| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Veterans Administration Puget Sound Health Care System and University of Washington (T.W.-H., D.R.M., R.L.M., C.W.W., A.M.M., D.D.R.), Seattle, Washington 98195; and Loma Linda University (S.M.Y.), Loma Linda, California 92350
Address all correspondence and requests for reprints to: Dr. Dennis D. Rasmussen, Research Service (151), Veterans Administration Medical Center, American Lake, Tacoma, Washington 98493. E-mail: drasmuss{at}u.washington.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adiposity, especially visceral adiposity, increases with advancing age in humans (4), nonhuman primates (5, 6), and rats (7), as do plasma insulin and leptin levels. These changes in adiposity, insulin, and leptin levels are often associated with detrimental metabolic consequences, such as glucose intolerance, insulin resistance, diabetes, dyslipidemia, and hypertension (4, 5, 8). We have previously demonstrated that daily nocturnal melatonin administration to middle-aged male rats suppressed intraabdominal (commonly referred to as deep abdominal or visceral) fat, plasma leptin, and plasma insulin to youthful levels (9), suggesting that appropriate melatonin supplementation may potentially reverse some of the aging-associated metabolic changes. Melatonin administration did not alter the aging-related decline in plasma testosterone, T3, or insulin-like growth factor I concentrations, nor were plasma T4 levels significantly altered, suggesting that the actions of melatonin occurred independent of marked changes in gonadal, thyroid, or somatotropin regulation.
The current study was designed to begin investigating mechanisms that may mediate these responses, focusing on the effects of melatonin on food intake and energy expenditure (physical activity and body core temperature) as well as on body composition. In a subset of subjects, a cross-over design also allowed us to assess the reversibility of the melatonin-induced weight loss. Preliminary behavioral tests evaluated the possibility that the taste of melatonin in the drinking water was detectable or that effects of melatonin might induce a conditioned taste aversion.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two-bottle choice test
The presence of novel tastes can alter ingestive behavior in
rats. To test whether the rats could detect concentrations of melatonin
added to the drinking water in the current study, we performed a
two-bottle choice test. On 2 sequential days, nine individually housed
male rats (6 months of age) that had not previously received melatonin
treatment were each provided with two identical Liquidiet Feeding Tubes
(BioServ, Frenchtown, NJ) placed side by side after removal of the
standard cage-top water bottle. One tube contained 0.01% ethanol
vehicle in water, and the other contained 0.01% ethanol vehicle and
0.4 µg melatonin/ml water. On the first day, five of the nine rats
were presented melatonin solution in the drinking tube on the left, and
the remaining four rats were presented melatonin in the tube on the
right. On the second day, the left vs. right presentation of
melatonin vs. vehicle control solutions was reversed for
each rat. The drinking tubes were weighed at the start and end of both
24-h periods to determine consumption, and clean tubes and freshly
prepared solutions were provided each day. Consumption of
melatonin-containing water and vehicle control water on the 2 days were
each averaged for each rat to provide indexes of consumption of
melatonin-containing vs. vehicle control water independent
of bottle position.
Alternating days consumption test
To further resolve whether melatonin acutely altered water
consumption, we also performed an alternating days consumption test.
Middle-aged (10 months of age; n = 24) male rats that had not
previously received any melatonin treatment were housed two per cage;
each cage was provided water in a standard cage-top bottle equipped
with a ball-bearing sipper tube as the sole source of fluid. On the
first day of a 3-day test, the water bottle for each cage contained
0.01% ethanol vehicle. After 24 h, this water bottle was
exchanged for one containing 0.01% ethanol and 0.4 µg melatonin/ml
water, and after another 24 h, this water bottle was again
replaced with one containing only 0.01% ethanol vehicle. The fluid
consumption on each of these 3 sequential days was determined by
weighing the water bottles at the start and finish of each 24-h period.
Clean water bottles and freshly prepared solutions were provided each
day.
Conditioned saccharin taste aversion test
When consumption of a novel taste (conditioned stimulus) is
paired with exposure to a toxin, rats will avoid consumption of that
taste in the future. This learned response, known as conditioned taste
aversion, can be used to determine whether a treatment (i.e.
