| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
* Metabolic Research Laboratory, University of Navarra; and
Department of Endocrinology, Clínica Universitaria de Navarra, 31008-Pamplona, Spain
1Correspondence: Deptartment of Endocrinology, Clínica Universitaria de Navarra, 31008-Pamplona, Spain. E-mail: gfruhbeck{at}unav.es
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
Key Words: leptin • parabiosis • lipostatic factor • obesity • energy balance
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
. According to his model, changes in energy
balance sufficient to alter body fat stores are signaled via one or
more circulating factors that act in the brain to elicit compensatory
changes in order to match energy intake to energy expenditure. Two
decades later, Coleman (2)
in his classic parabiosis
studies provided evidence that such circulating signals exist. In those
experiments, two mice were surgically joined to permit vascular
anastomosis. By joining the circulatory systems of severely obese mice
of two different strains (ob/ob and db/db) to
wild-type animals and to each other, he found that the obese gene
(ob) mutation lacked the production of a circulating
anorexic factor whereas the diabetes gene (db) mutation
impaired the response to this factor.
The cloning in late 1994 of the ob gene (3)
demonstrated that it encodes a 16 kDa protein, termed OB protein or
leptin, synthesized mainly by fat cells and secreted into the
bloodstream. Moreover, correction of leptin deficiency in
ob/ob mice by exogenous administration caused a marked
reduction in food intake and a normalization of the obesity syndrome
(4
5
6)
. Subsequent studies determined that the
db mutation resides in the gene encoding the leptin receptor
(7
, 8)
. Therefore, leptin was thought to be the
long-sought blood-borne factor working as a negative feedback signal
critical to the normal control of food intake and body weight, which
explained all the observations made in the different parabiosis
studies.
|
HYPOTHESIS |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
|
PARABIOSIS EXPERIMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
. A longitudinal incision is made
along opposed adjacent flanks of each rodent and the skin is loosened
from connective tissue. The adjoining muscles on the lateral side of
the peritoneum of the two animals are then divided and the abdominal
muscles on the right side of one rodent are sutured to the abdominal
muscles on the left side of the other animal so that peritoneal
surfaces are connected. Thoracic muscles and peritoneal walls are
joined by shallow silk sutures to prevent fluid accumulating in the
dead space between the animals. The ventral edges of the incision are
joined by suture. Femurs are exposed by separating surrounding muscles.
Periosteum is scraped off the exposed surfaces, and the femurs are
joined by suturing twice around the bones. Deep sutures are made
between adjacent muscles to support the union and prevent animals from
pulling themselves apart. Likewise, scapulas are exposed by removing
the covering muscles, scraping off periosteum, suturing through the
bones twice, and joining the surrounding muscles to stabilize the
union. Finally, the dorsal edges of the skin incisions are joined.
Subsequent histological examinations reveal the development of fibrous
tissue and vascular, but not neural, connections between the two
animals (10)
.
Fifteen days after surgery, blood exchange in parabiotic pairs is
measured by a bolus injection of Evans blue dye into the jugular vein
of one of the rodents. Forty to 50 min later, a 100 µl blood sample
is collected by tail bleeding from both members of the pair. Exchange
is confirmed by blue coloration of serum collected from the noninjected
animal. In rodents, the rate of blood exchange between partners is
relatively slow, with 
1.0–2.8% of the blood volume of each animal
exchanging per minute (11
, 12)
. The ability of a factor to
successfully pass between parabionts is not determined by size, as
erythrocytes can exchange between parabiosed rats (10)
.
The slow rate of exchange determines which hormones produced in one
animal can deliver bioactivity to the partner, and only those compounds
with a relatively long half-life cross the union faster than they are
cleared from the circulation. When one member of a parabiosed pair of
rats receives an intravenous injection of dye, the dye concentration
reaches equilibrium after 2 h (11)
. The
extended period of time required for equilibration means there is a
concentration gradient for many nutrients and hormones that have a
short circulating half-life and are metabolized faster than they
exchange between parabiotic partners (12)
. Thus,
parabiosis poses a functionally complete barrier to the exchange of
short-lived circulating factors.
