Clinical Science (1995) 89, 375-381 (Printed in Great Britain)
375
Cholecystokinin is a satiety hormone in humans at
physiological post-prandial plasma concentrations
Anne BALLINGER, Lorraine McLOUGHLlN*, Sami MEDBAK* and Michael CLARK
Departments of Gastroenterology and *Chemical Endocrinology, The Medical College of
St Bartholomew's Hospital, London, U.K.
(Received 8 Marchj23 June 1995; accepted 29 June 1995)
1. Intravenous infusions of the brain/gut hormone,
cholecystokinin, have been shown to reduce food
intake in a subsequent test meal. However, in previous studies the doses administered were large and
likely to have produced plasma concentrations far in
excess of the normal post-prandial range.
2. In this study cholecystokinin-8 was infused intravenously to six healthy subjects in doses that reproduced physiological post-prandial concentrations.
Plasma concentrations of cholecystokinin were measured using a novel sensitive and specific radioimmunoassay. The effect of cholecystokinin-8 infusion on subsequent food intake in a standard test
meal was compared with the effect of saline infusion
in the same subjects.
3. Food intake (mean±SEM) was significantly less
during cholecystokinin (5092 ± 665 kJ) than during
saline infusion (6418 ± 723 kJ, P=0.03). During
cholecystokinin infusion, plasma concentrations
increased from 0.45± 0.06 pmol/I to 7.28± 2.43pmol/I
immediately before the meal. With saline infusion
there was no premeal increase in plasma cholecystokinin concentration.
4. This paper describes a novel radioimmunoassay
for measurement of plasma concentrations of cholecystokinin. Using this assay we have demonstrated
that cholecystokinin is important in control of satiety
in humans.
INTRODUCTION
Cholecystokinin (CCK) is present in the brain
and in the endocrine cells of the duodenum and
jejunum. It is released into the bloodstream. from
the gut in response to the presence of food 10 the
intestinal lumen. CCK is a key hormone in the
physiological processes that regulate food digestion
and absorption; it is the main mediator of postprandial gall bladder contraction and i~ stimulll:tes
pancreatic enzyme secretion, delays gastnc ~mpt~1Og
and regulates gastric acid secretion and intestinal
motility. In addition, several studies have shown
that exogenous administration of CCK reduces food
intake in a variety of species, including humans
[1-4]. However, it is unclear from these experiments
if the effects of CCK were pharmacological rather
than physiological since CCK was administered in
varying amounts and by different routes.
In an attempt to gain additional support for
CCK as a satiety peptide, food intake has been
measured in response to treatments that affect secretion of endogenous CCK. Intestinal CCK release is
thought to be mediated by a CCK-releasing peptide
secreted from the pancreas and possibly the duodenum [5]. This peptide is degraded by trypsin, and
oral administration of trypsin inhibitors increases
CCK release [6]. Administration of trypsin inhibitors before a test meal has been shown to reduce
subsequent food intake [7], presumably through
increased CCK release, although plasma concentrations of CCK were not measured in these
experiments. In previous work we have manipulated
endogenous release of CCK using an oral load
of L-phenylalanine, a potent releaser of CCK.
The study was designed such that L-phenylalanine
was administered before a test meal and plasma
CCK concentrations were already approaching
post-prandial levels when the meal was started.
We demonstrated that administration of Lphenylalanine caused release of CCK and was associated with a significant reduction in food intake
compared with placebo and D-phenylalanine, a
weak releaser of CCK [8]. These experiments are
thus indirect evidence that endogenous CCK is a
satiety peptide. In the present study we have
extended these findings and studied the effect of
exogenous CCK on food intake. However, unlike
the previous studies discussed above, the dose of
CCK was carefully adjusted to provide physiological post-prandial plasma concentrations. ..
It has been difficult to develop sensitive and
reliable RIAs for measurement of CCK in plasma.
Firstly, CCK shares a common C-terminal pentapeptide with gastrin and it is difficult to raise
specific antisera with no cross-reactivity with gas-
Key words: cholecystokinin, radioimmunoassay, satiety.
.
..
.
.