the unconditioned stimulus) is perceived as toxic. A 6-day, single
pairing, two-bottle saccharin taste-conditioning procedure (10) was
slightly modified for this study. The taste-conditioning apparatus
consisted of a standard cage with standard cage-top bottles with
ball-bearing sipper tubes containing water, 0.15% saccharin solution,
or both; water and saccharin consumption in the conditioning apparatus
was quantitated by weighing the bottles before and after each
conditioning or test trial. Individually housed middle-aged male rats
(10 months of age; n = 24) that had not previously received any
melatonin treatment were maintained under reversed light/dark
conditions (lights off at 0800 h and on at 1800 h) for 3
weeks before and then throughout the experiment. On day 1, the colony
water bottles were removed from the home cages at 0900 h. On day
2, the water-deprived animals were each placed individually in a
taste-conditioning apparatus at 0900 h and allowed 15-min
access to drinking water. All rats were then returned to their home
cages, and colony water bottles were replaced. The amount of water
consumed during the 15-min access period on this day provided the basis
for subsequently distributing the rats into two groups (a and b; n
= 12/group; see below) with comparable initial water consumption. On
day 3, the colony water bottles were again removed from the home cages
at 0900 h. On day 4, after this second 24-h water deprivation, the
rats were again each placed individually in a taste-conditioning
apparatus at 0900 h and allowed 15-min access to 0.15% saccharin
solution containing either 0.01% ethanol control vehicle (group a) or
0.01% ethanol and 0.4 µg melatonin/ml (group b). All rats were then
returned to their home cages equipped with water bottles. On day 5, the
colony water bottles were again removed from the home cages at
0900 h. On day 6, after a third 24-h water deprivation, the rats
were again each placed individually in a taste-conditioning apparatus
at 0900 h and presented with a 15-min choice between two identical
bottles, counterbalanced for position upon the cage top, containing
either control vehicle in water or a 0.15% saccharin solution in water
containing vehicle. The measure of taste conditioning was the
difference score between saccharin solution intake on day 6 after the
pairing of melatonin/vehicle with saccharin vs. day 4 before
the pairing. For example, if melatonin ingestion produced effects
perceived as toxic, saccharin ingestion on day 6 would be decreased
more in those rats for which saccharin taste was previously paired with
melatonin than in those rats for which saccharin had been paired with
vehicle alone.
Study design
The principal experiment in the current investigation employed a
cross-over design for a subset of the subjects. Before the start of the
experiment, 37 individually housed middle-aged (10 months) male Sprague
Dawley rats were weighed and divided into 2 groups of 18–19 each,
maintaining similar distribution of weights between the 2 groups. Both
groups were switched from pelleted to powdered food (Laboratory Rodent
Diet 5001, powdered formulation, Ralston Purina Co.) 2
weeks before initiating the experiment and then maintained on powdered
food throughout. At the start of the experiment, one of the groups
(MELATONIN) began receiving melatonin (0.4 µg/ml) in ethanol vehicle
(0.01%) in their daily drinking water, whereas the other group
(CONTROL) received vehicle alone. An additional group of 18 young rats
(YOUNG; 3 months of age at the start of the experiment) was
individually housed in the same room as the middle-aged rats and
received powdered food and ethanol vehicle in the drinking water. After
12 weeks of treatment, 9 rats from each of the 3 groups (YOUNG,
MELATONIN, and CONTROL) were killed. At this point the remaining 9
young rats (6 months of age) were removed from the study. At week 14,
the treatments for the remaining middle-aged MELATONIN and CONTROL
groups were crossed over, i.e. the rats that had received
melatonin during the first 14 weeks then received control treatment
(0.01% ethanol vehicle in their drinking water) and the rats that had
received control treatment then received melatonin treatment (0.4
µg/ml melatonin and 0.01% ethanol). After an additional 12 weeks,
these remaining middle-aged rats were killed.
Melatonin administration
Melatonin (Sigma, St. Louis, MO) was dissolved in
100% ethanol and stored in aliquots at -70 C. Fresh solutions of
drinking water containing 0.01% ethanol with or without 0.4 µg
melatonin/ml were prepared twice weekly. The water bottles were covered
with aluminum foil; clean bottles with fresh solutions were provided
twice weekly.
In a recent study using this same model of melatonin administration via drinking water, we determined that both control and melatonin-treated middle-aged male Sprague Dawley rats maintained in a 14-h light, 10-h dark cycle in our facility (i.e. animals and conditions essentially identical to the current study) drank more than 90% of their total daily water during the dark period (9). Assuming that the rats in the current study exhibited a comparable pattern and drank a consistent amount during each hour of the dark period, melatonin consumption in the current study (14–19 µg melatonin/day; 30–35 µg melatonin/kg BW·day) would correspond to approximately 1.3–1.7 µg/h (2.8–3.2 µg/kg BW·h) through the 10-h night.
Body weights (BW)
All rats were weighed twice weekly, on Mondays and Fridays,
between 0900–1100 h. These two weights were averaged to yield a weekly
weight for each animal.
Food consumption
Each rat was provided with approximately 40 g fresh
powdered chow between 1100–1200 h daily in a 6-oz powdered food feeder
jar (PFF6D, Allentown Caging, Allentown, NJ) secured in a stainless
steel hanging jar holder (HJH6D, Allentown Caging) modified by addition
of a stainless steel spring that anchored the feeder jar to the hanging
jar holder (preventing it from being tilted or dislodged).