Hypothalamic lesions
Early studies with brain-lesioned rats showed that appropriately
placed lesions in the hypothalamus cause hyperphagia and obesity in
single animals (13)
. Lesions made by the same technique
produced essentially similar results in parabiotic rats
(9)
. Moreover, Hervey was the first to report that the
production of obesity by lesioning the ventromedial hypothalamus (VMH)
in one member of a parabiotic union led to hypophagia and weight loss
in the unlesioned parabiont, which showed no interest in food despite
progressive emaciation that frequently culminated in death (Fig. 1
). At autopsy, unlesioned rats showed loss of all body fat stores and
atrophy of the gastrointestinal tract and liver with no other apparent
cause of death other than starvation (9)
. Hervey concluded
that a circulating satiety factor was produced by the lesioned
parabiont as body fat accumulated. Whereas this rat was rendered
insensitive to the factor by VMH destruction, the unlesioned parabiont
became hypophagic in response to the high level of the satiety signal
transmitted across the parabiotic union. As a further corroboration,
Hervey produced VMH lesions in the lean parabionts to find that eating
was quickly resumed, with body weight and body fat rebounding to values
threefold those of unoperated single littermates (9)
.
|
Parameswaran et al. (14)
created parabiotic rat pairs and
placed stimulating electrodes into the lateral hypothalamus of one of
the parabionts. Marked hyperphagia and obesity in the operated
parabiont were produced by three daily stimulation sessions of the
lateral hypothalamus. As the stimulated rat gained weight, the thin
parabiont ate progressively less in accordance with Hervey’s findings.
At the time of death, the stimulated animal showed a tremendous adipose
tissue mass expansion, whereas the thin partner was essentially devoid
of fat. Glucose, insulin, and glucagon concentrations of the thin
parabionts were not increased, suggesting that none of these factors
was responsible for the hypophagia of these rats (14)
.
Genetic obesity
In 1950 the appearance of a recessive mutant associated with
massive obesity was reported (15)
. The genetic defect in
the obese or ob/ob mouse is a single autosomal recessively
inherited disease manifested early in life that is also associated with
diabetes. Shortly after discovery of the ob/ob mouse, a
second recessively inherited form of obesity, called diabetes or
db/db mouse, was described (16)
. That
ob/ob and db/db mice are phenotypically identical
when expressed on the same genetic background yet genetically different
was initially interpreted as a genetic defect at two steps of the same
metabolic pathway (17)
. The search for the underlying
biochemical and physiological mechanism of these animal models of
early-onset obesity and diabetes syndromes was tackled by performing
parabiosis experiments.
When parabiont pairs of db/db with normal (+/+) mice were
produced, it was observed that the lean littermates had decreased food
intake, rapidly lost weight, became hypoinsulinemic and hypoglycemic,
and died of apparent inanition within 50 days after surgery (Fig. 2
), whereas the diabetic partners that had been food-restricted before
the union gained weight easily and became obese (18)
.
Similarly, when ob/ob mice were parabiosed with
db/db mice, the ob/ob partner lost weight,
experienced a dramatic adipose tissue mass reduction, exhibited
hypoinsulinemia and hypoglycemia, and finally died of starvation
whereas the db/db parabiont increased body weight (Fig. 2)
.
These findings pointed to the same kind of response in ob/ob
mice and lean control littermates; i.e., the existence of normal
satiety centers responsive to a circulating satiety factor (2
, 17)
. Both partners survive in unions of ob/ob with
normal mice, suggesting that the obese parabiont, unlike the
db/db mice, does not produce the satiety factor necessary to
turn off the normal partner’s eating drive (Fig. 3
). However, ob/ob mice in such pairs exhibited decreased food
intake and gained weight less rapidly than when parabiosed to other
ob/ob mice (2
, 17)
. Earlier, Hausberger
(19)
had reported that non-obese mice suppress the weight
gain of ob/ob littermates in parabiosis. However, when the
two animals were separated, the obese mouse rapidly regained weight.
Hausberger concluded that obesity was not caused by the inherent
qualities of adipose tissue itself but by the lack of a factor that
could be transmitted by successful parabiosis. Partners of parabiotic
unions between control +/+ mice lost some weight initially (Fig. 3)
,
but maintained normal plasma glucose and insulin concentrations despite
a slight tendency toward hypoglycemia (2
, 17)
. These
experiments suggested that the ob/ob mouse is unable to
produce a sufficient satiety factor to regulate its food consumption,
whereas the db/db mouse produces the factor in excess but
cannot respond to it because of a defective satiety center. Consistent
with this explanation, parabiosis of either ob/ob or
db/db mice with rodents of their same strain is not lethal
to the different parabionts but does not prevent the obesity–diabetes
syndrome (Fig. 4
).