Abbreviations: CCK, cholecystokinin; DAB, 1,4-diaminobutane; G-I?s, sulphated gastrin-I?; G-I?ns, non-sulphated gastrin-I?; NTS, nucleus tractus solltarius; TFA, trlfluoroacetk
acid.
. I Ch arterhouse Square,
Correspondence:
Dr Anne Ballinger, Department of Gastroenterology, 4th Floor Science Block, The Medical College of St Bartholomew 'H
5 osprtat,
London EC IM6BQ, U.K.
376
A. Ballinger et al.
trin. Secondly, CCK circulates in very low concentrations and a sensitive assay is needed to detect
these levels. Finally, CCK is highly heterogeneous
and assays must be able to detect both small and
larger forms, all of which are biologically active. A
number of RIAs have been described for the
measurement of CCK, although very few antisera
[9-11] have been demonstrated to satisfy all these
requirements. A bioassay has also been described
for measurement of CCK [12]. This assay depends
on the ability of isolated rat pancreatic acini to
release amylase in response to CCK stimulation.
Th~ assay is sensitive and specific for CCK but,
unlike most RIAs, it is extremely limited in the
number of samples which can be processed in a
single assay. We have found that, if samples are
measured in triplicate, about 12 can be measured in
a single assay taking a whole day. In the present
study we have measured CCK using a recently
developed RIA that is sensitive and measures all the
biologically active forms. A degree of crossreactivity with gastrin has been overcome by an
extraction technique that differentially extracts CCK
and gastrin from plasma. This is a novel method
not previously described for measurement of CCK.
SUBJECTS AND METHODS
Eating studies
The study was approved by the Ethics Committee
of the City and Hackney Health Authority. Subjects
gave their informed consent before participating in
the study. In preliminary experiments plasma CCK
concentrations were measured in seven healthy subjects after eating a fatty meal. The meal had a total
energy value of 4701 kilojoules (kJ) and consisted of
1 pint of whole milk (1596 kJ), one cheese sandwich
(1992 kJ) and 50 g of chocolate (1113 kJ). Peak
plasma CCK concentrations in response to this
mea~ were ~.13 ± 2.1 pmol/l. Thus, in the eating
studies we aimed to reproduce this post-prandial
peak CCK response by an intravenous infusion
of CCK-8.
A randomized single-blind study was performed
in six subjects (mean age 31 years and 4 months).
All subjects (four men, two women) were within the
normal weight range (mean body mass index
23:2kg/m 2 , range 21.1-24.7) for their age, sex and
height,
Two infusions separated by at least 7 days were
performed in each subject. All the experiments were
performed at 20.00 hours after an 8-h fast from
midday, at which time subjects had eaten lunch of
their choice, and this was the same on each infusion
day.
In all experiments two intravenous catheters were
placed, one into each forearm. One catheter was
used to take blood samples at to-min intervals for
subsequent measurement of CCK. Through the
other catheter saline or synthetic sulphated CCK-8
Table I. Components of the test meal and their energy value
Component food
Energy value (kJ)
350 g savoury meat
500 g rice
Two slices of bread
Five slices of cake
Six squares of chocolate
One packet of crisps (28 g)
1536{IOOg
527jlOOg
314jslice
565jslice
I84jsquare
53 Ijpacket
(CCK-8s, Sincalide, Squibb Diagnostics) was infused
in random order. The infusion was continued until
the subject stopped eating. Five micrograms of
CCK was diluted in 100 ml of normal saline and the
infusion rate calculated to provide 40 ng h - 1 kg - 1 of
CCK-8. Preliminary experiments in four healthy
volunteers had demonstrated that this infusion rate
resulted in plasma CCK concentrations similar to
those seen after a meal. In the placebo experiments
the infusion rate was identical for saline. CCK-8
dissolved in saline appears as a clear colourless
solution which cannot be distinguished from saline;
thus, subjects were unable to tell which infusion
they were receiving.
Twenty minutes after the start of the infusion
subjects were given 200 ml of water to drink, and
5 min later subjects were presented with a standard
test meal of known energy content, far in excess of
the amount that they were likely to eat. A further
200 ml of water was given to be drunk with the
~eal. Fluid intake was strictly controlled as preVIOUS work has shown that gastric volume loads
potentiate and magnify the inhibition of food intake
produced by CCK infusions [4, 13].
