energy homeostasis in obesity. Nat. Rev. Endocrinol.
11, 489–501 (2015).
28. Sawchenko, P. E. Toward a new neurobiology of energy
balance, appetite, and obesity: the anatomists weigh
in. J. Comp. Neurol. 402, 435–441 (1998).
29. Elmquist, J. K., Maratos-Flier, E., Saper, C. B. &
Flier, J. S. Unraveling the central nervous system
pathways underlying responses to leptin.
Nat. Neurosci. 1, 445–450 (1998).
30. Luquet, S., Perez, F. A., Hnasko, T. S. & Palmiter, R. D.
NPY/AgRP neurons are essential for feeding in adult
mice but can be ablated in neonates. Science 310,
683–685 (2005).
31. Bumaschny, V. F. et al. Obesity-programmed mice are
rescued by early genetic intervention. J. Clin. Invest.
122, 4203–4212 (2012).
32. Loh, K., Herzog, H. & Shi, Y.‑C. Regulation of energy
homeostasis by the NPY system. Trends Endocrinol.
Metab. 26, 125–135 (2015).
33. Ollmann, M. M. et al. Antagonism of central
melanocortin receptors in vitro and in vivo by agouti-
related protein. Science 278, 135–138 (1997).
34. Hahn, T. M., Breininger, J. F., Baskin, D. G. &
Schwartz, M. W. Coexpression of Agrp and NPY in
fasting-activated hypothalamic neurons.
Nat. Neurosci. 1, 271–272 (1998).
35. Tatemoto, K. Neuropeptide Y — a novel brain peptide
with structural similarities to peptide YY and pancreatic
polypeptide. Nature 296, 659–660 (1982).
36. Fan, W., Boston, B. A., Kesterson, R. A., Hruby, V. J. &
Cone, R. D. Role of melanocortinergic neurons in
feeding and the agouti obesity syndrome. Nature
385, 165–168 (1997).
37. Poggioli, R., Vergoni, A. V. & Bertolini, A. ACTH-(1–24)
and α-MSH antagonize feeding behavior stimulated by
κ opiate agonists. Peptides 7, 843–848 (1986).
38. Harris, J. I. & Lerner, A. B. Amino-acid sequence of the
α-melanocyte-stimulating hormone. Nature 179,
1346–1347 (1957).
39. Cone, R. D. Anatomy and regulation of the central
melanocortin system. Nat. Neurosci. 8, 571–578
(2005).
40. Broberger, C., Johansen, J., Johansson, C.,
Schalling, M. & Hökfelt, T. The neuropeptide Y/agouti
gene-related protein (AGRP) brain circuitry in normal,
anorectic, and monosodium glutamate-treated mice.
Proc. Natl Acad. Sci. USA 95, 15043–15048 (1998).
41. Zhao, R., Chen, H. & Sharp, B. M. Nicotine-induced
norepinephrine release in hypothalamic
paraventricular nucleus and amygdala is mediated by
N‑methyl-d‑aspartate receptors and nitric oxide in the
nucleus tractus solitarius. J. Pharmacol. Exp. Ther.
320, 837–844 (2007).
42. Appleyard, S. M. et al. Proopiomelanocortin neurons
in nucleus tractus solitarius are activated by visceral
afferents: regulation by cholecystokinin and opioids.
J. Neurosci. 25, 3578–3585 (2005).
43. Appleyard, S. M. et al. Visceral afferents directly
activate catecholamine neurons in the solitary tract
nucleus. J. Neurosci. 27, 13292–13302 (2007).
44. Larsen, P. J., Tang-Christensen, M., Holst, J. J. &
Ørskov, C. Distribution of glucagon-like peptide‑1 and
other preproglucagon-derived peptides in the rat
hypothalamus and brainstem. Neuroscience 77,
257–270 (1997).
45. Mimee, A., Kuksis, M. & Ferguson, A. V. α‑MSH
exerts direct postsynaptic excitatory effects on NTS
neurons and enhances GABAergic signaling in the
NTS. Neuroscience 262, 70–82 (2014).
46. Wang, D. et al. Whole-brain mapping of the direct
inputs and axonal projections of pro-
opiomelanocortin (POMC) and agouti-related
peptide (AgRP) neurons. Front. Neuroanat. 9, 40
(2015).
47. Cone, R. D. The central melanocortin system and
energy homeostasis. Trends Endocrinol. Metab. 10,
211–216 (1999).
48. Friedman, J. M. & Mantzoros, C. S. 20 years of leptin:
from the discovery of the leptin gene to leptin in our
therapeutic armamentarium. Metabolism 64, 1–4
(2015).
49. Hillebrand, J. J., de Wied, D. & Adan, R. A.
Neuropeptides, food intake and body weight
regulation: a hypothalamic focus. Peptides 23,
2283–2306 (2002).
50. Bouret, S. G., Draper, S. J. & Simerly, R. B. Trophic
action of leptin on hypothalamic neurons that regulate
feeding. Science 304, 108–110 (2004).
51. Simon, J. J. et al. Neural signature of food reward
processing in bulimic-type eating disorders. Soc. Cogn.