|
|
|
Forced overfeeding
Tube feeding experiments also support the concept of a circulating
satiety factor. When excess caloric intake is provided to one animal by
intragastric intubation, the combined spontaneous intake falls below
the baseline level of food consumption of control pairs, demonstrating
the development of hypophagia in both parabionts (11)
. In
addition, fat pad weights are significantly smaller in the non-tube-fed
parabiont, an indication that caloric intake is below that needed to
maintain body fat stores (Fig. 5
). The mechanism of this weight loss appeared to be a 5–10% reduction
in caloric intake by the lean parabionts (12)
. Although
this degree of anorexia is modest compared with that reported for the
partners of VMH-lesioned rats, it is striking in view of the
hyperphagia expected for the 60% reduction in body fat content induced
in the lean parabionts (12)
. When overfeeding is
discontinued, the body composition of both partners returns to normal,
thus showing that the changes observed are not due to nonspecific
effects of parabiosis (20)
. Therefore, overfeeding one
animal leads to depression of food intake in the other via a humoral or
ergostatic, i.e., an energy stabilizing, mechanism. Measurement of
circulating concentrations of glucose, free fatty acids,
ß-hydroxybutyrate, insulin, corticosterone, and growth hormone
indicated that none of these factors can account for the loss of body
fat (12)
. In further studies of overfed rats, Harris et
al. (21)
elegantly demonstrated the existence of a serum
factor that directly inhibited lipogenesis.
|
Because in gavage overfeeding food is delivered directly to the
stomach, numerous gastrointestinal, hepatic, or pancreatic factors must
be considered as potential candidates for the satiety factor. Taking
into account the exchange kinetics of parabiosis experiments, it is
unlikely that the satiety factor that passes from the obese to the lean
animal is a short-acting gastrointestinal meal termination signal.
Experimental confirmation of this reasoning was provided by training
normal-weight rats in parabiosis to consume their total daily chow diet
during a 2 h period (22)
. When the rats were placed
in partitioned cages and one parabiont was fed 2 h ahead of the
other, food intake of the parabiont fed second was unaffected. Because
circulating meal-related satiety signals in the rodent fed first should
still have been elevated, it can be inferred that these signals do not
pass to the other animal. Therefore, the satiety induced in the
partners of experimentally obese parabiotic rats must be related to the
increased body fat content of the obese animals rather than to
signaling elicited by the filling of the gastrointestinal system. There
are various likely sites for the origin of an ergostatic factor. The
most obvious seems to be the expanding fat depot; it is conceivable
that as adipose tissue mass enlarges, a factor that acts as a sensing
hormone or ‘lipostat’ in a negative feedback control from adipose
tissue to hypothalamic receptors informs the brain about the abundance
of body fat, thereby allowing feeding behavior, metabolism, and
endocrine physiology to be coupled to the nutritional state of the
organism.
|
UNEXPLAINED FINDINGS |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
Pancreatic islet transplants
Strautz (23)
has shown that transplantation of
isolated pancreatic islets from normal to obese ob/ob mice
stabilizes the rate of weight gain and reduces both hyperglycemia and
hyperinsulinemia. This study implied that the missing satiety factor
may be of pancreatic origin. Similar islet transplant studies
undertaken in another genetic model of obesity, the polygenic New
Zealand obese (NZO) mouse, further support the existence of an islet
factor. These mice become obese early in life, have increased
adiposity, and are both hyperglycemic and hyperinsulinemic
(24)
. It has been shown that when NZO mice are implanted
with islets isolated from normal albino mice, the weight gain of the
obese rodents is reduced and both plasma glucose and insulin
concentrations are significantly lowered (25)
.
Furthermore, glycemia and insulinemia could be decreased to normal
concentrations by intraperitoneal (i.p.) implantation of pancreatic
islets from albino mice that had undergone a selective destruction of
ß cells by streptozotocin (26)
. It would seem,
therefore, that the genetic lesion responsible for the NZO syndrome
lies within the islets of Langerhans and not within the ß cell.
Hyperleptinemia has been reported in NZO mice although they do not
appear to carry mutations in either leptin or leptin receptor genes
(5
, 27)
. These rodents do not decrease food intake in
response to peripheral leptin administration, although
intracerebroventricular infusion of leptin causes a decrease in food
consumption whereas energy expenditure is unchanged, thus leading to a
negative energy balance (5)
.
Adrenalectomy
The obesity of both leptin-deficient and leptin receptor-deficient
rodents does not progress after adrenalectomy, showing that removal of
the adrenal cortex is sufficient to impair the development of these
genetic models of obesity (28)
. In the absence of
glucocorticoids of adrenal origin, food intake normalizes, muscle mass
increases, hyperglycemia abates, and insulin resistance disappears. The
reduced fat deposition that occurs after adrenalectomy has been shown
to be due to a large decrease in the efficiency of energy utilization
associated with a restoration of brown fat activity (29)
.