The test meal (Table 1) was designed to include
foods enjoyed by all the volunteers and was presented on a buffet tray so that subjects were free to
eat as much as they liked. The components of the
meal were individually weighed before and after the
meal and therefore the amount (grams) consumed
could be calculated. Individual items in the meal
were all bought as easily prepared convenience
foods in which the energy content per 100 g was
clearly stated on the packaging. Therefore, from the
weight of food eaten, the number of kJ consumed
could be calculated. The subjects were free to eat for
as long as they wanted; the times at which subjects
started and finished eating was recorded. A 5-min
period followed after a subject indicated that he or
she had stopped eating. The food was removed at
the end of this 5-min period if no further food was
consumed.
Collection of plasma samples
Venous blood samples were collected into cold
lithium heparin tubes and kept on ice until centrifuge~ at 2000g at 4°C for to min. The plasma
obtained was transferred into plastic tubes containing 150/11 of glycine-hydrochloride (glycine-HCI;
Cholecystokinin and satiety
213 mmoljl glycine in 1mol/l hydrochloric acid) per
ml of plasma. Samples were flash frozen by placing
into liquid nitrogen and subsequently stored
at -20°e.
Extraction of plasma samples
Before extraction, plasma samples were allowed
to thaw after the addition of 10III of formic acid per
ml of plasma. Samples were centrifuged at 2000 g for
10min and the supernatant measured and transferred to a separate tube. Plasma was then extracted
and concentrated by adsorption onto octadecylsilylsilica cartridges (C- 18 Sep-pak, Millipore,
Hertfordshire, O.K.). Cartridges were prepared with
sequential washes of Sml of methanol-formic acid
(methanol-water-formic acid; S0:19:1, by vol.), Sml
of water-formic acid (99:1, v/v) and finally 5 ml of
saline-formic acid (154 mmoljl saline-glycine-HCIformic acid; 975:15:10 by vol.). After loading plasma
samples the cartridges were again washed with 5 ml
of saline-formic acid and Sml of water-formic acid
before eluting adsorbed peptides with 2 ml of
methanol-formic acid into a round-bottomed plastic
tube. The column was washed with a further S ml of
methanol-formic acid before re-equilibration with
water-formic acid to prevent contamination of the
next sample by the previous sample. Extracts were
dried under nitrogen at 50°C and then reconstituted
with 500 III of assay buffer (0.5 g of human serum
albumin in 100 ml of 0.05 mol/l phosphate pH 7.4)
containing Phenol Red as indicator. If required, the
pH was corrected to 7.4 by the addition of 1.1 mol/l
NaHC0 3 . Tubes were spun at 2000gfor lOmin and
400 III of supernatant removed and transferred to
two 3-ml tubes for immediate use in the assay. Thus
each tube contained 200 III of plasma extract in
assay buffer and plasma samples were therefore
assayed in duplicate.
Radioimmunoassay of CCK
CCK antibodies were obtained by immunization
of rabbits with sulphated CCK 26-33 (CCK-Ss,
Sigma, UK.) coupled to thyroglobulin. The highest
titre of antibodies was obtained from the second
rabbit after the fourth immunization (antiserum
CE2-4). The final titre of antiserum used in the
assay was 1:24000. CCK-Ss labelled with 125 1 by
the Bolton and Hunter method (Amersham International, Amersham, U.K.) was used as tracer with
lO00c.p.m. in each assay tube. CCK-Ss was used as
antigen and unextracted standard curves obtained
in each assay by dilution of CCK in assay buffer. In
addition, an extracted standard curve was obtained
after addition of CCK-S to 3 ml of 'CCK-free'
plasma (prepared by collecting the first eluate from
loading acidified plasma onto Sep-pak cartridges)
and extraction performed as described above.