Approximately 10–15 g food remained when the feeder jar was removed
and another jar provided at the same time on the subsequent day.
Providing only slightly more food than that consumed each day reduced
the amount of food in the jar (i.e. the jar was well under
half full when initially provided each day), minimizing and usually
eliminating spillage. Food consumption was measured by weighing the jar
plus food before feeding and then subtracting the weight of the jar
plus food after 24-h consumption on Mondays, Wednesdays, and Fridays.
These three values were averaged as an index of daily consumption for
the week. Data from days when spillage was detectable were not included
(this occurred on <2% of days of monitored food consumption).
Cumulative food intake was calculated by multiplying average daily food
intake by 7, then summing the calculated weekly averages throughout the
treatment period. Relative food intake was calculated as grams of daily
food intake per g BW. Feed efficiency was calculated as grams of BW
change per g food eaten throughout that time period. Bulk powdered chow
was stored in a room temperature covered bin to minimize the
variability in weight of daily aliquots due to condensation and/or
evaporation that can occur with refrigerated storage.
Locomotor activity
Home cage spontaneous ambulatory locomotor activity was measured
by sequential infrared beam breaks using an Opto-Varimex Mini animal
activity monitoring system (Columbus Instruments, Columbus, OH). In
this system, an array of 15 infrared emitters is placed along 1 side of
the home cage, and an array of 15 receivers is placed along the other;
movement of the rat interrupts the infrared beams, generating digital
signals that are tabulated by computer. For this study only sequential
breaks of different beams (primarily corresponding to ambulatory
locomotion) were tabulated, excluding sequential repetitive breaks of
the same beam (commonly associated with grooming or eating). Activity
was monitored continuously during the last 4 h of the light
period, the entire 10-h dark period, and the first 6 h of the next
light period, for a total of 10 h of light and 10 h of
darkness. One hour before and then throughout this time, the colony
room door remained closed, and the rats were not disturbed. Two rats
were independently monitored each day during weeks 7–10 of
treatment.
Core body temperature
Body temperature was measured with a digital Precision
thermometer (model 4600) with 400 series small semiflexible vinyl
rectal/esophageal Thermistor probe (YSI, Inc., Yellow
Springs, OH), providing 0.01 C resolution and a time constant of 1.4
sec. Each rat was removed from the cage at 1000–1100 h and held on an
investigator’s arm; another investigator inserted the lubricated
Thermistor probe 5 cm into the colon, and temperature was recorded
after 30 sec. The measurement caused no apparent distress and was
completed less than 45 sec after removing the rat from the home cage.
The temperature of each rat was determined once per week during weeks
7–9 of the experimental trial, alternating the order of determinations
in sequential weeks; the three estimates for each rat were averaged to
provide a single value for core body temperature.
Tail blood collections
On each of the 3 days preceding blood sample collection, each
rat was transferred in its home cage to the adjoining blood-sampling
room, exposed to a mock blood-sampling procedure, and returned to the
colony. On the day of sample collection, the distal tip (<1 mm) of the
tail was snipped with sharp surgical scissors, and blood was gently
"milked" into a heparinized capillary tube; plasma was then
separated by centrifugation and stored at -70 C. Spontaneous bleeding
stopped immediately after sample collection; subsequent removal of the
coagulate allowed an additional sample to be collected on the same day
without an additional cut. The rats were unrestrained; they remained in
their home cages throughout the blood collection procedures and did not
appear distressed.
Collection of 350-µl blood samples for characterization of plasma melatonin levels at the midpoint of the light period and at the midpoint of the dark period was performed during week 6. The samples at the midpoint of the light period and those at the midpoint of the dark period were both collected on the same day, with the light period sample collected first. For samples collected during the dark period, lights remained off in the colony room; each rat was exposed to dim light (60–100 lux) in the blood-sampling room for less than 3 min.
Collection of 40-µl blood samples for characterization of plasma corticosterone levels in the early morning (1–2 h after lights on) and late afternoon (0–1 h before lights off) was performed during week 10 of the cross-over study. The early morning and late afternoon samples were both collected on the same day, with the morning sample collected first.
Decapitation, preparation for DEXA
The rats were killed by decapitation 2–3 h after the start of
the light period; blood was collected, and plasma was stored at -70 C.
The carcasses were rapidly frozen, ventral side down, on large blocks
of dry ice and buried under powdered dry ice, then stored in sealed
plastic storage bags at -20 C. The carcasses were positioned during
freezing to provide consistent dual energy x-ray absorptiometry (DEXA)
analysis (see below); the front legs were extended rostrally, the rear
legs were extended caudally, and the tail was turned rostrally to lie
along the body.