Affect. Neurosci. http://dx.doi.org/10.1093/scan/
nsw049 (2016).
52. Muenzberg-Gruening, H., Qualls-Creekmore, E., Yu, S.,
Morrison, C. D. & Berthoud, H.‑R. Hedonics act in
unison with the homeostatic system to unconsciously
control body weight. Front. Nutr. 3, 6 (2016).
53. Denis, R G. P. et al. Palatability can drive feeding
independent of AgRP neurons. Cell Metab. 22,
646–657 (2015).
54. Avena, N. M., Rada, P. & Hoebel, B. G. Underweight
rats have enhanced dopamine release and blunted
acetylcholine response in the nucleus accumbens while
bingeing on sucrose. Neuroscience 156,
865–871 (2008).
55. Jerlhag, E. et al. Ghrelin administration into tegmental
areas stimulates locomotor activity and increases
extracellular concentration of dopamine in the nucleus
accumbens. Addict. Biol. 12, 6–16 (2007).
56. Meguid, M. M. et al. Hypothalamic dopamine and
serotonin in the regulation of food intake. Nutrition 16,
843–857 (2000).
57. Legrand, R., Lucas, N., Breton, J., Déchelotte, P. &
Fetissov, S. O. Dopamine release in the lateral
hypothalamus is stimulated by α‑MSH in both the
anticipatory and consummatory phases of feeding.
Psychoneuroendocrinology 56, 79–87 (2015).
58. Barson, J. R., Morganstern, I. & Leibowitz, S. F.
Similarities in hypothalamic and mesocorticolimbic
circuits regulating the overconsumption of food and
alcohol. Physiol. Behav. 104, 128–137 (2011).
59. Norgren, R., Hajnal, A. & Mungarndee, S. S. Gustatory
reward and the nucleus accumbens. Physiol. Behav.
89, 531–535 (2006).
60. Broberger, C. Brain regulation of food intake and
appetite: molecules and networks. J. Intern. Med.
258, 301–327 (2005).
61. Mutt, V. in Advances in Metabolic Disorders Vol. 11 (eds
Luft, R. & Levine, R.) 1–545 (Academic Press, 1988).
62. Wren, A. M. & Bloom, S. R. Gut hormones and appetite
control. Gastroenterology 132, 2116–2130 (2007).
63. Janssen, S. & Depoortere, I. Nutrient sensing in the
gut: new roads to therapeutics? Trends Endocrinol.
Metab. 24, 92–100 (2013).
64. Kojima, M. et al. Ghrelin is a growth-hormone-
releasing acylated peptide from stomach. Nature 402,
656–660 (1999).
65. Feinle-Bisset, C. Upper gastrointestinal sensitivity to
meal-related signals in adult humans — relevance to
appetite regulation and gut symptoms in health,
obesity and functional dyspepsia. Physiol. Behav. 162,
69–82 (2016).
66. Degen, L., Matzinger, D., Drewe, J. & Beglinger, C. The
effect of cholecystokinin in controlling appetite and
food intake in humans. Peptides 22, 1265–1269
(2001).
67. Batterham, R. L. et al. Gut hormone PYY(3–36)
physiologically inhibits food intake. Nature 418,
650–654 (2002).
68. Drucker, D. J. Biologic actions and therapeutic
potential of the proglucagon-derived peptides.
Nat. Clin. Pract. Endocrinol. Metab. 1, 22–31 (2005).
69. Flanagan, D. E. et al. The influence of insulin on
circulating ghrelin. Am. J. Physiol. Endocrinol. Metab.
284, E313–E316 (2003).
70. Müller, T. D. et al. Ghrelin. Mol. Metab. 4, 437–460
(2015).
71. Nakazato, M. et al. A role for ghrelin in the central
regulation of feeding. Nature 409, 194–198 (2001).
72. Wynne, K. & Bloom, S. R. The role of oxyntomodulin
and peptide tyrosine-tyrosine (PYY) in appetite control.
Nat. Clin. Pract. Endocrinol. Metab. 2, 612–620
(2006).
73. Cowley, M. A. et al. Leptin activates anorexigenic
POMC neurons through a neural network in the
arcuate nucleus. Nature 411, 480–484 (2001).
74. Date, Y. et al. The role of the gastric afferent vagal
nerve in ghrelin-induced feeding and growth hormone
secretion in rats. Gastroenterology 123, 1120–1128
(2002).
75. Koda, S. The role of the vagal nerve in peripheral PYY3–
36‑induced feeding reduction in rats. Endocrinology
146, 2369–2375 (2005).
76. Cui, R. J., Li, X. & Appleyard, S. M. Ghrelin inhibits
visceral afferent activation of catecholamine neurons in
the solitary tract nucleus. J. Neurosci. 31, 3484–3492
(2011).
77. Rehfeld, J. F. Cholecystokinin as satiety signal.
Int. J. Obes. (Lond.) 5, 465–469 (1981).
78. Wang, J. D. & Levin, P. A. Metabolism, cell growth and
the bacterial cell cycle. Nat. Rev. Microbiol. 7,
822–827 (2009).