The binding of antiserum CE2-4 to sulphated
gastrin-17 (G-17s), non-sulphated gastrin 17
(G 17ns), non-sulphated gastrin 34 (Sigma, U.K.),
CCK-33 and CCK-39 (a gift from V. Mutt,
Karolinska Institute, Stockholm, Sweden), nonsulphated CCK-S (CCK-Sns) and CCKl-21 (Sigma,
UK.) was determined by addition of peptide to
assay buffer. In addition, extracted standard curves
were obtained by addition of peptide to 3 ml of
CCK-free plasma. Binding of antiserum to CCK
fragments 30-33 (Sigma), 26-29 sulphated and 2629 non-sulphated (Peninsula Laboratories Europe)
was also determined in assay buffer.
Incubation was for a total of 96 h. A 50-Ill volume
of antiserum was added to the samples or CCK-S
standards and incubated for 24 h at 4°C, at which
time 50 III of tracer was added and incubation
continued for a further 24 h. Separation of bound
from free labelled CCK was achieved using a second
antibody. A 50-Ill volume of 1:200 normal rabbit
serum followed by 50 III of 1:12 donkey anti-rabbit
serum (IDS, Tyne and Wear, U.K.) was added to
samples and standards and incubation continued for
a further 4S h. At the end of this time tubes were
centrifuged and the supernatant removed. The pellet
containing the bound 125-1-labelled CCK was
counted in a gamma-counter.
Separation of forms of CCK by HPLC
Separation of plasma forms of CCK by HPLC
was performed according to the methods described
by Beardshall et al. [9]. A dynamax C 18 reversed
phase HPLC column (4.6 mm x 25 em, 51lm Microsorb particle size) was equilibrated with 'buffer A'
containing 0.1% aqueous trifluoroacetic acid (TFA)
and 0.01% 1,4-diaminobutane (DAB). The CCK
forms were eluted by increasing concentrations of
'buffer B' containing 75% acetonitrile in 0.1% TFA
and 0.01% DAB. The following gradient was used:
0-5min, 100% buffer A; 5-10min, 27% buffer B; 10SO min, 50% buffer B; So-II0 min, SO% buffer B; and
110-120 min, 100% buffer B. The elution rate was
I ml/min and I-ml fractions were collected into
tubes and dried under nitrogen at 50°C before
reconstituting with assay buffer for immediate use in
the assay. HPLC columns were calibrated with
sulphated CCK-8, CCK-33 and CCK-39.
Statistical analysis
Results are expressed as means ± SEM. Food
intake during CCK and saline infusion was compared using a paired t-test.
RESULTS
Radioimmunoassay of CCK
The detection limit of the assay, defined as the
smallest concentration of CCK in the assay tube
that could be differentiated from zero hormone
concentration with 95% confidence, was 2.25 pmoljl.
A. Ballinger et al.
378
60 , - - - - - - - - - - - - - - - - - - - - - - - - - - - ,
50
Table 2. Cross-reactivity with antiserum CE2-4 of CCK peptides,
CCK fragments and gastrin after extraction of plasma
Peptide
]40
.8
~ 30
J!
20
10
O'------J---'----L----'----L--'--"--"--L----'-----L---'---~_'___'__L_'------J__'___l
0.5
50
500
Concn. (pmoljl)
5000
50000
Fig. I. Radioimmunoassay of sulphated cholecystokinin (CCK-8s) in
assay buffer (.) and extracted from hormone-free plasma spiked
with CCK-8s (0). Also shown are serial double dilutions of sulphated
gastrin (G-17s) in assay buffer CA.) and extracted from plasma spiked with
G-17s (.). Recovery of CCK-8 after extraction ofplasma with Sep-pak CIS
cartridges was 77%. In contrast, there was very low recovery ofG-17.
Thus, when 3 ml of plasma was concentrated by
using Sep-pak cartridges, less than 1pmol/l CCK-8like immunoreactivity could be detected.
Recovery of CCK-8 added to hormone-free
plasma before extraction was 76.7% ± 1.5%. Recovery of CCK-33 and CCK-39 was 78% and 71%
respectively. Intra- and inter-assay coefficients of
variation were determined by 13-fold measurements
of extracted hormone-free plasma spiked with
3 pmol/l or 30 pmol/l CCK. Intra-assay variation
was 15% and 10% for samples containing 3 pmol/l
and 30 pmol/l respectively. The inter-assay variation
of samples containing 30 pmol/l was 13%. All the
samples were assayed in one run. Fifteen plasma
samples were extracted with each Sep-pak cartridge.