Body composition and adiposity measurements
Body composition (fat and lean mass) was measured in frozen rat
carcasses by DEXA (QDR 1500, Hologic, Inc., Waltham, MA),
using the Rat Whole Body software package (version 5.67, Hologic, Inc.). DEXA provides a reliable method for quantification of rat
whole body composition, as we (7) and others (11, 12) have previously
described. The system is based on the differential attenuation of low
(70 keV) and high (140 keV) energy x-rays by the tissues in the scan
area. Fat mass detected by DEXA consists primarily of adipose tissue,
but lean mass includes organs, tendons, cartilage, blood, and body
water in addition to skeletal muscle. A tissue calibration scan, using
calibration phantoms, was performed each half-day, and measurement of a
separate frozen rat carcass was repeated each day as a control for
scanning variations. Intrascan coefficients of variation were 5.0% for
fat mass and 0.6% for lean mass. The interscan coefficients of
variation for fat and lean masses were 6.0% and 1.0%, respectively.
The sum of individual animals’ fat mass plus lean mass plus bone
mineral content was highly correlated (r = 0.999) with direct
measurement of BW.
After DEXA, the carcasses were thawed to 4 C, and all intraabdominal fat depots were dissected. An incision was made at the ventral midline, exposing the abdominal cavity, and the gastrointestinal tract was severed above the stomach and at the base of the colon and removed, exposing the internal fat pads. The omental fat depot was carefully separated from the pancreas, and the mesenteric fat depot was removed from the length of the gut, with blood vessels and connective tissue included. Paired retroperitoneal, perirenal, and epididymal fat pads were then removed, with kidneys, adrenals, testes, and epididymis dissected away. The fat pads were weighed immediately after dissection to avoid evaporative weight loss. Estimates of intraabdominal and peripheral fat were calculated by summing intraabdominal fat pad (omental plus mesenteric plus retroperitoneal plus perirenal plus epididymal) weights. This figure was subtracted from the total body fat determined by DEXA analysis to provide an estimate of peripheral fat.
Adrenal and thymus weights
At the time of dissection of fat pads, the thymus gland and both
adrenal glands were also removed and weighed.
Glucose measurement
Plasma glucose concentrations were determined by the hexokinase
enzymatic method with spectrophotometric quantitation, using the
glucose (HK) kit (procedure 16-UV) from Sigma. All samples
were quantitated in a single assay. The sensitivity of the assay was
1.0 mg/dl, and the intraassay coefficient of variation was 2.0%.
RIA
Melatonin was quantitated as previously described (13), using
the Guildhay antiserum, which has been well characterized for
determination of melatonin concentrations in rat plasma (14). Leptin
and insulin were quantitated with rat leptin and insulin RIA kits from
Linco Research, Inc. (St. Charles, MO). Total
corticosterone was quantitated as previously described (15). For each
hormone, all samples were assayed in a single RIA. Intraassay
coefficients of variation were all less than 10%. Limits of detection
for the melatonin, leptin, insulin, and corticosterone assays were 11.5
pg/ml, 0.5 ng/ml, 0.1 ng/ml, and 5 ng/ml, respectively.
Statistical analyses
Behavioral tests were analyzed by statistical tests appropriate
to each; these are described in Results. Other comparisons
were made by one-way ANOVA, except as noted. For the cross-over study,
results were compared between treatment groups by one-way ANOVA, and
compared within treatment group before vs. after the
cross-over by one-way ANOVA. Results are reported as the mean ±
SEM, with the level of significance set at
P < 0.05. Data from the YOUNG group are reported for
reference, but were not included in the statistical analyses of
melatonin treatment effects. The statistical software package used was
StatView version 4.57 for Windows (Abacus Concepts, Inc., Berkeley, CA;
and SAS Institute, Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Alternating days consumption test
During the first 24 h of the alternating days consumption
test, 24 previously untreated rats each drank 36.1 ± 1.0 ml water
containing 0.01% ethanol vehicle. When the water bottles contained
both ethanol vehicle and 0.4 µg/ml melatonin for the next 24 h,
the rats drank 37.8 ± 1.0 ml. When the water bottles were again
replaced with bottles containing only vehicle in water for 24 h,
the rats drank 36.8 ± 0.7 ml. The volumes consumed on these 3
consecutive days were not significantly different by repeated measures
ANOVA. Thus, addition of melatonin to the sole source of drinking water
at a dosage of 0.4 µg/ml did not alter the volume of drinking water
consumed over a 24-h period. Furthermore, the addition of melatonin to
the drinking water for 24 h did not alter the volume of drinking
water consumed over the subsequent consecutive 24-h period after
removal of the melatonin.