79. Rice, K. C. & Bayles, K. W. Molecular control of
bacterial death and lysis. Microbiol. Mol. Biol. Rev. 72,
85–109 (2008).
80. Keller, L. & Surette, M. G. Communication in bacteria:
an ecological and evolutionary perspective. Nat. Rev.
Microbiol. 4, 249–258 (2006).

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 23

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

81. Freter, R., Brickner, H., Fekete, J., Vickerman, M. M. &
Carey, K. E. Survival and implantation of Escherichia
coli in the intestinal tract. Infect. Immun. 39,
686–703 (1983).
82. Flint, H. J., Scott, K. P., Louis, P. & Duncan, S. H. The
role of the gut microbiota in nutrition and health. Nat.
Rev. Gastroenterol. Hepatol. 9, 577–589 (2012).
83. Allweiss, B., Dostal, J., Carey, K. E., Edwards, T. F. &
Freter, R. The role of chemotaxis in the ecology of
bacterial pathogens of mucosal surfaces. Nature 266,
448–450 (1977).
84. Stephen, A. M. & Cummings, J. H. The microbial
contribution to human faecal mass. J. Med. Microbiol.
13, 45–56 (1980).
85. Kotzé, S. et al. Spontaneous bacterial cell lysis and
biofilm formation in the colon of the Cape Dune mole-
rat and the laboratory rabbit. Appl. Microbiol.
Biotechnol. 90, 1773–1783 (2011).
86. Roager, H. M. et al. Colonic transit time is related to
bacterial metabolism and mucosal turnover in the gut.
Nat. Microbiol. 1, 16093 (2016).
87. Kau, A. L., Ahern, P. P., Griffin, N. W., Goodman, A. L.
& Gordon, J. I. Human nutrition, the gut microbiome
and the immune system. Nature 474, 327–336
(2011).
88. Kau, A. L. et al. Functional characterization of IgA-
targeted bacterial taxa from undernourished
Malawian children that produce diet-dependent
enteropathy. Sci. Transl Med 7, 276ra224 (2015).
89. Slack, E. et al. Innate and adaptive immunity
cooperate flexibly to maintain host-microbiota
mutualism. Science 325, 617–620 (2009).
90. Wu, G. D. et al. Linking long-term dietary patterns
with gut microbial enterotypes. Science 334,
105–108 (2011).
91. Faith, J. J. et al. The long-term stability of the human
gut microbiota. Science 341, 1237439 (2013).
92. Vrieze, A. et al. Impact of oral vancomycin on gut
microbiota, bile acid metabolism, and insulin
sensitivity. J. Hepatol. 60, 824–831 (2014).
93. Tyakht, A. V. et al. Human gut microbiota community
structures in urban and rural populations in Russia.
Nat. Commun. 4, 2469 (2013).
94. Yatsunenko, T. et al. Human gut microbiome viewed
across age and geography. Nature 486, 222–227
(2012).
95. Palmer, C., Bik, E. M., DiGiulio, D. B., Relman, D. A. &
Brown, P. O. Development of the human infant
intestinal microbiota. PLoS Biol. 5, e177 (2007).
96. Bolnick, D. I. et al. Individual diet has sex-dependent
effects on vertebrate gut microbiota. Nat. Commun. 5,
4500 (2014).
97. Pevsner-Fischer, M. et al. Role of the microbiome in
non-gastrointestinal cancers. World J. Clin. Oncol. 7,
200–213 (2016).
98. Haque, T. R. & Barritt, A. S. 4th. Intestinal microbiota
in liver disease. Best Pract. Res. Clin. Gastroenterol.
30, 133–142 (2016).
99. Aron-Wisnewsky, J. & Clement, K. The gut
microbiome, diet, and links to cardiometabolic and
chronic disorders. Nat. Rev. Nephrol. 12, 169–181
(2016).
100. Knip, M. & Siljander, H. The role of the intestinal
microbiota in type 1 diabetes mellitus. Nat. Rev.
Endocrinol. 12, 154–167 (2016).
101. Cotillard, A. et al. Dietary intervention impact on gut
microbial gene richness. Nature 500, 585–588
(2013).
102. Kleiman, S. C. et al. The intestinal microbiota in acute
anorexia nervosa and during renourishment:
relationship to depression, anxiety, and eating
disorder psychopathology. Psychosom. Med. 77,
969–981 (2015).
103. Woting, A. & Blaut, M. The intestinal microbiota in
metabolic disease. Nutrients 8, 202 (2016).
104. Furet, J. P. et al. Differential adaptation of human gut
microbiota to bariatric surgery-induced weight loss:
links with metabolic and low-grade inflammation
markers. Diabetes 59, 3049–3057 (2010).
105. Million, M. et al. Correlation between body mass
index and gut concentrations of Lactobacillus reuteri,
Bifidobacterium animalis, Methanobrevibacter smithii
and Escherichia coli. Int. J. Obes. (Lond.) 37,
1460–1466 (2013).
106. Monira, S. et al. Gut microbiota of healthy and
malnourished children in Bangladesh. Front.