Repeated use of one cartridge did not result in any
carry-over or loss of activity.
CCK standard curves obtained in assay buffer
and hormone-free plasma were parallel (Fig. 1),
indicating that extracted plasma did not interfere in
the binding between peptide and antibody.
Compared with CCK-8s the cross-reactivity towards G-17s was 10% when standard curves were
obtained in assay buffer. However, cross-reactivity
was virtually eliminated when G-17s was added to
hormone-free plasma and samples extracted (Fig. 1).
This demonstrates that, using this method, G-17s is
poorly extracted from plasma and thus the antisera
can be used for the specific measurement of CCK in
plasma. Similarly, G-34 showed 13% cross-reactivity
when standard curves were obtained in assay buffer,
however no cross-reactivity was detected when
standard curves were obtained with extracted
plasma. Table 2 summarizes cross-reactivity of
various CCK forms, CCK fragments and gastrin as
compared with CCK-8.
Fig. 2 shows the HPLC profiles of CCK-like
immunoreactivity in post-prandial plasma. Antiserum CE2-4 measures four CCK forms; CCK-33
was the major form found in human plasma. The
peak eluting after CCK-39 was thought to be CCK-
CCK 26-33 sulphated
CCK 26-33 non-sulphated
CCK 33
CCK 39
CCK fragment 26-29 non-sulphated
CCK fragment 26-33 sulphated
CCK fragment 30-33
CCK 1-21
Gastrin-17 sulphated
Gastrin-17 non-sulphated
Gastrin-H non-sulphated
Potency relative to sulphated CCK-8 (%)
100
II
41
H
<0.2
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
Fig. 2. Elution profiles from analytical reversed phase HPLC of
CCK-like immunoreactivity (CCK-L1) concentratedfrom plasma by
Sep-pak C-18 cartridges after stimulation by a fatty meal. The HPLC
column was eluted by a gradient of 0.1% trifluoroacetic acid-acetonitrile
(
). Fractions of Iml were collected and CCK-Iike immunoreactivity
was measured with antiserum CE2-4.
10,----------------
-,
3
Subject
Fig. 3. Food intake (kJ) during a test meal in six subjects during
saline (.) or CCK-8 (0) infusion. The infusion was begun 2Smin
before the start of the test meal and continued throughout the meal. Five
minutes before the meal was presented subjects drank 200 ml of water.
Energy intake was significantly less (P = 0.03) during CCK infusion than
during saline infusion.
58 because of its similar position when compared
with the elution profile in a previous study [9].
Eating studies
Fig. 3 shows the energy intake in each subject
with saline or CCK-8 infusion. Energy intake was
Cholecystokinin and satiety
10 r - - - - - - - - - - - - - - - - - - - - - ,
8
0_
10
o
t
Infusion
10
20
Time (min)
t
30
Meal
Fig. 4. Plasma CCK concentrations duringsaline(.) or CCK-8 (0)
infusion in six healthy human subjects. The arrows indicate the times
that the infusion and then the meal were started. At 20 min (not shown)
subjects drank 200 ml of water.
significantly less during CCK (5092 ± 665 kJ) than
during saline infusion (6418 ± 723 kJ, P=0.03). Infusion of CCK did not specifically reduce intake of
one type or" food component. In two subjects ~he
energy intake during CCK infusion was less owmg
to a reduced intake of rice (predominantly carbohydrate containing); in two others the difference was
predominantly due to a reduced intake of meat
(59% of energy content as fat) and in one other
energy intake as fat and carbohydrate were reduced
equally.
Basal plasma CCK concentrations were similar
before CCK and saline infusion. During saline infusion plasma CCK concentrations did. not rise ~ntil
10min after beginning the meal. Dunng CCK infusion, plasma concentrations increased from
0.45 + 0.06 pmol/l at baseline to 7.28 ± 2.43pmol/l at
20min, i.e. immediately before the meal (Fig. 4) No
subject experienced any nausea or other adverse
effects during the CCK infusion. The mean time f<,>r
food consumption during CCK infusion was 18 rmn
(individual times 15, 16.5, 17.5, 19,20, 21 min). The
last blood sample was taken 15min after beginning
the meal, and therefore plasma concentrations of
CCK at the end of the meal were not measured.