Conditioned saccharin taste aversion test
On the conditioning day (day 4) of the conditioned taste aversion
test, the 12 rats allowed access to saccharin solution containing
control vehicle (group a) each consumed 11.0 ± 0.8 ml in 15 min,
whereas the 12 rats allowed access to a saccharin solution containing
vehicle plus 0.4 µg melatonin/ml (group b) each consumed 11.9 ±
0.8 ml in 15 min (providing an acute oral dosage of 4.8 ± 0.3
µg melatonin, or 8.7 ± 0.5 µg melatonin/kg BW). The rats that
drank saccharin solution containing vehicle alone on the conditioning
day (group a) each drank 2.4 ± 1.0 ml less saccharin solution
when allowed access to a choice between saccharin solution
vs. water on the test day (day 6); the rats that drank
saccharin solution containing melatonin on the conditioning day (group
b) drank 2.1 ± 1.2 ml less saccharin solution on the test day.
These difference scores between saccharin solution intake on the
conditioning day vs. that on the subsequent test day were
not significantly different between the melatonin-conditioned
vs. vehicle control-conditioned rats by Student’s
t test. Thus, addition of 0.4 µg/ml melatonin to the
drinking water did not alter the volume of acute water plus saccharin
consumption by water-deprived rats (on day 4), and pairing of melatonin
consumption with the taste of saccharin did not induce conditioned
taste aversion.
Plasma melatonin levels
Daytime plasma levels of melatonin were low for both middle-aged
groups (CONTROL, 11.5 ± 0.00 pg/ml; MELATONIN, 14.1 ± 2.6
pg/ml). Melatonin treatment resulted in significantly
(P < 0.001) increased levels during the dark period
(150.5 ± 19.2 pg/ml), whereas nighttime melatonin was not
increased in the CONTROL group (24.1 ± 8.8 pg/ml) relative to
daytime levels. In the YOUNG untreated rats, low levels of plasma
melatonin during the daytime (20.0 ± 6.3 pg/ml) were
significantly increased at night (50.0 ± 4.7 pg/ml).
BW, weight change, and food consumption
The BW of MELATONIN animals (Fig. 1
and Table 1) decreased 3% (16 ±
2 g weight loss) over the initial 12 weeks of study compared with
a 3.6% increase (19 ± 3 g weight gain) in CONTROL rats.
There were no significant effects of melatonin treatment on relative
food intake (grams of daily food intake per g BW; Fig. 1), daily food
intake (Table 1), or cumulative food intake (Table 1). Feed efficiency
(grams of BW change per g cumulative food intake) was negative in
MELATONIN and positive in CONTROL animals (Table 1).
|
|
) of the MELATONIN group was 19%
greater than that of CONTROL rats (P < 0.05). Compared
with YOUNG animals, activity was 35% lower for CONTROL and 23% lower
for MELATONIN rats.
|
Body composition and adiposity
The MELATONIN group had 20% less intraabdominal fat than CONTROL
animals (P < 0.05; Table 2), whereas there was no significant
effect of melatonin treatment on total or peripheral fat. The
proportion of total fat located in the intraabdominal cavity was
decreased by 16% (P < 0.01) in MELATONIN compared
with CONTROL animals (Table 2), consequently resulting in an increased
proportion of fat located peripherally. Whereas melatonin treatment of
middle-aged rats resulted in significantly less absolute lean body mass
(grams), when lean mass was expressed relative to total BW (percent
lean mass), there was no difference due to treatment (Table 2).
|
),
although these differences reached significance only for the large
epididymal fat pad, which was 27% smaller (P < 0.01)
after 12 weeks of melatonin treatment. Compared with YOUNG rats,
middle-aged rats had nearly twice as much total body fat (
73 or
77 g compared with 40 g), whereas lean body mass was only
20–30% greater in MELATONIN and CONTROL groups, respectively.
|
) were
lower (P < 0.05) in MELATONIN than CONTROL rats
(decreased by 33% and 25%, respectively). The plasma leptin and
insulin levels in MELATONIN rats were not different from levels in the
YOUNG group (leptin, 3.1 ± 0.4 ng/ml; insulin, 2.2 ± 0.2
ng/ml). Plasma glucose concentrations were not significantly altered by
melatonin treatment (Table 2). Basal corticosterone concentrations
(measured during the morning) were 2.5 times higher in MELATONIN
compared with CONTROL rats (P < 0.001) and were not
different from those in YOUNG animals (78.6 ± 11.3 ng/ml). There
was no effect of treatment on afternoon (peak) corticosterone
levels.
Adrenal and thymus weights
There were no significant differences in adrenal or thymus weight
between MELATONIN (0.030 ± 0.003 and 0.196 ± 0.017 g,
respectively) and CONTROL (0.030 ± 0.004 g and 0.205 ±
0.015 g, respectively) rats. Adrenal weights of MELATONIN and CONTROL
rats were not different from those of YOUNG rats (0.030 ± 0.002
g), but the average weight of the thymus gland was greater (0.377
± 0.026 g) in the YOUNG group than in the MELATONIN or CONTROL
group.