Microbiol. 2, 228 (2011).
107. Everard, A. et al. Cross-talk between Akkermansia
muciniphila and intestinal epithelium controls diet-
induced obesity. Proc. Natl Acad. Sci. USA 110,
9066–9071 (2013).
108. Ley, R. E., Turnbaugh, P. J., Klein, S. & Gordon, J. I.
Microbial ecology: human gut microbes associated
with obesity. Nature 444, 1022–1023 (2006).
109. Walters, W. A., Xu, Z. & Knight, R. Meta-analyses of
human gut microbes associated with obesity and IBD.
FEBS Lett. 588, 4223–4233 (2014).
110. Vijay-Kumar, M. et al. Metabolic syndrome and
altered gut microbiota in mice lacking toll-like receptor
5. Science 328, 228–231 (2010).
111. Blanton, L. V. et al. Gut bacteria that prevent growth
impairments transmitted by microbiota from
malnourished children. Science http://dx.doi.
org/10.1126/science.aad3311 (2016).
112. Liou, A. P. et al. Conserved shifts in the gut microbiota
due to gastric bypass reduce host weight and
adiposity. Sci. Transl Med. 5, 178ra141 (2013).
113. Korem, T. et al. Growth dynamics of gut microbiota in
health and disease inferred from single metagenomic
samples. Science 349, 1101–1106 (2015).
114. Thaiss, Christoph, A. et al. Transkingdom control of
microbiota diurnal oscillations promotes metabolic
homeostasis. Cell 159, 514–529 (2014).
115. Zarrinpar, A., Chaix, A., Yooseph, S. & Panda, S. Diet
and feeding pattern affect the diurnal dynamics of the
gut microbiome. Cell Metab. 20, 1006–1017 (2014).
116. Liang, X., Bushman, F. D. & FitzGerald, G. A.
Rhythmicity of the intestinal microbiota is regulated
by gender and the host circadian clock. Proc. Natl
Acad. Sci. USA 112, 10479–10484 (2015).
117. Leone, V. et al. Effects of diurnal variation of gut
microbes and high-fat feeding on host circadian clock
function and metabolism. Cell Host Microbe 17,
681–689 (2015).
118. Thaiss, C. A., Zeevi, D., Levy, M., Segal, E. & Elinav, E.
A day in the life of the meta-organism: diurnal rhythms
of the intestinal microbiome and its host.
Gut Microbes 6, 137142 (2015).
119. Mukherji, A. et al. Shifting eating to the circadian rest
phase misaligns the peripheral clocks with the master
SCN clock and leads to a metabolic syndrome. Proc.
Natl Acad. Sci. USA 112, E6691–E6698 (2015).
120. Fetissov, S. O. & Meguid, M. M. Serotonin delivery
into the ventromedial nucleus of the hypothalamus
affects differently feeding pattern and body weight in
obese and lean Zucker rats. Appetite 54, 346–353
(2010).
121. Stunkard, A. J., Grace, W. J. & Wolff, H. G. The night-
eating syndrome: a pattern of food intake among
certain obese patients. Am. J. Med. 19, 78–86 (1955).
122. Rajpal, D. K. et al. Selective spectrum antibiotic
modulation of the gut microbiome in obesity and
diabetes rodent models. PLoS ONE 10, e0145499
(2016).
123. Cani, P. D. et al. Changes in gut microbiota control
metabolic endotoxemia-induced inflammation in high-
fat diet-induced obesity and diabetes in mice.
Diabetes 57, 1470–1481 (2008).
124. de Zwaan, M., Marschollek, M. & Allison, K. C. The
night eating syndrome (NES) in bariatric surgery
patients. Eur. Eat. Disord. Rev. 23, 426–434 (2015).
125. Tremaroli, V. et al. Roux‑en‑Y gastric bypass and
vertical banded gastroplasty induce long-term
changes on the human gut microbiome contributing to
fat mass regulation. Cell Metab. 22, 228–238
(2015).
126. Power, M. L. & Schulkin, J. Anticipatory physiological
regulation in feeding biology: cephalic phase
responses. Appetite 50, 194–206 (2008).
127. Hempfling, W. P. & Mainzer, S. E. Effects of varying the
carbon source limiting growth on yield and
maintenance characteristics of Escherichia coli in
continuous culture. J. Bacteriol. 123, 1076–1087
(1975).
128. Bäckhed, F. et al. The gut microbiota as an
environmental factor that regulates fat storage. Proc.
Natl Acad. Sci. USA 101, 15718–15723 (2004).
129. Yun, A. J., Lee, P. Y., Doux, J. D. & Conley, B. R. A
general theory of evolution based on energy efficiency:
its implications for diseases. Med. Hypoth. 66,
664–670 (2006).
130. Rabot, S. et al. Germ-free C57BL/6J mice are resistant
to high-fat-diet-induced insulin resistance and have
altered cholesterol metabolism. FASEB J. 24,
4948–4959 (2010).
131. Fleissner, C. K. et al. Absence of intestinal microbiota
does not protect mice from diet-induced obesity. Br.