However as determined from the individual meal
times the last blood sample was taken within 6 min
of ~al termination in all subjects. It is unlikely
that plasma concentrations of CCK would have
risen much above the physiological post-prandial
range during this last period.
DISCUSSION
The present study describes a sensitive. and spe~i­
fie radioimmunoassay for CCK. CCK circulates rn
plasma in multiple molecular forms. Biologically
active forms possess an identical C-terminal heptapeptide and a sulphated tyrosine at position 7.
Some workers have reported that the larger CCK
forms, e.g. CCK-58, are predominant in plasma
[14, 15], although other studies, including the mo~t
recent work, have found CCK-8 to be the predomi-
379
nant circulating form [16, 17]. Extraction of CCK
from the small intestine has been subject to experimental artefact [18], and this may also occur with
plasma extraction and account for some of the
differences in reported studies.
Antiserum CE2-4 binds to all the biologically
active forms but shows no binding to the smaller
non-biologically active forms. In addition, there is
no binding to CCK fragments that may arise during
breakdown of larger CCK forms. Antibodies
showed some binding with the non-sulphated form
of CCK-8, but this circulates in low concentrations
in plasma and thus is unlikely to interfere substantially with the present assay. Binding of antiserum
CE2-4 to CCK-58 has not been specifically tested as
neither the natural or synthetic peptide is freely
available. However, in other studies when testing
has been possible CCK-58 has been found to be far
less immunoreactive [19, 20] and bioactive than
CCK-8 [20]. This is thought to be because CCK-58
adopts a stable tertiary structure that interferes with
both antibody and receptor binding.
On the HPLC elution profile shown in this study
a peak was seen corresponding to the expected
position of CCK-58, and thus antiserum CE2-4 does
react although this cannot be quantified relative to
CCI(-8. Post-prandial plasma levels obtained using
this assay are similar to those measured by <,>u~­
selves and others using the bioassay [8, 12]. ThIS IS
supportive evidence that only the biologically active
forms of CCK are measured with this radioimmunoassay. After Sep-pak extraction plasma antibodies show virtually no cross-reaction with gastrin,
demonstrating that gastrin is poorly extracted from
plasma using this method.
We have shown that an infusion of CCK-8 that
reproduces post-prandial plasma concentrations significantly decreases food intake in a test ~eal
consisting of mixed food types. In two previous
studies an intravenous infusion of CCK-33 has been
given to reproduce post-prandial. concentr~ti<:>ns
[21, 22]. These studies have provided conflicting
results regarding the satiety effects of CCK.
In the first study a test meal was given to healthy
subjects and there was no reduction in food intake
during CCK infusion compared with placebo [21].
However, banana slices were used as the sole component of the test meal, and this may not represent
a physiological situation. Previous stu.dies have
shown that subjects will only eat a certain amount
of one food type, e.g. protein, but will eat an<:>th~r
food, e.g. carbohydrate, if offered [23]; this IS
referred to as sensory-specific satiety. The energy
intake in this previous study was only about half of
that usually consumed when a test meal of mixed
nutrients is given [7, 8, 24], and this observation
would also suggest that sensory-specific satiety was
a complicating factor.
In the second study the same group of workers
administered an identical dose of CCK-33 and using
visual analogue scales measured the effects of CCK
380
A. Ballinger et al.
on feelings of hunger, desire to eat, fullness and
prospective feeding intentions [22]. A test meal was
not given, but the authors found that CCK infusion
induced significant decreases in hunger feelings,
desire to eat and feeding intentions and that fullness
tended to be increased. The conclusions from this
study were thus in contrast to those of the first
study and the reasons for this are not clear.
In our study intravenous CCK was given with an
initial volume oral load of 200 ml of water. Previous
studies have suggested that gastric stimulation interacts with CCK to reduce food intake. Experiments
in both animals and humans have shown that
gastric loads potentiate and magnify the inhibition
of food intake produced by CCK infusions [4, 13].