BW, weight change, and food consumption before and after
cross-over
BWs were significantly different between treatment groups at
cross-over (P < 0.05), but were not significantly
different (P < 0.07) at the end of the study (Fig. 4). The nine rats initially treated with
MELATONIN that were then crossed over to CONTROL treatment for an
additional 12 weeks rapidly gained weight (+44 ± 3 g),
whereas the nine CONTROL rats that were crossed over to MELATONIN
rapidly lost weight (-24 ± 4 g), reversing their
precross-over weight trends (Fig. 4). The final weights of rats that
were crossed from MELATONIN to CONTROL treatment were similar to the
weights of CONTROL rats before the cross-over (551 ± 9 and
546 ± 13 g, respectively); the final weights of rats that
were crossed over from CONTROL to MELATONIN (522 ± 10 g)
were similar to those of the MELATONIN group before cross-over
(507 ± 9 g).
|
) did not change for either group. Feed
efficiency (grams of BW change per g cumulative food intake) was
negative in MELATONIN and positive in CONTROL animals before
cross-over, and this relationship was reversed after cross-over
(Table 3).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The two-bottle choice test demonstrated that rats exhibited no altered
preference for water containing 0.4 µg/ml melatonin relative to water
that did not contain melatonin. Work originally conducted by Garcia and
colleagues (17) on the concept of bait shyness has demonstrated that
rats exhibit innate aversion to novel tastes. The results of the choice
test thus suggest that the rats could not taste the melatonin present
in the drinking water. The alternating days consumption test
demonstrated that addition of this dosage of melatonin to the drinking
water also had no effect on the volume of water consumed over a 24-h
period as well as no rebound or delayed effect on the volume of water
consumed over the 24-h period after removal of the melatonin. Finally,
the conditioned saccharin taste aversion test demonstrated that acute
(15-min) oral consumption of a relatively large dose of melatonin
(i.e. an amount comparable to 33% of the melatonin
consumed over an entire 24-h period during the experimental trial) did
not induce a conditioned taste aversion and thus was not perceived as
toxic. The complementary results of these three experiments thus
suggest that addition of melatonin to the drinking water in this study
did not directly alter water consumption, could not be tasted by the
rats, and did not produce effects perceived by the rats as toxic.
Melatonin supplementation via the drinking water increased nocturnal plasma melatonin levels, but did not alter basal plasma melatonin levels during the daytime, consistent with melatonin’s short (20-min) biological half-life (18) and with our previous demonstration (9) that middle-aged rats drank more than 90% of their water plus melatonin during the dark phase. The melatonin treatment employed in this experiment thus reestablished a nocturnal pattern of melatonin exposure. The nocturnal melatonin levels resulting from this treatment protocol were higher than those in untreated young animals, although the single sample at a single time point chosen for nocturnal tail bleeding may have missed the spontaneous melatonin peak in the young animals. Preliminary results from a follow-up study reveal that a 50% lower melatonin dose produced nocturnal plasma melatonin levels in middle-aged male rats that were identical to nocturnal plasma melatonin levels in untreated young rats, and the effects on intraabdominal fat and plasma insulin and leptin levels were similar to those in the current study (16).
Melatonin treatment restored morning plasma corticosterone levels to those of young animals, although afternoon peak levels were not changed. This apparent time of day effect may be due to the fact that most of the daily melatonin exposure occurred at night, immediately before the time of the morning blood sample collection. The increased morning plasma corticosterone levels may have been due to either increased secretion or decreased clearance. The restoration of youthful morning corticosterone levels was achieved without affecting adrenal and thymus weights, sensitive indicators of chronic stress or glucocorticoid excess. Thus, it appears that melatonin supplementation may have altered some aspects of hypothalamo-pituitary-adrenal regulation in a manner not associated with a chronic stress response.
The ability of melatonin to decrease BW was observed within 2 weeks of treatment, after which weights tended to stabilize at the lower level rather than continuing to decline. In previous studies (9) we have demonstrated that this weight loss in response to melatonin treatment of middle-aged rats was sustained when melatonin treatment was continued for 1 yr, into old age (22 months). In the current study we have demonstrated that this weight loss was achieved despite the apparent lack of effect of melatonin treatment on food intake. These patterns were repeated in the cross-over portion of the study; the BWs of animals initially receiving the melatonin treatment rapidly began to increase when switched to the control treatment, and the BWs of control animals rapidly decreased in response to melatonin treatment, without corresponding changes in food intake. Animals that initially received melatonin regained weight aggressively when switched to control treatment, gaining almost twice the amount of weight lost before the cross-over, whereas animals initially receiving control treatment and then crossed over to melatonin lost the same amount of weight as that lost by the opposite group during the initial melatonin treatment. This similarity between BWs and weight changes in each group at the end of the melatonin (or control) treatment may suggest that these treatments influenced the animals to defend different BW set-points.