J. Nutr. 104, 919–929 (2010).
132. Woting, A. et al. Alleviation of high fat diet-induced
obesity by oligofructose in gnotobiotic mice is
independent of presence of Bifidobacterium longum.
Mol. Nutr. Food Res. 59, 2267–2278 (2015).
133. Crenn, P. et al. Net digestive absorption and adaptive
hyperphagia in adult short bowel patients. Gut 53,
1279–1286 (2004).
134. Messing, B. et al. Intestinal absorption of free oral
hyperalimentation in the very short bowel syndrome.
Gastroenterology 100, 1502–1508 (1991).
135. Monteiro-Sepulveda, M. et al. Jejunal T cell
inflammation in human obesity correlates with
decreased enterocyte insulin signaling. Cell Metab.
22, 113–124 (2015).
136. Morita, C. et al. Gut dysbiosis in patients with
anorexia nervosa. PLoS ONE 10, e0145274 (2015).
137. Crawford, P. A. et al. Regulation of myocardial ketone
body metabolism by the gut microbiota during
nutrient deprivation. Proc. Natl Acad. Sci. USA 106,
11276–11281 (2009).
138. Labouré, H., Van Wymelbeke, V., Fantino, M. &
Nicolaidis, S. Behavioral, plasma, and calorimetric
changes related to food texture modification in men.
Am. J. Physiol. Regul. Integr. Comp. Physiol. 282,
R1501–R1511 (2002).
139. Anini, Y. et al. Comparison of the postprandial release
of peptide YY and proglucagon-derived peptides in the
rat. Pflügers Archiv. 438, 299–306 (1999).
140. Pavlov, I. P. The Work of the Digestive Glands (Charles
Griffin Co. Ltd, 1902).
141. Gerspach, A. C., Steinert, R. E., Schönenberger, L.,
Graber-Maier, A. & Beglinger, C. The role of the gut
sweet taste receptor in regulating GLP‑1, PYY, and
CCK release in humans. Am. J. Physiol. Endocrinol.
Metab. 301, E317–E325 (2011).
142. Furness, J. B., Rivera, L. R., Cho, H.‑J., Bravo, D. M. &
Callaghan, B. The gut as a sensory organ. Nat. Rev.
Gastroenterol. Hepatol. 10, 729–740 (2013).
143. Cox, H. M. et al. Peptide YY is critical for
acylethanolamine receptor Gpr119‑induced activation
of gastrointestinal mucosal responses. Cell Metab. 11,
532–542 (2010).
144. Perez-Burgos, A. et al. Psychoactive bacteria
Lactobacillus rhamnosus (JB‑1) elicits rapid frequency
facilitation in vagal afferents. Am. J. Physiol.
Gastrointest. Liver Physiol. 304, G211–G220 (2013).
145. Berthoud, H. R., Kressel, M., Raybould, H. E. &
Neuhuber, W. L. Vagal sensors in the rat duodenal
mucosa: distribution and structure as revealed by in
vivo DiI-tracing. Anat. Embryol. (Berl.) 191, 203–212
(1995).
146. Ghatei, M. A., Ratcliffe, B., Bloom, S. R. &
Goodlad, R. A. Fermentable dietary fibre, intestinal
microflora and plasma hormones in the rat. Clin. Sci.
(Lond.) 93, 109–112 (1997).
147. Camilleri, M., Madsen, K., Spiller, R., Van
Meerveld, B. G. & Verne, G. N. Intestinal barrier
function in health and gastrointestinal disease.
Neurogastroenterol. Motil. 24, 503–512 (2012).
148. Neunlist, M. et al. The digestive neuronal–glial–
epithelial unit: a new actor in gut health and disease.
Nat. Rev. Gastroenterol. Hepatol. 10, 90–100 (2012).
149. Hamilton, M. K., Boudry, G., Lemay, D. G. &
Raybould, H. E. Changes in intestinal barrier function
and gut microbiota in high-fat diet-fed rats are
dynamic and region dependent. Am. J. Physiol.
Gastrointest. Liver Physiol. 308, G840–G851 (2015).
150. Bruce-Keller, A. J. et al. Obese-type gut microbiota
induce neurobehavioral changes in the absence of
obesity. Biol. Psychiatry 77, 607–615 (2015).
151. Ukena, S. N. et al. Probiotic Escherichia coli Nissle
1917 inhibits leaky gut by enhancing mucosal
integrity. PLoS ONE 2, e1308 (2007).
152. Ewaschuk, J. B. et al. Secreted bioactive factors from
Bifidobacterium infantis enhance epithelial cell barrier
function. Am. J. Physiol. Gastrointest. Liver Physiol.
295, G1025–G1034 (2008).
153. Mariadason, J. M., Catto-Smith, A. & Gibson, P. R.
Modulation of distal colonic epithelial barrier function
by dietary fibre in normal rats. Gut 44, 394–399
(1999).
154. Wall, R. et al. in Microbial Endocrinology: The
Microbiota-Gut–Brain Axis in Health and Disease (eds
Lyte, M. & John Cryan, F.) 221–239 (Springer New
York, 2014).