The reversal of the obesity–diabetes syndrome after adrenalectomy is
obviously attained without the involvement of leptin and provides
evidence for an important role of glucocorticoids in the development of
the pathophysiological characteristics of obesity. It is striking that
consumption of either a high-glucose or a high-fat diet prevents the
aforementioned effects of food intake reduction and body weight gain
associated with adrenalectomy, indicating that factors other than
adrenal secretions, which are influenced by diet composition
(particularly with highly palatable diets), mediate the development of
obesity (30
, 31)
.
Adrenalectomy, however, does not restore infertility, indicating that
not all metabolic and endocrine alterations associated with genetic
models of obesity can be explained by a single factor. Furthermore, the
role of glucocorticoids of adrenal origin in the phenotypic expression
differs depending on the experimental model of obesity. Whereas
adrenalectomy reduces the hyperphagia characteristic of
ob/ob mice, in obesity mediated by gold thioglucose
treatment adrenalectomized animals develop severe anorexia, exhibit
progressive weight loss, and die within several weeks of the surgical
intervention (32)
.
Ovariectomy
The role of estrogens in the development and topography of adipose
tissue has been documented for decades (33)
. Castration of
female obese rodents has a profound effect on body fat, increasing the
amount and changing its distribution as well as increasing food intake
and decreasing the activity of the sympathetic nervous system.
Estrogenic replacement by either peripheral or direct hypothalamic
application reverses the effects after ovariectomy. Additional evidence
for the critical role of estrogen signaling in white adipose tissue
expansion comes from knockout experiments where mice lacking estrogen
receptor 
exhibit increased fat pad weight together with augmented
adipocyte size and number (34)
. Thus, estrogen seems to be
an integral part of the feeding and reproductive systems. It is
noteworthy that adrenalectomy prevents the obesity after ovariectomy
just as it does in leptin deficiency and VMH-lesioned induced obesity
(35)
.
Gavage overfeeding
The time lag observed in the manifestation of some of the
different biological effects that occur in tube feeding parabiotic
experiments deserves some consideration. On the one hand, the changes
in body composition of partners of obese rats are not apparent until
overfeeding has continued for at least 23 days (12)
, which
implies that the lipid-depleting factor is produced in response to a
substantial increase in body fat of the tube-fed rat. On the other
hand, the effects on food appetite take place in an acute way. If a
lipostatic factor produced by the forced-fed animal is responsible for
the anorexic effect of the ad libitum-fed parabiont, a gradual decline
in food intake in the non-tube-fed parabiosed animal would be expected
as fat pads enlarge. However, food intake drops within the first 2 days
(when no evident changes in body composition have taken place) and
shows no further decline as fat depots are filled (11)
.
Although this divergence in the time-related effect on food consumption
and body composition may appear paradoxical, it can be explained by
assigning leptin a more dynamic role in whole-body physiology with a
simultaneous participation in both short- and long-term events. Based
on a somewhat static view of the hormone, leptin’s function was seen
initially only as informing the brain about the abundance of body fat,
acting as a sensing hormone in a negative feedback control from adipose
tissue to the hypothalamus (3
4
5
6)
. However, as would be
predicted of a factor playing a key role in energy balance, expression
of the ob gene is subject to quick regulation. Without a
parallel decrease in body fat stores, short-term fasting induces a
marked fall in ob mRNA, and consequently in circulating
leptin concentrations, that is rapidly reversed upon refeeding
(36)
. Thus, some metabolic and hormonal actions of leptin
precede its effects on appetite or body weight, operating independently
of changes in food intake and fat size stores (37)
.
Replication of early parabiosis studies
Coleman’s findings with ob/ob and lean parabiosed mice
showing that the obese parabiont ate less, gained less weight, and had
a marked insulin reduction (2)
were confirmed in a recent
study by Harris (38)
. However, the reduction in carcass
weight of ob/ob partners of lean mice vs. their controls was
not statistically significant and the reduction in carcass fat from 56
to 48% of carcass weight was smaller than anticipated
(38)
. In principle, these observations are all consistent
with leptin produced by the lean parabiont being carried in the
circulation to the obese partner. In fact, measurement of leptin
confirmed the presence of this hormone in the blood of the
ob/ob mice parabiosed to lean littermates (38)
.