Feed efficiency, a measure of metabolic function (19, 20), was negative in the melatonin-treated animals and positive in controls; this effect was immediately reversed when the treatments were crossed over. These results suggest that melatonin may alter either metabolism or the ability to store excess energy as body fat, and that this effect is readily reversible. This finding is consistent with the observation that melatonin injections increased the norepinephrine-stimulated metabolic rate in mice (21).
Aging is associated with decreased levels of physical activity (22, 23). We found that melatonin increased the spontaneous physical activity of middle-aged rats by 19%, toward levels in young animals. In a follow-up study (to be reported elsewhere) we recently determined that similar melatonin treatment again increased (P < 0.05) spontaneous locomotor activity in middle-aged male rats by 18%. Although a specific mechanism for this effect of melatonin is not clear, melatonin treatment has been found to reverse the decreased physical activity of old mice (24), of rats with suppressed endogenous melatonin secretion due to continuous light exposure (25), and of mice infected with murine retrovirus (26). Recent work (20) has shown that small changes in physical activity (e.g. fidgeting) can have large effects on the amount of weight gained due to caloric excess in humans. It is thus reasonable to speculate that similar small increases in physical activity of melatonin-treated middle-aged rats in the present study contributed to the decreased BW of these animals.
Aging is also associated with decreased body core temperature and blunted circadian rhythms of body temperature (27, 28, 29, 30). Melatonin treatment of middle-aged rats in this study increased daytime body core temperature to a level not different from that in young adult rats. The increased body temperature of the melatonin-treated rats may have been a consequence of increased physical activity in this group, and although great care was taken to minimize stress during the recording of body temperatures, it is possible that temperature changes may have been mediated by physical activity and/or stress-related responses specifically associated with the handling of the animals. Treatment with melatonin or an analog has been shown to restore blunted circadian rhythms of body temperature in old rats (27) or rats exposed to continuous light (25) and to acutely increase body core temperature in intact rats (31). The effects of acute and chronic daily melatonin treatment on thermoregulation have been investigated most frequently in experiments using photoperiodic animals or pinealectomy (see Ref. 29 for review). Melatonin has been shown to both increase and decrease body temperature; these inconsistencies may be explained by differences in species, age, and mode or time of day of melatonin treatment. In the present study the decreased body temperature of control middle-aged animals might be related to either decreased physical activity or decreased metabolism, highly interrelated variables. It is significant that decreased activity and/or metabolism have been shown to precede development of obesity and increased intraabdominal adiposity in both older rats and older humans (22, 23, 28, 32, 33).
Melatonin treatment decreased the amount of intraabdominal fat in middle-aged male rats, with no apparent changes in the level of total adiposity. There are a number of possible explanations for this observation, all speculative. One may be related to the increased core body temperature induced by melatonin. Lipoprotein lipase (an enzyme required for the uptake of fatty acids into adipocytes) has been shown to have increased activity when incubated at lower temperatures (34). This observation has led to the hypothesis that accumulation of visceral adipose tissue may be increased when body core temperatures are decreased (35). Observations of decreased core temperature in genetically obese mice (36) are compatible with this largely untested hypothesis. An alternate hypothesis, supported by human experimental trials, suggests that moderate increases in physical activity and exercise selectively decrease visceral adipose tissue (8, 37, 38, 39). Thus, the 19% increase in physical activity observed with melatonin treatment in the current study may have been at least in part responsible for the 20% decrease in intraabdominal fat stores. Alternatively, the increased activity and body temperature may simply reflect increased autonomic activity, which directly regulates metabolism (especially thermogenesis in brown fat, which is increased by sympathetic stimulation). A final potential explanation for the selective decrease in intraabdominal fat is suggested by studies in which peripheral leptin administration to rats was shown to selectively decrease visceral fat and improve insulin sensitivity (40); it is plausible that melatonin administration to the middle-aged rats in the current study may have increased sensitivity to endogenous leptin, resulting in selectively decreased intraabdominal fat and decreased plasma insulin.
Both leptin and insulin levels tend to parallel adipose tissue mass (5, 7, 41, 42), although they may be regulated differently by diet, exercise, age, and treatment. Although we have no direct measures of leptin or insulin sensitivity, it is reasonable to assume that the tissue responsiveness to these hormones increased, consistent with the decreases in circulating levels. Pinealectomy has recently been reported to cause glucose intolerance and a reduction in adipose tissue glucose transporter content (43) in young rats, although it should be noted that in another study pinealectomy and melatonin treatment were reported to have no effect on glucose homeostasis and insulin responsiveness (44). Overall, our results suggest that insulin and leptin sensitivity may be decreased by the gradual decline in endogenous circulating melatonin with aging, and that appropriate replacement with exogenous melatonin may reverse these declines.