155. Yano, J. M. et al. Indigenous bacteria from the gut
microbiota regulate host serotonin biosynthesis. Cell
161, 264–276 (2015).
156. Reigstad, C. S. et al. Gut microbes promote colonic
serotonin production through an effect of short-chain
fatty acids on enterochromaffin cells. FASEB J. 29,
1395–1403 (2015).
157. Lee, H. H., Molla, M. N., Cantor, C. R. & Collins, J. J.
Bacterial charity work leads to population-wide
resistance. Nature 467, 82–85 (2010).

24 | JANUARY 2017 | VOLUME 13 www.nature.com/nrendo

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

158. Chimerel, C. et al. Bacterial metabolite indole modulates
incretin secretion from intestinal enteroendocrine L cells.
Cell Rep. 9, 1202–1208 (2014).
159. Hubbard, T. D. et al. Adaptation of the human aryl
hydrocarbon receptor to sense microbiota-derived
indoles. Sci. Rep. 5, 12689 (2015).
160. Fetissov, S. O. et al. Expression of hypothalamic
neuropeptides after acute TCDD treatment and
distribution of Ah receptor repressor. Regul. Pept.
119, 113–124 (2004).
161. Bogunovic, M. et al. Enteroendocrine cells express
functional Toll-like receptors. Am. J. Physiol. Gastrointest.
Liver Physiol. 292, G1770–G1783 (2007).
162. Palazzo, M. et al. Activation of enteroendocrine cells
via TLRs induces hormone, chemokine, and defensin
secretion. J. Immunol. 178, 42964303 (2007).
163. Beutler, B. & Rietschel, E. T. Innate immune sensing
and its roots: the story of endotoxin. Nat. Rev.
Immunol. 3, 169–176 (2003).
164. Mukherji, A., Kobiita, A., Ye, T. & Chambon, P.
Homeostasis in intestinal epithelium is orchestrated
by the circadian clock and microbiota cues transduced
by TLRs. Cell 153, 812–827 (2013).
165. Zhu, X., He, L. & McCluskey, L. P. Ingestion of bacterial
lipopolysaccharide inhibits peripheral taste responses
to sucrose in mice. Neuroscience 258, 47–61 (2014).
166. Hosoi, T., Okuma, Y., Matsuda, T. & Nomura, Y. Novel
pathway for LPS-induced afferent vagus nerve
activation: possible role of nodose ganglion. Autonom.
Neurosci. 120, 104–107 (2005).
167. Bret-Dibat, J. L., Bluthe, R. M., Kent, S., Kelley, K. W.
& Dantzer, R. Lipopolysaccharide and interleukin‑1
depress food-motivated behavior in mice by a vagal-
mediated mechanism. Brain Behav. Immun. 9,
242–246 (1995).
168. Topping, D. L. & Clifton, P. M. Short-chain fatty acids
and human colonic function: roles of resistant starch
and nonstarch polysaccharides. Physiol. Rev. 81,
1031–1064 (2001).
169. Canfora, E. E., Jocken, J. W. & Blaak, E. E. Short-chain
fatty acids in control of body weight and insulin
sensitivity. Nat. Rev. Endocrinol. 11, 577–591 (2015).
170. Koh, A., De Vadder, F., Kovatcheva-Datchary, P. &
Bäckhed, F. From dietary fiber to host physiology:
short-chain fatty acids as key bacterial metabolites.
Cell 165, 1332–1345 (2016).
171. Chambers, E. S., Morrison, D. J. & Frost, G. Control of
appetite and energy intake by SCFA: what are the
potential underlying mechanisms? Proc. Nutr. Soc.
74, 328–336 (2015).
172. Chambers, E. S. et al. Effects of targeted delivery of
propionate to the human colon on appetite regulation,
body weight maintenance and adiposity in overweight
adults. Gut 64, 1744–1754 (2015).
173. Samuel, B. S. et al. Effects of the gut microbiota on
host adiposity are modulated by the short-chain fatty-
acid binding G protein-coupled receptor, Gpr41. Proc.
Natl Acad. Sci. USA 105, 16767–16772 (2008).
174. Wanders, A. J. et al. Effects of dietary fibre on
subjective appetite, energy intake and body weight: a
systematic review of randomized controlled trials.
Obes. Rev. 12, 724–739 (2011).
175. Cani, P. D. et al. Gut microbiota fermentation of
prebiotics increases satietogenic and incretin gut
peptide production with consequences for appetite
sensation and glucose response after a meal.
Am. J. Clin. Nutr. 90, 1236–1243 (2009).
176. Tennoune, N. et al. Bacterial ClpB heat-shock protein,
an antigen-mimetic of the anorexigenic peptide
α-MSH, at the origin of eating disorders. Transl
Psychiatry 4, e458 (2014).
177. Tennoune, N. et al. Sex-related effects of nutritional
supplementation of Escherichia coli: relevance to
eating disorders. Nutrition 31, 498–507 (2015).
178. Lee, S. et al. The structure of ClpB: a molecular
chaperone that rescues proteins from an aggregated
state. Cell 115, 229–240 (2003).