In addition, the recent study found that lean partners lost more fat
than the obese animals (37% vs. 14%, respectively) in the absence of
significant changes in food intake (38)
. Despite the 37%
reduction in body fat, it is surprising that neither leptin expression
nor circulating leptin concentrations were decreased in lean mice
parabiosed to ob/ob mutants. These findings suggest that in
lean ob/ob pairs, the ob/ob mice respond to
leptin delivered by the lean parabiont, whereas the lean partner
responds to a circulating signal, presumably originating in the
ob/ob parabiont, that maintains leptin expression at
inappropriate levels for the degree of adiposity of the lean animal
(38)
. Increased secretion of leptin cannot be attributed
to the ob/ob partner acting as a sink that absorbs all of
the OB protein and thus stimulates protein production, because
circulating leptin concentrations in lean mice of ob/ob
parabionts were the same as those in lean controls.
Many parabiotic pairs were reported to be lost to hypothermia in the
ob/ob partners (38)
. Body temperature lability
of these rodents in parabiosis shows impaired thermoregulation.
Although the amount of leptin exchanging between lean and
ob/ob partners was adequate to normalize serum insulin
concentrations and decrease food intake and weight gain, it was not
enough to raise body temperature (38)
. These observations
together with the thermogenic effect elicited in ob/ob mice
by administration of high leptin concentrations (6)
indicate that large amounts of the OB protein are required to induce
hyperthermia.
Parabiosis of ob/ob to db/db mice confirmed the
previous results of Coleman (2)
in which ob/ob
partners experienced a rapid weight loss, exhibited a reduced food
intake, and became hypoglycemic (39)
. In the more recent
study by Harris (39)
, special attention was directed to
monitoring changes in body composition, which had not been addressed
previously. The weight of ob/ob mice parabiosed to
db/db rodents was halved and body fat was reduced by 60%.
The loss of fat was associated with a substantial reduction in food
intake. Despite this extremely high rate of fat catabolism, lean tissue
of ob/ob mice was maintained, demonstrating a
tissue-specific energy mobilization. Surprisingly, parabiosis of the
db/db mouse to the ob/ob partner caused in the
former a small, but significant, reduction in carcass fat and a
significant 30% increase in carcass protein (39)
.
Altogether, these findings suggest that leptin reaching the
ob/ob mice through the parabiotic union leads to the release
of another circulating factor that promotes protein conservation in the
leptin-deficient mice. When this unknown circulating growth factor is
carried back to the db/db partner, it seems to induce
protein deposition in the leptin receptor-deficient rodent.
The elevation of rectal temperature in ob/ob, but not db/+,
partners of db/db mice confirmed previous observations that
leptin corrects hypothermia but does not induce hyperthermia in an
animal that maintains a normal body temperature (6
, 40)
.
When exogenous treatment with leptin became feasible, replication of
the classic parabiosis experiments was carried out to verify that
leptin was the long-sought circulating factor (41)
.
Twice-daily i.p. injections of recombinant murine leptin (50 µg/day)
resulted in extremely high serum concentrations of leptin at
intermittent intervals (200–700 ng/ml). The first conclusion obtained
was that the circulating half-life of recombinant murine leptin is
36 min (41)
. The rate of protein clearance determined
would allow leptin to exchange between parabiosed mice, but would not
exchange fast enough to reach equilibrium if there was a large
difference in concentrations of leptin in the two members of the pair.
The failure of leptin to reach equilibrium was confirmed by measurement
of the circulating protein, which showed that 2 h after injection,
the OB protein was present at higher concentrations in the treated
member of the parabiosed pair than in its partner (41)
. It
was further observed that leptin treatment of one member of a
parabiosed pair of ob/ob mice reduced serum insulin, food
intake, and body fat in both partners (Fig. 6
). The injected parabiont lost more fat than its partner, whereas body
temperature was increased only in the injected mouse. The normalization
of serum insulin concentrations was apparent in all mice, independent
of the change in body fat content, indicating that improved insulin
sensitivity is a primary effect of leptin. No effect on serum
corticosterone concentrations or adrenal weight was observed. Whereas
leptin administration to single ob/ob mice has been shown to
increase the weights of reproductive organs (42)
,
injection of recombinant leptin to ob/ob parabiosed mice had
no effect (41)
.
|
The differences observed in physiological effects between the treated
parabiont and its partner despite the high circulating concentrations
attained illustrate a gradation in the response to leptin. The decrease
in serum insulin concentrations is the most sensitive response and is
corrected in both the injected mouse and its partner. Changes in body
fat content are greater in the injected animal than in the partner. The
increase in body temperature is the least sensitive of the effects
measured, being evident only in the leptin-treated parabiont
(41)
. However, the possibility that these graded effects
of leptin on different biological functions are linked to the
intermittent, twice-daily injections of the protein as opposed to a
physiological, continuous slow release also has to be considered
(43)
.
