liver indicating an increase in nutrient
absorption during a meal would allow
larger meals to be eaten but the subse-
quently higher than normal levels of
nutrients in the intestines and/or blood
would trigger other types of receptor and
maintain abnormally long inter-meal
intervals, thus leading to a maintenance of
normal daily intakes.
Another limitation to the acceptance of
an important role for the liver is the lack
of evidence of nerve endings in the
parenchyma of the liver; although such
anatomical evidence has been obtained for
the guinea-pig, it has not been forthcoming
for other species, despite careful searching
with light and electron microscopy.
However, Berthoud et al. (1992) recently
have found nerve endings only in the porta
hepatis in the rat (entrance of the portal
vein in to the liver) and not in the
parenchyma. It seems, therefore, that in
most species the supply of metabolites is
monitored at the entry to the liver rather
than throughout the whole liver. This
would account for the lack of effect of
infusions of propionate deep into the liver:
Leuvenink et al. (1997) found that while
infusion of propionate into the mesenteric
or hepatic portal vein of sheep both
stimulated insulin secretion, only the
former caused a reduction in food intake,
which explains why Anil and Forbes
(1980) found intake to be depressed when
they infused sodium propionate into the
mesenteric vein of sheep, while de Jong et
al. (1981) observed no significant depres-
sion in intake when similar rates of
infusion were given deep into the portal
area in goats. Hitherto this had been
ascribed to some mysterious difference
between the two species!
Infusions into the general circulationMayer envisaged that the CNS monitored
the concentration of glucose in the blood;
when it was below a certain threshold,
feeding started while when it was above a
threshold the meal was terminated. Several
pieces of evidence suggested that the
receptors responsible were in the ventro-
medial nuclei of the hypothalamus but,
while there might be hypothalamic sensi-
tivity to blood glucose concentration,
significant fluctuations in this concentra-
tion would mean that the body had failed
in its task of maintaining a relative
constancy of nutrient supply to the brain.
Subsequent to Mayer’s ‘glucostatic theory’
has been the realization that there are
receptors in intestines and liver that
normally prevent over- or undereating, and
the focus of attention has moved away
from the concentration of glucose in the
general circulation as a major factor in the
control of intake. However, an acute
deficiency of oxidizable substrate supply to
the brain is sensed, both by the hypo-
thalamus and, more importantly, by the
hindbrain, as evidenced by increased feed-
ing when these parts of the CNS are treated
locally with 2-deoxyglucose (2DG) which
blocks glucose transport into cells and
therefore mimics underfeeding.
If, despite the controls already reviewed
in the digestive tract and liver, blood con-
centrations of important metabolites are
excessively high or low, then they may be
sensed and used as emergency signals to
prevent extreme over- or undereating.Adipose tissueMetabolites in blood must be either
utilized or stored by tissues, such as
adipose and muscle, or else excreted by the
kidneys. In the adult animal, by far the
greater proportion of glucose and fatty acid
uptake, once maintenance requirements
have been met, is by adipose tissue. The
relative constancy of body weight suggests
that there is a signal from adipose tissue
which is integrated with those other
signals already mentioned and which alters
feeding behaviour in a subtle manner to
adjust for any changes in body fat content.
This is not to say that there is a fixed ‘set-
point’ for body fat content because there is
greater fatness in animals fed on diets with
high energy concentrations than in those
fed low-energy foods. Rather, body fat326 J.M. Forbes