lations could dysregulate AgRP neurons
and cause long-term metabolic changes.
An alternative hypothesis is that the
olfactory system has direct effects on en-
ergy metabolism independent of its func-
tion in odor detection. It has long been
noted that olfactory structures express
an unusually high density of receptors
for metabolic hormones, such as insulin,
leptin, and ghrelin, and further that the
olfactory bulb has enhanced access to the
circulation resulting from an incomplete
blood-brain barrier (11). Although these
observations are often cited to explain the
fact that food deprivation can increase ol-
factory sensitivity, it is also possible that
the presence of these metabolic receptors
signals a broader function for olfactory tis-
sues in energy homeostasis. In this regard,
recent studies have identified several olfac-
tory receptors in mice that “moonlight” as
metabolic sensors in the periphery. These
include OLFR558, which senses isovalerate
in the intestine (12), and OLFR78, which
senses lactate in the carotid body (13). It
is conceivable that evolution could have
co-opted the olfactory system, with its ex-
traordinary ability to detect complex pat-
terns of chemosensory signals in the form
of odorants, to monitor internal nutrients
in a similar way (see the figure).
These speculations aside, a number of
basic questions about the mechanisms un-
derlying the connection between olfaction
and metabolism remain to be addressed.
For example, it is unclear whether the
metabolic responses to manipulations of
the olfactory system (2, 3) are due to spe-
cific changes in the sensing of food odors,
and, if so, which food odors are involved.
It is likewise unknown how detection of
food odors is communicated from the ol-
factory bulb to the hypothalamic circuits
that regulate energy balance, and what
specific information these signals convey
(e.g., nutrient content, toxicity, availabil-
ity). In this regard, odor information re-
ceived in the olfactory bulb is transmitted
along two anatomically distinct pathways:
One mediates learned responses to odors
(via the piriform cortex) and the other
controls innate behaviors (via the cortical
amygdala). The relative contribution of
these two pathways to the modulation of
hypothalamic circuits by the smell of food
has not been tested, but a role for the in-
nate pathway would suggest that certain
food odors are evolutionarily hardwired to
elicit physiologic responses. If so, it would
be fascinating to know why these odorants
have been selected by evolution to act as
A final question regards the extent
to which these observations in rodents
and invertebrates apply to human biology. Compared to mice, humans express
fewer functional odorant receptor genes
(1100 in mice versus 390 in humans) and
have a proportionally smaller olfactory
bulb. Humans also lack a functional vom-eronasal organ, an accessory olfactory
structure used by many animals to detect
pheromones (14). It has been speculated
that these changes were driven in part by
the evolution of color vision in primates,
which may have enabled the visual system
to co-opt functions previously performed
by olfaction. In this context, is it plausible
that olfactory responses in mice would
be conserved in humans? Although it is
true that humans rely less on smell than
many animals, the commonly held view
that human olfaction is feeble or in some
sense vestigial is based more on folklore
than scientific evidence (14). Indeed, smell
plays a critical role in the perception of
food flavor in humans, and decreased enjoyment of food is a common disturbance
in people suffering from olfactory deficits
(15). Whether such individuals also experience metabolic impairments independent
of changes in food intake is unknown, but
recent work in rodents suggests this question is worth investigating. j
REFERENCES AND NOTES
1. I.P.Pavlov, The Work of the Digestive Glands (Charles
Griffin, London, 1902).
2. C. E. Riera et al ., Cell Metab.26, 198 (2017).
3. K. Tucker,J.M.Overton,D.A.Fadool, J.Neuroendocrinol.
24, 1087 (2012).
4. K.Nagai et al., Auton. Neurosci. 185,29(2014).
5. S. Libert etal .,Science 315, 1133 (2007).
6. J. Alcedo, C. Kenyon, Neuron 41, 45 (2004).
7. E. D. Smith et al ., BMC Dev. Biol. 8, 49 (2008).
8. Y. Chen et al., Cell 160, 829 (2015).
9. J. N. Betley et al ., Nature 521, 180 (2015).
10. Y. Mandelblat-Cerf et al ., eLife 4, e07122 (2015).
11. A.K.Julliard et al., Front. Physiol. 8,468(2017).
12. N. W.Bellono et al., Cell 170,185(2017).
13. A. J. Chang et al ., Nature 527, 240 (2015).
14. J. P. Mc Gann, Science 356, eaam7263 (2017).
15. I.Croy, S.Nordin, T.Hummel, Chem. Senses 39,185
Our work is supported by grants from the American Federation
for Aging Research to J.L.G.; the New York Stem Cell Foundation,
Rita Allen Foundation, and American Diabetes Association
to Z.A.K.; and the U.S. National Institutes of Health to J.L.G.
(R35GM11982) and Z.A.K. (DP2DK10953, R01NS094781,
R01DK106399). We thank J. Carlson for discussions.
Hormones, metabolites Blood vessel
peptide 1 receptor
Olfactory sensory neuron
10 NOVEMBER 2017 • VOL 358 ISSUE 6364 719
Connecting olfaction to metabolism
Detection of food odors by sensory neurons in the olfactory epithelium is rapidly communicated to
hypothalamic neurons that control hunger, thereby modulating whole-organism metabolism. Neurons in the
olfactory epithelium and bulb can also monitor hormones and nutrients in the blood, which may enable the
olfactory system to regulate energy balance independently of odor perception.