In Coleman’s classic experiments, both +/+ and ob/ob mice
died of starvation when parabiosed to db/db mice (2
, 17
, 18)
. When the nature of the genetic defects in these mice
was established and exogenous treatment with leptin became available,
it was realized that leptin alone is not involved in the effects of
parabiosis, since leptin-treated wild-type or ob/ob mice do
not die from starvation (44)
. Therefore, the
hyperleptinemia in db/db mice may be accompanied by a high
concentration of another anorectic factor that is more powerful than
leptin itself or acts synergistically with leptin, potentiating its
effects in mice. The integrity of the leptin signaling system appears
to be necessary for the anorectic effect of the putative factor since
otherwise food intake would have to be limited in db/db mice
themselves. Potential explanations lie in the possibility that leptin
and the other factor share the same receptor, that db/db
mutants are also deficient in the receptor of this factor, or that
leptin plays a permissive role in the development of the anorectic
effect of the putative factor.
Adipose-derived satiety activity
After the demonstration that recombinant leptin behaves as a
prototype adipose satiety factor (4
5
6)
, only Weigle’s
group (45
, 46)
has attempted to characterize the satiety
activity of native adipose tissue. Initially, it was reported that
medium conditioned by adipose tissue from db/db mice
contains immunoreactive leptin and can suppress food intake for a 24-h
period after i.p. injection in ob/ob mice (45)
.
The subsequent study focused on a more complete characterization of
adipose-derived satiety activity (46)
. It was concluded
that adipose-derived satiety activity is not fully explained by leptin.
Adipocytes may secrete other factors that augment leptin action or
secrete leptin in a form that has greater biological activity. In
addition to leptin and insulin, other molecules secreted by adipocytes
may be involved in the feedback loop that communicates the status of
the body’s energy reserves to the brain. A leptin cofactor might be
produced constitutively by adipose tissue and secreted as a complex
with leptin, or production of the putative leptin cofactor might be
regulated by a variety of metabolic or hormonal molecules
(46)
. Variability or relative deficiency of the putative
leptin cofactor could account for a variable or diminished ability of
leptin to elicit satiety and curtail weight gain. In view of the
striking redundancy of hypothalamic pathways and neurotransmitter
systems controlling feeding, the existence of presently uncharacterized
molecules involved in the feedback loop between adipose tissue and the
central nervous system might be predicted.
Restoration of reproductive performance
Leptin quickly proved to play an important role in reproductive
physiology (47
, 48)
. Whereas sterility was a
well-recognized feature in ob/ob mice, exogenous
administration of leptin to these mice was shown to increase the weight
of ovaries and uterus, thus showing a trophic action of leptin on
gonadal function. Long-term injections of leptin have been reported to
correct the sterility of female (49)
and male
(50)
adult ob/ob mice, which does not appear to
be a consequence of weight change per se since weight loss in control
ob/ob animals due to food restriction did not ameliorate
infertility (49
, 50)
. In addition, leptin has been shown
to accelerate the onset of puberty in normal mice. Normal prepubertal
female mice injected with leptin experienced an earlier maturation of
the reproductive tract accompanied by a precocious onset of classic
pubertal signs like vaginal opening, estrus, and cycling
(51)
. In accordance with these findings, leptin is
increased in both boys and girls before the appearance of other
reproductive hormones related to puberty (36
, 47
, 48)
.
Therefore, leptin signals the adequacy of energy stores and seems to be
needed for the initiation of puberty and establishment of secondary
sexual characteristics by interacting with different target organs in
the hypothalamic-pituitary-gonadal axis.
Surprisingly, a genetically leptin-deficient women who entered puberty
in her late 20s (52)
and a female with congenital
lipodystrophy who had extremely low leptin concentrations but underwent
a normal reproductive progression (53)
both suggest the
existence of alternative mechanisms regulating reproductive
performance. Thus, in the absence of a crucial factor such as leptin,
other factors may be stimulated to rescue the reproductive system. This
plausible explanation is actually supported by studies showing that
leptin-deficient ob/ob males bred on a mixed C57Bl/6J and
BALB/cJ genetic background (54)
and ob/ob males
and females backcrossed for 10 generations to the BALB/cJ background
(48)
are fertile. Furthermore, genes that allow
reproduction in severely undernourished male, but not female, mice have
been shown to be present in wild house mice and presumably persist in
some laboratory mouse strains (55
, 56)
. Altogether, these
observations suggest that leptin is not essential for reproduction when
other, as yet unidentified factors are present.