179. Haange, S. B. et al. Metaproteome analysis and
molecular genetics of rat intestinal microbiota reveals
section and localization resolved species distribution
and enzymatic functionalities. J. Proteome Res. 11,
5406–5417 (2012).
180. Panaro, B. L. et al. The melanocortin‑4 receptor is
expressed in enteroendocrine L cells and regulates the
release of peptide YY and glucagon-like peptide 1 in
vivo. Cell Metab. 20, 1018–1029 (2014).
181. Ericson, M. D., Schnell, S. M., Freeman, K. T. &
Haskell-Luevano, C. A fragment of the Escherichia coli
ClpB heat-shock protein is a micromolar melanocortin
1 receptor agonist. Bioorg. Med. Chem. Lett. 25,
5306–5308 (2015).
182. Maaser, C. et al. Crucial role of the melanocortin
receptor MC1R in experimental colitis. Gut 55,
1415–1422 (2006).
183. Nwokolo, C. U., Freshwater, D. A., O’Hare, P. &
Randeva, H. S. Plasma ghrelin following cure of
Helicobacter pylori. Gut 52, 637–640 (2003).
184. Lluch, J. et al. The characterization of novel tissue
microbiota using an optimized 16S metagenomic
sequencing pipeline. PLoS ONE 10, e0142334
(2015).
185. Amar, J. et al. Energy intake is associated with
endotoxemia in apparently healthy men. Am. J. Clin.
Nutr. 87, 1219–1223 (2008).
186. Harte, A. L. et al. High fat intake leads to acute
postprandial exposure to circulating endotoxin in
type 2 diabetic subjects. Diabetes Care 35, 375–382
(2012).
187. Huang, Q.‑H., Hruby, V. J. & Tatro, J. B. Role of central
melanocortins in endotoxin-induced anorexia. Am.
J. Physiol. Regul. Integr. Comp. Physiol. 276,
R864–R871 (1999).
188. Dwarkasing, J. T., Marks, D. L., Witkamp, R. F. & van
Norren, K. Hypothalamic inflammation and food
intake regulation during chronic illness. Peptides 77,
60–66 (2016).
189. Cani, P. D. et al. Endocannabinoids — at the
crossroads between the gut microbiota and host
metabolism. Nat. Rev. Endocrinol. 12, 133–143
(2016).
190. Di Marzo, V. et al. Leptin-regulated endocannabinoids
are involved in maintaining food intake. Nature 410,
822–825 (2001).
191. Jo, Y. H., Chen, Y. J., Chua, S. C. Jr, Talmage, D. A. &
Role, L. W. Integration of endocannabinoid and leptin
signaling in an appetite-related neural circuit. Neuron
48, 1055–1066 (2005).
192. Nicolaidis, S. Metabolic and humoral mechanisms of
feeding and genesis of the ATP/ADP/AMP concept.
Physiol. Behav. 104, 8–14 (2011).
193. Dzamko, N. L. & Steinberg, G. R. AMPK-dependent
hormonal regulation of whole-body energy
metabolism. Acta Physiol. 196, 115–127 (2009).
194. Colldén, G., Mangano, C. & Meister, B. P2X2
purinoreceptor protein in hypothalamic neurons
associated with the regulation of food intake.
Neuroscience 171, 62–78 (2010).
195. Zilberter, Y., Zilberter, T. & Bregestovski, P. Neuronal
activity in vitro and the in vivo reality: the role of
energy homeostasis. Trends Pharmacol. Sci. 31,
394–401 (2010).
196. Mächler, P. et al. In vivo evidence for a lactate
gradient from astrocytes to neurons. Cell Metab. 23,
94–102 (2016).
197. Le Blay, G., Michel, C., Blottière, H. M. & Cherbut, C.
Prolonged intake of fructo-oligosaccharides induces a
short-term elevation of lactic acid-producing bacteria
and a persistent increase in cecal butyrate in rats.
J. Nutr. 129, 2231–2235 (1999).
198. Langhans, W. Hepatic and intestinal handling of
metabolites during feeding in rats. Physiol. Behav. 49,
1203–1209 (1991).
199. Silberbauer, C. J., Surina-Baumgartner, D. M.,
Arnold, M. & Langhans, W. Prandial lactate infusion
inhibits spontaneous feeding in rats. Am. J. Physiol.
Regul. Integr. Comp. Physiol. 278, R646–R653
(2000).
200. Frost, G. et al. The short-chain fatty acid acetate
reduces appetite via a central homeostatic
mechanism. Nat. Commun. 5, 3611 (2014).
201. De Vadder, F. et al. Microbiota-generated metabolites
promote metabolic benefits via gut–brain neural
circuits. Cell 156, 84–96 (2014).
202. Breton, J. et al. Elevated plasma concentrations of
bacterial ClpB protein in patients with eating
disorders. Int. J. Eat. Disord. 49, 805–808
(2016).
203. Alexander, K. L., Targan, S. R. & Elson, C. O.
Microbiota activation and regulation of innate and
adaptive immunity. Immunol. Rev. 260, 206–220
(2014).