A dose of CCK that failed to inhibit food intake
when given alone would suppress food intake when
given in combination with an intragastric volume
load. In rats both intragastric saline loads and
coeliac artery infusions of CCK-8 increase the firing
rate of afferent vagal fibres. In addition, CCK
pretreatment significantly enhances the response of
these fibres to subsequent gastric loads [25]. These
results suggest that CCK inhibits food intake, at
least in part, by its ability to mimic and amplify
vagal afferent responses to gastric distension.
The presence of CCK-containing neurons and
CCK receptors in the CNS has prompted investigators to question the role of central CCK in
control of eating. Experiments in animals suggest
that peripherally injected CCK does not cross the
blood-brain barrier and does not exert a direct
action in the CNS [26]. However, it seems likely
that there is a functional link between peripherally
and centrally CCK-activated neuronal circuits in the
control of satiety. The satiety effect of peripherally
administered CCK is blocked by severing the gastric
branches of the vagus nerve [27]. Gastric fibres of
the vagus terminate in the nucleus tractus solitarius
(NTS), lesions of which abolish the satiety action of
peripherally administered CCK [28]. Ascending
fibres which project from the NTS to the hypothalamus, amygdala, preoptic area and olfactory
bulbs have been identified as the essential projection
relaying the CCK signal. Finally, lesions of the
paraventricular nucleus of the hypothalamus, a key
area in the control of food intake and energy
balance, abolish the inhibition of feeding induced by
peripheral CCK [29]. Therefore, at least for exogenously administered peripheral CCK, animal
work suggests that the effects are mediated by both
peripheral and central CCK pathways.
In the present study CCK was measured for
15min post-prandially, and the peak plasma
concentration of CCK may well have occurred after
this time. The CCK infusion, together with endogenous CCK, may have produced venous plasma
values slightly above the normal post-prandial
range. However, all subjects finished eating within
21 min, and thus during this time CCK concentrations were likely to have been near to the normal
physiological range. Therefore, the results of this
experiment in which CCK has been administered to
reproduce physiological concentrations support the
results of previous work in which release of endogenous CCK has been manipulated. The present
study provides supportive evidence that CCK is
important in the control of satiety in humans.
ACKNOWLEDGMENTS
This study was supported by the National
Association for Crohn's and Colitis, the Magdalene
Hughes Endowment Fund and the Ernie Whitelaw
Research Fund. A.B.B. is supported by the Joint
Research Board of St Bartholomew's Hospital.
REFERENCES
I. Gibbs J. Young RC, Smith GP. Cholecystokinin decreases food intake in rats.
J Comp Physiol Psychol 1973; 84: 488-95.
2. Gibbs J, Falasco JD, McHugh PRo Cholecystokinin decreases food intake in
rhesus monkeys. Am J Physiol 1976; 230: 1),-18.
3. Kissilieff HR, Pi Sunyer FX, Thornton J, et al. C-terminal octapeptide of
cholecystokinin decreases food intake in man. Am J Clin Nutr 1981; 34: 154-60.
4. Schwartz GJ, Netterville LA, McHugh PR, Moran TH. Gastric loads potentiate
inhibition of food intake produced by a cholecystokinin analogue. Am J Physiol
1991; 261: RI141-6.
5. Miyasaka K, Guan D, Liddle RA, Green GM. Feedback regulation by trypsin:
evidence for intraluminal CCK-releasing peptide. Am J Physiol 1989; 257:
G17>-81.
6. Calam J, Bojarski JC, Springer q, et al. Raw soya-bean flour increases
cholecystokinin release in man. BrJ Nutr 1987; 58: 17),-9.
7. Hill AJ, Peikin SR, Ryan CA, Blundell JE. Oral administration of proteinase
inhibitor II from potatoes reduces energy intake in man. Physiol Behav 1990;
48: 241-6.
8. Ballinger AB, Clark ML. l-phenylalanine releases cholecystokinin and is
associated with reduced food intake in humans: evidence for a physiological
role of CCK in control of eating. Metabolism 1994; 43: 73>-8.
9. Beardshall K, Deprez P, Playford R, Alexander M, Calam J. Effect of
chymotrypsin on human cholecystokinin release: use of clostrtpain in the
validation of a new radioimmunoassay. Regul Peptides 1992; 40: 1-12.