Presence of torpor in food-restricted ob/ob and
fatless transgenic mice
Torpor is a state of physical inactivity, reduced core body
temperature, and decreased metabolic expenditure. It has been
documented that mice can enter torpor when there is a quiet
environment, food scarcity, and a low ambient temperature
(57)
. During deep torpor, mice maintain a core body
temperature down to a minimum of 16–19°C. Owing to sufficient food
supply and adequate room temperatures, deep torpor is rarely seen in
laboratory mice. The ob/ob mouse is an exception, entering
torpor even when well fed and housed at room temperature
(58)
. To study the role of adipose tissue and leptin in
the regulation of entry into torpor, researchers have used two
leptin-deficient animal models: massively obese ob/ob mice
and virtually fat-depleted A-ZIP/F-1 transgenic mice (59)
.
The A-ZIP/F-1 phenotype has virtually no white fat and a reduced amount
of brown adipose tissue (BAT) (60)
. Thus, even though both
are hypoleptinemic, ob/ob and A-ZIP/F-1 mice provide an
important contrast: the former has massive triglyceride stores and the
latter exhibits very low energy reserves. A-ZIP/F-1 mice down-regulate
their metabolic rate early in fasting and then go into deep torpor to
conserve energy (59)
. In fasted ob/ob mice,
leptin replacement prevented deep torpor and the modest hypothermic
state. In contrast, in fasted A-ZIP/F-1 mice, leptin treatment did not
prevent torpor. These findings suggest that torpor in rodents is
induced by both leptin-dependent and -independent mechanisms
(59)
. In the setting of large energy stores leptin
administration (e.g., in leptin-treated ob/ob mice), the OB
protein plays a key role in torpor induction. However, the occurrence
of profound torpor in food-restricted mice with no white fat tissue
that is not preventable by leptin administration indicates another
factor may be important in the adaptation of mice to starvation. This
role has been ruled out for thyroid hormones and
ß3 agonists, important regulators of basal
metabolic rate (59)
.
Adaptation of food ingestion to environmental temperature
Mice maintain thermal balance over a wide range of ambient
temperatures, from thermoneutrality down to mild cold
(61)
. Thus, mice can adjust their energy expenditure over
a fivefold range in order to maintain thermal balance. At the same
time, mice are also able to adjust their food intake over this same
fivefold range so as to maintain their fat stores while living at a
wide range of ambient temperatures (61)
. The nature of the
signal that informs a mouse that it should match its energy intake to
its energy expenditure is unknown, as is the nature of the
neuropeptidergic pathways that control the thermal balance feeding
system. Changes in leptin concentrations are not involved, since
mutants that lack either leptin or leptin receptors can still adjust
their food intake in accordance with acclimation temperature
(62)
. The principal site of thermogenesis in mice
acclimated to different ambient temperatures below thermoneutrality is
BAT. Some insight into the control system underlying thermal balance
feeding has come from studies with BAT-ablated mice (61
, 63
64
65)
. These mice carry a transgene containing diphtheria
toxin A chain (DTA) linked to the uncoupling protein-1 promoter
(UCP-DTA mice) (66)
. In these mice, the development of
obesity was predicted on the basis of the expected deficit in energy
expenditure for the observed level of thermogenesis. However, UCP-DTA
mice also become hyperphagic, a consequence of BAT ablation that was
not anticipated by any known function of BAT. Moreover, it has been
shown that lack of UCP1-mediated thermogenesis does not in itself
induce obesity or hyperphagia (67)
. Thus, other underlying
mechanisms for the hyperphagia of UCP-DTA mice must be sought. It has
been proposed that UCP1-expressing brown adipocytes secrete a satiety
factor in inverse relation to environmental temperature and sympathetic
stimulation as well as being able to operate independently of changes
in leptin concentrations (61
, 65)
. The generation of this
signal would be maximal at thermoneutrality, progressively suppressed
by norepinephrine as acclimation temperature decreases and sympathetic
nervous system activity increases, and minimal in a cold environment in
which sympathetic activity reaches its maximum level. UCP-DTA mice are
predicted to secrete this factor normally at thermoneutrality but
presumably lose the ability to secrete it at usual animal house
temperatures (22–24°C). Thus, at ambient temperature,
UCP1-expressing brown adipocytes die; UCP-DTA mice are unable to
produce the putative satiety factor and consequently become hyperphagic
(61
, 65)
. The main function of this unknown factor would
be (mediated by a decrease in its concentration) to promote an increase
in food intake to match the increasing energy expenditure as
environmental temperature decreases.
|
FUTURE PERSPECTIVES |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
. Since then, remarkable progress has been
made in the application of genetics to understand the control of body
weight (68)
.