The demonstrated dissociation between melatonin-induced BW and intraabdominal fat changes vs. the lack of corresponding changes in food intake, both during the initial treatment period and in response to treatment cross-over, is consistent with evidence that peripheral administration of leptin at a dosage that did not alter food intake did decrease visceral adiposity and BW (40). Although the current results suggest that the effects of melatonin treatment are independent of effects on food intake, it should nonetheless be noted that facilitation of satiety may have prevented a compensatory increase in eating that would presumably have maintained BW changes similar to those in control rats.
Intracerebroventricular melatonin administration has been shown to acutely decrease plasma insulin levels and increase insulin sensitivity in rats (45), suggesting potential central mediation of the demonstrated metabolic effects of melatonin supplementation during aging. However, a recent preliminary report (46) suggests that direct peripheral effects of melatonin on adipocytes may also have a role. In situ perfusion of a sc fat pad with low, physiological levels of melatonin was reported to block fatty acid transport via a melatonin receptor-mediated mechanism (46). As leptin is produced in adipocytes in response to nutrient cycling (47), it is reasonable to suggest that melatonin’s action to block fatty acid transport into the adipocytes could not only suppress lipogenesis but also acutely decrease leptin release into the circulation. If the sensitivity to leptin is increased by melatonin treatment, as suggested above, then even these lowered leptin levels could potentially facilitate energy expenditure, a response to leptin that can be dissociated from leptin’s actions to suppress food intake (40).
In summary, we have shown that supplementation of melatonin in middle-aged male rats mimics some youthful energy regulatory responses, decreasing BW, intraabdominal adiposity, and plasma insulin and leptin concentrations while increasing core body temperature, physical activity, and plasma corticosterone levels. These results suggest that aging-associated decreases in endogenous melatonin secretion may alter energy regulation in middle age, resulting in increased BW and adiposity and their associated detrimental metabolic consequences. However, it is important to note that extrapolating from these findings to species other than the rat is not currently warranted, as melatonin effects on BW and adiposity have been found to be variable even in distinctly photoperiodic species (48, 49, 50), and the effects of daily melatonin administration on energy balance in nonphotoperiodic diurnal animals or humans have not been determined. Finally, although several potential mediating mechanisms have been suggested here, it is also important to note that melatonin, which is both water soluble and lipophilic, has access to essentially all cells and intracellular compartments in the body and has been demonstrated to exert effects not only at the cell membrane but also at the intracellular and genomic levels (51). Accordingly, the mechanisms by which melatonin administration may reverse or prevent aging-associated metabolic changes are under further investigation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 6, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  D. Acuna-Castroviejo A new guest playing with bone and fat Am J Physiol Regulatory Integrative Comp Physiol, June 1, 2007; 292(6): R2206 - R2207. [Full Text] [PDF]
|
|
|
|
|
|
M. Sanchez-Hidalgo, Z. Lu, D.-X. Tan, M. D. Maldonado, R. J. Reiter, and R. I. Gregerman Melatonin inhibits fatty acid-induced triglyceride accumulation in ROS17/2.8 cells: implications for osteoblast differentiation and osteoporosis Am J Physiol Regulatory Integrative Comp Physiol, June 1, 2007; 292(6): R2208 - R2215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Bertrand, B. Faivre, and G. Sorci Do carotenoid-based sexual traits signal the availability of non-pigmentary antioxidants? J. Exp. Biol., November 15, 2006; 209(22): 4414 - 4419. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. R. Buff, C. D. Morrison, V. K. Ganjam, and D. H. Keisler Effects of short-term feed deprivation and melatonin implants on circadian patterns of leptin in the horse J Anim Sci, May 1, 2005; 83(5): 1023 - 1032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. I. C. Alonso-Vale, S. Andreotti, S. B. Peres, G. F. Anhe, C. das Neves Borges-Silva, J. C. Neto, and F. B. Lima Melatonin enhances leptin expression by rat adipocytes in the presence of insulin Am J Physiol Endocrinol Metab, April 1, 2005; 288(4): E805 - E812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Prunet-Marcassus, M. Desbazeille, A. Bros, K. Louche, P. Delagrange, P. Renard, L. Casteilla, and L. Penicaud Melatonin Reduces Body Weight Gain in Sprague Dawley Rats with Diet-Induced Obesity Endocrinology, December 1, 2003; 144(12): 5347 - 5352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Kreier, A. Yilmaz, A. Kalsbeek, J. A. Romijn, H. P. Sauerwein, E. Fliers, and R. M. Buijs Hypothesis: Shifting the Equilibrium From Activity to Food Leads to Autonomic Unbalance and the Metabolic Syndrome Diabetes, November 1, 2003; 52(11): 2652 - 2656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. N. Anisimov Effects of Exogenous Melatonin--A Review Toxicol Pathol, |