10. Jansen JBM), Lamers CBHW. Radioimmunoassay of cholecystokinin in human
tissue and plasma. Clin Chim Acta 1983; 131: 30),-16.
II. Shafmayer A, Werner M, Becker HD. Radioimmunological determination of
cholecystokinin in tissue extracts. Digestion 1982; 24: 146-54.
12. Liddle RA, Goldfine ID, Rosen MS, Taplitz RA, Williams JA. Cholecystokinin
bioactivity in human plasma: molecular forms, responses to feeding and
relationship to gallbladder contraction. J Clin Invest 1985; 75: 1144-52.
13. Muurahainen NE, Kissileff HR, Lachaussee J, Pi-Sunyer FX. Effect of a soup
preload on reduction of food intake by cholecystokinin in humans.
Am J Physiol 1991; 260: R672-80.
14. Eberlein GA, Eysselein YE, Hesse WH, Goebell H, Schaefer M, Reeve JR.
Detection of cholecystokinin-58 in human blood by inhibition of degradation.
Am J Physiol 1987; 253: G477-82.
15. Eysselein YE, Eberlein GA, Hesse WH, et al. Molecular variants of
cholecystokinin after endogenous stimulation in humans: a time study.
Am J Physiol 1990; 258: G951-7.
16. Cantor P, Rehfeld JF. The molecular nature of cholecystokinin in human
plasma. Clin Chim Acta 1987; 168: 153-8.
17. Pohl M, Seebeck J, Nustede R, Mutt Y, Creutzfeldt W, Schmidt WE. Molecular
forms of cholecystokinin - characterization of their relative bioactivities,
immunoreactivities and distribution in human plasma. Gastroenterology 1994;
106: A127.
18. Turkelson CM, Solomon TE. Molecular forms of cholecystokinin in rat
intestine. Am J Physiol 1990; 259: G364-71.
19. Turkelson CM, Reeve JR, Solomon TE. Low immunoreactivity of canine
cholecystokinin-58. Gastroenterology 1990; 99: 646-51.
20. Hocker M, Schmidt WE, Wilms HM, et al. Measurement of tissue
cholecystokinin concentration by bioassay and specific radioimmunoassay:
characterization of the bioactivity of CCK-58 before and after tryptic cleavage.
Eur J Clin Invest 1990; 20: S4).-50.
Cholecystokinin and satiety
21. Lieverse RJ, Jansen ]BM], van de Zwan A, Samson L, Masclee AAM, Lamers
CBHW. Effects of a physiological dose of cholecystokinin on food intake and
postprandial satiation in man. Regul Peptides 1993; 43: 83-9.
22. Lieverse RJ, Jansen JBMJ, Masclee AAM, Lamers CBHW. Satiety effects of
cholecystokinin in humans. Gastroenterology 1994; 106: 1451--4.
23. Rolls BJ, Rolls ET, Rowe EA, Sweeney K. Sensory specific satiety in man.
Physiol Behav 1981; 27: 137--42.
24. Silverstone T, Goodall E. The clinical pharmacology of appetite suppressant
drugs. Int JObes 1984; 8: 22-33.
25. Schwartz GJ, McHugh PR, Moran TH. Integration of vagal afferent responses
to gastric loads and cholecystokinin in rats. Am J Physiol 1991; 261: R6+-9.
381
26. Passaro E, Debas H, Oldendorf W, Yamada T. Rapid appearance of
intraventricularly administered neuropeptides in the peripheral circulation.
Brain Res 1982; 241: 335-40.
27. Smith GP, Jerome C, Gibbs J. Abdominal vagotomy blocks the satiety effect
of cholecystokinin in the rat. Science (Washington DC) 1981; 213:
103&-7.
28. Crawley IN, Schwaber JS. Abolition of the behavioral effects of cholecystokinin
following bilateral radiofrequency lesions of the parvocellular subdivision of the
nucleus tractus solitarius. Brain Res 1984; 295: 289-99.
29. Crawley IN. Kiss ]Z. Paraventricular nucleus lesions abolish the inhibition of
feeding induced by systemic cholecystokinin. Peptides 1985; 6: 927-35.