The application of transgenic technologies in life science research has
become widespread. Within the field of bioenergetics and metabolism,
their application had been expected to provide definitive evidence for
many longstanding metabolic hypotheses and theories. The expectation is
generally founded on one of the prevailing benefits of transgenic
technologies—the ability to decipher the importance of a specific
protein under in vivo conditions. During the elaboration of this
hypothesis, identification of a series of new molecules implicated in
obesity and adipose tissue development has been published. The gene
Lpin1 has been shown to encode a novel nuclear protein,
which has been named lipin (69)
. The identification of
lipin has revealed a new factor required for normal adipose tissue
development and metabolism. Elucidation of the molecular function of
lipin will likely lead to new insights into these processes. This novel
family of nuclear proteins contains at least three members in mammalian
species as well as homologs in distantly related organisms from human
to yeast. The human ortholog LPIN1 is a potential candidate
gene for lipodystrophy, a heterogeneous group of disorders with unknown
genetic determinants (except for LMNA, which is responsible
for Dunnigan-type familial partial lipodystrophy).
A variety of adipocyte-derived molecules have been proposed as
potential mediators of the resistance to insulin associated with
obesity. Recently, the discovery of a novel hormone, which the
researchers named resistin (for resistance to insulin), was reported
(70)
. Resistin is specifically expressed and secreted by
adipocytes, apparently in proportion to fat pad size. It impairs
glucose tolerance and insulin action, thus linking obesity to diabetes.
Resistin seems to be part of an emerging new family of secreted
proteins with a tissue-specific pattern of expression and probably
common signaling characteristics. Two other members of the family
resistin-like molecule and ß (RELM and ß) have been cloned
(71)
. RELM is expressed in white adipose tissue
(apparently in the stromal vascular constituents rather than in
adipocytes), mammary gland, heart, lung, and tongue; it has unknown
biological functions. RELMß is expressed and secreted in the
gastrointestinal tract only, especially in the colon, and at lower
levels in the cecum and ileum. The expression of RELMß is higher in
proliferative epithelial cells and is markedly up-regulated in tumors,
suggesting a role for this protein in proliferation.
We were still trying to digest all the information derived from these
findings when the outcome of knocking out the glucose transporter gene
GLUT4 was published this year (72)
. These knockout mice
have normal growth and adipose tissue mass despite markedly impaired
insulin-stimulated glucose uptake in adipocytes. Although GLUT4
expression is preserved in muscle, these rodents develop insulin
resistance in muscle and liver, manifested by decreased biological
responses and impaired activation of phosphoinositide-3-OH kinase.
Therefore, adipose-selective depletion of GLUT4 in mice leads to
impaired glucose tolerance and insulin resistance with preserved
adipose mass. Insulin resistance occurs secondarily in muscle and
liver, as evidenced by defective proximal signaling and reduced
physiological responses. Moreover, the insulin resistance cannot be
accounted by changes in circulating free fatty acids, triglycerides, or
leptin or by changes in tumor necrosis factor expression in adipose
tissue. Thus, selective down-regulation of GLUT4 and glucose transport
in adipose tissue can cause insulin resistance and thereby increase the
risk of developing diabetes.
The pace of advance is likely to accelerate as functional genomics and the human genome project expand and mature. The reasoning underlying the hypothesis of the existence of further leptin-like products relies on the fact that nature has carved a web of factors essential to ensure its own survival. Basic physiological insight like the one provided by this hypothesis is sorely needed in order to enhance the ‘targeted’ identification of further satiety factors, fat controllers, or lipostatic hormones. Future studies aimed at identifying the aforementioned yet unknown factors may cast new light on the causes underlying human obesity and broaden our knowledge of the pathophysiology of this multifactorial disease.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 28, 2001.
Revision received May 24, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONHYPOTHESISPARABIOSIS EXPERIMENTSUNEXPLAINED FINDINGSFUTURE PERSPECTIVESREFERENCES |
|---|
knockout mice. Proc. Natl. Acad. Sci. USA 97,12729-12734This article has been cited by other articles:
|