204. Garidou, L. et al. The gut microbiota regulates
intestinal CD4 T cells expressing RORγt and controls
metabolic disease. Cell Metab. 22,
100–112 (2015).
205. Lindner, C. et al. Diversification of memory B cells
drives the continuous adaptation of secretory
antibodies to gut microbiota. Nat. Immunol. 16,
880–888 (2015).
206. Christmann, B. S. et al. Human seroreactivity to gut
microbiota antigens. J. Allergy Clin. Immunol. 136,
1378–1386 (2015).
207. Mankarious, S. et al. The half-lives of IgG subclasses
and specific antibodies in patients with primary
immunodeficiency who are receiving intravenously
administered immunoglobulin. J. Lab. Clin. Med.
112, 634–640 (1988).
208. Fetissov, S. O. et al. Autoantibodies against
appetite-regulating peptide hormones and
neuropeptides: putative modulation by gut
microflora. Nutrition 24, 348–359 (2008).
209. Fetissov, S. O. et al. Autoantibodies against
neuropeptides are associated with psychological
traits in eating disorders. Proc. Natl Acad. Sci. USA
102, 14865–14870 (2005).
210. Karaiskos, D. et al. Psychopathological and
personality features in primary Sjogren’s syndrome
— associations with autoantibodies to
neuropeptides. Rheumatology 49, 1762–1769
(2010).
211. François, M. et al. Ghrelin-reactive immunoglobulins
and anxiety, depression and stress-induced cortisol
response in adolescents. The TRAILS study. Prog.
Neuropsychopharmacol. Biol. Psychiatry 59, 1–7
(2015).
212. Garcia, F. D. et al. Anti-neuropeptide Y plasma
immunoglobulins in relation to mood and appetite
in depressive disorder. Psychoneuroendocrinology
37, 1457–1467 (2012).
213. Lucas, N. et al. Effects of rabbit anti‑α‑melanocyte-
stimulating hormone (α‑MSH) immunoglobulins on
α‑MSH signaling related to food intake control.
Neuropeptides 48, 21–27 (2014).
214. Lucas, N. et al. Anti‑α‑melanocyte-stimulating
hormone autoantibodies in patients with eating
disorders and melanocortin 4 receptor signaling.
Eur. Neuropsychopharmacol. 24, S704–S705
(2014).
215. Takagi, K. et al. Anti-ghrelin immunoglobulins
modulate ghrelin stability and its orexigenic effect
in obese mice and humans. Nat. Commun. 4, 2685
(2013).
216. François, M. et al. High-fat diet increases ghrelin-
expressing cells in stomach, contributing to obesity.
Nutrition 32, 709–715 (2016).
217. Hamze Sinno, M. et al. Regulation of feeding and
anxiety by α‑MSH reactive autoantibodies.
Psychoneuroendocrinology 34, 140–149
(2009).
218. McFall-Ngai, M. et al. Animals in a bacterial world,
a new imperative for the life sciences. Proc. Natl
Acad. Sci. USA 110, 3229–3236 (2013).
219. Maffei, M. et al. Leptin levels in human and rodent:
measurement of plasma leptin and ob RNA in obese
and weight-reduced subjects. Nat. Med. 1,
1155–1161 (1995).
220. Everard, A. et al. Responses of gut microbiota and
glucose and lipid metabolism to prebiotics in
genetic obese and diet-induced leptin-resistant
mice. Diabetes 60, 2775–2786 (2011).
221. Alcock, J., Maley, C. C. & Aktipis, C. A. Is eating
behavior manipulated by the gastrointestinal
microbiota? Evolutionary pressures and potential
mechanisms. BioEssays 36, 940–949
(2014).
222. Thompson, J. A., Oliveira, R. A. & Xavier, K. B.
Chemical conversations in the gut microbiota.
Gut Microbes 7, 163–170 (2016).
223. Jacobi, C. A. et al. Quorum sensing in the probiotic
bacterium Escherichia coli Nissle 1917 (Mutaflor)
— evidence that furanosyl borate diester (AI‑2) is
influencing the cytokine expression in the DSS
colitis mouse model. Gut Pathog. 4, 1–10
(2012).
224. Wynendaele, E. et al. Quorum sensing peptides
selectively penetrate the blood–brain barrier. PLoS
ONE 10, e0142071 (2015).
225. Berg, R. D. The indigenous gastrointestinal
microflora. Trends Microbiol. 4, 430–435 (1996).
226. Chapelot, D., Aubert, R., Marmonier, C.,
Chabert, M. & Louis-Sylvestre, J. An endocrine and
metabolic definition of the intermeal interval in
humans: evidence for a role of leptin on the prandial
pattern through fatty acid disposal. Am. J. Clin.
Nutr. 72, 42–431 (2000).
Acknowledgements
The author is grateful to all colleagues who participated in
and inspired the reviewed research and to his wife being a
fervent and pertinent critic.
Competing interests
The author is a co‑founder of TargEDys.

NATURE REVIEWS | ENDOCRINOLOGY VOLUME 13 | JANUARY 2017 | 25

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.